Layered metal organosulfides: Hydrothermal synthesis, structure and magnetic behaviour of the spin-canted magnet Co(1,2-(O2C)(S)C6H4)

Article abstract of DOI:10.1039/b111339a

Reaction of cobalt(II) salts with the thiosalicylate dianion under hydrothermal conditions yields green lamellar Co((O₂C)(S)C₆H₄) which displays canted antiferromagnetism.

Full text of DOI:10.1039/b111339a

Layered metal organosulfides: hydrothermal synthesis, structure and
magnetic behaviour of the spin-canted magnet Co(1,2-(O2C)(S)C6H4)
Dale Cave,a José-Miguel Gascon,a Andrew D. Bond,a Simon J. Teatb and Paul T. Wood*a
a University Chemical Laboratory, Lensfield Rd, Cambridge, UK CB2 1EW
b CLRC Daresbury Laboratory, Warrington, UK WA4 4AD. E-mail: ptw22@cam.ac.uk
Received (in Cambridge, UK) 12th December 2001, Accepted 5th April 2002
First published as an Advance Article on the web 19th April 2002
Reaction of cobalt(II) salts with the thiosalicylate dianion
under hydrothermal conditions yields green lamellar
Co((O2C)(S)C6H4) which displays canted antiferromagnet-
ism.
unit and therefore there is only one metal environment. The
cobalt(II) ion is trigonal bipyramidal with an S2O3 donor set
(Fig. 1) coming from four different ligands. Sulfur ligands
occupy one of the axial and one of the equatorial sites whilst the
O and S donor atoms of the one chelating ligand occupy apical
(S) and equatorial (O) sites forming a six-membered ring. Each
ligand is coordinated to four metal atoms with Co–S–Co and
Co–O–Co angles of 97.38(3) and 101.90(8)°, respectively. The
carboxylate bridge has a syn–syn configuration but with a
dihedral angle of 49.8° between the Co1b–O1–C7 and Co1c–
O2–C7 planes. The metal ions form the core of a ligand bilayer
(Fig. 2) with the polar groups pointing inwards. The non-polar
groups form an essentially flat surface hence the interlayer
interactions will be relatively weak with no stabilising inter-
digitation as we have observed previously.6d It is therefore not
surprising that the crystals form as very thin plates with
hydrophobic surfaces.
Recrystallisation of metal chalcogenides was among the earliest
uses of superheated solution chemistry.1 Since then there has
been much interest in the preparation of new and more complex
chalcogenides using superheated methanol,2,3a water,2,3b am-
monia4 and ethylenediamine.5 In one case this has produced a
cluster containing the methyl tellurolate (CH3Te2) anion,3a but
there has been no successful synthesis of such a material starting
from organochalcogenide anions. This is in contrast to the
rapidly increasing number of solvothermally-prepared coor-
dination solids based on oxygen6 or nitrogen7 donor ligands.
We report here what we believe to be the first example of the
hydrothermal preparation of a metal-organochalcogenide ex-
tended lattice.
Reaction of CoCl2·6H2O (101 mg, 0.424 mmol) with
1,2-(HO2C)(HS)C6H4 (135 mg, 0.858 mmol) and KOH (1.68
ml of a 1 M solution in H2O, 1.68 mmol) in water (10 ml) at 200
°C for 15 h followed by cooling to room temperature over 4 h
yields very fine dark green plates in 73% yield. Larger crystals
can be grown by the addition of a mineraliser such as NaBr (up
to 5 g per reaction) and an identical product can also be obtained
from the reaction of Co(OH)2 with ligand salt. The phase purity
of the bulk material was confirmed by comparison of its powder
diffraction pattern with that calculated from the single crystal
study. X-Ray structural analysis† shows the compound to be
monoclinic with an asymmetric unit consisting of one formula
Previous studies on the chemistry of the Co(II)/thiosalicylate
8
system under ambient conditions yielded Co(LH)2(H2O)2 or
the Co(III)-containing ion9 CoL332, hence the high temperature
conditions have led to new chemistry. The small difference in
ligand field stabilisation energies for different high spin Co(II
)
geometries means that many compounds may be prepared with
this ion which do not exist for other transition metals. Reactions
of other divalent ions with thiosalicylate salts do not yield any
isostructural products, even in the case of Zn(II), which is the ion
most likely to mimic Co(II) chemistry.
There are three types of ligand bridging which might give rise
to magnetic superexchange interactions. In order of strength
these are the Co–S–Co, Co–O–Co and Co–OCO–Co pathways.
The sulfur-bridged superexchange lattice consists of isolated
Co–S–Co–S chains parallel to the crystallographic c axis (Fig.
3). Addition of the single atom oxygen bridges increases the
connectivity to give a rectangular net parallel to the bc plane.
Inclusion of the magnetically less important OCO bridges gives
a triangular sheet structure. Three-dimensional ordering is only
possible at temperatures when the dipolar interaction between
layers becomes important. Magnetisation measurements‡ show
a steady increase in susceptibility with decreasing temperatures
before a sharp anomaly centred at 9 K characteristic of long-
Fig. 1 The metal–ligand building blocks showing (a) all the ligands
coordinated to one metal and (b) all the metals coordinated to one ligand.
Selected bond distances (Å) and angles (°): Co1–O1 1.991(2), Co1–O1B
2.269(2), Co1–O2C 1.989(2), Co1–S1 2.3515(7), Co1–S1A 2.2996(7); O1–
Co1–O1B 78.10(8), O1–Co1-O2C 102.66(8), O1–Co1–S1 92.05, O1–Co1–
S1A 117.15(6), O1B–Co1–O2C 79.93(7), O1B–Co1–S1 169.82(5), O1B–
Co1–S1A 83.69(5), O2C–Co1–S1 105.06(6) O2C–Co1–S1A 132.41(6),
S1–Co1–S1A 98.77(3), Co1–S1–Co1A 97.38(3).
Fig. 2 Packing diagram parallel to [010].
1050
CHEM. COMMUN., 2002, 1050–1051
This journal is © The Royal Society of Chemistry 2002

term and therefore lacks an orbital contribution to its magnetic
moment and is also much less likely to undergo a significant
degree of canting.
Application of the Curie–Weiss law to the high temperature
( > 100 K) data gives a remarkably large value of the Weiss
constant (q = 2585(3) K) which is consistent with the low
value of m at room temperature compared to the spin only value
of 3.87 mB. This is indicative of strong antiferromagnetic
coupling as the lack of first-order angular momentum (vide
supra) for the singlet ground term rules out the possibility that
such a large deviation is due to single ion anisotropy.
Comparison with typical values we have observed for q ≈ 250
K in Co–O–Co bridged compounds with similar bridging angles
shows that this strong interaction must be mediated by sulfur
bridges. The greater covalency of Co–S bond vs. the Co–O bond
leads to greater delocalisation of spin density and therefore
stronger superexchange.10 Since the Co–S superexchange
lattice is comprised of isolated chains we should expect to see a
broad maximum in the susceptibility vs. temperature graph
coinciding with short-range ordering. Mean Field Theory is not
a particularly good model for low dimensional systems but can
give us a qualitative estimate of the exchange coupling; when it
is combined with an expression11 for the temperature at which
the susceptibility maximum should occur we obtain the result
Tmax = 1.9¡q¡. The susceptibility maximum will therefore occur
at a much higher temperature than could be measured.
We wish to thank EPSRC and the EU ERASMUS program
for funding.
Fig. 3 The one-atom superexchange pathways within a single layer.
range ordering (Fig. 4). The magnetisation does not begin to
saturate before 5 K but it is clear from its small magnitude that
this is not a simple ferromagnetic state. As there is only one
cobalt environment we may eliminate ferrimagnetism as a
possibility and the source of the spontaneous magnetisation
must therefore be a small degree of canting in an essentially
antiferromagnetic system. This was confirmed by the variation
of magnetisation with temperature for the ordered phase in
different applied fields. The sample magnetisation is higher in
lower measuring fields, which is characteristic of spin canting.
Hysteresis measurements (Fig. 5) show a small coercive field of
45 G and that the sample magnetisation is not saturated at 5 T.
The ferromagnetic interlayer interaction is a dipolar one across
a relatively large gap of approximately 13 Å and is therefore
small and only becomes significant at low temperatures. Canted
antiferromagnetism is usually due to single ion anisotropy
therefore high spin, octahedral Co(II) with large spin–orbit
coupling due to its 4T1g ground term is a good candidate for this
effect. Trigonal bipyramidal high spin Co(II) has a 4A1 ground
Notes and references
†
Crystal data: C7H4O2SCo, M = 211.09, monoclinic, space group C2/c,
a = 27.890(2), b = 7.6926(6), c = 6.7497(5) Å , b = 103.895(2)°, V =
1405.8(2) Å3, Z = 8, Dc = 1.995 g cm23, m = 2.669 mm21, l = 0.6904
Å (Daresbury Station 9.8), crystal dimensions 0.12 3 0.10 3 0.01 mm.
Total number of observed [F2 > 2s(F2)] and independent reflections 1865,
1505 (Rint = 0.0217). Full-matrix least-squares on F2 gives R1 = 0.0385
and wR2 = 0.0974.
CCDC reference number 176424. See http://www.rsc.org/suppdata/cc/
b1/b111339a/ for crystallographic data in CIF or other electronic format.
‡ Magnetisation measurements: Data were recorded on a Quantum design
MPMS-XL magnetometer. Temperature scans were recorded between 4.5
and 300 K using a cooling and measuring field of 500 G. Magnetisation
hysteresis was studied between 5 T at 5 K. Sample diamagnetism was
corrected for by employing Pascal’s constants.10
1 A. Rabenau, Angew. Chem., Int. Ed. Engl., 1985, 24, 1026.
2 W. S. Sheldrick, J. Chem. Soc., Dalton Trans., 2000, 3041.
3 (a) B. K. Das and M. G. Kanatzidis, Inorg. Chem., 1995, 34, 1011; (b)
J. A. Hanko and M. G. Kanatzidis, Angew. Chem., Int. Ed., 1998, 37,
342.
Fig. 4 The thermal evolution of the magnetic moment.
4 P. T. Wood, G. L. Shimek and J. W. Kolis, Chem. Mater., 1996, 8,
721.
5 P. T. Wood, W. T. Pennington and J. W. Kolis, J. Am. Chem. Soc., 1992,
114, 9233.
6 (a) S. O. H. Gutschke, A. M. Z. Slawin and P. T. Wood, Chem.
Commun., 1995, 2197; (b) A. Distler and S. C. Sevov, Chem. Commun.,
1998, 959; (c) C. Livage, C. Egger, M. Nogues and G. Ferey, J. Mater.
Chem., 1998, 8, 2743; (d) S. O. H. Gutschke, D. J. Price, A. K. Powell
and P. T. Wood, Angew. Chem., Int. Ed., 2001, 40, 1920.
7 (a) P. T. Wood, W. T. Pennington, J. W. Kolis, B. Wu and C. J.
O’Connor, Inorg. Chem., 1993, 32, 129; (b) O. M. Yaghi and H. L. Li,
J. Am. Chem. Soc., 1995, 117, 10401; (c) D. Hagrman, P. J. Hagrman
and J. Zubieta, Angew. Chem., Int. Ed., 1999, 38, 3165.
8
R. R. Jha, Asian J. Chem., 2001, 13, 1048.
9 N. Duran, W. Clegg, L. Cucurull-Sanchez, R. A. Coxall, H. R. Jimenez,
J.-M. Moratal, F. Lloret and P. Gonzalez-Duarte, Inorg. Chem., 2000,
39, 4821.
10 O. Kahn, Molecular Magnetism, Wiley-VCH, Weinheim, 1993.
11 L. J. de Jongh and A. R. Miedema, Adv. Phys., 1974, 23, 1.
Fig. 5 Expansion of the magnetic hysteresis at 4.5 K in the low field region.
Insert: field sweep between 5 T.
CHEM. COMMUN., 2002, 1050–1051
1051