Magnetic Interactions in Layered Nickel Alkanethiolates - American

277-8581, Japan, and. Surface Physics DiVision, Saha Institute of Nuclear Physics, 1/AF, Bidhannagar, Kolkata 700064, India. ReceiVed: NoVember 13...
0 downloads 0 Views 152KB Size
1868

2007, 111, 1868-1870 Published on Web 01/18/2007

Magnetic Interactions in Layered Nickel Alkanethiolates Neena Susan John,† G. U. Kulkarni,*,†,‡ Ayan Datta,§ Swapan K. Pati,*,‡,§ F. Komori,| G. Kavitha,† Chandrabhas Narayana,† and M. K. Sanyal⊥ Chemistry and Physics of Materials Unit, DST Unit on Nanoscience, and Theoretical Sciences Unit, Jawaharlal Nehru Center for AdVanced Scientific Research, Jakkur P. O, Bangalore 560064, India, Institute for Solid State Physics, UniVersity of Tokyo Kashiwa-shi, Chiba 277-8581, Japan, and Surface Physics DiVision, Saha Institute of Nuclear Physics, 1/AF, Bidhannagar, Kolkata 700064, India ReceiVed: NoVember 13, 2006

Layered nickel alkanethiolates are found to exhibit antiferromagnetic coupling along the S-Ni-S chain with the ordering temperature increasing linearly with decreasing alkyl chain length. Theoretical calculations are performed to study the origin of magnetic moments and interactions in these layered systems.

Metal alkanethiolates, both cyclic and layered, have been an interesting topic to chemists due to the versatile bonding patterns the thiol groups exhibit in these structures.1 They are also considered as model systems for the investigation of sulfidic active centers in biomolecules.2 Although extensive literature is available on the structure of layered and cyclic metal thiolates,1 there are very few studies on the properties and their correlation with the structure. Thus, the liquid crystalline nature of alkanethiolates of Ag, Pd, Ni, and Cu is well studied, but the nonavailability of single crystals has made the structureproperty correlation a difficult task.3,4 We considered it interesting to investigate magnetic interactions in such low-dimensional systems exhibiting mesoscopic ordering. We sought to carry out a combined experimental and theoretical study of nickel alkanethiolates of different interlayer spacings, conventionally thought to be diamagnetic,4 at low enough temperatures where new magnetic interactions possibly emerge. Indeed, layered compounds of Ni such as azides,5 hydroxides,6 framework sulfates,7 and supramolecular complexes8 are reported to exhibit unusual magnetic properties at low temperatures. In this communication, we report for the first time the structural transition giving rise to superexchange phenomena and chainlength-dependent magnetic ordering. Such unique behavior is realized due to the presence of various energy scales of interactions prevailing in the system. Nickel alkanethiolates were prepared by adding an equimolar solution of triethylamine and n-alkanethiol (purity >98%) in absolute ethanol to a solution of nickel chloride in the molar ratio of 2:1. Estimates of the chemical composition from elemental analysis suggested the molecular formulas of the * To whom correspondence should be addressed. E-mail: kulkarni@ jncasr.ac.in (G.U.K.); [email protected] (S.K.P.). † Chemistry and Physics of Materials Unit, Jawaharlal Nehru Center for Advanced Scientific Research. ‡ DST Unit on Nanoscience, Jawaharlal Nehru Center for Advanced Scientific Research. § Theoretical Sciences Unit, Jawaharlal Nehru Center for Advanced Scientific Research. | University of Tokyo Kashiwa-shi. ⊥ Saha Institute of Nuclear Physics.

10.1021/jp0675072 CCC: $37.00

Figure 1. (a) XRD pattern of nickel alkanethiolates with different chain lengths. n is the number of carbon atoms present in the alkyl chain. A schematic of the bilayer is shown in the inset. Hollow circles represent nickel, and dark circles, sulfur. (b) A linear variation of the d(001) spacing with the chain length. The d(001) spacing is a measure of the bilayer thickness. The other axis gives the variation of the temperature corresponding to the maxima of the susceptibility data with the number of carbons in the bilayer.

monomers as Ni(SR)2, with R being the alkyl group (see the Supporting Information). X-ray diffraction patterns from the powder samples shown in Figure 1 exhibit a series of (00k) reflections typical of a lamellar bilayer structure.3 We observe from Figure 1b that the d(001) spacing varies linearly from 13.53 to 43.5 Å as the alkyl chain length increases from 3 to 15 methylene units with a slope of 1.279 Å per methylene unit (closely matching the value of 1.24 Å of an all-trans conformer) and an intercept of 5.39 Å originating from the inorganic backbone. The proposed bilayer structure is depicted in the schematic in Figure 1, with Ni in a square planar coordination.9 Thus, there are two contributions to the bilayer thickness, twice the thickness of the inorganic slab consisting of Ni and S and twice the alkanethiol chain length. Infrared spectra showed the symmetric and antisymmetric methylene stretches at 2849 and 2918.5 cm-1, respectively, confirming an all-trans conformation of the alkyl chain.10 UV-visible spectra at room temperature show signatures corresponding to ligand-to-metal charge transfer transitions in these complexes. The absence of prominent d-d © 2007 American Chemical Society

Letters

J. Phys. Chem. C, Vol. 111, No. 5, 2007 1869

Figure 3. Frontier molecular orbitals for the cluster, Ni2(SR)2, R ) butyl. As can be seen, charge transfer occurs between the Ni 3d orbitals through 3p orbitals of sulfurs. Atom basin: Ni, blue (cannot be seen due to the large contribution from its orbitals to the MOs); S, yellow.

Figure 2. FC (circles) and ZFC (stars) as a function of temperature for nickel butanethiolate at the 100 Oe field. The solid line is the theoretical fit to the ZFC data (see text). Inset A: magnetization values at 11 K for three magnetic field strengths (diamonds, experimental) and calculated M from Brillouin functions for S ) 1 and S ) 2. The low-temperature part of the ZFC and FC are shown in inset B for clarity. The solid line shows the theoretical fit for the ZFC curve. Note the upward rise of the ZFC data below 4 K.

transitions is suggestive of a square planar geometry for Ni(II)11 (see Supporting Information for IR and UV-vis spectra). Magnetic measurements on various nickel thiolates were carried out using a SQUID magnetometer down to 2 K. Fieldcooled (FC) susceptibility (χ) data of all thiolates showed a rise in χ as the temperature is lowered. A plot of the temperature corresponding to the maxima of the susceptibility data with increasing chain length in thiolates is shown in Figure 1b. As T is lowered below room temperature, the susceptibility increases from the diamagnetic value, for all of the thiolates, showing a maximum at some T value which increases with reduction of bilayer thickness. For a detailed and complete understanding, we consider the smallest chain system, nickel butane thiolate, as it may be a model system to study the possibility of interesting magnetic properties arising from intralayer coupling as well as interlayer interactions. The temperature variations of both field-cooled (FC) and zerofield-cooled (ZFC) magnetic susceptibility data for nickel butanethiolate at 100 Oe are shown in Figure 2. Both FC and ZFC data exhibit an upward rise in χ as the temperature is lowered below 30 K. The system orders antiferromagnetically with a short-range order, with χ showing an exponential drop below ∼11 K, as seen in the ZFC curve. The short-range order is identified with the point where the FC data deviate from the ZFC data. Short-range order is generally associated with low-dimensional antiferromagnetic systems.12 A dimer model corresponding to two Ni2+ spin-1 does not fit the data below 20 K. Since the lowest excitation in a spin-1 chain consists of magnons, we use a parabolic dispersion to obtain χ ) A exp(-∆/KBT)/(KBT)τ, where ∆ is the gap, KB and T are the Boltzmann constant and temperature, respectively. The exponent τ relates to the extended nature of the antiferromagnetic interactions along the NiS2 chain at low temperatures. By fitting the experimental data with the above form of χ, we obtained the values ∆ ) 10.5 K and τ ) 0.45 (see the solid line in Figure 2). Since an S ) 1 system corresponds to a Haldane chain, the exchange coupling can be readily obtained (∆ ) 0.41 J).13 While the exponent τ is exactly 1 for a dimer, an infinitely extended structure should give a value close to 0.5.14 Clearly, the estimated value of τ gives an idea of the optimal length scales associated with the exchange processes in this system. The magnetization values at the

transition temperature (T ) 11 K) are shown in inset A for three H-field strengths. It may be noteworthy that, even at H ) 0.1 T, the spins are not completely polarized and the associated Brillouin function corresponds to a chain of particles with spin little smaller than unity. This suggests that the geometry associated with Ni2+ ions at lower temperature is just distorted square planar and not perfectly tetrahedral (spin-1). Note that the antiferromagnetic ordering temperature for longer thiolates was found to decrease with an increase in chain length, as given in Figure 1b. Longer chains compelled by their self-assembling nature15 seem to restrain from antiferromagnetic coupling along the NiS2 chain. Preliminary calculations on bilayer motifs have shown that there is a slight elongation in Ni-Ni distance with an increase in the alkyl chain due to the inductive effect of the R groups. Interestingly, at lower temperatures (below 4 K), we notice from inset B of Figure 2 that the susceptibility starts rising as T f 0, signaling a magnetic polarization. Using Gaussian 03 optimization at the DFT-based B3LYP/ LANL2DZ level,16 we have computed a few molecular orbitals for an optimized Ni2[S(CH2)3CH3]2 motif, as shown in Figure 3. As can be seen, the lowest unoccupied molecular orbital (LUMO) and the immediate higher lying MOs (LUMO+1, LUMO+2) correspond to the substantial charge transfer between Ni2+ ions leading to strong superexchange through the bridging S ions. For a better understanding of the structural changes leading to the new magnetic interactions observed in Figure 2, we performed low-temperature Raman measurements. Raman signals were collected at room temperature and 30 K on a custombuilt Raman spectrometer using a Jobin-Yovn 550 Triax monochromator and Nd:YAG (λ ) 532 nm) as the excitation source. Raman spectra at 300 K show δ(Eg) (S-Ni-S), δ(T2g) (SNi-S), ν(A1g) (Ni-S), and ν(T2g) (Ni-S-C) modes along with a librational mode around 112 cm-1, C-C-S deformation (246 cm-1), and C-C-C deformations (333 and 361 cm-1) in the range 100-400 cm-1, as reported earlier for a square planar structure of Ni2+.17,18 Figure 4 gives a comparison of Raman spectra at 30 and 300 K for the Ni-S vibrational region (see the Supporting Information for the assignments of the vibrational modes). In the case of 300 K, the Ni-S structure is square planar and hence the Raman spectra show that most of the modes associated with Ni-S are degenerate due to the high symmetry of the Ni-S unit. Upon cooling to 30 K, we observe an increase in the number of Raman modes. It is important to note that δ(E) and δ(T2) in the region 120-175 cm-1 show a marked change along with softening (decrease in frequency) of the modes. The increase in the number of Raman modes at low temperature could be due to lifting of the degeneracy of the corresponding modes of the asymmetric Ni-S unit.17 The presence of modes related to both the square planar and the

1870 J. Phys. Chem. C, Vol. 111, No. 5, 2007

Letters impurity moment contributing to the rise in χ at very low temperatures. To our knowledge, this is the first example of a bilayer metal alkanethiolate system exhibiting signatures of exchange processes of entirely different nature in a small temperature range. The present study offers new avenues for studying interactions in mesoscopic magnetic systems with structure-property relationships associated with mixed dimensionalities.

Figure 4. Raman spectra of nickel butanethiolate for the Ni-S vibrational region, 100-400 cm-1 window: (a) 300 K; (b) 30 K. The green curves are Lorentzian fits to the peaks.

Acknowledgment. The authors acknowledge Prof. C. N. R. Rao for useful discussions and encouragement. S.K.P. snd G.U.K. thank DST for financial support. G.U.K. thanks the India-Japan cooperative science program for a visiting fellowship. N.S.J. and A.D. thank CSIR for research fellowships. Supporting Information Available: Elemental analysis of nickel alkanethiolates, IR data, UV-vis spectra, low-temperature Raman measurement details, Raman spectra at 300 and 30 K in the Ni-S vibrational region, assignments of the vibrational modes, and complete author list for ref 16. This material is available free of charge via the Internet at http://pubs.acs.org. References and Notes

Figure 5. Ground state optimized structures for Ni4(SR)10, R ) butyl, in its room-temperature (square planar) and low-temperature (distorted square planar) geometries. Note that the calculation is performed on a finite sized cluster and terminal -SR groups are included to avoid the end-group effect.

distorted geometry at 30 K is clearly an indication of out-ofplane movement of Ni. The above observations imply that the magnetic moment arises due to a slight distortion in the NiS4 square planar geometry. Similar distortions have been observed in the case of tiara structures of nickel alkanethiolates.19 The square planar arrangement of Ni(SR)4 corresponds to the ground state geometry of the diamagnetic spin structure (S ) 0), while the stable geometry for the paramagnetic state is associated with a distorted arrangement of the Ni(SR)4 (1 > S > 0) unit (see Figure 5). In the nickel thiolates, the Ni2+ ions are bridged via S ions along the backbone and the backbones are held together by hydrocarbon spacers. At relatively higher temperatures (∼30 K), the exchange processes are along the -Ni-S-Nibackbone and are thus antiferromagnetic in nature. This exchange path dominates down to the ordering temperature (see Figure 1b). However, at even lower temperatures, the system may go through a dimensional crossover, whereby the exchange between the backbones through hydrocarbon spacers becomes dominant. Magnetic coupling mediated through space is most often considered to be ferromagnetic due to dipolar interactions,20 as can be seen in inset B of Figure 2, below 4 K. Such a feature may arise as a result of increased proximity of the neighboring layers due to the out-of-plane displacement of the alternate butyl chains in the distorted geometry compared to the square planar structure (Figure 5). This is expected to facilitate “through-space” ferromagnetic coupling between the layers. However, one may also consider the possibility of

(1) (a) Blower, P. J.; Dilworth, J. R. Coord. Chem. ReV. 1987, 76, 121. (b) Dance, I. G. Polyhedron 1986, 5, 1037. (2) (a) Kolodziej, A. F. Prog. Inorg. Chem. 1994, 41, 493. (b) Volbeda, A.; Charon, M. H.; Piras, C.; Hatchikian, E. C.; Frey, M.; Fontecilla-Camps, J. C. Nature 1995, 373, 580. (3) (a) Dance, I. G.; Fisher, K. J.; Banda, R. M. H.; Scudder, M. L. Inorg. Chem. 1991, 30, 183. (b) Thomas, P. J.; Lavanya, A.; Sabareesh, V.; Kulkarni, G. U. Proc.sIndian Acad. Sci., Chem. Sci. 2001, 113, 611. (c) Sandhyarani, N.; Pradeep, T. J. Mater. Chem. 2001, 11, 1294. (d) Lee, S. J.; Han, S. W.; Choi, H. J.; Kim, K. J. Phys. Chem. B 2002, 106, 2892. (4) (a) Jensen, K. A. Z. Anorg. Chem. 1944, 252, 227. (b) Higgs, T. C.; Ji, D.; Czernuszewicz, R. S.; Spartalian, K.; O’Connor, C. J.; Seip, C.; Carrano, C. J. J. Chem. Soc., Dalton Trans. 1999, 807. (5) Monfort, M.; Resino, I.; Ribas, J.; Evans, H. S. Angew. Chem., Int. Ed. 2000, 39, 191. (6) Taibi, M.; Ammar, S.; Jouini, N.; Fievet, F.; Molinie, P.; Drillon, M. J. Mater. Chem. 2002, 12, 3238. (7) Behera, J. N.; Gopalkrishnan, K. V.; Rao, C. N. R. Inorg. Chem. 2004, 43, 2636. (8) Waldmann, O.; Hassmann, J.; Mu¨ller, P.; Volkmer, D.; Schubert, U. S.; Lehn, J.-M. Phys. ReV. B 1998, 58, 3277. (9) Alemany, P.; Hoffmann, R. J. Am. Chem. Soc. 1993, 115, 8290. (10) Parikh, A. N.; Gillmor, S. D.; Beers, J. D.; Beardmore, K. M.; Cutts, R. W.; Swanson, B. I. J. Phys. Chem. B 1999, 103, 2850. (11) Storr, A.; Summers, D. A.; Thompson, R. C. Can. J. Chem. 1998, 76, 1130. (12) Pati, S. K.; Ramasesha, S.; Sen, D. In Magnetism: molecules to materials IV; Miller, J. S., Drillon, M., Eds.; Wiley-VCH: Weinheim, Germany, 2003; p 199. (13) White, S. R. Phys. ReV. Lett. 1992, 69, 2863. (14) Troyer, M.; Tsunetsugu, H.; Wurtz, D. Phys. ReV. B 1994, 50, 13515. (15) Love, J. C.; Estroff, L. A.; Kriebel, J. K.; Nuzzo, R. G.; Whitesides, G. M. Chem. ReV. 2005, 105, 1103. (16) Frisch, M. J.; et al. Gaussian 03, revision B05; Gaussian, Inc.: Pittsburgh PA, 2003. (17) Huang, Y. H.; Moura, I.; Moura, J. J. G.; LeGall, J.; Park, J. B.; Adams, M. W. W.; Johnson, M. K. Inorg. Chem. 1993, 32, 406. (18) Fijolek, H. G.; Grohal, J. R.; Sample, J. L.; Natan, M. J. Inorg. Chem. 1997, 36, 622. (19) Datta, A.; John, N. S.; Kulkarni, G. U.; Pati, S. K. J. Phys. Chem A 2005, 109, 11647. (20) Drillon, M.; Panissod, P. J. Magn. Magn. Mater. 1998, 188, 93.