|
1. |
Complex phosphates containing mono- and trivalent cations |
|
Russian Chemical Reviews,
Volume 71,
Issue 8,
2002,
Page 619-650
Lidia N. Komissarova,
Preview
|
|
摘要:
Russian Chemical Reviews 71 (8) 619 ± 650 (2002) Complex phosphates containing mono- and trivalent cations L N Komissarova, MG Zhizhin, A A Filaretov Contents I. Introduction II. Double phosphates containing mono- and trivalent cations III. Hydrogen phosphates containing mono- and trivalent cations IV. Fluoride phosphates containing mono- and trivalent cations V. Hydroxide phosphates containing mono- and trivalent cations VI. Other complex phosphates containing mono- and trivalent cations VII. Conclusion Abstract. structures, synthesis, the on data the surveys review The The review surveys the data on the synthesis, structures, thermal complex of properties physical and stability thermal stability and physical properties of complex phosphates phosphates containing phos- double from generated M containing MI and and MIII cations cations generated from double phos- phates replacement partial the by phates by the partial replacement M or I?H+ or PO PO3¡ 4 ?Xn7 4 [X , O n7=F7, OH , E 7, EOn¡ (VO3¡, MoO2¡) or S 4 sizes The , MoO4 ) or S2O23¡].]. The sizes of in role dominant a have ratio their and the of theM M I and andMIII cations cations and their ratio have a dominant role in the structure formation of complex phosphates. The prospects of the structure formation of complex phosphates. The prospects of the materials polyfunctional as phosphates complex of use the use of complex phosphates as polyfunctional materials are are discussed. 161 includes bibliography The discussed. The bibliography includes 161 references.references. I. Introduction In the last two decades, an extensive search has been carried out for new ferroelectric, piezoelectric, laser, luminescent and other materials, which can be applied in quantum electronics and fibre optics and used as sorbents and catalysts. In this connection, complex phosphates containing mono- and trivalent cations, in particular, compounds of the MI III(PO 3M 4)2 and MI3MIII 2 (PO4)3 types (MI are alkali metals), attract considerable interest.1 These compounds exhibit ionic conductivity as well as nonlinear optical and other interesting properties. Numerous luminophores pos- sessing excellent technical characteristics have been already pre- pared on the basis of inorganic phosphates.2, 3 The classical approach to the design of new materials involves the construction of more compositionally complex compounds and composites.Presently, new methods for the preparation of phosphates containing different combinations of mono- and polyvalent cations as well as procedures for the synthesis of complex phosphates in which the PO3¡ anions are combined 4 L N Komissarova, M G Zhizhin Department of Chemistry, MV Lomonosov Moscow State University, Leninskie Gory, 119992 Moscow, Russian Federation. Fax (7-095) 932 88 46. Tel. (7-095) 939 58 56 E-mail: komissarova@inorg.chem.msu.ru (L N Komissarova). Tel. (7-095) 939 52 48. E-mail: zhizhin@tech.chem.msu.ru (M G Zhizhin) A A Filaretov N S Kurnakov Institute of General and Inorganic Chemistry, Russian Academy of Sciences, Leninsky prosp.31, 119991 Moscow, Russian Federation. Fax (7-095) 954 12 79. Tel. (7-095) 939 58 56. E-mail: andfil@lamar.colostate.edu Received 24 April 2002 Uspekhi Khimii 71 (8) 707 ± 740 (2002); translated by T N Safonova #2002 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2002v071n08ABEH000728 619 619 625 633 641 646 648 with the F7, OH7, VO3¡ 4 , MoO24 ¡ or S2O23 ¡ anions attract considerable attention. Purposeful changes in the compositions of the cationic and anionic components of complex phosphates allow one to vary the dimensional characteristics of framework structures and inter- layer spaces in layered phases thus influencing various properties of complex phosphates containing MI and MIII cations, in particular, the ionic mobility, ion-exchange processes, catalytic properties and sensor characteristics (gas detection).Complex hydrogen phosphates containing the H+ ions and polyvalent cations are of obvious interest. The data on these types of complex phosphates have not been generalised. Such complex orthophos- phates involving the H+ ions and anions of different nature along with MI and MIII cations are the subject of the present review. Both the data published in the literature and the results of our studies are surveyed. The review also surveys the data on double orthophosphates containing mono- and trivalent cations. The latter compounds along with hydrogen orthophosphates with trivalent cations are parent compounds of complex phosphates under consideration.Each group of complex phosphates containing MI and MIII cations is characterised by different compositions. The targeted synthesis of new complex phosphates containing variable combi- nations of cations and anions and predictions of their physico- chemical characteristics require revealing the composition ± structure ± property relationships for different types of com- pounds, in particular, for complex double and hydrogen phos- phates containing MI and MIII cations, fluoride phosphates, hydroxide phosphates and phosphates containing other tetrahe- dral groups. II. Double phosphates containing mono- and trivalent cations Studies of the phase formation in the MI3PO4±MIIIPO4 systems (MI=Na, K;MIII=REE, Y, In, Sc) 4± 8 were started in 1970s in connection with a search for new laser, luminescent and other materials.Two types of compounds, viz., MI III(PO 3M MI III 4)2 (1) and 3M2 (PO4)3(2), were found.{ Compounds 1 are known for all rare-earth (Y, La ± Lu) and some other (Sc, In, Fe) elements, III { Data on double phosphates with more complex compositions, viz., Na6M3 (PO4)5 (3), where MIII=Dy, Ho or Tm, are limited to the conditions of their formation in the Na3PO4 ±MIIIPO4 binary systems. 4620 whereas compounds 2 are typical only of trivalent cations with the ionic radii rMIII40.85A. In studies of these compounds, primary attention has been given to their structures, polymorphic trans- formations, ionic conductivity and luminescent properties.1. Synthesis Phosphates with compositionsMI IIIÖPO4Ü2 andMI3MIII 3M 2 ÖPO4Ü3 are generally prepared by the solid-phase synthesis from stoichio- metric amounts of oxide MIII 2 O3, carbonate MI2CO3 and alkali- metal and ammonium hydrogen orthophosphates characterised by different degrees of replacement at temperatures from 400 to 1100 8C.9± 11 Different types of dielectric ceramics with such compositions are produced by hot pressing.12 The synthesis from aqueous solutions of the corresponding salts at normal 13 or high 14 pressure (hydrothermal method) is used more rarely. There are also procedures, which, strictly speaking, cannot be considered as preparative. As an example we refer to the reactions of cerium(IV) oxide with sodium and potassium hydrogen phos- phates or diphosphates 15 accompanied by oxygen liberation.Another procedure for the preparation of double phosphates involves calcination of mixtures of REE hydrogen phosphates with composition H3MIII(PO4)2 .2H2O with alkali-metal hydrox- ides or carbonates.16 Single crystals ofMI IIIÖPO4Ü2 were grown 3M under the hydrothermal conditions at high temperatures (T=600 ± 900 8C)14 or prepared by crystallisation from flux using an excess of alkali-metal halides (for example, KF).17 2. Structures and properties Compounds with composition MI IIIÖPO4Ü2 containing differ- 3M ent metal cations are similar in structure and physicochemical characteristics.The main structural feature common to all com- b a b a Figure 1. Projections of the K3Na(SO4)2 (a), b-K2SO4 (b), Na2CrO4 (c), and NaZr2(PO4)3 (NASICON) (d ) structures. Hereinafter, the structures are drawn using the atomic coordinates published in the original studies. a M, X X,M Yc X Y Y X Y X Y c X L N Komissarova,MG Zhizhin, A A Filaretov pounds of this group is the fact that they belong to the glaserite [K3Na(SO4)2], arcanite (b-K2SO4), sodium chromate (Na2CrO4) or NASICON{ [NaZr2(PO4)3] structural types (Fig. 1). The K3Na(SO4)2 (space group P3m1, a1=b1= 5.680, c1=7.309A),18 b-K2SO4 (space group Pnam, a2=7.456, b2=10.080, c2=5.776A) 19 and Na2CrO4 (space group Cmcm, a3=5.862, b3=9.251, c3=7.145A) 20 structural types are wide- spread among inorganic salts with the PO3¡ tetrahedral anions.1, 21, 22 All these salts belong to layered structures. In the crystals of these salts, the PO4 tetrahedra have different orienta- tions with respect to the cation positions (denoted asM, X and Y, see Fig. 1 ) 21 and can combine with each other to form numerous subtypes.The relationship between the parent structural types can be represented as the relation between the unit cell parameters of glaserite (a1, b1, c1) and those of arcanite (a2, b2, c2) and sodium chromate (a3, b3, c3): a2&c1, b2&a1 3, c2&a1 and a3&a1, b3&a1 3, c3&c1. pAAA It is assumed that the ratio between the radii of the cations is the governing factor responsible for the formation of glaserite- or arcanite-like structures.21 The glaserite-like structures are stable for compounds in which the difference between the radii of the M+ and M3+ cations [Dr(M+±M3+)] 23 occupying the M, X and Y positions is in the range 0.594Dr(M+±M3+)40.89A.21, 24 ± 27 However, compounds that are characterised by Dr(M+±M3+)< 0.59A and have glaserite-like structures are also known.Thus, Dr(M+±M3+) for Na3Fe(PO4)2 (glaserite structural type) is 0.38A.24 At Dr(M+±M3+)<0.59A, compounds generally crystallise in the arcanite structural type.19 ± 22 {NASICON stands for Na Super Ionic Conductors. b X X Y c d b M M b a 4 pAAA M XComplex phosphates containing mono- and trivalent cations The majority of representatives of the MI IIIOPO4U2 group 3M have structures of arcanite (b-K2SO4) or structures of an inter- mediate type (combinations of the b-K2SO4 and Na2CrO4 struc- tures).In particular, the former structural type is typical of K3MIII(PO4)2 (MIII=La ¡¾ Yb) 1, 17, 28 and the latter type is char- acteristic of K3In(PO4)2 (see Ref. 29) and b-Na3MIII(PO4)2 (MIII= La ¡¾ Lu, In).13, 14, 30 ¡¾ 33 These compounds belong to `arcanite-like phosphates'. In the structures of glaserite- and arcanite-like phosphates MI IIIOPO4U2, theMI andMIII cations occupy the X positions in 3M an ordered fashion. The Y positions are occupied only by mono- valent cations. However, the coordination polyhedra of the cations in the structures of MI IIIOPO4U2 differ from those in 3M the parent structural types due to distortions of the ideal b-K2SO4- and Na2CrO4-type crystal structures caused by displacements of the cations from the ideal positions and rotation of the PO4 tetrahedra. For example, the coordination sphere about potas- sium in the structures of K3MIII(PO4)2 (MIII=Y, La ¡¾ Yb) involves nine or ten oxygen atoms, whereas the coordination spheres about the REE atoms are formed by seven oxygen atoms.In Na3MIII(PO4)2, the coordination numbers of both Na+ and MIII vary from 6 to 8 ¡¾ 12. The NASICON structural type can be described as the {[MIII 2¢§x&x(PO4)3]37}3? framework whose channels are occupied by alkali-metal cations.21 The NaZr2(PO4)3 (NASICON) struc- tural type is widespread among superionic conductors, whereas double phosphates with composition Na3MIII(PO4)2 containing only REE of the end of the series (MIII=Tm, Yb, Lu) belong to this structural type.According to the stoichiometry of the frame- work, the structures of the latter compounds should be described by the Na4:5MIII 1:5OPO4U3 formula.34 Most of the sodium-containing phosphates Na3MIII(PO4)2 (MIII=Y, La ¡¾ Lu, In) have two or three crystal modifications each (Fig. 2, Table 1), viz., monoclinic low-temperature b and g modifications based on the arcanite structure (A type) or a combination of the b-K2SO4 and Na2CrO4 structures (arcanite- Table 1. Crystallographic characteristics of theMI3MIII(PO4)2 compounds. Compound MIII Space group Crystal system Subtype (structural type) La ¡¾Ce Pc21b Na3La(PO4)2 (A1) b-Na3MIII(PO4)2 ortho- rhombic Na3Nd(PO4)2 (A2) " Pr ¡¾Eu Pbc21 Cc Y, Gd ¡¾Lu Na3Nd(VO4)2 (A3) mono- clinic In Y, Nd7Er Na3Nd(PO4)2 (A2) P21/c Pbc21 g-Na3MIII(PO4)2 Na3In(PO4)2 (A5) "ortho- rhombic Na3Tm(PO4)2 (A4) " Tm¡¾ Lu (Pnm21) P3m1 trigonal La ¡¾ Er K3Na(SO4)2 (G) a-Na3MIII(PO4)2 R3c Tm¡¾ Lu NaZr2(PO4)3 (N) " 24 15.87 (Nd) ¡¾ 13.95 (Nd) ¡¾ 18.47 (Nd) ¡¾ 7 18.31 (Er) Pnmm 12 18.43 (Tm) ¡¾ 6.97 (Tm) ¡¾ 15.82 (Tm) ¡¾ 7 15.80 (Lu) 7.69 (La) ¡¾ 7.56 (Er) 21.88 (Tm) ¡¾ 7 21.80 (Lu) 7.048 In P21/m Na2CrO4 (C) Fe mono- clinic K3Na(CrO4)2 (G) " Y, La ¡¾Yb K3Nd(PO4)2 (A) " C2/c P21/m K3MIII(PO4)2 Lu trigonal P3K3Na(SO4)2 (G) In K3In(PO4)2 (A6) C2/c mono- clinic trigonal Sc P3K3Na(SO4)2 (G) aZ is the number of formula units per unit cell.621 T /8C 1200 G 1000 N 6 A2 800 A1 A2 C A4 A3 G 600 A5 Fe In Yb Er Dy Gd Sm Pr La Sc Ce Nd Eu Tb Ho Tm Lu Decrease in rMIII Figure 2. Occurrence of the structural types of the Na3MIII(PO4)2 compounds depending on the temperature and nature of theMIII cation. A1 ¡¾A5, arcanite-like subtypes (a combination of the b-K2SO4 and Na2CrO4 structures; G, N and C, the structural types of glaserite, NASICON and sodium chromate; N0, the NASICON-like phase with variable composition Na3O1axUMIII O2¢§xUOPO4U3 (04x40.2). Solid and empty circles denote polymorphic transformations and decom- position, respectively; the cross indicates incongruent melting.like structures, A1 ¡¾A5 subtypes) and high-temperature a mod- ifications derived from the structures of glaserite (G type) or NASICON (N structural type).20, 34 The polymorphism of sodium-containing phosphates is, apparently, associated with a small difference in the ionic radii of sodium and trivalent REE. In arcanite-like structures of b-Na3MIII(PO4)2, the proportion of the sodium-chromate struc- tural motif in a combination of the b-K2SO4 and Na2CrO4 structures increases on going from lanthanum [rVIII(La3+)= 1.18A] to lutetium [rVI(Lu3+)=0.85A] and indium Ref. c /AE b /AE a /AE b /deg 14, 32 33Melt Za Unit cell parameters 30 14.05 (Ce) 18.67 (Ce) 8 5.34 (Ce) 7 33 24 15.89 (Pr) ¡¾ 15.88 (Eu) 13.97 (Pr) ¡¾ 18.51 (Pr) ¡¾ 7 18.47 (Eu) 13.96 (Eu) 12 27.59 (Gd) ¡¾ 5.35 (Gd) ¡¾ 13.92 (Gd) ¡¾ 91.2 (Gd) ¡¾ 13 91.3 (Lu) 143.36 13.88 (Lu) 8.616 13.93 (Er) 15.86 (Er) 13 27 6.95 (Lu) 7 7 34 7 18.33 (Lu) 1 5.47 (La) ¡¾ 5.37 (Er) 6 9.15 (Tm) ¡¾ 9.11 (Lu) 2 8.634 32 90.30 5.455 91.44 13.873 5.034 4 9.074 24 2 9.632 (La) ¡¾ 5.660 (La) ¡¾ 7.514 (La) ¡¾ 90.55 (Lu) ¡¾ 28 90.92 (Yb) 7 7.417 (Yb) 7.731 5.613 (Yb) 7 9.324 (Yb) 3 9.611 29 N0+Na3PO4 (solid solution) 5.28 (Lu) 18.220 27.27 (Lu) 4 7.127 26 91.44 9.700 11.193 8 15.643 25 7.619 3 9.424 7 7622 [rVI(In3+)=0.76A] to form new subtypes in the temperature range T<700 ± 900 8C.This is in agreement with the general principle consisting in the fact that trivalent cations in arcanite- like structures [for example, in b-Na3La(PO4)2] occupy the XO8 polyhedra corresponding to the XO9 polyhedra in the structure of b-K2SO4, whereas these cations with smaller ionic radii tend to reduce their environment to octahedral, which corresponds to the XO6 polyhedron in the structure of sodium chromate.30, 33 The Na3Sc(PO4)2 and Na3Fe(PO4)2 compounds with the Sc3+ or Fe3+ cations possessing smaller ionic radii [rVI(Sc3+)=0.745A, rVI(Fe3+)=0.645A] than In3+ do not show polymorphism and crystallise in the glaserite structural type.24 As the temperature is increased (T>900 8C), sodium-con- taining phosphates Na3MIII(PO4)2 (MIII=La ± Er) are trans- formed from the b and g modifications into the a modification corresponding to the higher-symmetry glaserite structural type.The high-temperature a modifications of Na3MIII(PO4)2 (MIII=Tm±Lu) belong to the NASICON structural type.34 At high temperatures, a-Na3In(PO4)2 would also be expected to transform into the NASICON-like modification. Instead, this compound decomposes at T=910 8C to give a NASICON-like phase with variable composition Na3(1+x)In(27x)(PO4)3 (04x40.2) and a solid solution based on a-Na3PO4. Analogous processes were observed upon heating of the Na3Sc(PO4)2 and Na3Fe(PO4)2 compounds above 790 and 975 8C, respectively.24, 31 The nature and size of the REE ion have only a slight effect on the formation of a particular structural type of double potassium- containing phosphates K3MIII(PO4)2.28 However, a decrease in the ionic radius of the trivalent cation on going from ytterbium to lutetium also leads to the transformation from an arcanite-like to glaserite-like structure, which is accompanied by an increase in the symmetry of the crystal lattice from monoclinic [K3MIII(PO4)2 compounds with large MIII cations (Y, La ± Yb) 27, 28] to trigonal [typical of K3MIII(PO4)2, whereMIII=Lu or Sc 1, 26].In the structures of K3MIII(PO4)2 (MIII=La7Yb), two of three potassium atoms are surrounded by ten oxygen atoms each, whereas the third potassium atom is surrounded by nine oxygen atoms. Due to distortion of the initial arcanite structure, the X[6+3]O9 polyhedra are transformed into MIIIO7 seven-vertex polyhedra, viz., trigonal monocapped prisms with the MIII7O distances in the range of 2.20 ± 2.60A.In the virtually non- distorted glaserite-like structures of K3MIII(PO4)2 (MIII=Lu, Sc), the MIII cations have an octahedral environment formed by oxygen atoms. Apparently, the transformation from an arcanite- like to glaserite-like structure is associated not only with an increase in the difference between the MIII and K+ radii but also with the effect of the electronic structure of the MIII atom. Thus the K3Lu(PO4)2 compounds [Dr(K+±LuVIII 3á )= 0.58A] would be 4 expected to have arcanite-like structures stable at Dr(M+±M3+)<0.59A. However, Dr(K+±Lu3á VI ) is actually equal to 0.69A and the glaserite-like structural type is realised.Because of the larger ionic radius of ytterbium [Dr(K+±Yb3+)=0.57A] as compared to that of lutetium and the more labile f shell, the ytterbium cation tends to expand the nearest environment. As a consequence, K3Yb(PO4)2 crystallises in the b-K2SO4 structural type. However, the arcanite-like structure of K3Yb(PO4)2 is unstable and is readily destabilised under the action of different factors, in particular, upon the partial replacement of the PO3¡ ions by the bulkier VO3¡ 4 anions.28 The K3In(PO4)2 compound may be assigned to a new subtype of arcanite-like phosphates (A6) characterised by a combination of the b-K2SO4 and Na2CrO4 structural motifs (Fig. 3).29 Analysis of the crystal structure of K3In(PO4)2 demonstrated that the potassium cations displace the indium cations from the X posi- tions to the Y positions corresponding to the Y positions in the structure of sodium chromate (see Fig.1 c), as distinct from double phosphate Na3In(PO4)2 in which the Y positions are occupied exclusively by the Na+ cations. Hence, the larger potassium cation (compared to sodium) causes the indium ions to change an octahedral environment formed by the oxygen atoms L N Komissarova,MG Zhizhin, A A Filaretov K InO6 b PO4 a Figure 3. Structure of K3In(PO4)2 projected onto the ab plane. 4Ü2 from six PO4 tetrahedra for an octahedral environment formed by the oxygen atoms belonging to five PO4 tetrahedra. In the structures of arcanite-like double phosphates Na3MIII(PO4)2 (MIII= La ± Lu), the co-existence of the b-K2SO4 and Na2CrO4 motifs is achieved due to the presence of REE with different coordination numbers (6 ± 8), whereas the co-existence of these motifs in the K3In(PO4)2 structure is, apparently, associated with a substantial distortion of the polyhedra about the potassium atoms.Like the K3MIII(PO4)2 compounds (MIII=REE), K3In(PO4)2 does not possess polymorphism below 1000 8C. The only difference is that the REE compounds decompose prior to melting (T=800 ± 1000 8C), whereas K3In(PO4)2 melts con- gruently at a temperature above 1200 8C.35 Taking into account the difference in the K3In(PO4)2 and K3MIIIÖPO4Ü2 structures (MIII=Lu, Sc) and considering that the radii of the MIII cations decrease in the series Lu7In7Sc, it can be assumed that the size of theMIII cation is of secondary importance in the formation of a particular structural type in the group of the MI IIIÖPO 3M compounds.Double phosphates with composition MI III 3M2 ÖPO4Ü3 (type 2, MI=Li, Na, K; MIII=Fe, Sc, In, Cr) (Table 2) are typical of smallMIII cations (rM3+<0.8A).36 ± 48 The crystallographic char- acteristics of such compounds in which MIII is an REE atom (Ce ± Gd) are available in the literature.10, 11 For these com- pounds, the unit cell parameters were calculated assuming of trigonal system. However, these data are in doubt. Composites of the general formula MI3Ö1áxÜMIII 2¡xÖPO4Ü3 (0.074x40.5) are known. These composites belong to the region of solid solutions on the basis of high-temperature a modifications of Na4:5MIII 1:5(PO4)3 (MIII=Tm, Yb, Lu).34 The crystal structures of most of the MI III 3M2 ÖPO4Ü3 com- pounds were established, which made it possible to assign these compounds to the NASICON type.Complex phosphates of the general formula MxM02(PO4)3, where M=M+ (Li, Na) or M2+ (Sr, Ba) and M0=M2+ (Mg), M3+ (Sc, In, Fe), M4+ (Zr, Ti) or M5+ (Ta, Nb), belong to this structural type. The occupancies of the M positions can vary over a wide range (04x44). The exception is the low-temperature b modification of Na3In2(PO4)3, which has an unusual structure.45 Depending on the combination of the cations, the symmetry of the unit cell in the MI III 3M2 ÖPO4Ü3 compounds changes from trigonal to monoclinic.All structures have framework structures built from the MIIIO6 and PO4 polyhedra in a ratio of 2 : 3. The cavities in the framework are occupied by alkali-metal cations (Li+, Na+). The frameworks of the structures are penetrated by channels of two types, I and II, extended along the b axis of the unit cell (Fig. 4 a). The channel of type III are perpendicular to the channels I and II and intersect these channels. The smallest cross- sections of the channels I, II and III are 1.0, 1.1 and 0.9A, respectively. The large cavities in these channels (at the intersec- tions of the channels I and III and of the channels II and III) areComplex phosphates containing mono- and trivalent cations Table 2.Crystallographic characteristics of theMI III 3M2 (PO4)3 compounds. Crystal system Compound MIII b-Li3MIII 2 (PO4)3 III a-Li3M2 (PO4)3 III b-Na3M2 (PO4)3 III a-Na3M2 (PO4)3 monoclinic ""trigonal orthorhombic trigonal ""monoclinic trigonal monoclinic "trigonal " Fe Sc In Fe a Sc In a Fe a Sc a In Cr a Fe a Sc a In a Ce ±Gd trigonal Ce ±Gd K3MIII 2 (PO4)3 monoclinic Fe a The NASICON structural type. b The space groups were not unambiguously established. occupied by alkali-metal cations located in two positions, viz., M(1) and M(2) (Fig. 4 b). The analysis of the number of the polyhedra and their coordination environments demonstrated that the general crystal-chemical formula of the NASICON structural type can be written as [M(1),M(2)]3M(3)2(PO4)3.Taking the space group R3c (the highest symmetry for the compounds under consideration) as the basis, the M(1)O6 poly- hedron can be described as an elongated antiprism with the a a c I II bA M(1) A C D D M(1) C M(2) B x Figure 4. Structure ofMI III 3M2 ÖPO4Ü3 projected onto the ac plane (a) and the oxygen environment about the M(1) and M(2) positions occupied by alkali-metal cations (b). The M(3) position is omitted. 623 Z Ref. Unit cell parameters Space group c /AÊ b /AÊ a /AÊ b /deg P21/n P21/n P21/n R3c Pbcn R3c R3c R3c C2/c R3c C2/c Bb R3c see b 37 38 39 40 41 42 40 47 31, 45 46 43 44 31, 48 10, 11 44464666464466 90.5 90.0 90.0 77777114.16 7125.1 127.2 77 8.612 8.802 8.908 78.835 77712.772 78.740 9.109 77 10, 11 see b 3 7 7 12.005 12.273 12.290 22.459 8.827 23.268 21.808 22.340 6.583 21.619 8.724 8.928 22.100 14.44 (Ce) ± 14.20 (Gd) 14.41 (Ce) ± 14.27 (Gd) 16.691 8.562 8.853 8.592 8.316 12.400 8.432 8.727 8.927 12.439 8.643 15.070 16.100 8.966 9.18 (Ce) ± 8.86 (Gd) 9.30 (Ce) ± 8.86 (Gd) 16.303 36 118.4 9.463 C2/c 8 III 1:2Ca0:1.B y M(1)7O distances to the nearest six oxygen atoms equal to *2.6A (the other six O atoms are at considerable distances from the metal atom, viz., at *3.7A).The metal atom in the M(2) position is surrounded by ten oxygen atoms eight of which are involved in the nearest environment, whereas the remaining twoO atoms are somewhat remote. The analysis of the cross-sections of the channels showed that the cations with the radii of *1A (for example, Na+) could migrate within these channels. The cations can also pass through narrower regions, viz., through the ABC and ABD faces (see Fig. 4 b), 47 due to thermal displacements of the oxygen atoms. This fact is responsible for superionic con- ductivity of the MI III 3M2 ÖPO4Ü3 compounds.49 The cations can occupy the M(1) and M(2) positions in different fashions: (1) these positions are completely occupied [Na4:5MIII 1:5(PO4)3 or MI IIIÖPO4Ü2, where MIII=Tm, Yb, Lu];34 (2) only the M(2) 3M position is completely occupied (together with vacancies), for example, in b-Na3Sc2(PO4)3;44 (3) the M(1) and M(2) positions are statistically occupied (together with vacancies), for example, in a-Na3M2 ÖPO4Ü3, whereMIII=Sc, In.47, 48 The low-temperature b modification of Na3In2(PO4)3 is described by the crystal-chemical formula X(2)X(1)M(1)..M(2)2(PO4)3. Alluaudite-like phosphates have analogous struc- tures [the mineral alluaudite has the formula Na0:6Mn2á .Fe3á 2 Mg0:1(PO4)3]. The structures of indium compounds are characterised by the presence of the In2O10 dimers (bioctahedra) (Fig. 5 a), which form a framework structure together with the tetrahedral PO4 groups (Fig. 5 b).The analysis of the polyhedra in the crystal structure of b-Na3In2(PO4)3 led one to the conclusion that the complete occupancy of theM(2) position by a double- or triple-charged cation according to the crystal-chemical formula X(2)X(1)M(1)M(2)2(PO4)3 is a necessary condition for crystalli- sation of double phosphates in the alluaudite structural type. It should be noted that the ionic radius of the cation in the dimer [M(2) position] must be in the strictly specified range 0.744r40.76A. The radii r of the Fe2+ and In3+ cations fall within this range. Actually, the data on alluaudite-like phosphates Na2Fe3+Fe 2á 2 (PO4)3, NaCaMn2+Fe 22á(PO4)3 and NaCd. .In 3á 2 (PO4)3 are available in the literature.50 The fact is that the occurrence of alluaudite-like structures is determined by the size of the PO4 tetrahedron. Two of ten PO4 groups in the dimer are bridging and fix the distances between the vertices of the octahe- dra [d(O_O)&2.5A], which are equal to the length of the edge of the tetrahedron.As a result, this distance d[M(2)_M(2)] in the624 In a c Na In2O10 PO4 b a Figure 5. The In2O10 bioctahedron (a) and the structure of b-Na3In2(PO4)3 projected onto the ab plane (b). dimer must be approximately equal to 3.0A. Due to repulsion The sizes of the polyhedra in the structures of NASICON-like between the cations, this distance increases to 3.19A in phosphates allow one to predict the positions of the cations in the 2Fe3áFe22á(PO4)3, 3.35A in b-Na3In2(PO4)3 and 3.38A in Na NaCdIn2(PO4)3.The larger or smaller size of the cation as compared to In3+ and (i.e., r<0.74 and r>0.76A) will lead to destruction of the dimer as the structural block, which is con- firmed by the lack of information on alluaudite-like phases with compositionMI III 3M2 ÖPO4Ü3, whereMIII=REE, Sc, Fe, Cr. From the crystal-chemical viewpoint, the transformation of the alluaudite-like b phase into the NASICON-like a phase [which was established from the analysis of the structures of both modifications of Na3In2(PO4)3] involves the radical rearrange- ment of the atoms. It was demonstrated that the transformation from the b phase to the a phase is accompanied by an increase in the distance between the indium atoms, destruction of the In2O10 Table 3.Characterisation of the structures of different types of double phosphates containing mono- and trivalent cations. Compound Ratio MI:MIII 3 : 1 MI IIIÖPO4Ü2 3M MI=Na;MIII=REE, In, Ga, Sc, Al, Fe, Cr MI=K;MIII=REE, Sc, In, Fe 3 : 2 MI III 3M2 ÖPO4Ü3 MI=Li, Na;MIII= Yb ± Lu, Ga, Sc, Al, Fe Cr MI=Na,MIII=In L N Komissarova,MG Zhizhin, A A Filaretov a PO4 In In b In2O10 b dimers and formation of the isolated InO6 octahedra. This rearrangement leads to an increase in `porosity' [D(V/Z)b?a=8.1%] of the crystal structure due to the formation of rather large channels, which are statistically occupied by alkali- metal cations. The irreversibility of the polymorphic transforma- tion b-Na3In2(PO4)3? a-Na3In2(PO4)3 at T=820 8C was con- firmed by the DTA data.Hence, compounds with two compositions, viz., MI IIIÖPO4Ü2 and MI3MIII 3M 2 ÖPO4Ü3 , are most widespread among double phosphates containing mono- and trivalent cations. The structural types of these compounds are determined by the nature and size of theMIII cation as well as by the difference in the radii of the MI and MIII cations, which is most pronounced for indium compounds (Table 3). Investigations of the luminescence properties of REE phos- phates showed that theMI IIIÖPO4Ü2 compounds (MI=Na, K; 3M MIII=Y, La, Nd, Gd) are promising laser materials.27, 51 ± 54 Due to the unique luminescence properties (large absorption and luminescence cross-sections, rather large radiative lifetimes, weak concentration quenching of luminescence, low thresholds of generation, etc.), these crystals are quite competitive even with the recognised leader of this class of materials, viz., yttrium ± aluminium garnet YAG:Nd3+, which is traditionally used in compact lasers.Luminophores based on MI IIIÖPO4Ü2 have the 3M linear current characteristic of luminance at the cathode excitation in the current-density range of 0.1 ± 50 mA cm72 (for the known Zn2SiO4 ±Mn luminophore, the saturation of luminance takes place at the current load of 5 ± 10 mA cm72). The introduction of the second MIII cation (diluent) into these materials leads to a noticeable increase in the line width resulting in deterioration of the optical characteristics.This was observed for the Na3Nd1¡xMIII x (PO4)2 (MIII=Gd, Lu, Y) solid solutions.52 The introduction of the Ce3+ and Tb3+ cations into a matrix of the b-K2SO4 type leads to an increase in the quantum yield due to the energy transfer from Ce3+ to Tb3+.53 lattice and their mobility. Thus, a decrease in the size of the structure-forming MIII cation hinders the migration of mobile alkali-metal cations within the channels of the framework. It was experimentally found that the ionic conductivity of III Na3M2 ÖPO4Ü3 is weakened as the radius of the trivalent cation decreases. At 570 8C, s is equal to 5.061072, 9.861073 and 7.861073 S cm71 for Na3Sc2(PO4)3, Na3 Fe2(PO4)3 and Na3Cr2(PO4)3, respectively.21, 51 The dependence of the ionic conductivity of NASICON-like REE-containing phases on the temperature and degree of replacement of Zr byMIII was studied only for mixed phosphates Na1áxZr2¡xMIII x ÖPO4Ü3 (MIII=Gd, Tb, Dy, Y, Er, Yb, In; 04x42).55 For the Na2.4In1.4Zr0.6(PO4)3 compound, the maximum s value (1071 S cm71) was found at Structural type Building blocks Space group Crystal system Ratio MIII : PO3¡ 4 1 : 2 P3m1 [MIIIO8(7,6)]+[PO4] Pbc21, Pnmm Cc trigonal orthorhombic monoclinic [MIIIO6]+[PO4] monoclinic P21/m, C2/c [MIIIO7(6)]+[PO4] layered [MIII=La ± Er, In (a phase), Sc, Al, Fe, Cr] framework [MIII=Tm±Lu, In (b phase)] layered (MIII=La ± Er, Sc, Fe), framework (MIII=In) 2 : 3 P21=n, C2/c, Bb R3c framework [MIIIO6]+[PO4] monoclinic trigonal orthorhombic monoclinic Pbcn C2/c framework [MIII=In (b phase)] [MIII 2 O10]+[PO4]Complex phosphates containing mono- and trivalent cations 400 8C and the minimum value (261075 S cm71) was observed at room temperature.Of REE-containing composites, the Yb-doped analogue Na2.4Yb1.4Zr0.6(PO4)3 is the most competi- tive material (s is 8.561072 S cm71 at 400 8C). III. Hydrogen phosphates containing mono- and trivalent cations III III Complex hydrogen phosphates containing mono- and trivalent cations are characterised by the presence of hydrogen bonds between the oxygen and hydrogen atoms of the HPO4 groups. Although formally belonging to double phosphates containing mono- and trivalent cations, these compounds possess a series of specific features.Hence, it is reasonable to consider these com- pounds separately. Hydrogen phosphates containing mono- and trivalent cations are classified into four main types, viz., MIMIII(HPO4)2 . nH2O, MIxMy ÖHPO4ÜzÖPO4Üu . nH2O, MI y ÖH2PO4ÜzÖHPO4Üu . nH2O xM MI IIIÖH2PO4Ü and 2M ÖHPO4Ü2 . ÖH3PO4Ü . nH2O, where MI = Na, K, Rb, Cs, H3O, NH4 andMIII=Al, V, Fe, Ga, In. Phosphates MI III(OH)(HPO4)(PO4) containing the HPO4 3M groups along with the OH groups, which serve as bridges linking theMIIIO6 polyhedra in infinite chains, are known only for Al and Ga in combination with Na. These compounds will be considered in Section V. 1. Synthesis III The hydrothermal synthesis is one of conventional procedures for the preparation of hydrogen phosphates containing the MI and MIII cations.56 ± 75 The problems associated with the preparation of compounds with specified compositions consist in choosing the optimum conditions of the synthesis.The compositions of the resulting compounds depend substantially on the parameters of the synthesis (the temperature, the degree of filling of an autoclave or a tube, the concentrations of the starting reagents, the aggre- gate state, pH of the medium, duration of the synthesis). General procedures for the preparation of this group of compounds are lacking. One of the criteria for the existence of complex hydrogen phosphates containing different cations and anions is the presence of a strongly acidic medium (pH is no higher than 2 ± 2.5).All the aforesaid refer equally to the conditions of the hydrothermal synthesis of hydrogen phosphates with compositions MIMIIIÖHPO4Ü2 . nH2O (MI= Na, K, Rb, Cs, H3O, NH4; MIII=Al, Fe, V, In; n=0 ± 18) and MIxMy ÖHPO4ÜzÖPO4Üu . nH2O (MI= Rb, Cs; MIII = V, Ga, In; n = 0.5 ± 2). The MIMIIIÖHPO4Ü2 . nH2O compounds are prepared under rather mild conditions (145 ± 240 8C, see Table 4). In studies of the phase formation in the In2O3±MIH2PO4 ± H3PO4 ±HF and In2O3±MIHal ±H3PO4 systems under hydro- thermal conditions (T=200 8C), a series of indium hydrogen phosphates with composition MIIn(HPO4)2 (MI= K, Rb, Cs, NH4) were prepared. These compounds are stable at pH 0.5 ± 1.5.56 The RbIn(HPO4)2 compound was isolated as crystals of two forms (colourless prismatic and elongated bi- pyramidal) characterised by the same empirical formula.Crystals of hydrogen phosphate Na2In2(HPO4)4 .H2O were prepared in the synthesis of microporous borophosphates.57 Interestingly, the phosphate with the same composition was synthesised under analogous conditions in the absence of Na2B4O7 . 10H2O. Attempts to prepare a gallium-containing analogue according to the same procedure failed and led to the formation of new borophosphate with composition NaGaBP2O7(OH)3.57 Attempts to synthesise mixed rubidium gallium hydrogen phosphate were also unsuccessful in spite of the fact that the initial reagent ratio was varied over a wide range. Hydrogen phosphate Rb2Ga4(HPO4)(PO4)4 .0.5H2O crystallised simultaneously with GaPO4 (see Ref. 58) due, apparently, to high stability and low solubility of GaPO4 (the solubility product is 1610721). The field of chemistry dealing with the synthesis of porous compounds has been extensively developed in recent years. These 625 compounds have inorganic frameworks built from the phosphate tetrahedra. The cavities in these frameworks can capture different positively charged groups, including organic groupings. The preparation of these compounds under hydrothermal conditions is often accompanied by the formation of complex hydrogen phosphates as by-products. For example, prismatic crystals of NH4(Al17xGax)(HPO4)2 (x=0.36) comprising *5%± 10% of the total weight of the product were obtained under mild con- ditions (T=130 8C) from a mixture consisting of aluminium and gallium chlorides, tetramethylammonium hydroxide and a cobalt complex with composition [L-Co(en)3(H2PO4)3] (en is ethylenedi- amine).59 Two types of crystals were isolated from a mixture of RbOH± FePO4 .2H2O±Me2NH± MePO(OH)2±H3PO4 at 230 8C, viz., colourless elongated parallelepiped-shaped crystals of hydrogen phosphate RbFe(HPO4)2 (see Ref.60) and prismatic crystals of unknown structure. It should be noted that a powdered product obtained in the course of the synthesis had the same nominal composition as that of the crystals of RbFe(HPO4)2 but was the low-symmetry modification. Vanadium hydrogen phosphates MIV(HPO4)2 (MI=Rb, NH4) were synthesised under slightly modified conditions (T=240 8C).61 Only triclinic crystals of a-RbV(HPO4)2 were isolated in pure form.Attempts to obtain their monoclinic analogues b-RbV(HPO4)2 and b-(NH4)V(HPO4)2 in pure form failed even upon variation of the initial reagent ratio over a wide range. According to the results of chemical analysis, the compo- sition of the impurity phase is identical to that of b-(NH4)V(HPO4)2. This indicates that the phase formation of vanadium hydrogen phosphates, like that of indium hydrogen phosphates b1- and b2-RbIn(HPO4)2, is a complex process result- ing in co-crystallisation of different modifications of a compound under the same conditions. Attempts to prepare hydrogen phos- phates MIV(HPO4)2 (MI=K or Cs) led to the formation of two phases, viz., K(VIVO)VIII(HPO4)3(H2O)2 and CsV2(PO4)..(HPO4)2(H2O)2.62 The caesium indium derivative CsIn2(PO4). .(HPO4)2(H2O)2 (see Ref. 63) was isolated from aqueous solutions containing InCl3, H3PO4, CsVO3 and Me4NOH in a ratio of 2 : 6 : 1 : 1. It should be noted that mixed caesium indium hydrogen phosphate was not produced in the synthesis with the use of CsCl, CsOH and Cs2CO3 instead of CsVO3. Hydrogen phosphates containing the H2PO¡4 and HPO24 ¡ anions were prepared from solutions containing high concentrations of H3PO4 (53 M). The synthesis with the use of CsGa(H2PO4)4 as the starting compound in the presence of Cs2CO3 and an excess of H3PO4 afforded Cs2Ga(H2PO4)(HPO4)2 .H3PO4 .0.5H2O.64 2. Structures and properties 4 The MIMIII(HPO4)2 compounds possess a series of interesting structural features. These compounds were prepared for the Al, V, Fe, Ga and In cations in combination with K, Rb and Cs56, 60, 61, 65 as well as with the single-charged NHá and H3Oá cati- ons.56, 59, 61, 66 ± 68 The composition of the unit cell of mixed indium sodium hydrogen phosphate does not correspond to the MIMIII(HPO4)2 formula, which is associated with the incorpora- tion of water molecules in the crystal lattice of this compound. Two compounds whose compositions correspond to the Na2In2(HPO4)4 .H2O57 and Na4In8(HPO4)14(H2O)6 . 12H2O for- mulas were isolated.69 The crystallographic characteristic of mixed-cation hydrogen orthophosphates MIMIII(HPO4)2 and more complex sodium indium phosphates are given in Table 5.All MIMIII(HPO4)2 compounds have frameworks built from the vertex-sharing MIIIO6 octahedra and PO4 tetrahedra. The cavities and channels of the frameworks are occupied by mono- valent cations. Variations in the sizes of the MI and MIII cations give rise to different structural types. Thus, different types of frameworks can be formed, the polyhedra are distorted and the sizes of the channels and cavities are changed. As a consequence, the crystal lattice can incorporate water molecules, which partic- ipate in structure formation. It seems essential that in all struc- tures, the protonated vertices of the PO4 tetrahedra are directed626 Table 4. Hydrothermal synthesis of complex hydrogen phosphates. Compound Starting reagents (molar ratio) Characteristics of the crystals (colour) KAl(HPO4)2 .H2O colourless prismatic the same NH4(Al17xGax)(HPO4)2 (H3O)Fe(HPO4)2 RbFe(HPO4)2 pink needle-like colourless parallele- piped-shaped yellow-green b-(NH4)V(HPO4)2 b-RbV(HPO4)2 pale-green platelet-like yellow-green a-RbV(HPO4)2 Na2In2(HPO4)4 .H2O colourless platelet- like Al(OH)3, 1MKH2PO4, 1MH3PO4 (K : P=0.5) AlCl3 .6H2O, GaCl3, H3PO4, [L-Co(en)3(H2PO4)3], 7 Me4NOH, H2O [Co(en)3:Al : Ga : P : Me4NOH:H2O=0.25 : 1 : 1 : 2.86 : 2 : 90] Fe3(PO4)2 .3H2O, FePO4 RbOH, FePO4 . 2H2O, Me2NH, MePO(OH)2, H3PO4 [V2O5: Vmet : (NH4)2HPO4:H3PO4:H2O= =1 : 1 : 2 : 10 : 190] (Rb2V2O6: Vmet : Bu4NBr :H3PO4:H2O= =9 : 2 : 1 : 60 : 760) (Rb2V2O6: Vmet : PhMe3NBr :H3PO4:H2O= =1 : 1 : 0.5 : 15 : 190) Na2B4O7 .10H2O, Na2HPO4 . 12H2O, Inmet, HCl (Na : In : P : B=17 : 1 : 6 : 10) b Na2HPO3 .5H2O, In(NO3)3 .3H2O, H3PO3 Na4In8(HPO4)14(H2O)6 . 12H2O colourless hexagonal- KIn(HPO4)2 (NH4)In(HPO4)2 b2-RbIn(HPO4)2 prismatic transparent colourless bipyramidal white powder transparent colourless bipyramidal colourless needle-like CsIn(HPO4)2 Rb2Ga4(HPO4)(PO4)4 . 0.5H2O the same green CsV2(PO4)(HPO4)2(H2O)2 pale-yellow CsIn2(PO4)(HPO4)2(H2O)2 KFe3(H2PO4)6(HPO4)2 .4H2O In2O3, KH2PO4, H3PO4, HF (K : In : P : F=4.8 : 1 : 6.2 : 3) c In2O3, NH4Cl, H3PO4 (NH4 : In:P=8:1 : 2) c In2O3, RbH2PO4, H3PO4, HF (Rb : In : P : F=4.65 : 1 : 6.15 : 3) c In2O3, CsF, H3PO4 (Cs : In:P=3: 1 : 6) c Ga2O3, 3M RbH2PO4, 3M H3PO4 (Rb : P=0.5) Cs4V2O7, Vmet, H3PO4, H2O, PhMe3NCl (1 : 2 : 46 : 800 : 1.1) InCl3, H3PO4, CsVO3, Me4NOH, H2O (2 : 6 : 1 : 1 : 125) Fe2O3, H3PO4, K2CO3, H2O (Fe : P :K=1 : 4 : 1) colourless hexagonal- prismatic Cs2Ga(H2PO4)(HPO4)2 .H3PO4 .colourless needle-like . 0.5H2O CsGa(H2PO4)4, Cs2CO3, H3PO4 (Cs2O:Ga2O3=7.5 : 1) a Slow cooling to 25 8C for 24 h; b the degree of filling of the autoclave was 30%; c the degree of filling of the autoclave was 75%; d slow cooling to 275 8C for 65 h; e slow cooling with a rate of 20 8C per hour. Table 5. Crystallographic characteristics of theMIMIII(HPO4)2 compounds. Crystal system Compound Space Z group monoclinic KAl(HPO4)2 .H2O P1 P1 R3c P1 P1 P21/c 8 KAl(DPO4)2 .D2O " P21/c 8 "triclinic "trigonal monoclinic triclinic monoclinic triclinic monoclinic """triclinic trigonal P21/c 4366 P21/c 43 P21/c 43 P21/c 4 P21/c 4 P21/c 4 P21/c 422 (H3O)Fe(HPO4)2 NH4(Al0.64Ga0.36)(HPO4)2 a-(NH4)Fe(HPO4)2 RbFe(HPO4)2 b-(NH4)V(HPO4)2 a-(NH4)V(HPO4)2 b-RbV(HPO4)2 a-RbV(HPO4)2 KIn(HPO4)2 (NH4)In(HPO4)2 b2-RbIn(HPO4)2 CsIn(HPO4)2 Na2In2(HPO4)4 .H2O Na4In8(HPO4)14(H2O)6 .12H2O P1 P3c1 L N Komissarova,MG Zhizhin, A A Filaretov Synthesis conditions pH 727777771.5 10.5 177777 7 1 64 Unit cell parameters c /AÊ b /AÊ a /AÊ b (or a, b, g) /deg 10.040 10.037 5.191 7.109 9.838 8.160 5.201 7.173 5.210 8.831 9.623 9.665 9.705 5.330 100.98 100.97 94.81 a=65.01, b=70.25, g=69.01 a=93.46, b=85.58, g=89.47 794.8 a=65.1, b=70.7, g=69.6 94.4 a=109.6, b=110.3, g=65.3 115.72 116.15 116.60 93.81 a=98.710, b=98.953, g=60.228 7 16.254 16.223 14.448 9.252 14.159 52.75 14.398 9.458 14.330 7.188 9.457 9.596 9.528 14.796 9.2685 18.493 9.107 9.105 8.748 8.695 7.185 78.738 8.841 8.789 9.450 8.257 8.276 8.367 9.157 9.3013 9.4976 13.850 7 Ref.T /8C t /h 65 720 145 a 59 96 130 67 60 168 96 227 230 61 48 240 61 48 240 61 44 240 57 57 69 168 336 50 80 130 125 56 144 200 56 56 144 144 200 200 56 96 200 58 40 600 d 62 48 200 63 48 200 1 75 290 e Ref.65 65 67 59 68 60 61 66 61 61 56 56 56 56 57 69Complex phosphates containing mono- and trivalent cations inside the channels. This arrangement of the protons has an effect on the shape of the channels and their presence compensates the negative charge of the frameworks of these compounds. Two types of the AlO6 octahedra and four crystallographi- cally independent PO4 tetrahedra are the main polyhedra in the structure of KAl(HPO4)2 .H2O.65 The alternating AlO6 andHPO4 polyhedra are linked in chains along the a and c axis thus forming a framework (Fig.6). The size and shape of the cavities in the framework are affected by the potassium cations and protons involved in the HPO4 groups. The Al(1)O6 octahedron, which is formed by four oxygen atoms belonging to four HPO4 groups and two oxygen atoms of two cis-oriented water molecules, is strongly distorted. The Al(2)O6 octahedron is less distorted and is formed only by the oxygen atoms of the HPO4 groups. The Al(1)O6 octahedra are arranged along the a axis, whereas the Al(2)O6 octahedra are aligned along the c axis. The structure contains two types of the K cations. The coordination sphere about the K(1) cation is formed by the oxygen atoms of only the HPO4 groups, whereas the coordination sphere about the K(2) cation involves the oxygen atom of the H2O molecule along with the oxygen atoms of the HPO4 groups.The coordination numbers of the potassium atoms in both polyhedra are equal to 9. In the structure of the deuterated analogue KAl(DPO4)2 .D2O,65 seven types of hydrogen bonds were revealed. a H c KAlO6 PO4 Figure 6. Structure of KAl(HPO4)2 .H2O projected onto the ac plane. The difference in the size of the MI and MIII cations has a substantial effect on the structural features of hydrogen phos- phates MIMIII(HPO4)2 (MI=K, Rb, Cs, H3O, NH4; MIII = Fe, V, In), which crystallise in the monoclinic system. This is partic- ularly clearly seen in the case of indium compounds. For example, the unit cell volume in the family of indium hydrogen phosphates MIIn(HPO4)2 (MI=K, Rb, NH4, space group P21/c) increases gradually in the series K±NH4 ± Rb.However, the unit cell parameters are sharply changed on going to caesium, which has the largest ionic radius of all the alkali metals under consideration, while the space group (P21/c) is retained. The CsIn(HPO4)2 compound is more similar in crystallographic characteristics to hydrogen phosphates (H3O)Fe(HPO4)2, b-(NH4)V(HPO4)2 and b-RbV(HPO4)2 than toMIIn(HPO4)2 (MI=K, Rb, NH4). In the structure of (H3O)Fe(HPO4)2,67 the FeO6 octahedra are linked to each other through four PO4 tetrahedra. The block of the octahedra and tetrahedra forms a framework consisting of columns along the a axis. The cavities in the framework are 627 occupied by the oxonium cations.} The oxygen atom of the H3O+ cation is at distances of 2.90 ± 3.14Afrom the seven nearest oxygen atoms of the HPO4 groups.The structures of hydrogen phosphates b-MIV(HPO4)2 (MI=NH4, Rb) 61 and CsIn(HPO4)2 (see Ref. 56) differ from that of (H3O)Fe(HPO4)2 (see Ref. 67) in that the VO6 (as well as the InO6) octahedra within the channels in these structures are linked to each other only by two PO4 tetrahedra. On going from b-RbV(HPO4)2 to CsIn(HPO4)2, the oxygen polyhedron about the MIII atom is enlarged. Thus, the coordination numbers of rubidium and caesium are 7 and 12, respectively. The triclinic modification ofMIMIII(HPO4)2 (MI=Rb,NH4; MIII=Fe, V) 61, 66 and (NH4)(Al17xGax)(HPO4)2 (x=0.36) 59 have mixed {MIII[HPO4]2}3? frameworks of the MIIIO6 octahe- dra and PO4 tetrahedra.These skeleton frameworks contain large cavities occupied by monovalent cations. The sizes and shapes of the cavities allow mobile ions to migrate in different directions, which indicates that these compounds can exhibit ion-exchange properties.68 The MIIIO6 octahedra share all vertices with six nearest PO4 tetrahedra. Each tetrahedron shares three vertices with three adjacent octahedra, whereas the fourth proton-bearing vertex of each tetrahedron is not involved in the first coordination sphere of theM3+ cations. The functional role of the fourth vertex consists in linking the tetrahedra through hydrogen bonding to form groups of three polyhedra each. The fact that the V3+, Fe3+, Co3+ and Ni3+ cations have similar radii [rVI(MIII)=0.64, 0.645, 0.61 and 0.60A, respectively] suggests the existence of the triclinic modification of the MIMIII(HPO4)2 compounds with MI=Rb, NH4 andMIII=Co, Ni.The (NH4)Fe(HPO4)2 compound 68 exhibits antiferromag- netic properties and has two different magnetic subunits analo- gous to those found in Li3Fe2(PO4)3.70 It should be noted that an increase in the ionic radius of the MIII cation in the series Al ±Ga ±V± Fe in hydrogen phosphates with composition (NH4)MIII(HPO4)2 is accompanied by a substantial deformation of the FeO6 octahedra resulting in rotation of the PO4 tetrahedra about FeO6. This leads to doubling of the unit cell volume of the Fe-containing analogue (triclinic modification). The triclinic modification of the In-containing analogue was not found, which may be indicative of instability of this modification of hydrogen phosphates containing large MIII cations.The monoclinic modification of MIMIII(HPO4)2 differs from their triclinic modification primarily in the mode of linking of the octahedra and tetrahedra in infinite columns along the a axis. For example, the columns in the structure of monoclinic b-(NH4)V(HPO4)2 (Fig. 7 a) are linked in a framework through the PO4 tetrahedra, whereas these columns in the more complex structure of triclinic a-(NH4)V(HPO4)2 (Fig. 7 b) are linked along the [100] direction due to the presence of an additional VO6 octahedron. As a result, the structure b-(NH4)V(HPO4)2 contains channels of one type occupied by the ordered NHá4 cations, whereas two types of channels occupied by the NHá4 ions are present in the structure of a-(NH4)V(HPO4)2 (see Fig.7 a, b). The structure of RbFe(HPO4)2 belonging to the trigonal system 60 (Fig. 7 c) has a fragment typical of the triclinic modifi- cation of the MIMIII(HPO4)2 compounds. However, in the low- symmetry structure, the tetrahedra through which the octahedra are linked in columns, are `sunk' into the cavities between the octahedra, whereas the arrangement of the tetrahedra around the FeO6 octahedra in the trigonal structure corresponds to the symmetry 3. of structure The contains the RbFe(HPO4)2 Fe(1)2Fe(2)(HPO4)12 trimers involving two types of the octa- hedra, viz., Fe(1)O6 and Fe(2)O6.The trimers form helical columns along the c axis. Each HPO4 group is linked to one } The presence of the H3O+ cation was confirmed by IR spectroscopy (a narrow absorption band at 3392 cm71).628 (NHá4 )N NHá4 NHá4 b Figure 7. Structures of monoclinic b-(NH4)V(HPO4)2 (a) and triclinic a-(NH4)V(HPO4)2 (b) projected onto the bc plane and the structure of trigonal RbFe(HPO4)2 (c) projected onto the ac plane. Fe(1)O6 octahedron and one Fe(2)O6 octahedron within the trimer and to one Fe(1)O6 octahedron from the adjacent trimer so that the three-dimensional [Fe(1)2Fe(2)(HPO4)6]3? framework is formed. The tunnels extended along the [100] and [110] directions are occupied by two types of rubidium cations with the coordination numbers of 12 and 9.As can be seen from the projection of the structure of RbFe(HPO4)2 onto the ab plane, the layers of the Fe(1) and Rb(1) atoms alternate with the layers of the Fe(2) and Rb(2) atoms and interleaved with the HPO4 groups. All iron and rubidium atoms are located on threefold axes to form columns along the c axis with the 7Fe(1)Fe(2)Fe(1)Rb(2)Rb(1). .Rb(2)Fe(1)7 alternation. In RbFe(HPO4)2, the difference between the radii of the rubidium [rXII(Rb+)=1.73A] and iron [rVI(Fe3+)=0.645A] cations is*1.09A. The pair of the caesium [rXII(Cs+)=1.88A] and indium [rVI(In3+)=0.76A] cations as well as the pair of rubidium and vanadium [rVI(V3+)=0.64A] are characterised by similar values of Dr and, hence, the trigonal modifications of CsIn(HPO4)2 and RbV(HPO4)2 would be expected to occur.A decrease in Dr(M+±M3+) in the series Cs ±Rb ±K± Na (for one type of theM3+ atoms) leads to a decrease in the coordination number of the alkali-metal cation. As a consequence, a new structural type can be formed. This assumption was confirmed by the structures of indium hydrogen phosphates MIIn(HPO4)2, a b c c PO4 VO6 VO6 PO4 b c whereMI=K,Rb orNH4, which we have established recently. In the framework structures of these compounds, the octahedra and tetrahedra are linked in infinite columns along the [100] and [001] directions. These structures differ from that of CsIn(HPO4)2 in the mode of linking of the columns to form a framework. In addition, the alkali-metal atoms in these structures adopt a smaller coordi- nation number.Thus the coordination number of caesium in CsIn(HPO4)2 is 12, whereas the coordination number of potas- sium and rubidium in MIIn(HPO4)2 (M=K or Rb) is 9. In CsIn(HPO4)2, each column is linked to four analogous adjacent columns through four PO4 tetrahedra, whereas there are eight such interactions in the structures of MIIn(HPO4)2 (M=K, Rb) and all tetrahedra are shared by columns (Fig. 8). The PO4 tetrahedra in MIIn(HPO4)2 (M=K or Rb) share two vertices with the InO6 octahedra from one column and one vertex with the octahedron from the adjacent column. The fourth (protonated) oxygen vertex of each tetrahedron is directed inside the channel formed with the participation of three columns linked through the tetrahedra.These channels are occupied by the MI cations. The InO6 octahedra in the structure of KIn(HPO4)2 , unlike those in RbIn(HPO4)2 , are strongly distorted. These structural features of KIn(HPO4)2 suggest that the formation of this structural type is highly improbable for the sodium analogue. L N Komissarova,MG Zhizhin, A A Filaretov c a Fe(1)O6 Fe(2)O6 Rb(1) Rb(2)Complex phosphates containing mono- and trivalent cations Rb InO6 b PO4 a Figure 8. Structure of RbIn(HPO4)2 projected onto the ab plane. Sodium for which Dr(M+±M3+) reaches the maximum value (*0.23A) forms the Na2In2(HPO4)4 .H2O57 and Na4In8. .(HPO4)14(H2O)6 . 12H2O69 compounds with radically different structures.The Na2In2(HPO4)4 .H2O compound has the [In2(HPO4)4]3? framework built from two types of the InO6 octahedra and PO4 tetrahedra. This framework can be described as consisting of the characteristic [In(1)In(2)(HPO4)4] structural blocks. The InO6 octahedra of both types each share vertices with six PO4 tetra- hedra, which, in turn, each have three oxygen atoms in common with three adjacent octahedra. The remaining vertices of the HPO4 polyhedra occupied by the protonated oxygen atoms are involved in formation of five hydrogen bonds. The octahedra and tetrahedra are packed in layers parallel to the a axis. The frame- work of Na2In2(HPO4)4 .H2Ocontains two types of channels. The channels of one type are extended along the [100] direction and occupied by sodium cations and H2O molecules (Fig.9 a). The channels of another type are extended along the [010] direction and are empty. It should be noted that the mode of binding of the In, P and O atoms in Na2In2(HPO4)4 .H2O57 is analogous to that observed in the structures of Al2O3 (corundum) and TiIVTiIV(HPO4)4.71 The structure of Na4In8(HPO4)14(H2O)6 . 12H2O consists of the [In8(HPO4)14(H2O)6] layers formed by the PO4 tetrahedra and the In(1)O6 and In(2)O5OW octahedra. The [In8(HPO4)14(H2O)6] layers are packed in the ABAB order according to the trigonal symmetry P3c1, the channels along the [001] direction being lacking. The layers are linked to each other through the In(1)O6 octahedra to form a framework with channels along the crystallo- graphic a and b axis (Fig.9 b). These channels have a 12-mem- bered-ring shape in cross-section and include the Na+ cations and water molecules. The water molecules fulfil two functions. Some water molecules are involved in the first coordination sphere about the In(2) cation and the remaining water molecules occupy the channels of the structure (zeolite-type water). The partial occupancies of the sodium positions and the nearest oxygen positions are directly associated with the presence of abnormally short bonds (2.06 ± 2.19A) between these atoms. The Na7O interatomic distances in the NaOx polyhedron are close to the interatomic distances outside these polyhedra. This characteristic feature of the Na4In8(HPO4)14(H2O)6 .12H2O compound is also common to some hydrated sodium zeolites.72 III The second group of complex hydrogen phosphates of the general formula MIxMy ÖHPO4ÜzÖPO4Üu . nH2O (MI=Rb, Cs; MIII=V, Ga, In; n=0.5 ± 2) includes three compounds, viz., Rb2Ga4(HPO4)(PO4)4 . 0.5H2O, CsIn2(HPO4)2(PO4)(H2O)2 and a c b b a c Figure 9. Structure of Na2In2(HPO4)4 .H2O projected onto the bc plane (a) and structure of Na4In8(HPO4)14(H2O)6 . 12H2O projected onto the ac plane (b). CsV2(HPO4)2(PO4)(H2O)2. These compounds crystallise in two structural types (Table 6). In monoclinic Rb2Ga4(HPO4)(PO4)4 . 0.5H2O, 58 the GaOx, HPO4 and PO4 polyhedra form an open-type structure with channels containing the rubidium cations and water molecules (Fig.10 a). The presence of the HPO4 groups and water of crystallisation in the structure was confirmed by IR spectroscopy. The Rb(1) and Rb(2) cations are located in the MIO9 and MIO10 polyhedra, respectively. Four crystallographically non-equivalent gallium atoms have different coordination environments (an octahedron, two tetrahedra and a trigonal bipyramid). All GaOx polyhedra share vertices with the phosphate tetrahedra, which, in turn, are coordinated (except for the OH vertices) by the gallium atoms. Two types of channels are formed along the [100] direction. The protonated oxygen vertices of the PO4 tetrahedra are directed inside the channels of the first type occupied by the Rb(1) cations. The channels of the second type contain the Rb(2) cations and H2O molecules.The minimum O_O distances determining the sizes of the channels of the first and second types are 5.34 and 5.43A, respectively. The formation of an analogous structure is also highly probable for the RbAl analogue because the coordi- nation numbers of 6, 5 and 4 are typical of aluminium. The CsIn2(HPO4)2(PO4)(H2O)2 compound (centrosymmetri- cal structure) 63 has a rather complex structure. The octahedral environment about the In atoms is formed by five oxygen atoms of 629 H2O Na Na630 III Table 6. Crystallographic characteristics of theMIxMy (HPO4)z(PO4)u . nH2O andMIxMIII y (H2PO4)z(HPO4Uu . nH2O compounds. Crystal system Compound Space group monoclinic "" Rb2Ga4(HPO4)(PO4)4 .0.5H2O CsV2(HPO4)2(PO4)(H2O)2 CsIn2(HPO4)2(PO4) (H2O)2 (H3O)Al3(H2PO4)6(HPO4)2 .4H2O " (H3O)Fe3(H2PO4)6(HPO4)2 .4H2O "trigonal P21 P21/a P21/c C2/c C2/c KFe3(H2PO4)6(HPO4)2 .4H2O " C2/c C2/c P31c P31c (NH4)Fe3(H2PO4)6(HPO4)2 .4H2O " (NH4)0.33Al0.89(H2PO4)(HPO4) .2H2O (NH4)Fe3(H2PO4)3[PO3(OH)0.67O0.33]3 .6H2O " the PO4 tetrahedra and one oxygen atom of the water molecule. The HPO4 polyhedra share three vertices with the adjacent octahedra and their fourth protonated vertex is not involved in the first coordination sphere of the In3+ cation. The PO4 tetrahedron shares all vertices with the nearest InO6 octahedra. The InO6, PO4 and HPO4 polyhedra are linked to each other to form one-dimensional ribbons along the c axis (Fig.10 b). These ribbons are linked in a framework through the HPO4 groups. The large cavities in the framework are occupied by caesium cations (the coordination number of caesium is 11). Hydrogen phosphate CsV2(HPO4)2(PO4)(H2O)2 has an anal- ogous structure.62 However, the coordination polyhedron of caesium is enlarged (the coordination number of caesium is 13) due to the smaller ionic radius of vanadium. In this structure, the interactions between the ribbons somewhat resemble the inter- actions in the well-known structural type MVOPO4 with the difference that the MVO component is replaced by MIIIOH2 and additional phosphate groups are present between the ribbons. Taking into account the difference in the size of the vanadium and indium cations, it is reasonable to suggest that the structures characterised by an analogous topology of the framework must exist for the CsFe- and CsSc-containing compounds because the ionic radii of Fe and Sc are in the range between rVI(V3+) and rVI(In3+).a GaO5 GaO6 Ow Rb(2) Rb(1) GaO4 b c PO4 Figure 10. Structures of Rb2Ga4(HPO4)(PO4)4 . 0.5H2O (a) and CsIn2(HPO4)2(PO4)(H2O)2 (b) projected onto the bc plane. L N Komissarova,MG Zhizhin, A A Filaretov Z Ref. Unit cell parameters c /AE b /AE a /AE b /deg 58 62 63 73 74 75 76 77 78 8.207 6.370 10.180 17.126 17.609 17.50 17.647 16.488 16.862 5.061 10.004 6.580 16.722 16.797 17.01 16.845 8.905 9.151 244444462 91.8 97.0 97.9 90.9 90.6 90.9 90.9 77 21.643 17.812 18.092 9.437 9.528 9.604 9.611 77 III 4 The third type of complex hydrogen phosphates includes compounds of the general formula MIxMy OH2PO4UzOHPO4Uu .nH2O (MI=K, H3O, NH4; MIII=Al, Fe; n=2 ¡¾ 6) containing two types of hydrogen phosphate anions, viz., HPO2¢§ and H2PO¢§4 . The crystallographic characteristics of the monoclinic and trigonal structural types of these compounds are given in Table 6. These compounds are characterised by the ratio MI:MIII=1 : 3. Apparently, their stability is associated with the substantial difference in the radii of theMI andMIII cations. Hydrogen phosphates with composition MIMIII 3 OH2PO4U6..OHPO4U2 .4H2O (MI=K, H3O, NH4) 73 ¡¾ 77 and the ratio H2PO4 :HPO4=3 : 1 were obtained for aluminium and iron. In the structures of these compounds, water of crystallisation is not involved in the coordination sphere about the MIII cations. The structure of monoclinic phosphate with composition KFe3(H2PO4)6(HPO4)2 .4H2O (Fig. 11 a) 75 is characterised by a pronounced layered motif. The layers are composed of character- istic structural blocks, which are built from two different FeO6 octahedra linked through the phosphate groups of two types, viz., [HPO4]27 and [H2PO4]7 (Fig. 11 b). The interlayer space is occupied by water molecules. The layers are related by the symmetry elements of the crystal lattice (space group C2/c) so that the channels are not formed along the [001] direction.The layers contain cavities occupied by potassium atoms to form the b Cs InO6 PO4 b cComplex phosphates containing mono- and trivalent cations a b a c b b a Figure 11. Structure of KFe3(H2PO4)6(HPO4)2 .4H2O projected along the [010] direction (a) and projected onto the ab plane (b). KO12 polyhedra. The layers are linked through hydrogen bonds between the oxygen atoms of the phosphate groups and water molecules, which impart additional stability to the structure. Taking into account the data on the structures of the MIFe3(H2- PO4)6(HPO4)2 .4H2O compounds, where MI=K, H3O or NH4, and of their (H3O)Al analogue, compounds with such a compo- sition would be expected to exist also forMIII=Ga or V.The (NH4)0.33Al0.89(H2PO4)(HPO4) .2H2O (see Ref. 77) and (NH4)Fe3(H2PO4)3[PO3(OH)0.67O0.33]3 .6H2O (see Ref. 78) com- pounds are characterised by the presence of water molecules fulfilling different functions. The structures of these trigonal hydrogen phosphates differ in the number of formula units per unit cell (6 and 2, respectively).77, 78 Three crystallographically independentMIII cations occupy the virtually non-distorted MO6 octahedra in which theMIII7O distance increases in parallel with the ionic radius of MIII. The structural formula (NH4)0.33Al0.89. .(H2PO4)(HPO4)(H2O) .H2O reflects the different role of the water molecules in the crystal motif of this compound. Three independent OH groups belonging to the HPO4 [P(1)O4 tetra- hedron] and H2PO4 [P(2)O4 tetrahedron] groups and two water molecules (coordinated to the Al atom and the NH4 group) participate in hydrogen bonding.The NHa4 ions are surrounded by nine oxygen atoms most of which belong to the OH groups or H2O molecules. Trigonal NH4Al and NH4Fe phosphates have framework structures. In the crystal structures, the MIII(2)O6 and MIII(3)O6 octahedra are located on threefold axes and share all their vertices with the phosphate tetrahedra. The MIII(1)O6 octahedron has three vertices in common with the PO4 tetrahedra and the remaining three vertices are occupied by the oxygen atoms of water molecules. In the structure of NH4Fe phosphate (Fig. 12 a,b),78 the position of the protonated oxygen atom in the P(1) tetrahedron is statistically occupied by the hydrogen atom (OH)2/3O1/3. H2O K a a c �¢Ow �¢NH4 Figure 12.Structure of (NH4)Fe3(H2PO4)[PO3(OH)0.67O0.33]3 .6H2O projected onto the ac plane (a) and the structural block of two octa- hedra (b). The structural block consisting of two octaa, which are linked to nine tetrahedra through vertices (Fig. 12 b), is present in the crystal structures of many polytypic minerals, for example, of coquimbite, paracoquimbite and ferrinatrite Na3Fe(SO4)3(H2O)3. In the latter compound, such fragments formed by the Fe octahedra and SO4 tetrahedra are linked in chains parallel to the c axis. These chains are `threaded' on threefold axes of the trigonal unit cell (space group P3).Hence, the same repeated fragment is present in the crystal motifs of different classes of compounds. The degree of polymerisation of this fragment determines the struc- tural diversity of compounds. It should be noted that if the octahedra with the coordinates (1/3; 2/3; z) located on a threefold axis are linked through tetrahedra to analogous octahedra with the coordinates (2/3; 1/3; z) located on another threefold axis, a large number of structures analogous to NH4Al and NH4Fe phosphates are generated. These structures can be assigned to zeolite-type compounds. Compounds in which the MI-cation positions are vacant can also have analogous structures. In this case, the crystal-chemical formula can be written as &MIII(H2PO4)(HPO4)(H2O) .H2O.77 The fourth type of hydrogen phosphates, viz., MI IIIOH2PO4UOHPO4U2 .H3PO4 .0.5H2O, is known only for 2M gallium.64 This compound has a unique composition because it contains protonated phosphate groups of all three types, viz., HPO2¢§ 4 , H2PO¢§4 and H3PO4. The Cs2Ga(H2PO4)(HPO4)2 . H3PO4 . 0.5H2O compound crystallises in the tetragonal system with a rare non-centrosymmetrical unit cell (space group I41cd, a=20.168, c=18.287A, Z=16). In this compound, the gallium atom is in a slightly distorted octahedral environment formed by four oxygen atoms of the HPO4 groups and two oxygen atoms of the H2PO4 groups (Fig. 13 a). One H2PO4 group, two HPO4 groups and the GaO6 octahedron are involved in the typical [Ga(H2PO4)(HPO4)2]2¢§ ? structural block.These blocks are linked in infinite columns along the [001] direction with the identity period involving four GaO6 octahedra (due to the presence of a screw axis 41, Fig. 13 b). Each column is linked to four adjacent columns through hydrogen bonds involving theH3PO4 molecules. The crystal lattice contains large cavities occupied by water molecules and caesium cations. The structure contains two types of Cs atoms. The strongly distorted Cs(2)O11 polyhedron is formed by the oxygen atoms of the phosphate groups of all three types and by the H2O molecule, which is involved in formation of 631 b632 Cs(2) Cs(1) H2Ob a Figure 13. Structure of Cs2Ga(H2PO4)(HPO4)2 . H3PO4 . 0.5H2Oprojected onto the ab plane (a) and the fragment of the [Ga(H2PO4)(HPO4)2]?column in combination with the H3PO4 and H2O molecules and Cs+ cations (b).the Cs(2)7OW7Cs(2) chains. The Cs(1)O9 polyhedra are less distorted and are not linked to the water molecules. Analysis of the interatomic distances in the polyhedra of the Cs2Ga(H2PO4)(HPO4)2 .H3PO4 . 0.5H2O structure showed that the gallium position can, in principle, be occupied by the Al, Fe and V atoms whose ionic radii are not much different from that of gallium. These replacements can lead to distortion of the unit-cell symmetry but the main [MIII(H2PO4)(HPO4)2]2¡ ? structural block must persist. By contrast, the replacement of caesium with rubidium or potassium must lead to degradation of the structure due to the tendency of smaller cations to have smaller coordina- tion polyhedra.The above-considered structural features of hydrogen phos- phates containing mono- and trivalent cations, viz., the presence of the HPO4 and H2PO4 groups and water molecules and the mode of their binding, are responsible both for the temperature range of stability of hydrogen phosphates and the possibility of their use in the synthesis of other classes of compounds. decomposition Thermal of MIMIII(HPO4)2 . nH2O (MI=Na, K, Rb, Cs) is accompanied by the formation of pyrophosphatesMIMIIIP2O7. Decomposition of Table 7. Characterisation of different structural types of hydrogen phosphates containingMI and MIII cations. Type of compound III MIxMy (H2PO4)z(HPO4)u .nH2O 1:3 H2 PO4 :HPO4=3 : 1 monoclinic MI=K,NH4, H3O;MIII=Al, Fe III MIxMy (HPO4)z(PO4)u . nH2O MI=Rb, Cs;MIII=Ga, In, V III MIxMy (HPO4)z . nH2O MI=Na;MIII=In III MIxMy (HPO4)z . nH2O MI=Na, K, Rb, Cs, NH4, H3O; MIII=Al, Ga, In, V, Fe III MIxMy (H2PO4)z(HPO4)u(H3PO4)m . 2 : 1 nH2O (MI=Cs;MIII=Ga) a Ribbons, b columns, c helical columns. a hydrogen phosphates Ratio Type and ratio MI:MIII of phosphate groups 1 : 2 1 : 2 H2PO4 :HPO4=1 : 1 HPO4 : PO4=2 : 1 HPO4 : PO4=1:4 HPO4 1 : 1 HPO4 H2PO4 :HPO4 : :H3PO4=1 : 2 : 1 L N Komissarova,MG Zhizhin, A A Filaretov H2O Cs(1) H3PO4 Cs(2) c a RbFe(HPO4)2 is accompanied by the removal of the water molecule in the temperature range of 350 ± 550 8C to give crystal- line RbFeP2O7.Thermal decomposition of MIIn(HPO4)2 (MI=K, Rb, Cs) starts at about 280 8C and completes at 550 ± 650 8C to form MIInP2O7. Decomposition of (NH4)In. .(HPO4)2 at 800 8C affords a mixture of In4(P2O7)3 and In(PO3)3. Thermal decomposition of (H3O)Fe(HPO4)2 proceeds in two steps. In the first step (at 670 8C), H2O is removed and hydrogen pyrophosphate is formed. The latter decomposes at 870 8C to give Fe2P2O7 and Fe(PO3)2 . Heating of Na2In2. .(HPO4)4 .H2O results in the loss of a water molecule (*365 8C) followed by crystallisation of a non-identified phase, which melts at 1208 8C. Thermal decomposition of hydrogen phosphates of the sec- ond and third types is a complex process and does not afford single-phase products.In the case of Rb2Ga4(HPO4)(PO4)4 . 0.5H2O, the removal of water of crystallisation followed by dehydration was observed at 250 ± 600 8C. Dehydration of (NH4)0.33Al0.89(H2PO4)(HPO4) .2H2O proceeded successively. First, the water molecules and ammonium ions were eliminated at 320 8C and then the hydro- and dihydrogen phosphate anions Crystal system Structural type Space Structural blocks group [MIIIO6]+[HPO4] and [H2PO4] [MIIIO5w]+[HPO4] and [H2PO4] layered framework " C2/c P31c P21/c [MIIIO5w]+[HPO4] and [PO4] a P21/c [MIIIO4,5,6] +[HPO4] and [PO4] " P3c1 trigonal monoclinic "trigonal [InO6] and [InO5Ow]+[HPO4] " P1 [ IIIO6]+[HPO4] b M[MIIIO6]+[HPO4] c triclinic monoclinic trigonal tetragonal P21/c R3c I4 """chain 1cd [MIIIO6]+[HPO4], [H2PO4] and [H3PO4]Complex phosphates containing mono- and trivalent cations decomposed.In the final step, AlPO4 crystallised (tridymite modification). A comparison of different procedures for the synthesis and a knowledge of the structural features of hydrogen phosphates containing mono- and trivalent cations lead to the conclusion that the steric factors play the major role in the formation of different types of complex hydrogen phosphates (Table 7). The presence of various organic additives and the use (in some cases) of fluoride ions as a mineraliser have less substantial effects on the compositions of the resulting compounds as compared to the influence of the MI:MIII ratio and the difference in the sizes of these cations.The effect of theMI andMIII cations on stability of complex hydrogen phosphates in combination with the use of particular conditions for the formation of various protonated forms of phosphate groups in aqueous solutions provide the basis for the directed synthesis of new compounds. IV. Fluoride phosphates containing mono- and trivalent cations 4 Depending on the ratios of the cations (MI and MIII) and anions (F7 and PO3¡ 4 ), fluoride phosphates containing mono- and trivalent cations form several groups of compounds with compo- sitions MIMIIIFPO4, MI3MIII 2 F3(PO4)2, MI5MIIIF2(PO4)2 and (MI,MII,MIII)5F(PO4)3 (MI=Li, Na, K, NH4; MII are alkali- earth elements; MIII are p-, d- or f-elements).Four types of REE sodium fluoride phosphates with different compositions were found in studies of the phase formation in the ternary systems Na+,MIII//F7,PO3¡ (MIII=La, Nd, Gd, Er) by powder X-ray diffraction analysis.79, 80 Minerals belonging to fluoride phos- phates are known, in particular, amblygonite LiAlFPO4 ,81, 82 tavorite LiFe(OH,F)PO4 (see Ref. 83) and lacroixite NaAlFPO4 (see Ref. 84) as well as apatite-like deloneite NaCa3CeF(PO4)3 and belovite NaSr3CeF(PO4)3. In some fluoride phosphates, the fluoride ions are partially replaced with the hydroxide ions of a similar size. The OH:F ratio may vary; cf., Na3Fe2[(OH)2F](PO4)2 (see Ref. 85) and MIMIIIF17d(OH)dPO4 (MI=K, Rb, NH4; MIII=Al, Ga, In; d&0.3, 0.5).86 1.Synthesis III Fluoride phosphates of different types, viz., MIMIIIFPO4 (MI=Na, K, NH4; MIII=Al, Cr, Fe, Ga, In, Y, Gd± Ho), Na3Fe2[(OH)2F](PO4)2, Na 2 F3(PO4)2 (MIII=Al, V, Cr, Fe, 3M Ga), Na5MIIIF2(PO4)2 (MIII=Al, Cr, Ga), and (Na,MII,MIII)5. .F(PO4)3 (MII are alkali-earth metals; MIII are REE), were synthesised in the solid state by crystallisation from flux or by the hydrothermal method. The NaMIIIFPO4 compounds (MIII=Gd, Tb, Dy) were prepared by heating stoichiometric amounts of MIIIPO4 (MIII=Gd, Tb, Dy) and NaF in air at 550 8C (MIII=Gd, Tb) or 500 8C (MIII=Dy). Potassium fluoride phosphates KMIIIFPO4 (MIII=Cr, Fe) were prepared by the reactions of chromium (or iron) oxide withK2O±P2O5 melts in the presence of KF at 600 ± 800 8C.In pure form, these compounds are formed in theK2O:P2O5 range from 1 : 1 to 1 : 1.7 (MIII=Cr) or from 1 : 1.5 to 1 : 2 (MIII=Fe); otherwise the reactions afford the correspond- ing pyrophosphate KMIIIP2O7. III The solid-phase synthesis of fluoride phosphates Na3M2 F3(PO4)2 containing aluminium, chromium or iron was carried out with the use of two-, three- or four-component mixtures (NaF ±MIIIPO4, MIIIF3±MIIIPO4±Na3PO4, NaF ± MIIIPO4 ± NaCl ± ZnCl2, Na2 O±MIII 2 O3 ±P2O5 ± NaF).85, 87 ± 90 The reactions were carried out under an atmosphere of argon at T=600 ± 880 8C. Aluminium (iron) sodium fluoride phosphates were prepared by recrystallisation of fluoride phosphate glasses formed upon heating of the NaPO3±MIIIF3 or NaF ±MIIIPO4 ± MIIIF3 mixtures (MIII=Al, Fe) above the melting point followed by cooling in air to the crystallisation temperature.Aluminium- containing glass characterised by a large difference between the 633 4 crystallisation and glass-transition temperatures is more stable than iron-containing glass. Of fluoride phosphates of the MI IIIF2(PO4)2 family 5M (MIII=Al, Cr, Fe, Ga), only sodium-containing derivatives were prepared. The solid-phase synthesis of most of compounds with this composition was performed in the MIII 2 O3 ±Na2O± P2O5 ± NaF (or NaCl) system (MIII=Cr, Fe, Ga) at T&750 ± 1300 8C by crystallisation from flux.91 ± 95 The alumi- nium derivative was prepared according to two procedures, viz., by the solid-phase reaction in the AlF3 ± Al(OH)3 ±NaH2PO4 ± Na3PO4 system at T=1000 8C and precipitation from a solu- tion.96 In the latter case, AlF3 .3H2O was added to an aqueous solution of NaH2PO4 at 80 8C.The suspension was kept at T=60 8C for several days after which crystallisation started. Mixed sodium- and REE-containing fluoride phosphates were prepared with the use of two-, three- or four-component Na+,MIII//F7,PO3¡ mixtures (MIII=La, Nd, Gd, Er). The mixtures were annealed under an inert or reductive atmosphere and also in vacuo at a temperature from 750 to 1150 8C.79, 80 When synthesising REE fluoride phosphates in air, the duration of annealing of the samples must be rigidly controlled because the time of annealing has a substantial effect on the composition of the final product.Thus, annealing of samples for 12 ± 18 h (750 8C) gave rise to fluoride phosphate with specified composi- tion. More prolonged heating (72 h) afforded REE oxofluorides and double phosphates Na3MIII(PO4)2 (MIII=REE). Apatite-like REE fluoride phosphates were synthesised by the III solid-phase method fromMIICO3 (MII=Ca, Sr, Ba), ammonium hydrogen phosphate, REE oxide and an excess of NaF at *1200 8C.97 ± 99 Phases with variable composition LixMx Y1¡xF4x(PO4)1¡x (MIII=Gd ± Lu, Y; 04 x40.45) were prepared by heating stoichiometric amounts of double fluoride LiMIIIF4 and yttrium orthophosphate at 800 8C.100 Fluoride phosphates with different compositions can be obtained in the multicomponent MIF±MIII 2 O3 ±H3PO4±H2O systems (MI=Li, Na, K, NH4; MIII = Al, Fe, Ga, In, Y, Ho, Er) under the hydrothermal conditions.83, 85, 86, 101, 102 For exam- ple, LiFe[(OH),F]PO4 (synthetic mineral tavorite) 83 and ana- logues of natural iron phosphates, such as triphylite LiFePO4, sarcopside Fe3(PO4)2 and zwieselite Fe2FPO4,102 were obtained from the LiF ± Fe2O3±H3PO4±H2O system at &300 ± 500 8C and p&1.5 kbar.Attempts to prepare gallium potassium fluoride phosphate KGaFPO4, unlike analogous aluminium and iron compounds, failed although different modes of the solid-phase synthesis (Ga2O3±NH4H2PO4 ±KF mixture) and crystallisation from flux (anhydrous Ga2O3 ±KPO3 ±KF system) were used. Compounds in which the fluoride ions are partially replaced by the hydroxide groups, viz., MIMIIIF17d(OH)dPO4 (MI=K, Rb, NH4; MIII=Al, Ga, In; d&0.3; 0.5) 86, 103, 104 and Na3Fe2[(OH)2F](PO4)2,85 were prepared only under the hydro- thermal conditions.After the successful synthesis of a series of porous crystalline III aluminosilicates (zeolites) and aluminophosphates, attempts were made to prepare porous fluoride phosphates of the general formula AaMx Fy(PO4)z (MIII=Ga, Fe, In; A is organic amine). It was expected that these compounds would find use as catalysts, molecular sieves or ion exchangers. In studies of the phase formation in the NH4F±MIII 2 O3 ±H3PO4 ± amine ±H2O (MIII=Ga, Fe) and NH4F±CeIV(SO4)2±H3PO4 ± amine ±H2O systems at T=180 8C and p=18 bar, ammonium-containing fluoride phosphates NH4MIIIFPO4 (MIII=Ga, Fe) 105, 106 and NH4CeIVF2PO4 (see Ref.107) were obtained. When Ga2O3 was used, ammonium-containing fluoride phosphate with unusual composition (NH4)2Ga2F3(PO4)(HPO4) was obtained along with NH4GaFPO4.106 III Polycrystalline samples of the Na3M2 F3ÖPO4Ü2 compounds (MIII=V, Ga) were prepared under low-temperature hydro- thermal conditions (T=180 8C, p=18 bar). However, the final products were contaminated by intermediates produced in the synthesis. The procedure used for the synthesis of sodium vana-634 dium fluoride phosphate Na3V2F3(PO4)2 (see Ref. 87) differs from a conventional method. Vanadium oxide V2O5 was dissolved in phosphoric acid and reduced with an aqueous solution of N2H4 until the solution turned green (V3+ ions) and then NaF was added (the molar ratio Na :V=4.3 : 2).The resulting solution was heated first at 180 8C and 18 bar and then at 700 8C and 2 kbar after which it was slowly cooled to room temperature with a rate of 10 8C per hour to obtain finally fluoride phosphate Na3V2F3(PO4)2. The Na3Ga2F3(PO4)2 compound 87 was not iso- lated in pure form (it contained impurities of GaPO4 and GaF3). Aluminium fluoride phosphatesK3Al4F9(PO4)2 (see Ref. 108) and (NH4)Al2F(PO4)2 (see Ref. 109) were obtained by the hydro- thermal method as by-products in the studies of the MIF± MIII 2 O3 ±H3PO47H2O systems. 2. Structure and properties Sodium-containing fluoride phosphates with composition NaMIIIFPO4 are known only for MIII=Al, REE.In the struc- tures of the NaMIIIFPO4 compounds (MIII=Y, Gd ± Er), 110, 111 the MIII atoms are located in the [MIIIO6F2] eight-vertex poly- hedra (dodecahedra), the F atoms being in cis positions. The MIIIO6F2 polyhedra are linked in infinite chains through common edges composed of two oxygen atoms. Every second polyhedron of one chain shares the edge composed of two fluorine atoms with the polyhedron form the adjacent chain giving rise to layers. These parallel layers are linked to each other through the PO4 tetrahedra to form a framework (Fig. 14). The channels between the layers are occupied by sodium cations. The NaO8F2 coordination polyhedron is most preferable for sodium. The structural motif in the NaMIIIFPO4 compounds remains unchanged on going from Gd to Er and only the PO4 tetrahedra are slightly rotated. MIIIO6F2 PO4 a c Na b Figure 14.Fragment of the structure of NaMIIIFPO4 (MIII=Y, Gd ± Er). Considering the crystal structures of this type of compounds and taking into account that REE of the second half of the series (Gd ± Er) tend to adopt high coordination numbers (>7), the formation of analogous sodium-containing fluoride phosphates would be highly improbable in the case of small trivalent REE cations (Yb, Lu) as well as in the case of In and Sc cations for which the coordination number of 6 is most typical. The exception is the mineral lacroixite NaAlFPO4. 84 There is evidence for the formation of REE sodium fluoride phosphates with composition III Na3M2 F3ÖPO4Ü2 (MIII=Tm, Yb, Lu) } but the structures of these compounds remain to be established.The NaMIIIFPO4 compounds (MIII=Gd ± Er) melt incon- gruently. According to the DTA data, their melting points decrease in the series Gd± Er from 903 to 837 8C. In the case of stepwise annealing in air in the temperature range of 500 ± 850 8C, decomposition started already at 600 8C, proceeded through the intermediate formation of MIIIOF, Na3PO4 and MIIIPO4 and } Our data. L N Komissarova,MG Zhizhin, A A Filaretov afforded finally a mixture of double phosphate Na3MIII(PO4)2 and oxideMIII 2 O3. Potassium-containing fluoride phosphates KMIIIFPO4 are known for MIII=Al, Cr, Fe, In. In the crystal structure of KAlFPO4,101, 112 the aluminium atoms are located in the [AlO4F2] octahedra of two types, which differ in the positions of the F atoms (trans and cis).The structure of KInFPO4, unlike KAlFPO4, contains octahedra of only one type with a cis arrange- ment of the F atoms. In bothKMIIIFPO4 compounds (MIII=Al, In), the MIIIO4F2 octahedra are linked in parallel helical chains through fluoride bridges and these chains are linked to each other to form a framework. Fluoride phosphatesMIMIIIFPO4 (MIII=Fe, Cr, Ga) 113 ± 116 containing the large monovalentK+ orNHá4 cations are isostruc- tural to potassium titanyl phosphate KTiOPO4 (KTP).116 In the structures of MIMIIIFPO4 with MI=K+ or NHá4 , the PO4 tetrahedra and MIIIO4F2 octahedra are less distorted than the TiO6 octahedra in KTiOPO4. TheMIIIO4F2 octahedra are linked in chains extended in two mutually perpendicular directions (Fig.15). These chains are linked to each other through the PO4 tetrahedra to form a framework. The chains differ in that the PO4 tetrahedra in these chains are oriented in opposite directions. The cavities in the framework are occupied by the potassium or ammonium cations. a b c Figure 15. Chains of the MIIIO4F2 octahedra in the structure of MIMIIIFPO4 (MI=K, NH4, MIII=Cr, Fe, Ga). Two types of potassium atoms in the structures ofKMIIIFPO4 are in a complicated coordination environment formed by fluo- rine and oxygen atoms. The coordination numbers of the K(1) and K(2) atoms are 8 and 9, respectively.The structure of KTiOPO4 contains K atoms of one type with the coordination number of 9. The coordination polyhedra about the potassium atoms have one edge (F7O) in common. The cavities occupied by the potassium atoms are linked in pairs and form two types of helical channels perpendicular to each other. The crystallographic characteristics of theMIMIIIFPO4 compounds are given in Table 8. In recent years, the structural features and physical properties of KTP-type compounds have attracted considerable attention, which is associated with a search for new promising nonlinear optical materials. Taking into account the results of dielectric measurements and studies of conductivity, it was suggested that the structures of KTP-type compounds are changed at 0 8C due to disordering in the potassium sublattice.An analogous phenom- enon was found for fluoride phosphates KMIIIFPO4 (MIII=Al, Fe). Like in the structure of KTiOPO4, disordering in the potassium sublattice was observed in the structure of KFeFPO4. As a result, two additional positions are present along with two statistically occupied potassium positions. The occupancies of these four positions are 0.9, 0.92, 0.1 and 0.08. At a temperature of about7100 8C, the K atoms are partially ordered, two main and one additional positions with the occupancies of 1.0, 0.94 and 0.06, respectively, being retained. Disordering in the potassium sublattice disappeared as the temperature was lowered toComplex phosphates containing mono- and trivalent cations Table 8.Crystallographic characteristics of theMIMIIIFPO4 andMIMIIIF17d(OH)dPO4 compounds. Crystal system Space group Compound (mineral name) P1 C1 P1 LiAlFPO4 (amblygonite) triclinic "" C2/c C2/m C2/m monoclinic "" LiFe[(OH),F]PO4 (tavorite) NaAlFPO4 (lacroixite) NaYFPO4 NaMIIIFPO4 (MIII=Tb ± Er) KAlFPO4 KCrFPO4 KFeFPO4 orthorhombic """"tetragonal """orthorhombic Pn21a Pnna Pc21n P21nb Pna21 P43212 P43212 P43212 P43212 P212121 KInFPO4 KInF0.7(OH)0.3PO4 RbInF0.7(OH)0.3PO4 NH4InF0.7(OH)0.3PO4 (NH4)0.88(H3O)0.12 . . AlF0.67(OH)0.33PO4 " Pna21 8 12.717 " P21 2121 8 9.593 KGaF0.7(OH)0.3PO4 (NH4)0.93(H3O)0.07 . .GaF0.5(OH)0.5PO4 "" Pna21 Pna21 (NH4)FeFPO4 (NH4)GaFPO4 a Our data. 7173 8C.113, 114 Apparently, a more detailed study of this phe- nomenon will allow one to reveal correlations between the ionic conductivity and nonlinear optical properties of these com- pounds.The KMIIIFPO4 compounds (MIII=Cr, Fe) exhibit high thermal stability. The chromium compound is more stable and it melts congruently at a temperature above 1600 8C. For fluoride phosphate KFeFPO4, a phase transition was found at T=850 8C. At T>1150 8C, this compound decomposes with the loss of oxygen to form complex iron(II) phosphates, viz., KFePO4 and K2Fe2P2O7F2 (as black melts). The formation of complex fluoride hydroxide phosphates with composition MIMIII(OH)dF17dPO4 is typical of such triva- lent elements as Al, Fe, Ga and In in combination with lithium and large monovalent cations (K+, Rb+ and NHá4 ) (see Table 8). As examples we refer to synthetic tavorite LiFe[(OH),F]PO4 (see Ref.83) and the natural mineral amblygonite LiAl[(OH),F]PO4.81, 82 These compounds have framework struc- tures. The PO4 tetrahedra and MIIIO6 octahedra are the main structural elements of tavorite and amblygonite. These polyhedra are linked in chains along the [101] direction through the common (OH, F) vertices. These chains are linked in a three-dimensional framework through the PO4 tetrahedra (Fig. 16 a). The frame- work is penetrated by a series of rather broad channels occupied by lithium atoms. The environment about the lithium atom in the structures of tavorite and amblygonite deserves attention (Fig. 16 b).Thus, the structure of natural amblygonite contains two lithium positions with occupancies of 1/2, which are separated by a distance of 0.5A. These half-occupied positions are located in face-sharing five-vertex polyhedra which together form a distorted octahedron. In the structure of montebrasite (OH-containing analogue of amblygonite), the nearest environment about the lithium atoms { is formed by five oxygen atoms. This is attributed to the fact that { The lithium atoms statistically occupy two positions separated by a distance of 0.24A. Z Unit cell parameters a /AÊ 5.060 6.644 5.138 2424448 6.414 8.944 9.000 (Tb) ± 8.943 (Er) 8 12.522 8 12.612 6.346 8 10.656 8 12.847 9.351 9.345 9.437 9.447 9.416 888888 12.993 8 12.921 b /AÊ 5.160 7.744 5.307 8.207 6.930 6.923 (Er) 10.149 10.172 10.555 12.885 6.353 77779.563 115.47 106.11 6.985 (Tb) ± 6.506 (Tb) ± 105.98 (Tb) ± 106.21 (Er) 7777777777 6.302 9.742 6.468 6.440 the oxygen atom belonging to the AlO6 octahedron of one chain is involved in hydrogen bonding with the hydroxy group of the adjacent chain.In the structure of synthetic tavorite LiFe[(OH),F]PO4, the coordination polyhedron about the lithium atom is the axially elongated strongly distorted square pyramid a b c Figure 16. Structure of LiFe[(OH),F]PO4 projected onto the ac plane (a) and the [LiO4(O,F)] polyhedra (b).635 Ref. c /AÊ b (or a, b, g) /deg 82 81 83 7.080 6.910 7.422 a=109.9, b=107.5, g=97.9 a=90.3, b=117.3, g=91.0 a=67.5, b=67.7, g=82.0 6.885 6.469 84 110 111 6.468 (Er) 6.226 6.205 12.776 6.370 10.642 11.101 11.096 11.048 11.103 9.933 101 112 115 113 114 see a see a see a see a 103 86 104 10.431 9.981 77 105 106 10.640 10.415 77 a FFeO4F2 PO4 c Li 2.053 2.032 PO4 1.917 2.132 2.437 a F636 LiO4F whose base is formed by the O and F atoms located at virtually equal distances from the Li atom (1.92 ± 2.13A). The oxygen atom occupying the fifth vertex is at a distance of 2.44A from lithium.The compositions of the known indium fluoride hydroxide phosphates correspond to the formula MIInF0.7(OH)0.3PO4 (MI=K, Rb, NH4) (see Table 8). Apparently, the most stable structures are formed at the F7:OH7 ratio of 0.7 : 0.3. This conclusion is confirmed by the occurrence of analogous alumi- nium fluoride hydroxide phosphate (NH4)0.88(H3O)0.12AlF0.67. .(OH)0.33PO4,103 and gallium fluoride hydroxide phosphate KGaF0.7(OH)0.3PO4.86 The MIInF0.7(OH)0.3PO4 compounds (MI=K, Rb, NH4) are isostructural to KInFPO4 and RbIn(OH)PO4.117 The KGaF0.7(OH)0.3PO4 compound belongs to the KTiOPO4 structural type. In the structure of MIInF0.7(OH)0.3PO4, the fluoride and hydroxide ions statistically occupy one position.This replacement is quite reasonable and results from the similar sizes of these anions [r(F7)=1.15A, r(OH7)=1.18A]. The transformation of structurally similar fluorine-containing indium potassium phases KInF17d(OH)dPO4 (d=0 and 0.3) into hydroxide phos- phate KIn(OH)PO4 is accompanied by lowering of the symmetry of the crystal lattice from tetragonal to orthorhombic. Hence, even the partial replacement of theOHgroups by the fluoride ions in the structure of KIn(OH)PO4 leads to stabilisation of a more symmetrical structure, which is confirmed by X-ray diffraction analysis, electron diffraction study and the results of IR and Raman spectroscopy. Decomposition of KInF0.7(OH)0.3PO4 in air is a complex process giving rise to fluoride InF3 and indium oxofluoride InOF as intermediates, which are transformed into K3In(PO4)2, InPO4 and cubic In2O3.The structure of the (NH4)0.93(H3O)0.07GaF0.5(OH)0.5PO4 compound contains the GaO4(OH) trigonal bipyramid and the GaO4[F,(OH)]2 octahedron as the main building blocks linked to each other by two PO4 tetrahedra.106 These blocks are linked in layers. The channels extended along the [001] direction are occupied by the NHá4 cations. The replacement of one-half of fluoride ions in the (NH4)GaFPO4 compound by the hydroxide ions leads to a radical rearrangement of the fluoride phosphate structure. In the structure of the (NH4)0.93(H3O)0.07Ga- F0.5(OH)0.5PO4 compound,104 the bridging position linking the gallium cations is completely occupied by the hydroxy groups.The ammonium-containing compounds readily decompose to the corresponding orthophosphatesMIIIPO4.III Polymorphism was found in most of representatives of fluo- ride phosphates with composition Na3M2 F3ÖPO4Ü2 (MIII=Al, V, Cr, Fe) and several (up to three) polymorphic modification (a, b, g) were revealed.87 The relationships between the unit cell parameters of the a, b and g modifications are shown in Fig. 17. The high-temperature a modifications of Al, V, Cr and Fe fluoride phosphates are isostructural, whereas the b modifications stable at low temperatures are structurally different (b0 and b00) (Table 9). The unit cell volumes of the the b0 modifications are four times a a 0 a 000 b a 00 b 00 b 000 b 0 Figure 17.Relationship between the unit cell parameters of the a, b and g modifications MI III 3M2 F3ÖPO4Ü2.87 The a and b parameters correspond to the a modification, a 0 and b 0 correspond to the b0 modification, a 00 and b 00 correspond to the b00 modification and a 000 and b 000 correspond to the g modification. L N Komissarova,MG Zhizhin, A A Filaretov larger than those of the a modifications of the same compounds. The b0 polymorphic modification is typical of Al, Cr(III) and Ga fluoride phosphates. The unit cell volumes for the b00 modification of vanadium(III) and iron(III) fluoride phosphates are twice as large as those for the a modifications. The g modification was found only for an iron compound. The b and a modifications differ primarily in the arrangement of the sodium atoms.In the high-temperature tetragonal unit cell of the b0 modification (space group I4/mmm), the alkali-metal atoms are shifted only slightly as compared to the sodium positions in the unit cell of the a modification, whereas substantial shifts (by 0.3 ± 0.5A) are observed in the tetragonal structures of the b0 and b00 modifications (space group I42/mbc and I42/mnm, respectively). The volume of the orthorhombic unit cell of the g modification (T=7190 8C) of Na3Fe2F3(PO4)2 (space group Pbam) is twice as large as the unit cell volume of the b00 modification. The g and b00 modifications of Na3Fe2F3(PO4)2 differ primarily in the positions occupied by sodium cations and distortion of the polyhedra in the structures.The transformation g ? b00 is accompanied by deformation of the [Fe2F3O8]37 polyhedron and displacement of the Na+ cation by 0.7A. The tetragonal a modification of the Na3MIII 2 F3ÖPO4Ü2 com- pounds (MIII=Al, Fe, Cr) consists of the MIIIO4F2 octahedra and the isolated PO4 tetrahedra as the structural blocks (Fig. 18 a). The octahedra are linked in pairs through the fluoride bridges. These doubled octahedraMIII 2 O8F3 are linked in a mixed framework through the phosphate tetrahedra. The sodium atoms occupy the channels. The characteristic structural feature of these compounds is the statistical distribution of the alkali-metal atoms in the unit cell. Due to triple disorder, there are three modes of coordination of the sodium ions, all these atoms being located approximately on a circle with a diameter of 1.9 ± 2.2A(Fig.18 b). At room temperature, the b0-Na3Al2F3(PO4)2 compound crystallises in the tetragonal system (space group P42/mbc) but exhibits a superstructure. The partial replacement of the fluoride ions in the Na3Fe2F3(PO4)2 compound by the hydroxy groups leads to the formation of Na3Fe2[(OH)2F](PO4)2 (see Ref. 85) isostructural to b0-Na3V2F3(PO4)2 (see Table 9). In the structure of Na3Fe2[(OH)2F](PO4)2, the [Fe2O8F3] bioctahedra are deformed resulting in a doubling of the unit cell volume as compared to that of Na3Fe2F3(PO4)2. fluoride Ammonium-containing phosphate gallium (NH4)2Ga2F3(PO4)(HPO4) 106 (see Table 9) was assigned to the `pseudo-KTP' type related to the KTP type (KTiOPO4) (Fig.19 a). In compounds of the `pseudo-KTP' type, only three of four oxygen atoms of each PO4 tetrahedron are involved in the octahedral environment about gallium to form the GaO4F2 and b a F(1) MIII 2 O8F3 F(2) F(2) b a a b c Na(1) Na(2) Na(3) Figure 18. Structure Na3Al2F3(PO4)2. Overall view (a); a one-layer frag- ment projected onto the ab plane (b).Table 9. Synthesis conditions and crystallographic characteristics of the Na3M Compound b 0-Na3Al2F3(PO4)2 a-Na3Al2F3(PO4)2 b 00-Na3V2F3(PO4)2 b 0-Na3Cr2F3(PO4)2 a-Na3Cr2F3(PO4)2 g-Na3Fe2F3(PO4)2 b 00-Na3Fe2F3(PO4)2 a-Na3Fe2F3(PO4)2 Na3Fe2(OH)2F(PO4)2 Na3Ga2F3(PO4)2 (NH4)2Ga2F3(PO4)(HPO4) a Data obtained at T=7190 8C.Table 10. Synthesis conditions and crystallographic characteristics of the Na5MIIIF2(PO4)2 (MIII=Al, Ga, Cr, Fe) and Na57xFe(OdF17d)(PO4)2 compounds. Compound Na5AlF2(PO4)2 Na5GaF2(PO4)2 Na5CrF2(PO4)2 Na5FeF2(PO4)2 Na4.6Fe(O0.6F0.4)(PO4)2 Na4.5Fe(O, F)(PO4)2 (see a) Na4.5Fe(O,F)(PO4)2 (see b) a Modulated phase I stable at T=0 8C. b Modulated phase II stable at T=350 8C. Method of synthesis hydrothermal "solid-phase (from NaCl flux) hydrothermal solid-phase (under Ar) solid-phase (from flux) 27 ( orthorhombic g?b) 7 77 solid-phase (from NaCl flux) hydrothermal ""hydrothermal (with guanidine) Method of synthesis solid-phase (from flux) from aqueous solution solid-phase solid-phase (from flux) the same """" 2IIIF3(PO4)2 (MIII=Al, V, Cr, Fe), Na3Fe2(OH)2F(PO4)2 and (NH4)2Ga2F3(PO4)(HPO4) compounds.Synthesis conditions T /8C p /bar 2000 650 80 300 7 700 2000 650 7 700 7 950 7 7 tetragonal 700 80 300 20 200 18 180 18 180 T /8C Crystal system trigonal 730 60 """" 1000 730 850 ± 700 orthorhombic 850 ± 700 750 """ 750 750 Space group Z Crystal system Tm (Ttr) /8C 43 (b?a) P42/mbc tetragonal 7 I4/mmm " I4/mmm " 726 (sh.) 7 P42/mnm " 55 (b?a) " I4/mmm I4/mmm " 1045 (sh.) Pbam I4/mmm I4/mmm " 746 (sh.) 7 P42/mnm " I4/mmm " 701 (sh.) 7 Pna21 orthorhombic Unit cell parameters Z Space group a /AÊ P310.468 33 P310.483 P3 3 10.560 P3 33 10.576 P3 11.643 16 Pbca 15.514 Ibam 844 15.520 Bmcm 15.645 Bmcm Unit cell parameters b /AÊ a /AÊ 7 12.406 422442842424 7 6.211 7 6.206 7 9.047 7 12.666 7 6.341 12.803 a 12.756 a 7 9.037 7 6.399 7 9.050 7 6.291 7.702 12.497 c /AÊ b /AÊ 7 6.599 7 6.607 7 6.660 7 6.669 13.750 21.777 7.116 14.809 7.119 7.410 7.144 7.433 Ref.c /AÊ 87 10.411 88 10.407 87 10.418 87 10.705 87 10.608 89 10.613 10.602 a 87 87 10.668 90 10.679 85 10.679 87 10.646 106 9.846 Ref. 91 969691 92 93 94 95 95638 a b c Figure 19. Relationship between the `pseudo-KTP'-type structure of (NH4)2Ga2F3(PO4)(HPO4) (a) and the KTP-type structure of (NH4)GaF(PO4) (b).Layered fragments are shown. The ammonium groups are omitted. GaO3F3 octahedra, as distinct from compounds of the KTP type in which all oxygen atoms of the PO4 tetrahedra are involved in the MIIIO6 octahedra. The structure of the `pseudo-KTP' type can be transformed into the structure of the KTP type by rotation of the corresponding PO4 tetrahedra and GaO3F3 octahedra (see Fig. 19). In the KTP-type structures, the channels extended along the [100] direction have regular hexagonal cross-sections, whereas the channels along the [010] direction have distorted hexagonal cross-sections (Fig. 19 b). The framework structure of the (NH4)2Ga2F3(PO4)(HPO4) compound is more distorted as compared to the structure of (NH4)GaFPO4. Fluoride phosphates Na5MIIIF2(PO4)2 with MIII=Al, Cr, Ga 91, 92 crystallise in the trigonal system (Table 10), whereas an analogous iron compound is characterised by lower symmetry (orthorhombic).The Na5MIIIF2(PO4)2 compounds can be assigned to layered structures. These structures contain corru- gated layers parallel to the ab plane. The layers are built from the MIIIO4F2 octahedra and the PO4 tetrahedra (Fig. 20 a,b). One- third of all sodium atoms are located in these layers. The sodium atoms occupy from 6 (in the aluminium-containing compound) to 9 (in the chromium- and gallium-containing compounds) and 10 (in the iron-containing compounds) crystallographically inde- pendent positions most of which are statistically occupied. In the structures of these compounds, the alkali-metal atoms are located only in the centres of octahedra, which are formed either by oxygen atoms or together by fluorine and oxygen atoms (NaO6, NaO5F, NaO4F2, NaO3F3).For iron, not only fluoride phosphate with composition Na5FeF2(PO4)2 but also oxofluoride phos- phates with variable composition Na57xFe(O,F)(PO4)2 (x=0.4 and 0.5) are known. Modulations of the atomic positions were revealed in the structures of the latter compounds.93 ± 95 The reversible phase transition from one modulated phase to another (I?II) observed at T=257 8C is associated with slight displace- ments of the atoms and a change in the direction of the modu- lation vector. The investigation of polymorphism in Na4.5Fe(O,F)(PO4)2 (see Ref. 95) depending on the temperature (T<1000 8C) and pressure (p=16105 ± 1.261010 Pa) revealed the presence of one (at T&260 8C, p=101.3 kPa) and two [at p1=1.39(8) GPa and p2=4.52(32) GPa] polymorphic transfor- mations, respectively.The modulation effect in the structure of Na4.6Fe(O0.6F0.4)(PO4)2 has no influence on the ionic transport (s0&1.176104 ± 1.886104 S cm71, Ea=0.33 eV) and dielectric properties. III The difference in occupancies of the same positions by sodium atoms in the isotypic structures of fluoride phosphates Na3M2 F3ÖPO4Ü2 and Na5MIIIF2(PO4)2 (MIII=Al, Fe, Ga) is indicative of the possible ionic conductivity. At T=0 8C, the measured specific conductivity (s) of the Na5MIIIF2(PO4)2 com- pounds (MIII=Al, Ga) is approximately equal to 1077 S cm71 and it increases by only three orders of magnitude (to s&1074 S cm71) as the temperature is raised to 327 8C.91 As mentioned above, the Na+ ions can migrate only in some of the c b aF b b a Figure 20.Structure of Na5MIIIF2(PO4)2 (MIII=Al, Cr, Ga) projected along the [001] direction (a) and the one-layer fragment projected onto the ab plane (b). L N Komissarova,MG Zhizhin, A A Filaretov b b c a PO4 MIIIO4F2 Na MIIIO4F2 PO4Complex phosphates containing mono- and trivalent cations b Local motion Intercavity motion a F 1.39AÊ O z=0.848 Na(1) Na(2) z=0.356 Figure 21. Structure of Na3Al2F3(PO4)2 projected onto the ab plane. Different trajectories of motion of the Na+ ions are shown.III channels in the structures of Na3M2 F3ÖPO4Ü2. The Na3Al2F3(PO4)2 compound is characterised by a rather high activation energy (0.80 eV at 500 8C), which prevents the wide use of this fluoride phosphate as a ionic conductor. This com- pound is inferior to the recognised sodium conductors, such as compounds of the NASICON type (Ea&0.35 eV) or Na ± b-Al2O3 (Ea&0.14 eV). Based on the analysis of the sizes and geometry of the cavities occupied by the sodium atoms, several possible trajectories of motion of alkali cations in the structure were proposed, viz., the migration from one cavity to another through a distance of 2.72A and/or the local motion around a circle within each cavity. These trajectories are schematically shown in Fig.21.{ Fluoride phosphate K3Al4F9(PO4)2 (see Ref. 108) crystallises in another structural type characterised by the presence of the [Al4F9O8]n chains (Fig. 22 a). The structural block of the chain is formed by four vertex-sharing AlO2F4 octahedra with the cis arrangement of the oxygen atoms (Fig. 22 b). In the perpendicular direction, the chains are linked to each other through the PO4 tetrahedra to form a three-dimensional framework whose chan- nels are occupied by the potassium ions (Table 11). A change in { From J-M Le Meins, O Bohnke, G Courbion Solid. State Ion. 111 67 (1998). Table 11. Synthesis conditions and crystallographic characteristics of apatite-type fluoride phosphates. Crystal system Method of synthesis Compound T /8C K3Al4F9(PO4)2 (NH4)Al2F(PO4)2 2.72A Ê monoclinic orthorhombic tetragonal 180 180 800 LixMIIIY17x(PO4)17xF4x hydrothermal a "solid-phase (LiMF4 ±YPO4) hexagonal trigonal hexagonal "monoclinic 1000 1000 1200 1200 180 solid-phase """hydrothermal a (MIII=Gd ± Lu, Y; 04x40.45) (zircon type) Na2Ba6La2F2(PO4)6 Na3Ba4Nd3F2(PO4)6 Na2Ca6Sm2F2(PO4)6 Na2Ca6Eu2F2(PO4)6 (NH4)CeIVF2(PO4) a The pressure was 18 bar.a b a b b a c Figure 22. Structure ofK3Al4F9(PO4)2 projected onto the ab plane (a) and the mode of linkage of the AlO2F4 octahedra in the chain (b). the Al : F: P ratio from 1 : 1 : 1 corresponding to the KAlFPO4 compound to 2 : 4.5 : 1 corresponding to the K3Al4F9(PO4)2 com- pound is associated with a more complex mode of linkage of the octahedra in the chains in the latter compound.The data on apatite-type compounds containing MI, MII and MIII cations instead of (or along with) calcium are concerned primarily with sodium derivatives of particular REE (see Table 11). The parent compound of this family, viz., potassium fluoride phosphate Ca10F2(PO4)6 (fluoroapatite), crystallises in the hexagonal system (space group P63/m) (Fig. 23 a).118 In the unit cell of Ca10F2(PO4)6, six PO4 tetrahedra are located on mirror planes (z=1/4 and 3/4). The calcium cations occupy two crys- Z Unit cell parameters Space group a /AÊ 242 P21/m P212121 I41/amd 6.721 9.456 6.958 ± 7 6.900 9.939 9.786 9.390 9.385 6.660 21112 P6 P3 P63/m P63/m P21/m 639 OFK PO4 AlO2F4 Ref.b /AÊ c /AÊ b /deg 7 108 109 100 13.856 7.276 105.9 9.621 9.965 6.188 ± 7 6.055 7777 7777 97 97 98 99 107 7.442 7.281 6.895 6.893 5.875 7.177 114.31640 tallographically independent positions. The Ca(1) cations lie on threefold axes in the positions (1/3, 2/3, z) and form the CaO9 polyhedra in which six Ca7O distances have close values and are substantially shortened (2.43 ± 2.47A) as compared to the remain- ing three distances (2.83A). The Ca(2) cations are located on the planes m in the positions (x, y, 1/4) and form the Ca(2)O8F polyhedra in which six O atoms are at distances of 2.36 ± 2.53A, the F atom is located at a distance of 2.26A and two substantially remote O atoms are at distances of 3.27A.Two F7 ions lie on the 63 axis. In apatite-type structures, impurity cations (in particular, REE) are differently distributed over two crystallographically non-equivalent calcium positions (Fig. 23 b).} In the structure of Nd-doped fluoroapatite, the positions occupied by neodymium atoms depend on the mechanism of replacement.97 When neo- dymium was introduced as trifluoride NdF3, the Ca(1) and Ca(2) positions were equally replaced by Nd. By contrast, the use of Nd2O3 as a source of REE led to the structure in which neo- dymium atoms occupy predominantly the Ca(2) position. In the hexagonal structure of Na2Ba6La2F2(PO4)6, the La3+ and Na+ ions statistically occupy both positions, viz., Ca(1) and Ca(2).In the trigonal structure of Na3Ba4Nd3F2(PO4)6, two-third of all Nd3+ and Na+ ions occupy the Ca(1) positions, whereas the remaining atoms are disordered in the Ca(2) positions. Fluoride phosphates with composition Na2Ca6MIII 2 F2ÖPO4Ü6 (MIII= Sm, Eu),98, 99 like non-doped fluoroapatite, crystallise in the hexagonal system. The Eu3+ ions occupy the Ca(1) positions by approximately 75% and the alkali-metal ions are located only in the Ca(2) positions, whereas the opposite situation is observed in the structure of the samarium compound. In the latter structure, the sodium atoms are located only in the Ca(1) positions, whereas the REE ions replace the Ca(2) atoms by 78.5%.In fluoroapatite- type structures, the REE atoms have different coordination numbers (7 and 9). The REE compounds, which crystallise in the apatite structu- ral type, can be used as catalysts, ion-exchange phases and luminescent materials. The Na2Ca6Tb1.75Y0.257xEuxF2(PO4)6 compound was found to exhibit luminescence properties.119 The introduction of europium leads to weakening of radiation by the Tb3+ cations, which is indicative of the charge transfer from Tb3+ to Eu3+. } Figure 23 b was published by P K Rastsvetaeva, A P Khomyakov Kristallographiya 41 831 (1996); [Crystallogr. Rep. (Engl. Transl.) 41 789 (1996)]. Table 12. Characterisations of the structures of different types of complex fluoride phosphates with MI and MIII cations.Crystal system Type of compound Ratio MI:MIII monoclinic 3 : 4 MI III 3M4 F9(PO4)2 (MI=K,MIII=Al) trigonal hexagonal (2 : 4) ± (3 : 3) MI2áxMII 6¡2xMIII 2áxF2(PO4)6 (x=0, 1) (MI=Na;MIII=Sm, Nd, Eu) monoclinic 1 : 1 MIMIIIF(PO4) (MI=Na; MIII=Y, Gd ± Er) MIMIIIF(PO4) (MI=Li, Na, K, NH4, Rb; 1 : 1 MIII=In, Ga, Sc, Al, Fe, Cr) triclinic tetragonal orthorhombic tetragonal 3 : 2 MI III 3M2 F3(PO4)2 (MI=Na; MIII=Ga, Al, Fe, Cr, V) orthorhombic 5 : 1 MI IIIF2(PO4)2 (MI=Na; 5M trigonal orthorhombic MIII=Ga, Al, Fe, Cr) L N Komissarova,MG Zhizhin, A A Filaretov a F(1) Ca(2) Ca(2) Ca(2) Ca(1) Ca(1) Ca(1) Ca(1) Ca(2) Ca(2) Ca(2) F(1) b c b a P63/m P63 P3 IV P3I II III 1/3 2/3 2/3 1/3 Figure 23.Structure of Ca5F(PO4)3 projected along the [100] direction (a) and the schematic representation of the arrangement of the cations (1 ± 4) in the Ca(1) positions in the hexagonal and trigonal apatite-type structures ( I ± IV ) (b). (I) Ca5 F(PO4)3; (II) Sr5 F(PO4)3; (III ) NaSr3CeF(PO4)3; (IV) NaCa3. .CeF(PO4)3. Hence, there are several types of fluoride phosphates contain- ing mono- and trivalent cations, which differ in composition and structure (Table 12). Compounds belonging to the same type are either isostructural or structurally related. Variations in the type and ratio of the ions in the cation sublattice as well as a change of the anion composition in complex phosphates containing mono- and trivalent cations by partially replacing the PO4 tetrahedral groups with anions of different nature (F7) lead not only to the formation of compositionally different compounds with different MI:MIII : PO3¡ ratios but also to a radical rearrangement of the 4 Building blocks Space group P21/m [MIII 4 O8F9]?(chains)+[PO4] P3 [IIIMOxFy] (isolated polyhedra)+[PO4] P6; P63=m C2/m [MIIIO6F2]2?(layers)+[PO4] " P1, C1 P43212 [MIIIO4F2]?(chains)+[PO4] " P212121, Pc21n, Pnna, Pna21 [MIII 2 O8F3] (bioctahedra) + I4/mmm, P42/mnm, P42/mbc, I4/mmm Pbam +[PO4] P3, P3 [MIII4F2] (octahedra)+[PO4] Pbca, Ibam, Bmcm F(1) Ca(1) Ca(1) F(1) 1234 Structural type framework ""layeredComplex phosphates containing mono- and trivalent cations structure from layered (typical only of fluoride phosphates with a high content of alkali cations, MI:MIII=5 : 1) to framework consisting of (âMIII 2 O8F3ä and âMIII 4 O8F9ä) island fragments, chains or layers built from theMIIIOxFy polyhedra.V. Hydroxide phosphates containing mono- and trivalent cations Like fluoride-containing analogues, hydroxide phosphates are characterised by complex compositions combining tri- and mono- valent cations. The similarity between the effective ionic radii of the F7 and OH7 anions and the formation of bonds between these anions and the MIII cations of comparable strength are responsible for the occurrence of theMIMIII(OH)PO4 compounds compositionally similar to MIMIIIFPO4. However, hydroxide phosphates with compositions untypical of fluoride phosphates are also known: MI IIIÖOHÜÖPO4Ü2, MI3MIIIÖOHÜÖHPO4ÜÖPO4Ü, 4M MI1:5âMIII 2 O0:5ÖOHÜ0:5ÖPO4Ü2ÖH2OÜä .xH2O, MIMIII 2 ÖOHÜÖPO4Ü2 .2H2O and MIMIII 3 ÖOHÜ4ÖPO4Ü2 .2H2O. Some compounds of theMIMIII(OH)PO4 group as well as a series of phases with composition MIMIII 2 ÖOHÜÖPO4Ü2 .2H2O were found as minerals. Complex hydroxide phosphates of all these types are known forMIII cations with small ionic radii (Al, Ga, In, V, Fe, Mn). 1. Synthesis The mineral tavorite LiFe(OH,F)PO4 is the Fe-containing ana- logue of montebrasite LiAl(OH)PO4 (see Ref. 81) belonging to the group of amblygonite LiAl(OH,F)PO4. The ideal compositions of tavorite and amblygonite are described by the formulas Table 13.Hydrothermal synthesis of complex hydroxide phosphates with MIII cations. Compound Composition Starting reagents (molar ratio), pH MIMIII(OH)PO LiMn(OD)PO4 4 KIn(OH)PO4 MnPO4 .D2O, LiNO3 (Li :Mn=2 : 1) In2O3, K2HPO4(sol.) RbIn(OH)PO4 NH4In(OH)PO4 In2O3, 2.5M Rb2HPO4, 3M RbH2PO4 (Rb : P=1.5) In2O3, NH4H2PO4, NH4OH(sol.) (NH4 : In : P=8.1 : 1 : 2.41) MIMIII 2 (OH)(PO4)2 .2H2O KFe2(OH)(PO4)2 .2H2O (leucophosphite) (NH4)V2(OH)(PO4)2 .2H2O (NH4)Fe2(OH)(PO4)2 .2H2O (spheniscidite) K1.5[Al2P2O8.5(OH)0.5(H2O)] . xH2O FePO4 . nH2O, 2.5M K3PO4, pH 2 VCl4, H3PO4, H2O, 1,3-diamino- propane (1 : 7.4 : 1004 : 8.6) FeO(OH), H3PO4, HF, H2O, 1,3-diaminopropane (2 : 2 : 2 : 3 : 80) Al(OH)3, KH2PO4 (K :Al=3 : 2; 3 : 4; 1 : 1; 3 : 2) Rb1.5[Al2P2O8.5(OH)0.5(H2O)] .xH2O Al(OH)3, RbH2PO4 (Rb :Al=3 : 2; 3 : 1) MIMIII(OH)(PO Na4Al(OH)(PO4)2 4)2 Na4Ga(OH)(PO4)2 Al(NO3)3 .9H2O, H3PO3, NaOH, 1,4-diazabicyclo[2.2.2]octane, H2O, pH 10 Ga2O3, NH4H2PO4, NaOH (Na :Ga : P=3 : 1 : 2.4), pH 12 MI III(OH)(HPO4)(PO4) Na3 Al(OH)(HPO4)(PO4) 3M Al(OH)3, Na2HPO4 .2H2O, NaH2PO4 .2H2O (Na : P=1.5) Na3Ga(OH)(HPO4)(PO4) Ga2 O3, Na2HPO4 .2H2O, NaH2PO4 .2H2O (Na : P=1.5) a Slow cooling to 500 8C for 32 h; b slow cooling to 250 8C for 60 h; c unpublished data; d slow cooling to 500 8C for 66 h. LiFe(OH)PO4 (see Ref. 120) and LiAlFPO4 (see Ref. 81), respec- tively. Being a rare mineral from lithium pegmatites, tavorite is unstable under natural conditions.Various paragenetic associa- tions are typical of minerals of the leucophosphite group of the general formula MIMIII 2 ÖOHÜÖPO4Ü2 .2H2O, where MI=K, NH4 and MIII=Al, Fe. Leucophosphite KFe2(OH). .(PO4)2 .2H2O121, 122 by itself is a product formed by the action of organic solutions derived from bat guano with iron oxides in serpentine rocks or is generated in late stages of the hydrothermal transformation of triphylite in pegmatites. Tinsleyite 123 (Al analogue of leucophosphite) was found together with KFe2(OH). .(PO4)2 .2H2O in pegmatites. Spheniscidite (NH4)Fe2(OH). .(PO4)2 .2H2O124, 125 is formed by the actions of ammonium phosphate solutions of penguin guano with mica and chlorite minerals.Cyrilovite NaFe3(OH)4(PO4)2 .2H2O,126 which is the Fe-containing final member of the series of minerals belonging to the group of wardite Na(Al,Fe)3(OH)4(PO4)2 .2H2O, is fragmen- tary present in pegmatites and can occupy cavities of oxidation zones of ferrous precipitates. Most of the compounds under consideration were synthesised under hydrothermal conditions (Table 13).117, 122, 125, 127 ± 135 The simultaneous action of pressure and temperature leads to a substantial increase in solubility of the compounds under study in active aqueous media and is favourable for the preparation of complexMIII hydroxide phosphates as highly crystalline powders and well-faceted crystals, which cannot, in principle, be grown by alternative methods.Unlike Al- and Fe-containing phases of the MIMIII(OH)PO4 type, their indium analogues do not form minerals of their own. Single crystals of hydroxide phosphates KIn(OH)PO4 (see Ref. 127) and RbIn(OH)PO4 (see Ref. 117) were obtained under drastic conditions (900 and 550 8C, respec- tively) followed by stepwise cooling. It should be noted that these phases are stable at pH 7 ± 8. Attempts to prepare polycrystalline 641 Ref. Conditions of synthesis T /8C p /bar t /h 134 504 20 200 127 12 3000 900 a 117 2200 550 b 8 135 168 20 200 122 336 150 7 128 87 200 7 125 24 200 7 175 ± 225 7 7 129 225 7 7 129 132 96 20 200 96 96 400 160 see c see c 718 133 20 600 d 7 133 20 600 d 7642 samples of MIn(OH)PO4 (M=K, Rb) in pure form failed.These samples contained unconsumed In2O3 even when the synthesis was carried out for 2 ¡¾ 3 weeks. Under these conditions, the sodium analogue NaIn(OH)PO4 is not formed. The hydrothermal synthesis of hydroxide phosphates of the leucophosphite group is characterised by a wide variety of the starting reagents. In a number of studies aimed at preparing porous organometallic phosphate compounds, hydroxide phos- phates with composition MIMIII 2 OOHUOPO4U2 .2H2O were syn- thesised.125, 128 Ammonium leucophosphite was prepared not only by the hydrothermal method with the use of aqueous solutions but also by the low-temperature synthesis from hydrogel (T=150 8C, t=45 h, pH of the final solution was 6.52).124 The MI1:5aMIII 2 O0:5OOHU0:5OPO4U2OH2OUa .xH2O compounds 129 (MI=K, Rb), which are structurally similar to compounds of the leucophosphite group, are generated from mixtures of Al(OH)3 and MIH2PO4 in the temperature range of 175 ¡¾ 225 8C. The NaAl(OH)PO4 and (NH4)Al2(OH). .(PO4)2 . 2H2O compounds were prepared with the use of NaH2PO4 and (NH4)H2PO4, respectively.130, 131 Hydroxide phos- phates of this types are stable in weakly acidic and neutral media at pH 2.5 ¡¾ 7.5. The MI IIIOOHUOPO4U2 compounds (MI=Na; MIII=Al, 4M Ga) 132 are formed as transparent elongated prismatic crystals in an alkaline medium (pH 10 ¡¾ 12) under mild hydrothermal con- ditions in the presence of an excess of alkali cations (Na :Al53). The structurally related MI IIIOOHUOHPO4UOPO4U compounds 3M (MI=Na;MIII=Al, Ga) 133 were prepared from weakly alkaline media at high temperatures (600 8C) using stepwise cooling.The Na3Al(OH)(HPO4)(PO4) and Na3Ga(OH)(HPO4)(PO4) com- pounds crystallise as colourless prisms and thin platelet-like crystals, respectively. 2. Structures and properties There is an obvious structural analogy between all hydroxide phosphates, except for compounds belonging to the MIMIII(OH)PO4 group. The crystal chemistry of the latter com- pounds radically changes depending on the nature of the MI and MIII cations. Recent studies dealing with the determination of the crystal structures of complex hydroxide phosphates were aimed primarily at refining the assignment of a particular compound to the structural type (with a typical alternation of the cations and anions) and revealing the crystal-chemical features of the struc- tures of the compounds synthesised.Minerals with compositionMIMIII(OH)PO4, whereMI=Li, Na and MIII=Al, Fe, are isomorphous. These compounds are characterised by high mobility of alkali atoms in the channels of the structures. The structural feature of this group of complex phosphates is the presence of chains built from the vertex-sharing FeO6 or AlO6 octahedra linked by OH groups (Table 14). Mixed frameworks containing such chains are also typical of kizerite (MgSO4 .2H2O), titanite (CaTiOSiO4) and isostructural minerals (space group C2/c).81 In a manganese analogue of tavorite, viz., LiMn(OH)PO4,134 strong distortion of the MnO4(OH)2 octahedra gives rise to a substantial ferromagnetic component although this compound Table 14.Crystallographic characteristics of theMIMIII(OH)PO4 compounds. Z Crystal system Space group Compound (mineral name) triclinic ""orthorhombic tetragonal " C1 P1 C1 P212121 P43212 P43212 LiAl(OH)PO4 (montebrasite) LiFe(OH)PO4 (tavorite) LiMn(OD)0.442(OH)0.558PO4 KIn(OH)PO4 RbIn(OH)PO4 (NH4)In(OH)PO4 424888 L N Komissarova,MG Zhizhin, A A Filaretov belongs to antiferromagnetics. The elongated Mn7O bonds (O are the oxygen atoms of the PO4 groups) make the main contribution to the direction of the vector of the magnetic moment. As a result, the vectors of the magnetic moments of the MnO4(OH)2 octahedra belonging to the same chain are parallel, co-directed and virtually perpendicular to theM7OH7Mbond. This direction of the vectors of the magnetic moments is respon- sible for the ferromagnetic properties of the LiMn(OH)PO4 compound.134 On going from iron- and manganese-containing compounds [rVI(Fe3+)=0.645A, rVI(Mn3+)=0.65A] to indium com- pounds [rVI(In3+)=0.79A], the structural type changes attended with an increase in the symmetry from triclinic to orthorhombic and tetragonal (see Table 14).An increase in the ionic radius of alkali metal on going from potassium to rubidium also leads to an increase in the symmetry from orthorhombic to tetragonal. Complex indium hydroxide phosphates containing different monovalent cations have compositions in which the ratio MI:MIII : PO3¢§ 4 =1 : 1 : 1 is retained.The structure of orthorhom- bic indium hydroxide phosphate KIn(OH)PO4 is based on helical chains of the InO4(OH)2 octahedra sharing cis vertices.127 These chains are extended along the c axis (Fig. 24 a). The polyhedra about the potassium cations (KO7) can be described as distorted trigonal antiprisms. In the structures of NH4In and RbIn hydrox- ide phosphates, the presence of screw axes 43 leads to an ordered arrangement of the polyhedra in helical chains so that the channels of virtually square cross-section are formed along the [001] direction (Fig. 24 b). The NHa4 and Rb+ ions are surrounded by nine oxygen atoms each.In the structures of the MIIn(OH)PO4 compounds (MI=Rb and NH4), the indium octahedra are almost non-distorted. Thus, the difference in the In7O bond lengths is 0.003A and 0.059A in (NH4)In(OH)PO4 (see Ref. 135) and RbIn(OH)PO4 (see Ref. 117), respectively. The crystallographically independent indium atoms alternate along the helical chains and are linked to each other through the bridging oxygen atoms of the hydroxy groups, which are not involved in hydrogen bonding with the nearest environment. The presence of independent OH groups in the structures was confirmed by IR spectroscopy (n&3520 cm71). Each PO4 tetrahedron shares two vertices with the InO6 octahedra belonging to the same chain and shares two other vertices with the octahedra from the adjacent chain.The chains are linked through the tetrahedra to form a framework whose large cavities are occupied by single-charged cations. The framework of the KIn(OH)PO4 compound is similar to that of g-NaTiOPO4.136 The difference is observed in the positions of the alkali cations located in the tunnels of these structures. The fact that indium and scandium have similar ionic radii [rVI(Sc3+)=0.745A] as well as the fact that KIn(OH)PO4 and RbIn(OH)PO4 belong to different structural types indicate that related hydroxide phosphate KSc(OH)PO4 may occur. Appa- rently, the sodium-containing analogue with composition NaIn(OH)PO4 that is structurally similar to KIn(OH)PO4 does not occur. In this case, the steric factor is of considerable importance because the difference in the sizes of the potassium and sodium cations is much larger than that in the sizes of the potassium and rubidium cations. Ref.Unit cell parameters c /AE b /AE a /AE a, b, g /deg 81 120 134 127 117 135 91.31, 117.93, 91.77 109.29, 97.86, 106.32 89.5, 117.7, 86.4 777 7.019 5.110 7.123 11.245 11.179 11.177 7.708 7.283 8.187 9.339 77 6.713 5.340 6.719 9.277 9.400 9.423Complex phosphates containing mono- and trivalent cations a a b c (NHá4 )Nb a b octahedra in the structure of (NH4)In(OH)PO4 along the [110] direction (a) and the structure of (NH4)In(OH)PO4 projected onto the ab plane (b). Figure 24. Chains of the (OH)2 Hence, analysis of the structural types in the morphotropic series MIMIII(OH)PO4 demonstrates that the structures of these compounds are characterised by the presence of chains or ribbons, which are built from octahedra of the trivalent cations linked to each other through the OH vertices.Taking into account the structural and compositional sim- ilarity, theMIIn(OH)PO4 compounds (MI=K,Rb andNH4) can be compared with phases related to KTiOPO4. It is known that the TiO6 octahedra in the structure of KTiOPO4 are substantially distorted with alternation of short and long bonds along the 7O7Ti(1)7O7Ti(2)7O7 chain. The differences between the long and short Ti(1)7O and Ti(2)7O bonds are 0.275 and 0.363A, respectively.116 These differences in the bond lengths are of importance in determining the nonlinear optical properties of the KTP-type AIBIVOXVO4 compounds.However, weak distor- tion of the MIIIO6 and MIIIOxFy octahedra in the structures of complex indium phosphates, such asMIIn(OH)PO4 (MI=K, Rb, NH4), KInFPO4 and MIInF17d(OH)dPO4 (MI=K, Rb, NH4), and in the structures of MIMIIIF17d(OH)dPO4 (MI=K, NH4; MIII=Ga, Fe) is responsible for the fact that these compounds do not exhibit nonlinear optical properties. Two minerals, viz., wardite NaAl3(OH)4(PO4)2 .2H2O137 and cyrilovite NaFe3(OH)4(PO4)2 .2H2O,126 belong to hydroxide 643 phosphates with composition MIMIII 3 ÖOHÜ4ÖPO4Ü2 .2H2O. These compounds crystallise in the tetragonal system (space group P41212, Z=4).The unit cell parameters of wardite (a=7.030, c=19.040A) are close to those of cyrilovite (a=7.313, c=19.315A). The structures of these minerals are composed of the layers located perpendicular to the [001] direc- tion. The layers consist of infinite chains of the alternating cis- linked M(1)O6 and M(2)O6 octahedra extended along the a and b axes (Fig. 25). In the M(1)O6 octahedron, the M3+ cation is surrounded by three oxygen atoms of the PO4 tetrahedra, two oxygen atoms of the OH groups and the oxygen atom of the water molecule. In the M(2)O6 octahedron, the M3+ cation is sur- rounded by two oxygen atoms of the PO4 groups and four oxygen atoms of the OH group. The M(1)O6 octahedron is linked to two nearest M(2)O6 octahedra through bridging hydroxy groups.The M(2)O6 octahedron, in turn, is linked to four nearest strongly distorted M(1)O6 octahedra through four hydroxy bridges. The cavities in the layers are occupied by sodium ions (coordination number is 8). The PO4 tetrahedra share three vertices with three octahedra in the layer and one vertex with an octahedron from the adjacent layer, three vertices being involved in the M(1)O6 octahedron and one vertex being involved in the M(2)O6 octahe- dron. In the mixed framework of the octahedra and tetrahedra, two types of channels are extended along the [100] and [010] directions. These channels are occupied by zeolite-type water molecules. The presence of water of crystallisation and bridging OH group in the structures was confirmed by IR spectroscopy.Analysis of the interatomic distances in the MIIIO6 polyhedra in the NaMIII 3 ÖOHÜ4ÖPO4Ü2 . nH2O compounds and the radii of the MIII cations demonstrated that the octahedral positions can be occupied not only by the Al and Fe cations but also by the gallium and vanadium cations with no changes in the crystal structure. The occurrence of potassium analogues is highly improbable because of the small sizes of the cavities in the layers. Minerals of the leucophosphite of the general formula MIMIII 2 ÖOHÜÖPO4Ü2 .2H2O (MI=K, NH4; MIII=Al, Fe) form a continuous series of low-temperature solid solutions. Three minerals of this series are known, viz., leucophosphite KFe2(OH)(PO4)2 .2H2O, tinsleyite KAl2(OH)(PO4)2 .2H2O and spheniscidite (NH4,K)(Al,Fe)2(OH)(PO4)2 .2H2O. Besides, the related synthetic KAl, NH4Al, NH4Ga and NH4V ana- logues 138 ± 141 belong to this group (Table 15).Fe(1) b Fe(2) a Na Figure 25. Structure of NaFe3(OH)4(PO4)2 .2H2O projected onto the ab plane.644 Table 15. Crystallographic characteristics of minerals of the leucophosphite group of the general formula MIMIII 2 ÖOHÜÖPO4Ü2 .2H2O and the 4M MI III(OH)(PO4)2 andMI3MIII(OH)(HPO4)(PO4) compounds. Crystal system Compound (mineral) Space group KAl2(OH)(PO4)2 .2H2O KAl2(OH)(PO4)2 .2H2O (tinsleyite) P212121 P21/n P21/n KFe2(OH)(PO4)2 .2H2O (leucophosphite) (NH4)Fe2(OH)(PO4)2 .2H2O (spheniscidite) """""orthorhombic ""monoclinic LiNa2HAl(OH)(PO4)2 Na4Al(OH)(PO4)2 Na4Ga(OH)(PO4)2 Na3Al(OH)(HPO4)(PO4) orthorhombic monoclinic K1.5{Al2O0.5(OH)0.5(PO4)2(H2O)} .xH2O " (NH4)Al2(OH)(PO4)2 .2H2O " P21/n P21/n P21/n P21/n P21/n P21/n (NH4)V2(OH)(PO4)2 .2H2O " P21/n (H3O)Ga2(OH)(PO4)2 .2H2O " P21/n (NH4)Ga2(OH)(PO4)2 .2H2O " P21/n Pbcb Pbcm Pbcm C2/m Na3Ga(OH)(HPO4)(PO4) " C2/m a Our unpublished data. The compounds of this groups are characterised by the presence of theMIII 4 ÖOHÜ2ÖH2OÜ2ÖPO4Ü4 island fragment formed from four octahedra of two types, viz., M(1)O4(OH)2 and M(2)O4(OH)(H2O). The central core of this fragment consists of two edge-sharing M(1)O6 octahedra} with two anchored hydroxo groups. The M(1)O6 octahedra share edges containing two OH vertices with two M(2)O6 octahedra (Fig.26 a,b). The fragments consisting of four MIIIO6 octahedra are linked in a framework through the isolated PO4 tetrahedra. The channels of the structure are occupied by monovalent cations and water molecules. In each independent MIIIO6 octahedron, four oxygen atoms are at short distances from theM3+ ion (these distances differ by only 0.03 ± 0.04A), whereas two oxygen atoms are far remote from MIII. This distortion of the octahedra results from the fact that the oxygen atoms of theOHgroups and water molecules must satisfy the valence balance rule. The octahedral coordination of the MIII cations is retained but the MIII7O distances change as the ionic radius increases in the series Al ±Ga ±V± Fe.Analysis of the interatomic MIII7O distances in the MIIIO6 octahedra and the MI:MIII ratio showed that the compounds of all the above- mentioned trivalent cations (MIII=Al, Fe, V, Ga) in combina- tion with rubidium can also crystallise in the structural type MIMIII 2 ÖOHÜÖPO4Ü2 .2H2O. It is quite possible that In-containing analogues can also crystallise in this structural type. In all compounds with MI=NH4, the NHá4 cations are surrounded by four oxygen atoms giving rise to a tetrahedron within which two N7H_O hydrogen bonds occur (Fig. 26 b). 2 ÖOHÜÖPO4Ü2ÖH2OÜg .H2O. The water molecules in the structure of MIMIII 2 ÖOHÜÖPO4Ü2 .2H2O fulfil different functions. One H2O molecule is involved in the first coordination sphere of the MIII(2)O6 octahedron, whereas another water molecule is of a zeolite type and is not coordinated to any cation.As a result, the structural formula of these hydroxide phosphates can be written as followsMIfMIII It should be noted that the structures of all the above- considered minerals and their synthetic analogues are derived from the structure of GaPO4 .2H2O.139 In this structure, the positions of the hydroxonium ions correspond to the positions of the monovalent K+ and NHá4 cations according to the structural formula H3O{Ga2(OH)[PO4]2(H2O)} .H2O. The replacement of } The presence of the edge-sharingMIIIO6 octahedra in these structures is responsible for the magnetic properties of iron-containing representatives of this group. L N Komissarova,MG Zhizhin, A A Filaretov Z Unit cell parameters c /AÊ b /AÊ a /AÊ 9.828 9.543 9.4384 9.614 9.556 9.751 9.769 9.874 9.872 9.858 9.77 9.788 14.065 6.947 7.049 7.041 7.057 9.707 9.532 9.5589 9.577 9.572 9.658 9.664 9.737 9.738 9.688 9.64 9.703 14.089 14.660 14.661 7.054 7.164 9.205 9.602 9.495 9.553 9.617 9.782 9.756 9.819 9.823 9.803 9.68 9.689 6.948 15.279 15.407 15.277 15.432 44444444444448844 a b a c Figure 26.Structure of (NH4)Fe2(OH)(PO4)2 .2H2O projected onto the bc (a) and ac planes (b). The centrosymmetrical island fragments of the Fe octahedra are displayed (a); the positions of the NHá4 cations in the channels of the structure are shown (b).Ref. b /deg 138 123 129 130 131 121 122 124 125 128 139 140 141 132 see a 133 133 7 103.16 101.74 103.6 103.6 102.2 102.4 102.8 102.8 102.91 102.7 102.80 77796.73 96.64 H2O OH b c NH4Complex phosphates containing mono- and trivalent cations one-half of the hydroxide ions in theMI{MIII 2 (OH)(PO4)2(H2O)} . H2O compounds by the oxide ions gives rise to compounds with compositions MI1:5fMIII 2 O0:5ÖOHÜ0:5ÖPO4Ü2ÖH2OÜg . xH2O (MI=K, Rb;MIII=Al).129 2 O4 groups is common to hydroxide The presence of the MIII phosphates MIMIII 2 ÖOHÜÖPO4Ü2 .2H2O and their topological analogues deprived of OH groups, such as (NH4)Mo2. .P2O10 .H2O142 and RbMo2P2O10 .(17x)H2O.143 In the Al3(PO4)3 .H2O. en and Ga3(PO4)3 .H2O. en compounds (en is ethylenediamine), the MIII 2 O4 groups are involved in trigonal bipyramids and the organic cations (protonated ethylenediamine molecules) occupy the channels.144 Due to the specific ion- conducting properties, these compounds can be used as molecular sieves.145 4Ü42 ¡ ]? 2 )]? Hydroxide phosphates with composition Na4MIII(OH)(PO4)2 (MIII=Al, Ga) and the ratioMI:MIII=4 : 1 are also known (see Table 15). Let us consider the structures of these compounds using an aluminium compound as an example. The main struc- tural elements of the Na4Al(OH)(PO4)2 compound are the infinite [Al(OH)(PO4)2]? chains extended along the [001] direction (Fig.27 a). In these chains, the AlO4(OH)2 octahedra are trans- linked to each other through the bridging oxygen atoms of the hydroxy groups (Fig. 27 b). The adjacent octahedra in the chains are linked by two PO4 groups. Each PO4 group shares two vertices with the adjacent octahedra in the chain. The [Al(OH)(PO chains are held together primarily through interactions between two non-shared vertices of the phosphate groups and the sodium atoms located in the channels between the [Al(OH)(PO4Ü4¡ chains. The sodium atoms located between the PO4 tetrahedra in cis positions with respect to the hydroxyl hydrogen atoms have a lesser effect on binding of the chains. The PO4 tetrahedra and AlO6 octahedra are only slightly distorted.In the structure of Na4Al(OH)(PO4)2, the sodium atoms have different coordination numbers depending on their positions with respect to the [Al(OH)(PO4Ü42¡ ]? chains. Four of seven independ- ent sodium atoms are involved in the NaO6 octahedra. One NaO4+2 polyhedron is formed by four oxygen atoms located in the first coordination sphere (the Na7O distances are*2.495A) and two oxygen atoms at distances of 2.72 ± 2.99A. Finally, the coordination environment about two sodium atoms are five- vertex polyhedra with four short Na7O bonds (*2.3A) and one abnormally large distance (*2.88A). a b H(OH) b Na a cFigure 27. Structure of Na4Al(OH)(PO4)2 projected along the [001] direc- tion (a) and the [Al(OH)(PO4)42 ¡]? chain (b).645 In the mineral tancoite LiNa2HAl(OH)(PO4)2,141 the arrange- ment of the chains is analogous to that observed in the structure of Na4Al(OH)(PO4)2. The slight differences in the positions of the Li and H atoms in the structure of tancoite as compared to the positions of the sodium atoms replacing these atoms in the structure of Na4Al(OH)(PO4)2 cause changes in the symmetry of the structure, viz., the parameter a is decreased, whereas the parameter c is doubled. Taking into account the similarity of the ionic radii of aluminium, gallium and iron, gallium- and iron-containing ana- logues of Na4Al(OH)(PO4)2 with the same structure would be expected to occur. Actually, we have recently prepared isostruc- tural hydroxide phosphate Na4Ga(OH)(PO4)2.The structures of the Na3MIII(OH)(HPO4)(PO4) compounds (MIII=Al, Ga) are similar to that of Na4Al(OH)(PO4)2 with the only difference that the sodium atoms are partially replaced by the hydrogen atoms of the HPO4 groups (Fig. 28 a).133 The adjacent octahedra in the [Al(OH)(HPO4)(PO4)37]? chains are linked through the bridging HO groups and by the HPO4 and PO4 groups (Fig. 28 b). The structure is additionally stabilised by hydrogen bonds involving the HPO4 groups from the adjacent chains. The sodium atoms in the structure have different coordi- nation environments, viz., NaO5 , NaO6 and NaO8 . The coordi- nation polyhedron about the sodium atom with the coordination number of 8 is a distorted cube with four short, two medium and two long Na7O distances.The averageMIII7O and P7O bond lengths are comparable with the analogous distances in Na4Al(OH)(PO4)2. For Na3Al(OH)(HPO4)(PO4), these bond lengths are 1.911 and 1.536A, respectively. The chains consisting of the [M(EO4)2X] groups (X is a non- specific ligand), which were found in the structures of MI IIIÖOHÜÖPO4Ü2 and MI3MIIIÖOHÜÖHPO4ÜÖPO4Ü, are typical 4M of a large group of phosphate, sulfate and silicate minerals.132 For example, in the structures of the minerals jahnsite CaMnMg2. .[Fe(OH)(PO4)2]2 .8H2O,146 segelerite CaMg[Fe(OH)(PO4)2] . 4H2O and overite CaMg[Al(OH)(PO4)2] .4H2O (see Ref. 147), the [M(EO4)2X] chains are linked in layers through the MIIO2(H2O)4 octahedra. a c a b HPO4 PO4 Figure 28.Structure of Na3Al(OH)(HPO4)(PO4) projected onto the ac plane (a) and the [Al(OH)(HPO4)(PO4)3¡]? chain (b).646 Thermal stability of hydroxide phosphates containing mono- and trivalent cations depends on the mode of binding of water molecules in these structures. Thus, hydroxide phosphates MIIn(OH)PO4 (MI=K, Rb) lose water of crystallisation in the temperature range of 500 ± 700 8C. Decomposition of hydroxide phosphate KIn(OH)PO4 affords K3In(PO4)2, InPO4 and In2O3. Hydroxide phosphate (NH4)In(OH)PO4 decomposes in two steps at 180 ± 490 8C and 535 ± 690 8C. Crystallisation of InPO4 starts at 620 8C. Dehydration of leucophosphite KFe2(OH). .(PO4)2 .2H2O occurs in two steps at 300 ± 575 8C. First, the zeolite-type water is removed and then the H2O molecules located in the first coordination sphere about the iron atoms are elimi- nated.Dehydration of NaFe3(OH)4(PO4)2 . nH2Oproceeds in one step. The temperature ranges of the removal of the zeolite-type water and water of crystallisation overlap (the maximum temper- atures corresponding to the removal of these water molecules are 428 and 447 8C, respectively) and complete dehydration is achieved at 600 8C. In the case of the NH4V analogue of leucophosphite, water of crystallisation is eliminated at a substan- tially lower temperature (in the range of 45 ± 110 8C) and decom- position of the compound is completed at 320 8C. Dehydration of Na3Al(OH)(HPO4)(PO4) proceeds in one step with a maximum at *438 8C and ends in the formation of double aluminium phos- phate Na3Al(PO4)2. A comparison of the data on the structures of complex hydroxide phosphates containing mono- and trivalent cations revealed the main structural features of these compounds.The coordination polyhedra about the MIII cations are without exception octahedra. The MIIIO6 octahedra are linked in infinite chains through the bridging hydroxy groups. Depending on the ratio between theMI andMIII cations, the chains are either linked to each other through the phosphate tetrahedra or are held together by hydrogen bonds and MI_O interactions { to form framework, layered or chain structures (Table 16). For the above-considered compounds, two compositions with the ratios MI:MIII=1 : 2 or 1 : 1 (in the case of the medium and large monovalent K+, Rb+, NH4+ cations) are preferential.The compositions with 1 : 3 or 3 : 1 (4 : 1), whereMIII cations are either excessive or deficient with respect toMI cations, are typical only of sodium-containing aluminium and gallium hydroxide phos- phates, both the OH and HPO4 groups being involved in com- pounds with the ratio MI:MIII=3 : 1. It should be noted that water molecules play an important role in the formation of structures containing an excess of MIII cations with respect to MI. The H2O molecules are involved in the formation of the first coordination sphere about the MIII cations and occupy the { Considering the size factor, only sodium can serve as such monovalent cation. Table 16.Characterisations of the structures of different types of complex hydroxide phosphates with MI and MIII cations. Crystal system Type of compound Ratio MI:MIII tetragonal 1 : 3 MIMIII 3 (OH)4(PO4)2 .2H2O (MI=Na,MIII=Al, Fe) monoclinic 1 : 2 MIMIII 2 (OH)(PO4)2 .2H2O (MI=K, NH4, MIII=Al, Fe, Ga, V) 1 : 1 MIMIII(OH)PO4 (MI=Li, Na, K, NH4, Rb;MIII=Al, Fe, Mn, In) triclinic orthorhombic tetragonal monoclinic 3 : 1 MI III(OH)(HPO4)(PO4) 3M (MI=Na, (Li, Na), MIII=Al, Ga) orthorhombic 4 : 1 4)2 MI III(OH)(PO 4M (MI=Na,MIII=Al, Ga) L N Komissarova,MG Zhizhin, A A Filaretov channels and cavities in the structures as `guests', which are not bound to any cations. VI. Other complex phosphates containing mono- and trivalent cations 4 4 4 4 4 In addition to fluoride phosphates and hydroxide phosphates containing mono- and trivalent cations, other complex mixed- anion compounds are known.In such compounds, the PO3¡ anion is combined with VO3¡ 4 , MoO24 ¡, SO24 ¡ or S2O23 ¡. It is expected that mixed-anion phosphates will exhibit improved properties (such as sorption ability, luminescence, etc.) as com- pared to monoanion phases. The replacement of the tetrahedral PO4 group in the anion sublattices of complex phosphates by other groups enables one to vary not only the total negative charge of the compound, which makes it possible to change the number of cations, but also the size characteristics of the cavities in frame- work structures and of interlayer spaces in layered phases.Such structural changes may allow one to influence the catalytic properties and ionic mobility of compounds as well as ion exchange processes involving these compounds.148 ± 152 The partial replacement of the PO3¡ anions in compounds with composition K3MIII(PO4)2 (MIII=Y, La, Eu, Gd, Yb) by VO3¡ gives rise to continuous or limited solid solutions based on one of the structural types (arcanite or glaserite) typical of non- substituted double phosphates K3MIII(PO4)2 or vanadates K3MIII(VO4)2. The lack of new compounds is associated with a small difference in the sizes of the PO3¡ and VO3¡ anions. The formation of limited or continuous solid solutions is determined by the dissimilarity or similarity of the structural types of the starting double phosphates and vanadates.28 Phosphate vanadates K3MIII(PO4)x(VO4)27x (MIII=Sc, Y, La ± Lu; x=0 ± 2) exhibit sorption properties with respect to H2O, CO2 and organic compounds.The degree of sorption of H2O and CO2 from the gas phase depends on the PO3¡ 4 :VO34 ¡ ratio whose increase leads to preferential sorption of water molecules from a mixture of H2O and CO2.153 The degree of sorption correlates linearly with the conductivity and dielectric characteristics of the water-containing K3MIII(PO4)x(VO4)27x phases. III Complex phosphate vanadates K3M1¡yEuy(PO4Üx(VO4)2¡x (MIII=Sc, Y, La, Gd; x=0±2; y=0 ± 1) possess luminescent properties, which depend both on the concentration of the activator Eu3+ ions and the PO3¡ 4 : VO34 ¡ ratio.The formation of solid solutions, which crystallise in the same structural type, is evidenced by the similarity of their luminescence spectra. These spectra are characterised by the absence of complete concentra- tion quenching of luminescence as the concentration of activator ions is increased. The maximum intensity of luminescence was observed at the Eu3+ concentration corresponding to y=0.3. Building blocks Structural type Space group framework P41212 [MIIIO2(OH)4] uÁ [MIIIO3(OH)2Ow] (chains) +[PO4] " P21/n [MIII 4 (OH)2(H2O)2(PO4)4]3? (island fragments) [MIIIO4(OH)2]?(chains)+[PO4] " P1, C1 P212121 P43212 chain C2/m [MIIIO4(OH)2]?(chains)+ +[PO4] and [HPO4] Pbcm [MIIIO4(OH)2]?(chains)+[PO4] "Complex phosphates containing mono- and trivalent cations Phosphate vanadates K3M1¡yEuy(PO4Üx(VO4)2¡x (MIII=Sc, Y, Gd) possessing bright red luminescence are promising photo- luminescent and cathode-luminescent materials.154 ± 156 Complex molybdate phosphate Na2Y(PO4)(MoO4) 157 was prepared by crystallisation from flux in the study of the Y2O3 ± (NH4)2HPO4±Na2MoO4 system.This compound crystal- lises in the monoclinic system (space group C2/c, Z= 8, unit cell parameters a=13.928, b= 18.016, c=6.847A, b=119.62 8). The structure of Na2Y(PO4)(MoO4) contains chains (along the c axis) of the edge-sharing YO8 dodecahedra. The chains are linked in corrugated layers parallel to the ac plane through the PO4 (shared edges) and MoO4 (shared vertices) tetrahedra (Fig. 29 a).The sodium atoms are located in the interlayer space (Fig. 29 b). Molybdate phosphates of this type are unknown for REE. a b c a b c Figure 29. Structure of Na2Y(PO4)(MoO4) projected onto the ac plane (a) (MoO4 tetrahedra are omitted) and the arrangement of the Na atoms in the structure (b). Quite recently, sodium lanthanum phosphate thiosulfate with composition Na2La(PO4)(S2O3) was prepared by the reaction of La2O3 with Na3PO3S in an excess of Na2S4.158 This compound crystallises in the monoclinic system (space group P21/c, Z=4, unit cell parameters a=13.779, b=5.336, c=9.356A, b=89.97 8). In the crystal structure of Na2La(PO4)(S2O3), the LaO7S2 polyhedra are linked in chains through the apical sulfur atoms of the thiosulfate groups.The chains are linked in layers by the PO4 and S2O3 tetrahedra (Fig. 30 a, b). The negatively charged [La(PO4)(S2O3)]27 fragment is the major building block of each layer. The [La(PO4)(S2O3)]n layers are parallel to the bc plane. The terminal sulfur atoms of the thiosulfate ions are located within these layers. The structure contains three types of sodium atoms. The Na(1) and Na(2) atoms are located in the interlayer space occupying the centres of the octahedra, which are formed by the oxygen atoms belonging only either to the phosphate [Na(1)] or III a PO4 YO8 PO4 MoO4 YO8 Na a PO4 a c b b a c b Figure 30. Structure of Na2La(PO4)(S2O3) projected onto the bc plane (a) and the chains of the LaO7S2 polyhedra linked through the PO4 and S2O3 tetrahedra (b).thiosulfate [Na(2)] groups. The Na(3) atoms are in an octahedral environment formed by the oxygen atoms belonging to four S2O23 ¡ anions and one PO34 ¡ anion. A new structural type for the Na2La(PO4)(S2O3) compound would be expected to occur by analogy with the Na2Y(PO4). .(MoO4) compound combining the PO3¡ and MoO2¡ anions. However, analysis of the polyhedra demonstrated that the struc- ture of Na2La(PO4)(S2O3) is derived from glaserite K3Na(SO4)2 . The difference is that one-half of positions, which are occupied by the alkali-metal cations in the structures of glaserite-like double phosphates [for example, in K3MIII(PO4)2, where MIII=Lu, Sc], are vacant in the structure of Na2La(PO4)(S2O3). These changes in the cation composition lead to the appearance of a superstructure in Na2La(PO4)(S2O3) resulted from doubling of the unit cell parameter a corresponding to the parameter c in the glaserite structure.In addition, the incorporation of the S2O23 ¡ ion into complex phosphate has a substantial effect on thermal stability of the Na2La(PO4)(S2O3) compound. Thus, the decomposition tem- perature of the Na2La(PO4)(S2O3) compound is 313 8C, whereas this temperature for Na3La(PO4)2 is much higher (1000 8C). The data from powder X-ray diffraction analysis and the results of IR and Raman spectroscopy for the decomposition product of Na2La(PO4)(S2O3) (after TGA) demonstrated that sodium lan- thanum phosphate thiosulfate decomposes to form a compound of unknown composition containing the PO3¡ and SO2¡ ions in the anion sublattice.Recently, a series of new phases with composition Li37xFe2(PO4)(SO4)2 were prepared.159 In these compounds, x depends on the content of iron in oxidation states 2+ and 3+. The redox potential of the LiFe2(PO4)(SO4)2 compound (3.4 eV) is higher than that of its monoanion precursor Li3Fe2(PO4)3 647 LaO7S2 S2O3 LaO7S2 Na(1) S Na(2) Na(3) PO4 S2O3 4 4 4 4648 (2.8 eV), and this compound is a promising electrodic material for recharging lithium batteries. III The above-considered examples demonstrate that the MIxMy (PO4ÜÖEO4Ü2 compounds possess new or improved prop- erties (important for technical applications) as compared to monoanion complex phosphates.The total negative charge of the anionic portion of the compound can be changed by replacing the PO4 groups in the anion sublattices of complex phosphates by the tetrahedral EO4 groups (E=V, Mo, S), which allows one to change the number of cations and size characteristics of the cavities of framework structures and of the interlayer spaces in layered phases. VII. Conclusion 4 with MIII : PO3¡ the ratio or 4 +EO24 ¡) equal to 1 : 2, for example, MI3MIIIÖPO4Ü2, A comparison of the crystal structures of different groups of complex phosphates containing mono- and trivalent cations not only reveals the similarity of formal characteristics (composition, simplest structural polyhedra) among these compounds but also enables one to establish their structural features depending on the size factors (ionic radii of the MI and MIII cations), nature of the cations and anions, combinations of different anions, presence of water molecules and OH groups and hydrogen bonding.Compositions MIII : (PO3¡ MIMIII(HPO4)2 and MI IIIÖPO4ÜÖEO4Ü, are typical of simple 2M double phosphates, hydrogen phosphates and phosphates whose anionic portions contain the PO4 groups together with other EO4 tetrahedra (EO4=MoO4, S2O3). Simple hydroxide phosphates and fluoride phosphates have related structures corresponding to the formulas MIMIII(OH)PO4 and MIMIIIFPO4. A limited num- ber of compositions and structural types are known for simple double phosphates.A complication of the anionic portion of double phosphates and the presence of water molecules in these compounds lead to the formation of complex phosphates of different types (fluoride phosphates, hydrogen phosphates con- taining different HnPO34¡n anions and hydroxide phosphates), which differ both in composition and structure. In most represen- tatives of complex phosphates containing mono- and trivalent cations, the ionic radius of the MIII cation is no larger than 0.8A (Al, V, Fe, Cr, Sc, Ga, In). Compounds containing REE or Y are known only for double phosphates and fluoride phosphates. Changes in the ionic radii of the MI and MIII cations (particularly, in the REE series) both in compositionally different and analogous groups of compounds are accompanied by a change in the structural type.In the case of sodium derivatives, which are most prone to polymorphism, the structural changes associated with a change in the size of theMIII cation and synthesis conditions are most pronounced (see Fig. 1 and Table 3). The structures of complex phosphates of all the above- considered types are built from the MIIIOx, MIIIOx(H2O)y or MIIIOx7y[F,(OH)]y polyhedra (x=6 ± 8 and y=2 ± 4) and the PO4 orHnPO4 polyhedra (the EO4 tetrahedra can partially replace the PO4 tetrahedra). The polyhedra about the MIII cations with the ionic radii r40.8A are octahedra distorted to a greater or lesser extent. The vertices of the octahedra are occupied by the oxygen atoms of the PO4, HnPO4, EO4 or OH groups or water molecules.In the case of fluoride phosphates, the F atoms can also be involved in the coordination polyhedra. In the structures of double REE phosphates and fluoride phosphates, the REE atoms adopt the coordination numbers of 7, 8 and 9. 4 4 4 The partial replacement of the PO3¡ anions by the HnPO3¡n (n=1, 2), EOn¡ 4 , F7 or OH7 anions and changes in the ratios r(MI) : r(MIII), MI:MIII and MIII : PO4 (HnPO3¡n, n=0 ± 2) lead to the formation of more complex structural ensembles based on either framework or layered constructions. Layered structures are typical of compounds with a high content of theMI (orMI+H+) cations with respect to the MIII cations.This relationship is observed in most of the group of compounds considered above, viz., in double phosphates MI IIIÖPO4Ü2 , fluoride phosphates 3M L N Komissarova,MG Zhizhin, A A Filaretov MI IIIF2ÖPO4Ü2 , hydrogen phosphates MIMIII 5M 3 ÖH2PO4Ü6. .ÖHPO4Ü2 and phosphates containing an EO4 tetrahedron along with the PO4 tetrahedron, for example, in Na2Y(PO4)(MoO4) and Na2La(PO4)(S2O3). One-dimensional motifs are observed in com- pounds characterised by a high content of alkali-metal cations and the presence of hydroxy groups, for example, in MI IIIÖOHÜÖHPO4ÜÖPO4Ü andMI4MIIIÖOHÜÖPO4Ü2 . 3MA decrease in the MI:MIII ratio results in the formation of framework structures typical of most of representatives of these groups of compounds, in particular, of MI III 3M2 ÖPO4Ü3 , MIMIII(HPO4)2, MIMIII(OH)(PO4), and MIMIIIFPO4 MI III 3M2 F3(PO4)2.In framework structures, the MIIIOx or MIIIOx7y[F,(OH)]y polyhedra are linked to each other through either tetrahedral groups (in double and hydrogen phosphates) or fluoride and hydroxy groups (in fluoride phosphates and hydrox- ide phosphates, respectively). 4 The character of the structures depends substantially on the presence of the HnPO3¡n cations, OH groups and H2O molecules in the coordination sphere about the MIII cation, which is associated with the involvement of these groups and molecules in the formation of rather strong hydrogen bonds. 4 4 Complex phosphates can be used in the design of polyfunc- tional materials.Compounds of REE and Sc containing the PO3¡ anions together with the VO3¡ anions are of particular interest. The formation of solid solutions based of double REE- and Sc- containing phosphates and vanadates allows one to vary the luminescent, sorption and other properties of the materials rather accurately by changing the PO3¡ 4 : VO34 ¡ ratio. Phosphate vana- dates selective absorb H2O, CO2 and various organic molecules from the gas phase. Europium-doped compounds of this type containing Sc, Y and REE possess luminescent properties. It is also highly probable thatREEfluoride phosphates and phosphate vanadates will exhibit catalytic properties. Phosphate vanadates of REE can be used also in the construction of nonlinear optical materials.Phosphates with composition MI III 3M2 ÖPO4Ü3 (MI=Li, Na; MIII=Sc, In, Fe) are ionic conductors. Some representatives of hydrogen phosphates possess proton conductivity. A number of these compounds (with zeolite-type structures) are of interest as new ion exchange phases. The fact that most of the complex phosphates belong to dielectric compounds provides the basis for the preparation of dielectric materials for different purposes. * * * When this review was already in preparation, new data on the synthesis, crystal structures and properties of two groups of complex phosphates, viz., MIMIIIÖHPO4Ü2 and MIMIII 2 ÖOHÜ. .ÖPO4Ü2 .2H2O, were reported. Thus, we have expected that indium hydrogen phosphates MIIn(HPO4)2 (MI=Li, Na, Ag, Rb),160 scandium hydrogen phosphates MISc(HPO4)2 (MI=K, Rb, NH4) and iron hydrogen phosphate b-RbFe(HPO4)2 would occur, and this assumption was confirmed.All these compounds, except for KSc(HPO4)2 (space group Pnma), crystallise in the monoclinic system (space group P21/c). The above-listed hydro- gen phosphates either widen the range of representatives of the known structural type b-RbV(HPO4)2 (see Ref. 62), for example, b1-RbIn(HPO4)2 , b-RbFe(HPO4)2 and MISc(HPO4)2 , where MI=K, Rb, NH4, or form unique framework structures, for example,MIIn(HPO4)2 , whereMI=Li, Na, Ag. Our assumption that the structural group of the mineral leucophosphite must involve also rubidium hydroxide phosphate RbGa2ÖOHÜ. .ÖPO4Ü2 .2H2O was also confirmed.161 References 1.P P Mel'nikov, L N Komissarova Koord. Khim. 12 1299 (1986) a 2. I A Bondar', N V Vinogradova, L N Dem'yanets Soedineniya RZE. Silikaty, Fosfaty, Arsenaty, Vanadaty (Compounds of REE. Silicates, Phosphates, Arsenates and Vanadates) (Moscow: Nauka, 1983)Complex phosphates containing mono- and trivalent cations 3. I V Tananaev, V P Orlovskii Zh. Neorg. Khim. 31 1923 (1986) b 4. T A Slivko, I N Smirnova, F M Spiridonov, G V Zimina, N N Chudinova Zh. Neorg. Khim. 47 202 (2002) b 5. I Szczygiel, T Znamierowska J. Solid State Chem. 95 260 (1991) 6. G V Zimina, I N Smirnova, S I Kudryashova, F M Spiridonov, I F Poletaev Zh. Neorg. Khim. 35 2134 (1990) b 7. W Jungowska, T Znamierowska J. Solid State Chem. 95 265 (1991) 8.B N Litvin, V A Masloboev Redkozemel'nye Fosfaty (Rare-Earth Phosphates) (Leningrad: Nauka, 1989) 9. P P Mel'nikov, L N Komissarova Dokl. Akad. Nauk SSSR 256 878 (1981) c 10. A I Kryukova, I A Korshunov, E P Moskvichev, V A Mitrofanova, N V Vorob'eva, G N Kazantsev, O V Skiba Zh. Neorg. Khim. 21 2560 (1976) b 11. A I Kryukova, I A Korshunov, V A Mitrofanova, N V Vorob'eva, G N Kazantsev, O V Skiba Zh. Neorg. Khim. 22 2301 (1977) b 12. L N Komissarova, P P Mel'nikov, Sh B Ibragimov, E G Zaugol'nikova Dokl. Akad. Nauk SSSR 295 911 (1987) c 13. C Parent, G Demazeau, R Salmon, G Le Flem Rev. Chim. Miner. 16 548 (1979) 14. K-H Lii Eur. J. Solid State Inorg. Chem. 33 519 (1996) 15. C E Bamberger, P R Robinson, R L Sherman Inorg.Chim. Acta 34 L203 (1979) 16. T A Butuzova, Candidate Thesis in Chemical Sciences, Moscow State University, Moscow, 1983 17. M Toumi, L Smiri-Dogguy, A Bulou Eur. J. Inorg. Chem. 1545 (1999) 18. K Okada,U Ossaka Acta Crystallogr., Sect. B, Struct. Crystallogr. 36 919 (1980) 19. M T Robinson J. Phys. Chem. 62 925 (1958) 20. J K Nimmo Acta Crystallogr., Sect. B, Struct. Crystallogr. 37 431 (1981) 21. B I Lazoryak Usp. Khim. 65 307 (1996) [Russ. Chem. Rev. 65 287 (1996)] 22. V A Efremov, P P Melnikov, L N Komissarova Rev. Chim. Miner. 22 666 (1985) 23. R D Shannon Acta Crystallogr., Sect. A, Fund. Crystallogr. 32 751 (1976) 24. V A Morozov, B I Lazoryak, A P Malakho, K V Pokholok, S N Polyakov, T P Terekhina J. Solid State Chem. 160 377 (2001) 25.V B Kalinin, S Yu Stefanovich, in Khimiya Tverdogo Tela (Itogi Nauki i Tekhniki) [Solid State Chemistry (Advanced in Sciences and Engineering Series)] (Moscow: Izd. VINITI, 1992) p. 51 26. V A Efremov, P P Mel'nikov, L N Komissarova Koord. Khim. 17 467 (1981) a 27. H Y-P Hong, S R Chinn Mater. Res. Bull. 11 421 (1976) 28. L N Komissarova, A P Bobylev, A N Kirichenko, G Ya Pushkina, F M Spiridonov Zh. Neorg. Khim. 47 684 (2002) b 29. M G Zhizhin, A A Filaretov, A V Olenev, V V Chernyshev, F M Spiridonov, L N Komissarova Kristallografiya 47 800 (2002) e 30. F Mazzi, L Ungaretti Neues Jahrb. Miner. Monatsh. 2 49 (1994) 31. M G Zhizhin, L N Komissarova, V A Morozov, F M Spiridonov Phosph. Res. Bull. 11 102 (2000) 32. MG Zhizhin, V A Morozov, A P Bobylev, AMPopov, F M Spiridonov, L N Komissarova, B I Lazoryak J.Solid State Chem. 149 99 (2000) 33. R Salmon, C Parent,M Vlasse, G Le Flem Mater. Res. Bull. 13 439 (1978) 34. R Salmon, C Parent,M Vlasse, G Le Flem Mater. Res. Bull. 14 85 (1979) 35. A Arakcheeva, G Chapuis, M Zhizhin, in Aperiodic Structures: Satellite Meeting of 20th European Crystallographic Meeting (Krakow: EJB, 2001) p. 296 36. M Pintard-Scre pel, F d'Yvoire, J Durand Acta Crystallogr., Sect. C, Cryst. Struct. Commun. 39 9 (1983) 37. B A Maksimov, L A Muradyan, E A Genkina, V I Simonov Dokl. Akad. Nauk SSSR 288 634 (1986) c 38. E A Genkina, L A Muradyan, B A Maksimov Kristallografiya 31 595 (1986) d 39. D Tran Qui, S Hamdoune Acta Crystallogr., Sect.C, Cryst. Struct. Commun. 43 397 (1987) 40. C Masquelier, C Wurm, J Rodriguez-Carvajal, J Gaubicher, L Nazar. Chem. Mater. 12 525 (2000) 649 41. T Suzuki, K Yoshida, K Uematsu, T Kodama, K Toda, Z-G Ye, M Ohashi,M Sato Solid State Ion. 115 89 (1998) 42. E A Genkina, L A Muradyan, B A Maksimov, B V Merinov, S E Sigarev Kristallografiya 32 74 (1987) d 43. N Fanjat, J L Soubeyroux J. Magn. Magn. Mater. 104 933 (1992) 44. V A Efremov, V B Kalinin Kristallografiya 23 703 (1978) d 45. K-H Lii, J Ye J. Solid State Chem. 131 131 (1997) 46. E A Genkina, V B Kalinin, B A Maksimov, A M Golubev Kristallografiya 36 1126 (1991) d 47. B I Lazoryak, V B Kalinin, S Yu Stefanovich, V A Efremov Dokl. Akad. Nauk SSSR 250 861 (1980) c 48.J Winand, A Rulmont, P Tarte J. Mater. Sci. 25 4008 (1990) 49. J Goodenough, H Hong, J Kafalas Mater. Res. Bull. 11 203 (1976) 50. D Antenucci, G Miehe, P Tarte, W W Schmahl, A-M Fransolet Eur. J. Miner. 5 207 (1993) 51. M Kloss, B Finke, L Schwarz,D Haberlandl J. Lumin. 72 684 (1997) 52. M Vlasse, C Parent, R Salmon, G Le Flem, P Hagenmuller J. Solid State Chem. 35 318 (1980) 53. C Parent, P Bochu, A Daoudi, G Le Flem J. Solid State Chem. 42 190 (1982) 54. L N Zorina, Zh A Ezhova, V P Orlovskii, EMKoval', N P Soshchin Izv. Akad. Nauk SSSR, Neorg. Mater. 15 1116 (1979) e 55. Y Miyajima, T Miyoshi, J Tamaki, M Matsuoka, Y Yamamoto, C Masquelier,M Tabuchi, Y Saito, H Kageyama Solid State Ion. 124 201 (1999) 56. A A Filaretov, M G Zhizhin, A V Olenev, A A Gurkin, A P Bobylev, B I Lazoryak, L N Komissarova, V P Danilov Zh.Neorg. Khim. 47 1930 (2002) b 57. J-X Mi, Y-X Huang, S-Y Mao, X-D Huang, Z-B Wei, Z-L Huang, J-T Zhao J. Solid State Chem. 157 213 (2001) 58. K-H Lii Inorg. Chem. 35 7440 (1996) 59. S M Stalder, A P Wilkinson J. Mater. Chem. 8 261 (1998) 60. K-H Lii, L-S Wu J. Chem. Soc., Dalton Trans. 1577 (1994) 61. R C Haushalter, Z W Wang,M E Thompson, J Zubieta Inorg. Chim. Acta. 232 83 (1995) 62. R C Haushalter, Z W Wang,M E Thompson, J Zubieta Inorg. Chem. 32 3700 (1993) 63. S S Dhingra, R C Haushalter J. Solid State Chem. 112 96 (1994) 64. N Anisimova, N Chudinova, R Hoppe, M Serafin Z. Anorg. Allg. Chem. 623 39 (1997) 65. S Dick, U Gossner, A Weiss, C Robl, G Grossmann, G Ohms, M Mueller J.Solid State Chem. 132 47 (1997) 66. Z Bircsak,W T A Harrison Acta Crystallogr., Sect. C, Cryst. Struct. Commun. 54 1195 (1998) 67. I Vencato, E Mattievich, L deF Moreira, Y P Mascarenhas Acta Crystallogr., Sect. C, Cryst. Struct. Commun. 45 367 (1989) 68. O V Yakubovich Kristallografiya 38 43 (1993) d 69. M P Attfield, A K Cheetham, S Natarajan Mater. Res. Bull. 35 1007 (2000) 70. A Goni, L Lezama, A Espina, C Trobajo, J R Garcia, T Rojo J. Mater. Chem. 11 2315 (2001) 71. S Ekambaram, C Serre, G Ferey, S C Sevov Chem. Mater. 12 444 (2000) 72. D H Olsen J. Phys. Chem. 74 2758 (1970) 73. D Brodalla, R Kniep Z. Naturforsch., B Chem. Sci. 35 403 (1980) 74. W P Bosman, P T Beurskens, J M Smits, H Behm, J Mintjens, W Meisel, J C Fuggle Acta Crystallogr., Sect.C, Cryst. Struct. Commun. 42 525 (1986) 75. N Anisimova, N Chudinova,M Serafin Z. Anorg. Allg. Chem. 623 1708 (1997) 76. A Mgaidi, H Boughzala, A Driss, R Clerac, C Coulon J. Solid State Chem. 144 163 (1999) 77. O V Yakubovich, M S Dadashov Kristallografiya 33 848 (1988) d 78. P B Moore, T Araki Am. Miner. 64 587 (1979) 79. G V Zimina, I N Smirnova,M Yu Gorkovenko, F M Spiridonov, L N Komissarova, N I Kaloev Zh. Neorg. Khim. 39 1567 (1994) b 80. G V Zimina, I N Smirnova,M Yu Gorkovenko, F M Spiridonov, L N Komissarova Zh. Neorg. Khim. 40 1578 (1995) b 81. L A Groat,M Raudsepp, F C Hawthorne, T S Ercit, B L Sherriff, J S Hartman Am. Miner. 75 992 (1990) 82. V I Simonov, N V Belov Kristallografiya 3 428 (1958) d 83.E A Genkina, Yu K Kabalov, B A Maksimov, O K Mel'nikov Kristallografiya 29 50 (1984) d 84. S Lathi Am. Miner. 70 849 (1985)650 85. O V Yakubovich, M A Simonov, O K Mel'nikov Kristallografiya 29 484 (1984) d 86. W T A Harrison,M L F Phillips, G D Stucky Chem. Mater. 7 1849 (1995) 87. J-M Le Meins, M-P Crosnier-Lopez, A Hemon-Ribaud, G Courbion J. Solid State Chem. 148 260 (1999) 88. O V Yakubovich, O K Mel'nikov Kristallografiya 41 663 (1996) d 89. N S Slobodyanik, P G Nagornyi, Z I Kornienko, A A Kapshuk, V V Mitkevich Zh. Neorg. Khim. 34 2473 (1989) b 90. O V Yakubovich, O K Mel'nikov, V S Urusov Dokl. Akad. Nauk SSSR 342 615 (1995) c 91. D M Poojary, A Clearfield, V A Timofeeva, S E Sigaryov Solid State Ion.73 75 (1994) 92. P G Nagornyi, A A Kapshuk, Z I Kornienko, V V Mitkevich, S M Tret'yak Zh. Neorg. Khim. 35 839 (1990) b 93. R K Rastsvetaeva, B A Maksimov, V A Timofeeva Dokl. Akad. Nauk 350 499 (1996) c 94. A M Golubev, B A Maksimov, N E Klokova, O K Mel'nikov, V A Timofeeva, N I Sorokina, V I Simonov Kristallografiya 34 1574 (1989) d 95. B Maximov,M Sirota, S Werner,H Schulz Acta Crystallogr., Sect. B, Struct. Crystallogr. 55 259 (1999) 96. J Arlt,M Jansen, H Klassen, G Schimmel, G Heymer Z. Anorg. Allg. Chem. 547 179 (1987) 97. M Mathew, I Mayer, B Dickens, L W Schroeder J. Solid State Chem. 28 79 (1979) 98. M Toumi, L Smiri-Dogguy, A Bulou J. Solid State Chem. 149 308 (2000) 99. I Mayer, S Cohen J. Solid State Chem. 48 17 (1983) 100. D Zambon, M El-Ghozzi, D Avignant, N Ben Said, I Mansouri J. Chim. Phys. 88 1963 (1991) 101. O L Slovokhotova, G D Ilyushin, N S Triodina, O K Mel'nikov, L N Dem'yanets, R G Gerr, V G Tsirel'son Zh. Neorg. Khim. 32 103 (1991) b 102. O V Yakubovich, M A Simonov, E N Matvienko, N V Belov Dokl. Akad. Nauk SSSR 238 576 (1978) c 103. F Taulelle, T Loiseau, J Maquet, J Livage, G Fgrey J. Solid State Chem. 105 191 (1993) 104. G Fe rey, T Loiseau, P Lacorre, F Taulelle J. Solid State Chem. 105 179 (1993) 105. T Loiseau, Y Calage, P Lacorre, G Ferey J. Solid State Chem. 111 390 (1994) 106. T Loiseau, C Paulet, N Simon, V Munch, F Taulelle, G Ferey Chem. Mater. 12 1393 (2000) 107. R Yu, D Wang, T Takei, H Koizumi, N Kumada, N Kinomura J. Solid State Chem. 157 180 (2001) 108. W Massa, O V Yakubovich, O V Karimova, L N Dem'yanets Acta Crystallogr., Sect. C, Cryst. Struct. Commun. 51 1246 (1995) 109. L Yu,W Pang Gaodeng Xuexiao Huaxue Xuebao 12 1137 (1991) 110. M G Zhizhin, A V Olenev, F M Spiridonov, L N Komissarova, O G D'yachenko J. Solid State Chem. 157 8 (2001) 111. M A Strelkov, M G Zhizhin, F M Spiridonov, A P Bobylev, L N Komissarova Zh. Neorg. Khim. 46 1629 (2001 b 112. S J Kirkby, A J Lough, G A Ozin Z. Kristallogr. 210 956 (1995) 113. E L Belokoneva, O V Yakubovich, V G Tsirel'son, V S Urusov Dokl. Akad. Nauk SSSR 310 1129 (1990) c 114. E L Belokoneva, O L Slovokhotova,MYu Antipin, V GTsirel'son, Yu T Struchkov Dokl. Akad. Nauk 322 520 (1992) c 115. N S Slobodyanik, P G Nagornyi, Z I Kornienko, A A Kapshuk Zh. Neorg. Khim. 36 1390 (1991) b 116. G D Stucky,M L F Phillips, T E Gier Chem. Mater. 1 492 (1989) 117. K-H Lii J. Chem. Soc., Dalton Trans. 815 (1996) 118. P E Machie, R A Young J. Appl. Crystallogr. 6 26 (1973) 119. M Tachihante,D Zambon, J-C Cousseins Eur. J. Solid State Inorg. Chem. 33 713 (1996) 120. Powder Diffraction File Card 42-1412, JCPDS International Center of Diffraction Data, Swartmore, PA 121. P B Moore Am. Miner. 57 397 (1972) 122. S Dick, T Zeiske J. Solid State Chem. 133 508 (1997) 123. P Dunn Am. Miner. 69 374 (1984) 124. O V Yakubovich, M S Dadashov Kristallografiya 37 1403 (1992) d 125. M Cavellec, D Riou, G Ferey Acta Crystallogr., Sect. C, Cryst. Struct. Commun. 50 1379 (1994) L N Komissarova,MG Zhizhin, A A Filaretov 126. D Cozzupoli, O Grubessi, A Mottana, P F Zanazzi Miner. Petrol. 37 1 (1987) 127. J A Hriljac, C P Grey, A K Cheetham, P D VerNooy, C C Torardi J. Solid State Chem. 123 243 (1996) 128. V Soghomonian, L A Meyer, R C Haushalter, J Zubieta Inorg. Chim. Acta 275 ± 276 122 (1998) 129. R N Devi, K Vidyasagar J. Chem. Soc., Dalton Trans. 3841 (1999) 130. J Parise Acta Crystallogr., Sect. C, Cryst. Struct. Commun. 40 1641 (1984) 131. J J Pluth, J V Smith, J M Bennett, J P Cohen Acta Crystallogr., Sect. C, Cryst. Struct. Commun. 40 2008 (1984) 132. M P Attfield, R E Morris, I Burshtein, C F Campana, A K Cheetham J. Solid State Chem. 118 412 (1995) 133. K-H Lii, S-L Wang J. Solid State Chem. 128 21 (1997) 134. MA G Aranda, S Bruque, J P Attfield, F Palacio, R B Von Dreele J. Solid State Chem. 132 202 (1997) 135. A A Filaretov, M G Zhizhin, L N Komissarova, V P Danilov, V V Chernyshev, B I Lazoryak J. Solid State Chem. 166 362 (2002) 136. P G Nagornyi, A A Kapshuk, N V Stus', N S Slobodyanik Kristallografiya 35 634 (1990) d 137. L Fanfani, A Nunzi, P F Zanazzi Miner. Mag. 37 598 (1970) 138. S Dick, G Grossmann, G Ohms, M Mueller Z. Naturforsch., B Chem. Sci. 52 1447 (1997) 139. R C L Mooney Acta Crystallogr. 20 526 (1966) 140. T Loiseau, G Ferey Eur. J. Solid State Inorg. Chem. 31 575 (1994) 141. F C Hawthorne Tschermaks Mineral. Petrogr. Mitt. 31 121 (1983) 142. H E King, L A Mundi, K G Strohmaier, R C Haushalter J. Solid State Chem. 92 1 (1991) 143. A Leclaire,M M Borel,A Grandin, B Raveau J. Solid State Chem. 108 177 (1994) 144. J B Parise Inorg. Chem. 24 4312 (1985) 145. M S Dadashov, Candidate Thesis in Geological and Mineralogical Sciences, Moscow State University, Moscow, 1986 146. P B Moore, T Araki Am. Miner. 59 964 (1974) 147. P B Moore, T Araki Am. Miner. 62 692 (1977) 148. H Monma J. Catal. 75 200 (1982) 149. H Owada,K Yamashita, T Umegaki, T Kanazawa Solid State Ion. 35 401 (1989) 150. L Hench J. Am. Ceram. Soc. 74 1487 (1991) 151. A Laghzizil, A Bouhaouss, P Barboux, R Morineau J Livage Solid State Ion. 67 137 (1993) 152. A Laghzizil, S El Hajjaji, A Bouhaouss, M Ferhat Solid State Ion. 126 245 (1999) 153. A N Kirichenko, S N Bessonov, V I Yurchenko, A P Bobylev, L N Komissarova Neorg. Mater. 35 877 (1999) e 154. A N Kirichenko, E N Otkidach, F M Spiridonov, V F Kharsika, L N Komissarova Zh. Neorg. Khim. 44 1061 (1999) b 155. V F Kharsika, L N Komissarova, A N Kirichenko, E N Murav'ev, V P Orlovskii, A P Chernyaev Neorg. Mater. 37 981 (2001) e 156. V F Kharsika, L N Komissarova, A N Kirichenko, E N Murav'ev, V P Orlovskii, A P Chernyaev Neorg. Mater. 37 1130 (2001) e 157. M Ben Amara,M Dabbabi Acta Crystallogr., Sect. C, Cryst. Struct. Commun. 43 616 (1987) 158. M G Zhizhin, F M Spiridonov, L N Komissarova, P K Dorhout, in The 20th European Crystallographic Meeting (Book of Abstracts), Krakow, 2001 S8.M1.P2 159. A K Padhi, V Manivannan, J B Goodenough J. Electrochem. Soc. 145 1518 (1998) 160. A A Filaretov, L N Komissarova, V P Danilov, in XXXVIII Vserossiisk. Nauch. Konf. po Problemam Matematiki, Informatiki, Fiziki, Khimii i Metodam Prepodavaniya Estestvennonauchnykh Distsiplin (Tez. Dokl.) [XXXVIIIth All-Russian Scientific Confer- ence on Mathematics, Information Science, Physics and Chemistry Problems and on the Methods of Teaching of Natural Sciences (Abstracts of Reports)] (Moscow: Peoples' Friendship University of Russia, 2002) p. 27 161. L Beitone, T Loiseau, G Ferey Acta Crystallogr. Sect. C, Cryst. Struct. Commun. 58 103 (2002) a�Russ. J. Coord. Chem. (Engl. Transl.) b�Russ. J. Inorg. Chem. (Engl. Transl.) c�Dokl. Chem. (Engl. Transl.) d�Crystallogr. Rep. (Engl. Transl.) e�Inorg. Mater. (Engl
ISSN:0036-021X
出版商:RSC
年代:2002
数据来源: RSC
|
2. |
Barium borateβ-BaB2O4as a material for nonlinear optics |
|
Russian Chemical Reviews,
Volume 71,
Issue 8,
2002,
Page 651-671
Pavel P. Fedorov,
Preview
|
|
摘要:
Russian Chemical Reviews 71 (8) 651 ± 671 (2002) Barium borate b-BaB2O4 as a material for nonlinear optics P P Fedorov, A E Kokh, N G Kononova Contents I. Introduction II. Barium borates. The BaO ± B2O3 system III. Glass formation in the BaO ±B2O3 system IV. Methods of preparation of BaB2O4 V. Polymorphism and the crystal structure VI. Peculiarities of crystallisation of BaB2O4 melts VII. Growth of b-BaB2O4 single crystals VIII. Defects and impurities in crystals IX. Constitutional supercooling X. Properties and fields of application of b-BaB2O4 crystals XI. Conclusion Abstract. equilibria, phase polymorphism, structure, the on Data Data on the structure, polymorphism, phase equilibria, growth single of applications and properties techniques, growth techniques, properties and applications of single crystals crystals of the low-temperature modification of barium borate are ana- of the low-temperature modification of barium borate are ana- lysed.The bibliography includes 201 references. lysed. The bibliography includes 201 references. I. Introduction Barium borates possess many interesting and valuable properties. Interest in these compounds is due to not only pure research aspects, but also the possibility of modern technological applica- tions. Currently, barium borates are used in production of ceramic glazes, luminophors, oxide cathodes as well as additives to pig- ments for aqueous emulsion paints.1± 8 The BaO ± B2O3 system (a constituent of the BaO ± B2O3 ± SiO2 system) serves as the basis when synthesising special optical glasses.9±14 The history of studies of barium borate, BaB2O4, can be divided into two periods.The pioneering communication con- cerning the synthesis of BaB2O4 single crystals dates back to the 19th century.1, 2 The structure of anions in crystalline borates was established by X-ray structural analysis in the 1950s.3, 4, 15 The crystal structures of a number of barium borates including the high-temperature modification of BaB2O4 were established in the 1960s;16 ± 18 however, these results were of low reliability. Two crystalline modifications { of BaB2O4, viz., the high- temperature (a) and the low-temperature (b) forms were described and characterised by X-ray diffraction analysis by Levin and McMurdie.12 P P Fedorov A V Shubnikov Institute of Crystallography, Russian Academy of Sciences, Leninsky prosp.59, 119991 Moscow, Russian Federation. Fax (7-095) 135 10 11. Tel. (7-095) 330 78 74. E-mail: ppf@newmail.ru A E Kokh, N G Kononova Institute of Mineralogy and Petrography, Siberian Branch of the Russian Academy of Sciences, prosp. Acad. Koptuga 3, 630090 Novosibirsk, Russian Federation. Fax (7-383) 233 27 92. Tel. (7-383) 233 39 47. E-mail: kokh@mail.ru Received 14 March 2002 Uspekhi Khimii 71 (8) 741 ± 763 (2002); translated by AMRaevsky #2002 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2002v071n08ABEH000716 651 652 655 655 657 659 660 665 666 667 668 In the anions, oxygen atoms surrounding borons can form two types of coordination environment, viz., a tetrahedral and a trigonal coordination environment corresponding to the sp3- and sp2-hybridisation of the electron orbitals of the B atom, respec- tively.19 As a rule, both of them are realised in borate ions.Both the BO3 triangles and BO4 tetrahedra can form fused structures by sharing their vertices (bridging oxygen atoms), thus forming polyanions. The presence of polymeric anions in borate melts is responsible for hysteresis of properties (in particular, viscosity) and the glass-forming ability.8 The second period of studies of BaB2O4 dates back to the 1980s and is associated with the discovery of nonlinear optical properties in the single crystals of the low-temperature modifica- tion of barium borate, b-BaB2O4 (BBO){ (see Refs 20 ± 25).Its crystals are non-centrosymmetric, which is responsible for a number of specific properties of the compound.26, 27 This form of barium borate combines excellent nonlinear optical character- istics [large effective second harmonic generation (SHG) coeffi- cient, which is nearly 6 times higher than that of potassium dihydrophosphate crystals; large birefringence; wide transparency region (190 ± 3500 nm), high laser damage threshold], good mechanical properties and very low hygroscopicity. Comparison of the properties of BBO with those of other optically active nonlinear crystals shows that barium borate is the best candidate for fabrication of solid-state UV lasers operating at a wavelength of 200 nm.BBO single crystals are used in optoelectronics for laser radiation-to-radiation frequency conversion (in particular, visible-to-UV frequency upconversion) by the SHG and higher harmonic generation (up to the fifth harmonic) mechanisms. { In the literature, the high-temperature and the low-temperature mod- ifications of barium borate are, as a rule, designated as a-BaB2O4 and b-BaB2O4, respectively (in this review, we also use these notations). However, quite the reverse notations are sometimes used. { The terminology is somewhat ambiguous. The compound having a composition BaO .B2O3 is usually treated as a salt, Ba(BO2)2, of metaboric acid (HBO2) and called barium metaborate;2 sometimes, the same com- pound is called barium diborate (see, e.g., Ref.4). However, barium diborate has another composition, namely, BaO .2B2O3 .2 To match the crystal structure of BBO, it is more correct to represent its composition as Ba3(B3O6)2 .652 Recently, frequency conversion by the stimulated Raman scatter- ing (SRS) mechanism using BBO-crystal oscillators has been studied.28 ± 30 In the last decade, new crystals possessing nonlinear optical properties (including borates) were synthesised.31 ± 39 These are, e.g., LiB3O5 (LBO) } (see Ref. 31), KBe2BO3F2 (KBBF),32 (SBBO),33 LiCsB6O10 Sr2Be2B2O7 (CLBO),34 Ba2Be2B2O7 (TBO) andK2Al2B2O7 (KABO) 35 and BiBO3 (BIBO);36 however, BBO remains of considerable value. Nonlinear optical properties of various borate single crystals are due to electron delocalisation 37, 38, 40 ± 43 in complex boron ± oxygen anions.The simplest examples of such anions are provided by the [BO3]37 triangles and by two types of ring trimers, viz., the [B3O6]37 units constructed of three BO3 triangles with two of the three vertices shared and the [B3O7]57 units (two BO3 triangles linked by the BO4 tetrahedron). They are shown in Fig. 1. Peculiarities of the electronic structure of borate crystals deter- mine high polarisability of these compounds, which is comparable to that of aromatics. These crystals also have good mechanical properties and high laser damage threshold, which is typical of inorganic crystals.44a b c d B [B2O3] (mass%) O Figure 1. Structural units present in borate crystals.BO3 (a); BO4 (b); B3O6 (c) and B3O7 (d). Solid-state optoelectronics (photonics) requires high-quality crystals that contain no inclusions, twins and inhomogeneities, are free from thermal strain and variations of the refractive index, have a definite crystallographic orientation and are, as a rule, rather large (of the order of a centimetre). BBO single crystals can be grown by melt-solution (high-temperature flux) technique using the modified Kyropoulos and Czochralski methods 45, 46 or by direct crystallisation from supercooled melts. Crystal growth is complicated by high viscosity and glass-forming ability of melts. The polymeric structure of borate melts can be to some extent controlled by shifting the acid ± base equilibria.47, 48 Low yield of materials suitable for practical applications, undesired variations of the properties of BBO crystals and high cost of optical elements are due to the fact that growing such crystals is a technologically sophisticated procedure.Recently, interest to centrosymmetric single crystals of the high-temperature a-modification of BaB2O4 has quickened.49 These crystals are characterised by large birefringence, broad optical transparency region (190 ± 3500 nm) and can compete with calcite crystals that are widely used in linear optics. An extensive literature on BBO crystals has evolved. Despite this fact, a number of topical problems concerning both the basic } The notations widely used in the literature are given in parentheses.(nonstoichiometry, polymorphism, phase equilibria, thermody- namics) and applied (elucidation of the reasons responsible for characteristic optical and structural inhomogeneities of BBO crystals, optimisation of crystal growth conditions) aspects of the investigation and growth of these crystals are still to be solved. II. Barium borates. The BaO ±B2O3 system Studies of the phase equilibria in the BaO ± B2O3±H2O system at 30 8C (Fig. 2) revealed the existence of barium metaborate tetrahydrate, BaO . B2O3 .4H2O, and barium triborate heptahy- drate, BaO .3B2O3 .7H2O.2, 50 Some other hydrated barium borates were reported in the early studies.2 More recently,4 BaO . B2O3 .5H2O was synthesised, which is formed in dilute solutions and is thought to be metastable,5 as well as BaB2O4 .1.67H2O.51 Determination of the crystal structures of BaO . B2O3 .4H2O (see Ref. 52) and BaO . B2O3 .5H2O (see Refs 53 and 54) revealed the presence of isolated tetrahedral B(OH)4 groups in these crystals; therefore, the formulae of the compounds can be written as Ba[B(OH)4]2 and Ba[B(OH)4]2 .H2O, respec- tively. The crystalline Ba[B5O8(OH)] .H2O and Ba5[B20O33(OH)4] .H2O were hydro- thermally synthesised in the BaO ± B2O3±M2O±H2O (M=K, Rb, Cs) systems at 280 8C under a pressure of 70 atm.55 ± 58 The melting diagram of the BaO ± B2O3 system was studied by thermal analysis.59, 60 Using the `composition ± property' curve, six congruently melting compounds were found.These are 3 BaO . B2O3 , 2 BaO . B2O3 , BaO . B2O3 , B2O3 . (1/2)H2O+L B A 3.5 3.0~~ 1.0 0.5 BaO . 3 B2O 3 . 7 H2O + L BaO.B2O3 . 4H2O+L C L 0 1.0 0.5 [B2O3] (mass%) 1050 5 Figure 2. Phase equilibria in the BaO ± B2O3±H2O system at 30 8C (a) 2, 50 and crystallisation field of BaB2O4 . nH2O (open circles) in sodium borate solutions at 25 8C (b).5 Points B, C and D denote eutonics (a); the dashed line designates the boundary between the tetrahydrate and pentahydrate regions (b). P P Fedorov, A E Kokh, N G Kononova compounds Ba2[B5O8(OH)2]OH, BaO .2B2O3 , a BaO. 9H2O+L D 5.5 [BaO] (mass%) 5.0 b10 [Na2O] (mass%) ~~Barium borate b-BaB2O4 as a material for nonlinear optics BaO .3B2O3 and BaO .4B2O3 (for the generalised phase diagram of the BaO ± B2O3 system, see Ref.2). It is believed that super- cooling of melts occurred in these studies. Phase equilibria in the BaO ±B2O3 system have been studied in detail 12 by annealing and quenching techniques and by the X-ray phase and thermal analyses. Only four congruently melting compounds were found, namely, 3 BaO . B2O3 , BaO . B2O3 , BaO .2B2O3 and BaO .4B2O3 . A melt immiscibility region was observed near the boron oxide composition. Barium metaborate, BaO . B2O3, melts congruently at 1095 8C and forms eutectics at 905 and 889 8C.12 In addition to the high- temperature a-modification of BaO . B2O3, the low-temperature b-modification was found and characterised by X-ray powder diffraction.No transition temperature was reported;12 it was only assumed to lie between 100 and 400 8C. More recently, in studies of the BaO ± B2O3 ± SiO2 ternary system13 the temperatures of a number of phase transformations in the BaO ±B2O3 system were corrected (increased by*10 8C) while the melting temperature of BaO . B2O3 was found to be 1105 8C. The generalised phase diagram of the BaO ± B2O3 system (Fig. 3), which was proposed by Levin and reported in a hand- book,10 has been commonly accepted by crystal growth research- ers (see, e.g., Refs 11, 27, 46). Following Feigelson et al.,46 in Fig. 3 we added a horizontal line (dash-and-dot line) correspond- ing to the polymorphic transformation of BaO . B2O3 at 925 8C (see below).It should be noted that the boundaries of the immiscibility region (dashed line) do not match those reported earlier.12 The coordinates of the upper critical point of the immiscibility dome (1931 8C, 6.6 mol.% BaO) were determined in a thorough study by Hageman and Oonk.61 T /8C 1923 1383 BaO + L L 3 BaO. B2O3+L 1370 1300 a-BaO. B2O3+L 1100 Two immiscible liquids 925 915 1095b-BaO. B2O3+L BaO. 2B2O3+L BaO. 4 B2O3+L 910 889 L1+L2 900 869 879 BaO. 4 B2O3+L + 700 Figure 3. Phase diagram of the BaO± B2O3 system.10 3 BaO. B2O3+BaO BaO. 2 B2O3+ b-BaO. B2O3 BaO. 2 B2O3 BaO. 4 B2O3 80 40 20 100 B2O3 60 [B2O3] (mass%) 0 BaO The middle part of the phase diagram of the BaO ± B2O3 system was revised by HuÈ bner 62 who showed that not only the above-mentioned compounds, but also 2 BaO .5B2O3 , 2 BaO .B2O3 and 4 BaO . B2O3 are formed in the system. All of the newly found compounds were characterised by X-ray powder diffraction. The congruent melting temperature of 2 BaO . B2O3 (9205 8C) is close to the temperature of the eutectic between BaO . B2O3 and 3 BaO . B2O3 .12, 13 The temperature of the poly- morphic transformation of BaO . B2O3 is 9255 8C. HuÈ bner proposed the existence of homogeneity regions for both modifi- cations of BaO . B2O3 and suggested that this region for the high- temperature form is especially broad (Fig. 4). According to this b-BaO. B2O3+ 3BaO. B2O3 653 T /8C a L a + L 930 900 a+b a+[2BaO .B2O3] b b+[2BaO. B2O3] 870 72 802BaO. B2O3 76 [BaO] (mass%) BaO. B2O3 Figure 4. Middle part of the phase diagram of the BaO ±B2O3 system.62 version of the phase diagram, the low-temperature modification of b-BaB2O4 is not in the equilibrium with the B2O3-enriched melt; however, this contradicts the results obtained more recently.63, 64 HuÈ bner 65 also studied the phase equilibria in the BaO ± B2O3±Al2O3 system and reported the formation of continuous solid solutions in the composition range between 2 BaO . B2O3 and the low-temperature modification of BaO . B2O3 at 700 8C (Fig. 5). The formation of individual compound 2BaO . B2O3 was established based on the X-ray diffraction data obtained in melt crystallisation studies in a number of ternary systems.66 Compound 2 BaO .B2O3 crystallises in the monoclinic crystal system (Table 1) and its unit cell parameters are close to those of b-BaB2O4 in the monoclinic setting (the difference consists in doubling the parameter c). This structural type is possible only if barium and oxygen ions are inserted into the b-BaB2O4 lattice, which is accompanied by a dramatic increase in density. Single crystals of the 2 BaO .5B2O3 (Ba2B10O17) phase found by HuÈ bner 62 were grown by Stone et al.68 by slowly cooling a stoichiometric melt. The solid-phase synthesis was carried out at 750 8C and the melting temperature of the compound was reported to be 901 8C. The melt undergoes a glass transition on cooling with ease.The crystal structures of compounds BaO .2B2O3 (BaB4O7), BaO .4B2O3 (BaB8O13) and 2 BaO .5B2O3 (Ba2B10O17) were established (see Table 1).16, 18, 68 At 700 8C, compound BaO .4B2O3 undergoes a phase transition from the low-temper- ature orthorhombic to the high-temperature tetragonal modifica- tion which is stable up to the congruent melting temperature (889 8C).67 The heat of the transition is 2100 J mol71. The phase transition is accompanied by twinning of crystals. An IR spectroscopic study 75 revealed the presence of isolated BO3 triangles in the structure of compound 3 BaO . B2O3 which is thus an orthoborate. The generalised phase diagram of the BaO ±B2O3 system is shown in Fig. 6. Since the melting temperature of compound Ba2B10O17 (see Ref.68) is somewhat higher than the liquidus temperature for this composition (see Ref. 12), this compound seems to be stable. According to the version of the phase diagram proposed by Levin and McMurdie,12 an eutectic corresponding to the crystallisation of BaO .2B2O3 and BaO .4B2O3 should occur in this concentration range; however, no eutectic effects were observed. Probably, they simply overlooked the compound 2 BaO .5B2O3 . Earlier,60 a congruently melting compound of other stoichiometric composition was found in this concentration range. The `interrelation' between compounds 2 BaO . B2O3 and BaO . B2O3 is still a moot question. The melting temperature of 2 BaO . B2O3 reported by HuÈ bner is much lower than the liquidus654 Figure 5.Phase equilibria in the BaO ±B2O3±Al2O3 system at 700 8C.65 Open circles denote the compositions studied; full circles denote compounds; SSI and SSII designate the regions of solid solutions; 2 BaO.4B2O3 . Al2O3 (1), BaO .2B2O3 . Al2O3 (2), BaO . B2O3 . Al2O3 (3), 3 BaO .2B2O3 . 2Al2O3 (4) and 5 BaO .6B2O3 .2Al2O3 (5). Table 1. Crystallographic characteristics of anhydrous barium borates. Compound BaB813 BaB47 Ba2B1017 b-BaB24 a-BaB24 Ba2B25 Ba3B26 a At 725 8C. b Doped with neodymium. c Recalculated by the authors of this review using HuÈ bner's data.62 BaO. 4B2O3 2 BaO. 5 B2O3 BaO. 2B2O3 b-BaO. B2O3 5 SSI 2 BaO. B2O3 3 BaO. B2O3 4 BaO. B2O3 BaO 3 BaO.Al2O3 Space group Crystal system orthorhombic P2221 tetragonal a monoclinic triclinic P21/c P1 R3c trigonal trigonal R3c P2, Pm or P2/m 77 monoclinic trigonal orthorhombic B2O3 1 2 3 4 SSII BaO.Al2O3 [Al2O3] (mass%) Unit cell parameters c /A b /A a /A 17.38 17.352 13.20 13.21 13.268 8.56 8.55 8.630 10.56(1) 8.20(1) 13.01(1) 9.858(1) 9.990(1) 6.706(1) 12.532 12.529 12.519 12.547(6) 12.500(3) 12.517(3) b 7.2188(10) 7.2351 12.717 12.731 12.723 12.736(9) 12.6875(9) 12.708(3) b 39.000(4) 39.192 11.014(6) 12.684(5) 16.856(7) 21.70 c 13.82 12.74 c 13.43 14.81 P P Fedorov, A E Kokh, N G Kononova 2Al2O3 .B2O3 BaO. 6Al2O3 Z a, b, g /deg 884 b=104.95(17) 82 a=96.79(1) b=106.64(1) g=76.89(1) 18 18 18 18 18 18 18 18 19 b=99.82(3) 16 9Al2O3 .B2O3 Al2O3 Ref. Density /g cm73 2.91 2.927 2.899 18 67 67 3.57 3.54 16 68 22 69 70 44 71 72 3.74 777773.734 3.751 5.07 17 73 62 5.11 74Barium borate b-BaB2O4 as a material for nonlinear optics T /8C L 1234 1500 Melt cooling in the BaO ± B2O3 system causes glass forma- tion.9, 11, 14 The most stable glasses (ingots up to 50 g by mass) are fabricated by solidification of melts containing from 16 mol.% to 40 mol.% BaO, i.e., in the region of existence of the low-melting compounds BaO .2B2O3 , 2 BaO .5B2O3 and BaO .4B2O3 (see Fig. 6). As the cooling rate increases, the region of glass formation extends (up to 60.5 mol.% BaO for ingots of mass 0.1 g) and includes the composition BaB2O4.No transparent homogeneous glasses are formed on cooling of B2O3-enriched melts.At temperatures that are somewhat higher than the liquidus temperature, the equilibrium melt represents two immiscible liquids of different compositions. Similarly to the melt, on cooling each of the liquids can undergo a glass transition. Boron oxide can undergo a spontaneous glass transition.9 Glassy boron oxide can be readily obtained from boracic acid which melts at 176 8C. Long-term heating is accompanied by slow removal of water from the melt; however, trace amounts of water are hard to remove. The content of the residual water after heating 1 : 1 1370 1100 879 86980 2 : 1 1 : 2 2 : 5 1 : 4 1383 L1+L2 1000 925 a 910 901 500BaO B2O3 20 40 60 [B2O3] (mol.%) Figure 6.Generalised phase diagram of the BaO ±B2O3 system. The data were taken from Ref. 12 (1), Ref. 64 (2), Ref. 61 (3) and Ref. 68 (4). temperatures determined earlier 12, 59 and, hence, cannot corre- spond to the equilibrium congruent melting of this compound. However, it is rather close to the eutectic temperature. The authors of this review suggest incongruent melting of 2 BaO . B2O3 following a peritectic reaction (see Fig. 6). HuÈ bner's assumption 62 of the presence of the homogeneity region based on BaO . B2O3 is in agreement with the absence of eutectic effect in this concentration range.12 Figure 6 presents a version of the phase diagram containing the regions of solid solutions based on a-BaB2O4 and b-BaB2O4 shown according to HuÈ bner.The presence of these regions suggests a strong non- stoichiometry of BaB2O4 . If this is the case, the maximum on the liquidus curve should not coincide with the point corresponding to the 1 : 1 composition. It should also be noted that the densities of both the a- and b-modifications of BaB2O4 are much lower than the average density obtained by linear approximation of the data for compounds in the BaO ± B2O3 system. This anomaly can be to some extent compensated for by the formation of interstitial solid solutions. This issue as well as the thermal stability of 4 BaO . B2O3 call for further investigation.III. Glass formation in the BaO ±B2O3 system 4 : 1 3 : 1 655 at 1200 8C for 19 h is 0.127 mass %. Pumping anhydrous dini- trogen through the melt over a period of 5 h reduces the content of water down to 0.025 mass %. A nearly completely dehydrated material was prepared by long-term heating at a pressure of 1 Torr in a platinum crucible and then in a graphite crucible. The viscosity of the melt depends on the content of water; as this parameter decreases from 0.248 mass% down to 0.025 mass %, the temperature corresponding to a viscosity of 104 P increases from 537 to 550 8C. It is noteworthy that the viscosity of B2O3 at 450 8C (melting temperature) is much lower than that of silica at 1710 8C; however, glassy boron oxide is much more hard to crystallise compared to glassy silica.This was assumed to be due to abnormally high ratio of the bond energy to the melting temper- ature of glassy B2O3.No spontaneous devitrification of B2O3 at an atmospheric pressure was observed. Crystallisation of boron oxide requires heating at 225 ± 250 8C over a period of several days in a loosely plugged vessel. The sample slowly loses water and is transformed into a solid finely crystalline aggregate character- ised by a density of 2.42 g cm73 and a melting temperature lying between 460 and 470 8C. The content of B2O3 in the material thus obtained is 99.4% while the rest seems to be the bound water. An X-ray study of glassy B2O3 showed that its structure is a disordered network of [BO3]37 triangles (in which each oxygen atom is linked to two boron atoms) or distorted [BO4] tetrahedra. Glass formation is also typical of the systems comprising B2O3 and alkali metal oxides.The Na2O±B2O3 system is characterised by two glass formation regions14 corresponding to the concen- tration ranges 0 mol.% ± 8 mol.% (or 0 mol.% ± 30 mol.%) and 66.5 mol.% ± 71.5 mol.% Na2O. The second glass formation region coincides with the region of existence of low-melting compositions in the vicinity of the NaBO2±Na4B2O5 eutectic. A `belt'-like glass formation region was found in the Na2O± BaO ±B2O3 system at the B2O3 concentration of greater than 60 mol.%.76 IV. Methods of preparation of BaB2O4 There are four groups of methods for preparation of BaB2O4 .These are precipitation of crystal hydrates from aqueous solutions followed by dehydration; hydrolysis of alkoxides; anhydrous solid-phase syntheses (possibly) followed by melting and melt- solution syntheses. The last-named group also includes various single-crystal growth techniques which can be considered as specific methods of synthesis. 1. Precipitation of BaB2O4 . nH2O from aqueous solutions Methods for the synthesis of barium borate based on precipitation of crystal hydrates BaB2O4 . nH2O from aqueous solutions (aqueous methods) have been known for long.2±8 However, they are rarely employed for the preparation of charge that is used as raw material for the crystal growth. Aqueous methods for the synthesis of different metal borates share a number of features. During precipitation from aqueous solutions the borates are isolated as crystal hydrates.The content of the bound water depends on the reaction conditions. In addition, metal borates can be isolated in the amorphous state. Such borates are intermediate products of many syntheses; never- theless, they can undergo a transition into the crystalline state under particular conditions. Amorphous borates are, as a rule, highly reactive. Borates can contain water in the form of both hydroxyl ion and solvate molecules. This allows one to rationalise a complex process of dehydration of borates at elevated temperatures which involves elimination of individual solvate water molecules and hydroxyl groups, which is due to association of boron-containing anions.Usually, removal of water occurs in a stepwise manner in the temperature range from 100 to 200 8C. The residual water can be removed only at very high temperatures (*1000 8C and above). Often, dehydration of borates is accompanied by a specific656 effect of borate rearrangement 77 which was first reported by Nikolaev.78 Namely, dehydration causes a complete destruction of the crystal lattice and transition of the compound under study into the amorphous state; further heating leads to crystallisation accompanied by exothermic effect, which can be detected on thermograms. The synthesis of BaB2O4 by precipitation from aqueous solutions should be carried out in a strongly basic medium (pH >11) provided that variations of the pH value during crystallisation are small.This can be done by performing a reaction between a solution of boric acid and a solution of Ba(OH)2 (see Ref. 4) or BaCl2 .51 However, it is more convenient to use barium chloride and a mixture of sodium tetraborate and sodium hydroxide. BaCl2+Na2B4O7+2NaOH+3H2O= =BaB2O4 .4H2O;+2NaBO2+2 NaCl. To maintain the pH value at the desired level, a nearly twofold excess of the precipitating agent is required. The reaction results in a flaky precipitate which is transformed into crystalline BaB2O4 .4H2O after several hours. The crystallisation fields of BaB2O4 .4H2O and BaB2O4 .5H2O in sodium borate solutions at 25 8C for the reaction presented above were studied by Gode.5 The possibility of nearly complete precipitation of boron (in the form of barium borate) from solutions was demonstrated (see Fig.2 b). A draw- back of this method consists in poisoning the precipitate by sodium and chlorine impurities. The synthesis of barium borate from barium nitrate was proposed 8 Ba(NO3)2+Na2B4O7+2NaOH+3H2O= =BaB2O4 .4H2O;+2NaBO2+2NaNO3. It was proposed to remove the chloride (impurity in NaOH) and sodium ions using an ammonia solution (to prevent transforma- tion into polyborates). Drying at 400 ± 650 8C for 30 min resulted in an anhydrous product. To transform non-crystalline barium borate into the crystal- line BaO . B2O3 .4H2O, it was recommended 6 to grind the mate- rial in a boric acid solution in the presence of KOH.The yield of the desired product was low (*20%). Dehydration of the tetrahydrate BaO . B2O3 .4H2O begins at 70 and is completed at 140 8C, while dehydration of the pentahy- drate BaO . B2O3 .5H2 occurs in the temperature range from 50 to 450 8C.4 It should be noted that the melting temperatures of the dehydrated products (1015 and 925 8C, respectively) 4 are much lower than the equilibrium melting temperatures. Probably, this is due to the influence of the residual water. According to Lehmann et al.,51 dehydration of BaO . B2O3 . 1.67H2O occurs in the tem- perature range 90 ± 170 8C. Precipitation of the tetrahydrate BaB2O4 .4H2O from solu- tion followed by thermal decomposition following the reactions(1) BaCl2 .2H2O+2H3BO3=BaB2O4 .4H2O;+2 HCl, (2) BaB2O4 .4H2O=BaB2O4+4H2O was employed 79 ± 82 for the synthesis of BaB2O4 that is used as raw material (charge) for the melt growth of b-BaB2O4 crystals (the so- called chloride technique).Remelting or sintering BaB2O4 at 1095 8C for 3 h results in a coarse-grained material of higher purity grade compared to the charge obtained by solid-phase synthesis from barium carbonate. A thermogram exhibiting a clearly seen peak corresponding to the melting at 1100 8C has been reported.82 It should be noted that the synthesis of BaB2O4 by thermolysis of crystal hydrates allows reliable preparation of a product of stoichiometric composition. In addition, it seems that aqueous solutions favour purification of barium-containing reagents P P Fedorov, A E Kokh, N G Kononova (removal of impurities including strontium and calcium that strongly affect the stability of the crystalline modifications of BaB2O4).2. Syntheses using `soft chemistry' methods Recently, the so-called `soft chemistry' methods have been used for the synthesis of inorganic materials (see, e.g., Ref. 83). These methods can also be employed to synthesise barium borate. For instance, a stoichiometric mixture of boron triethoxide and barium isopropoxide was slowly hydrolysed with distilled water.84 The powder dried in vacuo at 60 8C was found to be a new g-modification of BaB2O4. 3. Anhydrous high-temperature syntheses Compound BaB2O4 can be directly synthesised by melting BaO and B2O3 at 1150 8C for 1 to 2 h.85 According to the results of differential thermal analysis (DTA) study, the reaction begins at 1060 8C.Barium borate (the starting material for the synthesis of single crystals) is usually prepared from a mixture of BaCO3 and B2O3 (see Refs 86 ± 89) or H3BO3 .64, 80, 90 ± 92 The syntheses are con- ducted using the following reaction equations (3) BaCO3+B2O3=BaB2O4+CO2:, (4) BaCO3+2H3BO3=BaB2O4+3H2O:+CO2:. Powder mixtures taken in the stoichiometric ratios are mechanically mixed and then heated until sintering and melting. In the case of reaction (3) the powders are sintered at 800 8C for 10 h, melted at 1200 8C for 10 min and then zone melted (both processes are carried out in a graphite boat in an argon atmos- phere) 88 or are melted in air at 1150 8C for 5 h. For the reaction with H3BO3 it was recommended 92 to use a stepwise synthesis with allowance for the reaction mixture to stay at 150 and 700 8C.Usually, syntheses are performed in platinum crucibles; how- ever, graphite crucibles can also be used in an inert atmosphere (argon).88 It seems likely that one can synthesise BBO from barium carbonate according to Berzelius by boiling a barium carbonate suspension in an aqueous boric acid (analogously to the synthesis of lead borate 4). However, this synthetic route was not studied experimentally. Barium borate BaB2O4 is also synthesised using other barium compounds (e.g., barium oxide,79 barium acetate 90, 91 and barium nitrate 12, 82) as starting reagents.BaO+2H3BO3=BaB2O4+3H2O:, Ba(CH3COO)2+2H3BO3+4O2= =BaB2O4+6H2O:+4CO2:, Ba(NO3)2+2H3BO3=BaB2O4+N2O5:+3H2O:. The reaction with barium acetate was carried out in a platinum crucible using stepwise heating;91, 92 finally, the product was kept at 1150 8C for 50 h. The reaction proceeds in air; however, it does not always happen that the reaction goes to completion under the crystal growth conditions. This leads to poisoning the growing crystals with carbon. In the reaction with barium nitrate the ground mixture of reagents was heated to 700 8C for 2 h and then repeatedly heated above this temperature.12 The following reaction was conducted 82 Ba(NO3)2+B2O3=b-BaB2O4+N2O5:.A ground mixture of reagents was repeatedly heated from 700 to 850 8C. It was also found that the reactions of BaCO3 or BaO with H3BO3 do not result in products that are suitable for the Czochralski growth of b-BaB2O4.Barium borate b-BaB2O4 as a material for nonlinear optics 4. Melt-solution crystallisation Barium borate can be synthesised by exchange reactions in melts. The first successful synthesis dates back to 1874.1, 79 The following reaction was used BaCl2+2NaBO2=BaB2O4;+2 NaCl. Sodium chloride having a lower melting temperature acts as solvent that is removed with water after cooling. Reactions analogous to the processes underlying the aqueous methods of synthesis 93, 94 can also be used for the synthesis of b-BaB2O4 H2O BaCl2+B2O3 BaB2O4;+2 HCl:, BaCl2+2H3BO3=BaB2O4;+2 HCl:+2H2O:.The excess barium chloride acts as flux. The reaction was carried out in a platinum crucible at 1000 8C for 2 h.93 The driving force of the synthesis of BaB2O4 is the difference between the heat of reaction and the energy of pyrohydrolysis of barium chloride (*78 kcal). A method of fabrication of a seed rod necessary for melt- solution growth of single crystals is as follows.94 BaB2O4 poly- crystals precipitated in the reaction with boric acid are ground, moulded, sintered in an oxygen atmosphere at 730 ± 850 8C, re- ground and then re-sintered at 850 ± 1095 8C. An extensive literature on the synthesis of single crystals using the systems of the BaB2O4 ± S or, more generally, BaO ± B2O3±S types (S is the melt playing the role of solvent) has evolved.Both equilibrium and nonequilibrium melt crystallisations are possible (the former corresponding to the primary crystallisation region of b-BaB2O4 in the phase diagrams). Among various solvents, Na2O is most widely used at present (see below). Melt-solution growth of single crystals is strongly affected by the impurities coming into the crystal from the solvent. These are the melt components (first of all, sodium), impurities contained in the starting reagents and accumulated within the growing crystal (e.g., potassium or strontium) and `extra' impurities that come from the material the crucible is made of (platinum). V. Polymorphism and the crystal structure In addition to the high- and low-temperature modifications of BaB2O4 (these two were mentioned above) there are some indications of the existence of at least one unstable low-temper- ature modification designated as T-BaB2O4 (see Ref.51), or g-BaB2O4 .84, 95 The X-ray diffraction patterns of the samples reported in Refs 51, 84 and 95 are close but not identical. The samples of this phase were obtained by dehydration of BaB2O4 . 1.67H2O (see Ref. 51) and by the interaction of boron triethoxide and barium isopropoxide. According to Yamaguchi et al.,84 g-BaB2O4 undergoes an irreversible transformation into a-BaB2O4 on heating to *600 8C. An IR spectroscopic study of g-BaB2O4 showed that, in contrast to the structure of the a- and b-forms of BaB2O4 (see below), the structure of this modification comprises infinite anionic chains constructed of [BO4] tetrahedra with their vertices shared (calcium, strontium and lithium metabo- rates have an analogous structure).} The crystal structure of a-BaB2O4 was established by Mighell and Perloff.17 It was found that the crystal is characterised by the trigonal system and centrosymmetric space group R3c (Fig.7). According to HuÈ bner,62 b-BaB2O4 single crystals have a monoclinic unit cell with the parameters a=11.133, b=12.67, c=8.381A and b=100.04 8; the space group is C2/c and Z=12. Coincidence of the results of X-ray powder diffraction studies 12, 62 indicates that the same modification of barium borate was investigated in both cases. } In addition, two unstable modifications of BaB2O4 were obtained during crystallisation of glasses in the BaO ±B2O3±Al2O3 system.65 657 However, HuÈ bner's 62 choice of the unit cell was questioned.69 According to Liebertz and Stahr,69 b-BaB2O4 crystals are uniaxial and trigonal (the crystallographic parameters are listed in Table 1).This discrepancy is of fundamental nature, since the space group C2/c is centrosymmetric while the space group R3c is non-centrosymmetric. Frohlich70 established the crystal structure of b-BaB2O4 and confirmed the results obtained by Liebertz and Stahr.69 Lu et al.22 also reported similar results; however, they believed that the b-BaB2O4 crystal is characterised by the space group R3. The results obtained by Lu et al.22 were used in the interpretation of the Raman spectra of b-BaB2O4 crystals.97 Re- determination of the crystal structure 44, 71 and the results of piezoelectric and electro-optical measurements 38, 98 and Raman spectroscopy studies 99, 100 confirmed that the low-temperature modification of barium borate is characterised by the space group R3c.A thorough study 85 of BaB2O4 single crystals confirmed the lack of central symmetry and revealed a slight monoclinic dis- tortion of the trigonal lattice.The parameters of the trigonal lattice of b-BaB2O4 obtained by different authors are slightly different (see Table 1). The high- and low-temperature modifications of barium borate exhibit pronounced structural similarity, viz., both of them have layered structures, contain nearly planar boron ± oxy- gen [B3O6]37 ring groups arranged perpendicular to the three-fold axis while their lattice parameters are related by simple formulae pÅÅÅ aa& 3ab , ca&(1/3) cb , Za=Zb .The [B3O6]37 groups are constituents of the structure of only one out of the three known crystalline modifications of metaboric acid, namely, metastable orthorhombic HBO2-III with a melting temperature of 176 8C.3 Therefore, the a- and b-modifications of BaB2O4 are salts of this acid. High-temperature Raman spectroscopy studies 99, 100 of the a- and b-forms of BaB2O4 revealed a change in the number of spectral lines in the frequency range of internal vibrations of the [B3O6]37 anions corresponding to the D63h and C63v symmetry, respectively.The B7O distances in and out of the ring are*1.40 and 1.32A, respectively. The structures of the a- and b-modifications of BaB2O4 are characterised by different types of coordination environment of barium ions. In the high-temperature a-modification there are two types of barium atoms. The coordination number of barium atoms occupying the first position is 6 while for Ba atoms in the second position it equals 9. In the low-temperature b-modifica- tion the Ba atoms are in irregular environment of eight oxygen atoms shared by three ring borate anions. The Ba ±O distances, d (Ba ± O), lie between 2.613 and 3.025A. The X-ray powder diffraction patterns of the a- and b-modifications of BaB2O4 are presented in Fig.8. Investigation of the nonlinear optical properties of BBO is a topical problem. In this connection it is appropriate to briefly outline the nature of chemical bonding in this com- pound.38, 40 ± 43, 97 In the [B3O6]37 ring anions, the sp2-hybridised boron atoms form s-bonds with oxygen atoms. The filled orbitals of the oxygen atoms and the vacant orbitals of the boron atoms involved in the rings interact with one another, thus forming the p-bonds. Positively charged Ba2+ ions and negatively charged [B3O6]37 groups interact electrostatically. In addition, the inter- molecular interaction occurs between electrically neutral layers. A number of the first-principle studies of the electronic structure and nonlinear optical susceptibilities of BBO have been reported.42, 43 The results of calculations are in good agreement with the experimental data.The energy states of the atoms constituting the [B3O6]37 anions contribute to both the upper part and the bottom of the conduction band. Different temperatures of the phase transition between the b- and a-modifications of BaB2O4 have been reported (100 ± 400 8C,12 880 8C,103 *900 8C 44 and 917 8C 104). Cur- rently, a value of 9255 8C reported by HuÈ bner 62 and confirmed658 c a c in the annealing and quenching studies 23, 105 has been commonly accepted. The phase transition in question is strongly hindered. An X-ray diffraction pattern obtained after annealing of coarse- grained (40 ± 50 mm), single-phase (b-phase) crystalline BBO at 940 8C for 4 h exhibited only traces of the a-phase.87 Annealing at higher temperatures is accompanied by gradual decrease in the intensity of the reflections of the b-phase and by simultaneous increase in the intensity of the reflections of the a-phase.After annealing at 960 8C for 2 h nearly *50% of the initial b-phase transformed into the a-phase. Analogous studies were carried out using single-phase (a-phase) crystalline BBO. After annealing at *800 8C for 7 h nearly*30% of the initial a-phase transformed a b BBa c a P P Fedorov, A E Kokh, N G Kononova c BBa a Figure 7. Projections of the trigonal crystal structures of a-BaB2O4 (space group R3c) 17 (a), b-BaB2O4 (space group R3c) 71 (b) and NaBO2 (space group R3c, a=11.875, c=6.4375A) 96 (c) on the plane passing parallel and perpendicular to the c axis (different scales were used).Triangles denote the BO3 groups. The unit cells are shown by solid lines. BNa into the b-phase. At lower temperatures, the phase transformation occurred very slowly. During measurements 106 of the thermal expansion coeffi- cients the BBO single crystals were overheated up to 983 8C at a heating rate of 1 deg min71 and to 1010 8C at a heating rate of 0.1 deg min71. Transition into the a-phase occurred after reach- ing these temperatures. According to Kimura and Feigelson,107 the b?a transfor- mation takes only 5 min on heating above 1000 8C. It should be noted that the samples studied contained Sr (*0.05%) and Ca (*0.03%); it is thought that Ca impurity causes stabilisation of the a-phase.108Barium borate b-BaB2O4 as a material for nonlinear optics Figure 8.X-Ray powder diffraction patterns (CuKa radiation) of a-BaB2O4 (a), b-BaB2O4 (b) and NaBaBO3 (c).101, 102 A number of in situ studies of the b>a transformation by high-temperature X-ray powder diffraction have been reported.82, 84, 106 Kozuki and Itoh106 heated BBO powder to 1050 8C but no phase transition was observed. Yoshimoto and Kimura 82 also did not observe the b?a and a?b phase transitions upon heating to 1000 8C. According to Yamaguchi et al.,84 transformation of b-BaB2O4 prepared by heating from the g-modification into a-BaB2O4 begins at 870 8C and is completed at 940 8C.The kinetics of this transformation have been studied. The most surprising results were reported by Kouta et al.80 who studied the transition in question and did not observe even minor indications of the phase transformation after heating b-BaB2O4 powder at 1090 8C for 5 h and subsequent cooling of the sample. The results of DTA study 80 showed that both modifications of BaB2O4 melt at nearly the same temperature, namely, 11000.5 8C (a-phase) and 10990.5 8C (b-phase). cooling rates (u=3.75, 5.0, 7.5, 10.0, 15.0 deg min71) on both the transition temperatures and reversibility of the phase transforma- tions has been reported.104 It was shown that the temperatures of the polymorphic transition of the low-temperature modification of BaB2O4 into the high-temperature form (917 8C) and the melting temperature of the latter (1092 8C) are virtually inde- pendent of the heating rate, whereas the reverse is observed on cooling. VI.Peculiarities of crystallisation of BaB2O4 melts BBO melts are characterised by high viscosity, tendency to super- cooling and by hysteresis of properties. The physical properties of BBO have been studied recentlly.81, 109 The viscosity of the melt Intensity (rel.u.) Intensity (rel.u.) Intensity (rel.u.) abc 30 20 10 A special study concerning the influence of the heating and 50 40 2y /deg 659 monotonicaly increases from 20 to 61 mPa cm71 as the temper- ature decreases from 1463 to 1271K and is*30 mPa cm71 at the melting temperature.The activation energy of the viscous flow for the melt is 0.95 eV. The time taken to achieve the equilibrium viscosity of the melt at 1463K is nearly 10 h. The density of the melt is a linear function of the temperature in the range between 1463 and 1271 K; it is described by the following equation r=(4.45370.000545)T It was found that r=3.707 g cm73 at the melting temperature. The thermal expansion coefficient of the melt is 1.561074 K71. The surface tension at the melting temperature is 0.35 N m71 (see Ref. 89). According to the results of our experiments data and to Itoh et al.79, the degree of supercooling of the melt (DT ) in a platinum crucible can be as high as 250 8C at cooling rates spanning the range from 4 to 15 deg min71.In this case crystallisation of b-BaB2O4 can occur at temperatures that are much higher than the temperature of the polymorphic transition. Crystallisation occurs randomly; there is a wide spread of the melt crystallisation temperature values but no correlations between this temperature and the cooling rate are observed. Mention has been made that the degree of supercooling of the melt can be as high as 300 8C.110 The problem of crystallisation of BBO (spontaneous crystallisation, seed crystallisation using a-BaB2O4 and b-BaB2O4 and crystal- lisation on a platinum wire) has been the subject of numerous studies. In a study 104 the degree of supercooling of the melt was 170 ± 200 8C. On cooling (u=3.75, 5.0, 7.5, 10.0, 15.0 deg min71), crystallisation occurred only at the temperature of the poly- morphic transition or at lower temperatures and resulted in the low-temperature modification of BaB2O4.The crystallisation temperatures were found to be Tc=917 8C at u=3.75 and 5 deg min71, Tc=887 8C at u=7.5 deg min71 and Tc=857 8C at u=15 deg min71. BBO melts can exist without any changes in the metastable state for long at temperatures that are higher than the temperature of the polymorphic transition and lower than the melting temper- ature. However, seeding causes rapid crystallisation the rate of which increases with temperature. If a-BaB2O4 is used as seed, the same modification of barium borate crystallises. Seeding with b-BaB2O4 causes crystallisation of the b-modification despite its instability under these conditions.Experiments on platinum seeded crystal growth showed 87 that co-crystallisation of b-BaB2O4 and a-BaB2O4 begins if DT540 8C. At smaller DT values, crystallisation of the a-phase occurs, whereas b-BaB2O4 mainly crystallises at temperatures below 1030 8C (i.e., at DT570 8C).82 The degree of supercooling achievable using graphite crucibles is substantially lower compared to platinum crucibles; in addition, crystallisation of both the a- and b-modifications of BaB2O4 in graphite crucibles is possible.107 This is thought to be due to some similarity between the trigonal structures of graphite and BBO. Thermodynamically, the possibility of direct melt crystallisa- tion of the low-temperature phase can be rationalised by taking into account strong supercooling of the melt and the existence of the temperature of metastable melting of the low-temperature phase (at this temperature the curves of the temperature depend- ences of the Gibbs thermodynamic potentials of the low-temper- ature phase and the melt intersect).The corresponding scheme is presented in Fig. 9. The temperature of metastable melting, Tb , of the low-temperature modification lies between the temperature of equilibrium melting, Ta , of the high-temperature modification and the temperature of polymorphic transformation, Tt . For the crystallisation of the b-phase to begin the melt temperature should be lower than Tb . Ignoring the difference between the heat capacities of the a- and b-modifications of BaB2O4 , the following relationship is valid:110660 G 123 Tt Tb Figure 9.Temperature dependences of the Gibbs thermodynamic poten- tials (G) of the melt (1 ) and of the a- (2) and b-modifications (3) of BaB2O4 .106 Tt is the temperature of polymorphic transformation, Tb is the temper- ature of metastable melting of the low-temperature modification and Ta is the temperature of the equilibrium melting of the high-temperature modification. T DSb b a DHt a DHa a DHb , DSt a DSa where DHand DS are the enthalpies and entropies of the reactions. By denoting d=DHt , DHa one gets 111 Tb a Od a 1UTaTt . dTa a Tt The estimates of the Tb value are strongly different (Table 2).From the standpoint of polymorphism, BaB2O4 presents a unique phenomenon, since both modifications (a and b) of this compound have virtually equal melting enthalpies and melting temperatures, while the enthalpy (and, correspondingly, entropy) of the a>b transition is very low.80, 82 On the other hand, here we deal with the first-order phase transition characterised by signifi- cantly different crystal structures, volumes and densities (in the temperature range from 20 to 1000 8C) of both phases. In this case, a correct thermodynamic description of the phase trans- formations requires that the heat capacities be taken into account (they must be measured) and, hence, the approximate relation- ships listed above cannot be used.Under the conditions in question the mutual stability of the polymorphous modifications strongly depends on the presence of impurities, defects, as well as on the mechanical strain and surface energy, so that the possibility of melt crystallisation of a particular modification is first of all determined by the kinetic considerations. Table 2. Melting temperatures (8C) and enthalpies (kJ mol71) of the a- and b-modifications of BaB2O4 . Tb Ta 1095 1078 67.784 1100 1050 11000.5 7 7 80 10990.5 7 7 109.45 T Ta Ref. DHb DHa 111 74.562 7 7 106 82 110.00 P P Fedorov, A E Kokh, N G Kononova n Hindered melt crystallisation of BBO can be rationalised based on structural considerations. The results of high-temper- ature Raman spectroscopy studies 99, 100, 112 of barium borate point to a dramatic difference between the spectra of the melt and crystalline b-BaB2O4 .The spectrum of supercooled melt exhibited not only narrow lines of [B3O6]37 anions, but also a number of broad bands corresponding to internal vibrations of long-chain [BO2]n¢§ anions. The contribution of the [B3O6]37 anions near the melting temperature is small; however, it increases as the degree of supercooling of the melt increases. Probably, melting of BBO is accompanied by rearrangement of the boron ¡¾ oxygen complexes (transformation of the [B3O6]37 ring anions into long anionic chains). Crystallisation is accompanied by the reverse rearrangement. High viscosity of barium borate melts and their tendency to glass formation are due to the presence of long anionic chains.The hysteresis of viscosity (see above) 81, 109 seems to be associated with polymerisation ¡¾ depolymerisation of [B3O6]37 anions. Structural rearrangement of the melt on heating was studied experimentally (by X-ray powder diffraction) and theoretically (by the molecular dynamics method).113, 114 It was found 82 that self-melt crystallisation of b-BaB2O4 single crystals is possible if the starting material used for the preparation of the melt corresponded to the b-phase and no melt overheating occurred. The character of crystallisation of the melt can be affected by the presence of fine particles and structural fragments of crystalline phases that play the role of nucleation and crystallisation centres.87 Overheating causes destruction of these centres.Graphite particles present in the melt can also play the role of crystallisation centres.107 Information on the influence of atmosphere on the character of crystallisation of BBO melts was questioned by some authors. However, an increase in the degree of supercooling of the melt in the case of crystallisation in an argon atmosphere compared to the crystallisation in air has been reported recently.115 VII. Growth of b-BaB2O4 single crystals Peculiarities of crystallisation of BBO melts determine two most widely used methods for the growth of b-BaB2O4 single crystals. These are the supercooled self-melt growth (under nonequilibrium conditions) and melt-solution growth techniques.In the latter case, both single crystal growth from the equilibrium primary crystallisation region and nonequilibrium crystallisation are pos- sible. The schemes of a number of most widely used single crystal growth techniques are presented in Fig. 10. 1. Self-melt growth of single crystals Self-melt growth techniques of b-BaB2O4 single crystals have the advantages that the growing crystals are not poisoned with flux and the crystallisation rates are high (by an order of magnitude higher compared to melt-solution crystal growth). However, in these cases crystallisation occurs under nonequilibrium conditions at large temperature gradients, which causes development of mechanical strain in and cracking of the crystal.In addition, the crystal quality strongly depends on the procedure employed for the preparation of the starting reagents and on the crystallisation conditions. These methods have not received wide application so far. Single crystals of b-BaB2O4 were Czochralski grown using a platinum wire seed.79, 80 Noteworthy is that only the a-modifica- tion was melt crystallised if the starting reagents were prepared by the carbonate technique [see reactions (3) and (4)]. The process using the starting reagents synthesised by the chloride method [see reactions (1) and (2)] led to direct crystallisation of the b-phase in a narrow range of the temperature gradient (G) values at the crystallisation front (G=150 ¡¾ 125 deg cm71). Variation of G (an increase up to 175 deg cm71 and a decrease down to 50 deg cm71) resulted in the crystallisation of the a-phase.A cracked single-crystal boule 7 mm in diameter and 18 mmBarium borate b-BaB2O4 as a material for nonlinear optics a 1 5 3 4 2 ce 7 Figure 10. Schematical illustrations of the most widely used experimental techniques for the melt and melt-solution growth of single crystals. Czochralski technique, T=const (a); Kyropoulos technique (immersion- seeded solution growth, ISSG), T decreases (b); top-seeded solution growth (TSSG) technique, T decreases (c); melt feed Czochralski techni- que (d); Stepanov technique, T=const (e); zone melting technique (f) and laser assisted pulling from pedestal (g). Crystal (1); melt (2); heater (3); crucible (4); seed (5); feeding charge (6); shape-forming element (7); laser heating (8) and polycrystal (9).long was obtained at a seed rotation rate of 20 rpm and a pulling speed of 0.6 mm h71.79 Improvement of the growth technique allowed obtaining single crystals 15 mm in diameter and 40 mm long without cracks at a seed rotation rate of 14 rpm and a pulling speed of 4 mm h71 .80 According to Kozuki and Itoh,106 the upper critical temper- ature for stable Czochralski self-melt growth of b-BaB2O4 crystals is 1050 8C. The starting material (charge) is prepared from barium nitrate and boron oxide. A single crystal 15 mm in diameter and 7 mm long was grown at a seed rotation rate of 10 rpm and a pulling speed of 3 mmh71.The vertical temperature gradient was only 10 deg cm71; however, a strong influence of the radial temperature gradient was revealed (in this case, the optimum value was 77 deg cm71, see Ref. 116). Yet another example of the Czochralski growth of b-BaB2O4 crystals has been reported.89 A b-BaB2O4 seed of specified orientation was placed in a platinum pipe 2 or 4 mm in diameter. The charge was prepared from barium carbonate. The predom- inant impurity was platinum (30 ppm). The temperature gradient was 300 deg cm71. Crystals 7 mm in diameter and 20 mm long were grown at a seed rotation rate of 15 rpm and a pulling speed of 5 mmh71. The same technique was also used for the crystallisa- bd 6 g f 8 9 661 (R=Al, Ga, tion of solid solutions b-Ba(B17xRx)2O4 04x40.1).116 2.Melt-solution growth of single crystals { a. Choice of solvent An appropriate solvent (flux) must provide a rather extended primary crystallisation region of the low-temperature modifica- tion of BaB2O4 and reduce the viscosity and glass-forming ability of the melt. As a consequence, the melts should exhibit a lower degree of polymerisation of the boron ± oxygen chains and an increased proportion of the [B3O6]37 ring anions. The solvent must be of low volatility, possess stable properties and provide very weak interaction with the crystal to exclude the possibility of poisoning. The melt should also not be an aggressive medium towards the material the crucible is made of. A melt growth technique of b-BaB2O4 single crystals from the starting reagents using a shift from the stoichiometric composition towards an excess of BaO or B2O3 (the so-called self-flux) has been reported.63, 103 However, the temperature ranges suitable for crystal growth in this manner are narrow (less than 20 8C).More often, an excess of B2O3 is used;120, 121 however, in this case the crystal growth is hampered due to high viscosity of the melt. A large number of compounds and their combinations have been considered as candidates for solvents. These are, e.g., Li2O,62, 103 BaCl2 ,103, 121 BaF2 ,66, 103, 121, 122 NaF,66, 90, 91, 123 LiF,66 NaCl,66, 124 ± 126 NaCl ±Na2O,125 PbO,127 Na2O± Nd2O3 ,72, 120 LiNbO3 ,120 PbF2±B2O3 ,120 CaF2 124 and Na2SO4 .124 Usually, the solvent is chosen based on the phase diagrams of corresponding systems.Some known regions of the phase dia- grams are presented in Fig. 11. The BaB2O4 ± NaCl, BaB2O4 ± { For basic principles of the methods for melt-solution (flux) crystal growth, see Refs 117 ± 119. a T /8C 1100 10955 L a-BaB2O4+L NaCl+L 900 92010 b-BaB2O4+L 700 7543 b-BaB2O4+NaCl NaCl 60 20 80 40 BaB2O4 [NaCl] (mol.%) b T /8C L 1000 L+a-BaB2O4 900 L+b-BaB2O4 890 800 BaB2O4 80 BaF2 60 [BaB2O4] (mol.%) Figure 11. Phase equilibria in the systems BaB2O4 ± NaCl (see Ref. 124) (a) and BaB2O4 ± BaF2 (see Ref. 122) (b).662 NaF and BaB2O4 ± BaF2 sections seem to be quasi-binary. The BaB2O4±Na2SO4 section is unstable, since in this case an exchange reaction occurs BaSO4+Na2B2O4 .BaB2O4+Na2SO4 The temperatures of the eutectic in the BaB2O4 ± NaCl system, determined by different authors,124, 125 virtually coincide (see Fig. 11 a); however, the compositions of the eutectic are appreci- ably different. Correspondingly, the concentration ranges of the growth of b-BaB2O4 crystals in this system are also different. Data on the phase equilibria along the BaB2O4 ± NaF (see Refs 66 and 123) and BaB2O4 ± BaF2 sections (see Refs 103 and 122) are contradictory. All solvents studied have certain drawbacks. The systems containing Li2O (see Ref. 103) and BaF2 (see Ref. 122) exhibit very narrow crystallisation regions of b-BaB2O4, while no such region at all was found in the system containing CaF2.124 The system with BaCl2 is characterised by fast hydrolysis.If fluoride melts are used, pyrohydrolysis also occurs in air,128 either rela- tively fast (BaF2) or slow (NaF). A number of additives (e.g., MgF2 and BaF2) do not favour reduction of the melt viscosity; besides, the melt containing MgF2 is prone to glass formation.66 The use of lead-containing compounds 120, 127 favours stabilisa- tion of the a-modification of BaB2O4.49 Potassium fluoride additives cause significant changes in the lattice parameters of BBO crystals, which seems to be due to isomorphous incorpo- ration of potassium.66 The BaO ± B2O3 ± BaF2 system is used as flux in the growth of rare-earth ferrite single crystals.129 Sodium fluoride is considered as candidate for flux.The BaB2O4 ± NaF section and the BaO ± B2O3 ± NaF system are characterised by broad regions of primary crystallisation of b-BaB2O4 66 and by substantial reduction of the viscosity of the melt, which permits crystallisation at higher rates. However, to make the best use of these advantages, the crystals should be grown in an inert gas atmosphere. Slow pyrohydrolysis of NaF in air during crystallisation causes a gradual transformation of the BaB2O4 ± NaF system into the BaB2O4±Na2O system. Na2O45, 46, 64, 86, 90, 91, 101 ± 103, 105, 121, 124, 130 ± 146 is, at present, most widely used for melt-solution growth of b-BaB2O4 single crystals. Melt-solution growth of single crystals in the BaB2O4 ± Na2B2O4 system has also been reported.103 The process is com- plicated by high viscosity of the melt and its tendency to super- cooling.A number of studies concerning variation of the melt composition in the BaO ± B2O3±Na2O system are avail- able.64, 134 ± 137, 143, 144 Phase equilibria in the BaB2O4±Na2O system were first studied by DTA105 (cited according to Refs 46 and 130). A compound Na2BaB2O5 was identified, which congruently melts at 846 8C. The temperatures of the eutectics are 755 and 573 8C. The phase diagram constructed by Huang and Liang 105 was used for the choice and `tuning' of the growth regimes of BBO crystals. Kaplun et al.133 proposed yet another version of the phase diagram of the BaB2O4±Na2O system. In this case, the liquidus curve in the BBO crystallisation region and the eutectic position match those reported by Huang and Liang;105 however, the compound Na2O.BaB2O4 was assumed to be incongruently melting at 8303 8C following a peritectic reaction resulting in a refractory compound that melts at temperatures above 1300 8C (this compound was assumed 133 to have a composition 3 BaO . B2O3). An analogous interpretation of the phase equilibria in this system was reported in other studies.143, 144 The generalised phase diagram of the BaB2O4±Na2O system is shown in Fig. 12. The region of primary crystallisation of b-BaB2O4 in the BaO ±B2O3±Na2O ternary system (Fig. 13) has been thoroughly studied by Nikolov and Peshev.64 The contradiction between the results of studies105, 133 was solved recently.102 In an investigation of the BaB2O4±Na2O section by DTA and visual-polythermal analysis (VPA) it was found (see Fig.12) that both strong thermal effects (ascribed 105 to P P Fedorov, A E Kokh, N G Kononova T /8C 1100 L 1000 L+a-BaB2O4 925 900 L+NaBaBO3 L+b-BaB2O4 800 745 700 b-BaB2O4+NaBaBO3+NaBO2 40 20 BaB2O4 [Na2O] (mol.%) Figure 12. Phase equilibria along the BaB2O4±Na2O section.102 The data were obtained by DTA, the mixtures were prepared by successive addition of Na2CO3 to remelted BaB2O4 (1); the data were obtained by DTA, the samples were prepared by solid-phase synthesis (2); the data were obtained by visual-polythermal analysis (3); other data were taken from Ref. 133 (4), Ref.105 (5), Ref. 64 (6) and Ref. 72 (7). 40 Na2O 30 30 800 20 20 850 900 10 60 40 30 50 Figure 13. Primary crystallisation region of the a- and b-modifications of BaB2O4 in the BaO ±B2O3±Na2O ternary system.64 Primarily crystallising phases: Na2B4O7 (1), NaBO2 (2), Na2BaB2O5 (3), Ba3B2O6 (4), a-BaB2O4 (5) and b-BaB2O4 (6). The dashed lines denote isotherms. the melting of Na2BaB2O5) and high-temperature equilibrium crystallisation (liquidus curve) manifest themselves in the concen- tration range from 30 mol.% to 50 mol.% Na2O. These results are in agreement with those reported by Kaplun et al.133 The medium-temperature effects were found 102 to be concentration dependent (this contradicts the results reported earlier 133), which is impossible in the case of stable quasi-binary section.X-Ray diffraction and Raman spectroscopy studies showed that the reaction of BaB2O4 with Na2O results in barium ortho- borate, NaBaBO3, whose structure comprises isolated [BO3] triangles and that compound Na2O. BaB2O4 does not exist. Single crystals of compound NaBaBO3 were grown by spontaneous crystallisation from melts of different compositions including a melt containing 62 mol.% BaB2O4 and 38 mol.% Na2O using a platinum loop seed. The compound melts at 1270 8C and under- goes a rapid degradation in air, being a good absorbent of water and carbon dioxide. The X-ray diffraction pattern of NaBaBO3 is 1234567 60 Na2O 123456 10[B2O3] (mol.%)Barium borate b-BaB2O4 as a material for nonlinear optics a I (rel.u.) bcn1 n3 400 800 0 Figure 14. Raman spectra of a NaBaBO3 single crystal (a), Na3BO3 powder (b) and 3Na2O.B2O3 melt (c).102 The fundamental vibration of the impurity carbonate ion is asterisked.shown in Fig. 8 c and its Raman spectrum is presented in Fig. 14 a. Both the a- and b-modifications of BaB2O4 have nearly identical thermodynamic characteristics.80, 82 This means that the liquidus curve (surface) corresponding to the polymorphic transformation of BaB2O4 should exhibit no kinks. The projection of the liquidus surfaces of a part of the BaO ± B2O3±Na2O system on the concentration triangle is shown in Fig. 15. In this system, compound NaBaBO3 is assumed to be congruently melting and the BaB2O4 ±NaBO2, NaBaBO3 ± NaBO2 and, probably, BaB2O4 ± NaBaBO3 sections are thought to be of eutectic character (the eutectics lie at 28 mol.% BaO, 50 mol.% B2O3 , 22 mol.% Na2O, 826 8C and at 22 mol.% BaO, 39 mol.% B2O3 , 39 mol.% Na2O, 830 8C).The region of primary crystallisation of NaBaBO3 can be determined by analysing the morphology of the primary (spontaneously grown) crystals. The primary crystals are isometric individuals in contrast to needle- shaped BaB2O4 and NaBO2 crystals. The ternary eutectic occurs at 745 8C and lies at 41 mol.% BaO, 41 mol.% B2O3 , 18 mol.% Na2O (or 70 mol.% BaB2O4 , 30 mol.% Na2O). The results obtained by different authors 64, 102 are in good agreement. By chance, the BaB2O4±Na2O section passes nearly exactly through the ternary eutectic in the BaB2O4 ± NaBaBO3 ±NaBO2 ternary system.Therefore, the left part of the BaB2O4±Na2O section looks like a binary system. It seems likely that weak high- temperature effects were overlooked 105 and that the non-variant effect of the L+NaBaBO3+NaBO2 binary eutectic was inter- preted to be the melting of a compound BaB2O4 . Na2O, while the co-crystallisation (co-melting) curve of NaBaBO3+NaBO2 was considered to be the liquidus curve. Studies of the phase diagram of the BaO ± B2O3±Na2O system are complicated by high reactivity and hygroscopicity of melts, by their ability to absorb carbon dioxide and by the n4 * 2 n2 1200 n /cm71 663 [Na2O] (mol.%) 950 900 [B2O 950 40 60 850 e2 850 950 900 1000 3] (mol.%) NaBO2 900 1100 1200 30 70 NaBaBO3 850 B2O3 20 1200 1100 E e3 850 900 10 1000 950 900 800 e1 850 900 Ba2B2O5 30 40 [BaO] (mol.%) 60 BaB2O4 Ba3B2O6 Figure 15.Projections of the liquidus surfaces in the BaB2O4 ± NaBaBO3 ±NaBO2 system on the concentration triangle.102 The eutectics are labelled e1 ± e3 and the ternary eutectic is labelled E. corroding action of the melts on the platinum crucibles. High viscosity of the melts and their tendency to supercooling, glass formation and appearance of metastable equilibria present a particular problem.102, 143 ± 145 According to the phase diagram of the BaB2O4±Na2Osection (see Fig. 12), the temperature range of crystallisation of b-BaB2O4 is *170 8C.The compositions suitable for growth of the low- temperature phase lie in the range 22 mol.% ± 30 mol.% Na2O. Usually, the composition with the Na2O content between 20 mol.% and 22 mol.% is used as charge. However, supercool- ing of the melt allows growth of large b-BaB2O4 crystals at aNa2O content of less than 5 mol.% at*1050 8C.121 The volatility of the melt increases along the BaB2O4±Na2O section, whereas its viscosity decreases as the concentration of Na2O increases.121, 132, 145 b. Crystal growth techniques The driving force of the melt-solution growth of BBO crystals is due to temperature reduction. Usually, vaporisation of the solvent is not used as a method for the preparation of a supersaturated solution.117 ± 119 The process can be performed using various equipment.Schematical illustrations of a number of crystal growth techniques are shown in Fig. 10. Only a few studies are concerned with the growth of (i) BBO single crystals by the travelling solvent method (a version of zone melting technique),131 (ii) single-crystal filaments by laser-assisted pulling from pedestal 120 and (iii) plates using the Stepanov technique.85, 92 The so-called top-seeded solution growth (TSSG) technique is most widely used.45, 46, 66, 86, 121, 130, 137 ± 140, 146 It allows bulk BBO crystals (up to 60 mmin diameter and more than 10 ± 15 mmlong) to be grown (Fig. 16). Several versions of this method are known. If crystal growth occurs on slow cooling of the melt without pulling the seed, one deals with the Kyropoulos technique.The Czochralski method involves pulling of the seed-growing crystal under isothermal conditions (see Fig. 10). In the case of melt-solution crystallisation the composition of the crystal differs from that of the melt; therefore, a feature of the Czochralski technique is continuous change in the composition of the melt and, correspondingly, in the crystallisation conditions. To avoid variations of the melt composition and other parameters and to maintain steady-state growth conditions, a method involv- ing melt feed with BBO charge was proposed.132 Often, a combination of the two above-mentioned growth techniques is employed. In this case, a slow pulling of the crystal is664 abc Figure 16.Photographs of a BBO crystal (a), axial slice 8 mm thick (b) and a nonlinear optical element of dimensions 868615 mm3 (c). Paper cell size is 565 mm. The crystal was grown by the authors of this review at the Crystal Growth Laboratory, Institute of Mineralogy and Petroghaphy, Siberian Branch of the Russian Academy of Sciences. combined with slow cooling of the melt. The major advantages of the TSSG technique with pulling of the crystal compared to analogous technique without pulling are as follows: � higher crystal growth rate; � the possibility of growing thick crystals in smaller cruci- bles;� prevention the crucible walls from touching with the growing crystal, which causes seed failure.Seed growth of bulk b-BaB2O4 crystals of good quality with and without pulling from Na2O flux has been reported.121 The seed rotation rate varied from 2 to 60 rpm while the pulling speed was 0.01 ± 0.02 mm h71. The high seed rotation and crystal pulling rates cause undesired changes in the shape of the interface from a convex interface to a concave one. Inclusion-free BBO crystals were grown at low pulling rates (of the order of 0.5 to 1.0 mm per day). As a rule, a melt of composition 75 mol.% BBO± 25 mol.% Na2O was used. The liquidus temperature for this composition is *886 8C. The starting reagents (BaCO3, H3BO3 and Na2CO3) were weighed, mixed and placed in a platinum crucible 75 mm in diameter. Then the crucible was heated to*1110 8C over a period of several hours until complete decomposition of carbonates.Seed crystals with different orientations were used for crystal growth. Orientation of the seed crystal affects not only the rate and quality of growth, but also the shape of the BBO crystal. A high P P Fedorov, A E Kokh, N G Kononova anisotropy of the growth rate of BBO crystals was reported. The rates of growth along the a and b axes were found to be higher than the rate of growth parallel to the c optical axis. Crystals grown parallel to the a and b axes or in other directions that strongly deviate from the c axis usually undergo spontaneous cracking in the (001) cleavage plane. This is due to anisotropy of the thermal expansion. The thermal expansion and thermal conductivity coefficient of BBO crystals measured along the c axis are by nearly an order of magnitude larger than those measured in the perpen- dicular direction.39 The radial and axial temperature gradients and the shape of the liquid ± solid interface are important parameters that define quality of the crystal. Crystals of the best quality are usually grown if the interface is flat or slightly convex.The shape of the interface depends on both the pulling speed and the seed rotation rate. Obtaining a flat or slightly concave interface requires that the pulling speed and the seed rotation rate be 0.5 ± 1.0 mm per day and 4 ± 8 rpm, respectively. As the seed rotation rate increases, the contribution of forced convection gradually becomes much larger than that of natural convection.If the pulling speed seed is too high, the growing crystals can lose contact with the solution. Search for methods of control of convective heat and mass transfer always has received considerable attention, since it is these processes that in most cases provide the possibility of growing particular types of crystals with preset properties. All methods of control can be divided into contact methods and contactless ones. The former are based on the action of a physical body placed in contact with the melt, solution or another crystallisation medium. Most of these methods involve mechan- ical rotation of the crystal and/or crucible, the use of various diaphragms, stirrers, shape-forming elements, etc.A number of contact methods have been applied to the growth of BBO crystals.139, 147, 148 Extreme sophistication of the mechanical devi- ces gave an impetus to search for simpler methods of control of the heat and mass transfer processes. This approach had some success. Contactless techniques are based on the action of physical fields (gravitational, electromagnetic and thermal) on the heat and mass transfer processes. In practice, all crystal growth techniques include (sometimes, unconsciously) a certain pattern of the heat field. In most cases these are static heat fields characterised by the symmetry axis of an infinite order, L?, and particular values of the axial and radial temperature gradients. Traditionally, crystals are grown in tube heating furnaces with cylindrical heat field symmetry L? (exclud- ing the edge effects).With the advent of precision heating furnaces (PHF),149 the possibility of controlling the heat field and the heat and mansfer during the crystal growth appeared. This is achieved by changing the heat field symmetry and by rotating the heat field. A version of dynamic control of the heating elements (HE) in such a furnace has been reported.150 ± 153 In practice, the growth of crystals under conditions of static, azimuthally distrib- uted heat field in the PHF is also of considerable interest.140 These conditions are created by either commutation or regular arrange- ment of the HE around the PHF muffle, which provides a radially distributed pattern for power delivered to the muffle walls and, hence, to the walls of the growth crucible.Experiments on the growth of BBO crystals 140 were carried out using a heating furnace with a total of 18 HE arranged evenly around the quartz muffle 110 mm in diameter with walls 4 mm thick. Switching on all series-connected HE in the furnace produced a heat field of cylindrical symmetry L?. Strictly speak- ing, in this case the heat field symmetry should be L18 ; however, it was considered as L? (within the limits of experimental error) due to the choice of an optimum distance between the HE and taking into account the effect of the heat field smoothing by the muffle and the crucible walls. Switching off each third HE changed the heat field symmetry in the furnace to six-fold symmetry L6. If three HE arranged at an angle of 120 8 with respect to one another were switched off, the heat field symmetry reduced to L3.InBarium borate b-BaB2O4 as a material for nonlinear optics a c b 8868C 8908C 8828C 8958C 8848C 8888C Figure 17. Polar diagrams of the heat fields of symmetry L? (a), L6 (b) and L3 (c) used for the growth of BBO crystals (at the top) and the corresponding convective patterns (at the bottom). The open circles denote the switched off heating elements. Fig. 17, we present the horizontal sections of the growth furnace and the polar diagrams of the heat fields of symmetry L?, L6 and L3 , as well as the typical convective patterns on the melt-solution surface, formed under the action of the heat fields of different symmetry.BBO crystals were grown on a 464 mm seed oriented parallel to the optical axis c. The growth procedure consisted in search for an equilibrium temperature by seed touching the melt-solution surface in the place where the rays of the convective star are brought to a point. After the seed started to spontaneously dissolve, the temperature in the furnace was lowered by 1 ± 2 degrees. The next touching of the melt surface was performed after 2 ± 3 h. After prolonged (12 ± 24 h) liquid ± seed contact the crys- tal was allowed to grow on cooling the furnace at a rate of 1 ± 3 degrees per day. The seed was rotated at 1 ± 2 rpm and moved upward at a rate of 0.3 ± 0.4 mm per day.The growth process continued until the crystal became visually much less transparent. This was interpreted as an indication of the constitu- tional supercooling, which induced a cellular crystal growth. The experimental data on the growth of BBO crystals under the action of steady-state heat fields of different symmetry are listed in Table 3. As can be seen, a change in the heat field symmetry from L? to L6 led to a slight increase in the `yield' coefficient (from 1.220.02 to 1.27). However, the `yield' coef- ficient increased by a factor of 1.7 to 2.07 g (kg deg)71 on going to L3 symmetry [cf. a theoretical value of *2.52 g (kg deg)71 obtained according to the liquidus equilibria in the BaB2O4 ± Na2O system]. Table 3. Experimental data on BBO crystals grown in a crucible 100 mm in diameter at initial charge of 1.5 kg in heat fields of different sym- metry.140 Run No.DT /8C Pulling /mm `Yield' coefficient a /g (kg deg)71 Heat field symmetry Crystal mass /g L? L? L? 1.20 1.23 1.24 1.27 2.07 212 227 198 179 319 117.5 122.9 106.4 94.2 103.0 L6 L3 12345 12.6 10.3 16.7 14.2 14.4 a The `yield' (or `mass production') coefficient is an increase in the mass of a crystal grown upon 1 8C temperature reduction at initial crucible charge of 1 kg. 665 The changes in the `yield' coefficient as function of the heat field symmetry and the observed character and evolution of convective flows on the melt surface were interpreted as fol- lows.140 As the heat field with L3 symmetry was created in the growth furnace, the convective regime in the bulk of the melt changed fundamentally.The fluid body divided into three azimu- thally distributed convective cells in which the thermal-gravita- tional flows moved up and down, thus `enveloping' nearly the whole bulk of the liquid phase and forming three intense surface flows merging at the centre. This convective regime provided a permanent renewal of the melt in the region near the crystallisa- tion interface, thus substantially delaying the onset of constitu- tional supercooling. In this case the `yield' coefficient amounted to 82% of its maximum theoretical value. This, in a way, character- ises the extent of involvement of the fluid body in the convection.VIII. Defects and impurities in crystals BBO crystals thus obtained were found to have a very small central defect region. These crystals can be used for fabrication of relatively large (up to 20mmand more) nonlinear optical elements and electro-optical cells oriented parallel to the optical axis c . Feigelson et al.130 reported the `yield' coefficients of 2.01 and 2.13 g (kg deg)71 obtained in the growth of BBO crystals from a crucible 55 mm in diameter with the corresponding initial charges of 380 and 290 g. An increase in the charge to 470 g caused reduction of the `yield' coefficient to 1.6 g (kg deg)71. This points to significant difference between the convective regimes in the crucibles 55 and 100 mm in diameter at comparable temperature gradients in the growth furnace. Recently,154 pioneering results concerning the growth of BBO crystals in rotating heat fields of different symmetry have been reported.Information on intrinsic nonstoichiometry and impurity defects in BBO crystals is scarce.155 Three types of radiation defects that exhibit UV absorption at 210, 250 and 310 nm were identified in g-irradiated BBO crystals. The formation of these defects is assumed to be associated with both cleavage of the B7O bonds and atomic displacements from the equilibrium positions.156 Growth defects were studied by chemical etching, optical microscopy, microprobe technique, X-ray topography and laser scanning tomography.46, 86, 121, 130, 157, 158 The main types of defects found in BBO crystals were inclusions (usually, they are interpreted as flux inclusions) and inclusion-induced dislocations. The appearance of gas bubbles is associated with the presence of trace amounts of non-decomposed carbonates in the charge.86 High density of dislocations due to solvent trapping was pointed out.121 It is assumed that removal of inclusions must favour a substantial improvement of the quality and optical characteristics of BBO crystals.The content of sodium in b-BaB2O4 crystals grown under different conditions (from melt-solution and from a stoichiomet- ric melt according to Czochralski) can vary from 120 to 680 ppm (0.012 mass%± 0.068 mass %).159 Transparent inclusion-free crystals contain from 0.012 mass% to 0.023 mass% Na.At a sodium content of greater than 0.035 mass %, the crystals are semi-transparent until becoming opal. Sodium impurity was found to be responsible for thermoluminescence of BBO crys- tals.159, 160 Among other impurities, an increased concentration of stron- tium was pointed out.161 Probably, this impurity incorporates into the BBO crystals from barium-containing starting reagents. Lead impurity causes stabilisation of the a-modification of BaB2O4 .49 A feature of BBO crystals is accumulation of sodium and potassium impurities.101 The critical concentration of sodium is 0.021 mass %. As the content of sodium exceeded this value, inclusions of the second phase and the scattering centers appear.101 The effective distribution coefficient of sodium between the crystal and melt is (2.5 ± 4.4)61073.666 From crystal chemical considerations it follows that alkali metal impurities can incorporate into the structure of b-BaB2O4 .As can be seen in Fig. 7, BaB2O4 and a sodium compound NaBO2 are structurally similar. They contain identical structural frag- ments (boron ± oxygen [B3O6]37 anions) stacked perpendicular to the c axis of the trigonal lattice and exhibit a similar packing motif. The difference is in the arrangement of cations; the number of sodium ions that provide electrostatic compensation of charge is twice as large as the number of barium ions. The trigonal unit cells also have close parameters (see Table 1).This suggests the possibility of heterovalent isomorphism simply described as Ba>2Na and incorporation of extra sodium ions into the b-BaB2O4 lattice. The authors of this review also assume that the major type of intrinsic point defects in b-BaB2O4 crystals are Frenkel pairs characterised by displacement of Ba ions from the equilibrium positions to the interstitial ones. Close values of the unit cell parameters allow epitaxial growth of NaBO2 on the b-BaB2O4 crystals. Even better is agreement between the lattice parameters of b-BaB2O4 and potassium metaborate, KBO2, which is isostruc- tural to the sodium-containing analogue. Potassium ions are readily incorporated into the BBO lattice (potassium fluoride melt was used as flux).66 The distribution coefficient of sodium in BBOcrystals is small, therefore this impurity can be removed by zone melting with ease.88 Low efficiency of this technique in the removal of stron- tium impurity from BBO crystals 107 indicates that the corre- sponding distribution coefficient approaches a unity.Inclusions in BBO crystals were found to be sodium enriched.46, 130 Identification of the phase that is formed at the crystallisation interface, thus blocking the process of crystal growth, by X-ray powder diffraction failed.66 However, a recent X-ray powder diffraction study of the particles with inclusions, picked up with a microscope, showed 102 that this phase has a composition NaBaBO3 . IX. Constitutional supercooling Morphological instability of the crystallisation front (Fig.18) occurs in a certain stage of melt-solution growth of BBO crystals (usually, as the crystal becomes *15 mm thick). It involves the development of a characteristic cellular substructure with quasi- hexagonal cells, which is accompanied by trapping of inclusions and the formation of a protrusion.46, 66, 130 The appearance of cellular substructure is the major reason for forced termination of the process of crystal growth. This phenomenon is associated with specific features of heat transfer during the growth process.130, 143 However, usually the morphological instability is due to the Deep cells Fluctuation instability of the interface Smooth interface Cells Figure 18. Qualitative picture of the development of morphological instability of the crystallisation front and the formation of cellular substructure.constitutional supercooling which has been studied in detail for binary systems.162 ± 166 Consider the growth of BBO crystals from a Na2O melt- solution. In this case, sodium piles up near the solidification front, thus causing the loss of its stability. Figure 19 presents a scheme illustrating the distribution of the melt components in the region adjacent to the solidification front for a steady-state growth of BBO crystals from Na2O melt-solution at the rate R. Since the amount of sodium incorporated into the crystal is very small, it is displaced by the growing crystal and is accumulated near the liquid ± solid interface.The sodium concentration jump appeared at the crystallisation front induces a diffusion flow of sodium ions in the melt, which is directed away from the liquid ± solid interface. The content of barium and boron in the melt is somewhat lower than in the crystal; therefore, these components are trapped by the crystal. This leads to the appearance of a region depleted of Ba and R S c (at.%) [B3O6]37 40 B 30 20 Ba Na Tc [Na2O] 12 BaO BaB2O4 Figure 19. Schemes illustrating the distributions of the B, Ba, Na and [B3O6]37 ring ion concentrations in the crystal ± melt system for a steady- state directional crystallisation occurring at a rate R (a), of the equilibrium crystallisation temperature of b-BaB2O4 (solid line) and of actual change in the temperature (dashed line) (b), and the hypothetical change in the melt composition with the distance from the solidification front (diffusion path) during the growth of BBO crystals (c).104 Shown are the constitutional supercooling region (b, hatched), the melt composition at the solidification front (c, 1) and the starting composition of the melt (c, 2).P P Fedorov, A E Kokh, N G Kononova ~~~~~~ a L B BaNa [B3O6]37l b TL l B2O3Barium borate b-BaB2O4 as a material for nonlinear optics B near the crystallisation front and induces diffusion flows directed from the bulk of the melt towards the interface. Accu- mulation of the excess sodium should reduce the equilibrium liquidus temperature. The integrated distribution curve of the equilibrium crystallisation temperature is presented in Fig.19 b while the possible shift of the figurative point corresponding to the melt composition for actual crystallisation is shown in Fig. 19 c. Since the BaB2O4±Na2O section is unstable, the line correspond- ing to the change in the melt concentration with distance from the solidification front (diffusion path) should deviate from this section, thus forming a zig-zag curve on the concentration triangle.167 More detailed analysis of the situation requires a consider- ation of different forms of boron present in the melt. As men- tioned above, melting of BaB2O4 causes destruction of the [B3O6]37 ring anions (i.e., this crystal is a typical example of compound which nearly completely dissociates on melt- ing).99, 100, 112 According to the molecular dynamics calcula- tions,168 the addition of Na2O to the melt leads to shortening of the [BO2]n¡ n polymer chains formed by [BO3]37 triangles and to an increase in the proportion of isolated [BO3]37 triangles, dimeric [B2O5]47 units, three-membered [B3O7]57 fragments and [B3O6]37 ring anions.Nevertheless, one can state that it is the low concentration of the [B3O6]37 ring anions near the crystal- lisation front that is the limiting factor for the growth of BBO crystals. It should be noted that steady-state melt-solution growth of BBOcrystals can be performed using the Czochralski method with melt feed 132 or by floating zone technique.131 Experiments carried out according to Kyropoulos (with slow cooling of the system) represent a kind of non-steady-state processes characterised by accumulation of sodium near the solidification front (Fig.20). The simplest constitutional supercooling criterion for solid- ification front stability is provided by the Tiller ± Chalmers criterion GD R < mC0Ök1 ¡ kÜ , where G is the temperature gradient in the melt at the solid- ification front, D is the diffusion coefficient of the solute, R is the crystal growth rate, m is the slope of the liquidus curve, C0 is the concentration of the solute in the solution and k is the partition coefficient. This criterion was proposed for steady-state processes in dilute binary systems 162, 163 and cannot be applied to the crystallisation of BBO.This criterion is equivalent to the boun- dary condition at the crystallisation interface while the integrated curve describing the increase in the equilibrium crystallisation temperature (liquidus temperature) with the distance from the solidification front of BBO follows the same pattern as in the case of binary systems. Therefore, the criterion for morphological stability of the solidification front of BBO (heat transfer is assumed negligible) can be written as follows GR D < mDC , where D* is the function of the diffusion coefficients of the components of the system in the melt (it varies from one composition to another), m is the directional derivative of the liquidus temperature taken along the tangent to the diffusion path at the point corresponding to the melt composition at the solid- ification front (this point is labelled 1 in Fig.19 c ) and DC is the change in the composition at the interface. An analogous expres- sion proposed166 for binary systems can also be used for non- steady-state processes. In any case the most important factors improving the stability of the solidification front against constitutional supercooling are an increase in the temperature gradient at the crystallisation front and a decrease in the crystal growth rate. a L S CNa ~~ ~~ ~~ b [Na2O] 2 1 BaO B2O3 BaB2O4 Figure 20. Schemes illustrating successive accumulation of sodium ahead of the crystallisation front for a non-steady-state growth of BBO crystal (a) and the change in the sodium content in the crystal (1) and in the melt (2) (b).There are two reasons for deterioration of the optical quality of flux grown BBO crystals. These are (i) the attainment of critical supersaturation with respect to sodium near the crystallisation front and (as a consequence) the appearance of a cellular structure with trapped inclusions and (ii) accumulation of sodium impurity above the equilibrium solubility (limiting concentration) in BBO followed by crystallisation of the second phase on cooling of the crystal. In both cases mixing of the melt for not only control of the heat convection, but also homogenisation of the melt composition is of paramount importance.X. Properties and fields of application of b-BaB2O4 crystals Crystals of the non-centrosymmetric low-temperature modifica- tion, b-BaB2O4, are widely used in nonlinear optics as nonlinear- optical crystals possessing large birefringence for the generation of optical harmonics and sum frequencies of laser radiation in the visible and UV regions. Widespread use of this material in quantum electronics is to a great extent due to relatively high optical damage threshold. Numerous applied 30, 169 ± 195 and mate- rials science studies 156, 157, 196 ± 201 of BBO were reported in the last decade. Development of compact solid-state sources of tunable UV, visible and IR radiation is topical. Currently, BBO-crystal para- 667 l l l l668 metric light oscillators (OPOs) pumped by the second and third harmonics of Nd:YAGlasers are most widely used.These devices generate tunable radiation in the 2628 and 3036 nm regions, respectively.192 The tuning range of BBO-crystal OPOs is limited to strong absorption of the idler wave radiation in the spectral region from 2200 to 3000 nm.44 The pioneering study of the characteristics of a BBO-crystal OPO pumped by the fourth harmonic of a Nd :YAG laser (l=266 nm) was reported by Bosenberg et al.193 A BBO-crystal OPO 25 mm long generated radiation in the spectral region between 330 and 1370 nm at a pulsed pump intensity exceeding 23 MW cm72. Another BBO- crystal OPO pumped by the fourth harmonic of a single-mode Nd:YAGlaser generated radiation in the region 302 ± 2248 nm at a pulsed pump energy of 20 mJ with a conversion efficiency of 6.3% at a wavelength of 340.2 nm at quadruple the oscillation threshold.192 The characteristics of a BBO-crystal Type-I OPO pumped by the fourth harmonic of a multimode Nd:YAG laser have been studied.194 The output was tunable over the region 300 ± 2340 nm.The efficiency ranged up to 15% near the peak of the tunability curve at double the oscillation threshold. The OPO configuration was based on a BBO crystal (aperture 668 mm, length 14 mm) cut for Type-I phase matching at an angle (y) of 38.3 8 with respect to the optical axis of the crystal. The above-mentioned applications of BBO crystals were based on the second-order optical nonlinearity.However, b-BaB2O4 also possesses third-order optical nonlinearity.30 In particular, it was found that picosecond laser pump of BBO at 1 mm excites efficient Stokes and anti-Stokes Raman parametric generation (RPG) and allows simultaneous SHG and multicom- ponent RPG in the visible and near IR regions. In particular, a BBO-crystal active element 15 mm long [excitation configuration a(cc)a] pumped at 1 mm re-emitted to all RPG components more than 35% of the energy (*1 mJ), the first Stokes line (l=1.1414 mm) accounting for *20%. The efficiency of pico- second RPG from b-BaB2O4 was also illustrated by an extended (up to the thirteenth component with l=0.5661 mm) wing of the multicomponent anti-Stokes spectral region.Based on high SHG efficiency of BBO crystals, an SRS experiment with the optically active element oriented parallel to the Type-I phasematching direction (ee-o) was carried out.30 The experiments showed the possibility of simultaneous conversion of the 1 mm pump of BBO into both the Stokes and anti-Stokes components and the second harmonic. It was found that excitation of the multicomponent RPG induced by the second harmonic generated by the BBO crystal occurs at a lower pump power, which is due to high SHG efficiency from BBO crystals and to dispersion of the Raman amplification coefficient of BBO. This discovery�simultaneous manifestation of the second- and third-order optical nonlinearities in BBO crystals � can be used for creation of novel types of b-BaB2O4-based laser frequency converters.Electro-optical application of BBO crystals as Pockels cells in the UV spectral region should be emphasised.195 Recently, BBO crystals have been on the commercial market. A large body of valuable information on the physicochemical properties of BBO crystals can be found in the Internet at URLs http://www.crystaux.com/betabbo.htm (website of DLK Crystal) and http://www.clevelandcrystals.com/BBOLBO.shtml (website of Cleveland Crystals, Inc.). The principal properties of BBO crystals are listed in Table 4. XI. Conclusion Based on the aforesaid, the following conclusions can be drawn. Polymorphism of BaB2O4 is a unique phenomenon. On the one hand, the high- and low-temperature modifications of this compound have substantially different structures. On the other hand, their thermodynamic characteristics are nearly identical. Studies of the mutual stability of these modifications and of the influence of different factors on their crystallisation and kinetics P P Fedorov, A E Kokh, N G Kononova Table 4.Properties of BBO single crystals. Parameter Optical transmission /nm (50% transmission at a thickness of 3 mm) Indices of refraction a at l 0.2660 nm 0.4358 nm 0.5461 nm 0.6328 nm 1.0642 nm Absorption coefficients a (cm71) at l 0.266 nm 1.0 nm 2.09 nm Laser damage threshold /GW cm72 at l=1.0642 nm 10 ns pulse 0.1 ns pulse (2.3 J) Nonlinear optical susceptibility /pm V71 Phasematching range (l /nm) Type I SHG Type II SHG Phasematching angle f (deg) at maximum nonlinear optical susceptibility and 25 8C 0.532 nm, type I SHG 0.532 nm, type II SHG 1.064 nm, type I SHG 1.064 nm, type II SHG Linear electro-optical coefficients g /pm V71 Hardness, Mohs Elastic stiffnesses /GPa c11 c12 c13 c14 c33 c44 c66 Young's modulus /GPa Poisson's ratio Thermal expansion coefficient,i 1076 (deg)71 Specific heat /J (g deg)71 Thermal conductivity /W (m deg)71 parallel to c axis perpendicular to c axis Resistivity /O cm Relative static dielectric constants k a The n0 and ne values are respectively given.b e-Ray. c o-Ray. d The d22 value. e The d15 value. f Given are the y and F values, respectively.g Listed are the r22 , r31 and r6 values, respectively. i Given are the aa and ac values, respectively. k Given are the e11/e0 and e33/e0 values, respectively. of transformations are of great importance. In this connection, investigations of the heat capacities of both modifications of BaB2O4 over a wide temperature range are also required. HuÈ bner's data 62, 65 on the stability of compound Ba2B2O5 and on the formation of BaB2O4-based solid solutions strongly contra- dict the phase diagram of the BaO ± B2O3 system proposed by Levin et al.12, 13 An indirect proof of variable composition of BBO is provided by small but, at the same time, noticeable variations of the lattice parameters of single crystals, determined by different authors. Both nonstoichiometry of BaB2O4 and the BaO ± B2O3±Na2O system in the BaO-enriched region (in particular, in the Ba3B2O6 ± NaBaBO3usmn; BaB2O4 concentration triangle) call for further investigation. Value 0.196 ± 2.2 1.7571; 1.6146 1.6868; 1.55638 1.6738; 1.5547 1.6673; 1.5500 1.6551; 1.5425 0.04 ± 0.15 0.001 ± 0.002 0.0085;b 0.07 c 0.0046 15 2.3;d <0.1 e 0.4096 ± 3+ 0.53 ± 3+ 47.6; 90 81.0; 0 22.8; 90 32.8; 0 2.70.4;*0; 0.055 4.5 123.8 60.3 49.4 12.3 53.3 7.8 31.8 39 0.58 4; 36 0.490.02 1.6 1.2 >1011 6.7; 8.1Barium borate b-BaB2O4 as a material for nonlinear optics Studies of the phase equilibria in the BaO ± B2O3±Na2O ternary system showed that the BaB2O4±Na2O section, which is most often used for growth of b-BaB2O4 single crystals from Na2O, is unstable and that compound Na2BaB2O5 does not exist.Compound NaBaBO3 was identified and structurally character- ised. The BaB2O4±Na2O section passes nearly exactly through the eutectic in the BaB2O4 ± NaBaBO3 ±NaBO2 ternary system. The BaB2O4±Na2O melt used for growth of BaB2O4 crystals is multicomponent and unstable. Further search for optimum growth conditions and better solvents seems to be appropriate. Improvement of methods of control of the heat and mass transfer during crystal growth including those involving variation of the heat field patterns is also topical. This review has been performed as part of the Research Project `Investigations of the Phase Diagrams, Search for Solvents and Growth of Nonlinear Optical Crystals in Heat Fields of Different Symmetry' carried out in the framework of the Integra- tion Programme `Novel Principles and Methods of Design and Targeted Synthesis of Chemical Compounds with Preset Proper- ties' in accord with the Decision of the Presidium of the Siberian Branch of the Russian Academy of Sciences signed on 09.04.2002.The authors express their gratitude to P Peshev (Institute of General and Inorganic Chemistry, Bulgarian Academy of Scien- ces, Sophia) for providing access to valuable information, to R M Zakalyukin, E A Tkachenko (A V Shubnikov Institute of Crystallography, Russian Academy of Sciences, Moscow) and S V Kuznetsov (M V Lomonosov Moscow State Academy of Fine Chemical Technology) for technical support and to Yu V Pisarevskij (A V Shubnikov Institute of Crystallography, Russian Academy of Sciences) and A M Yurkin (Institute of Mineralogy and Petrography, Siberian Branch of the Russian Academy of Sciences) for valuable help. References 1.R Benedikt Ber. Detsch. Chem. Ges. 7 703 (1874) 2. Gmelin Handbuch der Anorganichen Chemie (Berlin: Verlag Chemie, 1932) 3. Gmelin Handbuch der Anorganichen Chemie. Borverbindungen. Boroxide. Borsauren. Borate (Berlin; Heidelberg; New York: Springer, 1975) Bd 28, Teil 7 4. A D Keshan Sintez Boratov v Vodnom Rastvore i ikh Issledovanie (The Synthesis of Borates in Aqueous Solution and Investigation of Them) (Riga: Izd. Akad. Nauk Latv. SSR, 1955) 5.G K Gode, Doctoral Thesis in Chemical Sciences, Riga, 1974 6. USSR P. 1 611 861; Byull. Izobret. (45) 123 (1990) 7. M R Suvorova, E L Gabova, T V Sorokina, L N Kalitina, N I Berg, I A Leont'eva Tekhnol. Bornykh Soedin. 40 70 (1976) 8. USSR P. 1 024 866; Byull. Izobret. (23) 151 (1983) 9. H Rawson Inorganic Glass-Forming Systems (London; New York: Academic Press, 1967) 10. E M Levin, H F McMurdie, C R Robbins Phase Diagrams for Ceramics (Washington, DC: NBS, 1964) 11. N A Toropov, V P Bazarkovskii, V V Lapin, N N Kurtseva Diagrammy Sostoyaniya Silikatnykh Sistem. Spravochnik (State Diagrams of Silicate Systems. Handbook) (Leningrad: Nauka, 1969) 12. E M Levin, H F McMurdie J. Res. Nat. Bur. Stand. 42 131 (1949) 13. E M Levin, G M Urginic J.Res. Nat. Bur. Stand. 51 37 (1953) 14. O V Mazurin, M V Strel'tsina, T P Shvaiko-Shvaikoskaya Svoistva Stekol i Stekloobrazuyushchikh Rasplavov. Spravochnik (Properties of Glasses and Glass Forming Solutions. Handbook) (Leningrad: Nauka, 1975) Vol. 2 15. N I Leonyuk, L I Leonyuk Kristallokhimiya Bezvodnykh Boratov (Crystal Chemistry of Anhydrous Borates) (Moscow: Moscow State University, 1983) 16. S Block, A Perloff Acta Crystallogr. 19 297 (1965) 17. A D Mighell, A Perloff Acta Crystallogr. 20 819 (1966) 18. J Krogh-Moe,M Ihara Acta Crystallogr., Sect. B 25 2153 (1969) 19. H B Gray Electrons and Chemical Bonding (New York; Amsterdam: W A Benjamin, 1965) 20. C Chen, B Wu, A Jiang, G You Sci. Sinica, B 28 235 (1985) 669 21. C Chen, G Liu Annu.Rev. Mater. Sci. 16 203 (1986) 22. S Lu,M Ho, J Huang Wuli Xuebao (Acta Phys. Sinica) 31 948 (1982) 23. J-R Liang, Y-L Zhang, A-Z Huang Acta Chem. Sinica. 40 994 (1982) 24. D Y Tang, S T Ling J. Synt. Crystals 14 148 (1985) 25. A D Jiang, F Cheng J. Synt. Crystals 14 149 (1985) 26. L A Shuvalov, A A Urusovskaya, I S Zheludev, A V Zalesskii, S A Semiletov, B N Grechushnikov, I G Chistyakov, S A Pikin Sovremennaya Kristallografiya. Fizicheskie Svoistva Kristallov (Modern Crystallography. Physical Properties of Crystals) (Moscow: Nauka, 1981) Vol. 4 27. A A Blistanov Kristally Kvantovoi i Nelineinoi Optiki (Crystals of Quantum and Nonlinear Optics) (Moscow: Moscow Institute of Steel and Alloys, 2000) 28. V V Osiko, T T Basiev, in IX Natsional'naya Konferentsiya po Rostu Kristallov (Tez.Dokl.) [(The IXth National Conference on Growth of Crystals (Abstracts of Reports)] (Moscow: Institute of Catalysis, Russian Academy of Sciences, 2000) p. 7 29. T T Basiev Usp. Fiz. Nauk 169 1149 (1999) [Phys.-Usp. 42 1051 (1999)] 30. A A Kaminskii, S N Bagaev, A M Yurkin, A E Kokh, G Eikhler, Yu Findaizen Dokl. Akad. Nauk 367 468 (1999) a 31. C Chen, Y Wu, A Jiang, B Wu, G You, R Li, S Lin J. Opt. Soc. Am. B 6 616 (1989) 32. C Chen, Z Xu, D Deng, J Zhang, G K L Wong, Y Wu, N Ye, D Tang Appl. Phys. Lett. 68 2930 (1996) 33. C Chen, Y Wang, B Wu, K Wu,W Zeng , L Yu Nature (London) 373 322 (1995) 34. Y Mori, I Kuroda, S Nakajima, T Sasaki, S Nakai Appl. Phys. Lett. 67 1818 (1995) 35.N Ye, Z Weng, B Wu, C Chen Proc. SPIE-Int. Soc. Opt. Eng. 3556 156 (1998). 36. P Becker, J Liebertz, L Bohaty J. Cryst. Growth 203 149 (1999) 37. C Chen, Y Wu, R Li J. Cryst. Growth 99 790 (1990) 38. C Chen J. Synt. Cryst. (China) 30 36 (2001) 39. L I Isaenko, A P Eliseev Khim. Inter. Ustoich. Razvit. 8 103 (2000) b 40. R H French, J W Ling, F S Ohuchi, C T Chen Phys. Rev. B 44 8496 (1991) 41. W-D Cheng, J-X Lu Chinese J. Struct. Chem. 16 81 (1997) 42. J Lin, M-H Lee, Z-P Liu, C Chen, Ch J Pickard Phys. Rev. B 60 13 380 (1999) 43. D Xue,K Betzler, H Hesse,D Lammers Solid State Commun. 114 21 (2000) 44. D Elmeri, L Davis, S Velsko, E K Graham, A Zalkin J. Appl. Phys. 62 1966 (1987) 45. L K Cheng,W Bosenberg, C L Tang Prog.Cryst. Growth Charact. 20 9 (1990) 46. R S Feigelson, R J Raymakers, R K Route Prog. Cryst. Growth Charact. 20 115 (1990) 47. W Eitel The Physical Chemistry of the Silicates (Chicago: University of Chicago Press, 1954) 48. N I Leonyuk J. Cryst. Growth 174 301 (1997) 49. V P Soltsev, E G Tsvetkov, V A Gets, V D Antsygin J. Cryst. Growth 236 290 (2002) 50. U Sborgi Atti Reale Acad. Naz. Linc., Sec. 5 28 854 (1914) 51. H-A Lehmann, K Muhmel, Dzui-Fang Sun Z. Anorg. Allg. Chem. 355 238 (1967) 52. V B Kravchenko Zh. Strukt. Khim. 6 724 (1965) c 53. L Kutschabsky Acta Crystallogr., Serct. B 25 1811 (1969) 54. V I Simonov, O G Karpov, Ya Ya Shvirkst, G K Gode Kristallografiya 34 1292 (1989) d 55. D Yu Pushcharovsky, S Merlino, O Ferro, S A Vinigradova, O V Dimitrova J.Alloys Compd. 306 163 (2000) 56. O Ferro, S Merlino, S A Vinigradova, D Yu Pushcharovsky, O V Dimitrova J. Alloys Compd. 305 63 (2000) 57. A V Arakcheeva, S A Vinogradova, D Yu Pushcharovskii, M Khostettler, Zh Shapui, O V Dimitrova Kristallografiya 45 448 (2000) d 58. O V Dimitrova, in Kristallogenezis i Mineralogiya (Tez. Dokl.) [Crystallogenesis and Mineralogy (Abstracts of Reports)] (S Petersburg: S Petersburg University, 2001) p. 92 59. W Guertler Z. Anorg. Chem. 40 343 (1904) 60. F de Carli Atti Reale Acad. Naz. Linc., Sec. 6 5 44 (1927) 61. V B M Hageman, H A J Oonk Phys. Chem. Glasses 20 126 (1979) 62. K-H HuÈ bner Neues Jahrbuch Mineralog. Monatsh. 335 (1969)670 63. Jpn. P. 249 698 (A); Izobreteniya Stran Mira (Inventions of World States) (Moscow: All-Russian Research Institute of Patent Information, 1990) 64.V Nikolov, P Peshev J. Solid. State Chem. 96 48 (1992) 65. K-H HuÈ bner Neues Jahrbuch Mineralog. Abhandl. 112 150 (1970) 66. A B Bykov, V V Dozmarov, O K Mel'nikov Kristallografiya 39 720 (1994) 67. C B Robbins, E M Levin J. Res. Natl. Bur. Stand. A 73 615 (1969) 68. J L Stone, A Keszler, G Aka, A Kahn-Harari, T A Reynolds Proc. SPIE-Int. Soc. Opt. Eng. 4268 175 (2001) 69. J Liebertz, S Stahr Z. Kristallogr. 165 91 (1983) 70. R Frohlich Z. Kristallogr. 168 109 (1984) 71. K Ito, F Marumo,M Ohgaki, K Tanaka Res. Lab. Eng. Mater. Tokyo Inst. Technol. 15 1 (1990) 72. R Sole, V Nikolov,M C Pujol, J Gavalda, X Ruiz, J Massons, M Aguilo, F Diaz J.Cryst. Growth 207 104 (1999) 73. Powder Diffraction File Card 15-0862, JCPDS (Swartmore, PA: International Center of Diffraction Data) 74. Powder Diffraction File Card 44-0584, JCPDS (Swartmore, PA: International Center of Diffraction Data) 75. C E Weir, R A Schroeder J. Res. Nat. Bur. Stand. A 68 465 (1964) 76. I Wozniak, P F James Glass Technol. 25 98 (1984) 77. E M Shvarts, A E Dzene, in Boraty i Boratnye Sistemy (Borates and Borate Systems) (Ed. G Ya Slaidin) (Riga: Zinatne, 1978) p. 101 78. A F Nikolaev Fiziko-Khimicheskoe Izuchenie Prirodnykh Boratov (Physicochemical Study of Natural Borates) (Moscow: Izd. Akad. Nauk SSSR, 1947) 79. K Itoh, F Marumo, Y Kuwana J. Cryst. Growth 106 728 (1990) 80. H Kouta, Y Kuwana, K Ito, F Marumo J.Cryst. Growth 114 676 (1991) 81. S Imoto, S Kimura, Y Anzai, Y Kuwano J. Cryst. Growth 135 279 (1994) 82. N Yoshimoto, S Kimura J. Cryst. Growth 144 229 (1994) 83. D M Schleich Solid State Ion. 79/71 407 (1994) 84. O Yamaguchi, K Tominaga, K Shimizu Ceramurg. Int. 6 103 (1980) 85. L I Ivleva, D T Kiselev, Yu S Kuz'minov, V V Voronov, N M Polozkov Izv. Akad. Nauk SSSR, Neorg. Mater. 24 1153 (1988) 86. L K Cheng,W Bosenberg, C L Tang J. Cryst. Growth 89 553 (1988) 87. K L Ovanesyan, A G Petrosyan, G O Shirinyan Cryst. Res. Technol. 24 859 (1989) 88. T Katsumata, H Ishijima, T Sugano, M Yamagishi, K Takahashi J. Cryst. Growth 123 597 (1992) 89. H Kimura, T Numazawa, M Sato J. Cryst. Growth 165 408 (1996) 90.E G Tsvetkov, A M Yurkin Kristallografiya 43 1142 (1998) d 91. E G Tsvetkov, A A Tomilenko, G G Khranenko, A M Yurkin Kristallografiya 45 762 (2000) d 92. L I Ivleva, I G Gordadze, Yu S Kuz'minov, V V Voronov, V M Ivanovskaya Izv. Akad. Nauk SSSR, Neorg. Mater. 25 804 (1989) e 93. L H Brixner, K Babcock Mater. Res. Bull. 3 817 (1968) 94. Jpn. P. 63-18 755 (A); Izobr. Stran Mira 72 (MKI 130) (6) 13 (1990) 95. Powder Diffraction File Card 1-0861, JCPDS (Swartmore, PA: International Center of Diffraction Data) 96. M Marezio, H A Plettinger, W H Zachariasen Acta Crystallogr. 16 594 (1963) 97. J Q Lu, G X Lan, B Li, Y Y Yang, H F Wang, B C Wu J. Phys. Chem. Solids 49 519 (1988) 98. J Liebertz Z. Kristallogr. 182 307 (1988) 99. Yu K Voronko, A V Gorbachev, V V Osiko, A A Sobol, R S Feigelson, R K Route J.Phys. Chem. Solids 54 1579 (1993) 100. Yu K Voron'ko, A V Gorbachev, A A Sobol', L I Tsymbal Neorg. Mater. 30 646 (1994) e 101. A E Kokh, N G Kononova, P P Fedorov, V S Boyarkov, R M Zakalyukin,V A Chubarenko Kristallografiya 48 616 (2002) d 102. P P Fedorov, N G Kononova, A E Kokh, A A Sobol', Yu F Kargin, V S Boyarkov, R M Zakalyukin, E A Tkachenko Zh. Neorg. Khim. 47 1150 (2002) f 103. A Jiang, F Cheng, Q Lin, G Cheng, Y Zheng J. Cryst. Growth 79 963 (1986) P P Fedorov, A E Kokh, N G Kononova 104. A F Selevich, E E Vershel', in Aktual'nye Voprosy Sovremennoi Khimii (Tez. Dokl. 2-oi Respubl. Konf. Molodykh Uchenykh i Spetsialistov) [Topical Problems of Modern Chemistry (Abstracts of Reports of the Second Republican Conference of Young Scientists and Specialists)] (Minsk: Belorussia State University, 1991) p.33 105. Q-Z Huang, J K Liang Acta Phys. Sin. 30 559 (1981) 106. Y Kozuki, M Itoh J. Cryst. Growth 114 683 (1991) 107. H Kimura, R S Feigelson J. Alloys Compd. 234 187 (1996) 108. J Liebertz, R Frohlich Z. Kristallogr. 168 293 (1984) 109. X Huang, Y Zhao, K Lu, D Tang J. Cryst. Growth 165 413 (1996) 110. P P Fedorov, B P Sobolev Zh. Fiz. Khim. 62 896 (1988) g 111. G H Rao, J K Liang, Z Y Qiao,Q Z Huang CALPHAD, Comput. Coupling Phase Diagr. Thermochem. 13 169 (1989) 112. Yu K Voron'ko, A V Gorbachev, A B Kudryavtsev, V V Osiko, A A Sobol' Neorg. Mater. 28 1713 (1992) e 113. S Takeda, S Kimura, K Sugiyama High Temp.Mater. Process 10 209 (1992) 114. H Ogawa, Y Waseda J. Cryst. Growth 128 945 (1993) 115. Sangeeta, S C Sabharwal J. Cryst. Growth 222 427 (2001) 116. H Kimura, T Numazawa, M Sato J. Cryst. Growth 174 308 (1997) 117. K-T Vil'ke Vyrashchivanie Kristallov (Crystal Growing) (Leningrad: Nedra, 1977) 118. A A Chernov, E I Givargizov, Kh S Bagdasarov, L N Dem'yanets, V A Kuznetsov, A N Lobachev Sovremennaya Kristallografiya. Obrazovanie Kristallov (Modern Crystallography. Crystal Formation) (Moscow: Nauka, 1980) Vol. 3 119. V A Timofeeva Rost Kristallov iz Rastvorov-Rasplavov (Crystal Growth from Solution-Melts) (Moscow: Nauka, 1978) 120. D Y Tang, R K Route, R S Feigelson J. Cryst. Growth 91 81 (1988) 121. D Y Tang,W R Zeng, Q L Zhao J.Cryst. Growth 123 445 (1992) 122. A B Kaplun, E N Galashov, G D Vshivkova Neorg. Mater. 27 1723 (1994) e 123. MRoth, N Angert,MTseitlin, G Wang, T P J Han, H G Gallaher, N I Leonyuk, E V Koporulina, S N Barilo, L A Kurnevich, in The 3rd International Conference on Single Crystal Growth, Strength Problems, and Heat Mass Transfer (Abstracts of Reports), Obninsk, 1999 p. 67 124. Q Huang, J Luang J. Cryst. Growth 97 720 (1989) 125. D M Gualtieri, B H-T Chai,M H Randles J. Cryst. Growth 97 613 (1989) 126. A B Kaplun, A B Meshalkin Neorg. Mater. 31 1602 (1995) e 127. P F Bordui, G D Calvert, R Blachman J. Cryst. Growth 129 371(1993) 128. E G Rakov, V V Teslenko Pirogidroliz Neorganicheskikh Ftoridov (Pyrohydrolysis of Inorganic Fluorides) (Moscow: Energoatomizdat, 1987) 129.S N Barilo,A P Ges, S A Guretskii J. Cryst. Growth 108 309 (1991) 130. R S Feigelson,R J Raymakers,R K Route J. Cryst. Growth 97 352 (1989) 131. R Oude Hengel, F Fischer J. Cryst. Growth 114 656 (1991) 132. D Perlov, M Roth J. Cryst. Growth 137 123 (1994) 133. A B Kaplun, E N Galashov, G D Vshivkova, A B Meshalkin Neorg. Mater. 30 521 (1994) e 134. A M Luginets, S A Guretskii, A P Ges, A S Milovanov, L V Markova, V S Burak J. Cryst. Growth 162 89 (1996) 135. V Nikolov, P Peshev, Kh Khubanov J. Solid State Chem. 97 36 (1992) 136. V Nikolov, P Peshev J. Cryst. Growth 147 117 (1995) 137. RF P. 2 139 957; Byull. Izobret. (29) (1999) 138. A E Kokh, V A Mishchenko, V D Antsigin, A M Yurkin, N G Kononova, V A Guets, Y K Nizienko, A I Zakharchenko Proc.SPIE-Int. Soc. Opt. Eng. 3610 139 (1999) 139. A E Kokh, N G Kononova J. Cryst. Growth 198/199 161 (1999) 140. A E Kokh, N G Kononova, P W Mokruchnikov J. Cryst. Growth 216 359 (2000) 141. A E Kokh, V A Gets, A M Yurkin, O S Il'ina, E G Samoilova, in Tezisy Dokladov Vtoroi Konferentsii `Materialy Sibiri' (Abstracts of Reports of the Second Conference `Materials of Siberia') (Barnaul: Altai State Technology Institute, 1998) p. 145Barium borate b-BaB2O4 as a material for nonlinear optics 143. V I Tyurikov, E G Tsvetkov, G G Khranenko Poverkhnost' (5) 40 142. N G Kononova, A E Kokh, in IX Natsional'naya Konferentsiya po Rostu Kristallov (Tez. Dokl.) [The IXth National Conference on Crystal Growth (Abstracts of Reports)] (Moscow: Institute of Catalysis, Russian Academy of Sciences, 2000) p. 456 (2002) 144. E G Tsvetkov, V I Tyurikov Proc. SPIE-Int. Soc. Opt. Eng. 4268 186 (2001) 145. X Hong, K Lu, L Li, D Tang J. Cryst. Growth 193 610 (1998) 146. W R Bosenberg, R J Lane, C L Tang J. Cryst. Growth 108 394 (1991) 147. A Kokh J. Cryst. Growth 191 774 (1998) 148. RF P. 2 133 786; Byull. Izobret. (21) (1999) 149. A E Kokh, V E Kokh, V N Gets, N G Kononova Pribory Tekhn. Eksp. 1 153 (1998) 150. A E Kokh, V E Kokh,N G Kononova Pribory Tekhn. Eksp. 1 157 (2000) 151. A E Kokh, N G Kononova Solid-State Electron. 44 819 (2000) 152. RF P. 2 163 943; Byull. Izobret. (7) (2001) 153. RF P. 2 164 561; Byull. Izobret. (9) (2001) 154. A E Kokh, N G Kononova, V N Popov, P W Mokruchnikov Proc. SPIE-Int. Soc. Opt. Eng. 4268 161 (2001) 155. P A Morris J. Cryst. Growth 106 76 (1990) 156. N S Stel'makh, V E Kritskaya, G N Pirogova Neorg. Mater. 32 995 (1996) e 157. J Hou, Q Tan, Q Zhao, B Wu, C Chen Appl. Phys. Lett. 58 1149 (1991) 158. A Peter, K Polgar, E Beregi J. Cryst. Growth 209 102 (2000) 159. S C Sabharwal, Sangeeta,M Goswami, S K Kulkarni, B D Padalia J. Mater. Sci., Mater. Electron. 11 325 (2000) 160. S C Sabharwal, Sangeeta J. Cryst. Growth 187 253 (1998) 161. Q Tan, H Mao, S Lin, C Chen, S Lu, D Tang, T Ogawa J. Cryst. Growth 141 393 (1994) 162. B Chalmers Principles of Solidification (New York: Wiley, 1964) 163. M C Flemings Solidification Processing (New York: McGraw-Hill, 164. S R Corriel, G B McFadden, R F Sekerka Annu. Rev. Mater. Sci. 1974) 15 119 (1985) 165. J D Weeks, W Vansaarloos,M Grant J. Cryst. Growth 112 244 (1991) 166. P P Fedorov Neorg. Mater. 37 95 (2001) e 167. J S Kirkaldy, L C Brown Can. Metall. Q. 2 89 (1963) 168. Z Cheng, Y Lei, D Tang J. Cryst. Growth 183 227 (1998) 169. G C Bhar, S Das, U Chatterjee Appl. Opt. 28 2 (1989) 170. M Brown Opt. Lett. 23 1591 (1998) 171. C L Tang,W R Bosenberg , T Ukachi, R J Lane, L K Cheng Laser Focus World 10 107 (1990) 172. K Kato IEEE J. Quantum Electron. 22 1013 (1986) 173. S Imai, T Yamada, Y Fujimori, K Ishikawa Appl. Phys. Lett. 54 1206 (1989) 174. A Dubietis, G Tamosauskas, A Varanavicius, G Valiulis Appl. Opt. 39 2437 (2000) 175. P S Banks,M D Feit, M D Perry Opt. Lett. 24 4 (1999) 176. C Droz, H Kouta, Y Kuwano Opt. Rev. 6 97 (1999) 177. H Kouta, Y Kuwano Opt. Lett. 24 1230 (1999) 178. G C Bhar, U Chatterjee, A M Rudra, P Kumbkhakar Kvant. 179. J D Bhavalkar, Y Mao, H Po, A K Goyal, P Gavrilovic, Elektr. 29 800 (1999) [Quant. Electr. 29 800 (1999)] Y Conturie, S Singh Opt. Lett. 24 823 (1999) 180. K Matsubara, U Tanaka, H Imajo, M Watanabe, S Urabe J. Opt. Soc. Am. B 16 1668 (1999) 181. T Zhang, K Yamakawa Jpn. J. Appl. Phys. 39 91 (2000) 182. W Joosen, H J Bakker, L D Noordam, H G Muller, H B van Linden van den Heuvell J. Opt. Soc. Am. B 8 2087 (1991) 183. R Wu Appl. Opt. 32 971 (1993) 184. Z Y Min, R W Quandt, R Bersohn, H L Kim IEEE J. Quantum 185. W R Bosenberg,W S Pelouch, C L Tang Appl. Phys. Lett. 55 1952 Electron. 34 2409 (1998) (1989) 186. A Borsutzky, R Brunger, Ch Huang, R Wallenstein Appl. Phys. B 52 55 (1991) 187. Y Taira Jpn. J. Appl. Phys. 31 L682 (1992) 188. G C Bhar, S Das, U Chatterjee J. Phys. D 22 562 (1989) 189. G C Bhar, S Das, U Chatterjee Appl. Phys. Lett. 54 1383 (1989) 671 190. G C Bhar, P Kumbhakar, U Chatterjee, A M Rudra, Y Kuwano, H Kouta Appl. Opt. 37 7827 (1998) 191. G C Bhar, P Kumbhakar, U Chatterjee, A M Rudra, Y Kuwano, H Kouta Appl. Opt. 38 1802 (1999). 192. A Fix, T Schroder, R Wallenstein Laser Optoelectron. 23 106 (1991) 193. W R Bosenberg, L K Cheng, C L Tang Appl. Phys. Lett. 54 13 (1989) 194. N V Kondratyuk, A A Shagov, K L Demidchik, AMYurkin, Aleksandr E Kokh Kvant. Elektr. 30 253 (2000) [Quant. Electr. 30 253 (2000)] 195. V D Antsygin, V A Gusev, A M Yurkin Avtometriya 3 16 (1996) 196. V D Antsygin Avtometriya 1 24 (1998) 197. V T Adamiv, Ya V Burak, Ya O Dovgii, I V Kityk Kristallografiya 36 229 (1991) d 198. S-L Hong, B-C Wu Opt. Eng. 34 1738 (1995) 199. J D Beasley Appl. Opt. 33 1000 (1994) 200. H Kouta Appl. Opt. 38 545 (1999) 201. H Kouta, Y Kuwano Appl. Opt. 38 1053 (1999) a�Dokl. Chem. (Engl. Transl.) b�Chem. Sustain. Dev. (Engl. Transl.) c�Russ. J. Struct. Chem. (Engl. Transl.) d�Crystallogr. Rep. (Engl. Transl.) e�Inorg. Mater. (Engl. Transl.) f�Russ. J. Inorg. Chem. (Engl. Transl.) g�Russ. J. Phys. Chem. (
ISSN:0036-021X
出版商:RSC
年代:2002
数据来源: RSC
|
3. |
Isothiazoles (1,2-thiazoles): synthesis, properties and applications |
|
Russian Chemical Reviews,
Volume 71,
Issue 8,
2002,
Page 673-694
Rodislav V. Kaberdin,
Preview
|
|
摘要:
Russian Chemical Reviews 71 (8) 673 ± 694 (2002) Isothiazoles (1,2-thiazoles): synthesis, properties and applications R V Kaberdin, V I Potkin Contents I. Introduction II. Synthesis of isothiazoles III. Chemical transformations of isothiazoles IV. Areas of application of isothiazoles V. Conclusion Abstract. of chemistry the in achievements recent most The The most recent achievements in the chemistry of isothiazoles are generalised and systematised. The main applica- isothiazoles are generalised and systematised. The main applica- tions of isothiazole derivatives are surveyed. The bibliography tions of isothiazole derivatives are surveyed. The bibliography includes references 293 includes 293 references. I. Introduction Isothiazoles constitute a relatively novel class of heterocyclic compounds.Their ancestor, 1,2-thiazole, was first obtained in 1956.1 Other 1,2-azoles had been rather well studied by that time, both theoretically and with respect to their practical applications. Further rapid progress in the chemistry of isothiazole and intense studies into the synthesis and chemical conversions of its deriva- tives carried out in the 1960 ± 1990's were caused primarily by the extraordinarily broad range of useful properties manifested by various representatives of this class of compounds. Isothiazole- containing penicillins and cephalosporins competed successfully with ampicillin in their activities against gram-positive and gram- negative bacteria.2, 3 In the last decade of the 20th century, some isothiazole derivatives were found to be efficient in the treatment of Alzheimer's disease,4 as antiinflammatory, antithrombotic and anticonvulsive drugs 5, 6 and serine protease inhibitors;7, 8 physio- logically active compounds interacting with glutamate receptors were synthesised.9 Auxin transport inhibitors based on isothia- zoles 10 ± 14 are added to herbicides to enhance their activity.Isothiazole derivatives manifest synergistic effects when added to other biocidal preparations.15 ± 19 Compositions containing iso- thiazole derivatives are used as protectors for polymers, dyes, detergents 20 and leather goods,21 for decontamination of pig- ments, latexes, foodstuffs 22 and waste,23 as antifouling agents,24, 25 etc.Recently, isothiazoles have been patented for use in colour photography for stabilisation of photomaterials 26, 27 and as an inkbase for printers.28, 29 Industrial production of some isothiazolones, which represent broad-spectrum biocides and are safe for human beings, animals and environment, has begun.30 R V Kaberdin, V I Potkin Institute of Physical Organic Chemistry, National Academy of Sciences of Belarus, ul. Surganova 13, 220072 Minsk, Belarus. Fax (37-517) 284 16 79. Tel. (37-517) 284 16 00. E-mail: ifoch@ifoch.bas-net.by (R V Kaberdin) Tel. (37-517) 284 09 72. E-mail: potkin@ifoch.bas-net.by (V I Potkin) Received 3 June 2002 Uspekhi Khimii 71 (8) 764 ± 787 (2002); translated by R L Birnova #2002 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2002v071n08ABEH000738 673 673 684 690 691 The first data on the chemistry of isothiazoles were reviewed by Wooldridge et al.31, 32 Further progress in the synthesis and conversions of isothiazoles is documented in the reviews 33 ± 39 some of which are inaccessible for Russian readers as well as in the monographs 40 ± 45 where isothiazoles are considered only in con- junction with other heterocyclic compounds.The fundamental monograph by Houben-Weyl,46 which contains more than 500 references including some published in 1992, represents the most systematic and full generalisation of the experimental data on the chemistry of 1,2-thiazoles. The first review devoted to the physical and chemical properties of 1,2-azoles was published in Russian in 1979.47 In addition, isothiazoles and their benzoannelated ana- logues are discussed in `Comprehensive Organic Chemistry'.48 The necessity for systematisation and generalisation of the newest data on isothiazoles stems from the appearance, in the past decade, of a great number of publications devoted to the synthesis, conversions and applications of this class of compounds.The material described here covers original papers and patent data published over the period 1993 ± 2001 inclusive. In some cases, earlier sources concerned with the methods for isothiazole syn- thesis are cited for the sake of completeness. II. Synthesis of isothiazoles Isothiazole (1) was first prepared by the oxidation of 5-amino-1,2- benzoisothiazole (2) with an alkaline solution of potassium permanganate with subsequent decarboxylation of isothiazole- 4,5-dicarboxylic acid (3).1, 49 HO2C H2N N N N HO2C S S1 S3 2 This synthetic procedure has a purely historical significance. Later, isothiazoles were synthesised from simpler and more accessible compounds.1. Synthesis of isothiazoles based on cyclisation reactions The main methods for the construction of the isothiazole ring are based on cyclisation of compounds containing N7C7C7C7S fragments and heterocyclisation of compounds containing nitro- gen, sulfur and carbon atoms.674 a. Synthesis of isothiazoles from compounds containing the N7C7C7C7S group The construction of the isothiazole ring from compounds con- taining a preformed N7C7C7C7S fragment is the most attractive approach.Thus various 5-aminoisothiazole derivatives 4 can be prepared by the oxidation of 2-aminoalk-1-enethiocar- boxyamides 5.50 R S H2N I2, K2CO3, Et2O C CHC NH2 N H2N 5 R S4 R=Et, Pri, Bun, Bui, But, cyclo-C6H11. NR2 S Oxidation of 3-amino-3-(dialkylamino)thioacrylic acid amid- es 7 gives the 3,5-diaminoisothiazole derivatives 6.51 H2N oxidiser C CH C NH2 N H2N 7 R2N S6 R=Alk. A procedure for the synthesis of alkyl 3-methyl-5-ethoxycar- bonylaminoisothiazole-4-carboxylates (8) by the oxidation of substituted thioamides 9 with selenium dioxide has been devel- oped recently.52 Me RO2C S H2N SeO2 C C C NHCO2Et EtOH, 40 8C N EtO2CHN Me CO2R S8 9 R=Me, Et.b. Synthesis of isothiazoles from sulfur and compounds containing the C7C7C7N fragment The construction of the isothiazole ring from sulfur and com- pounds containing the C7C7C7N fragment is relatively little studied. This approach makes use mainly of nitriles.46 Thus the electrolysis of 3-aryl-2-phenylsulfonylacrylonitriles (10) in acetonitrile or DMF using a chemically active sulfur ± graphite electrode affords bis(5-arylisothiazol-3-yl) di- or tri- sulfides (11).53 Sn S N ArCH CC N N Ar Ar SO2Ph S S 11 10 Ar=Ph, 4-MeC6H4, 4-PhC6H4; n=2, 3. In the case of 2-phenyl-3-phenylsulfonylacrylonitrile (12), di(isothiazolyl) disulfide 13 containing the (Z)-2-cyano-2-phenyl- vinylthio group in position 3 of one of the isothiazole rings was obtained.53 Ph HC C C N Ph S S S S N PhSO2CH CC N N Ph S S Ph 13 12 It is believed that electrochemical synthesis of sulfur-contain- ing compounds consists in the elimination of the phenylsulfonyl group with concomitant addition of a polysulfide anion produced upon electroreduction of elemental sulfur.R V Kaberdin, V I Potkin According to the patent data,54 substituted a-amino ketones react with thionyl chloride to give the corresponding 4-hydroxy- isothiazoles. Thus the reaction of amino ketone 14 with SOCl2 in DMF yields isothiazole 15. When thionyl chloride was used in excess, the reaction products contained both isothiazole 15 and di(isothiazolyl) sulfide 16. SOCl2 PhCH2COCHPh DMF 14 NH2 Ph OH HO Ph Ph HO + N N N Ph S S S S 15 (62%) 16 The authors of the present review together with Yu A Olde- kop have demonstrated 55 ± 57 that heating of 2-nitropentachlo- robuta-1,3-diene (17) with sulfur resulted in its heterocyclisation and the formation of 4,5-dichloro-3-trichloromethylisothiazole (18).Cl CCl3 S, 190 ± 200 8C Cl2C CCl C CCl2 7SO2 N Cl NO2 17 S 18 This reaction offers a new approach to the construction of the isothiazole ring. It is known that the reaction of halobutenes and -butadienes with sulfur is used for the synthesis of the correspond- ing halogenothiophenes.58 In the case of nitrodiene 17, the reaction follows a different route: the heterocyclisation involves the nitro group and is accompanied by the evolution of sulfur dioxide. The yield of isothiazole 18 reaches 52%.Tetrachloro- thiophene (yield 4%) is the side product in this reaction. Appa- rently, the reaction has a radical character, since sulfur molecules exist in the form of reactive biradicals Sn at high temperature.59 c. Synthesis of isothiazoles using cycloaddition and condensation reactions The methods for the synthesis of isothiazoles using cycloaddition and condensation reactions of compounds containing a set of essential fragments are the most well-studied and attract, as before, the attention of investigators. Thus the 1,3-dipolar cycloaddition of nitrile sulfide 19 to dimethyl acetylenedicarboxylate (20) yields the isothiazole deriv- ative 21 the structure of which has been confirmed by X-ray diffraction analysis.60 R MeO2C + 7 N R C N S+MeO2C C C CO2Me 19 20 MeO2C 21 S R=The reactions of 1-cyano-3-(2-oxo-1,3,4-oxathiazol-5-yl)ada- mantane (22) and 1,4-bis(2-oxo-1,3,4-oxathiazol-5-yl)adaman- tane (23) prepared from the corresponding adamantanecarbox- amides and chlorosulfenylcarbonyl chloride with the alkyne 20 result in the decarboxylation of the oxathiazolone fragment.The reactions of the nitrile sulfides formed with the alkyne 20 proceed as 1,3-dipolar cycloadditions and yield 1-cyano-3-(4,5-dimeth- oxycarbonylisothiazol-3-yl)adamantane (24) and 1,3-bis(4,5- dimethoxycarbonylisothiazol-3-yl)adamantane (25), respec- tively.61675 Isothiazoles (1,2-thiazoles): synthesis, properties and applications O R2 CN CN R2 H2N O HN S HN N ClCOSCl 20, D N N HN O H2NC(O) S CONH2 N O S 29b 31 30 O R2 N S 22 CN HO BnO O O CO2Me R1= (29a), R2= (29b, 30, 31).HO OH BnO OBn CO2Me N S 24 O S CONH2 O N ClCOSCl 20, D The use of 1,3,5-trichloro-1,3,5,2,4,6-trithiatriazine (32) seems to be very promising for the synthesis of isothiazoles. It was found that ordinary allylic compounds of the type 33 react with the reagent 32 as two-carbon units to give 1,2,5- thiadiazoles 34 and as three-carbon units to give isothiazoles 35. The ratio of compounds 34 and 35 depends on the reaction conditions and the nature of the substituents.64 CONH2 O Cl O S N S N N 23 S S CO2Me Cl Cl S N32 N CO2Me R1 CH CHCH2R2 33 R2 R1 CH2R2 CO2Me + N N N CO2Me R1 N S 25 S 35 S 34 The reaction of nitrile sulfides of the thiophene series with dimethyl fumarate (26) gives substituted 4,5-dihydroisothiazoles 27.62 R O D RCONH2+ClCOSCl 70 ± 110 8C PhMe 7CO2 N O S Thus 1,3-diphenylpropene (33, R1=R2=Ph) reacts with two equivalents of compound 32 in boiling CCl4 to give 3,5- diphenylisothiazole 35 in 53% yield. The introduction of a substituent at the central carbon atom of the allylic system suppresses the formation of thiadiazoles.4-Methyl-3,5-diphenyl- isothiazole was obtained from 2-methyl-1,3-diphenylpropene without any admixtures of the corresponding thiadiazole. H MeO2C Ph Me C C 32 R Ph CH CCH2Ph H CO2Me + 26 7 N R C N S Ph Me S N MeO2C H H MeO2C S 27 The presence of terminal electron-withdrawing groups in the allylic derivatives 33 increases their reactivities with respect to the reagent 32.64 , , R= Me SMe Me SO2Me S S S In the case of non-symmetrical allylic compounds, the reac- tion proceeds regiospecifically and affords an isomer with the electron-withdrawing group adjacent to the sulfur atom of the ring.Thus the reaction of the trimer 32 with ethyl 3,3-dibenzyl- acrylate affords isothiazole 36 in 66% yield.64 Ph Bn CH2Ph 32 EtO2CHC C N EtO2C Bn The cycloaddition of 3,4,6-tri-O-benzyl-2,5-anhydro-D-allo- nonitrile N-sulfide (28) to dimethyl acetylenedicarboxylate (20) or dimethyl fumarate (26) followed by the oxidation of the reaction products with 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (DDQ) results in O-benzyl-3-(b-D-ribofuranosyl)isothiazoledi- carboxylate 29a, which was used for the preparation of the C-nucleoside 29b and isothiazolo[4,5-d ]pyrimidine analogues of the natural C-nucleosides pyrazofurin (30) and formycin (31).63 S 36 S N C BnO O R1 MeO2C 20 or 26+DDQ Trithiatriazine trichloride 32 can be used for the preparation of isothiazole derivatives from other heterocyclic compounds.It was shown 65 ± 67 that treatment of di- and trisubstituted N MeO2C OBn S 29a BnO28 furans 37a ± c with trithiatriazine trichloride 32 in boiling CCl4 results in their one-step conversion into 5-benzoylisothiazoles 38a ± c in high yields.676 Ph R R 32 Ph N Ph Ph O S 38a ± c O 37a ± c R = H (a), Ph (b), Br (c).2,5-Disubstituted furans 39a ± c give similar products 40a ± c (yields 50% ± 60%) with compound 32 upon boiling in CCl4 or toluene.65 R 32 N RC R R CCl4, D S O 40a ± c O 39a ± c R=4-MeC6H4 (a), 4-MeOC6H4 (b), But (c). It is assumed that heating of 1,3,5-trichloro-1,3,5,2,4,6-tri- thiatriazine (32) results in its fragmentation and the formation of the reactive monomer A which attacks the furan molecule to give the intermediate B. Recyclisation of the latter affords isothia- zoles.65 Cl S N N + 7 N S Cl 3 N S Cl S S A N Cl Cl 32 N N S Cl H S R R N 7HCl R R R R O O S 40 OB It has been shown 68 that fully substituted furans do not react with the reagent 32, since the elimination of HCl requires the presence of a hydrogen atom in the b-position of the starting furan.Yet another route to the synthesis of 5-acylisothiazoles 41 is based on the reaction of 2,5-disubstituted furans with ethyl carbamate, SOCl2 and pyridine.68 R2 NH2CO2Et, SOCl2 R1 N Py, PhH, D R2 R1 O O S 41 R1=R2=Ph, 4-MeC6H4, But; R1=4-MeOC6H4, R2=4-NO2C6H4. In contrast with 2,5-diphenylfuran (37a), the reaction of 2,5- diphenylthiophene (42) with the reagent 32 gives 5-benzoyl-3- phenylisothiazole (38a) in low yield. Presumably, the latter is formed upon oxidation of the intermediate 5-thiobenzoyl deriva- tive 43.69 Ph Ph 32 Ph Ph Ph CCl4 Ph N N S S S 42 O S 38a 43 Reactions of N-substituted 2,5-diphenylpyrroles 44a ± e with compound 32 afford mixtures of thiazolyl-substituted imines 45a ± e and 3,4-dichloropyrrole derivatives 46a ± e.The ratio of the reaction products depends on the nature of the substituent R.69 R V Kaberdin, V I Potkin Ph Cl Cl 32 + Ph Ph Ph CCl4 N Ph Ph NR S NRNR 44a ± e 46a ± e 45a ± e R=Ph (a), 4-MeOC6H4 (b), 4-O2NC6H4 (c), MeOC(O) (d), ButOC(O)H (e). Compounds 45b,e are formed as two geometrical isomers. Imine 45b is easily hydrolysed to give ketone 38a. The imine 45d formed from N-methoxycarbonylpyrrole (44d) is completely hydrolysed even at room temperature to give ketone 38a. It is of note that the reaction of 1-methyl-2,5-diphenylpyrrole (47) with the reagent 32 follows a different route and results in the formation of bi(1,2,5-thiadiazol-3-yl) 48; 5-benzoyl-3-phenyliso- thiazole (38a) is formed in only trace amounts.69 Ph N 32 S N +38a Ph Ph CCl4 N S N NMe Ph 48 47 Condensation of 2-dialkylaminomethyldithiomalonates 49, the reaction products of dithiomalonates 50 with formamide acetals, with hydroxylamine-O-sulfonic acid results in isothiazoles 51.70 C(S)OR1 H2C +R22 NCH(OR3)2 C(S)OR1 S 50 C R1O C(S)OR1 NH2OSO3H R22 NCH2CH N R1O C(S)OR1 49 S 51 R1, R3=Me, Et; R2=Me, Ph.Schulze et al.71 have developed two procedures for the syn- thesis of isothiazoles 52a ± c containing fluoro-substituted aryl groups.The first of them is based on the reaction of b-thiocyana- tocinnamaldehydes 53a,b with ammonium thiocyanate (54). Iso- thiazoles 52a,b were obtained in good yields (*60%). In this case, ammonium thiocyanate (54) acts as a source of ammonium required for ring closure. It is noteworthy that ammonium thiocyanate (54) was employed previously for the preparation of 4- and 5-alkylisothiazoles.72 The second route is based on the use of enamino thiones 55a ± c as starting compounds; their reaction with hydroxylamine- O-sulfonic acid gives isothiazoles 52a ± c. In this case, the yields reach 98%.CHO NH4SCN (54) Ar SCN 53a ± c N NMe2 Ar S 52a ± c NH2OSO3H Py, MeOH Ar S 55a ± c Ar=4-CF3C6H4 (a), 3-CF3C6H4 (b), 3-FC6H4 (c).Approaches to the synthesis of isothiazolium salts 56 were developed based on cinnamaldehydes 53a,b,d,e and enamino thiones 55d ± g.71Isothiazoles (1,2-thiazoles): synthesis, properties and applications CHO H2NAr2 HClO4 SCN Ar1 53a,b,d,e + N Ar2 Ar1 ClO¡ NHAr2 4 H2O2, AcOH S 56 HClO4 Ar1 S 55d ± g Ar1=4-CF3C6H4 (53a, 55f), 3-CF3C6H4 (53b, 55g), 4-FC6H4 (53d, 55d), 4-MeOC6H4 (53e, 55e); Ar2=Ph, 4-MeOC6H4. 4-Thiocyanato-1-aza-1,3-dienes 57 prepared from b-thiocya- nato a,b-unsaturated aldehydes 58 and arenesulfonohydrazides 59 undergo cyclocondensation to yield N-arylsulfonylisothiazol- 2-ylimines 60 and the corresponding N-arylsulfonyl-2-aminoiso- thiazolium salts 61.73 CH R1 R1 HCl NNHSO2Ar MeOH CHO+H2NNHSO2Ar 59 SCN R2 R2 57 SCN 58a ± e R1 + N 7 R2 NSO2Ar S 60 HClO4 base R1 HClO4 + ClO¡4N R2 NHSO2Ar S 61 R1=R2=Me (a); R1=Ph, R2=Me (b); R17R2=(CH2)n, n = 3 (c), 4 (d), 5 (e); Ar=Ph, 4-MeC6H4, 4-BrC6H4, 2,4,6-Me3C6H2, 2,4,6-Pri3C6H2.Thus, depending on the nature of the substituents R1, R2 and Ar, the reactions of unsaturated thiocyanato aldehydes with arenesulfono hydrazides afford different products. Isothiazoles are formed directly from compounds 58a,b, whereas alicyclic thiocyanato aldehydes 58c ± e yield stable hydrazones 57. Treat- ment of the latter with perchloric acid yields isothiazolium salts 61c ± e. A convenient procedure for the synthesis of N-acylisothiazo- lidinium trifluoroacetates 62, which includes treatment of N-acyl- amino sulfoxides 63 with trifluroacetic anhydride, has been developed.74 (CF3CO)2O N PhS(CH2)3NHCOR COR MeCN, 0 8C +SCF3CO¡263 O Ph 62 R=But, Ph.2. Synthesis of isothiazoles from other heterocyclic compounds Methods of synthesis of substituted isothiazoles from other heterocyclic compounds find wide application despite the fact that they belong to the most ancient ones. Thus 3,5-disubstituted isothiazoles 64a,b were prepared from isoxazoles 65a,b by the reaction with phosphorus pentasulfide in pyridine.75 677 Me HO Me HO P2S5 Py N N 4-RC6H4 4-RC6H4 S 64a,b O 65a,b R = H (a), MeO (b). The reaction of 2-imino-3,4-dihydro-2H-pyrrole 66 with sul- fur on heating is accompanied by dehydrogenation and incorpo- ration of the sulfur atom into the pyrrole ring resulting in the previously unknown thiazolylisothiazoles 67.76 Ph Ph Ph Ph CHPh S S N D Ph Ph N Ph N S N 67 66 The thermal fragmentation of 1,4,2-dithiazines in the presence of dienophiles aimed at the synthesis of 1,4-dithyines has been studied.74 It was found that 1,4,2-dithiazine derivatives 68 react with dimethyl acetylenedicarboxylate (20) on heating in o-dichlo- robenzene to give a mixture of dimethyl 4,5-dimethylthiophene- 2,3-dicarboxylate (69) and 3-(4-bromophenyl)-4,5-dimethyliso- thiazole (70).The latter is formed upon thermolysis of dithiazine 68 and is accompanied by a loss of the sulfur atom from position 4.77 Me S C6H4Br-4 MeO2C C C CO2Me 20 180 8C Me S N 68 Me Me C6H4Br-4 CO2Me + N Me Me CO2Me S 70 (34%) S 69 (20%) A new method for the synthesis of the isothiazole ring from 4-chloro-5-dicyanomethylidene-5H-1,2,3-dithiazole (71) pre- pared by the condensation of 4-chloro-1,2,3-dithiazole-5-thione with tetracyanoethylene oxide has been proposed recently.Treat- ment of the nitrile 71 with morpholine or benzyltriethylammo- nium chloride gives 3-morpholino- (72a) and 3-chloroisothiazole- 4,5-dicarbonitriles (72b), respectively.78 Cl S CN NC + N S CN NC O S O O HN N NC CN N Cl NC NC S 72a N S Cl NC BnEt3NCl +7 S 71 N NC S 72b A new simple procedure for the synthesis of isothiazoles from primary enamines and 4,5-dichloro-1,2,3-dithiazolium chloride (73) has been developed.Thus the reaction of methyl 3-amino- crotonate (74a) and 3-aminocrotononitrile (74b) with the salt 73 under mild conditions (20 8C, dichloroethane) afforded678 4-methoxycarbonyl-5-cyano-3-methylisothiazole (75a) and 4,5- dicyano-3-methylisothiazole (75b).79 Cl NH2 + N Me C CHR+ S Cl7 74a,b S 73 R=CO2Me (a), CN (b). Presumably, the conversion of enamine 74a into isothiazole 75a upon treatment with the salt 73 proceeds similarly to its conversion into isothiazole 76 under the action of thiophosgene used by Woodword et al. in the synthesis of colchicine back in 1965.80 NH2 CSCl2 Me C CH CO2Me 74a Me MeO2C N S 76 3. Synthesis of isothiazole derivatives by modification of compounds containing the isothiazole ring The syntheses of functionalised isothiazoles by exchange of substituents in the isothiazole ring or by transformation of the side chain of accessible derivatives are well documented.46 a.Substitution reactions Recently, Russian chemists have developed novel procedures for the synthesis of halogenated isothiazoles.81 ± 83 Thus bromo- and iodoisothiazoles 78 ± 83 have been synthesised from the readily available 3-hydroxyisothiazole (77).81 POBr3 OH H2SO4 NBS N Br S 77 PBr5 NBS, N-bromosuccinimide. Br I HIO4 I2 78 N S 80 Br Br2 oleum Br 79 Br HIO4 I2H2SO4 I R Me Cl N NC S 75a,b CO2MeCl Et3N Me C 72 HCl Cl H2N S Br N S 78 Br N S 79 Br I HIO4 I2 N I H2SO4 S 81 Br N S 82 Br N S 83 R V Kaberdin, V I Potkin The possibility of synthesis of alkenylhaloisothiazoles by palladium-catalysed cross-coupling of the halogen derivatives 80 and 83 with the terminal alkenes 84a ± d (the Heck reaction) has been studied.82 It was found that 3,4-dibromo-5-iodoisothiazole (83) does not react with alkenes under the reaction conditions.3-Bromo-4-iodoisothiazole 80 enters into the Heck reaction with the alkenes 84a ± c in the presence of the catalytic system Pd(OAc)2 ± NEt3 to give 4-alkenyl-3-bromoisothiazoles 85a ± c. 4,40-Bi(3-bromoisothiazole) (86) is the side product in this reac- tion; it is also formed in the absence of the alkene. Br I Pd(OAc)2 ± NEt3 +CH2 N MeCN, 100 8C CHR 84a ± c S 80 Br Br Br RHC CH + N N N S S 86 (9% ± 13%) S 85a ± c R=Ph (a, 60%), CO2Me (b, 29%), CN (c, 33%).3-Bromo-4-(1-oxopropan-2-yl)isothiazole (87) and biisothia- zolyl 86 were isolated from a complex mixture of reaction products of isothiazole 80 with allyl alcohol (84d). Me OHC Br Br HC I +86 (13%) CHCH2OH +CH2 N N 84d S 87 (6%) S 80 Under conditions of metal-complex catalysis, bromo- and iodoisothiazoles react with terminal alkynes.83 Thus tribromoiso- thiazole (82) reacts with phenylacetylene (88a), oct-1-yne (88b), propargyl alcohol (88c) and methoxymethylacetylene (88d) in the presence of (PPh3)2PdCl2 even at 20 8C to give the cross-coupling products (89a ± d).Br Br Br Br (PPh3)2PdCl2, CuI, NEt3 N N C RC Br +RC CH 88a ± d S 89a ± d S 82 R=Ph (a), n-C6H13 (b), CH2OH (c), CH2OMe (d). The highest yields (20% ± 56%) were obtained with 4 mol.% of (PPh3)2PdCl2 . 1,4-Diphenylbuta-1,3-diyne was isolated as a side product in the reaction of isothiazole 82 with phenylacetylene 88a.It should be noted that unsubstituted acetylene does not enter into the cross-coupling reaction with isothiazole 82 even at 80 8C; the reaction mixture contained only 3,4-dibromoisothiazole (79), the partial reduction product of the tribromide 82. Compound 79 did not react with phenylacetylene, i.e., only the bromine atoms in position 5 were active in the cross-coupling reaction. 5-Iodoiso- thiazoles were found to be the most reactive compounds.The reactions of 3,4-dibromo-5-iodoisothiazole (83) and 3-bromo-4,5- diiodoisothiazole (81) with terminal alkynes 88a,d occur in the presence of 2 mol.% of (PPh3)2PdCl2 and result in the corre- sponding 3,4-dibromo-5-alkynyl- (90) and 3-bromo-4-iodo-5- alkynylisothiazoles (91a,b).Isothiazoles (1,2-thiazoles): synthesis, properties and applications Br X Br X (PPh3)2PdCl2, CuI, NEt3, MeCN +RC CH N N C RC I 88a,d S 90, 91a,b S 81, 83 R X Compounds Br 90 91a 91b Ph I Ph I CH2 OMe The side reaction products of terminal alkynes with 3,4- dibromo-5-iodoisothiazole (83) and 3-bromo-4,5-diiodoisothia- zole (81), viz., the dibromide 79 and 3-bromo-4-iodoisothiazole (80), represent the reduction products of iodine in position 5.3-Bromo-4-iodoisothiazole (80) enter into cross-coupling reactions with phenylacetylene (88a) under conditions of metal- complex catalysis to give 3-bromo-4-(phenylethynyl)isothiazole (92).83 Br PhC C Br I +PhC CH N N (PPh3)2PdCl2, CuI, NEt3, MeCN 50 8C 88a S 92 (49%) S 80 3-Bromo-4-(3-methoxyprop-1-ynyl)-5-(phenylethynyl)isothi- azole (93), the first representative of isothiazoles with two alkynyl substituents, was synthesised by the reaction of 3-bromo-4-iodo- 5-(phenylethynyl)isothiazole (91a) with methoxymethylacetylene (88d) under similar conditions.83 Br C Br I MeOCH2C +MeOCH2C CH N N C PhC C PhC 88d S 91a S 93 (42%) 4,5-Dichloro-3-trichloromethylisothiazole (18) reacts smo- othly with various nucleophilic reagents resulting in the substitu- tion of the chlorine atom in position 5 of the isothiazole ring.Thus its reactions with sodium methoxide, ethoxide and isopropoxide afford the corresponding 5-alkoxy-4-chloro-3-trichloromethyli- sothiazoles (94a ± c) in 73%± 78% yields. The reaction of isothia- zole 18 with piperidine results in 4-chloro-5-piperidino-3- trichloromethylisothiazole (95).56 Cl CCl3 RONa N RO Cl CCl3 S 94a ± c R=Me (a), Et (b), Pri (c) N Cl Cl CCl3 S 18 HN N N S 95 (85%) The reaction of methyl 4,5-dichloroisothiazole-3-carboxylate (96) with sodium methoxide in methanol at 20 8C proceeds in a similar way and yields methyl 5-methoxy-4-chlorothiazole-3- carboxylate (97).84 At a higher temperature, the saponification of the ester 96 occurs.The reaction of sodium methoxide with 4,5-dichloroisothiazole-3-carboxylic acid piperidide (98) prepared by treatment of the ester 96 with piperidine yields compound 99, the substitution product of the methoxy group for the chlorine atom in position 5.84 Cl CO2Me N Cl S 96 Cl HN 96 Cl Cl MeO S 99 (32%) A vast variety of 4-aroyl- and 4-hetaroylisothiazoles were prepared from substituted 4-iodoisothiazoles.85 ± 88 This is exem- plified in the synthesis of 5-cyclopropyl-4-(2-methylsulfonyl-4- trifluoromethylbenzoyl)isothiazole (100) from 5-cyclopropyl-4- iodoisothiazole (101) by treatment with methylmagnesium bro- mide and 2-methylsulfonyl-4-trifluoromethylbenzoyl chloride (102).I F3C N + S 101 F3C N-Alkylisothiazolecarboxamides are synthesised from iodo- isothiazoles by treatment with CO and amines in the presence of palladium catalysts.89 Thus N-octyl-3-methylisothiazole-5-car- boxamide (103) was prepared by heating a 5-iodo-3-methyliso- thiazole (104) ± octylamine ± triphenylphosphine mixture in an atmosphere of CO (10 atm) in the presence of (PPh3)2PdCl2 (100 8C, 6 h).Me+Me(CH2)7NH2 N I S 104 Me(CH2)7NHC A convenient regiospecific method for the synthesis of iso- thiazole derivatives of the type 106 and 107 based on the reactions of thiazol-4-ylmagnesium bromide (prepared from 4-iodoisothia- zole (105) and ethylmagnesium bromide) with carbonyl com- pounds has been developed.90 679 Cl CO2Me MeONa N MeOH, 20 8C MeO S 97 (68%) C(O)N MeONa MeOH, 40 8C N S 98 C(O)N N SO2Me MeMgBr, THF, PhMe COCl 102 SO2Me O C N S 100 CO, (PPh3)2PdCl2 dioxane Me N O S 103680 R1CH(OH) R2=H N I S 106 1) EtMgBr 2) R1C(O)R2 N R1(O)C S 105 R2=OEt N S 107 R1=Ph, 4-ClC6H4, 2,4-Cl2C6H3, Ph; R2=H, OEt.For the first time, isothiazolyl gold(I) complexes were pre- pared by the reaction of isothiazol-5-yllithium with gold(I) chlor- ide or tetrahydrothiophene(pentafluorophenyl)gold; their protonation or alkylation gave the corresponding isothiazolinyli- dene complexes.91 b. Modification of amino, hydroxy, carboxy and other groups and functionalisation of the side chain in isothiazole derivatives In-depth studies into biological properties of functionalised iso- thiazoles carried out in the past decade were aimed at the develop- ment of methods for the synthesis of such compounds based on the known amino- and hydroxyisothiazoles, isothiazolecarboxylic acids and their chlorides, etc., and by functionalisation of the side chain.46 Heating of a 5-amino-4-chloro-3-methylisothiazole (108) ± (4-trifluoromethylphenyl)acetyl chloride (109) mixture in xylene gave N-isothiazol-5-ylamide 110.92 Me Cl 140 8C +F3CC6H4CH2COCl N H2N 109 S 108 Me Cl N F3CC6H4CH2C(O)HN S 110 A vast variety of novel useful N-isothiazol-5-ylamides have been prepared starting from structurally more complex acid chlorides and aminoisothiazoles containing various substituents in positions 3 and 4.92, 93 The methods for the synthesis of acylated 5-aminoisothiazoles of the type 111 and 112 have been developed.94 ± 96 R2 R3 O R3 R2 X O N N S R1 X N Y N R1 Z S 111 R4 112 R4 R1, R2, R3, R4=H, CN, NO2, CHO, alkyl, halogenoalkyl, alkoxyalkyl, alkoxy, alkylthio; X=alkylene, alkyleneoxy; Y, Z=O, S.For example, acylated 5-aminoisothiazole 113 was prepared from 4-chloro-5-[4-(4-cyanophenoxy)phenyl]acetylamino-3- methylisothiazole (114) by treatment with NaH in tetrahydro- furan and then with a twofold excess of allyloxycarbonyl chlor- ide.95 R V Kaberdin, V I Potkin Me Cl N HN 1) NaH, THF, 20 8C 2) ClCO2CH2CH CH2 3) H2O S C 4-NCC6H4OC6H4CH2 114 O O Me Cl C CHCH2O H2C N N S C 4-NCC6H4OC6H4CH2 113 O Carboxylic acids and acid chlorides of the isothiazole series are convenient starting compounds in the synthesis of isothi- azolecarboxamides.Thus a series of biologically active amides 116 have been synthesised from 3-methylisothiazole-5-carbonyl chlor- ide (115a) and 3,4-dichloroisothiazole-5-carbonyl chloride (115b) and various amines in the presence of triethylamine or pyri- dine.97 ± 101 R2 R1 Py or Et3N N +R3R4NH ClC(O) THF S 115a,b R2 R1 N R3R4NC(O) S 116 R1=H,R2=Me (115a); R1=R2=Cl (115b); R3=Ph, C6H5SO2, 4-ClC6H4, 1,2,4-triazolyl, pyrazolyl, piperazinyl; R4=H, Et. Amides 116 (R3=Het, R4=H) can be prepared by the reaction of 3,4-dichloroisothiazole-5-carboxamide with deriva- tives of the corresponding heterocylic compounds.101 5-Amino-3-methylisothiazole-4-carboxylic acid (117) is the starting compound in the synthesis of structurally more complex amides.Thus the reaction of the chloride 118 with 4-methylthio- aniline and subsequent acylation with acid chlorides yields 5-acyl- aminoisothiazole-4-carboxamides 119.102 Me Cl(O)C Me HO2C 1) 4-MeSC6H4NH2 2) RCOCl SOCl2 N N H2N H2N S 118 S 117 Me 4-MeSC6H4NH(O)C N RC(O)HN S 119 R=Me, Et, Ph. The reaction of 5-benzoylamino-3-methylisothiazole-4-car- boxylic acid (120) with p-chloroaniline in the presence of pyridine as the catalyst and alkoxycarbonyl chloride as the activating reagent gave 5-benzoylamino-3-methylisothiazole-4-carboxylic acid (121) p-chloroanilide.103 Me HO2C Py, ClCO2R +4-ClC6H4NH2 Me2CO EtOH N PhC(O)HN S 120Isothiazoles (1,2-thiazoles): synthesis, properties and applications Me 4-ClC6H4NH(O)C N PhC(O)HN S 121 (90%) R=Me, Et, Prn.3-Hydroxyisothiazole and its 4-halogeno derivatives react with acetals in the presence of toluene-p-sulfonic acid to give mixed acetals. For example, the reaction of isothiazole 77 with acetaldehyde dimethyl acetal results in 3-(1-methoxyethyloxy)iso- thiazole (122).104 Me OCH OH TsOH OMe +MeCH(OMe)2 N N 115 8C S 122 (23%) S 77 The reaction of 5-chloro-3-hydroxyisothiazole (123) with malonic dialdehyde bis(dimethylacetal) (124) in propionic acid yields a mixture of 5-chloro-3-(1,3,3-trimethoxypropyl)oxyiso- thiazole (125) and 1,3-bis(5-chloroisothiazol-3-yloxy)-1,3-dime- thoxypropane (126).105 OH EtCO2H N Cl +(MeO)2CHCH2CH(OMe)2 124 S 123 OCH(OMe)CH2CH(OMe)2 + N Cl S 125 OCH(OMe)CH2CH(OMe)O + N N Cl Cl 126 S S A great number of isothiazole derivatives of the general formula 127 have been prepared using a similar approach.OR O X1 Y Z(OR)2 N X2 127 S R=Alk(C17C18); X1=H, Br, Cl, Me; X2=H, Br, Cl; Y7Z=HC(CH2)nCH. 4,5-Disubstituted 3-hydroxyisothiazoles react with various allylic bromides to give alkenyl ethers of 3-hydroxyisothiazoles 128 ± 130.106 R5 O R1 R1 O O R4 N N R2 R2 R3 S S 129 128 O 681 O R1 N R2 Y S 130 R1, R2=H, Cl, alkyl(C1±C4); R3=H, alkyl, alkenyl, alkynyl, Ph, etc.; R4, various electron-withdrawing groups; R5=H, alkyl; Y=O, S, SO2.The condensation of isothiazole 77 and its 5-methylthio derivative 131 with arenesulfonyl chlorides 132 in the presence of a base yields the corresponding 3-arylsulfonyloxyisothiazoles 133.107 OH R2 + ClSO2 N R1 S 77, 131 132 R2 OSO2 N R1 133 S R1 = H (77), SMe (131); R2=H, 4-Me, 2-NO2, 4-NO2, 4-Br, 2-CO2Et. Aldehydes and ketones can be used for the functionalisation of isothiazoles. The reaction of the lithium derivative prepared by the metallation of 3-methyl-5-phenylisothiazole (134) with n-butyllithium in the presence of a lithium isopropyl- cyclohexylamide ±N,N,N0,N0-tetramethylenediamine system (LICA ±TMEDA) with carbonyl compounds leads to hydroxy derivatives 135.108 Me CH2CR1R2 OH BunLi N N +R1R2CO Ph Ph LICA7TMEDA S 135 S 134 R1=H,R2=Et, CH=CH2, Ph, CH=CHPh; R17R2 =(CH2)n, n=4, 5;R1=Ph; R2=Me, CH=CH2, Ph, CH=CHPh.The alkylation of a lithium derivative of isothiazole 134 with alkyl iodides or bromides gives compounds 136a ± c in up to 83% yields.109 Me CH2R 1) BunLi, LICA ±TMEDA 2) RX N N Ph Ph S 136a ± c S 134 R=Me, X = I (a); R=Bn, X=Br (b); R =Bu, X=Br (c). The condensation of 4-acetyl-5-chloro-3-methylisothiazole (137) with some aromatic and heterocyclic aldehydes involves the methyl group of the acetyl group rather than the methyl group in position 3 of the isothiazole ring to give a,b-unsaturated carbonyl compounds 138.110 Me RHC CH(O)C Me Ac B N N +RCHO Cl Cl S 138 S 137 R=2-ClC6H4, 4-ClC6H4, 4-Me2NC6H4, 2-furyl.682 (3-Chloro-4-cyanoisothiazol-5-yl)hydrazine (139) reacts with aromatic aldehydes to give the corresponding (3-chloro-4-cyano- isothiazol-5-yl)hydrazones 140.111 NC NC Cl Cl MeOH +ArCHO N N ArCH NHN H2NHN S 140 S 139 Ar=Ph, 4-O2NC6H4, 4-MeOC6H4, 4-ClC6H4, 2,4-Cl2C6H3, 3-O2NC6H4, 2,4-(O2N)2C6H3, 4-HOC6H4, 2-HOC6H4, 3-MeOC6H4, 2-MeOC6H4.Treatment of 4-cyano-3-hydroxy-5-methylthioisothiazole (141) with allyl bromide in the presence of potassium carbonate in DMF and subsequent treatment with iodine, potassium iodide and KOH yield the iodopropargyloxy derivative 142.112 Other cyanoisothiazole derivatives containing various alkyl, alkenyl and alkynyloxy(thio) groups in positions 3 and 5 were synthesised using a similar approach.112 OH NC 1) H2C=CHCH2Br, K2CO3, DMF 2) I2, KI, 10% KOH N MeS S 141 C I NC OCH2C N MeS S 142 (75%) 3,4-Dichloroisothiazole-5-carboxylic acid derivatives 143 (X=O, S; R is substituted alkyl, aryl and heterocyclic residues containing halogen, CN, NO2 , carbonyl, halogenalkylthio and other groups) have been synthesised recently from the acid chloride 115b.113 Cl Cl N C RX O S 143 4.Miscellaneous reactions for the synthesis of isothiazole The preparation of 3-substituted 4,5-diaminoisothiazoles used in the synthesis of novel heterocyclic systems (see Section III) is based on the conversions of 5-acetylamino-3-methyl-4-nitroiso- thiazole (144) which is readily obtainable from the commercially available 5-amino-3-methylisothiazole hydrochloride.114 Isothiazole 145 non-substituted in position 3 is formed upon oxidation of isothiazole 144 to carboxylic acid 146 and subsequent decarboxylation.Me O2N CrO3, H2SO4 N AcHN S 144 O2N CO2H O2N toluene, D N N AcHN AcHN S 145 S 146 4,5-Diaminoisothiazoles 147a,b are obtained from com- pounds 144 and 145 by deacetylation and reduction of the nitro group in the derivatives 148a,b.114 R V Kaberdin, V I Potkin R O2N NH3 N MeOH AcHN S 144, 145 R R H2N O2N Fe, HCl H2O, EtOH, D N N H2N H2N S 147a,b S 148a,b R=Me (a), H (b).5-(N-Alkylamino)-4-nitroisothiazoles 149a ± c were synthes- ised in high yields by the alkylation of sodium salts of 5-amino-4- nitroisothiazoles 148. The reduction of the nitro group in com- pounds 149a ± c gives 5-(N-alkylamino)-4-aminoisothiazoles 150a ± c.115 R1 O2N 1) NaH 2) R2X N H2N S 148 R1 R1 H2N O2N Fe HCl, EtOH, D N N R2HN R2HN S 149a ± c S 150a ± c R1=R2=Me (a); R1=H,R2=Me (b); R1=Me, R2=Bun (c). 4,5-Diamino-3-cyanoisothiazole (151) was obtained from 5-acetylamino-4-nitroisothiazole-3-carboxylic acid (146) in four steps which included the conversion of the carboxy group into the nitrile group and the reduction of the nitro derivative to diamine 151.115 CO2H O2N 1) SOCl2, D 2) NH3, MeOH N AcHN S 146 CONH2 O2N POCl3 85 8C N H2N S CN CN H2N O2N Fe HCl, EtOH, D N N H2N H2N S S 151 3-Acetamido-4-amino-5-(N-methylamino)isothiazole (152) is the starting compound in the synthesis of novel annelated hetero- cycles. This is prepared from ethyl 3-amino-5-(N-methylamino)- isothiazole-4-carboxylate (153) which is converted into the nitro derivative 154 in several steps. Selective removal of the acetyl protective group from the amino group in position 5 by treatment with methanolic ammonia and subsequent reduction of the 3-acetamido-5-(N-methylamino)-4-nitroisothiazole (155) formed completes the synthetic scheme.115, 116 NH2 HO2C NH2 EtO2C 1) KOH, D 2) HCl 1) H2O, D 2) Ac2O N N MeHN MeHN S S 153Isothiazoles (1,2-thiazoles): synthesis, properties and applications NHAc HNO3, TsOH CH2Cl2 N AcN S Me NHAc O2N NH3, MeOH N 0 8C AcNMe S 154 NHAc NHAc H2N O2N Fe N N HCl, EtOH, D MeHN MeHN S 152 S 155 4-Dibromoaminoisothiazoles 156a ± c used in the synthesis of the bisisothiazolopyrazine system are prepared by the reaction of with 157a ± c 4-aminoisothiazoles dibromoisocyanuric acid.117 ± 119 Br R2 R2 Br2N H2N N N Br N O N N R1 R1 S 156a ± c S 157a ± c R1=Br: R2=Me (a), Ph (b); R1=R2=Cl (c).A procedure for the synthesis of 3,4-dichloro-5-cyanoisothia- zole by the chlorination of the reaction product of carbon disulfide with NaCN in DMF120 has been improved.121 3,4-Dichloro-5- cyanoisothiazole synthesised (yield *88%) was converted into 3,4-dichloroisothiazole-5-carboxylic acid, which is employed in the synthesis of efficient plant growth regulators.Analogously, 3,5-dichloroisothiazole-4-carboxylic acid was prepared from 3,5- dichloro-4-cyanoisothiazole.122 4,5-Dichloroisothiazole-3-carboxylic acid (158) was prepared by hydrolysis of the trichloromethyl group of 4,5-dichloro-3- trichloromethylisothiazole (18) with fuming nitric acid and sub- sequent treatment of the reaction mixture with water.56 Cl Cl CO2H CCl3 1) HNO3 2) H2O N N Cl Cl S 158 (92%) S 18 The direction of reduction of the isothiazole 18 depends on the nature of the reducing agent.Boiling with zinc powder in ethanol results in the partial reduction of the trichloromethyl group and the formation of 4,5-dichloro-3-dichloromethylisothiazole (159). When tin dichloride is used, isothiazole 18 undergoes dechlorodi- merisation resulting in 1,2-bis(4,5-dichloroisothiazol-3-yl)tetra- chloroethane (160) in high yield.56 Cl CHCl2 Zn EtOH N Cl CCl3 Cl S 159 (53%) N Cl Cl CCl2 S 18 SnCl2 N Cl 2 S 160 The oxidation of 3-alkoxy-5-alkyl-4-cyanothioisothiazoles 161 under the action of Oxone (2KHSO3 .KHSO4 .K2SO4) 683 involves selectively the exocylic sulfur atom and affords the corresponding 5-alkylsulfonylisothiazoles 162 in high yields.123, 124 NC NC OR2 OR2 Oxone H2SO4 N N R1S R1O2S S 162 S 161 R1, R2=H, C17C12-alkyl, C27C7-alkenyl, C37C7-alkynyl, C37C7-cycloalkyl, Ph, PhO.The oxidation of the methylene group in substituted acet- amide 163 by dimethylformamide dimethylacetal (164), which can be regarded as a variant of the condensation reaction associated with the elimination of two MeOH molecules, gives enamino ketone 165.125 Me Cl O Me2NCH(OMe)2 164 N C 4-F3CC6H4O CH2 S NH 163 Me Cl O N C C 4-F3CC6H4O S NH CHNMe2 165 The chlorination of 5-[2-(N-tert-butyldimethylsilylamino)vi- nyl]isothiazoles 166 with N-chlorosuccinimide (NCS) proceeds selectively at the vinyl group resulting in the Z-isomers of isothiazole 167.126 R R R R NCS N N ButMe2SiHN ButMe2SiHN S 166 S 167 Cl R=4-ClC6H4, Ph.The monochlorination of related compounds with N-chloro- succinimide also proceeds in a selective manner. Isothiazoles 168 and 169 react with AgO3SCF3 to give dimeric silver(I) complexes the isothiazole fragment in which is linked to the metal centre through the nitrogen atom.127 Me Me N N N CH R(CH2)n (CH2)nN S 168 S 169 n=1, 2. R=[Et2N(CH2)2]2N, n=1; R=MeS, n=2. Treatment of selenolo[2,3-d ]isothiazole (170) with methyl- lithium yields a mixture of compounds including isothiazole 171.128 S S N CH MeSeHC MeLi N Se 171 170 Treatment of 5-substituted 4,5-dihydro-3-ethoxycarbonyliso- thiazolo[4,3-d ]pyrimidin-7(6H)-ones with hydrazine is accompa- nied by the opening of the dihydropyrimidinone ring and the formation of 4-amino-3-carbamoyl-5-hydrazinocarbonylisothia- zole (172).129684 CONH2 H2N N H2NNHCO S 172 It is of note that the ring opening reactions in bicylic isothiazole derivatives described above present purely theoretical interest.The reactions of 3,7-dichlorodiisothiazolo[4,5-b : 4050-e]pyr- azine (173) with benzylamine, aniline and morpholine are accom- panied by the opening of the pyrazine ring to give N,N0-bis- (5-amino-3-chloroisothiazol-4-yl)diazenes (174a ± c). The reac- tion of compound 173 with benzylamine gives 3,7-bis(benzyl- amino)bisisothiazolo[4,5-b : 4050-e]pyrazine (yield 5%) as a side product.130 Cl NR1R2 Cl N S HNR1R2 N S N N N N S N S Cl N 173 Cl NR1R2 174a ± c R1=H,R2=Bn (a, 25%); R1=H,R2=Ph (b, 25%); R17R2=(CH2)2O(CH2)2 (c, 28%).A mechanistic scheme for the synthesis of compounds 174a ± c has been proposed. Its first step consists in the addition of the amine at the C=N bond of the pyrazine ring. Interestingly, the reaction of compound 173 with MeONa in MeOH results in 5,6- dimethoxy-3-chloroisothiazolo[4,5-b]pyrazine, i.e., in the opening of one of the isothiazole rings. The difference in the behaviour of compound 173 in the reaction with the alkoxide anion and amines is explained by the difference in their basicities as well as by the presence of a mobile hydrogen atom in the amines which ensures their addition at the multiple bonds of the pyrazine ring.130 5. Synthesis of isothiazole 1,1-dioxides Methods for the synthesis, reactions and biological properties of isothiazole 1,1-dioxides were reviewed in 1997.131 Here we con- sider only the most recent publications.Thus a novel procedure for the synthesis of 4-amino-3,3-dimethyl-2,3-dihydroisothiazole 1,1-dioxide derivatives 175 based on substituted N-benzyl-N- sulfonylaminonitriles 176 has been developed recently. The cycli- sation was performed by treatment with sodium hydride in acetonitrile.132 Me H2N Me2C N SO2CH2R NaH X MeCN MeCH2 NC CH2C6H4X N R S 176 O O 175 (43% ± 75%) R=H, Me, Ph; X=H, 4-Cl, 3-Cl. Other 4-amino-2,3-dihydroisothiazole 1,1-dioxide deriva- tives,133 particularly compounds containing a cyano group in position 5, were prepared using analogous methods.134 ± 136 In the presence of bases, phenylmethanesulfonamide (177) reacts with dimethyl oxalate to give 4-hydroxy-5-phenylisothia- zol-3(2H)-one 1,1-dioxide (178).It is assumed that the linear methyl N-(benzylsulfonyl)oxalamidate 179, which is formed first, undergoes further cyclisation. A complex of dioxide 178 with DMF has been prepared.137 O ButONa, ButOH PhCH2SNH2+MeO2CCO2Me 7MeOH 177 O R V Kaberdin, V I Potkin O HO PhCH2SO2NHCCO2Me NH Ph O S 179 O O178 (55%) 6. Synthesis of dihydroisothiazolones Among representatives of this group, dihydroisothiazol-3-ones possessing high biological activities hold the greatest interest. The data on these compounds are summarised in the review.138 Some novel data on the synthesis of dihydroisothiazolones have been published recently.Thus the chlorination of 2-(N- methylcarbamoyl)ethanethiol (180) or bis[2-(N-methylcarb- amoyl)ethyl] disulfide (181) gave a mixture of 2,3-dihydro-2- methylisothiazol-3-one (182) and 5-chloro-2,3-dihydro-2-methyl- isothiazol-3-one (183).139 O O HSCH2CH2CONHMe Cl2 180 NMe NMe + Cl 5 ±20 8C (SCH2CH2CONHMe)2 S 183 S 182 181 Isothiazol-3-ones 182 and 183 can also be prepared by treat- ment of compounds 180 and 181 with thionyl chloride.140, 141 5-Chloro-2,3-dihydro-2-octylisothiazol-3-one (184) was syn- thesised by treatment of disulfide 185 with SO2Cl2.142 O SO2Cl2 NC8H17 Cl (C8H17NHCOCH2CH2S)2 185 S 184 The clathrate of compound 184 containing four molecules of 4,40-dihydroxydiphenylmethane manifests increased solubility in water. A procedure for the synthesis of isothiazol-3-one derivatives 187 from 3-alkylthio-2-cyano-3-mercaptoacrylamides 186 has been developed.143 O NC SR2 NC Cl2 or SO2Cl2 C C NR1 R2S R1NH(O)C SH S 187 186 R1, R2=H, lower alkyl, alkenyl, alkynyl, aryl, aralkyl, hetarylalkyl.The synthesis of isothiazol-3-ones devoid of highly toxic and carcinogenic nitrosoamines has been described in a patent. 144 III. Chemical transformations of isothiazoles Isothiazoles manifest aromatic properties and enter into the same reactions as other heteroaromatic compounds. Ring opening and transformations of the isothiazole ring, which give either func- tionalised alkenes or other heterocyclic compounds, deserve special attention.Addition reactions of isothiazoles leading to novel bisheterocyclic structures containing an isothiazole frag- ment are especially attractive. Oxidation reactions of the sulfur atom of the isothiazole ring yield reactive dioxides and present significant interest.64 1. Condensation reactions In recent years, isothiazole derivatives were used as starting compounds for the synthesis of several novel bicyclic compounds. Thus the reaction of 4,5-diamino-3-methylisothiazole (147a) with diethoxymethyl acetate affords 3-methylimidazo[4,5-d ]- isothiazole (188a) in low yield. Unsubstituted imidazo[4,5-d ]iso-Isothiazoles (1,2-thiazoles): synthesis, properties and applications thiazole could not be prepared from compound 147b.114 Isothia- zoles 147a,b easily undergo cyclisation with thiocarbonyldiimida- zole (189) to give unstable thiones 190a,b which are methylated in situ in alkaline media under the action of Me2SO4 to give 5-methylthioimidazo[4,5-d ]isothiazoles (191a,b) in good yields.Sodium salts of thiones 190a,b are also alkylated by benzyl and allyl bromides to give the corresponding imidazo[4,5-d ]iso- thiazoles 191c ± f in preparative yields.114, 115 Me H2N N H2N S 147a R1 H2N N H2N S 147a,bHN S NH 190a,b R1=R2=Me (191a); R1=H,R2=Me (191b); R1=Me, R2=Bn (191c) ; R1=Me, R2=All (191d); R1=H,R2 =Bn (191e); R1=H,R2=All (191f). 6-Alkyl-5-methylthioimidazo[4,5-d ]isothiazoles 192a ± c were prepared by treatment of 5-(N-alkylamino)-4-aminoisothiazoles 150a ± c with thiocarbonyldiimidazole (189) and subsequent methylation.The reactions of isothiazoles 150a,c with dieth- oxymethyl acetate result in compounds 193a,c.115 H2N R2HN S 150a ± c (EtO)2CHOAc 150a,c MeOCH2CH2OH R1=R2=Me (a); R1=H,R2=Me (b); R1=Me, R2=Bun (c). 3-Amino-6-methyl-5-methylthioimidazo[4,5-d ]isothiazole (192d) was obtained in good yield by successive treatment of 3-acetamido-4-amino-5-(N-methylamino)isothiazole (152) with thiocarbonyldiimidazole (189) and dimethyl sulfate.115 NHAc H2N N MeHN S 152 N MeS Me N Me N (EtO)2CHOAc N MeOCH2CH2OH S HN 188a N N C NH HN S 189 THF R1 R1 N 1) NaOH 2) Me2SO4 or R2Br N R2S N S S NH 191a ± f R1 R1 N 1) 189 2) NaOH 3) Me2SO4 N MeS N S RN2 192a ± c R1 N N S RN2 193a,c 1) 189, dioxane 2) NaOH 3) Me2SO4 NHAc NH2 N NH3 N MeS N MeOH, 100 8C S S Me N 192d 685 It should be noted, however, that not all 4,5-diamino-3-R- isothiazoles yield annelated products.Thus isothiazole 151 con- taining a cyano group in position 3 does not react with either diethoxymethyl acetate or thiocarbonyldiimidazole.115 Two approaches to the synthesis of 3,7-disubstituted bisiso- thiazolo[4,5-b:40,50-e]pyrazines 193a ± c, the first compounds con- taining a novel heterocyclic system, have been proposed.117 ± 119 It was found that the reaction of 3,5-dichloro-4-dibromoaminoiso- thiazole (156c) with a Cu(0) ± collidine system affords 3,7-di- chlorodiisothiazolo[4,5-b:40,50-e]pyrazine (193c) in 67% yield.117 Zlotin et al.118 suggested that this reaction has a radical character and studied the effect of UV light on isothiazoles 156a ± c. Irradiation of compounds 156a,c with a mercury lamp results in the cleavage of the N7Br bond and the formation of radicals 194a,c; their further transformations depend on the nature of the solvent used. Thus in dichloromethane, isothiazole 156c is con- verted into 4-amino-3,5-dichloroisothiazole virtually completely (157c). In CCl4, 3,7-disubstituted diisothiazolo[4,5-b:40,50-e]pyr- azines 193a,c (yields 58% and 76%, respectively) are the main reaction products of compounds 156a,c; their formation seems to proceed via the intermediate compounds 195a,c.N,N0-Di(iso- thiazol-4-yl)diazenes 196a,c (yields 37% and 13%, respectively) were identified as the side products.118 BrN R2 R2 Br2N hn N N 7Br R1 R1 S 194a ± c S 156a ± c Cl H2N R1=R2=Cl CH2Cl2 N Cl S 157c N N R2 R2 CCl4 N N R1 R1 7Br2 S S 196a ± c R2 R2 Br2N N CCl4, Br. N S N 7[R1] S Br R1 195a ± c 7Br, [7R1] ,7Br2 hn R2 N S N N S R2 N 193a ± c R1=Br, R2=Me (a), Ph (b); R1=R2=Cl (c). 5-Bromo-4-dibromoamino-3-phenylisothiazole (156b) mani- fests a similar behaviour upon irradiation with UV light; 3,7- diphenyldiisothiazolo[4,5-b : 40,50-e]pyrazine (193b) and N,N0- bis(5-bromo-3-phenylisothiazol-4-yl)diazene (196b) are the main reaction products.119 The isothiazolium salt 197 is condensed with cyclopenta- dienyl-, tert-butylcyclopentadienyl- and di-tert-butylcyclopenta- dienyllithium to give thialenes 198.145686 R2 MeNH R3 R1 Cl NC Li NC R1 + R2 NMe Ph S Ph FSO¡3R3 S 197 198 R1=R2=R3=H;R1=But, R2=R3=H; R1=R3=H,R2=But; R1=R3=But, R2=H.Reactivities of certain isothiazole 1,1-dioxide derivatives have been studied.146 It was found that 3-diethylamino-4-(4-meth- oxyphenyl)isothiazole 1,1-dioxide (199) reacts with sodium azide in ethanol to give a mixture of dihydroisothiazole oxides 200 and 201 and isothiazolotriazoles 202 and 203 in a ratio 72 : 12 : 9 : 7.146 NEt2 4-MeOC6H4 NaN3 N EtOH S O O 199 EtO EtO SO2 N SO2 N + +4-MeOC6H4 4-MeOC6H4 NEt2 NEt2 201 200 NEt2 4-MeOH4C6 N N N H N HN SO2 H SO2 N N + + O2S C6H4OMe-4 N N 4-MeOH4C6 NEt2 202 203 NEt2 Cycloaddition of the dioxide 199 to arylalkyl and aryl azides results inN-arylalkyl- orN-arylthiadiazabicyclo[3.1.0]hexene 204.The unstable cycloadducts 205, which are formed first (in some cases, they were isolated from the reaction mixture), easily eliminate the nitrogen molecule. H H SO2N RN RN3 SO2 N 199 RN N 7N2 NEt2 4-MeOH4C6 NEt2 204a ± e N 4-MeOH4C6 205 R=Ph (a), Bn (b), PhCH2CH2 (c), 4-MeOC6H4 (d), 4-O2NC6H4 (e). Thermal rearrangement of N-aryl- and N-phenyl-substituted bicyclic compounds 204a,c ± e affords 1,2-thiazete 1,1-dioxides 206a,c ± e, 1,2,6-thiadiazine 1,1-dioxides 207a,c ± e and pyrazole derivatives 208a,c ± e.Under optimum conditions, compound 207a was isolated in 51% yield. If the aziridine nitrogen atom has a benzyl substituent (com- pound 204b), the reaction follows a different route to afford the thiadiazine derivative 207b and pyrimidine 209.147 NR H RN SO2 4-MeOH4C6 SO2 N 4-MeOH4C6 N NEt2 Et2N 207a ± e 206a,c ± e R V Kaberdin, V I Potkin Ph N NR N N 4-MeOH4C6 4-MeOH4C6 209 208a,c ± e NEt2 NEt2 R=Ph (a), Bn (b), PhCH2CH2 (c), 4-MeOC6H4 (d), 4-O2NC6H4 (e). 4-Aryl-3-diethylamino-5-R-isothiazole 1,1-dioxides enter into cycloaddition reactions with diazoalkanes to give bicyclic com- pounds 210 with high regio- and stereoselectivity.The reaction with diazomethane yields a mixture of tautomeric pyrazolines 210 (R1=R2=H) and 211. Thermolysis of compounds 210 and 211 follows two routes. The loss of the nitrogen molecule leads to 2-thia-3-azabicyclo[3.1.0]hex-3-ene 2,2-dioxides 212. Elimination of SO2 and diethylcyanamide from the cycloaddition products results in the corresponding pyrazoles.148 R2 R3 R3 R3 R1 R1 SO2 N SO2 N SO2 N N N R2 N HN Ar Ar Ar NEt2 NEt2 NEt2 212 211 210 R1=R2=H, Me; R1=H,R2=CO2Et; R3=H, Me, Ph; Ar=Ph, 4-MeC6H4, 4-MeOC6H4. The readily available dioxide 199 enters into cyclocondensa- tion reactions with oxazolones 213a ± e and munchnones 214a ± h, which are prepared by cyclisation of (N-aroylamino)arylacetic acids 215a,b.149 The reactions with compounds 215a afford 4,6-diaryl-3-dieth- ylamino-3a,4-dihydro-3a-(4-methoxyphenyl)-6aH-pyrrolo[3,4-d ]- isothiazole 1,1-dioxides (216a ± e) (yields 60%± 74%) and their isomers 217a ± d (yields 5%± 13%).4,6-Diaryl-3-diethylamino- 3a-(4-methoxyphenyl)-5-methyl-3a-4-dihydropyrrolo[3,4-d ]iso- thiazole 1,1-dioxides 218a ± h are the main products in the reaction with munchnones 214a ± h.149 HO O O Ac2O toluene, 80 8C Ar2 Ar1 HN 215a SO2 N O 4-MeOH4C6 O 199 NEt2 7 Ar2 Ar1 HN + 213a ± e Ar2 H H Ar2 H SO2 N SO2 N N N + Ar1 NEt2 NEt2 Ar1 4-MeOH4C6 217a ± d H 4-MeOH4C6 216a ± e Ar2 Ar1 Compounds 216, 217 Ph Ph 4-ClC6H4 Ph 4-MeC6H4 Ph 4-ClC6H4 Ph 4-MeC6H4 Ph abcde687 Isothiazoles (1,2-thiazoles): synthesis, properties and applications HO O O Ac2O triethylamine result in the elimination of sulfur and recyclisation resulting in pyrrole-2-carboxylic acid derivatives 223.151 toluene, 80 8C Ar2 Ar1 R4 R3 R3 R2 Me N Et3N 215b X7 SO2 7S 7 R2 R1 R4 N R1H2C O 4-MeOH4C6 HN O N+S 222 199 NEt2 223 Ar2 Ar1 Me +N R1=CO2Me, CO2Et, CN; R2=H, Me, Ph; R3=H, Ph, 2-MeC6H4, 4-FC6H4, CO2Et; 214a ± h NMe; X=Cl, Br.R4 =Ph, N O, NHPh, SMe, 4-ClC6H4, N Ar2 SO2 N MeN Aminopyrroles 224 were obtained from isothiazolium salts 225 in a similar way (yields 70%± 90%).152 R4 Ar1 NEt2 N R3 R3 R2 R5 B X7 R4 H 4-MeOH4C6 218a ± h (63% ± 91%) R2 R1 N R1H2C Ar2 Ar2 Ar1 NH R5 N+S 225 224 Com- pound 218 Com- Ar1 pound 218 R1=H, alkenyl, aryl, NO2; R2, R3=H, substituted or unsubstituted alkyl, Ar, HetAr; R4, R5=H, heterosubstituted aryl or alkyl; X=Hal, ClO¡4 , BF¡4 , HSO¡4 , SO24 ¡, OH7, CF3SO¡3 .Ph 4-MeC6H4 Ph efg 4-MeOC6H4 h abcd Ph Ph 4-MeC6H4 Ph 4-MeOC6H4 Ph Ph Ph Ph 4-FC6H4 4-BrC6H4 4-NO2C6H4 The mechanism of formation of 3-aminopyrrole derivatives in the desulfurisation of isothiazolium salts has been studied.153 It was found that pyrrole 226 is formed from the salt 227 by a cascade of reactions via thioamide 228.153 C6H4Cl-4 + The cycloadducts 216 and 218 are decomposed at elevated temperatures or in alkaline media with the elimination of SO2 and diethylcyanamide to give 2,3,5-triarylpyrroles 219 and 1-methyl- 2,3,5-triarylpyrroles 220, respectively.149 4-MeOC6H4 Br7 H2N(O)CCH2N N S DBU 216 O 227 Ar2 Ar1 HN C6H4Cl-4 O N 4-ClH4C6 219 C6H4OMe-4 H2N(O)CH N N CONH2 S 180 ± 220 8C 218 HN O Ar1 Ar2 226 228 Me N 220 DBU, diazabicycloundecene.A convenient procedure for the synthesis of 2-phenylthiocar- bonylmethylidenedihydrothiazole derivatives 229 from isothiazo- lium chlorides has been developed.154 R1 R2 R2 S NaBH4 PhC(S)HC 4-Amino-3-carbamoyl-5-(ethoxycarbonyl)isothiazole is used in the synthesis of the isothiazolo[4,5-d ]pyrimidine derivatives 221.150 NH2(O)C CHCl3, EtOH +N Ph R3 N S R3 229 HN Cl7 N N N CH C6H4R S R1=Ac, EtCO; R2=CO2Et, CO2Me; R3=Me, Et.221 R=2-OMe, 2-Cl, 3-Br, 3-NO2. 2. Ring transformation reactions N-Arylisothiazolium salts 230 containing an activated methyl group in position 5 dimerise into thiadiazapentalenes 231 upon treatment with a base. The sulfur atom of one molecule of the salt is attacked by an activated methyl group of another molecule to give the intermediates 232. Subsequent oxidative closure of the ring via zwitterionic intermediates 233 results in thienothiadiaza- pentalenes 231. A competing reaction gives 5-(2-thienyl)isothia- zolium salts 234 with elimination of aniline.155 In recent years, some readily available isothiazolium salts have been used in the synthesis of other heterocyclic compounds, pyrrole derivatives, in particular.Thus the reactions of N-alkoxy- carbonylmethyl- and N-cyanomethylisothiazolium 222 salts with688 Me B +N Me Ar S X7 230 Ar N7 Me S Me Me 233 Me 72H Me Me 7ArNH2 Me Ar=Ph, 4-MeC6H4; X = Cl7, ClO¡4 ; B=(cyclo-C6H11)2NH. The reaction between the salts 230 and 235 catalysed by dicyclohexylamine in methanol or DMSO yields `mixed' thiadia- zapentalenes 236 upon oxidative cyclisation and salts 237 upon elimination of aniline.156, 157 Me + +N Me S Ar1 ClO¡4230 Ar2 N S S 236 Ar1=Ph, 4-MeC6H4; Ar2=Ph, 4-MeC6H4, 4-MeOC6H4. Photochemical reactions of several isothiazoles have been studied. Irradiation of 4-phenylisothiazole in benzene yielded small amounts of 4-phenylthiazole (238).When the photolysis was carried out in the presence of triethylamine, this rearrange- ment becomes predominant.158 Ph Ph hn N Et3N S H The effect of triethylamine and the solvent polarity on the photochemical reaction of 5-phenylisothiazole (239) has been studied.159 It was found that irradiation in the absence of triethyl- amine gives 4-phenylthiazole (238) as the main product (yield 15%). Other reaction products are represented by 3-phenyliso- thiazole (240) and 2-phenylthiazole (241). In the presence of Et3N, 5-phenylthiazole (242) is the main product (yield 14%). Ar N Me S Me S X7 +N Ar 232 Me Ar S +N Ar Ar N N S S Me 231 Ar S +N X7 S 234 Me (cyclo-C6H11)2NH N+ Ar2 S ClO¡4235 Ar1 Ar1 ClO¡4S +N N + Me S Me 237 Ph C7 N +N S SH 238 R V Kaberdin, V I Potkin Irradiation of 5-phenylisothiazole (239) in polar solvents makes this rearrangement more regioselective. The photolysis in methanol in the presence of Et3Ngives exclusively compounds 238 and 242 (yields 9% and 34%, respectively).In a more polar solvent (e.g., 2,2,2-trifluoroethanol), only thiazole 241 was iden- tified in the reaction mixture; in the presence and in the absence of Et3N, its yield was 42% and 32%, respectively.159 Ph Ph N N hn PhH + N + Ph S S S 241 240 238 N Ph N S 239 hn PhH, Et3N +238+240 Ph S 242 3. Ring opening reactions Under conditions of phase-transfer catalysis [PhH, H2O, tris(2,6- dioxaheptyl)amine], the reaction of 3,5-dimethylisothiazole (243) with acetylcobalttetracarbonyl generated in situ from Co2(CO)8 , CO and MeI is accompanied by ring opening and N-acylation, resulting in thione (244) (the ratio of the E- and Z-isomers is equal to 1 : 10).160 Me AcCo(CO)4 C(Me)NHCOMe (Me)C(S)CH N Me 244 (61%) S 243 Other N-acylated unsaturated thiones are synthesised from isothiazoles in a similar way.A convenient procedure has been developed for the synthesis of phenylthiocarbonylketene S,N-acetals (246) based on the reaction of 2-alkyl-3-alkylthio-5-phenylisothiazolium salts (e.g., the salts 245) with sodium borohydride.161 SEt SEt NaBH4 C PhC(S)HC Ph Me NHMe 246 I7 +N S 245 The reaction of 3-diethylamino-4-(4-methoxyphenyl)isothia- zole 1,1-dioxide (199) with organomagnesium compounds inTHF occurs via ring opening leading to a mixture of E- and Z-isomers of 3-substituted 2-arylpropeneamidines 247.162 NEt2 4-MeOC6H4 NH R 1) RMgBr, THF 2) H2O N C C C S H O O NEt2 C6H4OMe-4 247 199 R=Me, Et, Ph, HC C, MeC C, PhC C.The chemistry of the 4-nitroisothiazole-5(2H)-imine deriva- tives 248, particularly, their isomerisation and desulfurisation, has been discussed in the papers 163, 164 and a recent review.165 NO2 NAr ArN S 248 It has already been noted (Section III.2) that the photolysis of 4-phenylisothiazole in benzene gives small amounts of 4-phenyl- thiazole (238) as a result of a rearrangement; cyanothiol 249,689 Isothiazoles (1,2-thiazoles): synthesis, properties and applications which was trapped as benzyl sulfide 250, is the main photolysis product in the absence of a base.158 Ph reaction with 2,4,6-triethylbenzonitrile oxide to give the dihydro- isoxazole derivative 260.169 NEt2 4-MeOC6H4 CN Ph CN Ph 2,4,6-Et3C6H2CNO BnBr hn, PhH NH N C H2C S S SBn H SH H O 250 249 EtO O259 NEt2 4-MeOC6H4 4.Miscellaneous reactions of isothiazole derivatives The oxidation of 4,5-disubstituted isothiazoles with hydrogen N O NH S O 2,4,6-Et3C6H2 peroxide results in the corresponding isothiazol-3(2H)-one 1,1- dioxides 251.166 R1 O R1 H2O2 N NH AcOH R2 R2 S S O O251 OEt O 260 2,6-Dichlorobenzonitrile oxide reacts with isothiazolone (261a) at the double bond to give isoxazolecarboxanilide (262) as a result of transformation of the primary cycloadducts formed.Contrary to expectations, 2,4,6-triethylbenzonitrile oxide adds to the exocyclic carbonyl group of isothiazolone 261b to give the monoadduct 263 and the bisadduct 264.170 R1=Me, Ph, H; R2=Me, Ph, 4-BrC6H4, 4-MeOC6H4. PhHN(O)C O C6H3Cl2-2,6 Dioxides 253, 254 are the oxidation products of the isothiazo- lium salts 252.167 N NPh+2,6-Cl2C6H3CNO R R O Me R O 262 S 261a X7 OOH O N N Ar Me Ar Ar S S 2,4,6-Et3C6H2CNO O O NPh O O +N S 252 PhC(O) 254 253 S 261b R=Alk, Ph. 2,4,6-Et3C6H2 O O N Ph O NPh O NPh + N S O S Ph O 5-Bromo-3-diethylamino-4-(4-methoxyphenyl)-2,3-dihydro- isothiazole 1,1-dioxide (255) reacts smoothly with various organo- tin compounds in the presence of palladium catalysts to yield the corresponding 5-substituted derivatives 256.168NEt2 4-MeOC6H4 NEt2 4-MeOC6H4 O N 263 2,4,6-Et3C6H2 R4Sn NH NH 264 C6H2Et3-2,4,6 R Br S S O O O O 256 255 R=CH2=CH, Ph, HetAr, Ar. An efficient approach based on the Michaelis reaction is used in the synthesis of 5-substituted 3-amino-4-arylisothiazole 1,1- dioxides and their 4,5-dihydro derivatives.The addition of thiols, alcohols and trifluoroacetamide to 3-diethylamino-4-(4-meth- oxyphenyl)isothiazole 1,1-dioxide (199) in the presence of bases NEt2 4-MeOC6H4 R1SH N The reaction of the vinyl derivative 256 with nitrile oxides yields the cycloadducts 257a ± e, which are rearranged into 4,5- dihydro-5-(isoxazol-5-yl)isothiazole 1,1-dioxides 258a ± e (an equimolar mixture of cis- and trans-isomers) upon heating in DMSO.169 R1S S NEt2 4-MeOC6H4 O O ArCNO 265 R1=Me, Ph, cyclo-C6H11.NH CH H2C S NEt2 4-MeOC6H4 NEt2 4-MeOC6H4 O R2OH O256 N R2ONa N S R2O NEt2 NEt2 4-MeOC6H4 4-MeOC6H4 S O O Ar Ar O O 199 D, DMSO N NH 266 S S N N R2=Me, Et, Pri. O O O O O 258a ± e O257a ± e NEt2 4-MeOC6H4 N CF3CONH2 K2CO3, MeCN TEBA7Cl CF3COHN Ar=3,5-Cl2-2,4,6-Et3C6 (a), 2,6-Cl2C6H3 (b), 4-ClC6H4 (c), 2,4,6-Et3C6H2 (d), 4-NO2C6H4 (e). S O O TEBA± Cl, triethylbenzylammonium chloride. 267 3-Diethylamino-2,3-dihydro-5-(1-ethoxyvinyl)-4-(4-methoxy- phenyl)isothiazole 1,1-dioxide (259) enters into the cycloaddition690 occurs regiospecifically and yields a mixture of (4S,5S)- and (4R,5S)-diastereomers of 5-substituted 4,5-dihydroisothiazole 1,1-dioxides 265 ± 267.171 The reaction of the 5-bromo-derivative 255 with thiols or amines gives the substituted products of the bromine atom 268 and 269.168 NEt2 4-MeOC6H4 R1SH B NH R1S S O O 268 4-MeOC6H4 NEt2 R1=H, Me, Ph, Bn, NH Br , 4-MeC6H4, S O O N farnesyl 255 NEt2 4-MeOC6H4 R1R2NH NH R1R2N S O O 269 R1=4-MeC6H4, R2=H; R1=R2=Me.IV. Areas of application of isothiazoles As noted above, isothiazole derivatives manifest a broad spectrum of useful properties.Back in 1967, isothiazoles were used as starting compounds in the synthesis of highly efficient penicillins and cephalosporins,2, 3 which gave a strong impetus to systematic studies of the biological properties of isothiazole derivatives. In 1988, Mel'nikov who studied the production and application of pesticides and plant growth regulators drew attention to the broad potentialities of isothiazole derivatives.172 Recent years have been marked by numerous publications (including patents and patent applications) devoted to various applications of isothiazoles as efficient agrochemical and medicinal agents. It was found that 4-benzoyl- and 4-(hetaroyl)isothiazole derivatives manifest herbicidal activities.85 ± 88, 173 There is evi- dence of the use of substituted isothiazole-5-carboxamides 270,174 isothiazolylpyridines 271 175 and dihydroisothiazolone derivatives as herbicides.176 R1 R2 N Z R2 N N S N R1 O S 270 271 R1=Hal, MeO, FCH2O, F3CO; R2=CN, CO2H, CHO.R1=H, (un)substituted alkyl, (un)substituted acyl; R2=(un)substituted aryl, hetaryl; Z=(un)substituted C1±C4-alkylene. When used as herbicides, isothiazole derivatives manifest synergistic effects and are included in various compositions. Thus compounds 272 are auxin transport inhibitors and are used together with other herbicides in order to enhance their action.10 ± 14 C(Me) NNHCONHC6H3-5R1-3R2 MO2C N 272 S R1, R2=H, F, Cl; M=H, Na. R V Kaberdin, V I Potkin The role of herbicides in these compositions is played by different compounds, e.g., dicamba, glifosat, 2,4-D10 and herbi- cides based on phenoxypropionic 11 and phenoxybutyric acids.14 Some isothiazoles manifest high insecticidal activities.These include 4,5-dihydro-1H-pyrazole derivatives containing various isothiazole substituents in position 4,177, 178 1,2,3-triazole deriva- tives containing isothiazole substituents in position 3 ,179 etc. The cyanoisothiazole derivatives 161 manifest high insecticidal activ- ities against termites.180 Many isothiazole derivatives, particularly acylated 5-amino- isothiazoles,94, 95 isothiazolecarboxamides,96 ± 98 4-cyanoisothia- zoles,112, 123, 124 3,4-dichloroisothiazole-5-carboxylic acid and its derivatives,121 etc., manifest fungicidal activities.(3-Methyliso- thiazol-5-yl)(2,6-dinitro-4-trifluoromethylphenyl)amine (273) is used as an agrochemical.181Me NO2 N S HNNO2 F3C 273 Many heterocyclic and aromatic compounds containing var- ious isothiazole residues 182, 183 and 3-isothiazolone deriva- tives 184 ± 189 in their side chains manifest high fungicidal activities. Isothiazolones are broad-spectrum biocides; they are included in various compositions including commercial ones. They hold great promise for the technology and agriculture being non-harmful to humans, animals and the environ- ment.20, 23, 30, 190 ± 197 Isothiazole and isothiazolone derivatives are used in photog- raphy,198 ± 212 cosmetics,213 ± 215 and as dyes;216 ± 218 besides, they are components of synergistic bactericidal preparations 219 ± 225 and are used in the chemiluminescent analysis 226, 227 of radio- graphic visualisation.228 Many of them are antifouling agents.229 ± 235 A great number of isothiazole derivatives manifest high antimicrobial activities and are widely used for the protection of natural and technical materials, plants and other objects from harmful microorganisms.These include 3-alkoxyisothiazoles (particularly, compound 122),104 allylic ethers 128 ± 130,106 5-ami- noisothiazoles,236 3,5-diaminoisothiazole derivatives 6,51 isothia- zolecarboxamides 101 and the derivatives 125 and 126.105 The overwhelming majority of compounds endowed with antimicrobial activities were found among isothiazolone deriva- tives.237 ± 265 It was found that the addition of even the simplest isothiazolones to various compositions, e.g., 2-methyl- or 5-chloro-2-methyl-2,3-dihydroisothiazol-3-one, increase mark- edly their activities.266 ± 271 Some isothiazole derivatives manifest different types of bio- logical activities.Thus acylated aminoisothiazoles (compounds 165 125, 272 and 274 273) are used as insecticides and acaricides. Me N O NC CH2C(O)NH S 274 Compositions possessing bactericidal and fungicidal proper- ties were prepared from some dihydroisothiazolones.274 Cyano- isothiazoles 112 are used as fungicides and bactericides. Compositions comprising aminoacrylamides and trisubstituted isothiazoles possess insecticidal, fungicidal and herbicidal activ- ities.275 Compounds 275 have been patented as nematocides, insecticides and acaricides.276Isothiazoles (1,2-thiazoles): synthesis, properties and applications R N S 275CHCH2CH2SOn , n=0±2.R=CF2 N-Isothiazol-5-ylamides (e.g., compound 110 92) manifest an even broader range of pesticidal activities and possess nematoci- dal, insecticidal, miticidal and fungicidal properties. An intense search for therapeutic agents among isothiazole derivatives has been carried out over the past decade. Cephalo- sporins containing a (4-carboxy-3-hydroxyisothiazol-5-yl)thio- methyl group in position 3 (276) are used as medicinal drugs possessing a broad spectrum of antibacterial activities, particu- larly with respect to gram-negative bacteria.277 NOCH(SR1)CO2R2 OH O¡2 C CONH N S Bun4 N+ N Ph3CHN S S S O CO2CH2 276 C6H4OMe-4 Isothiazolyl aminoalkyl ketones were patented as myorelax- ants.278 5-Carboxy-3-phenylisothiazoles, particularly 4-amino-3- (3-trifluoromethylphenyl)isothiazole-5-carboxylic acid (277), were recommended for use as antiinflammatory agents.279 ± 281 C6H4CF3-3 H2N N HO2C S 277 Dihydroisothiazol-3-one 1,1,-dioxide derivatives inhibit ser- ine proteases. They have been recommended for the treatment of various inflammatory diseases and as antimetastatic agents.7, 8 Some aromatic and heterocyclic compounds containing isothia- zole residues inhibit phosphodiesterase and are widely used in the clinical practice as antiasthmatic drugs and in the treatment of diabetes, hypertension, allergic rhinites, nephrites and other diseases.282, 283 Compounds 278 and their salts have been patented as effective agents for the treatment of Alzheimer's disease.4 R1 N R2 S 278 R1=H, alkyl, cycloalkyl, (un)substituted phenyl; R2=substituted pyrrolidinyl.Trisubstituted isothiazole derivatives containing an amino acid or cyclobutene fragment in position 3 were recommended for the treatment of brain ischemia, Huntington's and Parkinson's diseases, epilepsy, Alzheimer's disease, schizophrenia, stress, anxiety and various memory disorders.284 5-Hydrazinoisothiazole derivatives act as immunodepres- sants,285 whereas 3-isothiazolone derivatives are included in the composition of weight reduction drugs.286 They are also used for modification of the interleukin-5 receptor.287 The isothiazole ring is the constituent of various physiologi- cally active compounds.Thus heterocyclic derivatives of urea containing an isothiazole substituent act as antagonists of 5-HT2C and 5-HT2B receptors and are used in the preparation of anti- depressant drugs and in the treatment of Alzheimer's disease, schizophrenia, etc.288 Compound 279 is a highly efficient antag- 691 onist of 5-HT2B receptors, although its activity is not too high.289 ± 291 2-Amino-3-(3-hydroxy-5-methylisothiazol-4-yl)pro- pionic acid (280), the isothiazole analogue of 2-amino-3-(3- hydroxy-5-methylisoxazol-4-yl)propionic acid (281, AMPA), blocks the activation of glutamate receptors which are crucial for brain ischemia, Alzheimer's disease, hypoglycemia, etc.292, 293 O HO2C OH Me HN HN N H2NMe N S S 280 279 Me HO2C OH N H2NMe O 281 V. Conclusion As can be seen from the material presented above, new original methods for the synthesis of many isothiazole derivatives have been developed in recent years. These include some previously unknown heterocyclisation reactions (e.g., the syntheses based on a-amino ketones and 2-nitropentachloro-1,3-butadiene), the syn- theses of isothiazoles from other, more accessible heterocyclic compounds (particularly, the syntheses based on trithiatriazine trichloride) as well as the syntheses of previously unknown fluoro- substituted isothiazoles.Studies aimed at the functionalisation of isothiazoles, e.g., cross-coupling reactions which made possible the synthesis of previously unavailable derivatives (e.g., isothi- azoles containing two acetylene groups) have received further development. The syntheses of biologically active amides, isothia- zolones and highly reactive 1,1-dioxides have been carried out. Isothiazoles were used as starting materials for the synthesis of novel, previously unavailable biheterocyclic structures, such as imidazoisothiazoles, diisothiazolopyrazines, isothiazolotriazines, substituted pyrroloisothiazole 1,1-dioxides and other compounds. Methods for the synthesis of various heterocyclic compounds (pyrazoles, thiadiazapentalenes and other hardly accessible com- pounds) from readily available isothiazole derivatives have been developed.High biological activities of isothiazoles is the main impetus for conducting in-depth studies in the field of isothiazole chem- istry. More than 50% of work cited in this review are patents devoted to the synthesis and applications of isothiazoles as efficient agrochemical and medicinal preparations. Obviously, the exceptionally broad spectrum of useful proper- ties of isothiazoles determines the expediency of further studies into their chemistry. We hope that the present review will help attract more attention to this field of chemistry. References 1. A Adams, R Slask Chem. Ind. (London) 42 1232 (1956) 2. R Raap, R G Micetich J.Med. Chem. 11 70 (1968) 3. US P. 3 268 523; Chem. Abstr. 65 18 597 (1967) 4. US P. 5 409 946; Chem. Abstr. 123 55 872 (1995) 5. US P. 5 411 977; Ref. Zh. Khim. 06 O 76P (1997) 6. US P. 5 378 729; Chem. Abstr. 123 33643 (1995) 7. US P. 5 550 139; Ref. Zh. Khim. 07 O 152P (1999) 8. WO PCT 95 18 797; Chem. Abstr. 123 340 140 (1995) 9. L Matzen, A Engesgaard, B Ebert,M Didriksen, B Frolund, P Krogsgaard-Larsen, J W Jaroszewski J. Med. Chem. 40 520 (1997) 10. Austr. P. 670 809; Ref. Zh. Khim. 22 O 391P (1997) 11. US P. 5 681 793; Ref. Zh. Khim. 07 O 394P (2000) 12. US P. 5 888 931; Ref. Zh. Khim. 02 O 475P (2000) 13. US P. 5 888 932; Ref. Zh. Khim. 04 O 402P (2000)692 14. US P. 5 888 937; Ref. Zh. Khim. 11 O 431P (2000) 15. Jpn. Appl.04-270 203; Chem. Abstr. 118 75 364 (1993) 16. US P. 5 591 760; Ref. Zh. Khim. 01 O 255P (1998) 17. US P. 5 489 588; Ref. Zh. Khim. 02 O 379P (1998) 18. US P. 5 648 086; Ref. Zh. Khim. 23 O 385P (1998) 19. US P. 5 559 083; Ref. Zh. Khim. 12 O 426P (1999) 20. BRD Appl. 19 548 710; Chem. Abstr. 127 118 601 (1997) 21. US P. 5 661 119; Chem. Abstr. 127 222 242 (1997) 22. US P. 5 922 745; Ref. Zh. Khim. 14 O 380P (2000) 23. US P. 5 658 467; Chem. Abstr. 127 224 778 (1997) 24. US P. 5 188 663; Chem. Abstr. 119 88 937 (1993) 25. US P. 5 683 686; Ref. Zh. Khim. 17 O 402P (1999) 26. Jpn. Appl. 07-140 620; Chem. Abstr. 123 241 899 (1995) 27. Jpn. Appl. 09-211 786; Chem. Abstr. 127 255 264 (1997) 28. BRD Appl. 4 426 623; Chem. Abstr. 124 204 936 (1996) 29.Br. Appl. 2 335 924; Ref. Zh. Khim. 09 N 162P (2000) 30. T Mulligam Spec. Chem. 16 161 (1996); Ref. Zh. Khim. 05 O 282 (1997) 31. R Slack, K R H Wooldridge Adv. Heterocycl. Chem. 4 107 (1965) 32. K R H Wooldridge Adv. Heterocycl. Chem. 14 1 (1972) 33. M Hori, H Shimizu Sulfur Rep. 4 313 (1986) 34. D M McKinnon Sulfur Rep. 15 317 (1994) 35. B Iddon Heterocycles 41 533 (1995) 36. J W Pavlik Mol. Supramol. Photochem. (Organic Photochemistry) 1 57 (1997); Chem. Abstr. 128 204 460 (1998) 37. V P Litvinov Usp. Khim. 68 817 (1999) [Russ. Chem. Rev. 68 737 (1999)] 38. Th L Gilchrist J. Chem. Soc., Perkin Trans. 1 2849 (1999) 39. M F Ward Annu. Rep. Prog. Chem., Sect. B 95 157 (1999) 40. A R Katrizky, J Lagowski Heterocyclic Chemistry (New York: Wiley) (Translated from the English; Moscow: Izd.Inostr. Lit., 1963) 41. L A Paquette Principles of Modern Heterocyclic Chemistry (New York: Benjamin, 1968) 42. K Turnbull, in Heterocyclic Chemistry Vol. 5 (London: Wiley, 1986) 43. T L Gilchrist The Chemistry of Heterocyclic Compounds (Translated from the English; Moscow: Mir, 1996) 44. R F Chapman, B J Peart, in Comprehensive Heterocyclic Chemistry II Vol. 3 (Ed. I Shinkai) (Oxford: Elsevier, 1996) 45. P A Bradley,D J Wilkins Progress in Heterocyclic Chemistry Vol. 11 (Eds G W Gribble, T L Gilchrist) (Oxford: Pergamon Press, 1999) 46. Methoden der Organischen Chemie (Houben-Weyl) Bd. EVIII(III) (Stuttgart: Georg Thieme Verlag, 1993) 47. S D Sokolov Usp. Khim. 58 533 (1979) [Russ.Chem. Rev. 58 289 (1979)] 48. M M Kempbell, in Comprehensive Organic Chemistry (Eds D Barton, W D Ollis) (Oxford: Pergamon Press, 1979) Vol. 4 49. A Adams, R Slack J. Chem. Soc. 3061 (1959) 50. R E Hackler, K W Burow Jr, S V Kaster, D I Wickiser J. Heterocycl. Chem. 26 1575 (1989) 51. M T Cocco, C Congiu, A Maccioni, V Onnis, M L Schivo, A De Logu Farmaco 49 137 (1994); Chem. Abstr. 121 205 259 (1994) 52. Pol. P. 175 889; Ref. Zh. Khim. 01 O 17P (2000) 53. A Kunugi, Md A Jabbar, K Mori, H Uno Elecrochim. Acta 44 4583 (1999) 54. Eur. Appl. 1 053 680; Ref. Zh. Khim. 06 O 158P (2001) 55. R V Kaberdin, V I Potkin, Yu A Ol'dekop Dokl. Akad. Nauk SSSR 300 1133 (1988) a 56. R V Kaberdin, V I Potkin, Yu A Ol'dekop Zh. Org. Khim. 26 1560 (1990) b 57.A I Verenich, A A Govorova, N MGalitskii, V I Potkin, R V Kaberdin, Yu A Ol'dekop Khim. Geterotsikl. Soedin. 399 (1992) c 58. A N Akopyan, A M Saakyan, Z A Dzhauari Arm. Khim. Zh. 21 (5) 414 (1968) 59. M G Voronkov, N S Vyazankin, E N Deryagina Reaktsii Sery s Organicheskimi Soedineniyami (Reactions of Sulfur with Organic Compounds) (Novosibirsk: Nauka, 1979) 60. J N Bridson, S B Copp,M J Schriver, S G Zhu,M J Zaworotko Can. J. Chem. 72 1143 (1994) 61. J N Bridson,M J Schriver, S G Zhu Can. J. Chem. 73 212 (1995) 62. M M Krayushkin,M A Kalik, A Ya Kudryavtseva Izv. Akad. Nauk, Ser. Khim. 1892 (1992) d 63. D K Buffel, L Meerpoel, S M Toppet, G J Hoornaert Nucleosides Nucleotides 13 719 (1994); Chem. Abstr. 121 83 859 (1994) R V Kaberdin, V I Potkin 64.X-G Duan, X-L Duan, C W Rees J. Chem. Soc., Perkin Trans. 1 2831 (1997) 65. X-L Duan, Ch W Rees, T-Y Yue Chem. Commun. 367 (1997) 66. X-L Duan, R Perrins, Ch W Rees J. Chem. Soc., Perkin Trans. 1 1617 (1997) 67. Ch W Rees, T-Y Yue J. Chem. Soc., Perkin Trans. 1 2247 (1997) 68. S M Laaman, O Meth-Cohn, Ch W Rees Synthesis 757 (1999) 69. X-L Duan, Ch W Rees J. Chem. Soc., Perkin Trans. 1 3189 (1997) 70. K Hartke, A Kraska Arch. Pharm. 326 (1) 57 (1993); Chem. Abstr. 118 191 686 (1993) 71. B Schulze, U Dietrich, K Illgen, J Siler Zh. Org. Khim. 30 1379 (1994) b 72. B Schulze, G Kirsten, S Kirrbach, A Rahm, H Heimgartner Helv. Chim. Acta 74 1059 (1991) 73. S Kirrbakh, K Myetze, R Kempe, R Meizinger, A Kol'berg, B Schulze Zh.Org. Khim. 32 1745 (1996) b 74. O Miyato, S Yamakawa,K Muroya, T Naito Heterocycles 51 1513 (1999) 75. Bhavana, S Hastak, B J Ghiya Ind. J. Heterocycl. Chem. 2 133 (1992); Chem. Abstr. 118 212 952 (1993) 76. L Capuano, V Hammerer, V Huch Liebigs Ann. Chem. 23 (1994) 77. M R Bryce, S Yoshida, A S Batsanov, J A K Howard J. Chem. Soc., Perkin Trans. 1 1157 (1997) 78. K Emayan, R F English, P A Koutentis, Ch W Rees J. Chem. Soc., Perkin Trans. 1 3345 (1997) 79. D Clarke,K Emayan, Ch W Rees J. Chem. Soc., Perkin Trans. 1 77 (1998) 80. R B Woodword The Harvey Lecture Series 59 (New York: Academic Press, 1965) p. 31 81. S G Zlotin, P G Kislitsyn, O A Luk'yanov Izv. Akad. Nauk, Ser. Khim. 1887 (1997) d 82. S G Zlotin, P G Kislitsyn, O A Luk'yanov Izv.Akad. Nauk, Ser. Khim. 534 (1998) d 83. S G Zlotin, P G Kislitsyn, O A Luk'yanov Izv. Akad. Nauk, Ser. Khim. 537 (1998) d 84. V I Potkin, N I Nechai, R V Kaberdin Izv. Akad. Nauk Belarusi, Ser. Khim. Nauk (4) 85 (1994) 85. WO PCT 97 38 987; Chem. Abstr. 127 346 385 (1997) 86. WO PCT 97 38 996; Chem. Abstr. 127 346 386 (1997) 87. WO PCT 97 38 988; Chem. Abstr. 127 358 855 (1997) 88. BRD Appl. 19 614 859; Ref. Zh. Khim. 20 O 483P (1998) 89. Jpn. Appl. 07-196 637; Chem. Abstr. 124 8805 (1996) 90. F Guilloteau, L Miginiac Synth. Commun. 25 1383 (1995) 91. H G Raubenheimer,M Desmet, G J Kruger J. Chem. Soc., Dalton Trans. 2067 (1995) 92. WO PCT 95 31 448; Chem. Abstr. 124 202 240 (1996) 93. J G Samaritoni, L Arndt, T J Bruce, J E Dripps, J Gifford, C J Hatton, W H Hendrix, J R Schoonover, G W Johnson, V B Hegde, S Thornburgh J.Agric. Food. Chem. 45 1920 (1997); Chem. Abstr. 126 302 652 (1997) 94. BRD Appl. 19 542 372; Ref. Zh. Khim. 23 O 408P (1998) 95. BRD Appl. 19 736 545; Ref. Zh. Khim. 12 O 375P (2000) 96. BRD Appl. 4 328 425; Ref. Zh. Khim. 14 O 398P (1997) 97. Jpn. Appl. 06-09 313; Chem. Abstr. 120 263839 (1994) 98. Jpn. Appl. 08-277 276; Chem. Abstr. 126 47213 (1997) 99. Jpn. Appl. 08-277 277; Chem. Abstr. 126 47214 (1997) 100. BRD Appl. 19 750 012; Ref. Zh. Khim. 09 O 383P (2000) 101. BRD Appl. 19 842 354; Ref. Zh. Khim. 03 O 367P (2001) 102. K Burak, Z Machon Pharmazie 47 492 (1992); Chem. Abstr. 120 298 527 (1994) 103.Pol. P. 167 891; Chem. Abstr. 125 58 496 (1996) 104. BRD Appl. 19 712 409; Ref. Zh. Khim. 20 O 326P (1999) 105. Eur. Appl. 542 468; Chem. Abstr. 119 139 214 (1993) 106. US P. 5 508 417; Chem. Abstr. 125 58 497 (1996) 107. X-P Yang, Z-M Li, L-X Wang, Y-H Li Chem. J. Chin. Univ. 19 228 (1998); Ref. Zh. Khim. 24 O 806 (1998) 108. A Alberola, L Calvo,M T R Rodriguez,M C Sanudo J. Heterocycl. Chem. 30 393 (1993) 109. A Alberola, L Calvo,M T R Rodriguez,M C Sanudo J. Heterocycl. Chem. 32 537 (1995) 110. U Lipnicka, Z Machon Acta Pol. Pharm. 50 235 (1993); Chem. Abstr. 121 35 413 (1994) 111. X-P Yang, Z-M Li, J Lui, H-S Chen Chin. J. Appl. Chem. 16 115 (1999); Ref. Zh. Khim. 22 Zh 351 (1999)Isothiazoles (1,2-thiazoles): synthesis, properties and applications 112.Can. Appl. 2 152 156; Chem. Abstr. 124 317 143 (1996) 113. BRD Appl. 1 975 0011; Ref. Zh. Khim. 11 O 410P (2000) 114. E E Swayze, L B Townsend J. Heterocycl. Chem. 30 1379 (1993) 115. E E Swayze, L B Townsend J. Org. Chem. 60 6309 (1995) 116. C J Shishoo, M B Devani, S Ananthan, V S Bhadti, C V Ullas J. Heterocycl. Chem. 25 759 (1988) 117. S G Zlotin, K S Chunikhin, M O Dekaprilevich Mendeleev Commun. 97 (1997) 118. S G Zlotin, A V Bobrov, K S Chunikhin Izv. Akad. Nauk, Ser. Khim. 1350 (1999) d 119. S G Zlotin, A V Bobrov Izv. Akad. Nauk, Ser. Khim. 957 (2000) d 120. US P. 3 341 547; Chem. Abstr. 68 114596 (1968) 121. Jpn. Appl. 05-59 024; Chem. Abstr. 119 117 242 (1993) 122. US P. 3 155 678; Chem. Abstr. 62 2778 (1965) 123.US P. 5 484 934; Ref. Zh. Khim. 23 O 373P (1997) 124. Eur. Appl. 604 978; Chem. Abstr. 121 134 107 (1994) 125. J G Samaritoni, J M Babcock,M L Schlenz,G W Johnson J. Agr. Food Chem. 47 3381 (1999); Ref. Zh. Khim. 08 O 401 (2000) 126. K Ohkata, Y Ohyama, K Akiba Heterocycles 37 859 (1994); Chem. Abstr. 121 57 558 (1994) 127. M Konrad, F Meyer, M Buechner, K Heinze, L Zsolnai Chem. Ber.-Recl. 130 95 (1997) 128. N V Onyamboko,M Renson, C Paulmier Bull. Soc. Chim. Belg. 96 407 (1987) 129. M S Mohamed Egypt. J. Pharm. Sci. 35 427 (1994); Chem. Abstr. 123 285 924 (1995) 130. A V Bobrov, B B Averkiev, S G Zlotin, M Yu Antipin Izv. Akad. Nauk, Ser. Khim. 1226 (2001) d 131. B Schulze, K Illgen J. Prakt. Chem., Chem.-Ztg. 339 1 (1997); Chem.Abstr. 126 117 919 (1997) 132. J L Marco, S Ingate Tetrahedron Lett. 38 4835 (1997) 133. S T Ingate, J L Marco, M Witvrouw, C Pannecouque, E DeClercq Tetrahedron 53 17795 (1997) 134. J L Marco, S T Ingate, P Manzano Tetrahedron Lett. 39 4123 (1998) 135. J L Marco, S T Ingate, C Jaime, I Bea Tetrahedron 56 2523 (2000) 136. J L Marco, S T Ingate, P M Chinchon Tetrahedron 55 7625 (1999) 137. S G Zlotin, A I Gerasyuto Izv. Akad. Nauk, Ser. Khim. 396 (1999) d 138. Sh Li, Zh Li Guangzhou Huagong 22 10 (1994); Chem. Abstr. 123 313 798 (1995) 139. BRD Appl. 19 652 531; Chem. Abstr. 127 135 791 (1997) 140. R Glinka, E Zak, K Walczynski Zesz. Nauk. Politech. Lodz., Technol. Chem. Spozyw. 708 (53) 51 (1995); Chem. Abstr. 124 86 868 (1996) 141.US P. 5 466 818; Ref. Zh. Khim. 24 O 436P (1997) 142. Sh Li, X Tan Jingxi Huagong (Fine Chem) 16 4 (1999); Ref. Zh. Khim. 22 O 382 (1999) 143. Jpn. Appl. 06-192 245; Chem. Abstr. 121 280 633 (1994) 144. US P. 5 420 290; Ref. Zh. Khim. 10 O 87P (1997) 145. K Hartke, C Ashry Pharmazie 51 638 (1996); Chem. Abstr. 125 300 769 (1996) 146. O Carugo, F Clerici, D Pocar Tetrahedron 49 9117 (1993) 147. F Clerici, F Galletti, D Pocar, P Rowersi Tetrahedron 52 7183 (1996) 148. F Clerici, T Ferrario, M L Gelmi, R Marrelli J. Chem. Soc., Perkin. Trans. 1 2533 (1994) 149. P Baggi, F Clerici, M L Gelmi, S Mottadelli Tetrahedron 51 2455 (1995) 150. S Abou El Maaty,M S Mohamed, F El Telebany Egypt. J. Pharm. Sci. 33 441 (1992); Chem. Abstr.121 57 440 (1994) 151. A Rolfs, J Liebscher Angew. Chem. 105 797 (1993) 152. Eur. Appl. 583 561; Chem. Abstr. 120 244661 (1994) 153. A Rolfs, P G Jones, J Liebscher J. Chem. Soc., Perkin Trans. 1 2339 (1996) 154. S H Kim, K Kim J. Heterocycl. Chem. 30 928 (1993) 155. B Friedrich, A Fuks, B Schulze Zh. Org. Khim. 30 1404 (1994) b 156. B Schulze, K Rosenbaum, J Hilbig, L Weber J. Prakt. Chem.- Chem. Ztg. 334 25 (1992) 157. B Schulze, U Obst, G Zahn, B Friedrich, R Cimirgalia, H-J Hofmann J. Prakt. Chem.-Chem. Ztg. 337 175 (1995) 158. J W Pavlik, P Tongcharoensirikul, N P Bird, A C Day, J A Barltrop J. Am. Chem. Soc. 116 2292 (1994) 159. J W Pavlik, P Tongcharoensirikul J. Org. Chem. 65 3626 (2000) 160. M de Wang, H Alper J. Organomet. Chem.451 169 (1993) 693 161. S H Kim, Y Y Lee, K Kim, J H Kim Bull. Korean Chem. Soc. 15 273 (1994); Chem. Abstr. 121 108 116 (1994) 162. E M Beccalli, F Clerici,M L Gelmi Tetrahedron 55 14975 (1999) 163. D M Argilagos, A Linden, R W Kunz, H Heimgartner Chimia 51 455 (1997); Ref. Zh. Khim. 13 Zh 264 (1998) 164. D M Argilagos,M J Trimino, A M Cabrera, A Linden, H Heimgartner Helv. Chim. Acta 80 273 (1997); Ref. Zh. Khim. 3 Zh 290 (1999) 165. D M Argilagos, R W Kunz, A Linden, H Heimgartner Helv. Chim. Acta 81 2388 (1998); Ref. Zh. Khim. 14 Zh 129 (1999) 166. B Unterhalt, C Langfermann,M Majdrpour, S Kirrbach, A Kolberg, B Schulze Pharmazie 53 764 (1998); Ref. Zh. Khim. 9 Zh 264 (1999) 167. B Schulze, A Noack, A Kolberg, in 6-ya Mezhdunarodnaya Konferentsiya `Khimiya Karbenov i Rodstvennykh Intermediatov' (Programma i Tez.Dokl.), S Peterburg, 1998 [The 6th International Conference `Chemistry of Carbenes and Related Intermediates' (Programme and Abstracts of Reports), S Petersburg, 1998] p. 137 168. F Clerici, E Erba, M L Gelmi,M Valle Tetrahedron 53 15 859 (1997) 169. F Clerici, M L Gelmi, R Soave, M Valle Tetrahedron 54 11 285 (1998) 170. E Coutouli-Argiropoulou, C Anastasopoulos J. Heterocycl. Chem. 33 731 (1996) 171. E M Beccalli, F Clerici,M L Gelmi Tetrahedron 55 2001 (1999) 172. N N Mel'nikov Zh. Vses. Khim. O-va im D I Mendeleeva 33 602 (1988) 173. BRD Appl. 19 614 858; Ref. Zh. Khim. 19 O 386P (1998) 174. Eur. Appl. 761 654; Chem. Abstr. 126 277 469 (1997) 175.US P. 5 438 033; Chem. Abstr. 123 1221 (1995) 176. BRD Appl. 19 620 135; Ref. Zh. Khim. 4 O 467P (1999) 177. US P. 5 324 837; Chem. Abstr. 121 205 339 (1994) 178. US P. 5 338 856; Chem. Abstr. 121 300 887 (1994) 179. US P. 6 096 898; Ref. Zh. Khim. 11 O 334P (2001) 180. Eur. Appl. 578 246; Chem. Abstr. 120 323 562 (1994) 181. WO PCT 93 19 054; Chem. Abstr. 120 107 030 (1994) 182. Eur. Appl. 538 231; Chem. Abstr. 119 160 256 (1993) 183. WO PCT 93 13 078; Chem. Abstr. 119 271 189 (1993) 184. Eur. Appl. 648 757; Chem. Abstr. 123 55 869 (1995) 185. Zh-M Li, H-S Chen, J Liu, Sh-Ih Lish Chem. J. Chin. Univ. 20 395 (1999); Ref. Zh. Khim. 20 Zh 277 (1999) 186. Br. Appl. 2 274 779; Chem. Abstr. 121 295 087 (1994) 187. Br. Appl. 2 304 574; Ref.Zh. Khim. 23 O 378P (1997) 188. US P. 5 668 083; Ref. Zh. Khim. 11 O 458P (1999) 189. Jpn. Appl. 06-227 912; Chem. Abstr. 121 295 094 (1994) 190. Can. Appl. 2 058 252; Chem. Abstr. 118 141 816 (1993) 191. Can. Appl. 2 069 453; Chem. Abstr. 121 29 268 (1994) 192. US P. 5 430 046; Ref. Zh. Khim. 2 O 287P (1997) 193. Jpn. Appl. 08-67 853; Chem. Abstr. 125 36 071 (1996) 194. WO PCT 9614092; Chem. Abstr. 125 95 454 (1996) 195. BRD Appl. 4 344 549; Ref. Zh. Khim. 21 O 398P (1997) 196. RF Appl. 96 122 055; Ref. Zh. Khim. 19 O 388P (2000) 197. US P. 5 595 575; Ref. Zh. Khim. 13 N 162P (1998) 198. Eur. Appl. 712 039; Chem. Abstr. 125 99 942 (1996) 199. Jpn. Appl. 04-194 928; Chem. Abstr. 118 90 712 (1993) 200. Jpn. Appl. 04-152 340; Chem.Abstr. 118 90 693 (1993) 201. Jpn. Appl. 06-03 780; Chem. Abstr. 121 22 412 (1994) 202. Jpn. Appl. 05-333 486; Chem. Abstr. 121 22 406 (1994) 203. Eur. Appl. 508 457; Chem. Abstr. 118 157 782 (1993) 204. Jpn. Appl. 04-177 239; Chem. Abstr. 118 70 018 (1993) 205. Jpn. Appl. 04-128 839; Chem. Abstr. 118 112 860 (1993) 206. Eur. Appl. 529 794; Chem. Abstr. 119 259 396 (1993) 207. Eur. Appl. 549 175; Chem. Abstr. 119 237 893 (1993) 208. Jpn. Appl. 04-260 036; Chem. Abstr. 118 136 118 (1993) 209. US P. 5 514 505; Chem. Abstr. 125 127 644 (1996) 210. Jpn. Appl. 08-201 979; Chem. Abstr. 125 288 680 (1996) 211. Jpn. Appl. 05-331 379; Chem. Abstr. 121 95 875 (1994) 212. Jpn. Appl. 09-133 977; Chem. Abstr. 127 57 952 (1997) 213. R Matissek GJT-Suppl.(7) 5; 8 (1987); Chem. Abstr. 108 137 671 (1988) 214. Czech. P. 276 924; Chem. Abstr. 121 163 678 (1994) 215. Dtsch. Lebensm.-Rundsch. 89 181 (1993); Chem. Abstr. 121 163 671 (1994) 216. Eur. Appl. 534 296; Chem. Abstr. 119 162 371 (1993)694 217. Eur. Appl. 545 420; Chem. Abstr. 119 162 375 (1993) 218. Jpn. Appl. 09-292 684; Chem. Abstr. 128 68 443 (1998) 219. Jpn. Appl. 07-207 090; Chem. Abstr. 123 230 303 (1995) 220. WO PCT 9 612 407; Chem. Abstr. 125 28 275 (1996) 221. US P. 5 591 759; Ref. Zh. Khim. 5 O 404P (1998) 222. US P. 5 523 020; Ref. Zh. Khim. 10 O 92P (1998) 223. Eur. Appl. 594 440; Chem. Abstr. 121 258 107 (1994) 224. US P. 5 552 423; Ref. Zh. Khim. 9 O 436P (1999) 225. Jpn. Appl. 08-198 714; Chem. Abstr. 125 240 850 (1996) 226.Jpn. Appl. 08-38 195; Chem. Abstr. 124 311 832 (1996) 227. Jpn. Appl. 08-38 196; Chem. Abstr. 124 311 833 (1996) 228. US P. 5 908 931; Ref. Zh. Khim. 07 O 47P (2000) 229. Jpn. Appl. 07-196 403; Chem. Abstr. 123 342 819 (1995) 230. Jpn. Appl. 06-40 812; Chem. Abstr. 121 29 270 (1994) 231. Eur. Appl. 706 974; Chem. Abstr. 124 324 873 (1996) 232. Jpn. Appl. 08-20 676; Chem. Abstr. 124 319 797 (1996) 233. US P. 5 670 055; Chem. Abstr. 127 298 502 (1997) 234. W W Bingaman, G L Willingham Int. Biodeter. Biodegr. 34 387 (1994); Chem. Abstr. 124 79 433 (1996) 235. US P. 6 010 693; Ref. Zh. Khim. 1 O 377P (2001) 236. BRD Appl. 19 654 147; Ref. Zh. Khim. 13 O 436P (1999) 237. Eur. Appl. 542 489; Chem. Abstr. 119 111 272 (1993) 238. Jpn. Appl. 05-163 107; Chem. Abstr. 119 175 864 (1993) 239. Jpn. Appl. 05-170 608; Chem. Abstr. 119 175 869 (1993) 240. Jpn. Appl. 05-286 815; Chem. Abstr. 120 127 766 (1994) 241. US P. 5 288 693; Chem. Abstr. 121 35 595 (1994) 242. Jpn. Appl. 06-87 705; Chem. Abstr. 121 76 166 (1994) 243. Jpn. Appl. 06-148 957; Chem. Abstr. 121 241 882 (1994) 244. Jpn. Appl. 06-227 916; Chem. Abstr. 121 295 095 (1994) 245. Jpn. Appl. 07-258 009; Chem. Abstr. 124 23 925 (1996) 246. Eur. Appl. 686 346; Chem. Abstr. 124 79 443 (1996) 247. Eur. Appl. 686 347; Chem. Abstr. 124 79 444 (1996) 248. W Anker Microb. Materialzerstoerung Materialschutz 151 (1995); Chem. Abstr. 124 236 544 (1996) 249. US P. 5 494 588; Chem. Abstr. 124 241 661 (1996) 250. Jpn. Appl. 08-165 206; Chem. Abstr. 125 161 120 (1996) 251. Jpn. Appl. 07-330 520; Chem. Abstr. 124 223 707 (1996) 252. Jpn. Appl. 07-277 910; Chem. Abstr. 124 48 313 (1996) 253. Jpn. Appl. 07-277 911; Chem. Abstr. 124 48 314 (1996) 254. Jpn. Appl. 08-239 694; Chem. Abstr. 126 3076 (1997) 255. US P. 5 670 529; Chem. Abstr. 127 283 428 (1997) 256. Jpn. Appl. 09-209 272; Chem. Abstr. 127 221 913 (1997) 257. Eur. Appl. 834 253; Chem. Abstr. 128 254 064 (1998) 258. Jpn. Appl. 10-27 381; Chem. Abstr. 128 174 208 (1998) 259. Eur. Appl. 827 939; Chem. Abstr. 128 196 442 (1998) 260. US P. 5 703 105; Chem. Abstr. 128 106 408 (1998) 261. Pol. P. 169 519; Ref. Zh. Khim. 17 O 376P (1997) 262. Eur. Appl. 774 444; Chem. Abstr. 127 55 587 (1997) 263. Jpn. Appl. 09-263 504; Chem. Abstr. 127 342 912 (1997) 264. BRD Appl. 19 651 351; Ref. Zh. Khim. 16 O 440P (1999) 265. BRD Appl. 19 734 244; Ref. Zh. Khim. 07 O 138P (2000) 266. Jpn. Appl. 04-290 806; Chem. Abstr. 118 118 948 (1993) 267. Jpn. Appl. 08-193 014; Chem. Abstr. 125 240 846 (1996) 268. Jpn. Appl. 09-151 101; Chem. Abstr. 127 132 269 (1997) 269. Jpn. Appl. 08-183 703; Chem. Abstr. 125 188 330 (1996) 270. Jpn. Appl. 08-183 704; Chem. Abstr. 125 188 331 (1996) 271. Jpn. Appl. 08-253 404; Chem. Abstr. 126 28 030 (1997) 272. J J Sheets, A Schmidt, J G Samaritoni, J M Gifford J. Agr. Food Chem. 45 4826 (1997) 273. BRD Appl. 19 542 372; Chem. Abstr. 127 50 631 (1997) 274. Jpn. Appl. 08-23 136 ; Chem. Abstr. 126 3075 (1997) 275. Jpn. Appl. 08-245 561; Chem. Abstr. 126 31 346 (1997) 276. US P. 5 952 359; Ref. Zh. Khim. 09 O 384P (2001) 277. R Hara, E Nakai, H Hisamichi, N Nagano J. Antibiot. 47 477 (1994) 278. US P. 5 338 857; Chem. Abstr. 123 83 354 (1995) 279. Eur. Appl. 578 477; Chem. Abstr. 120 261 340 (1994) 280. WO PCT 9 500 501; Chem. Abstr. 124 55 954 (1996) 281. Eur. Appl. 587 378; Chem. Abstr. 120 290 112 (1994) 282. M J Ashton, D C Cook, G Fenton, J-A Karlsson, M N Palfreyman, D Raeburn, A J Ratcliffe, J E Souness, S Thurairatham, N Vicer J. Med. Chem. 37 1696 (1994) 283. WO PCT 9 815 530; Chem. Abstr. 128 294 695 (1998) 284. Eur. Appl. 0 994 107; Ref. Zh. Khim. 19 O 117P (2000) R V Kaberdin, V I Potkin 285. U Lipnicka, Z Machon Acta Pol. Pharm. 54 207 (1997); Chem. Abstr. 127 346 328 (1997) 286. Jpn. Appl. 07-324 007; Chem. Abstr. 124 253 329 (1996) 287. R Devos, Y Guisez, G Plaetinck, S Cornelis, J Tavernier, J van der Heyden, L H Foley, J E Scheffler Eur. J. Biochem. 225 635 (1994) 288. WO PCT 94 14801; Chem. Abstr. 121 179 617 (1994) 289. O N Zefirova, N S Zefirov Usp. Khim. 70 382 (2001) [Russ. Chem. Rev. 70 333 (2001)] 290. I T Forbes, P Ham, D H Booth, R T Martin,M Thompson, G S Baxter, T P Blackburn, A Glen, G A Kennet,M D Wood J. Med. Chem. 38 2524 (1995) 291. I T Forbes, G E Jones, O E Murphy, V Holland, G S Baxter J. Med. Chem. 38 855 (1995) 292. O N Zefirova, N S Zefirov Zh. Org. Khim. 36 1273 (2000) b 293. L Matzen, B Ebert, T B Stensbol, B Frolund, J W Jaroszewski, P Krogsgaard-Larsen Bioorg. Med. Chem. 5 1569 (1997) a�Dokl. Chem. (Engl. Transl.) b�Russ. J. Org. Chem. (Engl. Transl.) c�Chem. Heterocycl. Compd. (Engl. Transl.) d�Russ. Chem. Bull., Int. Ed. (Engl.
ISSN:0036-021X
出版商:RSC
年代:2002
数据来源: RSC
|
4. |
The use of molecular-statistical methods for the calculation of thermodynamic characteristics of adsorption for identification of organic compounds by gas chromatography–mass spectrometry |
|
Russian Chemical Reviews,
Volume 71,
Issue 8,
2002,
Page 695-706
Alexey K. Buryak,
Preview
|
|
摘要:
Russian Chemical Reviews 71 (8) 695 ± 706 (2002) The use of molecular-statistical methods for the calculation of thermodynamic characteristics of adsorption for identification of organic compounds by gas chromatography ± mass spectrometry A K Buryak Contents I. Introduction II. Semiempirical molecular-statistical theory of adsorption III. The use of theoretical methods for identification of compounds Abstract. descrip- theoretical the on research of state current The The current state of research on the theoretical descrip- tion semiempirical by region Henry the in adsorption of tion of adsorption in the Henry region by semiempirical molec- molec- ular-statistical of identification to applied as procedures, ular-statistical procedures, as applied to identification of organic organic compounds considered.is mixtures, complex in compounds in complex mixtures, is considered. Various Various approaches potential atom ± atom the correcting to approaches to correcting the atom ± atom potential parameters parameters used of characteristics thermodynamic determine to used to determine thermodynamic characteristics of adsorption adsorption are the in involved calculations of Examples compared. are compared. Examples of calculations involved in the chromato- chromato- graphic spectrometric mass ± chromatography and graphic and chromatography ± mass spectrometric identification identification of are isomers including compounds organic real and model of model and real organic compounds including isomers are given. given.The bibliography includes 89 references The bibliography includes 89 references. I. Introduction Identification of organic compounds in mixtures is one of the most complicated problems of physical organic chemistry. The main difficulty arising in the work with mixtures is that simultaneous solution of two problems is required: the mixture should be separated into components and the components should be iden- tified. At present, the most informative and reliable method for identification of organic compounds is the chromatography ± mass spectrometry (GC-MS) technique.1 Modern instruments make it possible to analyse volatile compounds in the gas chromatography ± mass spectrometry mode (GC-MS) and non- volatile compounds, in the liquid chromatography ± mass-spec- trometry mode (LC-MS).2 The use of different ionisation techniques allows one to produce stable molecular ions, to vary fragmentation and to detect both positive and negative ions.2, 3 In recent years, the range of masses that can be determined has markedly extended due to application of new ionisation techni- ques.4 A significant advantage of mass spectrometry is the possibility to determine the masses of molecular and fragment ions in the high-resolution mode with an accuracy of millionths of atomic mass units.Thus, it is possible to determine almost unambiguously the elemental compositions of molecules.5 The only limitation that holds up extensive use of high-resolution mass spectrometers for identification of substances is their high cost, which is 5 ± 10 times as high as the cost of instruments for routine analysis.6 Chromatography ± IR spectroscopy or chromatogra- phy ±NMR spectroscopy could be possible alternatives to the A K Buryak Institute of Physical Chemistry, Russian Academy of Sciences, Leninsky prosp. 31, 119991 Moscow, Russian Federation.Fax (7-095) 335 17 78. Tel. (7-095) 330 19 29. E-mail: AKBuryak@icp.rssi.ru Received 12 February 2002 Uspekhi Khimii 71 (8) 788 ± 800 (2002); translated by Z P Bobkova #2002 Russian Academy of Sciences and Turpion Ltd DOI 10.1070/RC2002v071n08ABEH000711 695 696 701 GC-MS;7 however, the development of these techniques is held up by relatively low sensitivity of IR spectroscopy and insufficient speed of operation of NMR spectrometers.Even a highly potent method such as high-resolution mass spectrometry does not allow reliable identificaion of isomers in complex mixtures. This is due to the fact that differences in the fragmentation can hardly be attributed unambiguously to the structure of a particular isomer. Therefore, the use of chromato- graphic data for identification of substances becomes very impor- tant.8 The procedures for chromatographic identification of compounds based on the use of reference compounds, specific detectors, preliminary chemical reactions,9 chromatographic spectra 10 and dependences of retention indices on various molec- ular characteristics 11 have been developed fairly comprehensively and we do not consider them here. However, all these methods are much inferior to the GC-MS technique in labouriousness, infor- mation content and reliability.When discussing the GC-MS identification, one should note that the methods of structure determination based on mass- spectrometry data can be classified into three groups. (1) Library search which consists in comparing the spectrum of an unknown compound with the spectra deposited in databases or libraries and selecting those spectra that fit best the spectrum of the compound under study.12 (2) Methods for representing the compound structure on the basis of fragmentation rules.13 (3) Methods for substance identification consisting in deter- mination of structural features by statistical analysis.14 Successful utilisation of mass spectrometry and determination of the structures of organic compounds requires the use of empirical rules and regularities elucidated in the studies of large numbers of compounds belonging to different structural classes,13, 15 the presence of vast libraries of standard mass spectra (up to 300 thousand) of compounds, the use of high-end software for image recognition which resort to data of other spectroscopic methods 16 (not only experimental but also obtained by calcula- tions 17) for structure determination and advanced computing machinery.The use of GC-MS provides not only mass-spectro- scopic but also chromatographic information, namely, the reten- tion indices for various stationary phases.These values are sufficiently accurate because a mass spectrometer is a highly selective detector and chromatographic peaks are usually much better resolved in ionic mass chromatograms than in chromato- grams recorded with other detectors. Combination of chromatographic identification techniques with GC-MS to identify isomers in complex mixtures is usually resorted to in investigations carried out along two lines, namely, identification based on semiempirical molecular-statistical calcu-696 lations of retention values 18 and identification using retention indices calculated theoretically.19 Retention index is a qualitative characteristic of a molecule correlated with physicochemical parameters, for example, with the molecule polarisability or with the melting or boiling point of the substance.8, 11, 19 Numerous equations have now been found to relate retention parameters (calculated as retention indices) to various physicochemical properties of molecules, in particular, isomer molecules.20 Commonly, these equations make use of so- called increments which characterise separate fragments of a molecule that can be joined to form a new molecule.The main drawback of this approach is that assembling the increments gives an approximate geometrical structure rather than the real mole- cule. Therefore, fine intramolecular effects, especially those char- acteristic of isomers, can be taken into account only by introduc- ing special corrections, which are rather arbitrary.21 Generally, this situation resembles the situation faced by an attempt to obtain precise values for molecular polarisability in terms of additive schemes �¢ one too many corrections are needed to reach high accuracy. One more drawback of this method is the necessity of experimental determination of increments on the basis of a broad range of reference molecules.22 Extensive databases of experimen- tal retention indices have now been composed; they can be efficiently used to identify separate classes of organic mole- cules.23, 24 Such bases are also supplemented by theoretically calculated indic; however, these have not yet found wide use.Using the semiempirical molecular-statistical theory of adsorption, one can calculate the thermodynamic characteristics of adsorption (TCA) including the retention values for organic molecules of various classes.The theory is based on the use of atom ¡¾ atom potentials (AAPs) for the interaction between the sorbent and sorbate atoms. An important advantage of this theory is the fact that the calculations of retention indices are based on the real structures of molecules, determined from electron diffraction data or constructed with allowance for the regularities of variation of bond angles and bond lengths. In the case of isomers, the TCA calculations alone based on real data on the molecular structure permit reliable identification of compounds. II. Semiempirical molecular-statistical theory of adsorption 1. Scope of molecular-statistical calculations To date, molecular-statistical calculations of TCA have been carried out for a broad range of adsorbents and adsorbates.The calculations of TCA for inert gases, simple polar molecules and hydrocarbons on zeolites of various types 25 and on porous crystalline silica (silicalite) have been reported.26 ¡¾ 28 The molec- ular-statistical calculations have been developed most thoroughly for systems comprising graphitised thermal carbon black (GTCB) and hydrocarbons or their derivatives.29 Graphitised thermal black is a good model sorbent because it has a homogeneous planar surface the geometry and the chemical properties of which have been studied in sufficient detail. Vast experimental informa- tion on the adsorption of organic compounds of various classes on the GTCB surface has been accumulated.29, 30 The main reason why the molecular-statistical calculations for the adsorption on GTCB surface are being developed and improved is the fact that this process is exceptionally sensitive to the structure of adsorbed molecules. This creates conditions for the separation of geometric isomers.30 ¡¾ 32 In most cases, a combination of the molecular- statistical calculations and GC-MS data provides unambiguous identification of individual isomers in complex mixtures.18, 32 2.The procedure of molecular-statistical calculations Kiselev and Poshkus 29, 30 developed a semiempirical molecular- statistical method for the calculation of thermodynamic charac- teristics of the physical adsorption of gases on GTCB at very low surface coverages.Within the framework of this method, the A K Buryak 1=2exp (1) K1= 1 4p Fz00 Henry constant (K1) for the adsorption of quasi-rigid molecules is calculated using the equation ¢§ F0 sinYdYdc, kT OO 2pkT z z where F0 and F00 are the potential function F values for the interaction between the adsorbate and adsorbent molecules in a potential minimum and its second derivative with respect to the distance z from the center of gravity of the molecule to the adsorbent surface; Y and c are Euler angles, which define the orientation of the molecule relative to the adsorbent surface. The F00 andF0 values and the equilibrium distance z0 depend onYand c. The potential function F is presented as the sum of the AAPs (2) j F= A:::CGTCB : A CGTCB for the interaction of each atom of the adsorbed molecule (A) with each carbon atom of GTCB (CGTCB).X X The jA:::CGTCB potential is written in the Buckingham ¡¾ Corner form (6, 8, exp) (3) jA:::CGTS=7C6r767C8r78+Bexp(7qr), where C6 , C8 are the parameters of the dipole ¡¾ dipole and dipole ¡¾ quadrupole dispersion interaction potentials; r is the interatomic distance; B and q are parameters of the repulsion potential. The C8 and C6 values are estimated using approximate quantum-mechanical formulae. The parameter C6 is found using the Kirkwood ¡¾MuE ller relation aab (4) C6=76mec2 Oaa=waU a Oab=wbU, a where me is the mass of an electron; c is the velocity of light; aa , ab are polarisabilities of atoms a and b; wa and wb are diamagnetic susceptibilities of atoms a and b.The parameter C8 is estimated using the formula ¢§1 ab=wb , (5) aaab a12 C8= 45h2 32p2m aa=wa e ¢§1 a 1 a 1 aa=wa ab=wb 12 where h is Planck's constant. The repulsion parameter B is expressed in terms of the equilibrium interatomic distance r0 and other AAP parameters using the equilibrium condition for all the attraction and repulsion forces for r=r0 (6) dr a 0, dj rar0 (7) 3C6r20 . B=6C6 expOqr0U 1 a 4C8 qr70 The parameter q is taken to be 35.7 nm71. The equilibrium distance r0 is normally estimated using crystal parameters or data on the interaction of atoms of the adsorbate molecule with GTCB. In particular, the value r0=0.382 nm was accepted for C...CGTCB (this value was obtained from data on the graphite lattice proper- ties) and r0=0.341 nm was taken for H...CGTCB.The summation of j over the GTCB carbon atoms above the graphite basis plane is done using the Crowell approximation.30 The potential function i (8) pW Fi=7C6,i 2d 4 x pW ¢§C8;i 3d 6 x Fi for the interaction of an i-th atom of the molecule with the graphite basis plane is calculated from the equation zd ; 6 a zi d ; 4 a2BipW Oqzi a 1Ue¢§qz; q2The use of molecular-statistical methods for the calculation of thermodynamic characteristics of adsorption where W=0.38261073 nm72 is the number of carbon atoms per unit area of the graphite basis plane; d is the interplanar spacing in graphite equal to 0.356 nm; x is the Riemann zeta function; zi is the distance from the i-th atom to the plane passing through the centres of atoms of the outermost atomic layer of graphite. Relation (8) takes into account the contribution of only the outermost atomic layer of graphite to the repulsion energy.The differential molar changes of the intrinsic energy DU1 and entropy DS1 upon adsorption are calculated using the relations(9) DU1=RT2 d lnK1 dT , (10) DS R d 1 a lnK1 a T d lnTK1 ¢§ 1. H:::CGTCB A special program is used to calculate the TCA.33 The parameter C6 is first estimated from Eqn (4). However, the K1 and DU1 values found using the C6 value determined in this way differ markedly from the corresponding experimental data.29 Therefore, the AAPs for the interaction of atoms in the adsorbate molecule with GTCB (jH:::CGTCB and jC:::CGTCB for each atom of the molecule in different valence states) are further corrected using relations (4) ¡¾ (11) and the experimental K1 values for one or two simple representatives of the class of hydrocarbons under consid- eration.The j and jCOsp3U:::CGTCB values are corrected assuming that the parameters C8/C6 , r0 and q, estimated in terms of the properties of the adsorbate and the adsorbent taken separately, differ only slightly from the exact values and that the following equality holds: C6;C:::C (11) a C6KM ;C:::C, C6;C:::H CKM 6;C:::H where C6,C...C and C6,C...H are the values of parameters that satisfy experimental data; CKM 6;C:::C and C6KM ;C:::H are the C6 values estimated using the Kirkwood ¡¾MuE ller formula. If these assumptions are valid, the corrected potential functions j are related to the rough functions j* in the following way: (12) j=bj*, (13) F= j*=bF*, j=b C C A A GTCB GTCB X X X X where (14) 6;C:::C b=C6;C:::C a C6;C:::H .CKM CKM 6;C:::H The b coefficient is usually determined by comparing theK1 values calculated on the basis ofF* with those found experimentally. The equations for the interaction potentials of atoms in alkanes with GTCB corrected in this way (kJ mol71) have the form jCOsp3U:::CGTCB=71.38661073 r7672.14861075 r78+ (15) +1.896105exp(735.7r), (16) jH:::CGTCB=70.49861073 r7670.95061075 r78+ +3.606104exp(735.7r).The K1 and DU1 values calculated using these potentials for alkanes with known molecular structures are close to experimen- tal results. Assuming that the jH:::CGTCB potential of the hydrogen atom in a hydrocarbon molecule does not depend on the valence state of the car20. atom and that the r0 and q parameters are identical for carbon in any valence state, it was found that (17) jCOsp2U:::CGTCB=djCOsp3U:::CGTCB , where d is the parameter depending on the valence state of the carbon atom. 697 Using the experimental K1 values, it was found for ethylene thatjCOsp2U:::CGTCB=71.4861073 r7672.3061075 r78+ (18) +2.0266105exp(735.7r).Relying on the K1 values determined experimentally for acetylene adsorption, the potential jCOspU:::CGTCB can also be calcu- lated. These potentials were used to calculate the TCA for a large number of hydrocarbons: alkanes,34 cycloalkanes,29, 35 alkenes, alkynes 36 and aromatic hydrocarbons 37 on the GTCB surface. In the case of hydrocarbons with known molecular structures, the TCA values found in this way coincide with the corresponding experimental values to within the error of determination. The above-described procedure was used to determine the AAPs for the interaction of the carbon atoms in graphite with oxygen in two valence states (in ethers and ketones) 38 and to calculate the TCA for a number of oxygen-containing organic compounds on the GTCB surface.39 Guiochon and coworkers 40, 41 also used the molecular-stat- istical method for calculation of the TCA for various hydro- carbons and their chlorinated derivatives; however, these calculations were performed using uncorrected AAPs, therefore, the calculated and experimental TCA markedly differ.When calculating F for complex conjugated molecules, researchers attempted to take into account the anisotropy of graphite polar- isability.42 For chloro-derivatives of aromatic hydrocarbons, it was found that the order of elution of isomers out of the column is better described by a model that does not take into account the electrostatic interaction of the adsorbate molecule with the adsorbent surface.41 Apart from the organic compounds mentioned above, adsorp- tion of noble gases 29, 30 and nitrogen-,43 sulfur- and selenium- containing 44 compounds on the GTCB surface has also been studied.Some of the AAP parameters known to date are listed in Table 1. Satisfactory agreement of the experimental results with the theoretical results obtained using these parameters is attained in those cases where the mutual influence of atoms in the adsorbed molecule is insignificant. When the mutual influence of atoms is substantial, corrections should be applied to the AAPs. In some Table 1. Parameters of the AAPs for the interaction of various atoms with the GTCB carbon atom. Element b B C8 AAP parameter a C6 1618.0 2903.800 70.85 70.06403 70.07000 70.5672 70.9775 8491.0 9085.370 77 70.14530 70.15547 71.5750 71.6853 31189.489 30204.558 29219.626 70.38000 70.36800 70.35600 74.2180 74.0848 73.9516 0.95 0.92 0.89 3250.015 5086.273 26759.806 71.565 7 70.11005 70.17223 70.45069 71.1194 71.7519 73.8672 37501.136 36514.270 6194.358 70.45060 70.43878 70.26445 75.0740 74.9404 71.6318 0.76 0.74 7 Hydrogen Fluorine Carbon sp3 sp2 Chlorine see b ortho ortho-ortho Oxygen ketone ether Sulfur Bromine see b ortho Nitrogen aC6 is expressed in kJ m6 mol71, C8 is in kJ m8 mol71, and B is in kJ mol71.bAn atom not interacting with the neighbouring atoms in the molecule.698 cases, the calculations are hampered by the absence of precise structural parameters for sorbate molecules.45 Thus, fairly good agreement between the calculated and experimental TCA values for compounds with known molecular structures is attained if corrected AAP values are used to calculate the potential function of intermolecular interactions.However, for some molecules, TCA cannot be predicted with a satisfactory accuracy. For example, during identification of isomers in mixtures, the sequence of isomer emergence from the column with GTCB predicted by calculations does not coincide in some cases with the experimental one. This is due to the difference between the AAP parameters for atoms incorporated in the molecule in question and the parameters corrected using the reference molecules.The difference is caused by the change in the nearest environment or valence state of the atom, manifested as a change in polarisability. Therefore, to identify separate isomers in complex mixtures and to determine reliable TCA (especially, the Henry constants), one must find the precise values for the AAP parameters. 3. Correction methods for the parameters of the atom ± atom potentials Several methods for correcting the AAP parameters have been developed to date. a. Traditional method The approach proposed by the developers of the semiempirical molecular-statistical theory of adsorption 29 can be considered to be the first of the known methods for correcting the AAP parameters. This approach is based on determination of individ- ual AAPs (using reference molecules) for atoms forming the molecule under study, including chemically identical atoms in different valence states.In terms of the semiempirical molecular- statistical theory,29 the AAP parameters are determined from approximate quantum-mechanical formulae (see above) using some characteristics of adsorbate and adsorbent atoms, e.g., polarisability, diamagnetic susceptibility and equilibrium distan- ces. The atom ± atom potentials determined for the interaction of atoms of the adsorbate molecule with the adsorbent (GTCB) carbon atoms are corrected using the comparison of the theoret- ical and experimental Henry constants K1 for the adsorption of a representative of the class of compound studied. The correction is carried out using so-called reference molecules, whose adsorption has been studied most comprehensively.The obtained AAPs are assumed to be transferable, i.e., they allow correct prediction of the TCA for any molecule that contains the corresponding atoms. This \0procedure was used to determine and correct the AAP parameters for the adsorption of alkanes, alkenes, alkynes and aromatic hydrocarbons on GTCB; methane and ethane, ethylene, acetylene, and benzene, respectively, served as the reference molecules. In these calculations, three AAPs for carbon atoms corresponding to the sp3, sp2 and sp1 hybridisation states were found; they differ only by the correction factors, which are equal to 1.00, 1.07, and 1.27, respectively.Two AAPs for the oxygen atoms in ethers and ketones were found in the same way; they differ by the correction factors, which are 0.72 and 0.92, respec- tively.38 The AAPs for sulfur, selenium,44 fluorine,46 chlorine,47 and bromine 48 were determined. The corresponding correction factors are equal to 0.87, 0.89, 0.85, 0.95 and 0.76, respectively. Actually, this approach is similar to the calculation of polar- isabilities, diamagnetic susceptibilities, and the dipole moments of molecules in terms of additive schemes.49 The AAPs determined in this way were used to predict the TCA for a number of hydro- carbons and their derivatives, which are in good agreement with experimental data.36, 38, 46 However, for many molecules [both relatively simple ones such as ortho-halobenzenes,50 perhydroan- thracenes, perhydrophenanthrenes,51 heterocyclic nitrogen-con- taining compounds,52, 53 and fairly complex molecules such as molecules with abnormal geometric parameters (strained mole- A K Buryak cules), with bulky substituents, with several heteroatoms, and their unusual combinations], the TCA could not be correctly predicted.The main drawbacks of the above method are as follows. For each valence state (and, in the general case, for each type of the atom nearest environment), one must know a separate AAP, which has to be corrected on the basis of experimental data for reference molecules. However, the number of these AAPs may become too large (especially for carbon, oxygen and nitrogen atoms), which would complicate the calculation.In addition, this procedure does not take into account the changes in the AAPs caused by the influence of atoms not bound by valence bonds. b. Taking into account the dependence of the atom ± atom potentials on the bond angles The method proposed by Arkhipova 54 takes into account the effect of the change in the geometric parameters of a molecule on the carbon atom hybridisation and, correspondingly, on the AAP parameters. A pattern of dependence of AAP parameters on the deviation of the bond angles formed by carbon atoms from the ideal values corresponding to the sp3, sp2 and sp1 hybridisation states has been proposed. Using this dependence, one can deter- mine more precisely the AAPvalues for atoms that form non-ideal bond angles.The results obtained by this method for a series of strained molecules (for example, cyclopropane derivatives and some bicyclic compounds) are in good agreement with exper- imental data. An example illustrating the efficiency of this method is the correct (experimentally confirmed) prediction of the order of elution of C8H12 isomers, cyclobutadiene and bicyclo[3,3,0]oct- 1,5-ene, from a column with GTCB, which could not be deter- mined without applying corrections. However, this method also suffers from serious drawbacks: first, it is applicable only to molecules with a known geometry because corrections are based on structure differences; second, determining a correction requires the knowledge of, at least, two boundary potentials, because the correction provides some inter- mediate value of the AAP parameters; third, this method cannot be used to correct the AAP parameters that change under the influence of non-bonded atoms, which does not distort the geo- metric structure of the molecules.Other `AAP parameter ± atom property' relationships have also been constructed; the atom properties used in these relationships are, in particular, the polar- isability, the effective charge, the shielding constant, and the hybridisation state. The methods based on these dependences suffer from the above drawbacks.55 c. Variation of the parameters of atom ± atom potentials The variation of the AAP parameters is done in order to minimise the discrepancy between the experimental and calculated data for all molecules of the group chosen for the investigation; this can be done, for example, by the regularisation technique.56 A single parameter can also be varied.For example, in the calculation of AAP for a series of bromobenzene molecules, the role of this parameter was played by the equilibrium distance between the GTCB carbon atom and the bromine atom.48 However, this method allows one to attain the optimal agree- ment between the experimental and theoretical results only for the considered group of molecules. The transfer of parameters cor- rected for a given group of molecules to other (even `related') molecules usually results in a deviation between the experimental and calculated values for new molecules.For example, the AAP values corrected by varying the GTCB carbon ± halogen equili- brium distance are in good agreement with experimental data only for the molecules of halobenzenes containing no halogen atoms in ortho-positions.48 Otherwise, to describe experimental data with a satisfactory accuracy, additional corrections taking into account the mutual influence of atoms (corrections for ortho-effects) should be applied when calculating the AAPs.The use of molecular-statistical methods for the calculation of thermodynamic characteristics of adsorption d. The use of corrected initial data for the calculation of atom ± atom potentials It is known 57 that the mutual influence of atoms can change their polarisabilities.In calculations of the TCA, these changes are taken into account either by applying additional corrections to the equations for AAP48 or by using corrected polarisabilities. In our opinion, the use of corrected polarisabilities is most efficient for calculating the TCA of perhalogenated compounds, first of all, freons, for which isostructural fragments can hardly be chosen (or there are too many of them). The adsorption heats 58 of such compounds are described satisfactorily, within the experimental error, without introducing any corrections; however, for predict- ing the Henry constants, the correction of the AAPs is necessary. The method in which corrected initial data (first of all, polar- isabilities and diamagnetic susceptibilities) are used in the calcu- lation of the AAP parameters is of undisputed interest also due to the fact that a vast experimental base has now been accumulated and detailed additive schemes for polarisability calculations for molecules of diverse compounds,59 especially perhalogenated ones, have now been developed.A shortcoming of this method is the need to measure the polarisabilities, which requires a substantial amount of the material. In addition, one should borne in mind the relatively low accuracy of the polarisability values calculated according to additive schemes,60, 61 which can lead to greater errors in the calculations of the TCA. e. The introduction of corrections for ortho-effects Changes in the AAP parameters are due to the mutual influence of atoms, in particular, when they occupy ortho-positions.61 By varying these parameters, one can attain the correspondence between experimental and theoretical data for an ortho-substi- tuted molecule with known experimental TCA.The corrected AAPs can be used for the calculation of TCA for other molecules with atoms in ortho-positions. The applicability of this approach (the introduction of corrections for the ortho-effect) was demon- strated in relation to halobenzenes.48 Actually, this method for correcting the AAPs is similar to the method using reference molecules (and, correspondingly, the drawbacks are also the same) but applies to a more limited group of molecules. In some cases, additional corrections are required to bring the theoretical results in correspondence with the experiment, for example, ortho ± ortho-effect corrections.62 As an example, one can consider 1,2-dichlorobenzene and 1,2,3-trichlorobenzene molecules in which corrections for the ortho- and ortho ± ortho-effects are required. Cl Cl Cl Cl Cl Probably, it is expedient to apply such corrections to mole- cules that possess specific properties, for example, to adamantane in which 1-substituents are subject to the `cage effect', which is also involved in adsorption.The term `cage effect' implies the overlap of the electronic orbitals of the carbon atoms at the 1-positions which are not bound by valence bonds. This overlap brings about a substantial difference between the properties of 1- and 2-substituted isomers, which is manifested, in particular, in the chromatographic sepa- ration on phases having different polarity. The allowance made for the cage effect for chloro- 63 and amino-substituted adaman- tane isomers 64 permitted the prediction of the correct order of isomer elution from a column packed with GTCB, whereas traditional calculations 65 did not provide a correct prediction.The overlap of the orbitals of the bridgehead carbon atoms and their influence on the substituent (amino group) polarisability is illustrated by Fig. 1. 699 NH2 Figure 1. Cage effect in the 1-aminoadamantane molecule. f. The method of isostructural fragments The method of isostructural fragments 66 (IF) is based on the use of AAPs corrected for some single molecule or its fragment in the calculations made for other molecules containing the same frag- ment.For a molecule containing the isostructural fragment, the AAPs are corrected by processing the theoretical and experimen- tal values by the least-squares calculations, their agreement being attained by varying the correction factor. In doing this, the AAPs for atoms of one sort are changed, in order to decrease the total number of AAPs used in calculations. The IF is chosen on the basis of comparison of the geometrical structures of molecules of the group under interest; in particular, the specially obtained electron diffraction data, the known regularities of the change in the bond angles and bond lengths in groups of related molecules, or the results of quantum-chemical calculations can be compared.Usually, molecules with a geometrical structure (or its fragment) common to the given class of molecules are chosen as the IF. For example, in the case of cyclopropane derivatives, this is the cyclopropane molecule, and for ortho-substituted chlorobenzenes and naphthalenes, this is the 1,2-dichlorobenzene molecule. In fact, the IF method allows one to overcome the limitations of the atom ± atom approximation and takes into account the mutual influence of atoms and groups of atoms including the change in the polarisabilities and geometric parameters of the interacting fragments. g. The method of variation of the molecular geometry An important benefit of the semiempirical molecular-statistical theory of adsorption is the use of structural parameters of molecules in the TCA calculations.However, this benefit is simultaneously a drawback, because exact parameters of the molecular structure have been determined for a limited range of substances and, normally, they are lacking for most of isomers. An alternative to all the methods of AAP correction considered above is the method of variation of the geometric parameters of the molecule. This procedure has been used to predict the retention values for perhydroanthracenes and perhydrophenanthrenes.51 By using this variation, the researchers predicted the order of isomer elution from a GTCB column that coincided with the experimental results.However, the geometry variation, especially for molecules with an unknown structure, is legitimate only in the case of quantitative comparison of theoretical and experimental data, and, naturally, it is impossible to estimate the variation of the AAP parameters in terms of this method. 4. Comparison of various methods for the correction of the atom ± atom potential parameters Among the methods for correcting the AAP parameters consid- ered above, the following are most important: the method based on the use of corrected initial data (first of all, polarisability); the introduction of ortho-effect corrections and the IF method. The Henry constants calculated using the AAP parameters corrected by these methods for a number of molecules have been com- pared.66, 67 The results of comparative analysis confirm the good prospects of the IF method.For a number of chlorobenzenes with different degrees of substitution, the Henry constants were calculated using the AAP values corrected by various methods and the results were com-700 pared. The chlorobenzene molecules were chosen as the model series because they represent groups of isomers, and the sequence of elution of isomers from a GTCB column is an important criterion of the reliability of calculations. In addition, both experimental and calculated TCA determined by various methods are available for molecules of this series (Table 2).41, 47, 48, 68 Of all possible TCA that can be calculated using the semiempirical molecular-statistical adsorption theory, the Henry constants were chosen as the characteristics most sensitive to variations of the AAP parameters and the geometrical structures of molecules. It was shown 48, 50 that even for molecules containing no strained fragments or ortho-arranged atoms, the experimental and theo- retical values of the Henry constants can be brought in corre- spondence by correcting the AAP only in terms of the procedure of reference molecules.Therefore, corrections were applied to the already corrected potential, which was 5% smaller than the initial one.In the calculations in terms of the IF method, known AAPs tested in the calculations of TCA for chlorobenzenes were used as the basic parameters.The IF chosen were 1,2-di- and 1,2,3- trichlorobenzenes. Relying on the experimental data for these molecules (the Henry constants and their differences for isomers Table 2. Values lnK1 (cm3m72) for the adsorption of chlorobenzenes on GTCB at 473K obtained experimentally and calculated using the AAPs corrected by various methods.50 Molecule Experi- Calculations ment 1 2 3 4 50.98 1.08 1.06 0.98 1.15 1.06 70.44 70.44 70.44 70.44 70.44 0.98 1.01 1.06 2.25 0.98 1.01 1.06 2.25 0.98 1.10 1.06 2.25 2.23 Chlorobenzene 70.44 1,3-Dichlorobenzene 0.93 1,2-Dichlorobenzene 1.01 1,4-Dichlorobenzene 1.06 1,3,5-Trichloro 2.44 2.44 2.54 2.41 2.52 2.52 2.74 2.42 3.88 3.86 4.11 3.86 3.92 3.92 4.12 3.89 4.06 4.03 4.37 4.24 benzene 1,2,4-Trichloro- benzene 1,2,3-Trichloro- benzene 1,2,3,5-Tetrachloro- benzene 1,2,4,5-Tetrachloro- benzene 1,2,3,4-Tetrachloro- benzene 2.25 2.50 2.60 3.40 4.02 3.58 2.25 2.59 2.82 4.19 4.23 4.47 Notes.The correction of the AAPs by method 1 was carried out using the reference 1,4-dichlorobenzene molecule;48 the correction by method 2 was based on the experimental polarisability value determined in Ref. 60 for the chlorine atoms in 1,2-dichlorobenzene (in the calculations of lnK1 for other chlorobenzenes, these corrected AAP values were used only for ortho-arranged chlorine atoms); in method 3, the corrections for the ortho- and ortho ± ortho-effects were taken from Refs 48, 50; in the case of method 4, 1,2-di- and 1,2,3-trichlorobenzenes were chosen as the iso- structural fragments; the corrections for the AAPs based on the experi- mental Henry constants or the differences between the retention of neighbouring isomers is 0.92 for the fragment containing two chlorine atoms in ortho-positions and 0.91 for the fragment containing three atoms of this type; in the correction by method 5, the molecular geometry was varied assuming that ortho-chlorine atoms cause a deflection of the carbon7chlorine bond from the benzene ring plane; for the first five chlorobenzene molecules, the angles of deflection of the carbon7chlorine bonds from the benzene ring plane were taken from Ref.48, while for other isomers, the transferability of the angles was assumed, in particular, in calculations of the Henry constant for 1,2,3-trichlorobenzene, a value of 78 was taken, as for 1,2,4-trichlorobenzene, while in the case of 1,2,3,4- and 1,2,3,5-tetrachlorobenzenes, this value was 128, as for the 1,2,4,5- substituted isomer.A K Buryak closest in retention values), researchers obtained corrections for the AAPs to describe the interaction between the chlorine atom and the graphite carbon atom. In any case, the molecular- statistical calculations were carried out as described above (see Section II.2). The geometric parameters of molecules were found either from electron diffraction data or by combining the known parameters for other molecules of the given group.However, the introduction of corrections by some procedures does not always provide the best correspondence between the experimental and calculated data. For example, the use of the method of reference molecules, the method of correcting the polarisability values and the geometry variation give rise to inconsistency between the experimental and calculated orders of isomer elution out of a GTCB column, and for some isomers, the experimental and theoretical lnK1 values differ appreciably (see Table 2). The best results are provided by the IF method and by applying the ortho-effect corrections. The IF method appears to be more versatile and reliable because its application rules out the uncertainty related to the geometrical structures of molecules.Naturally, it is suitable for refining the parameters of AAPs not only for the molecules considered above. In our opinion, this method is especially advantageous for calculating the TCA of complex (in particular, isomeric) molecules containing heteroatoms and strained frag- ments. 5. Combination of the gas chromatography ± mass-spectrometry data with semiempirical molecular-statistical calculations When calculating retention indices, in particular, Kovac indices, an approximate rather than real structure of the molecule is actually used.8, 10 Therefore, in those cases where corrections for definite effects have not been applied, identical values are usually obtained for isomers. The corrections are not transferable: usu- ally, they are applicable only to the group of molecules for which they have been determined.When the retention indices are calculated within the framework of the semiempirical molecular- statistical adsorption theory, the real structure of the molecule is used.18, 30 To increase the efficiency of GC-MS identification, one should take into account the retention indices calculated in terms of this theory. Scheme 1 represents the pattern of usual GC-MS identification,1, 5 while Scheme 2 shows the identification proce- dure based on the molecular-statistical calculations.67 The first stage consists in molecular-statistical calculations performed to estimate tentatively the possible efficiency of the chromatographic column. The tentative calculations help to answer the key question of whether the isomers under consid- eration can be separated by chromatography using a GTCB column.Adsorption of a broad range of substances on this sorbent has now been studied, but their number, of course, is not Scheme 1 Data from gas chromato- graphy ± mass spectrometry Chromatographic methods Mass-spectromeric methods The use of reference Library search Structure representation compounds based on the fragmentation Retention indices pattern Structure- specific Molecular- statistical Calculation based on ionisation methods calculation incrementsThe use of molecular-statistical methods for the calculation of thermodynamic characteristics of adsorption Scheme 2 Tentative molecular-statistical calculations Evaluation of the possibility Selection of the chromatography of separation conditions Gas chromatography ± mass spectrometry experiment Identification by mass-spectromeric methods Check of the efficiency and selectivity of separation with other columns and sorbents Estimation of the efficiency and selectivity of separation using mass-spectrometric data Isomer identification based on molecular-statistical calculations Comparison of the hypothesised structure and the real fragmentation Comparison of experimental and theoretical retention indices Correction of calculations by the method of isostructural fragments commensurate with the number of known organic compounds; therefore, the tentative calculation of the TCA is a necessary item in the research.In addition, these calculations permit one to optimise the conditions of the chromatographic experiment planned, to choose the temperature, the carrier gas velocity, the adsorbent weight and the column length. The next stage is the GC-MS study in which individual isomers are to be separated and identified on the basis of differ- ences in the mass spectra or, if the spectra are identical, it is necessary to confirm that the substances being analysed belong to the group of isomers in question. The efficiency and selectivity of the chromatographic separation is evaluated on the basis of the observed overlapping mass spectra. The modern instrumentation and computer software for data processing allow this to be done in the mass chromatography and mass fragmentography modes.To verify the completeness of separation attained on a micropacked column with GTCB, it is expedient to record the chromatogram on a highly efficient capillary column most appropriate regarding the selectivity (in order to compare the results obtained with different columns). The next stage is conduction of molecular-statistical calcula- tions for isomers using corrections for the possible variations of the AAP parameters. In the simplest case, identification can be performed by comparing the sequence of isomer elution from a GTCB column observed experimentally with the calculated one. The experimental retention indices and those determined by molecular-statistical calculations can also be compared, which is generally preferred.In addition, the data on the molecular structure used in the calculations can be verified by examining fragmentation of the molecule, induced, for example, by an electron impact. Comparison of Schemes 1 and 2 shows that the identification procedure resorting to molecular-statistical calculations provides more reliable results. It is evident that combination of the results obtained in terms of the molecular-statistical theory with the results of a mass-spectrometric experiment would allow one to increase the efficiency, reliability and the information contents of the method and, hence, to solve the problem of isomer identifica- tion at a new level.Correlation equations that relate the Henry constant,69 the second virial coefficient of the intermolecular interaction,70 or the 701 heat of adsorption 71 to some characteristics of the molecule are usually considered as alternatives to the molecular-statistical calculations. The critical temperature and pressure or both,71 the molecular refraction,72 the dipole moments and the boiling points 73 are used most often. An advantage of such calculations is the possibility of changing the sorbent 71, 73 or specifying the degree of its inhomogeneity,74 which helps to elucidate the applicability of a sorbent to the separation and analysis of a particular group of substances. Good correlation coefficients are usually obtained for series of related molecules;69 however, the relationship between the molecular structure and the physico- chemical characteristics of adsorption remains obscure.There- fore, these methods are suitable only as auxiliary ones in the identification of isomers. An ab initio molecular-statistical method for calculation of the Henry constant has been developed.75 ± 80 The method makes use of the relationship between the structural characteristics of molecules and the energetic, geometric and topological 81 ± 84 properties involved in the adsorption. Unlike the traditional semiempirical molecular-statistical theory, the main idea of this approach has been the refusal of the additivity principle for AAP. Instead, the view of generalised charge of the molecule was introduced and a law for summing these charges was postulated. This theory was used to calculate the TCA for various types of hydrocarbons on GTCB and to determine the Kovac indices on a nonpolar stationary liquid phase (squalane).The calculated values are in good agreement with experimental data. This calculation method appears rather promising; however, it belongs to quantum chemistry and requires a separate review to be described. III. The use of theoretical methods for identification of compounds Mixtures of isomers can be separated into groups in conformity with the known numbers of isomers present. The simplest sit- uation arises when the mixture is known to contain all the possible isomers or when this can be determined from the results of a GC- MS experiment.When analysing mixtures of isomers with a `rigid' structure on a GTCB column, one should merely calculate the order of elution and thus identify the isomers. Such procedure has been carried out for the components of a mixture of endo- and exo- tetrahydrobicyclopentadienes. The endo-isomer has lower TCA and, correspondingly, it is the first to come out of the GTCB column. In this case, the fact that the mass spectra of the isomers are identical does not hamper the analysis.67 The isomers of 1-methyl-1,2-dicyclopropylcyclopropane 32 are more complex objects for identification, as their molecules are capable of internal rotation. The internal rotation angles are unknown; therefore, these isomers can be identified only on the basis of calculations performed for all the possible values of these angles; the differences between the Henry constants were calcu- lated and compared with the difference found in the experiment.Upon this comparison, it was found that the cis-isomer is the first to come out of the GTCB column. In this case, the molecular- statistical calculations represent the only possible method for identification because the mass spectra of the isomers are identi- cal. The isomer mixtures for which the composition can be suggested from a known chemical reaction pattern can reasonably be regarded as another group. The catalytic oxidative dimerisa- tion of methyl-, fluoro- and chlorobenzenes normally furnishes mixtures of six isomers containing one substituent in each benzene ring.Using molecular-statistical calculations, all the isomers of dimethyl- 85 and dichlorobiphenyls 86 were identified based only on the order of their elution from a GTCB column. The molec- ular-ion mass chromatograms are presented in Fig. 2. The follow- ing calculated values of lnK1 (cm3 m72) were obtained for dichlorobiphenyls at 475K and dimethylbiphenyls at 450 K:702 2,30 2,20 Substituent position 3.42 4.25 3.02 2.39 Methyl Chlorine Despite the fact that only traces of 2,20-substituted isomers were present, they were identified on the basis of mass spectra.87 In the case of difluorobiphenyls, the chromatogram exhibits only five peaks.To identify the isomers in conformity with Scheme 2, the Henry constants and the numbers of theoretical plates needed for the complete (n1) and partial (n2) resolution of each isomer pair were calculated. This gave the following results: Isomer pair 2,30 and 2,40 2,20 and 2,30 1500 22 n1 380 5 n2 For some isomers, only partial resolution is possible and this is confirmed by the chromatogram presented in Fig. 3 a. It can be seen that the mixture contains 2,20-difluorobiphenyl, which is the first to be eluted. To verify this, a chromatogram on a capillary column was obtained; however, in this case, too, the isomers could not be completely resolved (Fig. 3 b). The isomers cannot be identified using a mass spectrometry because they exhibit identical mass spectra.Therefore, comparison of the calculated and exper- imental retention values is the only way to confirm the presence of all isomers in the mixture. The results of calculations confirming that a mixture of difluorobiphenyls also contains all six isomers have been reported.87 a I (rel.u.) 750 600 450 300 1500 22 26 18 b I (rel.u.) 4000 3000 2000 10000 22 18 14 Figure 2. Molecular ion mass chromatograms of dimethylbiphenyl (a) and dichlorobiphenyl (b), m/z=182 and 222(224), respectively.85, 86 3,30 2,40 5.42 5.75 4.24 4.78 3,30 and 3,40 2,40 and 3,30 5100 22 1300 5 28 t /min 26 t /min 4,40 3,40 7.12 7.20 6.24 6.243,40 and 4,40 3500 900 a I (rel.u.) 3 1000 1 2 9 ~~ 500 3 b I (rel.u.) 1000 500 120 0 Figure 3.Chromatograms of a difluorobiphenyl isomer mixture obtained on a micropacked column withGTCB (a) and a capillary column (b).87 Isomer (position of the substituent): (1) 2,20; (2) 2,30; (3) 2,40; (4) 3,30; (5) 3,40; (6) 4,40. The advantages of calculations using the AAPs corrected by the IFmethod can be demonstrated in relation to the identification of trace impurities formed upon the catalytic oxidative dimerisa- tion of chlorobenzene.66, 88 The isostructural fragments used are presented in Fig. 4. The results of identification of the major components, dichlorobiphenyls, are presented above. The forma- tion of trichlorobiphenyls can easily be confirmed on the basis of mass-spectrometry data; however, it is impossible to identify individual isomers, as their mass spectra are identical.Figure 5 shows the mass chromatograms of the molecular ions of tri- and dichlorobiphenyls with m/z=256 and 226, respectively. It can be seen that the trichlorobiphenyls come out of the column with GTCB between two groups of dichlorobiphenyls. All the dichlorobiphenyls have been identified (see above), and their Henry constants can be used as the reference values in the identification of trichlorobiphenyl isomers. The results of calculations of the Henry constants for ortho-substituted tri- and dichlorobiphenyls are listed in Table 3. In this case, there is no need to carry out calculations for all the isomeric trichlorobiphen- yls, because only ortho-substituted isomers have retention times Cl Cl Figure 4.Isostructural fragments (marked by dashed lines) in molecules used for correction of the AAP parameters in calculations of the Henry constants of chlorobiphenyls. A K Buryak 5 6 4 15 t /min 2, 3 5 6 430 t /minThe use of molecular-statistical methods for the calculation of thermodynamic characteristics of adsorption a I (rel.u.) 360 300 240 180 120 600 38 36 b I (rel.u.) 400 300 200 1000 20 18 16 c I (rel.u.) 1600 1200 800 4000 20 18 16 Figure 5. Characteristic-ion mass chromatograms of di- and trichloro- biphenyls obtained from the chromatograms for total ion currents in the separation of a chlorobiphenyl mixture on a capillary (a) and micropacked (b, c) columns.66 The characteristic ion has m /z=256 (a, b), 226 (c).comparable with the retention times of dichlorobiphenyls. It can be seen from Table 3 that the elution of 3,30-dichlorobiphenyl from the GTCB column is preceded by the elution of 2,4,6-, 2,6,40- and 2,3,6-trichlorobiphenyls. The isomerwhose retention time is close to that of 2,30-dichlorobiphenyl is 2,3,20-trichlorobiphenyl, while the compound with a retention time similar to that of 3,30-dichlorobi- phenyl is 2,5,40-trichlorobiphenyl. It is noteworthy that the effi- ciency of separation of these isomers achieved on a microcolumn packed withGTCB is higher than that on a capillary column with a slightly polar liquid stationary phase.This outcome is attributable to the high `sensitivity' of the GTCB surface to the structural differences between the adsorbed molecules. In view of the large difference between the Henry constants for tri- and dichlorobiphen- yls (see Table 3), one can state that in this case, complete separation of the mixture has been attained and each chromatographic peak really corresponds to a single isomer. In the case of nonsymmetrical molecules, for example, 2,3,20- trichlorobiphenyl, two conformers can exist differing by the internal rotation angles. Since the structure of the adsorbed molecule is unknown, the calculations were carried out for the two possible conformations, syn and anti. The catalytic oxidative dimerisation of toluene 85 was found to give, in addition to dimethylbiphenyls, impurities with a molec- ular mass of 196.The chromatograms obtained using a GTCB 40 42 t /min 24 22 t /min 26 24 22 t /min Table 3. Henry constants calculated for tri- and dichlorobiphenyls at 500K and the internal rotation angles (a) used for the calculations.66 Molecule 2,6,20-Trichlorobiphenyl 2,30-Dichlorobiphenyl syn anti 2,3,20-Trichlorobiphenyl syn anti 2,5,20-Trichlorobiphenyl syn anti 2,6,30-Trichlorobiphenyl 2,4,20-Trichlorobiphenyl syn anti 2,40-Dichlorobiphenyl 2,4,6-Trichlorobiphenyl 2,6,40-Trichlorobiphenyl 2,3,6-Trichlorobiphenyl 3,30-Dichlorobiphenyl syn anti 2,5,40-Trichlorobiphenyl column (Fig. 6) show that the retention time of these compounds is close to the retention time of 2,4-dimethylbiphenyl.On the basis of analysis of the regular features of retention on GTCB, one can state that compounds with higher molecular masses can have I (rel.u.) 4000 3000 2000 1000 0 3 4 5 6 7 8 9 10 t /min I (rel.u.) 300 200 1000 5 6 7 8 9 10 t /min I (rel.u.) 300 200 1000 4 5 6 7 8 9 10 t /min Figure 6. Characteristic-ion mass chromatograms of dimethylbiphenyls (a) and impurities (b, c) obtained on a micropacked GTCB column.88 The characteristic ion has m /z=182 (a), 196 (b), 104 (c). 703 lnK1(cm3 m72) a /deg 2.88 65 53 3.34 2.91 3.38 2.97 127 60 120 60 120 3.15 3.48 3.62 60 3.70 3.30 3.94 4.15 4.32 4.57 60 120 53 60 60 60 4.85 4.50 4.87 38 142 53 abc704 shorter retention times only when their molecules are nonplanar, in particular, for large internal rotation angles.A dimethylbiphe- nylmethane or ortho-substituted trimethylbiphenyl structure can be proposed for these compounds. In this case, identification of the impurities using mass spectra confirmed the validity of the assumption. Analysis of the recorded mass spectra provides the conclusion that the isomer whose spectrum contains a large peak with m/z=104 (Fig. 7 a) corresponds to the group of dimethyl- biphenylmethanes. The mass spectrum of the other compound (Fig. 7 b) resembles most closely the spectrum of trimethylbi- phenyl but this assignment can hardly be confirmed on the basis of library search, as the published data are scarce.In order to identify this compound, molecular-statistical calculations of lnK1 (cm3 m72) were carried out for all possible trimethylbiphenyl isomers containing at least one methyl group in the ortho-position relative to the bond connecting the benzene rings. The results of these calculations are presented below. Substituent position 2,3,20 2,3,30 2,3,40 2,4,20 2,4,40 2,5,20 2,5,30 Yet another example of utilisation of the method in question for compound identification is analysis of an industrial mixture of chlorobiphenyls for the most toxic, `dioxin-like' isomers.66, 67, 89 The results of chromatographic separation of this mixture on micropacked and capillary columns showed that the degree of separation obtained on a micropacked column for a mixture containing tri- and tetrachlorobiphenyls is adequate for the molecular-statistical identification of isomers by comparison with characteristic-ion mass chromatograms of tri- and tetra- chlorobiphenyls (Fig.8). The results of calculation of the Henry I (rel.u.) 240 180 120 600 I (rel.u.) 300 240 180 120 600 30 Figure 7. Mass spectra of various impurities with identical molecular ions detected in the reaction mixture resulting from catalytic dimerisation of toluene.67, 88 Substituent position lnK1 lnK1 3.53 3.67 4.24 4.29 5.42 4.86 2,5,40 2,6,20 2,40 2,6,30 3,30 2,6,40 3.32 8.16 10.14 3.93 10.1 3.64 7.24 ab 70 50 90 110 130 150 170 190 m/z I (rel.u.) 0 I (rel.u.) 0 Figure 8.Molecular-ion mass chromatograms of trichloro- (m /z=256) and tetrachlorobiphenyls (m /z=292) obtained on a micropacked GTCB column (a) and on a capillary column (b).67, 89 constants for tetrachlorobiphenyl isomers were compared with this constant for 4,40-dichlorobiphenyl, present in the mixture and used as the standard. This comparison made it possible to find out that the retention times of all tetrachlorobiphenyls present in the mixture are shorter than that of the 4,40-substituted isomer, whereas the retention time of the `dioxin-like' 3,4,30,40-tetra- chlorobiphenyl is longer. Thus, the mixture under study does not contain the most toxic ortho-unsubstituted isomers.The gas chromatography ± mass spectrometry method considered in this review, which was used for mixture identification, appears to be the only reliable method known to date suitable to identify the belonging of isomers to a particular class of compound. Naturally, the examples cited do not cover all the possible fields of GC-MS application but they demonstrate that in combination with molecular-statistical calculations, this method provides the most reliable results. When discussing the semiempirical molec- ular-statistical theory, one should note one substantial limitation, namely, that the theory is applicable only to adsorption on the GTCB surface. Nevertheless, the experience accumulated for decades demonstrates high potency of this theory in complex situations, which often arise during chromatographic identifica- tion of organic compounds.References 1. R A Khmel'nitskii, G S Brodskii Khromato-Mass-Spektrometriya (Gas Chromatography/Mass Spectrometry) (Moscow: Khimiya, 1984) 2. J R Chapman Practical Organic Mass-Spectrometry (New York: Wiley-Interscience, 1985) 3. F W MacLafferty, in Proceedings of the 15th International Mass- Spectrometry Conference 2000, Barcelona, 2000 p. 12 4. S J Gaskell J. Mass Spectrom. 32 677 (1997) 5. P B Terent'ev, A P Stankyavichus Mass-Spektrometriya Biologi- cheski Aktivnykh Azotistykh Osnovanii (Mass Spectrometry of Biologically Active Nitrogenous Bases) (Vilnius: Mokslas, 1987) 6.G C Pimentel, J A Coonrod Opportunities in Chemistry Today and Tomorrow (Washington, DC: National Academy Press, 1987) a 20 10 b 15 10 5 * * * A K Buryak 292 256 30 t /min 292 256 20 t /minThe use of molecular-statistical methods for the calculation of thermodynamic characteristics of adsorption 7. G W Ewing Instrumental Methods of Chemical Analysis (New York: McGraw-Hill, 1985) 8. K I Sakodynskii, V V Brazhnikov, S A Volkov, V Yu Zel'venskii, E S Gankina V D Shatts Analiticheskaya Khromatografiya (Analytical Chromatography) (Moscow: Khimiya, 1993) 9. Yu S Drugov, A A Rodin Gazokhromatograficheskaya Identifika- tsiya Zagryaznenii Vozdukha, Vody i Pochvy (Gas Chromatographic Identification of Air, Water and Soil Pollution) (St Petersburg: Teza, 1999) 10.K A Gol'bert, M S Vigdergauz Vvedenie v Gazovuyu Khromatogra- fiyu (Introduction to Gas Chromatography) (Moscow: Khimiya, 1990) 11. E Leibnitz, H G Struppe (Eds) Handbuch der Gaschromatographie (Leipzig: Akademisch Verlagsgesellschaft Gesst und K G Porting, 1984) 12. Registry of Mass-Spectral Data P275, Wiley 13. N S Vul'fson, V G Zaikin, A I Mikaya Mass-Spektrometriya Organicheskikh Soedinenii (Mass Spectrometry of Organic Compounds) (Moscow: Khimiya, 1975) 14. V V Raznikov M O Raznikova Informatsionno-Analiticheskaya Mass-Spektrometriya (Informational Analytical Mass Spectrometry) (Moscow: Nauka, 1992) 15. I G Zenkevich, B V Ioffe Interpretatsiya Mass-Spektrov Organiche- skikh Soedinenii (Interpretation of Mass Spectra of Organic Compounds) (Leningrad: Khimiya, 1986) 16. M E Elyashberg Usp.Khim. 68 579 (1999) [Russ. Chem. Rev. 68 525 (1999)] 17. L A Gribov, V I Baranov, D Yu Zelentsov Elektronno-Koleba- tel'nye Spektry Mnogoatomnykh Molekul. Teoriya i Metody Rascheta (Electronic-Vibrational Spectra of Polyatomic Molecules. Theory and Calculation Methods) (Moscow: Nauka, 1997) 18. A K Buryak Izv. Acad. Nauk SSSR, Ser. Khim. 1995 (1990) a 19. M S Vigdergauz, L V Semenchenko, V A Ezrets, Yu N Bogoslovskii Kachestvennyi Gazokhromatograficheskii Analiz (Qualitative Gas Chromatography Analysis) (Moscow: Nauka, 1978) 20. I G Zenkevich Zh. Anal. Khim. 50 1048 (1995) b 21. I G Zenkevich Zh. Org. Khim. 30 1441 (1994) c 22.I G Zenkevich Zh. Fiz. Khim. 70 33 (1996) d 23. The Sadtler Standard GC Retention Index Library, Philadelphia, PA, 1986 Vol. 1 24. G Vernen,M Shanon (Eds)EVMPomogaet Khimii (Computers Help Chemistry) (Leningrad: Khimiya, 1990) 25. A A Lopatkin Teoreticheskie Osnovy Fizicheskoi Adsorbtsii (Theoretical Foundations of Physical Adsorption) (Moscow: Moscow State University, 1982) 26. A A Lopatkin Zh. Fiz. Khim. 63 2433 (1989) d 27. A A Lopatkin, in Fizicheskaya Khimiya. Sovremennye Problemy (Physical Chemistry. Current Problems) (Ed. Ya M Kolotyrkin) (Moscow: Khimiya, 1987) p. 5 28. A V Kiselev, A A Lopatkin, A A Shulga Zeolites 5 261 (1985) 29. N N Avgul', A V Kiselev, D P Poshkus Adsorbtsiya Gazov i Parov na Odnorodnykh Poverkhnostyakh (Adsorption of Gases and Vapours on Homogeneous Surfaces) (Moscow: Khimiya, 1975) 30.A V Kiselev, D P Poshkus, Ya I Yashin Molekulyarnye Osnovy Adsorbtsionnoi Khromatografii (Molecular Foundations of Adsorption Chromatography) (Moscow: Khimiya, 1986) 31. A K Buryak, A N Fedotov, A V Kiselev Vestn. Mosk. Univ., Ser 2, Khim. 26 568 (1985) e 32. A K Buryak, A V Ul'yanov, in Vserossiiskii Simpozium po Teorii i Praktike Khromatografii i Elektroforeza, Posvyashchennyi 95-letiyu Otkrytiya Khromatografii M S Tsvetom (All-Russian Symposium on Theory and Practice of Chromatography and Electrophoresis Dedicated to the 95th Anniversary of Discovery of Chromatography byM S Tsvet) (Samara: Samara University, 1999) p. 50 33. A Ya Afreimovich, Candidate Thesis in Chemical Sciences, Moscow State University, Moscow, 1972 34.A V Kiselev, D P Poshkus, A Ya Afreimovich Zh. Fiz. Khim. 42 2546 (1968) d 35. A V Kiselev, D P Poshkus, A Ya Afreimovich Zh. Fiz. Khim. 42 2553 (1968) d 36. A V Kiselev, D P Poshkus J. Chem. Soc., Faraday Trans. 2 72 950 (1976) 705 37. A V Kiselev, D P Poshkus, A J Grumadas J. Chem. Soc., Faraday Trans. 1 75 1281 (1979) 38. A V Kiselev, D L Markosian Chromatographia 17 526 (1983) 39. A V Kiselev, D L Markosyan Arm. Khim. Zh. 38 29 (1985) 40. M Gonnord, S Vidal-Madjar, G J Guiochon Chromatogr. Sci. 12 839 (1974) 41. S Vidal-Madjar, G J Guiochon, F Dondi J. Chrom. 291 1 (1984) 42. C Vidal-Madjar, E Bekassy-Molnar J. Phys. Chem. 88 232 (1984) 43. M Z Bobyleva, L A Dement'eva, A V Kiselev, N S Kulikov Dokl. Akad.Nauk SSSR 283 1390 (1985) f 44. A V Kiselev, P B Dallakyan Zh. Fiz. Khim. 59 1278 (1985) d 45. V I Nazarova, K D Shcherbakova Zh. Fiz. Khim. 73 2044 (1999) d 46. A K Buryak, D P Poshkus Izv. Akad. Nauk SSSR, Ser. Khim. 223 (1986) a 47. A K Buryak, P B Dallakyan, A V Kiselev Dokl. Akad. Nauk SSSR 282 350 (1985) f 48. A K Buryak, D P Poshkus Izv. Akad. Nauk SSSR, Ser. Khim. 12 (1989) a 49. V L Minkin, D A Osipov, Yu A Zhdanov Dipol'nye Momenty v Organicheskoi Khimii (Dipole Moments in Organic Chemistry) (Leningrad: Khimiya, 1967) 50. A K Buryak Izv. Akad. Nauk, Ser. Khim. 672 (1999) a 51. N S Kulikov Ads. Sci. Technol. 15 115 (1997) 52. A V Kiselev, E B Polotnyk, K D Sherbakova Chromatographia 14 478 (1981) 53.A V Kiselev, E B Polotnyuk, K D Shcherbakova Dokl. Akad. Nauk SSSR 266 892 (1982) f 54. E Yu Arkhipova, Candidate Thesis in Chemical Sciences, Moscow State University, Moscow,o1988 55. S N Yashkin, C B Kurbatova, D A Svetlov, A K Buryak Izv. Akad. Nauk, Ser. Khim. 731 (2000) a 56. A N Tikhonov, A V Gongarskii, V V Stepanov, A G Yagola Regulyarizuyushchie Algoritmy i Apriornaya Informatsiya (Regularising Algorithms and a Priory Information) (Moscow: Nauka, 1983) 57. B V Ioffe Refraktometricheskie Metody v Khimii (Refractometric Methods in Chemistry) (Leningrad: Khimiya, 1974) 58. A K Buryak, G I Berezin Izv. Akad. Nauk SSSR, Ser. Khim. 1721 (1989) a 59. A N Vereshchagin Polyarizuemost' Molekul (Polarisability of Molecules) (Moscow: Nauka, 1980) 60. I V Martynov, V I Martynov Zh. Vsesoyuz. Khim. O-va im. D I Mendeleeva 21 588 (1976) g 61. I V Martynov, V I Martynov Zh. Vsesoyuz. Khim. O-va im. D I Mendeleeva 21 536 (1976) g 62. A K Buryak Izv. Akad. Nauk, Ser. Khim. 1484 (1999) a 63. S N Yashkin, C B Kurbatova, A K Buryak Izv. Akad. Nauk, Ser. Khim. 792 (2001) a 64. S N Yashkin, O B Grigor'eva, A K Buryak Izv. Akad. Nauk, Ser. Khim. 938 (2001) a 65. C Yu Kudryashov, L A Onuchak, A V Voronkov, A K Buryak, I K Moiseev Izv. Akad. Nauk, Ser. Khim. 2021 (2000) a 66. A K Buryak Izv. Akad. Nauk, Ser. Khim. 681 (2000) a 67. A K Buryak, Doctoral Thesis in Chemical Sciences, Institute of Physical Chemistry, Russian Academy of Sciences, Moscow, 2000 68. E V Zagorevskaya, N V Ishenko, A V Kiselev, N V Kovaleva Ads. Sci. Technol. 2 212 (1985) 69. T R Rybolt,M T Epperson, H W Thomas, S E Clare, B M Manning, J T McClung J. Colloid Interface Sci. 173 202 (1995) 70. T R Rybolt, D L Logan,M W Milburn, H E Tomas, A B Waters J. Colloid Interface Sci. 220 148 (1999) 71. G I Berezin Dokl. Akad. Nauk SSSR 217 843 (1974) f 72. T R Rybolt, D L Hooper, J B Stensby, H E Tomas,M L Baker Jr J. Colloid Interface Sci. 234 168 (2001) 73. O R Meeks, T R Rybolt J. Colloid Interface Sci. 196 103 (1997) 74. C D Bruce, T R Rybolt, H E Tomas, T E Agnew, B S Davis J. Colloid Interface Sci. 194 448 (1997) 75. A M Dolgonosov Dokl. Akad. Nauk 338 39 (1994) f 76. A M Dolgonosov Zh. Fiz. Khim. 68 2187 (1994) d 77. A M Dolgonosov Zh. Fiz. Khim. 72 101 (1998) d 78. A M Dolgonosov Dokl. Akad. Nauk 358 355 (1998) f 79. A M Dolgonosov J. Phys. Chem. B 102 4715 (1998) 80. A M Dolgonosov Zh. Fiz. Khim. 75 391 (2001) d 81. A M Dolgonosov Zh. Fiz. Khim. 72 1281 (1998) dA K Buryak 706 82. A M Dolgonosov Dokl. Akad. Nauk 377 488 (2001) f 83. A M Dolgonosov Zh. Fiz. Khim. 76 346 (2002) d 84. A M Dolgonosov Zh. Fiz. Khim. 75 2002 (2001) d 85. A V Ul'yanov, Candidate Thesis in Chemical Sciences, Institute of Physical Chemistry, Russian Academy of Sciences, Moscow, 2000 86. A K Buryak, A V Ul'yanov, in Metody Fiziko-Khimicheskikh Issledovanii (Methods of Physicochemical Investigations) (Moscow: NIITEKhIM, 1990) 87. A K Buryak, A V Ul'yanov Izv. Akad. Nauk SSSR, Ser. Khim. 623 (1996) a 88. A K Buryak, A V Uleanov Balaton Symposium 099 on High- Performance Separation Methods, Siofok, Hungary, 1999 p. P-10 89. A K Buryak, A V Ul'yanov, in Tezisy Dokladov Mezhdunarodnogo Simpoziuma `Khromatografiya i Mass-spektrometriya v Analize Ob'ektov Okruzhayushchei Sredy' (Abstracts of Reports of International Symposium `Chromatography and Mass Spectrometry in Analysis of the Environment) (St Petersburg: Teza, 1994) p. 134 a�Russ. Chem. Bull., Int. Ed. (Engl. Transl.) b�Russ. J. Anal. Chem. (Engl. Transl.) c�Russ. J. Org. Chem. (Engl. Transl.) d�Russ. J. Phys. Chem. (Engl. Transl.) e�Moscow Univ. Bull. (Engl. Transl.) f�Dokl. Chem. (Engl. Transl.) g�Mendeleev Chem. J. (
ISSN:0036-021X
出版商:RSC
年代:2002
数据来源: RSC
|
|