Electrocatalytic Reduction of CO2 to Methanol by Iron Tetradentate Phosphine Complex Through Amidation Strategy

Article abstract of DOI:10.1002/cssc.201802929

The iron complex of tetradentate tris[2-(diphenylphosphino) ethyl]phosphine (PP₃), [Fe(PP₃)(MeCN)₂](BF₄)₂, was able to electrocatalytically reduce CO₂ to formate with a Faradaic efficiency (FE) of approximately 97.3 % in acetonitrile. Upon addition of diethylamine as a cocatalyst, electrocatalytic reduction to methanol was achieved with an FE of 68.5 %, and other products were formamide and formate. A mechanistic study suggested that the [FeH(PP₃)](BF₄) hydride complex was the active species in the electrocatalysis. Added amine as cocatalyst could react with CO₂ to form carbamate, which could then be reduced to formamide and further to methanol. By contrast, free CO₂ could only be reduced to formate as the end-product.

Full text of DOI:10.1002/cssc.201802929

DOI: 10.1002/cssc.201802929
Full Papers
Very Important Paper
Electrocatalytic Reduction of CO2 to Methanol by Iron
Tetradentate Phosphine Complex Through Amidation
Strategy
Jiaojiao Bi,[a, c] Pengfei Hou,[a, c] Fang-Wei Liu,[a, c] and Peng Kang*[a, b, c]
The iron complex of tetradentate tris[2-(diphenylphosphino)
ethyl]phosphine (PP3), [Fe(PP3)(MeCN)2](BF4)2, was able to elec-
trocatalytically reduce CO2 to formate with a Faradaic efficiency
(FE) of approximately 97.3% in acetonitrile. Upon addition of
diethylamine as a cocatalyst, electrocatalytic reduction to
methanol was achieved with an FE of 68.5%, and other prod-
ucts were formamide and formate. A mechanistic study sug-
gested that the [FeH(PP3)](BF4) hydride complex was the active
species in the electrocatalysis. Added amine as cocatalyst
could react with CO2 to form carbamate, which could then be
reduced to formamide and further to methanol. By contrast,
free CO2 could only be reduced to formate as the end-product.
Introduction
The use of CO2 as feedstock for production of fuels and chemi-
cals is a promising strategy for renewable energy storage and
reducing carbon emissions.[1] In the studies of electrochemical
CO2 reduction, CO and formate were the most reported prod-
ucts.[2] Methanol is a valuable chemical, liquid fuel, hydrogen
storage material, and chemical intermediate.[3] However, 6eꢀ/
6H+ reduction of CO2 to CH3OH remains rare, largely owing to
the high stability of formate and CO. Reports of electrochemi-
cal CO2 reduction to methanol predominately showed the use
of nanomaterial catalysts (Table S1 in the Supporting Informa-
tion). For example, Pd/SnO2 nanosheets produced CH3OH with
a Faradaic efficiency (FE) of 54.8%,[4] and RuO2/TiO2 nanotubes
with an FE of 60.5%.[5] In addition, use of electrodeposited
Cu2O resulted CH3OH with 36.3% FE,[6] whereas the FE drop-
ped to 43.2% for oxide-derived Cu[7] and to 15.9% for Cu-Au
alloy.[8] These nanocatalysts commonly suffered from low FE for
CH3OH, and the pathway to form CH3OH remained largely
unclear.
made significant advances (Table S2 in the Supporting Informa-
tion).[10] Huff and Sanford reported a ruthenium phosphine
complex that can hydrogenate CO2 to formate, methyl for-
mate, and CH3OH.[10d] Leitner and co-workers demonstrated
that a P3C ruthenium complex catalyzed the hydrogenation of
CO2 to CH3OH via either formic acid or methyl formate inter-
mediates. Recently, Sanford and co-workers used a Ru catalyst
in the presence of an amine to hydrogenate CO2 to CH3OH,[10g]
and formamide was a critical intermediate. Prakash and co-
workers utilized polyamine as CO2 capture agent and subse-
quently hydrogenated captured CO2 to CH3OH.[9i] Generation of
formamide could bypass the kinetically inert formate so that
the reduction could continue. So far, the homogeneous cata-
lytic systems in the literature are based on thermal catalysis,
which requires high-temperature and high-pressure conditions.
Herein, we used an electrochemical method to convert CO2 to
methanol with high selectivity under mild conditions.
Tetradentate phosphine is a potent scaffold in molecular cat-
alysis, and its Rh,[11] Co,[12] and Fe[13] complexes were reported
previously. Beller and co-workers reported an Fe complex
based on tris[2-(diphenylphosphino)ethyl]phosphine (PP3) for
hydrogenation of CO2 and bicarbonates to yield formate, alkyl
formate, and formamide in decent yields and turnover num-
bers.[14] However, PP3-based complexes have not been reported
in electrochemical CO2 reduction.[9a,14] Herein, we investigated
the PP3 iron complex 1MeCN (MeCN=acetonitrile) (Figure 1a) for
electrocatalytic reduction of CO2 to formate, and further reduc-
tion to CH3OH by using the amidation strategy (Scheme 1).
As a parallel route of electrochemical reduction, hydrogena-
tion of CO2 is also a heavily studied topic,[9] and recent re-
search on homogeneous CO2 hydrogenation to CH3OH has
[a] J. Bi, P. Hou, F.-W. Liu, Prof. P. Kang
Key Laboratory of Photochemical Conversion and Optoelectronic Materials
Technical Institute of Physics and Chemistry
Chinese Academy of Sciences
29 Zhongguancun East Rd, Beijing 100190 (P.R. China)
[b] Prof. P. Kang
School of Chemical Engineering and Technology
Tianjin University
Tianjin 300350 (P.R. China)
E-mail: kang.peng@tju.edu.cn
Results and Discussion
The crystal structure of [Fe(PP3)(MeCN)2](BF4)2 (1MeCN) revealed
an octahedral geometry with facial coordination of the PP3
ligand and two MeCN moieties coordinated as labile ligands
(Figure 1a). Upon switching to acetate as the counter-anion,
the crystal of the as-prepared 1OAc complex showed coordina-
[c] J. Bi, P. Hou, F.-W. Liu, Prof. P. Kang
University of Chinese Academy of Sciences
19A Yuquan Rd, Beijing 100049 (P.R. China)
Supporting Information and the ORCID identification number(s) for the
author(s) of this article can be found under:
https://doi.org/10.1002/cssc.201802929.
ChemSusChem 2019, 12, 1 – 8
1
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
assigned to a FeII/Fe0 couple. If scan rates were varied between
25 and 1000 mVsꢀ1, the presumed 1eꢀ FeII/FeI redox process
could not be resolved in CVs (Figure S1 in the Supporting In-
formation), suggesting that the reduction of FeII to Fe0 is a
concerted 2eꢀ process rather than a step-wise electron
transfer.
CVs at various scan rates (25–1000 mVsꢀ1) in MeCN under Ar
(Figure S1a in the Supporting Information) showed that the
peak currents ip at an E1/2 of ꢀ1.15 V varied linearly with the
square root of the scan rates u1/2 (Figure S1b,c in the Support-
ing Information), which is consistent with diffusional reduction
of solution species according to the Randles–Sevcik
Equation (1):
Figure 1. Crystal structure of (a) 1MeCN and (b) 1OAc. Selected bond lengths
[ꢁ]: 1MeCN: Fe1ꢀP1 2.1783(7); Fe1ꢀP2 2.3238(7); Fe1ꢀP3 2.2711(7); Fe1ꢀP4
2.3331(8); Fe1ꢀN1 1.963(2); Fe1ꢀN2 1.952(2). 1OAc: Fe1ꢀO1 2.0610(19); Fe1ꢀ
O2 2.0728(19). Full structure shown in Figure S7 in the Supporting
Information.
ip ¼ 0:4463ðF3=RTÞ1=2np3=2ADFe ½Feꢁu1=2
1=2
ð1Þ
in which F is the Faraday constant [Cmolꢀ1], R is the universal
gas constant [JKꢀ1 molꢀ1], np is the number of electrons trans-
ferred (2 for FeII/Fe0), T is the temperature [K], A is the elec-
trode area [cm2], DFe is the diffusion coefficient of the complex
[cm2 sꢀ1], and u is the scan rate [Vsꢀ1].[15]
Scheme 1. Electrochemical reduction of CO2 to methanol through amidation
strategy.
The electrochemical response of Fe hydride species was also
investigated. Upon 2eꢀ reduction of 1MeCN, a Fe0 species
formed and was then protonated by water to form a hydride
species [FeH(PP3)]+ (1H). There was adventitious water in the
tion by an OAcꢀ anion (Figure 1b), which could be a counter-
part of the formate-coordinated complex (1F).
1
electrolyte (ꢂ39 mm as determined by H NMR spectroscopy;
A cyclic voltammogram (CV) of 1MeCN in MeCN under Ar
(Figure 2) showed a quasi-reversible wave at a half-wave po-
tential E1/2 of ꢀ1.13 V versus normal hydrogen electrode (NHE),
and the peak-to-peak separation DEp was 51 mV. For compari-
son, DEp of Fc+/Fc (Fc=ferrocene) was 98 mV under the same
conditions (Figure 2, no internal resistance compensation),
nearly twice the reduction wave. Coulometry experiments
showed 2.5eꢀ passed per complex for the reduction of 1MeCN
(Figure S26a in the Supporting Information, background
charge uncorrected). CV and coulometry together suggested
that the reduction of 1MeCN was a 2eꢀ process and tentatively
Figure S25 in the Supporting Information). During the CV
under Ar (Figure 3a), a new irreversible anodic wave occurred
at Ep =ꢀ0.06 V versus NHE (Figure 3a) upon 2eꢀ reduction of
1MeCN. The new anodic wave was attributed to the oxidation of
1H to regenerate 1MeCN. The assignment of the anodic wave
Figure 3. Electrochemical investigations. (a) CVs of 1MeCN (2.5 mm) at a scan
rate of 100 mVsꢀ1 in MeCN under Ar (black) and CO2 (red). (b) CV of 1H
(2.5 mm) at a scan rate of 100 mVsꢀ1 in MeCN under Ar. (c) CPEs and prod-
uct distribution under CO2. (d) Total current density versus applied potential
in CPEs. Glassy carbon working electrode (0.071 cm2) for CV and carbon
cloth (0.50 cm2) for CPE, 1 atm CO2, nBu4NPF6 (0.1m), room temperature.
Figure 2. CV of 1MeCN under Ar with ferrocene(*) in MeCN. Conditions: 1MeCN
(2.5 mm) in nBu4NPF6/MeCN (0.1m), glassy carbon working electrode, Pt-
wire counter electrode, Ag/AgNO3 reference electrode, scan rate: 50 mVsꢀ1
no internal resistance compensation.
,
&
ChemSusChem 2019, 12, 1 – 8
www.chemsuschem.org
2
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
was confirmed by the CV of the chemically synthesized 1H
complex (Figure 3b), suggesting that electrochemically gener-
ated 1H is likely an intermediate in electrocatalysis.
A CV of 1MeCN under 1 atm CO2 (Figure 3a) exhibited a cata-
lytically enhanced current at the wave of FeII/Fe0, approximate-
ly 1.7 times that under Ar at 100 mVsꢀ1 scan rate. CVs in the
absence of 1MeCN exhibited no current change, indicating that
the reactivity was not owing to the glassy carbon electrode.
CVs of 1MeCN at various scan rates (25–1000 mVsꢀ1) under CO2
(Figure S2a in the Supporting Information) indicated that cata-
lytic current normalized by the scan rate increased upon de-
creasing the scan rate (Figure S2b in the Supporting Informa-
tion). These results indicated that 1MeCN is an efficient electroca-
talyst for CO2 reduction.
Controlled potential electrolyses (CPEs) were performed to
investigate the product distribution. Analyses of electrolyzed
solutions by NMR spectroscopy and GC revealed that the main
product was formate, with small amounts of CO (<1%) and H2
(2–10%) as gaseous products. Formate was not detected in
control experiments in the absence of 1MeCN. The formate FE
was between 90 and 98%, with the applied potential varied
between ꢀ1.0 and ꢀ1.6 V (vs. NHE), and the current density in-
creased as the potential became more negative. At ꢀ1.25 V,
1MeCN displayed the best formate FE of 97.3% (Figure 3c), in
agreement with the catalytic wave observed in the CVs. If the
applied potential was lowered to ꢀ1.6 V, the FE for H2 in-
creased from 2.7 to 10%. The current density increased as the
concentration of catalyst increased (Figure 4b), and the for-
mate FE remained essentially unchanged, indicating that the
catalyst concentration follows first-order kinetics and has little
effect on the product distribution (Figure 4a). During CPE for
11.5 h at ꢀ1.25 V with 1MeCN, the current density slowly de-
creased, probably owing to depletion of available protons (Fig-
ure S4 in the Supporting Information).[16]
Figure 5. CVs of 1MeCN (2.5 mm) in MeCN with added H2O (0–10 vol%) under
(a) Ar and (b) CO2. Scan rate: 100 mVsꢀ1, nBu4NPF6 (0.1m), room
temperature.
creased from 7.1 to 17.8%, which could be owing to water re-
duction at the carbon cloth electrode (Figure 4c). After adding
10 vol% water to the electrolyte, the current in the CPE was
more sustained than without added water (Figure S4 in the
Supporting Information).
Catalytic kinetics were evaluated by measurements of the
peak current ratio of icat and id, (icat is the peak current under
saturated CO2, id is the diffusion peak current under Ar).[2i] In
Equations (2)–(4), icat is given by Equation (3) with ncat =2 for
reduction of CO2 to formate and np =2 for FeII/Fe0:
½CO2ꢁ
Water can serve as a proton source for sustained CPE. Incre-
mental addition of water from 1–10 vol% under Ar did not sig-
nificantly change the catalytic current (Figure 5a), which sug-
gested that 1MeCN did not actively catalyze the H2 evolution re-
action. Addition of water under CO2 led to a slight current en-
hancement in the CVs (Figure 5b). The partial current densities
of both H2 and formate increased upon adding more water
from 1–10 vol% at ꢀ1.25 V (Figure 4d), and the H2 FE in-
ð2Þ
Rate ¼ ꢀd
¼ kcat½Feꢁ ¼ kCO ½Feꢁ½CO2ꢁ
2
dt
À
Â
1=2
1=2
icat ¼ ðncatFAÞ½FeꢁðkcatDFeÞ ¼ ðncatFAÞ½Feꢁ kCO DFe CO2ꢁÞ
2
ð3Þ
ð4Þ
icat
id
1=2
¼ 2:242ðkcatRTncat2=np FuÞ
3
in which the upper limit of the observed turnover frequency
(TOF) kcat under 1 atm CO2 was estimated as 2.8 sꢀ1 by using
CVs at the scan rate of 20 mVsꢀ1 (Figure S3 in the Supporting
Information),[2i] and this gives the upper limit for kcat because
the system was not under the “pure kinetic conditions”.
Adding diethylamine (HNEt2) significantly altered the reac-
tion pathways. Adding HNEt2 from 1 to 10 vol% under CO2 led
to strong catalytic current enhancement in the CVs (Figure 6b),
whereas hardly any change was observed under Ar (Figure 6a).
Upon addition of 1 vol% HNEt2 to CO2-saturated MeCN, the
TOF was calculated as 58.0 sꢀ1 by using the foot-of-the-wave
analysis (FOWA) method (Figure S5 in the Supporting Informa-
tion).[17] CPEs were performed at ꢀ1.25 V versus NHE in CO2-
saturated MeCN with 10 vol% added HNEt2. The main products
were formate, CH3OH, and N,N-diethylformamide (DEF) (Fig-
Figure 4. (a) FEs and (b) total current densities with varied 1MeCN concentra-
tion. (c) FEs and (d) total and partial current densities with varied amounts
of added H2O. Carbon cloth (0.50 cm2 geometric area), 1 atm CO2, nBu4NPF6
(0.1m), room temperature.
ChemSusChem 2019, 12, 1 – 8
www.chemsuschem.org
3
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
lyte was bubbled off by using Ar, forming a CO2-lean electro-
lyte for CPE. During CPE at ꢀ1.25 V versus NHE under Ar for
10 h, the FE of CH3OH reached 68.5%, higher than that under
CO2, the FE of formamide was 4.1%, and the FE of formate
dropped to 18% (Figure 7b). The current was stable over 10 h
(Figure S6 in the Supporting Information), and the observed
turnover number (TON) was 18 with an overpotential of 0.87 V
(CH3OH: E0 =ꢀ0.38 V vs. NHE, pH 7). This result suggests that
the carbamate is a critical species that leads to the formation
of CH3OH, and also indicates that free CO2 only leads to the
formation of formate. After the electrolysis, the X-ray photo-
electron spectroscopy (XPS) spectra of the carbon working
electrode showed no traces of Fe (Figure S26 in the Support-
ing Information), suggesting that the catalyst did not decom-
pose onto the electrode to form any nanomaterials.
Figure 6. CVs of 1MeCN (2.5 mm) in MeCN with added HNEt2 (0–10 vol%)
under (a) Ar and (b) CO2. Glassy carbon electrode (0.071 cm2), room
temperature.
ure 7a and Figure S23 in the Supporting Information). During
CPE for 12 h, FEs of both CH3OH and formate increased over
time, whereas the FE of DEF initially increased but then de-
creased, suggesting that DEF is an intermediate and can be
further reduced. At the end of CPE for 12 h, the FE of CH3OH
reached 55%, the FE of DEF dropped to 4%, and the FE of for-
mate reached 36%. H2 (<7%) was detected as the gaseous
product, and no CO was found (Figure 7a). If more 1MeCN was
added, the current density increased (Figure 7d), but the prod-
uct distribution showed only minor changes (Figure 7c).
In a series of control experiments, without HNEt2 or 1MeCN
,
no CH3OH was detected. With formate and HNEt2 both added
under Ar with 1MeCN, no CH3OH or DEF was found, suggesting
that formate is an end product that cannot be converted.
However, under Ar, DEF can be electrochemically reduced to
CH3OH with a FE of approximately 90% in the presence of
1MeCN at ꢀ1.25 V, suggesting that DEF is an important inter-
mediate product.
Relevant intermediates were investigated by using electro-
chemical and NMR spectroscopic techniques. Authentic 1H was
chemically synthesized (Figures S10 and S11 in the Supporting
To decrease the FE of formate, CO2 was purged into HNEt2-
containing electrolyte, and then excessive CO2 in the electro-
Figure 7. CPEs with added amine. (a) Product distribution over time under CO2. (b) Product distribution over time under Ar, by adding CO2 first, and then re-
moving CO2 by Ar purging. (c) Influence of catalyst concentration on product distribution under CO2. (d) Current densities versus catalyst concentration under
CO2. 1MeCN (2.5 mm), HNEt2 (10 vol%), ꢀ1.25 V versus NHE, carbon cloth (0.50 cm2), nBu4NPF6 (0.1m) in MeCN, Pt-wire counter electrode, Ag/AgNO3 reference
electrode, room temperature.
&
ChemSusChem 2019, 12, 1 – 8
www.chemsuschem.org
4
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
Information) and showed a hydride resonance at ꢀ12.47 ppm
1
in the H NMR spectrum. 1H was also electrochemically synthe-
sized by reducing 1MeCN (Figures S8 and S9 in the Supporting
Information) under Ar in MeCN, and the hydride resonance
(Figures S12 and S13 in the Supporting Information) matched
that of authentic 1H. 1H species showed high stability toward
added water, as shown in an NMR reaction (Figures S15 and
S16 in the Supporting Information), consistent with electro-
chemical observations with added water. Additionally, the reac-
tion of 1MeCN with NaH led to the dihydride complex
[Fe(H)2(PP3)][18] (1H2), identified by the dihydride resonance at
ꢀ10.8 ppm in the 1H NMR spectrum (Figure S17 in the Sup-
porting Information).[19] The hydride resonances of those spe-
cies are compared in Figure S14 in the Supporting Information.
Because the 1H2 complex was not observed upon electroreduc-
tion under Ar, it is probably not accessible by the electrochem-
ical method.
Complex 1H is a decent hydride donor. 1H reacted with CO2
and yielded formate, and the hydride resonance of 1H entirely
disappeared, suggesting that CO2 insertion into the FeꢀH
bond is quantitative (Figure S18 in the Supporting Informa-
tion). 1H was also found to react with carbonyl-containing
compounds in previous studies.[14b] However, 1H cannot react
with formate to form CH3OH in measurable quantities, proba-
bly because formate bears a negative charge and resists fur-
ther reduction.
Scheme 2. Proposed mechanism for the reduction of CO2 to formate by
1MeCN
.
Addition of HNEt2 is critical for formation of CH3OH. A new
carbamate species formed upon the reaction of HNEt2 and
1
CO2, and the H NMR spectrum showed that the ꢀCH2 proton
resonances of HNEt2 were shifted downfield by approximately
3.04 ppm (Figures S21 and S22 in the Supporting Information),
and the 13C NMR spectrum showed that the ꢀCH2 and ꢀCH3
resonances were shifted by approximately 2.51 and 2.03 ppm,
respectively (Figure S23 in the Supporting Information).[9i,10h,20]
Subsequent reduction of carbamate yielded formamide and
eventually led to CH3OH. This result suggests that carbamate
was the reactant rather than free CO2, distinct from previous
hydrogenation studies.[10]
Scheme 3. Proposed pathways for electrochemical reduction of CO2 by
1MeCN
.
Conclusions
The iron complex [Fe(PP3)(MeCN)2](BF4)2 (1MeCN), based on the
tetradentate phosphine ligand tris[2-(diphenylphosphino)
ethyl]phosphine (PP3), can catalyze the electrochemical reduc-
tion of CO2 to formate and CH3OH. Without amine added,
1MeCN reduced CO2 to formate with a Faradaic efficiency (FE) of
97.3%. With added amine as cocatalyst, CH3OH was formed
with an FE of 68.8%, higher than for reported heterogeneous
catalysts.[4–8] Mechanistic studies indicated that, in the presence
of amine, the formed carbamate species bypassed the kineti-
cally inert formate and enabled reduction to formamide and fi-
nally to CH3OH. The amidation strategy is a solid approach to
form CH3OH and could be further adopted with a variety of
metal complexes and even nanomaterial catalysts.
Parallel electroreduction mechanisms of the formate and
CH3OH routes are proposed in Schemes 2 and 3. In Scheme 2,
the formate pathway involves direct reduction of free CO2 to
formate. Under reductive conditions, 1MeCN is reduced by 2eꢀ/
1H+ to generate 1H. 1H then reacts with CO2 to form the for-
mate adduct [(PP3)Fe(HCO2)]+ (1F). The formate carbon comes
from CO2, as demonstrated by 13CO2 labeling, appearing at
163 ppm in the 13C NMR spectrum (Figure S19 in the Support-
ing Information). The formate is then dissociated to regenerate
1MeCN (Scheme 2), completing the formate pathway.
In the CH3OH pathway (Scheme 3), CO2 reacts with HNEt2 to
form carbamate, which is then attacked by the electrochemi-
cally generated 1H species to form DEF. DEF is further reduced
by 1H to yield CH3OH and release HNEt2. This mechanism is
quite different from previous cascade hydrogenation studies,
in which formate was an intermediate and was further convert-
ed to formamide in the presence of a Lewis acid.[10]
Experimental Section
Materials and methods
The PP3 was purchased from Acros, and other chemicals were pur-
chased from Energy Chemical if not mentioned otherwise. THF and
ChemSusChem 2019, 12, 1 – 8
www.chemsuschem.org
5
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

Full Papers
Synthesis of [Fe(PP3)H](BF4) (1H): 1MeCN (113.6 mg, 0.10 mmol)
was dissolved in MeCN (10 mL) under argon. NaBH4 (25 mg,
0.66 mmol) was added, and the mixture was heated to reflux
at 658C for 30 min. The reaction flask was transferred to the
glovebox, and the solid was filtered. By slow evaporation of
the solvent in a stream of Ar, a brownish solid appeared (yield
80%). 1H NMR (CD3CN, 400 MHz): d=6.90–7.8 (m, Ph), 2.15–
3.25 (m, CH2), 12.27 ppm (tdd, J=25.92 Hz, 19.44 Hz, 11.34 Hz,
FeꢀH); 31P NMR (CD3CN, 162 MHz): d=172.57 (dd, J=44.55 Hz,
23.49 Hz, Pa), 81.74 (dd, J=24.14 Hz, 10.85 Hz, Pb), 70.01 ppm
(dt, J=19.76 Hz, 11.99 Hz, Pc).
MeCN were of HPLC grade, freshly distilled, and stored over molec-
ular sieves in an Ar-filled glovebox. Distilled water was further puri-
fied by using a Master-S15 UV Water Purification system (Hetai,
China). Argon and CO2 (Huanyujinghui, Beijing, research grade)
were purchased at 99.998% purity with less than 3 ppm H2O and
used as received. Air-sensitive materials were prepared and manip-
ulated by using Schlenk techniques and in an Ar-filled glovebox
(Etelux, Lab2000, <1 ppm O2 and H2O). CD3CN was dried over
CaH2 and then distilled and stored over 4 ꢁ molecular sieves in an
Ar-filled glovebox. Tetrabutylammonium hexafluorophospate
(nBu4NPF6, Fluka, electrochemical grade) was dried at 608C under
vacuum for 12 h and stored in the glovebox. 13CO2 was purchased
from Beijing Gaisi Chemical Gases Center and 99% isotopically en-
riched in 13C. All other chemicals were used as received from com-
mercial vendors.
Synthesis of [Fe(PP3)(CH3COO)][Fe3(CH3COO)7] (1OAc): A suspen-
sion of Fe(CH3COO)2 (0.174 g, 1.0 mmol) in MeCN (10 mL) was
added to a suspension of PP3 (0.67 g, 1.0 mmol) in MeCN (20 mL)
in an Ar-filled glovebox. With the continuous addition of
Fe(CH3COO)2, the solid dissolved and changed into a pink solution.
The resultant solution was stirred at room temperature for 3 h. The
solvent was removed in a vacuum line to produce a bright-pink
solid, which was washed with n-hexane three times (3ꢂ10 mL). Re-
crystallization of the residue by slow diffusion of diethyl ether into
saturated acetonitrile solution gave a pink solid suitable for
crystallography.
NMR spectra were recorded on NMR spectrometers (Bruker
1
AVANCE-400, AVANCE-500). H NMR spectra were referenced to re-
sidual solvent signals. 31P chemical shifts were referenced to a
H3PO4 external standard. Iron complexes were prepared according
to the literature.[14] XPS spectra were obtained on a Kratos Analyti-
cal Axis UltraDLD spectrometer with monochromatized X-ray AlKa
radiation (1486.6 eV). Elemental analyses including C, H, and N
were measured by a VARIO EL cube Elemental Analyzer. Electro-
chemical experiments were performed by using a CHI 660E poten-
tiostat (CH Instruments, Shanghai). The three-electrode system con-
sisted of a glassy carbon working electrode, a coiled Pt-wire coun-
ter electrode, and an Ag/AgNO3 reference electrode (10 mm
AgNO3). All electrochemical experiments were performed in MeCN
with nBu4NPF6 (0.1m) as supporting electrolyte. For CV experi-
ments, working and counter electrodes were separated from the
reference electrode. Iron complex concentrations were generally at
2.5 mm in MeCN (5 mL). For CPE, carbon cloth (0.5 cm2 geometric
area) was used as working electrode instead of glassy carbon, and
the reference and counter electrodes were separated from the
working electrode by a glass frit.
Synthesis of [Fe(PP3)(H)2] (1H2): 1MeCN (113.6 mg, 0.10 mmol) was
dissolved in MeCN (10 mL) under an argon atmosphere. NaH
(16 mg, 0.66 mmol) was added, and the mixture was heated to
reflux at 608C for 30 min until a dark-red color appeared. The reac-
tion mixture was filtered in the glovebox, the filtrate was slowly
evaporated in a gentle stream of Ar, and a brown solid appeared
1
(yield 60%). H NMR (CD3CN, 400 MHz): d=6.90–7.8 (m, Ph), 2.15–
3.25 (m, CH2), 10.85 ppm (q, J=17.17 Hz, H).
Electrosynthesis of 1H: Controlled potential electrolysis was
performed in H2O/MeCN (5 mL, 1 vol% H2O) at ꢀ1.25 V [1MeCN
(2.5 mm), nBu4NPF6 (0.1m), 1 atm Ar, room temperature, no IR
compensation] in an airtight electrochemical cell with vigorous
stirring for 1 h. The solvent was removed in a vacuum line to
Gaseous analysis for CPE experiments was performed by using the
sample (1 mL) from the headspace of the electrochemical cell and
injecting it into an SRI-8610C MG#1 (SRI Inc, USA) with a molecular
sieve column and a helium ionization detector (HID). CH3OH analy-
sis was performed by using the sample (1 mL) from the liquid elec-
trolyte of the electrochemical cell and injecting it into a Shimadzu
BID-2010 Plus GC with flame ionization detector (FID). Formate and
DEF were measured on NMR spectrometers (Bruker AVANCE-400).
At each potential, the CPE measurements were performed at least
twice. The reported FEs are averaged values.
1
produce a brown solid. H NMR (CD3CN, 400 MHz): d=6.90–7.8
(m, Ph), 2.15–3.25 (m, CH2), 12.24 ppm (tdd, J=25.92 Hz,
19.44 Hz, 11.34 Hz, H); 31P NMR (CD3CN, 162 MHz): d=172.57
(dd, J=44.55 Hz, 23.49 Hz, Pa), 81.74 (dd, J=24.14 Hz, 10.85 Hz,
Pb), 70.01 ppm (dt, J=19.76 Hz, 11.99 Hz, Pc).
Acknowledgements
Syntheses
Synthesis of [Fe(PP3)(MeCN)2][BF4]2 (1MeCN): A colorless solution of
[Fe(H2O)6](BF4)2 (0.338 g, 1.0 mmol) in MeCN (5 mL) was added to a
suspension of PP3 (0.67 g, 1.0 mmol) in MeCN (20 mL) in an Ar-
filled glovebox. The resultant solution was stirred at room temper-
ature for 1 h. The solvent was removed in a vacuum line to pro-
duce a yellow solid, which was washed with n-hexane three times
(3ꢂ10 mL). The yield was 0.76 g (87%). Recrystallization of the
solid by slow diffusion of diethyl ether into saturated acetonitrile
This work was financially supported by the National R&D Pro-
gram of China (2016YFB0600901).
Conflict of interest
The authors declare no conflict of interest.
1
solution gave a yellow crystal suitable for crystallography. H NMR
Keywords: amidation · carbon dioxide · electrocatalysis · iron ·
methanol
(CD3CN, 400 MHz): d=6.90–7.8 (m, Ph), 2.15–3.25 ppm (m, CH2);
31P NMR (CD3CN, 162 MHz): d=156.9 (q, J=26.2 Hz, Pa), 68.05 (dd,
J=66.3 Hz, 38.4 Hz, Pb), 52.22 ppm (dd, J=38.2 Hz, 25.9 Hz, Pc); el-
emental analysis: calcd (%): C 55.41, H 5.06, N 5.21; found: C 55.34,
H 5.04, N 5.19.
[1] a) M. Eissen, J. O. Meteger, E. Schmidt, U. Schneidewind, Angew. Chem.
Int. Ed. 2002, 41, 414–436; Angew. Chem. 2002, 114, 402–425; b) T. Sa-
&
ChemSusChem 2019, 12, 1 – 8
www.chemsuschem.org
6
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!

Full Papers
kakura, J. C. Choi, H. Yasuda, Chem. Rev. 2007, 107, 2365–2387; c) G. A.
Olah, G. K. Prakash, A. Goeppert, J. Am. Chem. Soc. 2011, 133, 12881–
12898; d) A. Goeppert, M. Czaun, J. P. Jones, G. K. S. Prakash, G. A. Olah,
Chem. Soc. Rev. 2014, 43, 7995–8048; e) T. Gasser, C. Guivarch, K. Ta-
chiiri, C. D. Jones, P. Ciais, Nat. Commun. 2015, 6, 7958; f) K. Sordakis, A.
Tsurusaki, M. Iguchi, H. Kawanami, Y. Himeda, G. Laurenczy, Chem. Eur. J.
2016, 22, 15605–15608.
3834–3840; i) S. Kar, R. Sen, A. Goeppert, G. K. S. Prakash, J. Am. Chem.
Soc. 2018, 140, 1580–1583.
[10] a) C. Bianchini, D. Masi, A. Meli, M. Peruzzini, F. Zanobini, J. Am. Chem.
Soc. 1988, 110, 6411–6423; b) P. G. Jessop, Y. Hsiao, T. Ikariya, R. Noyori,
J. Am. Chem. Soc. 1996, 118, 344–355; c) C. A. Huff, M. S. Sanford, J. Am.
Chem. Soc. 2011, 133, 18122–18125; d) S. Wesselbaum, T. Stein, J. Klan-
kermayer, W. Leitner, Angew. Chem. Int. Ed. 2012, 51, 7499–7502;
Angew. Chem. 2012, 124, 7617–7620; e) J. R. Khusnutdinova, J. A. Garg,
D. Milstein, ACS Catal. 2015, 5, 2416–2422; f) N. M. Rezayee, C. A. Huff,
M. S. Sanford, J. Am. Chem. Soc. 2015, 137, 1028–1031; g) J. Kothan-
daraman, A. Goeppert, M. Czaun, G. A. Olah, G. K. Prakash, J. Am. Chem.
Soc. 2016, 138, 778–781; h) J. Schneidewind, R. Adam, W. Baumann, R.
Jackstell, M. Beller, Angew. Chem. Int. Ed. 2017, 56, 1890–1893; Angew.
Chem. 2017, 129, 1916–1919.
[11] a) G. Bianchini, C. Mealli, M. Peruzzini, F. Zanobini, J. Am. Chem. Soc.
1987, 109, 5548–5549; b) X. L. Luo, R. H. Crabtree, Inorg. Chem. 1989,
28, 3777–3779.
[12] C. Federsel, C. Ziebart, R. Jackstell, W. Baumann, M. Beller, Chem. Eur. J.
2012, 18, 72–75.
[13] a) C. Bianchini, A. Meli, M. Peruzzini, F. Vizza, F. Zanobini, P. Frediani, Or-
ganometallics 1989, 8, 2080–2082; b) C. Bianchini, E. Farnetti, M. Grazia-
ni, M. Peruzzini, A. Polot, Organometallics 1993, 12, 3753–3761.
[14] a) C. Federsel, A. Boddien, R. Jackstell, R. Jennerjahn, P. J. Dyson, R. Sco-
pelliti, G. Laurenczy, M. Beller, Angew. Chem. Int. Ed. 2010, 49, 9777–
9780; Angew. Chem. 2010, 122, 9971–9974; b) C. Ziebart, C. Federsel, P.
Anbarasan, R. Jackstell, W. Baumann, A. Spannenberg, M. Beller, J. Am.
Chem. Soc. 2012, 134, 20701–20704.
[15] P. Kang, T. J. Meyer, M. Brookhart, Chem. Sci. 2013, 4, 3497–3502.
[16] a) Z. Chen, J. J. Concepcion, M. K. Brennaman, P. Kang, M. R. Norris, P. G.
Hoertz, T. J. Meyer, Proc. Natl. Acad. Sci. USA 2012, 109, 15606–15611;
b) S. Zhang, P. Kang, M. Bakir, A. M. Lapides, C. J. Dares, T. J. Meyer, Proc.
Natl. Acad. Sci. USA 2015, 112, 15809–15814.
[2] a) M. D. Sampson, J. D. Froehlich, J. M. Smieja, E. E. Benson, I. D. Sharp,
C. P. Kubiak, Energy Environ. Sci. 2013, 6, 3748–3755; b) Y. Zhang, A. D.
MacIntosh, J. L. Wong, E. A. Bielinski, P. G. Williard, B. Q. Mercado, N.
Hazari, W. H. Bernskoetter, Chem. Sci. 2015, 6, 4291–4299; c) I. Azcarate,
C. Costentin, M. Robert, J. M. Saveant, J. Am. Chem. Soc. 2016, 138,
16639–16644; d) A. D. Chapovetsky, R. Haiges, M. K. Takase, S. C. Ma-
rinescu, J. Am. Chem. Soc. 2016, 138, 5765–5768; e) S. K. D. Sung, M.
Gil-Sepulcre, M. Nippe, J. Am. Chem. Soc. 2017, 139, 13993–13996; f) J.
Choi, T. M. Benedetti, R. Jalili, A. Walker, G. Wallace, D. L. Officer, Chem.
Eur. J. 2016, 22, 14158–14161; g) S. E. Witt, T. A. White, Z. Li, K. R.
Dunbar, C. Turro, Chem. Commun. 2016, 52, 12175–12178; h) N. Elgrishi,
M. B. Chambers, X. Wang, M. Fontecave, Chem. Soc. Rev. 2017, 46, 761–
796; i) R. Francke, B. Schille, M. Roemelt, Chem. Rev. 2018, 118, 4631–
4701.
[3] a) G. A. Olah, Angew. Chem. 2005, 117, 2692–2696; b) E. Alberico, M.
Nielsen, Chem. Commun. 2015, 51, 6714–6725.
[4] W. Y. Zhang, Q. Qin, L. Dai, R. X. Qin, X. J. Zhao, X. M. Chen, D. H. Ou, J.
Chen, T. T. Chuong, B. H. Wu, N. F. Zheng, Angew. Chem. Int. Ed. 2018,
57, 9475–9479; Angew. Chem. 2018, 130, 9619–9623.
[5] J. P. Qu, X. G. Zhang, Y. G. Wang, C. X. Xie, Electrochim. Acta 2005, 50,
3576–3580.
[6] A. Schizodimou, G. Kyriacou, Electrochim. Acta 2012, 78, 171–176.
[7] K. Zhao, Y. M. Liu, X. Quan, S. Chen, H. T. Yu, ACS Appl. Mater. Interfaces
2017, 9, 5302–5311.
[8] F. X. Jia, X. X. Yu, L. Z. Zhang, J. Power Sources 2014, 252, 85–89.
[9] a) C. C. Tai, T. Chang, B. Roller, P. G. Jessop, Inorg. Chem. 2003, 42, 7340–
7341; b) C. Federsel, R. Jackstell, M. Beller, Angew. Chem. Int. Ed. 2010,
49, 6254–6257; Angew. Chem. 2010, 122, 6392–6395; c) F. Liao, Y.
Huang, J. Ge, W. Zheng, K. Tedsree, P. Collier, X. Hong, S. C. Tsang,
Angew. Chem. Int. Ed. 2011, 50, 2162–2165; Angew. Chem. 2011, 123,
2210–2213; d) W. Wang, S. Wang, X. Ma, J. Gong, Chem. Soc. Rev. 2011,
40, 3703–3727; e) F. Studt, I. Sharafutdinov, F. Abild-Pedersen, C. F. Elkja-
er, J. S. Hummelshoj, S. Dahl, I. Chorkendorff, J. K. Norskov, Nat. Chem.
2014, 6, 320–324; f) F. Studt, M. Behrens, E. L. Kunkes, N. Thomas, S.
Zander, A. Tarasov, J. Schumann, E. Frei, J. B. Varley, F. Abild-Pedersen,
J. K. Nørskov, R. SchlÅgl, ChemCatChem 2015, 7, 1105–1111; g) R. Berg,
G. Prieto, G. Korpershoek, L. I. Wal, A. J. Bunningen, S. Laegsgaard-Jor-
gensen, P. E. Jongh, K. P. Jong, Nat. Commun. 2016, 7, 13057; h) B. An, J.
Zhang, K. Cheng, P. Jin, C. Wang, W. Lin, J. Am. Chem. Soc. 2017, 139,
[17] E. S. Rountree, B. D. McCarthy, T. T. Eisenhart, J. L. Dempsey, Inorg. Chem.
2014, 53, 9983–10002.
[18] E. L. Gutsul, N. V. Belkova, G. M. Babakhina, L. M. Epstein, E. S. Shubina,
C. Bianchini, M. Peruzzini, F. Zanobini, Russ. Chem. Bull. 2003, 52, 1204–
1206.
[19] C. Bianchini, F. Laschi, M. Peruzzini, F. M. Ottaviani, A. Vacca, P. Zanellot,
Inorg. Chem. 1990, 29, 3394–3402.
[20] L. Zhang, Z. Han, X. Zhao, Z. Wang, K. Ding, Angew. Chem. Int. Ed. 2015,
54, 6186–6189; Angew. Chem. 2015, 127, 6284–6287.
Manuscript received: December 14, 2018
Revised manuscript received: March 28, 2019
Version of record online: && &&, 0000
ChemSusChem 2019, 12, 1 – 8
www.chemsuschem.org
7
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
&
These are not the final page numbers! ÞÞ

FULL PAPERS
J. Bi, P. Hou, F.-W. Liu, P. Kang*
On the iron throne: The iron complex
of tetradentate tris[2-(diphenylphosphi-
no)ethyl]phosphine (PP3), [Fe(PP3)
(MeCN)2](BF4)2, can electrocatalytically
reduce CO2 to formate with a Faradaic
efficiency (FE) of approximately 97.3%
in acetonitrile. Upon addition of diethyl-
amine as a cocatalyst, electrocatalytic
reduction to methanol is achieved with
a FE of 68.5%.
&& – &&
Electrocatalytic Reduction of CO2 to
Methanol by Iron Tetradentate
Phosphine Complex Through
Amidation Strategy
&
ChemSusChem 2019, 12, 1 – 8
www.chemsuschem.org
8
ꢀ 2019 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ÝÝ These are not the final page numbers!