Effect of sp2 species in a boron-doped diamond electrode on the electrochemical reduction of CO2

Article abstract of DOI:10.1016/j.elecom.2020.106731

The effect of sp² carbon within a boron-doped diamond (BDD) electrode on the electrochemical reduction of CO₂ is investigated for the first time. With increasing sp² content, the Faradaic efficiency for the production of hydrogen (H₂) under constant current density increases while that for the production of formic acid (HCOOH) decreases. Moreover, favorable electrolysis conditions for producing HCOOH shift in the negative potential direction with increasing sp² content. This phenomenon is ascribed to the presence of adsorption sites on the sp² carbon. This work provides guidance for selectively controlling the reduction of CO₂ by adjusting the distribution of carbonaceous species in the BDD electrode.

Full text of DOI:10.1016/j.elecom.2020.106731

ElectrochemistryCommunications115(2020)106731
Contents lists available at ScienceDirect
Electrochemistry Communications
journal homepage: www.elsevier.com/locate/elecom
Effect of sp2 species in a boron-doped diamond electrode on the
electrochemical reduction of CO2
Jing Xu, Yasuaki Einaga⁎
Department of Chemistry, Keio University, 3-14-1 Hiyoshi, Yokohama 223-8522, Japan
A R T I C L E I N F O
A B S T R A C T
Keywords:
The effect of sp2 carbon within a boron-doped diamond (BDD) electrode on the electrochemical reduction of CO2
is investigated for the first time. With increasing sp2 content, the Faradaic efficiency for the production of
hydrogen (H2) under constant current density increases while that for the production of formic acid (HCOOH)
decreases. Moreover, favorable electrolysis conditions for producing HCOOH shift in the negative potential
direction with increasing sp2 content. This phenomenon is ascribed to the presence of adsorption sites on the sp2
carbon. This work provides guidance for selectively controlling the reduction of CO2 by adjusting the dis-
tribution of carbonaceous species in the BDD electrode.
Boron-doped diamond
sp2 carbon
CO2 reduction
1. Introduction
of electrolytes or of the working potential; there has been little in-
vestigation of the intrinsic characteristics of BDD electrodes.
Conversion of CO2 into value-added chemicals and fuels provides a
direct method for making use of the greenhouse gas CO2. The electro-
chemical reduction of CO2 has been comprehensively studied in recent
decades and large numbers of mechanisms have been proposed [1,2].
However, during electrochemical reduction, there is an unavoidable
competition between CO2 and proton (H+) generation in an aqueous
electrolyte. Since both the CO2 reduction reaction (CO2RR) and the
hydrogen evolution reaction (HER) consume electrons [Eqs. (1) and
(2)], it is not easy to achieve efficient CO2RR.
BDD is composed of carbon atoms with a sp3-bonded diamond
structure, in which some carbon atoms are replaced by boron atoms.
Following this replacement, the electrochemical properties change
dramatically from insulator to semiconductor [3,10,11]. Generally
speaking, the factors that affect the characteristics of the BDD are the
boron doping level [12], the presence or absence of sp2 carbon [13–15]
and the surface termination [16]. Amongst these, the presence of sp2
carbon has a direct effect on the band structure and the density of state
(DOS), and would be expected to have a strong influence on the elec-
trochemical properties [17].
CO2 + e− → CO2
2H+ + 2e− → H2
E0 = −1.9 V (vs. NHE)
E0 = 0 V (vs. NHE)
(1)
(2)
%
−
BDDs containing sp2 carbon are normally considered to be of low
quality, since they have a greater background current [13] and are
easier to corrode [18] than material with no sp2 carbon. Nevertheless,
sp2-containing BDD has performed very well in a number of applica-
tions, including the treatment of waste water [19], oxygen reduction
[20], and the generation of ozone [13]. The products of CO2RR on BDD
without sp2 carbon are mainly HCOOH [4]. However, CO2RR on sp2
carbon electrodes such as glassy carbon (GC), graphite, graphene and
HOPG has shown a tendency for CO to be the main product [21–24].
This difference in the products of CO2RR has been reported to be re-
lated to the absorbability of the electrode surface [25].
Boron-doped diamond (BDD) possesses a number of favorable
properties for an electrode, including a wide potential window, low
background current, chemical inertness and mechanical durability
compared to conventional metal electrodes [3]. With its wide potential
window, which can suppress HER during electrolysis, BDD is a pro-
mising candidate for CO2RR [4–6]. Effective CO2RR using a BDD
electrode has been demonstrated in previous studies in which for-
maldehyde was produced using sea-water [6], formic acid was pro-
duced using an alkaline solution [4], methanol was produced from an
amine solution [7], carbon monoxide was produced using a KClO4
electrolyte [8] and high carbon compounds were produced on metal-
modified BDDs [9]. However, these studies focused mainly on the effect
Herein, we have studied CO2RR on BDD electrodes with various
sp2/sp3 ratios, which were determined by Raman spectroscopy [14].
BDD containing 0.1% boron was selected as a candidate for further
investigation owing to its excellent performance for CO2RR compared
⁎ Corresponding author.
E-mail address: einaga@chem.keio.ac.jp (Y. Einaga).
https://doi.org/10.1016/j.elecom.2020.106731
Received 2 March 2020; Received in revised form 9 April 2020; Accepted 10 April 2020
Availableonline13April2020
1388-2481/©2020TheAuthors.PublishedbyElsevierB.V.ThisisanopenaccessarticleundertheCCBYlicense
(http://creativecommons.org/licenses/BY/4.0/).

J. Xu and Y. Einaga
ElectrochemistryCommunications115(2020)106731
Table 1
bubbled into the catholyte for 15 min at 50 sccm and the gaseous
products collected in an aluminum bag (CEK 3008–26401, GL Science).
The H2 and CO produced were analyzed by gas chromatography (GC-
2014, Shimadzu Corp.) while the HCOOH was analyzed by high-per-
formance liquid chromatography (HPLC, Prominence, Shimadzu
Corp.). The Faradaic efficiency was estimated from the yield of each
product, as described elsewhere [4].
B/C ratio and IG/IDiamond values of BDDs used in Fig. 4.
B/C ratio (%)
IG/IDiamond
no-sp2
mid-sp2
high-sp2
0.03
0.07
0.15
0.70
2.20
0
0
0
0
0
0.02
0.09
0.05
0.04
0.08
0.03
0.11
0.1
0.15
0.09
3. Results and discussion
3.1. Characterization of the BDDs
to BDD electrodes with other B/C ratios [5]. This study should help in
elucidating the electrocatalytic properties of the BDD electrode towards
the CO2RR.
Basic characterization results for the samples, such as the actual B/C
ratio, IG/IDiamond and the average grain size have been reported in a
previous study [14]. It is important to emphasise that the actual B/C
ratios of the three BDDs studied (0.08%, 0.06% and 0.07%) are roughly
equal to each other, and therefore any drastic change in electro-
chemical reduction could not be due to a difference in the boron con-
tent of the electrodes. The IG/IDiamond ratios (0, 0.09 and 0.11) were
obtained by Raman analysis. Each sample revealed a clear first-order
diamond peak at 1332 cm−1. A broad band around 1550 cm−1 was
found in the spectra of mid-sp2 and high-sp2 BDD, which confirms the
presence of sp2 sites (amorphous carbon and graphitic carbon). By
contrast, the no-sp2 BDD showed a negligible level of sp2 defects. Fur-
thermore, using SEM, the average grain size of the BDDs (13.0, 10.6 and
9.5 µm) was found to decrease with increasing sp2 content. Ad-
ditionally, there are more boundaries on sp2-containing BDD [26,27]
than in the sp2-free diamonds. This indicates that secondary nucleation
is extended by enhancing the sp2 character [17].
2. Experimental
2.1. Preparation of BDD working electrode
BDDs with different carbon sp2 levels were fabricated by MPCVD
(AX6500X, CORNES Technologies corp.). By modifying the gas flow
rates of the boron source B(CH3)3, the carbon source CH4 and the
carrier gas H2, BDDs with a boron to carbon ratio (B/C) of 0.1% and a
variety of sp2/sp3 ratios were deposited on silicon wafers as described
elsewhere [14]. Commercial GC electrodes (Tokai Corp.) were used for
comparison, and were smoothed by grinding before use.
The surface morphology was observed by SEM (JCM-6000, JEOL).
Raman spectra were recorded with an excitation wavelength of 532 nm
in ambient air at room temperature with an Acton SP2500 (Princeton
Instruments). The sp2/sp3 ratios (IG/IDiamond) estimated from the Raman
data are listed in Table 1. The samples were categorised as no sp2, mid
sp2 and high sp2 (details of this assignment are given in [14]). The
actual boron content of each BDD was estimated by Glow Discharge
Optical Emission Spectroscopy (GDOES) (GD-Profiler2, Horiba Ltd.)
with reference to a BDD sample whose boron content had already been
estimated by SIMS [13].
3.2. CO2RR on sp2-containing BDDs
3.2.1. LSV before and after CO2 bubbling
The electrocatalytic activity of the BDDs was examined in CO2-sa-
turated 0.5 M KCl solutions. The LSV curves before and after CO2
bubbling are shown in Fig. 1, in which the potential was swept between
−1 V and −3.5 V (vs. Ag/AgCl) at a rate of at 0.02 V s−1
.
Comparing the LSVs with increasing sp2 content after CO2 bubbling in
Fig. 1 (solid lines), two clear changes in the voltammograms are notice-
able. There is an obvious negative shift of the onset potential from −2.3 V
to −2.6 V and −3.4 V (vs. Ag/AgCl) with increasing sp2 content. On the
other hand, a sizeable decrease in the anodic current at −3.5 V was
2.2. Electrochemical reduction of CO2
The electrochemical measurements were conducted in a two-com-
partment flow cell separated with a Nafion membrane (NRE-212, 0.002
in thickness, Aldrich), as described in our previous research [4,5]. The
fabricated BDD, Pt plate and Ag/AgCl (in saturated KCl) were used as
the working electrode, counter electrode and reference electrode, re-
spectively. The geometric area of both the BDD and Pt electrode were
9.62 cm2. Prior to every electrochemical measurement, electrochemical
pretreatment using cyclic voltammetry (CV) in 0.1 M H2SO4 aqueous
solution was carried out to ensure that the surface termination was
consistent, as described in previous research [4]. Fundamental elec-
trochemical properties such as the potential window and the CV per-
formed in redox electrolytes were as shown in other reports [14].
The electrochemical reduction of CO2 was performed as follows.
The catholyte and anolyte were aqueous solutions of 0.5 M KCl and 1 M
KOH, respectively. Both were 50 mL solutions and were circulated se-
parately in the cell with a flow rate of 100 mL min−1 using separate
pumps. The catholyte was purged with N2 gas for 30 min at 200 sccm to
remove any dissolved oxygen and then bubbled with CO2 gas for 1 h at
500 sccm to saturate it with CO2. After N2 and CO2 saturation, linear
sweep voltammetry (LSV) measurements were carried out in the po-
tential range from −1 V to −3.5 V (vs. Ag/AgCl) at 0.02 V s−1. The
chronopotentiometry reductions were performed under a constant
current density of −2 mA cm−2 for 1 h. The chronoamperometry re-
ductions were at −2.2 V, −2.5 V, −2.7 V, −2.9 V and −3.4 V (vs. Ag/
AgCl) for 1 h. During the reduction, CO2 was continually bubbled into
the catholyte at a slow flow rate. After the electrolysis, N2 gas was
observed from approximately 45 mA cm−2 to
7
mA cm−2 and
1 mA cm−2 in samples with increasing sp2 content. The measured current
Fig. 1. Electrochemical performance of different electrodes by mean of an LSV
scan with a scan rate of 0.02 V s−1 in 0.5 M KCl aqueous electrolyte before and
after CO2 saturation.
2

J. Xu and Y. Einaga
ElectrochemistryCommunications115(2020)106731
Fig. 2. Faradaic efficiency of CO2RR on different electrodes and the corre-
sponding potential, under a constant current density of −2 mA cm−2 for a 1 h
reduction.
was composed of two parts: partial current from HER and partial current
from CO2RR. One possible reason for the negative shift in the potential
and decreased current is the diminished carrier density caused by the
increasing sp2/sp3 ratio [14], which also inhibits the CO2RR activity.
3.2.2. Reduction with constant current density
Fig. 2 shows the products of CO2RR and the corresponding reduc-
tion potentials on various BDD electrodes and also on a GC electrode.
The results were obtained at constant current density of −2 mA cm−2
for 1 h. It can be seen that the tendency towards HCOOH production
decreases with increasing sp2 content, while the tendency towards H2
production increased. The results show the reduction potential shifting
to more negative values with increasing sp2 content. The reduction
potential decreased from –2.5 V to −3.1 V and −3.6 V (vs. Ag/AgCl)
for no-sp2, mid-sp2 and high-sp2 electrodes, respectively, which to some
extent agrees with the LSV results in Fig. 1.
As indicated by previous studies, HCOOH is the main product of
CO2RR by BDD electrodes with little sp2 content [4,5]. This is also
shown by the data for no-sp2 BDD in Fig. 2, for which the Faradaic
efficiency was as high as 84% for HCOOH. The performance of mid-sp2
BDDs, at 36% efficiency for producing HCOOH, was not as high as for
the no-sp2 BDD electrode. Rather like GC electrodes, which contain sp2
over the entire surface and have 18% efficiency for HCOOH, high-sp2
BDD had a 16% HCOOH efficiency. However, H2 evolution increased
with increasing sp2 content, with efficiency values of 11% for no-sp2
BDD, 51% for mid-sp2 BDD, 68% for high-sp2 BDD and 71% for GC,
respectively. These results suggest that the amount of sp2 carbon pre-
sent has a significant influence on the selectivity of the CO2RR [28].
Fig. 3. Faradaic efficiency of generating (a) HCOOH, (b) H2 and (c) CO on
different electrodes under a constant potential from −4 V to −2.2 V (vs. Ag/
AgCl) respectively.
The Faradaic efficiency for production of CO on all electrodes was
lower than 10%. However, the high-sp2 BDD produced more CO than
the other BDDs in the potential range from −3.4 V to −2.5 V (vs. Ag/
AgCl). The highest Faradaic efficiency of high-sp2 BDD is 7% at a po-
tential of −2.9 V (vs. Ag/AgCl). The results indicate that a CO2RR
producing CO is easier to obtain on BDDs containing sp2 species than on
BDDs with no sp2 carbon.
3.2.3. Potential dependence
Fig. 3 shows the dependence of the Faradaic efficiency on the elec-
trode potential from −4 V to −2.2 V (vs. Ag/AgCl) for each electrode.
With increasing sp2 content, the optimized potential for maximum pro-
duction of HCOOH moves to a more negative potential. Furthermore, the
maximum Faradaic efficiency value for producing HCOOH decreases. As
can be seen in Fig. 3(a), the highest Faradaic efficiencies for HCOOH for
no-sp2, mid-sp2 and high-sp2 electrodes were obtained at constant po-
tentials of −2.5 V, −2.9 V and −3.4 V (vs. Ag/AgCl), respectively.
These values are in accordance with the onset potentials measured by the
LSV curves in Fig. 1. The Faradaic efficiencies were correspondingly
reduced to 78%, 62% and 12%, respectively.
3.3. Mechanism proposal
The presence of sp2 carbon clearly affects the rate of reduction of
CO2 under constant current, which is a competition between CO2RR
and HER. For BDD electrodes in the absence of sp2 carbon, HCOOH is
the major product, whilst H2 is a minor product. Our results indicate
that CO2RR performs much better than HER on a sp3 surface. This
phenomenon is ascribed to the wide potential window of BDD, allowing
the HER to be minimized and CO2RR to become the main reaction [4].
The yield of H2 is changed a great deal on varying the applied potential,
The minimum Faradaic efficiency for H2 also revealed a negative
potential shift. The no-sp2 BDD sample produced the least H2 at −2.5 V
(vs. Ag/AgCl), while the corresponding high sp2 and mid sp2 values
were −2.7 V and −2.9 V (vs. Ag/AgCl), respectively.
3

J. Xu and Y. Einaga
ElectrochemistryCommunications115(2020)106731
CRediT authorship contribution statement
Jing Xu: Conceptualization, Methodology, Investigation, Writing -
original draft, Visualization. Yasuaki Einaga: Conceptualization,
Validation, Writing - review & editing, Supervision.
Acknowledgement
This work is supported by JST (Japan Science and Technology
Agency – ACCEL (Strategic Basic Research Programs): Fundamentals
and Applications of Diamond Electrodes.
References
[1] N. Yang, S. Yu, J.V. Macpherson, Y. Einaga, H. Zhao, G. Zhao, G.M. Swain, X. Jiang,
Chem. Soc. Rev. 48 (2019) 157–204.
[2] D.D. Zhu, J.L. Liu, S.Z. Qiao, Adv. Mater. 28 (2016) 3423–3452.
[3] Y. Einaga, J. Appl. Electrochem. 40 (2010) 1807–1816.
[4] K. Natsui, H. Iwakawa, N. Ikemiya, K. Nakata, Y. Einaga, Angew. Chem. Int. Ed. 57
(2018) 2639–2643.
Fig. 4. Faradaic efficiency of HCOOH on electrodes with different boron
doping/sp2 content and the corresponding potentials under a constant current
density of −2 mA cm−2 for a 1 h reduction.
[5] J. Xu, K. Natsui, S. Naoi, K. Nakata, Y. Einaga, Diam. Relat. Mater. 86 (2018)
167–172.
[6] K. Nakata, T. Ozaki, C. Terashima, A. Fujishima, Y. Einaga, Angew. Chem. Int. Ed.
53 (2014) 871–874.
which indicates that the HER on this electrode is sensitive to potential.
By contrast, the BDD electrodes containing sp2 species produced more
H2 than HCOOH. This distribution of products is probably affected by
the adsorption sites present on sp2 carbon [29].
[7] P.K. Jiwanti, K. Natsui, K. Nakata, Y. Einaga, RSC Adv. 6 (2016) 102214–102217.
[8] M. Tomisaki, S. Kasahara, K. Natsui, N. Ikemiya, Y. Einaga, J. Am. Chem. Soc. 141
(2019) 7414–7420.
[9] P.K. Jiwanti, K. Natsui, K. Nakata, Y. Einaga, Electrochim. Acta 266 (2018)
414–419.
The product of CO2RR is a mixture of HCOOH and CO. On BDD
electrodes without sp2 species, HCOOH is the main outcome and CO is a
minor product. On the other hand, BDD electrodes containing sp2
produce a smaller amount of HCOOH and a relatively higher amount of
CO. It is suggested that the sp2 surface has a higher adsorption of
[10] A. Fujishima, Y. Einaga, T.N. Rao, D.A. Tryk (Eds.), Diamond Electrochemistry,
Elsevier B.V, Tokyo, 2005.
[11] J.P. Lagrange, A. Deneuville, E. Gheeraert, Diam. Relat. Mater. 7 (1998)
1390–1393.
[12] K. Schwarzová-Pecková, J. Vosáhlová, J. Barek, I. Šloufová, E. Pavlova, V. Petrák,
J. Zavázalová, Electrochim. Acta 243 (2017) 170–182.
[13] T. Watanabe, Y. Honda, K. Kanda, Y. Einaga, Phys. Status Solidi 211 (2014)
2709–2717.
%
−
CO2
and that affects the products of the CO2RR [25]. It has pre-
viously been suggested that CO is generated on adsorbing electrode
surfaces whilst HCOOH is produced on surfaces that do not show ad-
sorption [8,30].
[14] J. Xu, Y. Yokota, R.A. Wong, Y. Kim, Y. Einaga, J. Am. Chem. Soc. 142 (2020)
2310–2316.
[15] J. Li, C.L. Bentley, S.Y. Tan, V.S.S. Mosali, M.A. Rahman, S.J. Cobb, S.X. Guo,
J.V. Macpherson, P.R. Unwin, A.M. Bond, J. Zhang, J. Phys. Chem. C 123 (2019)
17397–17406.
The sp2 effect on the CO2RR is found not only on the 0.1% BDDs, but
also on other samples with different boron content. Fig. 4 shows the
Faradaic efficiency for the production of HCOOH on BDD electrodes
with various sp2 content and with different B/C ratios (as listed in
Table 1). For each boron doping level, HCOOH production efficiency
declined with increasing sp2 content.
[16] S. Kasahara, K. Natsui, T. Watanabe, Y. Yokota, Y. Kim, S. Iizuka, Y. Tateyama,
Y. Einaga, Anal. Chem. 89 (2017) 11341–11347.
[17] J.A. Bennett, J. Wang, Y. Show, G.M. Swain, J. Electrochem. Soc. 151 (2004)
E306–E313.
[18] T. Kashiwada, T. Watanabe, Y. Ootani, Y. Tateyama, Y. Einaga, ACS Appl. Mater.
Interfaces 8 (2016) 28299–28305.
[19] C.D.N. Brito, D.M. de Araújo, C.A. Martínez-Huitle, M.A. Rodrigo, Electrochem.
Commun. 55 (2015) 34–38.
4. Conclusion
[20] T.L. Read, S.J. Cobb, J.V. Macpherson, ACS Sensors 4 (2019) 756–763.
[21] K. Hara, A. Kudo, T. Sakata, J. Electroanal. Chem. 421 (1997) 1–4.
[22] N. Sreekanth, M.A. Nazrulla, T.V. Vineesh, K. Sailaja, K.L. Phani, Chem. Commun.
51 (2015) 16061–16064.
We have studied the influence of sp2 carbon in BDD electrodes on
the CO2RR. BDDs with different sp2/sp3 ratios were produced by
MPCVD and used as electrodes for CO2RR. The products were found to
demonstrate certain trends with increasing sp2 content. We observe
that, there is a sharp decline in the production of HCOOH with in-
creasing sp2 content. On the other hand, the yield of H2 was sig-
nificantly raised. The results for BDD electrodes with high sp2 content
were similar to those of a well-known sp2 electrode (GC). Moreover, our
results revealed a negative shift in the onset potential for the reduction
and a lower maximum yield of HCOOH with increasing sp2 content. It is
[23] N. Yang, S.R. Waldvogel, X. Jiang, A.C.S. Appl, Mater. Interfaces 8 (2016)
28357–28371.
[24] P. Han, X. Yu, D. Yuan, M. Kuang, Y. Wang, A.M. Al-Enizi, G. Zheng, J. Colloid
Interface Sci. 534 (2019) 332–337.
[25] A. Goeppert, M. Czaun, J.-P. Jones, G.K. Surya Prakash, G.A. Olah, Chem. Soc. Rev.
43 (2014) 7995–8048.
[26] J.V. Macpherson, Phys. Chem. Chem. Phys. 17 (2015) 2935–2949.
[27] T. Watanabe, S. Yoshioka, T. Yamamoto, H. Sepehri-Amin, T. Ohkubo,
S. Matsumura, Y. Einaga, Carbon 137 (2018) 333–342.
[28] D. Medeiros de Araújo, P. Cañizares, C.A. Martínez-Huitle, M.A. Rodrigo,
Electrochem. Commun. 47 (2014) 37–40.
suggested that the surface of sp2-containing BDDs could absorb CO2
%
−
[29] S. Garcia-Segura, E. Vieira dos Santos, C.A. Martínez-Huitle, Electrochem. Commun.
59 (2015) 52–55.
more easily than those without sp2 species, although further in-
vestigation is required. It is noteworthy that the selectivity of CO2RR
could be controlled by adjusting the distribution of carbonaceous spe-
cies within the BDD electrode.
[30] Y. Hori, H. Wakebe, T. Tsukamoto, O. Koga, Electrochim. Acta 39 (1994)
1833–1839.
4