From the Cover: Ultraefficient homogeneous catalyst for the CO2-to-CO electrochemical conversion期刊界 All Journals 搜尽天下杂志传播学术成果专业期刊搜索期刊信息化学术搜索

From the Cover: Ultraefficient homogeneous catalyst for the CO2-to-CO electrochemical conversion

Authors:

Cyrille Costentin Guillaume Passard Marc Robert Jean-Michel Savéant

Affiliation:

Université Paris Diderot, Sorbonne Paris Cité, Laboratoire d''Electrochimie Moléculaire, Unité Mixte de Recherche Université—Centre National de la Recherche Scientifique, 75205 Paris Cedex 13, France

Abstract:

A very efficient electrogenerated Fe⁰ porphyrin catalyst was obtained by substituting in tetraphenylporphyrin two of the opposite phenyl rings by ortho-, ortho''-phenol groups while the other two are perfluorinated. It proves to be an excellent catalyst of the CO₂-to-CO conversion as to selectivity (the CO faradaic yield is nearly quantitative), overpotential, and turnover frequency. Benchmarking with other catalysts, through catalytic Tafel plots, shows that it is the most efficient, to the best of our knowledge, homogeneous molecular catalyst of the CO₂-to-CO conversion at present. Comparison with another Fe⁰ tetraphenylporphyrin bearing eight ortho-, ortho''-phenol functionalities launches a general strategy where changes in substituents will be designed so as to optimize the operational combination of all catalyst elements of merit.The reductive conversion of CO₂ to CO is an important issue of contemporary energy and environmental challenges (–). Several low-oxidation-state transition metal complexes have been proposed to serve as homogeneous catalyst for this reaction in nonaqueous solvents such as N,N''-dimethylformamide (DMF) or acetonitrile (–).Among them, electrochemically generated Fe⁰ complexes have been shown to be good catalysts, provided they are used in the presence of Brönsted or Lewis acids (17 –19). More recent investigations have extended the range of Brönsted acids able to boost the catalysis of the CO₂-to-CO conversion by electrogenerated Fe⁰TPP (Scheme 1) without degrading the selectivity of the reaction. They have also provided a detailed analysis of the reaction mechanism (, ).Open in a separate window Scheme 1.Iron-based catalysts for CO₂-to-CO reduction.This is notably the case with phenol, which gave rise to the idea of installing prepositioned phenol groups in the catalyst molecule as pictured in Scheme 1 under the heading “CAT.” The result was indeed a remarkably efficient and selective catalyst of the CO₂-to-CO conversion (). At first blush, the comparison with the role of phenol in the case of FeTPP would entail attributing this considerable enhancement of catalysis to a local concentration of phenol much larger than can be achieved in solution. In fact, as analyzed in detail elsewhere (), the role played by the internal phenol moieties is twofold. They indeed provide a very large local phenol concentration, favoring proton transfers, but they also considerably stabilize the initial Fe⁰–CO₂ adduct through H bonding. Although the favorable effect of pendant acid groups has been noted in several cases (see ref. and references therein), this was, to our knowledge, the first time their exact role was deciphered. The difference in the role played by the phenol moieties takes place within the framework of two different mechanisms (see Scheme 2 for CAT and FCAT and Scheme 3 for FeTPP) (). With FeTPP, the first step is, as with CAT and FCAT, the addition of CO₂ on the electrogenerated Fe⁰ complex (et₁ in Schemes 2 and and3).3). The strong stabilization of the Fe⁰–CO₂ adduct formed according to reaction 1 (in Schemes 2 and and3)3) in the latter cases compared with the first has a favorable effect on catalysis, but one consequence of this stabilization is that catalysis then required an additional proton (reactions 2₁ and 2₂ in Scheme 2), the final, catalytic loop-closing step being the cleavage of one of the C–O bonds of CO₂ concerted with both electron transfer from the electrode and proton transfer from one of the local phenol groups (et₂ in Scheme 2). In the FeTPP case, the C–O bond-breaking step (reaction 2 in Scheme 3) is different: it involves an intramolecular electron transfer concerted with proton transfer and cleavage of the C–O bond. The catalytic loop is closed by a homogeneous electron transfer step (et₂ in Scheme 3) that regenerates the initial Fe^I complex.Open in a separate window Scheme 2.Mechanism for the reduction of CO₂ with CAT and FCAT.Open in a separate window Scheme 3.Mechanism for the reduction of CO₂ with FeTPP.The object of the present contribution is to test the idea that introduction of different substituents on the periphery of the porphyrin ring may improve the efficiency of the catalysis of CO₂-to-CO conversion. In such a venture, we will have to take into account both the overpotential at which the reaction takes place and the catalytic rate expressed as the turnover frequency as detailed in the following sections. Taking these two aspects simultaneously into consideration is essential in view of the possibility that substitution may improve one of the two factors and degrade the other, or vice versa. As a first example, we examined the catalytic performances of the FCAT molecule (Scheme 1), in which four of the eight phenol groups have been preserved in the same ortho-, ortho''- positions on two of the opposite phenyl rings, while the two other phenyl rings have been perfluorinated (the synthesis and characterization of this molecule is described in SI Text). A query that first comes to mind is as follows. The inductive effect of the fluorine atoms is expected to ease the reduction of the molecule to the Fe⁰ oxidation state, and thus to be favorable to catalysis in terms of overpotential. However, will this benefit be blurred by a decrease of its reactivity toward CO₂? Indeed, the same inductive effect of the fluorine atoms tends to decrease the electronic density on the Fe⁰ complex and might therefore render the formation of the initial Fe⁰-CO₂ adduct less favorable. Change in the rates of the follow-up reactions of Scheme 1 may also interfere. A first encouraging indication that the fluorine substitution has a globally favorable effect on catalysis derives from the comparison of the cyclic voltammetric responses of FCAT and CAT as represented in Fig. 1: the peak potential is slightly more positive for FCAT [−1.55 V vs. normal hydrogen electrode (NHE)] than for CAT (−1.60 V vs. NHE), whereas the apparent number of electrons at the peak at 0.1 V/s is clearly larger in the first case than in the second (120 vs. 80) (). However, a deeper analysis of the meaning of these figures in terms of effective catalysis is required. The mechanism of the reaction (Scheme 2) has been shown to be the same with FCAT as with CAT, and the various kinetic parameters indicated in Scheme 2, whose values are recalled in 26, ). Comparison of the two catalysts may then be achieved more rationally based on the determination, in each case, of the catalytic Tafel plots, which relates the turnover frequency (TOF) to the overpotential (η). The latter is defined, in the present case of reductive processes, as the difference between the apparent standard potential of the CO₂/CO conversion,

E_{{CO}_{2} /CO}^{0}

and the electrode potential, E:

η = E - E_{{CO}_{2} /CO}^{0} .

Large catalytic currents correspond to “pure kinetic conditions” in which a steady state is achieved by mutual compensation of catalyst transport and catalytic reactions. The cyclic voltammetry (CV) responses are then S-shaped independent of scan rate. They are the same with other techniques such as rotating disk electrode voltammetry and also during preparative-scale electrolyses. The fact that peaks instead of plateaus are observed at low scan rate, as in Fig. 1, derives from secondary phenomena related to the observed high catalytic efficiencies, such as substrate consumption, inhibition by products, and deactivation of the catalyst. These factors and the ways to go around their occurrence to finally obtain a full characterization of the mechanism and kinetics of the catalysis process are discussed in detail elsewhere (). Under pure kinetic conditions, the active catalyst molecules are then confined within a thin reaction-diffusion layer adjacent to the electrode surface. During the time where the catalyst remains stable, the TOF is defined asTOF = N_product/N_active?cat, where N_product is the number of moles of the product, generated per unit of time, and N_active?cat is the maximal number of moles of the active form catalyst contained in the reaction–diffusion layer rather than in the whole electrochemical cell (for more information on the notions of pure kinetic conditions, reaction–diffusion layer, and on the correct definition of TOF, see refs. , , 28). For the present reaction mechanism (Scheme 2) as well as for all reaction schemes belonging to the same category, the TOF–η relationships are obtained from the following equations (28, 29), using the notations defined in Scheme 2 and the data listed in TOF=TOFmax{1+ipl2FSCcat0kf2nd ETexp(α2fE)+ipl2FSCcat0Dcatk1[CO2]exp[FRT(E−E10)]},with

T O F_{\max} = \frac{1}{\frac{1}{k_{1} [{CO}_{2}]} + \frac{1}{k_{2}^{a p}}}, k_{2}^{a p} = \frac{k_{21} k_{22} [PhOH]}{k_{- 21} + k_{22} [PhOH]} .

The S-shaped catalytic wave is characterized by a plateau current i_pl that may be expressed as

i_{p l} = \frac{2 F S \sqrt{D_{c a t}} C_{c a t}^{0} \sqrt{k_{1} [{CO}_{2}]}}{1 + \frac{\sqrt{k_{1} [{CO}_{2}]}}{\sqrt{k_{2, a p}} (1 + \frac{\sqrt{k_{2, a p}}}{\sqrt{k_{1} [{CO}_{2}]}})} [1 + \frac{k_{2, a p}}{k_{2,2} C_{Z}^{0}}]},

[1]where S is electrode surface area;

C_{c a t}^{0}

and D_cat are concentration and diffusion coefficient of the catalyst, respectively.Open in a separate window Fig. 1.CV of 1-mM FCAT (Lower Left) and CAT (Lower Right) in neat DMF + 0.1 M n-Bu₄NPF₆ at 0.1 Vs⁻¹. The same, Upper Left and Upper Right, respectively, in the presence of 0.23 M CO₂ and of 1 M PhOH.

i_{p}^{0}

, the peak current of the reversible Fe^II/Fe^I wave is a measure of a one-electron transfer.

Table 1.

Kinetic characteristics of the reactions in Scheme 2 from ref.

Parameters for catalysis	FCAT	CAT
k₁(M^-1 ? s^-1)	3 × 10⁵	>5 × 10⁶
(k₂₁/k_?21)k₂₂(M^-1 ? s^-1)	2.5 × 10⁴	—
k₂₁(s^-1)	3 × 10⁴	7 × 10³
$k_{2}^{a p} (s^{- 1})$	2.1 × 10⁴	7 × 10³
α₂	0.3
[PhOH] (M)	$\log k_{f}^{2 n d E T}$ (cm s⁻¹)
0.3	−8.8	−9.4
0.5	−8.8	−9.4
1	−8.6	−9.3
2	−8.4	−9.0
3	−8.25	−8.8

Open in a separate windowThe

k_{2}^{a p}

value for FCAT is given for [PhOH] = 3 M. It is independent from the acid concentration in the case of CAT.The logTOF–η plots (Fig. 2) move upward as the phenol concentration increases. They are more favorable for FCAT than for CAT whatever [PhOH]. A more direct comparison between the two catalysts at [PhOH] = 3 M is shown in Fig. 3, where the results of preparative-scale electrolyses are also displayed within the same logTOF vs. η framework, pointing to the superiority of FCAT over CAT. This is confirmed by preparative-scale electrolyses. Fixed-potential electrolyses were performed at −1.08 and −1.14 V vs. NHE with 1 mM FCAT and CAT, respectively, using a carbon crucible as working electrode under 1 atm. CO₂ (0.23 M) in the presence of 3 M PhOH. The current density i_el/S_el is stable over 3 h with FCAT and 0.5 h with CAT and the production of CO is practically quantitative (faradaic yields of 100 ± 10% and 100 ± 5%, respectively, less than 1% H₂ in both cases). i_el/S_el= 0.5 and 0.3 mA/cm² with FCAT and CAT, respectively; S_el, the working electrode surface area of the preparative-scale electrode electrolysis, is much larger (20 cm²) than in CV experiments (0.07 cm²). The corresponding TOF value at the operated overpotential is calculated from TOF = (i_el/i_pl)TOF_max, in which i_pl is the plateau current given by Eq. 1. The TOF values thus obtained are 240 s⁻¹ (at η = 0.39 V) and 170 s⁻¹ (at η = 0.45 V) for FCAT and CAT, respectively. As seen in Fig. 3, they satisfactorily match the TOF–η relationships derived from CV taking into account inevitable imperfections in cell configuration leading to residual ohmic drop. Besides catalytic performances evaluated through logTOF–η relationship, durability is important in the evaluation of catalyst efficiency. It has been evaluated through estimation of the catalyst degradation over prolonged electrolysis. This estimation is based on recording CVs in the electrolysis solution during electrolysis. It turns out that (see SI Text for details) FCAT is more stable than CAT or simple FeTPP. Complete degradation of the initial 10⁻⁵ moles of catalyst is observed after the passage of 575, 200, and 290 coulombs for FCAT, CAT, and simple FeTPP, corresponding to 600, 210, and 300 catalytic cycles for FCAT, CAT and FeTPP, respectively.Open in a separate window Fig. 2.Catalytic Tafel plots for the two catalysts (see text) as a function of the concentration of phenol in the solution, in M, from bottom to top: 0.3, 0.5, 1, 2, 3.Open in a separate window Fig. 3.Benchmarking of all catalysts based on catalytic Tafel plots derived from CV experiments. See SI Text. Fig. 3 illustrates the ensuing benchmarking of all catalysts. In terms of preparative-scale electrolyses, the available information indicates that the stability of the catalysts is of the same order as for the two catalysts FCAT and CAT described here.

Table 2.

Comparison of FCAT and CAT with other catalysts of the CO₂/CO conversion

Reference	Solvent + acid $E_{{CO}_{2} /CO}^{0}$	Catalyst $E_{cat}^{0}$	$k_{1}^{a p} [{CO}_{2}] k_{2}^{a p}$	logTOF_max (s⁻¹)	logTOF₀ (s⁻¹)^*
27; this work	DMF +3 M PhOH	CAT	>5 × 10⁶	3.8	−6.0
27; this work	−0.69	−1.35	See 27; this work	3.8	−6.0	DMF +3 M PhOH	FCAT	>5 × 10⁶	4.0	−5.5
−0.69	−1.28	See 25	DMF +3 M PhOH	Fe⁰TPP	3.5 × 10⁴	4.5	−8.0		4.0	−5.5
−0.69	−1.43	^†				4.5	−8.0
21	DMF +0.1 M HBF₄	m-(triphos)₂Pd₂^‡	35	1.5	−7.4
21	−0.23	−0.76	^†	1.5	−7.4
20	CH₃CN +0.8 M CF₃CH₂OH	Re(bpy)(CO)₃(py)	875	2.9	−8.0
20	-0.65	−1.30	^†	2.9	−8.0
12	CH₃CN +1.4 M CF₃CH₂OH	Mn(bpytBu)(CO)₃Br^§	680	2.8	−9.8
12	-0.65	−1.40	^†	2.8	−9.8
22	CH₃CN	Ru^II(tpy⁻)(bpy⁻)^§	7.6	0.9	−10.8
22	−0.65	−1.34	^†	0.9	−10.8
22	CH₃CN	Ru^II(tpy⁻)(Mebim-py⁻)^§	59	1.8	−9.9
22	−0.65	−1.34	^†	1.8	−9.9
23	CH₃CN	N₂ = Mn(CO)₃^¶	5 × 10³	3.7	−7.0
23	−0.65	−1.28	^†	3.7	−7.0

Open in a separate windowPotentials in V vs. NHE, first-order or pseudo-first-order rate constants in s⁻¹.^*TOF at η = 0.^†

k_{2}^{a p}

k_{1}^{a p} [{CO}_{2}]

.^‡^§py = pyridine, tpy = 2,2'':6'',2''''-terpyridine, bpy =2,2''-bipyridine, Mebimpy = 2,6-bis(1-methyl benzimidazol-2-yl)pyridine.^¶Our conclusion is twofold. (i) The title iron porphyrin generated electrochemically under its Fe⁰ form (FCAT) operated in the presence of 3 M phenol in DMF appears to be the best homogeneous catalyst of the CO₂-to-CO conversion to date. This clearly appears after benchmarking of presently available catalyst of this reaction under the form of catalytic Tafel plots relating turnover frequency with overpotential (Fig. 3). Such plot allows optimization of the catalytic reaction by appropriately compromising between rapidity and energy costs. A further advantageous feature of FCAT is that it relies on one of the cheapest and most earth-abundant metals. (ii) Fluorine substitution in passing from CAT to FCAT was designed to favor catalysis in terms of overpotential thanks to the inductive effect of the fluorine substituents. At the same time it could have rendered the follow-up reactions less favorable, possibly annihilating the initial favorable effect of fluorine sub or even making catalysis globally less efficient than with CAT. The observation that this is not the case, and that the substitution has a global positive effect in this case, opens the route to the design and testing of further substituted molecules, which could become even more efficient catalysts of the CO₂/CO conversion. It should be particularly fruitful to use the prepositioned phenol functionalities to favor the formation and proton-coupled transformation of the initial Fe⁰–CO₂ adduct and to play with electron withdrawing substituents to improve the capabilities of the catalyst in terms of overpotential.See SI Text for experimental details and data treatment of the other catalysts in

Keywords:

solar fuels CO₂ reduction electrochemistry catalysis contemporary energy challenges

设为首页 | 免责声明 | 关于勤云 | 加入收藏