catalysts

Article
Reaction Mechanisms of CO2 Reduction to
Formaldehyde Catalyzed by Hourglass Ru, Fe,
and Os Complexes: A Density Functional Theory Study
Chunhua Dong 1,2,3 , Mingsong Ji 1 , Xinzheng Yang 1, *, Jiannian Yao 2 and Hui Chen 2, *
1 Beijing National Laboratory for Molecular Sciences, State Key Laboratory for Structural Chemistry of
Unstable and Stable Species, Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190, China;
dongchunhua@iccas.ac.cn (C.D.); lhjms@iccas.ac.cn (M.J.)
2 Beijing National Laboratory for Molecular Sciences, CAS Key Laboratory of Photochemistry,
Institute of Chemistry, Chinese Academy of Sciences, Beijing 100190, China; jnyao@iccas.ac.cn
3 College of Chemistry, Chemical Engineering and Material, Handan Key Laboratory of Organic Small
Molecule Materials, Handan University, Handan 056005, China
* Correspondence: xyang@iccas.ac.cn (X.Y.); chenh@iccas.ac.cn (H.C.); Tel.: +86-10-8236-2928 (X.Y.);
+86-10-6256-1749 (H.C.)

Academic Editors: Albert Demonceau, Ileana Dragutan and Valerian Dragutan

Received: 23 September 2016; Accepted: 21 December 2016; Published: 27 December 2016

Abstract: The reaction mechanisms for the reduction of carbon dioxide to formaldehyde catalyzed
by bis(tricyclopentylphosphine) metal complexes, [RuH2 (H2 )(PCyp3 )2 ] (1Ru ), [FeH2 (H2 )(PCyp3 )2 ]
(1Fe ) and [OsH4 (PCyp3 )2 ] (1Os ), were studied computationally by using the density functional theory
(DFT). 1Ru is a recently reported highly efficient catalyst for this reaction. 1Fe and 1Os are two
analogues of 1Ru with the Ru atom replaced by Fe and Os, respectively. The total free energy barriers
of the reactions catalyzed by 1Ru , 1Fe and 1Os are 24.2, 24.0 and 29.0 kcal/mol, respectively. With a
barrier close to the experimentally observed Ru complex, the newly proposed iron complex is a
potential low-cost catalyst for the reduction of carbon dioxide to formaldehyde under mild conditions.
The electronic structures of intermediates and transition states in these reactions were analyzed by
using the natural bond orbital theory.

Keywords: reaction mechanism; carbon dioxide; formaldehyde; homogeneous catalysis; ruthenium;

iron; osmium

1. Introduction
As an abundant and non-toxic C1 -building block, carbon dioxide can be reduced to various
chemicals, such as carbon monoxide [1], methanol [2], formaldehyde [3], acetals [4,5], formic acid [5],
formate [6,7], formamides [8], methylamines [8], formamidines [8], imines [3] and methane [9].
Recently, Beller and co-workers [10] reported the methylation of aromatic C–H bonds using CO2
and H2 with the assistance of a ruthenium triphos catalyst. In addition to the experimental studies,
there are some theoretical studies on catalytic reduction of CO2 in recent years. Pidko [11] used a
bis-N-heterocyclic carbene ruthenium CNC-pincer as catalyst and studied the mechanism of CO2
hydrogenation to formates by DFT method. Haunschild [12] reported the catalytic reduction of carbon
dioxide to methanol by using (Triphos)Ru(TMM) as catalyst. Musashi and Sakaki [13] reported a
theoretical study of cis-RuH2 (PH3 )4 -catalyzed hydrogenation of CO2 into formic acid.
As mild reducing agents, boron-containing compounds have been widely used in carbon dioxide
reduction [14,15]. Hazari and co-workers [16] reported the allene carboxylation with CO2 using a PSiP
pincer ligand supported palladium complex as the catalyst in the hydroboration of CO2 . Maron and
co-workers [17,18] reported the phosphine–borane-mediated hydroboration of CO2 to methanol using

Catalysts 2017, 7, 5; doi:10.3390/catal7010005 www.mdpi.com/journal/catalysts

Catalysts 2017, 7, 5 2 of 12

an organocatalyst 1-Bcat-2-PPh2 –C6 H4 (Bcat = catecholboryl). Their mechanistic study showed that
the simultaneous activation of both the reducing agent and CO2 plays an important role in the catalytic
reaction. Bouhadir and co-workers [19] reported the hydroboration of CO2 to methoxyboranes
with the ambiphilic phosphine–borane derivatives as catalysts. They also demonstrated that the
formaldehyde adducts are indeed the actual catalyst in the reaction. Wegner and co-workers [20]
reported a selective reduction of CO2 to methanol using Li2 [1,2-C6 H4 (BH3 )2 ] as the catalyst with
the presence of pinacolborane. Their work shows a novel transition-metal-free mode in which the
aromaticity plays an important role in the bidentate activation process.
Given the above progress, we can see that the reduction of carbon dioxide to formic acid and
its derivatives has been well studied. However, as a key step in the production of methanol from
CO2 and H2 , the catalytic reduction of CO2 to formaldehyde is rarely reported. Huang et al. [21]
studied the mechanistic details of nickel pincer-catalyzed reduction of CO2 to a methanol derivative
with catecholborane (HBcat), and found that formaldehyde is an inevitable intermediate, although it
was not observed in the experiment. Hazari and co-workers [22] reported a CO2 reduction reaction
using a nickel η 3 -cyclooctenyl complex supported by tridentate PSiP pincer ligand as the precatalyst,
and detected the characteristic peak (8.72 ppm) of formaldehyde in a 1 H-NMR spectrum. However,
the formaldehyde was mixed with miscellaneous unverifiable products without a clear yield. Hill and
co-workers [23] reported a selective reductive hydroboration of CO2 to a methanol equivalent,
CH3 OBpin, using B(C6 F5 )3 -activated alkaline earth compounds as catalysts. They also proposed
a mechanism with the formation of formaldehyde as a byproduct. Tzschucke and co-workers [24]
reported an electrocatalytic reduction of CO2 to formic acid and found an occasional formation of
formaldehyde using bipyridine iridium complexes with pentamethylcyclopentadienyl ligands as the
catalysts. In the above reactions, formaldehyde was only observed as byproducts with extremely
low yields.
In 2014, Bontemps and co-workers [3] reported the first unambiguous detection of formaldehyde
from the pinacolborane reduction of CO2 with a yield of 22% using the dihydride bis(dihydrogen)
bis(tricyclopentylphosphine) hourglass ruthenium complex [RuH2 (H2 )2 (PCyp3 )2 ] (Ru-1cyp) as the
catalyst precursor at room temperature in 24 h. Although the selectivity and yield of formaldehyde are
not ideal enough, the controllable generation of formaldehyde by reducing CO2 is still a significant
breakthrough. However, detailed mechanism of the above reaction, such as the structures of
rate-limiting states, is still missing. In this paper, we report a density functional theory (DFT) study of
the reaction mechanisms of the reduction of CO2 to formaldehyde catalyzed by [RuH2 (H2 )(PCyp3 )2 ]
(1Ru ) and its Fe and Os analogues, [FeH2 (H2 )(PCyp3 )2 ] (1Fe ) and [OsH4 (PCyp3 )2 ] (1Os ). The potentials
of 1Fe and 1Os as catalysts for the reaction were predicted accordingly. The relations between the
electronic structures and catalytic properties were analyzed using the natural bond orbital (NBO)
theory [25].

2. Results and Discussion

2.1. The Mechanism for the Reduction of Carbon Dioxide to Formaldehyde Catalyzed by the Ru Complex
Scheme 1 shows the whole catalytic cycle for the reduction of carbon dioxide to formaldehyde
catalyzed by 1Ru . Figure 1 shows the corresponding free energy profile. Figure 2 shows the optimized
structures of transition states in the catalytic cycle.
At the beginning of the reaction, the dissociation of H2 from Ru-1cyp forms the 10.5 kcal/mol
less stable catalyst [RuH2 (H2 )(PCyp3 )2 ] (1Ru ), which could attract a HBpin molecule and form a
15.8 kcal/mol more stable intermediate 1Ru-HBpin . The exchange of CO2 with HBpin in 1Ru-HBpin for
the formation of 2Ru is 14.8 kcal/mol uphill. Next, 3Ru is formed through transition state TS2,3-Ru
with a 5.5 kcal/mol barrier for the transfer of a hydride from Ru to the carbon in CO2 . The newly
formed formate group in 3Ru could easily rearrange through TS3,4-Ru and TS4,5-Ru for the formation
of a more stable intermediate 5Ru . Then, a HBpin molecule attacks 5Ru and forms a 2.6 kcal/mol
less stable intermediate 6Ru with a weak interaction between O and B. Next, 7Ru is formed quickly
Catalysts 2017, 7, 5 3 of 12

through TS6,7-Ru with the stretching of the B–H bond and the formation of the O–B and Ru–H bonds.
A pinBOCHO molecule is formed with the cleavage of the Ru–O bond through transition state TS7,8-Ru ,
Catalysts 2017, 7, 5  3 of 12 
which Catalysts 2017, 7, 5 
is 12.2 kcal/mol higher than 7Ru in free energy. The dissociation of pinBOCHO3 of 12  from 8Ru
for theis formed with the cleavage of the Ru–O bond through transition state TS
regeneration of 1Ru is 1.1 kcal/mol downhill. Then the dissociated pinBOCHO molecule
7,8‐Ru, which is 12.2 kcal/mol 
is formed with the cleavage of the Ru–O bond through transition state TS
higher than 7
can come back to Ru  in free energy. The dissociation of pinBOCHO from 8
1Ru and take a hydride from Ru to its carbonyl Ru 7,8‐Ru, which is 12.2 kcal/mol 
 for the regeneration of 1
carbon for the formation Ru is 
of 10Ru
◦ higher than 7 Ru in free energy. The dissociation of pinBOCHO from 8
1.1 kcal/mol downhill. Then the dissociated pinBOCHO molecule can come back to 1 Ru  for the regeneration of 1 Ru is 
Ru and take a 
(∆G = −8.3 kcal/mol). After a structural relaxation, 10Ru transforms to a 5.9 kcal/mol more stable
1.1 kcal/mol downhill. Then the dissociated pinBOCHO molecule can come back to 1
hydride  Ru and take a 
intermediate
hydride  11from 
from 
Ru  to  its  carbonyl  carbon  for  the  formation  of  10Ru  (ΔG°  =  −8.3  kcal/mol). 
Ru . Then
Ru  aits 
to  formaldehyde
carbonyl  molecule
carbon  for  the  forms withof 
formation  the 10formation of the Ru–OBpin After  a 
Ru  (ΔG°  =  −8.3  kcal/mol).  After  a 
bond and
structural  relaxation,  10Ru  transforms  to  a  5.9  kcal/mol  more  stable  intermediate  11Ru.  Then  a 
the breaking
structural of the Ru–OCH
relaxation,  10Ru2 OBpin and C–OBpin
  transforms  bonds through
to  a  5.9  kcal/mol  more  stable  transition state TS
intermediate  1111,12-Ru
Ru.  Then , which
a  is
formaldehyde molecule forms with the formation of the Ru–OBpin bond and the breaking of the Ru–
3.8 kcal/mol higher than 11 in free energy. The release of formaldehyde
formaldehyde molecule forms with the formation of the Ru–OBpin bond and the breaking of the Ru–
Ru
OCH2OBpin and C–OBpin bonds through transition state TS11,12‐Ru, which is 3.8 kcal/mol higher than  from 12 Ru for the formation
OCH
of 13Ru11 2OBpin and C–OBpin bonds through transition state TS
isRu in free energy. The release of formaldehyde from 12
3.0 kcal/mol downhill. Next, another HBpin molecule 11,12‐Ru, which is 3.8 kcal/mol higher than 
approaches 13Ru
Ru for the formation of 13 and forms a much
 is 3.0 kcal/mol 
Ru
11
more stableRu in free energy. The release of formaldehyde from 12Ru for the formation of 13Ru is 3.0 kcal/mol 
intermediate 15Ru with the formation of theRu and forms a much more stable intermediate 
downhill. Next, another HBpin molecule approaches 13 B–O bond through TS14,15-Ru . The Ru–O and
downhill. Next, another HBpin molecule approaches 13
15 Ru with the formation of the B–O bond through TS
Ru and forms a much more stable intermediate 
H–B bonds in 15Ru could break easily with the formation of. The Ru–O and H–B bonds in 15
14,15‐Ru a Ru–H bond through TS15,16-Ru Ru could 
, which is
15Ru with the formation of the B–O bond through TS14,15‐Ru. The Ru–O and H–B bonds in 15
break easily with the formation of a Ru–H bond through TS 15,16‐Ru, which is only 4.5 kcal/mol higher 
Ru could 
only 4.5 kcal/mol higher than 15Ru in free energy. Then the dissociation of pinBOBpin from 16Ru
break easily with the formation of a Ru–H bond through TS
than 15 15,16‐Ru, which is only 4.5 kcal/mol higher 
Ru in free energy. Then the dissociation of pinBOBpin from 16Ru regenerates 1Ru and completes 
regenerates
than 15 1RuRu and completes the catalytic cycle. It is worth noting
 in free energy. Then the dissociation of pinBOBpin from 16 that HBpinRucan
 regenerates 1 also attack 10Ru
 and completes 
the  catalytic  cycle.  It  is  worth  noting  that  HBpin  can  also  attack Ru10 Ru  and  form  a  stable  acetal 
and form a stable acetal
the  catalytic  cycle. 2CH
compound (pinBO)
compound
It  is  (pinBO)
worth  noting  that  CH
2 HBpin 
2 , which was observed in the experiment.
can  also  attack  10Ru  and  form  a  stable  acetal 
2, which was observed in the experiment. 
compound (pinBO)2CH2, which was observed in the experiment. 
H P
H
1Ru = P
H Ru
HH H
1Ru = Ru
P
H H
a = CO2; bP= HBpin; c = pinBOCHO; d = (pinBO)2CH2; e = HCHO; f = pinBOBpin
a = CO2; b = HBpin; c = pinBOCHO; d = (pinBO)2CH2; e = HCHO; f = pinBOBpin
TS2,3-Ru
(4.5)
TS
5 2,3-Ru TS7,8-Ru
(4.5) TS9,10-Ru
5 +4b TS3,4-Ru (0.4)
TS 7,8-Ru (-1.7)
0 TS TS TS9,10-Ru
1Ru +4b (-3.2)
TS
4,5-Ru 6,7-Ru (0.4)
3,4-Ru (-4.8) (-4.8) +3b (-1.7)
+3b
0 2
-5 (0.0)
1Ru
Ru (-3.2) TS4,5-Ru
+4b
TS6,7-Ru 9Ru TS11,12-Ru
(-1.0) 3Ru (-4.8)
+4b (-4.8) +3b +3b
(-2.2) (-10.4)
-5 +a 2Ru
(0.0) 4 6 +3b 9 Ru TS11,12-Ru
-10 +4b (-4.6) +4b Ru Ru
8 10
+4b (-1.0) 3Ru 5
+4b Ru (-5.8)+3b Ru +3b
(-2.2) Ru (-10.4)
+a (-6.0)
4Ru 6Ru 1 Ru +3b
+4b 7Ru (-9.0) (-8.3)
-10
-15 +4b +4b (-4.6) (-8.4)
5Ru +3b 8Ru
(-10.1) +3b
10Ru
TS14,15-Ru
+4b
(-6.0) (-5.8) (-11.8) +3b 1Ru +3b 11Ru+3b
+4b +4b
(-8.4) 7 Ru (-9.0) (-8.3) (-21.0)
-15 1Ru-HBpin +4b +3b +3b
(-10.1) (-14.2) 12 TS14,15-Ru
-20 +4b
+3b
(-11.8) +3b +c +3b 11Ru Ru
(-15.8)
1Ru-HBpin +3b +3b (-18.1) (-21.0)
+2b
(-14.2) 12Ru 13Ru 14
-20 +a +3b +c Ru +e
-25 (-15.8) +3b (-18.1)
+3b (-21.1) +2b TS15,16-Ru
+3b 13 (-22.6)
+a 1Ru-HBpin Ru 14 Ru +e (-30.5)
-25
-30 +3b TS15,16-Ru
+3b +3b +2b
(-21.1)
(-25.9)
1Ru-HBpin +e (-22.6) (-30.5)
+2b
-30 +2b 1Ru +3b +e +2b +e 16Ru 1Ru
-35 (-25.9)
+c (-31.9) +e +e 15Ru +2b(-31.1) (-31.8)
-35 +2b 1Ru
(-35.0)+e 16Ru 1Ru
+c +2b
(-31.9) 15Ru +2b (-31.8)
(-31.1) +2b
+d +2b
(-35.0) +e +e
+2b +2b +2b
+d +e +f
+2b +e +e
+e +f
Reaction Coordinate  
Reaction Coordinate  
Figure 1. Free energies profile of the reduction of carbon dioxide to formaldehyde catalyzed by 1Ru. 
1. Free energies profile of the reduction of carbon dioxide to formaldehyde catalyzedRuby
FigureFigure 1. Free energies profile of the reduction of carbon dioxide to formaldehyde catalyzed by 1 .  1Ru .

 
Figure 2. Optimized structures of TS2,3‐Ru (513i cm−1), TS3,4‐Ru (170i cm−1), TS4,5‐Ru (176i cm−1), TS6,7‐Ru 
 
Figure 2. Optimized structures of TS  (513i cm
cm), TS TS (170i cm
−1 −1), TS 1 (176i cm
−1 −1), TS6,7‐Ru 
Figure(127i cm −1), TS7,8‐Ru (209i cm−1),TS9,10‐Ru2,3‐Ru
 (363i cm −1), TS −1 ),3,4‐Ru
 (159i cm −
4,5‐Ru 15,16‐ cm−1 ),
2. Optimized structures of TS2,3-Ru (513i 11,12‐Ru
3,4-Ru ), TS
(170i14,15‐Ru
cm  (67i cm ) and TS
), TS−14,5-Ru (176i
(127i cm
Ru
−1
 (97i cm ), TS
−1 7,8‐Ru (209i cm ),TS9,10‐Ru (363i cm
−1 −1), TS11,12‐Ru (159i cm −1), TS14,15‐Ru (67i cm ) and TS
). Bond lengths are in Å. The cyclopentyl and pinacol groups are not shown for clarity. 
− 1 − 1 − 1
−1
− 1 15,16‐
TS6,7-Ru (127i cm ), TS7,8-Ru (209i cm ),TS9,10-Ru (363i cm ), TS11,12-Ru (159i cm ), TS14,15-Ru
Ru (97i cm−1). Bond lengths are in Å. The cyclopentyl and pinacol groups are not shown for clarity. 
−1
(67i cm ) and TS15,16-Ru (97i cm −1 ). Bond lengths are in Å. The cyclopentyl and pinacol groups are
not shown for clarity.
Catalysts 2017, 7, 5 4 of 12

Catalysts 2017, 7, 5  4 of 12 
By comparing all relative free energies shown in Figure 1, we can conclude that 1Ru-HBpin and
arecomparing 
TS9,10-Ru By  the rate-determining states
all relative free  of the catalytic
energies shown  reaction
in Figure 1,  we with a total free
can  conclude  energy
that  1Ru‐HBpinbarrier
  and  of
24.2 kcal/mol, which agrees well with the observed experimental reaction rate.
TS9,10‐Ru are the rate‐determining states of the catalytic reaction with a total free energy barrier of 24.2 
Compared with the mechanism proposed by Huang et al. for the reduction reaction of CO2
kcal/mol, which agrees well with the observed experimental reaction rate. 
to a methanol derivative with catecholborane (HBcat) catalyzed by a pincer nickel complex
Compared with the mechanism proposed by Huang et al. for the reduction reaction of CO 2 to a  [21],

methanol  derivative  with  catecholborane  (HBcat)  catalyzed  by  a  pincer  nickel  complex 
the catalytic reduction of carbon dioxide to formaldehyde reactions catalyzed by hourglass ruthenium [21],  the 
catalytic reduction of carbon dioxide to formaldehyde reactions catalyzed by hourglass ruthenium 
complex and pincer nickel complex have similar but not exactly the same pathways. There are two
complex and pincer nickel complex have similar but not exactly the same pathways. There are two 
oxygen atoms as possible bonding sites for HBpin in the nickel formate. However, there is only
oxygen atoms as possible bonding sites for HBpin in the nickel formate. However, there is only one 
one oxygen atom as the bonding site for HBpin in the ruthenium formate (5Ru to 6Ru in Scheme 1).
oxygen atom as the bonding site for HBpin in the ruthenium formate (5Ru to 6Ru in Scheme 1). Our 
Our calculations indicate that the six-membered ring structure is less likely to be formed in the
calculations indicate that the six‐membered ring structure is less likely to be formed in the hourglass 
hourglass ruthenium complexes.
ruthenium complexes. 

u
-R
12
,
11

 
Scheme 1. Catalytic cycle for the reduction of carbon dioxide to formaldehyde catalyzed by 1Ru. 
Scheme 1. Catalytic cycle for the reduction of carbon dioxide to formaldehyde catalyzed by 1Ru .
Catalysts 2017, 7, 5 5 of 12

Catalysts 2017, 7, 5  5 of 12 

2.2. The Mechanism for the Reduction of Carbon Dioxide to Formaldehyde Catalyzed by the Fe Complex
2.2. The Mechanism for the Reduction of Carbon Dioxide to Formaldehyde Catalyzed by the Fe Complex 
Scheme 2 shows the whole catalytic cycle for the reduction of carbon dioxide to formaldehyde
Scheme 2 shows the whole catalytic cycle for the reduction of carbon dioxide to formaldehyde 
catalyzed by the newly proposed Fe complex 1Fe
catalyzed by the newly proposed Fe complex 1 . Figures 3 and 4 show the corresponding free energy
Fe. Figures 3 and 4 show the corresponding free energy 
profile and optimized structures of key transition states, respectively. The reactions catalyzed byFe1 Fe
profile and optimized structures of key transition states, respectively. The reactions catalyzed by 1
andand 1Ruhave
1 Ru similar pathways but different rate-determining states.
 have similar pathways but different rate‐determining states. 

O
P P
H C H H
H
Fe O Fe O O
in H C
p H H TS
HB P P
H 3 ,4
5Fe 4Fe
-F
e

P O P
H H O
C H P = P(cyp)3
Fe O Fe C O
H H H
H P H Bpin O H
P
6Fe Bpin = B
3Fe

Fe
O
TS 6,7

TS2,3-
-Fe

O
H P H P
H

H C
OC

H O
Fe CHO Fe
pinBO
inO

H CO2 H H
H
Bp

P P
H H P H P P H
H H H2 HBpin H H
7Fe H 2Fe
Fe Fe Fe Bpin
H H H H H H
P P P
pin P
Fe-1cyp H P H
1Fe BO H
1Fe-HBpin
in CH H O
H Bpin Bp
Fe BO O Fe C
p in H
H O HBpin H
P H P O
Bpin
12Fe 8Fe Bpin

(pinBO)2CH2
HBpin
P P
H
H O H
H
Fe Fe C
O H H
H
Bpin H O
P P P
H O H Bpin
11Fe H C 9Fe
Fe
Fe H 0-
,1
HCHO O TS9
H
P Bpin

10Fe
 
Scheme 2. Catalytic cycle for the reduction of carbon dioxide to formaldehyde catalyzed by 1
Scheme 2. Catalytic cycle for the reduction of carbon dioxide to formaldehyde catalyzed byFe1. Fe .

Similar  to  the  reaction  catalyzed  by  1Ru,  a  −22.0  kcal/mol  more  stable  intermediate  1Fe‐HBpin  is 
Similar to the reaction catalyzed by 1Ru , a −22.0 kcal/mol more stable intermediate 1Fe-HBpin
formed at the beginning of the reaction catalyzed by 1Fe. The exchange of CO2 with HBpin in 1Fe‐HBpin 
is formed at the beginning of the reaction catalyzed by 1Fe . The exchange of CO2 with HBpin in
and the formation of 2Fe is 16.4 kcal/mol uphill. Next, 3Fe is formed quickly through transition state 
1Fe-HBpin and the formation of 2Fe is 16.4 kcal/mol uphill. Next, 3Fe is formed quickly through
TS2,3‐Fe for the formation of C–H bond and the rearrangement of hydrogen atoms. The rearrangement 
transition state TS2,3-Fe for the formation of C–H bond and the rearrangement of hydrogen atoms.
of  the  newly  formed  formate  group  and  hydrogen  atoms  in  3Fe  forms  a  slightly  more  stable 
The rearrangement
intermediate  4Fe.  of the newly
After  formedrelaxation, 
a  structural  formate group and hydrogen
4Fe  transforms  to  a atoms in 3Fe forms
2.0  kcal/mol  more a stable 
slightly
more stable intermediate
intermediate  Fe . Aftera aHBpin 
5Fe,  which 4attracts  structural relaxation,
molecule  transforms
4Feformation 
for  the  of to a 2.0
a  3.9  kcal/mol
kcal/mol  more
more  stable
stable 
intermediate 5 Fe , which attracts a HBpin molecule for the formation of a 3.9 kcal/mol
intermediate 6Fe with strong Fe–H and O–B interactions. Then, 7Fe is formed through a transition state  more stable
intermediate 6Fe with strong Fe–H and O–B interactions. Then, 7Fe is formed through a transition
TS6,7‐Fe with a free energy barrier of 20.8 kcal/mol with the breaking of the Fe–O and B–H bonds and 
state TS6,7-Fe with a free energy barrier
the formation of the O–B bond. 7 of 20.8 kcal/mol with the breaking
Fe is 12.3 kcal/mol less stable than 6 of the Fe–O and B–H bonds
Fe with a weak interaction between 

and the formation of the O–B bond. 7Fe is 12.3 kcal/mol

H and B. The dissociation of pinBOCHO from 7 less stable than Fe
Fe for the regeneration of 1 6Fe with a weak interaction
 is 3.6 kcal/mol downhill. 
Catalysts 2017, 7, 5 6 of 12

between H and B. The dissociation of pinBOCHO from 7Fe for the regeneration of 1Fe is 3.6 kcal/mol
Catalysts 2017, 7, 5  6 of 12 
downhill. Next, the dissociated pinBOCHO molecule comes back to 1Fe and forms a Fe–O bond
Catalysts 2017, 7, 5 
for the
6 of 12 
formation
Next,  ofthe 
a 7.3 kcal/molpinBOCHO 
dissociated  more stable intermediate
molecule  comes  8 Fe . After
back  to  1Fe aand 
structural
forms  a relaxation,
Fe–O  bond 8Fe for transforms
the 
Next,  the  dissociated  pinBOCHO  molecule 
to a 3.1formation of a 7.3 kcal/mol more stable intermediate 8
kcal/mol more stable intermediate 9Fecomes 
. Thenback  a to  1 Fe  and  forms 
formaldehyde a  Fe–O 
molecule bond 
is formed
Fe. After a structural relaxation, 8Fe transforms 
for  the 
with the
formationformation of a 7.3 kcal/mol more stable intermediate 8
to  a  of
3.1 the Fe–OBpin
kcal/mol  more bond
stable and the breaking
intermediate  of the
9Fe.  Then  . After a structural relaxation, 8
Fe–OCH2 OBpin
a Feformaldehyde  and C–OBpin
molecule  Fe transforms 
is  formed  bonds through
with  the 
to  a  3.1  kcal/mol 
formation  of  the  more  stable 
Fe–OBpin  intermediate 
bond  and  the  9Fe.  Then  of 
breaking  a  formaldehyde 
the  Fe–OCH molecule 
OBpin  and  is C–OBpin 
formed  with 
bonds the 
transition state TS9,10-Fe . The dissociation of formaldehyde from 10Fe leaves an 8.6 kcal/mol more 2
formation  of  the  Fe–OBpin  bond  and  the  breaking  of  the  Fe–OCH2OBpin  and  C–OBpin  bonds 
stable through transition state TS
intermediate 11Fe . Then,9,10‐Fe . The dissociation of formaldehyde from 10
another HBpin molecule approaches 11FeFe leaves an 8.6 kcal/mol 
and forms a 10.9 kcal/mol
through transition state TS
more  stable  intermediate  119,10‐Fe . The dissociation of formaldehyde from 10 Fe leaves an 8.6 kcal/mol 
Fe.  Then,  another  HBpin  molecule  approaches  11Fe  and  forms  a  10.9 
more stable intermediate
more  stable  12Fe 11
intermediate  with the formation of a B–O bond. The dissociation of newly formed
Fe.  Then,  another  HBpin  molecule  approaches  11Fe  and  forms  a  10.9 
kcal/mol more stable intermediate 12 Fe with the formation of a B–O bond. The dissociation of newly 
pinBOBpin molecule from 12 for
kcal/mol more stable intermediate 12
Fe the regeneration of the catalyst 1Fe is Fe9.5 kcal/mol uphill.
 with the formation of a B–O bond. The dissociation of newly 
formed pinBOBpin molecule from 12FeFe for the regeneration of the catalyst 1  is 9.5 kcal/mol uphill. 
formed pinBOBpin molecule from 12Fe for the regeneration of the catalyst 1Fe is 9.5 kcal/mol uphill. 

 
 
Figure 3. Free energies profile of the reduction of carbon dioxide to formaldehyde catalyzed by 1Fe. 
3. Free energies profile of the reduction of carbon dioxide to formaldehyde catalyzedFeby
Figure Figure 3. Free energies profile of the reduction of carbon dioxide to formaldehyde catalyzed by 1 .  1Fe .

 
 
Figure 4. Optimized structures of TS2,3‐Fe (460i cm ), TS3,4‐Fe (78i cm ), TS6,7‐Fe (191i cm−1) and TS9,10‐Fe 
−1 −1

Figure 4. Optimized structures of TS 2,3‐Fe (460i cm−1), TS3,4‐Fe (78i cm−1), TS6,7‐Fe (191i cm−1) and TS9,10‐Fe 
Figure(86i cm structures of TS2,3-Fe (460i cm−1 ), TS3,4-Fe (78i cm−1 ), TS6,7-Fe (191i cm−1 ) and
−1). Bond lengths are in Å. The cyclopentyl and pinacol groups are not shown for clarity. 
4. Optimized
(86i cm ). Bond lengths are in Å. The cyclopentyl and pinacol groups are not shown for clarity. 
−1
−1
TS9,10-Fe (86i cm ). Bond lengths are in Å. The cyclopentyl and pinacol groups are not shown
for clarity.
Catalysts 2017, 7, 5 7 of 12

Catalysts 2017, 7, 5  7 of 12 
By comparing all relative free energies shown in Figure 3, we can conclude that 1Fe-HBpin and
are the rate-determining states of the catalytic reaction with a total free energy
TS6,7-FeBy comparing all relative free energies shown in Figure 3, we can conclude that 1 barrier
Fe‐HBpin  and  of
24.0
TSkcal/mol, which is 0.2 kcal/mol lower than the free energy barrier of the reaction catalyzed
6,7‐Fe are the rate‐determining states of the catalytic reaction with a total free energy barrier of 

by24.0 kcal/mol, which is 0.2 kcal/mol lower than the free energy barrier of the reaction catalyzed by 
1Ru . Therefore, 1Fe is a potential low-cost and high efficiency catalyst for the reduction of CO2
to 1formaldehyde.
Ru. Therefore, 1Fe is a potential low‐cost and high efficiency catalyst for the reduction of CO2 to 

formaldehyde. 
Sabo-Etienne and Bontemps [26] recently reported the catalytic reduction of CO2 to
Sabo‐Etienne 
bis(boryl)acetal and  Bontemps with
and methoxyborane, [26] Fe(H)
recently  reported 
2 (dmpe) 2 used asthe  catalytic 
catalyst reduction 
precursor, of  CO2  to 
and pinacolborane
bis(boryl)acetal  and  methoxyborane,  with  Fe(H) (dmpe)   used  as  catalyst 
used as the reductant. For 1Ru , only one oxygen atom of CO2 is abstracted, and CO2 transforms
2 2 precursor,  and 
into
pinacolborane used as the reductant. For 1 Ru, only one oxygen atom of CO2 is abstracted, and CO2 
HCHO. For Fe(H)2 (dmpe)2 , one oxygen atom of CO2 is abstracted at the reduction step, then the
transforms into HCHO. For Fe(H)
second oxygen atom is removed and 2(dmpe)2, one oxygen atom of CO2 is abstracted at the reduction 
CO2 is transformed to methylene. This illustrates that iron
step, then the second oxygen atom is removed and CO
complexes can efficiently catalyze the reduction of CO22. is transformed to methylene. This illustrates 
that iron complexes can efficiently catalyze the reduction of CO2. 
2.3. The Mechanism for the Reduction of Carbon Dioxide to Formaldehyde Catalyzed by the Os Complex
2.3. The Mechanism for the Reduction of Carbon Dioxide to Formaldehyde Catalyzed by the Os Complex 
The reaction catalyzed by 1Os also has similar pathways to the reaction catalyzed by 1Ru
The  reaction  catalyzed  by  1Os  also  has  similar  pathways  to  the  reaction  catalyzed  by  1Ru  (see 
(see Scheme S1 on catalytic cycle for the reduction of carbon dioxide to formaldehyde catalyzed
Scheme S1 on catalytic cycle for the reduction of carbon dioxide to formaldehyde catalyzed by 1Os). 
by 1Os ). Figure 5 shows the calculated relative free energies in the reaction catalyzed by 1Os . Figure 6
Figure 5 shows the calculated relative free energies in the reaction catalyzed by 1Os. Figure 6 shows 
shows the optimized structures of Os-1cyp, 1Os , 4Os , TS7,8-Os and 8Os , in which the distances between
the optimized structures of Os‐1cyp, 1Os, 4Os, TS7,8‐Os and 8Os, in which the distances between the two 
the two hydrogen atoms (marked in Figure 6) are longer than those in corresponding Ru complexes.
hydrogen  atoms  (marked  in  Figure  6)  are  longer  than  those  in  corresponding  Ru  complexes.  As 
Asshown in Figure 5, the rate‐determining states in the reaction catalyzed by the osmium complex are 
shown in Figure 5, the rate-determining states in the reaction catalyzed by the osmium complex are
1Os-HBpin and TS2,3-Os
1Os‐HBpin and TS with a free energy barrier of 29.0 kcal/mol, which is 4.8 kcal/mol higher than
2,3‐Os with a free energy barrier of 29.0 kcal/mol, which is 4.8 kcal/mol higher than the 
thefree energy barrier of the reaction catalyzed by 1
free energy barrier of the reaction catalyzed byRu1.Therefore, 1
Ru .Therefore, 1Os could also catalyze the reduction
Os could also catalyze the reduction of 
of carbon dioxide to formaldehyde under harsher conditions. 
carbon dioxide to formaldehyde under harsher conditions.

P
H
1Os = H Os
H
P H
a = CO2; b = HBpin; c = pinBOCHO; d = (pinBO)2CH2; e = HCHO; f = pinBOBpin

15 TS2,3-Os
(9.8)
10
+4b TS3,4-Os
5 TS7,8-Os
(0.5) TS4,5-Os TS6,7-Os (-0.3) TS9,10-Os
0 (-3.4) (-2.7)
1Os 2Os 3Os +4b +3b
(-4.2)
TS11,12-Os
-5 (0.0) (0.5) (-0.3) 4Os +4b
G(kcal/mol)

+3b +3b
+a +4b 6Os 9 (-9.4)
+4b (-3.5) 5Os 8Os
Os
-10 +4b (-4.6) (-5.2) 10Os
+4b (-6.0) 7Os (-7.2) 1 +3b
+3b Os (-7.5)
-15 +4b +3b
(-10.3) +3b (-10.1)
+3b 11Os TS14,15-Os
+3b +3b (-14.1) (-22.4)
-20 +c
1Os-HBpin +3b 12Os
13
(-18.7) Os 14 +2b
-25 (-19.2) Os
+e TS
+a +3b (-21.1)(-22.3) 15,16-Os
+3b
-30 +3b +2b (-32.6)
1Os-HBpin +e
+e 1Os
-35 (-29.3) 1Os +2b 16
Os (-31.8)
+2b (-31.9) +e
15Os (-33.6) +2b
-40 +c +2b
(-36.3) +2b +e
+d +f
+2b +e
+e
Reaction Coordinate  
Figure 5. Free energies profile of the reduction of carbon dioxide to formaldehyde catalyzed by 1
Figure 5. Free energies profile of the reduction of carbon dioxide to formaldehyde catalyzed byOs1. Os .
Catalysts 2017, 7, 5 8 of 12
Catalysts 2017, 7, 5  8 of 12 

 
Figure 6. Optimized structures of Os‐1cyp, 1Os, 4 −1 ) and
Os, TS  (239i cm −1) and 8 . Bond lengths are in Å. 
Figure 6. Optimized structures of Os-1cyp, 1Os , 4Os , TS7,8‐Os
7,8-Os (239i cm Os8
Os . Bond lengths are in
The cyclopentyl and pinacol groups are not shown for clarity. 
Å. The cyclopentyl and pinacol groups are not shown for clarity.

2.4. Structures and Reactivity 
2.4. Structures and Reactivity
According 
According to  our  mechanistic
to our mechanistic  study,
study,  thethe  catalytic
catalytic  reaction
reaction  of 2HBpin  ++  CO22  → 
of  2HBpin → HCHO 
HCHO +  +
pinBOBpin  has 
pinBOBpin five  stages,
has five stages,  CO22  insertion 
insertion (S1), 
(S1), σ‐Bond 
σ-Bond metathesis 
metathesis (S2), 
(S2), pinBOCHO 
pinBOCHO insertion insertion (S3), 
(S3),
HCHO elimination (S4) and σ‐Bond metathesis (S5). 
HCHO elimination (S4) and σ-Bond metathesis (S5).
(S1) 
(S1) CO22  Insertion. 
Insertion. The  The reaction 
reaction in  in S1 
S1 is  [M]–H  +
is [M]–H +  CO22  → 
→ [M]–OOCH. 
[M]–OOCH. The  The most  significant 
most significant
structural change in S1 is the cleavage of M–H bond. Table 1 lists the natural population analysis 
structural change in S1 is the cleavage of M–H bond. Table 1 lists the natural population analysis
(NPA) charges of metal atoms (e), the Wiberg bond indices (WBIs) of the bonds, and the relative free 
(NPA) charges of metal atoms (e), the Wiberg bond indices (WBIs) of the bonds, and the relative free
energies between intermediates and transition states/intermediates (∆G‡‡/∆G
energies between intermediates and transition states/intermediates (ΔG ◦ ) in stages. We can see
/ΔG°) in stages. We can see 
that 2 Fe has the most negative NPA charge, the smallest WBI of the M–H bond and the lowest barrier 
that 2Fe has the most negative NPA charge, the smallest WBI of the M–H bond and the lowest barrier
for the cleavage of M–H bond among these three intermediates in S1. 
for the cleavage of M–H bond among these three intermediates in S1.
(S2) σ‐Bond Metathesis. The reaction in S2 is [M]–OOCH + HBpin → [M]–H + pinBOCHO. The 
(S2) σ-Bond Metathesis. The reaction in S2 is [M]–OOCH + HBpin → [M]–H + pinBOCHO.
most important structural change in S2 is the cleavage of M–O bond. We can see TS
The most important structural change in S2 is the cleavage of M–O bond. We can see7,8‐Os TS has the least 
7,8-Os has the
NPA charge, the smallest WBI of the M–O bond and the lowest barrier for the cleavage of M–O bond 
least NPA charge, the smallest WBI of the M–O bond and the lowest barrier for the cleavage of M–O
among these three transition states in S2. 
bond among these three transition states in S2.
(S3) pinBOCHO Insertion. The reaction in S3 is [M]–H + pinBOCHO → [M]–O–CH
(S3) pinBOCHO Insertion. The reaction in S3 is [M]–H + pinBOCHO → [M]–O–CH 2–OBpin. The 
2 –OBpin.
most 
The important 
most important structural 
structural change 
change in  S3  is  the 
in S3 M–H 
is the M–Hbond bondcleavage. 
cleavage.9Ru   has 
9Ru hasa amore 
morenegative 
negative NPA 
NPA
charge, a smaller WBI of the M–H bond and a lower barrier for the cleavage of M–H bond than 9
charge, a smaller WBI of the M–H bond and a lower barrier for the cleavage of M–H bond thanOs9 in  Os
S3.  The 
in S3. Fe Fe
The complexes 
complexes cannot 
cannotbe becompared 
comparedwith  withthe 
theRu 
Ruand 
andOs Oscomplexes 
complexes because  because they 
they have 
have a  a
different pathway in S3. Scheme 1 shows that 1
different pathway in S3. Scheme 1 shows thatRu1 and pinBOCHO first generate the intermediate 9
Ru and pinBOCHO first generate the intermediate Ru, 
then  form form
9Ru , then the  intermediate 
the intermediate 10Ru10
  through 
Ru through TS9,10‐Ru,  in  which 
TS9,10-Ru a  Ru–O 
, in which a Ru–Obond  bondis  formed. 
is formed.However, 
However, as 
shown 
as shown in  in
Scheme 
Scheme 2,  12,Fe1  combines 
Fe combines directly 
directly with  pinBOCHO 
with pinBOCHO and andtransforms 
transforms to  to
intermediate 
intermediate 8Fe, 8Fe
in ,
which an Fe–O bond is formed. 
in which an Fe–O bond is formed.
(S4) HCHO Elimination. The reaction in S4 is [M]–O–CH
(S4) HCHO Elimination. The reaction in S4 is [M]–O–CH 2–OBpin → HCHO + [M]–OBpin. The 
2 –OBpin → HCHO + [M]–OBpin.
most important structural change in S4 is the cleavage of the M–O bond. TS
The most important structural change in S4 is the cleavage of the M–O bond. TS
9,10‐Fe  has the lowest barrier 
9,10-Fe has the lowest
and the smallest WBI of the M–O bond among these three transition states in S4. There is no obvious 
barrier and the smallest WBI of the M–O bond among these three transition states in S4. There is no
relation between the energy barriers and the NPA charges of metal atoms in S4. 
obvious relation between the energy barriers and the NPA charges of metal atoms in S4.
(S5) σ‐Bond Metathesis. The reaction in S5 is [M]–OBpin + HBpin → [M]–H + pinBOBpin. The 
(S5) σ-Bond Metathesis. The reaction in S5 is [M]–OBpin + HBpin → [M]–H + pinBOBpin.
most important structural change in S5 is the cleavage of the H–B bond. 15
The most important structural change in S5 is the cleavage of the H–B bond.Os15  has a smaller WBI of the 
Os has a smaller WBI of
H–B 
the H–Bbond 
bondand and
a  lower 
a lower barrier  for for
barrier the 
thecleavage 
cleavageof ofH–B 
H–Bbond 
bondthan 
than15   in 
15RuRu inS5. 
S5.There 
There is 
is no  obvious 
no obvious
relation between the energy barriers and the NPA charges of metal atoms in S5. The Fe complexes 
relation between the energy barriers and the NPA charges of metal atoms in S5. The Fe complexes
cannot be compared with the Ru and Os complexes because they have a different pathway in S5. 
cannot be compared with the Ru and Os complexes because they have a different pathway in S5.
 
Catalysts 2017, 7, 5 9 of 12

Table 1. The natural population analysis (NPA) charges of metal atoms (e), the Wiberg bond
indices (WBIs) of the bonds, the relative free energies between intermediates and transition
states/intermediates (∆G‡ /∆G◦ ).

Stages Complexes e Bond WBI ∆G‡ /∆G◦ (kcal/mol)

2Fe −1.892 0.5876
0.4
TS2,3-Fe −1.782 0.5420
Stage 1 2Ru −1.538 M–H bond 0.6273
5.5
TS2,3-Ru −1.439 0.4674
2Os −1.540 0.6887
9.3
TS2,3-Os −1.518 0.5094
6Fe −1.209 0.3517
20.8
TS6,7-Fe −0.974 0.0287
Stage 2 7Ru −1.411 M–O bond 0.2704
12.2
TS7,8-Ru −1.104 0.0275
7Os −1.427 0.2995
10.0
TS7,8-Os −1.111 0.0271
1Fe −1.280 0.7154
−7.3
8Fe −1.279 0.2182
Stage 3 9Ru −1.867 M–H bond 0.6661
0.5
TS9,10-Ru −1.850 0.5757
9Os −1.832 0.7206
1.0
TS9,10-Os −1.808 0.6002
9Fe −0.890 0.4354
3.1
TS9,10-Fe −0.840 0.2609
Stage 4 11Ru −1.190 M–O bond 0.2583
3.8
TS11,12-Ru −1.184 0.2616
11Os −1.177 0.4049
4.7
TS11,12-Os −1.193 0.3061
11Fe −0.778 / /
−10.9
12Fe −1.290 0.5236
Stage 5 15Ru −1.478 0.5325
H–B bond 4.5
TS15,16-Ru −1.795 0.0863
15Os −1.483 0.4898
3.7
TS15,16-Os −1.742 0.0958

Overall, the above comparison of the important elementary steps of the reactions catalyzed by
different metal complexes indicate that the smaller corresponding WBIs of bonds, and the lower energy
barriers. We believe the reaction energy barriers are influenced by many factors. Due to the structural
complexity of the transition metal complexes, the reaction energy barrier is not only decided by the
electronic effect, but also relates to the sizes of metals, the binding strengths between metals and
ligands, and the steric effects.

3. Computational Details
All DFT calculations were performed using the Gaussian 09 suite of programs [27]
for the ωB97X-D functional [28,29] with the Stuttgart relativistic effective core potential and
associated valence basis sets for Ru (ECP28MWB, (8s7p6d2f1g)/[6s5p3d2f1g]) and Os (ECP60MWB,
(8s7p6d2f1g)/[6s5p3d2f1g]) [30,31], all-electron 6-31++G(d,p) basis set for Fe, the atoms coordinated
to metal atoms and the atoms in the reactant, and the 6-31G(d) basis set for all other atoms [32–34].
The ωB97X-D functional was also used in the previous theoretical modeling of hydrogenation of
Catalysts 2017, 7, 5 10 of 12

carbon dioxide by Fe complex [35] and the hydrogenation of dimethyl carbonate by Ru and Fe
complexes [36]. The calculation results in those studies are in agreement with the experimental
observations. In addition, high-level ab initio coupled cluster calibration study shows that ωB97X-D
performs well in barrier calculation of σ-bond activation promoted by Ru [37] and Fe [38] complexes.
All structures reported in this paper were fully optimized with solvent effect corrections using the
cavity-dispersion-solvent-structure terms in Truhlar and co-workers’ SMD solvation model for benzene
(ε = 2.2706) [39]. Tables and an xyz file giving solvent effect corrected absolute free energies, electronic
energies and atomic coordinates of all optimized structures are given in Supporting Information.
The thermal corrections were calculated at 298.15 K and 1 atm pressure with harmonic approximation.
An ultrafine integration grid (99, 590) was used for numerical integrations. The ground states of
intermediates and transition states were confirmed as singlets through a comparison with the optimized
high-spin analogues. Calculating the harmonic vibrational frequencies for optimized structures and
noting the number of imaginary frequencies (IFs) confirmed the nature of all intermediates (no IF)
and transition state structures (only one IF for each transition state). The latter were also confirmed to
connect reactants and products by intrinsic reaction coordinate (IRC) calculations. The 3D molecular
structure figures displayed in this paper were drawn by using the JIMP2 molecular visualizing and
manipulating program [40]. The NPA charges [41] and Wiberg indices [42] were obtained by using the
NBO 3.1 program.

4. Conclusions
In summary, the mechanistic insights of the reduction of carbon dioxide to formaldehyde catalyzed
by Ru, Fe and Os complexes are investigated by using DFT. The formation of C–H and Ru–O bonds
(TS9,10-Ru ), the cleavage of Fe–O bond and the formation of O–B bond (TS6,7-Fe ), and the formation of
C–H bond (TS2,3-Os ) are the rate-determining steps in the reactions catalyzed by the Ru, Fe, and Os
complexes with total free energy barriers of 24.2 (1Ru-HBpin → TS9,10-Ru ), 24.0 (1Fe-HBpin → TS6,7-Fe )
and 29.0 (1Os-HBpin → TS2,3-Os ) kcal/mol, respectively. Such barriers indicate that 1Fe is a potential
low-cost catalyst for the reduction of carbon dioxide to formaldehyde under mild conditions. With all
of our computational studies, we expect to predict the effects of different metals on the reaction and
provide useful information for the development of base metal complexes for the conversion and
utilization of carbon dioxide.

Supplementary Materials: Additional Supporting Information (Tables and an xyz file giving solvent effect
corrected absolute free energies, electronic energies and atomic coordinates of all optimized structures; Scheme on
catalytic cycle for the reduction of carbon dioxide to formaldehyde catalyzed by 1Os ) are available online at
www.mdpi.com/2073-4344/7/1/5/s1.
Acknowledgments: X.Y. acknowledges financial support from the 100-Talent Program of the Chinese Academy
of Sciences (CAS), and the National Natural Science Foundation of China (NSFC, 21373228, 21673250). H.C. is
supported by the NSFC (21290194, 21473215). C.D. is grateful for the financial support from Natural Science
Program of Handan University (16217).
Author Contributions: Xinzheng Yang and Chunhua Dong conceived and designed the computations;
Chunhua Dong performed the computations; Chunhua Dong, Mingsong Ji, Hui Chen and Jiannian Yao analyzed
the data; Chunhua Dong and Xinzheng Yang wrote the paper.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Kleeberg, C.; Cheung, M.S.; Lin, Z.; Marder, T.B. Copper-mediated reduction of CO2 with pinB-SiMe2 Ph via
CO2 insertion into a coppersilicon bond. J. Am. Chem. Soc. 2011, 133, 19060–19063. [CrossRef] [PubMed]
2. Rezayee, N.M.; Huff, C.A.; Sanford, M.S. Tandem amine and ruthenium-catalyzed hydrogenation of CO2 to
methanol. J. Am. Chem. Soc. 2015, 137, 1028–1031. [CrossRef] [PubMed]
3. Bontemps, S.; Vendier, L.; Sabo-Etienne, S. Ruthenium-catalyzed reduction of carbon dioxide to
formaldehyde. J. Am. Chem. Soc. 2014, 136, 4419–4425. [CrossRef] [PubMed]
Catalysts 2017, 7, 5 11 of 12

4. LeBlanc, F.A.; Piers, W.E.; Parvez, M. Selective hydrosilation of CO2 to a bis(silylacetal) using an anilido
bipyridyl-ligated organoscandium catalyst. Angew. Chem. Int. Ed. 2014, 126, 789–792. [CrossRef] [PubMed]
5. Moret, S.; Dyson, P.J.; Laurenczy, G. Direct synthesis of formic acid from carbon dioxide by hydrogenation in
acidic media. Nat. Commun. 2014, 5, 4017. [CrossRef] [PubMed]
6. Brown, N.J.; Harris, J.E.; Yin, X.; Silverwood, I.; White, A.J.P.; Kazarian, S.G.; Hellgardt, K.; Shaffer, M.S.P.;
Williams, C.K. Mononuclear phenolate diamine zinc hydride complexes and their reactions With CO2 .
Organometallics 2014, 33, 1112–1119. [CrossRef] [PubMed]
7. Dong, C.; Yang, X.; Yao, J.; Chen, H. Mechanistic study and ligand design for the formation of zinc formate
complexes from zinc hydride complexes and carbon dioxide. Organometallics 2015, 34, 121–126. [CrossRef]
8. Tlili, A.; Blondiaux, E.; Frogneux, X.; Cantat, T. Reductive functionalization of CO2 with amines: An entry to
formamide, formamidine and methylamine derivatives. Green Chem. 2015, 17, 157–168. [CrossRef]
9. Berkefeld, A.; Piers, W.E.; Parvez, M.; Castro, L.; Maron, L.; Eisenstein, O. Decamethylscandocinium-
hydrido-(perfluorophenyl)-borate: Fixation and tandem tris(perfluorophenyl)-borane catalysed deoxygenative
hydrosilation of carbon dioxide. Chem. Sci. 2013, 4, 2152–2162. [CrossRef]
10. Li, Y.; Yan, T.; Junge, K.; Beller, M. Catalytic methylation of C–H bonds using CO2 and H2 . Angew. Chem.
Int. Ed. 2014, 53, 10476–10480. [CrossRef] [PubMed]
11. Filonenko, G.A.; Smykowski, D.; Szyja, B.M.; Li, G.N.; Szczygiel, J.; Hensen, E.J.M.; Pidko, E.A. Catalytic
hydrogenation of CO2 to formates by a lutidine-derived Ru-CNC pincer complex: Theoretical insight into
the unrealized potential. ACS Catal. 2015, 5, 1145–1154. [CrossRef]
12. Haunschild, R. Theoretical study on the reaction mechanism of carbon dioxide reduction to methanol using
a homogeneous ruthenium(II) phosphine catalyst. Polyhedron 2015, 85, 543–548. [CrossRef]
13. Musashi, Y.; Sakaki, S. Theoretical study of ruthenium-catalyzed hydrogenation of carbon dioxide into
formic acid. Reaction mechanism involving a new type of sigma-bond metathesis. J. Am. Chem. Soc. 2000,
122, 3867–3877. [CrossRef]
14. Chong, C.C.; Kinjo, R. Catalytic hydroboration of carbonyl derivatives, imines, and carbon dioxide. ACS Catal.
2015, 5, 3238–3259. [CrossRef]
15. Bontemps, S. Boron-mediated activation of carbon dioxide. Coord. Chem. Rev. 2015, 308, 117–130. [CrossRef]
16. Suh, H.-W.; Guard, L.M.; Hazari, N. A mechanistic study of allene carboxylation with CO2 resulting in the
development of a Pd(II) pincer complex for the catalytic hydroboration of CO2 . Chem. Sci. 2014, 5, 3859–3872.
[CrossRef]
17. Courtemanche, M.-A.; Légaré, M.-A.; Maron, L.; Fontaine, F.-G. Reducing CO2 to methanol using frustrated
lewis pairs: On the mechanism of phosphine–borane-mediated hydroboration of CO2 . J. Am. Chem. Soc.
2014, 136, 10708–10717. [CrossRef] [PubMed]
18. Courtemanche, M.-A.; Légaré, M.-A.; Maron, L.; Fontaine, F.-G. A highly active phosphine–borane
organocatalyst for the reduction of CO2 to methanol using hydroboranes. J. Am. Chem. Soc. 2013, 135,
9326–9329. [CrossRef] [PubMed]
19. Declercq, R.; Bouhadir, G.; Bourissou, D.; Légaré, 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. [CrossRef]
20. Lu, Z.; Hausmann, H.; Becker, S.; Wegner, H.A. Aromaticity as stabilizing element in the bidentate activation
for the catalytic reduction of carbon dioxide. J. Am. Chem. Soc. 2015, 137, 5332–5335. [CrossRef] [PubMed]
21. Huang, F.; Zhang, C.; Jiang, J.; Wang, Z.-X.; Guan, H. How does the nickel pincer complex catalyze the
conversion of CO2 to a methanol derivative? A computational mechanistic study. Inorg. Chem. 2011, 50,
3816–3825. [CrossRef] [PubMed]
22. Suh, H.-W.; Guard, L.M.; Hazari, N. Synthesis and reactivity of a masked PSiP pincer supported nickel
hydride. Polyhedron 2014, 84, 37–43. [CrossRef]
23. Anker, M.D.; Arrowsmith, M.; Bellham, P.; Hill, M.S.; Kociok-Kohn, G.; Liptrot, D.J.; Mahon, M.F.;
Weetman, C. Selective reduction of CO2 to a methanol equivalent by B(C6 F5 )3 -activated alkaline earth
catalysis. Chem. Sci. 2014, 5, 2826–2830. [CrossRef]
24. Sypaseuth, F.D.; Matlachowski, C.; Weber, M.; Schwalbe, M.; Tzschucke, C.C. Electrocatalytic carbon
dioxide reduction by using cationic pentamethylcyclopentadienyl–iridium complexes with unsymmetrically
substituted bipyridine ligands. Chem.–Eur. J. 2015, 21, 6564–6571. [CrossRef] [PubMed]
Catalysts 2017, 7, 5 12 of 12

25. Glendening, E.D.; Reed, A.E.; Carpenter, J.E.; Weinhold, F. NBO Version 3.1; TCI, University of Wisconsin:
Madison, WI, USA, 1998.
26. Jin, G.; Werncke, C.G.; Escudié, Y.; Sabo-Etienne, S.; Bontemps, S. Iron-catalyzed reduction of CO2 into
methylene: Formation of C–N, C–O, and C–C bonds. J. Am. Chem. Soc. 2015, 137, 9563–9566. [CrossRef]
[PubMed]
27. Frisch, M.J.; Trucks, G.W.; Schlegel, H.B.; Scuseria, G.E.; Robb, M.A.; Cheeseman, J.R.; Scalmani, G.; Barone, V.;
Mennucci, B.; Petersson, G.A.; et al. Gaussian 09, Revision C.01; Gaussian, Inc.: Wallingford, CT, USA, 2010.
28. Chai, J.-D.; Head-Gordon, M. Systematic optimization of long-range corrected hybrid density functionals.
J. Chem. Phys. 2008, 128, 084106. [CrossRef] [PubMed]
29. Chai, J.-D.; Head-Gordon, M. Long-range corrected hybrid density functionals with damped atom–atom
dispersion corrections. Phys. Chem. Chem. Phys. 2008, 10, 6615–6620. [CrossRef] [PubMed]
30. Andrae, D.; Häußermann, U.; Dolg, M.; Stoll, H.; Preuß, H. Energy-adjusted ab initio pseudopotentials for
the second and third row transition elements. Theor. Chim. Acta 1990, 77, 123–141. [CrossRef]
31. Martin, J.M.L.; Sundermann, A. Correlation consistent valence basis sets for use with the
Stuttgart–Dresden–Bonn relativistic effective core potentials: The atoms Ga–Kr and In–Xe. J. Chem. Phys.
2001, 114, 3408–3420. [CrossRef]
32. Hehre, W.J.; Ditchfield, R.; Pople, J. Self-consistent molecular orbital methods. XII. Further extensions of
Gaussian-type basis sets for use in molecular orbital studies of organic molecules. J. Chem. Phys. 1972, 56,
2257–2261. [CrossRef]
33. Hariharan, P.C.; Pople, J.A. The influence of polarization functions on molecular orbital hydrogenation
energies. Theor. Chim. Acta 1973, 28, 213–222. [CrossRef]
34. Krishnan, R.; Binkley, J.S.; Seeger, R.; Pople, J.A. Self-consistent molecular orbital methods. XX. A basis set
for correlated wave functions. J. Chem. Phys. 1980, 72, 650–654. [CrossRef]
35. Yang, X. Hydrogenation of carbon dioxide catalyzed by PNP pincer iridium, iron, and cobalt complexes: A
computational design of base metal catalysts. ACS Catal. 2011, 1, 849–854. [CrossRef]
36. Yang, X. Metal hydride and ligand proton transfer mechanism for the hydrogenation of dimethyl carbonate
to methanol catalyzed by a pincer ruthenium complex. ACS Catal. 2012, 2, 964–970. [CrossRef]
37. Sun, Y.; Hu, L.; Chen, H. Comparative assessment of DFT performances in Ru- and Rh-promoted σ-bond
activations. J. Chem. Theory Comput. 2015, 11, 1428–1438. [CrossRef] [PubMed]
38. Sun, X.; Sun, X.; Geng, C.; Zhao, H.; Li, J. Benchmark study on methanol C–H and O–H bond activation by
bare [FeIV O]2+ . J. Phys. Chem. A 2014, 118, 7146–7158. [CrossRef] [PubMed]
39. Marenich, A.V.; Cramer, C.J.; Truhlar, D.G. Universal solvation model based on solute electron density and
on a continuum model of the solvent defined by the bulk dielectric constant and atomic surface tensions.
J. Phys. Chem. B 2009, 113, 6378–6396. [CrossRef] [PubMed]
40. Manson, J.; Webster, C.E.; Hall, M.B. JIMP2, Version 0.091, a Free Program for Visualizing and Manipulating
Molecules; Texas A&M University: College Station, TX, USA, 2006.
41. Reed, A.E.; Weinstock, R.B.; Weinhold, F. Natural population analysis. J. Chem. Phys. 1985, 83, 735–746.
[CrossRef]
42. Wiberg, K.B. Application of the pople-santry-segal CNDO method to the cyclopropylcarbinyl and cyclobutyl
cation and to bicyclobutane. Tetrahedron 1968, 24, 1083–1096. [CrossRef]

© 2016 by the authors; licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC-BY) license (http://creativecommons.org/licenses/by/4.0/).