Kinetics and mechanism of molybdenum-mediated acrylate formation from carbon dioxide and ethylene

Article abstract of DOI:10.1021/om100891m

A zerovalent molybdenum complex, [(Ph₂PCH₂CH ₂)₂PPh]Mo(C₂H₄)(N₂) ₂, was found to promote coupling of CO₂ and ethylene to afford a molybdenum(II) acrylate hy

Full text of DOI:10.1021/om100891m

520 Organometallics 2011, 30, 520–527
DOI: 10.1021/om100891m
Kinetics and Mechanism of Molybdenum-Mediated Acrylate Formation
from Carbon Dioxide and Ethylene
Wesley H. Bernskoetter* and Brian T. Tyler
Department of Chemistry, Brown University, Providence, Rhode Island 02912,
United States
Received September 15, 2010
A zerovalent molybdenum complex, [(Ph2PCH2CH2)2PPh]Mo(C2H4)(N2)2, was found to promote
coupling of CO2 and ethylene to afford a molybdenum(II) acrylate hydride complex. The identity of
the molybdenum(II) acrylate hydride complex was established by spectroscopy and reactivity stud-
ies. The mechanism of carbon dioxide functionalization with ethylene has been investigated by a
series of kinetic and isotopic labeling studies, in addition to the observation of a formally moly-
bdenum(0) carbon dioxide-ethylene intermediate along the reaction pathway. Acrylate formation
from the molybdenum(0) intermediate proceeds with a rate constant of 3.8(3) Â 10-5 s-1 and an
isotope effect of 1.2(2) for C2H4 vs C2D4 at 23 °C. Measuring rate constants of the CO2 reduction over
a 40 °C temperature range established activation parameters for acrylate formation of ΔS‡ = 1(7) eu
and ΔH‡ = 24(3) kcal/mol. The mechanism of CO2-ethylene coupling is proposed to proceed from a
molybdenum(0) carbon dioxide-ethylene adduct via rate-limiting oxidative C-C bond formation
followed by rapid β-hydride elimination from a metallalactone complex.
Introduction
Coupling carbon dioxide to olefins is another highly at-
tractive target for CO2 utilization, but one that has achieved
limited success thus far.6 Since Hoberg and Schaefers’s first
report of nickelalactone formation from CO2 and ethylene in
1983 (eq 1),7
The steady increase in demand for finite petroleum re-
sources,1 a primary carbon feedstock, has spurred consider-
able efforts to leverage renewable carbon sources as precur-
sors for commodity chemicals.2 Carbon dioxide in particular
has been touted as a promising target on the basis of its
incredible abundance, cheap availability, and relative ease of
transport.3 Unfortunately, the thermodynamic stability of
CO2 has led to its reticence in performing synthetically useful
transformations. One technique that has achieved success
is coupling CO2 with other small molecules to provide the
necessary reduction energy and functionalization agent from
one source.4 Among the leading examples of this methodol-
ogy is the production of urea from CO2/ammonia and carbo-
nates (and polycarbonates) from CO2/epoxides.5
transition metal chemists have pursued the production of
acrylic acid and other acrylates by coupling these unsaturates
for uses in superabsorbing polymers, coatings, and adhesives.8
Unfortunately Hoberg’s complex, as well as most titanium,9
zirconium,10 vanadium,11 palladium,12 rhodium,13 and other
nickel14 examples that followed it, proved resistant toward
β-hydride elimination reactions to form acrylate compounds.
*To whom correspondence should be addressed. E-mail: wb36@
brown.edu.
(1) United States Department of Enery. Energy Information Admin-
istration, International Energy Annual 2010. July 2010; EIA-0484.
(2) (a) Chemicals and Materials from Renewable Resources; Bozell, J.,
Ed.; ACS Symposium Series 784; American Chemical Society: Washington,
DC, 2001. (b) Patil, Y.; Tambade, P. J.; Jagtap, S. R.; Bhanage, B. M. Front.
Chem. Eng. China 2010, 4, 213.
(3) (a) Aresta, M.; Dibenedetto, A. Catal. Today 2004, 98, 455.
(b) Aresta, M. Carbon Dioxide Reduction and Uses as a Chemical Feedstock.
In Activation of Small Molecules: Organometallic and Bioinorganic
Perspectives; Tolman, W. B., Ed.; Wiley-VCH: Weinheim, 2006; pp 1-35.
(c) Aresta, M.; Dibenedetto, A. Dalton Trans. 2007, 2975.
(4) (a) Braunstein, P.; Matt, D.; Nobel, D. Chem. Rev. 1988, 88, 747.
(b) Shi, M.; Shen, Y.-M. Curr. Org. Chem. 2003, 7, 737.
(7) Hoberg, H.; Schaefer, D. J. Organomet. Chem. 1983, 251, C51.
(8) ICIS Market Research Report , Acrylic Acid Uses and Market
Data; Reported February 2010; http://www.icis.com/v2/chemicals/9074864/
acrylic-acid.html (accessed July 22, 2010).
(9) Cohen, S. A.; Bercaw, J. E. Organometallics 1985, 4, 1006.
(10) Alt, H. G.; Denner, C. E. J. Organomet. Chem. 1990, 390, 53.
(11) Hessen, B.; Meetsma, A.; Bolhuis, F.; Teuben, J.; Helgesson, G.;
Janger, S. Organometallics 1990, 9, 1925.
(12) Langer, J.; Fischer, R.; Goerls, H.; Walther, D. Eur. J. Inorg.
Chem. 2007, 2257.
(5) Sakakura, T.; Choi, J.-C.; Yasuda, H. Chem. Rev. 2007, 107, 2365.
(6) (a) Hoberg, H.; Jenni, K.; Kruger, C.; Raabe, E. Angew. Chem.,
Int. Ed. Engl. 1986, 25, 810. (b) Hoberg, H.; Jenni, K.; Angermund, K.;
Kruger, C. Angew. Chem., Int. Ed. Engl. 1987, 26, 153. (c) Behr, A. Angew.
Chem., Int. Ed. Engl. 1988, 27, 661. (d) Leitner, W. Coord. Chem. Rev.
1996, 153, 257.
(13) Aresta, M.; Quaranta, E. J. Organomet. Chem. 1993, 463, 215.
(14) (a) Hoberg, H.; Peres, Y.; Kruger, C.; Tsay, Y. H. Angew. Chem.,
Int. Ed. Engl. 1987, 26, 771, and references therein. (b) Hoberg, H.;
Ballesteros, A. J. Organomet. Chem. 1991, 411, C11. (c) Hoberg, H.;
Ballesteros, A.; Sigan, A.; Jegat, C.; Barhausen, D.; Milchereit, A.
J. Organomet. Chem. 1991, 407, C23.
r
pubs.acs.org/Organometallics
Published on Web 01/11/2011
2011 American Chemical Society

Article
Organometallics, Vol. 30, No. 3, 2011 521
Figure 1. Acrylate formations from nickelalactone complexes.
More recently Walther15 and Rieger16 have independently
surmounted this barrier to induce modest yields of acrylates
from nickelalactones by either ancillary ligand substitution
and chelate activation following CO2-ethylene coupling
or addition of strong electrophiles (Figure 1). The most
striking progress toward acrylate production from CO2 has
been reported by Carmona and co-workers,17a in which
trans-(PMe3)4Mo(C2H4)2 reduces carbon dioxide at mild
temperature and pressure to afford
for acrylate formation from CO2 and ethylene and experi-
mentally examine the kinetics and mechanism of this trans-
formation. Herein, we describe the preparation of a new
tridentate phosphine molybdenum(0) complex that functio-
nalizes carbon dioxide with ethylene to afford metal-bound
acrylates. The mechanism of acrylate formation was ex-
plored by kinetic and isotopic labeling studies and is sup-
ported by the direct observation of a key reaction interme-
diate.
Results and Discussion
CO2-Ethylene Coupling. Tetrakis(phosphine) molybde-
num and tungsten complexes have afforded some of the most
promising results for producing acrylates from CO2 and ethy-
lene; however these functionalizations occur with concomi-
tant loss of two ancillary ligands.17 This ligand extrusion
process may be required to activate the molybdenum center
for CO2 binding, limit the stability of the complex, or permit
dimerization of the acrylate hydride complex. To assess the
potential effects of ligand loss on CO2-ethylene coupling
and provide a thermally robust support, a commercially
available chelating ligand, (Ph2PCH2CH2)2PPh (Triphos),
was employed to stabilize low-valent molybdenum species
for CO2 reduction. Although, previous findings have not
augured well for the incorporation of chelating ligands on
molybdenum in CO2-ethylene coupling reactions,20 with
judicious selection of the other spectator ligands, there seems
little reason why suitable (Triphos)Mo complexes for acry-
late formation should not be feasible. To this end, our
laboratory began investigations into the preparation of low
oxidation state (Triphos)Mo species bearing ethylene and
weakly bound spectator ligands.
[(PMe3)2(C2H4)MoH(CO2CHdCH2)]2 (eq 2). This remark-
able transformation yields two acrylate ligands bridging
between two metal centers, although eventual acrylate dis-
placement required use of harsh bases. Following this initial
report, Carmona17b,c and others17d have described a small
family of closely related molybdenum and tungsten com-
plexes found to exhibit similar reactivity; however the me-
chanism of acrylate formation was not investigated.
The potential for producing valuable commodity chemi-
cals from carbon dioxide has sustained interest in the me-
chanism of CO2-ethylene coupling at transition metals since
Hoberg and Carmona’s early discoveries. Numerous com-
putational studies have explored the mechanism of metalla-
lactone formation at late transition metals18 (nickel in partic-
ular) and acrylate production at molybdenum complexes.19
While many similarities between these two reactions have
been predicted, consensus on some mechanistic features has
remained elusive (vide infra). Indeed, the mechanism for
CO2-ethylene coupling may well differ from early to late
transition metals. However, the practical application of this
mechanistic information toward improved complexes for
CO2 activation requires detailed understanding of the reac-
tion intermediates and rate-influencing event(s). Given this,
our laboratory has begun an effort to develop new platforms
Synthetic efforts found alkali metal reduction of (Trip-
hos)MoCl321 in the presence of excess dinitrogen and ethy-
lene afforded a new zerovalent complex, trans-(Triphos)Mo-
(N2)2(C2H4) (1), in good yield (eq 3). This bright yellow
complex was
(15) Fischer, R.; Langer, J.; Malassa, G.; Walther, D.; Gorls, H.;
Vaughan, G. Chem. Commun. 2006, 2510.
(16) Bruckmeier, C.; Lehenmeier, M. W.; Reichardt, R.; Vagin, S.;
Rieger, B. Organometallics 2010, 29, 2199.
(17) (a) Alvarez, R.; Carmona, E.; Cole-Hamilton, D. J.; Galindo, A.;
Gutierrez-Puebla, E.; Monge, A.; Poveda, M. L.; Ruiz, C. J. Am. Chem.
Soc. 1985, 107, 5529. (b) Alvarez, R.; Carmona, E.; Galindo, A.; Gutierrez,
E.; Marin, J. M.; Monge, A.; Poveda, M. L.; Ruiz, C.; Savariault, J. M.
Organometallics 1989, 8, 2430. (c) Galindo, A.; Pastor, A.; Perez, P.;
Carmona, E. Organometallics 1993, 12, 4443. (d) Collazo, C.; del Mar
Conejo, M.; Pastor, A.; Galindo, A. Inorg. Chim. Acta 1998, 272, 125.
(18) (a) Dedieu, A.; Ingold, F. Angew. Chem., Int. Ed. Engl. 1989, 28,
1694. (b) Sakaki, S.; Mine, K.; Taguchi, D.; Arai, T. Bull. Chem. Soc. Jpn.
1993, 66, 3289. (c) Sakaki, S.; Mine, K.; Hamada, T.; Arai, T. Bull. Chem.
Soc. Jpn. 1995, 68, 1873. (d) Papai, I.; Schubert, G.; Mayer, I.; Besenyei, G.;
Aresta, M. Organometallic 2004, 23, 5252. (e) Graham, D. C.; Mitchell, C.;
Bruce, M. I.; Metha, G. F.; Bowie, J. H.; Buntine, M. A. Organometallics
2007, 26, 6784.
characterized by infrared and multinuclear NMR spectros-
copy and reactivity studies. The 31P{1H} NMR spectrum of
the molybdenum(0) species in benzene-d6 exhibits doublet and
triplet resonances (2JP-P = 16.8 Hz) at 75.0 and 92.7 ppm,
1
respectively. H NMR spectroscopy reveals a broad signal
(20) Incorporation of a bidentate phosphine into Carmona’s molyb-
denum complexes did not produce clean acrylate formation, although
coupling was successful with bidentate tungsten species (ref 17c).
(21) George, A. T.; Kovar, R. A. Inorg. Chem. 1981, 20, 285.
(19) Schubert, G.; Papai, I. J. Am. Chem. Soc. 2003, 125, 14847.

522 Organometallics, Vol. 30, No. 3, 2011
Bernskoetter and Tyler
for the bound olefin at 1.44 ppm, and the assignment was
confirmed by preparation of the trans-(Triphos)Mo(N2)2-
decomposition prior to an observable change in isomer ratio.
Although the 2JP-P couplings were not well resolved for all
31P NMR signals, a maximum coupling constant of approxi-
mately 30 Hz determined from the peak widths discounts a
rigorously trans orientation of any two phosphine ligands.
The assignment of the isomers of 2 as dimolybdenum acry-
late hydride complexes was further supported by the obser-
1
(C2D4) (1-d4) isotopologue. H-13C HSQC NMR experi-
ments indicate a correlation between this peak and a res-
onance at 38.76 ppm in the 13C NMR spectrum. Solid-state
infrared analysis of 1 confirmed the presence of the two
dinitrogen ligands with bands at 2051 and 1981 cm-1. Addi-
tion of triphenylphosphine to 1 produced free ethylene and
trans-(Triphos)Mo(N2)2PPh3,22 further supporting the char-
acterization of the bis(dinitrogen) molybdenum ethylene
complex.
1
vation of two molybdenum-hydride resonances in the H
NMR spectrum at -6.73 and -7.15 ppm. Each metal-
hydride is coupled to three inequivalent 31P nuclei, resulting
2
in an eight-line splitting pattern, with JP-H constants ran-
Numerous molybdenum(0) and tungsten(0) bis(dinitrogen)
complexes have been prepared since Chatt’s early work to
utilize this motif in the reduction of N2 to ammonia.23 Typ-
ically these complexes are stabilized by four σ-donating ligands
such as phosphines or amines,24 although a few examples of
related bis(dinitrogen) compounds bearing additional π-acid
ligands, as in 1, have been reported using olefins,25 CO,26 and
additional N2 ligands.27 Complex 1 presented a particularly
attractive platform for studying the coupling of ethylene and
carbon dioxide at zerovalent molybdenum, as the chelating
Triphos ligand would minimize complications arising from
ancillary ligand loss and facile liberation of the dinitrogen
ligands should allow ready introduction of carbon dioxide
into the metal coordination sphere. Gratifyingly, the addi-
tion of a small excess of carbon dioxide (4 equiv) to 1 af-
forded a bridging acrylate hydride complex, [(Triphos)Mo-
(H)(CO2CHdCH2)]2 (2), over several hours at ambient
temperature (eq 4). The 31P{1H} NMR spectrum of 2 dis-
plays six signals, three
ging from 13.2 to 96.0 Hz.28
Examination of the infrared spectrum for the isomers of
complex 2 revealed a broad band at 1512 cm-1, consistent
with a bidentate carboxylate ligand.30 Employing 13CO2 in
the coupling reaction confirmed the origin of this IR band.31
Incorporation of 13CO2 also resulted in two enhanced reso-
nances at 189.9 and 192.3 ppm in the 13C{1H} NMR spec-
trum. Additionally, the absence of 1H NMR signals between
4.5 and 6.5 ppm suggests the acrylate olefins are bound to
molybdenum. With the data in hand, distinction between the
two coordination geometries, in which the acrylate ligands
bridge the two metal centers through the carboxylate frag-
ment or through one oxygen atom and the olefin, could not
be made. Complex 2 is therefore depicted in the bridging
carboxylate geometry on the basis of Carmona’s crystal-
lographically characterized complex.17b
The inability to obtain X-ray quality crystals of either
isomer of 2 has, unfortunately, limited more detailed analysis
of the coordination geometry about the molybdenum cen-
ters. Given this point of structural ambiguity, a reactivity
study was used to confirm the identity of 2 as a bridging
acrylate molybdenum hydride complex. Treatment of 2 with
triphenylphosphine resulted in immediate formation of the
red monomeric acrylate hydride species (Triphos)Mo(H)-
(PPh3)(CO2CHdCH2) (3) (Figure 2), which was unambigu-
ously characterized by multinuclear NMR spectroscopy and
elemental analysis. Benzene-d6 solutions of 3 exhibit 1H
NMR resonances at 4.68, 5.15, and 5.48 ppm assigned to
an unbound acrylic olefin.28 These signals display coupling
constants, splitting patterns, and chemical shift values quite
similar to those observed for free acrylic acid. Additionally, a
triplet of doublets of doublets is observed at -4.71 ppm,
establishing the presence of a molybdenum-hydride ligand.28
The 31P{1H} NMR spectrum of 3 displays a doublet of
doublets at 109.7 ppm for the two PPh2 fragments and two
doublet of triplet resonances at 47.7 and 99.4 ppm assigned
to the PPh and PPh3 groups, respectively. Solid-state infra-
red analysis revealed a band at 1519 cm-1, again consistent
with a bidentate acrylate ligand.30
appearing as a well-resolved doublet of doublets and three
as unresolved multiplets. The signals at 100.3, 92.8, and
81.0 ppm are slightly more intense (an approximate ratio of
1:1.3) than those at 96.0, 90.3, and 80.2 ppm.28 These data are
consistent with the formation of two similar energy diaster-
eomers, as may be expected for a dimeric complex formed
from two seven-coordinate metal centers.29 Attempts to in-
terconvert the two isomers by thermolysis at 85 °C resulted in
(22) George, A. T.; Hammund, H. H. J. Organomet. Chem. 1995, 503,
C1.
(23) Chatt, J.; Pearman, A. J.; Richards., R. L. Dalton. Trans. 1977,
1852.
(24) (a) Chatt, J.; Dilworth, J. R.; Richards, R. L. Chem. Rev. 1978,
78, 589. (b) Carmona, E.; Marin, J. M.; Poveda, M. L.; Atwood, J. L.; Rogers,
R. D. J. Am. Chem. Soc. 1983, 105, 3014. (c) Dadkhah, H.; Dilworth, J. R.;
Faiman, K.; Kan, C. T.; Richards, R. L.; Hughes, D. L. Dalton Trans. 1986,
245. (d) George, T. A.; De Bord, J. R. D.; Kaul, B. B.; Pickett, C. J.; Rose,
D. J. Inorg. Chem. 1992, 31, 1295. (e) Jimenez-Tenorio, M.; Puerta, M. C.;
Valerga, P.; Hughes, D. L. Dalton Trans. 1994, 2431. (f) George, T. A.; Rose,
D. J.; Chang, Y.; Chen, Q.; Zubieta, J. Inorg. Chem. 1995, 34, 1295.
(g) Hidai, M. Coord. Chem. Rev. 1999, 185, 99.
(25) Jackson, S. A.; Hodges, P. M.; Poliakoff, M.; Turner, J. J.;
Greveld, F. W. J. Am. Chem. Soc. 1990, 112, 1221.
(26) Goff, S. E. J.; Nolan, T. F.; George, M. W.; Poliakoff, M.
Organometallics 1998, 17, 2730.
(27) Anderson, S. N.; Richards, R. L.; Hughes, D. L. Chem. Commun.
Further confirmation of the identity of complex 3 was
obtained by independent preparation from trans-(Triphos)-
Mo(N2)2(PPh3) (Figure 2).22 Treating benzene-d6 solutions
of the previously reported bis(dinitrogen) molybdenum(0)
triphenylphosphine complex with equimolar excesses (6 equiv
of carbon dioxide and ethylene at ambient temperature over-
night afforded 1H and 31P{1H} NMR spectra identical to those
from PPh3 addition to 2. Most likely triphenylphosphine disso-
ciation from trans-(Triphos)Mo(N2)2(PPh3) provides access to
1982, 1291.
(28) See Supporting Information.
(29) For a representative paper addressing the stereochemistry of
seven-coordinate metals see: Fay, R. C. Coord. Chem. Rev. 1996, 154, 99.
(30) Nakamoto, K. Infrared and Raman Spectra of Inorganic and
Coordination Compounds, 4th ed.; Wiley: New York, 1986; pp 231-233.
(31) The red-shifted band for 13CO2-labeled 2 was obscured by bands
from the Triphos ligand near 1480 and 1430 cm-1
.

Article
Organometallics, Vol. 30, No. 3, 2011 523
Figure 2. Formation of 3 from CO2-ethylene coupling and PPh3 addition to 2.
Figure 3. Observation of intermediate 4 prior to formation of 2.
a reactive (Triphos)Mo0 center analogous to reactive inter-
mediates in the reduction of CO2 with complex 1 (vide infra).
However, the qualitatively slower acrylate formation and the
presence of minor amounts of free Ph3P formed during the
reaction render CO2-ethylene coupling from trans-(Triphos)-
Mo(N2)2(PPh3) somewhat less attractive for study than CO2
addition to 1.
converted slowly to the acrylate hydride species 2 over several
hours (vide infra). The molybdenum(0) carbon dioxide ethylene
complex exhibited a doublet resonance at 64.7 ppm and a triplet
resonance at 95.3 ppm (2JP-P = 6.1 Hz) in the 31P{1H} NMR
spectrum, which were assigned to the PPh2 and PPh fragments,
1
respectively. The H NMR spectrum displayed a broad reso-
nance at 0.21 ppm, attributed to the protons on the bound ethy-
lene. This assignment was again confirmed by deuterium label-
ing with the preparation of (Triphos)Mo(C2D4)(CO2) (4-d4).
The broad signal at 0.21 correlates to a single 13CNMRchemical
shift of 41.6 ppm in the 1H-13C HSQC NMR spectrum. Cool-
ing the sample to -20 °C did not decoalese this signal (limited
solubility of complex 4 at lower temperatures prohibited further
variable-temperature study), suggesting either that the inter-
change of the two ethylene CH2 units is a low-energy process or
that the ethylene binds in a symmetrically disposed position
coplanar with the Triphos ligand. The 1Hand13C chemical shifts
for the bound ethylene ligand are similar to those reported for a
range of six-coordinate molybdenum ethylene complexes,17c
with the proton signal in 4 shifted slightly upfield (∼0.5 ppm)
relative to the coordinatively saturated analogues. This small
perturbation in the chemical shift may arise from enhanced
reduction of the CdC bond in the Dewar-Chatt-Duncanson
model32 or the variation in metal coordination number.33
Further spectral analysis of the reaction intermediate was
obtained by preparation of (Triphos)Mo(C2H4)(13CO2)
(4-13C) from addition of excess 13CO2 to 1. Incorporation of
the 13C label resulted in an enhanced doublet of triplets signal
at 193.6 ppm in the 13C NMR spectrum with coupling
constants of 14.9 and 27.7 Hz (Figure 4). The magnitude of
The facile reductive functionalization of carbon dioxide to
acrylates is a relatively rare transformation dominated by a
small family of related zerovalent molybdenum and tungsten
complexes.17 The pathway for the formation of acrylate from
complex 1 likely bears a strong resemblance to the reactivity
of those molybdenum-mediated functionalizations reported
or inspired by Carmona.17 However, no experiments to probe
the mechanism or kinetics of this intriguing molybdenum-
promoted CO2-ethylene coupling have been reported.
Qualitatively, acrylate formation from complex 1 upon CO2
addition appears slower than the previously described tetrakis-
(phosphine) molybdenum bis(ethylene) complexes. The re-
ported synthetic procedures indicate modest yields of those
acrylate complexes may be obtained in a matter of minutes
with low pressures of CO2 for the monodenate phosphine (or
phosphate) molybdenum ethylene complexes,17 while 1 re-
quires many hours under similar conditions. Understanding
the origins of this slower CO2-ethylene coupling requires
elucidation of the reaction mechanism and identification of
the rate-influencing event(s). This motivation, along with
an interest in designing complexes to enhance acrylate prepara-
tion from CO2 and ethylene, prompted a series of experi-
ments to investigate the reaction coordinate for CO2-ethylene
coupling.
Observation of Reaction Intermediate. Monitoring the
addition of approximately 12 equiv of CO2 to a benzene-d6
solution of the bis(dinitrogen) molybdenum(0) ethylene com-
plex in situ by NMR spectroscopy revealed the formation of a
long-lived reaction intermediate characterized as (Triphos)Mo-
(C2H4)(CO2) (4) (Figure 3). Interestingly, this species was also
observed as an intermediate in the formation of the triphenyl-
phosphine acrylate hydride complex 3 from CO2 and ethylene
addition to trans-(Triphos)Mo(N2)2(PPh3). Complex 4 formed
rapidly from 1 in the presence of excess CO2 (∼15 min) and
2
this coupling is similar to the JC-P = 17.5 Hz coupling
34
reported for trans-(PMe3)4Mo(13CO2)2 and significantly
greater than the 3JC-P < 6 Hz couplings reported for more
commonly isolated metallalactone complexes.12,35 The labeling
(32) (a) Dewar, M. J. S. Bull. Soc. Chim. Fr. 1951, 18, C79. (b) Chatt, J.;
Duncaneon, L. A. J. Chem. Soc. 1963, 2339.
(33) For examples of other formally five-coordinate molybdenum(0)
complexes see: ref 24f and Niikura, F.; Seino, H.; Mizobe, Y. Organo-
metallics 2009, 28, 1112.
(34) Alvarez, R.; Carmona, E.; Marin, J. M.; Poveda, M. L.; Gutierrez-
Puebla, E.; Monge, A. J. Am. Chem. Soc. 1986, 108, 2286.

524 Organometallics, Vol. 30, No. 3, 2011
Bernskoetter and Tyler
Figure 4. Partial (a) 31P{1H} NMR and (b) 13C{1H} NMR spectra of 4-13C. Asterisk denotes minor formation of labeled 2.
experiment also alters the 31P{1H} NMR spectrum, with
4-13C splitting the two signals into a doublet of triplets (PPh)
and a doublet of doublets (PPh2), indicating only one 13CO2
molecule is incorporated into the intermediate (Figure 4).
Infrared spectroscopy confirms the presence of a CdO
fragment with a band at 1700 cm-1 for 4 in KBr, which
shifts to 1654 cm-1 for 4-13C, though the IR bands for
transition metal carbon dioxide complexes and metallalac-
tones do not have characteristically distinguishing fre-
quencies.7,9-14,34 Notably, no strong bands were observed
for either isotopologue of 4from 2200 to 1800 cm-1, suggesting
both dinitrogen ligands of 1 dissociate when binding carbon
dioxide. The observation of a pseudo-five-coordinate moly-
bdenum(0) complex was unexpected, given that 4 may be
observed in the presence of additional carbon dioxide, ethy-
lene, dinitrogen, and triphenylphoshine. However no spec-
troscopy signals indicative of a bound sixth ligand were
observed, and the use of relatively noncoordinating solvent
(benzene) tentatively supports assignment of the coordia-
tively unsaturated intermediate.33
these possibilities, as this technique reduces complications
from gases mixing into solution. To assess the influence of
CO2 on the formation of 4, two degassed samples of 1 in
benzene-d6 were treated with 4 and 12 equiv of CO2 by
calibrated gas bulb at -196 °C. After warming to ambient
temperature for 10 min, a 36% greater conversion to 4
(average of 3 trials) was observed by 31P{1H} NMR spec-
troscopy for the samples containing additional carbon diox-
ide. Likewise, parallel tube experiments with added dinitrogen
in addition to carbon dioxide exhibit lower conversions to 4
in the presence of N2.37 More significantly, upon formation
of complex 4, removal of excess CO2 and introduction of a
dinitrogen atmosphere at ambient temperature did not result
in reversion to 1 prior to formation of the acrylate hydride
species 2. These experiments are consistent with formation of
intermediate 4 by reversible dinitrogen loss from 1 followed
by slow, effectively irreversible, binding of carbon dioxide
with loss of a second N2 ligand (Figure 5).
Mechanistic Considerations of Acrylate Formation. The
observation of 4 en route to the bridging acrylate molybde-
num hydride complex experimentally verifies that simulta-
neous coordination of both ethylene and carbon dioxide is a
prerequisite for coupling at early transition centers. The re-
quirement of substrate precoordination for transition metal
mediated metallalactone or acrylate formation is an area that
has been investigated computationally for many years with
inconsistent conclusions.18,19 Early calculations by Dedieu18a
and others18d,e concluded that nickelalactone formation
proceeds without coordination of ethylene, instead occurring
The mechanism of carbon dioxide substitution with 1 was
briefly explored using a series of parallel NMR tube experi-
ments. The coordinative saturation of complex 1 implies that
the substitution process to form 4 is dissociative in nature;36
however, the concentration of CO2 or N2 may influence the
reaction rate if ligand dissociation is reversible. Parallel
experiments with stock solutions of 1 were used to investigate
(35) Sano, K.; Yamamoto, T.; Yamamoto, A. Chem. Lett. 1983, 115.
(36) Collman, J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R. G.
Ligand Substiution Processes. Principles and Applications of Organo-
transition Metal Chemistry; University Science Books: Sausalito, 1987; pp
235-278.
(37) The reduced partial pressure of CO2 produced by N2 addition
may also influence the rate of conversion to 4, obscuring firm conclu-
sions regarding inhibition by dinitrogen.

Article
Organometallics, Vol. 30, No. 3, 2011 525
Figure 5. Proposed mechanism of CO2 substitution with 1.
Figure 6. Proposed mechanism for the formation of acrylate.
via nucleophilic attack by incoming C2H4 on a bound
carbon dioxide ligand. Later, Sakaki18b,c and co-workers
supported CO2-ethylene coupling at similar nickel(0) com-
plexes with prerequisite coordination of both substrates.
ꢀ
More recently, Schubert and Papai have investigated the
coupling of carbon dioxide to (PMe3)4Mo(C2H4)2 with DFT
calculations and supported the argument that the C-C
bond forming event occurs with a metal-centered transition
structure.19 The characterization of the carbon dioxide
ethylene complex 4 provides the first direct evidence that
precoordination of the two unsaturates lies on the path to
C-C bond formation between CO2 and ethylene, and while
its implications may be less strong for nickel, the require-
ment is likely general for early transition metals centers
investigated to date.
Figure 7. Isotope effect for acrylate formation for C2H4 vs
C2D4. Asterisks denote deuterium sites.
mining the rate-limiting step for rationally optimizing metal
complexes for CO2-ethylene coupling, we proceeded to
further probe the kinetics of acrylate formation.
The preparation of intermediate complex 4 also serves as
an instrument to further study the kinetics and mechanism of
molybdenum-mediated acrylate formation from CO2 and
ethylene. Due to the acceleration in the formation of 4 with
added CO2, solutions of (Triphos)Mo(C2H4)(CO2) could be
generated cleanlypriortosubstantialformation of thebridging
acrylate hydride species by treatment of 1 with 12 equiv of
CO2.38 Monitoring the decay of the signal for complex 4
versus an internal standard in the 1P{1H} NMR spectrum at
23 °C afforded an observed rate constant of 3.8(3) Â 10-5 s-1
for acrylate formation. This corresponds to an activation
energy of 23.5(2) kcal/mol. As expected, control experiments
varying the equivalents of added CO2 did not alter the con-
version rate of 4 to 2. Previous computational studies of
acrylate formation from (PMe3)4Mo(C2H4)2,19 as well as the
experimental observation of numerous metallalactones,7,9-14
suggest the formation of 2 proceeds from complex 4 by
formally oxidative C-C bond formation between bound
CO2 and ethylene,39 followed by β-hydride elimination from
the resulting metallocycle, and finally a dimerization of the
Triphos molybdenum acrylate hydride complex (Figure 6).
Unfortunately, the DFT analysis could not distinguish the
rate-limiting step of acrylate formation from (PMe3)4Mo-
(C2H4)2 and CO2 as the computed barriers of several steps
were quite similar in energy.19 Given the importance of deter-
The modest stability of 4 as an intermediate during carbon
dioxide reduction, along with the absence of any other
detectable intermediates during formation of 2, implicates
two likely kinetic profiles for acrylate production. In one case
complex 4 is favored in a rapid equilibrium with the metal-
lalactone species followed by rate-limiting β-hydride elim-
ination and fast dimerization to afford 2. An alternative
kinetic profile is rate-limiting oxidative C-C bond coupling
from 4 followed by rapid β-hydride elimination and dimer-
ization.40 An isotope effect experiment was employed to dif-
ferentiate between these two reaction profiles by comparing
the rates of acrylate formation from 4 and 4-d4 (Figure 7).
Incorporation of deuterium into the bound ethylene pro-
duced only a minimal decrease in the observed rate constant,
affording an isotope effect of 1.2(2) at 23 °C. This small
isotope effect is inconsistent with a rate-limiting β-hydride
elimination from a metallacylic complex, which is expected
to produce a significant primary isotope effect.41
The activation parameters for acrylate formation from
complex 4 were also examined by measuring the observed
(40) A third possibility, rate-limiting dimerization, is deemed unlikely
given that the absence of observable acrylate hydride monomer necessi-
tates two rapid equilibria between 4-metallalactone and metallalactone-
acylate hydride monomer, both of which must favor the reactants. This
possibility is also excluded on the basis of the measured KIE and
activation entropy.
(41) (a) Agapie, T.; Schofer, S. J.; Labinger, J. A.; Bercaw, J. E. J. Am.
Chem. Soc. 2004, 126, 1304. (b) Agapie, T.; Labinger, J. A.; Bercaw, J. E.
J. Am. Chem. Soc. 2007, 129, 14281.
(38) Less than 15% formation of 2 was observed prior to complete
conversion of 1 to 4.
(39) The C-C bond forming event may also be viewed as a nucleo-
philic attack by ethylene on the electrophilic carbon dioxide.

526 Organometallics, Vol. 30, No. 3, 2011
Bernskoetter and Tyler
of CO2 by 1, although qualitatively slower than the handful
of similar transformations inspired by Carmona over two
decades ago, provided a platform for the first experimental
examination of the mechanism for this intriguing carbon
dioxide functionalization. Monitoring the conversion of 1 to
acrylate hydride complex 2 afforded direct characterization
of an intermediate molybdenum carbon dioxide ethylene
adduct 4, establishing precoordination of both substrates as
a likely requirement for CO2 functionalization by ethylene at
early transition metals. Kinetic analysis of the conversion of
4 to 2 revealed an isotope effect of 1.2(2) for functionaliza-
tion with C2H4 versus C2D4 and activation parameters of
ΔS‡ = 1(7) eu and ΔH‡ = 24(3) kcal/mol, consistent with
rate-limiting oxidative C-C bond formation during acrylate
formation. These findings suggested the electrophilicity and
coordination geometry of the metal center may dominate the
facility of acrylate formation from carbon dioxide and ethy-
lene. Further investigations to leverage this new mechanistic
data toward enhanced CO2-ethylene coupling as well as to
elucidate the influences of other coordination chemistry
factors are currently ongoing in our laboratory. Develop-
ment of improved methods for acrylate formation may lead
to new routes for carbon dioxide functionalization and fur-
ther our efforts to liberate acrylate-containing products from
the metal in a catalytically viable manner.
Figure 8. Eyring plot for the conversion of 4 to 2.
Table 1. Temperature Dependence of Rate Constants for Acrylate
Formationa
k (s-1
)
T (K)
1.3(4) Â 10-3
4.1(4) Â 10-4
2.1(3) Â 10-4
3.8(3) Â 10-4
6.2(3) Â 10-6
322.2
313.9
308.7
298.2
282.0
Experimental Section
a Rate constants measured for 0.030 M benzene-d6 solutions of 4 with
General Considerations. All manipulations were carried out
using standard vacuum, Schlenk, cannula, or glovebox techniques.
Ethylene and carbon dioxide were purchased from Corp Brothers
12 equiv of CO2.
rate constant for the reaction over a 40 °C temperature
range. The effect of temperature on the rate of molybdenum
acrylate hydride formation is depicted in Figure 7, and the
rate constants are reported in Table 1. From the Eyring plot
an activation entropy (ΔS‡) of 1(7) eu and an activation
enthalpy (ΔH‡) of 24(3) kcal/mol were computed. The near
zero ΔS‡ implicates a unimolecular transition structure for
the rate-limiting event, which along with the small isotope
effect for kC2H4/kC2D4 is consistent with oxidative C-C bond
formation as the slow step in acrylate formation. Rate-
limiting C-C bond formation for CO2-ethylene coupling
was unanticipated given that related coupling reactions out-
side group VI transition metals are arrested at metallalactone
structures with β-hydride elimination pathways often unob-
served. Our findings suggest that efforts to engender elec-
tron-rich early transition metal centers may enhance the rate
of acrylate formation. Clearly other features of the metal
coordination environment may also impact the facility of
acrylate production from CO2 and ethylene. These include
the ability to bind carbon dioxide and ethylene in a cis orienta-
tion, the availability of a vacant coordination site to ensure
rapid β-hydride elimination from the metallacycle,42 and the
chelation mode of the ancillary ligand.17c
˚
and stored over 4 A molecular sieves in heavy-walled glass vessels
prior to use. All other chemicals were purchased from Aldrich,
VWR, Strem, or Cambridge Isotope Laboratories. Solvents were
dried and deoxygenated using literature procedures.43
1H, 13C, and 31P NMR spectra were recorded on Bruker DRX
1
400 Avance and 300 Avance MHz spectrometers. H and 13C
chemical shifts are referenced to residual protio solvent signals;
31P chemical shifts are referenced to an external standard of
H3PO4. Probe temperatures were calibrated using ethylene
glycol and methanol as previously described.44 IR spectra were
recorded on Jasco 4100 FTIR and Metler Toledo React IR
spectrometers. Elemental analyses were performed at Robert-
son Microlit Laboratories, Inc., in Madison, NJ.
Preparation of [(Ph2PCH2CH2)2PPh]Mo(C2H4)(N2)2 (1). A
heavy-walled glass reaction vessel was charged with 9.36 g (2.04
mmol) of 0.5% sodium amalgam, 0.150 g (0.204 mmol) of
[(Ph2PCH2CH2)2PPh]MoCl3,21 and approximately 10 mL of
tetrahydrofuran. On a vacuum line approximately 4 equiv of
ethylene and 1 atm of dinitrogen were added at -196 °C. The
resulting reaction mixture was stirred at ambient temperature
for 18 h. The volatiles were removed in vacuo from the green
reaction mixture. The residue was extracted with toluene, filtered
through Celite, concentrated to approximately 5 mL, and recrys-
tallized at -35 °C to afford 0.135 g (93%) of 1 as a yellow
powder. Anal. Calcd for C36H37MoN4P3 0.5(C7H8): C, 62.37;
3
1
H, 5.43; N, 7.37. Found: C, 62.02; H, 5.39; N, 6.82. H NMR
Concluding Remarks
(23 °C, C6D6): δ 1.44 (br s, 4H, C2H4), 1.70 (m, 2H, PCH2), 2.03
(m, 2H, PCH2), 2.62 (m, 2H, PCH2), 2.82-3.03 (m, 2H, PCH2),
6.87-7.04 (m, 12H, C6H5), 7.08 (t, 7.7 Hz, 8H, C6H5), 7.32 (t, 7.7
Hz, 2H, C6H5), 7.46 (t, 8.0 Hz, 1H, C6H5), 7.65 (t, 7.7 Hz, 2H,
C6H5). 13C{1H} NMR (23 °C, C6D6): δ 30.17 (PCH2), 34.21
(PCH2), 38.76 (C2H4), 129.66, 129.72, 131.77, 131.97, 132.13,
A new (Ph2PCH2CH2)2PPh-supported bis(dinitrogen)
molybdenum ethylene complex has been prepared and found
to promote C-C bond coupling with carbon dioxide to
afford a binuclear acrylate hydride species. The reduction
(42) Collman, J. P.; Hegedus, L. S.; Norton, J. R.; Finke, R. G.
Decomposition of Metal Alkyls by β-Hydride Elimination. Principles
and Applications of Organotransition Metal Chemistry; University
Science Books: Sausalito; 1987; pp 386-388.
(43) Pangborn, A. B.; Giardello, M. A.; Grubbs, R. H.; Rosen, R. K.;
Timmers, F. J. Organometallics 1996, 15, 1518.
(44) Sandstrom, J. Dynamic NMR Spectroscopy; Academic Press:
€
New York, 1982.

Article
Organometallics, Vol. 30, No. 3, 2011 527
133.60, 139.05, 140.81(aryl). 31P{1H} NMR (23 °C, C6D6): δ
75.0 (d, 16.8 Hz, 2P, PPh2), 92.7 (t, 16.8 Hz, 1P, PPh). IR (KBr):
ν
(CHdCH2), 131.75 (CHdCH2) 123.48, 127.74, 127.90, 128.53,
128.91 131.53, 131.60, 133.61, 133.87, 134.54, 135.33, 135.44,
137.84, 138.39 (aryl) two quaternary signals not located. 31P-
{1H} NMR (23 °C, C6D6): δ 47.7 (dt, 12.1, 155.5 Hz, 1P, PPh),
99.4 (dt, 20.6, 155.5 Hz, 1P, PPh3), 109.7 (dd, 12.1, 20.6 Hz, 2P,
NtN = 2051, 1981 cm-1 (approximately 1:2 relative intensity).
Preparation of {[(Ph2PCH2CH2)2PPh]Mo(CO2CHdCH2)}2
(2). A heavy-walled glass reaction vessel was charged with 0.108
g (0.152 mmol) of 1 and approximately 5 mL of toluene. On a
vacuum line 12 equiv of carbon dioxide (335 Torr in 101 mL) was
added at -196 °C. The resulting reaction mixture was stirred at
ambient temperature for 25 h. The volatiles were removed in
vacuo from the brown reaction mixture. The residue washed
with pentane and dried to afford 0.096 g (90%) of 2 as a red-brown
solid, which was characterized without further purification.
Anal. Calcd for C37H37MoO2P3: C, 63.25; H, 5.31. Found: C,
PPh2). IR (KBr): νCdO = 1519 cm-1
.
Observation of [(Ph2PCH2CH2)2PPh]Mo(CO2)(C2H4) (4). A
J. Young NMR tube was charged with 0.015 g (0.021 mmol) of 1
and approximately 400 μL of benzene-d6. On a vacuum line,
12 equiv of carbon dioxide was admitted at -196 °C via cali-
brated gas bulb (162 Torr in 28.9 mL). The full conversion of 1 to
4 was monitored by NMR spectroscopy and occurred in ap-
1
proximately 20 min at 23 °C. H NMR (25 °C, C6D6): δ 0.21
1
62.99; H, 5.50. Spectral data for combined isomers: H NMR
(m, 4H, C2H4), 2.21 (m, 2H, PCH2), 2.62 (m, 2H, PCH2),
3.16-3.34 (m, 2H, PCH2), 3.59 (m, 2H, PCH2), 6.91 (m, 2H,
C6H5), 7.07-7.20 (m, 17H, C6H5), 7.25 (m, 4H, C6H5), 7.81 (m,
2H, C6H5). 31P{1H} NMR (23 °C, C6D6): δ 64.7 (d, 6.1 Hz, 2P,
(25 °C, C6D6): δ -7.15 (ddd, 13.2, 54.0, 96.0, Mo-H), -6.73
(ddd, 13.2, 64.9, 94.4 Hz, Mo-H), 1.12-3.20, 3.70, 3.99
(m, PCH2 and CHdCH2), 6.86, 6.95-7.34, 7.40, 7.61, 7.69,
8.15, 8.24 (m, C6H5). 31P{1H} NMR (23 °C, C6D6): δ 80.2 (dd,
14.5, 27.6 Hz), 81.0 (dd, 15.3, 28.5 Hz), 90.3 (m), 92.8 (dd, 16.8,
PPh2), 95.3 (t, 6.1 Hz, 1P, PPh). IR (KBr): νCdO = 1700 cm-1
.
Partial 13C NMR from 1H-13C HSQC: δ 27.72 (PCH2), 33.81
(PCH2), 41.60 (C2H4). Spectral data for 4-13C. 31P{1H} NMR
(23 °C, C6D6): δ 64.7 (dd, 6.1, 14.9, Hz, 2P, PPh2), 95.3 (dt, 6.1,
27.7 Hz, 1P, PPh). 13C{1H} NMR (23 °C, C6D6): δ 193.6 (dt,
28.4 Hz), 96.0 (m), 100.3 (m). IR (KBr): νCdO = 1512 cm-1
.
2-13C2: Partial 13C{1H} NMR (23 °C, C6D6): δ 189.9, 192.3
(CO2CHdCH2).
14.9, 27.7 Hz, CO2). IR (KBr): ν13CdO = 1654 cm-1
.
Preparation of (Triphos)Mo(H)(PPh3)(CO2CHdCH2) (3).
Method A: A heavy-walled glass reaction vessel was charged
with 1.54 g (0.330 mmol) of 0.5% sodium amalgam, 0.075 g
(0.010 mmol) of [(Ph2PCH2CH2)2PPh]MoCl3,21 0.029 g (0.011
mmol) of triphenylphosphine, and approximately 10 mL of tet-
rahydrofuran. On a vacuum line 4 equiv of ethylene (260 Torr in
28.9 mL) and 1.2 equiv of carbon dioxide (78 Torr in 28.9 mL)
were added via calibrated gas bulb at -196 °C. The resulting
reaction mixture was stirred at ambient temperature for 18 h.
The volatiles were removed in vacuo from the red-orange reac-
tion mixture. The residue was washed with pentane to remove
residue PPh3, extracted with toluene, filtered through Celite, con-
centrated to approximately 3 mL, and recrystallized at -35 °C
to afford 0.042 g (43%) of 3 as a microcrystalline red powder.
Method B: A J. Young NMR tube was charged with 0.015 mg of
General Procedure for the Determination of Kinetics of Acry-
late Formation. A J. Young NMR tube was charged with 410 mg
of a benzene-d6 solution of 1 of known concentration (ca. 0.03 M)
and a capillary of triphenylphosphite in benzene-d6 for use as an
integration standard. Then 12 equiv of carbon dioxide was
added via calibrated gas bulb at -196 °C. The reaction was
thawed, shaken, and inserted into a temperature-controlled
NMR probe. The reaction progress was monitored by 31P NMR
spectroscopy over greater than 2 half-lives beginning after all
observable quantities of 1 had converted to 4. The decay of the
resonances for 4 was converted to concentration and fitted to a
first-order plot of ln [4] versus time, which gave observed rate
constants as the slope. Example graphs may be found in the
Supporting Information.
General Procedure for Determination of Carbon Dioxide In-
fluence on Formation of 4. Two J. Young NMR tubes were
charged in parallel with a stock solution of 300 mg of a benzene-
d6 solution of 1 of known concentration (ca. 0.07 M). Then 4 and
12 equiv of carbon dioxide were added via calibrated gas bulb at
-196 °C, respectively. The reactions were thawed and shaken in
parallel for 10 min. The reactions were then frozen at -196 °C
and analyzed in series by 31P NMR spectroscopy.
22
trans-(Triphos)Mo(N2)2PPh3 and approximately 500 μL of
benzene-d6. On a vacuum line, 6 equiv (61 Torr in 28.9 mL) of
both carbon dioxide and ethylene were admitted to the sample
via calibrated gas bulb at -196 °C. The tube was warmed to
1
ambient temperature and agitated overnight. Analysis by H
and 31P{1H} NMR spectroscopy revealed conversion to 3
(>85% yield) along with a small quantity of free PPh3. Method
C: A J. Young NMR tube was charged with 0.008 g of 2 and
approximately 500 μL of benzene-d6. Addition of excess triphe-
nylphosphine (∼5 equiv) afforded 3 as the only observable
organometallic product by 1H and 31P{1H} NMR spectroscopy.
Anal. Calcd for C55H52MoO2P4: C, 68.47; H, 5.43. Found: C,
68.21; H, 5.71. 1H NMR (23 °C, C6D6): δ -4.71 (tdd, 74.0, 42.4,
14.0 Hz, 1H, Mo-H), 1.20 (m, 2H, PCH2), 1.49 (m, 2H, PCH2),
1.94 (m, 2H, PCH2), 2.51-2.64 (m, 2H, PCH2), 4.68 (dd, 2.0,
10.4 Hz, CHdCH2), 5.15 (dd, 10.4, 17.2 Hz, CHdCH2), 5.48
(dd, 2.0, 17.2 Hz, CHdCH2), 6.77 (m, 6H, C6H5), 6.85 (m, 6H,
C6H5), 6.90-7.20 (m, 12H, C6H5), 7.30 (m, 7H, C6H5), 7.38 (m,
3H, C6H5), 7.58 (m, 4H, C6H5), 7.79 (t, 7.6 H, C6H5). 13C{1H}
NMR (23 °C, C6D6): δ 26.24 (PCH2), 34.77 (PCH2), 123.15
Acknowledgment. We gratefully acknowledge funding
by Brown University and the U.S. Department of Energy-
National Energy Technology Laboratory (Grant DE-
FE0004498). We also acknowledge the U.S. Department
of Energy (Grant DE-SC0001556) for funds to support a
ReactIR instrument.
Supporting Information Available: Sample kinetic data and
NMR spectra. These data are available free of charge via the
Internet at http://pubs.acs.org.