Solvent dependence of the structure and magnetic ordering of ferrimagnetic manganese(III) meso-tetraphenylporphyrin tetracyanoethenide, [MNTPP]+[TCNE]?-·x(solvent). Evidence for orientationally disordered [TCNE]?-

Article abstract of DOI:10.1021/ic0011922

Tetraphenylporphinatomanganate(III) tetracyanoethenide, [MnTPP][TCNE], is the prototype of a growing family of linear chain (1-D) coordination polymers that magnetically order as ferrimagnets. [MnTPP][TCNE]·xS [S = PhMe (x = 2), 1,2-C₆H₄Me₂ (x = 1), 1,2-C₆H₄Cl₂ (x = 3), 1,2,4-C₆H₃Cl₃ (x = 2), and 1,3-C₆H₄Cl₂ (x = 2)] have been prepared and structurally and magnetically characterized. All form 1-D chain structures with intrachain Mn···Mn separations ranging from 9.202 to 10.218 A. The 173 K crystal structure of [MnTPP][TCNE]·2PhMe has been rerefined, revealing that the [TCNE]^?- is 2-fold-disordered and coordinated to Mn^III by a pair of trans cyano nitrogen atoms to form parallel one-dimensional chains. The two orientations of [TCNE]^?- are related by a 180° rotation about the diagonal axis joining the trans nitrogen atoms bound to Mn^III. The major form has an occupancy of 83.3(4)% with a Mn-N_TCNE distance of 2.328(3) A and a MnNC angle of 146.8(8)°. The minor form, with 16.7(4)% occupancy, has a Mn-N_TCNE distance of 2.176(15) A and a MnNC angle of 152.3(39)°. Lattice packing and molecular bonding imply static as opposed to dynamic disorder. The magnetic properties depend on the type and quantity of the solvent present in the structure. Desolvation via heating in n-octane (127 °C), n-dodecane (216 °C), and/or vacuum thermolysis (175 °C) leads to numerous different desolvated materials with differing magnetic properties. At higher temperatures the magnetic susceptibility can be fit by the Curie-Weiss expression, X ∝ (T - θ)^-1, with θ = 44, 52, 72, 55, and 77 K for the toluene, 1,2-xylene, 1,2-dichlorobenzene, 1,3-dichlorobenzene, and 1,2,4-trichlorobenzene solvates, respectively. The T_c's were taken as the maximum in 10 Hz X′(T) and are 7.8, 9.2, 11.3, 10.8, and 8.2 K for the PhMe, 1,2-xylene, 1,2-dichlorobenzene, 1,3-dichlorobenzene, and 1,2,4-trichlorobenzene solvates, respectively. Upon desolvation the T_c's increase for the PhMe, 1,2-xylene, 1,2,4-trichlorobenzene solvates and decrease for the 1,2- and 1,3-dichlorobenezene solvates. The compounds show one-dimensional ferrimagnetic exchange behavior at high temperatures with intrachain exchange of J/k_b = -63, -99, -234, -100, and -200 K for toluene, 1,2-xylene, 1,2-dichlorobenzene, 1,3-dichlorobenzene, and 1,2,4-trichlorobenzene solvates, respectively, as determined from fits to the Seiden expression, which models isolated 1-D interactions among alternating S = 2 classical and S = 1/2 quantum spins. For variation in the temperature at which the peak occurs per decade of frequency, φ, (ΔT_f/T_f)/Δ(log ω) is 0.167, 0.168, 0.066, 0.171, and 0.024 for toluene, 1,2-xylene, 1,2-dichlorobenzene, 1,3-dichlorobenzene, and 1,2,4-trichlorobenzene, respectively, typical of spin glass behavior. Since [TCNE] ^?- is only disordered for the PhMe solvate and all five solvates exhibit spin glass behavior, the spin glass behavior cannot be attributed to this disorder.

Full text of DOI:10.1021/ic0011922

Inorg. Chem. 2001, 40, 1915-1925
1915
Solvent Dependence of the Structure and Magnetic Ordering of Ferrimagnetic
Manganese(III) meso-Tetraphenylporphyrin Tetracyanoethenide,
[MnTPP]+[TCNE]•-‚x(solvent). Evidence for Orientationally Disordered [TCNE]•-
Wendy Hibbs,1a Durrell K. Rittenberg,1a Ken-ichi Sugiura,1a,b Brian M. Burkhart,1c,d
Brian G. Morin,1d,e Atta M. Arif,1a Louise Liable-Sands,1f Arnold L. Rheingold,1f
Muttaiya Sundaralingam,1c Arthur J. Epstein,1c,e and Joel S. Miller*,1a
Department of Chemistry, University of Utah, 315 S. 1400 E., Room 2020,
Salt Lake City, Utah 84112-0850, Department of Chemistry and Department of Physics,
The Ohio State University, Columbus, Ohio 43210-1106, and Department of Chemistry,
University of Delaware, Newark, Delaware 19716
ReceiVed October 31, 2000
Tetraphenylporphinatomanganate(III) tetracyanoethenide, [MnTPP][TCNE], is the prototype of a growing family
of linear chain (1-D) coordination polymers that magnetically order as ferrimagnets. [MnTPP][TCNE]‚xS [S )
PhMe (x ) 2), 1,2-C6H4Me2 (x ) 1), 1,2-C6H4Cl2 (x ) 3), 1,2,4-C6H3Cl3 (x ) 2), and 1,3-C6H4Cl2 (x ) 2)] have
been prepared and structurally and magnetically characterized. All form 1-D chain structures with intrachain
Mn‚‚‚Mn separations ranging from 9.202 to 10.218 Å. The 173 K crystal structure of [MnTPP][TCNE]‚2PhMe
has been rerefined, revealing that the [TCNE]•- is 2-fold-disordered and coordinated to MnIII by a pair of trans
cyano nitrogen atoms to form parallel one-dimensional chains. The two orientations of [TCNE]•- are related by
a 180° rotation about the diagonal axis joining the trans nitrogen atoms bound to MnIII. The major form has an
occupancy of 83.3(4)% with a Mn-NTCNE distance of 2.328(3) Å and a MnNC angle of 146.8(8)°. The minor
form, with 16.7(4)% occupancy, has a Mn-NTCNE distance of 2.176(15) Å and a MnNC angle of 152.3(39)°.
Lattice packing and molecular bonding imply static as opposed to dynamic disorder. The magnetic properties
depend on the type and quantity of the solvent present in the structure. Desolvation via heating in n-octane (127
°C), n-dodecane (216 °C), and/or vacuum thermolysis (175 °C) leads to numerous different desolvated materials
with differing magnetic properties. At higher temperatures the magnetic susceptibility can be fit by the Curie-
Weiss expression, ø (T - θ)-1, with θ ) 44, 52, 72, 55, and 77 K for the toluene, 1,2-xylene, 1,2-dichlorobenzene,
1,3-dichlorobenzene, and 1,2,4-trichlorobenzene solvates, respectively. The Tc’s were taken as the maximum in
10 Hz ø′(T) and are 7.8, 9.2, 11.3, 10.8, and 8.2 K for the PhMe, 1,2-xylene, 1,2-dichlorobenzene,
1,3-dichlorobenzene, and 1,2,4-trichlorobenzene solvates, respectively. Upon desolvation the Tc’s increase for
the PhMe, 1,2-xylene, 1,2,4-trichlorobenzene solvates and decrease for the 1,2- and 1,3-dichlorobenezene solvates.
The compounds show one-dimensional ferrimagnetic exchange behavior at high temperatures with intrachain
exchange of J/kb ) -63, -99, -234, -100, and -200 K for toluene, 1,2-xylene, 1,2-dichlorobenzene, 1,3-
dichlorobenzene, and 1,2,4-trichlorobenzene solvates, respectively, as determined from fits to the Seiden expression,
which models isolated 1-D interactions among alternating S ) 2 classical and S ) 1/2 quantum spins. For variation
in the temperature at which the peak occurs per decade of frequency, φ, (∆Tf/Tf)/∆(log ω) is 0.167, 0.168, 0.066,
0.171, and 0.024 for toluene, 1,2-xylene, 1,2-dichlorobenzene, 1,3-dichlorobenzene, and 1,2,4-trichlorobenzene,
respectively, typical of spin glass behavior. Since [TCNE]•- is only disordered for the PhMe solvate and all five
solvates exhibit spin glass behavior, the spin glass behavior cannot be attributed to this disorder.
Introduction
first predicted to consist of π/π [TCNE]•-/[MnTPP]+ interac-
tions,9 akin to [Fe(C5Me5)2][TCNE];4b however, the structure
The study of magnetically ordered molecule-based materials
is a growing area of contemporary chemistry, and establishment
of structure-function relationships is critical to its advance-
ment.2,3 [Fe(C5Me5)2]•+[TCNE]•- (TCNE ) tetracyanoethylene)
was the first organic-containing molecule-based magnet char-
acterized and has an ordering temperature Tc of 4.8 K.4
Improvements led to the discovery of V(TCNE)x‚y(solvent), the
first room-temperature organic magnet (Tc ≈ 400 K),5 and later
the magnetically ordered Mn, Fe, Ni, and Co analogues.6 More
recently, studies have focused on a class of metallomacrocycle/
[TCNE]•--based magnets exemplified by [MnTPP][TCNE]‚
2PhMe, 1a (H2TPP ) meso-tetraphenylporphyin).3a,7,8 1a was
(2) (a) Proceedings of the Conference on Ferromagnetic and High Spin
Molecular Based Materials; Miller, J. S., Dougherty, D. A., Eds.;
Molecular Crystals and Liquid Crystals 176; Gordon and Breach
Science: New York, 1989. (b) Proceedings of the Conference on
Molecular Magnetic Materials; Gatteschi, D., Kahn, O., Miller, J. S.,
Palacio, F., Eds.; NATO ARW Molecular Magnetic Materials; Kluwer
Academic: Dordrecht, 1991; E198. (c) Proceedings of the Conference
on the Chemistry and Physics of Molecular Based Magnetic Materials;
Iwamura, H., Miller, J. S., Eds.; Molecular Crystals and Liquid Crystals
232/233; Gordon and Breach Science: New York, 1993, 232/233. (d)
Proceedings of the Conference on Molecule-based Magnets; Miller,
J. S., Epstein, A. J., Eds.; Molecular Crystals and Liquid Crystals 271-
274; Gordon and Breach Science: New York, 1995. (e) Proceedings
of the Conference on Molecular-Based Magnets; Itoh, K., Miller, J.
S., Takui, T., Eds.; Molecular Crystals and Liquid Crystals 305/306;
Gordon and Breach Science: New York, 1997. (f) Turnbull, M. M.,
Sugimoto, T., Thompson, L. K., Eds.; American Chemical Society
Symposium Series; American Chemical Society: Washington, DC,
1996; Vol. 644. (g) Proceedings of the 6th Conference on Molecular-
Based Magnets; Kahn, O., Ed.; Molecular Crystals and Liquid Crystals
334/335; Gordon and Breach Science: New York, 1999.
(1) (a) University of Utah. (b) Current address: Osaka University. (c)
Department of Chemistry, The Ohio State University. (d) Current
address: Milliken Research Corporation, Spartanburg, SC 29304. (e)
Department of Physics, The Ohio State University. (f) University of
Delaware.
10.1021/ic0011922 CCC: $20.00 © 2001 American Chemical Society
Published on Web 03/13/2001

1916 Inorganic Chemistry, Vol. 40, No. 8, 2001
Hibbs et al.
1
determination7a revealed that S ) /2 [TCNE]•- is trans-µ-N-
Table 1. Infrared TCNE Stretching Data for [MnTPP][TCNE]‚xS
σ-bound to S ) 2 MnIII forming 1-D ‚‚‚D+A•-D+A•-‚‚‚ (D )
MnTPP; A ) TCNE) chains. This material is a ferrimagnet with
a critical temperature Tc of 18 K, as determined by extrapolating
the maximum slope of the low-field M(T) curve to M ) 0.7b,8
More recently, it is common to report Tc as the maximum in
the real component of the low-frequency (10 Hz) ac susceptibil-
ity.3a Further study of 1a revealed signs of complicated magnetic
dynamics suggestive of disorder that was not evident in the
crystal structure.10
ν
CN, cm-1 (m)
ν
CN, cm-1 (s)
compound
[TCNE]0 15
solvated
2259
R
â
solvated
R
â
2221
2143
2069
[Fe(C5Me5)2]•+[TCNE]•- 4b,a 2183
[TCNE]2- 15
2104
1a‚2PhMe
2195 2192 2192 2148 2147 2147
2193 2191 2192 2131 2134 2134
2196 2193 2192 2132 2135 2133
2200 2193 2193 2129 2135 2134
2195 2193 2192 2136 2135 2135
1b‚1,2-C6H4Me2
1c‚3(1,2-C6H4Cl2)
1d‚2(1,3-C6H4Cl2)
1e‚2(1,2,4-C6H3Cl3)
During the study of 1a, numerous other solvates as well as
their desolvated forms were prepared and studied. These
included [MnTPP][TCNE]‚xS [S ) PhMe, 1a (x ) 2); 1,2-
C6H4Me2, 1b (x ) 1); 1,2-C6H4Cl2, 1c (x ) 3); 1,3-C6H4Cl2,
1d (x ) 2); and, 1,2,4-C6H3Cl3, 1e (x ) 2)], where x is
determined from single-crystal X-ray analyses and is the number
of solvates when all solvent sites in the unit cell are fully
occupied. Herein, we report the structural and magnetic
consequences arising from these solvent substitutions as well
as from desolvation for this family of magnets.
a Unbound [TCNE]•-
.
porphyrin, to remove paramagnetic impurities. The [MnIIITPP][OAc]
salts were subsequently reduced to MnIITPP as MnIITPP(py) (py )
pyridine) by NaBH4, utilizing a literature method.11b TCNE was
obtained as a gift from O. Webster and was sublimed prior to use.
[MnTPP][TCNE]‚xS (S ) solvent) was prepared according to the
general method as reported by Summerville et al. with the exception
that the material was not dried in vacuo as described.9 All physical
measurements of solvated materials were performed immediately after
collecting the shiny black crystals on a glass frit in order to minimize
solvent loss.
Experimental Section
Desolvation was effected by either vacuum thermolysis (R method)
of freshly prepared 1 in a glass vial heated in a tube furnace under
dynamic vacuum at 175 °C for 12 h or by heating freshly prepared 1
(ca. 25 mg) to reflux (â method) in 7 mL of either n-octane (bp ) 127
°C) (âo) or n-dodecane (bp ) 216 °C) (âd) for 10 min followed by
filtering the hot mixture through a medium glass-frit funnel and
collecting the insoluble desolvated 1. Complete desolvation was
confirmed by thermogravimetric analysis-mass spectroscopy (TGA/
MS).
Synthesis. All manipulations involving [TCNE]•- were carried out
under nitrogen using standard Schlenck techniques or in a Vacuum
Atmospheres DriLab. Solvents used for the preparation of the [TCNE]•-
salts were predried and distilled from appropriate drying agents. H2-
TPP was prepared according to the Alder-Longo method.11a [MnIII-
TPP][OAc] complexes were prepared from H2TPP and Mn(OAc)2‚
4H2O;11b the Mn(OAc)2‚4H2O was predissolved in N,N-dimethyl-
formamide (DMF) and filtered before addition to the free-base
[MnTPP][TCNE]‚xPhMe, 1a, was prepared as previously reported
by reaction of MnIITPPpy and TCNE in toluene; however, the sample
was not dried until R and â desolvation.9 IR (cm-1) at room
temperature: 2195 (m), 2148 (s) for νCN. TGA/MS: 17.9% weight loss
at 108.5 °C corresponding to x ) 1.88. 1a was thermally annealed
according to the R and âo methods, and the νCN IR data are reported in
Table 1.
[MnTPP][TCNE]‚x(1,2-C6H4Me2), 1b, was prepared as described
above using 80.3 mg (0.11 mmol) of MnIITPPpy and 51.2 mg (0.40
mmol) of TCNE, both dissolved in 10 mL of boiling o-xylene and
filtered before addition. IR (cm-1) at room temperature: 2193 (m),
2131 (s) for νCN. Anal. Calcd (obsd) for [MnTPP][TCNE]‚1,2-C6H4-
Me2 [C58H38N8Mn]: C 77.24 (77.43%), H 4.25 (4.51%), N 12.43
(12.27%). TGA/MS: 13.93% weight loss at 164 °C corresponding to
x ) 1.04. Desolvation was performed according to the R and âd
methods, and the νCN IR data are reported in Table 1.
(3) Reviews. (a) Miller, J. S.; Epstein, A. J. Chem. Commun. 1998, 1319.
(b) Ovcharenko, V. I.; Sagdeev, R. Z. Russ. Chem. ReV. 1999, 68,
345. (c) Plass, W. Chem.-Ztg. 1998, 32, 323. (d) Miller, J. S.; Epstein,
A. J. Chem. Eng. News 1995, 73 (40), 30. (e) Miller, J. S.; Epstein,
A. J. Angew. Chem., Int. Ed. Engl. 1994, 33, 385. (f) Kinoshita, M.
Jpn. J. Appl. Phys. 1994, 33, 5718. (g) Miller, J. S.; Epstein, A. J.
AdV. Chem. Ser. 1995, 245, 161. (h) Caneschi, A.; Gatteschi, D. Progr.
Inorg. Chem. 1991, 37, 331. (i) Buchachenko, A. L. Russ. Chem. ReV.
1990, 59, 307. (j) Kahn, O. Struct. Bonding 1987, 68, 89. (k) Caneschi,
A.; Gatteschi, D.; Sessoli, R.; Rey, P. Acc. Chem. Res. 1989, 22, 392.
(l) Gatteschi, D. AdV. Mater. 1994, 6, 635. (m) Miller, J. S.; Epstein,
A. J.; Reiff, W. M. Acc. Chem. Res. 1988, 21, 114. (n) Miller, J. S.;
Epstein, A. J.; Reiff, W. M. Science 1988, 240, 40. (o) Miller, J. S.;
Epstein, A. J.; Reiff, W. M. Chem. ReV. 1988, 88, 201. (p) Miller, J.
S.; Epstein, A. J. In New Aspects of Organic Chemistry; Yoshida, Z.,
Shiba, T., Oshiro, Y., Eds.; VCH Publishers: New York, 1989; Vol.
237. (q) Kahn, O. Molecular Magnetism; VCH Publishers: New York,
1993.
[MnTPP][TCNE]‚x(1,2-C6H4Cl2), 1c, was prepared as described
above using 80.8 mg (0.11 mmol) of MnIITPPpy and 28.0 mg (0.22
mmol) of TCNE, both dissolved in 10 mL of boiling o-dichlorobenzene
and filtered before addition. IR (cm-1) at room temperature: 2196 (m),
2132 (s) for νCN. Anal. Calcd (obsd) for [MnTPP][TCNE]‚2.28(1,2-
C6H4Cl2) [C63.675H37.125Cl4.56N8Mn]: C 67.64 (67.74%), H 3.31 (3.58%),
N 9.91 (9.67%). TGA/MS: 32.23% weight loss at 153 °C corresponding
to x ) 2.17. Desolvation was performed according to the R and âo
methods, and the νCN IR data are reported in Table 1.
[MnTPP][TCNE]‚x(1,3-C6H4Cl2), 1d, was prepared as described
above using 96.8 mg (0.13 mmol) of MnIITPPpy and 55.3 mg (0.43
mmol) of TCNE, both dissolved in 10 mL of boiling m-dichlorobenzene
and filtered before addition. IR (cm-1) at room temperature: 2200 (w),
2129 (s) for νCN. Calcd (obsd) for [MnTPP][TCNE]‚2(1,3-C6H4Cl2)
[C62H36Cl4N8Mn]: C 68.33 (68.35%), H 3.33 (3.44%), N 10.28
(10.31%). TGA/MS: 27.43% weight loss at 123 °C corresponding to
(4) (a) Miller, J. S.; Calabrese, J. C.; Epstein, A. J.; Bigelow, R. W.; Zhang,
J. H.; Reiff, W. M. J. Chem. Soc., Chem. Commun. 1986, 1026. (b)
Miller, J. S.; Calabrese, J. C.; Rommelmann, H.; Chittipeddi, S.; Zhang,
J. H.; Reiff, W. M.; Epstein, A. J. J. Am. Chem. Soc. 1987, 109, 769.
(5) (a) Manriquez, J. M.; Yee, G. T.; McLean, R. S.; Epstein, A. J.; Miller,
J. S. Science 1991, 252, 1415. (b) Epstein, A. J.; Miller, J. S. In
Conjugated Polymers and Related Materials: The Interconnection of
Chemical and Electronic Structure, Proceedings of Nobel Symposium
NS-81; Oxford University Press: Oxford, U.K., 1993; p 475. Chim.
Ind. 1993, 75, 185, 257. (c) Miller, J. S.; Yee, G. T.; Manriquez, J.
M.; Epstein, A. J. In Conjugated Polymers and Related Materials:
The Interconnection of Chemical and Electronic Structure, Proceedings
of Nobel Symposium NS-81; Oxford University Press: Oxford, U.K.,
1993; p 461. Chim. Ind. 1992, 74, 845.
(6) Zhang, J. H.; Ensling, J.; Ksenofontov, V.; Gu¨tlich, P.; Epstein, A. J.;
Miller, J. S. Angew. Chem., Int. Ed. 1998, 37, 657.
(7) (a) Miller, J. S.; Calabrese, J. C.; McLean, R. S.; Epstein, A. J. AdV.
Mater. 1992, 4, 498. (b) Zhou, P.; Morin, B. G.; Epstein, A. J.;
McLean, R. S.; Miller, J. S. J. Appl. Phys. 1993, 73, 6569.
(8) Brinckerhoff, W. B.; Morin, B. G.; Brandon, E. J.; Miller, J. S.; Epstein,
A. J. J. App. Phys. 1996, 79, 6147.
(11) (a) Adler, A. D.; Longo, F. R.; Finarelli, J. D.; Goldmacher, J.; Assour,
J.; Korsakoff, L. J. Org. Chem. 1967, 32, 476. (b) Jones, R. D.;
Summerville, D. A.; Basolo, F. J. Am. Chem. Soc. 1978, 100, 446.
(c) Brandon, E. J.; Rittenberg, D. K.; Arif, A. M.; Miller, J. S. Inorg.
Chem. 1998, 37, 3376.
(9) Summerville, D. A.; Cape, T. W.; Johnson, E. D.; Basolo, F. Inorg.
Chem. 1978, 17, 3297.
(10) Wynn, C. M.; Gˆırtu, M.; Brinckerhoff, W. B.; Sugiura, K.-i.; Miller,
J. S.; Epstein, A. J. Chem. Mater. 1997, 9, 2156.

[MnTPP]+[TCNE]•-‚x(solvent)
Inorganic Chemistry, Vol. 40, No. 8, 2001 1917
Table 2. Summary of the Crystallographic Data for [MnTPP][TCNE]‚xS
S, x
PhMe, 2 (1a)
1,2-C6H4Me2, 1 (1b)
1,2-C6H4Cl2, 3 (1c)
1,3-C6H4Cl2, 2 (1d)
1,2,4-C6H3Cl3, 2 (1e)
formula
fw
space group
a, Å
b, Å
c, Å
C64H44MnN8
980.05
P1h
10.116 (1)
11.008 (1)
12.489 (2)
108.15 (1)
98.04 (1)
67.92(1)
1
1224.2
3.07
1.329
0.0569
C58H38MnN8
901.94
P1h
9.261 (3)
10.218 (2)
13.294 (3)
98.88 (2)
94.63 (2)
110.64 (2)
1
1150.71
3.212
1.301
0.0392
0.0571
C68H40Cl6MnN8
1236.79
P1h
9.572 (5)
12.879 (7)
12.949 (6)
92.60 (4)
106.11 (4)
104.67 (4)
1
1472.2
5.47
1.395
0.0543
C62H36Cl4MnN8
1089.72
P21/a
9.202(2)
25.6917(8)
10.686(3)
90
94.9635(17)
90
2
C62H34Cl6MnN8
1155.04
P1h
9.588 (1)
10.885 (9)
14.475 (1)
106.912 (7)
99.137 (7)
107.511 (0)
1
R, deg
â, deg
γ, deg
Z
V, Å3
µ, cm-1
2516.90
5.26
1.438
1327.17
6.01
1.450
F
calcd, g/cm3
R(F)a
0.0529
0.0403
Rw(F)b
Rw(F2)c
temp, °C
λ, Å
0.0885
0.1557
25
0.710 73
0.1088
-80
0.710 73
0.0970
25
0.710 73
-100
0.710 73
-80
0.710 73
a ∑(|Fo| - |Fc|)/∑|Fo|. b {∑w[Fo - Fc]2/∑w[Fo]2}1/2
.
c {∑w[Fo - Fc2]2/∑w[Fo ]2}1/2
.
2
2
x ) 2.05. Desolvation was performed according to the R and âo
methods, and the νCN IR data are reported in Table 1.
X-ray Structure Determination. The structure of 1a was solved
with X-ray diffraction intensities collected on a Syntex R3 diffracto-
meter using Mo KR radiation at 173 K7a (Table 2). A hemisphere (-12
e h e 13; -13 e k e 14; -16 e l e 0) of intensity data was collected
with 4.1° e 2θ e 55.0°, yielding a total of 5928 reflections of which
5022 were unique [3582 reflections g 2 σ(I)]. Crystal stability was
monitored by two standard reflections, which showed e1% variation
throughout the data collection. No absorption correction was performed.
These data were rerefined against the structure amplitudes, F, using
the program SHELXL9312a on a Silicon Graphics Iris4d Indigo.
Hydrogen atoms were located by difference Fourier mapping, and their
positional and thermal parameters were refined except for the methyl
hydrogens of the toluene, which were restrained to an ideal stereo-
chemistry.
[MnTPP][TCNE]‚x(1,2,4-C6H3Cl3), 1e, was prepared as described
above using 42.3 mg (0.06 mmol) of MnIITPPpy and 20.3 mg (0.16
mmol) of TCNE, both dissolved in 10 mL of boiling 1,2,4-trichlo-
robenzene and filtered before addition. IR (cm-1) at room tempera-
ture: 2195 (m), 2136 (s) for νCN. TGA/MS: 32.1% weight loss at 150
°C corresponding to x ) 2.07. Desolvation was performed according
to the R and âd methods, and the νCN IR data are reported in Table 1.
Table 1 reports the major IR peaks for the single-phase solvated crystals
used for structure analysis. In a typical preparation 1e has minor IR
peaks at 2202 (vw), 2182 (w), 2162 (w), and 2146 (m) in addition to
major peaks at 2195 (m) and 2136 (s). This is possibly due to rapid
partial desolvation or other structural defects in the lattice.
The refinement converged to an R(F) of 6.10% for data with I >
2 σ(I) and 9.25% for all data using 412 atomic parameters refined with
anisotropic thermal parameters. The asymmetric unit comprises7a one-
half [MnTPP]+, one-half [TCNE]•-, and one toluene solvate. However,
the difference density map (Figure 1) showed three significant peaks
(0.99, 0.34, and 0.33 e-/Å3) in the same plane as the [TCNE]•- into
which a rotated minor form of [TCNE]•- could be fit with two peaks
nearly coincident with existing atoms. The model was rerefined to
include an independent minor orientation’s positional and isotropic
thermal parameters with the sum of the occupancies constrained to unity
and corresponding major and minor bond lengths and angles restrained
to be the same value (Table 3).
The final model, including the minor [TCNE]•- orientation, consists
of 431 refined parameters with 10 restraints converging to an R(F) of
0.0569 for the 3584 reflections greater than 2 σ(I) and 0.0885 for all
5022 reflections. The final weighting scheme used is 1/(σ2(F2) +
0.8336P), where P ) [max(F2,0) + 2Fc2]/3, and the final goodness of
fit is 1.094. The highest features remaining in the difference synthesis,
0.20-0.28 e-/Å3, are in the bonds of the porphyrin ring. A Hamilton’s
R-factor test12b demonstrates that the modeled disorder is statistically
significant with a probability of 99.5%.
Physical Methods. The 2-300 K dc magnetic susceptibility was
determined on the equivalent of a Quantum Design MPMS-5XL 5 T
SQUID as previously described.11c Desolvated samples were loaded in
gelatin capsules, and solvated samples were loaded in an airtight Delrin
holder packed with oven-dried quartz wool, and a minimal amount of
solvent was added to prevent movement of the sample in the holder at
low temperature when the solvent became frozen. Delrin holders were
loaded and sealed within the first 2 min of filtering the solution in
order to minimize solvent loss. The samples were cooled in zero applied
field with data then taken upon warming. In addition to correction for
the diamagnetic contribution from the sample holder, core diamagnetic
corrections of -430, -60, -65.9, -77.8, -83.2, -84.3, and -106.5
× 10-6 emu/mol were used for MnTPP, TCNE, PhMe, 1,2-C6H4Me2,
1,2-C6H4Cl2, 1,3-C6H4Cl2, and 1,2,4-C6H3Cl3, respectively. High-field
((120 000 Oe) hysteresis experiments were performed on an Oxford
Instruments vibrating sample magnetometer. Infrared spectra ((1 cm-1
)
were obtained on a Bio-Rad FT-40 spectrophotometer in mineral oil
mulls between NaCl plates. The IR spectrometer was equipped with a
homemade heating system consisting of four button heaters controlled
with a type K Ni-Cr vs Ni-Al thermocouple and an Omega CN76000
temperature readout to enable the measurement of the temperature
dependence of the IR spectra from room temperature to 160 °C. The
temperature was determined to be ca. 4.5 K cooler at the sample via a
Fluke 80T IR temperature probe. The absolute temperatures are
estimated to be (3 K. The thermal properties were studied on a TA
Instruments model 2050 thermogravimetric analyzer (TGA) equipped
with a TA-MS Fison triple-filter quadrupole mass spectrometer to
identify gaseous products with masses less than 300 amu. The TGA
was located in a Vacuum Atmospheres DriLab under argon to protect
air- and moisture-sensitive samples. Samples were placed in an
aluminum pan and heated at 20 °C/min under a continuous 10 mL/
min nitrogen flow. Elemental analyses were performed by Atlantic
Microlabs, Norcross, GA.
Single crystals of 1b-e were grown from their respective solvents,
and their unit cells were determined on a CAD4 diffractometer using
Mo KR radiation. The crystallographic data are summarized in Table
2. Like 1a, 1b, 1c, and 1e belong to the P1h space group determined
from the lack of systematic absences and subsequent least-squares
(12) (a) Sheldrick, G. SHELXL93, Program for the Refinement of Crystal
Structures; University of Go¨ttingen: Go¨ttingen, Germany, 1994. (b)
Hamilton, W. C. Acta Crystallogr. 1965, 18, 502. The R-factor ratio,
R, for the current model has Rb)21,f)3151 ) 1.0510 for I > 2 σ(I) and
Rb)21,f)4589 ) 1.0510 for all data. In order for the disorder to be
statistically significant, R must be greater than the extrapolated tabular
value from the above reference, 1.0066 and 1.0042, respectively.

1918 Inorganic Chemistry, Vol. 40, No. 8, 2001
Hibbs et al.
case. The products appeared as shiny black needles or plates.
X-ray quality crystals were grown from solutions approximately
one-tenth as concentrated as noted in the Experimental Section.
In the case of 1,3-C6H4Cl2, the starting materials were each
dissolved in 5 mL of dichloromethane, filtered, and then reacted,
and 1 mL of 1,3-C6H4Cl2 was added to the resulting solution.
The porphyrin sponges prefer aromatic solvents with respect
to aliphatic solvents, and hence, the 1,3-C6H4Cl2 solvate was
isolated. The solvent content x varies with structure, and the
values reported herein are determined by the structure when all
solvent sites are occupied. The actual content may be less than
the x crystallographically determined. For samples studied by
magnetic susceptibility, the actual value of x was determined
from TGA analyses and used in analyses of the data. To ensure
minimal solvent loss, and hence determination of the intrinsic
physical properties of the fully solvated species, all studies were
performed immediately after collecting the crystals. In addition
to 1a-e, [MnTPP][TCNE]‚xS, compounds with S being chlo-
robenzene, m- and p-xylene, mesitylene, 1,2,4-trimethylbenzene,
and 1-chloronaphthalene, were also prepared, but crystals
suitable for single-crystal X-ray analyses could not be obtained.14
The magnetic properties of these materials were not studied
because a structure-function relationship could not be devel-
oped.
Figure 1. Difference electron density map in the plane of [TCNE]•-
.
Contours are at 0.1 e- Å3 (solid) beginning at zero e-/Å3.
refinement. For 1d, the P21/a space group was determined by the same
method. During data collection the intensities of several representative
reflections were measured as a check on crystal stability; there was no
loss of intensity during data collection. Equivalent reflections were
merged, and only those for which Io is greater than 2 σ(I) were included
in the refinement, where σ(Fo)2 is the standard deviation based on
counting statistics. Lorentz and polarization corrections and an empirical
absorption correction based on a series of ψ scans were applied to the
data. Intensities of equivalent reflections were averaged. The structures
were solved by the standard heavy-atom techniques with the MOLEN/
VAX package. Non-hydrogen atoms were refined with anisotropic
thermal parameters using SHELXL-93.12a The nonhydrogen atoms were
refined with anisotropic displacement coefficients U(H) ) 1.2U(C) or
1.5U(Cmethyl), and their coordinates were allowed to ride on their
respective carbons. The final cycle of full-matrix least-squares refine-
ment for 1b was based on 3003 observed reflections [Io > 2 σ(I)] and
313 parameters and converged with unweighted and weighted agree-
ment factors R(F) ) 0.0392 and Rw(F) ) 0.0571, respectively. The
final cycle of full-matrix least-squares refinement for 1c was based on
3834 observed reflections [Io > 2 σ(I)] and 364 parameters and
converged with unweighted and weighted agreement factors R(F) )
0.0543 and Rw(F2) ) 0.1557, respectively. The final cycle of full-matrix
least-squares refinement for 1d was based on 7968 observed reflections
[Io > 2 σ(I)] and 412 parameters and converged with unweighted and
weighted agreement factors R(F) ) 0.0529 and Rw(F2) ) 0.1088,
respectively. The final cycle of full-matrix least-squares refinement for
1e was based on 3381 observed reflections [Io > 2 σ(I)] and 349
parameters and converged with unweighted and weighted agreement
factors R(F) ) 0.0403 and Rw(F2) ) 0.0970, respectively. Solvent
disorder was found in 1b and 1c. The 1,2-xylene solvent is present in
a static 50% orientational disorder for 1b, and one of the 1,2-
dichlorobenzenes is disordered evenly over an inversion center for 1c.
Disorder in any form was not observed for 1d and 1e. [TCNE]•-
disorder was only observed for 1a.
Infrared Spectra. The νCN stretching frequencies for 1a-e
are 2196 ( 4 (m) and 2139 ( 10 (s) cm-1. Upon desolvation
the νCN’s occur at 2192 ( 1 (m) and 2134 ( 1 (s) except for
desolvated 1a, which occurs at 2147 cm-1 (Table 1). These
values are inconsistent with [TCNE]0 [2259 (s), 2221 (m) cm-1],
unbound [TCNE]•- [2183 (s), 2143 (m) cm-1], and [TCNE]2-
[2104 (s), 2069 (m) cm-1 15 (Table 1). However, the solvated
]
and desolvated materials displayed shifts in the cyano stretching
frequencies with respect to unbound [TCNE]•-,15 suggesting
coordination and reduction of TCNE. Variable temperature IR
(VTIR) spectra taken to 160 °C for 1a-e show a (4 cm-1 shift
in the low-frequency cyano stretch upon heating. When 1b-d
are heated to 160 °C, the lowest low-frequency cyano stretch
shifts to higher frequencies, while for 1a and 1e the high-
frequency cyano stretch shifts to lower frequencies. In each case,
the high-frequency cyano stretch shifts to 2189 cm-1 at 160
(13) (a) Goldberg, I.; Krupitsky, H.; Stein, Z.; Hsiou, Y.; Strouse, C. E.
Supramol. Chem 1995, 4, 203. (b) Krupitsky, H.; Stein, Z.; Goldberg,
I. J. Inclusion Phenom. Mol. Recognit. Chem. 1995, 20, 211. (c)
Goldberg, I. Mol. Cryst. Liq. Cryst. 1996, 278, 767. (d) Byrn, M. P.;
Curtis, C. J.; Hsiou, Y.; Kahn, S. I.; Sawin, P. A.; Tendick, S. K.;
Terzis, A.; Strouse, C. E. J. Am. Chem. Soc. 1993, 115, 9480. Byrn,
M. P.; Curtis, C. J.; Hsiou, Y.; Khan, S. I.; Sawin, P. A.; Terzis, A.;
Strouse, C. E. ComprehensiVe Supramolecular Chemistry; Atwood,
J. L., Davies, J. E. D., MacNicol, D. D., Vogtle, F., Eds.; Pergamon:
New York, 1996; Vol. 6, p 715.
(14) [MnTPP][TCNE]‚C6H5Cl. IR (cm-1) at room temperature: 2197 (m),
2155 (s), and 2137 (s) for νCN. Anal. Calcd (obsd) for C56H33ClN8-
Mn: C 74.05 (73.94%), H 3.66 (3.82%), N 12.34 (12.26%). [MnTPP]-
[TCNE]‚1.56(1,3-C6H4Me2). IR (cm-1) at room temperature: 2194
(m), 2147 (s), and 2138 (s) for νCN. Anal. Calcd (obsd) for
C
62.48H43.6N8Mn: C 78.06 (78.00%), H 4.57 (4.74%), N 11.66
(11.66%). [MnTPP][TCNE]‚0.76(1,4-C6H4Me2). IR (cm-1) at room
temperature: 2192 (m) and 2134 (s) for νCN. Anal. Calcd (obsd) for
C
56.08H35.6N8Mn: C 76.85 (76.43%), H 4.09 (4.37%), N 12.78
Results and Discussion
(12.51%). [MnTPP][TCNE]‚0.96(1,2,4-C6H3Me3). IR (cm-1) at room
temperature: 2195 (m) and 2138 (s) for νCN. Anal. Calcd (obsd) for
Synthesis. Typical of metalloporphyrins, [MnTPP][TCNE]
forms a plethora of solvates referred to as porphyrin sponges.13
[MnTPP][TCNE]‚xS compounds [S ) PhMe, 1a (x ) 2); 1,2-
C6H4Me2, 1b (x ) 1); 1,2-C6H4Cl2, 1c (x ) 3); 1,3-C6H4Cl2,
1d (x ) 2); and 1,2,4-C6H3Cl3, 1e (x ) 2)] were isolated from
the reaction of MnIITPPpy and TCNE, with each being dissolved
upon refluxing in the appropriate solvent (S) and filtered prior
to reaction. A 2- to 4-fold excess of TCNE was used in each
C58.64H39.62N8Mn: C 77.30 (76.90%), H 4.37 (4.56%), N 12.30
(12.01%). [MnTPP][TCNE]‚0.32(1,3,5-C6H3Me3). IR (cm-1) at room
temperature: 2196 (m), 2149 (s), and 2138 (m) for νCN. Calcd (obsd)
for C52.7H31.6N8Mn: C 76.10 (75.94%), H 3.83 (4.03%), N 13.47
(13.34%). [MnTPP][TCNE]‚1.1(1-chloronaphthalene). IR (cm-1) at
room temperature: 2197 (m) and 2142 (s) for νCN. Calcd (obsd) for
C61H35.7Cl1.1N8Mn: C 75.17 (74.77%), H 3.69 (3.81%), N 11.50
(11.54%).
(15) Dixon, D. A.; Miller, J. S. J. Am. Chem. Soc. 1987, 109, 3656.

[MnTPP]+[TCNE]•-‚x(solvent)
Inorganic Chemistry, Vol. 40, No. 8, 2001 1919
Table 3. Summary of Key Bond Distances and Separations for [MnTPP][TCNE]‚xS
1a,a major orientation
1a,b minor orientation
1b
1c
1d
1e
C-C, Å
1.426(7)c
2.328(3)d
146.8(8)e
55.6(3)f
2.000
10.116
13.269
14.932
1.390(36)c
2.176(15)d
152.3(39)e
53.9(10)f
1.407(5)
2.288(2)
167.0(2)
82.3
1.411(8)
2.384(4)
127.1(3)
29.9
2.008
9.572
18.673
20.832
21.135
12.879
13.967
17.886
12.949
13.803
18.114
16.898
17.844
23.265
1.420(4)
2.3374(16)
121.38(14)
28.9
2.0084
9.202
10.686
13.485
14.692
13.645
17.084
17.575
1.395(7)
2.334(3)
130.2(3)
36.8
2.011
9.588
20.486
22.201
23.029
10.885
12.149
16.529
14.475
16.043
18.588
14.879
15.373
20.860
MnNCN, Å
MnNC, deg
dihedral angle, deg
MnNpor, Å
Mn‚‚‚Mn, Å, intrachain
Mn‚‚‚Mn, Å, interchain in-registry
2.010
10.218
16.804
19.026
20.287
9.261
11.111
16.028
13.294
15.446
17.974
15.440
15.576
21.046
Mn‚‚‚Mn, Å, interchain out-of-registry
11.006
11.823
13.838
15.667
12.498
17.136
18.603
18.498
19.047
23.109
17.528
13.139
10.201
13.630
18.979
interchain separation, Å, in-registry
interchain separation, Å, out-of-registry
16.763
8.669
13.130
14.642
18.670
12.457
12.441
16.657
10.646
12.846
16.684
20.463
10.381
14.293
14.323
a 83.3(4)%. b 16.7(4)%. c Orientation-weighted average is 1.420 Å. d Orientation-weighted average is 2.305 Å. e Orientation-weighted average is
147.9°. f Orientation-weighted average is 55.4°.
°C, an overall lowering of between 4 and 11 cm-1. The effects
observed in the VTIR are irreversible upon cooling to room
temperature. As a result, desolvation was examined using either
boiling octane (âo) or dodecane (âd) as a desolvation media.
For 1a-e, the high-frequency cyano stretch of the â-prepared
materials are equivalent at 2192 ( 1 cm-1, while the low-
frequency cyano stretch for 1b-e is 2134 ( 1 cm-1 and for 1a
is 2147 cm-1. The νCN temperature dependence of 1a was also
monitored upon cooling the sample to 11 K, with only a slight
reversible shift to 2148 (s) and 2196 (m) cm-1 being observed.
Thermolysis under reduced pressure (R method) produced
materials with physical properties nearly identical to those
obtained via heating to reflux in n-octane (â method), and
therefore, they were deemed to be the same â-phase materials.
In contrast, different phases were observed for [MnTClPP]-
[TCNE] [H2TClPP ) meso-tetrakis(4-chlorophenyl)porphyrin].11c
While the high- and low-frequency νCN absorptions of 1d are
respectively greater and less than those for 1a-c and 1e, the
ν
CN absorptions for the desolvated materials are equivalent ((2
cm-1) to those of the desolvated materials for 1b-e. 1a exhibits
a shift in the higher νCN absorption from 2195 to 2192.5 cm-1
for solvated and desolvated materials, respectively, while the
low νCN remains unchanged. The lower frequency νCN absorp-
tions are indicative of stronger back-bonding to MnIII and MnIII-
NtC bonding. Therefore, the frequency shifts are attributed to
an intrachain structural rearrangement accompanied by the loss
of solvent, which would decrease or increase Mn‚‚‚Mn intra-
chain separation and thereby decrease or increase the MnNC
angle and the dihedral angle formed from the [TCNE]•- and
MnN4 mean planes. Verification of this requires structural
information on the desolvated materials, which has been elusive.
Thus, attempts to identify a correlation between the νCN
absorptions and the magnetic properties were unsuccessful.
TGA/MS. Thermogravimetric analysis (TGA) of solvated
materials showed a rapid loss of solvent in the range 100-165
°C, which was confirmed by the appearance of each solvent in
the mass spectrum. In the materials in which solvent loss
occurred above the boiling point of n-octane (bp 127 °C), the
âd method was carried out by refluxing in n-dodecane (bp 216
Figure 2. Thermogravimetric analysis plots of [MnTPP][TCNE]‚x(1,3-
C6H4Cl2), 1d, revealing x ) 2.05.
°C). A second weight loss above 300 °C is attributed to the
decomposition of [TCNE]•- based on characteristic TCNE
fragments, e.g., (CN)2CC (76 amu), (CN)2C (64 amu), and (CN)2
(52 amu) observed in the mass spectrum. For 1b, 1d, and 1e
where the solvent content determined from TGA slightly
exceeded the value of 2.00 determined from single-crystal X-ray
diffraction studies, the differences are ascribed to weighing
errors of the small samples used in the TGA studies. A typical
TGA plot for 1d is shown in Figure 2.
Structure. The structures of 1a-e were determined by single-
crystal X-ray analyses, and their unit cell and metrical param-
eters are summarized in Table 2. Although not isomorphous,
compounds 1a-c and e are structurally similar to each,
belonging to the P1h space group with Z ) 1. 1d, in contrast,
belongs to the P21/a space group with Z ) 2. The P1h space
group has been observed for most members of the [Mn-

1920 Inorganic Chemistry, Vol. 40, No. 8, 2001
Hibbs et al.
Figure 4. View down the a axis normal to the parallel chains, showing
the interchain interactions between chains I, II, III, and IV for
[[MnTPP][TCNE]‚2PhMe, 1a.
The important inter- and intrachain distances and angles are
listed in Table 3. The intrachain Mn‚‚‚Mn separation is the
greatest for 1b at 10.218 Å and least for 1d at 9.202 Å, and
these two separations are associated with the largest and smallest
[TCNE]•--MnN4 dihedral angles (82.3° and 28.9°) and MnNCTC
-
NE angles (167.0 and 121.4°), respectively. The in-registry,
interchain Mn‚‚‚Mn separations range from 10.646 to 20.486
Å.
The ORTEP atom-labeling diagram for 1a is presented in
Figure 3a, while the ORTEP atom-labeling diagram for 1d (and
1b and 1c) is presented in Figure 3b. For 1a-e the [TCNE]•-
is uniformly trans-µ-N-σ-bound to two MnIII’s forming parallel
1-D ‚‚‚D+A•-D+A•-‚‚‚ (D ) MnTPP; A ) TCNE) chains along
the crystallographic a axis for 1a and 1c-e and along the b
axis for 1b. The chains of 1a (1b, 1c, 1e) lie parallel to each
other (Figure 4), with contacts between chains I-II, I-III, and
I-IV. Chains I-II and I-IV are out-of-registry, while chains
I-III are in-registry. The key inter- and intrachain interactions
can be found in Figure 5 and Table 3. Compounds 1b, 1c, and
1e are structurally similar, and their important intra- and
interchain separations are summarized in Table 3. The parallel
chains of 1d exhibit the closest contacts between chains I-II,
I-III, II-III, and I-IV (Figure 6). The presence of a mirror
plane between the two porphyrins in the unit cell results in in-
registry and a herringbone arrangement of chains I-III, while
I-II and I-IV ()II-III) are out-of-registry (Figure 7). The
shortest in-registry interchain Mn‚‚‚Mn separations range from
10.686 (1d) to 13.269 Å (1a), while the shortest out-of-registry
interchain Mn‚‚‚Mn separations range from 9.261 (1b) to 13.645
Å (1d) for 1a-e. Starting from an imaginary line through the
Mn’s within a chain, the interchain separations for the in-registry
chains range from 10.646 (1d) to 20.463 Å (1e) and the out-
of-registry separation range from 8.669 (1b) to 12.846 Å (1d).
For 1a-c,e with the P1h space group the shortest interchain
separation lies between out-of-registry chains. However, for 1d
with the P21/a space group, the shortest interchain separation
is between the in-registry chains. Additional distances and angles
are summarized in Table 3.
Figure 3. (a) ORTEP labeling diagram (30%) of [MnTPP][TCNE]‚
2PhMe, 1a. The [TCNE]•- minor form is shown in hollow bonds, while
the major form is shown in solid bonds. (b) ORTEP labeling diagram
(30%) of [MnTPP][TCNE]‚2(1,3-C6H4Cl2), 1d. The solvents have been
omitted for clarity.
(porphyrin)][TCNE]‚xS family; however, P21/a has been ob-
served for [MnTXPP][TCNE]‚2(CH2Cl2) (X ) Cl,11c Br17) and
[MnTF4OMePP][TCNE]‚2PhMe18 [H2TF4OMePP ) tetrakis-
(2,3,5,6-tetrafluoro-4-methoxyphenyl)porphyrin].
The structures of the [MnTPP]+ cations are indistinguishable
and consistent with other related [MnTPP]+ cations in the
literature.3a,7a,11c,13,17-21 In each case the [MnTPP]+ cation is
planar with a MnNpyrrole bond distance of 2.007 ( 0.008 Å.
The intramolecular [TCNE]•- bond distances and angles are
indistinguishable from each other and to related com-
pounds.4b,15,17-20 The central CC bonds are 1.41 ( 0.01 Å,
consistent with a bond order of 1.5 as expected for [TCNE]•-.15
Solvent fills the cavities between the parallel chains, and it
is anticipated that desolvation would allow a collapse of the
intrachain stacking and/or a decrease in interchain separation.13
(16) Brandon, E. J.; Kollmar, C.; Miller, J. S. J. Am. Chem. Soc. 1998,
120, 1822.
(17) Rittenberg, D. K.; Sugiura, K.-i.; Sakata, Y.; Mikami, S.; Epstein, A.
J.; Miller, J. S. AdV. Mater. 2000, 12, 126.
(18) Rittenberg, D. K.; Sugiura, K.-i.; Sakata, Y.; Guzei, I. A.; Rheingold,
A. L.; Miller, J. S. Chem.sEur. J. 1999, 5, 1874.

[MnTPP]+[TCNE]•-‚x(solvent)
Inorganic Chemistry, Vol. 40, No. 8, 2001 1921
Figure 5. View parallel to out-of-registry chains I-II (a) and I-IV (b) and in-registry chains I-III (c) [MnTPP][TCNE]‚2PhMe, 1a.
Disorder in the orientation of [TCNE]0 has been observed
for several electron and charge-transfer complexes22 in which
the [TCNE]0 lies between donors, thereby maximizing the
overlapping electron orbitals. Because [TCNE]0 and [TCNE]•-
are nearly square (4.06 Å × 4.39 Å), the disorder can be
described as (i) an approximate 180° rotation about the central
CC bond or (ii) a 90° rotation about an axis perpendicular to
the TCNE plane, passing through the center of the TCNE
molecule. Disorder for [TCNE]•- is not well documented, but
should be identical to that reported for TCNE, and may be
expected to occur when either TCNE or [TCNE]•- is trans-µ-
bound to metal ions. The preservation of the metal coordination,
as observed for 1a, supports that the rotation is best described
as an approximate 180° rotation about the central CC bond.
[TCNE]•- is 2-fold orientationally disordered in the structure
of [MnTPP][TCNE]‚2PhMe, 1a (Figure 3a). The major form
(filled bonds) has a refined occupancy of 83.3(4)%, while the
minor form (hollow bonds) has an occupancy of 16.7(4)%. The
central CC bond is 1.426(7) and 1.390(36) Å, respectively, for
the major and minor orientations characteristic of a 1.5 bond
(20) (a) Brandon, E. J.; Sugiura, K.-i.; Arif, A. M.; Liable-Sands, A.;
Rheingold, A. L.; Miller, J. S. Mol. Cryst. Liq. Cryst. 1997, 305, 269.
(b) Brandon, E. J.; Burkhart, B. M.; Rogers, R. D.; Miller, J. S.
Chem.sEur. J. 1998, 4, 1938. (c) Brandon, E. J.; Arif, A. M.; Burkhart,
B. M.; Miller, J. S. Inorg. Chem. 1998, 37, 2792. (d) Brandon, E. J.;
Arif, A. M.; Miller, J. S.; Sugiura, K.-i.; Burkhart, B. M. Crystal Eng.
1998, 1, 97. (e) Sugiura, K.-i.; Mikami, S.; Tanaka, T.; Sawada, M.;
Manson, J. L.; Miller, J. S.; Sakata, Y. Chem. Lett. 1997, 1071.
(19) (a) Bo¨hm, A.; Vazquez, C.; McLean, R. S.; Calabrese, J. C.; Kalm,
S. E.; Manson, J. L.; Epstein, A. J.; Miller, J. S. Inorg. Chem. 1996,
35, 3083. (b) Sugiura, K.-i.; Arif, A.; Rittenberg, D. K.; Schweizer,
J.; O¨ hrstrom, L.; Epstein, A. J.; Miller, J. S.; Chem.sEur. J. 1997, 3,
138.

1922 Inorganic Chemistry, Vol. 40, No. 8, 2001
Hibbs et al.
Figure 6. View down the a axis normal to the parallel chains, showing
the interchain interactions between chains I, II, III, and IV for [MnTPP]-
[TCNE]‚2(1,3-C6H4Cl2), 1d.
order and assignment of [TCNE]•-.15 The major and minor
[TCNE]•- orientations have MnN distances of 2.328(3) and
2.176 (15) Å and a MnNC angle of 146.8(8)° and 152.3(39)°
for the major and minor orientations, respectively. Hence, the
orientation-weighted average distance and angle are 2.305 Å
and 147.9°, respectively. The dihedral angle between the MnN4
and the major and minor [TCNE]•- orientation mean planes
are 55.6(3)° and 53.9(10)°, respectively, and the orientation-
weighted average dihedral angle is 55.4°. These Mn-N
distances are comparable to those of other structures3a,7a,11c,17-20
and consistent with that of a high-spin six-coordinated MnIII.
Magnetic Data. The 2-300 K magnetic susceptibility was
determined for 1a-e and their R and â desolvated phases, and
the data are summarized in Table 4. The 300 K effective
moments range from 4.3 to 5.5 µB, averaging 4.8 µB. Values
less that 5.20 µB are due to antiferromagnetic coupling evident
at room temperature, while values higher than 5.20 µB are likely
due to errors in molecular weight due to the specific degree of
solvation, or conversely, a consequence of interchain coupling,
strong coupling with Tmin . 300 K (Table 4). The susceptibili-
ties of 1a 1a-R, and 1a-â can be fit to a Curie-Weiss expression
ø [ 1/(T - θ] in two linear regions. The data above ∼250 K
for 1a and >200 K for 1a-R and 1a-â were fit with θ values of
-267, -123, and -40 K for 1a, 1a-R, and 1a-â, respectively,
indicative of dominant antiferromagnetic coupling. Below ∼200
K the data can be fit with effective θ (θ′)11c values of 44, 41,
and 49 K, respectively, indicative of longer range ferromagnetic
coupling. Likewise, 1b-R (and 1b-â) has θ and θ′ values of
-16 (-20) and 40 (52 K), respectively. The susceptibility of
the remaining 10 compounds, as typified by [MnTPP][TCNE]‚
x(1,3-C6H4Cl2), 1d (Figure 8), can be fit only in one region
Figure 7. View parallel to out-of-registry chains I-II (a) and I-IV
(II-III) (b) and in-registry chains I-III (c) for MnTPP][TCNE]‚2(1,3-
C6H4Cl2), 1d.
(21) (a) Day, V. W.; Sults, B. R.; Tasset, E. L.; Marianelli, R. S.; Boucher,
L. J. Inorg. Nucl. Chem. Lett. 1975, 11, 505. (b) Cheng, B.; Cukiernik,
F.; Fries, P.; Marchon, J.-C.; Scheidt, W. R. Inorg. Chem. 1995, 34,
4627. (c) Guildard, R.; Perie, K.; Barbe, J.-M.; Nurco, D. J.; Smith,
K. M.; Caemelbecke, E. V.; Kadish, K. M. Inorg. Chem. 1998, 37,
973. (d) Landrum, J. T.; Hatano, K.; Scheidt, W. R.; Reed, C. A. J.
Am. Chem. Soc. 1980, 102, 6729. (e) Hill, C. L.; Williamson, M. M.
Inorg. Chem. 1985, 24, 3024. (f) Fleischer, E. B. Acc. Chem. Res.
1970, 3, 105. (g) Scheidt, W. R.; Reed, C. A. Chem. ReV. 1981, 81,
543. (h) Turner, P.; Gunter, M. J.; Hambley, T. W.; White, A. H.;
Skelton, B. W. Inorg. Chem. 1992, 31, 2297.
(22) (a) Saheki, M.; Yamada, H.; Yoshioka, H.; Nakatsu, K. Acta
Crystallogr., Sect. B 1976, 32, 662. (b) Maverick, E.; Trueblood, K.
N.; Bekoe, D. A. Acta Crystallogr., Sect. B 1978, 34, 2777. (c) Larsen,
F. K.; Little, R. G.; Coppens, P. Acta Crystallogr., Sect. B 1975, 31,
430. (d) Bernstein, J.; Trueblood, K. N. Acta Crystallogr., Sect. B
1971, 27, 2078. (e) Lee, D.; Wallwork, S. C. Acta Crystallogr., Sect.
B 1978, 34, 3604. (f) Renault, A.; Cohen-Addad, C. Acta Crystallogr.,
Sect. C 1986, 42, 1529.
with θ′ ranging from 44 (1a) to 104 K (1e-â), indicative of
stronger 2-D and 3-D ferromagnetic coupling. The θ′ values
for 1d can be enhanced by as much as 58% upon desolvation.
However, decreases in θ′ values were observed for 1b-R and
1b-â. Hence, thermolysis can either increase or decrease the
magnitude of the θ′ values. Minima in the øT(T), i.e., Tmin, are
observed for 1x, 1x-R, and 1x-â (x ) a, b) between 191 and
292 K, again indicating dominant antiferromagnetic coupling.23
1d has a Tmin at 275 K, and the remaining solvated and
desolvated materials did not display a Tmin within the measured
temperature range, but Tmin is expected to occur above 300 K
and is estimated by the Seiden expression (vide infra).
The 1-D intrachain coupling, Jintra, of the [MnTPP][TCNE]
series was determined from modeling øT(T) with the Seiden24

[MnTPP]+[TCNE]•-‚x(solvent)
Inorganic Chemistry, Vol. 40, No. 8, 2001 1923
Table 4. Summary of Magnetic Properties for [MnTPP][TCNE]‚xS
µeff, µΒ
θ, K
θ′, K
Tmin, K
Jintra, Ka
Tc,b K
φb
ø′′ b
Hc (2 K), kOe
M(2K,5T), emu Oe/mol
1a
4.65
4.97
4.72
4.30
4.92
4.86
5.28
4.80
5.05
5.50
4.90
4.65
4.50
4.79
4.32
-267
44
41
49
52
40
52
72
75
74
55
75
87
77
74
104
191
195
263
292
247
-63
-65
7.8e
11.1e
11.7e
9.2
11.6
15.3
11.3
9.5
0.167
0.109
0.103
0.168
0.099
0.003
0.066
0.170
0.151
0.171
0.16
6.4
8.9
9.6
7.9
9.1
11.3
8.1
8.2
8.4
8.8
8.3
8.6
6.6
8.1
8.3
22.9
23.2
22.4
25.4
22.3
22.6
25.6
27.0
26.6
26.8
26.5
26.8
26.5
27.0
27.1
14 600
16 100
11 940
12 300
11 500
12 700
15 750
13 800
12 200
11 800
12 000
11 800
13 000
11 800
11 300
1a-R
1a-â
1b
1b-R
1b-â
1c
1c-R
1c-â
1d
1d-R
1d-â
1e
-123
-40
-88
c
-99
-16
-20
-86
276
-99
c
c
c
c
c
c
c
c
c
700d
720d
625d
275
-234
-240
-210
-100
-217
-295
-200
-250
-360
9.8
10.8
9.5
650d
880d
600d
750d
1080d
9.7
0.157
0.024
0.17
8.2f
9.4
1e-R
1e-â
9.4
0.181
a Based on best fit of the Seiden model (H ) -2JSa‚Sb). b Temperature at which ø′′(T) at 10 Hz is at its maximal value. c Not observed. d Determined
from the Seiden model (see text). e Reported as M(T), M f 0, Tc ) 23, 24, 25 K, respectively, for 1a, 1a-R, and 1a-â at 1000 G field. f An
additional peak is observed at 9.4 K attributed to the presence of a desolvated phase.
Figure 8. Magnetic moment µeff and inverse susceptibility ø-1 as a function of temperature for [MnTPP][TCNE]‚2(1,3-C6H4Cl2), 1d, showing µeff
values of 5.50 (b), 4.90 (~), and 4.65 (∆) µB and θ values of 55 (O), 75 (0) and 87 (]) K, respectively, for 1d, 1d-R, and 1d-â.
expression, assuming H ) -2JSa‚Sb and g ) 2. These observed
negative Jintra/kB values of -63, -99, -234, -100 (Figure 9),
and -200 K for 1a-e, respectively (Table 4), are indicative of
antiferromagnetic coupling. Upon thermolysis the θ′ values and
Jintra increase for 1d and 1e but in general may either increase
or decrease as noted for 1a-c (Table 4). In cases where Tmin
could not be observed, it was estimated from a fit to the Seiden
model and ranges from 625 (1c-â) to 1080 K (1e-â) (Figure 9).
The Seiden model may not be valid for temperatures below
ca. 50 K because of increased 3-D spin correlations or for
systems in which the intrachain coupling is weak, likely because
of next-nearest-neighbor interactions. 1a-R, 1a-â, 1b, and 1b-R
deviations from the Seiden model occur below 150 K. The shape
and position of the minimum in øT(T), when present, was
modeled to estimate Jintra from the Seiden expression. The
coupling of the [TCNE]•- with MnIII is antiferromagnetic, Jintra
< 0, because of direct exchange. Although an inverse correlation
between θ′ and either the MnNCTCNE angle or the MnN4 and
[TCNE]•- mean plane dihedral angle has been proposed for
several ditoluene solvates of aryl-substituted [MnTPP][TC-
NE]’s,16 this correlation does not exist for this group of materials
in which only the solvents are altered.
Isothermal field-dependent magnetization experiments reveal
magnetizations at 50 kOe at either 2 or 5 K in the range 11 300
(1e-â) to 16 100 (1a-R) emu Oe/mol (Table 4). For an
1
3
(23) (a) Coronado, E.; Drillon, M.; Georges, R. In Research Frontiers in
Magnetochemistry; O’Connor, C. J., Ed.; World Scientific: Singapore,
1993; p 26. (b) Beltran, D.; Coronado, E.; Drillon, M.; Georges, R.
Stud. Inorg. Chem. 1983, 3, 583. (c) Drillon, M.; Coronado, E.; Beltran,
D.; Georges, R. Chem. Phys. 1983, 79, 449. (d) Verdaguer, M.; Julve,
M.; Michalowicz, A.; Kahn, O. Inorg. Chem. 1983, 22, 2624. (e)
Drillon, M.; Gianduzzo, J. C.; Georges, R. Phys. Lett. 1983, 96A, 413.
(24) Seiden, J. J. Phys. Lett. 1983, 44, L947.
antiferromagnetic Stot ) 2 - /2 ) /2 system, the expected
saturation magnetization Ms is 16 755 emu Oe/mol, and for a
ferromagnetic Stot ) 2 + 1/2 ) 5/2 system, Ms is expected to be
27 925 emu Oe/mol. The magnitude of the magnetization at 5
T is lower than expected for an antiferromagnetically coupled
1
S ) 2 - /2 system; however, on the basis of the slope of the

1924 Inorganic Chemistry, Vol. 40, No. 8, 2001
Hibbs et al.
of 45% upon desolvation, shifting from 7.8 K to 11.1 and 11.7
K for the R and â forms, respectively, with similar results seen
for 1b,c.
Tc for 1a was previously reported as ∼13 K, using a scaling
analysis with the additional reporting of a “shoulder” at ∼7 K,
which was explained as typical “reentrant” spin glass behavior.8
Solvated 1a has a Tc of 7.8 K; however, desolvation increases
Tc to 11.4 ( 0.3 K. Hence, the earlier report of Tc ≈ 13 K and
a shoulder at ∼7 K is now attributed to the presence of solvated,
partially solvated, and desolvated 1a. Although the magnetic
behavior of solvated and desolvated phases are reported herein,
the magnetic behavior of partially solvated phases was not
specifically studied. In contrast to 1a, the Tc of 1c-d decreased
upon desolvation and the Tc for 1b,e increased upon desolvation.
Thus, the change in magnetic behavior upon desolvation cannot
be predicted a priori. 1e has two peaks in ø′(T), with the higher
temperature peak occurring at the same temperature as both the
R and â desolvates. Therefore, the multiphase behavior is
suspected to be a result of partial desolvation or a phase change.
The Tc reported in Table 4 is that represented by the lower
temperature peak, 8.2 K, because it is uncharacteristic of either
the R or â desolvates. Each reported ø′(T) peak is associated
with a ø′′(T) peak consistent with a ferrimagnetic state as
observed for [MnOEP][MnTPP][C4(CN)6] (H2OEP ) octaeth-
ylporphyrin).10,25 These peaks range from 6.4 (1a) to 11.3 K
(1b-â). Frequency (ω) dependencies of the ac susceptibility, φ
) ∆Tf/(Tf ∆log ω), for the solvates range from 0.035 to 0.171,
consistent with typical spin glasses26 as reported for related
materials.4a,20 Very small φ values (<0.001) are typically noted
for ferromagnets and weak ferromagnets. However, disordered
spins systems display φ values that typically range from 0.01
to 0.1 such as those found in the alloys of PdMn, φ ) 0.013,
and NiMn, φ ) 0.018, and the superparamagnet a-(Ho2O3)-
(B2O3), φ ) 0.28.26 One source of spin glass behavior is
disorder,26 and as noted earlier, although [TCNE]•- is disordered
in 1a, magnets not exhibiting structural disorder also exhibit
spin glass behavior. Hence, the genesis of spin glass behavior
for this family of magnets is still under study.
Figure 9. Seiden model fit to the øT(T) data [MnTPP]+[TCNE]•-‚2(1,3-
C6H4Cl2) and the desolvated phases with Jintra ) -100, -217, -295
K for 1d, 1d-R, and 1d-â, respectively (H ) -2JSa‚Sb).
Figure 10. Magnetization as a function of field ((12 T) at 2 (b) and
5 (+) K for R-[MnTPP][TCNE], 1e-R.
Magnetic ordering is also evident by the presence of hysteresis
loop. 1d has a high coercive field, Hcr, of 25.6 kOe at 2 K and
a remanent magnetization of 7380 emu Oe/mol, typical for [Mn-
(porphyrin)][TCNE] systems.17 (At 5 K, Hcr is 2300 kOe and
the remanent magnetization is 7150 emu Oe/mol.) In addition
to hysteresis, these materials show metamagnetic-like behavior
such that above a metamagnetic-like critical field Hc (Table 4)
the exchange pathways are sufficient to sustain spin alignment.
This interesting effect is the subject of ongoing studies.17
M(H) curves, saturation was not evident. On the basis of the
magnitude of M, these materials are ferrimagnets.7b Unlike at 5
K, the 2 K M(H) rises gradually with increasing applied field
before rising more rapidly in a manner typical of a metamagnet
with a large critical field Hc. Critical fields range from 22.9 to
26.8 kOe for 1a-e and range from 22.3 to 27.1 kOe for
desolvated 1a-e (Table 4). The genesis of this behavior is the
subject of ongoing studies17 and is representative of complex
competing magnetic behaviors being present in this class of
materials.
The M(H) for 1e-R was measured to (120 kOe, and even at
this high-applied magnetic field 1e-R did not saturate (Figure
10). [For this experiment on a different instrument, even at 120
kOe, its magnetization was 13 450 emu Oe/mol, which is 80%
of expectation.] The Hc for 1e-R is 26.0 kOe. The low
magnetization at high field and the lack of saturation is attributed
to canting and spin glass behavior (vide infra).
Magnetic ordering temperatures Tc for 1a-e were determined
by the initial peak in the in-phase susceptibility at 10 Hz, ø′(T).
Tc ranges from 7.8 to 11.3 K for 1a-e, with the 1,2-
dichlorobenzene-solvate (1c) having the greatest Tc. The 10 Hz
peak for 1d at 10.8 K decreases to 9.5 and 9.7 (∼11% decrease)
upon R and â desolvation, respectively, with similar effects
observed for 1c. The 10 Hz peak for 1a increases by an average
A relationship between structure and magnetic properties
continues to be elusive. The data herein and that previously
reported3a,16 demonstrate a trend toward higher θ′ being associ-
ated with a smaller dihedral angle formed from the [TCNE]•-
and MnN4 mean planes. These angles are influenced by the size
of the solvent molecule or the porphyrin ring substituents. The
Mn‚‚‚Mn distances are the closest in 1c-e, which have large
chlorinated solvents. In these cases, space is made for these
solvents through larger interchain distances. 1d, however, does
not follow this trend, having a lower than expected θ′, 55 K,
(25) Wynn, C. M.; Gˆırtu, M.; Miller, J. S.; Epstein, A. J. Phys. ReV. 1997,
B56, 14050. Wynn, C. M.; Gˆırtu, M.; Miller, J. S.; Epstein, A. J. Phys.
ReV. B. 1997, 56, 315.
(26) Mydosh, J. Spin Glasses: An Experimental Introduction; Taylor and
Francois: London, 1993; pp 64-76.

[MnTPP]+[TCNE]•-‚x(solvent)
Inorganic Chemistry, Vol. 40, No. 8, 2001 1925
for its small dihedral angle of 28.9°. This may be due to the
unique space group of 1d, P21/a, compared to the P1h space
group of 1a-c,e.
Monbu-sho fellowship program (Monbu-sho is the Ministry of
Education, Science and Culture, Japan).
Supporting Information Available: A summary of the crystal-
lographic data, tables of fractional coordinates and isotropic thermal
parameters, anisotropic thermal parameters, bond distances and angles
for [MnTPP][TCNE]‚xS ) PhMe (x ) 2) (1a), 1,2-C6H4Me2 (x ) 1)
(1b), 1,2-C6H4Cl2 (x ) 3) (1c), 1,2,4-C6H3Cl3 (x ) 2) (1d), and 1,3-
C6H4Cl2 (x ) 2) (1e). This material is available free of charge via the
Internet at http://pubs.acs.org.
Acknowledgment. The authors appreciate the constructive
comments and insight provided by E. J. Brandon, C. M. Wynn,
and M. Gˆırtu. We also thank J. L. Manson for high-field
hysteresis experiments performed in collaboration with Oxford
Instruments, using a 12 T MagLab vibrating sample magne-
tometer. We also gratefully acknowledge the support in part
from the National Science Foundation, Grants CHE9320478 and
CHE9730984. K.-i.S. also acknowledges support from the
IC0011922