J O U R N A L O F C H E M I S T R Y Materials Communication Observation of a thermally induced spin crossover in a CdPS3 intercalate Christian N. Field, Marie-Laure Boillot* and Rene� Cle�ment* L aboratoire de Chimie Inorganique, Ba�timent 420, URA420, Universite� de Paris-Sud, 91405 Orsay, France solution of I at 110 °C for 20 h in a Pyrex ampoule sealed under vacuum. The dark red solids thus formed (IIa and IIb, Magnetic susceptibility measurements of the layered intercalate Cd1-xPS3 [FeIII(SalEen)2]2x (x=0.14) reveal that respectively) were washed with methanol until the supernatant liquor became colourless.the trapped species undergo a gradual thermally induced spin crossover. The intercalates IIa and IIb were characterised by their powder X-ray diVractograms which showed quite sharp 00l reflections corresponding to an interlayer spacing of 14.8 A° (6.5 A ° for CdPS3 and 11.5 A ° for the pre-intercalate).The diVractogram of IIb also exhibited the broadened 001 reflection of pure CdPS3. The reflections given by IIa are somewhat There has been in recent years a renewed interest in the highsharper than those of IIb, hence IIa is better crystallized. spin (HS) < low-spin (LS) crossover phenomenon exhibited Elemental analyses indicate that IIa retains 3% chloride while by transition metal complexes.1–3 Besides progress in the IIb only retains a negligible amount.On the other hand, the understanding of thermally induced spin crossovers, lightpresence of re-formed CdPS3 in IIb results in a lower complex induced spin switching has opened up perspectives in optical content.The P, S and Fe contents in IIa suggest a formula information technology.4,5 There are several reasons to investiclose to Cd0.86PS3[FeIII(SalEen)2]0.28 for the intercalate,† in gate intercalation of spin-transition complexes in layered agreement with the stoichiometry usually observed for guest systems. The environment of the active complex (solvent,2,6 species of that size.The IR spectra of IIa and IIb were almost counterion,2,7,8 intermolecular interactions,9 packing defects1) identical. They showed numerous bands slightly broader and is well known to exert a dramatic influence on the transition. weaker than those of salt I, but at essentially the same Therefore the eVect of sterically restraining host layers and wavenumbers, as well as a strong signal assignable to the u(PS3) reduced dimensionality may considerably aVect the transition.stretching modes of CdPS3, split into three components at A layered system can also be viewed as a storage medium 560, 580 and 605 cm-1. This set of characterisation data is which could provide new pathways to act on the spin states therefore consistent with the presence in IIa and IIb by irradiation of the host lattice and subsequent host to guest of nearly close-packed [FeIII(SalEen)2] cations lying in the energy transfer.10 Spin crossover in two complexes synthesized interlayer galleries.in situ in Y-zeolite cages or layered silicates have been The temperature dependence of the magnetic susceptibility reported.11–13 No evidence for intermolecular interactions x (per mole of Fe) of both intercalates, measured with a was found, but the zeolite host lattice appears to stabilize con- Quantum Design SQUID magnetometer over the range figurations that diVer from those in the usual solid state. We 4–400 K, is shown in Fig. 1 as a plot of xT versus T . These report here the thermally induced spin crossover of a results clearly demonstrate that the FeIII centres undergo a cationic [FeIII(SalEen)2]+ complex intercalated in the layered spin crossover between low-spin (LS) S=1/2 and high-spin diamagnetic CdPS3 thiophosphate.(HS) S=5/2, centered at a half-conversion temperature T1/2= The [FeIII(SalEen)2]Cl compound I, where SalEen is the 255 K. The transition is relatively smooth.No hysteresis was monoanion of the condensation product of salicylaldehyde observed when a full temperature cycle was followed. Focussing and N-ethylethylenediamine, was synthesised according to the on IIa, the experimental xT value on the low temperature procedure used for the NO3, BPh4 and PF6 analogues, using plateau (1.23–1.40 cm3 mol-1 K) is significantly higher than FeCl3·6H2O.7 The composition of I was verified by elemental that expected for the LS state of FeIII (xT#0.50 cm3 mol-1 K, analysis and IR spectroscopy. taken from I in the LS state; see below; this is ascribed to The CdPS3 layered compound14 is known to react with residual HS molecules (28%).In contrast, the crossover is solutions of certain ionic salts G+X- to give intercalation complete on the high temperature side.Comparing the two compounds of a general formula Cd1-xPS3G2x·(solvent)y .15 In intercalates, the crossover is very slightly more abrupt in IIa these compounds, charge balance is maintained by the loss of (DT 80=180 K)‡ than in IIb (DT 80=200 K), the completeness one Cd2+ ion from the intralayer region for every two G+ of the crossover is larger in IIa (HS residue at low T for ions that are inserted in the interlayer region. In the present IIb#36%), but the crossover temperatures are very similar. case, no reaction occurred upon treating CdPS3 with a meth- These features are consistent with the better crystallinity of anolic solution of I.A two-step procedure was therefore IIa, and hence the presence of fewer defects.16 The incomemployed.(i) A pre-intercalate Cd1-xPS3(Me4N)2x was prepared by treating CdPS3 (typically 200 mg) with Me4NCl (1 g) in dry methanol (20 ml ) at 20 °C for 1 day. As already seen † Analytical data for the intercalates. IIa: C, 21.12; H, 2.68; N, 4.60; Fe, 3.96; Cd, 24.96; P, 7.79; S, 23.03; Cl, 3.38. Calc. for for cobaltocenium intercalated into CdPS3, the Cd2+ ions Cd0.86PS3[Fe(salEen)2]0.28[(Me4N) (CdCl3)]0.10: C, 20.89; H, 2.57; N, extracted from CdPS3 form a sparingly soluble (Me4N) (CdCl3) 4.53; Fe, 4.16; Cd, 28.98; P, 8.22; S, 25.45; Cl, 2.78%.IIb: C, 12.09; H, diamagnetic salt.15 Extensive washing by methanol allowed 1.53; N, 2.52; Fe, 2.50; Cd, 37.40; P, 10.11; S, 30.46; Cl, 0.22. Calc. for dissolution of this impurity but caused partial decomposition Cd0.94PS3[Fe(salEen)2]0.12[(Me4N) (CdCl3)]0.01: C, 11.11; H, 1.3; N, of the pre-intercalate into CdPS3.(ii ) Two samples of the 2.37; Fe, 2.32; Cd, 37.38; P, 10.7; S, 33.14; Cl, 0.36%. above pre-intercalate, one moderately washed with methanol, ‡ DT 80 represents the temperature interval over which the spin conversion varies from 10 to 90%. the other one thoroughly, were then treated with a methanolic J.Mater. Chem., 1998, 8(2), 283–284 283in order to draw a general conclusion, the present result tells us that intercalation could be used as worthwhile strategy to pack active species, especially because it brings additional degrees of freedom. The authors are grateful to the Royal Society for a postdoctoral fellowship (to C. N. F.) under the European Science Exchange programme and to Dr Eric Rivie`re for the SQUID measurements.References 1 E.Ko� nig, G. Ritter and S. K. Kulshreshtha, Chem. Rev., 1985, 85, 219. 2 P.Gu� tlich, A. Hauser and H. Spiering, Angew. Chem., Int., Ed. Engl., 1994, 33, 2024. 3 O. Kahn,MolecularMagnetism, VCH Publishers, New York, 1993, ch. 4. Fig. 1 xT vs. T plot of [FeIII(SalEen)2]Cl ($) and of the CdPS3 4 P.Gu� tlich and A. Hauser, Coord.Chem. Rev., 1990, 97, 1. intercalates IIa (#) and IIb (6). x is expressed per mole of Fe in 5 M. L. Boillot, C. Roux, J. P. Audie`re, A. Dausse and all cases. J. Zarembowitch, Inorg. Chem., 1996, 35, 3975. 6 M. Sorai, J. Ensling, K. M. Hasselbach and P. Gu� tlich, Chem. Phys., 1977, 20, 1997. pleteness of the crossover at low temperature might also be 7 M.S. Haddad, M. W. Lynch, W. D. Federer and ascribed to Fe3+ impurities resulting from a slight decompo- D. N. Hendrickson, Inorg. Chem., 1981, 20, 123. sition of the complex, but the analytical data for Fe, C and N 8 A. M. Greenaway, C. J. O’Connor, A. Schrock and E. Sinn, Inorg. show good agreement with the expected composition Chem., 1979, 18, 2692.Fe(SalEen)2+. 9 J.-P. Martin, J. Zarembowitch A. Dworkin, The magnetic behaviour of the intercalated FeIII complex J. G. Haasnoot and F. Varret, Inorg. Chem., 1994, 33, 6325. 10 E. Lifshitz, R. Cle�ment, L. C. Yu-Hallada and A. H. Francis, should of course be compared to that of the starting chloride J. Phys. Chem. Solids, 1991, 52, 1081. I. The temperature dependence of xT measured for I is also 11 K.Mizuno and J. H. Lunsford, Inorg. Chem., 1983, 22, 3484. shown in Fig. 1. This compound undergoes a smooth but 12 Y. Umemura, Y. Minai, N. Koga and T. Tominaga, J. Chem. Soc., complete spin crossover centered at T1/2=320 K, more abrupt Chem. Commun., 1994, 893. than the intercalates (DT 80=148 K). 13 M. Nakano, S. Okuno, G. E. Matsubayashi, W.Mori and From these results, one might conclude that intercalation of M. Katada,Mol. Cryst. L iq. Cryst., 1996, 286, 83. 14 W. Klingen, R. Ott and H. Hahn, Z. Anorg. Allg. Chem., 1973, the FeIII complex does not suppress its spin crossover, but 396, 271. exerts a damping eVect. However the eVect of intercalation 15 R. Cle�ment, O. Garnier and J. Jegoudez, Inorg. Chem., 1986, 25, appears more positive if the intercalated complex is compared 1404. to other salts. Thus, [FeIII(SalEen)2]PF6 has been reported to 16 M. S. Haddad, W. D. Federer, M. W. Lynch and show a gradual spin crossover centered at T1/2=140 K only D. N. Hendrickson, Inorg. Chem., 1981, 20, 131. and the nitrate salt exhibits a very incomplete transition.7 Although other complexes and intercalates should be studied Communication 7/07887K; Received 3rd November, 1997 284 J. Mater. Chem., 1998, 8(2), 283&ndash