Cationic Zinc Hydride Catalyzed Carbon Dioxide Reduction to Formate: Deciphering Elementary Reactions, Isolation of Intermediates, and Computational Investigations

Article abstract of DOI:10.1002/chem.202005392

Zinc has been an element of choice for carbon dioxide reduction in recent years. Zinc compounds have been showcased as catalysts for carbon dioxide hydrosilylation and hydroboration. The extent of carbon dioxide reduction can depend on various factors, including electrophilicity at the zinc center and the denticity of the ancillary ligands. In a few cases, the addition of Lewis acids to zinc hydride catalysts markedly influences carbon dioxide reduction. These factors have been investigated by exploring elementary reactions of carbon dioxide hydrosilylation and hydroboration by using cationic zinc hydrides bearing tetradentate tris[2-(dimethylamino)ethyl]amine and tridentate N,N,N′,N′′,N′′-pentamethyldiethylenetriamine in the presence of triphenylborane and tris(pentafluorophenyl)borane.

Full text of DOI:10.1002/chem.202005392

Full Paper
doi.org/10.1002/chem.202005392
Chemistry—A European Journal
&
Homogeneous Catalysis
Cationic Zinc Hydride Catalyzed Carbon Dioxide Reduction to
Formate: Deciphering Elementary Reactions, Isolation of
Intermediates, and Computational Investigations
Raju Chambenahalli,[a] R. M. Bhargav,[a] Karl N. McCabe,[b] Alex P. Andrews,[a] Florian Ritter,[c]
Jun Okuda,[c] Laurent Maron,*[b] and Ajay Venugopal*[a]
Abstract: Zinc has been an element of choice for carbon di-
oxide reduction in recent years. Zinc compounds have been
showcased as catalysts for carbon dioxide hydrosilylation
and hydroboration. The extent of carbon dioxide reduction
can depend on various factors, including electrophilicity at
the zinc center and the denticity of the ancillary ligands. In a
few cases, the addition of Lewis acids to zinc hydride cata-
lysts markedly influences carbon dioxide reduction. These
factors have been investigated by exploring elementary re-
actions of carbon dioxide hydrosilylation and hydroboration
by using cationic zinc hydrides bearing tetradentate tris[2-
(dimethylamino)ethyl]amine and tridentate N,N,N’,N’’,N’’-pen-
tamethyldiethylenetriamine in the presence of triphenyl-
borane and tris(pentafluorophenyl)borane.
Introduction
involving ZnÀO and SiÀH bonds, regenerating the zinc hydride
catalyst accompanied by the displacement of silyl formate.[47]
Further studies have revealed that Lewis acid additives, such
as B(C6F5)3, if introduced in these catalytic reactions, promote
the reduction of silyl formates to silyl acetals and eventually to
methane.[38] Although these studies provide basic knowledge
of the elementary steps, there is ample opportunity to explore
the essential role and limits of the ZnÀH species and the Lewis
acid additives in CO2 reduction. Our preliminary investigations
on a cationic zinc hydride, [(Me6tren)ZnH]+ (Me6tren=tris[2-
(dimethylamino)ethyl]amine), for CO2 hydrosilylation in the
presence of BPh3 prompted us to explore the elementary reac-
tions in detail.[48] Herein, we provide new insights into CO2 hy-
drosilylation and hydroboration catalyzed by cationic ZnÀH
moieties. We show the effect of decreased coordination
number, resulting in the increased efficiency of hydrosilylation.
The effect of the Lewis acids B(C6F5)3 and BPh3 as additives are
also explored.
Molecular zinc compounds have received prominence as re-
agents for the reduction of unsaturated bonds.[1–4] They are
employed in the reduction of olefins,[5–8] imines,[9–11] aldehydes,
ketones,[12–18] and carbon dioxide.[19–30] Catalytic reduction of
carbon dioxide through catalytic hydrosilylation[31–34] and, to
some extent, hydroboration[35,36] with zinc catalysts has been
investigated.[37–44] Zinc acetate is known to hydrosilylate CO2 to
a mixture of reduced products: silyl formates, bis(silyl) acetals,
methoxysilanes, and methane.[45] The emergence of neutral
and cationic molecular zinc alkyls and hydrides have provided
more options for CO2 hydrosilylation. Due to inherent Lewis
acidity, alkyl zinc cations promote CO2 hydrosilylation through
activation of the C=O bond. Recent investigations have shown
that zinc hydrides catalyze CO2 hydrosilylation, resulting in se-
lective reduction to silyl formates or methoxysilanes.[43,46]
Mechanistic studies have revealed that molecular zinc hydrides
insert CO2 across ZnÀH bond. The resulting formate ZnÀ
OC(H)O reacts with the hydrosilane through s-bond metathesis
Results and Discussion
[a] R. Chambenahalli, R. M. Bhargav, A. P. Andrews, Dr. A. Venugopal
School of Chemistry, Indian Institute of Science Education and Research
Thiruvananthapuram, Vithura, Thiruvananthapuram 695551 (India)
E-mail: venugopal@iisertvm.ac.in
We recently reported that [(Me6tren)ZnH][B{3,5-C6H3(CF3)2}4]
(1a) catalyzed the reaction between CO2 and PhSiH3 in aceto-
nitrile at 508C, resulting in PhSi(OCHO)3 with a conversion of
95% in 2 h (Scheme 1).[48] The Lewis acid BPh3 was used as an
additive in equimolar quantity with respect to the zinc catalyst.
The conversion rates were temperature dependent, with the
[b] K. N. McCabe, Prof. L. Maron
LPCNO, UMR 5215, UniversitØ de Toulouse-CNRS, INSA, UPS
31077 Toulouse (France)
E-mail: maron@irsamc.ups-tlse.fr
[c] F. Ritter, Prof. J. Okuda
Institute of Inorganic Chemistry, RWTH Aachen University
Landoltweg 1, 52056 Aachen (Germany)
Supporting information and the ORCID identification number(s) for the
author(s) of this article can be found under:
Scheme 1. CO2 hydrosilylation catalyzed by compound 1a in the presence
https://doi.org/10.1002/chem.202005392.
of BPh3.
Chem. Eur. J. 2021, 27, 7391 –7401
7391
ꢀ 2021 Wiley-VCH GmbH

Full Paper
doi.org/10.1002/chem.202005392
Chemistry—A European Journal
rates being faster above 508C and lower below this tempera-
ture. Various solvents were screened for this transformation.
Chlorinated solvents, namely, chloroform and dichloromethane,
instantly converted [Me6trenZnH]+ into [Me6trenZnCl]+; the
latter does not further catalyze the reaction. Due to its ionic
nature, compound 1a is insoluble in hydrocarbon solvents.
The reaction was found to proceed only in polar aprotic sol-
vents, such as tetrahydrofuran and acetonitrile, with the effi-
ciency being superior in the latter. After screening various hy-
drosilane sources, it was observed that PhSiH3 was the only hy-
drosilane source that was active for this transformation under
the given reaction conditions. The possibility of interference
from the anion in the zinc hydride catalyst was excluded by
performing this reaction with different weakly coordinating
by B(C6F5)3. Compound 2a was characterized by means of mul-
tinuclear NMR spectroscopy and elemental analysis.
Unlike B(C6F5)3, the inability of BPh3 to abstract the hydride
from 1a can be reasoned based on their Lewis acid strengths.
To find experimental evidence for this interpretation, we aimed
to quantify and compare the Lewis acidities of BPh3 and
B(C6F5)3 with dicationic 2a by using the Gutmann–Beckett
method.[50,51] Accordingly, one equivalent of triethylphosphine
oxide (TEPO) was added to solutions of 2a, BPh3, and B(C6F5)3
in CD2Cl2, and the 31P NMR spectra of the resulting solutions
were recorded (Figure 1). A chemical shift value of d=71 ppm
was observed in the case of 2a, whereas shifts of d=73.2 and
77.2 ppm were observed for BPh3 and B(C6F5)3, respectively.
These observations suggest that 2a and BPh3 possess compa-
rable Lewis acidity, and hence, the hydride cannot be easily ab-
stracted from the zinc center. However, B(C6F5)3 has significant-
ly higher Lewis acidity relative to dicationic 2a, and hence, hy-
dride abstraction from the zinc center in 1a readily takes
place.
anions:
[B(C6F5)4]À.
[B(3,5-C6H3Cl2)4]À,
[B{3,5-C6H3Cl2(CF3)2}4]À,
and
BPh3 alone is known to catalyze CO2 hydrosilylation.[49] We,
however, observed that it took 100 h in CD3CN to consume
90% of PhSiH3 to give a mixture of hydrosilylation products,
suggesting the reactivity of BPh3 in catalyzing CO2 hydrosilyla-
tion is far lower in the absence of 1a (see the Supporting In-
formation). Performing catalysis with 1a alone as a catalyst
also resulted in 90% conversion of PhSiH3 to the respective
silyl formate in 150 h. We then considered using more Lewis
acidic B(C6F5)3 as an additive in place of BPh3. Although B(C6F5)3
alone does not catalyze the reaction, Parkin and co-workers[38]
have successfully used it as an additive in zinc hydride cata-
lyzed CO2 hydrosilylation. To our surprise, if B(C6F5)3 was em-
ployed in place of BPh3 as an additive, the catalytic reaction in
Scheme 1 was not observed under the given experimental
conditions. These observations infer that the Lewis acid addi-
tive does not directly impact catalysis, but rather influences
the active catalytic species.
Having fixed the reaction conditions, we set out to perform
a series of elementary reactions. To understand the role of the
Lewis acid additive, equimolar quantities of 1a and BPh3 were
dissolved in CD3CN. No reaction was observed, even with pro-
longed heating at 608C. However, the treatment of 1a with
B(C6F5)3 in CD3CN instantaneously led to the abstraction of the
hydride from the zinc center, resulting in [(Me6tren)Zn(CD3CN)]
[HB(C6F5)3][B{C6H3(CF3)2}4], which was further confirmed by
performing the reaction in tetrahydrofuran, leading to
[(Me6tren)Zn(C4H8O)][HB(C6F5)3][B{C6H3(CF3)2}4] (2a; Scheme 2).
Figure 1. Stacked 31P NMR spectra of TEPO and TEPO with compound 2,
BPh3, and B(C6F5)3.
In our previous investigations, we have established that 1a
readily reacts with CO2, resulting in the quantitative formation
of the zinc formate species, [(Me6tren)ZnOCHO][B{C6H3(CF3)2}4]
(3a). Based on this observation, we proceeded to investigate
the reaction between 1a and CO2 in the presence of BPh3. An
equimolar mixture of 1a and BPh3 in CD3CN was pressurized
with 1 bar of CO2 to obtain [(Me6tren)ZnOCHOBPh3]
[B{C6H3(CF3)2}4] (4a) instantaneously (Scheme 3). Compound 4a
could be independently prepared by treating zinc formate spe-
cies 3a with an equimolar amount of BPh3 (Scheme 3). The
structure of the cationic part in 4a was confirmed by prepar-
ing [(Me6tren)ZnOCHOBPh3][B{C6F5)4] (4b) through a method
analogous to that of the synthesis of 4a. Single crystals of 4b
were grown from a solution in diethyl ether, and X-ray diffrac-
tion studies elucidated the solid-state structure (Figure 2).
1
The disappearance of the ZnÀH signal in the H NMR spectrum
and the appearance of a characteristic doublet for [HB(C6F5)3]À
in the 11B NMR spectrum at d=À25.2 ppm, with a coupling
constant of 75 Hz, confirmed the abstraction of hydride in 1a
Scheme 2. Hydride abstraction from 1a with B(C6F5)3.
Chem. Eur. J. 2021, 27, 7391 –7401
www.chemeurj.org
7392
ꢀ 2021 Wiley-VCH GmbH

Full Paper
doi.org/10.1002/chem.202005392
Chemistry—A European Journal
Table 1. Selected bond lengths [Š] of compounds 3a (citation), 4b, and
5b.
3a
4b
5b
Zn1ÀN1
Zn1ÀN2
Zn1ÀN3
Zn1ÀN4
Zn1ÀO1
C1ÀO1
2.239(10)
2.175(11)
2.151(10)
2.150(9)
1.969(9)
1.272(11)
1.221(12)
–
2.181(3)
2.136(3)
2.099(3)
2.085(3)
2.033(3)
1.240(4)
1.264(4)
1.607(2)
2.176(2)
2.120(2)
2.111(2)
2.106(2)
2.0393(18)
1.231(3)
1.277(3)
1.557(3)
C1ÀO2
Scheme 3. a) Reactivity of compound 1 with CO2 and BPh3. b) Reactivity of
compound 3 with BPh3.
B1ÀO2
though 4b reacted with PhSiH3 to displace PhSiH2(OCHO) and
regenerate the zinc hydride catalyst, [(Me6tren)ZnH][B(C6F5)4]
(1b). Compound 5b reacted with PhSiH3, resulting in
PhSiH2(OCHO), and the dicationic zinc compound,
[Me6trenZn(CD3CN)][HB(C6F5)3][B(C6F5)4] (Scheme 4). From the
above observations, we concluded that the use of BPh3 along
with 1b resulted in the catalytic conversion of CO2 to PhSi(O-
CHO)3, whereas employing B(C6F5)3 with 1b resulted in the sto-
ichiometric conversion of CO2 into PhSiH2(OCHO). Having ex-
plored the elementary reactions (Scheme 5) in CO2 hydrosilyla-
tion mediated by 1a and 1b, we concluded that BPh3 was the
desired additive over B(C6F5)3. However, the exact role of the
former Lewis acid remained unexplained. We thus resorted to
computations to re-explore the elementary reactions in the
catalytic cycle.
Figure 2. Solid-state structures of cationic parts of 3a,[48] 4b, and 5b. Hydro-
gen atoms, except H1, are omitted for clarity.
If 1a was treated with CO2 in CD3CN in the presence of
B(C6F5)3, we observed the abstraction of the hydride from the
zinc center by B(C6F5)3, resulting in [(Me6tren)Zn(solvent)]
[HB(C6F5)3][B{C6H3(CF3)2}4], and no further reduction of CO2 was
observed under the given conditions (Scheme 4). This observa-
tion explains why CO2 hydrosilylation does not proceed with
B(C6F5)3 as an additive. However, the zinc formate, 3b, reacts
with B(C6F5)3, resulting in [(Me6tren)ZnOCHOB(C6F5)3][B(C6F5)4]
(5b). Single-crystal (SC) XRD studies established the structure
of 5b (Figure 2). Examining the bond lengths in 4b and 5b
(Table 1) shows that Lewis acid binding to the formate oxygen
elongates the ZnÀO bond from 1.969 to 2.033 Š in the case of
BPh3 and 2.039 Š in the case of B(C6F5)3. This elongation is pro-
portional to the strength of the Lewis acid.
Calculations were carried out on the reaction mechanism of
CO2 hydrosilylation with PhSiH3 catalyzed by 1 in the presence
of BPh3 at the DFT level (B3PW91; see the Supporting Informa-
tion for computational details). The initial step involves the
ZnÀH moiety (A) inserting CO2 via a four-membered transition
state (I) with an activation barrier of 14.3 kcalmolÀ1, resulting
in a monomeric zinc formate species, B (Figure 3). With a mod-
erate energy gain of 3.5 kcalmolÀ1, species B can dimerize to
C. Dimeric C is broken down with BPh3, resulting in D, and this
step is exothermic by 33.1 kcalmolÀ1. We ruled out the possi-
bility of silane-assisted dimer cleavage, since the induced stabi-
lization is only 12.6 kcalmolÀ1 (see the Supporting Information).
Species D then interacts with a silane molecule, leading to
four-membered s-bond metathesis, with an associated barrier
of 18.1 kcalmolÀ1. The transition state (II) appears to be fa-
vored due to the coordination of BPh3 to the carbonyl oxygen
Having obtained the formate adducts, 4b and 5b, of the
Lewis acids, we tested their reactivity with PhSiH3 in CD3CN. Al-
Scheme 4. Stoichiometric reactions of compound 3 with B(C6F5)3 and 5 with PhSiH3.
Chem. Eur. J. 2021, 27, 7391 –7401
www.chemeurj.org
7393
ꢀ 2021 Wiley-VCH GmbH

Full Paper
doi.org/10.1002/chem.202005392
Chemistry—A European Journal
Scheme 5. Schematic representation of the effect of Lewis acids on 1 and 3 in CO2 hydrosilylation.
Figure 3. Computed enthalpy profile for the hydrosilylation of CO2 catalyzed by 1 in the presence of BPh3.
(see Figures S54 and S55 in the Supporting Information). Fol-
lowing the intrinsic reaction coordinate, silyl formate is ob-
tained, along with the regeneration of the zinc hydride cata-
lyst. The overall cycle is an exothermic reaction by 70.7 kcal
molÀ1. From the reaction profile, it is proposed that silane acti-
vation is the rate-determining step. Interestingly enough, hy-
drosilylation appears to be an endothermic reaction in the ab-
sence of BPh3 because of the formation of a very stable (m-k2-
O,O) formate intermediate C.
We could not obtain conclusive experimental evidence for
the existence of compound 3 as a dimer in solution. Variable-
temperature NMR spectroscopy studies on 3 did not provide
evidence to prove the dimeric form in solution. To address
this issue, we then set out to prepare a cationic zinc hydride
with a decreased coordination number by using the tri-
dentate N,N,N’,N’’,N’’-pentamethyldiethylenetriamine (PMDTA).
[(pmdta)ZnEt][B(C6F5)4] (6) was employed as a precursor to
access [(pmdta)ZnOSiPh3][B(C6F5)4] (7), which was then treated
with excess PhSiH3 to obtain [(pmdta)ZnH][B(C6F5)4] (8;
Scheme 6). In another method, the treatment of polymeric
ZnH2 with Brønsted acid [(pmdta)H][B(C6F5)4] also led to the
formation of 8 (Scheme 6). Compounds 6–8 were characterized
by means of multinuclear NMR spectroscopic studies, elemen-
1
tal analysis, and SC-XRD studies (Figure 4). The H NMR spec-
trum of [(pmdta)ZnH][B(3,5-Me2-C6H3)4] (8b) recorded in
[D8]THF shows a characteristic ZnÀH signal at d=3.41 ppm.
SC-XRD studies revealed that 8 had a tetrahedral coordination
geometry at zinc with three sites occupied by the three nitro-
Chem. Eur. J. 2021, 27, 7391 –7401
www.chemeurj.org
7394
ꢀ 2021 Wiley-VCH GmbH

Full Paper
doi.org/10.1002/chem.202005392
Chemistry—A European Journal
Scheme 6. Synthesis of compounds 6, 7, and 8.
Figure 4. Solid-state structures of the cationic parts of 6, 7, and 8. Hydrogen atoms, except H1, in 8 are omitted for clarity.
gen atoms of PMDTA and the fourth corner occupied by a hy-
drogen atom with a ZnÀH distance of 1.469 Š (Figure 4). IR
spectroscopic studies provided a characteristic ZnÀH stretching
frequency at 1765 cmÀ1 (see the Supporting Information).
Compound 8 was pressurized with 1 bar of CO2 to form
[(pmdta)ZnOCHO][B(C6F5)4] (9) (Scheme 7). The 1H NMR spec-
trum of 9 recorded in CD3CN shows a single signal for the for-
mate CÀH at d=8.22 ppm, and the respective carbon shows a
single sharp signal at d=169.6 ppm in the 13C NMR spectrum
Figure 5. Solid-state structures of the cationic parts of 9 and 10. Hydrogen
atoms, except H1 and H2, are omitted for clarity.
(see the Supporting Information). Crystallographic studies re-
vealed that the cationic moiety in 9 existed as a dimer in the
solid state, with a distorted trigonal bipyramidal coordination
geometry at the zinc center (Figure 5). The central zinc metal is
coordinated to three nitrogen atoms of PMDTA, with an aver-
age ZnÀN distance of 2.129 Š, and two oxygen atoms with
ZnÀO distances of 1.966 and 2.085 Š (Figure 5). Treatment of
the dimeric formate complex with pyridine (py) resulted in the
formation of a mononuclear complex [(pmdta)ZnOCHO(py)][A]
(10), which was characterized by means of SC-XRD and found
Scheme 7. Synthesis of compounds 9 and 10.
Chem. Eur. J. 2021, 27, 7391 –7401
www.chemeurj.org
7395
ꢀ 2021 Wiley-VCH GmbH

Full Paper
doi.org/10.1002/chem.202005392
Chemistry—A European Journal
to contain a five-coordinate zinc center for A=B(C6H3-2,5-
Me2)4.
tions. The coordination of BPh3 allows cleavage of the dimer to
form H, which undergoes hydrosilylation with an associated
barrier of 16.2 kcalmolÀ1, indicating that CO2 insertion is the
rate-determining step of the reaction. Moreover, the highest
barrier in the process is slightly lower for compound 8 than
that for compound 1 (17.7 vs. 18.1 kcalmolÀ1), so that com-
pound 8 can be considered more active than that of 1.
The above studies prompted us to perform catalytic hydrosi-
lylation of CO2 with compound 8. As expected, compound 8
alone did not catalyze CO2 hydrosilylation in CD3CN at 608C,
even after 2 d. However, compound 8 and equimolar molar
amounts of BPh3 (1 mol% each) catalyzed the reaction be-
tween CO2 (1.5 bar) and PhSiH3 in CD3CN, resulting in 90%
conversion of SiÀH into SiÀOCHO. We further observed that
Ph2SiH2, which did not hydrosilylate CO2 in the presence of 1
and BPh3, reacted with CO2 in the presence of catalytic
amounts of 8 and BPh3. This observation proves that 8 is more
efficient than the reaction with 1 in the presence of BPh3. We
also observed that a combination of 8 and B(C6F5)3 did not cat-
alyze the hydrosilylation of CO2 under similar conditions and
even at 758C.
Furthermore, compound 9 was treated with two equivalents
of BPh3 and B(C6F5)3. Similar observations were made to those
observed for 3 to form [(pmdta)ZnOCHOBPh3][B(C6F5)4] (11)
and [(pmdta)ZnOCHOB(C6F5)3][B(C6F5)4] (12). Compound 12,
upon treatment with an excess of PhSiH3 in CD3CN, resulted in
[(pmdta)Zn(CD3CN)][HB(C6F5)3][B(C6F5)4] (13). Compounds 10–
12 were formed in situ in CD3CN and the structures were con-
firmed by means of multinuclear NMR spectroscopy studies.
These observations suggest that the decreased denticity of the
ancillary ligand stabilizes the dimeric form of the zinc formate
species. Hence, energy has to be supplied externally to break
the dimer and facilitate hydrosilylation.
Before proceeding to catalytic experiments, theoretical in-
vestigations were performed on the hydrosilylation of CO2 cat-
alyzed by 8 (Figure 6). The reaction mechanism is very similar
to that of A depicted in the reaction profile in Figure 3, and
hence, only the main differences between the two are dis-
cussed. The activation barrier for CO2 reduction to formate by
E is 17.7 kcalmolÀ1, which is slightly higher than that for A
(14.3 kcalmolÀ1). This is mainly due to the greater electronic
deficiency of the zinc center in E than that in A, as proven by
the zinc charges (+1.78 in E vs. +1.54 in A, whereas the
charges of the hydrides are À0.58 in E and À0.34 in A). Al-
though the formation of monomeric formate (F) is exothermic
(À0.7 kcalmolÀ1), dimerization is favored (À13.0 kcalmolÀ1), re-
sulting in dimer G, in agreement with experimental observa-
To experimentally prove that the role of BPh3, as explained
theoretically, is to break the formed dimers of compounds 3
and 9, further experiments have been considered. We per-
formed the hydroboration of CO2 by using 3 and 9 as catalysts
in the presence of bis(pinacolato)diboron (Hbpin; Scheme 8),
which is expected to play the dual role of serving as a Lewis
acidic center to facilitate monomeric catalyst formation and as
Figure 6. Computed enthalpy profile for the hydrosilylation of CO2 catalyzed by 8 in the presence of BPh3.
Chem. Eur. J. 2021, 27, 7391 –7401 www.chemeurj.org
7396
ꢀ 2021 Wiley-VCH GmbH

Full Paper
doi.org/10.1002/chem.202005392
Chemistry—A European Journal
own. Hydroboration was catalyzed by both compounds 1 and
8, which did not necessitate the addition of Lewis acids. Com-
pound 8 also proved to be a better catalyst for hydroboration
of CO2 with TON and TOF values of 94 and 42 hÀ1, respectively.
Scheme 8. CO2 hydroboration catalyzed by compounds 1 and 9.
Conclusion
a hydride transfer reagent. Whereas 5 mol% of 3 catalyzed the
hydroboration of CO2 selectively to OCHOBpin in 2 h at 608C,
a decreased catalyst loading of 1 mol% was required in the
case of 9. Computational studies revealed that this reaction
proceeded via a six-membered transition state involving ZnÀ
OC(H)ÀO and HBpin, with an activation barrier of 5.6 kcalmolÀ1
(Figure 7).
Experiments and DFT calculations on [(Me6tren)ZnH]+ and
[(pmdta)ZnH]+ have provided crucial evidence to understand
the elementary reactions operative in CO2 reduction catalyzed
by cationic zinc hydrides. The isolation of key intermediates
has re-enforced the theoretical propositions on the catalytic
cycles. Five-coordinate [(Me6tren)ZnH]+ alone shows moderate
reactivity towards CO2 hydrosilylation. Although decreased
ligand denticity is expected to exhibit higher reactivity, four-co-
ordinate [(pmdta)ZnH]+ alone did not catalyze this process.
This has been attributed to the formation of a stable zinc for-
mate dimer. The mild Lewis acid BPh3 was employed as an ad-
ditive to generate the monomeric zinc formate to facilitate hy-
drosilylation. B(C6F5)3 is an attractive choice as a strong Lewis
acid additive, but we have observed that it arrests the catalytic
cycle by the formation a stable [HB(C6F5)3]À anion. The use of
the hydroboration reagent HBpin further strengthened our
proposition of the role of the secondary Lewis acid in cationic
zinc hydride mediated CO2 reduction. Our findings validate the
role of cationic zinc centers in the activation of SiÀH and BÀH
bonds and reveal the importance of the second Lewis acidic
center for facile catalytic CO2 reduction.
Figure 7. Computed enthalpy profile for the hydroboration of CO2 catalyzed
by compound 3.
Experimental Section
General
Comparative data on catalytic CO2 hydrosilylation and hy-
droboration by compounds 1 and 8 are given in Table 2. Com-
pound 8 in the presence of BPh3 gives the highest TON of 273
with a TOF of 136.49 hÀ1, in comparison with previously report-
ed compound 1. This TOF is, so far, the best value observed in
the case of molecular zinc compounds.[46] It was also observed
that, in the absence of BPh3, compound 8 did not catalyze CO2
hydrosilylation, whereas compound 1 does, with a TON of 18
and TOF of 0.12 hÀ1. Compound 2 did not play any role in CO2
hydrosilylation either in the presence of compound 1 or on its
All reactions were performed under an argon atmosphere by using
standard Schlenk techniques or in a glove box under an argon at-
mosphere. Before use, glassware was dried at 2008C, and solvents
were dried, distilled, and degassed by using standard methods.[52,53]
Me6tren,[54]
[HNEt3][B{C6H3(CF3)2}4],
[(Me6tren)ZnH][A],
and
[(Me6tren)ZnOCHO][A] (A=B(C6H3(CF3)2)4/B(C6F5)4) were synthesized
and purified according to procedures reported in the literature.[48]
ZnEt2 was purchased from Sigma and used as received. [ZnH2]n,
[HNEt3][B(3,5-(CF3)2C6H3)4], and [HNEt3][B(3,5-(CF3)2C6H3)4] ([HNEt3]
[B(C6F5)4]) were synthesized according to procedures reported in
Table 2. Comparative data on catalytic CO2 hydrosilylation and hydroboration.[a]
Entry
Catalyst[b] ([mol%])
Additive ([mol%])
Hydride source
T [8C]
t [h]
Conversion [%]
TON[c]
TOF[d] [hÀ1
]
1
2
3
4
5
6
7
8
9
10
8a/b (5)
1a/b (5)
1a/b (5)
1a/b (1.66)
8a/b (0.33)
1a/b (5)
8a/b (5)
8a/b (5)
1a/b (5)
8a/b (1)
–
–
PhSiH3
PhSiH3
PhSiH3
PhSiH3
PhSiH3
Ph2SiH2
Ph2SiH2
Ph2SiH2
HBpin
60
60
60
45
60
24
150
100
2
0
90
92
98
91
0
0.0
18.0
18.4
58.8
273.0
0.0
0.0
19.6
19
0.00
0.12
0.18
29.4[48]
136.49
0
2 (5)
BPh3 (1.66)
BPh3 (0.33)
–
–
BPh3 (5)
–
2
60
60
24
24
24
6
0
0
60–75
60
98
95
94
0.82
3.16
42.00
–
HBpin
60
2
94
[a] A constant pressure of 1.5 bar of CO2 was maintained during catalysis. [b] With respect to SiÀH. [c] TON=turnover number. [d] TOF=turnover frequen-
cy.
Chem. Eur. J. 2021, 27, 7391 –7401
www.chemeurj.org
7397
ꢀ 2021 Wiley-VCH GmbH

Full Paper
doi.org/10.1002/chem.202005392
Chemistry—A European Journal
the literature.[55–57] BPh3 and B(C6F5)3 were purchased from Sigma– 89%). 1H NMR (CD3CN, 500 MHz): d=2.50 (s, 18H; NCH3), 2.68 (t,
3
Aldrich and purified by sublimation prior to use. PMDTA, PhSiH3,
Ph2SiH2, and Et3SiH were purchased from TCI Chemicals and dis-
tilled before use. Ph3SiOH was purchased from Sigma–Aldrich, puri-
fied by recrystallization, and dried in vacuo before use. CO2
(99.990% purity) was purchased from commercial sources. The
gases were passed through a column of molecular sieves dried
overnight at 2008C before performing the reactions by using
3J(H,H)=5 Hz, 6H; NCH2), 2.84 (t, J(H,H)=5 Hz, 6H; NCH2), 7.08 (m,
3H; p-C6H5), 7.17 (m, 6H; m-C6H5), 7.24 (m, 6H; o-C6H5), 8.44 ppm
(s, 1H; OCHO); 13C{1H} NMR (CD3CN, 126 MHz): d=47.4 (NCH3), 50.4
(NCH2), 57.2 (NCH2), 125.8 (p-C, C6H5), 127.8 (m-C, C6H5), 134.1 (o-C,
C6H5), 136.4 (i-C, C6F5), 138.3 (p-C, C6F5), 148.1 (m-C, C6F5), 150.0 (o-
C, C6F5), 169.1 ppm (s, OCHO); 11B NMR (CD3CN, 128 MHz): dÀ16.7
(s, B(C6F5)4), 2.5 ppm (br, OBPh3); 19F NMR (CD3CN, 471 MHz):
dÀ133.76 (o-CF), À163.94 (p-CF), À168.36 ppm (m-CF); elemental
analysis calcd (%) for C55H46B2F20N4O2Zn: C 52.35, H 3.67, N 4.44:
found: C 52.24, H 3.60, N 4.40.
1
Schlenk manifolds. H, 13C, 11B, and 19F NMR spectra were recorded
on a Bruker 500 MHz and DRX400 spectrometers. Chemical shifts
1
(d, in ppm) in the H and 13C NMR spectra were referenced to the
residual signals of the deuterated solvents. 11B NMR spectra were
referenced to the NaBH4 signal in D2O. 19F NMR spectra were refer-
enced to the CFCl3 signal. Abbreviations for NMR spectra: s (sin-
glet), d (doublet), t (triplet), q (quartet), quin (quintuplet), sext
(sextet), sep (septet), and br (broad). Samples for elemental analysis
were rigorously dried in vacuo at 10À3 mbar pressure. IR spectra
were recorded as KBr pellets by using AVATAR 360 FTIR and Pres-
tige-21 SHIMADZU FTIR spectrometers. Elemental analyses were
performed on an Elemental Vario Micro Cube machine. Crystals
were suspended in paraffin oil before being mounted on the X-ray
diffractometer. XRD data were collected on a Bruker Kappa Apex-II
charge-coupled device (CCD) diffractometer at 150 K and with
MoKa irradiation (l=0.71073 Š). SC-XRD data for 4b, 5b, 6b, 7, 8a,
8b, 9a, 10b, and 10c are reported in crystallographic information
files (CIF) accompanying this document. Details on data collection,
reduction, and refinement can be found in the individual CIFs.
Synthesis of 5b
A solution of B(C6F5)3 (0.098 mmol, 0.050 g) in THF (2 mL) was
slowly added to the solution of compound 3b (0.098 mmol,
0.100 g) in diethyl ether (2 mL). The reaction mixture was stirred
for 2 min and further concentrated to about 2 mL and layered with
n-pentane at room temperature to give 5b as colorless crystals in
24 h (0.130 g, 87%). 1H NMR (CD3CN, 500 MHz): d=2.53 (s, 18H;
3
3
NCH3), 2.75 (t, J(H,H)=5 Hz, 6H; NCH2), 2.88 (t, J(H,H)=5 Hz, 6H;
NCH2), 8.09 ppm (s, 1H; OCHO); 13C{1H} NMR (CD3CN, 126 MHz): d=
47.5 (NCH3), 49.8 (NCH2), 57.3 (NCH2), 136.4 (i-C, C6F5), 138.3 (p-C,
C6F5), 148.2 (m-C, C6F5), 150.0 (o-C, C6F5), 164.9 ppm (s, OCHO);
11B NMR (CD3CN, 128 MHz): d=À16.7 (s, B(C6F5)4), À4.4 ppm (br,
OBPh3); 19F NMR (CD3CN, 471 MHz): d=À133.76 (o-CF), À163.94 (p-
CF), À168.36 ppm (m-CF); elemental analysis calcd (%) for
C55H31B2F35N4O2Zn: C 43.12, H 2.04, N 3.66; found: C 43.00, H 1.99,
N 3.60.
Synthesis of 2b
A solution of B(C6F5)3 (0.086 mmol, 0.047 g) in THF (2 mL) was
slowly added to a solution of compound 1b (0.086 mmol, 0.100 g)
in THF (2 mL). The reaction mixture was stirred for 30 s. Volatile
compounds were removed under vacuum, followed by washing
the residue with pentane, yielding the product as an off-white
Synthesis of 6a
[Et3NH][B(C6F5)4] (0.48 mmol, 0.383 g) was added to a solution of
PMDTA (0.48 mmol, 0.100 mL) in diethyl ether (2 mL). The reaction
mixture was stirred for 2 min. A 1m solution of Et2Zn (0.48 mmol,
0.48 mL) was slowly added to the reaction mixture through a sy-
ringe at room temperature then stirred for 1 min. A white solid
precipitated out and was filtered subsequently. The residue was
dried under vacuum (10À3 mbar) at room temperature and washed
with pentane (5”3 mL) to give 6 as a pale brown powder. Color-
less crystals were obtained from the concentrated solution in di-
ethyl ether at À308C in 12 h (0.380 g, 85%). Because the crystals
were not of good quality, a similar procedure was repeated to give
[(pmdta)ZnEt][B(C6H3Cl2)4] from [Et3NH][B(C6H3Cl2)4]. Good-quality
crystals were grown from a solution in acetonitrile/pentane at
1
powder (0.130 g, 87%). H NMR (CD3CN, 500 MHz): d=2.53 (s, 18H;
3
3
NCH3), 2.75 (t, J(H,H)=5 Hz, 6H; NCH2), 2.88 (t, J(H,H)=5 Hz, 6H;
NCH2), 8.09 ppm (s, OCHO); 13C{1H} NMR (CD3CN, 126 MHz): d=47.5
(NCH3), 49.8 (NCH2), 57.3 (NCH2), 136.4 (i-C, C6F5), 138.3 (p-C, C6F5),
148.2 (m-C, C6F5), 150.0 (o-C, C6F5), 164.9 ppm (s, OCHO) (Figure S6
in the Supporting Information); 11B NMR (CD3CN, 128 MHz): d=
À16.7 (s, B(C6F5)4), À4.4 ppm (br, OBPh3); 19F NMR (CD3CN,
471 MHz): d=À133.76 (o-CF), À163.94 (p-CF), À168.36 ppm (m-
CF); satisfactory elemental analysis values for CHN could not be
obtained for this compound due the presence of trace amounts of
solvent molecules.
1
3
room temperature. H NMR (CD3CN, 500 MHz): d=0.14 (q, J(H,H)=
10 Hz, 2H; CH2Me3), 1.20 (t, J(H,H)=10 Hz, 3H; CCH3), 2.50 (s, 12H;
3
NCH3), 2.58 (s, 3H; NCH2), 2.74 (m, 4H; NCH2), 2.83 ppm (m, 4H;
NCH2); 13C{1H} NMR (CD3CN, 126 MHz): d=À3.1 (CH2Me), 13.1
(CCH3), 45.9 (NCH3), 47.4 ({NCH3}2), 54.4 (NCH2), 58.1 (NCH2), 136.4 (i-
C, C6F5), 138.3 (p-C, C6F5), 148.2 (m-C, C6F5), 150.0 ppm (o-C, C6F5);
11B NMR (CD3CN, 128 MHz): d=À16.66 ppm; 19F NMR (CD3CN,
471 MHz): d=À133.76 (brs, o-CF), À163.94 (t, p-CF), À168.36 ppm
(t, m-CF); elemental analysis calcd (%) for C35H28BF20N3Zn: C 44.40,
H 2.98, N 4.44; found: C 44.10, H 3.77, N 4.31.
Synthesis of 4b
Method A: Compound 1b (0.102 mmol, 0.100 g) and BPh3
(0.102 mmol, 0.025 g) in THF (5 mL) were pressurized with dry CO2
(1 bar). The reaction mixture was stirred at room temperature for
10 min. The volatile compounds were evaporated under vacuum
(10À3 mbar) at room temperature to give 4b as a colorless powder.
Compound 4b was further purified by recrystallization at 258C in
diethyl ether upon n-pentane layering (0.112 g, 85%).
Synthesis of 7
Method B: A solution of BPh3 (0.098 mmol, 0.023 g) in THF (2 mL)
was slowly added to a solution of compound 3b (0.0980 mmol,
0.100 g) in THF (2 mL). The reaction mixture was stirred at room
temperature for 5 min. All volatile compounds were evaporated
under vacuum (10À3 mbar) at room temperature to give 4b as a
colorless powder. Compound 4b was further purified by recrystal-
lization at 258C in diethyl ether upon n-pentane layering (0.108 g,
Compound 6 (0.317 mmol, 0.300 g) and Ph3SiOH (0.317 mmol,
0.088 g) were taken in THF (5 mL). The reaction mixture was stirred
for 20 h at room temperature. All volatile compounds were evapo-
rated under vacuum (10À3 mbar) at room temperature, washed
twice with n-pentane(2”5 mL), and dried to give 7 as a colorless
powder, which was further purified by recrystallization at À308C in
Chem. Eur. J. 2021, 27, 7391 –7401
www.chemeurj.org
7398
ꢀ 2021 Wiley-VCH GmbH

Full Paper
doi.org/10.1002/chem.202005392
Chemistry—A European Journal
THF upon n-pentane layering (0.425 g, 96%). 1H NMR (CD3CN,
500 MHz): d=2.34 (s, 6H; NCH3), 2.38 (s, 3H; NCH3), 2.40 (s, 6H;
NCH3), 2.50 (m, 2H; NCH2), 2.69 (m, 4H; NCH2), 2.81 (m, 2H; NCH2),
7.33 (brm, 9H; p-CH, m-CH, C6H5), 7.68 ppm (brm, 6H; o-CH, C6H5);
13C{1H} NMR (CD3CN, 126 MHz): d=44.73 (NCH3), 45.8 ({NCH3}2), 47.9
({NCH3}2), 54.74 (NCH2), 58.1 (NCH2), 128.4 (m-C, C6H5), 129.5 (p-C,
C6H5), 135.9 (o-C, C6H5), 136.4 (i-C, C6F5), 138.3 (p-C, C6F5), 142.8 (i-C,
C6H5), 148.2 (m-C, C6F5), 150.0 ppm (o-C, C6F5); 11B NMR (CD3CN,
128 MHz): d=À16.66 ppm; 19F NMR (CD3CN, 471 MHz): d=
À133.76 (brs, o-CF), À163.94 (t, p-CF), À168.36 ppm (t, m-CF); ele-
mental analysis calcd (%) for C51H28BF20N3OSiZn: C 51.34, H 3.21, N
3.52; found: C 51.20, H 3.28, N 3.60;
decantation and subsequently dried under reduced pressure. The
product was obtained as a colorless solid (421 mg, 627 mmol,
1
3
95%). H NMR (400 MHz, [D8]THF): d=1.17 (t, J(H,H)=7.2 Hz, 12H;
N-CH2CH3), 2.84–2.95 (m, 4H; N-CH2CH3), 2.96 (s, 4H; N-CH2CH2),
2.98–3.11 (m, 4H; N-CH2CH3), 3.73 (s, 4H; para-C6H3), 7.59 (s, 4H;
para), 7.80 ppm (s, 8H; ortho-C6H3); 13C{1H} NMR (101 MHz,
[D8]THF): d=8.9 (N-CH2CH3), 47.0 (N-CH2CH3), 51.6 (N-CH2CH2), 118.3
(para-C6H3), 125.6 (q, 1J(C,F)=272.2 Hz, CF3), 130.0 (qq, 2J(C,F)=
31.52 Hz, 3J(C,B)=2.81 Hz, meta-C6H3), 135.7 (ortho-C6H3),
162.9 ppm (q, 1J(B,C)=49.9 Hz, ipso-C6H3). 11B{1H} NMR (128 MHz,
[D8]THF): d=À6.50 ppm; 19F{1H} NMR (377 MHz, [D8]THF): d=
À63.44 ppm; elemental analysis calcd (%) for C41H36BF24N3Zn: C
44.65, H 3.29, N 3.81; found: C 44.35, H 4.37, N 7.71.
Synthesis of 8a
Synthesis of 9a
PhSiH3 (0.369 mmol, 0.045 mL) was added to a solution of 7
(0.167 mmol, 0.200 g) in THF (5 mL). The reaction mixture was
stirred for 24 h at room temperature. Traces of black precipitate
formed during the reaction and were removed by filtration. All vol-
atile compounds were removed under vacuum (10À3 mbar) at
room temperature, washed with n-pentane (3”5 mL), and dried to
give 8 as a colorless powder. The compound was recrystallized in
THF upon layering with n-pentane at À308C (0.126 g, 82%).
1H NMR (CD3CN, 500 MHz): d=2.34 (s, 6H; NCH3), 2.38 (s, 3H;
NCH3), 2.40 (s, 6H; NCH3), 2.50 (m, 2H; NCH2), 2.69 (m, 4H; NCH2),
2.81 (m, 2H; NCH2), 7.33 (brm, 9H; p-CH, m-CH, C6H5), 7.68 ppm
(brm, 6H; o-CH, C6H5); 13C{1H} NMR (CD3CN, 126 MHz): d=44.73
(NCH3), 45.8 ({NCH3}2), 47.9 ({NCH3}2), 54.74 (NCH2), 58.1 (NCH2),
128.4 (m-C, C6H5), 129.5 (p-C, C6H5), 135.9 (o-C, C6H5), 136.4 (i-C,
C6F5), 138.3 (p-C, C6F5), 142.8 (i-C, C6H5), 148.2 (m-C, C6F5),
150.0 ppm (o-C, C6F5); 11B NMR (CD3CN, 128 MHz): dÀ16.66 ppm;
19F NMR (CD3CN, 471 MHz): d=À133.76 (brs, o-CF), À163.94 (t, p-
CF), À168.36 ppm (t, m-CF); elemental analysis calcd (%) for
C33H24BF20N3Zn: C 43.14, H 2.63, N 4.57; found: C 43.02, H 2.55, N
4.50.
Compound 8 (0.092 mmol, 0.100 g) was dissolved in THF (5 mL).
The solution was pressurized with CO2 (1.5 bar) and stirred for
5 min at room temperature. All volatile compounds were evaporat-
ed under vacuum (10À2 mbar) at room temperature, washed with
n-pentane (2”5 mL), and dried to give 9 as a colorless powder.
The compound was recrystallized from a mixture of THF and n-
pentane at À308C (0.95 g, 96%). 1H NMR (CD3CN, 500 MHz): d=
2.42 (s, 3H; NCH3), 2.50 (s, 6H; NCH3), 2.54 (s, 6H; NCH3), 2.64 (m,
2H; NCH2), 2.71 (m, 2H; NCH2), 2.83 (m, 4H; NCH2), 8.22 ppm (s,
1H; OCHO); 13C{1H} NMR (CD3CN, 126 MHz): d=44.6 (NCH3), 46.4
({NCH3}2), 47.5 ({NCH3}2), 54.1 (NCH2), 57.5 (NCH2), 136.4 (i-C, C6F5),
138.3 (p-C, C6F5), 148.2 (m-C, C6F5), 150.1 (o-C, C6F5), 169.6 ppm (s,
OCHO); 11B NMR (CD3CN, 128 MHz): d=À16.66 ppm; 19F NMR
(CD3CN, 471 MHz): d=À133.76 (brs, o-CF), À163.94 (t, p-CF),
À168.36 ppm (t, m-CF); elemental analysis calcd (%) for
C34H24BF20N3O2Zn: C 42.42, H 2.51, N 4.36; found: C 42.22, H 2.48, N
4.26;
Synthesis of 9b
A solution of 8b (200 mg, 298 mmol) in THF was transferred to a
Schlenk tube. The reaction mixture was degassed by three freeze–
pump–thaw cycles and subsequently pressurized with CO2 (1 bar).
The solution was layered with n-pentane and stored at À308C.
Overnight, colorless crystals formed, which were isolated by de-
cantation and drying under high vacuum. The title complex was
obtained as a colorless solid (208 mg, 292 mmol, 98%). 1H NMR
(400 MHz, [D8]THF): d=2.10 (s, 24H; Ar-CH3), 2.30 (s, 12H; N-CH3),
2.17–2.52 (m, 11 H; N-CH2/N-CH3), 6.38 (s, 4H; para-C6H3), 7.01 (s,
8H; ortho-C6H3), 8.21 ppm (s, 1H; Zn-OCHO); 13C{1H} NMR
(101 MHz, [D8]THF): d=21.3 (Ar-CH3), 43.3 (N-CH3), 45.4 (N-CH3),
52.7 (N-CH2), 56.2 (N-CH2), 122.7 (para-C6H3), 132.1 (meta-C6H3),
Synthesis of 8b
A suspension of [ZnH2]n (53 mg, 792 mmol) in THF was added to a
solution of [(pmdta)H][B(3,5-Me2-C6H3)] (400 mg, 660 mmol) in THF
dropwise at room temperature. After gas evolution ceased (ap-
proximately 15 min), the reaction mixture was filtered and the sol-
vent volume was reduced in vacuo. Addition of n-pentane resulted
in the precipitation of a colorless solid, which was isolated by de-
cantation and subsequently dried under reduced pressure. Com-
pound 8b was obtained as a colorless solid (421 mg, 627 mmol,
1
95%). H NMR (400 MHz, [D8]THF): d=2.11 (s, 24H; Ar-CH3), 2.25 (s,
12H; N-CH3), 2.20–2.29 (m, 8H; N-CH2), 2.31 (s, 3H; N-CH3), 3.41 (s,
1H; Zn-H), 6.38 (s, 4H; para-C6H3), 7.01 ppm (s, 8H; ortho-C6H3);
13C{1H} NMR (101 MHz, [D8]THF): d=22.3 (Ar-CH3), 45.4 (N-CH3), 47.1
(N-CH3), 54.2 (N-CH2), 57.8 (N-CH2), 123.7 (para-C6H3), 133.1 (meta-
1
134.5 (ortho-C6H3), 164.2 (q, J(B,C)=49.2 Hz, ipso-C6H3), 171.2 ppm
(Zn-OCHO); 11B{1H} NMR (128 MHz, [D8]THF): d=À6.50 ppm; ele-
mental analysis calcd (%) for C84H120B2N6O4Zn2: C 70.54, H 8.46, N
5.88; found: C 69.25, H 8.45, N 5.72.
1
C6H3), 135.5 (ortho-C6H3), 165.7 ppm (q, J(B,C)=49.2 Hz, ipso-C6H3);
11B{1H} NMR (128 MHz, [D8]THF): d=À6.97 ppm; elemental analysis
calcd (%) for C41H60BN3Zn: C 73.38, H 9.01, N 6.26; found: C 72.44,
H 8.99, N 6.47.
Synthesis of 10a
Method A: [(pmdta)Zn(OCHO)]2[B(3,5-Me2-C6H3)4]2 (100 mg,
70 mmol) was dissolved in py (1 mL). The solution was layered with
n-pentane and stored at À308C overnight. The title compound
was obtained as a colorless solid (108 mg, 108 mmol, 98%).
Synthesis of 8c
A suspension of [ZnH2]n (53 mg, 792 mmol) in THF was added a so-
lution of [(pmdta)H][B(3,5-(CF3)2-C6H3)] (400 mg, 660 mmol) in THF
dropwise at room temperature. After gas evolution ceased (ap-
proximately 15 min), the reaction mixture was filtered and the sol-
vent volume was reduced in vacuo. Addition of n-pentane resulted
in the precipitation of a colorless solid, which could be isolated by
Method B: A solution of 8b (100 mg, 149 mmol) in py (1 mL) was
degassed by three freeze–pump–thaw cycles and pressurized with
CO2 (1 bar). The solution was layered with n-pentane and stored at
À308C overnight. The title compound was obtained as a colorless
solid (113 mg, 143 mmol, 96%). 1H NMR (400 MHz, [D8]THF): d=
Chem. Eur. J. 2021, 27, 7391 –7401
www.chemeurj.org
7399
ꢀ 2021 Wiley-VCH GmbH

Full Paper
doi.org/10.1002/chem.202005392
Chemistry—A European Journal
2.10 (s, 24H; Ar-CH3), 2.26 (s, 12H; N-CH3), 2.28 (s, 3H; N-CH3),
2.19–2.46 (m, 8H; N-CH2), 6.38 (s, 4H; para-C6H3), 7.01 (s, 8H;
ortho-C6H3), 7.18–7.39 (m, 2H; meta-C6H5N), 7.66 (t, J(H,H)=7.6 Hz,
7.59 (t, 3J(H,H)=5 Hz, 2H; p-C6H5), 7.77 (d, 3J(H,H)=5 Hz, 4H; o-
C6H5), 8.26 ppm (s, 2H; OCHO); 13C{1H} NMR (CD3CN, 126 MHz): d=
129.2 (i-C, C6H5), 129.4 (p-C, C6H5), 133.0 (o-C, C6H5), 135.8 (m-C,
C6H5), 160.4 ppm (OCHO)
3
1H; meta-C6H5N), 8.22 (s, 1H; Zn-OCHO), 8.53 ppm (dt, 3J(H,H)=
4.4 Hz, 4J(H,H)=1.7 Hz, 2H; ortho-C6H5N); 13C{1H} NMR (101 MHz,
[D8]THF): d=19.4 (Ar-CH3), 41.4 (N-CH3), 43.5 (N-CH3), 50.8 (N-CH2),
54.4 (N-CH2), 120.8 (para-C6H3), 121.6 (meta-C5H5N), 130.3 (meta-
C6H3), 132.6 (ortho-C6H3), 133.6 (para-C5H5N), 147.9 (ortho-C5H5N),
162.9 (q, 1J(B,C)=49.2 Hz, ipso-C6H3), 169.2 ppm (Zn-OCHO);
11B{1H} NMR (128 MHz, [D8]THF): d=À6.95 ppm; elemental analysis
calcd (%) for C47H65BN4O2Zn: C 71.08, H 8.25, N 7.05; found: C
69.76, H 7.88, N 7.48.
Hydroboration of CO2 with 1
An NMR tube equipped with a J. Young valve was charged with
compound 1 (0.005 g, 0.0041 mmol). [D3]Acetonitrile (0.500 mL)
was added, followed by HBpin (0.0125 mL, 0.086 mmol) and mesi-
tylene (0.004 mL, 0.028 mmol). The reaction mixture was degassed
three times, and a constant pressure of CO2 (1.5 bar) was main-
tained at 608C. The progress of the reaction was monitored by
1H NMR spectroscopy for the consumption of HBpin (95% in 6 h).
1H NMR (CD3CN, 500 MHz): d=1.31 (s, 12H; CH3), 2.26 (s, 3H; Mes-
CH3), 6.80 (s, 1H; Mes-CH), 8.41 ppm (s, 1H; OCHO); 13C{1H} NMR
(CD3CN, 126 MHz): d=24.9 (CH3), 85.8 (CCH3), 159.6 ppm (OCHO);
11B NMR (CD3CN, 128 MHz): d=22.63 ppm (OCHOBpin).
Synthesis of 10b
A solution of [(pmdta)ZnH][B(3,5-(CF3)2C6H3)4] (100 mg, 90 mmol) in
py (1 mL) was degassed by three freeze–pump–thaw cycles and
pressurized with CO2 (1 bar). The solution was layered with n-pen-
tane and stored at À308C overnight. The title compound was ob-
tained as
a
colorless solid (107 mg, 88 mmol, 97%). 1H NMR
Hydroboration of CO2 with 8
(400 MHz, [D8]THF): d=2.63 (d, J(H,H)=36.1 Hz, 12H; N-CH3), 2.66
(s, 3H; N-CH3), 2.84–3.18 (m, 8H; N-CH2), 7.40 (m, 2H; meta-C5H5N),
7.62 (s, 4H; para-C6H3), 7.79 (m, 1H; para-C5H5N), 7.84 (s, 8H;
ortho-C6H3), 8.33 (s, 1H; Zn-OCHO), 8.61 ppm (m, 2H; ortho-C5H5N);
13C{1H} NMR (101 MHz, [D8]THF): d=44.5 (N-CH3), 46.7 (br, N-CH3),
54.0 (N-CH2), 57.7 (N-CH2), 118.3 (para-C6H3), 125.0 (C5H5N-py), 125.6
(q, 1J(C,F)=272.2 Hz, CF3), 130.1 (qq, 2J(C,F)=31.5 Hz, 3J(C,B)=
2.8 Hz, meta-C6H3), 135.7 (ortho-C6H3), 137.6 (para-C5H5N), 150.7
(ortho- C5H5N), 162.9 (q, 1J(B,C)=49.9 Hz, ipso-C6H3), 173.0 ppm
(Zn-OCHO); 11B{1H} NMR (128 MHz, [D8]THF): d=À6.50 ppm;
19F{1H} NMR (377 MHz, [D8]THF): d=À63.39 ppm; elemental analy-
sis calcd (%) for C47H41BF24N4O2Zn: C 46.04, H 3.37, N 4.57; found: C
44.13, H 3.84, N 4.76.
An NMR tube equipped with a J. Young valve was charged with
compound 8 (0.008 g, 0.0081 mmol). [D3]Acetonitrile (0.500 mL)
was added, followed by HBpin (0.1175 mL, 0.81 mmol) and mesity-
lene (0.037 mL, 0.27 mmol). The reaction mixture was degassed
three times, and a constant pressure of CO2 (1.5 bar) was main-
tained at 608C. The progress of the reaction was monitored by
1H NMR spectroscopy for the consumption of HBpin (94% in 2 h).
1H NMR (CD3CN, 500 MHz): d=1.31 (s, 12H; CH3), 2.24 (s, 3H; Mes-
CH3), 6.80 (s, 1H; Mes-CH), 8.40 ppm (s, 1H; OCHO); 13C{1H} NMR
(CD3CN, 126 MHz): d=24.9 (CH3), 85.8 (CCH3), 159.6 ppm (OCHO);
11B NMR (CD3CN, 128 MHz): d=22.26 ppm (OCHOBpin).
Acknowledgements
Hydrosilylation of CO2 with PhSiH3 by 8 and BPh3
We thank the IISER Thiruvananthapuram for generous funding
and research facilities; the Alexander von Humboldt Founda-
tion for a short-term (May–July, 2018) research fellowship for
A.V. L.M. is a senior member of the Institut Universitaire de
France and thanks CalMip for computational facilities.
An NMR tube equipped with a J. Young valve was charged with
compound
8
(0.004 g, 0.0043 mmol) and BPh3 (0.001 g,
0.0043 mmol). [D3]Acetonitrile (0.500 mL) was added to this mix-
ture followed by PhSiH3 (0.0535 mL, 0.434 mmol) and mesitylene
(0.020 mL, 0.1446 mmol). The reaction mixture was degassed three
times, and a constant pressure of CO2 (1.5 bar) was maintained at
608C. The progress of the reaction was monitored by 1H NMR spec-
troscopy for the consumption of PhSiH3 (91% in 2 h). 1H NMR
(CD3CN, 500 MHz): d=2.26 (s, 3H; Mes-CH3), 6.81 (s, 1H; Mes-CH),
7.50 (t, 3J(H,H)=5 Hz, 1H; p-C6H5), 7.55 (t, 3J(H,H)=5 Hz, 2H; m-
Conflict of interest
The authors declare no conflict of interest.
3
C6H5), 7.88 (d, J(H,H)=5 Hz, 2H; o-C6H5), 8.20 ppm (s, 3H; OCHO);
13C{1H} NMR (CD3CN, 126 MHz): d=129.3 (i-C, C6H5), 129.4 (p-C,
C6H5), 133.6 (o-C, C6H5), 135.4 (m-C, C6H5), 159.9 ppm (OCHO);
chemical shift values corresponding to PhSi(OCHO)3 are given.
Keywords: carbon dioxide reduction · hydroboration
hydrosilylation · reactive intermediates · zinc hydride
·
Hydrosilylation of CO2 with Ph2SiH2 by compound 8 and
BPh3
[1] R. B. King, Encyclopedia of Inorganic Chemistry, Wiley-VCH, Weinheim,
2006.
[2] D. Seyferth, Organometallics 2001, 20, 2940–2955.
[3] S. Enthaler, ACS Catal. 2013, 3, 150–158.
An NMR tube equipped with a J. Young valve was charged with
[4] Y. Li, K. Junge, M. Beller, Zinc Catal. Appl. Org. Synth. 2015, 5–32.
[5] R. J. Wehmschulte, L. Wojtas, Inorg. Chem. 2011, 50, 11300–11302.
[6] D. Specklin, F. Hild, C. Fliedel, C. Gourlaouen, L. F. Veiros, S. Dagorne,
Chem. Eur. J. 2017, 23, 15908–15912.
[7] R. J. Procter, M. Uzelac, J. Cid, P. J. Rushworth, M. J. Ingleson, ACS Catal.
2019, 9, 5760–5771.
[8] A. Rit, T. P. Spaniol, J. Okuda, Chem. Asian J. 2014, 9, 612–619.
[9] P. Jochmann, D. W. Stephan, Chem. Eur. J. 2014, 20, 8370–8378.
[10] P. Jochmann, D. W. Stephan, Angew. Chem. Int. Ed. 2013, 52, 9831–
9835; Angew. Chem. 2013, 125, 10014–10018.
compound
8
(0.008 g, 0.0081 mmol) and BPh3 (0.002 g,
0.0081 mmol). [D3]Acetonitrile (0.500 mL) was added to this mix-
ture followed by PhSiH3 (0.050 mL, 0.271 mmol) and mesitylene
(0.037 mL, 0.271 mmol). The reaction mixture was degassed three
times and a constant pressure of CO2 (1.5 bar) was maintained at
608C for 8 h and at 758C for 10 h. The progress of the reaction
was monitored by 1H NMR spectroscopy for the consumption of
1
Ph2SiH2 (98% in 24 h). H NMR (CD3CN, 500 MHz): d=2.25 (s, 3H;
Mes-CH3), 6.8 (s, 1H; Mes-CH), 7.50 (t, J(H,H)=5 Hz, 4H; m-C6H5),
3
Chem. Eur. J. 2021, 27, 7391 –7401
www.chemeurj.org
7400
ꢀ 2021 Wiley-VCH GmbH

Full Paper
doi.org/10.1002/chem.202005392
Chemistry—A European Journal
[11] J. L. Lortie, T. Dudding, B. M. Gabidullin, G. I. Nikonov, ACS Catal. 2017,
7, 8454–8459.
[12] W. Sattler, S. Ruccolo, M. Rostami Chaijan, T. Nasr Allah, G. Parkin, Orga-
nometallics 2015, 34, 4717–4731.
[13] C. Boone, I. Korobkov, G. I. Nikonov, ACS Catal. 2013, 3, 2336–2340.
[14] M. J. C. Dawkins, E. Middleton, C. E. Kefalidis, D. Dange, M. M. Juckel, L.
Maron, C. Jones, Chem. Commun. 2016, 52, 10490–10492.
[15] Z. Mou, H. Xie, M. Wang, N. Liu, C. Yao, L. Li, J. Liu, S. Li, D. Cui, Organo-
metallics 2015, 34, 3944–3949.
[36] M. L. Shegavi, S. K. Bose, Catal. Sci. Technol. 2019, 9, 3307–3336.
[37] W. Sattler, G. Parkin, J. Am. Chem. Soc. 2012, 134, 17462–17465.
[38] M. Rauch, G. Parkin, J. Am. Chem. Soc. 2017, 139, 18162–18165.
[39] A. Rit, A. Zanardi, T. P. Spaniol, L. Maron, J. Okuda, Angew. Chem. Int. Ed.
2014, 53, 13273–13277; Angew. Chem. 2014, 126, 13489–13493.
[40] G. Feng, C. Du, L. Xiang, I. Del Rosal, G. Li, X. Leng, E. Y. X. Chen, L.
Maron, Y. Chen, ACS Catal. 2018, 8, 4710–4718.
[41] D. Specklin, C. Fliedel, C. Gourlaouen, J. C. Bruyere, T. AvilØs, C. Boudon,
L. Ruhlmann, S. Dagorne, Chem. Eur. J. 2017, 23, 5509–5519.
[42] J. C. Bruyere, D. Specklin, C. Gourlaouen, R. Lapenta, L. F. Veiros, A.
Grassi, S. Milione, L. Ruhlmann, C. Boudon, S. Dagorne, Chem. Eur. J.
2019, 25, 8061–8069.
[16] R. K. Sahoo, M. Mahato, A. Jana, S. Nembenna, J. Org. Chem. 2020, 85,
11200–11210.
[17] P. A. Lummis, M. R. Momeni, M. W. Lui, R. McDonald, M. J. Ferguson, M.
Miskolzie, A. Brown, E. Rivard, Angew. Chem. Int. Ed. 2014, 53, 9347– [43] F. Ritter, T. P. Spaniol, I. Douair, L. Maron, J. Okuda, Angew. Chem. Int. Ed.
9351; Angew. Chem. 2014, 126, 9501–9505.
2020, 132, 23535–23543.
[18] I. D. Alshakova, G. I. Nikonov, Synth. 2019, 51, 3305–3312.
[19] A. K. Wiegand, A. Rit, J. Okuda, Coord. Chem. Rev. 2016, 314, 71–82.
[20] N. J. Brown, J. E. Harris, X. Yin, I. Silverwood, A. J. P. White, S. G. Kazarian,
K. Hellgardt, M. S. P. Shaffer, C. K. Williams, Organometallics 2014, 33,
1112–1119.
[44] X. Wang, K. Chang, X. Xu, Dalt. Trans. 2020, 49, 7324–7327.
[45] Q. Zhang, N. Fukaya, T. Fujitani, J. C. Choi, Bull. Chem. Soc. Jpn. 2019, 92,
1945–1949.
[46] M. Tüchler, L. Gärtner, S. Fischer, A. D. Boese, F. Belaj, N. C. Mçsch-Zanet-
ti, Angew. Chem. Int. Ed. 2018, 57, 6906–6909.
[21] P. Jochmann, D. W. Stephan, Organometallics 2013, 32, 7503–7508.
[22] G. Ballmann, S. Grams, H. Elsen, S. Harder, Organometallics 2019, 38,
2824–2833.
[47] M. M. Deshmukh, S. Sakaki, Inorg. Chem. 2014, 53, 8485–8493.
[48] R. Chambenahalli, A. P. Andrews, F. Ritter, J. Okuda, A. Venugopal, Chem.
Commun. 2019, 55, 2054–2057.
[23] W. Sattler, G. Parkin, J. Am. Chem. Soc. 2011, 133, 9708–9711.
[24] D. M. Ermert, I. Ghiviriga, V. J. Catalano, J. Shearer, L. J. Murray, Angew.
[49] D. Mukherjee, D. F. Sauer, A. Zanardi, J. Okuda, Chem. Eur. J. 2016, 22,
7730–7733.
Chem. Int. Ed. 2015, 54, 7047–7050; Angew. Chem. 2015, 127, 7153– [50] M. A. Beckett, G. C. Strickland, J. R. Holland, K. Sukumar Varma, Polymer
7156.
1996, 37, 4629–4631.
[25] M. Rauch, Y. Rong, W. Sattler, G. Parkin, Polyhedron 2016, 103, 135–140.
[26] S. Schulz, T. Eisenmann, D. Schuchmann, M. Bolte, M. Kirchner, R. Boese,
[51] U. Mayer, V. Gutmann, W. Gerger, Monatsh. Chemie 1975, 106, 1235–
1257.
J. Spielmann, S. Harder, Zeitschr. Naturforsch. Sect. B. 2009, 64, 1397– [52] M. A. D. D. F. Schriver, The Manipulation of Air-Sensitive Compounds,
1400.
Wiley, 1986.
[27] A. Rit, T. P. Spaniol, L. Maron, J. Okuda, Angew. Chem. Int. Ed. 2013, 52,
4664–4667; Angew. Chem. 2013, 125, 4762–4765.
[28] R. Han, I. B. Gorrell, A. G. Looney, G. Parkin, J. Chem. Soc. Chem.
Commun. 1991, 10, 717–719.
[53] W. L. F. Armarego, C. L. L. Chai, Purification of Laboratory Chemicals
2013, Elsevier.
[54] M. Ciampolini, N. Nardi, Inorg. Chem. 1966, 5, 41–44.
[55] M. Brookhart, B. Grant, A. F. Volpe, Organometallics 1992, 11, 3920–
[29] K. Merz, M. Moreno, E. Lçffler, L. Khodeir, A. Rittermeier, K. Fink, K.
Kotsis, M. Muhler, M. Driess, Chem. Commun. 2008, 73–75.
[30] M. Rombach, H. Brombacher, H. Vahrenkamp, Eur. J. Inorg. Chem. 2002,
153–159.
3922.
´
[56] R. Anulewicz-Ostrowska, T. Klis, D. Krajewski, B. Lewandowski, J. Serwa-
towski, Tetrahedron Lett. 2003, 44, 7329–7331.
[57] A. J. Martínez-Martínez, A. S. Weller, Dalton Trans. 2019, 48, 3551–3554.
[31] F. J. Fernµndez-Alvarez, L. A. Oro, ChemCatChem 2018, 10, 4783–4796.
[32] S. Dagorne, Synth. 2018, 50, 3662–3670.
[33] F. J. Fernµndez-Alvarez, A. M. Aitani, L. A. Oro, Catal. Sci. Technol. 2014,
4, 611 –624.
[34] M. C. Lipke, A. L. Liberman-Martin, T. D. Tilley, Angew. Chem. Int. Ed.
2017, 56, 2260–2294; Angew. Chem. 2017, 129, 2298–2335.
Manuscript received: December 17, 2020
Revised manuscript received: January 11, 2021
Accepted manuscript online: January 18, 2021
Version of record online: February 24, 2021
´
[35] K. Kucinski, G. Hreczycho, Green Chem. 2020, 22, 5210–5224.
Chem. Eur. J. 2021, 27, 7391 –7401
www.chemeurj.org
7401
ꢀ 2021 Wiley-VCH GmbH