摘要:

Chemical Society Reviews Editorial Board Professor H. W. Kroto FRS (Chairman) Professor M. J. Blandamer Dr. A. R. Butler Professor E. C. Constable Dr. T. C. Gallagher Professor D. M. P. Mingos FRS Professor J. F. Stoddart FRS Consulting Editors Dr. G. G. Balint-Kurti Professor S. A. Benner Dr. J. M. Brown Dr. J. Burgess Dr. N. Cape Professor B. T. Golding Professor M. Green Professor A. Hamnett Dr. T. M. Herrington Professor R. Hillman Professor R. Keese Dr. T. H. Lilley Dr. H. Maskill Professor A. de Meijere Professor J. N. Miller Professor S. M. Roberts Professor B.H. Robinson Professor M. R. Smyth Dr. A. J. Stace Staff Editor Mr. K. J. Wilkinson University of Sussex University of Leicester University of St.Andrews University of Basel, Switzerland University of Bristol Imperial College London University of Birmingham University of Bristol Swiss Federal Institute of Technology, Zurich, Switzerland University of Oxford University of Leicester Institute of Terrestrial Ecology, Lothian University of Newcastle upon Tyne University of Bath University of Newcastle upon Tyne University of Reading University of Leicester University of Bern, Switzerland University of Sheffield University of Newcastle upon Tyne University of Gottingen, Germany Loughborough University of Technology University of Exeter University of East Anglia Dublin City University, Republic of Ireland University of Sussex Royal Society of Chemistry, Cambridge It is intended that Chemical Society Reviews will have the broad appeal necessary for researchers to benefit from an awareness of advances in areas outside their own specialities.Deliberate efforts will be made to solicit authors and articles from Europe which present a truly international outlook on the major advances in a wide range of chemical areas. It is hoped that it will be particularly stimulating and instructive for students planning a career in research. The articles will be succinct and authoritative overviews of timely topics in modern chemistry. In line with the above, review articles will not be overly comprehensive, detailed, or heavily referenced (ca. 30 references), but should act as a springboard to further reading.In general, authors, who will be recognized experts in their fields, will be asked to place any of their own work in the wider context. Review articles must be short, around 8-1 0 journal pages in extent. In consequence, manuscripts should not exceed 20-30 A4/American quarto sheets, this length to include text (in double line spacing), tables, references, and artwork. An Information to Authors leaflet is available from the Senior Editor (Reviews). Although the majority of articles are intended to be specially commissioned, the Society always considers offers of articles for publication. In such cases a short synopsis (including a selection of the literature references that will be cited in the review and a brief academic CV of the author), rather than the completed article, should be submitted to the Senior Editor (Reviews), Books and Reviews Department, The Royal Society of Chemistry, Thomas Graham House, Science Park, Milton Road, Cambridge CB4 4WF. @ The Royal Society of Chemistry, 1994 All Rights Reserved ’ No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form, or by any means, electronic or mechanical, photographic, recording, or otherwise, without the prior permission of the publishers. Typeset by Servis Filmsetting Ltd. Printed in Great Britain by Blackbear Press Ltd.

ISSN:0306-0012    

DOI:10.1039/CS99423FX001

出版商:RSC    

年代:1994

数据来源: RSC

ISSN:0306-0012    

DOI:10.1039/CS99423BX003

出版商:RSC    

年代:1994

数据来源: RSC

摘要:

Helical Poly(isocyanides) Roeland J. M. Nolte Department of Organic Chemistry NSR- Centre University of Nijmegen 6525 ED Nijmegen The Netherlands 1 Introduction One of the most intriguing dissymmetric shapes is the helix. It was developed in natural systems in the early stages of evolution and used as the structural motif for the molecules of life (DNA and RNA) and as an important conformational element that enforces long range order in other biomacromolecules e.g. enzymes. Helicity received attention in organic chemistry after the discovery of chirality at the end of the last century but molecules with extended helical structures have been described only recently. Examples are the copper phenanthroline-based helices reported by Lehn and the self-organized quadruple helices from amphiphilic molecules synthesized by Fuhrhop. In polymer chemistry helical architectures have been studied since the pioneering work of Natta Pino and others.2 Most isotactic polymers exist as short range helices in solution. These are dynamic rather than static structures and the direction of the helical twist is very sensitive to small changes in polymer side- chain structure and the type of solvent2 Polymers that maintain stable helical structures in solution as do bio-macromolecules are very rare but are of great interest since they can display optical activity due solely to main-chain conformation. Furthermore they may be used as versatile building blocks for the construction of novel chiral supramole- cular architectures. To date three examples of such helical polymers are known viz. polymers of isocyanides (1),3 poly (chloral) (2),4 and poly(methacry1ate esters) e.g. (3). Following suggestions by Millich6 we were able to demon-strate in 1974 that poly(t-butylisocyanide) (1; R = t-C,H,) can be resolved into left-handed and right-handed helices which do not racemize even at elevated temperatures. Subsequent studies have provided procedures for preparing helical poly(isocya- nides) by helix-sense selective polymerization and have given y3 (>C=N) \R (-CH,-O-) CCI3 I (-CH2-?-)” C=O ? Ph-C -Ph I Ph (3) Roeland J. M. Nolte is Pro- fessor of Organic Chemistry at the University of Nijrnegen The Netherlands. He was educated at the University of Utrecht (Ph.D. in 1973) and did post-doctoral work with D. J. Cram at U.C.L.A. His research interests include mole- cular recognition electron con- ducting molecular materials chiral polymers and metal-catalysed epoxidations. 11 insight into the mechanism of the polymerization rea~tion.~ The helical structure of a poIy(isocyanide) is the result of restricted rotation around the single bonds connecting the main carbon atoms (atropisomerism). A similar hindered rotation is observed in poly(chlora1) viz. around its carbon-oxygen bonds. This polymer can be prepared in a stable optically active form by anionic polymerization with e.g. Li-/3-cholestanoxide. Triphe- nylmethyl methacrylate and related bulky methacrylate esters yield optically active polymers when they are polymerized with initiators such as Li-(R)-( 1-phenylethyl)anilide or (-)-spar-teine-butyl lithium. The polymers are highly isotactic . The chiral carbon atoms in the polymer chains do not contribute to the optical activity as they are pseudo-chiral. The observed large optical rotations of the polymers are attributed to the presence of helical superstructures which again are the result of hindered rotation around single bonds The purpose of the present review is to highlight the most interesting features of helical poly(isocyanides). It may hope- fully serve as a starting point for further research in the field of atropisomeric polymers in particular with regard to the use of these polymers as building blocks for the construction of novel dissymmetric architectures similar to those found in nature. 2 Synthesis Millich was the first to develop a catalytic system for the polymerization of isocyanides.6 He used an acid-coated glass system in combination with a radical initiator or air. We discovered that simple nickel(I1) salts (NiC12.6H20; Ni(Acac) HAcac = acetylacetone) are very efficient catalysts for polymer- izing a wide variety of aliphatic and aromatic isocyanides including monomers with additional double or triple bonds with metal ligating functions with crown ether rings with donor-acceptor groups and with stable radical substituents [see (4)-( 16)].3,7The reactions should be carried out under aerobic conditions to prevent the formation of less reactive nickel(1) species as was shown by Novak.s Yields are moderate to excellent depending on the type of isocyanide. Molecular weights vary between 5000 and 250000. Interesting starting materials for the preparation of the monomers are amino acids and peptides. Their amino functions can be easily converted into isocyano functions as shown in Scheme 1. Polymerization with nickel chloride yields a new type of helical poly (amino acid) [e.g. COOR’ I COOR’ / H I H I. IR-Cc*-fkC- [R-C-N=C<J,I COOR‘ COOH Scheme I [>C=NxN-0.1 (7) 00H II H H II H H [>C =N-C-C-N-C-C-N-C-COOH] I I I (14) (L-Ala,L-His,L-Ser) see compound (14)] whose properties as enzyme mimics have been explored (see Section 6.4). Novak has recently reported that nickel complex (17) is a very efficient catalyst for the polymerization of is~cyanides.~.~ Even the bulky unreactive t-butyl isocyanide can be quantitatively polymerized by (1 7) within a few hours. Polymerizations with (17) unlike with nickel(@ chloride are living in nature2 and are characterized by narrow molecular weight distributions. As (17) is also a catalyst for the polymerization of butadiene to cis-1,4-~oly(butadiene) block copolymers can be prepared containing rigid helical poly(isocyanide) and elastomeric poly(butadiene) segments (see Scheme 2).9 Ito has found that palladium(r1) complexes [e.g.MePd(PPh-Me,),Br] catalyse the polymerization of 1,2-diisocyanoarenes to helical poly(2,3-quinoxalines) [see compound (15)]. O This pro- cedure offers a new way to prepare poly(heter0-aromatics). Very interestingly polymerization of 1,2-diisocyanobenzene with nickel(I1) chloride yields a polymer in which only one isocyano function is incorporated in the polymer backbone [see corn- pound (16)]. (9) (SiMe [>C=N \ /In0 CF3I CF3 (17) 3 Structure Space-filling molecular models indicate that a poly(isocyanide) molecule cannot adopt a planar structure. The unusual feature that each main-chain carbon atom carries a side-chain causes severe steric hindrance resulting in restricted rotation around the single bonds of the polymer backbone. As in the case of low molecular weight atropisomeric compounds two configu- HELICAL POLY(ISOCYAN1DES)-R J M NOLTE - m RNC Nt-catalyst H30+ MeOH R \ n block copolymer Scheme 2 rations are possible around each of the main chain single bonds viz R or S If these configurations are the same for all single bonds (meaning that the polymer is highly stereoregular or isotactic) a stable helix is formed This helix is right-handed (P)if the above-mentioned configurations are all S and left-handed (M) if they are all R (see Figure 1) Alternating R and S absolute configurations would lead to a zigzag (or syndiotactic) struc- ture However for steric reasons such an arrangement is not feasible in practice The following data support the helical structure of poly(isocyanides) Polymers of optically active isocyanides display optical rotations that are very different from those of the corresponding monomers or low-molecular weight model compounds This feature already observed by Millich6 in poly(a-phenylethyl isocyanide) and later confirmed by us in extensive studies,] suggests that an additional chiral element is present in the polymers In the circular dichroism spectra of these optically active polymers so-called exciton couplets -indicative of helices -are visible in the region from 240400 nm m A& Figure 1 Right-hanaed (A) and left-handed (B) poly(isocyanide) helices and their corresponding circular dichroism spectra (C and D respectively) These couplets are due to the n-zr* transitions of the imino chromophores in the polymer main chains In the UV-vis spectra these transitions are present as weak bands at approxi- mately 300 nm Circular dichroism calculations carried out by Huigel suggest that in the case of right-handed poly(isocya- nide) helices the couplets are 2-shaped and in the case of the left- handed helices they are S-shaped (see Figure 1 C and D) Definitive proof that polymers of isocyanides can adopt helical structures comes from resolution experiments Poly(t-butyl iso- cyanide) which has no chiral centres could be completely resolved into (+)-and (-)-rotating polymer fractions with the help of a chiral column consisting of glass beads coated with an insoluble high molecular weight polymer of optically active (9-s-butyl isocyanide The (+)-and (-)-rotating fractions were assigned to left- and right-handed helices respectively on the basis of a comparison between experimentally determined and calculated circular dichroism spectra Further support for a helical structure is provided by molecular orbital and molecular mechanics calculations Using the extended Huckel theory Kollmar and Hoffmannl showed that repulsion between the lone pairs of the imino groups in a poly(isocyanide) chain favours a departure from a planar structure Similar repulsive effects are known to be operative in poly(ketones) Calculations on the series of polymers (H-N=C() (CH,-N=C() and ((CH,),C-N=C() indicated that steric effects are more im- portant than electronic effects in determining the polymer structure For the hypothetical polymer (H-N=C() a broad range of helical conformations are available in contrast to ((CH,),C-N=C() for which the authors propose a quite stiff 1 helix According to the theoretical analysis the polymer with intermediate bulk (CH,-N=C() may adopt two helical struc- tures with different degrees of helicity Huige and Hezemans' have performed extensive molecular mechanics calculations using the consistent force-field method on various oligo- and poly(isocyanides) The hexadecamer of t-butyl isocyanide was calculated to have a helical middle section and disordered end sections The dihedral angle N=C-C=N in this middle section was found to be 78 6"and the number of repeat units units per helical turn was 3 75 The latter number is in agreement with circular dichroism calculations using Tinoco's exciton theory (3 fb-4 6) and De Voe's polarizability theory (3 81) l4 The molecular mechanics calculations further revealed that the less bulky polymers (1-C,H,-N=C() and (C,H,-N=C() form helical structures as well The polymer of methyl isocyanide was calculated to be disordered Another stereochemical feature of interest is syn-anti isomer-ism about the C=N double bonds of the monomeric units Molecular mechanics calculations suggest that the occurrence of both syn and anti structures in one polymer chain is energetically unfavourable However H- and ,C-NMR experiments indi- cate that the polymers very often display more than one signal for certain H or C atoms For instance the polymer of 4-methoxyphenyl isocyanide has two methoxy signals in the 'H- NMR spectrum at 6 3 25 and 3 75 ppm Addition of acid results in a decrease of the former signal and an increase of the latter one suggesting an equilibrium of the type shown in Figure 2 /" "\ R/ 'R (4 (b) Figure 2 syn-anti Isomerism in poly(isocyan1des) Random orientation of side chains (a) and the thermodynamically more favourable all syn- or all-anti configuration (which are identical) which is formed after the addition of acid (b) The exponents in the Mark-Houwink equations measured for poly(isocyanides) (e g a in [q]= K Mais 1 75 for poly(2-octyl- isocyanide)) suggest that the individual polymer molecules are rigid rods This conclusion however has been seriously ques- tioned by Green According to him the polymer chains contain defects which cause their persistence length to be relatively small l8 4 Polymerization Mechanism The polymerization of isocyanides proceeds very rapidly at room temperature -which is remarkable given the fact that so much steric bulk is introduced when the polymer chains are formed The driving force for the process undoubtedly is the conversion of a formal divalent carbon in the monomer into a tetravalent carbon in the polymer The heat of polymerization is considerable viz 81 4 kJ mo1-l l2 For the polymerization to occur at least two vacant cis-coordination sites are required at the nickel centre Thus the complex trans-Ni(Acac) (RNC) (1 8) is unable to start the polymerization reaction whereas (19) which is generated from (1 8) by acid is an effective catalyst H+Ni(Acac) (RNC) ====Ni(Acac)(RNC) + Acac-(18) (19) Recently Novak has presented similar evidence for the catalyst (17) whose activity can be blocked by the addition of cyanide ions l9 For the polymerization catalysed by nickel(I1) chloride a so-called merry-ga-round mechanism has been proposed This mechanism is based on kinetic measurements and experiments with optically active isocyanide monomers First step in the reaction is the formation of a square-planar nickel complex (20) (Scheme 3) Polymerization starts when a nucleophile attacks one of the coordinated isocyanide molecules of (20) This nucleophile may already be present in solution (e g a molecule of water or the counterion of Ni2 +),or may have been intention- ally added eg an amine R'NH Spectroscopic studies indicate that the nucleophile first coordinates to the nickel centre (21) and subsequently migrates to form complex (22) The plane of the carbene ligand C'(NHR')NHR in (22) is approximately perpendicular to the plane of the isocyanide carbon atoms and nickel The first C-C bond is formed when the carbon atom C1 of the carbene ligand attacks one of its neighbours C2NR or C4NR This process IS facilitated by the coordination of a new isocyanide CsNR Attack can occur on either C2or C4 and for an achiral amine or achiral isocyanide the chances for these attacks are equal In (23) attack has occurred on C2 When the reaction sequence continues in the direction C' C2 C3 C4 a left- handed helix is generated The opposite sequence will lead to a right-handed helix Each rotation around nickel adds one turn to the chain The reaction stops when the living chain-end accepts a proton Because the monomers are preorganized in the nickel complex only a slight rearrangement of bonds is required to form the polymer molecule This may explain why the tightly coiled helix is so easily formed 5 He1 ix-sense-select ive Polymerization As discussed in Section 3 polymers with an excess of one type of helix can be prepared by chromatographic resolution using an optically active column This method has proven to be successful for poly(t-butylisocyanide) but has turned out to be less appli- cable to other polymers of isocyanides Therefore other pro- cedures for preparing optically active helical poly(isocyanides) have been developed They are based on the mechanism of the polymerization reaction and will be discussed below 5.1 Chiral Isocyanides Incorporation of a homo-chiral isocyanide into a left-handed or a right-handed helix leads to different polymer species which are CHEMICAL SOCIETY REVIEWS 1994 R'NH I c -NHR' / 1*+ f(c>U Scheme 3 Table 1 Diastereoselective polymerization of chiral isocyanides with NiCl Polymer R* in R*NC [M]:y,(deg) Helix sense (R)-(CH,),CHCHzCH(CH3) +689 M (S)-CH,CO,CHzCH(CH,) -190 P (s)-cZH50c(0)(cH3) -355 P (S)-PhCH(CH,) -458 P (S)-H,C=CHCH(CzH,) +276 M (Ss)-ImCH,CH(COOH)NHCOCH(CH,) + 1025 M (Im = 4(5)-imidazolyl) diastereomers It is to be expected that these diastereomers are formed in unequal amounts A large number of optically active isocyanides have been synthesized and polymerized to test this hypothesis Some results are compiled in Table 1 The polymers display optical rotations which are different from those of the monomers and of model compounds such as the imines R*-N=CH-t-C,H9 The molar optical rotations in Table I are the sum of contributions from the polymers helices and the polymer side-chains The latter can be estimated from the model compounds For instance for R* = (9-PhCH(CH,) the difference between the optical rotation of the polymer (-458") and that of the model compound (-126") is negative (-332") suggesting that the contribution from the helix is laevorotatory By applying the relationship found for poly(t- HELICAL POLY(ISOCYAN1DES)-R J M NOLTE 3-c butylisocyanide) vzz an M-helix gives rise to a (+)-sign of optical rotation and a P-helix to a (-)-sign one can conclude that on polymerization (9-PhCH(CH,)NC preferentially forms a right-handed helix I e the polymerization reaction proceeds diastereoselectively A similar analysis has been carried out with the help of circular dichroism spectra in which case the shape of the exciton couplet of the optically active polymer is compared with the shapes of the couplet of (P)-and (M)-poly(t- butylisocyanide) In a few cases it has been possible to determine the degree of chiral induction vcz in the case of isocyanides R(CH,)CHNC (R = ethy1,n-hexy1,i-propy1,i-butyl)The high- est d e was measured for R = 1-propyl and amounted to 62% Novak and an associate have investigated whether the poly- merization of racemic mixtures of chiral isocyanides proceeds stereoselectivity z e whether the racemic monomer mixture is resolved into all-R and all-S polymer molecules l9 To this end enantiomerically pure (R)-or (3-PhCH(CH,)NC was polymer- ized with catalyst (1 7) The polymer subsequently produced displayed a narrow polydispersity index (PDI = MJM = 1 1) In contrast when the racemic monomer mixture was polymer- ized the polymer molecular weight distribution was found to be quite broad (PDI = 1 6-1 8) This result as well as the results from kinetic experiments and experiments with 3C-enriched isocyanide monomers were taken as evidence that the polymeri- zation of (R,S)-mixtures is only partly stereoselective 4 ,.c X 0 2+ /’/ /I Ni -c ’ L Figure 3 Model for predicting the helix sense in the polymerization of chiral isocyanides The mechanism of the polymerization reaction allows one to predict which helix is generated from an optically active isocya- nide In Figure 3 one of the first intermediates in the polymeriza- tion reaction is shown The nucleophile is denoted by X and the substituents on the chiral carbon atom are ranked according to their sizes vzz S (small) M (medium) and L (large) As mentioned in Section 4 the plane of ligand C’ is perpendicular to the nickel plane The substituent L is placed in such a way that the least steric hindrance occurs in the transition state of the first carbon-carbon bond formation reaction i e away from the nickel centre As depicted in Figure 3 attack will occur on C4 as this is the least sterically hindered side In this way a right- handed helix is formed This kind of reasoning has been applied to approximately 20 optically active monomers In most cases the predicted helix sense was found to be in agreement with the helix sense derived from optical rotation data and circular dichroism spectra From this result one may conclude that the process of helix selection takes place at the catalytic centre 5.2 Achiral Isocyanides Helix-sense-selective polymerization of achiral isocyanides has been achieved with optically active catalysts of type (24) Poly- merization of the sterically hindered t-butylisocyanide with (R)-(24) yielded a polymer with an excess of left-handed helices Likewise (S)-(24) gave a polymer with an excess of right-handed helices The e e values were estimated to be in the range of 45- 70% Less sterically encumbered monomers eg p-methoxy-R phenyl isocyanide were converted into optically inactive polymers by (R)-and (s)-(24) l9 Addition of an optically active initiator to a nickel(1r)-isocyanide complex generates a catalyst which can polymerize achiral isocyanides to optically active polymers Some results obtained for the polymerization of t-butyl isocyanide using Ni(CNR),(ClO,) and different chiral amines are presented in Table 2 2o Polymer samples displaying e e values up to 85% have been obtained by this procedure The helix sense that is induced by the chiral initiator can be predicted from the polymerization mechanism by using similar reasoning as des- cribed for the polymerization of chiral isocyanides (see Section 5 1) Table 2 Enantioselective polymerization of t-butyl isocyanide with Ni(CNR),(ClO,) and chiral ini tiatorsa Initiator e e (%) Helix sense (S)-C,H,CH(CH,)NH 7 P L-Prolinol 36 M ~-Alaninemethyl ester 47 P (S)-PhCH(CH,)NH 61 P (S)-PhCH(CH ,)NH Zh 83 P R = t C,H R = 2(t C,H,)C,H A third procedure for preparing optically active polymers from achiral isocyanides involves the use of bulky optically active co-monomers eg esters of (S)-2-isocyanoisovalericacid [(q-1-C,H,CH(COOR)NC (25)] In the presence of nickel(1r) salts these isocyanides slowly polymerize to give homopolymers with predominantly left-handed helices When an isocyanide (25) is mixed with an achiral isocyanide (see Table 3) and subsequently polymerized with nickel chloride polymer samples are obtained which consist mainly of the homopolymer of the achiral isocyanide This homopolymer has a high optical rotation and a helix sense opposite to that of the chiral co- monomer (Table 3 and Figure 4) The mechanism of this unusual reaction is probably as follows The bulky chiral isocyanide (25) is a slowly polymerizing monomer and forms an M-helix (Figure 5A) The achiral isocyanide is a fast-polymeriz- ing monomer and yields a racemic mixture of P and M helices (Figure 5B) When the chiral isocyanide is co-polymerized with the achiral one the former has a preference for inclusion into M-Table 3 Enantioselective polymerization of achiral isocyanides in the presence of optically active isocyanide(25) R in (RNC) -[aItl0tdeg) Helix sense Mw 44 000 42 000 35000 45 000 CHEMICAL SOCIETY REVIEWS 1994 AE/(L.mol-' cm-') AEX10 250 Unm 500 Figure 4 CD spectrum of the homopolymer from (25) (curve A) and the polymer obtained from 4-methoxyphenylisocyanide (curve B). Curve A reveals the presence of a left-handed helix curve B the presence of a right-handed helix. slowA MR*NC -999-M P M Figure 5 Mechanism of the enantioselective polymerization of achiral isocyanides in the presence of an optically active co-monomer. helices and retards the formation of these helices from the latter one. The P-helices continue to grow and eventually consume all the achiral monomer (Figure 5C). Ito and co-workers have described the enantioselective poly- merization of 1,2-diisocyanoarenes with optically active palla- dium(1r) catalysts to give helical poly(2,3-quinoxalines) e.g. (26). These polymers displayed high optical rotations and large Cotton effects in their CD spectra suggesting that they have helical conformations of one particular sense.* 6 Biomimetic Macromolecular Chemistry For a long time scientists have been fascinated by enzymes and intrigued by the fact that these biopolymers can fulfil so many (343I functions and catalyse such a diversity of reactions. Unravelling the principles that underlie the action of enzymes continues to be a challenge for chemists and biochemists and impressive progress towards this goal has been made in recent years. Much work is focused on the biopolymers themselves but also on low molecular weight model compounds which can give insight into the details of the catalytic processes involved. Comparatively little attention has been given to synthetic polymers as biomo- dels which is surprising given the fact that many molecules and molecular systems in nature have macromolecular dimensions. In this section some applications of polymers of isocyanides in the field of biomimetic chemistry will be discussed. These polymers are versatile building blocks for the construction of biomodels because they (i) are helical (ii) have a well-defined structure and (iii) can be synthesized in great variety from easily accessible amines and amino acids. 6.1 Bilirubin Binding Bilirubin (27) is the end product of heme catabolism in man and most animals and is ultimately excreted in bile.23 The compound has a low solubility in aqueous media. Almost all of the bilirubin transported in the blood is tightly bound to serum albumin. The nature of this binding is unknown. Salt bridges with charged residues of histidine and other amino acids as well as hydrogen bonds have been thought to be involved in the binding process. Two polymers of isocyanides poly(carby1histidine) (28) and poly(carby1histamine) (29) have been synthesized and used to study the type of binding interactions with bilir~bin.~~ These poly(isocyanides) have appreciably different pK(1m) values pKd(28)= 9.4 and pKd(29) = 5.2. Addition of bilirubin at the pH of blood plasma (pH = 7.3) to poly(carby1histidine) and poly(carby1histamine) leads to complex formation with the latter polymer but not with the former one. Bilirubin is an unstable compound but complexed to (29) it can be kept for prolonged periods of time. Spectroscopic and other studies indicate that approximately one molecule of bilirubin is bound per helical turn of poly(carby1histamine). At pH 7.3 bilirubin is a dianion and the observation that this dianion binds to the neutral (29) and not to the protonated (28) suggests that in the case of serum albumin salt bridges to charged histidine residues do not play a major role. Instead hydrogen bonding interac- tions are more likely to be operative viz. between the neutral imidazole groups and the lactam rings of bilirubin. 6.2 Artificial Ion Channel The unassisted transport of ions through cell membranes is very slow the reported permeation coefficient for Kf is 7 x cm~-'.~~Generally there are two ways by which nature facili- HoocT H H H H r ,CH,-CH-N=Cc 1 CH3 (28)R=COOH (26) (29) R= H HELICAL POLY(ISOCYAN1DES)-R. J. M. NOLTE tates ion transport across a bilayer membrane.25 One way is by carrier molecules e.g. the antibiotics Valinomycin and Nonac- tin. A second more frequently encountered mode is by proteins that form a transmembrane ion channel the archetype being Gramicidin A. Several attempts have been made to design and synthesize artificial systems that mimic the latter mode of ion transport. Notable are the studies by Fuhrhop Fyles and Lehn. We have been interested in constructing artificial ion channels by stacking ring-like molecules e.g.crown ether rings. The main problem is how to interconnect these macrocyclic riags. Piling them stepwise via lateral appendages is very difficult to achieve. We have solved this problem by anchoring the crown ether rings to a rigid poly(isocyanide) upp port.^ Isocyanides containing crown ether rings of different sizes were synthesized and poly- merized with nickel chloride to give polymers of type (1 1). As a result of the 1,-helical structure of the polymer backbone the crown ether rings in (1 1) are positioned on top of each other and form 4 channels which run parallel to the polymer helix axis (see Figure 6). The molecular weight of the polymers amounted to 4000-20000 which corresponds to channel lengths of 10 to 50 A. The metal ion binding properties of the channels were found to be greater than those of low-molecular-weight model com- pounds. This difference is explained by the fact that in the channels the metal ions can be sandwiched between consecutive crown ether rings which favours binding. In the model com- pounds such sandwiching is not possible. (B) Figure 6 Calculated structure of a poly(a-phenylethylisocyanide) with 18-crown-6 rings top view (A) and side view (B). (Reproduced with permission from reference 276). The ion channels were incorporated into bilayers of dihexade- cyl phosphate (DHP) vesicles and ion' transport across the vesicle bilayer was studied in the following way (Figure 7A). The dye 4-(2-pyridylazo)resorcinol mono sodium salt (PAR) was occluded in the inner aqueous compartment of the vesicles. This dye forms coloured complexes with cobalt@) ions. These ions were added to the vesicle dispersions and the increase in the absorption of the cobalt-PAR complex in the UV-vis was recorded as a function of time. In the presence of channels transport of cobalt ions was observed without channels this transport was very small or absent (Figure 7B). Permeability coefficients 'were determined to be in the range of 5 x 1Olo cms-l. The activation energy for cobalt ion transport was calculated from experiments carried out at different tempera- tures and amounted to E = 24 kJmol-l. This number is consistent with a pore mechanism for the ion translocation process (Figure 7A). A very similar value of E has been found for Gramicidin A (E = 20.5-22.5 kJ mol- l). Figure 7 Facilitated transport of cobalt ions across bilayers of dihexade- cyl phosphate (DHP) vesicles by channel compound (1 1) (A).Plots of the change in absorbance of the cobalt (11)-PAR complex at 5 10 nm versus time for vesicles with (upper curve) and without (lower curve) channel compound (1 1).The arrows indicate the addition of a reagent (Triton X-100) which destroys the vesicles (B). (Reproduced with permission from references 27a and 276.) 6.3 Cytochrome P-450 Mimic The Cytochrome P-450-dependent mono-oxygenases are mem- brane-bound enzymes which catalyse a great variety of reac- tions among which is the epoxidation of alkenes by molecular oxygen.,* The active centre of the enzymes contains an iron(Ir1) protoporphyrin IX and an axial ligand. After being reduced to iron@) this centre binds and cleaves molecular oxygen where- upon water and a high-valent iron-oxo complex are formed. The latter species transfers its oxygen to the substrate molecule. The electrons required in the process are provided by the cofactor NADPH via a coupled electron-transferring enzyme system. +02+2H++2e -yLy+H20 We have developed a synthetic model of Cytochrome P-450 which incorporates all the features of the natural The most important part is a microreactor of stabilized vesicles which holds the components of the catalytic system viz. a bilayer-bound manganese(m) porphyrin an electron donor colloidal Pt (incorporated in the interior of the vesicles) and H and an electron carrier (methylene blue) which shuttles elec- trons from the colloidal Pt to the metalloporphyrin (Figure 8). Figure 8 A membrane-bound Cytochrome P-450 mimic based on polymerized vesicles from isocyanosurfactant (30). Pt is colloidal platinum MBoxand MBredare the oxidized and reduced forms of the electron carrier methylene blue and MnP is manganese(II1) porph yrin. CHEMICAL SOCIETY REVIEWS 1994 H3C,+ CH N' Br-1 structure of cross links "'C-'C:"- C5 under C' ,etc I Scheme 4 The microreactor was constructed from the isocyano-amphi- phile (30) (Scheme 4) which was synthesized in the four steps from dimethylhexadecylamine 1 1 -bromoundecanol and L-ala- nine. On dispersal in water (30) forms closed vesicles with diameters of approximately 250 nm. These vesicles can be stabilized by polymerization of the isocyano functions in the bilayers with nickel capr~nate.~~ The polymerized vesicles retain their structure -as was shown by electron microscopy osmotic experiments and fluorescent techniques. The degree of polymer-ization of the isocyano surfactants within the bilayer was estimated to be approximately 75. Interestingly freeze-fracture electron micrographs of the polymerized vesicles of (30) provided direct evidence that the bilayer halves were cross- linked instead of the usual pattern of concave and convex half spheres observed for non-polymerized systems (Figure 9A) circles and ellipses were visible (Figure 9B). An explanation of the latter phenomenon is given in reference 29. The membrane-bound Cytochrome P-450 mimic was shown to epoxidize water-soluble (2,5-dihydrofuran) as well as water- insoluble (styrene) alkenes at room temperature with molecular oxygen as the oxidant. Turnover numbers are in the range of I .5-8 mol alkene/mol catalyst.h which is one hundredth of the activity of the natural enzyme system. 6.4 Protease Mimics The development of novel catalysts based on enzymes is cur- rently a topic of great interest. Many studies are dealing with catalytic systems mimicking pro tease^.^ O The reason for this choice is that the principles of protease action exemplified by Chymotrypsin and Elastase have been studied thoroughly and are now starting to be under~tood.~~ For the construction of an artificial proteolytic catalyst the following features are thought to be required (i) a nucleophile and a proton-transfer system organized in such a way to complement the structure of the substrate (ii) a water-soluble chiral platform to anchor the catalytic functions and to provide a substrate binding site and (iii) a hydrophobic microenvironment to mimic the hydropho- bic interior of the protease. We have synthesized a large series of optically active polymers of isocyanides containing imidazolyl carboxylic acid and hyd- roxymethyl functions in their side chains. These functions are also present in Chymotrypsin ('Charge relay system' feature i). The polymers were prepared in two different ways by homo- polymerization of isocyanides derived from alanylhistidylserine tripeptides (e.g. polymer (14)) and by copolymerization of isocyanides derived from the dipeptides alanylserine and alanyl- histidine e.g.(31). ()C=N-L- Ala-L-His-L-Ser) ()C=N-L-Ala-L-His) ()C=N-L-Ala-L-Ser) 1 Figure 9 Freeze-fracture electron micrographs of unpolymerized (A) and polymerized vesicles (B) from isocyanosurfactant (30). (Reproduced with permission from reference 29b). Polymers of type (14) and (31) are soluble in water and have an excess of one helix sense (feature ii).31 They were used as catalysts in the hydrolysis of achiral and chiral nitrophenyl esters. Extensive kinetic studies revealed that the hydroxymethyl functions had no appreciable effect on the catalysis. The homo- polymers and copolymers showed markedly higher activities than the corresponding low-molecular weight compounds. This enhancement was ascribed to cooperative effects involving interactions between imidazolyl groups and neighbouring imidazolyl and carboxylate groups (Scheme 5). The activities could be further enhanced by adding positively charged surfac- tants e.g.cetylpyridinium chloride or cetylundecyldimethylam- monium bromide. Negatively charged surfactants did not show any effect. Positively charged surfactants arrange themselves around the negatively charged polymer molecules (Figure 10) and in this way create a hydrophobic pseudophase which is favourable for catalysis (feature iii) viz. by changing the pK values of the imidazolyl groups and by increasing the concent- ration of substrate molecules in the vicinity of these groups.31 The polymer-surfactant complexes displayed small enantio- selectivities (kL/kD= 3) in the hydrolysis of chiral amino acid esters. Higher enantioselectivity ratios were obtained when polymerized isocyanosurfactants of type (10) were combined with free (i.e. not polymer-bound) tripeptide catalysts. In this HELICAL POLY(ISOCYAN1DES)-R J M NOLTE Scheme 5 Figure 10 Structure of the complex between a negatively charged poly(isocyanide) with tripeptide side chains and positively charged surfactant molecules case the enantioselectivity ratios amounted to kL/kD= 33 31 Apparently a polymer-anchored surfactant in combination with a free peptide catalyst is more effective in accomplishing enantioselectivity in the ester hydrolysis reaction than the com- bination of a polymer-anchored peptide catalyst and a free surfact ant 7 Concluding Remarks Compared to other polymers poly(isocyanides) have received little attention in the past despite the fact that this class of polymers has been known since the 1970s We hope to have shown in this article that poly(isocyanides) have a rich chemistry and deserve further study by organic chemists and polymer chemists New and interesting applications can be foreseen for these compounds in particular in the promising areas of biomi- metic chemistry and supramolecular chemistry Acknowledgment The author warmly thanks Dr Wiendelt Drenth Emeritus Professor of Organic Chemistry at the Univer- sity of Utrecht for his invaluable contributions to the field described in this article 8 References 1 ‘Frontiers in Supramolecular Organic Chemistry’ ed H -J Schneider and H Durr VCH Weinheim 1991 2 (a) M Farina Top Stereochem 1987 17 1 (6)G Wulff Angew Chem Int Ed Engl 1989,28,21 3 W Drenth and R J M Nolte Ace Chem Res 1979 12,30 4 K Ute K Hirose H Kashimoto K Hatada and 0 Vogl J Am Chem Soc 1991,113,6305 5 Y Okamoto and E Yashima Prog Polym Sci 1990,263 6 F Millich Chem Rev 1972,72 101 7 R J M Nolte and W Drenth in ‘New Methods for Polymer Synthesis’ ed W J Mijs Plenum Press New York 1992 273 8 T J Deming and B Novak Macromolecules 1991,24 326 9 T J Deming and B M Novak Macromolecules 1991,24,5478 and 6043 10 Y Ito E Ihara M Murakami and M Shiro J Am Chem SOC 1990 112,6446 11 P C J Kamer R J M Nolte,and W Drenth,unpublishedresults 12 R J M Nolte and W Drenth in ‘Recent Advances in Mechanistic and Synthetic Aspects of Polymerization’ ed M Fontanille and A Guyot D Reidel Publishing Company Dordrecht 1987,451 13 A J M van Beijnen R J M Nolte A J Naaktgeboren J W Zwikker W Drenth and A M F Hezemans Macromolecules 1983,16 1679 14 C J M Huige Thesis Utrecht 1985 5 (a)R J M Nolte A J M van Beijnen and W Drenth J Am Chem Soc ,1974,96,5932 (b)A J M van Beijnen R J M Nolte and W Drenth Reel Trav Chim Pays-Bas 1980,99 12 1 6 C Kollmar and R Hoffmann J Am Chem Soc 1990 112,8230 7 C J M Huige,A M F Hezemans R J M Nolte,and W Drenth Recl Trav Chim Pays-Bas 1992,112,33 8 (a) M M Green R A Gross F C Schilling K Zero and Ch Crosby 111 Macromolecules 1988 21 1839 (b)P C J Kamer W Drenth and W Drenth Polym Prepr 1989,30,418 19 TJ Deming and B M Novak J Am Chem Sue 1992,114,4400 and 7926 20 P C J Kamer R J M Nolte and W Drenth J Am Chem Soc 1988,110,6818 21 P C J Kamer M C Cleij R J M Nolte T Harada A M F Hezemans and W Drenth J Am Chem SOC 1988 110 1581 22 Y Ito E Ihara and M Murakami Angew Chem 1992,104,1508 23 (a) R Schmid in ‘Jaundice’ ed C A Goresky and M M Fischer Plenum New York 1975 (b) J Jacobsen Int J Peptide Protein Res 1977,9,235 24 J M van der Eijk R J M Nolte V E M Richters and W Drenth Biopolymers 1980 19,445 25 ‘MembraneTransport’,ed S L BontingandJ J H H M dePont Elsevier New York 1981 26 (a)J -H Fuhrhop U Liman and V Koesling J Am Chem Soc 1988 110,6840 (b)V E Carmichael P J Dutton,T M Fyles T D James J A Swan and M Zojaji J Am Chem SOC,1989 111 767 (c)M J Pregel L Jullien and J M Lehn Angew Chem ,1992 104 1695 27 (a)U F Kragten,M F M Roks,andR J M Nolte,J Chem Soc Chem Commun 1985,1275 (6)M F M Roks and R J M Nolte Macromolecules 1992 25 5398 28 R E White and H J Coon Annu Rev Biochem 1980,49,315 29 (a)J van Esch M F M Roks and R J M Nolte J Am Chem Soc 1986 108 6093 (b) M F M Roks H G J Visser J W Zwikker,A J Verkleij,and R J M Nolte J Am Chem Soc 1983 105,4507 30 (a) J Drenth Recl Trav Chim Pays-Bas 1980 99 185 (b) T Kunitake and Y Okahata Adv Polym Sci 1976 20 159 (c)H Dugas and C Penney ‘Bio-organic Chemistry A Chemical Approach to Enzyme Action’ Springer New York 1989 31 (a)H G J Visser R J M Nolte J W Zwikker and W Drenth J Org Chem 1985 50 3133 and 3138 (b)H G J Visser R J M Nolte and W Drenth Macromolecules 1985 18 1818 (c) M C Cleij Thesis Utrecht 1989

ISSN:0306-0012    

DOI:10.1039/CS9942300011

出版商:RSC    

年代:1994

数据来源: RSC

摘要:

Thin Film Diamond by Chemical Vapour Deposition Methods M. N. R. Ashfold, P. W. May, and C. A. Rego School of Chemistryl University of Bristol, Bristol BS8 I TS, U.K. N. M. Everitt Department of Aerospace Engineering, University of Bristol, Bristol BS8 1 TR, U.K. 1 Introduction Diamond, the sp3-bonded allotrope of carbon, has long held a special place in the hearts and minds both of scientists and the public at large. For the latter, the word diamond may conjure up images of 57 facetted brilliant gem stones, Amsterdam or mines in South Africa, wealth and special occasions. To the scientist, diamond is impressive because of its wide range of extreme properties. As Table 1 shows, by most measures diamond is ‘the biggest and best’: It is the hardest known material, has the lowest coefficient of thermal expansion, is chemically inert and wear resistant, offers low friction, has high thermal conductivity, is electrically insulating and optically transparent from the ultra- violet (UV) to the far infrared (IR).Given these many notable properties, it should come as no surprise to learn that diamond already finds use in many diverse applications including, of course, its use as a precious gem, but also as a heat sink, as an abrasive, and as inserts and/or wear-resistant coatings for cutting tools. Obviously, given its many unique properties it is possible to envisage many other potential applications for diamond as an engineering material, but progress in implement- ing many such ideas has been hampered by the comparative scarcity of natural diamond.Hence the long running quest for routes to synthesize diamond in the laboratory. At ambient temperatures and pressures graphite is the stable form of solid carbon. Its enthalpy of formation is a mere 2.9 kJ mol-lower than that of diamond, but a large activation barrier rules out simple thermal activation as a means of driving the graphite +diamond interconversion. The activation energy for graphitization at the { 1 10) diamond surface has been measured as 728 f50 kJ mol- l,l roughly the same as the vaporization Mike Ashfold is a Professor of Chemistry at the University of Bristol. He was born in Bridgwater, Somerset and obtained his B.Sc. and Ph.D. degrees at Birmingham University.There .followed three years as a Guy Newton Junior Research Fellow at P. May C. Rego M. Ashfold N. Everitt energy of graphite. Diamond is the densest allotrope of carbon (pdlamond= 3513 kg m- 3, cf. PgraphLte = 2260 kg m-at 293 K);thus, at high pressure, diamond must be the stable form of solid carbon. This is the scientific basis for the high pressure high temperature (HPHT) techniques (in which diamond is crystal- lized from metal solvated carbon at P -50-100 kBar and T-1800-2300 K) by which diamond has been synthesized commercially for over 30 years.2 -Table 1 Some of the outstanding properties of diamond Extreme mechanical hardness (-90 GPa) Strongest known material, highest bulk modulus (1.2 x 10l2Nm-*), lowest compressibility (8.3 x 10-m2 N-l) Highest known value of thermal conductivity at room temperature (2 x lo3 Wm-’ K-l ) Thermal expansion coefficient at room temperature (0.8 x K-l) is comparable with that of invar Broad optical transparency from the deep UV to the far IR region of the electromagnetic spectrum Good electrical insulator (room temperature resistivity is -1OI6 Q cm) Diamond can be doped to change its resistivity over the range 10-106 Q cm, so becoming a semiconductor with a wide band gap of 5.4 eV Very resistant to chemical corrosion High radiation hardness the University of Oxford, prior to a faculty appointment at Bristol in 1981,progression through the ranks andappointment to a Chair in 1992.He has received the Marlow (1987) and Corday-Morgan (1989) Medals of the Royal Society of Chemistry.Paul May was born in London, but spent his formative years in Redditch, Worcester. He obtained his B.Sc. from Bristol Univer- sity, then worked in industry for three years before returning to Bristol for his Ph.D. He currently holds a Ramsay Memorialpost- doctoral Fellowship. Christopher Rego was born in Barnet, England. He obtained undergraduate and graduate degrees from Bristol University, before taking up a post-doctoral fellowship at Exeter University working in theJield of microwave spectroscopy. He moved back to Bristol in 1992 as a D TI-sponsored Research Associate. Nicola Everitt is a Lecturer in Aerospace Materials at the University of Bristol. She was born in Dorset, England, and graduated with a B.Sc.in Materials Science from Bath University in 1985. A year with the Atomic Energy Authority, at Harwell, was followed by postgraduate research at Oxford University (Linacre College), the award of a D. Phil. in 1990 and, later in the same year, appointment to her present Lectureship post at Bristol. 21 22 World interest in diamond has been further increased by the much more recent discovery that it is possible to produce polycrystalline diamond films by a wide variety of chemical vapour deposition (CVD) techniques using, as process gases, nothing more exotic than a hydrocarbon gas (typically methane) in an excess of hydrogen. 5-7 CVD diamond can show mechani- cal, tribological, and even electronic properties comparable to those of natural diamond.There is currently much optimism that it will prove possible to scale CVD methods to the extent that they will provide an economically viable alternative to the traditional HPHT methods for producing diamond abrasives and heat sinks, whilst the possibility of coating large surface areas with a continuous film of diamond will open up whole new ranges of potential application for the CVD methods.8 In this Review we describe some of the various methods for growing CVD diamond, paying particular attention to our current (incomplete) understanding of the underlying chemistry. We also survey some of the more important surface analysis techniques used to characterize thin film diamond, and then consider how the deposition conditions influence the quality and the growth rates of the diamond films produced by the various CVD methods.Attention will then turn to the substrate mater- ials on which the films can be grown. We summarize current thinking regarding what physico-chemical properties of the substrate are important for nucleation, for subsequent film growth, and in determining whether the resulting diamond coating will be strongly adherent. The Review concludes with discussion of a few specific and novel applications (some rea- lized, some in the realms of reasonable expectation) of CVD diamond films. 2 The CVD Process Chemical vapour deposition, as its name implies, involves a gas- phase chemical reaction occurring above a solid surface, which causes deposition onto that surface.All CVD techniques for producing diamond films require a means of activating gas-phase carbon-containing precursor molecules. This generally involves thermal (e.g. hot filament) or plasma (D.C., R.F., or micro- wave) activation, or use of a combustion flame. Figure 1 illustrates three of the more popular experimental methods and gives some indication of typical operating conditions. (These and other deposition methods are discussed in greater detail in reference 9.) Whilst each method differs in detail, they all share features in common. For example, growth of diamond (rather than deposition of other, less well-defined, forms of carbon) normally requires that the substrate be maintained at a tempera- ture in the range 1000-1400 K, and that the precursor gas be diluted in an excess of hydrogen (typical CH, mixing ratio N 1-2 vol%).The resulting films are polycrystalline, with a morpho- logy that is sensitive to the precise growth conditions (see later). Growth rates for the various deposition processes vary consider- ably, and it is usually found that higher growth rates can be achieved only at the expense of a corresponding loss of film quality. ‘Quality’ here is a subjective concept. It is taken to imply some measure of factors such as the ratio of sp3(diamond) to sp2 bonded (graphite) carbon in the sample, the composition (e.g. C-C uersus C-H bond content) and the crystallinity. In general, combustion methods, such as oxyacetylene or plasma torches, deposit diamond at high rates (typically 100 -+ 1000 pm h-*, respectively), but often only over very small, localized areas and with poor process control leading to poor quality films.In contrast, the hot filament and plasma methods have much slower growth rates (0.1-10 pm h- l), but produce high quality films. However, the filament method often suffers from contami- nation problems, since metal boiled off the filament can be incorporated into the growing diamond film. One of the great challenges facing researchers in CVD diamond technology is to increase the growth rates to economically viable rates, (hundreds of pm h -l, or even mm h -*) without compromizing film quality. Another major problem that is receiving a lot of attention is CHEMICAL SOCIETY REVIEWS, 1994 Figure 1 Schematic diagram of three of the most commonly used types of diamond CVD apparatus.(a) Hot Filament reactor, (b) Microwave Plasma Enhanced CVD reactor, and (c) Oxyacetylene Torch. the mechanism of heteroepitaxial growth, that is, the initial stages by which diamond nucleates upon a non-diamond sub- strate. Several studies have shown that pre-abrasion of non- diamond substrates reduces the induction time for nucleation and increases the density of nucleation sites. Enhanced growth rates inevitably follow since formation of a continuous diamond film is essentially a process of crystallization, proceeding via nucleation, followed by three-dimensional growth of the various microcrystallites to the point where they eventually coalesce.However the precise manner in which the pre-abrasion enhances surface activity remains a topic of some debate. Suggested mechanisms range from the idea that the initial growth occurs on detritus from the (diamond) abrasive embedded in the substrate surface, to those which maintain that simply the presence of surface defects produced by mechanical damage is sufficient to provide suitable sites for diamond formation via a heterogeneous nucleation process. The abrasion process is usually carried out by mechanically polishing the substrate with an abrasive grit, usually diamond powder of 0.1-10 pm particle size, although other materials, such as Sic or sapphire have been used. Better abrasion uniformity (and hence subsequent diamond film uniformity) can be obtained if the substrate is placed into a slurry containing a mixture of the abrasive grit in a hydrocarbon medium, and is then ultrasonically agitated.What- ever the abrasion method, however, the need to damage the surface in such a poorly defined manner prior to deposition may severely inhibit the use of CVD diamond for applications in, say, the electronics industry (see later), where circuit geometries are frequently on a submicron scale. This worry has led to a search for more controllable methods of enhancing nucleation, such as THIN FILM DIAMOND BY CHEMICAL VAPOUR DEPOSITION METHODS-M N R ASHFOLD ET AL ion bombardment This is often performed in a microwave deposition reactor, by simply applying a negative bias of a few hundred volts to the substrate and allowing the ions to (1) damage the surface, (11) implant into the lattice, and (iii) form a carbide interlayer (see later) The exact mechanism by which this process enhances diamond nucleation is still not properly under- stood but recent evidencei0 shows that biasing the substrate increases both the concentration of H atoms close to the surface and the electron temperature of the plasma Figure 2 The mechanism of CVD involves dissociation of precursor gases (usually methane and hydrogen) by either electron impact (plasma methods) or, as shown here, thermal energy from a hot filament Atoms and reactive species diffuse to the substrate surface, where they absorb and coalesce to form a carbon film If the deposition conditions are favourable, the film is diamond Figure 2 provides a schematic of the diamond CVD process Here we assume a CHJH, source gas mixture, though we note in passing that many other hydrocarbon precursor gases have been investigated” l2 and that extra gases (e g 0,) are some- times added with a view to improving the resulting film quality and decreasing deposition temperatures Bachmann et a1 l2 have shown that successful synthesis of single-phase diamond is only feasible in a localized region of the C/H/O-gas-phase compositional diagram Most of this region straddles the line representing equal C and 0 atom concentrations, although it is of course also possible to grow diamond films using an oxygen- free gas mixture provided that there is a large excess of H atoms This shows that so long as the deposition conditions are energetic enough to produce complete dissociation of the parent gas molecules, the quality of deposited diamond depends only upon the ratio of carbon-to-hydrogen-to-oxygen within the reactor Activation of the gas, be it thermal or via electron bombardment, involves production of H atoms Our schematic shows these causing H-atom abstraction from the methane with the resulting formation of methyl radicals There is now a considerable body of evidence to suggest that methyl radicals are the dominant growth species in many variants of diamond CVD This evidence includes direct zn sztu dectection of CH, radicals in diamond CVD reactors by a variety of spectroscopic methods (e g mass spectrometry, resonance enhanced multiphoton ioni- zation (REMPI), and direct infrared absorption spectroscopy), the results of kinetic modelling both of the gas-phase chemistry and of observed film growth rates, and a variety of isotopic labelling studies The methyl radicals, along with other gas- phase species, will be transported towards the substrate by a combination of laminar, convective and/or diffusive flow mechanisms depending on the particular process conditions, whilst the relative species concentrations will continue to evolve as a result of further gas-phase reactions Transport back and forth across the boundary layer over the growing surface will be especially important with the higher pressure (e g torch) meth- ods, this can be expected to further influence the local gas-phase environment Deposition involves adsorption and desorption at the surface Diffusion across the growth surface may well result in nucleation and growth of the diamond film, whilst diffusion info the bulk substrate is often an alternative, generally undesir- able, competitive loss process We stated earlier that graphite, not diamond, was the stable form of solid carbon at ambient pressures and temperatures The fact that dzamondfilms can be formed by CVD techniques is inextricably linked to the presence of hydrogen atoms These are believed to play a number of crucial roles in the CVD process First, as Figure 2 suggests, they are intimately involved in the formation of carbon-containing radzcal species This is important, since stable hydrocarbon molecuks do not react with diamond to cause diamond growth Secondly, H-atoms termi- nate the ‘dangling’ carbon bonds on the growing diamond surface and prevent them from reconstructing to a graphite-like surface However, if these were the only functions of the H atoms, diamond growth would be unlikely to occur This conclusion is reached simply by considering the relative strengths of a C-H and a C-C bond The former is stronger, thus any approaching carbon atom would not be able to displace the surface bonded H-atom Fortunately, the H-H bond in molecular hydrogen is stronger than either Thus abstraction of a surface-bonded H-atom by a gas-phase H-atom leading to formation of a gas-phase hydrogen molecule is an exoergic process, so the H-atoms also ‘activate’ the growing surface by creating local vacant sites,14 see Figure 3 Given that the concentration of gas-phase H-atoms generally far exceeds that of any of the carbon-containing radical species, the most likely fate for any vacant site is re-termination by addition of another H-atom Occasionally, however, the colliding species will be a carbon-containing radical (e g CH,) A new C-C bond will be formed, thus providing the possibility of extending the diamond lattice Of course this extension of the carbon network need not have the correct diamond structure but, here again, the H-atoms play a helpful role Atomic hydrogen etches both diamond and graphite but, under typical CVD conditions, the rate of diamond growth exceeds its etch rate whilst for other forms of carbon (graphite, for example) the converse is true Thus it is that essentially pure diamond films can be grown The presence (and importance) of H-atoms has been verified by a number of spectroscopic methods Celii and Butler’ used three-photon REMPI to probe the spatial distribution of ground state H-atoms in a hot filamemt CVD reactor and to investigate the way in which the relative H-atom concentration -HI c-c /c-c>-c ‘c-c Figure 3 One of the possible reaction schemes for growth of diamond at a { 110)trough site Atomic hydrogen abstracts a hydrogen atom from the surface of diamond leaving a surface radical A methyl group then adds to the ‘dangling bond’, so adding carbon to the lattice A further hydrogen abstraction reaction and methyl addition, followed by an internal hydrogen elimination step closes the ring and propagates the diamond structure All the steps are reversible, but the process is driven to completion by the stability of the contiguous diamond lattice with respect to graphitic structures Similar mechanisms are believed to operate at other lattice sites, using either methyl, ethyl, or acetylene precursors varied with filament temperature and with CH,/H, mixing ratio.Recent modelling of these data points to the importance of surface-catalysed decomposition of H, on the hot filament in establishing the observed H-atom concentrations.’ Ground state H-atoms have also been monitored via their laser-induced fluorescence (LIF) and by in situ third harmonic ’ Plasma-enhanced CVD processes involve different excitation/ dissociation mechanisms: electron-molecule collisions play a much more significant role.These tend to yield much higher concentrations of electronically excited species, e.g. H-atoms with principal quantum number n = 3, which can be detected conveniently through observation of their Balmer-a emission in the red, and simple radical species like CH, OH, and C, which can also be detected by their spontaneous optical emission. Following these emissions can conceivably provide some mea- sure of process control, but the information such emissions appear to convey should be treated with circumspection.For example, both C, and CH emissions show strongly during growth of CVD diamond using combustion flames and plasma torches, but this does not necessarily imply that these species are important in the diamond growth mechanism. Analysis of the wavelength dispersed emission can provide a measure of the relative populations in the various rovibrational levels of the emitting electronic state. This population distribution often approximates a Boltzmann distribution, characterized by a ‘temperature’. However, given the short fluorescent lifetime of the emitting species, collisional thermalization prior to emission is unlikely and the ‘temperature’ deduced is thus more an indicator of the radical production mechanism than any charac- teristic of the bulk of the discharge. In situ, non-intrusive spectroscopic methods capable of probing ground state species are more likely to provide realistic measures of the plasma temperature. Examples of such techniques include LIF (though, given their lack of suitable fluorescing excited states this tech- nique is unsuitable for potentially important species like CH, radicals or stable molecular species like CH, and C2H,), REMPI, and, most recently, degenerate four-wave mixing spectroscopy.3 The CVD Diamond Film Figure 4 shows scanning electron micrographs (SEM) of poly- crystalline CVD diamond films grown on a single crystal silicon substrate using two different CH,/H, gas mixtures (CH, mixing ratios of 1% and 2.5% respectively) in a hot filament reactor.Such images, in themselves, do not prove whether the films are diamond but monitoring the attendant cathodoluminescence can provide supportive evidence. However, the most widely used technique, both for establishing that a film is indeed diamond and for providing some measure of the film quality, is laser Raman spectroscopy. The Raman spectrum of natural diamond shows a sharp, single peak centred at a wavenumber of -1332 cm-’. As Figure 5a shows, this feature also dominates the Raman spectra of good quality polycrystalline diamond films grown by CVD methods, though the peak linewidth is usually somewhat greater than for natural diamond and the linecentre is often found to be slightly shifted in wavenumber -both effects generally attributed to compressive stress in the film.19.20By way of contrast, the Raman spectrum of well-crystallized graphite shows a somewhat broader feature centred around 1580cm -l. When characterizing CVD ‘diamond’ films, the observation of any broad resonance around this higher wavenumber (such as can be seen, for example, in Figure 5c) is generally taken to indicate the presence of graphite-like non- diamond phases containing sp2-bonded carbon atoms.Raman spectroscopy is just one tool in an armoury of analytic techniques used to characterize CVD diamond films. Other spectroscopies that are useful for distinguishing diamond from alternative forms of carbon include electron energy loss spec- troscopy (EELS), Auger electron spectroscopy (AES), X-ray photoelectron spectroscopy (XPS), low energy electron diffrac- tion (LEED), and X-ray diffraction.Unfortunately, all are ex CHEMICAL SOCIETY REVIEWS, 1994 Figure 4 SEM of typical polycrystalline diamond films grown on Si by CVD. (a) was grown using 0.5% methane in hydrogen, which produces sharp angular crystals about 1 pm in size.” (b) was grown using 2.5% methane in hydrogen, which give rise to nanocrystalline diamond particles which aggregate into roughly spherical ‘cauli- flower’ structures. situ techniques, the implementation of any of which requires cessation of the deposition process, removal of the sample from the reactor and, frequently, some considerable sample prep- aration and analysis. The same criticisms apply to the use of secondary ionization mass spectrometry (SIMS) and to the technique of transmission electron microscopy (TEM) -argu-ably the ultimate route to establishing the detailed microstruc- ture of a CVD diamond film.SIMS does not allow one to distinguish between the different forms of carbon, but is ideally suited to mapping out the spatial distribution of different elements residing within a few atomic layers of a surface. Thus it finds particular use as a means of probing the interfacial region when CVD diamond films are grown on non-diamond sub- strates (after sectioning -see below). Monitoring atomic hydro- gen by SIMS is difficult, but not impossible.Given the expec- tation that hydrogen in CVD diamond films will be concentrated at the surface and in the grain boundaries, the effort involved in such a study might well be very rewarding. Other measurable properties which give some insight into the quality of CVD diamond films include hardness, coefficient of friction, density, bulk modulus, and thermal conductivity. All have been investigated, as a function of the growth conditions -most notably the CH,/H, mixing ratio. Hardness is tradition- ally measured by indent testing. Indentation, and scratch testing (a method of measuring friction), both have attendant complica- tions in the case of good quality CVD diamond films, not least because the diamond tip used to make the impression will have a hardness comparable to that of the film under test. Scratch testing is complicated by the fact that the as grown film is rough (recall the microcrystallinity evident in Figure 4), whilst indent measurements on thin films will only be reliable if proper regard is given to the deconvolution of contributions from the underly- THIN FILM DIAMOND BY CHEMICAL VAPOUR DEPOSITION METHODS-M.N R ASHFOLD ET AL. 1333 1332 (;)Laser Raman spectra measured by Knight and White.20 5Figure (a) 1332 11_______L_-_/--500 750 7000 7250 7500 7750 Wavenumber (cm-’) Natural diamond showing the characteristic sharp peak at 1332 cm -. (b) Good quality CVD diamond film grown on Si using 0.5% CH, in H, gas mixture.The peak at 520 cm-’ is due to Si. (c) Poor quality CVD diamond film grown using 1.5% CH, in H,. The broad hump at -1597 cm- is evidence that the film contains non-diamond phases, such as graphite or amorphous carbon. The diamond peak has also broadened considerably. (Reproduced, with permission, from J. Muter. Res., 1988,4, 385.) ing (almost inevitably softer) substrate. Such measurements have been reported;2 1.22 these show good quality CVD diamond films to have a hardness approaching that of natural diamond. We have already commented that diamond has a density well in excess of that of graphite or any other form of carbon. Using a sink-float method, Sat0 and KamoZ3 showed that CVD diamond films grown from lean CH,/H, mixtures (CH, mixing ratio <0.5%) have a density that is >99.5% of the correspond- ing value for natural diamond. However, they also found that the film density decreases with increasing partial pressure of CH, in the process gas mixture.By attributing this trend solely to progressive contamination of the grown film by disordered graphite, these workers were able to derive a lower limit to the fractional diamond content in films grown with CH, mixing ratios as high as 5%. Higher CH, mixing ratios have also been shown to have an adverse effect both on the bulk modulus and the thermal conductivity of CVD diamond films.23 As Figure 4 showed, the surface morphology is another property that is very sensitive to the gas mixing ratio.It also depends on the substrate temperature. Figure 6 provides a qualitative summary of the way both affect the surface morpho- logy. In essence we are here concerned with the way in which the growth rates vary with microcrystallite orientation. Under ‘slow’ growth conditions -low CH, partial pressure, low substrate temperature -triangular { 1 1 1 facets tend to be most evident, with many obvious twin boundaries. { 100) facets, appearing both as square and rectangular forms, begin to dominate as the relative concentration of CH, in the precursor gas mixture, and/or the substrate temperature, is increased. Both types of facet are clearly evident in the scanning electron micrograph shown in Figure 4a. A cross-section through this Figure 6 SchematlL diagram of film morphology as a function of deposition temperature and methane/hydrogen ratio for CVD between 30-80 Torr (adapted from ref.23). The regions correspond to different observed morphologies: (a) little or no deposition of either diamond or graphite, (b) { 1 1 1} faces predominate, (c) { 1 1 11 and { loo] faces appear with comparable frequency, (d) { 100)faces predominate, (e) { 100) faces predominate and diamond grains grow preferentially along the (100) axis, (f) films have a smooth surface composed of nanocrystals of diamond and disordered graphite, (g) film surface is composed of fibrous deposits of soot or disordered graphite growing vertically from the surface. Figure 7 Cross-sectional SEM of a 6.7 pm thick CVD diamond film showing columnar growth.” film shows the growth to be essentially columnar (Figure 7).At still higher CH, partial pressures the crystalline morphology disappears altogether; a film such as that shown in Figure 4b is an aggregate of diamond nanocrystals and disordered graphite. Obviously, the crystalline morphology of a CVD diamond film is an important consideration when it comes to potential applications. A film like that shown in Figure 4a might find use as a fine abrasive coating, but most of the envisaged uses for diamond films in optics, in thermal management applications, and as possible electronic devices require that the film surfaces be as smooth as possible. One can envisage (at least) two routes to this objective: one has either to identify growth conditions which naturally result in the formation of smooth films, or to optimize ways of ‘polishing’ away the surface roughness of the film as grown.Both concepts are presently the subject of intense research effort. We have already seen that the film structure and morphology is sensitive to the relative growth rates on the various crystallographic faces. Wild et af.,, have demonstrated that, at least in the case of silicon substrates, it is possible to set the deposition parameters in such a way that the CVD process is so biased in favour of growth on the { 100)facets that the surface CHEMICAL, SOCIETY REVIEWS, 1994 of the resulting film consists entirely of well aligned, coplanar only possible substrate material.What are the properties {loo} facets with the axis of growth normal to the substrate required of a substrate if it is to be suitable for supporting an surface. Figure 8 shows a scanning electron micrograph of such adherent film of CVD diamond? One requirement is obvious. a (100) textured CVD diamond film. Glass and co-~orkers~~ The substrate must have a melting point (at the process pressure) have demonstrated an alternative route to growing {100) tex- tured CVD diamond films on silicon by microwave plasma enhanced CVD. This relied on a three-step ‘growth’ process, the first two of which (carburization, followed by a negative biasing step at lower substrate temperature, the combined effect of which is believed to be the formation of an epitaxial P-SiC buffer layer) may be viewed as a special pretreatment of the substrate surface prior to ‘normal’ CVD growth.We now turn to consider the prospects for post-polishing. Mechanical polishing. the traditional method for preparing diamond gemstones, is one obvious, if uneconomic. possibility. A variant on this theme chemical-mechanical polishing- relies ~ on the fact that diamond reacts with iron at elevated tempera- tures. Mechanical polishing on a heated (-900 K) iron disc in the presence of atomic hydrogen leads to graphitization of surface diamond and subsequent dissolution of the carbon into the hot iron. The hydrogen serves to remove carbon from the iron and prevents saturation of the iron by carbon.A very recent and exciting extrapolation of this type of process is introduced at the end of this paragraph. Another route to surface smoothing involves use of a high energy pulsed laser (e.g. an excimer or a Nd-YAG laser). The actual polishing mechanism (or mechan- isms) remains the subject of some debate: selective ablation of ’high spots’ on the rough diamond surface has been suggested but so, too, has local oxidation, particularly in instances where the irradiation was carried out under an oxygen-containing atmosphere. Ion bombardment, and reaction with oxygen atoms, have also been proposed as polishing methods, but both tend to cause pitting at the grain boundaries. Recently Jin et ul.26 have demonstrated a new technique for smoothing free-standing diamond films which involves use of molten rare-earth metals.The diamond film is sandwiched between layers of the rare-earth metal (e.g. cerium or lanthanum), or a suitable alloy, and held for a few hours under an inert atmosphere of argon at a temperature (-1200 K) somewhat above the melting point of the metal. Carbon has high solubility in these metals at such temperatures. After the heat treatment both the residual (unreacted) and the reacted rare-earth metal is removed by acid etching, leaving a greatly smoothed (albeit thinner) diamond film. 4 The Substrate Most of the CVD diamond films reported to date have been grown on single crystal silicon wafers, but this is by no means the Figure 8 Scanning electron micrograph of a textured {loo] diamond film.(Reproduced, with permission, from Diamond und Reluted Muteriuls, 1993, 2, 158.) higher than the temperature window (1000-1400 K) required for diamond growth. This precludes the use of existing CVD techniques to diamond-coat plastics or low melting metals like aluminium. It is also helpful, though not essential, that the substrate be capable of forming a carbide. CVD of diamond on non-diamond substrates will usually involve initial formation of a carbide interfacial layer upon which the diamond then grows. Somewhat paradoxically, it is difficult to grow on materials with which carbon is ‘too reactive’, i.e. many of the transition metals (e.g. iron, cobalt, etc.) with which carbon exhibits a high mutual solubility.Hence the appeal of substrates like silicon, molybde- num, and tungsten materials which form carbides. but only as a localized interfacial layer because of their modest mutual solubility with carbon under typical CVD process conditions. The carbide layer can be pictured as the ‘glue’ which promotes growth of CVD diamond, and aids adhesion by (partial) relief of stresses at the interface. Turning now to the problem of stress, it is fairly obvious that it is desirable for the substrate material to have a low coefficient of thermal expansion. Diamond has one of the lowest coefficients of thermal expansion of any material (see Figure 9). Since the CVD growth process takes place at elevated temperatures it is almost always the case that upon cooling back to room tcmpera- ture, the substrate will have contracted more than the diamond film.As a result, the latter will be under compressive stress. This manifests itself in the broadening and shifting of the 1332 cm- Raman peak mentioned earlier.Clearly. da, the mismatch in the coefficients of thermal expansion for diamond and the substrate material, must be an important factor in determining whether it will be possible to grow an udherent CVD diamond film on that particular substrate. Figures 10 and 1 1, which show scanningelectron micrographs of CVD diamond ‘films’ that we have grown on quartz and on copper substrates illustrate some of the consequences of a non- zero Au. In both cases we have succeeded in growing a CVD diamond film, but the end result is very different.Consider first the case of the quartz (a form of silica). Growth of diamond on quartz proceeds via formation of a thin interfacial silicon carbide layer, however the substrate has a coefficient of thermal expansion much greater than that of diamond. Thus it is possible cu NI co Fe TI V Nb Cr Te Zr Mo W S13N4 SIC A1203 WC SI 5102 (fused) Sapphlre Graphlte Quartz Dlamond I#I,I, Thermal Expanslon Coefflclent at 300 K (IO-’K-‘) Figure 9 Bar graph showing typical values for the thermal expansion coefficient for a variety of substrate materials at 300 K. THIN FILM DIAMOND BY CHEMICAL VAPOUR DEPOSITION METHODS-M. N. R. ASHFOLD ET AL.Figure 10 SEM of a CVD diamond film grown on a flat quartz substrate. The film has grown and adhered but, upon cooling, compressive stresses resulting from the contraction of the quartz have caused the diamond film to crack into plates, with the plates riding over one another to relieve the stress. Note that cracking is often transgranular, indicating that grain boundaries are not in themselves an inherent source of weakness in polycrystalline diamond films. Figure 11 SEM of a CVD diamond film grown on Cu. The film has cracked and crazed into plates, which have partially delaminated. to grow thin adherent films of CVD diamond on pre-abraded quartz, but with increasing film thicknesses and/or areas the internal stresses become too great and the diamond film cracks into a number of ‘plates’ as shown in Figure 10.In order to alleviate the stress, these plates ride up over one another at their edges; each, however, remains well adhered to the substrate. Contrast this with growth on copper. In this case the substrate does not form a carbide and da is also very large (-1.5 x 10-K-I). Given these facts, it is perhaps surprising that diamond grows on copper at all. Nonetheless, Figure 11 confirms that it is indeed possible to grow a CVD diamond film on pre-abraded copper but, because there is no carbide layer providing the necessary ‘glue’ at the interface, it is not an adherent film: it will readily flake off as soon as the substrate is tilted. If we consider just carbon-substrate interactions, metals, alloys, and pure elements sub-divide into three classes2 exhibit- ing, respectively: (a) Little or no C solubility or reaction.These include metals such as Cu, Sn, Pb, Ag, and Au as well as non-metals, such as Ge, sapphire, diamond itself, and graphite, although in the latter case etching will occur concurrently with diamond growth. (b) C dzffusion. Here, the substrate acts as a carbon sink, whereby deposited carbon dissolves into the metal surface to form a solid solution. This causes large amounts of carbon to be transported into the bulk, leading to a temporary decrease in the surface C concentration, delaying the onset of nucleation. Metals where this is significant include Pt, Pd, Rh, Fe, Ni, and Ti. For substrates with a very high C diffusion rate, sample thickness becomes a significant parameter influencing the onset of nucleation; thin foils reach their carbon saturation more rapidly than thick (c) Carbide Formation.These include metals such as Ti, Zr, Hf, V, Nb, Ta, Cr, Mo, W, Fe, Co, Ni, Y, A1 and certain other rare earth metals. In some metals, such as Ti, the carbide layer continues to grow during diamond deposition and can become hundreds of pm thick. Such thick interfacial carbide layers may severely affect the mechanical properties, and hence the utility of CVD diamond coatings on these materials. Non-metals, such as B or Si, and Si-containing compounds such as SiO,, quartz, and Si3N, also form carbide layers. Substrates composed of carbide themselves, such as Sic, WC, and TIC are also particularly amenable to diamond deposition. Such problems with heteroepitaxial growth have ensured the continuing popularity of silicon as a substrate material.It has a sufficiently high melting point (1683 K), it forms a localized carbide layer and it has a comparatively low thermal expansion coefficient. Tungsten and molybdenum display similar virtues and are also widely used as substrate materials. Molybdenum, in particular, is finding increasing use, especially as a substrate for CVD diamond films grown by the high growth rate methods (e.g. when using a plasma torch), because of its ability to withstand thermal shock. A review such as this would not be complete without some mention of current progress towards the growth of single crystal (rather than polycrystalline) diamond films. This remains a major challenge but, given the potential reward, it is an area of great current activity.The need for, and the advantages of, single crystal diamond films are reasonably self-evident. Only the single crystal material will fully exhibit all of the extreme and unique properties we associate with diamond. Most of the potential electronic applications of diamond (see below) demand single crystal material. Devices made from homoepitax- ial diamond are actually expected to perform better than com- parable gemstone diamond devices with respect to properties such as dielectric breakdown strength and carrier mobility.29 To date, epitaxial growth of diamond films has been reported on diam~nd~~.~~and on a single crystal cubic boron nitride sub- strate~~~but the quest for heteroepitaxial growth on a cheaper, more readily available substrate material continues.What is required of such a substrate material? Obviously, the various properties discussed earlier but, in addition, one might look for a substrate that had a crystal structure and lattice parameters not dissimilar to that of diamond (cubic, with lattice spacing a = 356.7 pm). Nickel and copper are two such materials. Both have face-centred cubic structures and similar lattice constants (a = 352.4 and 361.5 pm, respectively). Unfortunately both have associated problems when it comes to CVD diamond growth.Diamond grows on copper, but does not adhere (recall Figure 11). Nickel, in contrast, is one of those metals like iron in which carbon has a high solubility. Nonetheless, local epitaxial growth of diamond microcrystals on Ni has been dem~nstrated,~, though reports of continuous films are rare. We conclude this section with mention of two alternative strategies which have been shown to yield oriented diamond films, i.e. films with a texture and a degree of crystalline order intermediate between that of a typical polycrystalline CVD diamond film and that of a genuine single crystal film. One is the bias-enhanced nucleation technique summarized earlier, by which it has proved possible to grow oriented diamond films, not only on single crystal p-Sic (another scarce substrate material) but also, after an appropriate in situ carburization pre-treat- ment, on single crystal silicon.2s Geis et al.33 have followed a different philosophy.They painstakingly prepare a mosaic of appropriately oriented diamond seed crystals to act as the ‘substrate’ in their CVD reactor. Subsequent homoepitaxial growth leads to the gradual coalescence of the individual growing islands and, ultimately, to a highly oriented CVD diamond film. 5 Present Applications and Future Prospects In this Review we have attempted to provide a snapshot of progress to date in some of the many aspects of research in the area of CVD diamond films. Of course, it is also appropriate to ask how this research effort is feeding through into the market- place.Two areas of application are beginning to show strongly. One comes under the general title of thermal management. Natural diamond has a thermal conductivity roughly four times superior to that of copper, and it is an electrical insulator: it should therefore come as little surprise to learn that CVD diamond is now being marketed as a heat sink for laser diodes and for small microwave integrated circuits. The natural extra- polation of this use in circuit fabrication ought to be higher speed operation, since active devices mounted on diamond can be packed more tightly without overheating. Reliability can be expected to improve also since, for a given device, junction temperatures will be lower when mounted on diamond.CVD diamond is also finding applications as an abrasive and as cutting tool insert^.,^,^^ In both this and the previous application, CVD diamond is performing a task that could have been fulfilled equally well by natural diamond if economics were not a consideration. However, there are many other applications at, or very close to, the market-place where CVD diamond offers wholly new opportunities. Wear resistant coatings are one such use. CVD diamond-coated drill bits, reamers, countersinks etc. are now commercially available for machining non-ferrous metals, plastics, and composite materials. Initial tests indicate that such CVD diamond-coated tools have a longer life, cut faster and provide a better finish than conventional tungsten carbide tool bits.The phrase ‘non-ferrous’ is worth emphasizing here since it reminds us of one of the biggest outstanding challenges in the application of diamond film technology -whether as a wear- resistant coating or as a fine abrasive. In any application where friction is important the diamond-coated tool bit will heat up and, in the case of ferrous materials (be it the tool substrate or the workpiece) the diamond coating will ultimately react with the iron and dissolve. Thermal mismatch is another potential problem with diamond-coated tool bits in situations where frictional heating is important. For example, attempts at extend- ing the effective working life of tungsten carbide tool bits by covering them with a wear-resistant film of CVD diamond have met with only limited success.This is not because of growth problems (it is perfectly possible to grow a film of CVD diamond on tungsten carbide substrates with 6% or 10% cobalt binder), but because the diamond film has a tendency to delaminate in order to relieve stresses arising as a result of the very different coefficients of thermal expansion of diamond and tungsten carbide. Hence the current interest in silicon nitride, a hard ceramic material with a smaller coefficient of thermal expansion, as a possible alternative substrate material for CVD diamond- coated cutting tools. Because of its optical properties, diamond is beginning to find uses in optical components, particularly as protective coatings for IR-optics in harsh environments.Most IR windows cur- rently in use are made from materials such as ZnS, ZnSe, and Ge, which, whilst having excellent IR transmission characteristics, suffer the disadvantage of being brittle and easily damaged. A thin protective barrier of CVD diamond may provide the answer, although it is more likely that future IR windows will be made from free-standing diamond films grown to a thickness of a few mm using improved high growth-rate techniques. How- ever, a major consideration when using polycrystalline CVD diamond films for optics is the flatness of the surface, since roughness causes attenuation and scattering of the transmitted IR signal, with subsequent loss of image resolution. Hence the current interest in techniques for smoothing diamond films we CHEMICAL SOCIETY REVIEWS, 1994 mentioned earlier.Several reviews considering the potential of CVD diamond in IR optics have appea~ed.~~.~~ The possibility of doping diamond and so changing it from being an insulator into a semiconductor opens up a whole range of possible electronic application^.^^^^^ However, there are a number of major problems that need to be overcome if diamond-based electronic circuits are to be achieved. Principal among these is the fact that CVD diamond films are polycrystal- line and hence contain grain boundaries, twins, stacking faults, and other defects, which all reduce the lifetime and mobilities of carriers. Active devices have been demonstrated using homoepi- taxially-grown diamond on natural or synthetic diamond sub- strates but, to date, there have been no corroborated reports of heteroepitaxial growth of device-quality diamond on non-diamond substrates.This remains a major limiting factor in the development of diamond devices. Nevertheless, the effect of grain boundaries and defects upon electronic carriers in the very best polycrystalline diamond films remains to be ascertained and, clearly, this possible route to active diamond devices cannot yet be ruled out. Another outstanding problem hindering potential diamond electronics is the inability to produce n-type doping. P-type doping is relatively straightforward, since addition of a few percent of B,H, to the CVD process gas mixture is all that is required to incorporate B into the lattice.However, the close packing and rigidity of the diamond lattice makes doping with atoms larger than C very difficult. This means that the dopants which are routinely used to n-dope Si, such as P or As, cannot easily be used for diamond, and so alternative dopants, such as Li are being investigated. One further difficulty that must be overcome if diamond devices are to be realized is the ability to pattern the diamond films into the required micron or even submicron geometries. Dry etching using 0,-based plasmas can be used, but etch rates are slow and the masking procedure complex. Alternative patterning methods include laser ablation,39 or selective nuclea- ti~n.~OThere are many variants of this latter process, but all involve trying to mask off certain areas of the substrate so allowing diamond to grow only in selected regions.A typical process scheme involves abrading a Si substrate, and then coating it with a thin layer of SiO,. This oxide layer is then patterned using standard photolithographic and etching tech- niques to expose areas of Si. CVD diamond is then grown, nucleating preferentially on the abraded Si areas rather than the oxide mask. Subsequent removal of the oxide mask in a chemical bath results in a patterned diamond film on Si, as shown in Figure 12. Figure 12 SEM showing patterned CVD diamond growth, using an oxy-acetylene flame, on a silicon substrate which had been pretreated as described in the text.(Reproduced, with permission, from J. Mafer.Res., 1992, 7, 2144.) THIN FILM DIAMOND BY CHEMICAL VAPOUR DEPOSITION METHODS-M. N. R. ASHFOLD ET AL. Despite these difficulties, CVD diamond-based devices are gradually beginning to appear, albeit with imperfect characteris- tics. Negative electron affinity cold cathode devices, piezoelec- tric effect devices, radiation detectors, and even the first field effect transistors have all been reported recently, with the likelihood that some of the simpler devices will become commer- cially available in the near future. Another interesting new development in diamond techno- logy4‘ is the ability to deposit CVD diamond onto the outer surfaces of metal wires or non-metallic fibres (see Figure 13).Figure 13 SEM of a section through a 25 pm diameter W wire that has been coated in diamond using hot filament CVD. Such diamond-coated fibres show increased stiffness and strength over the non-coated fibres, although quantitative mea- surements of coated fibre properties have yet to be presented. If growth rates can be increased to economically viable levels, such diamond fibres may find uses as reinforcing agents in metal- matrix composites, allowing stronger, stiffer and lighter load- bearing structures to be manufactured for use in, say, aerospace application^.^' Furthermore, etching out the metal core of the diamond-coated wire using a suitable chemical reagent yields free-standing diamond tubes, or hollow diamond fibres (see Figure 14).These too have potential applicatlons for reinforcmg smart composites, since the hollow cores may provide conduits for sealant, coolant, or sensors to be placed into the reinforced structure.Figure 14 SEM of hollow diamond fibre made by coating a 200 pm diameter W wire with a -20 pm thick film of diamond and then etching away the metal core with hot hydrogen peroxide solution for 1 hour. 6 Summary Most of the scientific research effort into CVD diamond techno- logy has been concentrated within the past five years yet, already, some of the more obvious applications, such as cutting tools and heat sinks, have reached the market-place. With the current rapid rate of progress, it should not be too long before this fledgling technology begins to make a significant impact in many areas of modern life.However, several issues need to be addressed before this can happen. Growth rates need to be increased (by one or more orders of magnitude) without loss of film quality. Deposition temperatures need to be reduced by several hundred degrees, allowing low melting point materials to be coated and to increase the number of substrates onto which adherent diamond films can be deposited. A better understand- ing of the nucleation process is required, hopefully leading to an elimination of the poorly controlled pre-abrasion step. Sub- strate areas need to be scaled up, again without loss of uniform-ity or film quality. For electronic applications, single crystal diamond films are desperately needed, along with reliable tech- niques for patterning and controlled n- and p-type doping.At present, there is a huge amount of work being done throughout the world on solving these issues, and progress is being made seemingly on a daily basis. If this continues, the future for CVD diamond looks bright indeed. Acknowledgements. Financial support from the Department of Trade and Industry is gratefully acknowledged, as is the help, advice, and encouragement offered by the many other members of the University of Bristol ‘Diamond Group’. P. W. M. also thanks the Ramsay Memorial Fellowship Trust and British Gas for funding. We are also grateful to Professors W. B. White (Pennsylvania State University), P.Koidl (Fraunhofer-Institut IAF, Freiburg), and J. T. Glass (North Carolina State Univer- sity), for supplying copies of Figures 5, 8, and 12 respectively. 7 References 1 G. Davies and T. Evans, Proc. R. Soc. London, 1972, A328,413. 2 J. E. Field, ‘The Properties of Diamond’, Academic, New York, 1979. 3 G. Davies, ‘Diamond’, Adam Hilger, Bristol, 1984. 4 ‘The Properties of Natural and Synthetic Diamond’, ed. J. E. Field, Academic Press, London, 1992. 5 S. Matsumoto, Y. Sato, M. Tsutsumi, and N. Setaka, f.Muter. Scr., 1982, 17,3106. 6 F. G. Celii and J. E. Butler, Ann. Rev. Ph+vs.Chem., 1991,42,643, and references therein. 7 ‘Diamond Films and Coatings’, ed. R. F. Davis, Noyes, Park Ridge, N.J., 1993. 8 Diamondand Reluted Muteriuls, ed. J.P. Gavigan, 1992, 1, 1007. 9 P. K. Bachmann and W. van Enckevort, Diumond and Reluted Materials, 1992, 1. 1021. 10 Y. Shigesato, R. E. Boekenhauer, and B. W. Sheldon. Appl. P1ij.s. Lett., 1993, 63, 314. 11 P. W. May, N. M. Everitt, C. G. Trevor, M. N. R. Ashfold, and K. N. Rosser, Appl. Surf. Sci., 1993,68, 299. 12 P. K. Bachmann, D. Leers, and H. Lydtin, Diumond and Relutrd Materials, 1991, 1, 1. 13 T. R. Anthony, Vucuum, 1990,41, 1356, and references therein. 14 J. C. Angus and C. C. Hayman, Science, 1988, 241, 913, and references therein. 15 F. G. Celii and J. E. Butler, Appl. Phys. Lett., 1989,54, 1031, 16 C. Wolden and K. K. Gleason, Appl. Phys. Lett., 1993,62, 2329. 17 L. Schafer, C.-P. Klages, U. Meier, and K. Kohse-Hoinghaus, Appl. Phys.Lett., 1991,58, 571, and references therein. 18 S. Williams, D. S. Green, S. Sethuraman, and R. N. Zare, f.hi. Cliem. Soc., 1992, 114. 9122; T. G. Owano, C. H. Kruger, D. S. Green. S.Williams, and R. N. Zare, Diumondund Reluted Muterial.\, 1993, 2, 661. 19 C. D. Clark, A. T. Collins, and G. S.Woods, in ref. 4, pp. 35-80, and references therein. 20 D. S. Knight and W. B. White, f.Muter. Res., 1988, 4, 385. 21 N. M. Everitt, A. M. Cock, P. W. May, K. N. Rosser, and M. N. R. Ashfold, ‘Proc. 3rd Int. Symp. Diamond Mater.’, Honolulu, May 1993, Electrochemical Society, Pennington, N.J., U.S.A., 1993, p. 955. 22 S.J. Bull and A. Matthews, Diamondund Related Materials, 1992,1, 1049. 23 Y. Sat0 and M. Kamo, in ref. 4.pp. 423-469, and references therein. CHEMICAL SOCIETY REVIEWS, 1994 24 C Wild, P Koidl, W Muller-Sebert, H Walcher, R Kohl, N Herres, R Locher, R Samlenski, and R Brenn, Diamond and Related Materials, 1993,2, 158 25 S D Wolter, B R Stoner, J T Glass, P J Ellis, D S Buhaenko, C E Jenkins, and P Southworth, Appl Phys Lett, 1993,62, 1215 26 S Jin, J E Graebner, M McCormack, T H Tiefel, A Katz, and W C Dautremont-Smith, Nature, 1993,362, 822 27 B Lux and R Haubner, ‘Proc 2nd Int Symp Diamond Mater ’, Washington D C ,May 199 1, Proceedings volumes 91-98, Electro-chemical Society, Pennington, N J U S A . 1991,p 314 28 S J Harris, D N Belton, A M Weiner, and S J Schmieg, J Appl Phys , 1989,66,5353 29 M I Landstrass, M A Plano, M A Moreno, S McWilliams, L S Pan, D R Kania, and S Han, Diamondand RelatedMaterials, 1993, 2, 1033 30 B V Spitsyn, L L Bouilov, and B V Derjaguin, J Cryst Growth, 1981,52,219 31 M Yoshikawa, H Ishida, A Ishitani, T Murakami, S Koizumi, and T Inuzuka, Appl Phys Lett, 1991, 58, 1387 and references therein 32 Y Sato, H Fujita, T Ando, T Tanaka, and M Kamo, Philos Trans R Sac London A 1993,342,225 33 M W Geis, H I Smith, A Argoitia, J Angus, G -H M Ma, J T Glass, J E Butler, C J Robinson, and R Pryor, Appl Phys Lett, 1991,58,2485 34 B Lux and R Haubner, Philos Trans R SOCLondon A, 1993,342, 297 and references therein 35 C A Klein, Diamond and Related Materials, 1993, 2, 1024 36 P Koidl and C -P Klages, Diamond and Related Materials, 1992, 1, 1065 37 M Seal, Diamond and Related Materials, 1992, 1, 1075 38 I M Buckley-Golder and A T Collins, Diamond and Related Materials, 1992, 1, 1083 39 C Johnston, P R Chalker, I M Buckley-Golder, P J Marsden, and S W Williams, Diamond and Related Materials, 1993, 2, 829 40 J A von Windheim and J T Glass, J Mater Res , 1992,7,2144 41 P W May, P G Partridge, C A Rego, R M Thomas, M N R Ashfold, K N Rosser, and N M Everitt, J Mater Sci Lett , 1993, 12, xxx

ISSN:0306-0012    

DOI:10.1039/CS9942300021

出版商:RSC    

年代:1994

数据来源: RSC

摘要:

Kirkwood-Buff Solution Theory Derivation and Applications Kenneth E. Newman Department of Chemistry The King‘s University College 9125 -50th Street Edmonton Alberta Canada T6B 2H3 1 Preamble The development of rigorous theories of the liquid state and of multi-component liquid mixtures has always been hampered by both the complexity of the statistical mechanics and the complex nature of intermolecular interactions. There are two central themes within chemical thermodynamics. The first deals with internal energy or enthalpy. As an example let us consider the internal energy E of a monatomic solid subject only to pairwise interactions. We can write for E (on a per atom basis) where N(r)equals the number of atoms at a distance r from an arbitrarily chosen central atom and E(r)is the pairwise interac- tion energy at a distance r. The summation is taken over all atoms. This equation needs to be modified to take into account the random structure of a liquid (or gas) composed of molecules of species 1 Q 1/2)p,Jo47+g *(r)E(r)dr (1.2) where p1 equals the number density (i.e. atoms per unit volume) of 1 and gll(r) is the so-called radial distribution function (rdf) the probability of finding a molecule of species 1 a distance r from the central atom relative to the probability of finding a molecule of species 1 a large distance from the central atom. In mixtures rdfs between different components can also be defined. For example in a mixture of i and j there are three rdfs g,,(r),g,(r) g,,(r). The rdf can be more rigorously defined than given here and it is also defined for molecules with angular dependence in their intermolecular potentials by suitable angu- lar averaging. The rdf is also in principle measurable by X-ray or neutron diffraction experiments. Figure 1 gives examples of diffraction-derived rdfs for argon and water. The function is zero at very small distances (less than the molecular diameter) and near the distance of closest approach there is a pronounced peak. The area under this first peak relates to the number of nearest neighbours around any given molecule. The much more pronounced peak for argon corresponds to approximately ten nearest neighbours unlike water which has only four or so. The ~~ ~~ ~ Dr. Ken Newman was born in London England and pursued his studies at the University of Durham and Newcastle-upon- Tyne where he received a Ph.D. in Physical Chemistry under the supervision of Professor Arthur Covington. Following periods at the Universitt? de Lausanne Simon Fraser University and the Universite de Sherbrooke he is presently Associate Professor of Chemistry at The King’s University College (a privately -funded Christian University) where he continues his research interests on the relationship be- tween the thermodynamic and structural properties of both electrolyte and non-electrolyte sohtions. 31 4 3 h cY2 6 1 I I I I 0 1 2 3 4 rlo Figure 1 Radial distribution functions of liquid argon (solid line) at T = 84.3 K P = 71 kPa and water (broken line) at T= 227 K P = 101 kPa. u equals 0.34 nm for argon and 0.282 nm for water. (Reproduced with permission from F. Franks ‘Water’ The Royal Society of Chemistry 1983.) function oscillates to unity after three or four molecular dia- meters. The small peaks for argon at roughly two and three molecular diameters correspond to lattice-like packing that occurs in a dense fluid. For water there is a peak at a value less than two diameters which is strong direct evidence for the tetrahedral structure of water. A knowledge of g l(r) is central to our understanding about the ‘structure’ of the fluid. It is clear that in some sense gll(r)and E(r) must be correlated since a strongly repulsive energy would imply a low value ofg ,(r) and vice versa. In practice the radial dependence E(r)is chosen either empirically or from quantum mechanical calculations and the variation of g l(r)is estimated using various different ‘closure’ approximations. Such approaches are referred to as ‘integral equations’. An example of the other major theme in thermodynamics the second law is the equation AGO = -RTlnK. In this case the equilibrium constant K gives us information about the molecu- lar nature the ‘structure’ of the system which is related to the free energy without recourse to knowledge about the interac- tions occurring in the system. Kirkwood-Buff (KB) theory’ is analogous to this idea in that it relates the thermodynamic parameters to the structure without recourse to knowledge of the interactions in the system. The ‘structure’ parameters which occur in this theory are the so-called Kirkwood-Buff integrals G,,= J,“ 4w2[g,,(r)-l]dr The theory is unable to obtain rdfs only their integrals. The results of KB theory are relatively straightforward although the original derivation is somewhat complicated being formulated in terms of matrix algebra. Our aim in this article is to demystify KB theory by mapping out its theoretical background and derivation as well as giving various examples of the utility of the approach. 2 Theory as derived by Hall In 1971 Hall2 rederived much of KB theory using a more intuitive molecular thermodynamic approach As a prelude to the full more formal derivation we present with certain modifi- cations the approach of Hall Let us consider a very dilute solution of a solute S in a mixture of 1 and 2 Let us consider two regions in solution one containing a molecule of S and its sphere of influence It is not necessary to define precisely this sphere of influence except to make sure that it contains all 1 and 2 that has been perturbed by S Such a region of solution is shown schematically in Figure 2a Let the number of molecules of 1 and 2 in this sphere of influence equal N; and N;respectively Let us also consider a region of solution exactly the same size and shape as the first but far from any molecule of S and thus containing only unperturbed solvent mixture 1 and 2 Let the number of molecules of A and B in this region equal N; and N2(Figure 2b) Figure 2 (a) Schematic diagram of a solute S and its region of influence N;' and N; are the number of molecules of 1 and 2 in this volume (b) Schematic diagram of volume identical to that in (a) which contains neither a solute molecule nor solute-perturbed solvent molecules N; and N; are the number of molecules of 1 and 2 in this volume It is clear that the difference N'; -Nb relates to the KB integral of 2 around S Let us now consider unit volume of the dilute solution of S with 1 and 2 in osmotic equilibrium with unit volume of the mixture 1 and 2 We may write for the constant temperature Gibbs- Duhem equation for the two solutions The p1 and pol refer to the number densities of 1 in the two solutions At this point Hall made the eminently reasonable assumption that the difference p1 -p; must relate immediately to the term Gsl Subtraction of equation (2 3) from (2 2) and substitution of (2 4) readily yields CHEMICAL SOCIETY REVIEWS 1994 Since we are dealing with very dilute solutions we may substi- tute for the limiting expression for osmotic pressure* which gives In the limit that ns goes to zero the term on the left hand side of equation (2 7) is simply the derivative of the standard chemical potential of S Thus This equation is readily extendable to any number of solvent components simply by adding further terms to the right hand side The equation which is readily obtainable from the full KB theory is applicable to the effect of change of solvent compo- sition on solute chemical potential which experimentally are measured as free energies of transfer primary medium effect salting in/out effects It has also been used in conjunction with transition state theory to discuss solvent effects in kinetics We will discuss the use of this equation below In order to develop equations that could be used for solutions that were not infinitely dilute Hall introduced the clever idea of considering what would happen if solute S were identical in all of its chemical properties to one of the solvent components e g component /3 but is in some way distinguishable It is clear that Gsl = G2 and Gs2 = G2 (2 9) We now wish to make the connection between the chemical potentials of S and 2 It can be readily shown that they are related by Substituting equations (2 9) and (2 10) into (2 7) we obtain This equation which has great utility in interpreting activity coefficients and osmotic pressure measurements is also obtain- able using the rigorous theory Its use will be discussed below 3 Kirkwood-Buff Theory The full derivation of Kirkwood-Buff theory arises out of Grand Canonical Ensemble statistical thermodynamics3 and as such is not very accessible to many chemists In this section we wish to present an outline of the derivation as well as some background theory on the ensemble so as to permit the reader to comprehend the original work In order to help the reader Figure 3 gives a flowchart of the ideas to be presented The original work was given in matrix algebra form for an n-component mixture and whilst it is an extremely elegant formu- lation it can be very daunting to the neophyte Thus we shall develop the equations explicitly for a binary mixture and simply point out the matrix nature of the equations at suitable points in the development In fact our initial development of Grand Canonical Ensemble statistical mechanics commences with a * Throughout this work the theoretical development 1s performed on a per molecule rather than d per mole basis Thus we use Boltzmann s constant kB rather than the gas constant R When comparisons are made with experimental ddtd on d per mole basis it is simply necessary to change ke for R and the resultant equations are automatically transformed into a per mole basis KIRKWOOD-BUFF SOLUTION THEORY DERIVATION AND APPLICATIONS-K E NEWMAN THEORETICAL SECTION THEME The Grand Canonical Ensemble for a pure The Grand Canonical Ensemble partition Calculation average properties between (N2)and Generalization of the ideas to include a - 1 I Relation between g 2(r) and the term I 1 I Conversion of the term (aN1/ap2)TVpr to (ap2/aN1)TVN2 1 Figure 3 Flowchart of various steps in theoretical development of Kirkwood-Buff theory pure liquid and is only generalized to a binary mixture as a second step 3.1 The Grand Canonical Ensemble Gibbs developed the ideas of statistical mechanical ensembles so as to forge a formal connection between the thermodynamic properties of a system and its mechanical properties By mecha- nical properties we mean those properties which are peculiar to the individual atoms and molecules such as velocity kinetic energy etc An ensemble is a conceptual collection of an extremely large number of systems each one constructed so as to be a replica at a thermodynamic level of the actual thermodyna- mic system of interest However since the numbers of molecules and the possible quantum (or classical) states are extremely large indeed the various systems will not be identical on the molecular level In order to forge the link between the mechanical and thermodynamic properties two postulates are necessary The first (often called the ergodic hypothesis) simply states that the longtime averages of the mechanical properties equal the ensemble averages The second hypothesis states that subject to the external constraints on the system all of the quantum (or classical) states are a prior1 equally probable The Grand Canonical Ensemble consists of systems all of volume Vimmersed in a bath at constant temperature Twith the REPRESENTATIVE EQUATIONS FIGURES ETC FIGURE 4 Equations Equations Equations such as 3 363 39 walls of each of the systems being permeable to the molecular species The ensemble thus acts as a reservoir of molecules and thus each system is characterized by the thermodynamic vari- ables V T and p (the chemical potential) See Figure 4 Note that the number of molecules N within the systems is not fixed but fluctuates around an average value (N) Similarly the energy of any system is not fixed It is important to note that the word 'fluctuation' which is frequently used in the context of the Grand Canonical Ensemble has a highly specific meaning It Figure 4 Schematic representation of a Grand Canonical Ensemble Each system is characterized by a fixed volume chemical potential and temperature However the walls between each system are perme- able to both energy and matter has nothing to do with time-dependent changes in a parameter. It simply means that different replicate systems will have differ- ent values of the parameter of interest. 3.2 The Grand Canonical Ensemble Partition Function As mentioned above for any system in the ensemble neither the total number of molecules Nnor the total energy Eis fixed. They fluctuate. As molecules pass in and out of any of the systems both N and the quantum state of the system j change. In this section we wish to explore the consequences of these fluctuating properties; in particular we wish to consider the expression for the probability that a given system is in a certain state i.e. the probability P(Nj)that the system contains exactly N molecules and is in the quantum state represented by the index j. By expanding out ln[P(Nj)] as a Taylor expansion in both Nand the energy E and considering two members of the ensemble which are widely separated and hence independent it can be shown that:4 where K,,8 and y are as yet undetermined constants. It turns out that = l/kgTand y = p/kBTwherep is the chemical potential. We require that the probability is correctly normalized i.e. the sum of the probabilities for all values of N andj equals unity. Thus we obtain from equation (3.1) and hence where Z is referred to as the Grand Canonical Ensemble partition function. Z is a function of both Tand p and is evidently a thermodynamic state function. The probability of a system being in the state (j,N)is thus 3.3 Calculation of Average Properties The average value of any set of replicate values is simply the value observed multiplied by the probability of occurrence and then summed over all of the set. Thus for N we may write It is interesting to compare equation (3.3) and (3.5). If we differentiate equation (3.3) with respect to y we obtain the dividend of equation (3S).(Several other important manipula- tions of the Grand Canonical Ensemble follow from this self- same property of differentiation of exponentials.) Thus we can write The distribution of Naround the mean value (N) is very narrow (see below for further details) and may be approximated by a Gaussian distribution whose variance 0%is given by where T is the total number of terms in the sum. As we shall see next w& is very closely related to (d(N)/i3p)T,,. It is also related closely to the radial distribution function; once we have extended the treatment to multicomponent systems we will have CHEMICAL SOCIETY REVIEWS 1994 all of the keys of the puzzle necessary to connect p to g(r) the basis of Kirkwood-Buff theory. The ensemble average (N2)can be written If we recast equation (3.5) in the form we readily obtain by differentiation with respect to y (at constant Tand V) If we now substitute equations (3.6) and (3.8) into equation (3.9) we obtain (having removed the ensemble average ( ) from around the N in the aN/+ term) Finally we can transform this equation through constant tem- perature thermodynamic equation dp = (V/N)dp to give (3.1 1) where K is the isothermal compressibility[= (+ l/N)(dN/ d~)~,~ This equation is one of the cele- = (-l/v)(dV/i?p)T,N]. brated equations of the Grand Canonical Ensemble. As an example of the use of this equation for liquid water at 298.15 K if we choose the volume to equal the molar volume then from the experimental values of K (4.53 x 10-lo Pa-’) we find uN= 1.93 x loll. This is the width of the distribution of values of Nwhose average value is obviously 6.022 x 1 Ot3. The fluctua- tions in N and hence in concentration are only one part in 3 x 10l2. 3.4 Generalizationto a Binary Mixture If we refer back to Section 3.1 we see that the changes necessary to accomodate a binary mixture 1 + 2 in place of a pure component in the setting up of the ensemble are insignificant. We require simply that the walls of each system are permeable to both components. The natural variables are in this case the temperature T,the volume V,and the two chemical potentials pl and p2.The equation for the partition function thus becomes Probability expressions occur in an analogous manner. How- ever crucial to the development of Kirkwood-Buff theory are terms of the form (N1N2)-(Nl)(N2),corresponding to co- variance terms in statistics which measure the correlation of fluctuations between 1 and 2. As an example if 1 and 2 were to have an affinity for each other then we would expect to see any positive fluctuation in N to be occurring simultaneously with a positive fluctuation in N,; this would be manifest as a positive co-variance. Here we begin to see some form ofchemically useful information emerging from the theory. Thus following the idea suggested above we obtain relatively straightforwardly The corresponding terms for (N:) and (Nf)are identical with those of equation 3.10. KIRKWOOD-BUFF SOLUTION THEORY DERIVATION AND APPLICATIONS-K E NEWMAN 3.5 Relationship between (N&) and g,,(r) The relationship between the composition fluctuation terms such as (N N2) and the radial distribution function g,2(r)have been dealt with at various levels of rigour in different texts In this review we will present a somewhat informal argument in order to limit the scope of this article somewhat Let us choose at random a molecule of species 1 in a mixture of 1 + 2 and consider the local (number) density of molecules of species 2 at a distance rp12(r) In a shell of thickness 6r the total number of molecules will be hr2plz(r)6r The number N ,of molecules of species 2 within a large volume Vdefined by the large distance R (R-+ a)from the reference molecule [ V = 4/3)nR3]is ~1 =21:477r2p12(r)dr (3 17) The number of molecules of species 1 in this region equals Vp and thus the product of molecules of species 1 and 2 within this region equals (N N ) VP,J:477r2pl2(r)dr = (NIN2) (3 18) The corresponding values of (N ) and (N2) are then given by and noting that p12(r) relates to the radial distribution function g*z(r) Since at large values of R p 2(r)should tend to pz (z e there is no correlation between 1 and 2 and glz(r) tends to unity) then the upper limit of equation 3 35 can be set to infinity The last equality comes from equation 1 3 where GI is the Kirkwood- Buff integral We now notice that (N1N2)-(N1)(N2) increases with V If we return to the above derivation as applied to the term (N:) we need to note one important difference When applied to molecules of species 2 around 1 equation 3 17 represents the number of molecules of species 2 in the volume If 1 and 2 are identical then we must add one molecule so as to include the central molecule Thus we write Repeating the above treatment for species 1 around species 1 yields (N:) -(Nl)2 = Vp;1:4nr2kl1(r) -l]dr + Vp = VpiG + Vp (3 23) 3.6 The Problems of Equations 3.14,3.15 and 3.16 These celebrated equations of Grand Canonical Ensemble sta- tistical mechanics link the conventional thermodynamic proper- ties to the ensemble fluctuations However they are somewhat problematical since virtually all thermodynamic measurements involve measuring a chemical potential as a function of compo- sition and not by fixing the second chemical potential as is required in these equations Thus our problem at this stage is simply to recast the equations to give terms of the form (aN,/ dpI)TYN,and not (aN2/8p1)Tvp2Using the rules of partial differentiation we may write for example The permuter rule of partial differentiation then gives Combining equations 3 24 and 3 25 and rearranging we obtain (at constant Tand V) where we have noted the definition of chemical potential implies that (apz)/dN,)N2equals (dp,>/aN2) Similar expressions are easily obtainable for (dlv,/ap,)p2and (dN,/i?p,) If we refer to the original work of Kirkwood and Buff written for an n-component mixture we find equation 3 26 written in determi- nant form [their equation 81 where A is the determinant whose factors are The determinant is obtained by removing row CL and column /3 from IA 1 In summary then for a binary mixture 1 + 2 there are three equations which relate the fluctuations to the conventional chemical potentials (equations such as 3 26) and three equa- tions which relate the fluctuations to the integrals of the radial distribution functions (equations such as 3 21 and 3 23) We now wish to solve equation 3 26 and the other two corresponding equations explicitly for the terms such as (i3pl/ aN2)N,and we readily obtain equations such as (3 29) which are the determinantal equations given by Kirkwood-Buff (their equation 9) At this point in the development we should note that the only problem with equations such as 3 29 is that they hold at constant volume and the final step is to recast them into constant pressure form and to put them into useful form 3.7 Algebraic Manipulation to Obtain Useful Equations It is perhaps useful at this stage to consider what we can expect from our equations They have all been derived at constant temperature and so the theory as cast can only refer to free energies and not to entropies (the temperature derivative of free energy) nor to enthalpies We have in equations such as 3 29 three independent expressions and thus would hope to be able to obtain equations for three useful parameters The Gibbs- Duhem expression tells us that for a binary mixture we have only one independent chemical potential and one independent partial molar volume A third parameter could be osmotic pressure (treating one of the two components as solute the other as solvent) However osmotic pressures can be converted into chemical potentials with the aid of a partial molar volume and the solution compressibility Thus the three 'naturally occur- ring' thermodynamic parameters can be considered to be either the set chemicalpotentialpartzal molar volume compressibility or the set chemicalpotentzal partial molar volume osmotic pressure For either of the two sets the resulting equations arise as different combinations of the terms Glz G,,,and G In order to convert equation 3 29 from a constant volume to a constant pressure expression we may write using the rules of partial differentiation (%) =(%) +(%) (*)(330)aN2 VTN aN2~TN TN N aN2 T VN The last term of the above equation can be rewritten as Equation 3 30 thus becomes [Kirkwood and Buffs equation 1 1) (&) =(&) +y2 (3 32)aN2 VTN aN2~TN KV There are of course two other very similar equations to 3 32 Final generation of the useful equations requires use of the Gibbs-Duhem equation It is readily shown by combining equations 3 23 and 3 33 and using the relationship V = N,V + N,V This then gives us the required expression for the isothermal compressibility in conjunction with equations of the form of 3 29 Similarly we can obtain from equations 3 23 and 3 33 N1(%) + N2(!?!?) VTN =".K (3 35)dN2 VTN This equation in conjunction with 3 34 gives the final expression for the partial molar volume with expressions such as 3 29 Finally with expressions for both compressibility and partial molar volume we can obtain an expression for (dp,/i?N,)~,N from equation 3 32 and such as 3 29 The osmotic pressure expression is obtained by conventional thermodynamic manipu- lation of the other quantities and is not derived explicitly here To summarize the equations of Kirkwood-Buff theory as applied to a binary mixture 3.8 Electrolyte Solutions One of the important constraints with the theory of electrolyte solutions is that of electroneutrality As well as requiring the well-known condition that in any given volume of solution the total charge on the cations plus the anions must equal zero it requires some other important conditions First in terms of Debye-Huckel ideas the charge on any given ion plus that of its ion atmosphere must equal zero Thus for the cation M in a 1 1+ electrolyte MX in solvent 1 we may write 6-8 with a similar equation for the anion X-which implies that GMM = Gxx In addition the solvent around the ions is corre- lated such that the total solvent correlated around a central ion is due to the solvent around the ion plus the solvent correlated with both the cations and anions that are around the central ion Hall6 has treated this problem by arguing that the solvent CHEMICAL SOCIETY REVIEWS 1994 correlations are much shorter range than the ion-ion interac- tions Thus he defines in dilute solution a slightly different KB integral G&l = 4rrrz1:[g~,(r)-l]dr (341) where R is chosen to be sufficiently large to include all ion- solvent interactions but sufficiently small to exclude ion-ion interactions Thus we write for GM1 with a similar equation for anion-solvent interactions If these equations are combined with equation 3 40 we can show that GM1 equals Gxl which is the normal way of expressing this other electroneutrality condition Many of the salt properties of interest refer to infinite dilution and it is of interest to consider the behaviour of the various KB parameters in this domain by using the Poisson-Boltzmann radial distribution which occurs in Debye-Huckel theory for the ion-ion parameters It has been showng that for a 1 1 electrolyte such as NaCl GMM is given by where A and Bare the conventional Debye-Huckel parameters with a the distance of closest approach and V,the excluded volume of the ions We note in passing that the limiting behaviour would arise if a more rigorous treatment were to be used in place of the Debye-Huckel approach We may note that GMM tends to infinity as the salt concentration tends to zero but that pMGMM equals -0 5 in the same limit Note also that pXGMX equals + 0 5 In order to apply Kirkwood-Buff theory to a 1 I electrolyte we simply replace the solute-solute term with GMM and the solute-solvent term with GMl As we shall see below this procedure whilst not proved rigorously guarantees that the derived equations have the correct form I e they exhibit ion additivity where required give the correct limiting form and the correct Debye-Huckel limiting behaviour For higher charge type electrolytes related expressions are obtained * More rigorous methods of treating electrolyte solutions are available but are generally formulated in terms of matrix algebra O 4 Comparison with Experiment 4.1 Introduction Equations 3 36-3 39 are the central results of Kirkwood-Buff theory with equation 2 11 being the corresponding equation from Hall's derivation It is very straightforward to obtain equation 3 39 from equation 2 11 by application of the Gibbs- Duhem relationship Pldll.1 + P2dP2 = 0 (4 1) We note that in the limit of pure I equation 3 36 yields which is the celebrated equation of Grand Canonical Ensemble which also arises directly from equations 3 11 and 3 23 We note in passing that for typical liquids the compressibilities are very small and G ,is very close to minus the molar volume (1/pl) For d dilute gas the value of G is simply twice the second virial coefficient Figure 5 shows the calculated values of GI,for a van der Waals fluid with parameters corresponding to carbon dioxide as a function of density for three temperatures At the critical point the compressibility and hence also G are infinite which is the origin of the maximum for the T/T,= 1 5 graph It is only at temperatures sufficiently high above the critical tempera- ture that we see a monatonic variation with density Before we discuss in detail the use of Kirkwood-Buff theory to interpret the thermodynamic properties of mixtures it is worth noting that the expression for the osmotic pressure (equation KIRKWOOD-BUFF SOLUTION THEORY DERIVATION AND APPLICATIONS-K E NEWMAN 063 1 -111 ,0 05 1 15 2 25 3 35 4 45 5 1/pl (dm3 mol-’) Figure 5 Variation of GI1 with density for van der Waals’ fluid corresponding to CO Top curve T/T = 1 5 middle curve T/Tc = 3 bottom curve T/Tc = 10 3 38) involves only solute-solute interactions The theoretical ideas concerning the use of osmotic equilibrium conditions to remove solvent interactions were developed by McMillan and Mayerl and has led to the phrase ‘McMillan-Mayer standard state’ whereby the solution of interest is in osmotic equilibrium with pure solvent Their theory is in many senses a precursor to KB theory and it is probably fair to say that it has found more acceptance than KB theory in that it concentrates attention on solute-solute interactions The beauty of KB theory is that through equations such as 2 11 one can work either in the McMillan-Mayer standard state (where dpl = 0) or for normal (otherwise known as Lewis and Randall) standard states where p1and pz are connected by the Gibbs-Duhem expression thus allowing the effect of solute-solvent interactions to be included in a rigorous fashion in any treatment of solutions 4.2 Binary Solvent Mixtures Ben-Naim13 was the first to consider the use of equations (3 36-3 39) to obtain the values of GI,(i,j = 1,2) in the binary inixture H,O + EtOH (1 + 2) Although the problem is in essence simple (we have three equations and three unknowns) there are many practical difficulties involved which relate to the requirements of highly precise data the need to curve fit data and then differentiate them so as to obtain partial molar volumes etc and then to combine the differentiated data so as to obtain the KB integrals The reader is referred to the original paper for specific details (cfthe work of Donkersloot14 which also includes an analysis of the HzO + EtOH system ) An interesting attempt at forging a connection between the other major theme in solution theory that of ‘integral equation’ approaches (see Preamble above) involves the use of the ‘Per- cus-Yevick’ closure approximation to estimate KB parameters for various model liquids l5 Donkersloot14 has explored the connection of thermodynamically obtained KB parameters with the information obtainable from X-ray and neutron diffraction data It turns out (see also ref 7) that the KB parameters arise very naturally in expressions for the diffraction experiment in the limit of zero scattering angle Thus either the thermodyna- mic data can be used to calibrate the diffraction experimental data or else the diffraction data can be used to estimate thermo- dynamic information In a homogenous liquid light scattering occurs due to composition fluctuations the same fluctuations manifest in KB theory To date most light scattering data have been interpreted in terms of association equilibria but recently efforts have been made to interpret light scattering results through KB theory l6 Work in the area of binary solvent mixtures discussed above has been somewhat hampered by a lack of intepretation of the values of the various KB parameters In particular there is criticism that the r2 weighting in the KB integral (equation 1 3) implies that it can be dominated by molecules beyond the first shell of neighbours l4 If this is the case then a simple ‘chemical’ interpretation is likely to prove to be daunting This problem will be discussed further in the section on molar volumes of electro- lyte solutions It seems likely that a value of G in a binary mixtuie considerably more positive than minus the molar volume of pure component 1 is evidence of strong self-associa- tion (see e g refs 13 and 14) Similar arguments could apply to GI where (as we will see below) in the limit of pure 2 G is close to minus the molar volume of 1 What is needed are analyses of systems that are well understood Examples could be hard spheres of different sizes very weakly interacting systems and mixtures of single H-bond donors and acceptors where conven- tional analysis using classical thermodynamics and association equilibria already give unequivocal results 4.3 Free Energies of Transfer The variation of the thermodynamic properties of a solute in dilute solution as the solvent composition is varied is of major importance throughout chemistry impinging as it does on solvent effects on solubility kinetics equilibria pH etc Much of the work in this area has involved the idea of preferential solvation where for example the composition of solvent close to the solute in a binary solvent mixture 1 + 2 will be different from the bulk solvent composition Thus for example if solvent component 1 is found preferentially close to the solute then we expect that the solute chemical potential will be lower in 1 than in 2 Preferential solvation can in principle be estimated spectro- scopically and here is apparently a method for obtaining ther- modynamic information for single ions (an idea not in accord with classical ideas about thermodynamics) Earlier work used the ideas of successive equilibria of discrete solvates but the rigorous development of such ideas did require extra-thermody- namic assumptions In order to use KB theory for this problem we need to explore the consequences of the electroneutrality conditions discussed above (Section 3 9) on the equations of interest If we start with equation 2 I1 for a ternary mixture of S + 1 + 2 and simply substitute the expression GM1 (and GMZ) in equation 3 42 for Gsl(and GS2) and GMM in equation 3 43 for Gss we obtain This equation has the required form for extrapolation to infinite dilution for a 1 1 electrolyte Equation 4 7 exhibits the required ion additivity relationship from which Hall6 defined a single ion medium effect Equations 4 7 and 4 8 have been used either in conjunction with spectroscopic data (NMR)ls or as a simple thermodynamic analysis l9 When combined with the Gibbs-Duhem expression (4 1) equation 4 8 becomes (4 9) where AG; is the free energy of transfer of ion M from pure 2 to+ solvent mixture of mole fraction x The solvent chemical potential p1 can be expanded out to give dPl = RTdIna = RTdlnx + RTlny (4 10) where a,,xl and y are respectively the activity mole fraction and mole fraction activity coefficient of 1 in the pure solvent mixture Providing that we have solvent mixture density data from which we can obtain an analytical expression for the mole fraction dependence of p1 by curve fitting and an analytical expression for yl then the only unknown on the right-hand side of equation 4 9 is Gk2 -Gkl Two possibilities have been explored First we can try to estimate Gk2 -Gkl from spectro- scopic data for the cation so as to allow integration of equation 4 9 We can do similarly for the anion and then addition of these two free energies of transfer would allow comparison with thermodynamically derived data The other option is not to separate the cation and anion terms in equation 4 7 but to develop a corresponding equation in differential form with respect to x1 and numerically to differentiate the experimental thermodynamic data so as to extract the term (Gk2 -G& + Gk2 -Gk,) directly l9 4.4 Partial Molar volumes The limiting partial molar volumes of a non-electrolyte solute at zero concentration is readily obtained from equation 3 37 (noting a change of nomenclature to a mixture of solute S in solvent 1) Lim(Js) = I/n + G -Gsl (411)PS -0 Note that the first two terms on the right-hand side of this equation equal KkBT If we refer to equation 2 8 and use the explicit pressure derivative we immediately obtain The reason for the difference between equations 4 11 and 4 12 is in the pressure derivative of the lnps term in equation 2 8 which simply equals minus the compressibility Virtually all experi- mental measurements of volumes refer to equation 4 11 As mentioned above the compressibilities of liquids are small far from the critical point and the difference between the two equations is then small Using the approach outlined above for free energies it has been shown19 that the corresponding equation for the partial molar volume of an ion M in solvent 1 is given by + Lim (GM) = -PM -+o which provides an unambiguous definition of single ion molar volumes Ion-solvent radial distribution functions have been obtained for a variety of ions in water by X-ray and neutron diffraction The results are generally interpreted in terms of a given number of water molecules N in a primary solvation shell If the total radius of ion plus solvation shell equals rand the total contribution of solvent beyond the first shell is negligible then it is clear that GG1 = Nipl -(4n/3)r3 (4 14) Using experimental values for the molar volumes of ions in water and knowing N it is possible to calculate Y for the above assumption It is found that for most cations the value of R is close to 1 A larger than the position of the first maximum in the radial distribution function (as it should be) which is strong evidence that (at least for ion-solvent interactions) long-range contributions to the KB parameters are negligible From equation 2 8 it is also apparent that the partial molar volume of the salt MX in a binary solvent mixture 1 + 2 equals CHEMICAL SOCIETY REVIEWS 1994 Providing partial molar volume data are available for the pure solvent mixture (or can be calculated from density data) as are compressibilities then in conjunction with free energy of transfer data (see Section 4 3 above) it is possible to extract (Gk + Gk,) and (Gh2 + Gk2)separately Figure 6 shows such a separation for the salt NaCl in H,O + MeOH l9 h cu gI 0 7--1 I I I I I I 0 02 04 06 08 10 Mole fraction MeOH(1) Figure6 Variation of G&l + Gg1 = Gsl (curve a) and GC2 + Gk2 = GS2 (curve b note change ofszgn)with mole fraction for solutions of NaCl (MX) at infinite dilution in methanol (1) + water (2) mixtures The effect of change of solvent composition on the rate of a chemical reaction has frequently been discussed in terms of free energies of transfer of reactants and transition state from one solvent mixture to another A + B In solvent I C' 1AGt" 1Act" .1 AG (4 16)A in solvent 1 + 2+ Cf The free energy of activation in solvent mixture 1 + 2 [ACi(1 + 2)]is thus given by dGf(1 + 2) -AGf(1) = dG,"(Cr) -dG,"(A)-dGy(B)(4 17) with a similar equation for volumes From a temperature-dependent kinetic study one can obtain the free energies of activation (dG*) and from a pressure dependence study the corresponding volumes of activation can be obtained Using an analysis similar to the one above it is possible to estimate the Kirkwood-Buff parameters (Gel -CAI-GBl) and (Gc2 -GAz -GB2) The use of these ideas to obtain important insight into the nature of the transition state and its solvation has recently been explored by Blandamer and co-workers 2o We note in passing that if the free energies of transfer of reactants could be measured directly by eg solubilities or electrode measurements then the KB parameters corresponding to the transition state could be obtained directly 4.5 Salt Activity Coefficients As was mentioned above the Poisson-Boltzmann radial distri- bution which arises in Debye-Huckel theory is a useful starting point for the application of Kirkwood-Buff theory to electro- lytes If equation 3 43 is substituted into equation 3 39 as discussed above we obtain after some simplification where If we convert equation 418 to activity coefficients we immedi- ately obtain with C = GM1 + V In the interests of simplicity we have not substituted the corresponding equation for GMl but treat C as a fitting parameter Figure 7 shows the best fit for activity coeffi- cients of KCI up to saturation with two unknowns 4,the distance of closest approach and C The value of A the Debye- Huckel slope is fixed at its normal value 061 1 04--$03-I 02-O'i f 0 02 04 06 08 1 12 14 16 18 2 Concentration' 5(mol dm3)' Figure 7 Observed (El) molar standard-state mean ionic activity coeffi- cients for KC1 in water and best fit to equation (4 20) with a = 2 95 8 and C = 2 95 cm3 mol-* (-) 4.6 Separation of KB Parameters for Electrolyte Solutions As with binary solvent mixtures (Section 4 2 above) all that is required to separate the three KB parameters for electrolyte solutions is three sets of thermodynamic data most likely molar standard-state activity coefficients density data so as to obtain partial molar volume and compressibilities We may note from equation 3 43 that for a 1 1 electrolyte the term for GMM always contains the term -0 5/PM and that once the expression has been inserted into equation 3 39 this term disappears from the expression for activity coefficients (equation 4 18) If we thus curve fit molar standard-state activity coefficients we obtain (C -GbM)which equals GM1 -GMM+ 0 5/PM This term may be expressed in terms of a power series We may rewrite equation for partial molar volumes (3 37) We thus substitute from the activity coefficient the expression for GMl -G into equation 4 19 which allows us to obtain (GI -GMM) It is then straightforward but somewhat tedious to substitute (GMl -G and (GMl -GMM)into the compressi- bility equation 3 36 to obtain an expression for GM1 and thence GMM and Gll (See ref 8 for details) In Figure 8 we show unpublished data for the separation of the three KB parameters for aqueous NaCl up to saturation 5 Postlude In this pedagogic review of the development and applications of KB theory no attempt has been made to be all encompassing with the literature Rather we give a somewhat biased and personal view of some of the more interesting and hopefully KIRKWOOD-BUFF SOLUTION THEORY DERIVATION AND APPLICATIONS-K E NEWMAN 39 I I 1 2 3 4 5 6 Concentration(mot dm-3) 064 0 Figure 8 Variation of the Kirkwood-Buff Parameters for NaCl in water as a function of concentration CMM + 0 5/PM (-) G (---) GrZ; + G:1 (---) fruitful avenues for the understanding of liquid mixtures There is one theoretical area that we have not touched upon namely that of direct correlation functions These functions closely related as they are to radial distribution functions are of central importance in a whole variety of liquid state and solution theories The only reason for not dealing with them in the meat of the text here is in the interests of brevity Direct correlation functions were introduced by Ornstein and Zernike in 1914 21 They argued that for a pure liquid the total correlation [g(r)-11 of molecule 1 on molecule 2 should be equal to a direct correlation C(Y) of molecule 1 on molecule 2 plus an indirect correlation of molecule 1 on all of the other molecules which were themselves correlated with 2 This problem can be treated in terms of a convolution integral 22 g(r) -1 = c(r) + pc(r)*[g(r)-I] (5 1) where * implies a convolution integral Such functions can be manipulated by Fourier Transform techniques and it can be shown that G = c,,+ PCllGll (5 2) where C = S47rrZc(r)dr (5 3) The application of Kirkwood-Buff-type theories which use the integrals C may be found in some of the other articles in the multi-author reference 8 6 References 1 J G Kirkwood and F P Buff J Chem Phys 1951 19,774 2 D G Hall Trans Faraday Soc 197 1,67,25 16 3 For a good review of the statistical mechanics of liquids and solutions see e g H L Friedman 'A Course in Statistical Mecha- nics' Prentice Hall New Jersey 1985 4 Reference 3 pp 14-16 5 For a more rigorous approach see e g Reference 3 pp 77-82 6 D G Hall J Chem Soc Faraday Trans 2 1972,68,25 7 J L Beeby J Phys C Solid State Phys 1973,6,2262 8 K E Newman in 'Fluctuation Theory of Mixtures' ed E Matteoli and G A Mansoori 1aylor and Francis New York 1990,373 9 K E Newman J Chem Soc Faraday Trans I 1989,85485 10 R J Perry H Cabezas Jr ,and J P OConnell Mol Phys 1988 63 189 1 1 Reference 3 p 112 12 W G McMillan and J E Mayer J Chem Phys 1945 13,276 13 A Ben-Naim J Chem Phys 1977,67,4884 14 M C A Donkersloot J Solution Chem ,1979,9,293 15 K Kojima T Kato and H Nomura J Solution Chem 1984 13 151 16 T Kato in 'Fluctuation Theory of Mixtures' ed E Matteoli and G A Mansoori Taylor and Francis New York 1990 p 227 CHEMICAL SOCIETY REVIEWS 1994 17 A. K. Covington and K. E. Newman Pure Applied Chem. 1979,51 Engberts I. M. Horn and P. Warrick Jr. J. Am. Chem. Soc. 1990 204 1. 112,6854. 18 A. K. Covington and K. E. Newman J. Chem. Soc. Faraday Trans. 21 L. S. Ornstein and F. Zernike Proc. Acad. Sci. (Amsterdam) 1914 I 1988,84 1393. 17 793. 19 K. E. Newman J. Chem. SOC. Faraday Trans. I 1988,84 1387. 22 Reference 3 pp. 137-140. 20 M. J. Blandamer N. J. Blundell J. Burgess H. J. Cowles J. B. F. N.

ISSN:0306-0012    

DOI:10.1039/CS9942300031

出版商:RSC    

年代:1994

数据来源: RSC

摘要:

Tetrathiafulvalenes as Building-blocks in Supramolecular Chemistry Tine JQrgensen Thomas Kruse Hansen and Jan Becher Department of Chemistry Odense University 5230 Odense M. Denmark Recently an overlap has emerged between two important fields of current preparative chemistry. These are the fields of supra- molecular chemistry and tetrathiafulvalene chemistry. It is the aim of this review to give an account of recent developments in this overlap-zone along with a brief background. I Conceptual Development of Tet rat h iafuIvaIe ne C hem ist ry‘ The investigation of tetrathiafulvalene (TTF) was initiated almost seventy years ago,2 but prior to 1970 only sporadic reports of TTF derivatives appeared in the literature and these concerned mainly non-targeted by-product^.^ From 1970 investigations were intensified following the observation by Wudl et uI.~that TTF forms a stable radical cation when treated with chlorine and in 1973 the first ‘organic metal’ TTF-TCNQ was discovered. In its neutral state TTF is an orange organic solid. However it readily loses electrons in the presence of oxidizing agents to form initially the purple radical cation and subsequently the yellow dication. Both species are aromatic in the Huckel sense and surprisingly stable. Soon after the discovery of TTFs’ donor-properties an important experiment was done in which TTF was mixed with an electron acceptor tetracyanoquinodimethane (TCNQ). The resulting black crystalline charge-transfer complex showed an electrical conductance which was several orders of magnitude higher than usual for organic solids. It also showed other characteristic features resembling metals thereby justifying the term ‘the first organic metal’. From now on the physical and Tine Jorgensen was born in Thisted Denmark in I965. She earned her Cand. Scient. degree in chemistry and biochemistry at Odense University in 199I. She is presently working in J. Becher ’s group in Odense as a Ph.D. student and is expected to receive her Ph.D. in February 1994. She has been working on functionalized tetrathia- fulvalene systems in M. R. Bryce’s group at Durham University and on catenate synthesis in J.-P. Sauvages group at Strasbourg. Thomas Kruse Hansen was born in Lem vig Denmark in I962 and received his Ph.D. in chemistry in 1992from Odense University where Professor Jan Becher was his supervisor. He continued working as a post-doctoral fellow in the same group until November 1993 when he was employedby Novo-Nordisk AIS. He Tine Jsrgensen Thomas Kruse Hansen 41 Figure 1 Oxidation of TTF affords stable cationic species. chemical investigation of TTF-derivatives became a new and important field which required interdisciplinary efforts world- wide. Many new physical terms have been coined as a result of these efforts. A particularly important result by Bechgaard and Jerome in 1980was the finding of the first organic superconduc- tor.6 Tetramethyltetraselenafulvalene hexafluorophosphate ([TMTSF],PF,) was found to be superconducting at 0.9 K at 12 kbar a property which had not previously been observed in any organic compound. The search for new organic conductors and superconductors is still active and an impressive number of TTF derivatives have been prepared and investigated. However the main emphasis of these investigations has been on the solid-state properties of the TTF-salts and only recently have several groups simultaneously initiated a study of TTF as a building-block in supramolecular and macromolecular chemistry. 2 Cation-sensit ive Tetrat hiaf ulvalenes During the past few years considerable attention has been paid to molecular systems containing a redox-active functionality and a host unit capable of cation binding. Such systems can be has been a visiting fellow in the research groups of Professor A. E. Underhill Wales Professor M. P. Cava Alabama and Dr. M. R. Bryce Durham. His research has been focused on combinations of supramolecular chemistry and TTF-chemistry. Jan Becher was born at Frederiksberg Denmark in 1939. He did graduate work and finished his thesis with Professor Ole Buchardt University of Copenhagen in I966. Research associate at N. Clauson-Kaas AIS 1966-68 and postdoctoral fellow at Synvar Research Institute Fulbright fellowship U.S.A. I968-69. Assistant professor 1969-89 awarded Dr. Scient. I986 and appointed docent (professor) in 1989 also at Odense University. Received the ‘Bjerrum Chemistry Award’ and gold medal in 1992for his work in organic syn- thesis and heterocyclic chemistry. Jan Becher is a member of the editorial board Journal of Heterocyclic Chemistry and of The Danish Natural Science Academy. His research interests cover hetero- cyclic chemistry especially sul- fur-containing heterocycles organic sulfur chemistry tetra- thiafulvalene chemistry macro- cyclic chemistry macrocyclic ligands and supramolecular chemistry. Jan Becher ~ ~~~ Table 1 Landmarks in TTF chemistry 1926 The first TTF derivative dibenzo-TTF was synthesized under a general study of five-membered ring systems2 1965 Deprotonation of 1,3-dithiolium salts afforded TTF derivatives for the first time.3 1970 First synthesis of parent TTF. TTF forms a stable purple radical cation on reaction with chlorine gas4 1973 First observation of metallic conductivity in an organic solid (TTF)(TCNQ).5 TCNQ = tetracyanoquinodimethane.Conductivity 500 (Q-cm)- l. 1980 Superconductivity observed in a TTF derivative tetramethyl tetraselenafulvalene hexafluorophosphate([TMTSF],PF,). 1980-TTF was derivatized extensively in the search for organic (super-)conductors. 1985-Macrocylic TTF-based systems investigated with the aim of making molecular devices sensor switches and shuttles. regarded as chemical sensors as well as redox-switchable ligands. Several systems incorporating aza- thia- and oxamacrocy- cles and -cryptates have been investigated; the redox-active unit has also been varied a great deal with ferrocenes and metal complexes frequently used. Figure 2 (a) Redox-detected sensor molecule; (b) redox-controlled ligand system. In a sensor system a 'transducer' built into the molecule responds to complexation in the host with a change in physical properties. Depending on the transducer this effect can be monitored by measuring shifts in colour redox-properties pH etc. A redox controlledlswitchable ligand system is essentially similar but the roles of the two parts are reversed so that a physical change of the transducer unit is imposed by an external stimulus e.g. electro-chemical or in other systems light in order to change the binding abilities of the ligand part and thereby to control uptake and release of guest molecules. In the TTF case it can be anticipated that oxidation of the redox-active transducer creates a positive charge which inductively decreases the binding ability of the attached host-unit towards cations. Whichever point of view is chosen it is evidently important to study a range of different host units 'transducer units' and linker typeslgeometries in order to establish their specificity of binding their sensitivity and the (electro-)chemical stability of such systems. 2.1 Planar TTF Derivatives 2.1.I Synthesis In 1985 a compound (7) was synthesized by Otsubo and co- workers,' in which a small crown ether was annelated to a TTF core. Since then TTF has been incorporated in a number of different macrocyclic systems,8 aiming at molecular sensors switches wires and shuttles in all cases exploiting the inherent electron-donor properties present in the TTF moiety. We have utilized the readily available 1,3-dithiole-2-thione- 4,5-dithiolate (1)9 as a key starting material in the synthesis of planar TTF-based sensor systems which generally consists of a CHEMICAL SOCIETY REVIEWS. 1994 central TTF moiety situated between two macrocycles. Reac- tion between dithiolate (1) and a range of electrophiles (2) generated a series of thiones (3) which were converted into the corresponding TTF systems (4) in a standard coupling reaction with trialkyl phosphites. O In Figure 4 seven TTF derivatives are shown which have been investigated as potential metal ion sensor systems. (4) Figure 3 General procedure for the preparation of planar macrocyclic TTF derivatives. (5)n = 1 (8)n = 4 (6)n =2 (9)n =5 (7)n = 3 n A sf .3S W WS II 00 Figure 4 Planar TTF containing macrocyclic systems. 2.1.2 'H-NMR Study of TTF-Crown Ether Derivatives.'O Two of the TTF derivatives with crown ethers as the host part (5) and (8) have been studied by 'H-NMR. The chemical shift positions of compounds (5) and (8) were measured in the presence of increasing amounts of sodium ions in order to gain insight into the changes taking place during complexation. The spectra are given in Figure 5. The spectrum of (5) (ring size equivalent to 9-crown-3) was independent of the salt concent- ration. In contrast the spectrum of (8)(equivalent to 18-crown- 6) showed a significant change in resonance position of the -SCH,CH,O-protons even for a relatively low ratio of Na+ ions per molecule. The fact that (5) did not show a shift in the NMR resonance positions lends support to the assumption that the changes observed for (8) were due to complex formation. 2.1.3 Cyclic Voltammetry Study of TTF-Crown Ether Derivatives A cyclic voltammetry study was undertaken with compounds (5)-(9) which revealed in all cases a two-electron reversible TETRATHIAFULVALENESAS BUILDING-BLOCKS IN SUPRAMOLECULAR CHEMISTRY-J. BECHER ET AL. c (8)+7eq Na+ d 1 3.0 2.9 2.8 1 I I i i 8 I I 2 Volt 0 10 I 3.9 3.8 3.7 3.6 3.5 3.4 3.3 3.2 3.1 PPm Figure 5 'H-NMR spectra of TTF-crown ether derivatives with and wlthout Na+ ions oxidation with E+(l)= 0.48 V and E+(z)= 0.64V. Similar experi- ments were performed with controlled amounts of alkali metal hexafluorophosphates added (in acetonitrile with 0.1 M TBAPF and a standard calomel electrode as reference). The observed shifts in Table 2 are maximum values achieved after addition of an excess of metal salts (approximately 250 equivalents). Table 2 Shifts in the first oxidation potentials on addition of metal ions +L1+ Na K+ Ag + >250equiv of mV mV mV mV -(5) 0 0 0 -(6) 0 0 0 -(7) 0 + 10 0 -(8) + 10 + 80 + 10 -(9) 0 + 15 + 60 (10) 0 0 0 + 170 (1 1) -50 -50 -60 + 90 The observed shifts were in all cases towards a more anodic (more positive) potential as expected for a repulsive coulombic interaction. Furthermore only E+(l,was shifted whereas EH2) remained unchanged. Compounds (8)and (9) showed the largest shifts and the selectivity profile appeared to be in good agree- ment with the general correlation between hole size in crown ethers and radius of alkali metal ions.' Thus (8) seems to have the highest affinity to sodium ions since this ion induces the largest shift whereas the largest shift for (9) was observed when potassium ions were added. Compared with an all-oxygen crown these sulfur-containing macrocycles had a rather low affinity. The magnitude of the shift was found to be dependent on the concentration of the alkali metal ion up to a certain limit. The cyclic voltammogram of (8) is shown in Figure 6. An explanation of the fact that only the first oxidation peak is influenced by alkali metal ions could be sought in the following model oxidation of the TTF unit produces a positive charge in close proximity to the ligand system which can be expected to repel the sodium ions thereby lowering the binding constant and literally 'pushing' the alkali metal ion out of the ring. It should however be emphasized that sodium ion exchange is very fast compared to the timescale of the CV as well as the NMR experiments. When the interaction with sodium ions is 4 5x10 0 Figure 6 Cyclic voltammogram of the TTF-crown ether derivative (8) excluded the second oxidation peak will naturally be identical to the neutral species as the CV indeed show. Scheme 1 summarizes the possible equilibria and suggests a principal route for the process occurring during a CV cycle The assumption that a complexed metal ion guest is expelled Kl L+Na+ LNa' L"+ Na' LNa"' Scheme 1 upon oxidation -and that it is taken up again after reduction confirms the theory such systems could serve as redox switch- able ligand systems and eventually be useful in transport studies. 2.1.4 Thiacrown Ether TTF Derivatives When similar experiments were performed with the sulfur crowned macrocyclic system (lO),l the cyclic voltammograms showed two reversible one-electron oxidation steps at E+(l)= 0.53 V and Ei(2)= 0.89 V (in dichloromethane SCE reference). No changes were observed when alkali metal ions were added. However large anodic shifts of the first oxidation wave (dEpal= + 170 mV) were induced by addition of silver perchlorate and in this case the second wave was affected as well (LIE, = 70 mV). Both oxidation waves (the peak values Epdl and E,,,) were changed but the reduction waves (E, and E,,,) remained unchanged (the peak values are in reversible cases related to the redox potential E+(l)= (E, + Epc1)/2. This indicates a different situation with respect to the kinetics of the system. Probably exchange is slower in this all-sulfur system compared with the previous example and the route outlined in Scheme 2 can be suggested for this system kl ' k2' LAg+ 7 L+.Ag+ 7 L++Ag+ L+Ag+ L+.+Ag+ 3L+++Ag+ +e +e Scheme 2 2.1.5 A Cyclic Voltammetry Study of a TTF-Cryptand In system (1 l) the TTF moiety was combined with two bicyclic systems affording the cryptand system (1 1) in good yield (Figure 4).The TTF unit was attached to an 18-crown-6 (O,N,) via an amide linker. Compared to the previous systems we have here a more flexible ligand likely to be capable of stronger binding due to the three-dimensional complexation sites. Larger binding constants may have a favourable effect on the electrochemical response to complexation and it should be noticed that com- plexation can take place in more than one geometrical fashion as proposed in Figure 7. Figure 7 Geometrically different complexation modes. 1 1 binding vs. intramolecular sandwich formation. In general the complexed metal ion can be expected to interact with the x-system and cause a change of the ligand system in at least two different ways both of which are relevant to the interpretation of the redox behaviour of the systems (i) The presence of a positive metal in the host will exert an inductive or through-bond eilect on the TTF system resulting in an increase of the oxidation potential (Eox). (ii) The complexation may induce an allosteric effect by a change of ligand conformation. A change in geometry could result in an increase in oxidation potential provided a bending of the TTF unit is forced on the system. Conformational changes involving rotation of the outer sulfur atoms are also likely to cause changes in redox behaviour. Furthermore if the ligand system can get in close contact with the TTF unit the resulting electron-electron repulsion could raise the energy of the HOMO orbital and consequently lower the oxidation potential. CHEMICAL SOCIETY REVIEWS. 1994 When Li ,Na ,or K ions are added to (1 1) both oxidation + + + waves are shifted to lower potentials while addition of silver ions results in a substantially different voltammogram. It is remarkable that the first oxidation wave is shifted towards a higher potential (dE+,,)= + 90 mV) whereas the second moves to a lower potential (LIE+(,)= -70 mV). A likely explanation is that the first electrochemical oxidation induces a conformatio- nal change due to the presence of the extra positive charge from the radical cation formed. In conclusion this final example of a planar TTF system showed a selective response to addition of metal cations follow- ing the trend of the previous examples; all these systems can be considered first-generation TTF-based sensor systems. V(16) Scheme 3 2.1.6 Modijxations in the TTF Unit Having established that it is possible to influence the oxidation potential of the TTF unit via crown ether-mediated electrostatic sodium ion interactions we undertook a study in collaboration with Dr. M. R. Bryce (University of Durham) investigating vinylogous TTF systems.14 Key intermediates in this type of chemistry are Horner-Wittig reagents of type (13) which can be generated from thiones of type (12). When (1 3) was mixed with a bis-aldehyde or ketone substrate in dry THF and reacted with LDA at room temperature we obtained 'extended TTF deriva- tives' of type (14)-(16). Alternatively the carbanion of (13) could be generated at -78 "C and subsequently treated with the aldehyde. Preliminary electro-chemical data suggest that these extended compounds are less sensitive to alkali ions than their TTF counterparts. Furthermore semi-empirical methods [on (14) and (16)] indicate that the HOMO orbitals in these cases are mainly localized in the centre of the molecules quite far away from the crown ether. This seems to be a likely explanation of the low sensitivity which is observed for (14)-(16). It also stresses the value of computational methods for predicting the sensiti- vity of donor systems. 2.2 Study of Distorted Macrocyclic TTF Derivatives 2.2.1 Synthesis In the previous sections we have in the main discussed the planar TTF derivatives with annelated macrocyclic moieties. A key step in their synthesis is the cyclization process which in principle can produce an unlimited number of oligomeric products. In Figure 3 only the monomeric (1 :1) product is shown and indeed by the use of high-dilution techniques it was possible to maxi- TETRATHIAFULVALENES AS BUILDING-BLOCKS IN SUPRAMOLECULAR CHEMISTRY-J BECHER ET AL mize the yield of this product For the pentaethylene and hexaethylene chains an efficient template effect (sodium ions were present) assured very high yields (> 80%) of the mono- meric product independently of dilution and solvent (DMF THF EtOH) However with shorter chains (varying in length from diethylene glycol to tetraethylene glycol) it was only possible to obtain 60-65O/0 of the monomeric products and only if the right mixing-technique and solvent were used In these cases higher oligomers were unavoidably formed as well (Figure 8) lS monomers dimers Figure 8 Formation of oligomers in the reaction of bis-nucleophiles with bis-electrophiles Scheme 4 When the reaction conditions were modified it was possible to isolate up to 25% of the dimeric (2 2) product In a procedure similar to the original synthesis of (17) by Mullen,' the dimeric products underwent an intramolecular coupling on treatment with triethylphosphite to produce novel TTF cage molecules (18)-(22) (Scheme 4),which incorporated oxygen [(18) and (22)] nitrogen [(20) and (21)] or sulfur atoms (19) in the side chains These molecules were constitutional isomers to the planar TTF crown ethers prepared by intermolecular coupling of the corresponding 1 1 adducts by now with the side chains running from one end of the TTF unit to the other In general the TTF unit of these compounds was distorted away from the planarity and the degree of distortion reflected the length and flexibility of the macrocyclic chains 2 2 2 Bis-TTF-macrocycles It should be noted that phosphite coupling normally proceeds in an intramolecular fashion giving rise to monomeric cage mole- cules We have observed intermolecular coupling in a few cases producing low yields of larger 'belt'-type macrocycles Recently Mullen et a1 'have reported a surprising reaction in which cage molecule (1 7) was electrolysed at a constant anodic potential followed by treatment with DMSO This procedure transformed the monomeric product (17) into a dimeric belt system in moderate yields (4653%) (Scheme 6) Applied to other monomeric TTF cages this method might lead to novel systems having the transducer and ligand units arranged in a geometrically different fashion In particular such bis-TTF systems might be useful for making redox-switched anion-sensors since they incorporate the possibility of generat- ing positive charges inside a macrocyclic cavity Scheme 5 Scheme 6 2 2 3 Cage Molecules as Potential Sensors In principle these TTF cage molecules could also bind metal cations to produce redox-active ligands In particular a CPK model of compound (22) looked attractive because the model appeared to be bowl-shaped with the TTF unit constituting the bottom and the glycol chains forming the rim However,judging from PDMS (Plasma Desorption Mass Spectrometry)'* and extraction experiment^'^ this molecule appeared to be a poor ligand toward alkali metal ions In Figure 9 the X-ray structures of (18) and (22) are given They clearly show that the cavity of compound (18) is very small and that the two oxygen atoms point out of the cavity The cavity size in (22) looked far more attractive but again the four oxygen atoms were pointing out of the cavity Although this might be a solid-state effect it appeared to be a likely explanation for the poor ligand proper- ties in solution Furthermore the bent TTF unit probably acted as a stretched 'spring' and made the glycol chains fairly inflexible 2 2 4 Factors aflecting the Oxidation Potentials Since tetrathiafulvalene derivatives are normally planar this series of compounds offered a unique opportunity for the study of oxidation potentials as a function of the degree of distortion and of the through-space effects of the heteroatoms in the macrocyclic chain by comparing the X-ray crystal structures with the experimental oxidation potentials and by using semi- empirical methods (in this case MNDO-PM3) This was done in order to achieve a better understanding of the factors influenc- ing various physical properties which may be helpful in a CHEMICAL SOCIETY REVIEWS. 1994 Figure 9 Ball and stick,and CPK view of molecules(18) and (22). Black carbon; medium shade sulfur; white oxygen. rational design of new compounds. The X-ray structures of compounds (17)-(21) revealed that the bending of the TTF unit was roughly similar. In contrast to ordinary TTFs which are red- orange these molecules were almost colourless. To a first approximation. oxidation of the less bent (22) (0.41 V:SCE) would be expected to be easier than oxidation of the very bent derivatives (1 7)-(21) because bending should cause a less efficient delocalization of the radical cation generated by oxi-dation. Indeed (17) (18) and (21) show oxidation potentials around 1.0 V. Because of its insolubility (19) was not studied. However (20) has an oxidation potential of only 0.27 V and is consequently more easily oxidized than (22) although it is significantly more bent. This fact was imposssible to explain using a model that assumed a simple correlation between oxidation potentials and bending. Previously it had been reportedz0 that the oxidation potentials of the planar TTF derivatives correlated quite well with the ionization potentials calculated by the MNDO program. This turned out to be true for the bent derivatives as well. The calculated ionization flavine electroactive unit NADH potentials correlated well with the experimental oxidation potentials of (17)-(22). Further insight was sought by calculat- ing the ionization potentials of tetramethylthio-TTF with increasing degrees of bending. This showed the expected increase in oxidation potentials with increased bending. Then. by calculating the ionization potential while keeping the tetrathio-TTF core fixed in the bent conformation and from this value subtracting the ionization potentials of the fully optimized structures with the appropriate side chains added we were left with a new value which could be called ‘the side-chain influence‘. It appeared that the glycol chains in (18) and (22) added (+ 0.18 eV; + 0.24 eV) to the ionization potentials whereas the bis-methylpyridine in (20) reduced the ionization potential ( -0.23 eV). The sulfur-nitrogen distance in com- pound (20) was very short (3.8A) and the effect could be thought of as an electron-electron repulsion raising the HOMO energy. Furthermore calculations on the face-to-face complex of TTF with benzene showed that the HOMO energy was raised as the distance between the two mystems decreased and the effect began to become significant when the distance fell below 4 A. In compound (17) the distance is 4.1 A. The trends in oxidation potentials could then be rationalized in terms of three funda- mental effects. Thefirst of these can be thought of as an inductive or through-bond effect the second as a HOMO localization due to bending and the third as a through-space effect in which electron-electron repulsion raises the HOMO energy. The above analysis suggests that TTF compounds with lower oxidation potentials may be made by forcing an aromatic system or the lone pairs of a heteroatom close ( <4 A) to the TTF T-system and at the same time preventing bending of the TTF moiety. In the present cage compounds the repulsion between the 7-system is relieved by bending the TTF part of the molecule. and the observed oxidation potentials are a consequence of the energetic balancing of these two effects. If however the ‘chain’ could be made sufficiently rigid and longer than the TTF molecule itself. it might be possible to achieve very low oxidation potentials. One such possibility could be to use anthracene or tetracene with sufficiently short connecting chains as linkers. 2.3 Sensor Systems based on Electroactive Units other than TTF 2.3.1 Introduction TTF is of course not the first system to have been used as the electroactive unit in sensor systems. A large variety of electro- responsive units have been used which can be divided into those containing reducible and those containing oxidizable centres. Both categories have been built into organic- and organometal- lic receptorihost assemblies to provide specific binding sites for cations anions. or neutral organic guest species. Since P. D. Beer has publishedz1 an excellent review on redox-active systems prior to 1989,we will only give a brief account of recent examples here in order to place the TTF-based systems in a proper context. guest cation / anion / organic Scheme 7 TETRATHIAFULVALENES AS BUILDING-BLOCKS IN SUPRAMOLECULAR CHEMISTRY-J BECHER ET AL 2 3 2 Sensor Systems with Reducible Electroactive Units Among the reducible electroactive units nitrobenzene- and quinone-based systems have been used extensively During reduction such systems become negatively charged As a conse- quence sensor models possessing reducible electroactive sites often show enhanced cation-binding after reduction With the presence of a positively charged cation in the host compartment a system will usually exhibit an anodic shift of the reduction potentials The magnitude of the anodic shifts is often strongly dependent on the charge/radius ratio of the alkali metal ions The shifts generally decrease in the order Li+ > Na+ > K+ reflecting the fact that ion pairing between metal cations and the respective reduced anion radical redox centre is stronger for smaller cations which posses a higher polarizing power R'OLN Cokel 1991 0 00 M+ Na+ K+ Greene 1990 /o 0 (0 oJ 0 Figure 10 Four sensor model systems containing different redox active units The anthraquinone-based sensor system made by Echegoyen et a1 22 in 1993 (Figure lo) had potential as a redox-switched shuttle for cations and electrons across membranes When the system was studied by cyclic voltammetry it showed two quasi- reversible redox pairs indicating that the two anthraquinone units acted independently and uncoupled First two radical anions were formed and then two dianions were generated After addition of 0 5-2 equivalents of sodium tetraphenyl borate two additional redox couples were observed simulta- neously at a more anodic potential corresponding to an easier reduction of the complex All four redox couples were resolved although this system was not expected to show a high binding affinity for sodium ions In most other electrochemical investi- gations with simple quinones only potential shifts (rather than a splitting into two couples) are observed as a function of the concentration of metal ion Distinct waves are normally only observed when the initial binding constant K of the neutral ligand system is large In such a case the differences of the formal redox potentials for the free ligand (p)and the complex (qmplex) can be used to calculate quantitative values of the electrochemical binding enhancement of metal cations according to the following equation 23 (K = binding constant for unreduced ligand K2 = binding constant for reduced ligand) It was possible to use equation 1 on the anthraquinone system to calculate the following binding enhancement values (for sodium ions) after reduction 1 4 x lo3 (K2/K1)and 7 5 x lo2 (K3/K2) K3 = the binding constant for doubly reduced ligand These enhancement values were calculated on the basis of the corres- ponding Et values for each of the waves and the enhancement was relatively large compared with structurally related systems This made the new system a promising redox-switchable ligand system since the enhancement value is of obvious importance in transport studies where cycling between reduction and oxi- dation should promote release and uptake of ions Another reducible redox unit was incorporated in Green's sensor system2 from 1990 This system resembled compound (7) having two 15-crown-S2-0 macrocycles as ligand part of the molecule Instead of TTF a Culll tetrathiolate unit (in such compounds a formal oxidation state III for copper is not unusual) was incorporated as the redox-active centre with either Na+ or K+ as counter ions Independently of the counterion a reversible reduction at -0 70 V and an irreversible oxidation at + 0 24 V were observed A relatively large shift in the reduction potentials was induced when an additional amount of alkali metal salt was added to the system Replacing tetrabutylammo- nium tetrafluoroborate by sodium tetrafluoroborate as support- ing electrolyte induced an anodic shift of 175 mV whereas smaller shifts were induced by using lithium or potassium salts The magnitude of the shifts of E+ and the order (Na+ > Li + > K+) indicated that a simple ion-pairing model alone could not explain the shifts and thus cation binding by the host was assumed to play a significant role in the processes In contrast to the previous anthraquinone example this system behaved in an electrochemically similar fashion compared to the TTF-based systems Instead of the appearance of a new resolved redox couple a shift in the position of the redox wave was induced by addition of alkali metal ions This reflected that the electrochemical steps 1 and 1' were averaged together Scheme 8 When K is small and the binding between the ligand system and the metal ion in question is weak only one wave is observed In such cases it was necessary to add a large excess of sodium cations in order to achieve the maximum shift The rather large shift value may be explained by the proximity of the cations bound in the host to the redox-active centre and the fact that dithiolene complexes are strongly delocalized allowing com- plexed cations to interact strongly with the redox-active centre 2 3 3 Sensor Systems with Oxidizable Electroactive Units The two remaining examples in Figure 10 both contain a ferrocene unit as the redox-active centre In the system reported by Gokel and co-workers2 in 199 1 ,ferrocene was linked to a 18- crown-N,-0 macrocycle forming a cryptand structure which proved to be an efficient sodium and potassium ion responsive assembly With only 0 25 equivalents NaClO added to the ligand system a second well-resolved redox couple appeared in the CV at a more positive potential (+ 188 mV) As the Na+ concentration was increased further the peak currents of the new redox couple also increased at the expense of the original couple which disappeared completely in the presence of 10 equivalents of Na+ cation The cathodic shift was indicative of a destabilization of the oxidized form of the complex compared to the reduced form This should allow the system to serve as a redox-switchable ligand because the oxidized form had a lower affinity for Na+ compared with the reduced form However it still possesed some affinity for Na+ as evidenced by the com- plete reversibility of the redox couple corresponding to the complex The binding ratio was calculated for Na+ K+ and Ca2+,using equation 1 Addition of lithium salts did not induce any new redox couples For sodium ions it was found that K,/ K2 = 1 5 x lo3 for potassium ions K1/K2= 1 2 x lo2 and for calcium ions K1/K2= 4 3 x lo4 The cyclic voltammograms after addition of 0 0 0 5 and 1 0 equivalents of Na+ are reproduced in Figure 11 -02 07 UV vs SSCE Figure 11 Cyclic voltammograms of Gokel’s ferrocene cryptand system after addition of (a) 0 equivalents (b) 0 5 equivalents dnd (c) I 0 equivalent of NaCIO The second ferrocene-based system shown in Figure 10 was made by Fabbrizzi et a1 26 in 1992 Their amido-linked podand system showed a surprising cathodic shift in the redox poten- tials when complexed with Ni2 + All cyclic voltammograms were made in aqueous solutions at pH 10 5 with NaClO present as electrolyte Before complexation oxidation took place at E+= +0 402 V (vs NHE) This peak was shifted -42 mV to E+= +O 360 V upon addition of Ni2+ Thus the close presence of a positively charged centre (Ni2 +)made it easier to form the positively charged ferrocenium subunit Electrostatic effects were invoked in order to explain this behaviour along with the fact that the system lost two protons during complexa- tion When the complex was formed the amide nitrogens would carry a formal negative charge leading to stabilization of the positively charged oxidized species The negative charge formally placed on the nitrogens would not be completely offset by the Ni2+ consequently oxidation would occur at a lower potential 3 A Molecular Shuttle with a TTF-station 3.1 Introduction As described in the previous sections the fundamental redox properties of TTF have been exploited in several sensor appli- cations A quite different type of system will be introduced here in which TTF’s ability to form charge-transfer complexes is as important to the principle as TTF’s unusual redox properties This new kind of device coined the ‘molecular shuttle’ by Stoddart et a1 ,27 works according to a principle by which two ‘stations’ are connected in a chain structure and encircled by a ‘shuttle’ A prerequisite for obtaining reasonable yields is that the shuttle and the station should be attracted to each other In the case where the electron-donating TTF serves as the CHEMICAL SOCIETY REVIEWS 1994 station the shuttle should be electron accepting This charge- transfer interaction has been demonstrated beautifully in the case2* where the acceptor (shuttle) is a cyclobis(paraquat-p- phenylene) tetracationic macrocycle and TTF is the donor (station) Mixing of the components afforded a crystalline charge-transfer complex whose structure was determined It turned out to have an interesting channel-type structure (Figure 12) where the tetracationic shuttle formed the channel and TTF occupied the central cavity The next step will involve the TTF being incorporated into a chain This type of chemistry was initially perfected with donors other than TTF but here only the TTF-containing version will be mentioned MeCN Figure 12 Crystal structure of the charge-transfer complex between TTF and the tetracationic macrocycle 3.2 The Rotaxane-based Shuttle containing a TTF Station The ‘shuttle’ made by Stoddart and co-w~rkers~~ was essentially a complicated rotuxane [The word rotaxane is derived from the Latin words rota meaning wheel and axis for axle ] One rotaxane they made incorporated a cyclobis(paraquat-p-phenylene)mac-rocyclic unit as the shuttle which moved on an axis containing one TTF station and two hydroquinol stations Since the shuttle was a good electron acceptor it was able to ‘thread’ the chain and was held in place by either one of the electron-donating stations (TTF or hydroquinol) by charge- transfer interactions Since TTF is a better electron donor (E+= + 0 5 V) then the hydroquinol units (E+= + 1 1 V) it is likely that the acceptor macrocycle will form a complex predo- minantly with the TTF unit The rotaxane shown in Figure 13 was then formed by blocking the two ends of the chain with 4- trityl-phenylether units which served as ‘stoppers’ Alternatively the shuttle could be assembled at the last step using a chain with the stoppers already in place The synthesis itself was an impressive example of ‘self-assembly’ where non-covalent forces were used to hold components together during assembly with relatively high yields as a result An NMR investigation of the fully assembled rotaxane showed that selective oxidation of TTF forced the shuttle away from TTF to one of the hydroqui- no1 units This was the result of the oxidized form of TTF (Ee a radical cation or a dication) being a less efficient donor than neutral TTF Furthermore the positive charge on TTF+ repelled the tetracationic acceptor macrocycle by coulombic forces Future applications can be envisaged following major developmental work in molecular data storage 4 Molecular Switches 4.1 Introduction The design synthesis and manipulation of molecular devices is the challenging subject of the field molecular electronics Over a TETRATHIAFULVALENES AS BUILDING-BLOCKS IN SUPRAMOLECULAR CHEMISTRY-J BECHER ET AL Figure 13 The TTF-containing 'shuttle system' 1 n a) rotaxane Figure 15 The TTF-containing switch off The system incorporated a TTF unit and a photosensitive unit The on/off function followed the oxidation state of the TTF moiety A Ru1l(bpy)Z system was chosen as the photosensitive + unit and it was attached to the TTF centre by a small vinyl linker b) shuttle system (Figure 15) Ruthenium 11 complexes of bipyridine are known to be fluorescent when exited at h = 450 nm giving rise to an Scheme 9 emission with h = 610 nm When the TTF present in the switch is ' \P Figure 14 A switch controls a function period of 60 years the scale of components has decreased from a typical device size of about 10-' m down to m However before further miniaturizing of components based upon semi- conductors is possible problems of over-voltage and heat exchange will have to be overcome A switch is an essential component in electronics which in principle can be almost any device (or ultimately a molecule) able to interchange between two different states The onlo#-component can be addressed by electric current radiation heat light etc 4.2 A TTF-controlled Molecular Switch Lehn and Goulle have recently reported an example29 of a molecular device capable of switching a fluorescent state on and Figure 16 The processes controlling the switch PHOTOSENSIBILJSER / REDOXACTNE CENTRE neutral the fluorescence is quenched by electron transfer from TTF to Ru2+* [E(TTF+/TTF) = 0 43 V E(Ru2+*/Ru+) = 0 84 V] When the TTF unit was oxidized to form the radical cation by applying a potential 100 mV higher than the first oxidation step of TTF the fluorescence was moderately restored However the intensity was only 30% compared to a control sample contain- ing Ru1I(bpy):+ It was not possible to return to the non- fluorescent state probably because of decomposition of the compound Similar results were obtained when ferrocene qui- none and hydroquinol were used as the redox active units Nevertheless this result is encouraging in the developing field of molecular devices 5 Supramolecular TTF-containing Wires made by Self Assembly The problem of addressing molecular electronic components is of great importance An obvious solution is to connect devices via 'electrical' wires of similar dimensions Recently a step towards TTF-based molecular wires was made by J~rrgensen et a1 30 By combining the TTF unit with lipophilic chains capped with large hydrophilic endgroups (similar to the gelforming arborol aggregates reported by Newkome3') Jsrgensen et a1 aimed at self-assembled structures where the conductivity was obtained by T-interactions between partly oxidized TTF reduchon NON-FLUORESCENT FLUORESCENT CHEMICAL SOCIETY REVIEWS 1994 OHJ NHO Figure 17 The molecular unit for ‘wire-formation’. molecules. In comparison with ordinary macroscopic wires the hydrophilic groups would resemble the insulation and the TTF- stack serve as the conducting copper core. A gel was indeed formed when the prepared compound (Figure 17) was dissolved in 25% EtOH in hot water. After cooling to room temperature an opaque orange-yellow gel was formed which could be oxidized to a green form by addition of aliquots of iodine in EtOH. Phase contrast light microscopy of the gel revealed textures of string-like structure with a length of several microns. These were also visible when crossed polarizing filters were used showing that the aggregates were ordered structures over distances of several microns. Transmission elec- tron microscopy on gels stained with phosphotungstic acid showed narrow string- or band-like structures with diameters ranging from 300-1000 I$. Single strings of the aggregates should have a diameter of around 50 I$ therefore the structures were interpreted to be superstructures consisting of several strings. UV+Vis)-NIR spectroscopy of the oxidized gel revealed a CT-band typical for stacked TTF molecules. Since the gross morphology is in accordance with a wire-type structure this is another encouraging result en route to commer-cial applications even though this currently may seem a remote goal. The field of molecular nanostructures is still young but it seems that an increased ability to control the self-assembly of such superstructures has already been achieved. QQ -f wire cable Figure 18 Self-assembly of the TTF derivative forms wires and cables. 6 A Bis-TTF-podand forms C6,-complexes The recently reported charge-transfer complex between a TTF derivative and c60 will be mentioned here as a final example of the versatility of TTF as a building block in supramolecular chemistry. Sugawara and co-workers3* synthesized a bis-BEDT-TTF derivative (Figure 20) that formed an intriguing tweezer-like structure when complexed with 0. Figure 19 Twin-donor TTF derivative. n Figure 20 ORTEP drawing32 of C surrounded by the twin donor viewed along the threefold axis of Ce0. As a result of the flexible ethylene bridges used to connect the two donor moieties the inter-planar angle was variable. In neutral crystals the two TTF’s were almost parallel taking a U-shaped conformation while the angle was 90”in C10 salts and the cavity of the podand-like host was increased even more when c60 was complexed. Acknowledgments. We wish to thank the Danish Natural Science Research Council (SNF) for support to TKH and Forskeraka- demiet for support to TJ. 7 References 1 Reviews on TTF chemistry (a)M. R. Bryce Chem. SOC. Rev. 1991 20,355; (6) G. Schukat A. M. Richter and E. Fanghandel Sulphur Reports 1987 7 155; (c) M. Narita and C. U. Pittman Synthesis 1976,489. 2 W. R. H. Hurtley and S. Smiles J. Chem. Soc. 1926,1821 and 2263. 3 H. Prinzbach H. Berger and A. Luttringhaus Angew. Chem. 1965 77,453. 4 F. Wudl G. M. Smith and E. J. Hufnagel J. Chem SOC.,Chem. Commun. 1970 1453. 5 J. P. Ferraris D. 0.Cowan V. Walatka and J. H. Perlstein J. Am. Chem. SOC. 1973,95,948. 6 K. Bechgaard and D. JCrome Sci. Am. 1982,247 50. 7 T. Otsubo and F. Ogura Bull. Chem. SOC. Jpn. 1985,58 1343. 8 (a) B. Girmay J. D. Kilburn A. E. Underhill K. S. Varma M. B. Hursthouse M. E. Harman J. Becher and G. Bojesen J. Chem. SOC. Chem. Commun. 1989 1406; (6) J. Becher T. K. Hansen N. Malhotra G. Bojesen S. Bswadt K. S. Varma B. Girmay J. D. Kilburn and A. E. Underhill J. Chem. SOC. Perkin Trans. I 1990 175. 9 K. S. Varma A. Bury N. J. Harris and A. E. Underhill Synthesis 1987,837. 10 T. K. Hansen T. Jsrgensen P. C. Stein and J. Becher J. Org. Chem. 1992,57,6403. 11 F. Vogtle H. Siege and W. M. Muller in ‘Host-Guest Chemistry l’ Topics Curr. Chem. 1980,98. 12 T. Jmgensen B. Girmay T. K. Hansen J. Becher A. E. Underhill M. B. Hursthouse M. E. Harman and J. D. Kilburn J. Chem. Soc. Perkin Trans. I 1992,2907. 13 R. Gasiorowski T. Jerrgensen J. Merller T. K. Hansen M. Pietrasz- kiewicz and J. Becher Adv. Mat. 1992,4 568. 14 T. K. Hansen M. R. Bryce and J. Becher manuscript in preparation. 15 T. K. Hansen T. Jerrgensen,F. Jensen P. H. Thygesen K. Christian- TETRATHIAFULVALENES AS BUILDING-BLOCKS IN SUPRAMOLECULAR CHEMISTRY-J BECHER ET AL sen M B Hursthouse M E Harman M A Malik B Girmay A E Underhill M Begtrup J D Kilburn K Belmore P Roepstorff and J Becher J Org Chem 1993,58 1359 16 J Rohrich P Wolf V Enkelmann and K Mullen Angew Chem 1988 100 1429 17 M Adam V Enkelmann H -J Rader J Rohrich K Mullen Angew Chem ,Int Ed Engl 1992,31,309 18 The PDMS method N Malhotra P Roepstorff T K Hansen and J Becher J Am Chem Soc 1990,112,3709 19 The picrate method introduced by Pedersen was used C J Pedersen J Org Chem ,1971,36,254 20 S Bswadt and F Jensen Synth Met 1989,32 179 21 P D Beer Chem Soc Rev 1989,18,409 22 L Echegoyen Y Hafez R C Lawson J de Mendoza and T Torres J Org Chem ,1993 58 2009 23 Only true for the two-wave situation see S R Miller D A Gustowski Z C Chen G W Gokel L Echegoyen and A E Kaifer Anal Chem ,1988,60,2021 24 M L H Green W B Heuer and G C Saunders J Chem SOC Dalton Trans 1990,3789 25 J C Medina T T Goodnow S Bott J L Atwood A E Itaifer and G W Gokel J Chem Soc ,Chem Commun 1991,290 26 G De Santis L Fabbrizzi M Licchelli P Pallavicini and A Perotti J Chem Soc ,Dalton Trans 1992 3283 27 P R Ashton R A Bissell N Spencer J F Stoddart and M S Tolley Synletr 1992,923 28 D Philp A M Z Slawin N Spencer J F Stoddart and D J Williams J Chem Soc Chem Commun 1991 1584 29 Veronique Goulle Ph D thesis Strasbourg 1992 30 M Jsrgensen T Bjsmholm P Sommer-Larsen L Lithen-Madsen L G Hansen K Schaumburg and K Bechgaard submitted to J Org Chem ‘Molecular Materials for Molecular Electronics’ Progress Report CISMI University of Copenhagen 1993 13 1 31 G R Newkome C N Moorefield G R Baker R K Behhera G H Escamilia and M J Saunders Angew Chem Int Ed Engl 1992,31,917 32 A Izuoka T Tachikawa T Sugawara Y Saito and H Shinohara Chem Lett 1992 1049

ISSN:0306-0012    

DOI:10.1039/CS9942300041

出版商:RSC    

年代:1994

数据来源: RSC

摘要:

Polyelectrolyte Materials -Reflections on a Highly Charged Topic John W. Nicholson Materials Group Laboratory of the Government Chemist Queens Road Teddington Middlesex TWl1 OL V U.K. and Dental Bio- Materials Unit King's College School of Medicine and Dentistry Bessemer Road London SE5 9PJ U.K. 1 Polyelectrolytes Polyelectrolytes are polymers having a multiplicity of ionizable groups that generally cause them to be water-soluble. When in solution they dissociate into polymeric ions and numerous small ions known as c0unterions.l The counterions tend to remain in the region of the polymeric ions because of electro- static attraction a phenomenon known as ion binding. There are numerous pieces of experimental evidence for ion binding such as the measured tendency of cations to travel in the same direction as a polyanion in an electric field. In aqueous solution the molecules of a polyelectrolyte can take up a number of conformations because rotation of the bonds making up the polymer backbone is reasonably unhin- dered. This ease of rotation makes polyelectrolyte molecules flexible. The size and shape of the conformation depends on the state of neutralization of the polyelectrolyte. When the charge on the polymer is low i.e. at low or zero degrees of neutralization the polymer adopts a compact coil shape. As neutralization pro- ceeds and the charge on the polymer increases electrostatic repulsion causes the conformation to change. The compact coil unfolds and at high levels of neutralization adopts an expanded rod-like conformation. Poly(acry1ic acid) for example with a degree of polymerization of 1000 adopts a coiled conformation at its natural pH that is essentially spherical and has a radius of 20 nm. At full neutralization it becomes a rod with a maximum extension of 250 nm2. Changes of this kind affect the solution viscosity which for the practical materials described in the current article has a profound effect on properties such as film formation or measured setting times. This review is particularly concerned with two sets of mater- ials of growing technical and commercial importance namely waterborne paints and polyelectrolyte cements for various medi- John Nicholson C.Chem. F.R.S.C. graduated with a B.Sc. in applied chemistry at Kingston Polytechnic (now University) in 1977 then worked for a just over a year at the Paint Research Association in Teddington. Between 1978 and 1983 he was at South Bank Polytechnic (also now University) London where he obtained his Ph.D. in 1981 then worked on polymer stability and degradation as an SERC Research Fellow. He joined the Materials Group of the Lab- oratory of the Government Chemist in 1983 rising to be-come its Head of Research in 1989. Since 1991 he has been part-time lecturer in dental ma- terials at King's College School of Medicine and Dentistry London and recently became visiting Professor of Chemistry at the University of Teesside. He has published some sixty papers and patents and two books. cal applications. Here we are less concerned with the compli- cated physical chemistry of polyelectrolytes and their solutions than with the practical aspects of how functional materials can be fabricated from these polymers having the necessary proper- ties of integrity adhesion and/or strength. 1.1 Polyelectrolytes and Water The key property of polyelectrolytes is their ability to dissolve in water. Water though is an unusual solvent for polymers and some of its features need to be considered briefly. Water is unique in being the only inorganic liquid to occur naturally on earth; it is also the only substance which can be found naturally in all three physical states solid liquid and ~apour.~It is the most readily available solvent and in princi- ple the least expensive. Yet its physical properties are very different from those of organic solvents. Its surface tension boiling point latent heat of evaporation and density are all very high both in comparison with other solvents and for its very small molar mass. Details of these physical properties together with those of other common organic solvents used in industry are given in Table 1.4 The reason that so many of these values of physical constants are anomalously high for water is that its molecules are highly associated. This associated nature arises from hydrogen bonds between adjacent molecules which in turn occur because of the high polarity of the water molecule. We know from simple observations that when it comes to solvents like dissolves like i.e. polar solvents dissolve polar (including ionic) substances while non-polar solvents dissolve non-polar substances. Water being highly polar therefore dis- solves polar and ionic compounds. Polymers are usually non-polar. Polyelectrolytes by contrast are more or less polar in character i.e. they contain some monomer units bearing polar or ionizable functional groups (See Table 2). Where it can be done neutralization of ionizable groups in a polymer will improve the strength of the hydrophilic attraction hence improve the water-solubility. For example the co-polymer of acrylic acid (89Y0)and methyl acrylate (1 1YO)is not water-soluble as such but becomes so on neutralization with sodium hydroxide. Table 1 Physical properties of water and other solvents Property Water Mineral spirits Acetone Xylene Molar mass 18 170 58 106 Boiling point/"C Flash point/"C 100 - 214.5 -12 56.5 -95 144 -25 Latent heat of 2.259 0.481 0.565 0.39 evaporation at bp/kJg-' Dipole moment /debyes 1.84 0.0 2.88 0.4 Densitylg 1.00 0.752 0.787 0.86 Dielectric constant 78 1.83 21.3 2.37 53 Table 2 Polar monomers for water solubility/dispersibility Monomer Functional group Structure ‘Vinyl alcohol’ Acrylic acid Acrylonitrile Hydroxyl Carboxylic acid Nitrile -(CHz -CH.OH)- -(CH2 -CH.CO,H)- -(CH2-CH.CN)- 1.1.1 Hydrophobic Interactions So far the discussion of solubility has been limited to a consideration of the attractive forces involved. But the polymers used do not necessarily experience only attractive forces. They generally have hydrocarbon backbones which are insoluble in water. It is only the polar functional groups that confer water solubility/dispersibility. Hydrocarbon backbones are hydro- phobic and there is thus a hydrophobic contribution to the interaction between the polymer and water. To understand how this hydrophobic interaction manifests itself we need to consider some simple thermodynamics. Energetically hydrocarbons and water are attracted to each other as thermochemical measurements demonstrate. The enthalpy change AH is negative. However hydrocarbons are none the less insoluble in water. This is because the free energy AG,,l,tion is against the process i.e.it is positive. From the Gibbs equation it follows that the 7‘ASsolutlonterm and hence itself must be negative. This in turn shows that the entropy of the final state (the proposed aqueous solution of hydrocarbon) has decreased relative to the initial two-phase state. In other words the proposed solution is more ordered than water itself. This result is attributed to the formation of an ordered cage structure of water molecules around the hydrocarbon molecule in which water has fewer degrees of freedom than in pure water itself. In liquid-liquid phase diagrams the hydrophobic interaction results in a lower critical solution temperature and in the remarkable result that raising the temperature reduces the solubility (Figure 1 a). In general solution behaviour represents a balance between the hydrophilic and hydrophobic nature of the macromolecule. This may result in so-called ‘closed solubi- lity loops’ (Figure 1b),in which the lower temperature behav- iour and lower critical solution temperature arise due to hydro- phobic effects while at higher temperatures solution behaviour becomes dominated by hydrophobic effects. 2 phases f 1 phase 1 0 1 Mole fraction of polymer Mole fraction of polymer Figure 1 Two possible phase diagrams for polyelectrolyte/water systems with significant hydrophobic interactions. (a) Solution behaviour of water-soluble polymer exhibiting a lower critical solution temperature only. (b) Solution behaviour of water-soluble polymer showing closed solubility loop in which lower-temperature behaviour is dominated by hydrophobic interactions and higher-temperature behaviour by strong hydrophilic effects. 2 Waterborne Coatings Waterborne coatings are becoming increasingly widely used because they cause reduced environmental impact compared CHEMICAL SOCIETY REVIEWS. 1994 with traditional solvent-borne paints. Though of long history the main impetus for their development is usually seen as the beginnings of environmental awareness in the late 1960s a trend which has continued ever since in the industrialized world. The first significant piece of legislation was the Los Angeles ‘Rule 66’ passed in July 1966 and aimed at reducing emission of a range of volatile organic compounds (VOCs) into the atmosphere. Among the main VOCs controlled by this legislation were alkenes aromatics and branched-chain ketones all of which are constituents of conventional low-solids solvent-borne paints. The major concern with these compounds was that they are photochemically active; in the special circumstances of the geographical location of Los Angeles solvent vapours did not readily disperse but lingered and concentrated with their photochemical reactions leading to the formation of severe smogs. Concern with such solvents was not confined to Los Angeles and by 1980 all of the US states had individual regulations similar to those of Rule 66. This environmental concern is by no means restricted to the USA. European countries too have enacted legislation con- trolling VOC emissions over the last decade or so. While this relatively benign effect on the environment is the main advantage of waterborne coatings it is by no means the only one. Others are listed in Table 3. Table 3 Advantages of waterborne coatings (i) Reduced risk of fire (ii) Reduced exposure of the workforce to potentially damaging organic vapours (iii) Easier waste disposal (iv) Liquid therefore can use conventional application methods e.g. spraying If water were the ideal paint solvent no technologist would ever use anything else. Clearly it also has disadvantages for this purpose. These are listed in Table 4. Table 4 Disadvantages of waterborne coatings (i) A tendency for films to remain water-sensitive (ii) Greater energy needed for drying due to high latent heat of evaporation of water (iii) Humidity affecting on drying rates (iv) Flash rusting on ferrous substrates (v) Difficulty in wetting greasy surfaces Hayward’ has also pointed out that there is a tendency to assume that the water is of negligible cost. This is not always the case and in many countries the cost of obtaining water of sufficient purity may exceed that of using the alternative organic solvents. Water must be available not only in large quantities and reliably but also of consistent quality. For many coating applications deionized water has become the standard carrier liquid with very low conductivities required (typically 1-50 ps/ cm) depending on the proposed end-use of the paint. In order to obtain water of such quality potable water free of organic matter and suspended solids is needed as feed for the deioniza- tion unit. Water quality also has to be maintained at the point of disposal. The Water Act of 1985 laid down legislation to cover the control of discharge of effluent to rivers and other surface waters. Factories discharging aqueous effluent must advise the authorities on the nature composition temperature volume and rate of discharge. They must keep comprehensive records of discharges and must ensure that effluent does not contaminate specified underground water. The manufacture of waterborne coatings inevitably generates aqueous effluent and care must be taken that the environmental advantages at the application stage are not lost at the point of waste disposal. POLYELECTROLYTE MATERIALS -REFLECTIONS ON A HIGHLY CHARGED TOPIC-J W NICHOLSON 2.1 Electrodeposition One of the major industrial uses of polyelectrolytes is in painting by electrodeposition This is the technique of applying coatings by means of an electric current The substrates must allow the passage of the electric current and hence function as electrodes In order to use polyelectrolytes in this way solvents of high dielectric constant are needed which in practice means the use of water In addition since the article to be coated (the workpiece) must be capable of acting as an electrode it must be (a) metallic and (b) previously uncoated Consequently the process is res- tricted to that of priming (I e initial coating) of metal compo- nents In practice this is not an undue limitation and large volumes of bulky items such as car bodies are routinely primed in this way Indeed it is now the only method of priming of car bodies used industrially The commercial development of electrodeposition began in the early 1960s In fact however the technology had existed in embryo form since 1919 when the first patent was granted for the coating of conductive articles using electric current * Later fundamental work was done on the deposition of latex paints and on coatings for the interiors of metal cans Despite this early work it was not until 1964 in a classic paper that Finn and Melllo described in detail a viable industrial process for electro- deposition using acceptable paint formulations The process was used for priming car bodies an application which still represents the largest use of electrodeposition technology The original version of the process as described by Finn and Mell,lo used anodic deposition That is to say the paint resins were negatively charged the polymers generally bearing ionized carboxylic acid functional groups to confer mobility in the electric field In turn the workpiece being coated was made the anode in the coating bath Nowadays however cathodic depo- sition has become more widely used In this process the work- piece is the cathode of the coating bath and the resins are positively charged typically with amino functional groups By the mid 1980s the change to cathodic deposition for priming automobile bodies was virtually complete Cathodic deposition offers a number of advantages over anodic deposition as an industrial coating technique These include better corrosion resistance at low film thicknesses and high deposition rate per unit of current (so-called 'good throw- ing power') An important change needed in altering the technology from anodic to cathodic deposition is that the counter-electrode in the cathodic process must be inert in order to prevent it becoming oxidized Typical counter-electrodes used are graphite or stain- less steel By contrast in anodic electrodeposition mild steel is acceptable as the counter-electrode 2 I I Physical Chemistry of Electrodeposition Electrodeposition consists essentially of two steps (1) the move- ment of the charged polyelectrolyte molecules towards the test piece under the influence of the electric field so-called electro- phoresis and (11) insolubilization at the electrode This latter step covers a complex sequence of events and typically occurs because the steps taken to solubilize the polymer are reversed For example as we have seen carboxylic acid polymers are more soluble when neutralized the extreme case being that they are insoluble in their acid form and only soluble when neutralized Under the influence of the electric current the polyanion travels to the anode dragging with it a high proportion of counterions These counterions become expelled from the polymer as they get close to the anode at which point the polymer becomes insoluble and is deposited Similar arguments apply for the more common process of cathodic deposition Polycations are typically solubilized by species such as quaternary ammonium salts and they too become insoluble at the electrode In this case insolubilization occurs because solubilizing counterion reacts with the excess of hydroxyl ions present in the vicinity of the cathode to generate the less soluble tertiary amine groups NR,H+ + OH-+NR + H,O In fact there is a preliminary step to the whole electrodeposi- tion process namely the establishment of a diffusion boundary layer at the metal surface l2 Fundamental studies using a rotating disk as the electrode have shown that this diffusion boundary layer takes a finite time to form and that until this layer is established a coherent film cannot be laid down Commercial coating processes use well-defined conditions I e potential difference of at least 80 V and current density of at least 1 mA/cm2 Time of deposition is typically quite short z e between 30 seconds and 4minutes Under these conditions the electrode reactions are effectively those of the electrolysis of water At the anode 2H,O- 0 + 4H+ + 4e-At the cathode 4H,O + 4e-+ 2H + 40H-Total reaction H,O+ H + to The cathode reaction generates twice as much gas by volume as does the anode reaction This seems important in determining the electrical resistance of the deposited film However at each electrode the rate at which gas is generated is low sothat bubbles do not build up in the film Instead the gases are able to escape by simply dissolving in the electrolyte solution Side-reactions may also occur and these cause varying levels of metal compounds to be deposited within the film The effect of these side-reactions is illustrated by data given in Table 5 Table 5 Comparison of films formed on different substrates9 Substrate Deposition Metal content of film (%) Coating weight (mg/Coulomb) Iron Anodic 0 05 10 0 Iron Cathodic 0 015 45 Aluminium Anodic 0 037 14 3 Aluminium Cathodic 0 200 32 Nickel Cathodic 0 002 42 Zinc Cathodic 0 094 40 Studies such as those from which these data were taken show that in anodic deposition ferrous substrates such as mild or galvanized steel undergo side-reactions that contribute to the insolubilization processes This is undesirable because the pro- ducts are coloured and discolour the deposited film When the workpiece is removed from the deposition bath it has two coats one the insoluble electrodeposited film the other a surface layer of still-soluble polymer which is effectively a dip- coat over the top of the main one This soluble coating is generally removed by rinsing with water after which the work- piece is stoved to bring about crosslinking of the electrodepo- sited film Practical paints for priming consist not only of polymer but also of pigments and extenders both of which must be deposited on the article being coated and at roughly the same rate as the polymer If not the composition of the coating bath will change with time and this in turn will change the composition of the paint being deposited In principle ensuring the deposition of two dissimilar dispersions at equal rates is a formidable feat of physical chemistry Fortunately in practice trial and error can usually be relied upon to yield formulations that are able to deposit binder and pigment at about the same rates The reason that pigment particles move under the influence of the electric field is that molecules of binder tend to become adsorbed onto their surfaces l3 As a result of this adsorbed layer the pigment particles travel in the same direction as the binder during electrodeposition and arriving at the surface retain their adsorbed layer of binder to become fully integrated into the coating On insolubilization the pigment remains embedded in the organic coating 2.2 Other Waterborne Paints Paints for application by electrodeposition are by no means the only type of waterborne coating available. There are the so-called 'emulsion paints' widely used domestically which are actually made not from an emulsion at all but from a latex i.e. a stable dispersion of polymer in water. These however are not generally considered to be polyelectrolytes even though indus- trial versions of them have been made and applied by electric current. Other paints based on polyelectrolytes are the water-reducible systern~.'~They often contain a small proportion of organic solvent to aid the dissolution process but in sufficiently small amounts that the paint conforms to the regulations concerning VOC emissions. Typically paints are made based on conven- tional solvent-soluble polymers that have been modified to confer water solubility such as natural drying oils alkyd resins or epoxy systems. Modification generally involves introducing the kind of polar functional groups which we have already seen that confer water-solubility. Drying oils for example are reacted with maleic acid to introduce carboxylic acid groups. When applied to a substrate these coatings 'dry' mainly by the same process as the parent oil i.e. by metal-catalysed reaction with atmospheric oxygen to lightly crosslink the film. An alternative approach used for epoxy resins is to introduce amine groups using the so-called Mannich reaction (Scheme 1) 0-CH,-CH-CH,\/ R ,N-CH \ CH-Scheme 1 Mannich reaction for the preparation of water reducible epoxy resins. Typically practical paints based on these polyelectrolytes contain in addition to pigment a range of additives designed to improve flow and aid the drying process. For maleinized drying oils these include the metal catalysts such as long-chain fatty acid salts of cobalt manganese and zirconium. Other additives such as surfactants and thickeners aid wetting of substrates and stabilize the pigment dispersion. Lastly for some polymers with carboxylic functional groups volatile amines or even ammonia are used to enhance water-solubility. These are effectively additives and are lost by evaporation as the paint film dries thus reducing the solubility of the film and making it less susceptible to disruption by moisture. 3 Dental and Medical Applications Within the realm of healthcare applications polyelectrolyte materials are mainly used in dentistry though their use is spreading to branches of medicine such as maxillo-facial and ear nose and throat (ENT) surgery. Two classes of polyelectro- lyte material are available to clinicians namely zinc polycar- boxylate cements and glass-polyalkenoate (so-called 'Glass- ionomer') cements; this latter class includes light-curable versions. Both classes fall into the category of acid-base cements. ' CHEMICAL SOCIETY REVIEWS 1994 Their setting involves reaction of acid groups on the polymer with powdered solid bases either modifiedzinc oxide or alumino- silicate glass. These solids are bases in the sense that they are proton acceptors even though they are not soluble in water.' The zinc polycarboxylate cement was developed as a dental material by Smith in 1968.16 In this first paper the good adhesion to tooth materials and also the low irritancy of the material were highlighted. The zinc polycarboxylate cement is now widely used for such clinical procedures as lining cavities prior to placement of the main filling material for attaching crowns to posts and for the adhesion of orthodontic brackets as part of the treatment for misalligned teeth. The components are a polycarboxylic acid usually poly(acry- lic acid) and a modified zinc oxide powder. The modified zinc oxide is prepared by mixing pure zinc oxide with small amounts of magnesium oxide and fusing the mixture at between 1100and 1200°C. This process reduces the reactivity of the zinc oxide towards the acid so that in clinical use the cement paste sets slowly enough to be mixed and placed. The heat treatment causes the zinc oxide to become slightly yellow in colour. This coloration is due to evaporation of oxygen to yield a non- stoichiometric substance corresponding to Zn + ylO,where x is less than or equal to 70 ppm." The zinc polycarboxylate cement sets rapidly to give a mater- ial typical properties for which are given in Table 6. Not all of the zinc oxide reacts so that when set the cement consists of a zinc polyacrylate matrix with unreacted particles embedded in it as reinforcing filler. Table 6 Properties of zinc polycarboxylate dental cement2 Property Typical components/values Liquid 4W5% polyacid Polymer Usually poly(acry1ic acid) Powder:liquid ratio 2.5-3 1 Setting time/min 2.54 Compressive strength/MPa 80-100 The setting process involves reaction of carboxylic acid groups with zinc oxide to form the zinc carboxylate units. The structure of these units is complex with a range of types with subtly different chelating character being present as shown using FTIR spectroscopy'* (see Figure 2). Covalent structures of this type are also known to occur in variety of monomeric zinc carboxylates.' Despite the presence of these chelated zinc carboxylate struc- tures zinc polycarboxylate cements behave like thermoplastic 1-C'v ,//'-(A) IONIC (C) CHELATING BIDENTATE (B) BRIDGING BIDENTATE (0) ASYMMETRIC UNIDENTATE (E)CHELATE BIDENTATE &membered ring Figure2 Possible modes of bonding of zinc ions to carboxylate groups in zinc polycarboxylate dental cements. POLYELECTROLYTE MATERIALS REFLECTIONS ON A HIGHLY CHARGED TOPIC-J W NICHOLSON composites at least in terms of the effect of molar mass of the polyacid on fracture toughness 2o These results can be recon- ciled with the spectroscopic evidence for chelation if the covalent structures postulated are intramolecular rather than crosslinking Water is retained in these cements when they set and it has a number of functions for these cements First it is the solvent for the setting reaction secondly it is one of the products of the neutralization reaction between zinc oxide and poly(acry1ic acid) and finally it occupies co-ordination sites around the metal ions and hydration sites around the polyanion In addi- tion it may also act as plasticizer I e low molar mass component that reduces the rigidity of the bulk polymeric structure The invention of the glass-poly(a1kenoate) dental cement occurred soon after that of the zinc polycarboxylate and was achieved essentially by replacing zinc oxide with an aluminosili- cate glass powder of the type used in the now obsolete dental silicate cement Initially as for the zinc polycarboxylate the acid used was poly(acrylic) but since this time a variety of other acids have been employed including poly(ma1eic) and acrylic acid/itaconic acid co-polymers Glass-poly(a1kenoates) were given the trivial name ‘Glass- ionomer’ the name by which they still tend to be known to the dental profession These cements have more varied properties than zinc polycarboxylates because the glass can be modified in a variety of ways to alter setting speeds and final properties However like the zinc polycarboxylate they set fairly rapidly with the inclusion of water to form materials that can be shaped and finished by the clinician Their uses include all those of zinc polycarboxylates plus others such as gum-line cavity filling and so-called fissure-sealing between the cusps of teeth as a protec- tive measure to prevent dental decay 22 Typical properties of glass-poly(a1kenoates) are given in Table 7 Table 7 Properties of glass-poly(a1kenoate) dental cements Property Typical components/values Liquid 4-5% polyacid with up to 5% (+ )-tartaric acid Polymer Poly(acry1ic acid) poly(ma1eic acid) or itaconic/acrylic acid co-polymers Powder liquid ratio 2-3 5 1 Setting time/min 2 54 Compressive strength/MPa 160-1 80 Compared with zinc polycarboxylates glass-poly(a1kenoates) have improved opacity hence can be used as restorations in the front teeth where aesthetics is important They also release fluoride since fluoride as CaF is a constituent of the glass and this protects the natural tooth material in the region immedi- ately surrounding the glass-poly(a1kenoate) restoration Early on in their development a number of factors were identified that influence the speed of the setting reaction and the final strength of the cement These include molar mass of polymer concentration of acid solution powder liquid ratio and the presence or absence of chelating agents such as (+)-tartaric acid 22 In general increasing all of these parameters except the concentration of ( +)-tartaric acid reduces the setting time and increases the compressive strength of the set cement Glass-poly(a1kenoate) cements undergo gradual maturation processes that have no parallel in the setting of zinc polycar- boxylates For example compressive strength tends to increase gradually with time up to a maximum value The ratio of bound to unbound water also increases this being defined as the ratio of water that may be removed by chemical dessication (e g by storage for 24 over silica gel at elevated temperatures) to that which remains in the cement during this treatment Zinc poly- carboxylate experiences no real change in this ratio over a period of up to 6 months,2 whereas glass-poly(a1kenoates) show a significant increase in bound to unbound water over a similar period The setting reactions of glass-poly(a1kenoates) are as follows (I) Decomposition of the glass under the influence of the +polyacid leading to release of Ca2 ions and A13 + species these latter probably being in the form of complex oxyanions containing several aluminium atoms 23 (11) Rapid reaction of the Ca2+ ions with the polyacid chains followed by slower reaction of A13+ species gradually released from the anionic complex (111) Gradual hydration of the inorganic fragments released in step (1) to yield a matrix of increasing strength and resistance to dessication 24 These steps are illustrated in Figure 3 PolyacidLIQUID hydrogen’\ Polyanions .ions POWDER-/ \ SI IMgel@ t Calcium and Aluminium pdysalts Figure 3 Schematic representation of the setting of glass-polyalkenoate cements Improvements in strength and wear properties of glass-polyalkenoates have been sought by such means as the inclusion of finely divided silver alloy or by fusing the glass with silver alloy prior to mixing forming a ceramic-metal hybrid known as a cermet The success or otherwise of these approaches is the subject of continuing research Another recent attempt to develop cements of this type has been to use the polymer poly(viny1 phosphonic acid) as the cement former This carries the promise of improved clinical adhesion to tooth surfaces and of better translucency than polycarboxylic acid cements These cements which are not poly(a1kenoates) but poly(phosphonates) are at an early stage of development but initial findings concerning their properties are promising One disadvantage of glass-poly(a1kenoates) is that they are sensitive to moisture in the early part of their life soon after placement This may result either in washing out of still soluble species from the cement by saliva or in patients who breathe through the mouth in desiccation and inhibition of the setting reaction To overcome this sensitivity clinicians tend to cover freshly placed cement with an impervious layer of varnish or petroleum jelly An alternative approach has been the development of hybrid glass-polyalkenoates involving the incorporation of photopoly-merizable components The resulting materials can be cured by irradiation with visible light They are very new and have only become available within the past 3-4 years 26 The use of photochemical means of curing limits the depth of individual layers of material that can be used because light can penetrate to only a given depth Such materials must therefore be used in thin layers (2 mm maximum) only This restricts their use to cavity lining or certain limited areas of restoration such as the edges of incisors Light-cured glass-ionomers consist of a complex mixture of components These are (1) poly(acry1ic acid) or a modified poly(acry1ic acid) (11) a photocurable monomer such as hydroxyethyl methacrylate HEMA or a photocurable side- chain grafted onto the poly(acry1ic acid) (in) an ion-leachable glass and (iv) water Consequently they set by a number of competing reactions to give complex structures The initial setting reaction is the photochemical polymeriza- tion Depending on the precise details of the formulation this process is either a co-polymerization of the HEMA with the polymer side-chains or homopolymerization of the functional groups in the side-chains Subsequently the acid-base reac-tions typical of glass-poly(a1kenoates) take place This process is designed to be slower than in conventional glass-poly(a1ke- noates) it is also probably impeded by the development of photopolymerized network and by the presence of organic molecules in the aqueous mixture Very recently these cements have been shown to exhibit properties of weak hydrogels a feature attributed to the pres- ence of HEMA in the cement 26 Cements allowed to mature dry for 24 hours were stronger in compression than those kept in water (70-80 MPa compared with 50-60 MPa for those in water) More significantly those stored in water had become plastic rather than brittle and had swollen by between 3 and 5% an expansion that might prove significant given the close toler- ances required for filling materials in clinical dentistry These are new materials though and should not be written off because of disadvantages with the earliest commercial products Improve- ments to them are highly likely given the intense research currently going into their development Recently glass-poly(a1kenoate) cements have begun to be used outside dentistry Following the discovery of their excellent biocompatibility when in direct contact with bone cements are being studied for a variety of surgical uses including in ortho- paedic surgery for fixing artificial joints,’ in oral and maxillo- facial surgery for repairing damaged regions of bone and in ear nose and throat surgery for repair of the chain of ossicles in the inner ear and for fixing cochlea implants in place 27 These materials thus seem to have an exciting future in wider fields of medicine as well as in dentistry The scope for tuning their properties by subtle changes in the chemistry of their consti- tuents coupled with their excellent biocompatibility makes them likely to become materials of choice for a variety of clinical procedures in the future 4 Conclusions This review has covered the chemistry of polyelectrolyte mater- ials that are in the forefront to two major technical develop- ments namely environmentally acceptable industrial paints and improved medical and dental materials In both fields external trends continue to drive developments in directions that favour polyelectrolytes Environmental legislation is making water- based paints increasingly attractive because they conform so readily to emission requirements yet are presented to the user in the familiar way as liquids that dry out to give solid functional films Polyelectrolyte biomedical materials especially the glass- poly(alkenoates) are being used ever more widely as demand grows for materials that are biocompatible In dentistry as an ageing world population tends to retain more of its teeth for CHEMICAL SOCIETY REVIEWS 1994 longer and also demands increasingly high standards in aes- thetic tooth repair polyelectrolytes are assuming greater importance In medicine surgical techniques for repair and replacement are becoming increasingly sophisticated again fuelled by an ageing world population demanding higher standards Here too polyelectrolytes through their blandness and compatibility with living tissue have an increasingly important r61e to play Both applications rely heavily on the contribution of chemistry Chemists often working as part of interdisciplinary teams and essential to provide the necessary new and exciting chemistry will be central to continuing success in both fields 5 References 1 A D Wilson and J W Nicholson Acid-Base Cements Their Biomedical and Industrial Applications Cambridge University Press 1993 2 F Oosawa ‘Polyelectrolytes’ Marcel Dekker New York 197 1 3 F Franks ‘Water’ Royal Society of Chemistry London 1983 4 L 0 Kornum J Oil Col Chem Assoc 1980,63 103 5 L J T Hughes and D B Fordyce J Polym Sci 1956,22 509 6 L Valentine J Oil Col Chem Assoc ,1984,67 157 7 G R Hayward in ‘Waterborne Coatings Surface Coatings -3’ ed A D Wilson J W Nicholson and H J Prosser Elsevier Applied Science Publishers Barking 1990 Chapter 9 8 W P Davey US Patent 1294627 1919 9 H -J Streitberger and R P Osterloh in ‘Surface Coatings- 2’ ed A D Wilson J W Nicholson and H J Prosser Elsevier Applied Science Publishers Barking 1988 Chapter 2 10 S R Finn and C C Mell J Oil Col Chem Assoc 1964,47,219 I1 H U Schenck and J Stoelting J Oil Col Chem Assoc ,1980,63 482 12 F Beck Prog Org Coatings 1976,4 1 13 TEntwistle J Oil Col Chem Assoc 1972,55,480 14 Australian Oil & Colour Chemists’ Association ‘Surface Coatings’ Volume 1 2nd edition Chapman and Hall London 1983 15 A D Wilson Chem Soc Rev 1978,7,265 16 D C Smith Brit Dent J 1968 125,381 17 N N Greenwood and A Earnshaw ‘The Chemistry of the Ele- ments’ Pergamon Press Oxford 1984 18 J W Nicholson,P J Brookman,O M Lacy,G S Sayers,andA D Wilson J Biomed Muter Res 1988,22,623 19 R C Mehrotra and R Bohra ‘Metal Carboxylates’ Academic Press London and New York 1983 20 R G Hill and S A Labok J Muter Sci ,1991,26,67 21 J W Nicholson S J Hawkins and E A Wasson J Muter Sci Muter Med 1993,4 32 22 A D Wilson and J W McLean ‘Glass Ionomer Cement’ Quintes- sence Publishers Chicago 1988 23 E A Wasson and J W Nicholson Br Polym J 1990,23 179 24 E A Wasson and J W Nicholson J Dent Res 1993,72,481 25 E A Wasson Clin Muter 1993 12 18 1 26 H M Anstice and J W Nicholson J Mater Sci Muter Med 1992,3,447 27 R T Ramsden R C D Herdman and R H Lye J Laryngol Otol 1992,106,949

ISSN:0306-0012    

DOI:10.1039/CS9942300053

出版商:RSC    

年代:1994

数据来源: RSC

摘要:

Magnetic Field Gradients in NMR Friend or Foe? Timothy J. Norwood Department of Chemistry Leicester University University Road Leicester L El 7RH U.K. 1 Introduction Anyone who has used an NMR spectrometer is familiar with magnetic field gradients. In the form of inhomogeneities in the static field of the magnet they cause broadening of the spectral lines often hindering analysis. Although this somewhat negative manifestation of the phenomena is the first experience of most spectroscopists a growing appreciation of their practical uses has resulted in some of the most important developments in NMR methodology over recent years often extending its uses far beyond its traditional role as an analytical tool for chemists. The exploitation of the properties of magnetic field gradients is at the heart of magnetic resonance imaging (MRI); this tech- nique is now widely used as an aid to clinical diagnosis in many hospitals and is proving to be a promising tool for addressing chemical problems. Magnetic field gradients are also fundamen- tal to NMR techniques for studying diffusion which have proved invaluable for studying the microscopic structure of heterogeneous systems. A further use for magnetic field gradi- ents that has come to maturity over the past few years is in replacing phase cycling in many multiple-pulse experiments enabling cleaner spectra to be obtained in shorter times. In this article we examine the place of magnetic field gradients in pulsed Fourier transform NMR from two perspectives as a problem and as a solution to problems. 2 Magnetic Field lnhomogeneity NMR is concerned with the precession of nuclear magnetic moments in an applied magnetic field B,. In a perfectly homo- geneous magnetic field the precessional frequency of the nuclear spins the Larmor frequency w is given by w = y(l -up (1) where y is the magnetogyric ratio and u is the shielding constant which is responsible for chemical shift. In a conventional one- dimensional NMR spectrum nuclear spins in different chemical environments will each give rise to a peak at a characteristic frequency W. If B is not homogeneous but exhibits spatiaIIy Tim Norwood did his BSc. at King's CoIIege London (1983) an M.Sc. at the University of British Columbia (1985) and his Ph.D. at Cambridge Un- iversity with Professor Laurie Hall (1989). After two post- doctoral years with Dr. I. D. Campbell at Oxford University he took up his present position as a lecturer in physicaI chemistry at Leicester Univer- sity. His research interests include the design and appli- cation of NMR spectroscopic and imaging techniques for ad- dressing biological problems. dependent magnetic field inhomogeneity dB,(v) then the effec- tive Larmor frequency ueR will become wem= w + y( 1 -u)dB,(r). (2) Since a<< I and as it is usually the case that dB,(r)c<B equation 2 can be reduced to weK= w + ydB,(r). (3) It is clear from this equation that as the magnetic field varies across the sample so will the Larmor frequency of the nuclei. At equilibrium the nuclear magnetization of a sample gives rise to a static net-magnetization vector in the direction of B (by convention the z-axis). To observe a free induction decay the magnetization vector must first be rotated into the xy-plane; this is achieved by applying a pulse of coherent radiofrequency radiation. By adjusting the power length and phase of the pulse it is possible to determine both the angle the pulse will rotate the nuclear magnetization through and the axis in the xy-plane about which it will be rotated; for example a 900,pulse will rotate the magnetization through 90" about the x-axis. After the application of a 90; pulse to the equilibrium magnetization the net-magnetization vector will be found along the y-axis. The subsequent evolution of this transverse magnetization vector in both homogeneous and inhomogeneous magnetic fields the resulting free induction decays (fids) and Fourier transformed spectra are sketched in Figure 1. In a homogeneous magnetic field the magnetization vector precesses at a single frequency and gives rise to a free induction decay that decays solely due to transverse relaxation. In an inhomogenous magnetic field the magnetization vector has to be subdivided into a number of smaller vectors each precessing at a different frequency. These vectors spread out (dephase) as they evolve resulting in a reduction in the size of the net transverse magnetization of the sample causing the free induction decay to disappear more rapidly and the line in the NMR spectrum to broaden. In the extreme the inhomogeneity-broadened lines of a spectrum may coalesce into a featureless hump. In some applications magnetic field gradients are used to deliberately dephase the transverse magnetization. Magnetic field variations across a sample can have two main origins inhomogeneities in the applied magnetic field and variations in magnetic susceptibility across the sample. In a x time -X fid spectrum x X Figure 1 The evolution resulting free induction decay (fid) and spec- trum of a transverse magnetization vector in (a) a homogeneous and (b) an inhomogeneous magnetic field. 59 homogeneous liquid in an NMR tube magnetic susceptibility variations arise at the solution-glass-air and solution-air inter- faces Since NMR tubes have axial symmetry the resulting variations in the magnetic field are also symmetric and can be counteracted by applying small additional magnetic field gradi- ents across the sample using the spectrometer's shim coils However this compensation measure may fail if the sample is either particularly large or heterogeneous Large samples are often encountered in in vivo imaging or spectroscopic studies of living systems and those encountered in applications to geology and material science as well as biological systems are often extremely heterogeneous The microscopic but often extremely strong magnetic field gradients generated in heterogeneous samples as a result of variations in magnetic susceptibility can obliterate any features in the spectrum and are impervious to shimming The problems presented by the presence of unwanted magne- tic field gradients can be tackled in two ways by using echo- based experiments designed to reverse their effects and by using zero-quantum coherence which is unaffected by magnetic field gradients 3 Counteracting the Effects of Magnetic Field Gradients Echos Some NMR measurements are more susceptible to the effects of unwanted magnetic field gradients than others In a perfectly uniform magnetic field the transverse relaxation time T2can be measured from the free induction decay or from peaks in the spectrum since the natural line width at half peak height is equal to (1/rT2) However transverse relaxation time measurements are particularly vulnerable to variations in the magnetic field since even gradients that are not immediately obvious from the appearance of the spectrum can cause large errors in their apparent values For example gradients causing as little as 0 1 Hz line broadening will result in an error of 50% for a T2of 1 6 s or 20% for a T2 of 0 64 s a 90' I I 'I; I Acquisition b ad Figure 2 (a) Pulse sequence for the spin echo experiment (b) Vector model representation of the evolution of transverse magnetization during a spin echo Transverse relaxation times can be measured independent of the effects of magnetic field gradients by using an experiment based on a spin echo,' Figure 2 This experiment is explained using the vector model in the figure The transverse magnetiza- tion excited by the first pulse is subsequently allowed to evolve due to its chemical shift scalar couplings (not included in this example) and magnetic field inhomogeneity during T In this example the last factor has caused the transverse magnetization to dephase entirely at the end of this period The 180" pulse often known as a refocusing pulse rotates all the nuclear magnetiza- tion vectors through 180" about the y-axis This rearranges the positions of the dephased vectors such that while they would CHEMICAL SOCIETY REVIEWS 1994 previously have continued to spread out subsequent evolution Tduring the second period will cause them to come back together reversing the evolution due to chemical shift and magnetic field gradients during the first Tperiod Consequently all that will have happened by the end of this sequence of events when the free induction decay is acquired is decay of the magnetization due to T2 and evolution due to any scalar couplings present The intensity of the observed signal will therefore be given by 427) = Z(O)exp(-27/T2) (4) Consequently T2 can be readily calculated from a series of measurements at different values of T Spin echos are incorpor- ated into many NMR experiments to reverse the effects of magnetic field inhomogeneity A number of two-dimensional experiments based upon the use of echos produce F spectra that are unaffected by magnetic field inhomogeneity 2-The 2D-J experiment,2 which corre- lates the chemical shift of a spin in F2 with its multiplet in Fl is perhaps the best known example However in all of these experiments if the magnetic field gradients present are strong signal may also be lost due to diffusion (see below) Besides distorting Tvalues this effect may also broaden the lines of the F spectra produced by these experiments In the case of T2 measurements diffusion-attenuation can be overcome by rep- lacing the single spin echo with a train of closely spaced 180" pulses5 differing in phase from the excitation pulse by 90" 4 High Resolution Spectroscopy in the presence of Magnetic Field Gradients Zero- Quantum Coherence Many of the samples of interest in geological and material science applications of NMR are by nature heterogeneous and consequently high resolution spectra cannot be obtained from them using conventional methods In this context the possibili- ties opened up by exploiting the properties of zero-quantum coherence are particularly exciting since zero-quantum spectra are unaffected by magnetic field gradients The signal observed in the conventional NMR spectrum arises from single-quantum coherence A single-quantum coher- ence is a phase coherence between two states for which AM= f1 where AM = CkAmk (5) and Amk is the change in magnetic quantum number of a spin k between the two states concerned Those spins that have 'flipped' between the two states are said to be active in the coherence Single-quantum coherence can be detected because it has a net magnetization vector associated with it Figure 3A An energy level diagram of a scalar coupled two-spin system is sketched in Figure 4 In addition to the four allowed single- quantum transitions it can be seen that there are two additional transition possibilities a double-quantum transition for which AM = f2 and a zero-quantum transition for which AM = 0 The coherences to which these transitions give rise are 'spin forbidden' since AM# f I and consequently they cannot be excited by the application of a single non-selective pulse to the equilibrium magnetization of a spin system Neither can they be observed since they have no net magnetization associated with them Figure 3B However these multiple-quantum coher-ences6-can be studied indirectly typically in a two-dimen- sional experiment Figure 5 multiple-quantum coherence is excited indirectly using a sequence of pulses and delays and detected indirectly following its conversion into observable single-quantum coherence In such experiments the extent of multiple-quantum evolution during the evolution period t is reflected in the phase and amplitude of the observed single- quantum coherence By repeating the experiment whilst syste- matically incrementing t and then Fourier transforming the MAGNETIC FIELD GRADIENTS IN NMR FRIEND OR FOE?-T LVI \ I X 4-J 0 Figure 3 Vector model representations of (a) a single-quantum coher- ence of an uncoupled spin k and (b) a two-spin multiple-quantum coherence between a pair of coupled spins k and I within the limitations of this model it may correspond to either a zero- or double-quantum coherence Energy level dtag ram Energy I akal Spectrum s1 s4 Z S2 S3 D c,Jkl 0-Figure 4 Energy level diagram and spectrum for a pair of coupled spins k and I Allowed transitions and peaks are indicated by continuous lines while forbidden transitions and peaks are indicated by dashed lines A transition is allowed when AM = f 1 and is forbidden otherwise a and /3 correspond to the two states (rn = ++ and -+ respectively) of a spin-+ nucleus Zero- single- and double-quantum transitions are denoted by Z S and D respectively Preparation Evolution 1 Detection Figure 5 Pulse sequence for multiple-quantum spectroscopy Since multiple-quantum coherence violates the rules for allowed transitions it is excited and detected indirectly using a sequence of pulses and delays multiple-quantum coherence evolves during the evolution period t A multiple-quantum spectrum can be obtained in a two- dimensional experiment by repeating the experiment while systemati- cally incrementing t and then Fourier transforming the data with respect to t Typically T = (&J) Even orders of coherence are excited when 4 1 = x and odd when 4 I = y A given order of coherence can be selected by phase cycling 42 and +R In the presence of a magnetic field gradient only zero-quantum coherence will be observed since all other orders will be dephased during t J NORWOOD resulting data matrix with respect to t a multiple-quantum spectrum can be obtained Multiple-quantum coherence is of interest here because of its sensitivity to magnetic field gradients When modified to include multiple-quantum coherence equa- tion 3 becomes It can be seen from equation 6 that a coherence's sensitivity to magnetic field gradients will be proportional to its order Thus while homonuclear double-quantum coherence will be twice as sensitive as single-quantum coherence to a magnetic field gradi- ent zero-quantum coherence will be completely unaffected by it Consequently the zero-quantum spectrum will always be at high resolution regardless of any magnetic field gradients present Unlike echo-based techniques zero-quantum spectra are unaffected by diffusion From equation 5 it can be seen that a minimum of two spins are required to excite a zero-quantum coherence For the pulse sequence given in Figure 5 to be effective a given pair of spins must either have a mutual scalar coupling or both must have scalar couplings to a common third spin for a coherence to be excited between them Therefore a zero- or double-quantum coherence cannot be excited from water for example Multiple- quantum spectra do not have the appearance of conventional spectra From equation 6 it can be seen that a two-spin zero- quantum coherence will have a frequency that is the difference in Larmor frequencies of its two active spins This is illustrated for a two-spin system in Figure 4 Furthermore scalar couplings which will only be observed to spins 1not active in the coherence (and not between those spins k active in it) will also correspond to linear combinations of those found in the conventional single- quantum spectrum JMQC = CkArnkJkl (7) In the case of a two-spin system (Figure 4),since both spins must be active in either a zero- or double-quantum coherence both spectra will only contain a singlet Consequently the zero-quantum spectrum (and multiple- quantum spectra in general) will have an unfamiliar appearance and cannot be interpreted in the conventional manner However alternative methods for analysing zero-quantum spectra have been developed * An intriguing feature of multiple-quantum coherence is high- lighted by the vector model representation given in Figure 3B This picture can represent a zero-quantum coherence between two spins k and I which we know from equation 6 will precess at the frequency (Wk -w1) However no vector in this picture actually evolves at this frequency the antiphase k-spin vectors evolve at wk fJkr/2 and the antiphase /-spin vectors evolve at WI fJkJ2 The properties of a multiple-quantum coherence only become apparent when it is converted into single-quantum coherence for detection the efficiency of this process depends upon the difference in phases of the k and I spin antiphase vectors and consequently on (wk-wI) for a zero-quantum coherence The zero-quantum spectrum of an A(MK)X spin system -the four aromatic protons of salicylic acid -acquired in an inhomo- geneous magnetic field with the pulse sequence sketched in Figure 5 is given in Figure 6c The zero-quantum spectrum is clearly at high resolution even though the conventional spec- trum acquired under the same conditions Figure 6b is little more than a broad featureless hump The zero-quantum spec- trum is symmetrical about the centre the zero-frequency peak does not arise from zero-quantum coherence but IS an artifact arising from magnetization that was longitudinal during the evolution period There is a zero-quantum coherence between each pair of adjacent protons which occurs at the difference of their Larmor frequencies in the conventional high resolution spectrum Figure 6a CHEMICAL SOCIETY REVIEWS 1994 a X 9 I 85 I 8 75 PPm / -1 3 2 1 0 1 2 3 PPm C Ti Figure 6 500 MHz 'H spectra of the aromatic protons of salicylic acid (a) Conventional (single-quantum) spectrum acquired in a homo- geneous magnetic field (b) Conventional spectrum acquired in the presence of a magnetic field gradient (c) Zero-quantum spectrum acquired in the presence of the same magnetic field gradient as (b) 5 Measurement of Diffusion In the physical sciences self-diffusion data have been shown to be a valuable source of information on molecular organization and phase structure while in the context of magnetic resonance imaging diffusion-contrast has been found useful as an aid to the clinical diagnosis of a number of diseases As has already been noted diffusion through a magnetic field gradient can cause attenuation of the signal obtained in a spin echo experiment Whilst this phenomenon is a problem in the context of T2measurement it has been turned to good use in a series of NMR experiments designed to study diffusion Tradi- tionally self-diffusion coefficients were measured using radioac- tive tracer techniques Although extremely accurate such stu- dies take days or even weeks for a single component and require isotopic labelling NMR techniques have proved to be an attractive alternative since they can provide accurate diffusion coefficients of multicomponent systems in a matter of minutes and without the need for isotopic substitution The most com- monly used NMR experiment for measuring diffusion the pulsed gradient spin echo (PGSE) method,lO-l is sketched in Figure 7a It consists of a spin echo and a pair of magnetic field a 90; 180; 1 Acquisition 1 23 on t-t-Gradient G +&+ 7 -6* 4 b 1 2 3 4 b Figure 7 (a) Pulsed gradient spin echo (PGSE) pulse sequence for measuring diffusion (b) Vector model representation of the evolution of diffusing transverse magnetization during a PGSE experiment gradient pulses one either side of the 180" pulse The static magnetic field B is assumed to be homogeneous How PGSE works can best be appreciated by considering one of the criteria for echo formation in the spin echo expenment For evolution during the first half of the spin echo to be completely reversed by the end of the second half of the experiment the frequencies of individual nuclei must remain constant throughout However if a magnetic field gradient is present the frequency of the nuclei will vary with their position in the direction of the gradient Consequently any motion such as that due to diffusion in the direction of the gradient will result in a change in frequency The effect of this on echo formation is illustrated in Figure 7b Three vectors are considered a does not move in the direction of the gradient during the experiment and so its frequency remains constant and it ends up along they-axis bmoves in the direction of increasing magnetic field strength between the two halves of the experiment and consequently its frequency is greater during the second half and as a result it 'overshoots' the y-axis c moves in the direction of decreasing magnetic field strength between the two halves of the experiment resulting in a decrease in frequency as a consequence of which it 'undershoots' the y-axis Clearly diffusion will result in a dispersion of vectors about the y-axis instead of alignment along it and consequently the overall intensity of the echo will be reduced The intensity of the echo produced in the PGSE experiment is given by where G is the intensity of the magnetic field gradient S is the duration ofeach gradient pulse A is the time between the start of the two gradient pulses and D is the diffusion coefficient It can be seen from equation 8 that a plot of lOg,[@T)] against s2(A -6/3)will have a gradient of -(YG)~Dfrom which D can be calculated Data are usually acquired for several values of G while keeping all other parameters constant In heterogeneous systems the diffusing molecules may encounter a barrier to their motion for example particles in a suspension or cell walls in biological tissue Under these circum- stances diffusion is said to be restricted Restricted diffusion can be detected with the PGSE experiment I1-l2 If diffusion is measured with a short echo time the diffusing molecules will not have moved very far and therefore few will have experienced any restriction consequently the apparent value of the diffusion coefficient will be close to the unrestricted value As the echo time of the PGSE experiment is increased so the fraction of molecules that experience restriction will also increase and will therefore have moved over small distances than they otherwise would have consequently the apparent value of the diffusion coefficient will decrease A typical plot showing the variation of the apparent value of the diffusion coefficient as a function of echo time is sketched in Figure 8 By modelling such data it is possible to determine values for the parameters associated with restriction For example from the diffusion of water in biologi- cal tissue cell size and permeability can be determined Pulse sequences for measuring diffusion have been incorpor- ated into a number of magnetic resonance imaging experiments to produce images in which the signal intensity reflects the diffusion coefficient The environment of the diffusing molecules as well as their concentration can be mapped using such techniques l3 Diffusion coefficient maps can also be generated from such data Diffusion-contrasted imaging zn uzvo has been shown to be useful in determining the regions of the brain affected by stroke and multiple sclerosis We have noted that the sensitivity of a coherence to magnetic field gradients is proportional to its order Consequently by using a higher order of coherence the sensitivity of the PGSE experiement to diffusion can be increased allowing diffusion coefficients to be determined more accurately or a lower gradient strength to be used l4 A multiple-quantum PGSE experiment can be constructed by placing PGSE minus the first 90" pulse into the evolution period of the multiple-quantum experiment sketched in Figure 5 MAGNETIC FIELD GRADIENTS IN NMR FRIEND OR FOE?-T Diffusion coeff ictent unrestricted diffusion restricted diffusion \ loge(Diffusiontime) + Figure 8 Variation of the apparent diffusion coefficient as a function of the diffusion time (A in Figure 7) for unrestricted and restricted diffusion The use of pulsed magnetic field gradients in both this and other applications has one major problem associated with it When a magnetic field gradient is switched on or off the change in magnetic flux induces electrical currents in nearby metal I e in the probe and magnet These eddy currents can persist for seconds and have the effect of distorting the spectrum some- times beyond recognition This problem can be overcome by using gradients which are shielded so that their effects are not felt outside their immediate volume,' these are only now becoming widely available 6 Imaging In the two decades since its inception imaging has become not only the most important application of magnetic field gradients in NMR but one of the most important applications ofNMR as a whole familiar to most people as magnetic resonance imaging -or MRI The word 'nuclear' is omitted so as not to distress patients' Although imaging can be implemented in many different ways the earliest method remains the best for explaining the fundamental principles Consider a sample consisting of two tubes of water Figure 9 In a uniform magnetic field a single peak is observed in the NMR spectrum since all the molecules will precess at the same frequency equation 1 regardless of their spatial location If a linear magnetic field gradient is applied across the sample in this example in the x-direction then analogously to equation 3 equation 1 will become Sample Spectrum 0-Figure 9 The principle of one-dimensional NMR imaging The appli- cation of a magnetic field gradient across a sample will give rise to a frequency offset proportional to its spatial location in the direction of the gradient thus making the spectrum a profile of the sample in the direction of the gradient J NORWOOD where G is the strength of the magnetic field gradient and x is the spatial coordinate Clearly the Larmor frequency of the water will now increase linearly across the sample as a function of its x-coordinate and the spectrum Figure 9 will become a one- dimensional image or profile of the sample \ Image \ reconstruction \ V \ \\---I>-\--Figure 10 The principles of constructing a two-dimensional NMR image using back projection A number of profiles of the sample are obtained with different orientations which are then used to construct a two-dimensional image Profiles with different orientations are obtained by acquiring spectra in the presence of magnetic field gradients in different directions (indicated by arrows) an x-gradient will result in an x-profile while a y-gradient will result in a y-profile of the sample The principles of two-dimensional imaging are illustrated in Figure 10 Here by using combinations of two gradients G and Gv,a series of one-dimensional images are obtained in different directions in the xy-plane A two-dimensional image of the object can be reconstructed from these one-dimensional projec- tions using a technique known as back projection reconstruc- tion as illustrated in the figure Clearly the greater the number of profiles used to construct an image the better defined it will be This method can be extended to three dimensions At the present the most widely used imaging experiments are based on multi-dimensional Fourier transformation rather than back projection The intensity of a given element of an NMR image can reflect a number of factors depending on the structure of the pulse sequence used In addition to the concentration of the observed species the image intensity may also reflect its transverse or longitudinal relaxation times or its diffusion coefficient It is possible to produce separate images of different chemical spe- cies An NMR image of a guinea pig is given in Figure 11 here the contrast reflects the concentration of mobile protons and their transverse relaxation times Although as a result of the quality of the anatomical detail it provides and its non-invasive nature NMR imaging is most widely used as a clinical tool a number of chemical applications have also been explored Often the changes occurring within a system are observed indirectly An example is given in Figure 12 where the spatial progression of a change in a chemical equili- brium through a gel is monitored The chemical equilibrium the chelation of aqueous copper by ethlenediaminetetraacetic acid is pH-dependent Initially the gel has a pH of 7 An acidic solution is added to the central well and diffuses out through the gel causing a change in pH as it does so and thus shifting the equilibrium towards free Cu2+ The image intensity reflects the concentration of the water which is the main observed species and its transverse and longitudinal relaxation times The free Cu2+ is particularly good at increasing the longitudinal relaxa- tion rate of the water which with the pulse sequence used here results in a region of low signal intensity Since a single Cu2 + ion Figure 11 'H NMR image of the body of a live guinea pig under anaesthetic. The spinal column runs across the top of the image while from right to left can be seen the heart liver and stomach. The oesophagus extends from above the heart to the stomach. Two reference tubes can be seen at the top of the image. Figure 12 'H NMR images illustrating the shift in the equilibrium of the chelation of aqueous copper by ethlenediaminetetraacetic acid (EDTA) in a gel under the influence of a pH gradient (see text). The left and right images were acquired 4 and 25 minutes respectively after the addition of acid to the central well. Traces through the two images are given at the bottom of the figure. can help relax many water molecules techniques of this type can enable changes to be observed in species that have too low a concentration to be observed directly. 7 Magnetic Field Gradient Pulses An Alternative to Phase Cycling A pulse sequence consists of a series of radiofrequency (r.f.) pulses and delays. Each pulse is assumed to rotate the nuclear magnetization through a specific angle about a specific axis. However in practice mis-setting of pulse angles r.f. inhomoge- neity across the sample and resonance offset effects all give rise to non-ideal behaviour. As a result of this not all of the nuclear magnetization will evolve as expected during an experiment. For CHEMICAL SOCIETY REVIEWS 1994 example although an ideal 180"pulse inverts all of the longitudi- nal magnetization of a sample in practice due to the effects noted above some will end up as transverse magnetization. The different routes that magnetization can follow through an experiment are known as coherence transfer pathways. In many multipulse experiments the magnetization may follow a number of coherence transfer pathways even under ideal conditions. Magnetization following unwanted coherence pathways may distort measurements and in the case multi-dimensional experi- ments give rise to unwanted peaks or artifacts in the resulting spectrum. Traditionally magnetization that follows unwanted coher- ence transfer pathways arising from either source is eliminated by using phase cycling.19 Phase cycling is implemented by repeating an experiment several times using different phases (the axes about which the nuclear magnetization is rotated) for some or all of the pulses and then combining the resulting free induction decays in such a way that signals arising from unwanted pathways cancel out. For example the inversion recovery pulse sequence for measuring longitudinal relaxation (T,)consists of 180; 7900 acquisition. (10) The 180" pulse inverts the equilibrium magnetization which subsequently returns to equilibrium during 7 due to TI relaxa-tion. The extent of relaxation is monitored with the 90" pulse which converts the longitudinal magnetization into observable transverse magnetization. Any transverse magnetization arising due to the non-ideal behaviour of the 180"pulse will of course decay due to transverse relaxation during T and this will intro- duce error into the measurement. Magnetization following this unwanted coherence transfer pathway can be eliminated by adding data acquired with equation 10 to that acquired with 180 ,T 900 acquisition. (1 1) In this latter pulse sequence the phase of the 180"pulse has been altered from x to -x. The result of this is that any transverse magnetization excited by the 180" pulse in the two cases will differ in phase by 180" and therefore undergo mutual cancel- lation when the free induction decays are coadded. In general the more pulses there are in an experiment the more phase cycling is necessary to eliminate unwanted coherence transfer pathways. Phase cycling has a number of drawbacks of which the most obvious is the additional time it takes to implement. Further- more phase cycling often works less than perfectly in practice particularly when intense signals are to be removed; the reasons for this can include instrumental instability sample spinning (if used) and an insufficient relaxation period for the magnetiza- tion to return to equilibrium between experiments. Dynamic range is also a frequent problem. The imperfections in phase cycling typically manifest themselves as artifacts in the spectrum or in the case of two-dimensional experiments as ridges of 'tl-noise' that may run right across the spectrum and can obscure peaks. Magnetic field gradients offer the spectroscopist an alterna- tive to phase cycling to remove unwanted magnetization. We have already noted that a magnetic field gradient can cause transverse magnetization to disappear completely Figure 1b and that this process can if desired subsequently be reversed by applying a refocusing pulse Figure 2; both of these features can be exploited to allow phase cycling to be replaced with magnetic field gradient pulses in NMR experiments. For example in the inversion recovery pulse sequence described above the phase cycling can be replaced by a single magnetic field gradient pulse during Tto dephase any transverse magnetization excited by the 180" pulse. The gradient pulse must of course be applied for sufficient time to completely dephase the transverse magnetiza- tion; this time can be minimized by maximizing the strength of the gradient used. In this example the use of gradient pulses MAGNETIC FIELD GRADIENTS IN NMR FRIEND OR FOE'-T J NORWOOD permits the experiment time to be halved since only a single transient needs to be acquired with each T value In many experiments a radiofrequency pulse may bring about a number of coherence transfer processes of which only one is desired For example in experiments similar to that sketched in Figure 7 zero- single- and double-quantum coherence may all be present during the evolution period t and all may subse- quently be detected following coherence transfer into observable single-quantum coherence although only one process is of interest A single process was traditionally isolated by using phase cycling However a single process can also be separated by exploiting the different sensitivities of different orders of coherence to magnetic field gradients and the fact that the pulse that brings about a coherence transfer process also acts as a refocusing pulse and gives rise to what is known as a coherence transfer echo 'Coherence transfer echos can be used to reverse the effects of magnetic field gradients in a similar way to spin echos 2o The main difference between a spin echo and a coher- ence transfer echo is that with the latter the coherence that is 0-9 8 7 6543210 1-2-3-.... 4-. ..*-. E1 BQ-Q-5-rephased after the refocusing/coherence transfer pulse is not the same as the one that was dephased before it The formation of coherence transfer echos after a 90" pulse is illustrated in Figure 13 where multiple to single-quantum coherence transfer pro- cesses are observed in the presence of a magnetic field gradient The echos arising from different coherence transfer processes occur at different times reflecting the fact that while each multiple-quantum coherence dephased at a different rate the observed single-quantum coherences all rephase at the same rate regardless of their history Clearly if the gradient were switched off at the moment a particular coherence transfer echo is formed only magnetization that has followed the correspond- ing coherence transfer pathway will be observed In practice this is achieved by using magnetic field gradient pulses For example a single gradient pulse applied during t will dephase all but the zero-quantum coherence and consequently only the zero to single-quantum coherence coherence transfer process will be observed Alternatively if identical gradient pulses are applied during evolution and mixing and detection periods only the single-quantum to single-quantum process will be observed analogously to the spin echo To observe the double- to single-quantum coherence transfer process selectively it is necessary to apply a magnetic field gradient pulse during the detection period that is twice as long as that applied during the evolution period An example of a spectrum acquired using gradients to select a particular coherence transfer process is given in Figure 14 21 a b 9 time -Figure 13 (a) Coherence transfer echos forming in the presence of a magnetic field gradient after a dephasing period of 200 ms and a 90" coherence transfer pulse The echos arise from the transfer of zero- single- double- and triple-quantum coherence to single-quantum coherence as indicated and occur at 0 200 400 and 600 ms respectively into the free induction decay (b) Conventional free induction decay acquired under the same conditions for comparison 6-7-8-9-987654321 PPm Figure 14 400 MHz 'H double-quantum filtered COSY spectrum of 8 mM angiotensin I1 in H,O Coherence transfer pathway selection was achieved using magnetic field gradient pulses no phase cycling wds used (Reproduced with permission from ref 21 R Hurd J Map Reson 1990,87,422 ) Since this spectrum was acquired with a double-quantum fil-tered COSY experiment only spins that can form double- quantum coherence (z e that are scalar coupled to at least one other spin) will be observed Consequently the water peak although dominating the conventional one-dimensional spec- trum is absent from the double-quantum filtered COSY spec- trum except for a insubstantial ridge of 1,-noise The cleanness with which the selection process has been achieved highlights a major advantage of using magnetic field gradient pulses if the same experiment were attempted using phase cycling without using any other form of solvent suppression one would be unlikely to see anything but water in the resulting spectrum The major disadvantage of using gradient pulses to select coherence transfer pathways is that up to 50% of the signal can be lost in some applications since in some contexts they can be more selective than phase cycling Signal losses due to diffusion- attenuation can be minimized by keeping pairs of dephasing and rephasing gradient pulses close together and the amplitude of the gradient used to the minimum 8 Conclusion In the early days of NMR magnetic field gradients were seen primarily as a problem often hindering spectral analysis How- ever over the past two and a half decades this perception has changed dramatically as the use of magnetic field gradients has made possible the measurement of diffusion spawned magnetic resonance image which is now a field in its own right and most recently made possible the elimination of phase cycling in many experiments allowing cleaner spectra to be obtained in shorter times Even where unwanted magnetic field gradients remain a problem particularly when large or heterogeneous samples are under study techniques have been developed which have en- abled many of the problems they pose to be overcome 66 Acknowledgements. It is a pleasure to thank Drs. Suzanne Gilbert and Mark Horsfield for their many helpful comments while preparing this article Dr. Steve Williams and the Univer- sity of London NMR Imaging Facility at Queen Mary and Westfield College for Figure 11 and Professor Laurie Hall Dr. Alan Fischer and the Herchel Smith Laboratory for Medicinal Chemistry for Figure 12. 9 References I H. Y. Carr and E. M. Purcell Phys. Rev. 1954,94,630. 2 R. Benn and H. Gunther Angew. Chem. Znt. Ed. Engl. 1983 22 350. 3 M. Gochin D. P. Weitekamp and A. Pines J. Magn. Reson. 1985 I63,431. 4 S. L. Due L. D. Hall and T. J. Norwood J. Mugn. Reson. 1990,89 273. 5 S. Meiboom and D. Gill Rev. Sci. Znstr. 1958 29 688. 6 G. Bodenhausen Progr. NMR Spectrosc. 1981,14 137. 7 R. R. Ernst G. Bodenhausen and A. Wokaun ‘Principles of Nuclear Magnetic Resonance in One and Two Dimensions’ Claren- don Press Oxford 1987. CHEMICAL SOCIETY REVIEWS 1994 8 T. J. Norwood Progr. NMR Spectrosc. 1992,24 295. 9 L. D. Hall and T. J. Norwood J. Chem. SOC. Chem. Commun. 1986 44. 10 E. 0.Stejskel and J. E. Tanner J. Chem. Phys. 1965,42,288. 11 P. Stilbs Progr. NMR Spectrosc. 1987 19 1. 12 J. E. Tanner and E. 0.Stejskel J. Chem. Phys. 1968,49 1768. 13 T. J. Norwood S. L. Duce and L. D. Hall J. Magn. Reson. Series A. 1993,102,370. 14 D. Zax and A. Pines J. Chem. Phys. 1983,78,6333. 15 P. B. Roemer W. A. Edelstein and J. S. Hickley ‘Book of Abstracts of the 5th Annual Meeting of the Society of Magnetic Resonance in Medicine’ Montreal August 19-22 1986 p. 1067. 16 P. Lauterbur Nature 1973,242 190. 17 M. A. Foster and J. M. S. Hutchison J. Biomed. Eng. 1985,7 171. 18 B. J. Balcom T. A. Carpenter and L. D. Hall Can. J. Chem. 1992 70 2693. 19 R. Freeman ‘A Handbook of Nuclear Magnetic Resonance’ Long- man Scientific & Technical 1988. 20 A. Bax P. G. De Jong A. F. Mehlkopf and J. Smidt Chem. Phys. Lett. 1980,69 567. 21 R. E. Hurd J. Mugn. Reson. 1990,87,422.

ISSN:0306-0012    

DOI:10.1039/CS9942300059

出版商:RSC    

年代:1994

数据来源: RSC

摘要:

Thermodynamic Properties of Additive-Surfactant-Water Ternary Systems R. De Lisi and S. Milioto Department of Physical Chemistry University of Palermo Via Archirafi 26 90123 Palermo Italy 1 Introduction Surfactants are often used in colloidal chemistry because of their amphiphilic nature. For example they stabilize naturally occur- ring dispersions of different types (emulsions suspensions foams) which are used in many technologies. Therefore an understanding of surfactant systems is important from both theoretical and practical points of view. In the last few years extensive structural kinetic and thermo- dynamic studies have been performed on surfactant-water systems. The thermodynamic properties investigated (volume heat capacity enthalpy Gibbs energy entropy compressibility etc.) have been interpreted using appropriate models. In addition thermodynamic studies have been devoted to water-surfactant-additive ternary systems so that a clear picture even if not complete of additive-surfactant interactions is available. In particular the solubilization process of additives in micelles has been studied2-I5 by considering the nature of both the additive and the surfactant. Moreover more attention has been paid to polar additives than to apolar ones because the latter have very low solubility in water. The effect of additives on 2 Water-Surfactant Systems The thermodynamic characterization of water-additive-surfac- tant ternary systems often requires the knowledge of the water- additive and water-surfactant systems. In spite of the very wide literature of solutes in water only in the last few years has interest grown in studies of surfactants in water. The usual approach to surfactant solutions is that of studying a property as a function of surfactant concentration (ms).At a given concentration called the critical micelle concentration (cmc) the property shows a more or less abrupt change in slope (see Figure 1). This experimental evidence is consistent with the association of monomer surfactant molecules into aggregates called micelles. The shape and the size of micelles depend on different parameters (temperature pressure concentration nature of the surfactant etc.). By assuming that the formation of the micelles corresponds to that of a new phase,23 in the micellar region the increase in mS leads to the increase in the concent- ration of the micellized surfactant while that of the unmicellized surfactant is constant and equal to the cmc. So the cmc represents the solubility of monomeric surfactant in water and the properties of micellization has been also ~t~died.~7~,~~,~~-~~ therefore it is easily correlated to the standard free energy of In this review thermodynamic results of water-additive- surfactant ternary systems are reported. Additives which essen- tially solubilize in the aqueous phase and additives which solubilize in the micellar phase are both briefly discussed. Among the former compounds which can affect the properties of the surfactant through their interactions with the solvent (strong inorganic electrolytes urea etc.) are distinguished from 8-cyclodextrin which interacts directly with the alkyl chain of the surfactant to form inclusion complexes. As far as additives solubilizing in the micellar phase are concerned homologous series of polar additives (alcohols for instance) are discussed. Attention is also paid to crown ethers whose solubilization in micelles strongly depends on the nature of the surfactant counterion. Finally the importance of a given experimental thermodyna- mic approach as well as that of theoretical models is emphasized. Rosario De Lisigraduated in Chemistry in 1967 and waspostdoc- toral fellow until 1970 at the Institute of Physical Chemistry University of Palermo. From 1970 to 1980 he taught Physics Physical Chemistry and Spec- troscopy and since 1980 he has been Professor of Physical Chemistry of Interfaces. During 1977-1980 he was In- vited Professor at the Univer- sity of Sherbrooke Canada. He worked on conductometric studies of interactions in solu-tion and on anomalous proton mobility. Since 1977 his re-search activity has focused on the thermodynamics of surfac-tant solutions. He has pub- lished some 60 papers on this topic. micellization (d,G"). Other thermodynamic properties for the micellization process can be derived from the dependence of d,G" on temperature and/or pressure. However since micelliza- tion is not a true phase-transition it occurs in a more or less wide range of concentration around the cmc and the uncertainties which affect the cmc are reflected in the derived properties. An accurate approach to the thermodynamics of surfactant solutions is based on the direct determination of surfactant properties as a function of ms. In the pre-micellar region the properties change almost linearly and give information on the solvent-monomer and monomer-monomer interactions. Just above the cmc the properties change strongly with mS owing to the transfer of the surfactant from water to the micelles. At high ms the properties tend to a constant value and deviations reflect micelle-micelle and monomer-micelle interactions. Examples are shown in Figure 1 where the apparent molar volume (VQ,s) Stefania Milioto graduated in Chemistry in 1985 and received her Ph.D. degree in 1989 from the University of Palermo Italy. She spent several months at the 'Centre des Recherche sur les Macromolecules (France) at the University of Saskatchew-an (Canada) and at the Oak Ridge National Laboratory (USA) before taking up her present position as Research Assistant in the Department of Physical Chemistry Univer-sity of Palermo in 1991. She has collaborated with Pro-fessor R. De Lisi on research into the thermodynamics of surfactant solutions since 1985. She has more than 40 papers to her credit. 67 68 1000 ,',,"',",,,','' ,,,,I,,ll 1111 11 4 262 261900 260 -2 i'800 Lo X \ b 259 0 \ v) 700 3" 258 37 600 257 256 0 01 02 03 04 05 06 07 08 ms/ mol kg-' Figure 1 Apparent molar volume and heat capacity of decyltrimethyl- ammonium bromide in water as a function of surfactant concent ration at 298 K and heat capacity (C S) of decyltrimethylammonium bromide in water are plotted as a function of ms To extract from the experimental data information on inter- actions within surfactant solutions theoretical models are needed However since the treatment of the water-surfactant system is outside the scope of this review we will only briefly summarize the most used models and quote references where full details can be found The pseudo-phase-transition model is currently used for surfactants with a low cmc According to this model," 24 the property of micellization is given by the difference between the partial molar property of the micellized and unmicellized surfac- tant at the crnc The two values are obtained by extrapolating at the cmc the trends in the post- and pre-micellar regions of a given thermodynamic property as a function of mS For surfactants having a high cmc value a mass action model accounts more satisfactorily for the thermodynamic properties The simplest model is the two-step mass action model for non- ionic surfactants proposed by Desnoyers and co-workers This model assumes that the monomers aggregate into monodis- perse micelles The concentration of unmicellized and micellized surfactant depends on the stoichiometry through the micelliza- tion constant (K,) and the aggregation number (N)From the fit of the resulting equation to the experimental data are derived K N and the corresponding property for the surfactant in the micellized and unmicellized form This model was applied also to ionic surfactants by taking into account the coulombic interactions 26 For ionic surfactants different thermodynamic properties have been treated16 with a model valid for mixed electrolytes which uses the two-step mass action model and assumes that the aggregation number and the fraction of counterions bonded to the micelle are independent of temperature and pressure Another mode114 is based on the electrostatic cell approach in which the entire micellar solution is divided into spherical cells each containing one micellar aggregate and amounts of both water and electrolyte consistent with the overall composition As a consequence the Gibbs energy is expressed in terms of hydrophobic and electrostatic interactions and of mixing of micelles 3 Water-Additive-Surfactant Systems As we mentioned above additives can solubilize in the micellar phase depending on their nature Therefore it is useful to discuss the phenomena according to the following scheme (I) Additives which do not penetrate micelles Although a third component added to a micellar solution does not solubilize in the micellar phase it can involve different interactions with surfactant which depend on the nature of the CHEMICAL SOCIETY REVIEWS 1994 additive Therefore thermodynamic properties are differently affected We can distinguish the following (a) Strong inorganic electrolytes (KBr KCl etc ) which interact with the micelles electrostatically (b) Highly polar non-ionic additives (methanol acetone urea acetonitrile etc )which affect the micellization process through their effect on the physicochemical properties of the solvent (c) Additives such as cyclodextrins -very different in nature and structure from the above additives -which can interact strongly with the surfactant alkyl chain because of the formation of inclusion complexes (u) Additives which penetrate micelles In the micellar aggregates three sites of solubilization can be identified the micellar core (highly hydrophobic) the micellar surface (highly hydrophilic) and the palisade layer (the region between the head group and the core) Therefore depending on the site of solubilization there are three classes of compounds (a) Apolar additives (e g alkanes) which are essentially solu- bilized in the micellar core Very few data are available2' for these compounds because their low solubilities in water produce experimental problems (b) Polar additives (medium alkyl chain-length alcohols nitriles nitroalkanes etc ) whose site of solubilization is the palisade layer (c) Complexes with inorganic ions which can solubilize at the micellar surface depending on the nature of the surfactant and/ or of the complexing molecule 'l9 27 Unfortunately thermo- dynamic data are available only for crown ethers s l9 4 Additives which do not Penetrate Micelles There is not an extensive literature reporting thermodynamic studies of additive-surfactant-water ternary systems where the additive is solubilized in the aqueous phase This is not fortui- tous since the relevant importance of surfactant systems is the high solubilizing power that micelles show towards additives whose solubility in water is low Properties of additives in surfactant aqueous solutions are quite scarce However data are available for the properties of surfactants in additive aqueous solutions 4.1 Additives in Surfactant Solutions For additives which do not penetrate micelles if ma is not high (in the molality scale) a small dependence of their properties is expected on the additive (ma)and surfactant concentrations Accordingly for methanol (MeOH) which is the most extensi- vely studied additive the apparent molar volume (V@,)in sodium dodecylsulfate (NaDS) micellar solutions does not change with m and ms V a is practically equal to that in water as observed also in sodium decanoate (NaDec) These values are smaller by 2 5 and 10 2 cm3 mol than the molar volume and the standard partial molar volume in octane respectively Similar results are obtained from the isoentropic compressibi- lity In NaDec," this property is close to that in water and smaller than that of the pure liquid Also for urea in DTAB micellar solutions both volume and heat capacity are scarcely dependent on ms and md and therefore consistent with the additive-surfactant interactions in the aqueous phase only These findings confirm that MeOH and urea cannot penetrate micelles More extensive studies deal with the enthalpies of transfer of additives from water to micellar solution From these infor- mation on the effect of the head group of both the additive and the surfactant can be drawn MeOH has been studied in non- ionic (dodecyldimethylamine oxide DDAO) cationic (dodecyl- trimethylammonium bromide DTAB) and anionic (NaDS) surfactant solutions Experimental data confirm that micelle- additive interactions are absent and that weak hydrophilic interactions in the aqueous phase are present in NaDS and DTAB while negligible in DDAO A different behaviour4 is THERMODYNAMIC PROPERTIES OF TERNARY SYSTEMS-R observed when the -OH group is replaced by the -CN group In fact while CH,CN is a cosolvent in DDAO it is a penetrating additive in DTAB because of the favoured interactions at the micellar surface This view is corroborated by data for nitro- methane in DTAB4 whose distribution constant is equal to that of the longer alkyl chain butanol Cyclodextrins are cyclic carbohydrates consisting of six seven or eight a-D-glucopyranose units called a-,/3-,and y-cyclodex- trins respectively Although the formation of inclusion com- plexes of cyclodextrins with unmicellized surfactants has been investigated,z0 their thermodynamic properties in micellar solu- tions are practically unknown As far as we know only the apparent molar volumes and heat capacities of /3-cyclodextrin (p-Cy) 0 005 rn as a function of rns in NaDS and DTAB are available The V vs rns trends in both surfactants are practically the same V@dincreases strongly with rns in the pre- micellar region and slowly decreases in the micellar region An opposite behaviour is observed for CGdIn fact by increasing rns Cad sharply decreases up to the cmc beyond which it increases Also for NaDS the CQ,vs rns curve shows a jump at ~0 Irn due to the NaDS postmicellar transition Wh*le it is easy to explain data in terms of the formation of inclusion complexes in the pre-micellar region this is not the case for the micellar region 4.2 Surfactants in Water-Additive Mixtures As a general rule at a fixed rn apparent molar properties of the surfactant (Y s) as a function of rns are similar to that in water shown in Figure 1 and the same theoretical models can be used A few thermodynamic properties of surfactants in inorganic strong electrolyte solutions are available Nevertheless these studies offer insights into the effect of these additives on the interactions in micellar solutions For example addition of NaBr to alkyltrimethylammonium bromide (as )involves an increase in osmotic coefficients attributed to reduced repulsive interactions between the micelles This effect is not evidenced by volume and compressibility data of sodium dodecanoate in NaCl solutions 21 In fact addition of NaCl affects these proper- ties for the unmicellized surfactant but not for the micellized surfactant This means that these properties are practically insensitive to micelle-micelle interactions Attention has been paid to the effect of the electrolyte on the micellar structural transition which surfactants sometimes undergo For example from volume and heat capacity data,22 the hexadecyltrimethylammonium bromide postmicellar transi- tion in KBr and KCl aqueous solutions was investigated An increase in the electrolyte content induces the transition which appears at low rn for KBr and high rnd for KCl As far as the non-ionic solutes are concerned the water- DTAB-urea system has been extensively investigated at 298 K At a given urea concentration the profiles of Y (volume heat capacity relative enthalpy) and the non-ideal contributions to the free energy and entropy as functions of rns are similar to those for the corresponding properties in water From a quanti- tative point of view by adding urea these profiles are progressi- vely shifted (towards higher or lower values depending on the property) up to 3 rn beyond which they are virtually unaffected Figures 2 and 3 show examples of the effect of urea on the trend of the apparent molar relative enthalpy (L@s) and the non-ideal Gibbs energy (G;') as a function of the 'normalized' concent- ration Inrns/cmc respectively Since the increase in the urea content acts in the same direction as temperature it was assumed that urea at 298 K breaks up the water molecules network in a similar manner as does temperature As the entropies of micellization as a function of urea concentration at different temperatures show (see Figure 4),this finding is still valid at lower temperatures but not at higher temperatures In fact the slopes of these trends are negative at 288 and 298 K and positive at 308 K the change in the sign occurring at about 300 K These results indicated that around 300 K there is a change in the urea effect on the water structure DE LISI AND S MILIOTO 69 0-P '9 -1 -2 --3 --4 -A ma =7 '~"~""*'"'""'~'''~ -1 0 1 2 3 4 In mS/cmc Figure 2 Effect of urea molality on apparent molar relative enthalpy VF surfactant concentration for dodecyltrimethylammonium bromide at 298 K 0 -5 dI 4 2 b 24 -10 \ dCN u -15 0 ma 3 -20 '""""""'""""' 05 10 15 20 25 30 35 In ms/cmc Figure3 Effect of urea molality on non-ideal Gibbs energy PS surfactant concentration for dodecyltrimethylammonium bromide at 298 K t T 308 K --6 r\ 0" 0 05 10 15 20 25 30 35 ma/ mol kg ' Figure 4 Effect of temperature on the dependence of the entropy of micellization of dodecyltrimethylammonium bromide on the urea molality In contrast to what is observed for DTAB in urea solutions for NaDec in water +MeOH mixtures 'by increasing alcohol concentration V of micellized surfactant increases regularly while that of unmicellized surfactant is a concave curve The former trend was ascribed to a decrease in the electrostriction of water molecules at the micellar surface The latter was assigned to concomitant effects due to (1) a maximum number of structure-promoted water molecules caused by the hydrophobic CHEMICAL SOCIETY REVIEWS 1994 surfactant; (ii) the decrease in hydrophobic and hydrophilic hydration of the surfactant. The effect of 8-cyclodextrin on the Y,,s vs. ms profiles6 is peculiar. In fact Va,sand Ca,sof NaDS and DTAB in water +/3-Cy mixture are strongly affected in the premicellar region and hardly affected in the postmicellar one. However these trends seem to be dependent on the nature of the surfactant. In fact V& of DTAB shows a maximum at x 0.01 m beyond which it strongly decreases up to the cmc and then slowly tends to the value in water. In the case of NaDS the maximum was not detected. C,,s of DTAB and NaDS are very similar. In the pre- micellar region they strongly decrease while those in water are essentially constant; in the post-micellar region they are con- stant and practically equal to those in water. These results are consistent with the formation of inclusion complexes with the alkyl chain of monomeric surfactant in the pre-micellar region. 5 Additives which Penetrate Micelles There are several ways for studying the solubilization of addi- tives in the micellar phases. The choice of the experimental approach to use depends on the nature of the system to be investigated and the information required. If the study is aimed at determining the distribution constant of the additive between the aqueous and the micellar phases different techniques (solu- bility gas chromatography vapour pressure for instance) yield directly the amount of the additive in the two phases. Moreover the distribution constant can be obtained from the dependence of the cmc on the additive concentration8 In -= [2.3Ks +-ma‘Icmc,cmc+a 55.5F where crnc, represents the crnc in water +additive mixture. P is the partition constant in the mole fraction scale; Ks is the Setchenov constant and Fis an ‘activity coefficient’. The numeri- cal coefficient 2 takes into account the dissociation of the I:1 surfact ant. If the distribution constant is determined as a function of temperature and/or pressure several thermodynamic properties of transfer (enthalpy heat capacity volume compressibility expansibility etc.) of the additive from the aqueous to the micellar phases can be calculated. However the uncertainty in the determination of the distribution constant and the small change in the intensive variables lead to unreliable derived properties and consequently only qualitative conclusions can be drawn. In addition even precisely determined properties of transfer cannot yield information on the additive-surfactant interactions in the aqueous and micellar phases. Standard thermodynamic properties are sensitive to solute-solvent inter- actions and by comparing a given property in different solvents insights into the nature of the interactions involved in the solubilization process can be obtained. For example,’ at 298 K the standard partial molar heat capacity of pentanol is 532 J K-mol-in water 191 J K-mol-in dimethylformamide 208 J K-’ mol-l inethylene glycol 136 J K-’ mol-l in octane while its molar value is 208 J K -mol -l. Consequently since a given property of the additive is expected to be different in the micellar and aqueous phases the extraction of the additive from the aqueous phase (obtained by increasing ms) affects the bulk property (Figure 5). Therefore a better approach to the thermo- dynamics of solubilization of an additive in micellar solutions is based on the determination of a given standard partial molar property as a function of ms. By fitting the bulk experimental data using an appropriate model both the distribution constant and the property of the additive in the micellar phase can be obtained. Often the apparent molar property at low additive concentration is analysed instead of the standard one since they are very close in value. In addition the study is less time- consuming since for a given ms,measurements as a function of the additive concentration are not needed. n fl I106 -4I $; 400 1lo5K 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 ms/ mol kg Figure 5 Apparent molar volume and heat capacity of pentanolO.05 rn in decyltrimethylammonium bromide aqueous solutions as a function of the surfactant concentration at 298 K. 5.1 Basic Models If the additive can be considered in a state of infinite dilution it behaves as a probe unaffecting the physicochemical properties of the micelles. Therefore to rationalize the investigated property as a function of ms the following contributions are almost sufficient. (i) Additive-surfactant interactions in the aqueous phase. (ii) Additive-surfactant interactions in the micellar phase. (iii) Distribution constant of the additive between the aqueous and the micellar phases. (iv) Shift of the micellization equilibrium due to the presence of the additive. (v) Structural variations (shape aggregation number degree of ionization changes) of the micelles. At present no models taking into account structural variations (point v) have been proposed. The simplest approach considers the points (i) through (iii) and is based on the pseudo-phase-transition model for both the micellization and the additive distribution processes.* Conse- quently the property of the additive in the aqueous (Yf)and micellar (Yb) phases contribute to the bulk property (Y,) through the fraction of the additive in the corresponding phases (Nf and Nb) where Nband Nf are given by Nb = K(ms -cmc) 1Nf = 1 +K(ms -cmc) 1 +K(ms -cmc) (3) In equation 3 K is the distribution constant of the additive between the aqueous and the micellar phases. If the cmc is sufficiently low then Y can be considered to be equal to the value in water (Y,). From equation 2 Yb can be obtained provided that Kand Y are known. The model has been applied to enthalpies of solution of various alcohols in hexadecyltri- methylammonium bromide micellar solutions. From these data the standard enthalpy of transfer of the additive from water to the micellar phases has been derived. Since the property of the surfactant in the micellized and unmicellized form are generally different the change of the cmc with the additive affects the property of the water +surfactant binary solvent. This effect is reflected in the bulk property of the additive. It is to be stressed that for a ternary system at fixed ma by increasing ms the monomer concentration is not constant even in the hypothesis of the pseudo-phase-transition model. Consequently the micellization shift contribution depends on ms and by neglecting it the quantities derived from equation 2 THERMODYNAMIC PROPERTIES OF TERNARY SYSTEMS-R are more or less reliable depending on the nature of both the additive and the surfactant This contribution was taken into account in the model proposed by Desnoyers and co-workers O who assumed a mass- action model for the micellization process and the pseudo- phase-transition model for the additive distribution and by De Lisi et a1 who considered the pseudo-phase-transition model for the micellization and a mass-action model for the distribution In spite of the difference in the models the resulting equations for Y as a function of ms are very similar where [m,]and [m]are the unmicellized surfactant concent- rations in the absence and in the presence of the additive respectively m is the additive concentration and AY is the property of micellization Following the approach discussed by De Lisi et a1 ,2 in the limit the additive concentration tends to zero the shift of the micellization equilibrium can be expressed as By introducing equations 3 and 5 into equation 4 the follow- ing equation is obtained from which Yb Yf and Kcan be obtained through non-linear regression By converting K into the partition constant the standard Gibbs energy of transfer (AG;)of the additive from the aqueous to the micellar phases can be calculated In order to solve equation 4 Desnoyers and co-workersi0 evaluate [m,],d Y the aggregation number and the constant of micellization (KIM)from the water-surfactant binary system (see Section 2) Then KMand the distribution constant have been correlated with ms and with the additive concentration in the aqueous and micellar phases From the resulting equations the distribution constant and Yb are calculated by successive approximations These models have been applied to different thermodynamic properties (volume heat capacity enthalpy compressibility) for Ovarious additives and surfactants 2-4 Other models from which it is difficult to extract the property of the additive in the micellar phase have been proposed by Johnson et a1 I4 and by Christian et a1 The first one is based on the electrostatic cell model for water +surfactant systems expanded in order to take into account the presence of the additive In other words one additional contribution to the Gibbs energy due to the distribution of the additive was con- sidered The second one is based on a two-step mass-action model for the micellization and a multi-step model for the distribution of the additive By assuming that the micelles are monodisperse and that the equilibrium constants of the different steps are equal (with the exception of that of the solubilization of the first molecule in the micelle) an equation correlating the aggregation number the constant of micellization and the various distribution constants was obtained 5.2 Examples of Applications In this paragraph some examples of properties derived from equation 4 applied to the experimental bulk properties of the additive in surfactant solutions are reported in order to show that information on additive-micelle interactions in connection with the nature of both the additive and the surfactant can be obtained In addition the above models can be also used for a DE LISI AND S MILIOTO 71 quantitative treatment of the transfer properties of the surfac- tant from water to water +additive mixtures 5 2 1 Polar Additives Polar additives in micellar solutions have been extensively explored However these studies are essentially limited to the determination of the distribution constant and are not systema- tic Data reported here were chosen to show the effect of the hydrophobic and hydrophilic moieties of both the additive and the surfactant on the thermodynamics of solubilization of additives in the micellar phase For this reason complete sets of thermodynamic properties of homologous series are needed In this review we refer essentially to medium alkyl chain length alcohols Comparisons between a given standard thermodynamic property of an additive in micellar phase (Yb) and that in different solvents give information on the nature of the additive- micelle interactions and therefore on the site of solubilization of the additive in the micelle For a homologous series of additives Yb generally changes linearly with the alkyl chain length (n,) of the additive where A and B represent the hydrophilic and the CH hydropho- bic contributions to Yb respectively The validity of equation 7 implies that an additional CH group does not affect the hydrophilic interactions and that it is equivalent to the other methylene groups Figure 6 shows molar volumes (P)and standard partial molar volumes in DTAB micellar phase (&) in water (V,) and in octane (V,,,) for some primary alcohols as functions of n As can be seen equation 7 is always satisfied the intercept and slope depending on the medium solvent The CH2 group contribution to volume in the micellar phase (1 6 7 cm3 mol- I) is practically equal to that in pure liquid alcohols (16 9 cm3 mol-l) and in octane (1 6 5 cm3 mol- l) but different from that in water (1 6 0 cm3 mol-l) If these results indicate that the environment of the CH group is hydrophobic in nature they do not indicate if the alcohol is solubilized in either the core or the palisade layer of the micelle The hydrophilic contribution in the micellar phase (23 9 cm3 mol- l) is equal to that in pure liquids (24 4 cm3 mol-') close to that in water (22 5 cm3 mol-') and smailer than that in octane (33 2 cm3 rno1-I) These findings suggest that alcohol solubilizes in the palisade layer of the micelle being involved in both hydrophilic and hydrophobic interactions A different behaviour is observed for alkanesi2 since V values in sodium 140 -r-200 120 190 100 180 '4 2 170 m 80 E 160 60 150 40 140 20 I I 1 I I I 0 2 4 6 8 10 "c Figure 6 Molar volumes (triangles) and standard partial molar volumes of alkanes (filled symbols) and alcohols (open symbols) in the micellar phase (squares) in water (circles) and in octane (cross) as a function of the additive tail at 298 K The micellar phase refers to sodium dodecanoate for alkanes and to dodecyltrimethylammoniumbromide for primary alcohols 72 CHEMICAL SOCIETY REVIEWS 1994 20 I I I I I I 10 IA -8 .. Ll 2 m6 E :..=*A 30--O-0 3 I7\ a2 4 0 L 4 I I I I I I 1-15'0 1 2 3 4 5 6 7 nc Figure 7 Standard Gibbs energy (squares) enthalpy (circles) and entropy (triangles) of transfer of nitroalkanes (open symbols) and primary alcohols (filled symbols) from the aqueous to the micellar phases of dodecyltrimethylammoniumbromide at 298 K dodecanoate micelles are equal to V* and higher than those in water (Figure 6) This pattern indicates that the site of solubiliza- tion of alkanes in micelles is the hydrocarbon liquid core The same information was derived from heat capacity data As far as other thermodynamic properties are concerned Gibbs energy enthalpy and entropy of transfer of alcohols from the aqueous to the micellar phases are available 28 The additi- vity rule (equation 7) for these properties is not expected to be valid because the aqueous phase is not a pure phase Of course if additive-surfactant interactions in the aqueous phase can be neglected then equation 7 can be used This is the case for Gibbs energy the CH group contribution to this property is -2 3 kJ mol-l for all surfactants 2-4 In the case of enthalpy and entropy (Figure 7) their plots vs n curves show maxima at n z 5 or 6 depending on the nature of the system The same profiles are obtained if the properties of transfer are corrected for the additive-surfactant interactions in the aqueous phase I e if the transfer from water to micelle is considered More direct information on the additive-micelle interactions are obtained through the solvation properties These were calculated for primary alcohols and alkanes in NaDS micelles2 and compared with those in water and octane These properties were interpreted in the same manner as those of transfer and the information obtained was consistent with that derived from other thermodynamic properties As mentioned above the nature of the polar head of the additive in the solubilization process in the micelle has a relevant effect since MeOH behaves like a co-solvent while nitromethane penetrates the micelles From homologous series of nitriles nitroalkanes and alcohols in DTAB4 the affinity of the micelles towards -CN and -NO2 groups is higher than that towards the -OH group while opposite behaviour was found for the methy- lene group While the transfer of the CH group from the aqueous to the micellar phases is always governed by the entropy that of the polar groups depends on their nature The transfer of -NO2 and -CN groups is governed by the entropy and that of the -OH group by the enthalpy 52 2 Crown Ethers Crown ethers are macrocyclic polyethers which with inorganic cations form stable complexes having mostly 1 1 stoichiometry the result of strong ion-dipole interactions They show a high degree of selectivity towards specific cations which depends on the radius of both the cation and the cavity of the crown ether l9 For example 18-crown-6 and 15-crown-5 form stable com- plexes with sodium ions whereas 12-crown-4 does not Studies of these additives in micellar solutions gave insights into the effect of the complexes' formation on their solubilization in micelles -2 0 01 02 03 04 05 m,/ mol kg Figure 8 Apparent molar volume corrected for that in water of pentanol (triangles) 18-crown-6 (circles) and urea (squares) in sur-factant solutions as a function of the surfactant concentration Filled symbols sodium dodecylsulfate open symbols dodecyltrimethylam- monium bromide The apparent molar volume and heat capacity of 18-crown-6 0 04 m as a function of NaDS concentration are typically trends of distribution Accordingly the V vs ms trend is similar to that of pentanol in NaDS (Figure 8) The distribution of crown was ascribed to the formation of a complex which can be adsorbed at the micellar surface In fact in DTAB micellar solutions where crown ethers cannot be complexed both V a and C are independent of ms such as it occurs for urea The properties for pentanol possess again the feature of a distribut- ing additive Therefore in order to obtain the distribution constant and the property of the complexed crown in the micellar phase equation 6 was modified to take into account the complexation equilibrium in the aqueous phase The K value agreed with that obtained from NMR techniques The Vbvalue is smaller than that in the aqueous phase counter to what is observed for other additives This pattern is consistent with the substitution at the micellar surface of uncomplexed sodium ion with the complexed one This process involves the formation of free sodium ions which causes a decrease in the volume 52 3 Transfer of Surfactants from Water to Water + Additives Mixtures The transfer properties dYas (w+w + a) = Yas(w + a) -Yas(w) (8) of surfactants from water to aqueous solutions of alcohols and l7crown ethers as a function of ms are available The shapes of these curves are characteristic of the system investigated For example for NaDec in water + propanol mixtures d V s(w -+ w + a) monotonically increases with ms at low md,whereas it shows a maximum at higher ma l7 Maxima are present also in the d VQS(w-+ w + a) and in the isoentropic compressibility of transfer of DTAB in water + pentanol mix- ture at different temperatures As a general feature by increas- ing the hydrophobicity of the alcohol the maximum is shifted towards lower ms values and its amplitude increases A similar behaviour was observed for the volume of transfer of NaDS from water to water + 18-crown-6 mixtures for DTAB d V s(w + w + a) is negative in the pre-micellar region and null in the post-micellar region (Figure 9) The effect of the size of the cavity of crown ethers was also investigated by studying NaDec in 12-crown-4 15-crown-5 and 18-crown-6 l9 Both the volume and the isoentropic compressibility of monomeric and micellized surfactant increase with crown concentration These results were quantitatively explained in terms of the extraction of the additive from the aqueous phase to the micelle which depends on the distribution constant In the case of crown ethers THERMODYNAMIC PROPERTIES OF TERNARY SYSTEMS-R DE LISI AND S MILIOTO 0 0 0 A A &A A A A A 0 0.1 0.2 03 0.4 0.5 ms/ mol kg-' Figure 9 Volume of transfer of sodium dodecylsulfate (circles) and dodecyltrimethylammonium bromide (triangles) from water to water + 18-crown-6 (filled symbols) and water +pentanol (open symbols) as a function of the surfactant concentration at 298 K the distribution constant is related also to the complexation constant which for Na+ is known to increase in the order 12-crown-4 15-crown-5 18-crown-6 Acknouledgernents The authors are grateful to the National Research Council of Italy (CNR Progetto Finalizzato Chimica Fine 11) and to the Ministry of Universlty and of Scientific and Technological Research (MURST) for financial support 6 References 1 'Surfactant Solutions New Methods of Investigation' ed R Zana M Dekker New York 1987 2 R De Lisi and S Milioto J Solution Chem 1988 17 245 and references therein 3 R De Lisi A Lizzio S Milioto and V Turco Liven J Solution Chem 1986 15,623 4 S Milioto and R De Lisi Thermochim Acta 1988 137 151 5 M Bakshi R Crisantino R De Lisi and S Milioto Langmuir submitted 6 M Bakshi R Crisantino R De Lisi and S Milioto manuscript in preparation 7 P Stilbs J Colloid Interface Sci ,1982 87 385 8 C Treiner A K Chattopadhyay and R Bury J Colloid Interface Sci ,1985,104 569 9 R De Lisi S Milioto and A Inglese J Solution Chem 1990 19 767 10 D Hetu A H ROUX and J E Desnoyers J Solution Chem ,1987 16,529 11 S D Christian E E Tucker and E H Lane J Colloid Interface Sci ,1981,84,423 12 E Vikingstad and H Holland J Colloidlnterface Sci ,1978,64,5 10 13 R De Lisi S Milioto and R E Verrall J Solution Chem ,1990,97 97 and references therein 14 I Johnson G Olofsson M Landgren and B Jonsson J Chem SOC Faraday Trans 1 1989,85,4211 15 R De Lisi S Milioto and A Inglese J Phys Chem 199 1,95,3322 16 L V Deaiden and E M Woolley J Phys Chem ,1987,91,2404 17 E Vikingstad J Colloid Interface Sci 1980 74 16 18 E Caponetti S Causi R De Lisi M A Flonano S Milioto and R Triolo J Phys Chem ,1992,96,4950 and references therein 19 E Vikingstad and J Bakken J Colloid Interface Sci ,1980,74 8 20 E Junquera G Tardajos and E Aicart Langmuir 1993 9 1213 and references therein 21 H Hrailand and E Vikingstad J Colloid Interface Sci ,1978,64,126 22 F Quirion and J E Desnoyers J Colloid Interface Sci ,1986 112 565 23 P Mukerjee Adv Colloid Interface SLI 1967 1 241 24 K Kale and R Zana J Colloid Interface Sci ,1977,61 312 25 J E Desnoyers G Caron R De Lisi D Roberts A Roux and G Perron J Phys Chem ,1983,87 1397 26 G Caron G Perron M Lindheimer and J E Desnoyers J Colloid Interface Sci 1985 106 324 27 G Calvaruso F P Cavasino C Sbriziolo and M L Turco Liven J Chem Soc Faradaj Trans 1993,89 1373

ISSN:0306-0012    

DOI:10.1039/CS9942300067

出版商:RSC    

年代:1994

数据来源: RSC