J. MATER. CHEM., 1994, 4(8), 1337-1341 1337 Ion Exchange of Ruthenium Cationic Complexes by a-Tin@) Bismonohydrogenphosphate Michael J. Hudsonfaand Andrew D. Workmanb a Department of Chemistry, University of Reading, Box 224, Whiteknights, Reading, Berkshire, UK RG6 2AD Department of Chemistry, University of Leicester, University Road, Leicester, Leicestershire, UK L€2 7RH The ion exchange from aqueous solution of some cationic ruthenium species by a-tin(rv) bismonohydrogenphosphate (SnP) has been investigated. There is a marked difference in behaviour according to the cationic species which are used. Thus the stable [Ru(NH,)~]~+ is only exchanged on the surface; ruthenium red [Ru,O,(NH,),,]~+ and ruthenium nitrosyl species were ion exchanged onto the surface from dilute solutions but, at higher concentrations, the host SnP, and its butylamine intercalation compound SnP-BuA, were delaminated to give disordered non-crystalline materials.In contrast, the reactive [Ru(NH,),]*' was rapidly extracted in an autocatalytic, topotactic reaction to give a polyphasic but microcrystalline intercalation compound, the layered structure of which was retained on heating. At high loadings there appears to be an interlamellar electron transfer resulting in the oxidation of ruthenium(i1) to ruthenium(iii) probably with tran~-[Ru(H,O),(NH,),1~+ as the dominant guest cation. There is much current interest in the ion exchange of metal- containing cationic species into layered hosts. The interest derives from the extraction of radionuclides and the prep- aration of microcrystalline pillared layered solids.One suitable host is the inorganic ion exchanger a-tin(1v) bismonohydrog- enphosphate monohydrate [Sn( HPO,),-H,O] (SnP), which is a layered (hydrogen) phosphate with a basal spacing of 0.78nm, an area of 21.4A2 per phosphate group, and a maximum cation exchange capacity (CEC) of 6.08 mmol 8-l for a monovalent ion.' Despite this high CEC, SnP is regarded as a poor ion exchanger because the free diffusion of counter- ions is limited by the strong Layer-layer interactions and the small passageways (ca. 2.6 A) that connect the interlayer cavities. Normally, only surface ion exchange will occur.' Cationic metal complexes are not known to intercalate directly into SnP or other phosphates and, consequently, alternatives to the direct intercalation have been investigated.For example, amine intercalation compounds have been prepared in order to separate the layers and to reduce the interlayer charge den~ity.~'~In certain cases delamination processes are used to encourage the ion exchange of cationic metal complexes in layered phosphates.' Highly dispersed and hydrated powders of layered phosphates obtained by the sonication of some amine intercalated phosphates may be used to extract large cations such as aqueous Ba2f.6 Unfortunately, both the delamination and sonication methods lead to ill-defined, amorphous products. In this study, the extraction of ruthenium-containing cations has been investigated and the reactivity of the ruthenium@) cation [Ru( NH3)6]2+ has been exploited to produce microcrystalline layered intercalation compounds.A new autocatalytic mechanism has been pro- posed in which labile ammonia ligands extract protons from the host, rendering SnP a stronger acid and a more effective ion exchanger. Experimental Synthesis SnP was synthesized according to the published pr~cedure.~ Its X-ray powder diffraction (XRD) pattern (Spectrolab series 3000 CPS-120 instrument, Ni-filtered Cu-Kcc =0.154051 nm) indicated a basal spacing of 0.78 nm and the absence of any impurity peaks. Hexaamminoruthenium(r1) chloride was prepared by the method of Fergusson and Love.8 The hexa- amminoruthenium(II1) chloride was prepared by the dropwise addition of chlorine water into a stirred solution of the ruthenium(I1) compound at 50 "C.The non-radioactive ruthenium nitrosyl compounds were prepared by the dilution of aqueous ruthenium nitrosyl nitrate (Johnson Matthey Materials) with nitric acid (4 mol drnp3) and alloming the solutions to age at least one month before use. Extraction Experiments All extractions were done in duplicate as batch experiments using degassed, doubly deionised water. The CEC was taken as 6.08 mmol H+ g-' of dry ion exchanger. The solutions were diluted for atomic absorption analysis using hydrochloric acid ( lo%, v/v) and lanthanum(Ir1) chloride penta hydrate (0.5%, w/v) so that the range of ruthenium concentration was 0-100mg dm-3.The mass balances between the amounts extracted from the solution into the solid phases were correct. Since the compounds were polyphasic, the analytical results were not used to calculate the composition of the conipound. Extraction studies for trace quantities of ruthenium nitrosyls were carried out by the Novel Absorbers Club of AEA Harwell Laboratory. The previously aged cocktail of radio-nuclides was at pH 6.4 and contained sodium nitrate (0.05 mol dmV3). The activity of the Io6Ru was 1.23 x lo', Bq g -I. Pyrolysis of SnP-[Ru( NH3)6]2'3+ The pyrolysed form of SnP intercalated with hexaammino- ruthenium@) dichloride SnP-[ RU(NH,)~]~/~ + was prepared as follows. SnP was shaken for 2 h with an [Ru(NH~)~]~+ solution containing 300% of the CEC of SnP.The rcsultant solid was then filtered through a number 4 sintered glass crucible and the exchange process repeated. The filtered sample was washed with doubly distilled water, ethanol and propanone. The air-dried product was heated in dry (sulfuric acid), oxygen-free nitrogen to 1000°C at a ramp rate of 10"C min-' and held at this temperature for 3 h. Scanning electron microscopy was carried out on a JEOL JXA 840 Scanning Microanalyser. Electron paramagnetic resonance (EPR) studies were carried out with a Bruker ESP-300 instrument. All spectra were recorded in the Y-band at 77 K. Results and Discussion Ion-exchange Studies Hexaamminoruthenium(rI1) Cation The direct extraction of [Ru(NH~)~]~+ proceeded only slowly to 14% of the maximum CEC even when 150% of the CEC was present in the initial solution.There was no change in the XRD pattern and clearly only surface ion exchange took place. If ammonia ligands are involved in the exchange, the relative inertness of the cation towards substitution and strong retention of the ammonia ligands may account for its lack of intercalation. The acidity (pK, 12.4)' does not provide an alternative mechanism for the reaction with SnP, which is also a weak acid. Ruthenium Red [RU~O~(NH~)~~]~+ In contrast to the hexaamminoruthenium(II1) cation, the ruthenium red cation was rapidly exchanged by SnP. However, in none of the samples was there any XRD evidence for the formation of microcrystalline intercalation compounds.There appeared to be two well defined stages in the extraction process, Fig. 1. In the first region AB' there was surface exchange possibly followed by delamination (B'C') and an additional stage (C'D') during which the ruthenium red exchanged on the newly formed surfaces. It is known that layered compounds with intercalated monoamines delaminate but those with diamines do not. Therefore, in order to enhance the rate of extraction, the butylamine intercalation (SnP-BuA) compound was used." This has a bilayer of amines with weak van der Waal bonds between the hydrophobic groups of the hydrocarbon chains. This weak intralayer bonding enables the compounds to delaminate. Interestingly, there was less extraction in the region AB than was the case for SnP and it appears that the protonated amine, which is on the surface, is less readily exchanged than the protons in the P-0-H groups.The initial extraction was also much slower with the SnP-BuA compound because the surface BuA groups are less readily exchanged than the protons of the surface PO-H groups." The decrease in the amount of ruthenium red exchanged from B to C or from B' to C' results from the loss of the exchanged ruthenium cations as the compound delaminates. Delamination is necessary for further exchange because the ruthenium red is able to bridge the layers, as do the diamines. As judged by the enhanced extraction, the delamination of SnP-BuA is greater than that of the SnP itself. The products were polyphasic with a microcrystalline unreacted SnP phase and an additional amorphous phase resulting from the delamination and exchange. It was con- sidered that one possible reason why ruthenium red was involved in delamination was that the cation contains exchangeable labile ammonia ligands.Thus the exchange could involve ammonia molecules, which were previously 1.6r , 1 I I 0 400 800 1200 1600 extraction Deriod/min Fig. 1 Extraction of the ruthenium red cation by SnP (a) and SnP-BuA (b) J. MATER. CHEM., 1994, VOL. 4 ligands, and thus enhance extraction compared with the hexaammino rut henium (111) cation. Nitrosyl-containing Cations Ruthenium nitrosyl complexes are formed in nitric acid media during the reprocessing of spent nuclear fuels.The ruthenium nitrosyl solution contained neutral species and cations of charges 1-3 of the general formula [Ru(NO)(H20)x(N03),-x]3-x.."In addition to the nitrosyl ligands, there were also nitrato and water ligands to maintain the overall octahedral ge~metry.'~,'~ The kinetic curve was of a similar shape to that of the ruthenium red, with point A corresponding to 0.15 mmol g-' and 25 min. The extraction under equilibrium conditions, Fig. 2, was a function of the concentration of the original solution. The region from the origin to A and B corresponds to surface extraction and the region beyond B to extraction and delamination. The reason for the increase from B to C is probably connected with the separation of the layer edges in the presence of the excess of cationic species followed by delamination and exchange.The decrease over the region CD may be associated with hydrolysis and loss of phosphate groups. As discussed more fully later, for the compounds with intercalated cationic ammineruthen- ium, when the products were heated ruthenium was retained and a tin(1v) pyrophosphate phase was obtained in which the ruthenium appeared to be covalently bound. Since radioactive (lo6Ru) ruthenium nitrosyl species are present in some aqueous effluents from nuclear industries, there was interest in establishing whether Sn P could extract them from solutions which contain trace quantities such that the concentrations are of the order of 10-12 times lower than those used above.The ruthenium nitrosyl species were indeed extracted by SnP. For example, an initial feed activity for Io6Ru of 73 Bq cmP3 was reduced to 20 within 1 h. Such extraction no doubt involves the surface phosphate groups rather than those in the interlamellar regions. Consequently, if these layered phosphates are to be used for the recovery of pollutants then attention should be given to preparing mate- rials with high surface areas. Hexaamminoruthenium(11) Cation The extraction of [Ru(NH,),I2+ and the resulting materials are quite different from those described above. The extraction curve for hexaamminoruthenium(11) chloride from a degassed, doubly deionised solution (with an initial concentration of ruthenium of 1.5xCEC) onto SnP is shown in Fig.3. Degassed water was used to restrict the conversion of ruthenium(11) to ruthenium(II1). The rate of extraction appears to be quite rapid with t, [the time for half of the ruthenium(11) to be extracted] of <5 min to a capacity of 1.66 mmol g-', 1.8r 7 'm 1.4 --E Eg 1.0-c02 c -$ 0.6 *' 0 100 200 300 CEC (divalent ion) in original soln. (%) Fig. 2 Extraction of some ruthenium nitrosyl cations 3. MATER. CHEM., 1994, VOL. 4 I -cn 1.5 E E ag 1.0 9 c W 1 1 1 50 100 150 200 tlmin Fig. 3 Extraction of the hexaamminoruthenium(r~) [Ru(NH,),]'+ cation (A) by SnP as a function of time which is 55% of the theoretical exchange capacity of 3.04 mmol g-' for a divalent ion.The differences between the extraction of the hexaamminoruthenium(11) and the other cations together with the materials that are formed may be rationalised on the basis of the lability of the ammonia ligands in the hexaamminoruthenium(i~) cation. The rate constant for the exchange of ammonia ligands with water in hexaammino- ruthenium(I1) cation has been measured', as 1.24 x dm3 mol-' s-' whereas the calculated half-life of the ligand- exchange process for the ruthenium(II1)-ammine complex, for example, is 3 years. Thus the ruthenium@) cation has readily exchangeable ammonia ligands which could extract protons from the host. This proton extraction could lead to a local expansion of the layer edges and intercalation of the complex cation.The fate of the ammonium ion is uncertain, but there was little observed change in pH. From the EPR evidence discussed below, it appears that the most likely guest cation is one in which one or more of the ammonia groups have been replaced by a water molecule. It has been suggested previously that trace amounts of sodium can act as catalysts for the intercalation of molecule^.^^ However, the two cata- lysed mechanisms are different because, in the second case, the sodium increases the interlayer distance only and not the effective acidity of the host. Thus the overall reversible mech- anism is: PO-H,,,p, +H,O +[RU(NH3)6J2' =[PO -NH4+](snp)+[Ru( NH3)5H20I2+ This exchange may be repeated to give trans-[Ru(NH,),)(H,O),]~+.It is also possible that the ionised phosphate group could act as a ligand or there could be extensive hydrogen bonding between the ionised phosphate groups and the bound ammonia or water ligands. The separate addition of ammonia to the solution containing the hexaammi- noruthenium(i1) cation did not accelerate the intercalation because SnP is hydrolysed by ammoniacal solutions to give tin@) oxide and phosphate groups. The loss of the phosphate groups reduces the CEC of SnP. In addition, [Ru(NH3)J2+ decomposes in ammoniacal solution, particularly under aero- bic conditions, to produce dark decomposition products.16-" Interlamellar Electron Transfer There are clear indications that there are secondary electron transfer reactions inside the interlamellar regions for the compounds with high loadings of the ruthenium@).The initial evidence comes from the XRD patterns of the intercalation compounds, which changed according to the initial concentrations of the solutions, Fig. 4. Compounds 2-4were biphasic materials in which the unreacted host do02 I I 20 15 10 5 28Idegrees Fig. 4 X-Ray powder diffraction data for the host (1) and the intercalation compounds: 1, SnP; 2, 5% CEC of A in original solution; 3, 10%; 4, 25%; 5, 50%; 6, 75%; 7, 100%; 8, 3Oo"'o (dOo2=0.78nm at 11.3" 20) and the intercalation compound (do,, =1.06 nm) coexist. At higher loadings there is a re-arrangement of the host and guest (sample 4) and a single new microcrystalline phase is formed for which do,, = 0.99 nm.The reasons for the differences in the intensities md pos- itions of the peaks may be related to changes in oxidation state as indicated by the EPR spectra. The first five samples (including the host) up to initial concentrations of 50% CEC were inactive, but from 70 to 300% CEC there were clear EPR signals with g, =3.066, gy=1.115, g, =2.057. The inac- tivity of the first set of samples with the lower initial concen- trations of ruthenium in solution indicate that ruthenium(i1) (spin-paired 4d6) is present. The EPR for the other higher loaded samples indicates that a distorted ruthenium(1ri) octahedral species of C,, symmetry is present (spin-paired 4d5)19320(cf. trans-[RuC1,(NH3), 1' 3.33, 1.54, 1.13; trans-[Ru(en),(NCS), JNCS 3.07, 1.11 and 2.25).The question that must be answered is: why does the electron transfer occur at high loadings and not at low loadings? The oxidation of ruthenium(r1) to ruthenium(rrr) is sensitive to pH with ruthenium@) favoured at low pH.21 The electron transfer involves the ruthenium(I1) cation and either the host or the interlamellar water. It appears, therefore, that although SnP is essentially a weak acid,22 the low pH in the interlamellar regions favours ruthenium(i1) for the lower amounts of exchange (up to 50% CEC and 0.6 mmol Ru g-'). Sample 5 in Fig. 4 corresponds to the half-exchanged material in which the ruthenium(r1) and ruthenium(iI1) species are able to coexist in the interlamellar regions. For a tervalent ion 8Ooh of the CEC is utilised.The reason why not all of the CEC is used is that the ruthenium species cover sites which are not used for ion exchange. The cross-sectional area for the ruthenium(rI1) ion is approximately 0.24 nm2 whereas the free area per phosphate is 0.21 (4) nm2. The estimated cross-sectional area of trans-[ Ru(H,O),(NH,),]~+ is 0.243 nm2 compared with 0.214 nm2 for each phosphate group. Thus some phosphate groups are covered by the guest cation. Thermal Properties The S~P-[RU(NH,),]~'~ compounds were evaluated by+ thermogravimetry (TG) and differential thermochemistry 1340 because it is essential for the nuclear industry that no ruthenium is lost, as lo6Ru04 for example, on heating the compounds. No evidence of loss of ruthenium was found.There are two endotherms (see Fig. 5): the first, at 85"C, is associated with loss of interlamellar water (9.4%, 2.8H20per formula mass); the second, at 430 "C,is related to the decompo- sition of the ruthenium complex and therefore involves loss of water and ammonia, the conversion of phosphate to pyrophosphate is also probably contained in this endo-therm.23,24 The exotherm at 560°C does not involve a mass loss and is associated with the binding of the ruthenium species with the tin pyrophosphate and recrystallisation of the phases so formed. The XRD pattern for this material contains comparatively sharp peaks and the material is more crystalline than SnP, SnP-BuA and the SnP-[ Ru( NH3)6]2'3+ intercalation compound, Fig.6. The peak assignments are shown in Table 1;essentially the material Fig. 5 Thermogravimetry (a) and differential thermal analysis (b) of the S~P-[RU(NH,),]~+ intercalation compound 100 80 h 60 40--20 &lddu0, 2Bldegrees Fig. 6 X-Ray powder diffraction pattern of the SnP-[Ru(NH,),J2+ intercalation compound which had been heated to 1000°C Table 1 X-Ray powder diffractogram of pyrolysed SnP-Ru( NH,), SnP,O, RuO, SnR" [hkr] d/nm I/Z, (%) d/nm I/I, (YO) d/nm 111 0.460 40 0.2217 4 0.458 0.435 110 0.317 100 101 0.2550 50 200 0.399 100 0.2243 10 0.397 210 0.357 30 0.2005 1 0.356 0.343 0.335 21 1 0.326 25 0.1685 30 0.321 220 0.282 30 0.1586 9 0.282 22 1 0.266 2 0.265 002 0.1552 4 "Pyrolysed SnP-Ru"'(NH,),.J. MATER. CHEhl., 1994, VOL. 4 is based on the tin pyr~phosphate~~ with bound ruthenium. There was no evidence for ruthenium(1v) oxide as a separate p hase.26 Morphology The scanning electron micrograph of the SnP-[ Ru(NH3),I3+, Fig. 7, shows that the layered structure is retained but that some of the flakes were curled up. Slow heating of the compound gave a similar morphology but rapid heating, Fig. 8, resulted in holes being formed in the layers by the molecules which had insufficient time to diffuse along the layers. In the synthesis of microcrystalline solids, slow heating rates are considered to be important. Conclusions The extraction appears to involve an ammonia ligand from the reactive ruthenium(11) complex and may be regarded as an autocatalytic, topotactic reaction.The peaks in the XRD patterns are not, however, sharp and there may be some amorphous material in the resultant compound. The reaction of the free ammonia group and the extraction of protons renders the SnP to be a strong acid and an effective ion exchanger. The reactive [Ru(NH3),I2+ was rapidly exchanged in an autocatalytic, topotactic reaction to give a polyphasic, microc- rystalline intercalation compound, the layered structure of which was retained on heating. At high loadings there appears Fig. 7 Scanning electron micrograph of the pyrolysed compound showing the flakey appearance Fig. 8 Scanning electron micrograph of the pyrolysed compound showing the hole caused by the rupturing of the layers of the host when the heating rate was too great J.MATER. CHEM., 1994, VOL. 4 1341 to be an electron transfer involving oxidation of ruthenium@) to ruthenium(m), with truns-[Ru(H2O),(NH,),13f as the resultant dominant guest cation. 9 10 11 D. Waysbort and G. Navon, J. Chem. SOC.D, 1971,1410, M. J. Hudson, E. Rodriguez-Castellon, P. Sylvester, A. Jimenez- Lopez and P. Olivera-Pastor, Hydrometallurgy, 1990,24, 77. C. G. J. Boswell and S. Soentono, J. Inorg. Nucf. Chom., 1981, This work was supported by the Department of the Environment as part of their Radioactive Waste Management 12 13 43, 1625. M. Nowak, Radiochem. Radioanal. Lett., 1971,8, 165. V. G. Shumkov, Jad. Energ., 1974,20,315. Programme.The results may be used in the formulation of Government policy but at this stage do not neces-sarily represent Government policy. Professor B. C. Gilbert, University of York is thanked for use of the EPR equipment. 14 15 16 P. C. Ford, J. R. Kuempel and H. Taube, Innorg. Chtpm., 1968, 7, 1976. G. Alberti, U. Costantino and G. P. Gupta, J. Inorg. Nud. Chem., 1978,40,87. F. M. Lever, Platinum Met. Rev., 1969, 13, 151. 17 F. M. Lever and A. R. Powell, Chem. SOC., Spec. Puhl., 1959, References 18 13, 135. F. M. Lever and A. R. Powell, J. Chem. SOC.A, 1969, 1477. 1 M. J. Hudson, E. Rodriguez-Castellon, P. Sylvester, A. Jimenez-Lopez and P. Olivera-Pastor, Hydrometallurgy, 1990,24, 77. 2 G. Alberti, U. Costantino, S. Alluli and M. A. Massucci, J. Inorg.Nucl. Chem., 1975,37, 1779. 3 M. J. Hudson, E. Rodriguez-Castellon, P. Olivera-Pastor, A. Jimenez-Lopez, P. Maireles-Torres and P. Sylvester, Can, J. Chem., 1989,67,2095. 4 M. J. Hudson, E. Rodriguez-Castellon, P. Olivera-Pastor, A. Jimenez-Lopez, P. Maireles-Torres and P. Sylvester, Can. J. Chem., 1989,67,2095. 5 G. Alberti, M. Casciola and U. Costantino, J. Colloid Interface Sci.,1985, 107, 256. 6 G. Alberti, U. Costantino, F. Marmottini, R. Vivani and C. Valentini, in Recent Developments in Ion Exchange 2, ed. M. J. Hudson and P. A. Williams, Elsevier, London, 1990. 19 20 21 22 23 24 25 26 J. B. Raynor, B. G. Jeliazkowa, J. Chem. Soc., Dalton Trans., 1982,1185. B. Z. Wan and J. H. Lunsford, Inorg. Chim. Actcr., 1982,65, L29. J. R. Pladziewicz, T. J. Meyer, J. A. Broomhead and El. Taube, Inorg. Chem., 1973,12,639. D. J. Jones, J. Penfold, J. Tomkinson and J. Roziere, J. Mol. Struct., 1989, 197, 113. C. H. Huang, 0.Knop, D. A. Othen, F. W. D. Woodhams and R. A. Howe, J. Can. Chem., 1975,53,79. E. Tillmans, W. Gilbert and W. H. Baur, J. Solid Staie Chem., 1973,7, 69. G. G. J. Boswell and S. Soetono, J. Inorg. Nucl. Chem., 1981, 43, 1625. Joint Committee on Powder Diffraction Standards, Card Number 21-1172,1971. U. Costantino and A. Gasparoni, J. Chromatogr., 1970,51,289. J. E. Fergusson and J. L. Love, Inorg. Synth., 1972,13,208. 7 8 Paper 31068 18H; Received 15th November, 1993