Octamethyl-1,1′-diphosphachromocene: Its spin distribution and oxidation

Article abstract of DOI:10.1021/om9705618

Octamethyl-1,1′-diphosphachromocene ((Tmp)₂Cr) was prepared from (Tmp)K and chromium dichloride in 81% yield. According to X-ray analysis, it has a distorted sandwich structure with the P atoms being bent away from chromium. Large ¹H, ¹³C, and ³¹P NMR shifts established the similarity to chromocenes (S = 1) and negative spin on the ligands. The signal splitting and temperature-dependent ¹H NMR data revealed that, on the ligand, the spin sits predominantly at the phosphorus and two carbon atoms. In the cyclic voltammetry, two oxidation steps and one reduction were visible. After chemical oxidation, [(TmP)₂Cr]⁺[B(C₆H₅) ₄]^- was isolated in 65% yield. Its EPR spectrum is in accord with a S = 3/2 species. The ¹H NMR spectrum resembled those of chromocenium ions. Its analysis and that of the temperature-dependent ¹H NMR spectra indicated a rather uniform spin distribution within the ligands. It was concluded that, in any case, the phosphorus may release spin to neighboring molecules.

Full text of DOI:10.1021/om9705618

4606
Organometallics 1997, 16, 4606-4610
Octa m eth yl-1,1′-d ip h osp h a ch r om ocen e: Its Sp in
Distr ibu tion a n d Oxid a tion
Robert Feher,1a Frank H. Ko¨hler,*,1b Franc¸ois Nief,*,1b Louis Ricard,1b and
Stefan Rossmayer1a
Anorganisch-chemisches Institut, Technische Universita¨t Mu¨nchen,
D-85747 Garching, Germany, and Ecole Polytechnique, CNRS URA 1499,
F-91128 Palaiseau, France
Received J uly 2, 1997X
Octamethyl-1,1′-diphosphachromocene ((Tmp)2Cr) was prepared from (Tmp)K and chro-
mium dichloride in 81% yield. According to X-ray analysis, it has a distorted sandwich
1
structure with the P atoms being bent away from chromium. Large H, 13C, and 31P NMR
shifts established the similarity to chromocenes (S ) 1) and negative spin on the ligands.
1
The signal splitting and temperature-dependent H NMR data revealed that, on the ligand,
the spin sits predominantly at the phosphorus and two carbon atoms. In the cyclic
voltammetry, two oxidation steps and one reduction were visible. After chemical oxidation,
[(Tmp)2Cr]+[B(C6H5)4]- was isolated in 65% yield. Its EPR spectrum is in accord with a S
3
1
) /2 species. The H NMR spectrum resembled those of chromocenium ions. Its analysis
1
and that of the temperature-dependent H NMR spectra indicated a rather uniform spin
distribution within the ligands. It was concluded that, in any case, the phosphorus may
release spin to neighboring molecules.
In tr od u ction
magnetic exchange between the stacks when the donor
capability of phosphorus leads to an interaction with
the metal of the sandwich of a neighboring stack. We
have embarked on tuning the paramagnetic sandwich
by synthesizing octamethyl-1,1′-diphosphachromocene,
((Tmp)2Cr5 ). Subsequently, we studied its oxidation to
[Tmp2Cr]+, the building block necessary for stacks, and
the spin distribution within both species.
Phospholyl ligands are susceptible of π bonding to
transition metals very much like cyclopentadienyl (Cp),
thus yielding phosphametallocenes.2 As for the parent
metallocenes, they have never ceased to attract atten-
tion. In particular, paramagnetic metallocenes have
experienced a remarkable renaissance, because they
have led to novel molecule-based magnetic materials of
the type [metallocene]+[TCNE]-.3 These materials
undergo a transition from the desirable ferromagnetic
to paramagnetic behavior below 10 K. In order to raise
this temperature, improving the interaction between the
building blocks has been attempted, for instance, by
varying the metal M and the radical anion.3 Another
possibility to tune these materials consists in changing
the ligands of the metallocene. We have outlined in
detail previously4 how the number and type of the
substituents at Cp would localize the unpaired electron
spin density within the ligand at certain carbon atoms,
which in turn could improve the magnetic interaction.
A pronounced effect is to be expected from replacing a
CH fragment of Cp by a heteroelement, and a promising
candidate in this regard is the phospholyl ligand. Since
compounds like [metallocene]+[TCNE]- form stacks that
are parallel, the interaction within the stack might be
improved by changing the spin distribution within the
π ligand upon replacing Cp* by a phospholyl ligand. In
addition, the phospholyl may open a new pathway of
Resu lts a n d Discu ssion
The existence of phosphachromocenes was established
by Nixon’s pioneering synthesis of (t-Bu2C2P3)2Cr.6a
Since the yield was lower than 1%6b we chose the more
readily accessible Tmp- ligand.7 It has the additional
advantage of being very similar to the parent phospholyl
-
PC4H4 so that theoretical studies8 could be used as
guidelines.
The neutral octamethyl-1,1′-diphosphachromocene,
(Tmp)2Cr, was synthesized in 81% yield from Tmp- as
red air-sensitive crystals, as illustrated in Scheme 1.
The structure, which was determined by X-ray analysis,
is similar to that of 1,1′-diphosphaferrocenes9 (Figure
1). The ligands are fully staggered, which results in the
molecular symmetry C2h. The phosphorus atom is bent
away from chromium in such a way that the planes C2,
P1, C5 and C2, C3, C4, C5 include an angle of 4.1°, and
the distance of chromium from the plane C2, C3, C4,
C5 is 1.76 Å. This distance is longer than that found
for similar iron derivatives9 by 0.13 Å, a value that was
X Abstract published in Advance ACS Abstracts, September 15, 1997.
(1) (a) Technische Universita¨t Mu¨nchen. (b) Ecole Polytechnique.
(2) Mathey, R. New J . Chem. 1987, 11, 585-593. Nixon, J . F. Chem.
Soc. Rev. 1995, 319-328. For bent phosphametallocenes, see: Nief,
F.; Ricard, L.; Mathey, F. Organometallics 1989, 8, 1473-1477.
Baudry, D.; Ephritikhine, M.; Nief, F.; Ricard, L.; Mathey, F. Angew.
Chem., Int. Ed. Engl. 1990, 29, 1485-1486. Gosink, H.-J .; Nief, F.;
Ricard, L.; Mathey, F. Inorg. Chem. 1995, 34, 1306-1307.
(3) Miller, J . S.; Epstein, A. J . Angew. Chem., Int. Ed. Engl. 1994,
33, 385-415. Gatteschi, D. Adv. Mater. 1994, 6, 635-645.
(4) Blu¨mel, J .; Hebendanz, N.; Hudeczek, P.; Ko¨hler, F. H.; Strauss,
W. J . Am. Chem. Soc. 1992, 114, 4223-4230.
(5) Acronym derived from bis(η5-tetramethylphospholyl)chromium.
(6) (a) Bartsch, R.; Hitchcock, P. B.; Nixon, J . F. J . Organomet.
Chem. 1988, 356, C1-C4. (b) Cloke, F. G. N.; Flower, K. R.; Hitchcock,
P. B.; Nixon, J . F. J . Chem. Soc., Chem. Commun. 1995, 1659-1660.
(7) Gradoz, P.; Baudry, D.; Ephritikhine, M.; Nief, F.; Mathey, F.
J . Chem. Soc., Dalton Trans. 1992, 3047-3051.
(8) Kostic, N. M.; Fenske, R. F. Organometallics 1983, 2, 1008-1013.
Guimon, C.; Gonbeau, D.; Pfister-Guillouzo, G.; de Lauzon, G.; Mathey,
F. Chem. Phys. Lett. 1984, 104, 560-567. Su, M.-D.; Chu, S.-Y. J . Phys.
Chem. 1989, 93, 6043-6051.
(9) De Lauzon, G.; Deschamps, B.; Fischer, J .; Mathey, F.; Mitschler,
A. J . Am. Chem. Soc. 1980, 102, 994-1000.
S0276-7333(97)00561-X CCC: $14.00 © 1997 American Chemical Society

Octamethyl-1,1′-diphosphachromocene
Organometallics, Vol. 16, No. 21, 1997 4607
Ta ble 1. NMR Sign a l Sh ifts of (Tm p )2Cr a n d [(Tm p )2Cr ]+[B(C6H5)4]-
nucleus
Ca2/5
shifta
P
C2/5
C3/4
Ca3/4
Hâ2/5
Hâ3/4
(Tmp)2Feb
14.1 (24.4)
(Tmp)2Crd
δc
-58.5
94.0 (55.2)
28.8
-65.2
56.8
95.1 (<1)
12.7 (<1)
1.65 (9.4)
1.83 (<0.5)
δexptl
δpara
δdip e
δcon
-173
-115
50.6
-376
-471
53.7
-525
12.9
-1.2
-2.7
1.5
245
232
-2.6
235
34.0
-39.4
32.4
-6.9
39.3
-41.2
-6.7
-166
-122.0
-34.5
[(Tmp)2Cr]+ f
δexptl
δpara
43.1
41.5
24.3
26.1
a
In ppm. δexptl, δpara, δdip, and δcon are the experimental (relative to TMS), the paramagnetic, the dipolar, and the contact shift,
respectively. In C6D6 at 298 K. c Coupling to 31P (Hz) in parentheses. 1H NMR in C6D6 at 298 K; 13C and 31P NMR at 413 and 368 K,
respectively; solvent, anisole. e Calculated values (cf. Supporting Information). f In CD3NO2 at 298 K; resonances of [B(C6H5)4]- at 7.34,
6.99, and 6.83 ppm (ratio 2/2/1).
b
d
Sch em e 1a
Ch a r t 1. Liga n d π Or bita ls In volved in Sp in
Deloca liza tion
a
Key (a) K, dimethoxyethane; (b) CrCl2(THF).
other signal whose width would allow C,H coupling to
be observed is that near 30 ppm. Since no coupling was
detected, this signal must be assigned to a ring-C rather
than to the second methyl-C atom. Hence, in general,
the signals of the ring-C and CR11 atoms are found at
low and high frequency, respectively, relative to the
reference signals of diamagnetic (Tmp)2Fe (cf. Table 1).
This pattern has been previously observed for chro-
mocenes4,12 and is indicative of a negative spin density
on the ligands.
Further distinction of the nuclei in positions 2/5 and
3/4 of (Tmp)2Cr is of relevance because it is related to
the spin density distribution within the Tmp ligands
which we are interested in. As far as the ligand π
orbitals are concerned, the spin may reside in the
symmetric and antisymmetric (with respect to the C2
axis) e1-type MOs (molecular orbitals) πs and πa, respec-
tively, which are shown in Chart 1. For chromocene,
the πs- and πa-containing MOs are degenerate, while
they are split for (Tmp)2Cr as a result of the perturba-
tion introduced by the phosphorus atoms.8 This means
that for chromocene both πs and πa equally contribute
to the spin delocalization, while for (Tmp)2Cr one of
these MOs dominates, and this must be reflected in the
NMR spectra. Actually, the squared coefficients of the
2pz AOs (atomic orbitals) (qualitatively given in Chart
1 by different lobe diameters) are related to the hyper-
fine coupling constants A(X) expected for the nucleus
X,13 and these are proportional to the NMR contact
shifts δcon. When this is combined with the negative
overall spin on the Tmp ligand, the 31P NMR signal of
F igu r e 1. NMR spectra11 and molecular structure of
(Tmp)2Cr. 13C NMR spectrum (top) at 413 K; solvent
anisole. 1H NMR spectrum (bottom) at 298 K; solvent, C6D6.
S and X are the solvent and impurities, respectively; scale
in ppm. Selected distances (Å) and angles (deg): Cr-P1
2.3642(4), Cr-C2 2.170(1), Cr-C3 2.205(1), Cr-C4
2.207(1), Cr-C5 2.172(1), P1-C2 1.787(1), P1-C5
1.792(2), C2-C3 1.421(2), C3-C4 1.416(2), C4-C5
1.411(2); C2-P1-C5 89.05(7).
also found for the couple Cp2Cr/Cp2Fe and that follows
from the electron count.10
The structure in solution was confirmed by the 13C
and 1H NMR spectra (Figure 1), which showed the
expected number of signals. In addition, a 31P NMR
signal was observed (Table 1). It is worth noting that
the relaxation which is reflected by the signal width
does not allow nuclear coupling with 31P to be observed.
The assignment of one of the 13C NMR signals is based
on the fact that the signal near 10 ppm is a quartet due
to the one-bond C,H coupling of a CH3 group. The only
(11) R and â designate nuclei which are separated from the ring by
one and two bonds, respectively.
(12) (a) Ko¨hler, F. H. J . Organomet. Chem. 1976, 113, 11-32. (b)
Ko¨hler, F. H.; Doll, K.-H.; Pro¨ssdorf, W. J . Organomet. Chem. 1982,
224, 341-353. (c) Ko¨hler, F. H.; Doll, K.-H. Z. Naturforsch. B: Chem.
Sci. 1982, 37, 144-150. (d) Ko¨hler, F. H.; Geike, W. J . Organomet.
Chem. 1987, 328, 35-47. (e) Blu¨mel, J .; Hofmann, P.; Ko¨hler, F. H.
Magn. Reson. Chem. 1993, 31, 2-6. Note that inverted shift signs were
used in a-d.
(13) Karplus, M.; Fraenkel, G. K. J . Chem. Phys. 1961, 35, 1312-
1323. Yonezawa, T.; Kawamura, T.; Kato, H. J . Chem. Phys. 1969,
50, 3482-3492.
(10) Haaland, A. Acc. Chem. Res. 1979, 12, 415-422.

4608 Organometallics, Vol. 16, No. 21, 1997
Feher et al.
(Tmp)2Cr should have a negative shift if the πa orbital
dominates. In contrast, the shift should be positive if
the πs orbital dominates strongly. The large negative
31P NMR signal shift favors the πa orbital, and hence
the signal of C3/4 must be shifted to a lower frequency
than that of C2/5.
The 13C NMR signal pattern is striking in that the
splitting is much larger than that known for any
substituted chromocene so far. Because the signal
splitting increases with the separation of the πa- and
πs-containing MOs, we conclude that the spin localiza-
tion within the ligands of (Tmp)2Cr, mainly at P and
C3/4, is much more pronounced than for chromocenes.
1
As for the H NMR, all of the protons of Tmp are â
protons and, thus, their signal shift is strongly influ-
enced by hyperconjugative spin transfer from the pz
orbitals of the neighboring ring-C atoms.4 Given the
negative spin at C2-C5, we therefore expect two signals
at the low-frequency side. The signal of Hâ3/4 should
be more shifted than that of Hâ2/5, again following the
relative size of the coefficients of πa in Chart 1. In
contrast, Figure 1 shows one signal at either frequency
side. In order to explain this observation, dipolar shifts,
F igu r e 2. Temperature dependence of the 1H NMR signal
shifts of (Tmp)2Cr represented as a ϑ vs T plot (see text).
δ
dip, and σ delocalization must be taken into account.4,12
Calculation of the dipolar shifts (Table 1) yields δdip
-
(Hâ2/5) ) -6.9 and δdip(Hâ3/4) ) -6.7 at 298 K, which
are too small to account for the experimental data.
Rather, σ delocalization should be the reason. It shifts
all signals to high frequency, and consequently,
δ
para(Hâ2/5) becomes positive while δpara(Hâ3/4) remains
negative. The NMR data of other chromocenes clearly
show that σ delocalization significantly contributes to
the signal shifts.14 Note that (Me4C5H)2Cr,12c which is
the CH analogue of (Tmp)2Cr, has a very similar 1H
NMR spectrum, although the signal splitting is smaller
(as expected from a smaller perturbation).
Once the signal assignment has been established, the
experimental shifts can be converted to δpara values by
referencing to the corresponding signals of (Tmp)2Fe.
Finally δpara - δdip yields δcon, and these data are
collected in Table 1. It is gratifying that, in retrospect,
the δcon values confirm the reasoning which led to the
πa orbital (Chart 1) as being mainly responsible for the
spin distribution within the ligands.
F igu r e 3. Cyclic voltammogram of (Tmp)2Cr (1.6 × 10-3
mol-1 in propionitrile, 25 °C). Supporting electrolyte, 0.1
M n-Bu4NPF6; scan rate, 200 m Vs-1, scale relative to Cp2-
Co/Cp2Co+.
and ϑ the reduced paramagnetic signal shift). This
behavior has been observed previously.16
With these facts in mind, the 1H NMR spectra of
(Tmp)2Cr were recorded between 243.5 and 374.0 K. The
data was converted to ϑ values, as mentioned above,
and plotted against T (Figure 2). In contrast to simple
substituted chromocenes, ϑ for (Tmp)2Cr is essentially
constant with T. This implies that only one of the πa/
πs-containing MOs (i.e., πa cf. above) determines the spin
distribution within the ligand. This in turn confirms
the strong perturbation which is introduced on going
from Cp2Cr to (Tmp)2Cr.
The concentration of spin at certain ligand nuclei can
be verified independently by temperature-dependent
NMR studies. Suppose that the MOs relevant for the
spin distribution contain the ligand orbitals πa and πs,
as mentioned (Chart 1), and that the energy separation
between them is comparable to the thermal energy with
the πa-containing MO being more important than the
other. On increasing the temperature, the population
of the MOs is changed in such a way that the πs
contribution becomes more important. As a conse-
quence, the spin should increase on C2/5 and decrease
on C3/4 and the NMR signal shifts should no longer
follow the Curie law which would require δpara T-1 or
In the cyclic voltammetry experiments at 25 °C,
(Tmp)2Cr underwent a reversible oxidation at 0.56 V
(∆Ep ) 70 mV) to the monocation, a quasi-reversible
reduction at -1.07 V (∆Ep ) 95 mV) to the monoanion,
and a second, irreversible oxidation at Ep ) 2.23 V; here
a
∆Ep is the separation Ep - Ep between the anodic and
a
c
cathodic peak potential, respectively, and all potentials
are relative to the potential of the internal couple Cp2-
Co/Cp2Co+. The potential of the first two electron
transfers was about 0.3 V higher than that for Cp2Cr
in the same solvent.17 This must be ascribed to the
δ
paraT/298 ) ϑ ) constant (T is the absolute temperature
(14) For alkylated chromocenes, the 13C NMR signal shifts of the
ring-C atoms (C1-C5) are less shifted than those of the respective next-
neighbor carbons of the substituent (CR).11a-c The reason is that on
C1-C5 the spin resulting from π and σ delocalization has opposite
signs while it has the same sign at CR. The same phenomenon was
encountered previously for other paramagnetic sandwich compounds.15
(15) Hebendanz, N.; Ko¨hler, F. H.; Mu¨ller, G.; Riede, J . J . Am. Chem.
Soc. 1986, 108, 3281-3289. Ko¨hler, F. H.; Metz, B.; Strauss, W. Inorg.
Chem. 1995, 34, 4402-4413.
(16) Ko¨hler, F. H.; Geike, W. J . Magn. Reson. 1983, 53, 297-302.
Ko¨hler, F. H.; Cao, R. D.; Manlik, G. Inorg. Chim. Acta 1984, 91, L1-
L2. Note that the shift signs were changed since this publication.
(17) (a) Atzkern, H.; Bergerat, P.; Fritz, M.; Hiermeier, J .; Hudeczek,
P.; Kahn, O.; Kanellakopulos, B.; Ko¨hler, F. H.; Ruhs, M. Chem. Ber.
1994, 127, 277-286. (b) Lemoine, P.; Gross, M.; Braunstein, P.;
Mathey, F.; Deschamps, B.; Nelson, J . H. Organometallics 1984, 3,
1303-1307.

Octamethyl-1,1′-diphosphachromocene
Organometallics, Vol. 16, No. 21, 1997 4609
Figu r e 4. X-band EPR spectrum of [(Tmp)2Cr]+[B(C6H5)4]-
(powder at 135 K).
F igu r e 6. Temperature dependence of the 1H NMR signal
shifts of [(Tmp)2Cr]+[B(C6H5)4]- represented as a ϑ vs T
plot (see text).
cases, the NMR signal shifts should follow the Curie
law. This was confirmed by temperature-dependent
proton NMR measurements. Figure 6 shows that the
reduced paramagnetic signal shifts are almost constant,
as expected. It follows that, unlike in (Tmp)2Cr, the spin
density must be distributed more evenly on the ligand
atoms of [(Tmp)2Cr]+, including phosphorus. Both
(Tmp)2Cr and its cation are, therefore, susceptible to
transfering spin to neighboring molecules via the P
donor.
F igu r e 5. 1H NMR spectrum of [(Tmp)2Cr]+[B(C6H5)4]-
in CD3NO2 at 303 K. S and X are the solvent and impurity,
respectively; scale in ppm.
combined effect of the heteroatom and the methyl
substituents.17b
Exp er im en ta l Section
Chemical oxidation with [(MeC5H4)2Fe]+[B(C6H5)4]-
gave [(Tmp)2Cr]+[B(C6H5)4]- as an air-sensitive red
powder in 65% yield. Because it is a Cr(III) sandwich
complex it must have three unpaired electrons, and this
is reflected in both the EPR and NMR spectra. [(Tmp)2-
Cr]+ salts are soluble in polar solvents, whereas they
decompose readily when the solvent is a good donor.
Therefore, the solid-solution EPR spectra showed vary-
ing amounts of impurities. A clean spectrum was
obtained from [(Tmp)2Cr]+[B(C6H5)4]- powder (Figure
4); the surprisingly good resolution might be due to the
separation of the paramagnetic cations by the bulky
anions. The features centered at 160, 180, 215, 330, and
530 mT are similar to what is expected for S ) 3/2 ions
having an (almost) isotropic g factor and a zero-field
The synthetic work and the characterization were carried
out under purified dinitrogen by using combined Schlenk/
cannula techniques with dry, oxygen-free solvents. The mass
spectrum was obtained with a Varian MAT 311 spectrometer.
The EPR spectrum was measured with a J eol J ES RE 2X
spectrometer working at 9.04 GHz; MnO diluted in MgO was
used for calibration. The NMR spectra were run with Bruker
MSL 300 and J eol J NM GX 270 spectrometers. The equip-
ment used for the cyclic voltammetry measurements has been
described previously.20 Traces of water were removed from
the solution by passing it through a column with activated
-
alumina integrated in the electrolysis cell, and Cp2Co+PF6
was added as an internal standard. The elemental analyses
were carried out in Inorganic Microanalytical Laboratory at
Garching, Germany.
Bis(tetr a m eth ylp h osp h olyl)ch r om iu m ((Tm p )2Cr ). A
solution of 1.16 g (6.5 mmol) of Me4C4PK7 in 80 mL of THF
was prepared at 25 °C, and 0.59 g (3.0 mmol) of CrCl2(THF)
was added. The mixture, which became dark red-brown
immediately, was stirred overnight, the solvent was stripped,
and the residue was extracted with diethyl ether. Filtration
via cannula and prolonged cooling to -35 °C gave 0.8 g (80.7%
relative to CrCl2(THF)) of red crystals of (Tmp)2Cr, mp 157-8
°C. MS (m/ z): 330 (100, M+), 315 (6, M+ - CH3), 190 (4, M+
- C8H12P), 165 (4, M2+). Anal. Calcd for C16H24CrP2: C, 58.18;
H, 7.32; Cr, 15.75; P, 18.75. Found: C, 58.02; H, 7.43; Cr,
15.62; P, 18.24.
Bis(tetr a m eth ylp h osp h olyl)ch r om iu m Tetr a p h en ylb-
or a te. ([(Tm p )2Cr ]+[B(C6H5)4]-). A 1.43 g (2.7 mmol)
amount of [(CH3C5H4)2Fe]+[B(C6H5)4]- was added quickly to
a stirred solution of 0.89 g (2.7 mmol) of (Tmp)2Cr in 50 mL of
THF. The blue solid ferrocenium salt disappeared within 1 h
while a red precipitate was formed. The solvent was removed
via cannula, and the solid was washed with 50 mL portions of
hexane until the solution was colorless. After drying in vacuo,
splitting smaller than 5 cm-1 18
.
The 1H NMR spectrum (Figure 5) showed two methyl
signals near 26 and 43 ppm and three phenyl reso-
nances in the usual range (cf. Table 1). Thus, upon
oxidation the mean paramagnetic signal shift changes
the sign. This corresponds to passing, for instance, from
(EtMe4C5)2Cr (δ(Hâav) ) -9.112b) to [EtMe4C5)2Cr]+
(δHâav ) 3.14), which both have negative spin density
on their ligands. The reason for the change of signs may
be an increase of the σ delocalization already mentioned,
while the dipolar signal shifts should be small owing to
the presumably small g-factor anisotropy.19 Compared
to (Tmp)2Cr, the proton signal splitting of [(Tmp)2Cr]+
3
is smaller. This is a general feature of S ) /2 metal-
locenes and follows from the fact that spin density
resides in both the πa- and πs-containing MOs. In such
(18) Collison, D.; Mabbs, F. E. J . Chem. Soc., Dalton Trans. 1982,
1565-1574.
(19) Kurland, R. J .; McGarvey, B. R. J . Magn. Reson. 1970, 2, 286-
(20) Atzkern, H.; Hiermeier, J .; Ko¨hler, F. H.; Steck, A. J . Orga-
nomet. Chem. 1991, 408, 281.
301.

4610 Organometallics, Vol. 16, No. 21, 1997
Feher et al.
Ta ble 2. Cr ysta l Str u ctu r e Da ta
BCrP2: C, 73.96; H, 6.83; Cr, 8.01; P, 9.54. Found: C, 73,21;
H, 6.91; Cr, 7.81; P, 9.67.
formula; mol wt
cryst description
cryst syst
C16H24CrP2; 330.31
orange plate; 0.35 × 0.35 × 0.25 mm
monoclinic
X-r a y Str u ctu r e Deter m in a tion . Crystals of (Tmp)2Cr
were grown from a diethyl ether solution of the compound by
cooling. The crystallographic details are collected in Table 2.
The structure was solved and refined by using the Enraf-
Nonius MOLEN package. The molecule occupies a crystal-
lographic symmetry center. The Cr and P atoms were located
from a Patterson map; all other atoms were obtained from a
standard Fourier synthesis. The hydrogens are occupying the
two possible conformations and were included as fixed contri-
butions in the final stages of the least-squares refinement
while using anisotropic temperature factors for all other atoms.
A non-Poisson weighting scheme was applied with a p factor
of 0.08.
space group
P21/n (No. 14)
cell constants
a ) 7.800(1) Å, b ) 12.514(1) Å
c ) 8.821(1) Å, â ) 109.34(1)°
V ) 812.41(3) Å3, Z ) 2
dcalcd ) 1.350 g cm-3
Enraf-Nonius CAD4
Lorentz-polarization
-150 ( 0.5 °C
diffractometer
corrections
temp
radiation
Mo KR (λ ) 0.710 73 Å),
graphite monochromator
µ ) 8.7 cm-1
abs coeff
F(000)
θ range
hkl ranges
348
2° e 2θ e 60.0°
h ) 0-10, k ) 0-17, l ) -12 to 11
Ack n ow led gm en t. Support from the Fonds der
Chemischen Industrie (Frankfurt), the Ecole Polytech-
nique, and the CNRS is gratefully acknowledged.
no. of reflns collected
no. of reflns included
no. of params refined
goodness of fit
R, Rw
2629 total, 2361 unique
2
2
2094 with Fo > 3.0σ(Fo
)
88
1.37
0.028, 0.061
(|Fo| - |Fc|)2,
Su p p or tin g In for m a tion Ava ila ble: Text giving the
details of the calculation of the dipolar shifts, tables containing
positional parameters, bond lengths, bond angles, and least-
squares planes, and an ORTEP diagram (8 pages). Ordering
information is given on any current masthead page.
minimization function
where w ) 4F2/σ2(F2)
4Fo2/σ2(Fo2),
least-squares function
with σ2 (F2) ) σ2(I) + (pF2)2
[(Tmp)2Cr]+[B(C6H5)4]- was obtained as a dark red microcrys-
talline powder (yield 1.15 g, 65%). Anal. Calcd for C40H44
-
OM9705618