Highly efficient palladium-catalysed carbon dioxide hydrosilylation employing PMP ligands

Article abstract of DOI:10.1039/C8DT03777A

A series of zero-valent palladium complexes featuring diphosphinometal ligands is reported. The metalloligand in the PMP-type framework (M = Li^I, Cu^I, Zn^II) acts as a weak to medium acceptor ligand for palladium. A Pd⁰→Zn^II bond was observed in the solid state and further confirmed by NBO analysis. The heterobimetallic Pd/Zn complex 2-Zn displayed excellent activity in chemoselective CO₂ hydrosilylation producing silyl formate (TOF_1/2 = 3000 h^?1).

Full text of DOI:10.1039/C8DT03777A

Dalton
Transactions
View Article Online
View Journal | View Issue
PAPER
Highly efficient palladium-catalysed carbon
dioxide hydrosilylation employing PMP ligands†
Cite this: Dalton Trans., 2019, 48,
1017
Patrick Steinhoff, Melanie Paul, Julian P. Schroers and Michael E. Tauchert
*
A series of zero-valent palladium complexes featuring diphosphinometal ligands is reported. The metallo-
ligand in the PMP-type framework (M = LiI, CuI, ZnII) acts as a weak to medium acceptor ligand for
palladium. A Pd0→ZnII bond was observed in the solid state and further confirmed by NBO analysis. The
heterobimetallic Pd/Zn complex 2-Zn displayed excellent activity in chemoselective CO2 hydrosilylation
producing silyl formate (TOF1/2 = 3000 h−1).
Received 18th September 2018,
Accepted 14th December 2018
DOI: 10.1039/c8dt03777a
rsc.li/dalton
metalloligands which coordinate only weakly via an acceptor
s-orbital. To this end, we developed DPPTPA (bis(diphenylpho-
Introduction
Z-Type ligands provide an increasingly popular approach to sphino)trispyridylamine). This neutral pre-ligand was used to
fine-tune transition metals (TM) for homogeneous catalysis.1–3 produce a series of [(PMP)Rh(CO)Cl]n+ complexes by chemo-
The valence-neutral coordination of these ligands allows the selective κ4N-coordination of LiI, CuI or ZnII cations to its TPA-
formation of complexes with uncommonly high coordination scaffold (Fig. 1).22
numbers for a given oxidation state, while their acceptor pro-
Pd→M (M = CuI, ZnII) interactions are thought to play an
perties simultaneously modify the electronic properties of the important role in palladium catalysed cross-coupling
coordinated metal.4 Chelating “ambiphilic” ligands in which reactions.23–29 Surprisingly, only very limited structural data
Lewis basic L-type donor functionalities are combined with for Pd0→M interactions (LiI, CuI) are available, and no
Lewis acidic Z-type acceptor functionalities have been success- example of a Pd0→ZnII interaction has been reported (see the
fully applied in catalysis, including challenging transform- ESI; Fig. S2†).24–26,30–36
ations like N2 fixation5 or harnessing CO2 as a versatile C1
To produce zero-valent palladium complexes featuring LiI,
building block.6,7 Lu and coworkers used GaIII as an acceptor CuI and ZnII as metalloligands, we substituted the PPh2-
ligand to generate a nickel catalyst (Fig. 1) for carbon dioxide groups of the DPPTPA ligand with PCy2-groups. In this paper,
hydrogenation with unprecedented high formate turnovers.6 we report on the application of this new DCPTPA ligand (bis
Iwasawa and coworkers employed an AlIII-ligand to produce a (dicyclohexylphosphino)trispyridylamine) to produce zero-
palladium catalyst with the highest activity reported so far for valent palladium complexes featuring PMP-type ligands and
carbon dioxide hydrosilylation yielding silyl formates.7 The
TM → Z bond in these catalysts is typically formed by an inter-
action of an occupied d-orbital (TM) with an unoccupied
p-orbital at the acceptor atom (Z).3,8–12 Additionally, the π-type
interaction with an unoccupied d-orbital (Z) has been
described.3 The manipulation of the ligand’s acceptor pro-
perties has successfully been used to control catalytic
activity.6,8,13–16
Strong acceptor ligands may stabilize low oxidation states to
an extent that oxidative addition reactions are inhibited.17–21
We were motivated to extend the toolbox of acceptor ligands to
Institute of Inorganic Chemistry, RWTH Aachen University, Landoltweg 1A,
D-52074 Aachen, Germany. E-mail: Michael.Tauchert@ac.rwth-aachen.de
†Electronic supplementary information (ESI) available: Experimental section,
details of XRD analysis, NMR spectra, and xyz-files for calculated structures.
CCDC 1855254 (1), 1569435 (2-Zn) and 1815009 (2-Li). For ESI and crystallo-
graphic data in CIF or other electronic format see DOI: 10.1039/c8dt03777a
Fig. 1 Various ligand scaffolds featuring metalloligands.
This journal is © The Royal Society of Chemistry 2019
Dalton Trans., 2019, 48, 1017–1022 | 1017

View Article Online
Paper
Dalton Transactions
their application in highly efficient and chemoselective CO2
hydrosilylation. The Pd0→M interaction in these complexes is
analysed by computational methods.
Results and discussion
The pre-ligand DCPTPA was synthesised by palladium cata-
lysed phosphorylation37 of the TPA derivative Br2-TPA 22,38 with
HPCy2 and was isolated in 81% yield (Scheme 1). The reaction
of DCPTPA with Zn(NTf2)2 in MeOH produced the desired
PZnP2+ ligand 1. Single crystals suitable for X-ray diffraction
analysis were grown from MeOH. The crystal structure of 1
confirmed the anticipated κ4N-coordination of ZnII to the TPA-
scaffold of DCPTPA. Two independent complexes of 1 were
observed in the solid state. The P,P-distances of 4.311 and
4.542 Å are similar to established diphosphinoligands like trip-
typhos (4.34–4.63 Å)39 and xantphos (4.05–4.16 Å).40 The 31P
NMR spectrum showed a singlet resonance at δ 32.1. The
coordination of ZnII to the DCPTPA ligand made its methylene
hydrogens Ha and Hb diastereotopic (Fig. 2). The resulting AB-
Fig. 2 1H NMR resonances of DCPTPA methylene groups in 2-Zn (top,
THF-d8), 1 (middle, CD2Cl2) and DCPTPA (bottom, CD2Cl2).
been made for P→B interactions in the diphosphinoborane
ligand (o-iPr2PC6H4)2BPh.41
The reaction of PZnP2+ ligand 1 with [CpPd(C3H5)] gave the
Cs-symmetric complex 2-Zn (31P NMR δ 58.1) in 80% selectivity
(Scheme 2). The κ4N-coordination of the ZnII ion was
suggested by an AB-type resonance in the 1H NMR spectrum at
δ 4.48 and 4.62 (J = 16.8 Hz, Fig. 2). Single crystals for XRD
analysis were grown from THF/HMDS. The crystal structure of
2-Zn confirmed the formation of a heterobimetallic complex
(Scheme 2). 2-Zn displayed a ZnII cation coordinated in a
trigonal bipyramidal fashion (τ5 = 0.79),43 with the pyridine
ligands in the equatorial position, and the N3 bridgehead and
palladium in the apical positions. The P-donors coordinate in
trans-configuration to the palladium atom. The weak κ1O-
coordination of the NTf2-anion via its SO2-group (d(Pd,O) =
2.4366(12) Å) made the configuration of the palladium center
distorted square planar (τ4 = 0.21).44 The observed Pd,Zn-
distance of 2.4652(2) Å in 2-Zn is smaller than the sum of the
covalent radii (2.61 Å).45 The Formal Shortness Ratio (FSR) is
0.98.46 Structural data for molecular Pd,Zn-interactions is
limited (see ESI, Fig. S2†). The Pd,Zn-distance in 2-Zn is
similar to those reported in metal rich Pd,Zn-cluster
1
type resonance in the H NMR spectrum (δ 4.51 and 4.44, J =
17.5 Hz) was found to be diagnostic of the κ4N-coordination of
a metal to the DCPTPA ligand (vide infra). It has also been
observed for the DPPTPA ligand.22 The MeOH ligand in
1 could be removed in vacuo without any significant change in
the 1H or 31P NMR spectrum. With regard to the significant
downfield shift of 1 compared to free DCPTPA (31P NMR δ 5.6),
we suggest that the MeOH ligand dissociates in solution and
the free coordination site at ZnII is occupied by one of the two
P-donors, which exchange rapidly. Similar observations have
Scheme 1 Synthesis of pre-ligand DCPTPA (top) and PZnP2+ ligand 1
(bottom). ORTEPs of 1 shown at 50% probability level. Only one of two
independent complexes was depicted. Two non-coordinating NTf2-
anions and DCPTPA hydrogens omitted for clarity. Selected bond
lengths (Å) and angles (°): P1–P2 = 4.311, Zn1–O1 = 2.074(2), N3–Zn1–
Scheme 2 Synthesis (left) and crystal structure of 2-Zn (right). ORTEPs
shown at 50% probability level. One non-coordinating NTf2-anion, one
solvent THF and H atoms omitted for clarity. Selected bond lengths (Å)
and angles (°): Pd1–Zn1 = 2.4652(2), Pd1–P1 = 2.2961(4), Pd1–P2 =
2.2888(4), Pd1–O1 = 2.4366(12), Zn1–Pd1–O1 = 151.62(4), N4–Zn1–Pd1
= 172.48(3), P1–Pd1–P2 = 156.214(15).42
O1
= 145.14(9), N2–Zn1–N4 = 104.5(1), N1–Zn1–N2 = 113.33(10),
N3–Zn1–N2 = 79.53(10).42
1018 | Dalton Trans., 2019, 48, 1017–1022
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
compounds ([(tBuNC)2Pd(ZnCp*)4]: 2.428(1)–2.484(1) Å,31 [Pd
(ZnCp*)4(ZnMe)4]: 2.417(1)–2.459(1) Å,32 and [Cp*Pd(ZnCp*)3]:
2.369(1)–2.384(1) Å (ref. 30)), but is shorter than the PdII→ZnII
interaction reported by Chen (2.5786(3) Å; ESI, Fig. S2†).24 To
the best of our knowledge, 2-Zn is the first crystallographically
characterised example featuring a dative Pd0 → ZnII bond.
Complex 2-Zn could also be produced in a one-pot reaction
from DCPTPA, Zn(NTf2)2 and [CpPd(C3H5)] in 61% isolated
yield (Scheme 2).
Attempts were made to synthesize PMP ligands PLiP+ and
PCuP+. However, while LiNTf2 was readily dissolved in the
presence of one equiv. of DCPTPA (CD2Cl2), no significant
change in the 31P and 1H NMR spectra could be observed.47
No changes were observed in the 31P NMR spectrum when
DCPTPA was reacted with two equiv. of [Cu(MeCN)4]OTf in
C6D6. No AB-type resonance was detected in the 1H NMR
spectrum.
Fig. 3 Left: Crystal structure of 2-Li. ORTEPs shown at 50% probability
level. NTf2-anions, one solvent toluene, H atoms and atom sites with
minor occupancies (25.5%) omitted for clarity. Selected bond lengths (Å)
and angles (°): Pd1–Li1 = 2.665(9), Pd1–P1 = 2.2612(13), Pd1–P2 =
2.2705(13), N3–Li1–Pd1
= 162.6(6), P1–Pd1–P2 = 157.27(5). Right:
Superposition of 2-Li (light grey) and mirrored 2-Zn (dark grey).42
Complex 2-Li was produced in a one-pot reaction of
DCPTPA, [CpPd(C3H5)] and LiNTf2 in benzene (Scheme 3) and
7
displayed a singlet resonance in the 31P and Li NMR spectra
benzene produced complexes 2-Zn and 2-Li, respectively. The
reaction of Zn(NTf2)2 with 2-Li resulted in the formation of 2-
Zn.
(δ 39.8 and δ 4.0 in benzene-d6). Single crystals of 2-Li suitable
for XRD analysis were produced from toluene/hexane. The
crystal structure of 2-Li (Fig. 3) was found to be very similar to
that of 2-Zn and featured a Pd,Li-distance of 2.665(9) Å. This
distance is equal to the sum of the covalent radii (2.67 Å)45
and within the range of the only other three reported examples
Complexes 2-Zn, 2-Cu and 2-Li were analysed by compu-
tational methods (see the ESI†). For NBO/NLMO analyses of
the Pd→M interactions, model complexes 3-M were used, in
which the PCy2-groups of the DCPTPA ligand were truncated
to PMe2-groups (bis(dimethylphosphino)trispyridylamine:
DMPTPA). For complex [(dmptpa)ZnIIPd0]2+ 3-Zn, a profound
interaction between a donor dx2−y2-orbital of palladium and an
acceptor s-orbital of zinc was observed (Table 1, Fig. 4). The
corresponding NLMO is similar to the donor NBO with a
major contribution from the occupied palladium dx2−y2-orbital
(89.0%) and a “delocalisation tail” between the metals from
the vacant zinc s-orbital (9.0%) (Fig. 4).
(see ESI, Fig. S2†) with
a
short Pd,Li-distance
(2.457–2.737 Å).33–35,48 The FSR is 1.02.46 No NTf2-coordination
to the palladium center was observed.
The reaction product of DCPTPA, [CpPd(C3H5)] and
(CuOTf)2·C6H6 displayed poor solubility in benzene, THF and
DMF. Satisfactory solubility was realised for [(dcptpa)CuIPd0][B
(C6F5)4] 2-Cu, which was produced in a one-pot reaction
(Scheme 3) with the addition of one equiv. of K[B(C6F5)4] (31P
NMR δ 45.8).
As a reference for catalysis, we attempted to synthesise the
metalloligand free complex [(dcptpa)Pd0] 2. The reaction of
DCPTPA with [CpPd(C3H5)] in Et2O led to the precipitation of a
yellow solid. CHN combustion analysis agreed with the compo-
sition of [(dcptpa)Pd] 2. However, 31P NMR spectroscopy
revealed a mixture of complexes in which all phosphorus
donors were coordinated to a palladium center (δ 26.9–44.5).
The reaction of 2 with one equivalent of Zn(NTf2)2 or LiNTf2 in
The HOMO−2 molecular orbital of 3-Zn showed a similarly
distorted dx2−y2-type orbital at palladium (Fig. 4).
The atom–atom net linear NLMO/NPA bond order of the
Pd0→ZnII bond is 0.256, which is of the same order of magni-
tude as the P–Pd bonds in 3-Zn (0.258 and 0.304). A similar
situation was found for [(dmptpa)CuIPd0]+ 3-Cu (see ESI,
Fig. S4†). However, the interaction is weaker with only 2.6% of
Table 1 NBO/NLMO analysis of the Pd→M interaction for model com-
plexes 3-Zn, 3-Cu and 3-Lia
Occ.
Pdd
Occ.
Md
Bond
e
ΔEb
% Pdc
% Mc
ΔEij
orderf
3-Zn
3-Cu
3-Li
93.3
27.8
11.8
89.0
95.3
96.9
9.0
2.6
1.4
1.819
1.908
1.940
0.584
0.348
0.167
0.215
0.359
0.444
0.256
0.115
0.078
a Structure optimisation: Turbomole 7.0.1, BP86/def-SV(P); NBO ana-
lysis: Gaussian 09/NBO 6.0, BP86/6-31G(d), MWB28 (Pd), MDF10 (Cu,
Zn). b NBO stabilizing energy associated with the Pd→M interaction, in
kcal mol−1 c Contribution of the donor/acceptor NBO to the NLMO.
.
d Occupancy of the donor/acceptor NBO. e Energy difference between
donor/acceptor NBO in Eh. f Atom–atom net linear NLMO/NPA bond
order.
Scheme 3 One-pot syntheses of complexes 2 M.
This journal is © The Royal Society of Chemistry 2019
Dalton Trans., 2019, 48, 1017–1022 | 1019

View Article Online
Paper
Dalton Transactions
Table 2 Hydrosilylation of CO2
Entrya
Catalyst
Additive
Yieldb
1
2
3
2-Zn
2-Zn
2
Without CsOPiv
—
—
0.1 mol% Zn(NTf2)2
—
0.5 mol% Zn(NTf2)2
—
—
24
100
11
57
68
0
12
9
4c
5d
6d
7e
8
2
2-Zn
2-Zn
2-Zn
2-Li
9
2-Li
0.4 mol% LiNTf2
—
0.1 mol% (CuOTf)2·C6H6
0.1 mol% Zn(NTf2)2
8
6
0.5
0
10
11
12
2-Cu
DCPTPA
DCPTPA
a Catalyst (3.3 µmol), CsOPiv (33 µmol), PhSiMe2H (3.3 mmol), DMF
(0.2 mL), CO2 (1 bar). b 1H NMR yield based on C6Me6 as an internal
standard. c 2 and Zn(NTf2)2 stirred in 0.2 mL DMF for 4 h at r.t. before
the addition of CsOPiv and PhSiMe2H. d Kinetics 0.05 mol% 2-Zn,
0.5 mol% CsOPiv. e 0.01 mol% 2-Zn.
Fig. 4 Computational analysis of the Pd0 → ZnII interaction in 3-Zn. (a)
Donor NBO (Pd); (b) acceptor NBO (Zn); (c) corresponding NLMO; (d)
molecular orbital HOMO−2.49
(Fig. 1).7 Complex 2-Li gave comparable results (9%) to 2
(11%), regardless of additionally added LiNTf2 (Table 2,
entries 8, 3 and 9). Utilisation of 2-Cu as a catalyst gave 6%
conversion (Table 2, entry 10). Cu51,52 and Zn53,54 are known to
catalyze the hydrosilylation of carbon dioxide. However, con-
version using DCPTPA and (CuOTf)2·C6H6 or DCPTPA and Zn
(NTf2)2 as a catalyst was negligible (0.5% and 0%; Table 2,
entries 11 and 12).
the copper s-orbital compared to 95.3% of the palladium
d-orbital contributing to the NLMO (see ESI, Fig. S4†). In line
with this observation is a lower NBO stabilizing energy of
27.8 kcal mol−1 compared to 3-Zn (93.3 kcal mol−1) and a
decreased Pd0→CuI bond order of 0.115 (Table 1). In the case
of [(dmptpa)LiIPd0]+ 3-Li, a NLMO is formed from an occupied
palladium dz2-orbital (96.9%) with a small contribution of
1.4% from a vacant lithium s-orbital (see ESI, Fig. S5†). A
small NBO stabilizing energy (11.8 kcal mol−1) and a decreased
Pd0→LiI bond order of 0.078 suggest that the Pd0→LiI inter-
action in 3-Li is negligible.
Conclusions
Complexes 2-Zn, 2-Cu, 2-Li and 2 were tested for catalytic
activity in the hydrosilylation of CO2 with PhSiMe2H (Table 2).
Complex 2-Zn (0.1 mol%) yielded 24% of the hydrosilylation
product within 30 min at 30 °C (Table 2, entry 1). The yield
increased to 100%, when 1 mol% of CsOPiv was used as an
additive to 2-Zn (Table 2, entry 2). The catalytic performance is
poor (11%) for complex 2 but increased to 57% in the presence
of Zn(NTf2)2 (Table 2, entries 3 and 4), suggesting that 2-Zn is
the catalytically active species. Decreasing the load of 2-Zn to
0.05 mol% with time resolved monitoring of the reaction (1H
NMR) resulted in 68% conversion within 30 min, and a TOF1/2
(turnover frequency after 50% conversion) of 3000 h−1
(Table 2, entry 5; ESI, Table S4†). No reaction was observed
under these conditions in the presence of an additional
0.5 mol% of Zn(NTf2)2 (Table 2, entry 6).
In summary, we introduced a series of zero-valent palladium
complexes featuring PMP-type ligands (M = LiI, CuI, ZnII)
which are conveniently formed in a one-pot synthesis from
pre-ligand DCPTPA and the appropriate metal precursors.
NBO analysis suggests weak to medium Pd0→M (M = CuI, ZnII)
interactions via an acceptor s-orbital of the metalloligand.
Complex [(dcptpa)ZnIIPd0](NTf2)2 2-Zn displayed excellent
activity (TOF1/2 = 3000 h−1) and selectivity in the hydrosilyl-
ation of carbon dioxide.55
Conflicts of interest
There are no conflicts to declare.
Lowering the loading of 2-Zn to 0.01 mol% gave the same
TOF1/2 as for 0.1 mol% and 12% conversion within 30 min
(Table 2, entry 7). To the best of our knowledge, this is one of
Acknowledgements
the highest TOFs reported for this transformation50 (ESI, We are grateful for the financial support by the Funds of the
Table S5†), significantly surpassed only by Baba’s monometal- Chemical Industry (fellowship to MET and PS). Simulations
lic copper catalyst yielding polymethylsiloxane formate (–Si were performed with computing resources granted by RWTH
(Me)(OCHO)O–)n as the product51 and Iwasawa’s Al/Pd catalyst Aachen University under project rwth0245. We thank K.-N.
1020 | Dalton Trans., 2019, 48, 1017–1022
This journal is © The Royal Society of Chemistry 2019

View Article Online
Dalton Transactions
Paper
Truong for collection of X-ray diffraction data, M. Konrad for 22 P. Steinhoff, R. Steinbock, A. Friedrich, B. G. Schieweck,
experimental assistance during a group internship, and Prof.
Dr I. M. Oppel and Prof. Dr J. Okuda for generous support.
C. Cremer, K.-N. Truong and M. E. Tauchert, Dalton Trans.,
2018, 47, 10439–10442.
23 K. Böck, J. E. Feil, K. Karaghiosoff and K. Koszinowski,
Chem. – Eur. J., 2015, 21, 5548–5560.
24 R. J. Oeschger and P. Chen, Organometallics, 2017, 36,
1465–1468.
Notes and references
1 G. Bouhadir and D. Bourissou, Chem. Soc. Rev., 2016, 45, 25 R. J. Oeschger and P. Chen, J. Am. Chem. Soc., 2017, 139,
1065–1079. 1069–1072.
2 H. Kameo and H. Nakazawa, Chem. – Asian J., 2013, 8, 26 R. J. Oeschger, D. H. Ringger and P. Chen, Organometallics,
1720–1734. 2015, 34, 3888–3892.
3 R. C. Cammarota, L. J. Clouston and C. C. Lu, Coord. Chem. 27 R. Álvarez, Á. R. de Lera, J. M. Aurrecoechea and A. Durana,
Rev., 2017, 334, 100–111. Organometallics, 2007, 26, 2799–2802.
4 A. Amgoune and D. Bourissou, Chem. Commun., 2011, 47, 28 J. del Pozo, G. Salas, R. Álvarez, J. A. Casares and
859–871. P. Espinet, Organometallics, 2016, 35, 3604–3611.
5 J. S. Anderson, J. Rittle and J. C. Peters, Nature, 2013, 501, 29 A. B. González-Pérez, R. Álvarez, O. N. Faza, Á. R. de Lera
84–87.
and J. M. Aurrecoechea, Organometallics, 2012, 31, 2053–
6 R. C. Cammarota, M. V. Vollmer, J. Xie, J. Ye, J. C. Linehan,
2058.
S. A. Burgess, A. M. Appel, L. Gagliardi and C. C. Lu, J. Am. 30 T. Bollermann, K. Freitag, C. Gemel, R. W. Seidel and
Chem. Soc., 2017, 139, 14244–14250. R. A. Fischer, Organometallics, 2011, 30, 4123–4127.
7 J. Takaya and N. Iwasawa, J. Am. Chem. Soc., 2017, 139, 31 K. Freitag, M. Molon, P. Jerabek, K. Dilchert, C. Rösler,
6074–6077.
R. W. Seidel, C. Gemel, G. Frenking and R. A. Fischer,
Chem. Sci., 2016, 7, 6413–6421.
32 T. Cadenbach, T. Bollermann, C. Gemel, M. Tombul,
I. Fernandez, M. V. Hopffgarten, G. Frenking and
R. A. Fischer, J. Am. Chem. Soc., 2009, 131, 16063–16077.
8 D. You, H. Yang, S. Sen and F. P. Gabbaï, J. Am. Chem. Soc.,
2018, 140, 9644–9651.
9 W. Guan, G. Zeng, H. Kameo, Y. Nakao and S. Sakaki,
Chem. Rec., 2016, 16, 2405–2425.
10 H. Kameo, Y. Hashimoto and H. Nakazawa, 33 M. D. Fryzuk, B. R. Lloyd, G. K. B. Clentsmith and
Organometallics, 2012, 31, 3155–3162. S. J. Rettig, J. Am. Chem. Soc., 1994, 116, 3804–3812.
11 M. Sircoglou, S. Bontemps, G. Bouhadir, N. Saffon, 34 M. D. Fryzuk, B. R. Lloyd, G. K. B. Clentsmith and
K. Miqueu, W. Gu, M. Mercy, C.-H. Chen, B. M. Foxman, S. J. Rettig, J. Am. Chem. Soc., 1991, 113, 4332–4334.
L. Maron, O. V. Ozerov and D. Bourissou, J. Am. Chem. Soc., 35 H.-O. Fröhlich, R. Wyrwa and H. Görls, J. Organomet.
2008, 130, 16729–16738. Chem., 1993, 456, 7–12.
12 S. Bontemps, M. Sircoglou, G. Bouhadir, H. Puschmann, 36 C. M. Fafard, C.-H. Chen, B. M. Foxman and O. V. Ozerov,
J. A. K. Howard, P. W. Dyer, K. Miqueu and D. Bourissou,
Chem. – Eur. J., 2007, 14, 731–740.
13 M. V. Vollmer, J. Xie, R. C. Cammarota, V. G. Young Jr.,
Chem. Commun., 2007, 43, 4465–4467.
37 M. Murata and S. L. Buchwald, Tetrahedron, 2004, 60,
7397–7403.
E. Bill, L. Gagliardi and C. C. Lu, Angew. Chem., Int. Ed., 38 D. Schnieders, M. Merkel, S. M. Baldeau and B. Krebs, Z.
2018, 57, 7815–7819. Anorg. Allg. Chem., 2004, 630, 1210–1214.
14 H. Yang and F. P. Gabbaï, J. Am. Chem. Soc., 2015, 137, 39 M. E. Tauchert, D. C. M. Warth, S. M. Braun, I. Gruber,
13425–13432.
A. Ziesak, F. Rominger and P. Hofmann, Organometallics,
2011, 30, 2790–2809.
40 S. Hillebrand, J. Bruckmann, C. Krüger and M. W. Haenel,
Tetrahedron Lett., 1995, 36, 75–78.
41 S. Bontemps, G. Bouhadir, P. W. Dyer, K. Miqueu and
D. Bourissou, Inorg. Chem., 2007, 46, 5149–5151.
42 CCDC 1855254 (1), 1569435 (2-Zn) and 1815009 (2-Li)†
contain the supplementary crystallographic data for this
paper.
43 A. W. Addison and T. N. Rao, J. Chem. Soc., Dalton Trans.,
1984, 1349–1356.
44 L. Yang, D. R. Powell and R. P. Houser, Dalton Trans., 2007,
955–964.
45 B. Cordero, V. Gómez, A. E. Platero-Prats, M. Revés,
J. Echeverría, E. Cremades, F. Barragán and S. Alvarez,
Dalton Trans., 2008, 2832–2938.
15 R. C. Cammarota and C. C. Lu, J. Am. Chem. Soc., 2015,
137, 12486–12489.
16 F. Inagaki, K. Nakazawa, K. Maeda, T. Koseki and
C. Mukai, Organometallics, 2017, 36, 3005–3008.
17 C. M. Conifer, D. J. Law, G. J. Sunley, A. J. P. White and
G. J. P. Britovsek, Organometallics, 2011, 30, 4060–4066.
18 T. Schindler, M. Lux, M. Peters, L. T. Scharf, H. Osseili,
L. Maron and M. E. Tauchert, Organometallics, 2015, 34,
1978–1984.
19 D. Schuhknecht, F. Ritter and M. E. Tauchert, Chem.
Commun., 2016, 52, 11823–11826.
20 P. Steinhoff and M. E. Tauchert, Beilstein J. Org. Chem.,
2016, 12, 1573–1576.
21 W. K. Walker, B. M. Kay, S. A. Michaelis, D. L. Anderson,
S. J. Smith, D. H. Ess and D. J. Michaelis, J. Am. Chem. Soc.,
2015, 137, 7371–7378.
46 L. J. Pauling, J. Am. Chem. Soc., 1947, 69, 542–553.
This journal is © The Royal Society of Chemistry 2019
Dalton Trans., 2019, 48, 1017–1022 | 1021

View Article Online
Paper
Dalton Transactions
47 A singlet resonance was observed in the 7Li NMR spectrum at
δ 3.49. LiNTf2 is insoluble in CH2Cl2 in the absence of
DCPTPA. The chemical shift of LiNTf2 in THF-d8 is δ, −0.61.
48 F. Armbruster, T. Augenstein, P. Oña-Burgos and F. Breher,
Chem. – Eur. J., 2013, 19, 17899–17906.
49 Jmol: an open-source Jawa viewer for chemical structures in
3D, http://www.jmol.org.
50 F. J. Fernández-Alvarez and L. A. Oro, ChemCatChem, 2018,
10, 4783–4796.
51 K. Motokura, D. Kashiwame, N. Takahashi, A. Miyaji and
T. Baba, Chem. – Eur. J., 2013, 19, 10030–10037.
52 L. Zhang, J. Cheng and Z. Hou, Chem. Commun., 2013, 49,
4782–4784.
53 M. Rauch and G. Parkin, J. Am. Chem. Soc., 2017, 139,
18162–18165.
54 W. Sattler and G. Parkin, J. Am. Chem. Soc., 2012, 134,
17462–17465.
55 One of the referees asked for an analysis of the reaction
mechanism with computational methods to elucidate the
role of the Pd0→ZnII interaction. Calculations using 3-M,
Me3SiH and formate as a model were performed (see the
ESI† and ref. 6). A plausible reaction pathway was found in
agreement with intermediates observed by Lu in CO2
hydrogenation with their Ga/Ni catalyst (Fig. 1). In this ten-
tative mechanism CO2 reacts with the hydride complex
[(dmptpa)MPd0H−I]n+ (7-M) giving the formate complex
[(dmptpa)MPd0(OCHO)]n+ (6-M). 6-M regenerates 7-M by
reaction with Me3SiH and simultaneous extrusion of silyl
formate (see ESI; Scheme S2†). This step would be in agree-
ment with the observed beneficial effect of the additive
CsOPiv on catalysis (see Table 2 entries 1 and 2). NBO/
NLMO analysis showed a substantial stabilization of the
hydride coordination in 7-Zn by the zinc ligand (see ESI;
Fig. S12†). This makes the generation of 7-Zn from 6-Zn
less endergonic (ΔG298.15 = +5.0 kcal mol−1), compared to
the transition 6-Li/7-Li (ΔG298.15 = +6.6 kcal mol−1), where
no substantial stabilization of the hydride coordination by
the lithium ligand was found (see ESI, Fig. S13†).
1022 | Dalton Trans., 2019, 48, 1017–1022
This journal is © The Royal Society of Chemistry 2019