Comparative reactivity of Zr- and Pd-alkyl complexes with carbon dioxide

Article abstract of DOI:10.1021/om401082k

Structure/reactivity trends and DFT studies reveal mechanistic differences and parallels for the carboxylation of Zr and Pd alkyls. CO₂ reacts with Cp₂ZrMe(C₆D₅Cl)⁺ >10 ⁴ faster than with Cp₂ZrMe₂, yielding monoacetate products in both cases. These reactions proceed by insertion mechanisms in which Zr- - -O interactions activate the CO₂. In contrast, CO₂ reacts readily with [(PO-ⁱPr)PdMe ₂]^- (PO-ⁱPr^- = 2-P ⁱPr₂-4-Me-C₆H₃SO₃ ^-) to yield [(PO-ⁱPr)PdMe(OAc)]^- but not with (PO-ⁱPr)PdMe(L) species. Carboxylation of [(PO-ⁱPr) PdMe₂]^- occurs by direct S_E2 attack of CO ₂ at the Pd-Me_trans-to-P group, and the nucleophilicity of the Pd-Me group controls the reactivity. However, the S_E2 process is accelerated by a Li⁺- OCO interaction when Li⁺ is present.

Full text of DOI:10.1021/om401082k

Communication
pubs.acs.org/Organometallics
Comparative Reactivity of Zr− and Pd−Alkyl Complexes with Carbon
Dioxide
Ka-Cheong Lau, Benjamin J. Petro, Sebastien Bontemps, and Richard F. Jordan*
́
Department of Chemistry, The University of Chicago, 5735 South Ellis Avenue, Chicago, Illinois 60637, United States
S
* Supporting Information
ABSTRACT: Structure/reactivity trends and DFT studies reveal
mechanistic differences and parallels for the carboxylation of Zr and
Pd alkyls. CO2 reacts with Cp2ZrMe(C6D5Cl)+ >104 faster than with
Cp2ZrMe2, yielding monoacetate products in both cases. These
reactions proceed by insertion mechanisms in which Zr- - -O
interactions activate the CO2. In contrast, CO2 reacts readily with
[(PO-iPr)PdMe2]− (PO-iPr− = 2-PiPr2-4-Me-C6H3SO3 ) to yield
−
[(PO-iPr)PdMe(OAc)]− but not with (PO-iPr)PdMe(L) species. Carboxylation of [(PO-iPr)PdMe2]− occurs by direct SE2
attack of CO2 at the Pd−Metrans‑to‑P group, and the nucleophilicity of the Pd−Me group controls the reactivity. However, the SE2
process is accelerated by a Li+- - -OCO interaction when Li+ is present.
he reaction of metal alkyls with CO2 to yield metal
carboxylates plays a prominent role in strategies for the
Scheme 2
T
conversion of CO2 to value-added chemicals.1 These reactions
can proceed by insertion mechanisms involving M- - -OCO
interactions1b,c,2 or by direct SE2 attack of CO2 at the metal alkyl
carbon (Scheme 1).1g−j,o Here we explore the reactions of CO2
Scheme 1
with two prototypical Zr and Pd LxMRn+ olefin polymerization
(n−1)+
catalysts and the corresponding LxMR2
species. These
studies provide insight into the mechanistic differences and
parallels in the carboxylation of early- and late-transition-metal
alkyls.
Figure 1. Calculated free-energy profile for the reaction of Cp2ZrMe-
Cp2ZrR+ cations generated from Cp2ZrR2 precursors are
effective olefin polymerization catalysts.3 Wendt reported that
Cp2ZrMe{MeB(C6F5)3} reacts with CO2 to form [{Cp2Zr(μ-
OAc)}2][MeB(C6F5)3]2, which was characterized by X-ray
diffraction and conversion to [Cp2Zr(OAc)(THF-d8)][MeB-
(C6F5)3].2 In principle, this reaction could proceed by reaction of
CO2 with the Cp2ZrMe+ cation or with the MeB(C6F5)3− anion
to generate B(C6F5)3(OAc)−,4 followed by Me/OAc exchange.
To probe the reactivity of Cp2ZrMe+ directly, we investigated the
carboxylation of [Cp2ZrMe(C6D5Cl)][B(C6F5)4] (1).5 The
reaction of 1 with 1 atm of CO2 in C6D5Cl at 20 °C occurs
within the time of mixing to form a gelatinous product assigned
as [Cp2Zr(OAc)][B(C6F5)4]; addition of THF-d8 gives [Cp2Zr-
(OAc)(THF-d8)][B(C6F5)4] (Scheme 2). DFT computations
show that this reaction proceeds by initial formation of the κ1-O-
bound CO2 adduct Cp2Zr(Me)(CO2)+,6 followed by insertion
via a four-center transition state with a barrier of 16.4 kcal/mol
(Figure 1).
(C6H5Cl)+ with CO2.
In contrast, Cp2ZrMe2 (2) was reported to react only slowly
with CO2 (50 atm of CO2, 80 °C, 24 h) to yield an insoluble ill-
defined product that liberates acetic acid and methane upon
reaction with H2SO4.7 We found that 2 reacts very slowly with
CO2 (1 atm, 100 °C, C6D5Cl, t1/2 = ca. 30 h) to yield
Cp2Zr(Me)(OAc)8 with no NMR-observable intermediates
(Scheme 2). DFT computations show that this reaction also
proceeds by an insertion mechanism, but an intermediate CO2
adduct was not found (Figure 2). Lateral attack of CO2 (ΔG⧧ =
33.2 kcal/mol) is preferred over central attack (ΔG⧧ = 34.4 kcal/
mol).
Received: November 6, 2013
Published: November 18, 2013
© 2013 American Chemical Society
6895
dx.doi.org/10.1021/om401082k | Organometallics 2013, 32, 6895−6898

Organometallics
Communication
Figure 2. Calculated free-energy profile for the reaction of Cp2ZrMe2
(2) and CO2.
The >104-fold difference in the reactivities of 1 and 2 with CO2
reflects the greater electrophilicity of the Zr center in 1, which
results in a stronger Zr−OCO interaction and greater activation
of the CO2 for nucleophilic attack by the Zr−Me group. The
Zr- - -O distance in the transition state for CO2 insertion of 1
(2.166 Å) is much shorter than that for 2 (2.464 Å, lateral; 2.357
Å, central), consistent with a stronger Zr−O interaction.
Figure 3. Molecular structure of {[Li(THF)2][(PO-iPr)PdMe2]}2 (3a).
Hydrogen and solvent atoms are omitted. Selected bond lengths (Å)
and angles (deg): Pd(1)−C(14) 1.986(6), Pd(1)−C(15) 2.038(6),
Pd(1)−P(1) 2.309(2), Pd(1)−O(2) 2.197(4); C(14)−Pd(1)−C(15)
85.1(2), O(2)−Pd(1)−P(1) 87.52(12).
Neutral (PO)PdR(L) complexes that contain o-phosphinoar-
enesulfonate ligands ([PO]−) catalyze the insertion copoly-
merization of ethylene and polar monomers.9 These species do
not react with CO2 under conditions where they are thermally
stable. For example, (PO-iPr)PdMe(L) species ([PO-iPr]− = 2-
appears at δ −2.4, outside of the normal range for solvent-
separated Li(THF)n X− ion pairs (δ −0.5 to −1.5; e.g. δ −0.6 for
+
[Li(THF)n][B(C6F5)4]).13 These results and the previous
observation that B(C6F5)3 binds to a sulfonate oxygen of (PO-
3,5-tBu)PdMe(OEt2)10 suggest that 3a exists as a tight ion pair
with Li+ binding to a sulfonate oxygen in THF-d8. 3a is thermally
unstable in the solid state and in THF-d8, decomposing at 20 °C
to ethane, Pd0, and unidentified products.
PiPr2-4-Me-C6H3SO3 ; L = py, THF) do not react with CO2
−
(23−50 °C, 2.5−5.6 atm) in THF-d8. Similarly, the sulfonate-
bridged dimer {(PO-3,5-tBu)PdMe}2 ([PO-3,5-tBu]− = 2-
10
P(3,5-tBu2-C6H3)2-4-Me-C6H3SO3 ) is unreactive with CO2
in CD2Cl2. This lack of reactivity may reflect an inability of CO2
to displace the THF, py, or μ-sulfonate ligands, a high barrier to
insertion of a putative (PO)PdMe(CO2) species, or a high
barrier to an SE2 process. DFT calculations show that
displacement of THF from (PO-iPr)PdMe(THF) by CO2 is
endergonic by 10.9 kcal/mol, insertion of (PO-iPr)PdMe(κ1-
OCO) has a high barrier (ΔG⧧ = 33.8 kcal/mol), and overall
conversion to (PO-iPr)Pd(OAc)(THF) is endergonic by 9.2
kcal/mol. An SE2 transition state was not found. The lack of
reactivity of neutral (PO)PdMe species is consistent with
previous reports that L2PdMe2 complexes are unreactive with
CO2,11 and (PCP)PdR pincer complexes only react with CO2 at
elevated temperature when the PCP ligand is a strong donor.1g,h,o
One potential approach to enhancing the reactivity of
(PO)PdMe species with CO2 is to make the complex anionic,
to increase the nucleophilicity of the Pd−Me group.12 The
dimethyl complex {[Li(THF)2][(PO-iPr)PdMe2]}2 (3a) was
synthesized from Li[PO-iPr] and (COD)PdMe2 (Scheme 3).
−
Sequestration of Li+ from 3a by Crypt211 (1 equiv per Pd) in
THF-d8 affords [Li(Crypt211)][(PO-iPr)PdMe2] (3b). The 1H
and 7Li NMR (δ −0.7) resonances of 3b are shifted from those of
3a, and 3b is markedly more soluble in THF-d8 and more
thermally stable in comparison to 3a.
Complex 3a reacts irreversibly with CO2 (0.36 − 2.27 atm) in
THF-d8 at 20 °C to afford the acetate complex Li[(PO-iPr)-
PdMe(OAc)] (4a) in 77% yield (Scheme 4).14 4a was
Scheme 4
independently synthesized from 3a and CH3CO2H. The IR
νCO bands for 4a (1575, 1433 cm−1) were identified by
comparison of the IR spectra of 4a and Li[(PO-iPr)PdMe-
(O13C(O)Me)] (4a-13C1), which was prepared from 4a and
CH313CO2H. The ΔνCO value (νCO,asym − νCO,sym = 142 cm−1)
indicates that both oxygen atoms of the OAc unit are bound to
metals.15 Attempts to obtain crystalline samples of 4a were
unsuccessful.
Scheme 3
However, the analogous complex {[Li(THF)][(PO-iPr)Pd-
(Me)OC(O)An]}2 (6; An = 9-anthracenyl), prepared from 3a
and AnCO2H, was characterized by X-ray diffraction (Figure 4).
6 crystallizes as a dimer. The anthracenoate units are bound in a
κ1,κ1 fashion to Pd and a Li+ ion. Each Li+ also binds a THF and a
sulfonate oxygen from each [(PO-iPr)Pd(Me)OC(O)An]− unit,
similar to the structure of 3a. The IR νCO values for 6 (1595, 1430
cm−1; ΔνCO = 165 cm−1) are similar to those for 4a, consistent
with a similar carboxylate coordination mode for the two species.
The 1H and 13C NMR spectra of 3a each contain two doublets
for the Pd−Me groups, which were assigned by HMQC-NMR. In
the solid state, 3a adopts a dimeric structure in which two
(PO-iPr)PdMe2− anions are linked by Li(THF)2+ ions that bind
sulfonate oxygens (Figure 3). The Pd−Metrans‑to‑P bond is 0.05 Å
longer than the Pd−Mecis‑to‑P bond, due to the strong trans
influence of the phosphine. PGSE-NMR data suggest that 3a is
monomeric in THF-d8. The 7Li NMR resonance of 3a in THF-d8
6896
dx.doi.org/10.1021/om401082k | Organometallics 2013, 32, 6895−6898

Organometallics
Communication
Figure 4. Molecular structure of {[Li(THF)][(PO-iPr)Pd(Me)OC-
(O)An]}2 (6). Hydrogen and solvent atoms are omitted. Selected bond
lengths (Å) and angles (deg): Pd(2)−C(34) 2.006(6), Pd(2)−O(7)
2.062(5), Pd(2)−O(11) 2.150(5), Pd(2)−P(2) 2.186(2), C(48)−
O(7) 1.259(8), C(48)−O(8) 1.223(8); C(34)−Pd(2)−O(7) 86.2(3).
Figure 6. 1H NMR monitoring of the reaction of CO2 (1.06 atm) with
3a in THF-d8 at 20 °C. Unlabeled resonances: δ 1.79 (m, THF), 1.72 (s,
THF-d7), 0.25 (t, Pd−Me of 5), 0.19 (s, CH4).
Scheme 5
7
The Li NMR resonance of 4a in THF-d8 appears at δ 0.0.
Addition of Crypt211 to 4a gives [Li(Crypt211)][(PO-iPr)-
1
PdMe(OAc)] (4b). The H NMR OAc resonance of 4b is
shifted from that of 4a (4b, δ 2.01; 4a, δ 1.84), and the ΔνCO
value for 4b (237 cm−1) is consistent with κ1-OAc coordination.
Addition of Li+ to 4b regenerates 4a. These results suggest that
Li+ ion-pairs with the sulfonate and acetate oxygens of 4a in
THF-d8, as shown in Scheme 4.
The reaction of 3a and CO2 was monitored by 31P and H
1
NMR. No intermediates were observed. Upon exposure to CO2,
the 31P resonance of 3a decreased in intensity over time and that
for 4a grew in, and only resonances of 4a and byproduct 5 were
observed at the end of the reaction (Figure 5). In the 1H NMR
resulting in collapse of the 31P coupling but no line broadening of
the Pd-Me3a,trans‑to‑P resonance. The second-order rate constant
for the 3a/4a exchange estimated by line shape analysis is 1.5(8)
× 104 M−1 s−1 at 20 °C.
The kinetics of the reaction of 3a with CO2 were measured by
1H NMR by monitoring the appearance of the OAc resonance of
4a. The reaction is first order in 3a and CO2: i.e., rate = k2[3a]
PCO with k2 = 2.2(1) × 10−4 atm−1 s−1 at 20 °C.
2
In classic work, Darensbourg reported that Na+ accelerates the
insertion of CO2 into the W−R bond of cis-W(CO)4LR− (R =
Me, Et, Ph; L = CO, PMe3, P(OMe)3).1b The observation of
significant Li+ ion pairing for 3a, 4a, and 6 suggests that Li+ might
play a role in the carboxylation of 3a. Indeed, 3b, in which the Li+
is sequestered by Crypt211, reacts irreversibly with CO2 to afford
4b at a rate that is ca. 6 times slower than the reaction of 3a under
similar conditions.16 Interestingly, the exchange between 3b and
4b is also much slower than the corresponding 3a/4a exchange,
indicating that Li+ plays a role in the latter process.
Figure 5. 31P NMR monitoring of the reaction of CO2 (2.27 atm) with
3a in THF-d8 at 20 °C. The signal for 5 is at δ 27.9.
spectrum (Figure 6), the Pd-Me3a,cis‑to‑P resonance of 3a initially
appeared as a broad singlet and the Pd-Me3a,trans‑to‑P resonance
appeared as a sharp singlet. As the reaction proceeded, singlets
for the Pd−Me (broad) and Pd−OC(O)Me (sharp) groups of 4a
emerged, and the signals of 3a disappeared. These results are
consistent with the conversion of 3a to 4a and concurrent
reversible exchange of 3a and 4a.
The 3a/4a exchange can be explained by the mechanism in
Scheme 5, which involves displacement of OAc− from 4a by the
nucleophilic Pd−Metrans‑to‑P group of 3a to generate a methyl-
bridged intermediate, followed by displacement of 3a from the
intermediate by OAc−. This process permutes the Pd-Me3a,cis‑to‑P
and Pd-Me4a groups, resulting in broadening of these 1H
resonances, but simply shuttles the Pd-Me3a,trans‑to‑P groups
between equivalent positions of different (PO)PdMe units,
The observations that anionic 3a reacts readily with CO2 while
neutral (PO)PdMe(L) species do not and that Li+ accelerates the
carboxylation of 3a are consistent with an SE2 mechanism in
which Li+ can play a role. To probe the mechanism further, DFT
studies with implicit THF solvation (IEF-PCM) were carried out
−
on (PO-iPr)PdMe2 with and without Li+. For the free
−
(PO-iPr)PdMe2 anion, the lowest-energy pathway involves
SE2 attack of the CO2 carbon at the Pd−Metrans‑to‑P group with
ΔG⧧ = 31.3 kcal/mol, to generate a C−H σ-bound acetate
intermediate that relaxes to the O-bound acetate product (Figure
7). The lowest-energy pathway with inclusion of Li+ involves a
Li+-assisted SE2 transition state, in which a sulfonate oxygen
6897
dx.doi.org/10.1021/om401082k | Organometallics 2013, 32, 6895−6898

Organometallics
Communication
ACKNOWLEDGMENTS
This work was supported by the National Science Foundation
(Grants CHE-0911180 and CHE-1048528).
■
REFERENCES
■
(1) (a) Miyashita, A.; Yamamoto, A. J. Organomet. Chem. 1973, 49,
C57. (b) Darensbourg, D. J.; Hanckel, R. K.; Bauch, C. G.; Pala, M.;
Simmons, D.; White, J. N. J. Am. Chem. Soc. 1985, 107, 7463.
(c) Darensbourg, D. J.; Groetsch, G.; Weigreffe, P.; Rheingold, A. L.
Inorg. Chem. 1987, 26, 3827. (d) Arafa, I. M.; Shin, K.; Goff, H. M. J. Am.
Chem. Soc. 1988, 110, 5228. (e) Allen, O. R.; Dalgarno, S. J.; Field, L. D.;
Jensen, P.; Willis, A. C. Organometallics 2009, 28, 2385. (f) Darensbourg,
D. J.; Grotsch, G. J. Am. Chem. Soc. 1985, 107, 7473. (g) Johnson, M. T.;
Johansson, R.; Kondrashov, M. V.; Steyl, G.; Ahlquist, M. H. G.; Roodt,
A.; Wendt, O. F. Organometallics 2010, 29, 3521. (h) Schmeier, T. J.;
Hazari, N.; Incarvito, C. D.; Raskatov, J. A. Chem. Commun. 2011, 47,
1824. (i) Fan, T.; Chen, X.; Lin, Z. Chem. Commun. 2012, 48, 10808.
(j) Ostapowicz, T. G.; Holscher, M.; Leitner, W. Chem. Eur. J. 2011, 17,
10329. (k) Aresta, M.; Dibenedetto, A. Dalton Trans. 2007, 2975.
(l) Correa, A.; Martin, R. Angew, Chem. Int. Ed. 2009, 48, 6201.
(m) Schmeier, T. J.; Nova, A.; Hazari, N.; Maseras, F. Chem. Eur. J. 2012,
18, 6915. (n) Darensbourg, D. J.; Kyran, S. J.; Yeung, A. D.; Bengali, A.
A. Eur. J. Inorg. Chem. 2013, 4024. (o) Johansson, R.; Jarenmark, M.;
Wendt, O. F. Organometallics 2005, 24, 4500.
Figure 7. Calculated free-energy profile for the reaction of (PO-iPr)-
PdMe2− and CO2 with implicit THF solvation.
anchors Li+ in a position where it can bind a CO2 oxygen,
activating the CO2 for nucleophilic attack by the Pd-Metrans‑to‑P
group (Figure 8). The initial product is again a C−H σ-complex,
(2) Hill, M.; Wendt, O. Organometallics 2005, 24, 5772.
(3) Janiak, C. Coord. Chem. Rev. 2006, 250, 66.
(4) (a) Li[MeB(C6F5)3] reacts with CO2 to yield Li[B(C6F5)3(OAc)].
(b) Simona, M.; Baird, M. C. Can. J. Chem. 2006, 84, 225.
(5) Wu, F.; Dash, A. K.; Jordan, R. F. J. Am. Chem. Soc. 2004, 126,
15360.
Figure 8. Calculated free-energy profile for the reaction of Li[(PO-iPr)-
PdMe2] and CO2 with implicit THF solvation.
(6) Castro-Rodriguez, I.; Nakai, H.; Zakharov, L.; Rheingold, A. L.;
Meyer, K. Science 2004, 305, 1757.
(7) (a) Wailes, P. C.; Weigold, H.; Bell, A. P. J. Organomet. Chem. 1972,
34, 155. (b) Gambarotta, S.; Strologo, S.; Floriani, C.; Chiesa-Villa, A.;
Guastini, C. Inorg. Chem. 1985, 24, 654.
(8) Suzuki, H.; Takiguchi, T.; Kawasaki, Y. Bull. Chem. Soc. Jpn. 1978,
51, 1764.
which rearranges to an O-bound acetate product in which Li+
binds sulfonate and acetate oxygens, similar to the structure of 6.
The presence of Li+ lowers the barrier by 5.2 kcal/mol and
increases the exergonicity by 13.1 kcal/mol. Similar results were
obtained with explicit THF coordination (two per Li+). While
further work is needed to fully elucidate the role of Li+, these
results show that Li+ accelerates this carboxylation reaction by
binding a CO2 oxygen.
Previously, Wendt proposed that insertion of CO2 by
zirconocene alkyls is promoted by the presence of an open
coordination site and an electrophilic metal center, implicating
the importance of initial coordination of the CO2 to Zr, whereas
(PCP)Pd alkyls react by an SE2 process that is promoted by
electron-rich Pd centers.1g,o,2 The results reported here support
this general picture, clarify several aspects of the carboxylation of
zirconocene alkyls, and show that the presence of an anionic
charge promotes the SE2 carboxylation of Pd alkyls. Further-
more, in parallel to the Zr system, the SE2 process can be
accelerated by a Li+- - -OCO interaction when a suitably
positioned Li+ counterion is present.
(9) Review: Nakamura, A.; Anselment, T.; Claverie, J.; Goodall, B.;
Jordan, R. F.; Mecking, S.; Rieger, B.; Sen, A.; van Leeuwen, P.; Nozaki,
K. Acc. Chem. Res. 2013, 46, 1438.
(10) Cai, Z.; Shen, Z.; Zhou, X.; Jordan, R. F. ACS Catal. 2012, 2, 1187.
(11) (a) Pushkar, J.; Wendt, O. F. Inorg. Chim. Acta 2004, 357, 1295.
(b) Ariyananda, P. W. G.; Yap, G. P. A.; Rosenthal, J. Dalton Trans. 2012,
41, 7977.
(12) Anionic metal hydrides exhibit enhanced reactivity with CO2:
(a) Darensbourg, M. Y.; Ludwig, M.; Riordan, C. G. Inorg. Chem. 1989,
28, 1630. (b) Schmeier, T. J.; Dobereiner, G. E.; Crabtree, R. H.; Hazari,
N. J. Am. Chem. Soc. 2011, 133, 9274. (c) Suh, H.-W.; Schmeier, T. J.;
Hazari, N.; Kemp, R. A; Takase, M. K. Organometallics 2012, 31, 8225.
(13) Reich, H. J.; Borst, J. P.; Dykstra, R. R.; Green, D. P. J. Am. Chem.
Soc. 1993, 115, 8728.
(14) The reaction of 3a with CO2 also produces small amounts of
methane, ethane, Pd0, and Li[(PO-iPr)PdMe(κ1-PO-iPr)] (5).
(15) Deacon, G. B.; Phillips, R. J. Coord. Chem. Rev. 1980, 33, 227.
(16) Li+ does not promote carboxylation of (PO)PdMe(L) species.
ASSOCIATED CONTENT
* Supporting Information
Text, figures, tables, and CIF files giving experimental and
computational details and crystallographic data. This material is
available free of charge via the Internet at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Author
*E-mail for R.F.J.: rfjordan@uchicago.edu.
■
Notes
The authors declare no competing financial interest.
6898
dx.doi.org/10.1021/om401082k | Organometallics 2013, 32, 6895−6898