Solvent effects and activation parameters in the competitive cleavage of C-CN and C-H bonds in 2-methyl-3-butenenitrile using [(dippe)NiH]2

Article abstract of DOI:10.1021/ja8000216

The reaction of [(dippe)NiH]₂ with 2-methyl-3-butenenitrile (2M3BN) in solvents spanning a wide range of polarities shows significant differences in the ratio of C-H and C-CN activated products. C-H cleavage is favored in polar solvents, whereas C-C cleavage is favored in nonpolar solvents. This variation is attributed to the differential solvation of the transition states, which was further supported through the use of sterically bulky solvents and weakly coordinating solvents. Variation of the temperature of reaction of [(dippe)NiH]₂ with 2M3BN in decane and N,N-dimethylformamide (DMF) allowed for the calculation of Eyring activation parameters for the C-CN activation and C-H activation mechanisms. The activation parameters for the C-H activation pathway were ΔH^? = 11.4 ± 5.3 kcal/mol and ΔS^? = -45 ± 15 e.u., compared with ΔH^? = 17.3 ± 2.6 kcal/mol and ΔS ^? = -29 ± 7 e.u. for the C-CN activation pathway. These parameters indicate that C-H activation is favored enthalpically, but not entropically, over C-C activation, implying a more ordered transition state for the former.

Full text of DOI:10.1021/ja8000216

Published on Web 06/10/2008
Solvent Effects and Activation Parameters in the Competitive
Cleavage of C-CN and C-H Bonds in 2-Methyl-3-Butenenitrile
Using [(dippe)NiH]2
Brett D. Swartz,† Nicole M. Reinartz,† William W. Brennessel,† Juventino J. Garc´ıa,‡
and William D. Jones*,†
Department of Chemistry, UniVersity of Rochester, Rochester, New York 14627, and Facultad de
Qu´ımica, UniVersidad Nacional Auto´noma de Me´xico, Me´xico City, Me´xico D. F. 04510
Received January 2, 2008; E-mail: jones@chem.rochester.edu
Abstract: The reaction of [(dippe)NiH]2 with 2-methyl-3-butenenitrile (2M3BN) in solvents spanning a wide
range of polarities shows significant differences in the ratio of C-H and C-CN activated products. C-H
cleavage is favored in polar solvents, whereas C-C cleavage is favored in nonpolar solvents. This variation
is attributed to the differential solvation of the transition states, which was further supported through the
use of sterically bulky solvents and weakly coordinating solvents. Variation of the temperature of reaction
of [(dippe)NiH]2 with 2M3BN in decane and N,N-dimethylformamide (DMF) allowed for the calculation of
Eyring activation parameters for the C-CN activation and C-H activation mechanisms. The activation
parameters for the C-H activation pathway were ∆Hq ) 11.4 ( 5.3 kcal/mol and ∆Sq ) -45 ( 15 e.u.,
compared with ∆Hq ) 17.3 ( 2.6 kcal/mol and ∆Sq ) -29 ( 7 e.u. for the C-CN activation pathway.
These parameters indicate that C-H activation is favored enthalpically, but not entropically, over C-C
activation, implying a more ordered transition state for the former.
Introduction
The cleavage of C-C σ-bonds has many potentially important
applications in synthesis and industry, leading to increased
exploration of this area of organometallic chemistry. Many
studies have shown that C-C cleavage can be accomplished
through the insertion of a transition metal center into certain
types of C-C bonds.1 One notable industrial process employing
a C-C cleavage application is the 3-step adiponitrile (AdN)
process developed by DuPont (Figure 1),2 in which C-CN
cleavage by a homogeneous nickel catalyst plays a critical role.
The AdN process starts with the reaction of HCN, butadiene,
and a Ni(0) phosphite catalyst producing a 2:1 mixture of
3-pentenenitrile (3PN) and 2-methyl-3-butenenitrile (2M3BN).1
The desirable, linear 3PN is simultaneously isomerized to the
terminal olefin 4-pentenenitrile (4PN) under the reaction condi-
tions and a second HCN addition then gives AdN using a Ni(0)
catalyst and a Lewis acid cocatalyst.2–4 The branched isomer
2M3BN is capable of being catalytically isomerized to the
desirable linear isomer 3PN beginning with η2-coordination of
nickel to 2M3BN (1) followed by the formation of a π-allylic
intermediate (2) formed by cleavage of the C-CN bond (Figure
Figure 1. General sequence of the DuPont adiponitrile (AdN) process.
2), permitting rearrangement to 3PN.2–7 There has been much
recent work investigating the mechanistic aspects of the
conversion of 2M3BN to 3PN using Ni(0) catalysts via
8–11
activation and rearrangement of the C-CN bond.
Vogt and others performed studies using a variety of chelating
Xantphos ligands (Figure 3) with rigid backbones that showed
(5) (a) Garcia, J. G.; Brunkan, N. M.; Jones, W. D J. Am. Chem. Soc.
2002, 124, 9547. (b) Garcia, J. J.; Jones, W. D. Organometallics 2000,
19, 5544.
† University of Rochester.
‡ Universidad Nacional Auto´noma de Me´xico.
(1) For a recent review of these topics, see: Topics in Organometallic
Chemistry. ActiVation of UnreactiVe Bonds and Organic Synthesis;
Murai, S., Ed.; Springer-Verlag: Berlin, 1999.
(6) Brunkan, N. M.; Jones, W. D. J. Organomet. Chem. 2003, 683, 77.
(7) Brunkan, N. M.; Brestensky, D. M.; Jones, W. D. J. Am. Chem. Soc.
2004, 126, 3627.
(2) McKinney, R. J. In Homogeneous Catalysis; Parshall, G. W., Ed.;
Wiley: New York, 1992; pp 42-50. Abbreviations used: 2-methyl-
3-butenenitrile (2M3BN); 2-methyl-2-butenenitrile (2M2BN); 2-pen-
tenenitrile (2PN); 3-pentenenitrile (3PN); 4-pentenenitrile (4PN).
(3) Tolman, C. A.; Seidel, W. C.; Druliner, J. D.; Domaille, P. J.
Organometallics 1984, 3, 33.
(8) Van der Vlugt, J. I.; Hewat, A. C.; Neto, S.; Sablong, R.; Mills, A. M.;
Lutz, M.; Spek, A. L.; Mu¨ller, C.; Vogt, D. AdV. Synth. Catal. 2004,
346, 993.
(9) Wilting, J.; Muller, C.; Hewat, A. C.; Ellis, D. D.; Tooke, D. M.;
Spek, A. L.; Vogt, D. Organometallics 2005, 24, 13.
(10) Chaumonnot, A.; Lamy, F.; Sabo-Etienne, S.; Donnadieu, B.; Chaudret,
B.; Barthelat, J. C.; Galland, J. C. Organometallics 2004, 23, 3363.
(4) McKinney, R. J.; Nugent, W. A Organometallics 1989, 8, 2871.
9
8548 J. AM. CHEM. SOC. 2008, 130, 8548–8554
10.1021/ja8000216 CCC: $40.75
2008 American Chemical Society

Competitive Cleavage of C-CN and C-H Bonds
A R T I C L E S
Scheme 1. Proposed Mechanism Including C-C vs. C-H
Activation Intermediates
Figure 2. Conversion of 2M3BN to 3PN via π-allyl intermediate.
Figure 3. Structures of xantphos catalysts and derivatives used for 2M3BN
isomerization.
a good selectivity toward the isomerization of 2M3BN to 3PN.8
The POP-Xantphos ligand, while not showing the highest
conversion of the ligands examined, did show the best selectivity
possibly due to its large natural bite angle which increases the
driving force for reductive elimination. During a spectroscopic
study of the reductive elimination step associated with this
reaction the addition of ZnCl2 to the reaction enabled the
observation and isolation of the C-CN cleavage intermediate
[(xantphos)(η3-allylmethyl)Ni(CN-ZnCl2)].9
Sabo-Etienne et al. presented a mechanism for the isomer-
ization of 2M3BN to 3PN using a Ni(PMe3)2 catalyst that was
supported by DFT calculations.10 According to these calcula-
tions, there are five steps involved in the mechanism beginning
with the coordination of the Ni-catalyst to the CdC bond of
2M3BN. The C-CN bond was shown to be cleaved, forming
a σ-allylic species that was further isomerized into a π-allylic
species. This intermediate then rearranged further into another
σ-allylic species and finally, the C-CN bond was reformed
giving the Ni-3PN complex.
tool for controlling linear/branched ratios in the isomerization
that has not been reported previously.
Results and Discussion
Stoichiometric Reaction of [(dippe)NiH]2 with 2M3BN. A
solution of [(dippe)NiH]2 was treated with 2 equiv of 2M3BN
in THF-d8 at -20 °C, turning the deep red solution to yellow-
brown with the evolution of H2 gas. The reaction was monitored
Reactions of [(dippe)NiH]2 with a variety of aryl, het-
eroaryl, and alkyl cyanides have demonstrated the formation
of an η2-nitrile complex of nickel(0), which undergoes
oxidative addition via either C-CN or C-H cleavage to form
a nickel(II) complex.5–7 More recently, the reaction of
[(dippe)NiH]2 with allyl cyanide showed the quantitative
generation of the η2-coordinated olefin complex, which was
later converted to a mixture of the C-CN cleavage product
[(dippe)Ni(η3-allyl)(CN)] and the C-H cleavage products
cis- and trans-[(dippe)Ni(η2-C,C-crotononitrile)].6,7
primarily through 31P {1H} NMR spectroscopy and initially
2
showed two pairs of doublets at δ 72.0 and 66.1 with JP-P
)
59 Hz and δ 71.2 and 66.3 with 2JP-P ) 61 Hz associated with
the η2-coordination of the Ni(0) complex to either face of the
prochiral olefin 2M3BN (1a, 1b). After four hours at room
temperature, a very broad singlet was observed at δ54 for
complex 2 as the original doublets became smaller. This singlet
converted over 1 day to two new pairs of doublets at δ 70.9
2
and 67.6 with JP-P ) 53 Hz (5) and δ 70.5 and 67.0 with
2JP-P ) 53 Hz (6). Two additional pairs of doublets at δ 70.1
The use of [(dippe)NiH]2 as a catalyst has provided mecha-
nistic insight into a number of C-CN cleavage reactions
including those in aryl-5 and alkyl-12 nitriles. The experiments
reported here describe the stoichiometric and catalytic isomer-
ization of 2M3BN with [(dippe)NiH]2 under changing solvent
and temperature conditions as a means of further investigating
the mechanistic details of this system, and show a remarkable
2
and 62.9 with JP-P ) 58 Hz (8) and δ 69.7 and 62.8 with
2JP-P ) 58 Hz (9) were also observed, giving a total of four
products in all. Based on this data, 1H NMR spectroscopy, X-ray
crystallography, independent synthesis, GC analysis, and by
analogy to previous studies showing the π-allylic intermediate
of the allyl cyanide system,7 the mechanism illustrated in
Scheme 1 is proposed. It is interesting that with [Ni(dippe)]
both isomers 1a and 1b are observed, whereas earlier studies
of binding of 2M3BN to L2Ni0 have shown only one set of
signals in the 31P NMR spectra.
The mechanism shown in Scheme 1 presents two possible
pathways for bond activation resulting in the formation of
π-allylic intermediates based on C-CN activation (2, observed)
or C-H activation (7, not observed). These intermediates
undergo further rearrangement to form either η2-coordinated,
nonconjugated 3PN (3) or η2-coordinated, conjugated 2-methyl-
(11) (a) Acosta-Ram´ırez, A; Mun˜oz-Herna´ndez, M.; Jones, W. D.; Garc´ıa,
J. J. J. Organomet. Chem. 2006, 691, 3895. (b) Acosta-Ram´ırez, A.;
Flores-Gaspar, A.; Mun˜oz-Herna´ndez, M.; Are´valo, A.; Jones, W. D.;
Garc´ıa, J. J. Organometallics 2007, 26, 1712. (c) Ates¸in, T. A.; Li,
T.; Brennessel, W. W.; Lachaize, S.; Garc´ıa, J. J.; Jones, W. D. J. Am.
Chem. Soc. 2007, 129, 7562. (d) Acosta-Ram´ırez, A.; Mun˜oz-
Herna´ndez, M.; Jones, W. D.; Garc´ıa, J. J. Organometallics 2007, 26,
5766.
(12) Garc´ıa, J. J.; Arevalo, A.; Brunkan, N. M.; Jones, W. D. Organome-
tallics 2004, 23, 3997.
9
J. AM. CHEM. SOC. VOL. 130, NO. 26, 2008 8549

A R T I C L E S
Swartz et al.
2-butenenitrile (2M2BN) (8 and 9). 3 undergoes irreversible
C-H cleavage to produce the more stable, conjugated 2-pen-
tenenitrile (2PN) (5 and 6). The final ratio of this reaction
showed 89% linear isomers (5 and 6) and 11% branched isomers
(8 and 9). Neither π-allyl hydride (4 or 7) nor olefin complex
3 were observed directly over the course of the reaction.
Characterization of 2. The first species observed in the C-CN
pathway is the π-allylic intermediate 2. Independent synthesis
of 2 was accomplished through the reaction of bis-cycloocta-
dienyl nickel ((cod)2Ni) with methyl-allyl bromide followed by
the addition of dippe and then NEt4CN. 31P{1H} NMR
spectroscopy indicated the presence of a very broad singlet at
δ54, the same as observed in the reaction of 2M3BN with
[(dippe)NiH]2.
The 1H NMR spectrum is consistent with η3-coordination of
the allylic unit to nickel.6 Three allylic resonances for 2 were
observed as a quartet at δ 5.1, a multiplet at δ 3.9 and a multiplet
at δ 2.2. The latter multiplet overlaps with the isopropyl C-H
from the dippe ligand. As 2 was consumed, all of these
resonances were converted into those of the other isomers.
Characterization of 3-6. During the stoichiometric reaction
of [(dippe)NiH]2 with 2M3BN, neither 3 nor 4 were observed
via NMR spectroscopy. 3 was independently synthesized,
however, by the reaction of [(dippe)NiH]2 with trans-3PN.
31P{1H} NMR spectroscopy of 3 shows a pair of doublets at δ
Figure 4. ORTEP drawing of 6 showing 30% probability ellipsoids.
Selected bond distances (Å): N1-C19 (1.148(3)), C17-C18 (1.446(3)),
Ni1-C17 (1.940(2)), Ni1-C18 (1.962(2)). Selected angles in degrees:
P1-Ni1-P2 (92.21(2)), C17-Ni1-C18 (43.49(9)).
2
67.5 and 66.7 with JP-P ) 65 Hz that were observed to
disappear within ∼1 h at room temperature, forming primarily
2. 5 and 6 were also observed in small quantities, identified by
their characteristic doublets at δ 70.9 and 67.6 with 2JP-P ) 53
Hz and δ 70.5 and 67.0 with 2JP-P ) 53 Hz, respectively. The
1H NMR spectrum of 3 showed two olefinic resonances as a
quartet of doublets at δ 5.6 and a triplet of doublets at δ 4.9.
The CH2 resonance was identified as part of an overlapping
multiplet from δ 1.5-2.1 and the CH3 resonance was observed
to overlap with the isopropyl CH3 groups of dippe at δ 0.8-1.1.
The presence of trans-3PN was also confirmed through GC
analysis after quenching with CHCl3.
After one day at room temperature in THF-d8, 2 was mostly
converted into 5 and 6 (89%) and 8 and 9 (11%). Independent
synthesis of 5 and 6 from the corresponding nitrile and
[(dippe)NiH]2 confirmed these assignments. A yellow crystal
of 6 suitable for X-ray diffraction was obtained. Selected bond
lengths and angles for the complex are given in Figure 4. The
olefinic carbons lay approximately in the P2Ni square plane with
similar Ni-C bond lengths. The olefin bond length is similar
to those in the allyl-cyanide and acrylonitrile systems and the
C-N bond length is similar to that in the allyl-cyanide system.7
The 1H NMR spectrum of 6 showed two olefinic resonances as
a doublet of triplets at δ 5.5 and a doublet at δ 4.5. The CH2
resonance was identified as part of a multiplet at δ 1.5-2.3
and the CH3 resonance was observed as a triplet at δ 0.6. The
presence of the free olefin trans-2PN or cis-2PN was confirmed
by GC analysis following treatment of 5 or 6, respectively, with
CHCl3.
Figure 5. ORTEP drawing of 10 showing 30% probability ellipsoids.
Selected bond distances (A): Ni1-C1 (2.089(7)), Ni1-C2 (2.036 (7)),
Ni1-C3 (2.032(7)), C1-C2 (1.410(9)), C2-C3(1.424(8)). Selected angles
-
in degrees: P1-Ni1-P2 (90.44(6)), C3-Ni1-C1 (72.8(3)). BPh4 is left
out for clarity.
confirmed these assignments and showed no evidence for a
reversible equilibrium once 8 or 9 was formed (i.e., 8 and 9 do
1
not interconvert). The H NMR spectrum of 8 showed one
olefinic resonance as a doublet of doublets at δ 4.9. The two
CH3 groups were observed as a doublet at δ 1.5 and a multiplet
overlapping with CH3 groups from dippe at δ 0.8-1.3. The
presence of the corresponding free olefin Z- or E-2M2BN was
also confirmed by GC analysis following treatment of 8 or 9,
respectively, with CHCl3.
The independent synthesis of 7 started with the formation of
(dippe)Ni(η3-CH2CHC(Me)(CN))+BPh4- (10) from the reaction
of [(dippe)NiH]2, NaBPh4, and 1-bromo-3-cyano-2-butene. A
yellow, needle-like crystal suitable for X-ray diffraction was
obtained and selected bond lengths and angles are given in
Figure 5. The allylic bond lengths are similar to those in the
allyl-cyanide and acrylonitrile systems and the C-N bond length
is similar to the allyl-cyanide system.7 Although a crystal
structure could be obtained for complex 10, it proved to be
unstable in solution, preventing the characterization via NMR
spectroscopy.
Reactions of cis- and trans-2PN with [(dippe)NiH]2 indicated
that there is no reverse equilibrium once 5 or 6 are formed.
Complex 4 was not observed in intermediate NMR spectra but
is implied as a potential intermediate based on the prior reactant
and subsequent products formed during the reaction.
Characterization of 7-9. The 31P{1H} NMR spectrum of 8
and 9 showed two sets of doublets at δ 70.1 and 62.9 with 2JP-P
2
) 58 Hz and δ 69.7 and 62.8 with JP-P ) 58. Independent
synthesis of 8 or 9 from [(dippe)NiH]2 and pure Z- or E-2M2BN
9
8550 J. AM. CHEM. SOC. VOL. 130, NO. 26, 2008

Competitive Cleavage of C-CN and C-H Bonds
A R T I C L E S
a
Table 1. Solvent Study Data for the Catalytic Isomerization of 2M3BN to Other Isomers by [Ni(dippe)H]2
% products
trans-2PN
entry
solvent
dielectric constant (ꢀ)
% conv
Z-2M2BN
E-2M2BN
cis-2PN
trans-3PNb
linear:branchedc
1
2
3
4
5
6
7
8
9
decane
benzene
2
2.3
5
7.5
9.2
10
96.9
95.9
96.9
87.2
67.4
99.6
99.6
98.6
100
2.6
3.8
3.1
5.7
8.3
9
22.3
8.9
24.3
16.4
4.9
9.2
9
14.7
30.5
50.6
70
31
73.9
73.4
4
6.1
9.9
5.5
8.6
6
9.3
1.6
6.8
0.8
0.7
82.5
66.8
78.7
65.5
49d
12.4:1
6.7:1
7.3:1
3.9:1
1.6:1
1:1.5
1:12
1.5:1
1:27.3
1:14.9
6.3
3.7
5.5
4.5
5.7
1.1
4.8
0.6
0
2,2,5,5-tetramethyl THF
THF
trifluoro-toluene
di-t-butyl ketone
acetone
pivalonitrile
CH3CN
DMF
22.5d
5
21
21.1
36.6
38.3
47.1d
2.2
5.4
10
95.8
a Reaction conditions: [2M3BN] ) 1 mM; [Ni] ) 0.105 mM; equiv 2M3BN ) 10; T ) 100 °C; t ) 180 min. b A total of 1-2% of cis-3PN was
seen in all samples. c Linear: branched ratio calculated with linear products cis-2PN, trans-2PN, cis-3PN, trans-3PN, and 4PN vs branched products
Z-2M2BN and E-2M2BN. d About 2% 4-PN was also observed.
This precursor was then treated with LiBEt3H in THF to form
7. Although 7 was not directly observed, the organic products
of this reaction were observed via GC following reaction of
the organometallic products with CHCl3 (which produces
(dippe)NiCl2 and the corresponding free olefin), indicating that
the reaction forms primarily 8 and 9 (80%), but also 1 (20%).
Therefore, the step 1 f 7 must be at least partially reversible,
but the following step (7 f 8 + 9) is not. This reaction further
supports the characterization of 7 and 10 in conjunction with
the crystal structure.
Catalytic Reaction of [(dippe)NiH]2 and 2M3BN in Different
Solvents. Catalytic isomerization of 2M3BN was examined by
injection of 2M3BN into a preheated, septum-capped reaction
tube containing [(dippe)NiH]2 and solvent under a positive
pressure of dinitrogen (see Supporting Information for picture).
This procedure provided reproducible results by ensuring that
oxygen would not contact the reaction solution, which was
independently demonstrated to cause erratic behavior in reaction
times and product distributions.
The reaction of [(dippe)NiH]2 with a 10-fold excess of
2M3BN in THF at 100 °C produced an instantaneous color
change from deep-red to red-orange. After three hours under
an atmosphere of N2, the color was observed to change from
red-orange to yellow-brown with a yellow precipitate. Analysis
of the resultant solutions by GC indicate a reaction preference
toward the linear C-CN activation products 3PN and 2PN
versus the branched C-H activation products E- and Z-2M2BN.
A similar trend was observed in the additional nonpolar solvents
studied. In sharp contrast, a preference toward the branched
C-H products was observed in any other polar solvent studied
(Table 1). Confirmation of GC peaks was determined using
authentic samples of each nitrile.
Decane showed the highest linear to branched ratio (12.4:1),
whereas polar solvents such as acetonitrile gave the lowest linear
to branched ratio (1:27). In decane, the dominant isomer formed
was trans-3-pentenenitrile (82%), although some trans-2-pen-
tenenitrile (6%) and cis-2-pentenenitrile (4%) were formed via
olefin isomerization. In acetonitrile, the branched isomers
E-2M2BN (74%) and Z-2M2BN (24%) constituted the majority
of the products. All of the other solvents produced results within
the range mentioned above. In some instances, small amounts
of 4PN were observed, likely resulting from the isomerization
of 3PN.
dielectric constant. One possible explanation for this trend is
that the distribution of charge in intermediate π-allyl hydride 7
(Scheme 1) is such that it can interact with solvent to a greater
extent than intermediate π-allyl cyanide 2. A more polar solvent
might then be capable of stabilizing the transition state leading
to 8 or 9 more than that to 3, which would result in preferential
C-H activation. The high polarity associated with the nickel-
cyanide bond, however implies that C-C cleavage should be
expected to be favored in polar solvents, and argues against
this explanation. A second consideration is the effect of possible
solvent coordination to the 16-electron nickel(0) complex.
Although some of these solvents have been shown to coordinate
to metal complexes,12,13 this is likely not a significant factor in
this reaction based on the use of trifluorotoluene as a nonco-
ordinating solvent (entry 5), which follows the trend based on
dielectric constant. Further evidence for a general solvation
model is provided by entries 3, 6, and 8 where solvents with
sterically bulky, electron-donating substituents were used.
Tetramethyl THF and di-t-butyl ketone fit the trend based on
dielectric constant fairly well (entries 3 and 6), whereas
pivalonitrile (entry 8) seems to be somewhat different in its
solvation abilities than its neighbors of similar dielectric effect.
A possible explanation is that as the bulk of the solvent
increases, it leads to weakened solvation of 7 relative to 2,
thereby resulting in the more favorable formation of products
derived from C-CN cleavage.
An alternative consideration that would be consistent with
the greater amount of C-H activation seen in polar solvent
would be a deprotonation mechanism for the conversion of 1
into 8 and 9. In an attempt to look for such a pathway, 0.75
equiv piperidine (pKb ) 2.8) or pyridine (pKb ) 8.8) was added
to several of the catalytic reactions. As shown in Table 2, added
base has little effect on the linear to branched ratio in decane
or THF. A slight change in ratio is seen in di-t-butylketone,
producing twice as much branched product as without piperi-
dine. The effect of added base was greatest in pivalonitrile,
which shifted the product ratio by a factor of 6 toward C-H
activation products when piperidine was present. Consequently,
it can be concluded that a deprotonation pathway may constitute
a reasonable pathway leading to branched products in polar
solvents. In nonpolar solvents an ionic pathway is still disfavored
even with base present.
Determination of Molecularity. Previous experiments involving
the isomerization of 2M3BN into 3PN have shown this reaction
The observed selectivity data appear to be strongly governed
by solvent effects. The initial trials included solvents lacking
sterically bulky groups (entries 1, 2, 4, 7, and 9) and appear to
follow a trend of linear to branched product ratio based on
(13) Mindiola, D. J.; Waterman, R.; Jenkins, D. M.; Hillhouse, G. L. Inorg.
Chim. Acta 2003, 345, 299.
9
J. AM. CHEM. SOC. VOL. 130, NO. 26, 2008 8551

A R T I C L E S
Swartz et al.
Table 2. Isomerization of 2M3BN by [Ni(dippe)H]2 in the Presence
Alternatively, if the rotation places the allylic hydrogen endo
to the nickel then C-H activation can occur giving 2M2BN
products (8, 9).
of Basea
linear: branched
ratio with
added base
linear:branched
ratio without
added base
dielectric
constant
solvent/additive
decane/pyridine
decane/piperidine
THF/piperidine
di-t-butyl ketone/piperidine
pivalonitrile/piperidine
2
2
7.5
10
12:1
9:1
4:1
1:3
1:4
12:1
12:1
4:1
1:1.5
1.5:1
21.1
a Reaction conditions: [2M3BN] ) 1 mM; [Ni] ) 0.105 mM; equiv
2M3BN ) 10; T ) 100 °C; t ) 180 min; 0.75 equiv base.
A similar type of mechanism has been proposed for the
isomerization of 2M3BN into 3PN by Chaumonnot et al.10
beginning with η2-coordination of the Ni(0) fragment to 2M3BN
followed by C-CN activation to a σ-allyl species and isomer-
ization to the π-allylic species. A C-H activation pathway was
not calculated in these studies. The pathway was supported by
DFT calculations and these sequences are consistent with the
activation parameters determined both in this study and in the
study with allyl cyanide.7 DFT calculations are underway on
the current system to try to determine the origin of the
differences in these thermodynamic parameters.
to be both zero order and first order with regards to substrate.9,14
A series of reactions were run with 2M3BN (1-9.5 mM) to
determine the molecularity of the system. The method of initial
rates was used as a result of catalyst deactivation by conversion
to (dippe)Ni(CN)2 over the course of the reaction. The resulting
data show a first order dependence on [2M3BN] (see Supporting
Information). Additionally, variation of nickel concentration
(0.0105 mM-0.105 mm) also showed a first order dependence
of the initial rate indicating that the reaction is overall second
order (eq 1).
It is also worth noting that under typical industrial conditions,
neat 2M3BN (a polar solvent) is used during catalysis. Could
use of a nonpolar solvent lead to improvements in selectivity,
thereby avoiding the second branched-to-linear isomerization
in the DuPont Adiponitrile process?15 In an attempt to evaluate
this possibility, the catalyst Ni[P(O-p-tolyl)3]3 was synthesized16
and reacted with 2M3BN in decane under the same reaction
conditions as used in Table 1. Surprisingly, only 36% conversion
was seen and the products were exclusively branched. Conse-
quently, verification of this hypothesis seems unlikely, although
it would be preferable to test under true industrial conditions.
rate ) k[2M3BN][catalyst]
(1)
Temperature Dependence of C-CN and C-H Activation in
2M3BN Reaction. It was previously reported in the π-allylic
cyanide system that C-C cleavage becomes more competitive
kinetically with C-H cleavage with an increase of temperature.7
The temperature dependence of the isomerization of 2M3BN
was also investigated. Two series of reactions were run at
various temperatures using either decane or N,N-dimethylfor-
mamide (DMF) as solvents, these solvents being chosen to
examine regimes in which the dominant reaction pathway is
either C-CN (93%) or C-H (90%) cleavage (Table 3). In
general, the linear to branched product ratio increases by about
a factor of 3 with increasing temperature in both nonpolar decane
and polar DMF.
Rate constants (Table 4) were determined for each reaction
temperature and the activation parameters for C-CN and C-H
bond cleavage were obtained from Eyring plots (Figure 6, Table
5). While the enthalpic difference between C-CN and C-H
activation shows the latter to be more favorable by about 6 kcal
mol-1, the unfavorable entropy of activation for C-H activation
results in comparable ∆Gq values for these reactions under
typical catalytic conditions. These values for ∆Sq suggest a more
ordered transition state in the C-H mechanism, as seen in the
earlier study of C-H vs C-CN activation of allyl cyanide by
[(dippe)NiH]2.7 Perhaps this order is associated with the
deprotonation of 1 by an external base or solvent molecule.
Summary and Conclusions
Reactions of [(dippe)NiH]2 with 2M3BN in different solvents
and at different temperatures gives different linear to branched
product ratios in the isomerization. As the solvent polarity
increases, the ratio of C-H activated branched products to
C-CN activated linear products increases, apparently as a result
of better solvation of the C-H activation transition state vs the
C-CN transition state, or because of the appearance of a
deprotonation pathway, since the ground-state is the same for
both reactions. The activation parameters indicate a smaller ∆Hq
for the C-H oxidative cleavage by ∼6 kcal/mol, while the ∆Sq
value was ∼15 eu more negative, suggesting a more tightly
bound and more ordered transition state in the C-H activation
step.
Experimental Section
The reaction of [(dippe)NiH]2 with 2M3BN begins with the
η2-coordination of the reactive (dippe)Ni(0) fragment to one of
two faces of the olefin in 2M3BN (1a, 1b) as observed by
31P{1H} NMR spectroscopy. The C-C bond then rotates so
CN is able to interact with the nickel center (eq 2, 1a shown).
General Considerations. All manipulations were performed
under a nitrogen atmosphere, either on a high-vacuum line using
modified Schlenk techniques or in a Vacuum Atmospheres Cor-
poration glovebox. Decane, octane, benzene, and tetrahydrofuran
were distilled from dark purple solutions of sodium/benzophenone
ketyl. Acetone and acetonitrile were dried over calcium hydride
and 4A molecular sieves and distilled before use. All other solvents
were dried using 4A molecular sieves and distilled before use.
[(dippe)NiH]2 was synthesized according to the previously reported
procedure.17 trans-3-Pentenenitrile, cis-2-pentenenitrile, and 4-pen-
tenenitrile were purchased from Aldrich Chemical Co. 2-Methyl-
3-butenenitrile, Z-2-methyl-2-butenenitrile, and E-2-methyl-2-
butenenitrile were supplied by DuPont. trans-2-Pentenenitrile was
(14) Valle´e, C.; Vale´rio, C.; Chauvin, Y.; Niccolai, G. P.; Basset, J.-M.;
Santini, C. C.; Galland, J.-C.; Didillon, B. J. Mol. Cat. A: Chem. 2004,
214, 71.
(15) For an example of direct hydrocyanation of butadiene to 3PN, see:
Bini, L; Mueller, C.; Wilting, J.; von Chrzanowski, L.; Spek, A. L.;
Vogt, D J. Am. Chem. Soc. 2007, 129, 12622.
(16) Gosser, L. W.; Tolman, C. A. Inorg. Chem. 1970, 9, 2350.
(17) Vicic, D. A.; Jones, W. D. J. Am. Chem. Soc. 1997, 119, 10855.
9
8552 J. AM. CHEM. SOC. VOL. 130, NO. 26, 2008

Competitive Cleavage of C-CN and C-H Bonds
A R T I C L E S
a
Table 3. Temperature Dependence Data for the Catalytic Isomerization of 2M3BN to Other Isomers by [Ni(dippe)H]2
% products
entry
solvent
T (°C)
t (min)
% conv
Z-2M2BN
cis-2PN
E-2M2BN
trans-3PN
trans-2PN
linear:branchedb
1
2
3
4
5
6
7
8
9
10
decane
decane
decane
decane
decane
DMF
DMF
DMF
DMF
DMF
100
90
80
70
60
100
90
80
70
60
7
16
17
30
60
16
16
30
60
60
94
2.1
2.7
1.7
3
3.9
17.1
17.7
17.5
18.3
16.8
2.4
2.1
1.4
1
0.8
0
0.7
0.4
0.2
0.2
3.9
6.2
4.2
87.7
85.6
90.5
86.9
82.9
5.6
7.4
6.3
2.5
2.5
3.8
3.4
2.2
1.5
1.3
0.7
1
0.6
0.2
0.2
15.7:1
10.4:1
15.9:1
8.4:1
94.1
86.2
71.1
71.3
95.8
91.8
97.2
99
7.6
11.1
76.5
73.1
75.2
78.8
80.3
5.7:1
1:14.9
1:10.1
1:12.5
1:33.5
1:34
99
a Reaction conditions: [2M3BN] ) 1 mM; [Ni] ) 0.105 mM; %V 2M3BN ) 10; equiv 2M3BN ) 10. b Linear:branched ratio calculated with linear
products cis-2PN, trans-2PN, cis-3PN, trans-3PN, and 4PN vs. branched products Z-2M2BN and E-2M2BN.
Table 4. Initial Rates for C-CN and C-H Activation of 2M3BN by
[Ni(dippe)H]2 in Decane and DMF
T (°C)
k
init(C-CN), decane, s-1
k
init(C-H), DMF, s-1
100
90
80
70
60
0.233
0.142
0.0575
0.0296
0.0167
0.202
0.170
0.0699
0.0558
0.0346
synthesized by modification of a previously reported method.18
[(dippe)Ni(η3-CH2CHCHMe)]+BPh4- was prepared by modifica-
tion of a previously reported method,19 substituting methyl-allyl
bromide for allyl-bromide. cis-3-Pentenenitrile was unable to be
synthesized and no commercial samples were available, but its GC
retention time was estimated based upon relative retention time
patterns of the other cis/trans isomers of the products (Z isomers
elute before E, but in the same order). All other chemicals, filter
aids, and chromatographic materials were used as received.
All reactions were run in modified Schlenk flasks (see Supporting
Information) under a positive pressure of nitrogen in an aluminum
heating block at temperatures ranging from 25 to 100 °C ( 0.2°.
The temperature was controlled with an Omega CSC32 benchtop
controller. Samples were collected in 10 µL aliquots and chemically
quenched with chloroform to destroy any remaining catalyst and
stored at -20 °C until analysis. NMR data (1H and 31P{1H}) were
recorded on a Bruker Avance 400 instrument. Chemical shifts are
given in δ and referenced to residual solvent peaks.
Figure 6. Eyring Plot of C-CN(() and C-H(9) activation.
Table 5. Thermodynamic Parameters for C-CN and C-H
a
Activation of 2M3BN by [Ni(dippe)H]2
∆Hq (kcal/mol)
∆Sq (e.u.)
∆Gq at 100 °C (kcal/mol)
C-CN
C-H
17.3 ( 2.6
11.4 ( 5.3
-29 ( 7
-45.3 ( 15
28.2
28.3
a Errors calculated based on 95% confidence limits.
from dark-red to yellow-brown as 3, 5, 6, 8, or 9 formed
respectively. The reaction was monitored while warming at intervals
of 10 °C up to 55 °C via 31P{1H} and 1H NMR spectroscopy.
Aliquots (5 µL) were collected, quenched with CHCl3 (25 µL) and
analyzed via GC. 1H NMR of 3 (400 MHz, THF-d8): δ 5.1 (dq, 1
H, CH3CHCHCH2CN), 4.9 (q, 1 H, CH3CHCHCH2CN), 1.5-2.1
Stoichiometric Isomerization of 2M3BN. [(dippe)NiH]2 (0.015
g, 0.023 mmol) was dissolved in THF-d8 in a septum-capped NMR
tube and cooled to -50 °C. 2M3BN (4.8 µL, 0.047 mmol) was
added via syringe. The reaction was monitored via variable
temperature 31P NMR spectroscopy at 10° increments from -50
to 30 °C. Observations are described in the text.
i
(m, 10 H, pr CH, CH3CHCHCH2CN and PCH2CH2P), 0.8-1.1
i
(m, 27 H, pr CH3 and CH3CHCHCH2CN). 31P{1H} NMR of 3
(400 MHz, THF-d8): δ 67.5 and 66.7 (d). 1H NMR of 6 (400 MHz,
THF-d8): δ 5.5 (dt, 1 H, CH3CH2CHCHCN), 4.5 (d, 1 H,
CH3CH2CHCHCN), 1.5-2.3 (m, 10 H, ipr CH, CH3CH2CHCHCN
and PCH2CH2P), 0.8-1.3 (m, 24 H, ipr CH3), 0.6 (t, 3 H,
CH3CH2CHCHCN. 31P{1H} NMR of 6 (400 MHz, THF-d8): δ 70.5
Preparation of 2. [(dippe)Ni(η3-CH2CHCHMe)]+BPh4- (0.199
g, 0.286 mmol) and NEt4CN (0.055 g, 0.352 mmol) were dissolved
in 30 mL THF and stirred for 3 h producing a dark-red solution
with white precipitate. The solution was filtered through celite and
the solvent was removed under vacuum leaving 0.053 g of an orange
1
and 67.0 (d). H NMR of 8 (400 MHz, THF-d8): δ 5.3 (q, 1 H,
residue (37% yield). H NMR (400 MHz, THF-d8): δ 5.07 (q, 1
CH3CHC(CH3)(CN)), 2.1 (septet, 4 H, ipr CH), 0.8-1.7 (m, 34 H,
ipr CH3, CH3CHC(CH3)(CN), CH3CHC(CH3)(CN) and PCH2CH2P.
31P{1H} NMR of 8 (400 MHz, THF-d8): δ 70.1 and 62.9 (d).
Preparation of 10. [(dippe)NiH]2 (0.02 g, 0.031 mmol) and
NaBPh4 (0.015 g, 0.044 mmol) were dissolved in THF at 25 °C.
1-Bromo-3-cyano-2-butene (0.015 g, 0.094 mmol) was added, and
the solution was stirred for 3 h and allowed to sit overnight forming
a yellow-orange solid. The solution was decanted and the remaining
solid extracted with diethyl ether. The ether solution was dried
leaving 0.02 g of yellow crystals (29% yield). This complex was
too unstable to allow for spectroscopic characterization other than
through low temperature X-ray crystallography.
1
H, allylic CH2CHCHMe), 3.91 (m, 1 H, allylic CH2CHCHMe),
i
2.2 (m, 6H, pr CH and allylic CH2CHCHMe), 1.72 (dd, 4 H,
i
i
PCH2CH2P), 1.26 (dd, 12 H, pr CH3), 1.17 (dd, 12 H, pr CH3).
The allylic methyl group was obscured by the phosphine resonances.
31P{1H} NMR (400 MHz, THF-d8): δ 54 (v br s).
Preparation of 3, 5, 6, 8, and 9. [(dippe)NiH]2 (0.0125 g, 0.0194
mmol) was dissolved in THF and placed in a septum-capped NMR
tube at -30 °C. trans-3PN, trans- or cis-2PN, or E- or Z-2M2BN
(3.7 µL, 0.038 mmol) was added to the reaction changing the color
(18) Lohaus, G. Org. Synth. 1966, 46, 23.
(19) Tenorio, M. J.; Puerta, M. C.; Salcedo, I.; Valerga, J. J. Chem. Soc.,
Dalton Trans. 2001, 653.
Reaction of 10 with LiBEt3H. 10 (0.01 g, 0.139 mmol) was
dissolved in 1 mL THF yielding a yellow solution. An excess of
9
J. AM. CHEM. SOC. VOL. 130, NO. 26, 2008 8553

A R T I C L E S
Swartz et al.
LiBEt3H (0.2 mL, 1.68 mmol) was added and the solution was
stirred for 30 min. A white precipitate was observed to form. The
solution was filtered through alumina, chemically quenched with
chloroform, and analyzed via GC. The results are described in the
text.
Catalytic Isomerization of 2M3BN in Different Solvents. [(dip-
pe)NiH]2 (0.068 g, 0.105 mmol) was dissolved in 1.5 mL decane
(or selected solvent) and 0.300 mL octane (GC internal standard)
and added to a modified Schlenk flask under N2 (see Supporting
Information). The reaction mixture was sealed and placed in an
aluminum heating block at 100 °C for 3 h. Once the reaction
reached 100 °C, 2M3BN (0.205 mL, 2.5 mmol) was added via
syringe. After 3 h, the reaction was chemically quenched with 2
mL CHCl3 to destroy any remaining catalyst and a sample was
extracted and stored at -20 °C until analysis by GC.
Catalytic Isomerization of 2M3BN at Different Temperatures.
[(dippe)NiH]2 (0.068 g, 0.105 mmol) was dissolved in 1.5 mL
decane (or DMF) and 0.3 mL octane in a modified Schlenk flask
under N2 (see Supporting Information). The reaction mixture was
placed in an aluminum heating block, subjected to positive nitrogen
pressure via a needle, and heated to the required temperature
(25-100 °C). 2M3BN (0.205 mL, 2.5 mmol) was added and
samples were collected at specific intervals via syringe. The samples
were quenched with 20 µL CHCl3 and stored at -20 °C until
analysis. This reaction was followed and quantified by GC. See
Supporting Information for additional details regarding the kinetic
determinations. Reactions run with added base (piperidine or
pyridine) were performed under identical conditions; 0.79 mmol
piperidine (7.8 µL) or pyridine (6.4 µL) were added prior to addition
of 2M3BN. The reaction was analyzed as described above.
Catalytic Isomerization of 2M3BN using Ni[P(O-p-tolyl)3]3. This
reaction was run using 0.226 g (0.203 mmol) Ni[P(O-p-tolyl)3]3
and 0.205 mL (2.02 mmol) 2M3BN in 1.5 mL solvent with 0.3
mL octane added as internal standard. The reaction was carried
out and analyzed as described above.
Acknowledgment. We acknowledge the U.S. Department of
Energy, grant #FG02-86ER13569, for their support of this work,
and NSF for a travel grant (INT0102217).
Supporting Information Available: GC traces for reaction
products, pictures of reaction apparatus, and X-ray data collec-
tion parameters for 6 and 10. This material is available free of
charge via the Internet at http://pubs.acs.org.
JA8000216
9
8554 J. AM. CHEM. SOC. VOL. 130, NO. 26, 2008