Kinetic and thermodynamic studies of the reactivity of (trimethylsilyl) diazomethane with HMo(CO)3(C5R5) (R = H, Me). Estimation of the Mo-N2CH2SiMe3 bond strength and experimental determination of the enthalpy of formation of (trimethylsilyl)diazomethane

Article abstract of DOI:10.1021/om800336p

The rates of reaction of N₂CHSiMe₃ with HMo(CO) ₃Cp (Cp = η⁵-C⁵H₅) in heptane obey the rate law -d[HMo(CO)₃Cp]/dt = k[HMo(CO)₃Cp][N ₂CHSiMe₃] (k = 0.035 ± 0.01 M^-1 s ^-1 at 0 °C; Δ H^?= 11.7 ± 2.0 kcal/mol and ΔS^? = -22.0 ± 3.0 cal/(mol K)). Isotopic scrambling between DMo(CO)₃Cp and N₂CHSiMe ₃ occurs at a rate faster than the overall reaction. Reversible 1,2-addition to form the tightly bound intermediate [Me₃SiCH ₂N_β=N_α^δ+]][ ^δ-M(CO)₃Cp] is proposed as the first step of the reaction. Spectroscopic and computational data support this formulation. The contact ion pairs can undergo heterolytic cleavage to ions or homolytic cleavage to radicals, and the solvent influence on k_obs (THF > toluene > heptane) is interpreted in terms of this model. The enthalpy of this reaction has been measured by solution calorimetry at 272 K in THF: ΔH = -11.6 ± 1.2 kcal/mol. These data, together with computed organic reaction energies allow estimation of the bond strength between the three-electron donors · N₂CHSiMe₃ and · Mo(CO) ₂Cp to be 25 ± 5 kcal/mol stronger than the two-electron Mo-CO bond. Coordination of N₂CHSiMe₃ to the complexes M(PR₃)₂(CO)₃ (M = Mo, W; R = Cy, ⁱPr; Cy = cyclohexyl; iPr = isopropyl) alters the course of reaction with HMo(CO)₃Cp. The stoichiometric reaction of Me₃SiCH= N=NMo(PⁱPr₃)₂(CO)₃ with 2 equiv of HMo(CO)₃Cp produces SiMe₄, Mo(N₂)(P ⁱPr₃)₂(CO)₃, and [Mo(CO) ₃Cp]₂. In the presence of excess N₂CHSiMe ₃ this reaction is catalytic and has been used to experimentally measure the heat of hydrogenation of N₂CHSiMe₃ to N ₂ and SiMe₄ by 2 equiv of HMo(CO)₃Cp. The derived enthalpy of formation of N₂CHSiMe₃ (5.8 ± 3.0 kcal/mol) is in reasonable agreement with high-level theoretical calculations. X-ray crystal structure data are reported for W(CO) ₂(N₂CH₂SiMe₃)Cp: triclinic, space group P1, a = 6.3928(7) A, b = 10.6551(12) A, c = 10.8766(12) A, α = 100.632(2)°, β= 96.254(2)°, V = 721.32 A³, Z = 2.

Full text of DOI:10.1021/om800336p

Organometallics 2008, 27, 4873–4884
4873
Kinetic and Thermodynamic Studies of the Reactivity of
(
Trimethylsilyl)diazomethane with HMo(CO) (C R ) (R ) H, Me).
3
5
5
Estimation of the Mo-N CH SiMe Bond Strength and
2
2
3
Experimental Determination of the Enthalpy of Formation of
Trimethylsilyl)diazomethane
(
†
†
†
,‡
George C. Fortman, Derek Isrow, James E. McDonough, Paul von Ragué Schleyer,*
,
‡
§
,§
,|
Henry F. Schaefer III,* Brian Scott, Gregory J. Kubas,* Tam a´ s K e´ gl,*
,
|
,†
Ferenc Ungv a´ ry,* and Carl D. Hoff*
Department of Chemistry, UniVersity of Miami, Coral Gables, Florida, 33126, Department of Chemistry
and Center for Computational Chemistry, UniVersity of Georgia, Athens, Georgia 30606, Structural
Inorganic Chemistry Group, Chemistry DiVision, MS J514, Los Alamos National Laboratory, Los Alamos,
New Mexico, 87545, and Department of Organic Chemistry, UniVersity of Pannonia, Veszpr e´ m, H-8201,
Hungary
ReceiVed April 15, 2008
5
The rates of reaction of N
d[HMo(CO) Cp]/dt ) k[HMo(CO)
2.0 kcal/mol and ∆S ) -22.0 ( 3.0 cal/(mol K)). Isotopic scrambling between DMo(CO)
CHSiMe occurs at a rate faster than the overall reaction. Reversible 1,2-addition to form the tightly
bound intermediate [Me
2
CHSiMe
3
with HMo(CO)
3
Cp (Cp ) η -C
5 5
H ) in heptane obey the rate law
-
1
-1
q
-
(
3
3
Cp][N
2
CHSiMe
3
] (k ) 0.035 ( 0.01 M
s
at 0 °C; ∆H ) 11.7
Cp and
q
3
N
2
3
δ+ δ-
3
SiCH
2
N
ꢀ
dN
R
][ Mo(CO)
3
Cp] is proposed as the first step of the reaction.
Spectroscopic and computational data support this formulation. The contact ion pairs can undergo
heterolytic cleavage to ions or homolytic cleavage to radicals, and the solvent influence on kobs (THF >
toluene > heptane) is interpreted in terms of this model. The enthalpy of this reaction has been measured by
solution calorimetry at 272 K in THF: ∆H ) -11.6 ( 1.2 kcal/mol. These data, together with computed
organic reaction energies allow estimation of the bond strength between the three-electron donors · N
and · Mo(CO) Cp to be 25 ( 5 kcal/mol stronger than the two-electron Mo-CO bond. Coordination of
CHSiMe to the complexes M(PR
alters the course of reaction with HMo(CO)
2 3
CHSiMe
2
i
i
N
2
3
3
)
2
(CO)
3
(M ) Mo, W; R ) Cy, Pr; Cy ) cyclohexyl; Pr ) isopropyl)
Cp. The stoichiometric reaction of Me SiCHd
(CO) , and
this reaction is catalytic and has been used to
experimentally measure the heat of hydrogenation of N CHSiMe to N and SiMe by 2 equiv of
HMo(CO) Cp. The derived enthalpy of formation of N CHSiMe (5.8 ( 3.0 kcal/mol) is in reasonable
agreement with high-level theoretical calculations. X-ray crystal structure data are reported for
3
3
i
i
NdNMo(P Pr
Mo(CO) Cp]
3
)
2
(CO)
3
with 2 equiv of HMo(CO)
3
Cp produces SiMe
4
, Mo(N
2
)(P Pr
3
)
2
3
[
3
2
. In the presence of excess N
2
CHSiMe
3
2
3
2
4
3
2
3
j
2 2 2 3
(N CH SiMe )Cp: triclinic, space group P1, a ) 6.3928(7) Å, b ) 10.6551(12) Å, c ) 10.8766(12)
3
W(CO)
Å, R ) 100.632(2)°, ꢀ ) 96.254(2)°, V ) 721.32 Å , Z ) 2.
Introduction
bond strength between metal centers and the three-electron donor
•
NNR have not been reported. In fact, there are relatively few
The first transition-metal complex containing a bound diazo
group was prepared according to eq 1 by Bisnette and King
•
reports for the enthalpy of binding of NO to metal complexes.
The enthalpies of the reactions shown in eqs 2 and 3 have been
1
more than 40 years ago.
2
measured and yield estimates of the Cr-NO and Mo-NO bond
+
-
strengths at 70 and 83 kcal/mol, respectively.
NaMo(CO) Cp + [ArN ] [BF ] fMo(CO) (N Ar)Cp +
3
2
4
2
2
•
•
CO + NaBF (1)
4
Cr(CO) Cp * + NO f Cr(CO) (NO)Cp* + (CO) (2)
3
2
At that time the authors noted that the relationship between
Mo(CO)2(NO)Cp and Mo(p-N2C6H4OCH3)(CO)2Cp (Cp ) η -
t
•
t
Mo([ Bu]Ar) + NO f Mo(NO)(N[ Bu]Ar)
3
(3)
5
3
C5H5) “appears to be especially close.” This statement first
highlighted the similarity of the NO and NNR ligands’ binding
properties to metal complexes. Quantitative estimates of the
These data show the high strength of the formal three-electron
•
•
•
bond between a metal and NO. Comparison of bond strengths
•
•
between NO and NNR with metal complexes is of relevance
to the conversion of dinitrogen to amines, since a possible first
step in any such scheme may be formation of LnM-NNR.
†
University of Miami.
University of Georgia.
Los Alamos National Laboratory.
University of Pannonia.
‡
•
•
§
Unfortunately NtNR is unstable with respect to loss of R,
|
and thus direct calorimetric measurements such as those made
•
(
1) King, R. B.; Bisnette, M. B. J. Am. Chem. Soc. 1964, 86, 5694.
with NtO cannot be made. To circumvent this problem, the
1
0.1021/om800336p CCC: $40.75  2008 American Chemical Society
Publication on Web 09/06/2008

4
874 Organometallics, Vol. 27, No. 19, 2008
Fortman et al.
reaction of metal hydrides with diazoalkanes could be used to
evaluate the thermochemistry of the formal insertion of N2 into
the Mo-R bond to form a metal diazo complex, provided the
reaction is sufficiently clean to allow for a thermochemical
study.
3
Herrmann reported that HW(CO)3Cp reacts with diaz-
omethane to produce the diazo complex shown in eq 4 as the
major reaction product.
HW(CO) Cp + N CH f W(CO) (N CH )Cp + CO (4)
3
2
2
2
2
3
One other obstacle in the task of evaluating the thermochem-
istry of organometallic diazo complexes is the uncertainty in
experimental data on the enthalpy of formation of diazomethane
A minor reaction pathway, however, produced a metal alkyl as
shown in eq 5.
5
itself. Recent computational results suggest an enthalpy of
HW(CO) Cp + N CH f W(CO) (CH )Cp + N (5)
3
2
2
3
3
2
formation of ∼65 kcal/mol for CH2N2(g). This is considerably
6
If conditions could be found such that the enthalpy of eq 4 could
be measured, it would provide, in principle, an inroad to
estimating the strength of the W-(NNCH3) bond in
W(CO)2(N2CH3)Cp. Furthermore, the difference between eqs
different than existing experimental data of ∼50 kcal/mol.
For practical organometallic chemistry N2CHSiMe3 provides
a useful starting material and an alternative to the more
hazardous diazomethane. On the basis of the process pioneered
7
by Seyferth and co-workers, heptane solutions of N2CHSiMe3
4
and 5 results in the reaction shown in eq 6.
are commercially available and stable for long periods of time.
Reactions analogous to that shown in eq 4 were originally
W(CO) (CH )Cp + N f W(CO) (N CH )Cp + CO (6)
3
3
2
2
2
3
8
reported by Lappert for the trimethylsilyl derivative. A stable
This reaction corresponds to the formal insertion of N2 into the
metal-alkyl bond to form a diazenido complex, which is also
accompanied by the elimination of CO. The energetics of
reactions such as that shown in eq 6 are of clear relevance to
catalytic and photochemical attempts to incorporate N2 into the
production of organic amines or other organonitrogen compounds.
N2CHSiMe3 adduct of Co(I) was recently reported by Caulton
9
2
and co-workers, while the structure and reactions of Cp*2Ti(η -
N2CHSiMe3) (Cp* ) C5Me5) were investigated by Bergman
1
0
and co-workers. The catalytic conversion of (trimethylsilyl)-
diazomethane to the corresponding ketene by dicobalt octac-
1
1
arbonyl was recently reported, as well as a mechanistic study
4
•
Analysis of the CO insertion reaction was reported earlier
utilizing the thermochemical cycle shown in eq 7: The enthalpy
of the reaction of N CHSiMe with Cr(CO) C R (R ) H,
2
3
3 5 5
1
2
CH3).
This paper reports thermodynamic, kinetic, and computational
data for the reaction of HMo(CO)3Cp (HMp) with N2CHSiMe3
to yield Mo(CO)2(N2CH2SiMe)3Cp and CO. Also reported is a
catalytic reaction in which N2CHSiMe3 is coordinated to
i
M(CO)3(PR3)2 (M ) Mo, W; R ) Cy, Pr) and subsequently
reacts with 2 equiv of HMo(CO)3Cp. The bound diazo ligand
is hydrogenated to form molecular nitrogen and TMS (TMS )
Si(CH3)4), while the molybdenum hydride compound is con-
verted to the dimer [Mo(CO)3Cp]2 (Mp2). In addition to probes
of the mechanisms of these reactions, reaction energetics relative
to the insertion of dinitrogen into the metal-alkyl bond are
investigated from both theoretical and experimental vantages.
Finally, crystallographic structure data are reported for the
complex W(CO)2(N2CH2SiMe3)Cp.
of carbonyl insertion was dominated by the organic radical
addition (step 7-ii), which is exothermic by ∼12 kcal/mol and
corresponds closely to the observed net enthalpy of reaction.
The M-C bond strengths for the alkyl and acyl derivates (steps
Results
7
-i and 7-iii) roughly cancel. A corresponding scheme for
analysis of N2 insertion is shown in eq 8: In spite of the
isoelectronic nature of CO and N2, both the energetics and the
structures of the products are different for these two cycles.
Kinetics of the Reaction of N2CHSiMe3 with HM(CO)3-
C5R5) (M ) Mo,W; R ) H, CH3). Reaction of HMp with
excess N2CHSiMe3 results in complete consumption of metal
(
•
Formation of the acyl radical C(dO)CH3 is exothermic, and
its stability can be experimentally determined. Diazenido radicals
such as NNCH3 are unstable and are best analyzed computa-
tionally. Another salient difference between eqs 7 and 8 is that
the enthalpies of steps 7-i and 7-iii were found to largely cancel
while the same cannot be expected for steps 8-i and 8-iii.
(5) Dixon, D. A.; de Jong, W. A.; Peterson, K. A.; McMahon, T. B. J.
Phys. Chem. A. 2005, 109, 4073.
•
(
6) (a) http://webbook.nist.gov. (b) Foster, M. S.; Beauchamp, J. L.
J. Am. Chem. Soc. 1972, 94, 2425. (c) Hassler, J. C.; Setser, D. W. J. Am.
Chem. Soc. 1965, 87, 3793.
(
7) Seyferth, D.; Dow, A. W.; Menzel, H.; Flood, T. C. J. Am. Chem.
Soc. 1968, 90, 1080.
8) Lappert, M. F.; Poland, J. S. Chem. Commun. 1969, 1061.
(9) Ingleson, M. J.; Pink, M.; Caulton, K. G. J. Am. Chem. Soc. 2006,
128, 4248.
(
(2) (a) Capps, K. B.; Bauer, A.; Sukcharoenphon, K.; Hoff, C. D. Inorg.
Chem. 1999, 38, 6206. (b) Johnson, A. R.; Baraldo, L. M.; Cherry, J. P. F.;
Tsai, Y. C.; Cummins, C. C.; Kryatov, S. V.; Rybak Akimova, V.; Capps,
K. B.; Hoff, C. D.; Haar, C. M.; Nolan, S. P. J. Am. Chem. Soc. 2001, 123,
(10) Polse, J. L.; Kaplan, A. W.; Andersen, R. A.; Bergman, R. G. J. Am.
Chem. Soc. 1998, 120, 6316.
(11) (a) Ungv a´ ri, N.; Kegl, T.; Ungv a´ ry, F. J. Mol. Catal. A 2004, 219,
7. (b) Tuba, R.; Ungvary, F. J. Mol. Catal. A 2003, 203, 59.
(12) Fortman, G. C.; Kegl, T.; Li, Q.-S.; Zhang, X.; Schaefer, H. F.,
III; Xie, Y.; King, R. B.; Telser, J.; Hoff, C. D. J. Am. Chem. Soc. 2007,
129, 14388.
7
271.
3) Herrmann, W. A. Angew. Chem. 1975, 87, 358; Angew. Chem., Int.
Ed. Engl. 1975, 14, 355.
4) Nolan, S. P.; Lopez de la Vega, R.; Mukerjee, S. K.; Hoff, C. D.
Inorg. Chem. 1986, 25, 1160.
(
(

ReactiVity of N
2
CHSiMe
3
with HMo(CO)
3
(C
5
R
5
)
Organometallics, Vol. 27, No. 19, 2008 4875
Table 1. Relative Product Distribution Based on Solvent for the
Reaction of HMo(CO)3Cp with N2CHSiMe3
Table 2. Rate Constants for Reaction of Metal Hydrides and
a
(Trimethylsilyl)diazomethane
k (M- s-1
1
)
product distribn (%)
HMLn
T (°C)
solvent
a
b
solvent
heptane
toluene
Mo-N2CH2SiMe3 Mp-CH2SiMe3 [Mo(CO)3Cp]2
HMo(CO)3Cp
HMo(CO)3Cp
HMo(CO)3Cp
HMo(CO)3Cp
DMo(CO)3Cp
HMo(CO)3Cp
HMo(CO)3Cp*
HW(CO)3Cp
0
0
0
13
13
28
28
28
heptane
toluene
THF
heptane
heptane
heptane
heptane
heptane
0.035 ( 0.002
0.56 ( 0.02
80
95
10
5
10
0
2.0 ( 0.3
0.085 ( 0.005
0.082 ( 0.005
0.29 ( 0.01
0.011 ( 0.001
0.0039 ( 0.0002
THF
100
0
0
0
0
100
c
Mo(PR3)2(CO)3/tol
a
b
c
i
i
Mo ) Mo(CO)2Cp. Mp ) Mo(CO)3Cp. R ) Pr, Cy; Pr )
CH(CH3)2; Cy ) C6H11.
a
Reactions were performed under pseudo-first-order conditions with
N2CHSiMe3 in excess and obey the rate law -d[HMLn]/dt
)
k[HMLn][N2CHSiMe3]. First-order dependence on the metal- hydride
complex was found for all data. First-order dependence in N2CHSiMe3
was confirmed for HMo(CO)3Cp in heptane only; other metal hydride
data are for comparative purposes.
M) yielded the same derived rate constants within ∼10%
experimental error. The rates of reaction of HMp and DMp in
heptane were also found to be the same within experimental
error. As discussed later, this is attributed to H/D exchange
between DMp and the diazoalkane. The rates of reaction of
5
HMo(CO)3Cp* (Cp* ) η -C5(CH3)5) and HW(CO)3Cp were
found to be slower than the reaction of HMo(CO)3Cp. Derived
rate constants for eq 9 are summarized in Table 2.
HMp + N CHSiMe f Mo(CO) (N CHSiMe )Cp + CO
2
3
2
2
3
(
9)
Figure 1. First-order plots of ln(A - A
of HMo(CO) Cp (0.01 M) with N CHSiMe
toluene, and THF at 0 °C.
∞
) versus time for the reaction
3
2
3
(0.056 M) in heptane,
An Eyring plot of the data in heptane solution is shown in
the Supporting Information (Figure SF-1; ∆H ) 11.7 ( 2 kcal/
q
q
mol and ∆S ) -22.0 ( 3 cal/(mol K)).
hydride and is readily followed by quantitative FTIR spectro-
Reaction of DMo(CO)3Cp with N2CHSiMe3. Attempts were
made to probe the kinetic isotope effect by investigating
reactions of HMo(CO)3Cp and DMo(CO)3Cp (85 ( 10%
isotopic enrichment) under comparable conditions. As shown
in Figure SF-2 in the Supporting Information, these data yielded
virtually identical rates and appeared to show no kinetic isotope
effect within experimental error. The negligible value of the
kinetic isotope effect k(H)/k(D) ) 1.04 ( 0.05 prompted NMR
investigations of the reaction to see if H/D exchange occurred
as shown in eq 10.
13
scopic studies. The product distribution of the reaction of HMp
with N2CHSiMe3 was found to mainly depend on the solvent
in which the reaction was carried out. This relationship between
product distribution and solvent is shown in Table 1.
The reaction of HMo(CO)3Cp with N2CHSiMe3 was found
to be first order in metal complex in all three solvents studied.
1
4
Under pseudo-first-order conditions with excess N2CHSiMe3
the rates were also seen to have a solvent dependence as shown
in Figure 1. The plots of ln[A - A∞] versus time were linear
through at least 2 half-lives.
D-Mo(CO) Cp + N)N)CHSiMe h H-Mo(CO) Cp +
The rate of reaction in toluene was roughly 16 times faster
while the reaction in THF was observed to be approximately
3
3
3
N)N)CDSiMe (10)
3
5
7 times faster than the measured rate in heptane. As shown in
In the limiting case that H/D exchange was slow compared to
the rate of eq 9, the use of D-Mp with 85% labeling would
produce the Mo(N2CHDSiMe3) complex in the ratio of 85:15
Table 1, the reaction also produces more diazo complex as
product on going from heptane to toluene to THF. No significant
perturbation of the reaction rate was observed when the reaction
was performed under an atmosphere of CO (as opposed to an
atmosphere of Ar). The expected first-order dependence on
N2CHSiMe3 was confirmed in heptane solution, where pseudo-
first-order reactions at different [N2CHSiMe3] (0.056 and 0.385
(HD:HH). That is, the product distribution would be independent
of the presence of excess N2CHSiMe3 and, conversely, solely
dependent upon the extent to which the H-Mp was isotopically
labeled as D-Mp. However, if the H/D exchange is extremely
rapid compared to the rate of eq 9, then full randomization of
H/D would be expected. This would result in a statistical mixture
of products as shown in eq 11.
(
13) Products were identified qualitatively and quantitatively by FTIR
bands as assigned in Table ST-1 in the Supporting Information and verified
by comparison to authentic samples of recrystallized Mp2 and Me3SiCH2Mp
(
independently prepared by the reaction of NaMo(CO)3Cp and
D-Mo(CO) Cp + N CHSiMe (excess) f N CDSiMe +
3
2
3
2
3
ClCH2SiMe3). Peak positions assigned to the diazo compounds are in
agreement with those originally reported by Lappert and Poland.
8
Mo(CO) CH SiMe + Mo(CO) CHDSiMe +
2
2
3
2
3
(
14) The pseudo-first-order conditions were typically in the range of a
Mo(CO) CD SiMe (11)
5
.7-10-fold excess of N2CHSiMe3, and the average concentration during
2
2
3
the run was utilized. For a 5.75-fold excess, during 2 half lives the final
Control experiments showed that, once formed, the diazo
complex did not undergo exchange with DMo(CO)3Cp. Experi-
ments were performed in heptane at 286 K by the addition of
varying ratios of N2CHSiMe3 to DMo(CO)3Cp, allowing the
reaction to go to completion and then evacuating the reaction
N2CHSiMe3 will drop to 5.0. The average of 5.38 will thus vary by ∼0.38/
5
.38 ≈ 7% during the course of the run. Somewhat surprisingly, this did
not show up as a systematic deviation in fitting data to a straight line in
ln[A - A∞] versus time plots in comparison to reactions in which a 10-fold
excess of ligand was used. The lower amounts were used to slow down the
reaction and also conserve the amount of diazoalkane used.

4
876 Organometallics, Vol. 27, No. 19, 2008
Fortman et al.
1
5
solution to dryness. Samples were then redissolved in C6D6
1
and analyzed by H NMR and FTIR spectroscopy. As shown
in Figure SF-3 in the Supporting Information, the (H,H):(H,D):
(
D,D) product ratios were found to depend upon the relative
concentrations of the reagents, implying that exchange as shown
in eq 10 was in fact competitive with the overall reaction itself.
More detailed analyses of these data (including both kinetic and
1
6
equilibrium isotope effects) was not attempted beyond the
mechanistic observation of H/D exchange being more rapid than
the overall reaction.
Spectroscopic Detection of an Intermediate Complex in
Heptane Solution. Typical spectroscopic data for the overall
reaction in heptane solution are shown in Figure SF-4 in the
Supporting Information. These data show a smooth decay in
the bands assigned to HMo(CO)3Cp and concomitant buildup
of bands of the major product Mo(CO)2(N2CH2SiMe3)Cp, as
well as the minor products Mo(CO)3(CH2SiMe3)Cp and
2 2 2 3
Figure 2. ORTEP drawing of W(CO) (N CH SiMe )Cp. Selected
bond distances (Å) and angles (deg): W-N1 ) 1.862(4), N1-N2
) 1.207(6), N2-C8 ) 1.471(7), C8-Si1 ) 1.898(6); W1-N1-N2
) 174.3(4), N1-N2-C8 ) 117.4(5), N2-C8-Si ) 110.6(4).
[
Mo(CO)3Cp]2. Expansion of the spectroscopic data during
reaction in heptane at low temperature reveals bands at 2017
and 1930 cm- assigned to an intermediate complex (see Figure
SF-5 in the Supporting Information). Following a rise time
during the first spectrum, the bands near 2017 and 1930 cm-
reach a maximum and then decay during the course of the
1
by slow recrystallization from heptane, and the structure is
shown in Figure 2 (full structural data are available as
Supporting Information). The W-N1-N2 angle of 174.3° and
1
2
the N1-N2-C8 angle of 117.4° are consistent with sp and sp
-
1
reaction. A plot of the absorbance at 2017 cm
for the
hybridizations on N1 and N2, respectively, and are consistent
with the formulation WdNdNCH2SiMe3.
-
1
intermediate versus 2030 cm
for HMo(CO)3Cp showed
reasonable linear behavior over the course of the reaction, as
shown in Figure SF-6 in the Supporting Information. At higher
concentrations of N2CHSiMe3 the bands assigned to the
intermediate were observed to increase. These data are in
keeping with formation of a preequilibrium complex between
N2CHSiMe3 and HMp which goes on to products as shown in
eq 12.
Reaction of 2 HMo(CO)3Cp with N2CHSiMe3 in the
i
Presence of M(PR3)2(CO)3 (M ) Mo, W; R ) Cy, Pr):
Catalytic Hydrogenation to SiMe4 and [Mo(CO)3Cp]2. Co-
ordination of N2CHSiMe3 to M(PR3)2(CO)3 (M ) Mo, W; R
i
18
)
Cy, Pr) was investigated. N2CHSiMe3 was found to be a
better donor ligand than dinitrogen in terms of the equilibrium
shown in eq 13.
k1
(N )Mo(PR ) (CO) + N)NCHSiMe h
2
3 2
3
3
(
13)
HMp + N CHSiMe {k2} ″I″ f Products
\
(12)
2
3
(Me SiHC)N)N)Mo(PR ) (CO) + N
3 3 2 3 2
The relatively strong binding of N2CHSiMe3 to this complex
prompted an investigation of how this would alter the nature
of reactivity of N2CHSiMe3 with HMo(CO)3Cp. Surprisingly,
the reaction occurred rapidly according to eq 14
As discussed later, the intermediate, “I”, is formulated as
δ+ δ-
[
Me3SiCH2-Nꢀ)NR ][ Mo(CO)3Cp].
Thermodynamics of Reaction of N2CHSiMe3 with
HMo(CO)3Cp. The enthalpy of eq 9 was measured in THF by
solution calorimetry at -1 °C. The resultant value of -9.5 (
(
Me SiHCN )Mo(PR ) (CO) +
3 2 3 2 3
1
.1 kcal/mol based on solid HMo(CO)3Cp yields a value with
2
HMp f (N )Mo(PR ) (CO) + Mp + SiMe (14)
2 3 2 3 2 4
all species in solution of -11.6 ( 1.2 kcal/mol. Estimated data
in toluene solution were found to be in agreement with this
value.
In stoichiometric reactions, the dinitrogen complex is observed
as the final product of reaction 14. However, since this is readily
displaced by the stronger ligand N2CHSiMe3, as shown in eq
Structure of W(CO)2(N2CH2SiMe3)Cp. The structure of
W(CO)2(N2CH3)Cp has been reported by Hillhouse and Her-
1
3, reactions in which excess N2CHSiMe3 and HMp are
1
7
rmann. Determination of the structure of the analogous
trimethylsilyl-substituted derivative was of interest to see if any
significant differences occurred which might be indicative of
bond strength differences. Within experimental error there are
no significant differences. Crystals of this complex were grown
combined in the presence of trace amounts of M(PR3)2(CO)3
yield the catalytic hydrogen transfer reaction shown in eq 15.
Such catalysis was also observed for other coordinatively
(15) (a) The complex Mo(CO)2(N2CH2SiMe3)Cp shows a singlet at 2.92
ppm, and the complex Mo(CO)2(N2C(H)(D)SiMe3)Cp shows a broad
unresolved triplet at 2.90 ppm (in C6D6). However, analysis of these data
required evacuation, which leads to some side products. Since N2CHSiMe3
was added in hexane and the reaction done in heptane, NMR analysis
required evacuation to dryness and redissolution in C6D6. During concentra-
tion some decomposition was observed to occur, particularly for
Mo(CO)3(CH2SiMe3)Cp. The alkyl product can react further with
unsaturated complexes of formula M(PR3)2(CO)3 (M ) Mo,W;
R ) Cy, Pr).
i
31
HMo(CO) Cp. (b) Jones, W. D.; Bergman, R. G. J. Am. Chem. Soc 1979,
3
1
01, 5447.
16) Bullock, R. M. In Transition Metal Hydrides; Dedieu, A., Ed.; VCH:
New York, 1991; p 263.
17) Hillhouse, G. L.; Haymore, B. L.; Herrmann, W. A. Inorg. Chem.
979, 18, 2423.
(
(18) The enthalpies of this and related reactions have been measured
and will be reported separately: Fortman, G. C.; Isrow, D.; McDonough,
J. E.; Weir, J. J.; Kubas, G. J.; Scott, B.; Hoff, C. D. Manuscript in
preparation.
(
1

ReactiVity of N
2
CHSiMe
3
with HMo(CO)
3
(C
5
R
5
)
Organometallics, Vol. 27, No. 19, 2008 4877
Scheme 1.
Computed Energies (B3LYP/6-311+G**)
W(CO)2(N2CH2SiMe3)Cp are very similar. Conversely, the
charge distribution shows a notable difference between them,
as molybdenum carries a negative charge and tungsten carries
a partial positive charge in the analogous complexes. Accord-
ingly,NBOanalysisrevealssomedifferenceinthemetal-nitrogen
bond polarization: the bond is 56% polarized toward nitrogen
in Mo(CO)2(N2CH2SiMe3)Cp and 59% polarized toward nitro-
gen in W(CO)2(N2CH2SiMe3)Cp. The largest contributions in
the metal-nitrogen NBOs involve the valence dxz natural atomic
orbitals (NAOs) of the transition metals and the 2pz NAOs of
nitrogen. The Wiberg bond indices are 1.295 for
Mo(CO)2(N2CH2SiMe3)Cp and 1.325 for W(CO)2(N2CH2-
SiMe3)Cp.
Relevant to the Thermochemical Cycle for Hydrogenation of
a
2 3
N CHSiMe
a
All values in kcal/mol.
The M-N2CHSiMe3 bond dissociation enthalpy according
to eq 18 was calculated to be 74.3 kcal/mol (M ) Mo) and
8
3.3 kcal/mol (M ) W).
M(CO) (N CH SiMe )Cp f M(CO) Cp + N CH SiMe
Enthalpy of Hydrogenation of N2CHSiMe3. The reaction
shown in eq 15 provided a means to determine the enthalpy of
formation of N2CHSiMe3. This is possible due to the fact that
the enthalpy of eq 16 has previously been reported.
•
•
2
2
2
3
2
2
2
3
(
18)
This compares to that computed for loss of CO as shown in eq
9.
Mp + H f HMp ∆H ) + 6.3 ( 1.0 kcal/mol (16)
2
2
1
Pure HMo(CO)3Cp was used as the limiting reagent in the
solution calorimetric studies. The selective nature of the catalysis
•
•
M(CO) Cp f M(CO) Cp + CO
(19)
3
2
i
by M(P Pr3)2(CO)3 (M ) Mo, W) was verified, in that
The enthalpy of eq 19 is computed to be 52.0 kcal/mol for M
Mo and 59.8 kcal/mol for M ) W. The computed absolute
quantitative conversion to only Mp2 and SiMe4 was observed.
Production of these products was confirmed by NMR spectros-
copy. Values for the enthalpy of reaction were independent of
whether the Mo or W analogue was used as a catalyst. Adding
the measured enthalpy of the reaction shown in eq 15) -80.6
)
bond strength data seem somewhat higher than expected, on
the basis of the OC-M(CO)5 bond strengths, which are 40 and
2
1
4
6 kcal/mol for Mo and W, respectively. Of direct relevance
to this work is the difference between eqs 19 and 18:
(
2.0 kcal/mol, to the reaction in eq 16 yields the enthalpy of
the reaction in eq 17 to be equal to -74.3 ( 3.0 kcal/mol.
•
•
M(CO) Cp + N CH SiMe f
3
2
2
3
N CHSiMe (soln) + H (gas) f SiMe (soln) + N (gas)
2
3
2
4
2
M(CO) (N CH SiMe )Cp + CO(20)
2
2
2
3
(17)
The computed enthalpies of eq 20 are -22.3 kcal/mol for M
The enthalpies of mixing of both N2CHSiMe3 and SiMe4 in
toluene are expected to be less than 1 kcal/mol and largely
cancel. The enthalpy of formation of N2CHSiMe3 is calculated
as 5.8 ( 3.0 kcal/mol, on the basis of the standard enthalpy of
) Mo and -23.5 kcal/mol for M ) W and reflect the computed
bond strength differences between CO and the diazo group. The
computed difference for Mo of -22.3 kcal/mol is in good
agreement with an independent analysis based on experimental
data of -25 ( 5 kcal/mol, as discussed later. Thus, while the
individual Mo-CO and Mo-N2CHSiMe3 bond strengths may
be somewhat higher than experimental data, the difference
between the two ligands is relatively close.
1
9
formation of SiMe4 of -68.5 kcal/mol.
Computed Enthalpy of Hydrogenation of N2CHSiMe3. The
experimental value for ∆H of eq 17 in the solution phase (-74.3
(
3.0 kcal/mol) can be compared to the computational result
of -67 kcal/mol derived from B3LYP/6-311+G** in the gas
Computed Structure and Homolytic Bond Cleavage of
HM(CO)3Cp Complexes. Computation of potential reaction
pathways of HM(CO)3Cp and N2CHSiMe3 required initial
calculation of the structure and energy of the hydrides them-
selves. Surprisingly, two minima were found for HM(CO)3Cp
potential energy hypersurfaces for Mo and W and are shown in
Figure SF-7 of the Supporting Information. The lowest energy
structures (MoH-e and WH-e) have the hydrido ligand in an
equatorial position. This structure corresponds closely to the
phase, as shown in Scheme 1 above. Step i in this scheme is
simple dissociation of H2, and step ii is addition of H to
NdNdC(H)SiMe3 to yield NdNR, where R ) CH2SiMe3. Step
iii is a key step, since it shows that dissociation of CH2SiMe3
from NdNCH2SiMe3 is considerably more exothermic than that
computed for loss of H from NdNH, for which values of -3.8
and -8 kcal/mol have been reported.
Computed Structure and Bond Strengths of M(CO)2-
•
•
•
•
•
•
2
0
22
(
N2CH2SiMe3)Cp (M ) Mo, W). Computed structures for
M(CO)2(N2CH2SiMe3)Cp (M ) Mo, W) at the BP86 level are
shown in Figure 3.
neutron diffraction structure reported for H-Mo(CO)3Cp* and
is taken as the ground state. A second (local) minimum was
found in which the hydrido ligand was located in an axial
position directly opposite and below the C5H5 ring system
(MoH-a and WH-a). MoH-a and WH-a possessed a free
energy ∼7.8 (Mo) and 9.7 (W) kcal/mol above the ground-
state configuration. We are not aware of computations of this
second structural pattern. The transition state connecting the
two minima has a low barrier, and fluxional behavior of this
relatively ”soft” potential energy surface would be expected.
The theoretical bond lengths for the W complex are in good
agreement with the actual structure shown in Figure 2. The
computed geometries of Mo(CO)2(N2CH2SiMe3)Cp and
(
19) (a) Steele, W. V. J. Chem. Thermodyn. 1983, 15, 595. (b) Davalos,
J. Z.; Baer, T. J. Phys. Chem. A 2006, 110, 8572. (c) http://webbook.nist.gov.
•
(
20) A computed value of 3.8 kcal/mol for addition of H to N2 (a) Gu,
J.; Xie, Y.; Schaefer, H. F. J. Chem. Phys. 1998, 108, 8029. A computed
•
value of 8 kcal/mol for addition of H to N2: (b) Matus, M. H.; Arduengo,
A. J.; Dixon, D. A. J. Phys. Chem. 2006, 110, 10116.
(21) Smith, G. P. Polyhedron 1988, 7, 1605.

4
878 Organometallics, Vol. 27, No. 19, 2008
Fortman et al.
2 2 2 3 2 2 2 3
Figure 3. Computed structure of Mo(CO) (N CH SiMe )Cp (left) and W(CO) (N CH SiMe )Cp (right). Bond lengths are given in angstrom;
natural charges are given in italics.
Figure 4. Structures of possible transition states describing the proton transfer and of the intermediate containing a loose Mo-N bond
values given in Å).
(
-
+
The charge distributions show minor differences in the cases
of the two metals, and molybdenum is found to be more negative
than tungsten, as in the earlier complexes. The results of the
harmonic vibrational frequency analyses are in reasonable
agreement with the available experimental data, seen in Table
ST-2 of the Supporting Information. The strengths of the
metal-hydrogen bonds in the four minima were also examined,
and the results are summarized in Table ST-3 of the Supporting
Information. The differences in Wiberg bond indices are
consistent with the enthalpies and free energies associated with
the bond dissociation.
Computed Intermediates and Transition States in the
Reaction of HMo(CO)Cp with N2CHSiMe3. The seminal
isomer involved in the initial proton transfer process from the
hydrido complex to (trimethylsilyl)diazomethane, which is
discussed later, is the MoH-e isomer shown in Figure SF-7.
The MoH-e isomer is calculated to be the global minimum.
Computed structures on the reaction pathway are shown in
Figure 4. Two slightly different orientations of the diazoalkane
dissociation to the ion pair [Mo(CO)3Cp] [N2CH2SiMe3] . The
lowest energy pathway going from MoH-e to the proposed
intermediate Mo-INT through Mo-TS1 was calculated to
possess a barrier of 17.5 kcal/mol. Proton transfer through the
transition state Mo-TS2 possessed a barrier of 19.9 kcal/mol.
The structure of Mo-TS3 in Figure 4 corresponds to the
transition state for dissociation of this ion pair. Computed
solvated energies using the PCM method are summarized in
Table 3. The computed energies of dissociation to ion pairs of
2
5.8, 25.2, and 23.6 kcal/mol in heptane, toluene, and THF are
in qualitative agreement with experimental observations. For
q
heptane solution at 298 K a value of ∆G ) 18.3 ( 5.0 kcal/
q
mol (from experimental data ∆H ) 11.7 ( 2.0 kcal/mol and
S ) -22.0 ( 3.0 cal/(mol K)) was calculated. While the
q
∆
experimental value is 7.5 kcal/mol lower than the computed
value, the agreement is within the errors expected for calcula-
tions (and experiment) on such difficult systems. The ratio of
rates based on computed solvent activation parameters gives
closer agreement: the computed relative rates at 298 K are ca.
1:3:40, compared to experimental values at 273 K of ca. 1:16:
57 for heptane, toluene, and THF, respectively. The computa-
(
Mo-TS1 and Mo-TS2), which are shown in Figure 4, were
observed in the reaction leading to Mo-INT. The structure of
Mo-INT is proposed to be a polar coordinate bond capable of

ReactiVity of N
2
CHSiMe
3
with HMo(CO)
3
(C
5
R
5
)
Organometallics, Vol. 27, No. 19, 2008 4879
Table 3. Computed Gibbs Free Energies and Gibbs Free Energies
of Activation (kcal/mol) for Reaction of HMo(CO)3Cp with
observed. The computed structure of this intermediate is shown
in Figure 4 as Mo-INT; rapid H/D exchange could presumably
occur from either of the two transition states shown. Separation
of the ion pair and recombination at NR would result in
formation of the diazo complex, which is the major reaction
product. This corresponds roughly to the reaction shown in eq
a
N2CHSiMe3 as a Function of Solvent
b
q c
q d
‡ e
phase
∆G1°
∆G
∆G2
∆G° + ∆G2
gas phase
heptane
toluene
THF
15.7
16.4
16.5
17.1
17.5
17.9
19.9
18.2
10.9
26.6
25.8
25.2
23.6
9.4
8.7
6.5
1
, but in which the arene diazonium salt is replaced with the
alkyl diazonium cation formed by protonation of (trimethylsi-
a
b
Solvated free energies computed using the PCM method. Refers to
lyl)diazomethane. Following migration from Nꢀ to NR, loss of
the ∆G° value for the reaction of HMo(CO)3Cp and N2CHSiMe3 to give
•
c
q
CO yields the final product in which the N2CH2SiMe3 is now
Mo-INT of Figure 4. Refers to the ∆G value for the reaction
a formal three-electron-donor ligand.
HMo(CO)3Cp and N2CHSiMe3 to give Mo-INT through Mo-TS1.
Refers to the ∆G value for Mo-TS3 through Mo-INT of Figure4.
Total computed free energy to the transition state for dissociation to
d
e
‡
The increased tendency to dissociate to ion pairs as more
coordinating solvents are used is in keeping with the fact that
on going from heptane to toluene to THF the reaction becomes
both faster and shows fewer side products.
ion pairs.
tions provide a model for the transition state, and the computed
data are in reasonable accord with experiment.
The products of diazoalkane protonations have long been
postulated to depend on the rate of escape from the solvent cage
of initial ion pairs formed. In studies of the reactions in alcohol
of diazoalkanes with benzoic acid, the product ratio has been
interpreted in terms of the relative rates of escape of the ions
from the initial ion pair, with ester product arising from a tight
Discussion
One goal of this work was to determine the thermochemical
barriers to N2 insertion, as shown in eq 8, comparable to those
already reported for CO insertion, as shown in eq 7. In addition,
we sought to compare the bonding capabilities of NO and N2R
as ligands. For N2 insertion to be favorable, a strong bond
between the metal and the N2R fragment is required. As
discussed earlier, unlike OCR or NO, the N2R fragment itself
is unstable. A blend of theory and experiment is required to
analyze the bonding. In order to start investigations in this area,
we initially sought conditions to achieve a clean reaction of
HMo(CO)3Cp with N2CHSiMe3 to yield a single product.
Proposed Mechanism for Reaction of HMo(CO)3Cp and
N2CHSiMe3. The lack of an inhibiting influence of CO on the
rate of reaction discounted steps involving ligand displacement
of carbon monoxide by a diazo group. However, it does suggest
either H atom transfer, as shown in eq 21, or proton transfer,
as shown in eq 22, as the initial step of the reaction.
2
4
ion pair and ether product from escape. This ratio was found
to not be temperature dependent but to depend critically on
solvent polarity. The role of cage effects in photochemical
reactions of [W(CO)3Cp]2 and PR3 ligands has been recently
2
5
studied.
Retention of the initial ion pair (favored in more nonpolar
solvents), however, could result in electron transfer to yield
radical products, as shown in eq 23.
-
+
•
[
M(CO) Cp] [N CH SiMe ] f M(CO) Cp +
3 2 2 3 3
•
N CH SiMe (23)
2
2
3
This corresponds to the net reaction for H atom transfer and is
kinetically indistinguishable from eq 21 occurring directly. The
fact that the computed activation energy for eq 21 is only
marginally higher than the observed activation energy supports
the close balance between these two processes. Computational
•
•
HM(CO) Cp + N CHSiMe f M(CO) Cp + N CH SiMe
3
2
3
3
2
2
3
(21)
•
results suggest that formation of a free N2CH2SiMe3 radical
would be expected to result in rapid expulsion of N2, as shown
in eq 24.
HM(CO) Cp + N CHSiMe f
3
2
3
-
+
[
M(CO) Cp] [N CH SiMe ] (22)
3 2 2 3
•
•
N CH SiMe f CH SiMe + N
2
(24)
Theoretical calculations discussed earlier showed that the C-H
2
2
3
2
3
•
bond strength formed in N2CH2SiMe3 (+53 kcal/mol) is weaker
•
•
The radicals CH2SiMe3 and Mo(CO)3Cp would then be free
to form the alkyl product if they combine directly. Reaction of
than the bond broken in H-Mo(CO)3Cp (+66 kcal/mol) by ∼13
kcal/mol. The experimentally observed enthalpy of activation
of +11.7 ( 2 kcal/mol is lower than this value but overlaps
within experimental error. Therefore, eq 21 cannot be ruled out.
Key findings were that the H/D exchange was faster than
the overall reaction itself and that both the rate and product
distribution were solvent dependent. A plausible mechanism
consistent with all observations discussed in the Results is shown
in Scheme 2. The first step in Scheme 2 involves formation of
an initial adduct by rapid proton transfer from the metal hydride
•
26
CH2SiMe3 and HMo(CO)3Cp in solution would be expected
•
to rapidly form SiMe4 and Mo(CO)3Cp. Combination of two
•
Mo(CO)3Cp radicals will rapidly form [Mo(CO)3Cp]2. The
solvent dependence on reaction products as well as spectroscopic
detection of an intermediate appears to be in accord with these
observations.
The mechanism in Scheme 2 is proposed to involve a polar
+
addition of H in a reversible first step. This might be expected
2
3
to depend on the pKa of the metal hydride. In comparison to
HMo(CO)3Cp (pKa ) 6.4; H2O), HMo(CO)3Cp* (pKa ) 9.6;
H2O) and HW(CO)3Cp (pKa ) 8.6; H2O) are both weaker
to the carbene carbon and formation of the proposed key
intermediate complex. The intermediate can be viewed either
1
as a tight ion pair or as an η -diazo complex bonded to Nꢀ, as
shown in Scheme 2. The reversible nature of the protonation/
deprotonation step accounts for the H/D scrambling that is
(
24) (a) Baer, T.; Bauer, S. H. J. Am. Chem. Soc. 1970, 92, 4773. (b)
More O’Ferrall, R. A.; Kwok, W. K.; Miller, S. I. J. Am. Chem. Soc. 1964,
6, 24–5553.
8
(
22) Brammer, L.; Zhao, D.; Bullock, R. M.; McMullan, R. K. Inorg.
(25) Cahoon, J. F.; Kling, M. F.; Schmatz, S.; Harris, C. B. J. Am. Chem.
Soc. 2005, 127, 12555.
(26) Franz, J. A.; Linehan, J. C.; Birnbaum, J. C.; Hicks, K. W.; Alnajjar,
M. S. J. Am. Chem. Soc. 1999, 121, 9824.
Chem. 1993, 32, 4819.
(
03
23) Zollinger, H. Diazo Chemistry; VCH: New York, 1995; Vol. 2, p
3

4
880 Organometallics, Vol. 27, No. 19, 2008
Fortman et al.
Scheme 2. Proposed Mechanism for the Reaction of HMo(CO)Cp with N
Dimer Products
2
CHSiMe
3
, Producing Diazo, Alkyl, and Metal-Metal
2
7
acids. Investigations of these two other hydrides showed that
they reacted more slowly. At T ) 28 °C HMo(CO)3Cp* was
found to react approximately 26 times more slowly and
HW(CO)3Cp approximately 75 times more slowly than
HMo(CO)3Cp. It is worth noting that in reactions of
HMo(CO)3Cp* both alkyl and metal-metal dimer side products
are formed, as was the case for HMo(CO)3Cp. However,
reactions of HW(CO)3Cp with N2CHSiMe3 in heptane produced
proposed initial protonation/deprotonation of the diazo group,
as shown in eq 25. The increased basicity of coordinated
exclusively W(CO)3(CH2SiMe3)Cp as
W(CO)3Cp]2 was not produced in any detectable amount.
Reaction of HMo(CO)3Cp with N2CHSiMe3 in the
a side product.
2
8
[
diazoalkanes has been reported by Eisenberg. The strong π
i
basicity of Mo(P Pr3)2(CO)3 has been demonstrated by Kaim
2
9
i
and co-workers, where the methylpyrazine cation was found
to form a stable complex. Thus, it is plausible that the
methyldiazonium cation would remain bound to the complex
Mo(P Pr ) (CO) and that this protonation/deprotonation equi-
librium is, first, rapidly established and, second, more favorable
for the bound diazo relative to the free group.
Presence of Mo(P Pr3)2(CO)3, M ) Mo, W). The reaction of
Mo(CO)3Cp with N2CHSiMe3 was observed to be much more
i
rapid in the presence of M(P Pr3)2(CO)3, in spite of the new
i
i
3
2
3
steric barriers presented by the M(P Pr3)2(CO)3 fragment sur-
rounding the N2CHSiMe3. The observed increased rate of
reaction may be due to the coordination of the N2CHSiMe3 to
the electron-rich metal center which in turn could shift the
(
(
28) Woodcock, C.; Eisenberg, R. Organometallics 1985, 4, 4.
29) (a) Bruns, W.; Hausen, H. D.; Kaim, W.; Schulz, A. J. Organomet.
(
27) Kristjansdottir, S. S.; Norton, J. R. In Transition Metal Hydrides;
Chem. 1993, 444, 121. (b) Bruns, W.; Kaim, W.; Waldhoer, E.; Krejcik,
M. Inorg. Chem. 1995, 34, 663.
Dedieu, A., Ed., VCH: New York, 1991; p 322.

ReactiVity of N
2
CHSiMe
3
with HMo(CO)
3
(C
5
R
5
)
Organometallics, Vol. 27, No. 19, 2008 4881
i
a
Scheme 3. Proposed Steps in the Catalytic Hydrogenation of N
2
3 3 3 3 2
CHSiMe by HMo(CO) Cp in the Presence of Mo(CO) (P Pr )
a
•
A step not shown is combination of 2 Mo(CO)
3
Cp radical to yield [Mo(CO)
3
Cp]
2
3
(Mp ) Mo(CO) Cp).
Scheme 4. Thermochemical Cycle Used To Calculate the
Enthalpy of the Homolytic Bond Formation between
Scheme 5. Thermochemical Cycle Used To Estimate the
Enthalpy of Insertion of Molecular Nitrogen into the Mo-R
•
•
a
a
Mo(CO)
3
Cp and N
2
CHSiMe
3
3 3
Bond of Mo(CO) (CH )Cp
a
a
All values are in kcal/mol.
All values are in kcal/mol.
puted difference between the Mo-CO and Mo-NNCHSiMe3
bonds of 22.3 kcal/mol discussed earlier.
Due to the steric factors, the reaction channel involving
migration from Nꢀ to NR which leads to the diazo complex as
shown in Scheme 2 would be blocked when the diazoalkane is
coordinated. A plausible proposed mechanism for the catalytic
process is summarized in Scheme 3.
This can also be compared to data for eq 2 (∆H ) -33.2 (
•
1.8 kcal/mol), where the NdO radical displaces CO in the
•
corresponding Cr(CO)3Cp* radical. There is a larger gap in
•
energy between NdO and CO for Cr than between
Thermochemistry of the Formation of Mo(CO)2-
N2CH2SiMe3)Cp. Data generated in this work was used to
•
(
N2CH2SiMe3 and CO for Mo. Work is in progress to expand
estimate the enthalpy of eq 26 to be -25 ( 5 kcal/mol.
data in this series to allow more accurate assessment of bond
strength trends in these and related complexes.
•
•
The data generated here can be further used to estimate the
energy required for N2 insertion into the Mo-R bond for
R-Mo(CO)3Cp, as shown in eq 8. In spite of the strong
Mo-diazo bond formed, the insertion is predicted to be uphill
by ∼42 ( 5 kcal/mol, as shown in Scheme 5. This estimate
was made using the H3C-Mo BDE of 47 kcal/mol in
Mo(CO) Cp + N CH SiMe f
3
2
2
3
Mo
(
CO) (N CH SiMe )Cp + CO (26)
2
2
2
3
This value was calculated using the thermochemical cycle shown
in Scheme 4. The combination of the Mo-H bond strength in
3
0
•
HMo(CO)3Cp of 66 kcal/mol and the enthalpy of H addition
to N2CHSiMe3 of -52.8 kcal/mol (computed as shown in
Scheme 1) with the experimental net enthalpy of reaction of
31
Mo(CO)3(CH3)Cp and the computed endothermic addition of
an alkyl radical to N2 of 20 kcal/mol as shown in Scheme 1,
together with the value of -25 ( 5 kcal/mol estimated from
Scheme 4 for the enthalpy of displacement of CO by
-
11.6 kcal/mol allowed for the estimation of the enthalpy of
displacement of CO by the diazenido radical fragment. This
value overlaps within experimental error the independently com-
•
N2CH2SiMe3.
(
31) Nolan, S. P.; Lopez de la Vega, R.; Mukerjee, S. L.; Gonzalez,
(
30) Landrum, J. T.; Hoff, C. D. J. Organomet. Chem. 1985, 282, 215.
A. A.; Zhang, K.; Hoff, C. D. Polyhedron 1988, 7, 1491.

4
882 Organometallics, Vol. 27, No. 19, 2008
Fortman et al.
3
6
This suggests that prior photochemical loss of CO from
through cell that has been described in detail elsewhere. NMR
spectra were obtained on a Bruker AVANCE 400 MHz spectrom-
eter. MS spectra were obtained on a VG MASSLAB TRIO-2
instrument using FAB techniques.
R-Mo(CO)3Cp would result in a nearly thermoneutral reaction,
since the Mo-CO bond is ∼40 kcal/mol. In this regard, nitrile
3
2
insertion has been recently reported to occur in the photo-
chemical/thermal reaction sequence shown in eq 27. This
Kinetic Studies of the Reaction of HM(CO) Cp with
3
N
2
3
CHSiMe . In a typical procedure, a solution of 0.1 g of
HMo(CO) Cp in 35 mL of heptane was prepared in a Schlenk tube
3
under argon to make a 0.0012 M solution. The solution was then
loaded into a syringe in the glovebox. A Hamilton gastight syringe
fitted with a valve was filled with 1.00 mL of a 2.0 M solution of
2 3
N CHSiMe in hexanes. Both syringes were taken from the
glovebox. The hydride solution was transferred under an argon
atmosphere to a 50 mL thermostated glass reactor with a magnetic
stir rod. The system was connected to a Schlenk line and also to
the Perkin-Elmer 2000 FTIR microscope system by means of a
small-diameter Teflon tube. Following temperature equilibration
approximately 1 mL of the hydride solution was transferred under
argon to the FTIR cell and a stock solution spectrum recorded.
observed NtCR insertion in this photochemically generated
nitrile complex provides hope that the goal of analogous
photochemical NtN insertion may not be out of reach. In the
case of N2 insertion, the final product would presumably be an
η -diazo complex rather than the η binding motif observed for
the nitrile product in eq 27.
2 3
The reaction was started by injection of the N CHSiMe solution,
and periodically small samples were removed for analysis by
transfer to the flow-through cell. Similar procedures under a CO
atmosphere were performed with careful ventilation.
NMR Studies of H/D Exchange. In the glovebox a Schlenk
3
tube was charged with 0.2057 g of DMo(CO) Cp, which was then
dissolved in 25.0 mL of heptane (0.035 M) under argon. One
milliliter of this solution was used for taking an FTIR spectrum,
and to three different Schlenk tubes was added 8.0 mL of this
solution. These tubes were removed from the glovebox and
connected to a Schlenk line. All three Schlenk tubes were placed
in an ice bath at 0 °C. After temperature equilibration 1.0 mL of a
1
2
Enthalpy of Formation of N2CH2SiMe3. The catalytic
hydrogenation of N2CHSiMe3 by HMo(CO)3Cp outlined in
Scheme 3 allowed reasonably accurate determination of the
enthalpy of formation of this reactive diazo compound in
solution to be 5.8 ( 3.0 kcal/mol. Key aspects of this
measurement were that the determination relied on the purified
organometallic reagent HMo(CO)3Cp and that the catalysis by
2
2 3
.0 M solution of N CHSiMe in hexanes was added to one of the
i
Mo(P Pr3)2(CO)3 cleanly led to the products SiMe4 and
tubes. To the other two tubes 0.100 and 0.025 mL of the diazo
solution were added, respectively. All three flasks were allowed to
stand for 1 h at 0 °C. At that time the reaction was judged to be
complete by FTIR spectra, and all volatile material was removed
[
Mo(CO)3Cp]2. This line of approach, in which an organic
compound too reactive to be isolated in neat form can be
addressed by its reaction with organometallic reagents of known
thermochemistry which are easier to purify and characterize
thermochemically, may find an increased use in the future.
6 6
from each flask by vacuum. C D (3.0 mL) was added to the solid
residue in each flask, and the material dissolved with no trace of
1
residual solids. Each solution was analyzed by FTIR and H NMR
spectroscopy. In addition to the expected major product
Mo(CO) (N CH D SiMe )Cp, FTIR bands showed both the
Experimental Section
2
2
x
2-x
3
known [Mo(CO)
3
Cp]
2
3 2 3
and Me SiCH Mo(CO) Cp and unidentified
General Procedures. Unless stated otherwise, all operations
were performed in a Vacuum Atmospheres glovebox under an
atmosphere of purified argon or utilizing standard Schlenk tube
techniques under argon. THF, toluene, and heptane were purified
by distillation under argon from sodium benzophenone ketyl into
flame-dried glassware and then degassed. Solutions of (trimethyl-
silyl)diazomethane in hexanes (2.0 M) were purchased from Aldrich
minor additional products. These unidentified products were at-
tributed to reactions that most likely occurred upon concentration
during the evacuation of the sample and were judged to not
significantly influence the NMR analysis of the major products
(
Mo(CO)
Crystal Growth and Structure of W(CO)
Crystals of W(CO) (N CH SiMe )Cp suitable for structure deter-
mination were prepared by dissolving 0.1002 g of HW(CO) Cp in
0 mL of toluene (0.0030 M). To this solution was added 0.200
mL of a 2.0 M N CHSiMe solution in hexanes. After the addition
of N CHSiMe the solution turned from a clear orange to a dark
2 2 x 3
(N CH D2-xSiMe )Cp).
2
2 2 3
N CH SiMe Cp.
3
3
and used without further purification. HMo(CO)
3
Cp, HW-
(M ) Mo, W; R
Pr, Cy) were prepared by standard literary procedures. The
2
2
2
3
3
3
34
(
)
CO)
3
Cp, HMo(CO)
3
Cp*, and M(PR
3
)
2
(CO)
3
3
i
35
1
metal hydrides were further purified by sublimation at ca. 60 °C
2
3
and 10- Torr. The complexes M(CO)
2
N CH SiMe Cp (M ) Mo,
2 2 2 3
2
3
red. The reaction was allowed to continue for a period of
approximately 80 min, and the product was characterized by FTIR
spectroscopy. The major product of the reaction was
W) were characterized by FTIR, NMR, and mass spectroscopy, as
well as X-ray crystallography for M ) W. FTIR and NMR data
were found to be in agreement with those originally reported by
8
W(CO)
of W(CO)
2
(N
3
2
CH
(CH
2
SiMe )Cp. A side reaction also led to the formation
3
SiMe )Cp. Peak assignments for these compounds
3
Lappert and Poland. A sample of Me
3
SiCH
2
3
Mo(CO) Cp was
prepared by reaction of NaMo(CO)
3
Cp and ClCH
2
SiMe and shown
3
2
can be seen in Table ST-1 in the Supporting Information. The
solution was then evaporated to dryness and the residue redissolved
in 4.0 mL of heptane. A portion of the W(CO) (N CH SiMe )Cp
2 2 2 3
to have spectroscopic properties identical with those detected in
solution-phase studies. FTIR data were obtained on a Perkin-Elmer
System 2000 spectrometer fitted with a microscope and flow-
remained undissolved. The solution was removed and filtered into
a glass tube under argon. The remaining undissolved portion was
dissolved in another 4.0 mL of heptane and used to characterize
the product by IR spectroscopy. The glass tube was sealed under
(
32) Suzuki, E.; Komuro, T.; Okazaki, M.; Tobita, H. Organometallics
007, 26, 4379.
33) King, R. B. Organometallic Synthesis; Academic Press: New York,
961; Vol. 1, p 165.
2
(
1
(
(
34) Kubas, G. J.; Kiss, G.; Hoff, C. D. Organometallics 1991, 10, 2870.
35) Wasserman, H. J.; Kubas, G. J.; Ryan, R. R. J. Am. Chem. Soc.
(36) Ju, T. D.; Capps, K. B.; Roper, G. C.; Hoff, C. D. Inorg. Chim.
Acta 1998, 270, 488.
1
986, 108, 2294.

ReactiVity of N
2
CHSiMe
3
with HMo(CO)
3
(C
5
R
5
)
Organometallics, Vol. 27, No. 19, 2008 4883
vacuum and placed in a freezer (∼-20 °C) for a period of 2 weeks.
The tube was opened in the glovebox; the mother liquor was
removed by syringe and replaced with a small amount of degassed
mineral oil. The tube was then sealed again and stored until
mounting for structure determination.
Calorimetric Measurement of the Enthalpy of Reaction of
HMo(CO)
3
Cp with N
2
CH
2
SiMe
3
in THF. In the glovebox a
CHSiMe in
solution of 350 mL of THF and 10 mL of 2.0 M N
2
3
hexanes was prepared in a Schlenk flask. The solution was
transferred to the calorimeter. Also in the glovebox a glass ampule
Crystal Structure Determination. Due to air sensitivity the
crystal was mounted from a pool of mineral oil under an argon gas
flow. The crystal was placed on a Bruker P4 diffractometer with
3
was filled with 0.2204 g of HMo(CO) Cp (0.90 mmol). The ampule
was sealed under vacuum and was then placed inside the calorimeter
and allowed to come to thermal equilibrium at 272 K under stirring.
1
k CCD and cooled to 203 K using a Bruker LT-2 temperature
Upon equilibration the calorimeter was electrically calibrated by
heating a standardized resistor with a known current for typically
device. The instrument was equipped with a sealed, graphite-
monochromated Mo KR X-ray source (λ ) 0.710 73 Å). A
hemisphere of data was collected using ꢁ scans, with 30 s frame
exposures and 0.3° frame widths. Data collection and initial
indexing and cell refinement were handled using SMART
software. Frame integration, including Lorentz-polarization cor-
rections, and final cell parameter calculations were carried out using
6
0 s. The system was then allowed to reestablish thermal equilib-
rium over a period of (on average) 15-20 min. The reaction was
then initiated, and the thermogram was recorded electronically at
3
7
10 s intervals. The calorimeter was once again left to reach thermal
equilibrium. Once equilibrium was reached, the calorimeter was
electrically calibrated. The average of the two electrical calibrations
was used to evaluate the heat given off in the reaction. Six total
runs were averaged together to obtain a final value of -9.5 ( 1.1
kcal/mol. Correction for the endothermic enthalpy of solution of
3
8
SAINT software. The data were corrected for absorption using
3
9
the SADABS program. The decay of reflection intensity was
monitored via analysis of redundant frames. The structure was
solved using direct methods and difference Fourier techniques. All
hydrogen atom positions were idealized and rode on the atom they
were attached to. The final refinement included anisotropic tem-
perature factors on all non-hydrogen atoms. Structure solution,
refinement, graphics, and creation of publication materials were
solid HMo(CO) Cp (+2.1 ( 0.1 kcal/mol in THF) yielded a final
3
value with all species in THF solution of -11.6 ( 1.2 kcal/mol.
After all runs were completed, the final products of the several
runs of calorimetry were verified by FTIR spectroscopy. The main
peaks at the end of the experiment were assigned to excess
4
0
performed using SHELXTL.
Reaction of HM(CO) Cp with N
of M(PR (CO) Complexes (M ) Mo, W; R ) Pr
the glovebox 0.2044 g of solid Mo(PCy (CO) (0.27 mmol) was
dissolved in 10.0 mL of toluene to which 0.2 mL of a 2.0 M solution
of N CHSiMe in hexanes (0.4 mmol) was added. An FTIR
spectrum was run and showed peaks at 2066 cm due to free
CHSiMe
(∼0.13 mmol) as well as bands at 1948 (w), 1844 (s),
-1
N
2
CHSiMe
3 2
(2065 cm ) and the final product of Mo(CO) -
3
2 2 3
CH SiMe in the Presence
i
(
N
2
CH SiMe )Cp. Trace amounts of Mo(CO) (CH SiMe ) were
2
3
2
2
3
3
)
2
3
3
, Cy). In
also seen in the spectrum. Both products’ peak assignments can be
seen in Table ST-1 of the Supporting Information. Also, small peaks
3
)
2
3
-1
at 1862 and 1804 cm were seen and are believed to be due to
2
3
unknown decomposition products.
-
1
Calorimetric Measurement of the Enthalpy of Hydrogena-
N
2
3
-1
tion of N
2
CH
2
SiMe
CHSiMe
mL of toluene and 0.400 mL of 2.0 M N
prepared. To this solution was added 0.2080 g of W(CO)
0.36 mmol). One milliliter of this solution was used for recording
3
by HMo(CO)
. In the glovebox a solution of 10.0
CHSiMe in hexane was
3
Cp in the Presence of
1
833 (sh) cm assigned to Mo(PCy
mmol). In addition, a small band at 1866 cm was present, due
to the known and unreactive complex Mo(PCy (CO) , which is a
common contaminant of the air-sensitive Mo(PCy (CO) . In a
second Schlenk tube a solution of freshly sublimed HMo(CO) Cp
0.0811 g 0.33 mmol) was prepared in 10 mL of toluene and then
added to the solution of Mo(PCy (CO) (N CHSiMe ). An FTIR
spectrum run within minutes of mixing showed new bands attributed
to [Mo(CO) Cp] . No bands due to HMo(CO) Cp, Mo(CO)
CH SiMe )Cp, or free N CHSiMe were seen. The FTIR data
for Mo(PCy (CO) (N CHSiMe ) was largely unchanged. Bands
assigned to the known complexes Mo(PCy (CO) and
Mo(PCy (CO) (N ) were present, in keeping with the calculated
stoichiometry of the reaction: HMp (0.33 mmol)
Mo(PCy (CO) (N CHSiMe ) (0.27 mmol) + N CHSiMe (0.13
mmol) f Mp (0.165 mmol) + Mo(PCy (CO) (N CHSiMe
(CO) (N )/Mo(PCy (CO) (0.035
(0.165 mmol) + SiMe (0.165 mmol). In separate
the production of SiMe was proven
3 2
) (CO)
3
(N
2
CHSiMe
3
) (∼0.27
i
-
1
W(CO)
3
(P Pr
3
)
2
N
2
3
2
3
3
)
2
4
i
3
3 2
(P Pr )
3
)
2
3
(
3
an FTIR spectrum, and 4 mL was loaded under argon into the cell
of a Setaram-C-80 Calvet microcalorimeter. The solid-containing
compartment of the calorimeter was loaded with 0.0206 g of freshly
(
3
)
2
3
2
3
3
sublimed HMo(CO) Cp. The assembled calorimeter cell was taken
3
2
3
2
-
from the glovebox and loaded into the calorimeter. After the system
reached temperature equilibration, the reaction was initiated and
the thermogram indicated a rapid reaction which returned cleanly
to baseline with no thermal signal indicative of any substantial
secondary reactions occurring.
(
N
2
2
3
2
3
3
)
2
3
2
3
3
)
2
3
3
)
2
3
2
2
+
Following a return to the baseline, the cell was taken back into
3
)
2
3
2
3
2
3
the glovebox, and a sample analyzed by FTIR spectroscopy showed
2
3
)
2
3
2
3
)
i
the presence of W(CO)
Mo(CO) Cp] , and W(CO)
whenever handling W(CO)
3
(P Pr
3
)
2
(N
the last of which is present
. The spectral peaks of these
2 3 2 3
CHSiMe ), N CHSiMe ,
(
0.235 mmol) + Mo(PCy
3
)
2
3
2
3
)
2
3
i
[
3
2
4 3 2
(P Pr )
mmol) + N
experiments performed in C
2
4
i
3
(P Pr
3
)
2
D
6 6
4
compounds are given in Table ST-1 of the Supporting Information.
The reported enthalpy of reaction is the average of five measure-
by NMR spectroscopy. Addition of CO to the reaction mixture
described above showed IR bands attributable to free N CHSiMe
generated from Mo(PCy (CO) (N CHSiMe ) as well as conversion
of all other products except Mp and Mo(PCy (CO)
Repetition of this reaction at room temperature with varying
ratios of reagents showed that both Mo(PCy (CO) and
as well as the corresponding W derivatives
produced clean hydrogenation of N CHSiMe from HMp in toluene,
even at 4/1 ratios of free N CHSiMe to bound N CHSiMe . No
detectable amount of Mo(CO) (N CH SiMe )Cp was observed in
the room-temperature reactions.
2
3
ments and yielded an average value based on solid HMo(CO)
of -37.2 ( 0.9 kcal/mol. Correction for the endothermic enthalpy
of solution of solid HMo(CO)
3
Cp
)
3 2
3
2
3
2
3
)
2
4
.
3
Cp (+3.1 ( 0.1 kcal/mol in toluene)
yielded a final value with all species in toluene solution of -40.3
3
)
2
3
i
( 1.0 kcal/mol of reacting HMo(CO) Cp -80.6 ( 2.0 kcal/mol of
3
3 2 3
Mo(P Pr ) (CO)
N
2
CHSiMe
3
reacting. The derived enthalpy of reaction was found
i
2
3
to be the same when Mo(CO)
place of W(CO)
tion.
3
(P Pr
3 2 2 3
) (N CHSiMe ) was used in
2
3
2
3
i
3
(P Pr
3
)
2
(N
2
CHSiMe ) as catalyst for the hydrogena-
3
2
2
2
3
Computational Details. Full geometry optimizations have been
performed at the density functional level of theory without any
symmetry constraints using the Gaussian 03 suite of programs.
(
(
(
37) SMART-NT 4; Bruker AXS, Inc., Madison, WI 53719, 1996.
38) SAINT-NT 5.050; Bruker AXS, Inc., Madison, WI 53719, 1998.
39) Sheldrick, G. SADABS, first release; University of G o¨ ttingen,
4
1
G o¨ ttingen, Germany.
40) SHELXTL Version 6.10; Bruker AXS, Inc., Madison, WI 53719;
001.
(
(41) Frisch, M. J. Gaussian 03, Revision B.05; Gaussian, Inc., Pittsburgh,
PA, 2003.
2

4
884 Organometallics, Vol. 27, No. 19, 2008
Fortman et al.
4
7
For all of the calculations the gradient-corrected exchange functional
(PCM) with the dielectric constant values ꢀ
0
) 1.92, 2.38, 7.58
42
developed by Becke was utilized in combination with a correlation
for heptane, toluene, and tetrahydrofuran, respectively. Thermo-
chemistry corrections were taken from gas-phase frequency calcula-
tions at 298.15 K and 1 atm.
4
3
functional developed by Perdew and denoted as BP86. For
molybdenum and tungsten the valence triple-ꢂ SDD basis set
following the (8s7p6d1f) f [6s5p3d1f] contraction pattern is
44
utilized with the corresponding relativistic effective core potential.
Acknowledgment. We dedicate this article to Professor
R. Bruce King of the University of Georgia and thank
Professor Joel Liebman, the University of Maryland, Balti-
more County, and Dr. Manuel Temprado for helpful dis-
cussions. Support of this work by the National Science
Foundation (Grant Nos. CHE 0615743 and CHE 0716718)
and by the Hungarian Scientific Research Fund (Grant No.
OTKA NK 71906) is gratefully acknowledged.
4
5
For the other atoms the 6-311G(d,p) basis set was used. The
density fitting basis sets were generated automatically from the AO
primitives by the Gaussian 03 program. The stationary points were
characterized by frequency calculations in order to verify that they
have zero imaginary frequencies for equilibrium geometries and
one imaginary frequency for transition states. The NBO analyses
were carried out on the stationary points using the NBO 3.1
4
6
program as implemented in Gaussian. To estimate the effect of
the solvent, single-point calculations on the gas-phase optimized
structures were carried out using the polarized continuum model
Supporting Information Available: Tables and figures giving
experimental and computed vibrational spectroscopic data, com-
puted minimum energies of HM(CO)
kinetic and spectroscopic data and a CIF file giving crystallographic
data for the structure of W(CO) (N CH SiMe )C . This material
3
Cp for M ) Mo, W, and
(
(
(
42) Becke, A. D. Phys. ReV. A 1988, 38, 3098.
43) Perdew, J. P. Phys. ReV. B 1986, 33, 8822.
5 5
H
44) Dolg, M.; Stoll, H.; Preuss, H.; Pitzer, R. M. J. Phys. Chem. 1993,
2
2
2
3
9
7, 5852.
is available free of charge via the Internet at http://pubs.acs.org.
(45) Krishnan, R.; Binkley, J. S.; Seeger, R.; Pople, J. A. J. Chem. Phys.
1
980, 72, 650.
OM800336P
(46) Glendening, E. D.; Reed, A. E.; Carpenter, J. E.; Weinhold, F. NBO
Version 3.1.
(47) Miertus, S.; Scrocco, E.; Tomasi, J. Chem. Phys. 1981, 55, 117.