Remarkable effects of axial π* coordination on the Cr-Cr quadruple bond in dichromium paddlewheel complexes

Article abstract of DOI:10.1021/ja9912675

It is well-known that donation of electron density into the σ* orbital of a Cr-Cr quadruple bond causes major lengthening of the Cr-Cr distance, and there is some prior evidence that a similar lengthening is caused by dative interaction with the π* orbitals. Some molecules have now been made that allow a definitive assessment of this axial π* effect. A molecule has been designed to ensure that there is axial donation into the π* orbitals but not onto the σ* orbital; ligands have been used in which the donor atoms are tethered to the bridging ligands in such a way that they can reach only the π* orbitals but not the σ* orbital. The ligands used for this purpose are the anions of 2,6-di(phenylimino)piperidine (DPhIP) and 2,2′-dipyridylamine (dpa). In the compound Cr₂(DPhIP)₄ four imino nitrogen lone pairs are suitably positioned to donate to the π* orbitals and the Cr-Cr bond length is 2.265(1) A. For direct comparison, the compound Cr₂(PhIP)₄ (PhIP is the anion of 2-(phenylimino)piperidine) was made and found to have a Cr-Cr distance of 1.858(1) A. In this case the ligand is very similar to DPhIP except that it has no donor nitrogen atoms available for axial π* donation. Thus, the cumulative effect of donation from four nitrogen atoms is very large, namely, 0.4 A in the Cr-Cr distance. The Cr₂(dpa)₄ molecule occurs in three different crystalline compounds, in all of which there are slightly different conformations, but the same Cr-Cr distance, 1.94 ± 0.01 A; these may be compared to that in the compound Cr₂(mpa)₄ (1.87 A) in which the bridging is quite similar but there are no tethered additional donor atoms.

Full text of DOI:10.1021/ja9912675

6856
J. Am. Chem. Soc. 1999, 121, 6856-6861
Remarkable Effects of Axial π* Coordination on the Cr-Cr
Quadruple Bond in Dichromium Paddlewheel Complexes
F. Albert Cotton,*,† Lee M. Daniels,† Carlos A. Murillo,*,†,‡ Isabel Pascual,† and
Hong-Cai Zhou†
Contribution from The Laboratory for Molecular Structure and Bonding, Department of Chemistry,
P.O. Box 30012, Texas A&M UniVersity, College Station, Texas 77842-3012, and Escuela de Qu´ımica,
UniVersidad de Costa Rica, Ciudad UniVersitaria, Costa Rica
ReceiVed April 20, 1999
Abstract: It is well-known that donation of electron density into the σ* orbital of a Cr-Cr quadruple bond
causes major lengthening of the Cr-Cr distance, and there is some prior evidence that a similar lengthening
is caused by dative interaction with the π* orbitals. Some molecules have now been made that allow a definitive
assessment of this axial π* effect. A molecule has been designed to ensure that there is axial donation into the
π* orbitals but not onto the σ* orbital; ligands have been used in which the donor atoms are tethered to the
bridging ligands in such a way that they can reach only the π* orbitals but not the σ* orbital. The ligands used
for this purpose are the anions of 2,6-di(phenylimino)piperidine (DPhIP) and 2,2′-dipyridylamine (dpa). In the
compound Cr2(DPhIP)4 four imino nitrogen lone pairs are suitably positioned to donate to the π* orbitals and
the Cr-Cr bond length is 2.265(1) Å. For direct comparison, the compound Cr2(PhIP)4 (PhIP is the anion of
2-(phenylimino)piperidine) was made and found to have a Cr-Cr distance of 1.858(1) Å. In this case the
ligand is very similar to DPhIP except that it has no donor nitrogen atoms available for axial π* donation.
Thus, the cumulative effect of donation from four nitrogen atoms is very large, namely, 0.4 Å in the Cr-Cr
distance. The Cr2(dpa)4 molecule occurs in three different crystalline compounds, in all of which there are
slightly different conformations, but the same Cr-Cr distance, 1.94 ( 0.01 Å; these may be compared to that
in the compound Cr2(mpa)4 (1.87 Å) in which the bridging is quite similar but there are no tethered additional
donor atoms.
Introduction
case2 π donation was found to be predominant. In this case
molecules of Cr2(O2CCPh3)4 have benzene axially coordinated
with the ring planes perpendicular to the Cr-Cr axis. The Cr-
Cr distance is 2.26 Å, which may be compared with a value of
1.97 Å found for the Cr2(O2CCH3)4 molecule in the gas phase.3
Calculations showed clearly that the lengthening was attributable
to donation of electron density from the benzene e1g orbitals to
the π* orbitals of the Cr2 unit. A subsequent attempt to develop
the concept in more detail had only limited success, but did
provide further support.4 More recently, we have reported5 that
in certain molecules there are weak interactions between
o-fluorine atoms on ligands and the chromium atoms that seem
to be responsible for modest increases (0.07-0.11 Å) in the
Cr-Cr distance. From the molecular structures, it was not easy
to say whether these interactions were of π* or σ* character,
or both. In any case, these results reawakened our interest in
axial interaction of the π* type and motivated us to seek ligands
that would be capable of introducing such interactions in an
explicit and major way.
A fundamental and recurring question concerning the M-M
distances in complexes of the type M2L4‚Xn, where M is one
of a number of transition metal atoms, L is a bridging bidentate
ligand (e.g., RCO2-), and X is an axial ligand (of which there
may be 0, 1, or 2), is the extent to which these distances are
influenced by the identity of M, the nature of L, and the presence
of X.1 It is for compounds of chromium1b that this question
arises in its most acute form, since Cr-Cr distances range from
1.83 to 2.60 Å depending on L, X, and n, all within the common
structural motif of the paddlewheel arrangement of the four L
ligands (I).
(1) (a) Cotton, F. A.; Walton, R. A. Multiple Bonds Between Metal Atoms,
2nd ed.; Oxford University Press: Oxford, England, 1993. (b) Cotton, F.
A.; Walton, R. A. Multiple Bonds Between Metal Atoms, 2nd ed.; Oxford
University Press: Oxford, England, 1993; Chapter 4. (c) Cotton, F. A.;
Walton, R. A. Multiple Bonds Between Metal Atoms, 2nd ed.; Oxford
University Press: Oxford, England, 1993; Chapter 7.
(2) Cotton, F. A.; Feng, X.; Kibala, P. A.; Matusz, M. J. Am. Chem.
Soc. 1988, 110, 2807.
(3) Ketkar, S. N.; Fink, M. J. Am. Chem. Soc. 1985, 107, 338.
(4) Cotton, F. A.; Daniels, L. M.; Kibala, P. A. Inorg. Chem. 1992, 31,
1865.
It has been argued1b that while the Cr-Cr distances are
certainly sensitive to the identity of the ligand L, they are much
more sensitive to axial ligation. In most previous discussions,
an axial ligand has been assumed to be a σ-donor, but in one
† Texas A&M University.
‡ University of Costa Rica.
(5) Cotton, F. A.; Murillo, C. A.; Pascual, I. Inorg. Chem. 1999, in press.
10.1021/ja9912675 CCC: $18.00 © 1999 American Chemical Society
Published on Web 07/09/1999

Effects of Axial π* Coordination on Cr-Cr
J. Am. Chem. Soc., Vol. 121, No. 29, 1999 6857
Experimental Section
Thus, in this report we are concerned with new, and we
believe definitive, examples that show how donation of electrons
to the π* orbitals can weaken and thus lengthen the Cr-Cr
bond. The examples we present here are quite different from
anything previously reported. Here the donors are nitrogen atoms
appended to the bridging ligands in such a way that their lone
pair electrons are positioned where the distal lobes of the π*
orbitals are expected to be, as shown schematically in II.
General Procedures. All manipulations were carried out under
nitrogen using standard Schlenk techniques. Solvents were purified by
conventional methods, and were freshly distilled under nitrogen prior
to use. Anhydrous CrCl2 was purchased from Strem Chemicals and
stored in a drybox; MeLi (1.0 M in THF/cumene) was purchased from
Aldrich and used as received; Hdpa (2,2′-dipyridylamine) was obtained
from Aldrich and sublimed before use; HDPhIP, 2,6-di(phenylimi-
no)piperidine, and HPhIP, 2-(phenylimino)piperidine, were synthesized
following published procedures.8,9 Infrared spectral data were recorded
on KBr pellets using a Perkin-Elmer 16 PC FT-IR spectrometer; NMR
spectra were recorded on a Varian XL-200 spectrometer. Elemental
analyses were performed by Canadian Microanalytical Services Ltd.;
they were satisfactory.
Preparation of Cr2(DPhIP)4 (1). The compound HDPhIP (1.06 g,
4.0 mmol) was dissolved in THF (5.0 mL) and cooled to -78 °C. Then
MeLi in THF/cumene (1.0 M, 4.2 mL) was added dropwise. Bubbles
quickly formed, and a pale yellow solution was obtained. Anhydrous
CrCl2 (0.26 g, 2.0 mmol) was then added through a solids addition
tube. The resulting yellow suspension was then stirred at room
temperature for 3 h. A yellow solid was collected by a filtration and
washed with THF (5.0 mL) and hexanes (3 × 10 mL). Additional
product was obtained from workup of the filtrate. The filtrate and THF
wash solution were combined and mixed with hexanes (30 mL),
resulting in a yellow solid. This solid was dissolved in CH2Cl2 (10.0
mL) and filtered to remove LiCl; the resulting reddish solution was
evacuated to dryness under vacuum, leaving a yellow product. The
solids were combined; crystallization from warm THF afforded 1‚2THF
The diagram II shows a situation in which donation from
the filled lone pair orbitals on appropriately tethered nitrogen
atoms into the empty π* lobes of the chromium atoms is
maximal. Two deviations from this situation can lessen the
interaction: (1) The Cr‚‚‚N distances may increase as a result
of increases in the N-C-N angles. (2) The lone-pair orbitals
may become “misdirected” as the torsional angles Cr-N1-
C-N2 deviate from 0°, that is, as the N2 atom deviates from
the plane defined by Cr-N1-C. To avoid confusion of these
angles with the torsional angles about the Cr-Cr bonds, we
shall call these “direction angles”. We shall comment further
on the role of these factors after the actual compounds and their
structures have been described.
1
as a yellow crystalline solid (0.850 g, 65.5%). H NMR (CD2Cl2, δ):
6.73 (m, 8H), 5.71 (m, 2H), 2.77 (m, 4H), 2.58 (m, 2H). IR (KBr,
cm-1): 1655 (w), 1623 (w), 1579 (w), 1492 (w), 1440 (s), 1370 (m),
1346 (m), 1319 (w), 1261 (w), 1217 (s), 1192 (m), 1072 (w), 1026
(m), 905 (w), 866 (w), 803 (m), 762 (m), 699 (s), 601 (w), 511 (m),
420 (w). Recrystallization of 1 from THF/hexanes generated red crystals
of 1‚THF. Recrystallization of 1‚THF or 1‚2THF from either CH2Cl2/
hexanes or CH2Cl2/diethyl ether afforded red crystals of 1‚CH2Cl2.
Preparation of Cr2(PhIP)4 (2). The compound HPhIP (0.38 g, 2.2
mmol) was dissolved in toluene (15 mL), deprotonated by MeLi at
-78 °C, and added to CrCl2 (0.14 g, 1.1 mmol). The reaction mixture
first became yellow and then orange. This orange suspension was
refluxed for 1 h, then stirred at room temperature overnight. The
resulting dark red solution was filtered through Celite to remove LiCl
and layered with hexanes. Needle-shaped yellow (1 mm) crystals grew
The complexes with which this work has been carried out
are shown schematically as 1-4. Compound 2 is a reference
1
in two weeks. Yield: 0.065 g (15%). H NMR (C6D6, δ): 7.12 (t,
2H), 6.89 (t, 1H), 6.52 (d, 2H), 2.86 (t, 2H), 2.62 (t, 2H), 1.43 (m,
4H). IR (KBr, cm-1): 1593 (m), 1571 (m), 1547 (vs), 1522 (w), 1518
(w), 1508 (w), 1488 (vs), 1458 (m), 1438 (w), 1425 (w), 1400 (w),
1389 (w), 1376 (w), 1355 (m), 1327 (w), 1317 (w), 1274 (s), 1261 (s),
1234 (s), 1178 (w), 1155 (w), 1117 (s), 1088 (m), 1065 (m), 1026 (m),
959 (m), 909 (w), 892 (w), 856 (w), 800 (s), 756 (m), 704 (s), 695
(m), 670 (w), 661 (w), 618 (w), 594 (w), 522 (w), 492 (w), 474 (w),
448 (w), 420 (w), 412 (w).
Preparation of Cr2(dpa)4 (3). The compound Hdpa (0.34 g, 2.0
mmol) dissolved in 15 mL of THF was deprotonated with an equivalent
amount of MeLi. To the colorless suspension, at -78 °C, were added
0.24 g (1.5 mmol) of anhydrous CrCl2 through a solids addition tube.
The suspension quickly turned red. It was kept at low temperature to
avoid conversion to Cr3(dpa)4Cl2,9 which is green in color. After 2 h
of stirring, the red suspension was filtered while cold. A red solid was
isolated, washed with THF, and dried under vacuum. It was recrystal-
lized from CH2Cl2/hexanes as red crystals of 3‚2CH2Cl2. It was also
crystallized from a solution in THF/hexanes as red needle-shaped
compound for 1. In 2, where the ligand is the 2-phenyliminopi-
peridinate anion, PhIP, no axial π* interactions can occur,
whereas in 1, where the ligand is the 2,6-di(phenylimino)-
piperidinate anion, DPhIP, four such interactions occur, two at
each end of the molecule. Compound 4, whose structure has
previously been reported6 is a reference for compound 3, which
has also been reported7 earlier in one crystal form (3‚DMF)
and is now reported here in two more.
1
crystals of 3. Crystalline yield: 0.32 g (56%). H NMR (CD2Cl2, δ):
8.16 (d, 1H), 7.52 (d, 1H), 7.42(d, 1H), 7.29 (m, 3H), 6.97 (d, 1H),
6.01 (d, 1H). IR (KBr, cm-1): 1603 (vs), 1580 (vs), 1560 (s), 1487-
1421 (br, vs), 1376 (s), 1284 (s), 1255 (s), 1172 (m), 1016 (s), 986
(m), 946 (w), 921 (w), 877 (w), 835 (w), 769 (s), 747 (s), 731 (s), 535
(6) Cotton, F. A.; Niswander, R. H.; Sekutowski, J. C. Inorg. Chem.
1978, 17, 3541.
(7) Edema, J. J. H.; Gambarotta, S.; Meetswa, A.; Spek, A. L.; Smeets,
W. J. J.; Chiang, M. Y. J. Chem. Soc., Dalton Trans. 1993, 789.
(8) Elvidge, J. A.; Linstead, R. P.; Salaman, A. M. J. Chem. Soc. 1959,
208.
(9) Bredereck, H.; Bredereck, K. Chem. Ber. 1961, 2779.

6858 J. Am. Chem. Soc., Vol. 121, No. 29, 1999
Table 1. Crystal Data and Structure Refinementa
Cotton et al.
complex
chem formula
fw
space group
a, Å
b, Å
c, Å
1‚THF
1‚2THF
C76H74Cr2N12O2
1291.47
P1h
14.198(2)
14.638(7)
18.12(1)
108.97(2)
109.53(4)
95.27(3)
3270(3)
2
1‚CH2Cl2
C69H66N12Cr2Cl2
1238.24
I2/a
27.197(3)
16.467(8)
29.51(1)
90
110.43(2)
90
12385(7)
8
2
3
3‚2CH2Cl2
C72H72N12Cr2O
1225.42
I2/a
C44H52N8Cr2
796.94
P1h
C40H32N12Cr2
784.775
C2/c
18.100(4)
11.706(2)
16.595(3)
90
102.48(3)
90
3433(1)
4
C42H36N12Cl4Cr2
954.63
C2/c
27.2495(8)
16.486(2)
29.413(2)
90
110.416(5)
90
9.8649(7)
10.1338(8)
20.561(3)
77.59(2)
80.50(2)
81.14(1)
1964.9(3)
2
23.966(6)
9.581(2)
19.106(4)
90
104.763(5)
90
a, deg
â, deg
γ, deg
V, Å3
12384(1)
8
4242(2)
4
Z
T, K
213(2)
0.71073
1.315
4.10
0.058/0.075
0.132/0.155
213(2)
0.71073
1.312
3.90
0.060/0.072
0.147/0.159
213(2)
0.71073
1.333
4.90
0.070/0.091
0.160/0.189
213(2)
0.71073
1.347
6.00
0.049/0.056
0.117/0.126
213(2)
0.71073
1.518
6.85
0.069/0.100
0.133/0.156
213(2)
0.71073
1.495
8.12
0.051/0.056
0.125/0.132
radiation λ, Å
F(calcd), g cm-3
µ(Mo KR), cm-1
R1b/R1c
wR2b/wR2c
2
2
aR1 ) ∑(|Fo| - |Fc|)/∑|Fo|. wR2 ) [∑w[(Fo2 - Fc2)2/∑[w(Fo )2]]1/2; w ) 1/[σ2(Fo ) + (a‚P)2 + b‚P], P ) [max(Fo2 or 0) + 2(Fc2)]/3. bDenotes
the value of the residual considering only the reflections with I > 2σ(I). cDenotes value of the residual considering all the reflections.
Table 2. Selected Bond Lengths (Å) and Angles (deg) for
Cr2(DPhIP)4 in 1·THF
Table 3. Selected Bond Lengths (Å) and Angles (deg) for
Cr2(PhIP)4, 2
Cr(1)-Cr(2)
Cr(1)-N(2)
Cr(1)-N(6)
Cr(1)-N(8)
Cr(1)-N(12)
2.2652(9)
2.063(3)
2.103(3)
2.060(3)
2.124(3)
Cr(1)-Cr(1A)
Cr(1A)-N(1)
Cr(1)-N(3)
1.858(1)
2.057(3)
2.050(3)
Cr(2)-N(3)
Cr(2)-N(5)
Cr(2)-N(9)
Cr(2)-N(11)
2.111(3)
2.065(3)
2.097(3)
2.063(3)
Cr(1)-N(2)
2.050(3)
2.053(3)
Cr(1A)-N(4)
Cr(1)-Cr(1A)-N(1)
Cr(1A)-Cr(1)-N(3)
N(1)-C(1)-N(2)
98.55(8) Cr(1A)-Cr(1)-N(2)
94.77(8) Cr(1)-Cr(1A)-N(4)
116.7(3)
92.67(8)
96.16(8)
116.0(3)
N(2)-Cr(1)-Cr(2)-N(3) 9.1(1) N(6)-Cr(1)-Cr(2)-N(5)
N(8)-Cr(1)-Cr(2)-N(9) 7.6(1) N(12)-Cr(1)-Cr(2)-N(11) 7.4(1)
7.6(1)
N(3)-C(12)-N(4)
N(1)-Cr(1A)-Cr(1)-N(2) 0.1(1) N(4)-Cr(1A)-Cr(1)-N(3) 3.4(1)
Cr(1)‚‚‚N(1)
Cr(1)‚‚‚N(7)
2.745(3)
2.809(3)
Cr(2)‚‚‚N(4)
Cr(2)‚‚‚N(10)
2.945(3)
2.738(3)
Table 4. Selected Bond Lengths (Å) and Angles (deg) for
Cr2(DPhIP)4 in 1·2THF
Cr(1)-N(2)-C(7)-N(1) 6.3(4) Cr(2)-N(5)-C(24)-N(4)
Cr(1)-N(8)-C(41)-N(7) 2.7(4) Cr(2)-N(11)-C(58)-N(10) 2.2(4)
0.9(4)
Cr(1)-Cr(2)
Cr(1)-N(2)
Cr(1)-N(6)
Cr(1)-N(8)
Cr(1)-N(12)
2.155(1)
2.034(3)
2.128(3)
2.073(3)
2.112(3)
Cr(2)-N(3)
Cr(2)-N(5)
Cr(2)-N(9)
Cr(2)-N(11)
2.115(3)
2.056(3)
2.097(3)
2.046(3)
(m), 518 (w), 446 (m), 410 (m). Mass spectroscopy (FAB+, NBA as
matrix, m/z): 784, [Cr2(dpa)4]+; 614, [Cr2(dpa)3]+; 565, [Cr(dpa)3]+;
392, [Cr(dpa)2]+; 273, [Cr2(dpa)]+.
N(2)-Cr(1)-Cr(2)-N(3) 10.9(1) N(6)-Cr(1)-Cr(2)-N(5)
N(8)-Cr(1)-Cr(2)-N(9) 12.1(1) N(12)-Cr(1)-Cr(2)-N(11) 10.1(1)
13.1(1)
Crystallographic Studies. Data collection for all crystals was carried
out on a Nonius Fast area detector diffractometer with each crystal
mounted on the tip of a glass fiber under a stream of nitrogen. All data
sets were collected at -60 °C. Cell parameters were obtained by least-
squares refinement of 250 reflections ranging in 2θ from 15° to 41°.
Laue groups and centering conditions were confirmed by axial images.
Data were collected using 0.2° intervals in æ for the range 0° < æ <
220° and 0.2° intervals in ω for two different regions in the range 0°
< ω < 72°. In this way, nearly a full sphere of data was collected.
The highly redundant data sets were corrected for Lorentz and
polarization effects, and for absorption.
The positions of the chromium atoms and their first coordination
spheres were determined by direct methods and refined by using the
program SHELXL-93. All non-hydrogen atoms were found by suc-
cessive iterations of least-squares refinement followed by Fourier
syntheses and, during the final cycles, were refined anisotropically.
Hydrogen atoms were placed in idealized positions, and a common
thermal parameter was refined. In 2‚2THF, the disordered phenyl rings
were modeled in two orientations each, and refined at half occupancy
for each orientation.
Cr(1)‚‚‚N(1)
Cr(1)‚‚‚N(7)
2.734(3)
3.057(3)
Cr(2)‚‚‚N(4)
Cr(2)‚‚‚N(10)
2.881(3)
2.935(3)
Cr(1)-N(2)-C(7)-N(1) 10.8(4) Cr(2)-N(5)-C(24)-N(4)
Cr(1)-N(8)-C(41)-N(7) 9.4(4) Cr(2)-N(11)-C(58)-N(10) 0.4(4)
2.5(4)
it can also support linear trimetallic chains.11 Control of
stoichiometry and reaction temperature are crucial to the product
distribution. When the molar ratio of LiDPhIP and CrCl2 is 2:1,
and the reaction is carried out in THF at room temperature, 1
is formed exclusively. If the molar ratio of LiDPhIP and CrCl2
is higher than 2:1, and the reaction is carried out at reflux
temperature, a mixture of 1 and a green trichromium complex
is formed.12 Compound 1 is moderately soluble in THF, and
very soluble in CH2Cl2, but insoluble in benzene or acetonitrile.
It can be recrystallized from THF/hexanes as red crystals of
1‚THF and from hot THF as orange crystals of 1‚2THF.
Compound 1 can also be recrystallized from CH2Cl2/hexanes
or CH2Cl2/ether as red block-crystals of 1‚CH2Cl2. The solutions
of 1 are air sensitive, but the crystalline forms of 1 are
surprisingly stable. For example, 1‚THF as red crystals has been
kept in air for two months without noticeable change in
appearance or crystal structure.
Crystallographic data for 1‚THF, 1‚2THF, 1‚CH2Cl2, 2, 3, and 3‚
2CH2Cl2 are given in Table 1. Selected bond distances and angles for
1‚THF (structural parameters of 1‚CH2Cl2 are very similar to those of
1‚THF) are given in Table 2. Selected bond distances and angles for
1‚2THF, 2, and 3 (structural parameters for 3‚2CH2Cl2 are very similar
to those of 3) are found in Tables 3, 4 and 5, respectively.
(10) Cotton, F. A.; Daniels, L. M.; Murillo, C. A.; Pascual, I. Inorg.
Chem. Commun. 1998, 1, 1.
Results and Discussion
(11) Cotton, F. A.; Daniels, L. M.; Murillo, C. A.; Pascual, I. J. Am.
Chem. Soc. 1997, 119, 10223.
(12) The green trichromium complex will be described elsewhere.
Remarks Regarding Syntheses. The ligand DPhIP has three
nitrogen donors. It is structurally similar to the dpa ligand,10 so

Effects of Axial π* Coordination on Cr-Cr
J. Am. Chem. Soc., Vol. 121, No. 29, 1999 6859
1‚2THF [Cr-Cr (Å): 2.155(1)]
Table 5. Structural Parameters of 1‚THF and 1‚2THF Related to Axial Coordination
1‚THF [Cr-Cr (Å): 2.2652(9)]
direction angles (deg)
Cr‚‚‚N (Å)
direction angles (deg)
Cr‚‚‚N (Å)
Cr(2)-N(11)-C(58)-N(10) 2.2(3)
Cr(2)‚‚‚N(10) 2.738(3)
Cr(1)-N(2)-C(7)-N(1)
Cr(2)-N(5)-C(24)-N(4)
10.8(4)
2.5(4)
Cr(1)‚‚‚N(1)
Cr(2)‚‚‚N(4)
2.734(3)
2.881(3)
Cr(1)-N(2)-C(7)-N(1)
Cr(1)-N(8)-C(41)-N(7)
Cr(2)-N(5)-C(24)-N(4)
6.3(4)
2.7(4)
0.9(4)
Cr(1)‚‚‚N(1)
Cr(2)‚‚‚N(7)
Cr(2)‚‚‚N(4)
2.745(3)
2.809(3)
2.945(3)
Cr(2)-N(11)-C(58)-N(10) 0.4(3)
Cr(2)‚‚‚N(10) 2.935(3)
Cr(1)-N(8)-C(41)-N(7)
9.4(4)
Cr(2)‚‚‚N(7)
3.057(3)
∑
12(2)
11.24(2)
∑
23(2)
11.61(2)
The preparation methods for dichromium paddlewheel com-
plexes with various bidentate ligands can vary greatly depending
on the solubility and basicity of the ligand. Carboxylate ligands
can be deprotonated with NaOMe, and can react with Cr2(CH3-
COO)4‚2H2O in ethanol or aqueous solution. Hydroxypyridine
can be neutralized with butyllithium in THF and reacts with
Cr2(CH3COO)4. Chromocene is also an excellent starting
material, when the ligand is insufficiently soluble.13 When a
formamidinate-type ligand is used, CrCl2 can be allowed to react
with the lithiated ligand in THF.14 HPhIP is structurally similar
to a formamidine, and it is even soluble in petroleum ether.
Instead of THF, toluene was used for the reaction between CrCl2
and LiPhIP. The reaction mixture was refluxed for 1 h to ensure
a complete conversion. A toluene solution of 2 was obtained
by a single filtration which removed LiCl and a small amount
of an impurity.
The most remarkable thing we have observed in the prepara-
tive chemistry concerning compound 1 is the structural diversity.
We have obtained it in three crystalline forms: 1‚THF, 1‚CH2-
Cl2, and 1‚2THF. The first two (which are isomorphous) are
red, whereas the last one is orange. The color difference, red
vs orange, is striking and immediately prompts the question of
how the “same” molecule can have two quite different colors
when only the kind or number of solvent molecules is changed.
It is well-known1c that Rh2(O2CCH3)4‚S2 compounds vary a
great deal in color as the solvent molecules, S, are varied, but
this is a case where the solvent molecules are directly coor-
dinated to the Rh2(O2CCH3)4 molecules in axial positions. As
we shall see shortly, in all of the three differently solvated forms
of compound 1, the solvent molecules are simply interstitial,
having no specific interaction with the chromium atoms.
The Evidence for π* Interactions. We begin with 1‚THF,
which crystallizes in space group I2/a with all atoms in general
positions. No significant intermolecular interactions were found.
A drawing of the molecule in 1‚THF is shown in Figure 1. There
are four DPhIP anions bridging a dichromium unit. Each ligand
uses two of its three nitrogen atoms to coordinate equatorially
with the dichromium unit. The third nitrogen is dangling, but
located near one of the two chromium atoms in an off-axis
position. The directions of the four pendant arms of the ligands
alternate around the dichromium unit. The distances between
the pendant nitrogen atoms and the closest chromium atoms
are listed in Table 2. The striking feature of molecule 1 in 1‚
THF is the Cr-Cr distance of 2.265(9) Å, which is the longest
Cr-Cr distance in any paddlewheel complex supported only
by nitrogen donor ligands.
We see a group orbital composed of the two nitrogen lone pairs,
which matches the symmetry of the dyz orbital of a chromium
atom, interacts with the dyz orbital, and results in a half-occupied
antibonding orbital. This antibonding orbital then interacts with
the dyz orbital of the other chromium atom to form a destabilized
π-bonding orbital. The dxz orbital of the other chromium atom
in a molecule of Cr2(DPhIP)4 also interacts with the other two
nitrogen atoms in exactly the same way. These interactions pull
the Cr atoms apart causing the elongation of the Cr-Cr bond.
An alternative way to formulate the situation is to regard the
pendant nitrogen atoms as donating into the π* orbital in the
Cr-Cr bond, thus canceling some of the π bonding. Table 6
shows that the sum of the direction angles of 1 in 1‚THF is
only 12(2)°, so the lone pair orbitals of nitrogen atoms point
almost directly to the empty π* lobes of the chromium atoms.
In addition, the Cr‚‚‚N distances are short; they vary from 2.74
to 2.95 Å, with a sum of 11.24 Å for the four ligands in the
molecule. This is the explanation we propose for the long Cr-
Cr distance of 2.265(1) Å that is found in 1‚THF.
To check on the correctness of this explanation other
experiments were carried out. The first one was the following.
We prepared a compound as similar as possible to 1, except
that the pendant nitrogen atoms that we claim are responsible
for the lengthening of the Cr-Cr bond are absent. This is
compound 2, and its actual structure is shown in Figure 3. The
molecular dimensions are listed in Table 3, and it can be seen
that there is an enormous change in the Cr-Cr distance, which
has shrunk from 2.265(1) Å to 1.858(1) Å, while no other
significant changes have occurred.
While we believe that the case is proven by the results just
presented, we are pleased that there is more supporting evidence.
One such piece is especially welcome because it came unex-
pectedly. The reader will recall that in addition to 1‚THF and
its isomorph 1‚CH2Cl2, both of which are red solids, we obtained
the orange solid 1‚2THF. Let us now look at the structure of
1‚2THF. Superficially, the molecule of 1 present closely
resembles the one found in 1‚THF, but (Table 4) a critical
difference is immediately obvious: the Cr-Cr distance is about
0.10 Å shorter, namely, 2.155(1) Å compared to 2.265(1) Å.
How can the “same” molecule show such a structural difference
in two different crystal forms? Once again, we see the exquisite
sensitivity of the Cr-Cr bond to axial interactions at play, only
now we are looking at π* instead of σ* interactions.
According to our previous reasoning in explaining the
difference between the Cr-Cr distances in 1‚THF and 2 we
look for a difference in the axial interactions occurring in the
molecule 1 in 1‚THF and 1‚2THF, and as Table 5 shows, we
find exactly what is needed to explain the Cr-Cr bond length
difference. While the sum of the four Cr‚‚‚N distances is only
slightly longer in 1‚2THF the nitrogen lone pairs are far more
“misdirected” than they are in 1‚THF. Therefore, there is much
less donation into the Cr2 π* orbitals and the Cr-Cr distance
is much less elongated, namely by only about 0.3 Å instead of
0.4 Å.
We attribute the remarkable elongation of the Cr-Cr bond
to the interaction between the lone pair of pendant nitrogen
atoms and the dxz and dyz orbitals of the two chromium atoms.
One way to look at this interaction is shown in Figure 2, which
is a schematic molecular orbital diagram showing the interac-
tions of a metal dyz orbital and two such nitrogen lone pairs.
(13) Cotton, F. A.; Thompson, J. L. Inorg. Chem. 1981, 20, 1292.
(14) Cotton, F. A.; Ren, T. J. Am. Chem. Soc. 1992, 114, 2237.

6860 J. Am. Chem. Soc., Vol. 121, No. 29, 1999
Cotton et al.
Figure 1. A drawing of the molecular structure of [Cr2(DPhIP)4] in
1‚THF. Thermal displacement ellipsoids are drawn at 50% probability.
Hydrogen atoms have been omitted for clarity. The axial π* interactions
are also shown.
Figure 3. A drawing of Cr2(PhIP)4, 2. Ellipsoids are drawn at the
50% probability level. Hydrogen atoms have been omitted for clarity.
Figure 4. 1H NMR spectrum of Cr2(DPhIP)4 in CD2Cl2 showing a
signal shifted from the normal aromatic region (6.5-7.0 ppm) to 5.7
ppm. The signal at 5.3 ppm is from the solvent.
Figure 2. Schematic MO diagram showing the effect of axial
coordination from an appendage of a bridging ligand.
Our reasoning so far suggests an obvious question. If it is
only molecular packing forces that cause the axial π* interac-
tions to change in going from 1‚THF to 1‚2THF, will the
structural difference not vanish in solution? To answer this
question we made solutions in CD2Cl2 from both kinds of
1
crystals and showed that they have identical H NMR spectra.
The spectrum is shown in Figure 4. Moreover, the colors of
the two solutions were visually indistinguishable. In fact we
were able to learn a little more about the structure in solution
1
from the H NMR spectrum. It appears that it must be rather
similar to that shown in Figure 1 for 1‚THF. Another view of
that structure is shown as a space-filling model in Figure 5.
Here it can be seen that on each end of the molecule the phenyl
groups in the pendant NCC6H5 groups have a very particular
spacial relationship to each other. They are close (3.44 Å from
C(6) to C(36)), almost parallel, but offset so that only two
hydrogen atoms of each one, those on C(36), C(37) and C(5),
C(6), lie right over the plane of the other ring. In such positions
and at such close distances we should expect these four protons
(eight altogether for the whole molecule) to experience a distinct
upfield shift relative to the others. That is exactly what Figure
4 shows. Most of the aromatic resonances lie in their usual
region, 6.5-7.0 ppm, but there is a signal at 5.7 ppm due to
the eight that are shifted by the diamagnetic anisotropy of the
neighboring phenyl rings. The intensity ratio observed agrees
within experimental error with that expected, 4:1. It may also
Figure 5. Space filling model of Cr2(DPhIP)4 in 1‚THF showing the
shielding of two protons on each end phenyl ring by the opposite phenyl
group.
be noted that because of the unsymmetrical relationship of each
DPhIP ligand to the Cr2 unit, there are three methylene signals.
There is still more experimental evidence to adduce in favor
of the influence of axial π* interactions on Cr-Cr bond lengths.
For this we turn to the Cr2(dpa)4 molecule, 3, whose structure
has been determined in three crystalline forms, 3, 3‚2CH2Cl2,
and previously7 3‚DMF (DMF ) dimethylformamide). Our
results for 3 are shown in Figure 6 and Table 6. Results for all
three cases are found in Table 7, where the structural features
pertaining to axial interaction are presented. The Cr-Cr distance

Effects of Axial π* Coordination on Cr-Cr
J. Am. Chem. Soc., Vol. 121, No. 29, 1999 6861
Table 6. Selected Bond Lengths (Å) and Angles (deg) for 3
Cr(1)-Cr(2)
Cr(1)-N(20A)
Cr(1)-N(40A)
1.943(2)
2.068(5)
2.075(5)
Cr(2)-N(10A)
Cr(2)-N(50A)
2.062(5)
2.040(5)
N(10A)-Cr(2)-Cr(1)-N(20A)
Cr(1)‚‚‚N(30A)
4.1(2)
2.751(6)
11.4(6)
N(50A)-Cr(2)-Cr(1)-N(40A)
Cr(2)‚‚‚N(60A)
6.0(2)
3.089(6)
Cr(1)-N(20A)-C(7)-N(30A)
Cr(2)-N(50A)-C(46)-N(60A)
40.4(7)
Table 7. Structural Parameters of 3, 3‚2CH2Cl2, and 3‚DMF Related to Axial Coordination
3 [Cr-Cr (Å): 1.943(2)] 3‚2CH2Cl2 [Cr-Cr (Å): 1.940(1)]
3‚DMFa [Cr-Cr (Å): 1.935]
direction angles (deg)
Cr-N (Å)
direction angles (deg)
Cr-N (Å)
direction angles (deg)
Cr-N (Å)
11.4(6)
11.4(6)
40.4(7)
40.4(7)
2.751(6)
2.751(6)
3.089(6)
3.089(6)
24.7(4)
24.7(4)
37.1(4)
37.1(4)
2.820(3)
2.820(3)
3.022(4)
3.022(4)
23.4
23.4
32.0
32.0
2.865
2.865
2.955
2.955
∑
104(3)
11.68(3)
∑
124(2)
11.68(2)
∑
111
11.64
a See reference 7.
nitrogen atoms in 3 are about the same distance away from the
chromium atoms (the sum of the four CrCN distances is ca.
11.6 Å compared to 11.2 Å in 1‚THF and 11.6 Å in 1‚2THF)
but the nitrogen lone pair orbitals are enormously more
misdirected. In Cr2(dpa)4 the sum of the direction angles varies
from 104° to 124°, depending on the crystal form, whereas in
1‚THF and 1‚2THF these values were 12° and 23°, respectively.
Thus, we would expect that for 3 the Cr-Cr bond lengthening
effect would be relatively small and it is (0.07 Å), whereas in
1 it rises as high as 0.40 Å.
Concluding Remarks. The work reported here addresses
(definitively in our opinion) the question of how axial interac-
tions which inject electron density largely, if not entirely, into
the π* orbitals affect the Cr-Cr distance. It is clear that the
effect can be very large (ca. 0.4 Å), or smaller, depending on
how well placed and oriented the donor atoms are.
Figure 6. A drawing of Cr2(dpa)4, 3. Ellipsoids are drawn at the 50%
probability level. Hydrogen atoms have been omitted for clarity. The
possible axial π* interactions are also shown.
Acknowledgment. We are grateful to the National Science
Foundation for financial support.
Supporting Information Available: Tables of crystal-
lographic data including diffractometer and refinement data,
atomic coordinates, bond lengths, bond angles, and anisotropic
displacement parameters (PDF). An X-ray crystallographic file,
in CIF format, is also available. This material is available free
of charge via the Internet at http://pubs.acs.org.
is almost invariant in these three structures, being virtually
constant at 1.940 ( 0.005 Å. This may be compared with the
value in the reference compound 4, where it is 1.87 Å. This
elongation can also be attributed to axial π* interactions.
It can be seen in Figure 6 that in Cr2(dpa)4 there are axial π*
interactions qualitatively similar to those in the molecule of 1,
but as Table 7 shows, they are much weaker. The pendant
JA9912675