Double silyl migration converting ORe[N(SiMe2CH2PCy2)2] to NRe[O(SiMe2CH2PCy2)2] substructures

Article abstract of DOI:10.1021/ic020424p

The reaction of (R₂PCH₂SiMe₂)₂NM (PNP^RM; R = Cy; M = Li, Na, MgHal, Ag) with L₂ReOX₃ [L₂ = (Ph₃P)₂ or (Ph₃PO)(Me₂S); X = Cl, Br] gives (PNP^Cy)ReOX₂ as two isomers, mer,trans and mer,cis. These compounds undergo a double Si migration from N to O at 90° C to form (POP^Cy)ReNX₂ as a mixture of mer,trans and fac,cis isomers. Additional thermolysis effects migration of CH₃ from Si to Re, along with compensating migration of halide from Re to Si. DFT calculations on various structural isomers support the greater thermodynamic stability of the POP/ReN isomer vs PNP/ReO and highlight the influence of the template effect on the reactivities of these species.

Full text of DOI:10.1021/ic020424p

Inorg. Chem. 2002, 41, 5615−5625
Double Silyl Migration Converting ORe[N(SiMe2CH PCy2)2] to
2
NRe[O(SiMe2CH PCy2)2] Substructures
2
,†
Oleg V. Ozerov,† He´le`ne F. Gerard,‡ Lori A. Watson,† John C. Huffman,§ and Kenneth G. Caulton*
Department of Chemistry and Molecular Structure Center, Indiana UniVersity,
Bloomington, Indiana 47405, and LSDSMS (UMR 5636), UniVersite´ de Montpellier 2,
34095, Montpellier Cedex 5, France
Received June 25, 2002
The reaction of (R PCH SiMe2)2NM (PNPRM; R ) Cy; M ) Li, Na, MgHal, Ag) with L2ReOX [L2 ) (Ph3P)2 or
2
2
3
(Ph3PO)(Me2S); X ) Cl, Br] gives (PNPCy)ReOX as two isomers, mer,trans and mer,cis. These compounds undergo
2
a double Si migration from N to O at 90 °C to form (POPCy)ReNX as a mixture of mer,trans and fac,cis isomers.
2
Additional thermolysis effects migration of CH from Si to Re, along with compensating migration of halide from Re
3
to Si. DFT calculations on various structural isomers support the greater thermodynamic stability of the POP/ReN
isomer vs PNP/ReO and highlight the influence of the template effect on the reactivities of these species.
Introduction
In this work, we report how certain bonds in these ligands
can be reactive to publicize how both N-Si and the Si-C
bonds can be cleaved under frequently encountered condi-
tions.
The design of ligands intended for use in “aggressive”
environments (strongly oxidizing or reducing or acidic
conditions, high temperature, or high flux of energetic
photons) obviously has to anticipate and avoid bonds
vulnerable to such conditions. This becomes increasingly
challenging when a polydentate ligand is multifunctional or
when co-ligands or solvent incorporate functionality that can
react with the chelate. Finally, polydentate ligand synthesis
is based on certain bonds being “easily” formed. These bonds
often are the result of nucleophilic/electrophilic combination,
and the resulting bonds are often polar (single bonds) or
reactive (e.g., CdN double bonds).
A very attractive class of ligands has been widely applied
in transition metal chemistry where a multifunctional and
monoanionic ligand is useful. This ligand class is formed
from commercially available HN(SiMe2(CH2Cl))2 and nu-
cleophilic phosphides R2P- (Scheme 1).1 These ligands
permit the incorporation of variable steric and electronic
factors, and they incorporate numerous useful NMR spectral
probes. Here, we will abbreviate N(SiMe2CH2PR2)2 as PNPR.
Results
Ligand Synthesis. The ligand employed in this study was
synthesized by using a modification of Fryzuk’s preparation1
of analogous compounds to obtain an improved yield. In the
original preparation, two-thirds of the R2PLi salt acted as a
nucleophile in the formation of two P-C bonds. The
remaining one-third of the R2PLi merely functioned as a base
to deprotonate [(PNPR)H] in situ and was ultimately lost as
R2PH during the workup. We found that repeating the initial
sequence in situ on one-third of the initial scale can convert
valuable R2PH formed in the reaction mixture to additional
chelate product (Scheme 1). To investigate ways of introduc-
ing the PNPCy ligand into the Re coordination sphere, we
sought to explore various PNPCy starting materials (Scheme
2). [(PNPCy)Li] (1a) typically contained approximately one
molecule of Et2O after crystallization. Recrystallization in
the presence of TMEDA afforded [(PNPCy)Li‚TMEDA] (1b).
Other simple derivatives of (PNPCy) can be prepared in a
straightforward manner. Cation exchange2 between [(PNPCy)-
Li‚Et2O] or [(PNPCy)Li‚TMEDA] and NaOCMe2Et in benzene/
* To whom correspondence should be addressed. E-mail: caulton@
indiana.edu.
† Department of Chemistry, Indiana University.
§ Molecular Structure Center, Indiana University.
‡ Universite´ de Montpellier. Current address: Laboratoire de Chimie
Theorique, Case, 137, Universite Pierre et Marie Curie, 4 Place Jussieu,
75252 Paris Cedex 6, France.
(2) (a) Lochmann, L.; Trekoval, J. J. Organomet. Chem. 1979, 179, 123.
(b) Lochmann, L.; Trekoval, J. Collect. Czech. Commun. 1988, 53,
76. (c) Lochmann, L. Eur. J. Inorg. Chem. 2000, 1115. (d) Ahlbrecht,
H.; Schneider, G. Tetrahedron 1986, 42, 4729.
(1) (a) Fryzuk, M. D.; Carter, A.; Westerhaus, A. Inorg. Chem. 1985, 24,
642. (b) Coalter, J. N.; Watson, L. A.; Caulton, K. G. Unpublished
results.
10.1021/ic020424p CCC: $22.00 © 2002 American Chemical Society
Published on Web 09/20/2002
Inorganic Chemistry, Vol. 41, No. 21, 2002 5615

Ozerov et al.
Scheme 1
This decomposition pathway of [(PNPR)H] leading to the
formation of R2PMe was further confirmed by the formation
of a known compound, [RuHCl(CO)(tBu2PMe)2], isolated
from the reaction of [(PNPt-Bu)Li] in MeOH (Scheme 3).
The identity of [RuHCl(CO)(tBu2PMe)2] was confirmed by
comparison of 1H, 31P, and 13C NMR data with the data for
an authentic sample.3
Scheme 2
Synthesis of Complexes. (a) From ReIII. Several com-
monly used Re starting materials were screened for their
suitability for the introduction of the PNPCy ligand. Because
of the lack of easily available ReI complexes devoid of π-acid
ligands such as CO, we focused our attention on higher Re
oxidation states. Two ReIII complexes were chosen: [(PMe2-
Ph)3ReCl3]4 (8) and [(PPh3)2(MeCN)ReCl3]5 (9). [(PMe2-
Ph)3ReCl3] (8) proved to be unreactive in salt metathesis
attempts. It was inert toward 4, 6, and 7 even on prolonged
exposure at high temperature. No reaction occurred between
8 and 1a at ambient temperature in benzene or THF, whereas
at higher temperature, [(PNPCy)H] was produced along with
unidentified byproducts (Scheme 4). The protonation of
(PNPCy) anion in the reaction with 8 occurred more readily
with 1b and especially 2. In the reactions of 1-7 with 9,
multiple products were formed in all cases. [(PNPCy)H] was
detected in the reactions of 1 or 2 with 9.
(b) From ReV. ReV starting materials proved to be much
more amenable to substitution by the PNPCy sources (Scheme
4). [(PPh3)2ReOHal3]6 (10, Hal ) Cl; 11, Hal ) Br) and
[(PPh3O)(Me2S)ReOHal3]7 (12, Hal ) Cl; 13, Hal ) Br)
reacted with anionic PNPCy compounds to yield correspond-
ing [(PNPCy)ReOHal2] complexes (14, Hal ) Cl; 15, Hal )
Br). The reaction mixtures from the reactions with 6
contained several soluble silver phosphine complexes, as was
the case in reactions of 6 with 9. The reaction of 7 with
10-13 produced mixtures of unidentified products. 1 and 4
(or 5) performed the best and provided 14 (15) in excellent
pentane mixtures afforded crystalline [(PNPCy)Na] (2). Unlike
the Li derivative, [(PNPCy)Na] is very poorly soluble in
pentane, benzene, and ether and is moderately soluble in
THF. [(PNPCy)K] (3) can be similarly prepared by cation
exchange with KOR (R ) menthoxide), but the apparently
high solubility of [(PNPCy)K] precluded its isolation. Reac-
tions between [(PNPCy)Li] or [(PNPCy)Na] and anhydrous
MgHal2 provide [(PNPCy)MgHal] (4, Hal ) Cl; 5, Hal )
Br), which could be isolated as a solid or used in situ. We
were unable to prepare [(PNPCy)2Mg], presumably because
of the prohibiting steric bulk of the PNPCy ligand. [(PNPCy)-
Ag] (6) was prepared from [(PNPCy)Na] and AgOTf in THF.
[(PNPCy)H] (7) was prepared by protonation of the Li or Na
salt with 1 equiv of anhydrous HCl or Et3NHCl and was
used in situ.
Si-C and Si-N Bond Scission by Protolysis. It was
found here that pincer ligands of the general formula HN-
(SiMe2CH2PR2)2 [or its anionic form LiN(SiMe2CH2PR2)2],
undergo Si-C cleavage, as well as cleavage of the Si-amide
bond in the presence of excess alcohol. Four equivalents of
alcohol can be consumed in the protolysis of 7 (Scheme 3).
For example, reaction of 7 with phenol (ratio 1:5.3) in
benzene is complete within 14 h at ambient temperature and
6 h at 60 °C (see Supporting Information), and forms Cy2-
PMe and SiMe2(OPh)2 as confirmed by GC-MS and NMR
analyses. The third product is assumed to be NH3. Similar
reactivities are observed with other PNP ligands: HN(SiMe2-
(3) Gill, D. F.; Shaw, B. L. Inorg Chim. Acta 1979, 32, 19.
(4) (a) Parshall, G. W. Inorg. Synth. 1977, 17, 111. (b) Douglas, P. G.;
Shaw, B. L. Inorg. Synth. 1977, 17, 64.
(5) Rouschias, G.; Wilkinson, G. J. Chem. Soc. A 1967, 993.
(6) (a) Johnson, N. P.; Lock, C. J. L.; Wilkinson, G. J. Chem. Soc. 1964,
1054. (b) Johnson, N. P.; Lock, C. J. L.; Wilkinson, G. Inorg. Synth.
1967, 9, 145. (c) Parshall, G. W. Inorg. Synth. 1977, 17, 110.
(7) In the original preparation, the compounds were assigned as (PPh3)(Me2-
SO)ReOX3.8a Later, it was shown that these compounds are, in fact,
(Ph3PO)(Me2S)ReOX3.8b (a) Grove, D. E.; Wilkinson, G. J. Chem.
Soc. A 1966, 1224. (b) Bryan, J. C.; Stenkamp, R. E.; Tulip, T. H.;
Mayer, J. M. Inorg. Chem. 1987, 26, 2283.
t
CH2PBut2)2 is decomposed to Bu2PMe and SiMe2(OR′)2
upon treatment with excess alcohol.
5616 Inorganic Chemistry, Vol. 41, No. 21, 2002

Double Silyl Migration in ORe[N(SiMe2CH2PCy2)2]
Scheme 3
Scheme 4
spectra, located in an otherwise desolate region, can be
informative. Consistent with the mer,trans assignment, 14a
and 15a each displays a single resonance in the 31P NMR
1
spectrum and a single Si-Me resonance in the H NMR
spectrum. Each 14b and 15b displays a single 31P NMR
resonance and two Si-Me resonances of equal intensity in
1
the H NMR spectrum. This is consistent with the mer,cis
assignment; however, a fac geometry for the PNPCy ligand
for 14b and 15b cannot be ruled out on the basis of these
data alone. The following experiment was useful in elucidat-
ing this issue. Reaction of 14 with MgBr2 in C6D6/Et2O (Re/
Mg ) 1:1) yields products of halide scrambling. Both
isomers of each 14 and 15 are observed in the mixture by
31P NMR spectroscopy, along with three other singlets that
can be assigned to [(PNPCy)ReOClBr] (16). The fact that
each of these three signals is a singlet is consistent with only
the mer arrangement of the PNPCy ligand because any isomer
of 16 with the fac-PNPCy arrangement should display in
equivalent phosphines by 31P NMR spectroscopy. The
preference for the mer-PNP geometry is further supported
by the computational results (vide infra).
Thermolytic Si-N Bond Scission and Silyl Migration.
14 and 15 are thermally unstable and undergo a remarkable
N/O transposition reaction at elevated temperatures (Scheme
6). Heating separate samples of 14 and 15 at 80-100 °C in
C6D6 for 24 h results in the formation of the new complexes
17 and 18, respectively. Two isomers of each 17 and 18 are
observed and the exact ratio depends on the handling history.
The interconversion between the isomers at ambient tem-
perature is slow in solution and does not occur in the solid
yield as judged by NMR spectroscopy. 12 (13) is a more
suitable precursor than 10 (11) because of the easier
separation from byproducts.
(c) Stereoisomers. Green complexes 14 and 15 were
isolated as mixtures of mer,trans (14a, 15a) and mer,cis (14b,
15b) isomers (Scheme 5). The PNPCy ligand offers certain
NMR spectroscopic probes to determine the symmetry of
PNP complexes. In addition to the 31P NMR signals, the
1
number of signals from the Si-Me groups in the H NMR
Scheme 5
Inorganic Chemistry, Vol. 41, No. 21, 2002 5617

Ozerov et al.
Scheme 6
state. 17a and 18a each display a single resonance in the
31P NMR spectrum and a single Si-Me resonance in the 1H
NMR spectrum. This is consistent with the mer,trans
geometry. 17b and 18b each display a single resonance in
the 31P NMR spectrum and two Si-Me resonances of equal
heating a sample of 14 for 2 days at 90 °C produced a
mixture of 14a, 14b, 17a, and 17b. These results imply that
not only the transformation of 14 into 17 but also the
interconversion between the isomers of 14 and 17 is possible
at elevated temperatures in the solid state.
1
intensity in the H NMR spectrum. This is consistent with
Methyl/Halide Interchange. Even more remarkable is the
fact that the nitrido complexes 17 and 18 are not the final
thermodynamic products. NMR analysis of C6D6 solutions
of 17 that were heated at 80-110 °C for several days
revealed the formation of two new products (Scheme 7).
These two compounds displayed AB patterns in their 31P
NMR spectra with almost identical chemical shifts and JPP
coupling constants (20a: 24.7, 22.8 ppm, JPP ) 168 Hz.
20b: 24.8, 22.7 ppm, JPP ) 166 Hz). From the fact that
each of these new compounds shows three Si-Me signals
of equal intensity, we conclude that one of the methyl groups
on the Si atom is no longer present on the pincer ligand
backbone in 20. Indeed, 13C NMR analysis of the mixture
revealed the presence of two resonances at -14.8 and -15.5
ppm, each appearing as a triplet due to coupling to the two
P atoms (JCP ) 5.5 Hz).8 13C DEPT experiments demon-
both the fac,cis and the mer,cis geometries. A halide
scrambling experiment (Scheme 6) produced two isomers
of each 17, 18, and 19. A singlet resonance in the 31P NMR
spectrum corresponded to 19a (mer-POP), but the other
isomer of [(POPCy)ReNBrCl], 19b, displayed an AB pattern
with a JPP typical of cis-coordinated phosphines, supporting
a fac-POP geometry. An X-ray diffraction study confirmed
a fac,cis geometry for 17b (Figure 1) and a mer,trans
geometry for 18a (Figure 2). The preference for the observed
isomers is also supported by computational evidence (vide
infra). The thermal transformation of 14 (15) into 17 (18)
also occurs in the solid state, albeit more slowly. Thus,
heating a sample of 14 for 5 days at 115 °C produced 17a
at 95% conversion with virtually no 17b detected. However,
Figure 1. ORTEP28 view of 17b drawn with 50% probability ellipsoids.
Figure 2. ORTEP28 view of 18a drawn with 50% probability ellipsoids.
Hydrogen atoms are omitted for clarity.
Hydrogen atoms are omitted for clarity.
5618 Inorganic Chemistry, Vol. 41, No. 21, 2002

Double Silyl Migration in ORe[N(SiMe2CH2PCy2)2]
Scheme 7
Scheme 8
strated that each of these signals corresponds to a CH3 group.
The magnitude of the JPP coupling constant rules out a cis
arrangement of the phosphine arms. It is consistent with the
transoid geometry similar to that observed in the structure
of 18a. We therefore assign these new products as 20a and
20b, the two possible diastereomers for the proposed
structure: the Cl substituent on the Si atom is either syn or
anti to Me on Re. We could not achieve more than 50%
conversion to 20 even after 8 days of heating 17 at 90 °C,
so we could not isolate 20 in a pure form. Heating 17 at
higher (120 °C) temperatures in toluene-d8 did not provide
for a higher conversion, so it is likely that, at high
temperatures, 17 and 20 exist in an equilibrium that is
temperature-dependent. A similar thermal transformation
occurs with the dibromo analogue 18, but it requires longer
reaction times and even after 7 days at 90 °C, the concentra-
tion of the proposed 21 was quite small.
in particular, employ alcohols, water, and acids as solvents
or reagents.11 It is evident then that the resistance of the
ancillary ligands to such conditions must be established. We
found that alcohols and water easily disassemble the PNP
ligand framework. The PNPCy moiety can be regarded as a
derivative of the (Me3Si)2N parent in which two of the
hydrogens are substituted by a Cy2P group. The cleavage of
the N-Si bonds is then to be expected because hexameth-
yldisilazane is a well-known silylation reagent.12 The cleav-
age of the Si-CH2PR2 bond also has precedence. Facile silyl-
group cleavage from a carbon atom attached to a phosphorus
substituent has been reported for Ph2PCH2SiMe2H (Scheme
9).13 The overall protolytic decomposition of the PNP
framework is rationalized in terms of the kinetic lability of
these Si-N and Si-C bonds and the high thermodynamic
affinity of Si to form Si-O bonds.
Displacement of the POPCy Ligand by Et3P. The
reduction of ReV oxo complexes by deoxygenation with
trialkylphosphines is a common technique;5-7,9 however,
complexes 14 and 15 show no reaction with Et3P at ambient
temperature. At 80 °C in the presence of 3 or more
equivalents of Et3P, 14 (15) is quantitatively (NMR) con-
verted into 22 (23)10 and free POPCy (24) in several hours
(Scheme 8). Thus, heating 14 (15) with a simple phosphine
results in N-to-O silyl migration and displacement of a
tridentate ligand! Control experiments showed that 17b (18b)
reacts with an excess of Et3P at ambient temperature in the
time of mixing to form 22 (23) and 24, whereas 17a (18a)
requires longer reaction times and elevated temperature.
Evidently, excess Et3P displaces POPCy from 17 and 18
formed in situ from 14 or 15. [(Et3P)3ReNHal2] (22, 23)10
and 24 were identified on the basis of NMR and mass-
spectrometry data.
(10) (a) Chatt, J.; Garforth, J. D.; Johnson, N. P.; Rowe, G. A. J. Chem.
Soc. 1964, 1012. (b) Chatt, J.; Falk, C. D.; Leigh, G. J.; Paske, R. J.
J. Chem. Soc. A 1969, 2288.
Discussion
(11) (a) Baudry, D.; Ephritikhine, M.; Felkin, H. J. Organomet. Chem. 1982,
224, 363. (b) Kelle Zeiher, E. H.; DeWit, D. G.; Caulton, K. G. J.
Am. Chem. Soc. 1984, 106, 7006. (c) Jones, W. D.; Maguire, J. A.
Organometallics 1987, 6, 1301. (d) Carr, S. W.; Fowles, E. H.;
Fontaine, X. L. R.; Shaw, B. L. J. Chem. Soc., Dalton Trans. 1990,
573. (e) Smith, K. J.; Ondracek, A. L.; Gruhn, N. E.; Lichtenberger,
D. L.; Fanwick, P. E.; Walton, R. A. Inorg. Chim. Acta 2000, 300-
302, 23. (f) Abel, E. W.; Bennett, M. A.; Wilkinson, G. J. Chem.
Soc. 1959, 3178. (g) Esteruelas, M. A.; Werner, H. J. Organomet.
Chem. 1986, 303, 221. (h) Gruenwald, C.; Gevert, O.; Wolf, J.;
Gonzalez-Herrero, P.; Werner, H. Organometallics 1996, 15, 1960.
(12) (a) Cossy, J.; Pale, P. Tetrahedron Lett. 1987, 28, 6039. (b)
Firouzabadi, H.; Karimi, B. Synth. Commun. 1993, 23, 1633. (c)
Karimi, B.; Golshani, B. J. Org. Chem. 2000, 65, 7228.
Protolytic Decomposition. Many common preparative
methods in the chemistry of later transition metals, and Re
(8) Me groups in the following ReV complexes were reported to resonate
in the negative region of the 13C NMR spectrum: Cp*Re(O)MeX (X
- 9c
) Cl, alkyl);9a,b trans-[(R3P)2Re(O)Me2]+ BF4
. (a) Herrmann, W.
A.; Floeel, M.; Herdtweck, E. J. Organomet. Chem. 1988, 358, 321.
(b) Herrmann, W. A.; Felixberger, J. K.; Anwander, R.; Herdtweck,
E.; Kiprof, P.; Riede, J. Organometallics 1990, 9, 1434. (c) Hoffman,
D. M.; Wierda, D. A. Polyhedron 1989, 8, 959.
(9) (a) Conry, R. R.; Mayer, J. M. Inorg. Chem. 1990, 29, 4862. (b)
Seymore, S. B.; Brown, S. N. Inorg. Chem. 2000, 39, 325. (c)
Chakraborty, I.; Bhatacharyya, S.; Banerjee, S.; Dirghangi, B. K.;
Chakravorty, A. J. Chem. Soc., Dalton Trans. 1999, 3747. (d)
Bhatacharyya, S.; Chakraborty, I.; Dirghangi, B. K.; Chakravorty, A.
Inorg. Chem. 2001, 40, 286.
(13) (a) Brost, R. D.; Bruce, G. C.; Grundy, S. L.; Stobart, S. R. Inorg.
Chem. 1993, 32, 5195. (b) Brost, R. D.; Bruce, G. C.; Stobart, S. R.
J. Chem. Soc., Chem. Commun. 1986, 21, 1580. (c) Eaborn, C.; Retta,
N.; Smith, J. D. J. Chem. Soc., Dalton Trans. 1983, 905.
Inorganic Chemistry, Vol. 41, No. 21, 2002 5619

Ozerov et al.
Zr, Ta, or Mo.18 The rearrangement of 14 (15) into 17 (18)
is thermodynamically favorable and irreversible, but it
apparently has a high activation energy, as the transformation
is slow even at elevated temperatures. The thermodynamic
driving force behind the silyl migration is doubtless the
formation of strong Si-O bonds. It is likely that the process
occurs as a nucleophilic attack of the oxygen of the oxo
group on the Si center. Presumably, this demands a cis
geometry of the oxo and amido groups and an approximate
syn orientation of the O and Si atoms. The high activation
energy might be due to the deformation of the PNP
framework required to bring the Si atom into the proximity
of the oxo group. Perhaps, dissociation of a phosphine arm
is necessary to accomplish this step. The favorable thermo-
dynamics of the migration of a closely related silyl group in
a PNP ligand from an amide nitrogen to an acyl oxygen have
been reported;19 in that case, the reaction is even faster than
those we report.
Solid-State Structures of 17b and 18a. An X-ray
diffraction study (Figure 1) revealed that 17b exhibits an
approximate tetragonal pyramid geometry about the Re
center, with the nitrido ligand occupying the apical position,
if the Re-O interaction is not taken into account. The Re-O
distance of 2.856(3) Å is quite long and can only be
considered as incipient or weak bonding. However, the sum
of the van der Waals radii of O (1.42 Å) and Re (using the
literature20 value for Pt, 1.75 Å) is 3.17 Å, which ensures
that the 2.856(3) distance does involve some interaction.
Disilyl ethers are weak O-donors, and the Re-O bond is
further weakened by the strong trans influence of the nitrido
ligand. The weak interaction that we believe exists between
the Re and O atoms is undoubtedly made possible by the
chelate effect. Computational results help address this issue
(vide infra). The angles between the nitrido ligand and each
of the basal ligands are greater than 90° as is typical for
metal nitride structures.21 The angle between the cis-
phosphines of 102.97(3)° is larger than the angle of 80.73-
(3)° between cis-chlorides, presumably a consequence of the
steric bulk of the phosphine groups. The Re-N bond distance
of 1.651(3) Å is typical for rhenium nitrido complexes.21
Scheme 9
Influence of the Cation on the Reactivity of PNPCy
Salts. The difference in behavior of Li, Mg, Na, and Ag
derivatives toward various Re starting materials can be
justified by the following arguments. All of the donor atoms
of the PNPCy ligand in [(PNP)Ag] are covalently bound to
the silver atom, which makes it a poor PNPCy transfer
reagent. In addition, the propensity of silver to form mixed-
halide bridged phosphine complexes is well-documented,14
and it complicates the reactions where the transfer of the
PNPCy ligand to Re does occur. In the series of 1, 2, 4, and
5, the difference in reactivity can be justified by the
differences in the relative basicities and nucleophilicities of
these compounds. Hexamethyldisilazide salts are known as
poorly nucleophilic Brønsted bases because of the bulk of
the anion;15 introduction of the voluminous Cy2P groups only
increases steric encumbrance. It is possible that the first step
in the introduction of the PNPCy moiety into the Re
coordination sphere occurs via coordination of one of the
phosphine arms and not of the amido fragment. Indeed, 9-13
are known to react easily with neutral ligands such as
phosphines by ligand displacement because of the lability
of the uncharged ligands. Where such ligand displacement
is kinetically retarded, such as in the case of 8, secondary
modes of reactivity can be explored by the PNPCy compound.
Deprotonation of a coordinated phosphine is one alternative
pathway, which is apparently operative in the reactions of 8
with 1 and 2 where we observed the formation of [(PNP-
Cy)H]. It might appear that the basicity of a disilylamido is
insufficient to remove a proton from a weakly acidic
coordinated PhPMe2.16 Yet, most likely, the deprotonation
is thermodynamically driven by the formation of LiCl or
NaCl and not by the protonation of a stronger base, i.e., 1
or 2 assists in the E2 elimination process. It is also well-
documented that sodium amides and alkyls act as stronger
bases than their Li counterparts2 and that TMEDA complex-
ation increases the basicity of alkyllithiums.17 The Mg
derivative can be expected to be an even milder, less basic
reagent, and 4 simply does not react with 8 at all.
(18) (a) (i) O’Donoghue, M. B.; Davis, W. M.; Schrock, R. R. Inorg. Chem.
1998, 37, 5149. (ii) Freudlich, J. S.; Schrock, R. R.; Davis, W. M. J.
Am. Chem. Soc. 1996, 118, 3643. (iii) Schrock, R. R.; Seidel, S. W.;
Schrodi, Y.; Davis, W. M. Organometallics 1999, 18, 428. (b) Me3-
Si-substituted amines are often used for the introduction of an amido
ligand into a Ti coordination sphere, generally accompanied by
elimination of Me3SiCl from a Me3Si-substituted amine and TiCl4.
For Me3Si-N vs Ti-Cl metathesis, see: (i) Scollard, J. D.; McCo-
nville, H. D.; Payne, N. C.; Vittal, J. J. Macromolecules 1996, 29,
5241. (ii) Guerin, F.; McConville, H. D.; Payne, N. C. Organometallics
1996, 15, 5085.
Thermolysis of Re Complexes. There is precedent for
cleavage of N-Si bonds when amide N is coordinated to
(14) (a) Sanghani, D. V.; Smith, P. J.; Allen, D. W.; Taylor, B. F. Inorg.
Chim. Acta 1982, 59, 203. (b) Attar, S.; Alcock, N. W.; Bowmaker,
G. A.; Frye, J. S.; Bearden, W. H.; Nelson, J. H. Inorg. Chem. 1991,
30, 4166. (c) Baker, L.-J.; Bowmaker, G. A.; Camp, D.; Effendy;
Healy, P. C.; Schmidbaur, H.; Steigelmann, O.; White, A. H. Inorg.
Chem. 1992, 31, 3656.
(19) Fryzuk, M. D.; MacNeil, P. A. J. Am. Chem. Soc. 1984, 106, 6993.
(20) Bondi, A. J. Phys. Chem. 1964, 68, 441.
(15) (a) Stork, G.; Uyeo, S.; Wakamatsu, T.; Grieco, P.; Labovitz, J. J.
Am. Chem. Soc. 1971, 93, 4945. (b) Rathke, M. W. Org. Synth. Coll.
1988, 6, 598.
(16) (a) pKa of HN(SiMe3)2 was determined to be 25.8 in THF: Fraser R.
R.; Mansour, T. S.; Savard, S. J. Org. Chem. 1985, 50, 3232. (b)
Acidity of coordinated phosphines can be approximated by using the
data on acidity of phosphine oxides: Streitwieser, A.; Juaristi, E. J.
Org. Chem. 1982, 47, 768.
(21) (a) Nugent, W. A.; Mayer, J. M. Metal-Ligand Multiple Bonds;
Wiley: New York, 1988. (b) Cundari, T. R. Chem. ReV. 2000, 100,
807. (c) Chatt, J.; Falk, C. D.; Leigh, G. J.; Paske, R. J. J. Chem. Soc.
A 1969, 2288. (d) Forsellini, E.; Casellato, U.; Graziani, R.; Magon,
L. Acta Crystallogr. B 1982, 38, 3081. (e) Neuhaus, A.; Veldkamp,
A.; Frenking G. Inorg. Chem. 1994, 33, 5278. (f) Abram, U.; Lang,
E. S.; Abram, S.; Wegmann, J.; Dilworth, J. R.; Kirmse, R.; Woollins,
J. D. J. Chem. Soc., Dalton Trans. 1997, 623. (g) For Tc nitrido
complexes, closely related to 17 and 18, see: Marchi, A.; Marvelli,
L.; Rossi, R.; Magon, L.; Uccelli, L.; Bertolasi, V.; Ferretti, V.;
Zanobini, F. J. Chem. Soc., Dalton Trans. 1993, 1281.
(17) (a) Rausch, M. D.; Ciappenelli, D. J. Organomet. Chem. 1967, 10,
127. (b) Chadwick, S. T.; Rennels, R. A.; Rutherford, J. L., Collum,
D. B. J. Am. Chem. Soc. 2000, 122, 8640.
5620 Inorganic Chemistry, Vol. 41, No. 21, 2002

Double Silyl Migration in ORe[N(SiMe2CH2PCy2)2]
+
Figure 3. Calculated (DFT) optimized structures and energies (kcal/mol) of isomeric isoelectronic (H3P)2ReOCl2 and (H3P)2ReNCl2.
Table 1. Selected Bond Distances and Angles in the
The solid-state structure of 18a shows a different disposi-
Crystallographically Determined Structures of 17b and 18a and in the
Calculated Optimized (DFT) Structures of T4b and T4c
tion of the ligands about Re (Figure 2). The POPCy ligand
in 18a adopts a mer geometry. The sum of the angles about
the O atom in the structures of both 17b and 18a is almost
exactly 360°. This is consistent with the sp2-hybridized lone
17b, XRD T4c, DFT 18a, XRD T4b, DFT
distances, Å
Re-N
1.651(3)
2.3907(7)
2.4190(7)
2.4128(8)
2.4276(8)
2.856(3)
1.648
2.428
2.428
2.392
2.392
2.809
1.664(2)
2.5503(3)
2.5600(3)
2.4611(7)
2.4331(7)
2.739(2)
1.643
2.423
2.434
2.417
2.417
3.013
Re-Cl(Br)1
Re-Cl(Br)2
Re-P1
pair of the O atom being suitably directed for the interaction
with the site trans to the nitrido ligand. However, the
N-Re-O angle is closer to 180° in 18a [177.56(12)°] than
in 17b [165.39(12)°], and the Re-O distance of 2.739(2) Å
is shorter than in 17b. This apparently stronger Re-O
interaction in 18a as compared to 17b might be due to the
different geometric restrictions imposed by the chelate
framework in these structures. The Re-N distance of 1.664-
(2) Å in 18a is slightly longer than that in 17b, and it might
reflect stronger binding of the O trans to the nitrido ligand
in 18a. The key geometric parameters for the structures of
17b and 18a are listed in Table 1.
Re-P2
Re-O
angles, degrees
P1-Re-P2
102.97(3)
Cl(Br)1-Re-Cl(Br)2 80.73(3)
98.45
85.45
95.22
162.73(2)
154.209(10) 146.10
156.78
N-Re-P1
N-Re-P2
N-Re-Cl(Br)1
N-Re-Cl(Br)2
N-Re-O
92.47(9)
96.07(9)
110.85(9)
103.44(9)
165.39(12)
101.31(9)
95.92(9)
103.40(8)
102.39(8)
177.56(12)
101.58
101.58
106.81
107.08
178.08
95.22
109.85
109.85
159.15
are slightly larger than 90° and differ by less than 10° in T1
and T2. The isomers with trans phosphines (T1a, T2a) are
C2V-symmetric and more stable than the isomers of Cs
symmetry with cis-phosphines (T1b, +9.7 kcal/mol; T2b,
+4.5 kcal/mol). The preference for the trans isomer is twice
as large for the oxide T1a as for the nitride T2a.
(b) Structures of [(PNPH)ReOCl2] (T3) and [(POPH)-
ReNCl2] (T4). The experimental PNPCy ligand was modeled
as N(SiH2CH2PH2)2 (PNPH). An analogous model used for
the POPCy ligand was O(SiH2CH2PH2)2 (POPH). Four
structures were examined for both T3 and T4. The T4
structures can be deduced from the T3 structures by
exchanging O and N. These optimized structures are shown
in Figure 4. Because all of these species are isomers, a unique
energy reference can be used to compare them.
Computational Results
(a) Monodentate Analogues. To obtain better insight
into the relative stabilities of the species studied experimen-
tally, compounds [ReOCl2(PH3)2]+ (T1) and [ReNCl2(PH3)2]
(T2) were studied using DFT (B3PW91) methods. This
allowed for an examination of the structures and their relative
energies when no cyclic strain between the phosphine ligands
and no additional interaction (Re‚‚‚O or Re‚‚‚N) are present.
Both T1 and T2 adopt tetragonal pyramidal structures, with
the multiply bonded atom (O and N, respectively) occupying
the apical site (Figure 3). Two different structures are
possible: the phosphine ligands can be either trans to each
other (T1a, T2a) or cis to each other (T1b, T2b). The
structures of T1 and T2 are very similar. The Re-O and
Re-N bonds differ by less than 0.01 Å (1.645 vs 1.651 Å).
The angles between the apical ligand and the basal ligands
All T4 structures are more stable than any T3 structure,
in agreement with experiment. Isomerization of T3 to T4,
forming a POPH complex, is thus thermodynamically favored.
Inorganic Chemistry, Vol. 41, No. 21, 2002 5621

Ozerov et al.
distance in the T4 analogue, whereas the Re-O bond in the
T3 species is approximately the same length as the Re-N
bond in the T4 species. This shows that the influence of the
strength of the Re-N bond in T3 is much stronger than that
of Re-O in T4, as suggested by a simple Lewis structure
analysis of these complexes (dative interaction for Re‚‚‚O
in T4 but a single bond in T3).
Two isomers were obtained with a fac arrangement of the
PNPH (POPH) in T3 (T4). The structures T3c and T4c
(analogue of 17b, 18b), with both P atoms trans to Cl, are
quasi-isoenergetic with T3b and T4b, respectively, which
shows that structures with a fac arrangement are also
energetically accessible. Structure T3d has only one P atom
trans to Cl; the other phosphine is trans to O. Even though
the Re-N distance is quite short, T3d is a high-energy
species. A T4 analogue of T3d with one P trans to N and O
interacting trans to Cl could not be found: dissociation of
the O atom (Re‚‚‚O ) 4.14 Å) allows coordination of both
P and Cl in the basal sites. Consequently, T4c and T4d
exhibit similar coordinations, except that O interacts with
the vacant site in T4c and not in T4d. The energy difference
between these two species can thus be used to evaluate the
strength of the Re‚‚‚O interaction when it is trans to N.
Because T4c is only 3.7 kcal/mol more stable than T4d, this
feature confirms the hypothesis of a very weak Re‚‚‚O
interaction.
(c) Rationale for Structural Preferences. The basic
structures of the monodentate complexes T1 and T2 show a
modest preference for the trans arrangement of the phos-
phines; yet T3b and T3c are isoenergetic, and T4b and T4c
are likewise isoenergetic. This indicates that the PXPH
backbone (X ) N, O) favors the fac geometry.
The most important factor determining the relative stabili-
ties of the isomers of T3 and T4 structures is the strong
trans influence of the oxo or nitrido ligand.21 A bond to a
ligand trans to oxo or nitrido will be considerably weakened.
This is evident from the fact that, in all cases, the bond to a
donor atom trans to an oxo or a nitrido group is substantially
longer than an analogous bond in an equatorial position.
Therefore, those structures should be more stable that have
the weakest ligand trans to a multiply bonded atom. In a
sense, the bond to such a ligand is sacrificed to the strong
trans influence ligand so that the bonds to other ligands have
the opportunity to contribute the greatest thermodynamic
gain. For the T3 series, the best “sacrificial” ligand is the
chloride; the T3a structure is the most stable. For the T4
series the coordination of the weakest ligand, the Si-O-Si
oxygen, trans to the nitrido group gives more stable
structures. Indeed, equivalents of T4b (17a, 18a) and T4c
(17b, 18b) are the structures observed experimentally.21g The
analogues of T3a (14b, 15b) and T3b (14a, 15a) are the
two isomers observed in the experiment for the Re oxo
compounds. It should be noted, however, that the introduction
of real alkyl substituents on the PNP ligand instead of the
model Hs will destabilize the T3a (repulsion between
eclipsed Si-Me groups) and the T3c (cis Cy2P groups)
structures. Thus, PNPCy analogues of T3a and T3b should
Figure 4. Calculated (DFT) optimized structures and energies (kcal/mol)
of isomeric isoelectronic [(H2PCH2SiH2)2N)]ReOCl2 and H2PCH2SiH2)2O)]-
ReNCl2.
The globally most stable isomer is T4b (analogue of 17a,
18a), with a mer arrangement of the POP ligand and O trans
to the N atom. The Re‚‚‚O distance is very long (3.01 Å),
suggesting a very weak interaction of the O with the metal
center. The Re-O distance is much shorter in T4a, yet T4a
is the least stable T4 isomer by 13.7 kcal/mol. This shows
that the influence of the strength of the Re-O bond on the
overall stability of the structure is minimal. In contrast, T3a
(analogue of 14b, 15b) is the most stable structure for the
T3 complexes, and it is 5.9 kcal/mol more stable than isomer
T3b (analogue of 14a, 15a). The Re-N bond in the T3
complexes is much shorter than the corresponding Re‚‚‚O
5622 Inorganic Chemistry, Vol. 41, No. 21, 2002

Double Silyl Migration in ORe[N(SiMe2CH2PCy2)2]
1
instrument (400 MHz H, 162 MHz 31P, 101 MHz 13C, 59 MHz
be closer in energy and more stable than T3c, in accord with
the experimental observations.
6Li), or a Varian Unity Inova instrument (500 MHz 1H, 126 MHz
13C). GC-MS spectra were recorded on a Hewlett-Packard 5890/
5971 series GC-MS.
(d) Correlation with the Experimental Results. The
computed structures T4b and T4c correlate closely with the
experimental results of the X-ray diffraction studies of 18a
and 17b, respectively (see Table 1). Slight differences in the
geometrical parameters can be ascribed to steric factors,
which are absent in the theoretical model. The somewhat
greater differences in the Re-O bond distances between
theory and experiment occur because any very weak bonding
(e.g., Re-O) is more difficult to model accurately with DFT.
On the basis of the structural and computational studies,
it is easy to understand why Et3P reacts much faster with
the T4c-type structures (17b, 18b) than with the T4b-type
structures (17a, 18a). The oxygen atom is coordinated weakly
in both cases, and presumably, the barrier to its dissociation
is negligible. However, in the T4c-type structure, oxygen
dissociation (giving a T4d structure) can lead to a sterically
available empty site, whereas in the T4b-type structure, even
with the oxygen dissociated, the backbone of the POP ligand
will block access to the empty site. Moreover, X-ray data
(vide supra) support a stronger Re-O interaction in 18a than
in 17b. Et3P can easily bind to the empty site of a T4d
structure, facilitating the loss of the POP ligand. To create
an available empty site, T4b would have to dissociate a
phosphine arm first, clearly a higher activation barrier process
than dissociation of the weakly bound disilyl ether moiety.
From the thermodynamic point of view, apparently having
three phosphine donors in 22 or 23 is preferred to having
two phosphine donors and a weak donor ether in 17 and 18,
despite the unfavorable entropy change.
Computational Details. The calculations were carried out using
the Gaussian 98 set of programs23 within the framework of DFT at
the B3PW91 level of the theory.24a,b Re was represented using the
Hay-Wadt relativistic electron core potential for the 46 innermost
electrons and its associated basis set.24c Si, P, and Cl were also
represented with the Los Alamos ECPs and their associated basis
set,25 completed by an additional d polarization function.26 The
6-31G** basis set was used for H, C, N, and O.27 Full geometry
optimizations without symmetry constraints were carried out using
the default algorithm. The nature of all minima was checked by
analytical frequencies calculation. Evaluation of the Gibbs energy
was done using the frequency calculation results and does not affect
the relative stabilities of the various species. Results are thus given
in term of nuclear plus electronic energies.
Preparation of 1-3. (PNPCy)Li‚Et2O (1a). Cy2PH (11.0 mL,
54.4 mmol) was dissolved in 100 mL of THF. To this solution
was added n-BuLi in hexanes (34.0 mL of 1.6 M solution, 54.4
mmol), which caused the appearance of the yellow color of Cy2-
PLi. The solution was stirred for 15 min, and then (ClCH2SiMe2)2-
NH (3.95 mL, 18.1 mmol) was added via syringe. The mixture
was stirred for 15 min, during which it became practically colorless.
Another portion of n-BuLi (11.3 mL of 1.6 M solution, 18.1 mmol)
was added, the mixture turned yellow and was stirred for 15 min,
and then (ClCH2SiMe2)2NH (1.32 mL, 6.04 mmol) was added via
syringe. The mixture was stirred for 15 min, and then the third
portion of n-BuLi (3.80 mL of 1.6 M solution, 6.04 mmol) was
added; the mixture turned yellow and was stirred for 15 min, and
then (ClCH2SiMe2)2NH (0.44 mL, 2.0 mmol) was added via
syringe. The mixture was stirred for 15 min, the volatiles were
removed, and the residue was twice triturated with heptane. Then,
it was extracted with pentane and filtered. The filtrate was stripped
to dryness, triturated with heptane, extracted with pentane, and
filtered. The filtrate was stripped to dryness. The residue was then
crystallized from ether at -30 °C, producing three fractions totaling
12.9 g (74%) of pure product. For the preparation of the sodium
salt 2, the combined washings and supernatants were treated with
NaOCMe2Et (0.66 g, 6.0 mmol) in benzene/pentane (20 mL) and
Conclusion
This paper began with the assertion that those bonds that
are conveniently formed in ligand syntheses might be, for
the same reason (e.g., polarity), vulnerable to attack once
they are incorporated into a metal complex.22 Because the
reaction times and temperatures employed in this report are
relatively extreme, the message of this work is not altogether
pessimistic: chemical reactivity that proceeds at or below
25 °C or in a short period of time at elevated temperature is
both attractive and realizable without rearrangement of the
ligand backbone. Nevertheless, the reactions reported here
are interesting, and even surprising at first sight, because of
the extensive rearrangement of the molecules. The idea that
oxide or halide ligands can be nucleophiles for internal
“displacement” warrants attention.
(22) Giesbrecht, G. R.; Shafir, A.; Arnold, J. Chem. Commun. 2000, 2135.
(23) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.; Robb,
M. A.; Cheeseman, J. R.; Zakrzewski, V. G.; Montgomery, J. A., Jr.;
Stratmann, R. E.; Burant, J. C.; Daprich, S.; Millam, J. M.; Daniels,
A. D.; Kudin, K. N.; Strain, M. C.; Farkas, O.; Tomasi, J.; Barone,
V.; Cossi, M.; Cammi, R.; Mennucci, B.; Pomelli, C.; Adamo, C.;
Clifford, S.; Ochterski, J.; Petersson, G. A.; Ayala, P. Y.; Cui, Q.;
Morokuma, K.; Malick, D. K.; Rabuck, A. D.; Raghavachari, K.;
Foresman, J. B.; Cioslowski, J.; Ortiz, J. V.; Baboul, A. G.; Stefanov,
B. B.; Liu, G.; Liashenko, A.; Piskorz, P.; Komaromi, I.; Gomperts,
R.; Martin, R. L.; Fox, D. J.; Keith, T.; Al-Laham, M. A.; Peng, C.
Y.; Nanayakkara, A.; Gonzalez, C.; Challacombe, M.; Gill, P. M. W.;
Johnson, B.; Chen, W.; Wong, M. W.; Andres, J. L.; Gonzalez; Head-
Gordon, M.; Replogle, E. S.; Pople, J. A. Gaussian 98, revision A.7;
Gaussian, Inc.: Pittsburgh, PA, 1998.
(24) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648. (b) Perdew, J. P.;
Wang, Y. Phys. ReV. B 1992, 45, 13244. (c) Hay, P. G.; Wadt, W. R.
J. Chem. Phys. 1985, 82, 299.
(25) Wadt, W. R.; Hay, P. G. J. Chem. Phys.1985, 82, 184.
(26) Ho¨llwarth, A. H.; Bo¨hme M. B.; Dapprich, S.; Ehlers, A. W.; Gobbi,
A.; Jonas, V.; Ko¨hler, K. F.; Stegman, R.; Veldkamp, A.; Frenking,
G. Chem. Phys. Lett. 1993, 208, 237.
Experimental Section
General Considerations. All manipulations were performed
using standard Schlenk techniques or an argon-filled glovebox
unless otherwise noted. Solvents were distilled from Na/benzophe-
none or CaH2, as appropriate. LiN(SiMe2CH2PCy2)2‚Et2O (1a) and
HN(SiMe2CH2PCy2)2 were prepared by modification of established
procedures.1 All other reagents were used as received from
commercial vendors. 1H NMR chemical shifts are reported in ppm
relative to protio impurities in the deuterated solvents. 31P spectra
are referenced to external standards of 85% H3PO4 (at 0 ppm). NMR
spectra were recorded with a Varian Gemini 2000 instrument (300
MHz 1H, 121 MHz 31P, 75 MHz 13C), a Varian Unity Inova
(27) Hariharan, P. C.; Pople, J. A. Theor. Chem. Acta 1973, 28, 213.
(28) Albers, M. O.; Ashworth, T. V.; Oosthuisen, H. E.; Singleton, E. Inorg.
Synth. 1987, 26, 68.
(29) ORTEP-3 for Windows: Farrugia, L. J. J. Appl. Crystallogr. 1993,
30, 565.
Inorganic Chemistry, Vol. 41, No. 21, 2002 5623

Ozerov et al.
allowed to stand for 48 h. A colorless crystalline precipitate of
(PNPCy)Na formed over this time. It was separated by decantation,
washed with benzene, and dried in vacuo to give 0.75 g (5% based
on the initial amount of Cy2PH). (PNPCy)Na can also be obtained
as a crystalline precipitate from equimolar amounts of (PNPCy)Li
PO)(Me2S)ReOCl3 (0.975 g, 1.50 mmol) were placed in a flask
and treated with 50 mL of benzene. The reaction mixture was stirred
for 18 h, and then the volatiles were removed under vacuum. The
residue was triturated with heptane, extracted with pentane, and
filtered. The filtrate was stripped to dryness and treated with 10
mL of cold Et2O, and the flask was stored in a freezer at -30 °C
for 3 h. Then, the supernatant was carefully decanted, and the green
residue was washed with cold Et2O and pentane and dried in vacuo
to give 0.480 g (40%) of the product. The combined washings were
stripped to dryness and treated analogously to give the second crop
of the product, 0.200 g (16%). Alternatively, the product can be
obtained by crystallization from various ether/pentane mixtures at
low temperature. The product is obtained as a mixture of C2V and
Cs isomers; the ratio is poorly reproducible and is dependent on
the handling history.
1
and NaOCMe2Et in benzene in >75% yield. H NMR (C6D6): δ
0.50 (s, 12H, SiMe2), δ 0.78 (d, JPH ) 4 Hz, 4H, CH2), δ 1.12 [t,
4.5H, 3JHH ) 7 Hz, O(CH2CH3)2], δ 1.26 [br m, 20H, P(C6H11)2],
δ 1.64 [br t, 8H, JHH ) 12 Hz, P(C6H11)2], δ 1.77 [br s, 8H,
3
P(C6H11)2], δ 1.90 [br s, 8H, P(C6H11)2], δ 3.30 [q, 3H, JHH ) 7
Hz, O(CH2CH3)2]. 13C{1H} NMR (101 MHz, C6D6, 20 °C): δ 7.3
(s, SiMe2), δ 11.4 (d, JPC ) 24 Hz, CH2), δ 15.4 [s, O(CH2CH3)2],
δ 26.9 [s, P(4-C6H11)2], δ 27.9 [d, JPC ) 9 Hz, P(3/5-C6H11)2], δ
28.0 [d, JPC ) 9 Hz, P(3/5-C6H11)2], δ 30.3 [d, JPC ) 9 Hz, P(2/
6-C6H11)2], δ 30.6 [d, JPC ) 9 Hz, P(2/6-C6H11)2], δ 34.4 [br s,
P(1-C6H11)2], δ 65.8 [s, O(CH2CH3)2]. 31P{1H} NMR (C6D6): δ
-9.3 (very br s).
Method 2 [from (Ph3PO)2ReOCl3 (10)].6 (PNPCy)Li (1a) (0.950
g, 1.50 mmol) and (Ph3PO)2ReOCl3 (1.250 g, 1.50 mmol) were
placed in a flask and treated with 25 mL of benzene. The reaction
mixture was stirred for 5.5 h. During this time, the mixture became
dark green. Elemental sulfur (0.112 g, 3.50 mmol) was added to
the reaction mixture. It was stirred at ambient temperature for 18
h and then for 10 min at 55 °C. The volatiles were then removed
under vacuum, and the residue was triturated with heptane, extracted
with pentane, and filtered. The filtrate was stripped to dryness,
extracted with pentane, and filtered again. From this solution, 0.84
g (70%) of the product was obtained in the same manner as
described for method 1.
(PNPCy)Li‚TMEDA (1b). (PNPCy)Li‚TMEDA (1b) can be
obtained by recrystallizing (PNPCy)Li‚Et2O (1a) from ether or
pentane in the presence of >1 equiv of TMEDA. 1H NMR (C6D6):
δ 2.07 (br s, 12H, NCH3), 1.94 (br s, 4H, NCH2), 1.6-2.0 (several
multiplets, 20H, Cy), 1.10-1.40 (several multiplets, 20H, Cy), 0.78
(br s, 4h, PCH2Si), 0.50 (s, 12H, SiCH3). 13C NMR (C6D6): δ 57.2
(s, TMEDA), 46.6 (s, TMEDA), 34.8 (d, J ) 10 Hz, P-CH), 30.7
(d, J ) 10 Hz, CH2 of Cy), 30.4 (d, J ) 10 Hz, CH2 of Cy), 28.0
(d, J ) 9 Hz, CH2 of Cy), 27.8 (d, J ) 9 Hz, CH2 of Cy), 27.1 (s,
CH2 of Cy). 31P NMR (C6D6): δ -7.9 (br s).
(PNPCy)Na (2). 1H NMR (C6D6): δ 1.0-2.1 (several broad
multiplets, 40H, Cy), 0.80 (br s, 4H, PCH2Si), 0.51 (s, 12H, SiCH3).
31P NMR (C6D6): δ -12.1 (br s).
C
2W Isomer (14a). 1H NMR (C6D6): δ 2.54 (br t, 4H, Cy), 2.43
(br d, J ) 11 Hz, 4H, Cy), 2.02 (br d, J ) 13 Hz, 4H, Cy), 1.71
(br d, J ) 12 Hz, 8H, Cy), 1.05-1.60 (several multiplets, 28H, Cy
and PCH2Si), 0.27 (s, 12H, SiCH3). 13C NMR (C6D6): δ 36.9 (t, J
) 10 Hz, P-CH), 29.3 (br s, CH2 of Cy), 29.2 (br s, CH2 of Cy),
27.7 (br t, J ) 5 Hz, CH2 of Cy), 27.6 (br t, J ) 6 Hz, CH2 of Cy),
26.6 (br s, CH2 of Cy), 10.6 (br t, J ) 6 Hz, PCH2Si), 5.2 (br s,
SiCH3). 31P NMR (C6D6): δ 8.9 (s).
(PNPCy)K (3). 3 was prepared in situ by an exchange reaction
in C6D6 between 1b (13.5 mg, 20 µmol) and potassium menthoxide
1
(3.9 mg, 20 µmol). H NMR (C6D6), selected resonances: δ 0.79
(d, 3 Hz, 4H, PCH2Si), 0.44 (s, 12H, SiCH3). 31P NMR (C6D6): δ
-8.6 (s).
1
(PNPCy)MgBr‚0.75THF‚0.25dioxane (5) [Attempted Prepara-
tion of (PNPCy)2Mg]. (PNPCy)Na (1.22 g, 2.12 mmol) was dissolved
in 20 mL of THF/Et2O and MgBr2‚Et2O (0.280 g, 1.06 mmol) was
added. The mixture was stirred for 2 h, and then 0.5 mL of 1,4-
dioxane was added. The mixture was stirred for 1 h, and then the
volatiles were removed. The residue was extracted with ether and
filtered. The filtrate was stripped, and the white residue was
recrystallized from ether/pentane at -30 °C to give 0.62 g (0.85
mmol, 40%) of (PNPCy)MgBr‚0.75THF‚0.25dioxane. 1H NMR
(C6D6): δ 3.76 (br, 3H, THF), 3.37 (s, 2H, 1,4-dioxane), 1.85 (br
t, 8H, Cy), 1.50-1.75 (m, 16H, Cy), 1.10-1.45 (several multiplets,
19H, Cy and THF), 0.73 (d, J ) 10 Hz, P-CH2-Si), 0.40 (s, 12H,
SiCH3). 31P NMR (C6D6): δ -12.5 (s).
Cs Isomer (14b). H NMR (C6D6): δ 2.98 (br d, J ) 11 Hz,
2H, Cy), 2.83 (m, 2H, Cy), 2.14-2.24 (m, 6H, Cy), 2.06 (br d, J
) 12 Hz, 2H, Cy), 1.69 (br, 4H, Cy), 1.63 (br d, J ) 12 Hz, 4H,
Cy), 0.95-1.57 (several multiplets, 26H, Cy and PCH2Si), 0.80
(dvt, 2JHH ) 14 Hz, JPH ) 7 Hz, 2H, PCH2Si), 0.56 (s, 6H, SiCH3),
0.30 (s, 6H, SiCH3). 13C NMR (C6D6): δ 37.7 (t, J ) 13 Hz,
P-CH), 36.7 (t, J ) 9 Hz, P-CH), 28.62 (br s, CH2 of Cy), 28.58
(br s, CH2 of Cy), 28.5 (br s, CH2 of Cy), 28.4 (br s, CH2 of Cy),
27.8 (br t, J ) 5 Hz, CH2 of Cy), 27.8 (2 overlapping t, J ) 5 Hz,
CH2 of Cy), 27.3 (br t, J ) 5 Hz, CH2 of Cy), 26.6 (br s, CH2 of
Cy), 26.4 (br s, CH2 of Cy), 12.2 (br s, P-CH2-Si), 5.1 (br s,
SiCH3), 4.7 (br s, SiCH3). 31P NMR (C6D6): δ -0.4 (s). Elem anal
(mixture of isomers). Calcd for C30H60Cl2NP2ReSi2‚0.5C4H10O: C,
44.53; H, 7.59; N, 1.62. Found: C, 44.58; H, 7.53; N, 1.62.
(PNPCy)ReOBr2 (15). (PNPCy)ReOBr2 was synthesized analo-
gously to (PNPCy)ReOCl2 (14), and abotained in similar yields,
using (Ph3PO)(Me2S)ReOBr3 (13)7 or (Ph3PO)2ReOBr3 (11)6 as the
Re source. All samples of (PNPCy)ReOBr2 appear as ca. 1:1 mixture
of isomers when dissolved in C6D6 at ambient temperature.
(PNPCy)Ag (6). (PNPCy)Na (0.700 g, 1.22 mmol) was dissolved
in 20 mL of THF, and AgOTf (0.310 g, 1.22 mmol) was added.
The mixture was stirred in the dark for 3 h, and then the volatiles
were removed. The residue was extracted with ether and filtered
to remove NaOTf. The filtrate was stripped, and the white residue
was recrystallized from ether/pentane at -30 °C to give 0.60 g
1
(75%) of (PNPCy)Ag. H NMR (C6D6): δ 1.78-1.88 (br dd, 8H,
C
2W Isomer (15a). 1H NMR (C6D6), selected resonances: δ 0.22
Cy), 1.49-1.71 (several multiplets, 16H, Cy), 0.98-1.34 (several
multiplets, 20H, Cy), 0.92 (m, 4H, PCH2Si), 0.53 (s, 12H, SiCH3).
13C NMR (C6D6): δ 34.1 (q, J ) 5 Hz, P-CH), 30.8 (br, CH2 of
Cy), 30.0 (br, CH2 of Cy), 27.15-27.50 (m, CH2 of Cy), 26.5 (br
s, CH2 of Cy), 13.8 (br s, PCH2Si), 7.5 (s, SiCH3). 31P NMR
(C6D6): δ 19.4 (two doublets, JPAg ) 434 and 500 Hz).
(s, 12H, SiCH3). 13C NMR (C6D6): δ 37.9 (t, J ) 11 Hz, P-CH),
29.6 (br s, CH2 of Cy), 29.4 (br s, CH2 of Cy), 27.8 (br t, CH2 of
Cy), 26.7 (br s, CH2 of Cy), 12.5 (t, J ) 7 Hz, PCH2Si), 5.1 (br s,
SiCH3). 31P NMR (C6D6): δ 6.6 (s).
Cs Isomer (15b). 1H NMR (C6D6), selected resonances: δ 0.57
(s, 6H, SiCH3), 0.26 (s, 6H, SiCH3). 13C NMR (C6D6): δ 39.0 (t,
J ) 13 Hz, P-CH), 37.8 (t, P-CH), 29.4 (br s, CH2 of Cy), 29.3
(br s, CH2 of Cy), 28.9 (br s, CH2 of Cy), 28.8 (br s, CH2 of Cy),
(PNPCy)ReOCl2 (14). Method 1 [from (Ph3PO)(Me2S)-
ReOCl3]7 (12). (PNPCy)Li (1a) (0.930 g, 1.47 mmol) and (Ph3-
5624 Inorganic Chemistry, Vol. 41, No. 21, 2002

Double Silyl Migration in ORe[N(SiMe2CH2PCy2)2]
27.2-28.0 (4 triplets, CH2 of Cy), 26.6 (br s, CH2 of Cy), 26.4 (br
s, CH2 of Cy), 13.2 (br t, PCH2Si), 4.5 (br s, SiCH3). 31P NMR
(C6D6): δ -3.5 (s).
(C6D6): δ 15.1 (s). Elem anal. Calcd for C30H60Br2NP2ReSi2‚
0.25C5H12: C, 40.23; H, 6.80; N, 1.50. Found: C, 40.09; H, 6.90;
N, 1.53.
(POPCy)ReNBr2 (Cs) (18b). We were unable to isolate the pure
Cs isomer, so we report only partial NMR data from observations
of the mixture of isomers in situ. 1H NMR (C6D6), selected
resonances: δ 0.34 (s, 12H, SiCH3), -0.06 (s, 12H, SiCH3). 31P
NMR (C6D6): δ 16.7 (s).
(PNPCy)ReOBrCl (16). (PNPCy)MgBr (5) (19.5 mg, 32.0 µmol)
and (Ph3PO)(Me2S)ReOCl3 (12)7 (21.1 mg, 32.0 µmol) were mixed
in C6D6 and allowed to react for 10 h. The distribution of products
was analyzed by NMR spectroscopy. Alternatively, (PNPCy)-
ReOBrCl can be observed by treating (PNPCy)ReOCl2 with MgBr2‚
Et2O in C6D6. Three isomers of 16 were observed by 31P NMR
(C6D6): δ 6.9 (s), -0.7 (s), -6.1 (s).
(POPCy-Cl)ReNClMe (20). 20a and 20b can be observed in situ
upon prolonged (>2 days) heating of 14 or 17 C6D6 solutions at
90 °C.
(POPCy)ReNCl2 (17). Both isomers can be observed in situ by
heating <30 mg samples of (PNPCy)ReOCl2 (14) in 0.6 mL of C6D6
for 24 h at 90 °C. (POPCy)ReNCl2 (C2V) (17a). Solid (PNPCy)-
ReOCl2 (14) (9.0 mg, 11 µmol) was placed into a J. Young tube.
The tube was placed into a 115 °C oil bath and was kept at this
temperature for 5 days. During this time, the solid changed color
from green to orange. The solid was then dissolved in C6D6, and
the NMR analysis showed ca. 95% conversion to exclusively
(POPCy)ReNCl2 (C2V) (17a). However, after 2 days at ambient
temperature in solution, signals of the Cs isomer appeared as well.
1H NMR (C6D6): δ 2.76 (m, 8H, Cy), 1.93 (br d, J ) 12 Hz, 4H,
Cy), 1.72 (br d, 8H, J ) 11 Hz, Cy), 1.07-1.62 (several multiplets,
28H, Cy and PCH2Si), 0.14 (s, 12H, SiCH3). 13C NMR (C6D6): δ
33.8 (t, J ) 12 Hz, P-CH), 30.2 (d, J ) 15 Hz, CH2 of Cy), 29.2
(br s, CH2 of Cy), 28.2 (br t, J ) 5 Hz, CH2 of Cy), 27.9 (br t, J
) 5 Hz, CH2 of Cy), 26.9 (br s, CH2 of Cy), 10.2 (br t, P-CH2-
Si), 2.7 (br s, SiCH3). 31P NMR (C6D6): δ 18.9 (s).
1
Major Isomer. H NMR (C6D6): 2.81 (br, 2H, Cy), 2.58 (br,
2H, Cy), 2.27 (br, Cy), 1.47-2.1 (several multiplets, 26H, Cy),
0.9-1.3 (several multiplets, 16H, Cy and PCH2Si), 0.81 (t, J ) 14
Hz, 2H, PCH2Si), 0.37 (s, 6H, SiCH3), -0.05 (s, 6H, SiCH3). 13C
NMR (C6D6): δ -14.8 (t, JCP ) 5.5 Hz, Re-CH3). 31P NMR
(C6D6): δ AB system (JPP ) 168 Hz): 24.7 (d), 22.8 (d).
1
Minor Isomer. H NMR (C6D6), selected resonances: δ 0.39
(s, 3H, SiMe), 0.38 (s, 3H, SiMe), 0.07 (s, 3H, SiMe). 13C NMR
(C6D6), selected resonances: δ -15.5 (t, JCP ) 5.5 Hz, Re-CH3).
31P NMR (C6D6): δ AB system (JPP ) 166 Hz) 24.8 (d), 22.7 (d).
(POPCy-Br)ReNBrMe (21). 21 can be observed in situ upon
prolonged (>2 days) heating of 15 or 18 C6D6 solutions at 90 °C.
Only a small fraction of 21 is present in the mixture (ca. 5%), and
further heating does not improve the conversion. Only one isomer
can be reliably observed in the 31P NMR spectrum; 1H NMR
resonances are obstructed by 18a and 18b and cannot be assigned.
31P NMR (C6D6): δ AB system (JPP ) 162 Hz) 22.5 (d), 21.2 (d).
POPCy (24), [(Et3P)3ReNCl2] (22),10 [(Et3P)3ReNBr2] (23).
Upon codissolution of a sample of a mixture of isomers of [(POPCy)-
ReNCl2] (17) with an excess (>3 equiv) of Et3P in C6D6 at ambient
temperature, the signals due to 17b disappear, and POPCy (24) along
with [(Et3P)3ReNCl2] (22) can be observed. Further heating of the
mixture for 10 h at 80 °C results in full conversion of 17a to 22
and 24. The same results are obtained upon heating 14 with an
excess of Et3P for 24 h at 80 °C. The dibromo analogues 23 and
24 can be prepared in the same manner starting either from 15 or
18. The solids derived from resulting solutions were analyzed by
mass spectrometry (EI).
(POPCy)ReNCl2 (Cs) (17b). (PNPCy)ReOCl2 (14) (100 mg, 120
µmol) was dissolved in 0.7 mL of C6D6 and heated for 24 h at 90
°C. During this time, yellow X-ray-quality crystals of (POPCy)-
ReNCl2 (Cs) (17b) precipitated out of the brown-orange solution.
These crystals were separated by filtration, washed with pentane,
and dried. Yield: 45 mg (45%). The supernatant contained mostly
(POPCy)ReNCl2 (C2V) (17a). 1H NMR (C6D6): δ 2.81 (br, 2H, Cy),
2.58 (br, 2H, Cy), 2.27 (br, Cy), 1.47-2.1(several multiplets, 26H,
Cy), 0.9-1.3 (several multiplets, 16H, Cy and PCH2Si), 0.81 (t, J
) 14 Hz, 2H, PCH2Si), 0.37 (s, 6H, SiCH3), -0.05 (s, 6H, SiCH3).
31P NMR (C6D6): δ 18.6 (s).
(POPCy)ReNBr2 (18). Both isomers can be observed in situ by
heating <30-mg samples of (PNPCy)ReOBr2 (15) in 0.6 mL of C6D6
for 24 h at 90 °C.
1
POPCy (24). H NMR (C6D6), selected resonances: δ 0.37 (s,
12H, SiMe), 0.69 (d, 4H, J ) 2.5 Hz, SiCH2P). 31P NMR (C6D6):
δ -13.0 (s). MS (EI): 471 (M+ - Cy), 388 (M+ - 2Cy).
[(Et3P)3ReNCl2] (22). 1H NMR (C6D6): δ 2.28 (m, 6H, PCH2),
2.03 (pseudo quintet, J ) 8 Hz, 6H, PCH2), 1.93 (m, 6H, PCH2),
1.11 (pseudo quintet, 18H, PCH2CH3), 1.02 (dt, 9H, JPH ) 14 Hz,
JHH ) 8 Hz, PCH2CH3). 31P NMR (C6D6): δ -4.8 (d, J ) 15.5
Hz), -12.7 (t, J ) 15.5 Hz). MS (EI): 505, 507, 509 (M+ - Et3P),
387, 389, 391 (M+ - 2Et3P), 118 (Et3P+).
(POPCy)ReNBr2 (C2W) (18a). (PNPCy)ReOBr2 (15) (370 mg, 405
µmol) was dissolved in 3 mL of toluene and was heated for 24 h
at 90 °C. During this time, the color changed from green to orange.
A small amount of insolubles was filtered off, and the reaction
vessel was additionally rinsed with a small amount of benzene and
filtered. The combined filtrates, ca. 5 mL, were layered with 5 mL
of pentane and allowed to stand at ambient temperature for 2 days.
Orange X-ray-quality crystals of (POPCy)ReNBr2 (C2V) (18a)
precipitated. They were washed with pentane and dried to give 0.220
g (55%) of pure product. The combined washings and the
supernatant were cooled to -30 °C for 2 days. The second crop of
crystals precipitated (0.140 g, 35%). The isolated crystals contained
1 equiv of benzene per Re. A solvent-free sample could be prepared
by recrystallization from CH2Cl2/pentane at low temperature and
[(Et3P)3ReNBr2] (23). 1H NMR (C6D6): δ 2.40 (m, 6H, PCH2),
2.09 (pseudo quintet, J ) 8 Hz, 6H, PCH2), 1.98 (m, 6H, PCH2),
1.06-1.14 (overlapping multiplets, 27H, PCH2CH3. 31P NMR
(C6D6): δ -12.6 (d, J ) 15.5 Hz), -17.6 (t, J ) 15.5 Hz). MS
(EI): 593, 595, 597, 599 (M+ - Et3P), 475, 477, 479, 481 (M+
-
2Et3P), 118 (Et3P+).
1
prolonged drying under vacuum. H NMR (C6D6): δ 3.06 (br t,
Acknowledgment. This work was supported by the
Department of Energy.
4H, Cy), 2.79 (br d, 4H, J ) 11 Hz, Cy), 1.91 (br d, J ) 13 Hz,
4H, Cy), 1.72 (br d, 8H, J ) 11 Hz, Cy), 1.40-1.65 (several
multiplets, 20H, Cy), 1.1-1.3 (several multiplets, 8H, Cy and PCH2-
Si), 0.13 (s, 12H, SiCH3). 13C NMR (C6D6): δ 34.6 (t, J ) 12 Hz,
P-CH), 30.3 (br s, CH2 of Cy), 29.4 (br s, CH2 of Cy), 28.1 (br t,
J ) 5 Hz, CH2 of Cy), 28.0 (br t, J ) 5 Hz, CH2 of Cy), 26.9 (br
s, CH2 of Cy), 11.1 (br, P-CH2-Si), 2.9 (br s, SiCH3). 31P NMR
Supporting Information Available: Full crystallographic
details (.cif file) and alcoholysis reactions. This material is available
free of charge via the Internet at http://pubs.acs.org.
IC020424P
Inorganic Chemistry, Vol. 41, No. 21, 2002 5625