Preparation and characterization of the disjoint diradical 4,4′-bis(1,2,3,5-dithiadiazolyl) [S2N2C-CN2S2] and its iodine charge transfer salt [S2N2C-CN2S2][I]

Article abstract of DOI:10.1021/ja952144x

Condensation of oxamidrazone with sulfur dichloride in acetonitrile affords 4,4′-bis(1,2,3,5-dithiadiazolium) dichloride in moderate yield. Reduction of this salt with triphenylantimony yields the diradical 4,4′-bis(1,2,3,5-dithiadiazolyl) [S₂N₂C-CN₂S₂], which has been isolated and characterized in the solid state as its dimer [S₂N₂C-CN₂S₂]₂. The diradical is disjoint, and ab initio molecular orbital methods confirm a very small energy gap (<0.5 kcal/mol) between the triplet and diradical singlet states, regardless of the torsion angle about the central C-C bond. In accord with these theoretical predictions the ESR spectrum of the diradical consists (in CHCl₃ at 273 K) of a simple five-line pattern (a_N = 0.50 mT, g = 2.011), i.e., there is no observable exchange coupling between the two centers. In the solid state, the dimer [S₂N₂C-CN₂S₂]₂ forms a slipped stack structure, with a mean intradimer S-S distance of 3.078 ? and mean interdimer S- - -S contact of 3.761 ?. Cosublimation of the diradical with iodine produces the charge-transfer salt [S₂N₂C-CN₂S₂][I], orthorhombic space group Ccmm, a = 11.909(3) ?, b = 3.271(2) ?, c = 19.860(6) ?, Z = 4 (at 293 K). In this structure the heterocyclic rings form perfectly superimposed and evenly spaced stacks along the y direction, with channels of disordered iodines. The iodine-doped material is metallic at ambient temperatures, with a single-crystal conductivity of 460 S cm^-1 at 300 K; variable temperature conductivity and magnetic measurements reveal a phase transition near 270 K, with the onset of semiconducting behavior. Transport data for the neutral and doped materials are discussed in the light of Extended Hückel band calculations.

Full text of DOI:10.1021/ja952144x

330
J. Am. Chem. Soc. 1996, 118, 330-338
Preparation and Characterization of the Disjoint Diradical
4,4′-Bis(1,2,3,5-dithiadiazolyl) [S2N2C-CN2S2] and Its Iodine
Charge Transfer Salt [S2N2C-CN2S2][I]
C. D. Bryan,1a A. W. Cordes,*,1a J. D. Goddard,1b R. C. Haddon,*,1c R. G. Hicks,1b
C. D. MacKinnon,1b R. C. Mawhinney,1b R. T. Oakley,*,1b T. T. M. Palstra,1c and
A. S. Perel1c
Contribution from the Department of Chemistry and Biochemistry, UniVersity of Arkansas,
FayetteVille, Arkansas 72701, Guelph-Waterloo Centre for Graduate Work in Chemistry,
Department of Chemistry and Biochemistry, UniVersity of Guelph, Guelph, Ontario N1G 2W1,
Canada, and AT&T Bell Laboratories, 600 Mountain AVenue, Murray Hill, New Jersey, 07974
ReceiVed June 29, 1995. ReVised Manuscript ReceiVed NoVember 7, 1995X
Abstract: Condensation of oxamidrazone with sulfur dichloride in acetonitrile affords 4,4′-bis(1,2,3,5-dithiadiazolium)
dichloride in moderate yield. Reduction of this salt with triphenylantimony yields the diradical 4,4′-bis(1,2,3,5-
dithiadiazolyl) [S2N2C-CN2S2], which has been isolated and characterized in the solid state as its dimer [S2N2C-
CN2S2]2. The diradical is disjoint, and ab initio molecular orbital methods confirm a very small energy gap (<0.5
kcal/mol) between the triplet and diradical singlet states, regardless of the torsion angle about the central C-C bond.
In accord with these theoretical predictions the ESR spectrum of the diradical consists (in CHCl3 at 273 K) of a
simple five-line pattern (aN ) 0.50 mT, g ) 2.011), i.e., there is no observable exchange coupling between the two
centers. In the solid state, the dimer [S2N2C-CN2S2]2 forms a slipped stack structure, with a mean intradimer S-S
distance of 3.078 Å and mean interdimer S- - -S contact of 3.761 Å. Cosublimation of the diradical with iodine
produces the charge-transfer salt [S2N2C-CN2S2][I], orthorhombic space group Ccmm, a ) 11.909(3) Å, b ) 3.271(2)
Å, c ) 19.860(6) Å, Z ) 4 (at 293 K). In this structure the heterocyclic rings form perfectly superimposed and
evenly spaced stacks along the y direction, with channels of disordered iodines. The iodine-doped material is metallic
at ambient temperatures, with a single-crystal conductivity of 460 S cm-1 at 300 K; variable temperature conductivity
and magnetic measurements reveal a phase transition near 270 K, with the onset of semiconducting behavior. Transport
data for the neutral and doped materials are discussed in the light of Extended Hu¨ckel band calculations.
Introduction
have discovered that oxidation of dithiadiazolyl radicals, e.g.,
1 (E ) S; R ) H, Ph), 2, and 3, with iodine affords stacked
charge-transfer (CT) salts.8,9 Doping of the radicals away from
the half-filled energy band leads to improved transport proper-
In the past decade the molecular and solid state properties of
heterocyclic thiazyl radicals have been actively investigated.2
We have concentrated our efforts on derivatives of the 1,2,3,5-
dithiadiazolyl and -diselenadiazolyl ring systems 1 (E ) S, Se),3
with the intent of using these neutral π-radicals as templates
for the design of low-dimensional (stacked) molecular conduc-
tors.4 A wide variety of mono-,5 di-,6 and trifunctional7 radicals
have been prepared and characterized, and their solid state
features have been probed with a combination of experimental
and theoretical techniques. Several of the selenium variants
have been found to be small band gap semiconductors (e.g., 2
and 3, E ) Se).6a,b Metallic behavior has yet to be observed in
the neutral radicals, a consequence of the Peierls instability
associated with a half-filled energy band. More recently we
(5) (a) Cordes, A. W.; Haddon, R. C.; Hicks, R. G.; Oakley, R. T.; Palstra,
T. T. M. Inorg. Chem. 1992, 31, 1802. (b) Cordes, A. W.; Chamchoumis,
C. M.; Hicks, R. G.; Oakley, R. T.; Young, K. M.; Haddon, R. C. Can. J.
Chem. 1992, 70, 919. (c) Davis, W. M.; Hicks, R. G.; Oakley, R. T.; Zhao,
B.; Taylor, N. J. Can. J. Chem. 1993, 71, 180. (d) Cordes, A. W.; Glarum,
S. H.; Haddon, R. C.; Hallford, R.; Hicks, R. G.; Kennepohl, D. K.; Oakley,
R. T.; Palstra, T. T. M.; Scott, S. R. J. Chem. Soc., Chem. Commun. 1992,
1265. (e) Cordes, A. W.; Bryan, C. D.; Davis, W. M.; DeLaat, R. H.;
Glarum, S. H.; Goddard, J. D.; Haddon, R. C.; Hicks, R. G.; Kennepohl,
D. K.; Oakley, R. T.; Scott, S. R.; Westwood, N. P. C. J. Am. Chem. Soc.
1993, 115, 7232.
(6) (a) Cordes, A. W.; Haddon, R. C.; Oakley, R. T.; Schneemeyer, L.
F.; Waszczak, J. V.; Young, K. M.; Zimmerman, N. M. J. Am. Chem. Soc.
1991, 113, 582. (b) Andrews, M. P.; Cordes, A. W.; Douglass, D. C.;
Fleming, R. M.; Glarum S. H.; Haddon, R. C.; Marsh, P.; Oakley, R. T.;
Palstra, T. T. M.; Schneemeyer, L. F.; Trucks, G. W.; Tycko, R.; Waszczak,
J. V.; Young, K. M.; Zimmerman, N. M. J. Am. Chem. Soc. 1991, 113,
3559. (c) Cordes, A. W.; Haddon, R. C.; Hicks, R. G.; Kennepohl, D. K.;
Oakley, R. T.; Palstra, T. T. M.; Scheneemeyer, L. F.; Scott, S. R.;
Waszczak, J. V. Chem. Mater. 1993, 5, 820.
(7) (a) Cordes, A. W.; Haddon, R. C.; Hicks, R. G.; Oakley, R. T.; Palstra,
T. T. M.; Schneemeyer, L. F.; Waszczak, J. V. J. Am. Chem. Soc. 1992,
114, 5000. (b) Cordes, A. W.; Haddon, R. C.; Hicks, R. G.; Kennepohl, D.
K.; Oakley, R. T.; Schneemeyer, L. F.; Waszczak, J. V. Inorg. Chem. 1993,
32, 1554.
(8) (a) Cordes, A. W.; Haddon, R. C.; Hicks, R. G.; Kennepohl, D. K.;
MacKinnon, C. D.; Oakley, R. T.; Palstra, T. T. M.; Perel, A. S.; Scott, S.
R.; Schneemeyer, L. F.; Waszczak, J. V. J. Am. Chem. Soc. 1994, 116,
1205. (b) Bryan, C. D.; Cordes, A. W.; Fleming, R. W.; George, N. A.;
Glarum S. H.; Haddon, R. C.; Oakley, R. T.; Palstra, T. T. M.; Perel, A.
S.; Schneemeyer, L. F.; Waszczak, J. V. Nature 1993, 365, 821. (c) Bryan,
C. D.; Cordes, A. W.; Haddon, R. C.; Hicks, R. G.; Oakley, R. T.; Palstra,
T. T. M.; Perel, A. S.; Scott, S. R. Chem. Mater. 1994, 6, 508.
X Abstract published in AdVance ACS Abstracts, December 15, 1995.
(1) (a) University of Arkansas. (b) University of Guelph. (c) AT&T Bell
Laboratories.
(2) (a) Scherer, O. J.; Wolmersha¨user, G.; Jotter, R. Z. Naturforsch. 1982,
87b, 432. (b) Wolmersha¨user, G.; Schnauber, M.; Wilhelm, T. J. Chem.
Soc., Chem. Commun. 1984, 573. (c) Dormann, E.; Nowak, M. J.; Williams,
K. A.; Angus, R. O.; Wudl, F. J. Am. Chem. Soc. 1987, 109, 2594. (d)
Wolmersha¨user, G.; Wortmann, G.; Schnauber, M. J. Chem. Res., Synop.
1988, 358. (e) Banister, A. J.; Rawson, J. M. In The Chemistry of Inorganic
Ring Systems; Steudel, R., Ed.; Elsevier: Amsterdam, 1992; p 323. (f)
Parsons, S.; Passmore, J. Acc. Chem. Res. 1994, 27, 101. (g) Preston, K.
F.; Sutcliffe, L. H. Magn. Reson. Chem. 1990, 28, 189.
(3) Cordes, A. W.; Haddon, R. C.; Oakley, R. T. In The Chemistry of
Inorganic Ring Systems; Steudel, R., Ed.; Elsevier: Amsterdam, 1992; p
295.
(4) (a) Haddon, R. C. Nature (London) 1975, 256, 394. (b) Haddon, R.
C. Aust. J. Chem. 1975, 28, 2343.
0002-7863/96/1518-0330$12.00/0 © 1996 American Chemical Society

Preparation and Characterization of [S2N2C-CN2S2]
J. Am. Chem. Soc., Vol. 118, No. 2, 1996 331
Results and Discussion
Synthesis. The most versatile synthetic route to the dithia-
and diselenadiazolyl rings involves condensation of a per-
silylated amidine with sulfur or selenium dichloride (eq 1; E )
S, Se). The amidines are prepared by the “Sanger” reaction,
i.e., nucleophilic addition of lithium bis(trimethylsilyl)amide to
a nitrile followed by transmetalation with Me3SiCl.13,14 Initial
attempts to synthesize diradical 4 followed this approach.
However, the addition of LiN(SiMe3)2 to cyanogen did not
proceed to give the expected persilylated oxamidinesinstead
the major product was N,N′-bis(trimethylsilyl)carbodiimide.
Apparently the lithium amide reacts Via substitution rather than
addition to give lithium cyanide and cyanobis(trimethylsilyl)-
amine, which then tautomerizes to its more stable symmetric
isomer, the carbodiimide (eq 2).15
ties, and in some cases, room-temperature metallic conductivity
has been observed. However, like the neutral materials, these
CT salts are susceptible to charge density wave (CDW)-driven
instabilities, and all the conductive materials studied to date fall
into insulating states at low temperatures.
Suppression of charge density waves in the stacked structures,
to afford a metallic ground state, requires the design of materials
with more two- and three-dimensional electronic structures. At
the simplest level of theory the dimensionality of the electronic
structure and consequent transport properties are directly related
to both intra- and interstack overlap. Extended Hu¨ckel band
calculations have shown that crystal orbital dispersion along
the stacking direction is considerable. Lateral, interstack overlap
is, by comparison, considerably weaker but has been shown to
have important effects on the properties of the neutral radicals
by influencing (i) the magnitude of the band gap in the undoped
materials5c and (ii) the stability of the metallic state in the doped
compounds.9 On the basis of previous structure-property
correlations, interstack overlap can be enhanced in derivatives
bearing very small R groups in the 4-position of the heterocyclic
ring. In this context the directly linked diradical 4 has been a
synthetic target for some time. The isomeric 1,3,2,4-dithia-
diazolyl diradical 5 has been isolated but, to date, no structural
information has been reported.10 The difficulties encountered
with this latter material may be a consequence of the known
tendency of the 1,3,2,4-dithiadiazolyls to isomerize to the more
stable 1,2,3,5-isomers.11 In addition to displaying potentially
strong multidimensional interactions in the solid state, diradical
4 itself also provides an intriguing opportunity to study the extent
of intramolecular electronic (exchange) effects arising from the
presence of two directly linked radical units. A preliminary
report of the synthesis and structure of 4 has been published.12
In this paper we present the results of both theoretical and
experimental investigations on the molecular and solid state
properties of the diradical 4, as well as the preparation and solid
state properties of the highly conducting charge-transfer salt
[4][I] formed by the reaction of the diradical with iodine.
Subsequent synthetic efforts focused on the unsubstituted
oxamidine. A wide variety of N-substituted oxamidines are
known but conspicuously absent from this list is the unsubsti-
tuted prototypal derivative.16 We first attempted to generate
oxamidine Via the Pinner methodology (eq 3).17 The reaction
of diethyl oximidate18 with a mixture of ammonium chloride
and ammonia gas afforded impure oxamidine dihydrochloride
(identified by mass spectrometry) that was contaminated with
small amounts of ammonium salts. Surprisingly, however, this
compound did not react with sulfur dichloride at room temper-
ature, and heating the reaction mixture only caused extensive
decomposition.
In view of the difficulties encountered in preparing the
dication of 4 from either a persilylated or unsubstituted
oxamidine, we sought other potential precursors. The compound
N,N′-diaminooxamidine, or “oxamidrazone” (6) was identified
(13) Boere´, R. T.; Oakley, R. T.; Reed, R. W. J. Organomet. Chem.
1987, 331, 161.
(9) Bryan, C. D.; Cordes, A. W.; Fleming, R. M.; George, N. A.; Glarum,
S. H.; Haddon, R. C.; Oakley, R. T.; Palstra, T. T. M.; Perel, A. S. J. Am.
Chem. Soc. 1995, 117, 6880.
(10) (a) Parsons, S.; Passmore, J.; Schriver, M. J.; White, P. S. J. Chem.
Soc., Chem. Commun. 1991, 369. (b) Parsons, S.; Passmore, J.; White, P.
S. J. Chem. Soc., Dalton Trans. 1993, 1499.
(11) (a) Burford, N.; Passmore, J.; Schriver, M. J. J. Chem. Soc., Chem.
Commun. 1986, 140. (b) Aherne, C.; Banister, A. J.; Luke, A. W.; Rawson,
J. M.; Whitehead, R. J. J. Chem. Soc., Dalton Trans. 1992, 1277. (c)
Passmore, J.; Sun, X. Inorg. Chem., in press.
(12) Bryan, C. D.; Cordes, A. W.; Haddon, R. C.; Hicks, R. G.; Oakley,
R. T.; Palstra, T. T. M.; Perel, A. S. J. Chem. Soc., Chem. Commun. 1994,
1447.
(14) Edelmann, F. T. Coord. Chem. ReV. 1995, 137, 403.
(15) Gordestov, A. S.; Vostokov, V. P.; Sheludyalkova, S. V.; Dergunov,
Y. I.; Mironov, V. F. Russ. Chem. ReV. 1982, 51, 485.
(16) One claim has been made for the preparation of oxamidine as its
phosphate salt [Ito, M. Chem. Abstr. 1973, 78, 57800a], but in our hands,
the reaction of cyanogen with ammonia and phosphoric acid was unsuc-
cessful.
(17) Gautier, J. A.; Miocque, M.; Farnoux, C. C. In The Chemistry of
Amidines and Imidates; Patai, S., Ed.; John Wiley & Sons: London, 1975;
pp 283-348 and references cited therein.
(18) Diethyl oximidate has been prepared using a modified literature
procedure (see the Experimental Section): Behun, J. D. Chem. Abstr. 1961,
60, 5342c,d.

332 J. Am. Chem. Soc., Vol. 118, No. 2, 1996
Bryan et al.
Scheme 1
Figure 1. Packing of [4]2 in the yz plane.
as a promising candidate. It possessed the necessary C2N4
central core and also had reactive N-N bonds which might be
cleaved under oxidizing conditions, e.g., reaction with SCl2.19
Oxamidrazone has traditionally been prepared by addition of
hydrazine to cyanogen,20 but we have found that the reaction
of diethyl oximidate with 2 mol equiv of anhydrous hydrazine
in hot ethanol provides a much easier route (eq 4). To our great
satisfaction we then discovered that the dication of 4 could be
synthesized by the reaction of oxamidrazone 6 with SCl2
(Scheme 1); addition of an excess of SCl2 to 6 in acetonitrile
thus afforded [4][Cl]2 in 16% yield.21 The dication was isolated
in pure form by metathesis of the crude dichloride with NOPF6
to yield the hexafluorophosphate salt [4][PF6]2. Reduction of
the dichloride by well-established procedures (triphenylantimony
in refluxing acetonitrile) produced the dimer [4]2, which was
purified by vacuum sublimation at 140 °C/10-2 Torr in a
gradient tube furnace to afford blue/black needles in 52% yield.
Oxidative doping of 4 with iodine was effected with the
cosublimation strategy used in the preparation of other doped
dithiadiazolyls (e.g., 1, R ) H, Ph; 2, E ) S).8,9 Thus, heating
a 1:1 mixture of iodine and diradical 4 in an evacuated tube
along a carefully monitored temperature gradient from 150 to
110 °C afforded lustrous black crystalline needles of the CT
salt [4][I] (Scheme 1).
Crystal Structures. Details of the crystal structure solution
of [4]2 were reported in an earlier communication.12 The
structure is monoclinic, space group P21/c, with a ) 6.7623(17)
Å, b ) 11.5180(8) Å, c ) 8.3834(14) Å, â ) 110.20(2)°, Z )
4 (at 293 K) and consists of cofacial diradical dimers as the
structural building block. The dimers form stacks parallel to
the x axis, with neighboring dimers within the stacks translated
parallel to the central C-C bond vector, producing stacks that
are skewed by approximately 18° (as determined by the mean
S-S- -S bond angles). The packing of dimeric stacks produces
a herringbone-like motif in the yz plane (Figure 1). The
intradimer S-S distances (mean value 3.078 Å) and interdimer
Figure 2. Packing of [4][I] in the xz plane. Intermolecular contacts
are defined in the text.
S- -S distances (mean value 3.761 Å) are comparable to those
found in the parent monofunctional dithiadiazolyl 1 (E ) S, R
) H).5e,8a Neighboring stacks are slightly out-of-register and
closely packed, producing a very tight web of interstack contacts
(see Figure 1). These distances were defined in our earlier
communication.12
Crystals of [4][I] belong to orthorhombic space group
Ccmm,22 a ) 11.909(3) Å, b ) 3.271(2) Å, c ) 19.860(6) Å,
Z ) 4 (at 293 K). Molecules of 4 in [4][I] lie on a mirror
plane and are centrosymmetric. The crystal structure consists
of columns perfectly superimposed and regularly spaced mol-
ecules of 4 (as opposed to dimers) interspersed by columns of
iodines, with a herringbone packing motif (Figure 2) qualita-
tively similar to that found for the iodine-doped diradical [3][I]
(E ) S) rather than the linear ribbons found in [2][I]. At
3.271(3) Å, the interannular repeat along the y direction is
considerably closer than the interplanar separations in [2][I]
(3.415(2) Å) and [3][I] (3.487(3) Å). The closer spacing in
the present case could arise from the absence of an intervening
benzene ring (present in [2][I] and [3][I]), which would tend to
impose an interannular separation near the graphite value of
3.35-3.37 Å.23 The crystal symmetry requires that molecules
within alternate stacks centered at, for example x ) 0, 0.5, 1,
1.5, etc.; neighboring stacks are thus out-of-register. As
observed in [2][I] and [3][I], the iodines are disordered. The
disorder model used places the dominant fractional iodine
(19) The oxidative cleavage of N-N bonds by sulfur dichloride as a
route to SN-containing heterocycles has been used elsewhere. (a) Bestari,
K. T.; Cordes, A. W.; Oakley, R. T. J. Chem. Soc., Chem. Commun. 1988,
1328. (b) Roesky, H. W.; Mu¨ller, T. Chem. Ber. 1978, 111, 2960.
(20) Hergenrother, P. M. J. Macromol. Sci. 1973, A7, 573.
(21) This reaction is in contrast to the condensation of SCl2 with
oxamidoxime, which gives the fused ring compound bis(thiadiazole). See:
Komin, A. P.; Street, R. W.; Carmack, M. J. Org. Chem. 1975, 40, 2749.
(22) Ccmm is a nonstandard setting of Cmcm; in Cmcm the a and b cell
dimensions would interchange.
(23) Laves, F.; Baskin, Y. Z. Kristallogr.; Kristallgeom., Kristallphys.,
Kristallchem. 1956, 107, 337.

Preparation and Characterization of [S2N2C-CN2S2]
J. Am. Chem. Soc., Vol. 118, No. 2, 1996 333
Table 1. Mean Observed Intramolecular Distances (Å) in [4]2 and
[4][I]
the two unpaired electrons can be localized to separate groups
of atoms.27 The two singly occupied molecular orbitals
(SOMOs) are linear combinations of the localized orbitals 7 (a
and b). These can be compared with those of the well-known
tetramethyleneethane (TME) diradical, i.e., 8 (a and b).28-30
[4]2
[4][I]
S-S
S-N
N-C
C-C
2.088(2)
1.630(5)
1.331(7)
1.488(7)
2.088(11)
1.628(12)
1.335(16)
1.473(24)
centers in the plane of its two nearest neighbor rings, in a pattern
reminiscent of that found for [2][I] and [3][I].24
A comparison of the internal structural distances with those
of the undoped diradical dimer [4]2 are presented in Table 1. It
is noteworthy that no significant differences are observed
between the intramolecular S-S and S-N distances in the two
structures. Typically, such changes are related to degree of CT
by virtue of the fact that the SOMO, a strongly antibonding
distibution over the S-S and S-N linkages, is unoccupied in
the cationic state and filled in the neutral (radical) state.5e Thus
the lack of structural differences between [4]2 and [4][I] is in
contrast to those noted between 2 and 3 and their iodine CT
salts9 and suggests a significantly smaller degree of charge
transfer from radical to iodine in the present case. The relatively
short close cell repeat (see above) in comparison to those
observed in [2][I] and [3][I] also points to smaller degree of
charge transfer in the present system. Infrared spectroscopy
provides further substantiation of this conclusion. On the basis
of previous work,5e the band at 543 cm-1 in [4][Cl]2 can be
assigned to an intrannular S-S vibration. As expected this shifts
and splits into two bands at 495 and 507 cm-1 in the dimer
[4]2. The IR spectrum of [4][I], however, is severely weakened
by scattering (the compound is conductive); there is, nonetheless,
a weak line at 508 cm-1 which may correspond to the intrannular
S-S stretching mode. If this assignment is correct, then the
near equality of frequencies observed for [4]2 and [4][I] also
support a very low degree of charge transfer from 4 to iodine.25
We cannot, however, go much further than to provide these
qualitative assessments of the degree of charge transfer. For
[2][I] we were able to identify the low-temperature superlattice
by X-ray diffraction and, hence, assign the degree of charge
transfer as [2]+0.5[I]-0.5. In the case of [4][I] we have been
unable to observe the CDW-driven superlattice formation by
low-temperature crystallography, although clearly the magnetic
and conductivity measurements (Vide infra) signal its onset.
On the basis of the number and magnitude of the lateral
interannular S- - -S contacts, the structure appears to be highly
one-dimensional. The closest contact, which is well outside of
the standard van der Waals separation26 for two sulfurs (3.60
Å), is the S1- - -S1′ contact that spans adjacent heads across
the bridging iodine channels (see Figure 2). The only other
contact under 4 Å is the lateral S2- - -S2′ contact (3.985 Å).
Electronic Structure of 4. Diradical 4 is a rare example of
a disjoint diradical, i.e., one in which the molecular orbitals for
Since these localized orbitals are symmetry equivalent, the
configurations of the four possible states arising can be described
using the (a + b) and (a - b) combinations as31
diradical triplet
diradical singlet
out-of-phase zwitterionic singlet
in-phase zwitterionic singlet
3(a + b)(a - b)
1(a + b)2 - 1(a - b)2
1(a + b)(a - b)
1(a + b)2 + 1(a - b)2
For disjoint diradicals exchange interactions between the two
centers are small, and the diradical triplet and singlet states
which arise are essentially degenerate (as are the two zwitter-
ionic states). Recent calculations on TME suggest the diradical
triplet is lower in energy, by 1.0-1.5 kcal mol-1, than the
diradical singlet.32
In order to probe the energetic differences between the two
diradical states possible for 4, we have carried out a series of
ab initio molecular orbital calculations.33 The predicted bond
lengths for the planar and perpendicular forms of 4 as well as
the dimer can be found in Table 2. Full details of all geometries
are provided in the supporting information. Comparison of the
geometries corresponding to the diradical triplet and singlet
states reveals that they are structurally very similar. This is a
direct result of the nature of the two nearly degenerate SOMOs
[(a + b) and (a - b)]. They are both C-C nonbonding and
S-S and S-N antibonding, and so single occupation of both,
or double occupation of only one, leads to little change in
geometry. For both states the lack of interaction between the
carbons is also manifested by the absence of any torsional
preference (rotation angle τ about the C-C bond); the geom-
etries of planar (τ ) 0°) and perpendicular (τ ) 90°) structures
differ by no more 0.006 Å in any bond length and less than 1°
in any bond angle. Rotation about the C-C bond is also
relatively unhindered; the diradical singlet exhibits a barrier of
ca. 0.2 kcal mol-1. The triplet, on the other hand, is found to
have a minimum at τ ) 43.5° which is 0.59 kcal mol-1 more
stable than either the planar or perpendicular minima. The
(27) (a) Borden, W. T.; Davidson, E. R. J. Am. Chem. Soc. 1977, 99,
4587. (b) Pranata, J. J. Am. Chem. Soc. 1992, 114, 10537. (c) Rajca, A.
Chem. ReV. 1994, 94, 871.
(28) (a) Dowd, P.; Chang, W.; Paik, Y. H. J. Am. Chem. Soc. 1986,
108, 7416. (b) Dowd, P.; Chang, W.; Paik, Y. H. J. Am. Chem. Soc. 1987,
109, 5284.
(29) Du, P.; Borden, W. T. J. Am. Chem. Soc. 1987, 109, 930.
(30) Borden, W. T. In Diradicals; Borden, W. T., Ed.; J. Wiley and
Sons: New York, 1982; p 24.
(31) (a) Salem, L. Electrons in Chemical Reactions; First Principles;
Wiley Interscience: Chichester, U.K., 1982; Chapter 3. (b) Salem, L.;
Rowland, C. Angew. Chem., Int. Ed. Engl. 1972, 11, 92.
(32) Nachtigall, P.; Jordan, K. D. J. Am. Chem. Soc. 1993, 115, 270.
(33) Using SCF methods we have also examined briefly one (the out-
of-phase) combination of the two higher-lying zwitterionic singlet states
and established that it lies much higher in energy than the diradicals. The
in-phase zwitterionic singlet was not examined in any detail. A smaller
basis set configuration interaction calculation placed it an energy similar
to that of the out-of-phase zwitterionic singlet, as expected.
(24) For more general discussions of charge transfer in iodine containing
low-dimensional conductors, see: (a) Coppens, P. In Extended Linear Chain
Compounds; Miller, J. S., Ed.; Plenum Press: New York, 1982; pp 333-
356. (b) Marks, T. J.; Kalina, D. W. In Extended Linear Chain Compounds;
Miller, J. S., Ed.; Plenum Press: New York, 1982; pps 197-331. (c)
Kerte´sz, M.; Vonderviszt, F. J. Am. Chem. Soc. 1982, 104, 5889.
(25) We have also recorded the FT Raman spectra of [4]2 and [4][I] and
have encountered interpretational difficulties very similar to those observed
with [HCN2S2] and [HCN2S2]6[I]1.1 (see ref 8a). [4][I] fluoresces very
strongly and exhibits only a single, very broad band near 93 cm-1. This
-
could be attributed to an Ix species and some assignment made on the
basis of past experience (ref 24a,b). However, in [4]2, there are no less
than three bands in this region (at 182, 135, and 85 cm-1). These correspond
to interannular (interplate) S- - -S modes. We cannot therefore assign the
observed band(s) in [4][I] (and hence provide some clue as to the nature of
-
the Ix species) because we cannot identify them as I- - -I or S- - -S (or
even S- - -I) modes.
(26) Bondi, A. J. Phys. Chem. 1964, 68, 441.

334 J. Am. Chem. Soc., Vol. 118, No. 2, 1996
Table 2. Calculated bond distances (Å) for 4 and [4]2
Bryan et al.
S-S
S-N
N-C
C-C
Planar (τ ) 0°) Triplet (D2h)
CEP-31G*
6-31G*
6-311G*
2.075
2.067
2.081
1.641
1.632
1.628
1.322
1.311
1.309
1.520
1.499
1.502
Planar (τ ) 0°) Open Shell Singlet (D2h)
CEP-31G*
6-31G*
6-311G*
2.075
2.067
2.082
1.641
1.632
1.629
1.322
1.310
1.309
1.518
1.497
1.500
Twisted (τ ) 90°) Triplet (D2d)
CEP-31G*
6-31G*
6-311G*
2.071
2.062
2.077
1.646
1.636
1.633
1.320
1.309
1.307
1.515
1.497
1.498
Twisted (τ ) 90°) Open Shell Singlet (D2d)
CEP-31G*
6-31G*
6-311G*
2.072
2.063
2.077
1.646
1.637
1.633
1.320
1.309
1.307
1.513
1.495
1.496
Figure 3. ESR spectrum of [4] at 273 K (in CHCl3).
Closed Shell Singlet Dimer (D2h)
interpretations and have established that both reduction and
oxidation of the radicals are electrochemically reversible.35
However, observation of a reversible reduction wave requires
that the experiment be performed on the radical itself rather
than its oxidized counterpart, i.e., the dithiadiazolium cation. If
the latter is used, the reduction is irreversible, as the anion so
formed reacts rapidly with excess cation in the bulk solution.
The low solubility of the dimer [4]2 in organic media has
prevented us from carrying out electrochemical measurements
on the diradical itself. In seeking electrochemical data on 4
we have therefore been restricted to CV experiments on [4]2+
(as its PF6- salt in acetonitrile). The CV waves for this system
are very similar in position and shape to those recently reported
CEP-31G*
6-31G*
6-311G*
2.061
2.054
2.065
1.635
1.624
1.621
1.323
1.312
1.310
1.517
1.499
1.501
Table 3. Relative Energies (kcal mol-1) of Planar (τ ) 0°) and
Twisted (τ ) 90°) 4
state, τ
CEP-31G*
6-31G*
6-311G*
singlet, 0°
singlet, 90°
triplet, 0°
0.53
0.00a
1.03
0.54
0.00a
0.28
0.46
0.77
0.00a
0.03
0.50
0.58
triplet, 90°
a CEP-31G*, -89.302 980; 6-31G*, -1883.420 043; 6-311G*,
-1883.580 500.
for the benzene-bridged dications [2]2+ and [3]2+ 36
i.e., a
,
reversible cation/radical wave at 0.68 V (Vs SCE) and an
irreversible radical/anion wave near -0.80 V (Vs SCE).
Analysis of the dication/diradical wave revealed a peak-to-peak
separation (98 mV) slightly larger than those observed for the
benzene-bridged species. This may be simply a concentration
effect or may indicate two slightly interacting one-electron
transfers rather than a single two-electron step.37 In accord with
the electrochemical work, and the conclusions of the ab initio
studies described above, the ESR spectrum of 4 indicates
virtually no exchange interactions between the two radical
sites.38,39 At 273 K (in CHCl3) a simple five-line pattern with
g ) 2.011 and aN ) 0.50 mT is observed (Figure 3). At 303
K there is a slight broadening of the spectrum and the incipient
appearance of features associated with the onset of exchange
coupling. This slight temperature dependence may be associated
with subtle variations in the magnitude of exchange coupling
with torsional motion about the C-C bond, although the
theoretical results described above suggest that such changes
would be small. In our previous studies on the ESR spectra of
energy difference between the triplet and diradical singlet states,
as displayed for various levels of theory in Table 3, is very
small. The planar diradical singlet is lowest in energy, with
the corresponding triplet only very slightly higher in energy
(ca. 0.5 kcal mol-1); allowance for torsion brings the two states
even closer in energy.
Dimerization of the diradical, and consequent full orbital
occupancy, effects only slight internal structural changes. By
comparing the predicted bond lengths (Table 2) with those from
the X-ray data (Table 1), we see that the isolated dimer treated
theoretically has slightly shorter S-S, S-N, and N-C bonds,
with the most noticeable contraction being in the N-C distance.
By contrast the calculated C-C bonds are slightly longer. The
intermolecular S-S distance is predicted to be 3.096 Å at the
6-311G* SCF level, which is consistent with the experimental
data (mean value 3.078 Å). These findings mirror the previously
found results for the prototypal radical [HCN2S2] (1, E ) S, R
) H) system and its dimer.5e Investigation of the binding
energies show that the dimer is unbound at the SCF level with
respect to two triplet diradicals. However, with the inclusion
of electron correlation (Via MP2), it is bound by about 17 kcal
mol-1 relative to the triplet monomers. As seen in the prototypal
system,5e the binding energy is probably overestimated at the
MP2 level and would be expected to decrease somewhat at
higher levels of Møller-Plesset theory.
Cyclic Voltammetry and ESR Spectra. The electronic
structure of dithiadiazolyls 1 can be described in terms of a
singly occupied molecular orbital (SOMO) that is nodal at
carbon. The near invariance of many physical properties, e.g.,
ionization potentials and ESR hyperfine coupling constants, to
the nature of the R group in the 4-position can be understood
in that light.34 Recent cyclic voltammetric (CV) studies on a
series of monofunctional derivatives have led to similar
(34) (a) Boere´, R. T.; Oakley, R. T.; Reed, R. W.; Westwood, N. P. C.
J. Am. Chem. Soc. 1989, 111, 1180. (b) Cordes, A. W.; Goddard, J. D.;
Oakley, R. T.; Westwood, N. P. C. J. Am. Chem. Soc. 1989, 111, 6147.
(35) (a) Boere´, R. T.; Moock K.; Parvez, M. Z. Anorg. Allg. Chem. 1994,
620, 1589. (b) Aberne, C. M.; Banister, A. J.; Gorrell, I. B., Hansford, M.
I. Hauptman, Z. V.; Luke, A. W.; Rawson, J. M. J. Chem. Soc., Dalton
Trans. 1993, 976.
(36) Boere´, R. T.; Moock, K. H. J. Am. Chem. Soc. 1995, 117, 4755.
(37) (b) Polcyn, D. S.; Shain, I. Anal. Chem. 1966, 38, 370. (b) Ammar,
F.; Save´ant, J. M. J. Electroanal. Chem. 1973, 47, 215. (c) Bard, A. J.;
Faulkner, L. R. In Electrochemical Methods; John Wiley and Sons: New
York, 1980; p 232.
(38) (a) Weil, J. A.; Bolton, J. R.; Wertz, J. E. In Electron Paramagnetic
Resonance; John Wiley and Sons: New York, 1993; Section 6.4. (b)
Glarum, S. H.; Marshall, J. H. J. Chem. Phys. 1967, 47 1374. (c) Brie`re,
R.; Dupeyre, R. M.; Lemaire, H.; Morat, C.; Rassat, A.; Rey, P. Bull. Soc.
Chim. Fr. 1965, 3290.
(39) Bencini, A.; Gatteschi, D. In EPR of Exchange Coupled Systems;
Springer-Verlag: New York, 1990; p 187.

Preparation and Characterization of [S2N2C-CN2S2]
J. Am. Chem. Soc., Vol. 118, No. 2, 1996 335
Figure 4. Evolution of frontier π-orbitals (EHMO) of 1 (R ) H) (a) to those of 4 (b) and [4]2 (isolated) (c) to the block band structure (d) of a
single stack of [4]2 set at the observed spacing in the crystal structure. The band gap in the density of states (DOS, states/eV‚cell)) diagram (e)
corresponds to the smallest (π7/π8)+ - (π7/π8)- difference.
2 and 3 we noted stronger exchange interactions;6a,b indeed at
60 °C the spectrum was near the limiting case for which Jex
>
aN. The more extensive coupling observed for these latter
systems presumably arises from the ability of the benzene
bridge, particularly in the 1,3-substituted derivative, to mediate
coupling.
Band Structure Calculations. In our previous studies on
dithiadiazolyl-based materials we have tried to link solid state
structure to transport properties by means of Extended Hu¨ckel
band structure calculations. Our intent has been to interpret
the degree of crystal orbital dispersion parallel and perpendicular
to the stacking direction and to correlate this dispersion with
the calculated valence-to-conduction band gap.40 The present
system proved more complicated than was originally anticipated,
as the Extended Hu¨ckel method leads to a misordering of levels
in the band gap region. The origin of this difficulty is illustrated
in Figure 4, which shows the evolution of the π-orbital manifold
of a single dithiadiazolyl radical 1 into the molecular orbitals
of the diradical 4, its dimer [4]2, and the final density of states
(DOS) for a one-dimensional stack of dimers. The progress of
the radical SOMO (ø4) through to the valence and conduction
bands can be easily monitored in this diagram. Note, however,
that the next highest π-orbital, ø5, is sharply split in the diradical
and a component of this interaction, π9-, tracks across and
crosses into the normal valence band region. On the basis of
our ab initio calculations, we know that this occurrence is an
artifact of the Extended Hu¨ckel method; the states arising from
π9 are, in reality, much higher lying.41 In contrast to the simple
picture in Figure 4 the dimer [4]2 is predicted, at the ab initio
Figure 5. Density of states (DOS, states/(eV‚cell)) for a one-
dimensional stack of [4]2 (left) and the full three-dimensional structure
(right). The band gaps specified correspond to the smallest (π7/π8)+
(π7/π8)- difference.
-
-
level, to have π9 as LUMO+3. If we take this factor into
account and assume that the band arising from the π9- orbitals
should be at higher energy than the EHMO results indicate,
then the calculated band gap, i.e., the (π7/π8)+ to (π7/π8)-
difference, for the one-dimensional structure is 1.40 eV. Figure
5 compares the DOS for a one-dimensional stack with that of
the full three-dimensional structure. Band spreading in the latter,
as a result of lateral interactions, is clearly quite pronounced. If
we again use the (π7/π8)+ to (π7/π8)- difference, as outlined
(41) The problem stems from the use (in the band calculations) of a
double-ú basis set, which slightly overemphasizes the internal C-C overlap
and, hence, the orbital splitting between π9 and π10 (Figure 4). We know
explicitly from our ab initio work on [4]2 that the occupied states associated
with π9 are, in reality, more high lying than the EHMO method suggests.
(40) Strictly speaking, such an approach requires that the unit cell vectors
be orthogonal. For structures in which this does not hold, the argument is
only of qualitative value.

336 J. Am. Chem. Soc., Vol. 118, No. 2, 1996
Bryan et al.
Figure 6. Density of states (DOS, states/(eV‚cell)) for a one-
dimensional stack of [4] in [4][I]. The energy difference specified
corresponds to the π7/π8 bandwidth. The hatched line at -8.2 eV is
included for reference purposes only; it corresponds to the Fermi level
for zero charge transfer.
Figure 8. Magnetic susceptibility of [4][I] as a function of temperature.
diamagnetism of -97 × 10-6 ppm emu mol-1 and a vanishingly
small paramagnetic contribution. Above 250 K paramagnetism
begins to rise, and carriers are still being generated at 400 K
(the high-temperature limit of the experiment). This behavior
is similar to that reported for [2][I] and [3][I], although the
maximum conductivity for present system is significantly higher.
This we attribute to the close intracolumnar spacing (the stack
repeat) and the consequently strong intermolecular overlap along
the stacking direction. The high metal-insulation transition TMI
near 270 K is likewise a consequence of the absence of crystal
orbital dispersion lateral to the stacking direction.
Summary
The disjoint nature of the diradical 4 has been confirmed by
the results of ab initio studies and ESR spectroscopy. In spite
of a well-developed network of intermolecular interactions,
which collectively lead to a small (by comparison with related
systems) calculated band gap, the solid state dimer [4]2 is a
diamagnetic insulator. The CT salt [4][I] exhibits the highest
conductivity yet reported for a dithiadiazolyl-based conductor.
The material is, however, more one-dimensional than, for
example, [2][I], and the lack of lateral interactions, coupled with
the strong intrastack contacts, leads to a greater susceptibility
to a CDW-driven phase transition to an insulating state, i.e., a
higher TMI. One of our future goals is to produce materials in
which the metallic state is stable at lower temperatures; this
will require derivatives in which intrastack interactions are
reduced and interstack interactions enhanced. Experiments to
this end are in progress.
Figure 7. Conductivity of [4][I] as a function of temperature.
above, the calculated band gap is found to be near 1.04 eV, a
value not far from that calculated for 3.6b
Band structure calculations on the doped structure [4][I]
confirm the qualitative conclusions drawn from the crystal-
lographic results; the electronic structure is highly one-
dimensional. The interpretation of the density of states data is
again complicated by crossover of states, as shown in the DOS
and block band diagrams (Figure 6) for a single stack of
diradicals based on the lattice separation found in [4][I].
Analysis of the crystal orbitals reveals a conduction bandwidth
of nearly 3.7 eV. This dispersion arises from interactions along
the stacking direction (Y). Dispersion calculations perpendicular
to the stacking direction (Γ to Σ and Γ to Z) reveal dispersion
of less than 0.1 eV.
Magnetic and Conductivity Measurements. Magnetic
susceptibility measurements on the dimer [4]2 confirm that it is
diamagnetic. Pressed pellet conductivity measurements point
to a value of <10-6 S cm-1. These results are similar to those
obtained for other stacked polyfunctional dithiadiazolyls. By
contrast the CT salt [4][I] is highly conductive, with a single-
crystal conductivity of 460 S cm-1 at 300 K. The temperature
dependence of the conductivity (Figure 7) shows weakly metallic
behavior at room temperature, with the onset of a phase
transition near 270 K; below this temperature conductivity is
weakly activated. Magnetic susceptibility measurements on
[4][I] (Figure 8) are consistent with the conductivity data. At
low temperature the material is diamagnetic, with a measured
Experimental Section
General Procedures and Starting Materials. Lithium bis(tri-
methylsilyl)amide, triphenylantimony, anhydrous hydrazine, chloro-
trimethylsilane, sulfur dichloride (Aldrich), cyanogen (Matheson),
nitrosyl hexafluorophosphate (Aesar), tetrabutylammonium hexafluo-
rophosphate (Fluka), and ethanol (95%, Fisher) were obtained com-
mercially. The solvents acetonitrile (Fisher, HPLC grade), chloroben-
zene (Fisher), and methylene chloride (Fisher) were distilled from P2O5
under a nitrogen atmosphere. Sulfur dichloride was also freshly distilled
prior to use; iodine (Aldrich) was resublimed before use. The gas phase
doping reactions with iodine were performed in an ATS series 3210
three-zone tube furnace, linked to a series 1400 temperature control
system. 1H NMR spectra were recorded on a Varian Gemini 200 MHz
NMR spectrometer; chemical shift values were internally referenced
to TMS or the residual proton signals of the solvents (δ, CHCl3).
Infrared spectra (Nujol mulls, CsI optics) were recorded on a Nicolet
20SX/C FTIR spectrometer at 2 cm-1 resolution. X-band ESR spectra

Preparation and Characterization of [S2N2C-CN2S2]
J. Am. Chem. Soc., Vol. 118, No. 2, 1996 337
were recorded on a Varian E-109 spectrometer with DPPH as a field
marker. Cyclic voltammetry was performed on a PAR 273A electro-
chemical system (EG&G Instruments) using a two-compartment cell
with Pt working, counter, and quasi-reference electrodes. Electrolyte
solutions contained 2-5 mg of [4][PF6]2 in 20 mL of 0.1 M tetra-n-
butylammonium hexafluorophosphate in acetonitrile. Potentials were
scanned from -2 to +1 V with respect to the quasi-reference electrode,
which was later referenced to a standard Ag/AgCl electrode. The half-
wave potentials are reported with reference to SCE. Mass spectra (70
eV, EI) were obtained on a Kratos MS890 spectrometer. Elemental
analyses were performed by MHW Laboratories, Phoenix, AZ 85018.
Preparation of Oxamidrazone 6. Freshly sublimed diethyl oxim-
idate (25.2 g, 0.175 mol) and hydrazine (12.0 mL, 0.38 mol; CAUTION!
extremely toxic compound) were dissolved in 130 mL of absolute
ethanol. The mixture was heated at 50 °C for 16 h, producing an off-
white microcrystalline precipitate. The solid was filtered in air, rinsed
with ethanol and diethyl ether, and air-dried to afford pure 6, yield
19.5 g (96%), mp 177-80 °C dec. The compound was recrystallized
from hot water as a hydrate; the water of hydration was removed by
heating the material at 110 °C in Vacuo for 16 h. Infrared spectrum
(4000-400 cm-1): 3352 (m), 3281 (w, br), 3137 (m, br), 2950 (s),
292 (s, br), 2845 (s), 1623 (m), 1570 (w), 1465 (s), 1377 (m), 1320
(w), 1262 (m), 1120 (w), 1030 (m), 876 (s), 739 (m) cm-1
. Mass
Reaction of Cyanogen with LiN(SiMe3)2. Cyanogen (9.23 g, 0.177
mol) was slowly condensed onto a frozen solution of LiN(SiMe3)2 (56.0
g, 0.345 mol) in 250 mL of toluene. The flask was allowed to warm
to room temperature, producing a light brown slurry. The reaction
was stirred for 2 h, during which time the mixture darkened consider-
ably. Neat Me3SiCl (40.2 g, 0.38 mol) was added, and the mixture
was gently heated overnight. A light brown solid was filtered off, and
the toluene was removed in Vacuo. Slow vacuum distillation (with a
30-cm Vigreux column) of the residue afforded 31.0 g of a colorless
liquid, bp 28-30 °C/10-2 Torr. The infrared and 1H NMR spectra of
this compound were identical with those of authentic N,N′-bis-
(trimethylsilyl)carbodiimide (Hu¨ls America). The yield was 95% based
on cyanogen. Subsequent reactions, which were performed with a 1:1
cyanogen:LiN(SiMe3)2 stoichiometry and without addition of chloro-
trimethylsilane, also afforded the carbodiimide as the major product.
spectrum: m/e 116 (M+, 100%), 100 ((M - NH2)+, 9%), 85 ((M -
N2H3)+, 17%), 70 (C2H4N3+, 18%), 58 (M2+, 36%), 43 (CH3N2+, 57%).
Anal. Calcd for C2H8N6: C, 20.69; H, 6.94; N, 72.37%. Found: C,
20.60; H, 6.80; N, 72.33%.
Preparation of 4,4′-Bis(1,2,3,5-dithiadiazolium) Dichloride, [4][Cl]2.
A solution of freshly distilled SCl2 (20.0 mL, 0.314 mol) in 80 mL of
acetonitrile was added slowly (over a 30-min period) to a slurry of
oxamidrazone (6.39 g, 55.0 mmol) in 200 mL of acetonitrile cooled in
an ice/water bath. After the SCl2 addition was complete the mixture
was allowed to warm to room temperature and stirred for 2 h. During
this time gentle gas evolution was observed. The resulting orange-
brown solid so obtained was filtered, rinsed copiously with acetonitrile,
and dried in Vacuo to afford crude 4,4′-bis(1,2,3,5-dithiadiazolium)
dichloride [4][Cl]2. The crude dication was slurried in 40 mL of sulfur
monochloride (S2Cl2) and refluxed overnight to consume any unreacted
ammonium salts. The mixture was cooled to room temperature and
filtered off. The orange solid was then slurried in 30 mL of
dichloromethane, and chlorine gas was rapidly bubbled through the
solution for 15 min to convert any binary sulfur-nitrogen species to
(CH2Cl2 soluble) S3N3Cl3. The remaining orange solid was filtered
off, rinsed copiously with dichloromethane, and dried in Vacuo to afford
[4][Cl]2, yield 2.55 g (16% based on oxamidrazone). Infrared spectrum
(2000-200 cm-1): 1695 (w), 1462 (vs), 1377 (s), 1367 (m), 1354 (w),
1341 (w), 1302 (s), 1261 (m), 1247 (m), 1094 (m, br), 1021 (m, br),
880 (s), 851 (s), 831 (w), 821 (s), 801 (m), 723 (w), 619 (s), 559 (w),
Preparation of Diethyl Oximidate. Sodium cyanide (40.0 g, 0.81
mol) was dissolved in a mixture of 200 mL of absolute ethanol and
350 mL of water, and the solution was cooled to -15 °C in a dry
ice/CCl4 bath. Chlorine gas was slowly bubbled through the solution,
and simultaneously, a solution of potassium hydroxide (1.5 g) in 100
mL of water was added dropwise. The reaction gradually turned
yellow, then orange and solidified after approximately 45 min. The
chlorine and KOH addition were halted, and the mixture was allowed
to warm to room temperature over a 3-h period. The solution was
extracted with three 40-mL portions of diethyl ether, the extracts were
dried over K2CO3, and the ether solution was evaporated to a
concentrated solution (ca. 100 mL) containing diethyl oximidate and
ethanol. The remaining ethanol was removed in Vacuo, leaving a
semisolid light brown mass. Vacuum sublimation at 30-35 °C/10-2
Torr afforded pure diethyl oximidate as white rodlike crystals, yield
34.0 g (68%). The product can also be isolated be direct crystallization
from the concentrated ethanol solution. Infrared spectrum (4000-400
cm-1): 3337 (s), 2958-2839 (vs), 1636 (vs), 1479 (w), 1463(vs), 1384
(m), 1377 (m), 1366 (m), 1284 (vs), 1174 (m), 1158 (w), 1116 (s),
543 (s) cm-1
.
Preparation of [4][PF6]2. Nitrosyl hexafluorophosphate (2.26 g,
12.9 mmol) and [4][Cl]2 (1.80 g, 6.4 mmol) were heated in 20 mL of
acetonitrile for 30 min, causing the dichloride salt to dissolve. The
solvent was removed in Vacuo and the residual yellow/brown solid
pumped dry overnight to remove traces of NOCl. This material was
triturated with 30 mL of 5:1 chlorobenzene/acetonitrile and the resulting
slurry filtered to give crude [4][PF6]2 as a white powder, yield 1.99 g
(99%). Recrystallization from 1:1 mixture of chlorobenzene/acetonitrile
afforded colorless crystalline blocks which were extremely air and
moisture sensitive. Infrared spectrum (2000-200 cm-1): 1705 (w),
1577 (w), 1460 (s), 1375(s), 1366 (s), 1300 (w), 1262 (w), 1208 (w),
1123 (s), 924 (m), 840 (vs, br), 742 (m), 723 (w), 638 (w), 594 (w),
562 (s), 524 (w), 414 (w) cm-1. Anal. Calcd for C2F12N4P2S4: C,
4.82; N, 11.25%. Found: C, 5.06; N, 11.44%.
Preparation of 4,4′-Bis(1,2,3,5-dithiadiazolyl) 4. Solid triphenyl-
antimony (3.40 g, 9.62 mmol) was added to a slurry of [4][Cl]2 (2.55
g, 9.13 mmol) in 50 mL of acetonitrile, immediately producing a black
solid. The mixture was gently refluxed overnight and then cooled.
The black precipitate was filtered off, rinsed with acetonitrile, and dried
in Vacuo to afford crude dimer [4]2, yield 1.89 g. Vacuum sublimation
at 140 °C/10-2 Torr gave metallic blue-black needles of [4]2, yield 0.97
g (51%). Mp 275-280 °C. Infrared spectrum (2000-200 cm-1): 1460
(s), 1376 (s), 1294 (m), 1264 (m), 1244 (s), 1080 (m, br), 820 (w), 804
(s), 793 (s), 784 (s), 559 (m), 507 (s), 495 (m), 421 (m) cm-1. Mass
spectrum: m/e 208 (M+, 84%), 182 ((M - NS)+, 97%), 144 ((M -
S2)+, 11%), 130 ((M - NS2)+, 31%), 116 ((M - N2S2)+, 14%), 104
(M2+, 45%, 64 (S2+, 100%), 46 (NS+, 80%). Anal. Calcd for
C2N4S4: C, 11.53; N, 26.90%. Found: C, 11.92; N, 26.98%.
Preparation of [4][I]. Resublimed [4]2 (0.210 g, 1.01 mmol) and
iodine (0.126 g, 1.00 mmol) were sealed in an evacuated (10-3 Torr)
Pyrex tube (25-mm diameter, 25-cm length). The contents of the tube
were heated to 150 °C for 1 week, during which time metallic black
needles grew at the far end of the tube (at 110 °C). The crystals were
harvested in air but stored under an inert atmosphere. Yield of [4][I]
1083 (vs), 1016 (s), 1005 (s), 857 (s), 798 (w), 715 (w), 483 (w) cm-1
.
1H NMR (δ, CDCl3): 8.30 (s, NH, 2H), 4.25 (q, CH2, 4H), 1.36 (t,
CH3, 6H). Anal. Calcd for C6H12N2O2: C, 49.99; H, 8.39; N, 19.43%.
Found: C, 49.73; H, 8.12; N, 19.63%.
Preparation of Oxamidine Dihydrochloride. Diethyl oximidate
(9.82 g, 68.1 mmol) and ammonium chloride (7.45 g, 139 mmol) were
slurried together in 60 mL of acetonitrile. Ammonia gas was slowly
bubbled through the mixture for 90 min and the flask then sealed and
the contents stirred at room temperature for 2 days to give a brick-red
precipitate and a brown solution. The solid was filtered in air, rinsed
with acetonitrile and dried in Vacuo, yield 11.0 g. Infrared spectrum
(4000-400 cm-1): 3100 (s, br), 2950 (vs), 1654 (s), 1603 (s), 1460
(s), 1377 (s), 1252 (m), 1175 (m), 1060 (w), 858 (w), 795 (w), 723
(w), 680 (m), 590 (m) cm-1. Mass spectrum: m/e 86 (C2H6N4+, M+,
29%), 70 ((M - NH2)+, 16%), 44 (MH22+, 100%).
Reaction of Oxamidine Dihydrochloride with Sulfur Dichloride.
Oxamidine dihydrochloride (0.75 g, 4.7 mmol) was slurried in 20 mL
of acetontrile. To this was added an excess of SCl2 (2.0 mL). The
mixture was stirred at room temperature for 1 h, after which no reaction
had occurred. The mixture was then heated at gentle reflux, causing
the mixture to turn dark green. The oxamidine dihydrochloride was
consumed, producing a small amount of a green solid which was filtered
off, rinsed with acetonitrile, and dried in Vacuo. The infrared spectrum
of this solid was not consistent with that of the desired dication [4]2+
.
In addition the material did not react with triphenylantimony in
acetonitrile.

338 J. Am. Chem. Soc., Vol. 118, No. 2, 1996
Table 4. Crystal data for [4][I]
Bryan et al.
0.80 (N), and 0.532 (S). All geometries were optimized using gradient
methods. The symmetry restrictions of D2h and D2d were imposed for
most calculations.
The band structure calculations were carried out with the EHMACC
suite of programs45 using the parameters discussed previously.6a,46 The
off-diagonal elements of the Hamiltonian matrix were calculated with
the standard weighting formula.47
Magnetic Susceptibility Measurements. The magnetic suscepti-
bilities as a function temperature were measured using a SQUID
magnetometer operating at 1 T.
Conductivity Measurements. Four-point conductivity measure-
ments along the highly conducting axes were performed with a Keithley
236 unit. Wires were attached to the pads with gold paint. The
measurements were carried out at constant voltage with an excitation
voltage of 0.1 V. The conductivity was ohmic over at least 2 orders
of magnitude.
formula
fw
a, Å
S4N4C2I
335.19
11.909(3)
3.271(2)
19.860(6)
773.6(7)
Ccmm
b, Å
c, Å
V, Å3
space group
Z
4
temp, K
µ, mm-1
R(F), Rw(F)a
293
5.07
0.059,0.078
a R ) [∑||Fo| - |Fc||]/[∑|Fo|]; Rw ) {[∑w||Fo| - |Fc|| ]/
2
2
[∑(w|Fo| )]}1/2
.
80 mg (23%). Anal. Calcd for C2N4S4I: C, 7.17; N, 16.71; I, 37.86%.
Found: C, 7.25; N, 16.87; I, 37.68%. Infrared spectrum: as observed
for other highly conductive CT iodides,9 [4][I] scatters IR radiation
very strongly; the IR spectrum of [4][I] is very weak and poorly
resolved, revealing only three distinct bands at 1283, 819, and 508
Acknowledgment. Financial support at Guelph was provided
by the Natural Sciences and Engineering Research Council of
Canada (NSERC) and at Arkansas by the National Science
Foundation (EPSCOR program). C.D.B. acknowledges a DOE/
ASTA Traineeship, and both R.G.H. and C.D.M. acknowledge
NSERC postgraduate scholarships.
cm-1
.
X-ray Measurements. All X-ray data were collected on an ENRAF-
Nonius CAD-4 diffractometer with monochromated Mo KR (λ )
0.710 73 Å) radiation. Crystals of [4][I] were mounted on a glass fiber
with silicone rubber. Data were collected using a θ/2θ technique. The
structures were solved using direct methods refined by full-matrix least
squares which minimized ∑w(∆F)2. A summary of crystallographic
data is provided in Table 4.
Computational Methods. Ab initio predictions were made using
the 1995 version of the GAMESS suite of programs.42 For the triplet
state restricted open shell (ROHF) methodology was used. The
diradical singlet was described by the GVB-PP method, a two-
configuration SCF approach, starting from the ROHF orbitals for the
corresponding triplet. The basis sets used in this study consisted of
the Pople split-valence sets, 6-31G* and 6-311G*,43 and the SBK CEP-
31G set included within GAMESS,44 with the latter altered by the
addition of d-type polarization functions with coefficients of 0.75 (C),
Supporting Information Available: Tables of optimized
(ab initio) geometries of 4 and [4]2, crystal data, structure
solution and refinement, atomic coordinates, bond lengths and
angles, and anisotropic thermal parameters and an ORTEP
diagram for [S2N2C-CN2S2][I] (8 pages). This material is
contained in many libraries on microfiche, immediately follows
this article in the microfiche version of the journal, can be
ordered from the ACS, and can be downloaded from the
Internet; see any current masthead page for ordering information
and Internet access instructions.
JA952144X
(42) Schmidt, M. W.; Baldridge, K. K.; Boatz, J. A.; Elbert, S. T.;
Gordon, M. S.; Jensen, J. J.; Koseki, S.; Matsunaga, N.; Nguyen, K. A.;
Su, S.; Windus, T. L.; Dupuis, M.; Montgomery, J. A. J. Comput. Chem.
1993, 14, 1347.
(43) (a) Dill, J. D.; Pople, J. A. J. Chem. Phys. 1975, 56, 5255. (b) Francl,
M. W.; Pietro, W. J.; Hehre, W. J.; Binkley, J. S.; Gordon, M. S.; DeFrees,
D. J.; Pople, J. A. J. Chem. Phys. 1982, 77, 3654. (c) Krishnan, R.; Binkley,
J. S.; Seeger, R.; Pople, J. A. J. Chem. Phys. 1980, 72, 650. (d) Hariharan,
P. C.; Pople, J. A. Theor. Chim. Acta 1973, 28, 213.
(44) (a) Stevens, W. J.; Basch, H.; Krauss, M. J. Chem. Phys. 1984, 81,
6026. (b) Stevens, W. J.; Basch, H.; Krauss, M.; Jasien, P. Can. J. Chem.
1992, 70, 612. (c) Cundari, T. R.; Stevens, W. J. J. Chem. Phys. 1993, 98,
5555.
(45) EHMACC, Quantum Chemistry Program Exchange, Program No.
571.
(46) Basch, H.; Viste, A.; Gray, H. B. Theor. Chim. Acta 1965, 3, 458.
(47) Ammeter, J. H.; Burghi, H. B.; Thibeault, J. C.; Hoffmann, R. J.
Am. Chem. Soc. 1978, 100, 3686.