Ph2PCH2CH2B(C8H14) and Its Formaldehyde Adduct as Catalysts for the Reduction of CO2with Hydroboranes

pubs.acs.org/IC

Article

Ph₂PCH₂CH₂B(C₈H₁₄) and Its Formaldehyde Adduct as Catalysts for

the Reduction of CO₂with Hydroboranes

́

̃

Alberto Ramos,* Antonio Antinolo, Fernando Carrillo-Hermosilla, and Rafael Fernandez-Galan

Cite This: https://dx.doi.org/10.1021/acs.inorgchem.0c01152

Read Online

sı

*

Metrics & More

Article Recommendations

Supporting Information

ACCESS

ABSTRACT: We study two metal-free catalysts for the reduction of CO₂with four

diﬀerent hydroboranes and try to identify mechanistically relevant intermediate

species. The catalysts are the phosphinoborane Ph₂P(CH₂)₂BBN (1), easily accessible

in a one-step synthesis from diphenyl(vinyl)phosphine and 9-borabicyclo[3.3.1]-

nonane (H-BBN), and its formaldehyde adduct Ph₂P(CH₂)₂BBN(CH₂O) (2),

detected in the catalytic reduction of CO₂with 1 as the catalyst but properly prepared

from compound 1 and p-formaldehyde. Reduction of CO₂with H-BBN gave mixtures

of CH₂(OBBN)₂(A) and CH₃OBBN (B) using both catalysts. Stoichiometric and

kinetic studies allowed us to unveil the key role played in this reaction by the

formaldehyde adduct 2 and other formaldehyde−formate species, such as the

polymeric BBN(CH₂)₂(Ph₂P)(CH₂O)BBN(HCO₂) (3) and the bisformate macro-

cycle BBN(CH₂)₂(Ph₂P)(CH₂O)BBN(HCO₂)BBN(HCO₂) (4), whose structures

were conﬁrmed by diﬀractometric analysis. Reduction of CO₂with catecholborane

(HBcat) led to MeOBcat (C) exclusively. Another key intermediate was identiﬁed in the reaction of 2 with the borane and CO₂, this

being the bisformaldehyde−formate macrocycle (HCO₂){BBN(CH₂)₂(Ph₂P)(CH₂O)}₂Bcat (5), which was also structurally

characterized by X-ray analysis. In contrast, using pinacolborane (HBpin) as the reductant with catalysts 1 and 2 usually led to

mixtures of mono-, di-, and trihydroboration products HCO₂Bpin (D), CH₂(OBpin)₂(E), and CH₃OBpin (F). Stoichiometric

studies allowed us to detect another formaldehyde−formate species, (HCO₂)BBN(CH₂)₂(Ph₂P)(CH₂O)Bpin (6), which may play

an important role in the catalytic reaction. Finally, only the formaldehyde adduct 2 turned out to be active in the catalytic

hydroboration of CO₂using BH₃·SMe₂as the reductant, yielding a mixture of two methanol-level products, [(OMe)BO]₃(G, major

product) and B(OMe)₃(H, minor product). In this transformation, the Lewis adduct (BH₃)Ph₂P(CH₂)₂BBN was identiﬁed as the

resting state of the catalyst, whereas an intermediate tentatively formulated as the Lewis adduct of compound 2 and BH₃was

detected in solution in a stoichiometric experiment and is likely to be mechanistically relevant for the catalytic reaction.

turn, milder reaction conditions (25−90 °C, 1−5 atm of CO₂),

which facilitate monitoring of the reactions to gain valuable

mechanistic information and even enable the use of metal-free

catalysts.^4b,d,f,i

Turning our attention to the catalytic hydroboration of CO₂

to the methanol level, Guan and co-workers reported the ﬁrst

example back in 2010, using a Ni complex with a pincer ligand

and catecholborane (HBcat) as the reductant.⁶In 2013,

Fontaine and co-workers reported the ﬁrst example of a metal-

free catalyst for the reduction of CO₂to the methanol level

using an ambiphilic phosphinoboronate catalyst (I in Scheme

1) and up to four diﬀerent commercial hydroboranes: HBcat,

pinacolborane (HBpin), 9-borabicyclo[3.3.1]nonane (H-

INTRODUCTION

■

In the last decades, there has been increasing concern among

the scientiﬁc community to develop more eﬃcient CO₂

capture and conversion technologies in order to reduce the

concentration of greenhouse gas in the atmosphere and, at the

same time, use it as a C1 feedstock for value-added chemicals.¹

One of the most interesting and challenging transformations to

achieve is an eﬃcient six-electron reduction of captured CO₂

to produce methanol, which could then be used as a fuel. This

is the paradigm of a circular economy coined as “Methanol

Economy”, initially proposed by Olah and collaborators.²From

an industrial point of view, hydrogen would be the ideal

reductant, generating water as the only byproduct. However,

only a few homogeneous catalysts based on transition metals

have been reported to date for the aforementioned trans-

formation, operating under harsh reaction conditions, with low

eﬃciencies and usually requiring stoichiometric amounts of

additives.³Alternatively, hydroboranes⁴and hydrosilanes^4a,e,i,5

can also be employed as reductants. Although less appealing

from an industrial point of view because of the generation of

undesired B- or Si-containing byproducts, their use allows, in

Received: April 21, 2020

https://dx.doi.org/10.1021/acs.inorgchem.0c01152

Inorg. Chem. XXXX, XXX, XXX−XXX

A

Inorganic Chemistry

pubs.acs.org/IC

Article

Scheme 1. P/B-Containing Catalysts for the Reduction of CO₂to the Methanol Level with Hydroboranes

a

Table 1. Catalytic Hydroboration of CO₂with HBBN

b

c

entry

catalyst

cat. load (mol %)

T (°C)

t (h)

conv (%)

TON A

TON B

1

2

3

4

5

6

7

8

Ph₂PCHCH₂

1

2

1

25

60

25

60

25

60

22

0.5

1

3

1

12

1

16

20

0.5

1

3

1

7

12

0.2

0.5

3

1

7

>99

47

80

100

51

83

40

63

100

50

82

>99

23

45

100

60

>99

98

5.0

9.6

14.6

14.8

220.5

92.0

6.6

2.5

1

0.1

0.05

1

2.5

1

0.1

1

14.3

64.6

228.5

663.0

322.0

764.0

91.4

9.2

19.0

70.8

167.0

324.0

92.3

8.8

263.0

15.0

9.5

8.9

9.6

6.6

77.0

9.5

9

10

11

12

13

14

15

16

17

18

19

20

1.1

2.5

1

0.1

14.5

16.1

18.0

97.5

17.5

17.1

71.0

177.0

341.0

37

63

a

Reaction conditions: J. Young valve NMR tube containing ca. 0.6 mL C₆D₆solutions of H-BBN (0.2 mmol), catalyst and internal standard

b

c

[Si(SiMe₃)₄, 0.01 mmol], and CO₂(1 atm). Based on the relative integrals of H-BBN and products in the ¹¹B NMR. Calculated according to the

number of C−H bonds formed in the reduction products by integration of the corresponding signals relative to the internal standard in the H

NMR spectra.

1

BBN), and BH₃·SMe₂.⁷Since then, a limited number of metal-

free catalysts have been reported, mostly involving strong bases

or ambiphilic species.^7,8Among them, a few phosphinoborane

systems turned out to be eﬃcient, easily accessible catalysts for

this reaction. As mentioned above, Fontaine and co-workers

developed a series of phosphinoboronate catalysts, which

turned out to be very active,^7,8g,lespecially when using BH₃as

the reductant. Mechanistic studies revealed the crucial role of

the formaldehyde adducts of phosphineboronate species (II in

Scheme 1)^8gdetected during the catalysis that, when used as

catalysts themselves, eliminated the induction period pre-

viously observed for the reactions catalyzed by the parent

B

https://dx.doi.org/10.1021/acs.inorgchem.0c01152

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

pubs.acs.org/IC

Article

phosphinoboronates. Wang and Stephan reported that simple

phosphines (III in Scheme 1) were capable of reducing CO₂to

a mixture of boron formate, acetal, and boron methoxide using

H-BBN as both the reductant and Lewis acid counterpart for

the phosphine to form an intermolecular frustrated Lewis pair

(FLP).⁸ⁱThe same authors also reported intramolecular P/B

FLPs, generated by ring expansion in the reactions of

phosphine-derived carbenes and H-BBN (IV in Scheme 1)

as active catalysts on the reduction of CO₂with HBpin, HBcat,

and BH₃·SMe₂.^8hFinally, Cantat and co-workers reported the

synergistic eﬀects of ferrocene-based phosphinoboranes (V in

Scheme 1) in the catalytic reduction of CO₂to the methoxide

level with H-BBN.⁹

In order to ﬁll the gap left by the phenylene and

ferrocenylene phosphinoboranes, we decided to focus on a

simpler ethylene-linked phosphinoborane, Ph₂P(CH₂)₂BBN

(1),¹⁰easily accessible in an atom-economical uncatalyzed

process from commercial reagents (Ph₂PCHCH₂and H-

BBN), which, to our knowledge, had not been tested as a

catalyst for hydroboration of CO₂. We also used the four most

common commercial hydroboranes as reductantsH-BBN,

HBcat, HBpin, and BH₃·SMe₂obtaining in most cases

methoxide derivatives as the major (or only) reduction

products. Additionally, the formaldehyde adduct Ph₂P-

(CH₂)₂BBN(CH₂O) (2), detected as an intermediate in the

catalytic reduction with H-BBN, was tested as a catalyst as well

with all of the diﬀerent boranes. Stoichiometric and kinetic

experiments helped us to identify other important intermedi-

ates and gain some mechanistic insights for these trans-

formations. These studies not only conﬁrmed the crucial role

played by the formaldehyde adduct 2 but also that of other

formaldehyde−formate derivatives detected during the catal-

ysis, some of them structurally characterized for the ﬁrst time.

number (TON) of 779 was measured after 12 h at 60 °C (63%

of the borane consumed, entry 8), whereas a more pronounced

activity decay was observed for compound 1 at 0.1 mol % load

(entry 14). Another interesting trend observed in the reactions

conducted at 60 °C, especially at lower catalyst loads, was the

dramatic activity decay after 1 h (i.e., the total TON is already

585 at a catalyst load of 0.05 mol % after 1 h in entry 7),

suggesting catalyst deactivation, which will be commented on

in more detail in the kinetic studies section.

Stoichiometric Reactions with H-BBN and CO₂. In

order to ascertain the transformations undergone by the

catalyst during the hydroboration reaction and identify

possible intermediates or active species involved during the

catalysis, we monitored the reaction using 5 mol %

Ph₂PCHCH₂as the catalyst by multinuclear NMR and were

able to detect several species. Initially, after 20 min under a

CO₂atmosphere (1 atm) at 25 °C, we detected what we

propose to be the diphenyl(vinyl)phosphine−H-BBN adduct,

with signature signals in the ³¹P and ¹¹B NMR at δ_P5.3 ppm

and δ_B−13.7 ppm, respectively, which was also observed as an

intermediate in the preparation of the phosphinoborane 1

reported by Tilley and collaborators.^10cFurther conﬁrmation

of this hypothesis came from the presence of three multiplets

in the ¹H NMR spectrum between δ_H5.1 and 6.5 ppm

attributed to the three vinylic protons. Interestingly, no traces

of the reduction products, the acetal A or the methoxide B,

were detected in the latter spectrum. The sample was then

heated at 60 °C for 1.5 h, depleting most of the borane (89%

consumption by ¹¹B NMR) and forming substantial amounts

of A and B. The initial diphenyl(vinyl)phosphine−H-BBN

adduct had completely disappeared, and the main active

species was the new adduct (H-BBN)Ph₂P(CH₂)₂BBN (1-H-

BBN). The nature of 1-H-BBN was fully conﬁrmed by the

stoichiometric reaction between compound 1 and H-BBN at

room temperature (Scheme 2), slowly reaching an equilibrium

after 24 h, which is clearly shifted toward the adduct (ratio of

ca. 9:1 1-H-BBN/1). The NMR characterization of this

independently prepared solution of 1-H-BBN matched nicely

with the signals observed in the previous spectra during the

catalytic reaction, with those of the ³¹P and ¹¹B NMR spectra,

at δ_P8.4 ppm and δ_B86.1 and −15.1 ppm, the latter two

ascribed to tri- and tetracoordinated boron, respectively, being

the most revealing. The initial sample was further heated at 60

°C for a total time of 2.3 h until full consumption of H-BBN.

At this point, a new intermediate was detected and postulated

as the formaldehyde adduct Ph₂P(CH₂)₂BBN(CH₂O) (2;

Scheme 2), with signals at δ_P−0.4 ppm and δ_B−1.2 ppm,

RESULTS AND DISCUSSION

■

Catalytic Reduction of CO₂with HBBN. As stated above,

we synthesized the phosphinoborane 1 according to a slightly

modiﬁed literature procedure,¹⁰by the direct reaction of

equimolar amounts of Ph₂PCHCH₂and H-BBN in toluene at

100 °C for 3 h. Compound 1 was isolated in very good yields

(up to 76%, ca. 0.5 g), and its NMR characterization matched

well with the reported data in the literature, so no further

comments in this respect will be made (see the Supporting

Information). Because H-BBN reacts directly with

Ph₂PCHCH₂, even at room temperature (slowly), we decided

to test initially both diphenyl(vinyl)phosphine (entries 1−8,

Table 1) and compound 1 (entries 9−14, Table 1) for the

reduction of CO₂with H-BBN. As can be seen in Table 1, the

results and trends found after modiﬁcation of the time,

temperature, and catalyst loads for both catalysts were rather

similar. Thus, in all instances, mixtures of the double and triple

hydroboration products, the acetal CH₂(OBBN)₂(A) and the

methoxide CH₃OBBN (B), were obtained in diﬀerent

proportions, with B being favored at long reaction times and

lower catalyst loadings, in good agreement with the results

reported with phosphines as catalysts.⁸ⁱIn contrast, the

formate derivative, HCO₂BBN, was never detected. As

expected, for the same catalyst load, increasing temperature

also decreased the reaction time (i.e., total consumption of

borane took roughly 1 day at 25 °C but only 3 h at 60 °C at 1

mol % catalyst load). In terms of activity, we noticed slightly

better results with diphenyl(vinyl)phosphine because catalysts

loads down to 0.05 mol % were tolerated and a total turnover

respectively, and a diagnostic doublet at ca. δ_H4.8 ppm (²J_PH

≈

2 Hz) for methylene protons of the formaldehyde unit in the

¹H NMR spectrum. Compound 2 could be selectively prepared

from 1 and p-formaldehyde, and its full characterization in

both solution and solid state will be commented on later.

Finally, right after the addition of more H-BBN and

repressurization with CO₂, we detected a new species with

signature signals at δ_P26.0 ppm and δ_H8.74 and 4.63 ppm,

respectively, assigned to the formate (HCO₂) and form-

aldehyde (CH₂O) units, which could be identiﬁed after

independent stoichiometric reactions (vide infra) as the

macrocyclic compound BBN(CH₂)₂(Ph₂P)(CH₂O)BBN-

(HCO₂)BBN(HCO₂) (4; Scheme 2), with two formate

moieties and one formaldehyde group. Further monitoring of

this reaction for 24 h at room temperature led again to

depletion of the borane and full transformation into the

C

https://dx.doi.org/10.1021/acs.inorgchem.0c01152

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

pubs.acs.org/IC

Article

compound 2 in very good yields (up to 89%, 0.33 g). The

structure of compound 2 was unambiguously established after

X-ray diﬀraction analysis of a single crystal (Figure 1).

Scheme 2. General Reactivity of Ph₂PCHCH₂toward H-

BBN and CO₂

Figure 1. Molecular structure of compound 2 with ellipsoids at 30%

probability level and H atoms omitted for clarity. Only one of the two

crystallographically independent molecules found in the asymmetric

unit is shown. Selected bond lengths (Å) and angles (deg): C1−O1

1.387(3), O1−B2 1.550(3), B2−C3 1.624(3), C3−C2 1.522(3),

C2−P1 1.793(2), P1−C1 1.829(2); C3−B2−O1 106.7(2), C1−P1−

C2 104.9(1), C12−P1−C18 110.0(1), C4−B2−C8 104.8(2), C1−

C2−C3−O1 2.13(8).

̅

Compound 2 crystallized in the triclinic space group P1,

with two crystallographically independent molecules in the

asymmetric unit, essentially with the same parameters, so only

one of them will be discussed. As can be seen, compound 2 is a

six-membered zwitterionic heterocycle with a chair conforma-

tion. The PPh₂and BBN groups are linked together by

ethylene (CH₂CH₂) and formaldehyde (CH₂O) units, which

display roughly a planar arrangement, as denoted by a C1−

C2−C3−O1 torsion angle of only 2.13(8)°, with the P and B

atoms folding out of this plane in an almost antiparallel

disposition. These two atoms also display somewhat distorted

tetrahedral coordination geometries, as expected for phospho-

nium/borate groups. Two new single P−C and O−B bonds

have been formed, as denoted by the corresponding P1−C1

and O1−B2 distances, of 1.829(2) and 1.550(3) Å,

respectively. The chair arrangement of the heterocycle diﬀers

from that of the formaldehyde adduct of o-(C₆H₄)(PPh₂)-

(Bcat),^8lwhich shows a half-chair conformation similar to that

of cyclohexene.

The NMR spectra for compound 2 are consistent with the

solid-state structure allowing the usual dynamic processes, fast

in the NMR time scale, such as free rotation of Ph rings or

chair ﬂipping, which give rise to an apparent symmetry plane

that accounts for the presence of ﬁve signals for the BBN group

and four signals for the Ph groups in the ¹³C NMR spectrum.

The formaldehyde moiety was evident in the ¹H and ¹³C NMR

spectra with the presence of two doublets, at 4.87 and 60.4

ppm (¹J_PC= 47.5 Hz), respectively. The signals in the ³¹P and

¹¹B NMR spectra match well with those found in the catalytic

reaction commented on above and will not be further

discussed. Because formaldehyde adducts of ambiphilic species

have been proven to play a key role as catalysts for the

hydroboration of CO₂, as shown by us (boron amidinate

system)^8aand Fontaine and co-workers (phosphinoboronate

compounds),^8g,lwe also decided to test compound 2 as the

catalyst for hydroboration of CO₂with all of the boranes used

in this study. In the ﬁrst place, the results obtained were similar

to those obtained for compound 1 when H-BBN was used as

the reductant (entries 15−20, Table 1), but shorter reactions

times were often required for reaction completion, especially at

hydroboration products A (minor) and B (major), whereas

compound 4 remained as the only active species detected in

solution. It should also be mentioned that, in a separate

experiment, we tested the lack of reactivity of phosphinobor-

ane 1 toward CO₂, in agreement with the experimental and

computational results reported by Fontaine and co-workers for

related phosphineboronate derivatives.^7,8g,l

Formation of the formaldehyde adduct 2 can be explained

by the reaction of the phosphinoborane 1 with the acetal A

originating from double hydroboration of CO₂with H-BBN

catalyzed by 1 itself. As previously computed by Fontaine,

bisboryl acetal species may be in equilibrium with free

formaldehyde and bisboryl ether.^8lThus, free formaldehyde

would be immediately trapped by the ambiphilic P/B system

(in our case, compound 1) to give a formaldehyde adduct.

Evidence, and subsequent trapping, of transient formaldehyde

was also provided by Sabo-Etienne and co-workers using Ru

catalysts and HBcat as the reductant.¹¹However, a more

convenient and eﬃcient approach to prepare 2, as stated

above, involves the use of p-formaldehyde [CH₂O]_n, a

synthetic route also employed by Fontaine and co-workers to

prepare formaldehyde adducts of phosphinoboronates of the

formula o-(C₆H₄)(PPh₂)[B(OR)₂].^8g,lThus, mild heating (60

°C) of suspensions containing 1 and p-formaldehyde led to

D

https://dx.doi.org/10.1021/acs.inorgchem.0c01152

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

pubs.acs.org/IC

Article

25 °C (only 12 h, entry 15). A more detailed analysis

accounting for these results will be given in the next section,

discussing the kinetic studies.

signals for compound 4 in the ¹H NMR spectrum was roughly

1:1, and a sharp signal in the ³¹P NMR spectrum at 26.0 ppm

was also attributed to 4.

Deﬁnitive structural conﬁrmation for compounds 3 and 4

came from diﬀractometric studies. Fortunately, single crystals

of 3 for an X-ray diﬀraction analysis could be obtained from

the ﬁrst stoichiometric reaction. Compound 3 crystallized in

the monoclinic P2₁/c space group, displaying a polymeric

structure with a helical arrangement and a pseudo-C₂

symmetry (Figures 2 and S32). As can be seen, the structure

The formation of compound 4 formally requires the capture

of two boron formate units derived from the monohydrobora-

tion of CO₂by H-BBN by the formaldehyde adduct 2

(Scheme 2). In order to prove this hypothesis, several

stoichiometric experiments were carried out. In the ﬁrst

place, we mixed equimolar amounts of 2 and H-BBN in a C₆D₆

solution and exposed the mixture to a CO₂atmosphere (ca. 1

bar) at room temperature in an NMR tube. Monitoring the

reaction, we noticed that new signals due to species still

containing a formaldehyde group (δ_Hca. 4.6 ppm) and a

1

formate group (δ_Hca. 8.7 ppm) appeared in the H NMR

spectra, which were also associated with two new broad signals

in the ³¹P NMR spectra at 25.8 and 24.2 ppm. Interestingly,

the integral ratio of the formaldehyde/formate signals in the

¹H NMR spectra was neither 2:1, as would be expected for a

species containing one unit of each group, nor 1:1, as would be

expected for a species with two (equivalent) formate groups

and one formaldehyde. Instead, the relative integral ratio takes

an intermediate value that decreases as the reaction progresses:

2:1.4 after 20 min and 2:1.7 after 15 h. A parallel intensity

increase of the signal at 25.8 ppm with respect to the one at

24.2 ppm was also noticed in the ³¹P NMR spectra. This

evidence led us to propose that up to two species, one bearing

one formate group, BBN(CH₂)₂(Ph₂P)(CH₂O)BBN(HCO₂)

(3), and the other bearing two formate groups, 4, were present

in solution (Scheme 2). Compound 3 progressively crystallized

from the latter mixture from the early stages of the reaction

(vide infra), but its almost negligible solubility after

crystallization precluded solution NMR characterization.

After 15 h, apart from small amounts of 3 and 4, only traces

of compound 2 and the boron methoxide B are present in

solution. The fact that only one set of signals was observed in

the ¹H NMR spectra for the formaldehyde and formate groups

of compounds 3 and 4, together with the presence of two

broad signals in the ³¹P NMR spectra of these two compounds,

might be due to rapid dissociation/coordination equilibria in

solution of the boron formate units both within and between

compounds 3 and 4. Unfortunately, the extremely low

solubility of both species in the usual organic solvents

precluded a variable-temperature NMR study to support our

hypothesis. In order to attest that the formate units came from

CO₂, the experiment was reproduced using ¹³CO₂, and a

Figure 2. Molecular structure of compound 3 (polymeric structure),

with ellipsoids at the 30% probability level and H atoms omitted for

clarity. Selected bond lengths (Å) and angles (deg): B1−O1 1.621(6),

B1−O3 1.495(7), O1−C1 1.264(5), C1−O2 1.256(6), O2−B2

1.630(5), B2−C19 1.608(6), C19−C20 1.518(7), C20−P1 1.801(5),

P1−C10 1.812(5), C10−O3 1.401(6); O1−B1−O3 105.4(4), O1−

C1−O2 120.8(4), O2−B2−C19 104.8(3), C10−P1−C20 106.2(2).

of 3 can be constructed from that of 2 after cleavage of the O

→ B dative bond, followed by the attachment of another BBN

fragment to the O atom (O3 in Figure 2) and a formate group

(HCO₂) to the original BBN moiety present in 2. This would

leave a tricoordinate B atom (B1) with unquenched Lewis

acidity, which accounts for the polymeric structure found in

the solid state by donor−acceptor interactions with the O

atom of a formate unit from a diﬀerent monomer (O1). This

type of polymeric structure in BBN derivatives is not unusual.

In fact, the parent phosphinoborane 1 also forms a polymeric

framework with P−B interactions in the solid state, accounting

as well for its low solubility.^10cAs a consequence of this

polymeric structure and the dative O → B bond formed, there

is a symmetric charge delocalization in the formate unit, with

almost identical C−O bond distances of ca. 1.26 Å, similar to

the values measured for the macrocyclic tetramer

[HCO₂BBN]₄.¹²Interestingly, the B−O distances between

the formate unit and the BBN groups are slightly longer (ca.

1.63 Å) than those found in the mentioned formate tetramer,¹²

the formaldehyde−formate zwitterions R₃P(CH₂O)BBN-

1

doublet in the H NMR spectrum at 8.76 ppm (¹J_CH= 206.9

Hz), together with a broad signal in the ¹³C NMR spectrum at

172.5 ppm, was easily attributed to the H¹³CO₂groups of ¹³C-

3 and ¹³C-4. Small amounts of ¹²CH₃OBBN (B) were also

1

detected in the H and ¹¹B NMR spectra. The absence of an

8i

isotopically enriched product (¹³C-B) suggests that direct

hydroboration of the formaldehyde unit in 2 also competes

with the formation of 3 and 4, but the reaction is much slower

at room temperature, favoring the formation of 3 and 4. In

order to obtain compound 4 selectively, we decided to carry

out a stoichiometric reaction involving 2 equiv of H-BBN,

compound 2 and CO₂. Again, monitoring the reaction by

NMR spectroscopy revealed the progressive formation of

compounds 3 and 4 in solution from the early stages of the

reaction. Finally, after 3 days at room temperature, only

compound 4 and the boron methoxide B remained in solution

alongside a considerable amount of crystalline precipitate of 4.

As expected, the integral ratio between HCO₂and H₂CO

t

(O₂CH) (R = Bu, p-C₆H₄Me) or related BBN carboxylate

derivatives (1.53−1.59 Å).¹³This is indicative of weak O → B

dative interactions¹³and may account for the proposed

formate cleavage and rearrangement equilibria proposed in

solution for 3 and 4.

Crystals of compound 4 were grown from a reaction crude

in toluene at −20 °C under a CO₂atmosphere. It crystallized

in the monoclinic P2₁/c space group as a 14-membered

zwitterionic macrocycle (Figure 3), as opposed to the open-

chain polymeric structure displayed by compound 3. The

tendency of boron formate derivatives to form macrocycles in

the solid state was already evident in HCO₂BBN itself, which

crystallized as a 16-membered macrocycle tetramer.¹²In fact,

E

https://dx.doi.org/10.1021/acs.inorgchem.0c01152

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

pubs.acs.org/IC

Article

concentration of active species (Figures 4, 5, and S25 −S28).

The most outstanding feature in the experiments conducted at

25 °C was the presence of an induction period of ca. 4 h for

the formation of A and B with catalysts Ph₂PCHCH₂and 1.

This period was absent for catalyst 2, which accounts for the

shorter reaction times observed (ca. 12 h) compared to the

other two catalysts (18−20 h). Induction periods for the

hydroboration of CO₂have been also detected with other

catalysts such as phosphines (with H-BBN),⁸ⁱphosphinobor-

onates (with BH₃),^8gand boryl amidinates (with H-BBN).^8aIn

the latter two cases, these periods disappeared using the

corresponding formaldehyde adducts as catalysts. We also

followed the concentration of active species present in solution

during catalysis at 25 °C with diphenyl(vinyl)phosphine and 1.

We noticed that the rapid increase of TON for the formation

of products A and B, which took place after 4 h, coincided with

the formation of compound 4, which started to appear around

the same time. In contrast, the only active species detected

during the ﬁrst 4 h were the adducts diphenyl(vinyl)-

phosphine−H-BBN (using Ph₂PCHCH₂as the catalyst) and

1-HBBN (using 1 as the catalyst). In both cases, compound 2

started to appear at the late stages of the catalytic reaction,

after ca. 14−15 h. By the end of the reaction, mixtures

containing similar amounts of compounds 2 and 4 were

obtained. In contrast, in catalysis with the formaldehyde

adduct 2 at 25 °C, compound 4 was present from the

beginning practically as the only active species during the ﬁrst

1 h, when the activity was higher. After that period, the

concentration of 4 dropped, and mixtures of 1-HBBN, 2, and 4

were obtained. These experiments revealed a correlation

between the concentration of 4 and the activity of the

catalysts. This was further conﬁrmed after plotting instant

turnover frequency (TOF) values for the formation of

products [d(TON A + B)/dt] versus time (Figure 5) because

the highest TOF values were achieved when the concentration

of 4 was highest. Kinetic studies performed at 60 °C for the

mentioned catalysts showed much shorter induction periods

when using compounds 1 and diphenyl(vinyl)phosphine (<5

min), whereas, once more, there was no induction period for

the formaldehyde adduct 2. Again, a correlation between the

concentration of compound 4 and the activity of the catalyst

was conﬁrmed by the TOF versus time plots. However,

compound 4 was a ﬂeeting species in all cases at 60 °C, which

lasted only a few minutes in solution at the beginning of the

reaction, but its presence accounts for total TON values of up

to 40 (i.e., 40% conversion to products) and instant TOF

values (A + B) of up to 680 h⁻¹(with catalyst 2), roughly

matching with the highest concentration of 4 in solution. After

this time, compound 4 disappeared and only the adduct 1-H-

BBN was detected. The activity dramatically decreased in all

three instances up until the end of the monitored reaction

(>90% conversion in all cases). All of these facts suggest an

involvement of the bisformate derivative 4 in the reduction of

CO₂with H-BBN.

Figure 3. Molecular structure of compound 4, with ellipsoids at the

30% probability level and H atoms and solvent molecules (toluene)

omitted for clarity. Selected bond lengths (Å) and angles (deg): B1−

O1 1.668(3), O1−C11 1.236(3), C11−O2 1.267(3), O2−B2

1.550(3), B2−O3 1.571(3), O3−C20 1.251(3), C20−O4 1.243(3),

O4−B3 1.661(3), B3−O5 1.479(3), C29−O5 1.405(3); O5−B3−O4

102.6(2), O4−C20−O3 123.1(2), O3−B2−O2 103.9(2), O2−C11−

O1 121.9(2), O1−B1−C2 102.2(2), C1−P1−C29 109.3(1).

the macrocycle of 4 is made up of two HCO₂BBN units linked

together, trapped by the formaldehyde adduct 2 after cleavage

of the O → B dative bond. All three B atoms (borate groups),

as well as the P atom (phosphonium group), display

tetrahedral coordination environments. There is signiﬁcant

charge delocalization within the formate groups, more

symmetrically in the O4−C20−O3 unit, with distances of

1.243(3) and 1.251(3) Å, whereas there is slightly more

double-bond character for the C11−O1 couple [1.236(3) Å]

than for the other pair, C11−O2 [1.267(3) Å], in the other

formate group. Two of the B−O bonds between two of the

BBN groups and the formate fragments are signiﬁcantly

elongated (B1−O1 and B3−O4 of ca. 1.66 Å), even more so

than those of compound 3, implying even weaker O → B

dative interactions. However, the other two, B2−O2 and B2−

O3, with distances of 1.550(3) and 1.571(3) Å, respectively,

fall in the range found for related carboxylate−BBN

compounds,^8i,12,13which suggest some covalent character.

Again, the weak nature of the dative interactions in B1−O1

and B3−O4 may explain the higher symmetry observed in the

¹H NMR spectra in solution. Assuming facile cleavage of these

dative bonds in solution to form open structures, as well as

formate−BBN intermolecular exchange reactions, fast on the

NMR time scale, may average the two formate chemical

environments, giving rise to a single resonance. We should

note that two structures of formaldehyde−formate zwitterionic

derivatives, both open-chain monomers, were reported by

Wang and Stephan from the reaction of phosphines, H-BBN,

and CO₂.⁸ⁱHowever, the presence of an extra BBN unit in

compound 1 allowed us to isolate and characterize not only the

formaldehyde adduct 2 but also a formaldehyde−monoformate

species with a polymeric structure (3) and, for the ﬁrst time, a

formaldehyde−bisformate compound with a macrocyclic

structure (4). In the next section, kinetic experiments will

shed more light on the role of these species in the catalytic

hydroboration of CO₂with H-BBN.

Some valuable mechanistic information for the hydro-

boration of CO₂using phosphinoborane 1 as the catalyst

could be gathered by combining the results from these kinetic

studies and the stoichiometric reactions commented on before.

In the ﬁrst place, the adduct 1-H-BBN, generated in the

presence of an excess of reactant H-BBN, will react with a

molecule of CO₂like an intermolecular FLP to generate Int-I

(Scheme 3), undetected in the reaction, which leads to free

boron formate, also undetected. Int-I will react in the presence

Kinetic Studies for the Reduction of CO₂with H-BBN.

A series of kinetic studies were carried out for the

hydroboration of CO₂using diphenyl(vinyl)phosphine, 1,

and 2 as catalysts (1 mol %) at 25 and 60 °C. In these

experiments, not only was formation of the reduction products,

acetal A or methoxide B, monitored over time but also the

F

https://dx.doi.org/10.1021/acs.inorgchem.0c01152

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

pubs.acs.org/IC

Article

●

Figure 4. (a, c, e, and g) Total TON (A + B, blue ), and TON for the formation of A (red ) and CH₃OBBN (B, yellow ) versus time (h). (b,

●

d, f, and h) Distribution of the active species 1-H-BBN (blue ), 2 (red ), and 4 (yellow ) versus time for the reduction of CO₂with H-BBN

using catalysts 1 and 2 at 25 and 60 °C (1 mol % catalyst).

of more H-BBN to ﬁnally give compounds A (acetal) and B

(methoxide) in a very slow fashion, hence the induction time

detected. This reactivity resembles that proposed for the

reduction of CO₂with phosphines as catalysts.⁸ⁱAnother

possibility is that compound 1 will act as an intramolecular

FLP, trapping a molecule of CO₂to give zwitterionic

intermediate Int-II. The reaction of Int-II with H-BBN will

also generate free boron formate. In any event, the low

activities displayed by 1 or 1-H-BBN, linked to the initial

induction periods (ca. 4 h at room temperature), indicate that

these are not the main active species for the reduction of CO₂.

Indeed, the activity boost starts with the formation of

macrocycle 4, which originates from compound 2, as we

have already commented on. The formation of 2 can be

explained by the reaction of acetal A (as a source of “free”

formaldehyde) with compound 1, generating bisboryl ether

O(BBN)₂as the byproduct. Once the formaldehyde adduct 2

is formed, an undetected open-chain intermediate must be

obtained after cleavage of the O−B bond in the heterocycle

and subsequent formation of a new one with an incoming H-

G

https://dx.doi.org/10.1021/acs.inorgchem.0c01152

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

pubs.acs.org/IC

Article

−1

●

Figure 5. Plots comparing d(TON A + B)/dt (h ; blue ) and millimoles of 4 (red ) versus time (h) for the catalytic reduction of CO₂with H-

BBN using catalysts 1 (a and c) and 2 (b and d) at 25 and 60 °C (1 mol % catalyst).

give formate derivative 3 as a monomer. This reaction pathway

for the generation of formate derivatives would be much faster

than the previous ones (through Int-I and Int-II) because of

the enhanced hydridic character of zwitterionic Int-III. In fact,

it is known that borohydrides can react with CO₂: Mizuta and

co-workers proved that small amounts of NaBH₄catalyze the

reduction of CO₂to the methanol level using BH₃as the

reductant,¹⁴whereas Cummins and co-workers showed that

the stoichiometric reaction of NaBH₄with 3 equiv of CO₂

generates the trisformate compound Na[HB(OCHO)₃].¹⁵

Further reaction of 3 with more H-BBN and CO₂will give

bisformate 4, detected during catalysis at both 25 and 60 °C.

Because free HCO₂BBN was not detected, we propose that

adducts 3 and 4 “trap” formate molecules, enhancing their

reactivity toward the second hydroboration reaction to give

acetal A and regenerate the formaldehyde adduct 2. Another

possible pathway, as evidenced by NMR and the solid-state

structures of 3 and 4, would consist of fast equilibria between

compounds 2−4 facilitated by fast boryl formate dissociation/

coordination equilibria proposed for 3 and 4. This would imply

an active role for 2 as a catalyst for the hydroboration of boryl

formate, directly leading to acetal A through equilibria with

formate−formaldehyde species 3 and 4. In this sense, Fontaine

and co-workers calculated plausible formate adducts (un-

detected under experimental conditions) in fast equilibrium

with the parent catalyst using o-(C₆H₄)(PPh₂)(Bcat) in their

mechanistic study of the hydroboration of CO₂using HBcat.^8l

In this reaction, the corresponding boryl formate was also

undetected and assumed to react very fast with further

equivalents of HBcat due to the action of the catalyst. One

could assume that, in the early stages of the reaction with

compound 2 as the catalyst, p(CO₂) would still be high

enough in the NMR tube so that CO₂diﬀusion in the solution

would not be the limiting step and the transformation of 2 into

4, as well as the formation of acetal A and methoxide B, would

Scheme 3. Proposed Mechanism for the Reduction of CO₂

with H-BBN Using Compound 1 as the Catalyst

BBN molecule (Int-III in Scheme 3). A similar intermediate

was computed for the hydroboration reaction of CO₂using the

formaldehyde adduct of a boron amidinate as the catalyst.^8a

Then, hydride attack of Int-III toward a CO₂molecule will

H

https://dx.doi.org/10.1021/acs.inorgchem.0c01152

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

pubs.acs.org/IC

Article

a

be fast. Alternatively, the presence of thermodynamically

favored formate adduct 4 could be viewed as a consequence of

the high activity of the formaldehyde adduct 2 as the active

catalyst in the ﬁrst and second hydroboration reactions.

However, when the pressure of CO₂drops and most of the

borane is consumed at 60 °C, we conjectured that the

formation of compound 4 would be less favored and the

competing formaldehyde reduction reaction between 2 and H-

BBN would be operative instead, giving methoxide B and

compound 1, which, in the presence of excess H-BBN, would

be mainly in the form of adduct 1-H-BBN. In contrast, the

concentration of this adduct in the late stages of the reaction at

25 °C is almost negligible, meaning that the reduction of the

formaldehyde unit of 2 is less favored at this temperature, and

mixtures of 2 and 4 are obtained at these stages instead. In

order to further verify the existence of this formaldehyde

hydroboration reaction between 2 and H-BBN, NMR-scale

kinetic experiments were carried out using ¹³CO₂and a higher

catalyst load of compound 2 (2.5 mol %) to better monitor the

formation of catalytic intermediates at 25 and 60 °C (Figures

S29 and S30). In both experiments, we detected the formation

Table 2. Catalytic Hydroboration of CO₂with HBcat

cat. load

(mol %)

T

conv

b

c

entry

catalyst

(°C) t (h)

(%)

TON C

1

2

3

4

5

6

7

8

Ph₂PCHCH₂

1

60

14

0.5

1

15

9

0.5

1

15

9

1

2

2.5

1

44

70

10.7

22.0

1

90

73.2

0.1

2.5

1

0.1

66

88

71

23.2

32.0

60.6

82.5

9

10

11

100

a

Reaction conditions: J. Young valve NMR tube containing ca. 0.6

mL C₆D₆solutions of HBcat (0.2 mmol), catalyst and internal

b

standard [Si(SiMe₃)₄, 0.01 mmol], and CO₂(1 atm). Based on the

relative integrals of HBcat and products in the ¹¹B NMR spectra.

1

of B in the H NMR spectra, which can only be obtained by

c

Calculated according to the number of C−H bonds formed in the

reduction products by integration of the corresponding signals relative

the direct reaction of H-BBN with the formaldehyde group of

2, and ¹³C isotopic enrichment in the active species 4.

Compound ¹³C-4 was identiﬁed by the presence of a doublet

of doublets for the formaldehyde fragment in the region 4.5−

5.0 ppm in the ¹H NMR spectra with a high ¹J_CHcoupling (ca.

149 Hz) and doublets at δ_Pca. 26 ppm and δ_Cin the range

55−58 ppm in the ³¹P and ¹³C NMR spectra, respectively, with

¹J_PCof ca. 70 Hz. However, the ¹³C isotopic enrichment in the

formaldehyde unit of 4 in the experiment at 25 °C was only ca.

30% after 14 h, whereas it was ca. 100% after 15 min at 60 °C,

the time required to full conversion in both cases. This is in

line with our previous results showing that direct hydro-

boration of the formaldehyde group of 2 takes place very

slowly at 25 °C compared to the formation of acetal (A) or

methoxide (B). It is worth noting that Fontaine and co-

workers performed similar isotopic labeling experiments with

diﬀerent outcomes: using the formaldehyde adducts of

phosphinoboronates o-C₆H₄(PR₂)(Bcat) as catalysts, HBcat

as the reductant, and ¹³CO₂at 70 °C, they did not ﬁnd ¹³C

incorporation in the catalyst, meaning that the formaldehyde

unit remained unscathed during catalysis.^8g

Catalytic Reduction of CO₂with HBcat. We also carried

out the hydroboration of CO₂using HBcat to give MeOBcat

(C) with compounds 1 and 2 as catalysts at 60 °C (Table 2).

We did not detect the presence of other reduction products in

these reactions, such as formate or acetal derivatives, which is

in line with the results reported by other authors using HBcat

for the catalytic hydroboration of CO₂.^6,7,8g,h,lPh₂PCHCH₂

was also tested as the catalyst in a control experiment (entry

1). As can be seen, no reaction occurred at 60 °C for 14 h with

1 mol % catalyst load, which was not surprising regarding the

lesser Lewis acidity of HBcat with respect to H-BBN. Thus, it

seems evident from this experiment that (i) an intermolecular

FLP between the phosphine and HBcat is not active for CO₂

hydroboration as opposed to the intramolecular FLP systems

o-(C₆H₄)(PPh₂)[B(OR)₂] reported by Fontaine;^7,8g,l(ii)

direct hydroboration of the vinyl fragment of Ph₂PCHCH₂

does not take place under the reaction conditions, as expected

for less reactive dialkoxyboranes;¹⁶otherwise, we would expect

the formation of an intramolecular FLP akin to that described

by Fontaine that should be active in the hydroboration of CO₂.

1

to the internal standard in the H NMR spectra.

As for compounds 1 and 2, they turned out to be rather active

for this transformation but less active than in the reductions

with H-BBN, as indicated by the TON values, only up to 82.5

for compound 2 (1 mol %, entry 10) and 73.2 for compound 1

under the same conditions (entry 5). Moreover, catalyst loads

down to 0.1 mol % were not tolerated (entries 6 and 11). The

main diﬀerence observed between the catalytic activities of 1

and 2 was again related to the induction periods at early stages

of the transformation. Thus, at a catalyst load of 1 mol %, there

was no trace of the reduction product C after 1 h of using 1 as

the catalyst (entry 4), whereas a TON of 60.6 was already

measured with catalyst 2 at the same time (entry 9). This

induction period disappeared with increasing catalyst load (2.5

mol %), although compound 2 was still slightly more active

than 1.

Stoichiometric Reactions with HBcat and CO₂. During

one of the catalytic runs with compound 2 as the catalyst (1

mol %), we detected only one P-containing species, displaying

a sharp singlet ca. 24 ppm in the ³¹P{¹H} NMR spectrum, after

1

1 h at 60 °C. Signals in the H NMR spectrum at ca. 8.6 and

4.8 ppm suggested another formaldehyde−formate derivative

similar to 3 and 4. However, after 15 h at the same

temperature, a very broad signal was now observed in the

³¹P{¹H} NMR spectrum, also at ca. 24 ppm, and no downﬁeld

signals in the range expected for formate protons were present

in the ¹H NMR but only a singlet at δ_H4.9 ppm attributed to a

formaldehyde fragment. In an attempt to trap any of these

intermediates, we ﬁrst carried out the reaction between

equimolar amounts of compound 2 and HBcat. Interestingly,

a new species was identiﬁed, with a very broad signal in the

³¹P{¹H} NMR spectrum at ca. 17 ppm and a singlet in the ¹H

NMR at 5.0 ppm ascribed to the formaldehyde fragment. This

species is possibly an adduct involving interaction between the

O atom of the formaldehyde group of 2 and the B center of the

HBcat molecule. Unfortunately, this compound could be

neither isolated nor crystallized for structural conﬁrmation.

However, Fontaine and co-workers also detected P-containing

I

https://dx.doi.org/10.1021/acs.inorgchem.0c01152

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

pubs.acs.org/IC

Article

Scheme 4. Preparation of Compound 5

species with resonances at δ_Pca. 14 ppm in the reaction of the

formaldehyde adduct of o-C₆H₄(PⁱPr₂)(Bcat) with HBcat in

the absence of CO₂.^8gAlthough the mentioned species could

not be isolated, they proposed it to be a Lewis adduct between

one of the O atoms bound to B and HBcat, as well.

the catalytic reduction of CO₂with HBcat using catalyst 2.

Mechanistically speaking, the formation of 5 would proceed via

Lewis adduct formation between 2 and HBcat, detected

previously in the stoichiometric reaction, which is presumably

an open-chain species similar to Int-III in Scheme 3. The

adduct 2−HBcat, in the presence of CO₂, will give a formate

derivative akin to monomer 3 in Scheme 3 with the formate on

the BBN moiety. At this point, the formate−formaldehyde

species would be trapped by another molecule of 2 to give 5,

instead of reacting with more borane and CO₂to give a

compound similar to 4. This is likely due to the lesser Lewis

acidity of HBcat compared to H-BBN.

Compound 5 was also detected in the sluggish reaction

between compound 1, HBcat (2 equiv), and CO₂at room

temperature, together with considerable amounts of methoxide

C (ratio of ca. 1:1 5/C). This indicates that a mechanism

similar to that described for the reduction of CO₂with H-BBN

and 1 as the catalyst (vide supra) is also operative. Thus, the

undetected formaldehyde adduct 2 must be initially formed in

the reaction of phosphinoborane 1 and the acetal formed in

the double hydroboration of CO₂with HBcat [CH₂(OBcat)₂],

which, in the presence of excess HBcat and CO₂, gives rise to

compound 5.

The structure of compound 5 was unambiguously conﬁrmed

by an X-ray diﬀraction study of a single crystal. This 16-

membered zwitterionic macrocycle crystallized in the ortho-

rhombic system Pnma (Figure 6). It consists of two molecules

of the open-chain isomer of the formaldehyde adduct 2 after

cleavage of the dative O → B bond, symmetrically trapping a

formate unit (through two new O → B bonds between the O

atoms of the formate and the BBN groups) and a Bcat moiety

(through the O atoms of the formaldehyde units). The C₂

symmetry of this macrocycle is also evident in the asymmetric

unit, as only half the molecule is present and the other half can

be built up by C₂rotation. Thus, there is symmetrical charge

delocalization within the formate group, with C11−O2/O2′

distances of 1.250(4) Å. The B−O distances between the Bcat

fragment and formaldehyde groups, of 1.445(5) Å, are slightly

shorter than that measured in the formaldehyde adduct o-

C₆H₄(PPh₂)(Bcat)(CH₂O), of 1.473(3) Å,^8land consistent

with covalent B−O bonds. In contrast, the B−O distances

between the BBN fragments and formate group, of 1.612(3) Å,

can be regarded as true dative O → B interactions (>1.6 Å).¹³

Kinetic Studies for the Reduction of CO₂with HBcat.

Kinetic studies were performed at 60 °C using compounds 1

and 2 as catalysts (1 mol %), following the formation of the

methoxide product C (Figure 7). An induction period of ca. 1

h was clearly observed with catalyst 1, similar to that found in

the reduction with H-BBN with the same catalyst. This was

already evident in the catalytic runs commented on before (see

Additionally, we carried out the reaction of the form-

aldehyde adduct 2 and HBcat (1 equiv) under 1 atm of CO₂,

yielding a new macrocyclic derivative formulated as (HCO₂)-

{BBN(CH₂)₂(Ph₂P)(CH₂O)}₂Bcat (5) and small amounts of

methoxide C after only 15 min at room temperature (Scheme

4). Compound 5 crystallizes after a few hours under a CO₂

atmosphere from the reaction crude. The formation of 5

follows from the reaction of two molecules of 2 with one

molecule of both HBcat and CO₂. Interestingly, when we

carried out the reaction with higher HBcat/2 ratios (up to 4:1)

in the presence of CO₂, compound 5 was still the only P-

containing species detected, with increasing amounts of

methoxide C as a side product. Some structural evidence for

compound 5 was already found in the solution NMR spectra of

reaction crudes: the relative integrals of the methylene protons

of the two CH₂O groups (δ_H4.72 ppm) and that of the

1

formate group (δ_H8.79 ppm) in the H NMR spectrum was

4:1, in agreement with the stoichiometry shown in Scheme 4.

The latter groups also gave rise to signature signals in the ¹³C

NMR spectrum: a doublet at 56.0 ppm (J_PC= 72.1 Hz) and a

singlet at 173.9 ppm, attributed to the formaldehyde groups

and formate fragment, respectively. The chemically equivalent

P atoms in 5 displayed a singlet in the ³¹P{¹H} NMR spectrum

at 23.7 ppm, whereas two broad signals in the ¹¹B NMR

spectrum at 19.4 (1B) and 10.8 (2B) ppm were assigned to the

Bcat and two BBN fragments, respectively. To conﬁrm that the

formate group originated from the hydroboration of CO₂with

HBcat, an isotopic labeling experiment was carried out using

¹³CO₂, HBcat, and compound 2. In situ NMR characterization

of the reaction mixture conﬁrmed the presence of (¹³CHO₂)-

{BBN(CH₂)₂(Ph₂P)(CH₂O)}₂Bcat (¹³C-5), with a diagnostic

1

doublet in the H NMR spectrum at 8.79 ppm with a J_CHof

207.2 Hz and a sharp singlet at 173.9 ppm for the isotopically

enriched H¹³CO₂group. We should note tha small amounts of

both ¹²CH₃OBcat (C) and ¹³CH₃OBcat (¹³C-C) were also

1

detected in the H NMR spectrum of the reaction crude. The

unlabeled methoxide C must come from direct hydroboration

of the formaldehyde fragment in 2. This reaction competes

with the formation of 5, although it is less favored at room

temperature. On the other hand, the isotopically enriched ¹³C-

C is a genuine byproduct of the hydroboration of ¹³CO₂

mediated by compound 2. From all of the latter ﬁndings, we

propose that formation of the macrocyclic compound 5 must

be thermodynamically favored, even under an excess of borane,

and is therefore the intermediate detected in the early stages of

J

https://dx.doi.org/10.1021/acs.inorgchem.0c01152

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

pubs.acs.org/IC

Article

values reached with catalyst 1 compared to those reached with

2 (maximum TOF values of 15 h⁻¹for 1 vs 325 h⁻¹for 2;

Figure 7b).

Catalytic Reduction of CO₂with HBpin and BH₃·SMe₂.

HBpin was also used for the catalytic hydroboration of CO₂

with compounds 1 and 2 as catalysts (Table 3). Overall, the

activity shown by these catalysts was lower than that with the

other boranes because of the lesser reactivity of HBpin, as was

already observed with other FLP-like P/B catalysts.^7,8hIn fact,

much longer reaction times (days) were required to reach

conversions of >85% under reaction conditions similar to those

employed with previous hydroboranes (see, for example,

entries 7 and 12). Moreover, mixtures of the mono-, di-, and

trihydroboration products, HCO₂Bpin (D), CH₂(OBpin)₂

(E), and CH₃OBpin (F), were obtained in most cases. This

lack of selectivity in the reactions of HBcat with CO₂was also

observed not only with other metal-free P/B catalysts^8hbut

also with transition-metal-based catalysts under certain

conditions^11a,c,12and is probably related to the lesser reactivity

and higher steric hindrance of HBpin compared to HBcat,

hampering the second and third hydroboration reactions. A

control experiment with diphenyl(vinyl)phosphine (1 mol %)

was also conducted (entry 1) and only trace amounts of

formate D were detected after 14 h at 60 °C. Similar results in

terms of activity and product distribution were obtained with

catalysts 1 and 2 at 60 °C; i.e., conversions of 33 and 38% were

achieved after 1 h at 60 °C employing 1 mol % catalyst load of

1 and 2, respectively (entries 5 and 13). As expected,

increasing catalyst load also increased conversion; for example,

under the same reaction conditions as before but using 2.5 mol

% catalyst loads of 1 and 2, 62% conversion was obtained in

both cases (entries 3 and 9). Likewise, increasing reaction time

at 60 °C led to higher selectivity for the methoxide product F

(entry 7). However, there are subtle changes in the activity of 1

and 2 at 25 °C, more speciﬁcally at short reaction times,

because compound 1 only produced a trace amount of formate

D (TON of 0.4 after ca. 20 min, entry 4), whereas compound

2 was much more active, also selectively generating formate D

(TON value of 15.9 in ca. 10 min). This behavior is

reminiscent of that found with the previous boranes, for

which induction times were observed with catalyst 1 that were

not present at all with 2.

Figure 6. Molecular structure of compound 5, with ellipsoids at the

30% probability level and H atoms and solvent molecules (toluene)

omitted for clarity. Selected bond lengths (Å) and angles (deg): C2−

B1 1.626(4), B1−O2 1.612(3), O2−C11 1.250(4), P1−C1 1.792(2),

P1−C12 1.807(2), C12−O5 1.391(3), O5−B2 1.445(5), B2−O3

1.486(7), B2−O4 1.495(3); C2−B1−O2 104.4(2), O2−C11−O2′

121.4(6), C1−P1−C12 108.6(1), O5−B2−O5′ 103.0(5).

entry 4 in Table 2). After that time, there was a moderate

activity boost to ﬁnally reach TON values of 61 after 16 h (ca.

70% conversion). Conversely, catalyst 2 was highly active from

the beginning, already reaching a TON value of 50 in the ﬁrst

45 min. The activity decayed signiﬁcantly after the ﬁrst 1 h to

reach a TON of 83 after 13.3 h (ca. 90% conversion). In order

to account for these results, we tried to monitor the

concentrations of the intermediates and active species during

catalysis. Pleasingly, we were able to establish a correlation

again between the concentration of the bisformaldehyde−

formate species 5 and the high initial TON (and TOF) values

for catalysis with compound 2 (Figure S31c). In fact, total

consumption of 5 occurred after 45 min at 60 °C, ﬁtting nicely

with the period of highest activity in the catalysis.

Unfortunately, we were not able to identify other intermediates

during the catalysis with 2. As for catalyst 1, we were not able

to detect any of the intermediates identiﬁed in the

stoichiometric studies, so the activity boost after 1 h could

not be associated with compound 5 this time. We have

previously established that the stoichiometric reaction of

compound 1 with HBcat in the presence of CO₂led to

mixtures of compounds 5 and C at room temperature. It seems

that the formation of compounds 2 and 5 is not favored under

catalytic conditions at 60 °C, or the latter are too short-lived to

be detected by NMR. This would also explain the lower TOF

We tried to follow the formation of possible active species or

resting states during the catalytic hydroboration of CO₂with

HBpin with catalysts 1 and 2. In the ﬁrst case, several peaks

were observed in the ³¹P NMR spectrum after 1 h at 60 °C,

but only one, at δ_P−11 ppm, remained after 48 h, which is

probably ascribed to a resting state of the catalyst or a

Figure 7. (a) TON for the formation of C versus time (h). (b) d(TON C)/dt (h⁻¹) versus time (h) for the reduction of CO₂with HBcat using

●

catalyst 1 (blue ) and catalyst 2 (red ) (1 mol % catalyst).

K

https://dx.doi.org/10.1021/acs.inorgchem.0c01152

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

pubs.acs.org/IC

Article

a

Table 3. Catalytic Hydroboration of CO₂with HBpin

b

c

entry

catalyst

cat. load (mol %)

T (°C)

t (h)

conv (%)

TON D

TON E

TON F

1

2

3

4

5

6

7

8

Ph₂PCHCH₂

1

60

25

60

25

60

14

0.5

1

0.3

1

3

45

62

<1

33

60

94

50

62

23

41

87

38

71

1.5

9.2

7.9

0.4

10.9

4.1

1

2

2.5

1

4.3

7.4

4.2

7.4

10.1

33.0

3.7

8.9

21.9

88.2

8.3

5

1

48

0.5

1

0.2

2

96

1

15

2.5

1

4.7

5.6

15.9

27.3

7.2

5.7

9

7.8

9.5

10

11

12

13

14

2.4

13.3

10.4

26.2

49.0

10.6

42.3

1

11.4

a

Reaction conditions: J. Young valve NMR tube containing ca. 0.6 mL C₆D₆solutions of HBpin (0.2 mmol), catalyst and internal standard

b

c

[Si(SiMe₃)₄, 0.01 mmol], and CO₂(1 atm). Based on the relative integrals of HBcat and products in the ¹¹B NMR spectra. Calculated according

to the number of C−H formed in the reduction products by integration of the corresponding signals relative to the internal standard.

Scheme 5. Stoichiometric Reaction between 2, HBcat, and CO₂

a

Table 4. Catalytic Hydroboration of CO₂with BH₃·SMe₂

b

c

entry

catalyst

cat. load (mol %)

T (°C)

t (h)

conv (%)

TON G

TON H

1

2

3

4

5

6

7

Ph₂PCHCH₂

2

1

0.1

25

60

25

60

14

1

0.2

0.5

1

2

3

7.7

62

65

86

95

13

161.0

201.2

254.0

285.8

373.0

15.8

17.5

21.0

22.0

20.0

1

a

Reaction conditions: J. Young valve Schlenk ﬂask with a stir bar, containing ca. 0.8 mL C₆D₆solutions of BH₃·SMe₂(0.5 mmol), catalyst and

b

internal standard [Si(SiMe₃)₄, 0.01 mmol], and CO₂(1 atm). Based on the relative integrals of BH₃·SMe₂and products in the ¹¹B NMR.

c

Calculated according to the number of C−H bonds formed in the reduction products by integration of the corresponding signals relative to the

internal standard.

deactivated species. The nature of this species could not be

ascertained because it was not detected in subsequent

stoichiometric reactions. In contrast, a single peak was detected

in the ³¹P NMR spectrum, at δ_P26 ppm after 0.2 h at 25 °C

using catalyst 2. When we carried out the stoichiometric

reaction between equimolar amounts of HBpin and compound

2, in the presence of CO₂(1 atm), we obtained, after a few

minutes, a major product with a signature singlet at δ_P26 ppm

in the ³¹P NMR spectrum that we formulate as the formate−

formaldehyde compound (HCO₂)BBN(CH₂)₂(Ph₂P)-

(CH₂O)Bpin (6) and small amounts of acetal E, methoxide

F, and unreacted 2 (Scheme 5). The match in the ³¹P NMR

spectrum between compound 6 and the species detected in the

reaction with catalyst 2 at 25 °C led us to assume that both are

L

https://dx.doi.org/10.1021/acs.inorgchem.0c01152

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

pubs.acs.org/IC

Article

the same species. However, compound 6 turned out to be less

stable than its BBN or Bcat counterparts 3−5 because small

amounts of 2 were always present in the NMR spectra even

after workup attempting to isolate it as a pure species. Crystals

of 6 suitable for a diﬀractometric analysis could not be

obtained either, but the NMR data collected allowed us to

conﬁdently conﬁrm the monoformate−monoformaldehyde

nature of this compound. Thus, the integral ratio between

the formate proton [δ_H8.95 ppm (C₆D₆) or 8.32 ppm

(CD₂Cl₂)] and the methylene protons of the formaldehyde

[δ_H4.99 ppm (C₆D₆) or 5.11 ppm (CD₂Cl₂)] was roughly 1:2,

in good agreement with the proposed structure. The rest of the

NMR data were also in line with our proposal, with the most

outstanding features being two broad signals observed in the

¹¹B NMR spectrum, at δ_B22.3 and 1.9 ppm, in the expected

regions for a tricoordenate borate ester (OBpin) and a

tetracoordinate borate (CH₂)(O)BBN fragment. Thus, it

seems that the higher steric demands of the Bpin group

probably preclude further reaction with another molecule of 2

to give a macrocycle akin to 5.

Finally, we also tested BH₃·SMe₂for the catalytic hydro-

boration of CO₂with catalysts 1 and 2. As commented on in

the Introduction, Fontaine and co-workers used this reductant

for the ﬁrst time for this type of reaction,⁷employing

phosphinoboronates as catalysts, with excellent results in

terms of activity. Because initial tests at NMR scale showed

poor conversions, we decided to use Schlenk ﬂasks equipped

with J. Young stoppers instead to make sure CO₂was not the

limiting reagent. First, the control experiment with diphenyl-

(vinyl)phosphine showed negligible activity at 25 °C for 14 h

(entry 1, Table 4). Likewise, catalyst 1 turned out to be almost

inactive even at 60 °C (entry 2). This is not at all surprising

because two of the phosphinoboronates of the formula o-

(C₆H₄)(PPh₂)[B(OR)₂] reported by Fontaine also showed

little to no activity in the catalytic reduction of CO₂with

BH₃.^8gIn contrast, the formaldehyde adduct 2 proved to be

very active for this transformation to yield mixtures of

methoxide products [(OMe)BO]₃(G, major product) and

B(OMe)₃(H, minor product; entries 3−7). As expected,

increasing temperature led to higher activities, rising

conversions from 62% at 25 °C to 95% at 60 °C after 1 h

with a catalyst load of 1 mol % (entries 3 and 6). Higher

activities, in terms of TOF values, were measured at shorter

reaction times. Thus, total TON values of 218.7 and 305.8

were calculated after 0.2 and 1 h at 60 °C, respectively (average

TOF values of 1093.5 and 305.8 h⁻¹, entries 4 and 6). Catalyst

loads down to 0.1 mol % were tolerated, keeping the same

concentration of borane but with lower conversions (13%) and

total TON values of 393 after 1 h at 60 °C (entry 7).

agreement with the structural proposal for this adduct. One of

the most diagnostic signals of this species was a very broad

resonance in the range 2.2−1.3 ppm in the ¹H NMR spectrum,

assigned to the BH₃unit, which became a sharp singlet at δ_H

1

1.83 ppm in a H{¹¹B} NMR experiment. A broad multiplet

1

due to J(³¹P−¹¹B) was also observed at δ_P19.3 ppm in the

³¹P{¹H} NMR spectrum, matching with that observed in the

catalytic reaction. The ¹¹B and ¹¹B{¹H} NMR spectra were

also very informative, displaying both a broad downﬁeld signal

at δ_Bca. 85 ppm, in the region expected for a tricoordinate

boron of a CH₂BBN fragment, and an upﬁeld quadruplet of

doublets (¹¹B) at δ_B−38.6 ppm for the tetracoordinate boron

1

of the BH₃moiety, which turned into a doublet in the H-

decoupled experiment (¹J_BP= 57 Hz). The latter signals were

also detected as low-intensity resonances in the ¹¹B NMR

spectrum of the catalytic experiment.

We also carried out other stoichiometric experiments in an

attempt to trap or detect intermediates relevant to the catalytic

reduction of CO₂with BH₃. In the ﬁrst place, the reaction of

compound 2 (0.07 mmol) with BH₃·SMe₂(0.10 mmol) in the

presence of CO₂was conducted in an NMR tube. The NMR

spectra recorded immediately after exposure to a CO₂

atmosphere showed complete conversion of the borane to

reduction products G and H in a 2:1 ratio, as well as a mixture

of compounds 2 and 1-BH₃, which obviously follows from

reaction of the formaldehyde group with BH₃to give 1 and

subsequent formation of a Lewis adduct between 1 and

another molecule of BH₃. Because we were not able to detect

other intermediates, we decided to monitor the reaction

between 2 and BH₃·SMe₂in the absence of CO₂. We noticed

that complete reduction of the formaldehyde unit in 2 to give

1-BH₃takes ca. 3 h, generating small amounts of reduction

products, G and H, but in an inverse ratio, 1:2, together with

unreacted borane. At short reaction times (10 min), we were

able to detect a plausible Lewis adduct between 2 and BH₃,

most likely by interaction between the O atom in 2 and the B

center of BH₃. Some diagnostic resonances support this

1

formulation, such as a singlet at 4.75 ppm in the H NMR

spectrum, ascribed to the methylene protons of the CH₂O unit

or a multiplet in the ³¹P{¹H} NMR spectrum at 24.1 ppm,

resembling that found for 1-BH₃, suggesting ³¹P−¹¹B coupling.

Moreover, this adduct was the major P-containing species after

10 min, with small amounts of 2 and 1-BH₃. Finally, we also

carried out the reaction between 1, BH₃·SMe₂(molar ratio

1:3), and CO₂. As expected, the formation of methanol-level

products G and H was practically negligible after 1 h at room

temperature, in stark contrast to the experiment with

compound 2. However, after that time, reduction products

started to form sluggishly, yielding a mixture of G and H in a

10:1 ratio after 22 h. The only P-containing species detected

throughout the reaction was adduct 1-BH₃. From all of these

stoichiometric experiments, we can deduce that 1-BH₃is not

completely inactive for the catalytic reduction of CO₂with

BH₃but is much less active than compound 2. We also

observed that, once more, reduction of the formaldehyde unit

in 2 was a competitive reaction in this transformation, which

transformed 2 into the almost inactive 1-BH₃, detected as the

resting state of the catalyst at the end of the catalysis. A

plausible Lewis adduct formed between 2 and BH₃was

detected as an intermediate in the stoichiometric reactions

conducted but not isolated because of its ﬂeeting existence in

solution.

In our case, not only was the formaldehyde adduct 2 a very

active catalyst for the reduction of CO₂with BH₃, but also the

phosphinoborane 1 turned out to be essentially inactive under

the same conditions, ruling out an FLP-type mechanism in

which the CO₂molecule is trapped by the phosphinoborane

and the borane reduces the activated CO₂molecule. After

analyzing the NMR spectra at the end of the catalytic run in

entry 6, we noticed a major P-containing species in the ³¹P

NMR spectrum, displaying a broad signal at δ_P19.3 ppm,

attributed to the Lewis adduct between phosphinoborane 1

and BH₃(1-BH₃). In order to conﬁrm this hypothesis, we

conducted the reaction between compound 1 and BH₃·SMe₂

to obtain the expected adduct, (BH₃)Ph₂P(CH₂)₂BBN (1-

BH₃), in high yields. The NMR spectra of 1-BH₃are in good

M

https://dx.doi.org/10.1021/acs.inorgchem.0c01152

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

pubs.acs.org/IC

Article

Accession Codes

CONCLUSIONS

■

CCDC 1995178−1995181 contain the supplementary crys-

tallographic data for this paper. These data can be obtained

free of charge via www.ccdc.cam.ac.uk/data_request/cif, or by

emailing data_request@ccdc.cam.ac.uk, or by contacting The

Cambridge Crystallographic Data Centre, 12 Union Road,

Cambridge CB2 1EZ, UK; fax: +44 1223 336033.

Overall, in this paper we have proven that the formaldehyde

adduct of phosphinoborane 1 and its formaldehyde−formate

derivatives play an essential role in the catalytic hydroboration

of CO₂with four diﬀerent hydroboranes, using 1 as the

catalyst. Thus, the a priori more obvious intramolecular FLP

pathway for the direct reduction of a CO₂molecule once

trapped by the phosphinoborane turns out to be a less active

route. This usually leads to induction periods that can last for

hours at room temperature, the time required for the sluggish

formation of enough amounts of acetal, the source of free

formaldehyde, to react with phosphinoborane to give the

formaldehyde adduct 2. In fact, the main operative pathway for

the catalytic hydroboration of CO₂involves this adduct, which

forms formaldehyde−hydroborane intermediates, detected in

some instances but not isolated in this work. The enhanced

hydridic character of these adducts enables the ﬁrst hydro-

boration of CO₂to boryl formate, which always requires a

catalyst. Free boryl formate, undetected in reactions with H-

BBN and HBcat, is readily trapped by the formaldehyde

adduct to form formaldehyde−formate derivatives, some of

them fully characterized for the ﬁrst time in this work, which

mediate in the hydroboration of boryl formate to bisboryl

acetal. Deactivation pathways often involve reduction of the

formaldehyde moiety in the late stages of the catalysis when

most of the borane and/or CO₂have been consumed, thus

regenerating the nearly inactive phosphinoborane catalyst.

In terms of the catalyst activity (TON and TOF values

measured per the number of C−H bonds formed in the

products) with diﬀerent secondary boranes tested under the

same reaction conditions (NMR scale), the order goes as

follows: H-BBN > HBcat > HBpin. This is indeed the expected

order, taking into account the reactivity of the boranes

themselves. The only diﬀerences observed using phosphino-

borane or its formaldehyde adduct as the catalyst with these

boranes are the induction periods at the beginning of the

reactions, detected for the former, thus reducing the activity in

the early stages of the catalysis. In terms of selectivity, the

methoxide derivative is the only product detected in the

catalysis with HBcat, probably because of the lesser steric

demands of this borane, whereas mixtures of acetal and

methoxide are obtained for H-BBN. Reductions with the less

active and bulky HBpin often lead to mixtures of the three

products (formate, acetal, and methoxide), but at short

reaction times, the formate can be obtained almost selectively

with the formaldehyde adduct as the catalyst. BH₃·SMe₂, tested

under diﬀerent conditions (Schlenk scale), shows an activity

comparable to that of H-BBN but only when the formaldehyde

adduct is used as the catalyst because the phosphinoborane 1

shows almost negligible activity. A mixture of two methoxide-

level products is always obtained with this active borane.

AUTHOR INFORMATION

Corresponding Author

■

́

Alberto Ramos − Departamento de Quimica Inorganica,

́

Organica y Bioquimica, Centro de Innovacion en Quimica

Avanzada (ORFEO-CINQA), Universidad de CastillaLa

Mancha, E-13071 Ciudad Real, Spain; orcid.org/0000-

0001-7993-7864; Email: Alberto.Ramos@uclm.es

Authors

Antonio Antin

́

olo − Departamento de Quimica Inorganica,

̃

́

Organica y Bioquimica, Centro de Innovacion en Quimica

Avanzada (ORFEO-CINQA), Universidad de CastillaLa

Mancha, E-13071 Ciudad Real, Spain; orcid.org/0000-

0002-4417-6417

́

Fernando Carrillo-Hermosilla − Departamento de Quimica

́

Inorganica, Organica y Bioquimica, Centro de Innovacion en

́

Quimica Avanzada (ORFEO-CINQA), Universidad de

CastillaLa Mancha, E-13071 Ciudad Real, Spain;

orcid.org/0000-0002-1187-7719

́

Rafael Fernandez-Galan − Departamento de Quimica

́

Inorganica, Organica y Bioquimica, Centro de Innovacion en

́

Quimica Avanzada (ORFEO-CINQA), Universidad de

CastillaLa Mancha, E-13071 Ciudad Real, Spain;

orcid.org/0000-0001-5832-6247

Complete contact information is available at:

https://pubs.acs.org/10.1021/acs.inorgchem.0c01152

Author Contributions

The manuscript was written through contributions of all

authors. All authors have given approval to the ﬁnal version of

the manuscript.

Notes

The authors declare no competing ﬁnancial interest.

ACKNOWLEDGMENTS

We gratefully acknowledge ﬁnancial support from the

Ministerio de Economia y Competitividad (MINECO),

■

́

Spain (Grants CTQ2016-77614-P and CTQ2016-81797-

REDC). We also thank Dr. D. Elorriaga for invaluable

assistance in X-ray structure analysis. A.R. acknowledges a

postdoctoral contract funded by the “Plan Propio de I + D + i”

of the Universidad de CastillaLa Mancha.

REFERENCES

■

ASSOCIATED CONTENT

■

(1) Aresta, M. Carbon Dioxide as Chemical Feedstock; Wiley-VCH

Verlag GmbH & Co. KGaA, 2010; p 394.

sı

* Supporting Information

(2) (a) Olah, G. A.; Goeppert, A.; Prakash, G. K. S. Beyond Oil and

Gas: The Methanol Economy; Wiley-VCH Verlag GmbH & Co. KGaA,

2009; p 334. (b) Olah, G. A.; Goeppert, A.; Prakash, G. K. S.

Chemical Recycling of Carbon Dioxide to Methanol and Dimethyl

Ether: From Greenhouse Gas to Renewable, Environmentally Carbon

Neutral Fuels and Synthetic Hydrocarbons. J. Org. Chem. 2009, 74,

487−498.

The Supporting Information is available free of charge at

https://pubs.acs.org/doi/10.1021/acs.inorgchem.0c01152.

General remarks, synthetic protocols, and NMR spectra

for new compounds and stoichiometric reactions,

general protocols and NMR spectra for the catalytic

reduction of CO₂with hydroboranes, kinetic experi-

ments, and X-ray crystal determination (PDF)

(3) (a) Sordakis, K.; Tang, C.; Vogt, L. K.; Junge, H.; Dyson, P. J.;

Beller, M.; Laurenczy, G. Homogeneous Catalysis for Sustainable

N

https://dx.doi.org/10.1021/acs.inorgchem.0c01152

Inorg. Chem. XXXX, XXX, XXX−XXX

Inorganic Chemistry

pubs.acs.org/IC

Article

Hydrogen Storage in Formic Acid and Alcohols. Chem. Rev. 2018,

118, 372−433. (b) Kar, S.; Kothandaraman, J.; Goeppert, A.; Prakash,

G. K. S. Advances in catalytic homogeneous hydrogenation of carbon

dioxide to methanol. J. CO2 Util. 2018, 23, 212−218. (c) Dong, K.;

Razzaq, R.; Hu, Y.; Ding, K. Homogeneous Reduction of Carbon

Dioxide with Hydrogen. Top. Curr. Chem. 2017, 375, 23.

(d) Bernskoetter, W. H.; Hazari, N. Reversible Hydrogenation of

Carbon Dioxide to Formic Acid and Methanol: Lewis Acid

Enhancement of Base Metal Catalysts. Acc. Chem. Res. 2017, 50,

1049−1058. (e) Klankermayer, J.; Wesselbaum, S.; Beydoun, K.;

Leitner, W. Selective Catalytic Synthesis Using the Combination of

Carbon Dioxide and Hydrogen: Catalytic Chess at the Interface of

Energy and Chemistry. Angew. Chem., Int. Ed. 2016, 55, 7296−7343.

(f) Wang, W.-H.; Himeda, Y.; Muckerman, J. T.; Manbeck, G. F.;

Fujita, E. CO₂Hydrogenation to Formate and Methanol as an

Alternative to Photo- and Electrochemical CO₂Reduction. Chem. Rev.

2015, 115, 12936−12973. (g) Alberico, E.; Nielsen, M. Towards a

methanol economy based on homogeneous catalysis: methanol to H₂

and CO₂to methanol. Chem. Commun. 2015, 51, 6714−6725.

(4) (a) Wang, X.; Xia, C.; Wu, L. Homogeneous carbon dioxide

reduction with p-block element-containing reductants. Green Chem.

2018, 20, 5415−5426. (b) Fontaine, F. G.; Rochette, E. Ambiphilic

Molecules: From Organometallic Curiosity to Metal-Free Catalysts.

Acc. Chem. Res. 2018, 51, 454−464. (c) Dagorne, S.; Wehmschulte, R.

Recent Developments on the Use of Group 13 Metal Complexes in

Catalysis. ChemCatChem 2018, 10, 2509−2520. (d) Fontaine, F.-G.;

Stephan, D. W. Metal-free reduction of CO₂. Curr. Opin. Green Sust.

Chem. 2017, 3, 28−32. (e) Fontaine, F.-G.; Courtemanche, M.-A.;

51, 2107−2110. (g) Declercq, R.; Bouhadir, G.; Bourissou, D.;

́

Legare, M.-A.; Courtemanche, M.-A.; Nahi, K. S.; Bouchard, N.;

Fontaine, F.-G.; Maron, L. Hydroboration of Carbon Dioxide Using

Ambiphilic Phosphine −Borane Catalysts: On the Role of the

Formaldehyde Adduct. ACS Catal. 2015, 5, 2513−2520. (h) Wang,

T.; Stephan, D. W. Carbene-9-BBN ring expansions as a route to

intramolecular frustrated Lewis pairs for CO₂reduction. Chem. - Eur.

J. 2014, 20, 3036−3039. (i) Wang, T.; Stephan, D. W. Phosphine

catalyzed reduction of CO₂with boranes. Chem. Commun. 2014, 50,

́

7007−7010. (j) Legare, M.-A.; Courtemanche, M.-A.; Fontaine, F.-G.

Lewis base activation of borane-dimethylsulfide into strongly reducing

ion pairs for the transformation of carbon dioxide to methoxyboranes.

Chem. Commun. 2014, 50, 11362−11365. (k) Das Neves Gomes, C.;

́

Blondiaux, E.; Thuery, P.; Cantat, T. Metal-free reduction of CO₂

with hydroboranes: two efficient pathways at play for the reduction of

CO₂to methanol. Chem. - Eur. J. 2014, 20, 7098−7106.

́

(l) Courtemanche, M.-A.; Legare, M.-A.; Maron, L.; Fontaine, F.-G.

Reducing CO₂to methanol using frustrated Lewis pairs: on the

mechanism of phosphine-borane-mediated hydroboration of CO₂. J.

Am. Chem. Soc. 2014, 136, 10708−10717. (m) Sokolovicz, Y. C. A.;

́

Nieto Faza, O.; Specklin, D.; Jacques, B.; Lopez, C. S.; dos Santos, J.

H. Z.; Schrekker, H. S.; Dagorne, S. Acetate-catalyzed hydroboration

of CO₂for the selective formation of methanol-equivalent products.

Catal. Sci. Technol. 2020, 10, 2407−2414.

(9) Tlili, A.; Voituriez, A.; Marinetti, A.; Thuery, P.; Cantat, T.

Synergistic effects in ambiphilic phosphino-borane catalysts for the

hydroboration of CO₂. Chem. Commun. 2016, 52, 7553−7555.

(10) (a) Greenacre, V. K.; Ansell, M. B.; Roe, S. M.; Crossley, I. R.

Synthesis, Structures and Coordination Chemistry of Singly Bridged

Phosphane-Boranes with Coordinately Unsaturated Platinum Group

Metals. Eur. J. Inorg. Chem. 2014, 2014, 5053−5062. (b) Vergnaud, J.;

Grellier, M.; Bouhadir, G.; Vendier, L.; Sabo-Etienne, S.; Bourissou,

D. Synthesis and Reactivity of Ruthenium Arene Complexes

Incorporating Novel Ph₂PCH₂CH₂BR₂Ligands. Easy Access to the

́

Legare, M.-A.; Rochette, E Design principles in frustrated Lewis pair

catalysis for the functionalization of carbon dioxide and heterocycles.

Coord. Chem. Rev. 2017, 334, 124−135. (f) Bontemps, S. Boron-

mediated activation of carbon dioxide. Coord. Chem. Rev. 2016, 308,

117−130. (g) Chong, C. C.; Kinjo, R. Catalytic Hydroboration of

Carbonyl Derivatives, Imines, and Carbon Dioxide. ACS Catal. 2015,

5, 3238−3259. (h) Chakraborty, S.; Bhattacharya, P.; Dai, H.; Guan,

H. Nickel and iron pincer complexes as catalysts for the reduction of

carbonyl compounds. Acc. Chem. Res. 2015, 48, 1995−2003.

Four-Membered Ruthenacycle [(p -cymene)RuCl(κ^{C , P}

-

CH₂CH₂PPh₂)]. Organometallics 2008, 27, 1140−1146. (c) Fisch-

bach, A.; Bazinet, P. R.; Waterman, R.; Tilley, T. D. β -

Phosphinoethylboranes as Ambiphilic Ligands in Nickel-Methyl

Complexes. Organometallics 2008, 27, 1135−1139.

(11) (a) Bontemps, S.; Vendier, L.; Sabo-Etienne, S. Ruthenium-

catalyzed reduction of carbon dioxide to formaldehyde. J. Am. Chem.

Soc. 2014, 136, 4419−4425. (b) Bontemps, S.; Sabo-Etienne, S.

Trapping formaldehyde in the homogeneous catalytic reduction of

carbon dioxide. Angew. Chem., Int. Ed. 2013, 52, 10253−10255.

(c) Bontemps, S.; Vendier, L.; Sabo-Etienne, S. Borane-mediated

carbon dioxide reduction at ruthenium: formation of C₁and C₂

compounds. Angew. Chem., Int. Ed. 2012, 51, 1671−1674.

́

(i) Fontaine, F.-G.; Courtemanche, M.-A.; Legare, M.-A. Transition-

metal-free catalytic reduction of carbon dioxide. Chem. - Eur. J. 2014,

20, 2990−2996.

́

(5) Fernandez-Alvarez, F. J.; Aitani, A. M.; Oro, L. A. Homogeneous

catalytic reduction of CO₂with hydrosilanes. Catal. Sci. Technol. 2014,

4, 611−624.

(6) Chakraborty, S.; Zhang, J.; Krause, J. A.; Guan, H. An Efficient

Nickel Catalyst for the Reduction of Carbon Dioxide with a Borane. J.

Am. Chem. Soc. 2010, 132, 8872−8873.

́

(7) Courtemanche, M.-A.; Legare, M.-A.; Maron, L.; Fontaine, F.-G.

A highly active phosphine-borane organocatalyst for the reduction of

CO₂to methanol using hydroboranes. J. Am. Chem. Soc. 2013, 135,

9326−9329.

(12) Espinosa, M. R.; Charboneau, D. J.; Garcia de Oliveira, A.;

Hazari, N. Controlling Selectivity in the Hydroboration of Carbon

Dioxide to the Formic Acid, Formaldehyde, and Methanol Oxidation

Levels. ACS Catal. 2019, 9, 301−314.

́

(8) (a) Ramos, A.; Antinolo, A.; Carrillo-Hermosilla, F.; Fernandez-

̃

̈

(13) Lang, A.; Noth, H.; Schmidt, M. Synthesis of Structures of

́

Galan, R.; Rodríguez-Dieguez, A.; García-Vivo, D. Carbodiimides as

catalysts for the reduction of CO₂with boranes. Chem. Commun.

2018, 54, 4700−4703. (b) Yang, Y.; Yan, L.; Xie, Q.; Liang, Q.; Song,

D. Zwitterionic indenylammonium with carbon-centred reactivity

towards reversible CO₂binding and catalytic reduction. Org. Biomol.

(Acyloxy)boranesSynthesis of Structures of (Acyloxy)boranes. Chem.

Ber. 1995, 128, 751−762.

(14) Fujiwara, K.; Yasuda, S.; Mizuta, T. Reduction of CO₂to

Trimethoxyboroxine with BH₃in THF. Organometallics 2014, 33,

6692−6695.

̀

Chem. 2017, 15, 2240−2245. (c) von Wolff, N.; Lefevre, G.; Berthet,

(15) Knopf, I.; Cummins, C. C. Revisiting CO₂Reduction with

NaBH₄under Aprotic Conditions: Synthesis and Characterization of

Sodium Triformatoborohydride. Organometallics 2015, 34, 1601−

1603.

́

J. C.; Thuery, P.; Cantat, T. Implications of CO₂Activation by

Frustrated Lewis Pairs in the Catalytic Hydroboration of CO₂: A View

Using N/Si⁺Frustrated Lewis Pairs. ACS Catal. 2016, 6, 4526−4535.

(d) Sau, S. C.; Bhattacharjee, R.; Vardhanapu, P. K.; Vijaykumar, G.;

Datta, A.; Mandal, S. K. Metal-Free Reduction of CO₂to

Methoxyborane under Ambient Conditions through Borondiformate

Formation. Angew. Chem., Int. Ed. 2016, 55, 15147−15151. (e) Yang,

Y.; Xu, M.; Song, D. Organocatalysts with carbon-centered activity for

CO₂reduction with boranes. Chem. Commun. 2015, 51, 11293−

(16) (a) Beletskaya, I.; Pelter, A. Hydroborations catalysed by

transition metal complexes. Tetrahedron 1997, 53, 4957−5026.

(b) Obligacion, J. V.; Chirik, P. J. Earth-abundant transition metal

catalysts for alkene hydrosilylation and hydroboration. Nat. Rev.

Chem. 2018, 2, 15−34.

́

11296. (f) Ho, S. Y. F.; So, C.-W.; Saffon-Merceron, N.; Mezailles, N.

Formation of a zwitterionic boronium species from the reaction of a

stable carbenoid with borane: CO₂reduction. Chem. Commun. 2015,

O

https://dx.doi.org/10.1021/acs.inorgchem.0c01152

Inorg. Chem. XXXX, XXX, XXX−XXX

Article Doi

Ph2PCH2CH2B(C8H14) and Its Formaldehyde Adduct as Catalysts for the Reduction of CO2with Hydroboranes

DOI: 10.1021/acs.inorgchem.0c01152

Source and publish data:

Authors:

Article abstract of DOI:10.1021/acs.inorgchem.0c01152

Full text of DOI:10.1021/acs.inorgchem.0c01152

Products guided by the article

R&D Labs maybe for 102-24-9

Relevant to this article

Hot Product