Decomposition of methylnickel(II) amido, alkoxo, and alkyl complexes by β-hydrogen elimination: A comparative study

Article abstract of DOI:10.1021/om900400s

Thermal decomposition of structurally related amido, alkoxo, and alkyl complexes of type [Ni(Me)(EC(H)RR′)(dippe)] (EC(H)RR′ = cyclo-NC₄H₈ (pyrrolidino), N(CH₂Ph) ₂, OCH₂Ph, OCH(Me)Ph, and CH₂

Full text of DOI:10.1021/om900400s

Organometallics 2009, 28, 6515–6523 6515
DOI: 10.1021/om900400s
Decomposition of Methylnickel(II) Amido, Alkoxo, and Alkyl Complexes
by β-Hydrogen Elimination: A Comparative Study†
ꢀ
Inmaculada Matas, Juan Campora,* Pilar Palma, and Eleuterio Alvarez
ꢀ
´
ꢀ
Instituto de Investigaciones Quımicas, CSIC-Universidad de Sevilla, c/ Americo Vespucio,
49, 41092, Sevilla, Spain
Received May 15, 2009
Thermal decomposition of structurally related amido, alkoxo, and alkyl complexes of type
[Ni(Me)(EC(H)RR0)(dippe)] (EC(H)RR0 = cyclo-NC4H8 (pyrrolidino), N(CH2Ph)2, OCH2Ph,
OCH(Me)Ph, and CH2CH2Ph) takes place through formally analogous processes involving
β-hydrogen elimination and reductive elimination of methane, cleanly affording corresponding
Ni(0) imine, aldehyde, ketone, and olefin complexes, [Ni(η2-EdCRR0)(dippe)]. Kinetic studies on
these decomposition reactions show that, in spite of their similarity, they involve substantial
mechanistic differences.
Introduction
elimination from alkoxo or amido intermediates can also
represent an unproductive side reaction in some other catalytic
reactions such as alcohol or amine arylations.7
Late transition metal alkoxo and amido complexes resemble
the corresponding alkyl derivatives in many of their proper-
ties.1 Thus, the presence of hydrogen atoms at the βpositions of
alkoxo and amido ligands renders them prone to β-hydrogen
elimination, one of the most characteristic decomposition
routes of transition metal alkyls. In many cases, attempts to
prepare late transition alkoxides lead to hydrides or to reduced
species without any detectable intermediates, apparently due to
facile β-hydrogen elimination processes.2 This process plays a
central role in many of the catalytic applications of late
transition metal alkoxides and amides. For instance, this is a
key step in the mechanism of aerobic oxidation of alcohols to
aldehydes2,3 and other environmentally friendly reactions that
may have important industrial applications in the near
future.4,5 It is also relevant because it constitutes the micro-
scopic reversal of the insertion of aldehydes, ketones, or imines
into metal-hydride bonds, which occurs in certain types
of hydrogenation reactions.6 At the same time, β-hydrogen
Although β-hydrogen elimination from transition metal
alkyls has been extensively studied,8 direct observation of
this process in well-defined monomeric alkoxide9 complexes
is uncommon and even less usual for the corresponding
amido derivatives.10 The shortfall of appropriate model
compounds for studying β-elimination in late transition
metal alkoxo and amido complexes stems from the difficulty
of their synthesis and isolation, which is due not only to this
very same process but also to their hydrolytic sensitivity or to
other decomposition routes.1,2,9a In fact, the tendency of
well-characterized transition metal alkoxides to undergo
β-hydrogen elimination has proven to be not as high as earlier
results anticipated. For instance, in his seminal work on the
decomposition of platinum methoxo complexes,9a Bryndza
and co-workers showed that, although the decomposition
rate decreases in the order Pt(OMe)2(dppe) > Pt(OMe)(Et)-
(dppe) > PtEt2(dppe), the mixed alkyl-alkoxide produces
more ethylene than ethane, indicating that the ethyl group is
† Dedicated to the memory of Prof. Tatiana A. Stromnova.
*Corresponding author. E-mail: campora@iiq.csic.es.
~
(7) (a) Hartwig, J. F.; Richards, S.; Baranano, D.; Paul, F. J. Am.
(1) (a) Bryndza, H.; Tam, W. Chem. Rev. 1988, 88, 1163. (b) Fryzuk,
M. D.; Montgomery, C. D. Coord. Chem. Rev. 1989, 95, 1. (c) Fulton, J. R.;
Holland, A. W.; Fox, D. J.; Bergman, R. G. Acc. Chem. Res. 2002, 35, 44.
(2) (a) Kaesz, H. D.; Saillant, R. B. Chem. Rev. 1972, 72, 231. (b)
Schunn, R. A. In Transition Metal Hydrides; Muetterties, E. L., Ed.: Marcel
Dekker: New York, 1971. (c) For a recent example, see: Campbell, A. N;
Chem. Soc. 1996, 118, 3626. (b) Vorogushin, A. V.; Huang, X. H.;
Buchwald, S. L. J. Am. Chem. Soc. 2005, 127, 8146.
(8) (a) Cross, R. J. In The Chemistry of the Metal Carbon Bond;
Hartley, F. R., Patai, S., Eds.; Wiley: New York, 1985; Vol. 2, pp 559-624.
(b) Collman, J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R. G. Principles and
Applications of Organotransition Metal Chemistry, 2nd ed.; University
Science Books: Mill Valley, 1987; pp 383-388.
ꢀ
Gagne, M. R. Organometallics 2007, 26, 2788.
(3) (a) Stahl, S. S. Angew. Chem., Int. Ed. 2004, 43, 3400. (b) Stahl, S. S.
Science 2005, 309, 1824. (c) Popp, V.; Stahl, S. S. Chem.;Eur. J. 2009, 15,
2022. (d) Komiya, N.; Nakae, T.; Sato, H.; Naota, T. Chem. Commun. 2006,
4829. (e) Privalov, T.; Linde, C.; Zetterberg, K.; Moberg, C. Organometallics
2005, 24, 885.
(4) Beller, M. Eur. J. Lipid Sci. Technol. 2008, 110, 789.
(5) (a) Podhajsky, S. M.; Sigman, M. S. Organometallics 2007, 26,
5680. (b) Zhao, J.; Hartwig, J. F. Organometallics 2005, 24, 2441. (c)
Taguchi, K.; Nakagawa, H.; Hirabayashi, T.; Sakaguchi, S.; Ishii, Y. J. Am.
Chem. Soc. 2004, 126, 72.
(9) (a) Bryndza, H. E.; Calabrese, J. C.; Marsi, M.; Roe, C.; Tam, W.;
Bercaw, J. E. J. Am. Chem. Soc. 1986, 108, 4805. (b) Blum, O.; Milstein, D.
J. Am. Chem. Soc. 1995, 117, 4582. (c) Ritter, J. C. M.; Bergman, R. G.
J. Am. Chem. Soc. 1998, 120, 6826. (d) Blum, O.; Milstein, D.
J. Organomet. Chem. 2000, 593-594, 479. (e) Zhao, J.; Hesslink, H.;
Hartwig, J. F. J. Am. Chem. Soc. 2001, 123, 7220. (f) Fafard, C. M.; Ozerov,
O. V. Inorg. Chim. Acta 2007, 360, 286. (g) Smythe, N. A.; Grice, K. A.;
Williams, B. S.; Goldberg, K. I. Organometallics 2009, 28, 277.
ꢀ
(10) (a) Hartwig, J. F. J. Am. Chem. Soc. 1996, 118, 7010. (b) Campora,
ꢀ
J.; Matas, I.; Palma, P.; Alvarez, E.; Graiff, C.; Tiripicchio, A. Organome-
(6) (a) Noyori, R. Angew. Chem., Int. Ed. 2002, 41, 2008. (b) Clepham,
S. E.; Hadzovic, A.; Morris, R. H. Coord. Chem. Rev. 2004, 248, 2201.
tallics 2007, 26, 3840. (c) Pawlikowski, A. V.; Getty, A. D.; Goldberg, K. I.
J. Am. Chem. Soc. 2007, 129, 10382.
r
2009 American Chemical Society
Published on Web 10/28/2009
pubs.acs.org/Organometallics

6516 Organometallics, Vol. 28, No. 22, 2009
Matas et al.
slightly more prone to undergo β-elimination than the meth-
oxide. More recently, Hartwig showed that β-hydrogen elim-
ination is much slower for trans-[Ir(OR)(CO)(PPh3)2]9e than
for analogous alkyls.11 However, so far there are no studies
comparing the rates of β-elimination of structurally analogous
alkyl, alkoxo, and amido compounds. As a consequence, there
remains some uncertainty on the relative tendencies of these
compounds to experience β-elimination.
Scheme 1
Recent work in our research group has focused on the
study of the synthesis and properties of methylnickel alkoxo
and amido complexes of type [Ni(Me)(ECHRR0)(dippe)]
(dippe = 1,2-bis(diisopropylphosphino)ethane, E = NR00,
1; O, 2 ).10b,12 In the course of this work, we discovered that
some amido complexes (ECHRR0 = pyrrolidino, 1a; diben-
zylamido, 1d) are thermally labile and evolve into the
corresponding Ni(0) η2-imine derivatives 3a and 3d with
concurrent production of methane.10b As shown in Scheme1,
compounds 3 arise from β-hydrogen elimination from the
amido ligand, followed by reductive coupling of the resulting
hydride and the methyl group. To gain more insight into the
nature of this process, we decided to continue our studies on
the decomposition of these compounds and compare their
behavior with the related alkoxo and alkyl derivatives. In this
contribution, we show that formally analogous decomposi-
tion processes can ensue for the three types of compounds,
leading to the corresponding Ni(0) η2-aldehyde, ketone, and
olefin derivatives. However, beyond the superficial similarity
of these reactions, kinetic analyses reveal important differ-
ences in their mechanisms.
Scheme 2
31P-containing product. In contrast, primary amido com-
pounds 1b and 1c decompose less cleanly, affording mixtures
that contain Ni(dippe)2, NiMe2(dippe), and the correspond-
ing η2-imine complexes even at 60 °C, identified by their
characteristic 31P{1H} spectra. Conditions allowing their
selective decomposition could not be found, and therefore
their study was not pursued any further.
Results and Discussion
Thermolysis of Amido, Alkoxo, and Alkyl Complexes and
Characterization of the Ni(0) Products. We have recently
developed a methodology well suited for the preparation of
alkoxo and amido complexes of type [Ni(Me)(ECH-
RR0)(dippe)] (dippe = 1,2-bis(diisopropylphosphino)ethane,
E = NR00, 1; O, 2 ).12a Our method is based on the reaction of
thefluoro complex [Ni(Me)(F)(dippe)] withthe corresponding
lithium alkoxides and amides and relies on the low solubility of
lithium fluoride in organic solvents. This reaction is highly
selective and provides ready and very convenient access to
spectroscopically pure and halide-free solutions of these com-
pounds.
Alkoxide derivatives 2 are stable in solution at room
temperature for prolonged periods of time. Methoxide deri-
vative 2a can be heated up to 50 °C in THF for 24 h without
noticeable decomposition, whereas heating for longer peri-
ods of time gradually leads to the formation of a complex
mixture containing NiMe2(dippe) and Ni(dippe)2, which
resemble those obtained in the thermolysis of the related
hydroxide and tert-butoxide derivatives.10b In contrast,
upon heating the THF solutions of 2b or 2c above 50 °C,
these compounds undergo clean transformation into the
corresponding η2-benzaldehyde (4b) or η2-acetophenone
(4c) complexes. Like η2-imine derivatives, these compounds
are easily identified by their characteristic 31P{1H} spectra,
which consist of AB patterns with large JAB constants (ca. 70
Hz). Similarly to compound 3b, the NMR spectra of com-
pounds 4 are sharp at room temperature and there is no
evidence of fluxional behavior. Therefore, the previously
reported10b fluxionality of the imine complex 3a, which
undergoes rapid π:η2/σ:κ-N coordination shift, appears to
be rather unique. However, π:η2/σ:κ-O isomerization of
transient iridium η2-ketone complexes has recently been
invoked in order to explain the racemization of chiral alk-
oxide derivatives.13 The 13C resonances of the carbonyl
carbon atoms are shifted upfield by 114 (4b) and 117 ppm
(4c) from their positions in the free ligands, as a consequence
As described previously,10b amido complexes 1 are ther-
mally unstable and gradually decompose in solution at room
temperature. This process can be readily monitored by 31P
NMR directly using samples generated in THF, avoiding
further manipulation of these compounds, which are extre-
mely sensitive to water traces (Scheme 2). The pyrrolidino
derivative is the least stable and disappears from the solution
within 4 h, affording 3a and minor amounts of other nickel
complexes, including NiMe2(dippe) and Ni(dippe)2, which
probably are formed by a competing disproportionation
process. However, slightly above room temperature
(40 °C), β-hydrogen elimination becomes the dominant pro-
cess, and 3a is produced in a selective manner. Decomposition
of the dibenzylamido derivative 1d is slower (half-life ≈ 1 day
at room temperature), but cleanly affords 3d as the only
(11) (a) Schwartz, J.; Cannon, J. B. J. Am. Chem. Soc. 1974, 96, 2277.
(b) Bergwall, C.; Parsons, E. J. J. Organomet. Chem. 1994, 483, 73.
ꢀ
(12) (a) Campora, J.; Matas, I.; Palma, P.; Graiff, C.; Tiripicchio, A.
ꢀ
Alvarez, E.; Graiff, C.; Tiripicchio, A. Organometallics 2007, 26, 5712.
Organometallics 2005, 24, 2827. (b) Campora, J.; Matas, I.; Palma, P.;
ꢀ
(13) Macgregor, S. A.; Vadivelu, P. Organometallics 2007, 26, 3651.

Article
Organometallics, Vol. 28, No. 22, 2009 6517
of the substantial dfπ* back-donation.14 Similarly, the 1H
aldehyde signal of 4b appears at δ 6.03 ppm, almost 4 ppm
upfield to that in free benzaldehyde. It is interesting to
compare these coordination shifts with those of the analo-
gous signals of the imine derivative and the styrene complex 6
(see below). The R-methyne resonances of 3d and 6 have
nearly identical upfield shifts of ca. 87 ppm relative to those
of the free molecules, which suggests that PhCH2NdCHPh
and styrene have similar π-acceptor strengths. On the other
hand, the coordination shift of the imine carbon of 3a, 98
ppm based on its estimated chemical shift in 3,4-dihydro-2H-
pyrroline (δ 169 ppm), suggests that this unstable heterocycle
has acceptor strength intermediate between those of aceto-
phenone or benzaldehyde and the acyclic imine. A set of closely
related Ni(dippe)(η2-imine) complexes, recently reported by
J. J. Garcıa and co-workers, containing fluorinated imine
´
ligands show nearly identical coordination shifts of ca. 100
ppm.15 Therefore, 13C coordination shifts indicate that π-
acceptor capabilities decrease in the order CdO > CdN g
CdC. Unfortunately, it is not possible to support these con-
clusions with IR frequencies of CdE bond stretches, as these
Figure 1. Crystal structure of compound 4b. Selected bond
distances (A) and angles (deg) (one of two independent
molecules): Ni-O1, 1.8726(12); Ni-C15, 1.9323(18); Ni-P1,
2.1319(5); Ni-P2, 2.1740(5); O1-C15, 1.345(2); C15-C16,
1.476(2); O1-Ni-C15, 41.36(7); O1-Ni-P2, 116.12(4);
C15-Ni-P1, 112.00(6); P1-Ni-P2, 90.92(2); O1-C15-H15,
121.7; O1-C15-C16, 119.81(15); H15-C15-C16, 112.2.
are shifted to the crowded spectral region below 1300 cm-1
,
where their assignment becomes uncertain.16
The crystal structures of 4b and 4c are shown in Figures 1
and 2. The unit cell of the benzaldehyde complex contains
two crystallographically independent molecules, but these
show no significant differences. As expected, the nickel atom
is formally three-coordinated, with one of the positions
occupied by the η2 carbonyl ligand, which lies in the co-
ordination plane.17 As a result of its interaction with the
metal center, the aldehyde or the ketone ligand loses its
planarity. Thus, in complex 4c the sum of the angles
O1-C1-C2, O1-C1-C3, and C2-C1-C3 is smaller than
360° by ca. 10°. In spite of this, the phenyl ring lies coplanar
to the formal CdO bond, indicating that the latter retains
some degree of double-bond character. Bond distances and
angles are very similar in the two compounds and resemble
those observed in related Ni(0) ketone or aldehyde com-
plexes described in the literature.18 The two compounds
exhibit virtually identical C-O bond lengths (1.345(2) for
4b and 1.3434(2) A for 4c) that lie between those of the single
C-O bond in oxiranes (ca. 1.45 A) and the double bond in
aromatic ketones or aldehydes (1.221 A).19 These distances
are slightly longer than those reported for similar Ni(η2-
CdO) complexes (ranging between 1.32 and 1.33 A), for
example, 1.331(6) A in the closely related benzophenone
complex [Ni(η2-Ph2CO)(dtbpe)] (dtbpe = 1,2-bis(di-tert-
butylphosphino)ethane), reported by Hillhouse.18d
Figure 2. Crystal structure of compound 4c. Selected bond
distances (A) and angles (deg): Ni-O1, 1.8531(8); Ni-C1,
1.9606(11); Ni-P1, 2.1382(3); Ni-P2, 2.1764(3); O1-C1,
1.3434(2); C1-C3, 1.4923(16); C1-C2, 1.5176(17); O1-
Ni-C1, 41.13(4); O1-Ni-P2, 111.26(3); C1-Ni-P1,
116.18(4); P1-Ni-P2, 91.484(12); O1-C1-C2, 116.74(11);
C2-C1-C3, 117.62(11); O1-C1-C3, 116.46(10).
Since decomposition of transition metal alkyls by β-hydro-
gen elimination is well documented, we decided to undertake
the synthesis of the related phenethyl derivative 5 (Scheme 3).
The synthesis of such a compound is not a trivial task, since
nickel bis(alkyl) complexes with β-hydrogen atoms have usual-
ly low thermal stability and, in addition, mixed alkylnickel
derivatives are rather unusual compounds.20 Thus, we explored
the application of the fluoride exchange methodology to the
synthesis of this alkylnickel derivative. Due to their easy
preparation, we favored the use of alkylmagnesium reagents
instead of their lithium counterparts. Diphenethylmagnesium
was generated by dioxane addition to a solution of the
phenethyl Grignard reagent in Et2O and then reacted with
(14) Tolman, C. A.; English, A. D.; Manzer, L. E. Inorg. Chem. 1975,
14, 2353.
~
ꢀ
(15) Iglesias, A. L.; Munoz-Hernandez, M.; Garcıa, J. J. J. Organo-
´
met. Chem. 2007, 692, 3498.
(16) A similar trend has been deduced for [Ni(η2-CdX)(NtCtBu)2]
based on their isocyanide IR NtC stretch bands. For example, for
(CF3)2CdX complexes, ν(CtN) = 2188 cm-1 (X = O); 2168 (X =
1
NMe); 2165 (X = CH2; estimated). Ittel, S. D. Inorg. Chem. 1977, 16,
2589.
(17) Huang, Y.-H.; Gladysz, J. A. J. Chem. Ed. 1988, 65, 298.
(18) (a) Countryman, R.; Penfold, B. R. J. Chem. Soc. D: Chem.
Commun. 1971, 1598. (b) Tsou, T. T.; Huffman, J. C.; Kochi, J. K. Inorg.
Chem. 1979, 18, 2311. (c) Kaiser, J.; Sieler, J.; Walther, D.; Dinjus, E.; Golic,
L. Acta Crystallogr. B: Struct. Crystallogr. Cryst. Chem. 1982, 38, 1584.
(d) Mindiola, D. J.; Waterman, R.; Jenkins, D. M.; Hillhouse, G. L. Inorg.
Chim. Acta 2003, 345, 299. (e) Ogoshi, S.; Ueta, M.; Arai, T.; Kurosawa, H.
J. Am. Chem. Soc. 2005, 127, 12810.
ꢀ
(20) Campora, J. Nickel-Carbon s-Bonded Complexes. In Compre-
(19) CRC Handbook of Chemistry and Physics on CD-ROM; Linde, D.
R., Ed.; CRC Press: Boca Raton, FL, 2009.
hensive Organometallic Chemistry III, Vol. 8; Crabtree, R., Mingos, M.,
Eds.; Canty, A., Vol. Ed.; Elsevier: Oxford, 2006.

6518 Organometallics, Vol. 28, No. 22, 2009
Matas et al.
Scheme 3
Ni(dippe)(Me)(F). This method leadstothe mixed alkyl, which
was isolated in high yield (80%) as an orange crystalline solid.
The NMR spectra of this product reveal the presence of
variable amounts of a minor species, identified as the secondary
alkyl isomer, 50. The 5:50 ratio is variable and depends strongly
on the experimental conditions, but they mutually interconvert
in THF, attaining a 6:1 equilibrium ratio within 1 h at room
temperature.
Figure 3. Crystal structure of compound 50. Selected bond dis-
tances (A) and angles (deg): Ni-C15, 1.987(4); C16-C18,
1.489(5); Ni-C16, 2.039(3); C16-C17, 1.537(5); Ni-P2,
2.1666(9); Ni-P1, 2.2075(9); C15-Ni-C16, 88.70(16); C15-
Ni-P2, 89.25(12); C16-Ni-P1, 93.49(10); P1-Ni-P2, 88.40(3).
The presence of the primary phenethyl group in the main
product 5 is confirmed by its 1H and COSY NMR spectra,
which show characteristic signals at δ 1.48 and 3.48 ppm for
the two connected CH2 groups. The 13C resonance of the
nickel-bound CH2 (δ 19.9 ppm) group is partially hidden by
the signals of the phosphine ligand. As expected for a
complex containing two alkyl groups bound to Ni, the
31P{1H} spectrum displays two nearby doublets separated
by a few ppm (δ 72.7, 75.4 ppm; 2JPP = 6 Hz). In contrast, 50
gives rise to two 31P signals separated by 15 ppm (63.5,
78.4 ppm; 2JPP= 10 Hz), evincing the dissimilarity between
the alkyl groups. The 1H spectrum of this compound shows
multiplets with 1:3 relative intensities at δ 4.33 and 1.70 ppm,
corresponding to the CH-CH3 unit. The structure of 50 was
confirmed by its X-ray structure (Figure 3), established from
a single crystal selected from the isomeric mixture. This
chiral compound crystallizes as racemic crystals containing
the two enantiomers. As can be seen, the orientation of the
secondary alkyl group minimizes the steric interactions,
projecting the methyl group above and the phenyl group
below the coordination plane. Noteworthy, the Ni-C(H)-
(Me)(Ph) bond (2.039 A) is slightly longer than the Ni-Me
bond (1.987 A), owing to the larger steric hindrance in the
former. The Ni-P bonds show significant differences as well,
as the bond cis to the secondary alkyl group is somewhat
longer than the other (2.2075(9) vs 2.1666(9) A). This
difference is probably the cause of the relatively large
separation of the 31P NMR signals. No other remarkable
aspects in this structure are worthy of comment.
As expected, the alkyl complexes 5/50 readily experience a
β-elimination process. Upon heating a solution of these
compounds at 60 °C for 1.5 h, the η2-styrene complex 6 is
formed, which is easily isolated as a yellow crystalline solid in
65 % yield (Scheme 3). The X-ray structure of this com-
pound, shown in Figure 4, is typical for a η2 Ni(0) olefin
complex.
Relative Stability of the Ni(0) Complexes. Ni(0) complexes
undergo facile ligand exchange reactions, often reversibly.
Equilibrium constants associated with these equilibria can be
used as an indication of the relative binding capabilities of
the involved ligands to the Ni(0) center. In our system, these
data are of interest since the stability of the η2-heteroolefin
Scheme 4
complexes could provide a driving force for the above
mentioned decomposition processes. Thus, we studied sev-
eral ligand exchange reactions by titrating samples of the
suitable Ni(0) complex (3c, 4b, 4c, or 6) with increasing
amounts of the desired heteroolefin and monitoring the
reaction mixtures by 31P{1H} at 25 °C. The main results of
this study are summarized in Scheme 4.
In general, these ligand exchange reactions proceed
cleanly at room temperature, leading to full or partial
displacement of the π-bonded heteroolefin. The imine com-
plex 3a is an exception since 3,4-dihydro-2H-pyrroline is
known to be an unstable molecule.21 Benzaldehyde reacts
rapidly with an equimolar amount of the ketone (4c) and the
imine (3d) complexes, quantitatively yielding the corre-
sponding aldehyde complex 4b. It also displaces styrene from
6, but in this case the reaction is not quantitative. The latter
equilibrium can be studied more accurately by investigating
the opposite reaction, i.e., by titrating 4b with styrene. The
corresponding equilibrium constant (first entry in Scheme 4)
confirms that benzaldehyde is also stronger ligand than
styrene, but the difference is not as marked as for the ketone
or the imine.
In order to quantify the relative binding strength of the
ketone ligand, we studied the reaction of 4b with acetophe-
none. Small amounts of compound 4c can be detected when
4b is treated with a large exces of acetophenone (>10 equiv),
(21) Kraus, G. A.; Neuenschwander, K. J. Org. Chem. 1981, 46, 4791.

Article
Organometallics, Vol. 28, No. 22, 2009 6519
Figure 4. Crystal structure of compound 6. Selected bond dis-
tances (A) and angles (deg): Ni-C1, 1.9619(13); C1-C2,
1.4264(18); Ni-C2, 1.9801(12); C2-C3, 1.4777(19); Ni-P1,
2.1600(4); Ni-P2, 2.1483(4); C1-Ni-C2, 42.42(5); C1-Ni-P2,
111.81(4); C2-Ni-P1, 113.83(4); P1-Ni-P2, 91.931(14).
but the reaction is sluggish and difficult to use for quantita-
tive purposes. Although titrations of 6 with acetophenone or
with the imine ligand (N-benzylidenebenzylamine)22 proceed
slowly as well, these reactions provided reliable results when
the samples were allowed to equilibrate in a thermostatic
bath for g24 h prior to each measurement. As shown in
Scheme 4, the Keq values determined for these two equilibria
are small (within the 10-2-10-3 range) and indistiguishable
within the experimental uncertainty. Therefore, the relative
stability of the Ni(η2-CdX) linkage decreases in the order 4b
(benzaldehyde) > 6 (styrene) . 4c (ketone) ≈ 3d (imine).
The different relative stabilities of the ketone and the alde-
hyde complexes is particularly noteworthy, since these two
compounds show very similar spectroscopic properties. This
suggests that the relative binding capability of these hetero-
olefins is dominated by steric factors, which is confirmed
by the similar binding capability displayed by sterically
alike ligands (i.e., benzaldehyde/styrene; acetophenone/
N-benzylidenbenzylamine). However, it is likely that the
origin of the small but noticeable difference between benzal-
dehyde and styrene might be electronic in nature.
Kinetics and Mechanism of the Decomposition Reactions of
Amido, Alkyl, and Alkoxo Complexes. The observation of
clean decomposition reactions of complexes 1d, 2b,c, and 5
prompted us to investigate their kinetics in order to gather
additional information regarding their mechanisms. Due to
the high hydrolytic and/or thermal sensitivity of these com-
pounds, we chose to avoid sample manipulation as much as
possible and studied the decomposition processes in clean
samples in THF, monitoring the reaction progress by 31P{1H}
NMR. Although the transformation of pyrrolidine complex
1a into pyrroline 3a is selective above 40 °C, preliminary
experiments showed that at higher temperatures it becomes
too fast for accurate monitoring by the described techniques.
Figure 5. First-order plots for the thermolyses of compounds
1d (up) and 5 (bottom) at different temperatures.
Decompositions of the amido and alkyl complexes 1d and
5 obey simple first-order rate laws for over 3 half-lives
(Figure 5). In order to determine their activation parameters,
both reactions were studied at different temperatures. The
corresponding Eyring plots are shown in Figure 6 together
with the corresponding activation parameters. The decom-
position rate of the dibenzylamido derivative 1d experiences
a dramatic increase with the temperature, narrowing the
interval available for NMR monitoring (297 to 310 K). The
half-life of 1d is only 3 min at the highest temperature, but at
room temperature is more than 100 min. Experiments aimed
to measure the rate at sub-ambient temperatures met with
problems due to the very long times required to observe
substantial transformation into 3d and to other competing
processes. The narrowness of the accessible temperature
range severely limits the accuracy of ΔHq and ΔSq, which
can be considered as merely indicative. However, it can be
concluded that the transformation is characterized by a
strongly positive value of the activation entropy.
31P monitoring of the decomposition of the 5/50 mixture of
alkyl complexes showed that the two isomers equilibrate
above ambient temperature within the first minutes or
seconds of the experiment. Estimated rate constants for this
exchange process indicate that it is at least one order of
magnitude faster than the formation of product 6. On the
other hand, the 5/50 equilibrium ratio rapidly increases with
temperature in such a way that at 305 K it is 14:1 and at
31
higher temperatures the P signal of 50 becomes undetect-
able. In contrast with the amido 1d, the decomposition of 5 is
characterized by a small and slightly negative activation
entropy, and therefore the measurements could be extended
to a wider temperature range (297-335 K).
As stated previously (Scheme 1), the mechanisms of these
two reactions must involve at least two steps: β-hydrogen
elimination leading to intermediate methylnickel hydrides
and decomposition of this intermediate by reductive elim-
ination to afford methane and the corresponding Ni(0) η2
olefin or imine complex. Although free activation energies
(22) Reaction of 6 with N-benzylidenebenzylamine produces a sec-
ond product besides 3d, which is characterized by an AX set of
resonances in the 31P{1H} spectrum, with δA = 66.4, δB = 64.8, and
JAB = 65.6 Hz and relative intensity 0.5 with regard to the former. The
similarity of the parameters of these signals to those of 3d (δA = 67.5,
δB = 66.2, and JAB = 69.3 Hz) suggests that both products might be
geometric isomers arising from the Z or E configurations of the imine
ligand. The equilibrium constant reported in Scheme 4 corresponds to
the equilibrium involving 6 and 3d.

6520 Organometallics, Vol. 28, No. 22, 2009
Matas et al.
Scheme 5
into the ketone complex 4c is initially zero order on the
complex as well, but after a few minutes gradual decay of the
reaction rate becomes noticeable. This transformation is also
quantitative, according to 31P spectra. Figure 7 shows over-
laid zero-order plots for the decomposition of 0.06 M
samples of the two compounds in THF, at 60 °C. As can
be seen, the initial apparent zero-order rate constants
are very similar for the two compounds (apparent k ≈ 7 ꢀ
10-4 mol L-1 min-1, half-life ≈ 40 min). Zero-order decom-
Figure 6. Eyring plots for the decompositions of the amido
complex 1d (b, 297-310 K) and alkyl 5 (9, 297-335 K).
Activation parameters are shown in the insets.
are fairly similar for both decomposition processes at room
temperature, this is due to the mutual compensation of very
different activation enthalpy and entropy values, which
indicates that the corresponding mechanisms must involve
significant differences. The rapid equilibration of the pri-
mary and secondary alkyl compounds 5 and 50 is mechani-
stically significant, as it clearly indicates that a reversible
β-elimination must precede the irreversible reductive elim-
ination of CH4, which, as shown in Scheme 5, acts as the rate-
limiting step. Decompositions of transition metal alkyls by
β-elimination often have product release as their rate-limit-
ing step.8a A near-zero ΔSq is quite typical of this mechan-
ism, as it comprises the algebraic sum of the activation
parameter corresponding to the rate-limiting step and the
thermodynamic entropies associated with the preceding
equilibria, which probably have different signs and tend to
cancel mutually.
The first-order kinetics observed in the thermolysis of the
amido 1d is compatible with the generic mechanism depicted
in Scheme 1. The only remarkable feature of this process is its
positive activation entropy, which suggests that some sig-
nificant mechanistic differences with the decomposition of
the alkyl complex might exist. Although a positive value of
ΔSq suggests that some ligand dissociation (i.e., Ni-N
heterolysis23 or Ni-P scission24) might take place at the
rate-determining step, at this point we lack any additional
evidence to support such a proposal.
3
3
position kinetics was consistently observed with 2b, in both
THF and C6D6, but the value of the apparent rate constants
showed some variability depending on the origin of the
sample. For instance samples of 2b prepared from isolated,
crystalline material decomposed about 2 times slower than
samples generated in situ in THF.
The observed zero-order rate law, the similarity of the
initial decomposition rates of the two compounds, and the
variability of this rate indicate that an external agent cata-
lyzes this process. The observed rates must be proportional
to the concentration of this agent, which depends on the
sample history. The nature of the catalyst is currently
unknown, but it must be some substance present in small
amounts, since isolated and analytically pure samples ex-
perience this decomposition process. The catalyst is prob-
ably similar for 2b and 2c, although in the latter case it
gradually loses its activity over time, leading to the observed
decrease of the reaction rate. Bergman has shown that
β-hydrogen elimination from iridium alkoxides of the type
[Ir(Cp*)(Ph)(OCH2R)(PMe3)] is catalyzed by small amounts
of the cationic species [Ir(Cp*)(Ph)(PMe3)]þ, which is
generated from the starting material, the triflate
[Ir(Cp*)(Ph)(OTf)(PMe3)].25 Since low coordination capa-
bility anions were not used in the synthesis of 2, the
involvement of cationic, coordinatively unsaturated species
appears unlikely. Apart from small amounts of hydroxide
[Ni(Me)(OH)(dippe)], the NMR spectra of the alkoxo com-
pounds used in this study lack signals attributable to other
compounds, and this has led us to suspect that a paramag-
netic Ni(I) species, similar to the amido derivative
[Ni(NHC6H3-2,6-iPr2)(dtbpe)] (dtbpe = 1,2-bis(di-tert-
butylphosphino)ethane), reported by Hillhouse,26 could be
responsible for the observed behavior. A small amount of
this impurity would not alter significantly the analytical
data, since its composition only differs by one methyl from
In contrast with 1d and 5, samples of 2b decay linearly with
time (in either THF or C6D6); that is, the reaction exhibits
zero-order dependency on the complex. At 60 °C this kinetic
law is obeyed until the complex is quantitatively converted
into 4b. The transformation of the secondary alkoxide 2c
(23) Heterolysis of the M-X bond has been proposed to take place
during the decomposition of some late transition metal alkoxides (see
refs 9d,f,g). However such heterolytic processes are often characterized
by negative activation entropies. See also: Kawataka, F.; Kayaki, Y.;
Shimizu, I.; Yamamoto, A. Organometallics 1994, 13, 3517.
(24) Phosphine ligand dissociation is a relatively common feature of
β-hydrogen elimination mechanisms. See for example: (a) Ozawa, F.;
Ito, T.; Yamamoto, A. J. Am. Chem. Soc. 1980, 102, 6450. (b) Komiya, S.;
Morimoto, Y.; Yamamoto, A.; Yamamoto, T. Organometallics 1982, 1, 1528.
(c) Nuzzo, R. G.; McCarthy, M. C.; Whitesides, G. M. Inorg. Chem. 1981,
20, 1312. (d) McCarthy, T. J.; Nuzzo, R. G.; Whitesides, G. M. J. Am. Chem.
Soc. 1981, 103, 3396. (e) Alexanian, E. J.; Hartwig, J. F. J. Am. Chem. Soc.
2008, 130, 15627.
(25) Ritter, J. C. M; Bergman, R. G. J. Am. Chem. Soc. 1998, 120,
6826.
(26) Mindiola, D. J.; Hillhouse, G. L. J. Am. Chem. Soc. 2001, 123,
4623.

Article
Organometallics, Vol. 28, No. 22, 2009 6521
(25 mol %) of [Ni(μ-H)(dippe)]2, a well-known Ni(I) hydrido
compound.29 This cannot be the active catalyst, since it is
diamagnetic and its presence would not have passed unnoticed
in the spectra of 2b or 2b. However, it caused the decomposition
rate to become nearly doubled (from k ≈ 3 ꢀ 10-4 to 8 ꢀ 10-4).
This effect is too small to support an active role of this
compound as catalyst, but it suggests that the dimer could
act as a source of catalytically active species. Although these
experiments give no evidence of what might be the active
catalyst, they are compatible with our proposal of a paramag-
netic Ni(I) complex.
Figure 7. Zero-order plots of the decomposition of 2b and 2c in
THF (0.06 M) at 60 °C.
Conclusions
Monomeric methylnickel amido, alkoxo, and alkyl com-
plexes bearing hydrogen atoms in the β position can undergo
formally analogous decomposition reactions, eliminating
methane and selectively affording the corresponding Ni(0)
imine, ketone, aldehyde, and alkene complexes through a
sequence of β-elimination and reductive elimination reac-
tions. The process is highly selective for the secondary amido
complexes 1a and 1d, as well as for the alkoxides 2b,c and the
alkyl 5. However, formation of an η2-aldehyde product from
the methoxide 2a has not been observed, and decomposition
of the primary amido complexes 1b and 1c involves compe-
ting β-elimination and disproportionation mechanisms, aff-
ording mixtures of products. Nickel(0) η2-heteroolefin and
olefin products have been isolated and fully characterized.
Ligand exchange reactions show that the thermodynamic
stability of the Ni(η2-CdX) linkage decreases in the order 4b
(benzaldehyde) > 6 (styrene) . 4c (acetophenone) ≈ 3d
(N-benzylidenebenzylamine). Steric effects probably have a
major role in determining this trend.
Kinetic measurements on the decompositions of com-
pounds 1d, 2b,c, and 5 have uncovered dissimilarities that
point to significant differences in the mechanisms of these
decomposition reactions. Nickel amido 1d and the alkyl
complex 5 spontaneously evolve following simple first-order
kinetic rate laws. However, their respective activation en-
tropies display very different values, large and positive for
the former and slightly negative for the latter, suggesting that
different reaction steps control the reaction rate. The facile
isomerization of the primary alkyl derivative 5 to its second-
ary isomer 50 indicates that in this case β-elimination is
reversible and the reaction rate is controlled by the irrever-
sible reductive elimination of methane. The decomposition
rates of the alkoxides 2b and 2c display zero-order depen-
dency on the alkoxide, which points to a catalytic process.
The identity of the catalyst is unknown, but we propose that
it could correspond to Ni(I) alkoxides related to compounds
2. This conclusion is supported by the inhibition of the
decomposition reaction by TEMPO and its acceleration by
the binuclear Ni(I) hydride [Ni(μ-H)(dippe)]2.
Scheme 6. Proposed Mechanism for the Decomposition of
Alkoxide Complexes 2b (R = H) or 2c
that of the main species. A plausible decomposition mecha-
nism involving a Ni(I) alkoxide is suggested in Scheme 6. The
rate-limiting β-elimination process would take place on a
Ni(I) complex [Ni(OR)(dippe)], and the resulting hydride
would then be transferred to the Ni(II) compound to afford
methane and regenerate the Ni(I) catalyst.
In order to provide some support to the mechanism out-
lined in Scheme 6, we carried out additional experiments with
2b in THF at 60 °C. The first experiment monitored the
decomposition in three solutions of different concentration
prepared from a single sample of this material. The apparent
decomposition rate constants were observed to be propor-
tional to the initial concentration (0.021 M, k = 3 ꢀ
10-4 mol L-1 min-1; 0.043 M, k = 5 ꢀ 10-4 mol L-1 min-1
;
3
3
3
3
0.18 M, k = 3.3 ꢀ 10-3 mol L-1 min-1). This is expected if the
3
3
reaction is catalyzed by some impurity, as the catalyst concen-
tration will increase together with that of the substrate. The
second experiment studied the effect of TEMPO (2,20,6,60-
tetramethylpiperidinyloxy) on the reaction rate. TEMPO is a
radical-trapping agent extensively used as a diagnostic tool for
radical mechanisms,27 but its ability to trap transition metal
hydride species has also been shown.28 More interestingly,
TEMPO reacts with a Ni(I) compound such as [NiCl(dtbpe)]
to yield a stable Ni(II) adduct.18d Thus, TEMPO is likely to act
as a trap for either hydride or Ni(I) intermediates. Addition of
5 mol % of TEMPO to a 0.06 M solution of 2b significantly
inhibits the decomposition process at 60 °C (the half-life of 2b
increases from 40 min to ca. 15 h), whereas using 1 equiv
completely suppresses it. In the third experiment, the decom-
position of 2b was monitored in the presence of 0.25 equiv
The different mechanistic features detected for amido (1d),
alkoxo (2b,c), and alkyl (5) derivatives makes it difficult to
directly compare their tendencies to undergo β-elimination
reaction. However, it seems that while the alkyl and amido
derivatives spontaneously experience β-hydrogen elimina-
tion, alkoxides will not unless a catalyst is present. The
reasons for this behavior are not easy to ascertain, but it
appears obvious that the feasibility of these decomposition
processes bear no direct relation to the thermodynamic
stability of the Ni(η2-CdX) fragment that is generated in
the last term, since in this case the decomposition of the
(27) Lawrence, L. M.; Whitesides, G. M. J. Am. Chem. Soc. 1980,
102, 2493.
ꢀ
(28) Elia, C.; Elyashiv-Barad, S.; Sen, A.; Lopez-Fernandez, R.;
ꢀ
Albeniz, A. C.; Espinet, P. Organometallics 2002, 20, 4249.
(29) Vicic, D. A.; Jones, W. D. J. Am. Chem. Soc. 1997, 119, 10855.

6522 Organometallics, Vol. 28, No. 22, 2009
Matas et al.
benzoxide derivative 2b to the benzaldehyde complex would
be expected to be more favorable than those of the
R-methylbenzoxide 2c or dibenzylamido derivative 1d, respec-
tively, leading to the acetophenone and imine complexes.
More definitive answers to these questions will require further
mechanistic investigations.
PCHMeMe), 19.1 (d, 2JCP = 3 Hz, PCHMeMe), 19.6 (d, 2JCP = 7
2
Hz, PCHMeMe), 19.7 (d, JCP1 = 7 Hz, PCHMeMe), 21.8 (pt,
CP
J* ≈ 22 Hz, CH2), 23.9 (d, JCP = 4 Hz, PCHMe2), 24.1 (d,
1JCP = 4 Hz, PCHMe2), 24.9 (dd, 1JCP = 19 Hz, 3JCP = 4 Hz,
PCHMe2), 26.4 (s, PhCOCH3), 81.8 (d, JCP = 23 Hz, Ph-
2
COCH3), 122.7 (s, CarHp), 124.1 (s, CarHo), 153.0 (s, Car). 31P{1H}
=
NMR (C6D6, 121 MHz): δ 65.0 (d, 2JPP = 71 Hz), 69.1 (d, 2JPP
71 Hz).
Synthesis of [Ni(Me)(C2H4Ph)(dippe)] (5/50). A 0.15 M solu-
tion of diphenethylmagnesium, Mg(CH2CH2Ph)2, was pre-
pared by adding 6.4 mL of dioxane (75 mmol, 2 equiv) to a
freshly prepared solution of the corresponding Grignard re-
agent (75 mL, 0.35 M, 37.5 mmol). The resulting precipitate of
MgBr2(dioxane)2 was removed by filtration. An aliquot
was taken, hydrolyzed, and titrated with a standard 0.1 N HCl
solution to determine the reagent concentration. Then 1.21 mL
(0.18 mmol) of the Mg(CH2CH2Ph)2 solution was added to a
cooled (-78 °C) solution containing 129 mg (0.36 mmol) of
complex [Ni(Me)(F)(dippe)] in 3 mL of THF, causing an
immediate color change from orange to yellow. After the
solution reached room temperature the solvent was removed
under reduced pressure, and the resulting residue was extracted
with 5 mL of pentane. Yellow crystals of product 5, containing
some 50, were obtained in 79% yield by concentrating and
cooling this solution to -20 °C. Anal. Calcd for C23H44NiP2:
Experimental Section
General Considerations. All preparations were carried out
under an oxygen-free nitrogen atmosphere by conventional
Schlenk techniques. Solvents were rigorously dried and de-
gassed before use. Microanalyses were performed by the Micro-
analytical Service of the Instituto de Investigaciones Quımicas
´
(Sevilla, Spain). Infrared spectra were recorded on a Bruker
Vector 22 spectrometer, and NMR spectra on Bruker DRX 300
and 400 MHz spectrometers. 1H and 13C{1H} resonances of the
solvents were used as internal standard, but the chemical shifts
are reported with respect to TMS. 31P resonances are referenced
to external 85% H3PO4. Compounds 1a-d, 2a-c, and 3a-d
were prepared or generated in THF solution as described
previously.10b,12
Synthesis of [Ni(η2-OdCHPh)(dippe)], 4b.
Alkoxide
[Ni(Me)(OCH2Ph)(dippe)] (2b) (133 mg, 0.3 mmol) was dis-
solved in THF (5 mL). The resulting solution was heated at
80 °C for 3 h. The solvent was then removed under vacuum, and the
residue extracted with hexane (5 mL). Crystallization at -20 °C
gave the product as a dark orange solid. Yield: 65 mg, 50%. This
compound can also be prepared directly from [Ni(Me)-
(F)(dippe)], without isolating the alkoxide 2b: a solution of
[Ni(Me)(F)(dippe)] in THF is treated with an equimolar amount
of lithium benzoxide in the same solvent, and the mixture is
heated at 80 °C for 3 h. The same workup is applied to isolate the
product. Anal. Calcdfor C21H38NiOP2: C, 59.05; H, 8.97. Found:
C, 58.85; H, 8.84. 1H NMR (C6D6, 50 °C, 400 MHz): δ 0.43 (dd,
3H, 3JHP = 15.4 Hz, 3JHH = 7.1 Hz, PCHMeMe), 0.59 (pt, 3H,
1
C, 62.61; H, 10.05. Found: C, 62.42; H, 10.01. Complex 5: H
NMR (C6D6, 400 MHz): δ 0.52 (dd, 3H, 3JHP = 9.3, 3.6 Hz, Ni-
CH3), 0.87 (m, 12H, PCHMeMe), 1.08 (m, 12H, PCHMeMe),
1.48 (m, 2H, CH2CH2Ph), 1.92 (m, 4H, PCHMe2), 3.18 (m, 2H,
3
CH2CH2Ph), 7.10 (t, 1H, JHH = 6.9 Hz, CarHp), 7.28 (t, 2H,
3
3JHH = 6.9 Hz, CarHm), 7.55 (d, 2H, JHH = 7.0 Hz, CarHo).
13C{1H} NMR (C6D6, 75 MHz): δ 3.4 (dd, 2JCP = 70, 20 Hz, Ni-
CH3), 18.3 (s, PCHMeMe), 19.5 (d, 2JCP = 5 Hz, PCHMeMe),
1
19.9 (m, CH2CH2Ph), 20.9 (m, CH2), 24.4 (dd, JCP = 18 Hz,
3JCP = 3 Hz, PCHMe2), 37.2 (s, CH2CH2Ph), 124.2 (s, CarHp),
126.3 (s, CarHm), 126.9 (s, CarHo), 150.4 (d, 4JCP = 7 Hz, Car).
31P{1H} NMR (C6D6, 162 MHz): δ 72.7 (d, 2JPP = 6 Hz), 75.4
2
(d, JPP = 6 Hz). Complex 50: 1H NMR (C6D6, 400 MHz)
3JHP ≈ JHH ≈ 9.0 Hz, PCHMeMe), 1.02 (m, 12H, PCHMeMe),
3
δ-0.01 (m, 3H, Ni-CH3), 1.70 (pt, 3H, J* = 6.5 Hz, CH(CH3)Ph),
4.33 (m, 1H, CH(CH3)Ph), 7.06 (t, 1H, 3JHH = 7.6 Hz, CarHp),
7.38 (t, 2H, 3JHH = 7.6 Hz, CarHm), 7.48 (d, 2H, 3JHH = 7.5 Hz,
CarHo). 31P{1H} NMR (C6D6, 162 MHz): δ63.5 (d, 2JPP =10Hz),
78.4 (d, 2JPP = 10 Hz).
1.21 (m, 6H, PCHMeMe), 1.44 (m, 1H, PCHMe2), 1.63 (m, 1H,
PCHMe2), 1.77 (m, 2H, PCHMe2), 6.03 (s, 1H, PhCHO), 7.04
(m, 1H, CarHp), 7.18 (m, 2H, CarHm), 7.69 (d, 2H, 3JHH = 6.9 Hz,
CarHo). 13C{1H} NMR (C6D6, 75 MHz): δ 16.9 (s, PCHMeMe),
2
18.5 (d, JCP = 6 Hz, PCHMeMe), 18.8 (s, PCHMeMe), 19.3
(s, PCHMeMe), 19.6 (s, PCHMeMe), 19.8 (d, JCP = 7 Hz,
Synthesis of [Ni(η2-CH2dCHPh)(dippe)], 6. A solution of
125 mg (0.28 mmol) of dialkyls 5/50 in 4 mL of THF was heated
at 60 °C for 1.5 h. The solvent was then removed under vacuum,
and the residue extractedwith 5 mLof diethyl ether. Product6 was
obtained as orange crystals after crystallization from this solution
at -20 °C. Yield = 65 %. Anal. Calcd for C22H40NiP2: C, 62.14;
H, 9.48. Found: C, 61.96; H, 9.26. 1H NMR (C6D6, 500 MHz): δ
0.43 (dd, 3H, 3JHP = 15.5 Hz, 3JHH = 7.2 Hz, PCHMeMe), 0.65
(dd, 3H, 3JHP = 10.5 Hz, 3JHH = 7.0Hz, PCHMeMe), 0.83-0.95
2
PCHMeMe), 21.2 (pt, J* ≈ 21 Hz, CH2), 23.3 (dd, JCP
=
1
CP
18Hz, 3JCP = 5 Hz, PCHMe2), 23.9 (d, 1JCP = 13 Hz, PCHMe2),
=
25.2 (dd, 1JCP = 19 Hz, 3JCP = 5 Hz, PCHMe2), 78.0 (d, 2JCP
19 Hz, PhCHO), 123.1 (s, CarHp), 123.4 (s, CarHo), 151.6 (s, Car).
31P{1H} NMR (C6D6, 121 MHz): δ 69.7 (d, 2JPP = 66 Hz), 70.9
(d, 2JPP = 66 Hz).
Synthesis of [Ni(η2-OdCMePh)(dippe)], 4c. A solution of 137
mg (0.3 mmol) of alkoxide Ni(dippe)(Me)(OCH(CH3)Ph) (2c)
in 5 mL of THF was heated at 50 °C for 6 h. The solvent was then
removed under vacuum, and the residue extracted with 5 mL of
hexane. The product was obtained as a dark orange solid after
crystallization from this solution at -20 °C. Yield: 60 mg, 45%.
Similarly to 4b, this compound can also be obtained directly
from [Ni(Me)(F)(dippe)], without isolation of the alkoxide.
Anal. Calcd for C22H40NiOP2: C, 59.89; H, 9.14. Found: C,
3
3
(m, 9H, PCHMeMe), 1.01 (dd, 3H, JHP = 14.8 Hz, JHH
=
7.1 Hz, PCHMeMe), 1.09 (m, 6H, PCHMeMe), 1.21 (m, 2H,
CH2), 1.52 (m, 1H, PCHMe2), 1.70 (m, 1H, PCHMe2), 1.76 (m,
1H, PCHMe2), 2.42 (m, 1H, PhCHdCHH), 2.68 (m, 1H,
PhCHdCHH), 4.26 (m, 1H, PhCHdCH2), 6.90 (t, 1H, 3JHH
=
7.0 Hz, CarHp), 7.14 (m, 2H, CarHm), 7.26 (d, 2H, 3JHH = 7.68 Hz,
CarHo). 13C{1H} NMR (C6D6, 75 MHz): δ 17.2 (s, PCHMeMe),
18.7 (d, JCP = 2 Hz, PCHMeMe), 18.9 (d, JCP = 8 Hz,
2
2
1
59.24; H, 8.78. H NMR (C6D6, 300 MHz): δ 0.43 (dd, 3H,
2
PCHMeMe), 19.3 (s, PCHMeMe), 19.5 (d, JCP = 8 Hz,
3JHP = 19.0 Hz, 3JHH = 7.1 Hz, PCHMeMe), 0.55 (dd, 3H, 3JHP
= 12.0 Hz, JHH = 7.0 Hz, PCHMeMe), 0.88 (m, 3H, PCH-
2
PCHMeMe), 19.6 (d, JCP = 6 Hz, PCHMeMe), 19.8 (d,
3
1
2JCP = 8 Hz, PCHMeMe), 21.0 (d, JCP = 19 Hz, CH2), 21.6
MeMe), 1.01 (m, 9H, PCHMeMe), 1.21 (dd, 3H, 3JHP = 7.0 Hz,
3JHH = 3.7 Hz, PCHMeMe), 1.26 (dd, 3H, JHP = 7.0 Hz,
(d, 1JCP = 19 Hz, CH2), 23.8 (dd, 1JCP = 14 Hz, 3JCP = 4 Hz,
PCHMe2), 24.7 (d, JCP = 14 Hz, PCHMe2), 24.8 (d, JCP =
3
3JHH =3.6Hz, PCHMeMe), 1.41 (m, 1H, PCHMe2), 1.58(m, 1H,
PCHMe2), 1.76 (m, 2H, PCHMe2), 2.12 (dd, 3H, 4JHP = 8.6, 2.7
Hz, PhCOCH3), 7.08 (t, 1H, 3JHH = 7.0 Hz, CarHp), 7.25 (t, 2H,
1
1
1
14 Hz, PCHMe2), 25.4 (dd, JCP = 15 Hz, JCP = 4 Hz,
3
2
PCHMe2), 31.0 (d, JCP = 21 Hz, PhCHdCH2), 50.2 (d,
3
3JHH = 7.6 Hz, CarHm), 7.88 (d, 2H, JHH = 7.6 Hz, CarHo).
2JCP = 18 Hz, PhCHdCH2), 120.4 (s, CarHp), 123.9 (s, CarHo),
149.5 (d, 4JCP = 5 Hz, Car). 31P{1H} NMR (C6D6, 162 MHz): δ
63.1 (d, 2JPP = 64 Hz), 76.4 (d, 2JPP = 64 Hz).
13C{1H} NMR (C6D6, 75 MHz): δ 15.5 (d, 1JCP = 20 Hz,CH2),
17.4 (s, PCHMeMe), 18.2 (d, 2JCP = 6 Hz, PCHMeMe), 18.9 (s,

Article
X-ray Crystal Structure Analyses of 4b, 4c, 50, and 6. Crystals
Organometallics, Vol. 28, No. 22, 2009 6523
(benzaldehyde, acetophenone, styrene, or N-benzylidenbenzy-
lamine) was then added in successive additions at room tem-
perature using a 10 μL syringe (up to 40 equiv). 31P{1H} NMR
spectra were recorded, ensuring that the mixture had reached
equilibrium at the NMR probe temperature (25 °C). When the
reaction was observed to proceed slowly, the NMR tube was
stored in a Schlenk tube under nitrogen and kept for g24 h in a
thermostatic water bath at 25 °C. Equilibrium constants were
determined for each reagent addition by integrating the reso-
nances corresponding to each Ni(0) complex. Final values are
averages of at least five independent measurements.
of these compounds, obtained from hexane, were coated with
dry perfluoropolyether and fixed to the goniometer head in a
cold nitrogen stream (100(2) K). Reflections were collected from
a Bruker-Nonius X8Apex-II CCD diffractometer. Data reduc-
tion was made with SAINT and corrected for Lorentz-polariza-
tion effects and absorption with the multiscan method applied
by SADABS.30,31 The structures were solved by direct methods
(SIR-2002)32 and refined against all F2 data by full-matrix least-
squares techniques (SHELXL97).33
Crystal Data for 4b:. C21H38OP2Ni, Mw = 427.16; orange
prism, 0.17 ꢀ 0.16 ꢀ 0.09 mm; monoclinic, space group P21/c (no.
14), a = 15.5249(3) A, b = 15.0436(4) A, c = 20.5056(5) A, β =
Kinetic Measurements of β-Hydrogen Eliminations. In the case
of the amido complex 1d, a 0.145 M solution of 1d in THF was
generated in situ as previously reported, and LiF was separated by
centrifugation.10b Aliquots of 0.6 mL of THF were transferred to
NMR tubes charged with an external standard (a sealed capillary
tube containing a solution of PPh3 in C6D6). These were placed into
the NMR probe, previously stabilized at the selected temperature
(297, 302, 305, or 310 K), and the β-hydrogen elimination was
monitored by 31P{1H} NMR until completion of the process.
108.9560(10)°, V = 4529.37(19) A3, Z = 8, Fcalcd = 1.253 g cm-3
,
.
λ(Mo KR1) = 0.71073 A, F(000) = 1840, μ = 1.005 mm-1
Collected reflections, 43 071; θ range, 5.74° < 2θ < 61.10°;
independent reflections [R(int) = 0.0412], 13 525. R1 = 0.0374
[I > 2σ(I )], and wR2 = 0.0903 for all data. Goodness-of-fit on F2,
1.039, number of parameters, 451.
Crystal Data for 4c:. C22H40OP2Ni, Mw = 441.19; yellow
prism, 0.20 ꢀ 0.12 ꢀ 0.11 mm; monoclinic, space group P21/c
(no. 14), a = 7.83830(10) A, b = 19.1152(4) A, c = 15.9885(3) A,
β = 101.7950(10)°, V = 2344.99(7) A3, Z = 4, Fcalcd = 1.250 g
Observed first-order rate constants (s-1): 1.1 ꢀ 10-4; 5.2 ꢀ 10-4
;
1.4 ꢀ 10-3; 3.5 ꢀ 10-3, respectively. For the decomposition of the
hydrolytically unsensitive compound 5, the solutions were prepared
from weighed samples (20 mg) of the crystalline complex in 0.6 mL
of THF. Observed rate constants (s-1): 297 K, 8.0 ꢀ 10-6; 305 K,
3.0 ꢀ 10-5; 320 K, 1.3 ꢀ 10-4; 335 K, 8.2 ꢀ 10-4. In the case of the
alkoxides 2b and 2c, kinetics were performed by both of these
methods at 333 K, using THF or C6D6 as a solvent.
cm-3, λ(Mo KR1) = 0.71073 A, F(000) = 952, μ = 0.972 mm-1
.
Collected reflections, 19 498; θ range, 6.64° < 2θ < 61.10°;
independent reflections [R(int) = 0.0276], 7085. R1 = 0.0270
[I > 2σ(I)], and wR2 = 0.0715 for all data. Goodness-of-fit on F2,
1.045, number of parameters, 235.
Crystal Data for 50:. C23H44NiP2, Mw = 441.23; yellow prism,
0.44 ꢀ 0.22 ꢀ 0.19 mm; orthorhombic, space group Pna21
(no. 33), a = 17.5114(13) A, b = 9.5310(6) A, c = 14.5532(8) A,
V = 2428.9(3) A3, Z = 4, Fcalcd = 1.207 g cm-3, λ(Mo KR1) =
0.71073 A, F(000) = 960, μ = 0.936 mm-1. Collected reflections,
21 250; θ range, 4.66° < 2θ < 61.12°, independent reflections
[R(int) = 0.0464], 6399. Goodness-of-fit on F2, 1.063, number of
parameters, 235.
Kinetic Experiments on the Decomposition of 2b. In these
experiments, a semiquantitative approach was used, since the
specific value of the rate constant proved variable. To this end,
solutions of compound 2b in THF were prepared at the selected
concentration. This solution was divided into three aliquots,
and each of them placed in a sealed NMR tube. These were then
placed into a thermostatic bath at 60 °C. Each sample was
removed at a different time and its 31P NMR spectrum recorded.
In order to analyze the rate dependency on the concentration,
three solutions (0.013, 0.026, and 0.108 M) were prepared and
simultaneously subjected to this procedure. To study the influ-
ence of added TEMPO or [Ni(μ-H)(dippe)]2, the solution was
previously divided in two parts; the additive was introduced in
one of them, and the other part was used as a control experi-
ment. Aliquots were taken from each of these solutions, trans-
ferred to NMR tubes (3 each), and subjected to the same
protocol.
Crystal Data for 6:. C22H40NiP2, Mw = 425.19; yellow block,
0.47 ꢀ 0.31 ꢀ 0.30 mm; triclinic, space group P1 (no. 2), a =
8.6934(6) A, b = 9.9152(7) A, c = 14.3854(9) A, R = 84.411(2)°,
β= 84.620(2)°,γ= 68.354(2)°,V= 1144.81(13) A3,Z=2,Fcalcd
=
1.233 g cm-3, λ(Mo KR1) = 0.71073 A, F(000) = 460, μ =
0.990 mm-1. Collected reflections, 18 002; θ range, 4.42° < 2θ <
61.04°, independent reflections [R(int) = 0.0223], 6730. Goodness-
of-fit on F2, 1.089, number of parameters, 226.
Exchange Reactions of Ni(0) Complexes with Organic Com-
pounds. A ca. 20 mg sample of the suitable Ni(0) complex (4b,
22 mg; 4c, 18 mg; or 6, 16 mg) was dissolved in 0.6 mL of
deuterated benzene in a NMR tube. The organic compound
Acknowledgment. Financial support from the DGI
(Project CTQ2006-05527/BQU) and Junta de Andalucıa
´
(Project P06-FQM-01704) is gratefully acknowledged.
I.M. thanks a research fellowship from the Consejo
Superior de Investigaciones Cientıficas (I3P program).
´
(30) Bruker. Apex 2, version 2.1; Bruker AXS Inc.: Madison, WI, 2004.
(31) Bruker. SAINT and SADABS; Bruker AXS Inc.: Madison, WI,
2001.
(32) Burla, M. C; Camalli, M.; Carrozzini, B.; Cascarano, G. L.;
Giacovazzo, C.; Polidori, G.; Spagna, R. J. Appl. Crystallogr. 2003, 36,
1103.
Supporting Information Available: X-ray crystallographic file
in CIF format is available free of charge via the Internet at
http://pubs.acs.org.
€
(33) Sheldrick, G. M. SHELXL97; University of Gotingen: Germany,
1997.