Porous boron nitride for combined CO2 capture and photoreduction

Article abstract of DOI:10.1039/c9ta02793a

Porous and amorphous materials are typically not employed for photocatalytic purposes, like CO₂ photoreduction, as their high number of defects can lead to low charge mobility and favour bulk electron-hole recombination. Yet, with a disordered nature can come porosity, which in turn promotes catalyst/reactant interactions and fast charge transfer to reactants. Here, we demonstrate that moving from h-BN, a well-known crystalline insulator, to amorphous BN, we create a semiconductor, which is able to photoreduce CO₂ in the gas/solid phase, under both UV-vis and pure visible light and ambient conditions, without the need for cocatalysts. The material selectively produces CO and maintains its photocatalytic stability over several catalytic cycles. The performance of this un-optimized material is on par with that of TiO₂, the benchmark in the field. For the first time, we map out experimentally the band edges of porous BN on the absolute energy scale vs. vacuum to provide fundamental insight into the reaction mechanism. Owing to the chemical and structural tunability of porous BN, these findings highlight the potential of porous BN-based structures for photocatalysis particularly solar fuel production.

Full text of DOI:10.1039/c9ta02793a

View Article Online
View Journal
Journal of
Materials Chemistry A
Materials for energy and sustainability
Accepted Manuscript
This article can be cited before page numbers have been issued, to do this please use: R. Shankar, M.
Sachs, L. Francás, D. Lubert-Perquel, G. Kerherve, A. Regoutz and C. Petit, J. Mater. Chem. A, 2019, DOI:
10.1039/C9TA02793A.
Volume
Number 12
28 March 2018
Pages 4883-5230
6
This is an Accepted Manuscript, which has been through the
Royal Society of Chemistry peer review process and has been
accepted for publication.
Journal of
Materials Chemistry A
Materials for energy and sustainability
rsc.li/materials-a
Accepted Manuscripts are published online shortly after acceptance,
before technical editing, formatting and proof reading. Using this free
service, authors can make their results available to the community, in
citable form, before we publish the edited article. We will replace this
Accepted Manuscript with the edited and formatted Advance Article as
soon as it is available.
You can find more information about Accepted Manuscripts in the
Information for Authors.
Please note that technical editing may introduce minor changes to the
text and/or graphics, which may alter content. The journal’s standard
Terms & Conditions and the Ethical guidelines still apply. In no event
shall the Royal Society of Chemistry be held responsible for any errors
or omissions in this Accepted Manuscript or any consequences arising
from the use of any information it contains.
ISSN 2050-7488
COMMUNICATION
Zhenhai Wen et al.
An electrochemically neutralized energy-assisted low-cost
acid-alkaline electrolyzer for energy-saving electrolysis
hydrogen generation
rsc.li/materials-a

Page 1 of 26
Journal of Materials Chemistry A
View Article Online
Porous boron nitride for combined CO capture and photoreduction
DOI: 10.1039/C9TA02793A
2
Ravi Shankara, Michael Sachsb, Laia Francàsb, Daphné Lubert-Perquelc, Gwilherm
Kerherved, Anna Regoutzd, and Camille Petita*
aBarrer Centre, Department of Chemical Engineering, Imperial College London, South
Kensington Campus, Exhibition Road, London SW7 2AZ, United Kingdom
bDepartment of Chemistry, Imperial College London, South Kensington Campus, Exhibition
Road, London SW7 2AZ, United Kingdom
cLondon Centre for Nanotechnology and Department of Materials, Imperial College London,
South Kensington Campus, Prince’s Consort Road, London SW7 2BP, United Kingdom
dDepartment of Materials, Imperial College London, South Kensington Campus, Prince’s
Consort Road, London SW7 2BP, United Kingdom
*Corresponding author:
E-mail: camille.petit@imperial.ac.uk; Phone: +44 (0)20 7594 3182 (C. Petit)
1

Journal of Materials Chemistry A
Page 2 of 26
View Article Online
DOI: 10.1039/C9TA02793A
Abstract
Porous and amorphous materials are typically not employed for photocatalytic purposes, like
CO2 photoreduction as their high number of defects can lead to low charge mobility and
favour bulk electron-hole recombination. Yet, with a disordered nature can come porosity,
which in turns promotes catalyst/reactant interactions and fast charge transfer to reactants.
Here, we demonstrate that moving from h-BN, a well-known crystalline insulator, to
amorphous BN, we create a semiconductor, which is able to photoreduce CO2 in a gas/solid
phase, under both UV-vis and pure visible light, ambient conditions, without the need for
cocatalysts. The material selectively produces CO and maintains its photocatalytic stability
over several catalytic cycles. The performance of this un-optimized material is on par with
that of TiO2, the benchmark in the field. For the first time, we map out experimentally the
band edges of porous BN on the absolute energy scale vs. vacuum to provide fundamental
insight into the reaction mechanism. Owing to the chemical and structural tunability of
porous BN, these findings highlight the potential of porous BN-based structures for
photocatalysis and particularly solar fuels production.
Keywords: boron nitride, photocatalysis, CO2 photoreduction, solar fuel, porous materials.
2

Page 3 of 26
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C9TA02793A
Introduction
With a push towards sustainable energy technologies, the share of renewables is projected
to significantly increase in the years and decades to come.1 Solar fuel production via
photocatalysis represents one such energy generation route, allowing to directly take
advantage of sunlight. Solar fuels include H2 derived from water splitting and carbon-based
molecules (e.g. CH4, CO, CH3OH, etc…) derived from CO2 reduction. Literature on CO2
photoreduction catalysts is comparatively less extensive than on H2 and O2 generation,
owing to the significant thermodynamic and kinetic limitations of CO2 reduction. In fact, about
40 years after the work of Inoue et al. (1979), TiO2 remains the reference material in the
field.2 Researchers have looked at expanding the range of viable photocatalysts with works
reported on derivatives of TiO2, metal oxides, oxynitrides, carbon nitride, metal-organic
frameworks and composites thereof.3-16 Overall, these studies concentrate on optimizing
light absorption, preventing charge recombination, enhancing material robustness and
controlling surface-catalysed reaction mechanisms and selectivity. For the most part, they
focus on metal-based crystalline materials with low porosity. The latter feature prevents the
development of gas phase CO2 photoreduction, which would allow integrated CO2 capture
and conversion set-ups. In addition, the usage of rare metals either as part of the
photocatalyst or as co-catalysts limits large-scale deployment.
Here, taking into account the advent of graphitic carbon nitride and other carbonaceous
materials, we turn our attention to porous boron nitride (BN).17-26 Porous BN exhibits key
features of an ideal photocatalyst, yet it has never been studied for CO2 photoreduction. The
compelling features of BN include its metal-free nature as well as its tuneable chemistry and
structure that a priori allow for a tailored band structure.27-29 Porous BN is the amorphous
analogue of hexagonal boron nitride (h-BN), a wide band gap insulator (~5.5 eV).30 Doping
of this robust material and its utilisation in composites have been employed to reduce the
band gap and allow the formation of heterojunctions, respectively.31-34 Today, the only
reports of BN-based materials for CO2 photoreduction are on carbon-doped BN35-38, which
3

Journal of Materials Chemistry A
Page 4 of 26
View Article Online
have since been shown to be unstable in the presence of moisture owing to the bDreOaI: k10i.n10g39o/Cf9TA02793A
the B-C bond.39, 40
Herein, we report a porous BN material as both CO2 adsorbent and CO2 photoreduction
catalyst under both UV and visible light. We explain how the structure and chemistry of the
materials control its adsorptive, optoelectronic, photophysical, paramagnetic and
photocatalytic properties. Further, we delve deeper into the photochemistry of porous BN by
determining, and aligning the position of the conduction (CB) and valence (VB) band edge
as well as the Fermi level on the absolute energy scale vs. vacuum to elucidate the CO2
photoreduction reaction mechanism. Prior to this study, the band edges of porous BN were
yet to be obtained experimentally. This work paves the way for the creation of a new CO2
photocatalyst platform.
Results and discussion
We synthesised porous BN based on the previous work of Marchesini et al. (2017), whose
approach relies on a bottom-up pyrolytic synthesis at elevated temperatures using a mixture
of boric acid, melamine and urea.41 To gain insight into the morphology and structure of the
material, we collected high resolution transmission electron microscopy (HRTEM) and
scanning electron microscopy (SEM) images, as well as X-ray diffraction (XRD) patterns
(Figures 1a-c). Both techniques confirm the low degree of crystallinity of porous BN as
opposed to h-BN. As seen on the HRTEM images, the range of stacked 2D layers is limited
in porous BN (Figure 1b) and the SEM images depict a densely packed pore structure
(Supplementary Figure 1). In fact, the XRD patterns confirm the turbostratic nature of the
material with only broad peaks at 2θ values of 26° and 44°, corresponding to the (002) and
(100) planes, respectively. This contrasts with the sharp peaks for h-BN that indicate long-
range 3D/graphitic order. The d-spacing calculated from the TEM images and XRD peak (002)
for h-BN and porous BN is 3.4 Å and 3.5 Å, respectively.42, 43 44 The d-spacing in porous BN
is only relevant to the pseudo-crystalline regions of the material. To characterize the structural
4

Page 5 of 26
Journal of Materials Chemistry A
features of the material at the micro/nano-scale, we analysed the specific surface areaViaewnAdrticle Online
DOI: 10.1039/C9TA02793A
porosity through nitrogen sorption at -196 °C (Figure 1d for isotherms, Supplementary Table
1 for textural parameters). The porous BN sample displays a BET equivalent specific surface
area of 1585 m2 g-1 and total pore volume of 1.10 cm3 g-1, which is considerably larger than
those for h-BN (3 m2 g-1 and 0.006 cm3 g-1). The type I/IV isotherm with a type H3/H4 hysteresis
loop observed for porous BN indicates the presence of slit-shaped micro- and mesopores.
The high surface area of porous BN is key for interfacial phenomena, enhancing access to
active CO2 absorption and reduction sites. This is particularly important in gas phase schemes
like the one proposed here, where CO2 reduction is limited by the ability of the molecules to
be adsorbed on the catalyst, close to the catalytic sites.
5

Journal of Materials Chemistry A
Page 6 of 26
View Article Online
Figure 1 I Structure and morphology of porous BN sample compared to h-BDNO.I:(1a0).1039/C9TA02793A
HRTEM image of the as-synthesized porous BN sample highlighting the porous morphology
through the absence of stacked layers and the presence of wide crystalline domains, (b)
HRTEM image of h-BN, taken for comparison, illustrating the high degree of crystallinity with
a large number of stacked, well-ordered 2D layers, (c) Powder XRD patterns of porous BN
and h-BN, (d) N2 sorption isotherms for porous BN and h-BN measured at -196 °C.
We characterised the chemical structure and bonding types of the porous BN through
Raman spectroscopy, Fourier transform infrared spectroscopy (FT-IR), EDX, and X-ray
photoelectron spectroscopy (XPS) (Figure 2 and Supplementary Figures 2 and 3). Again, we
used h-BN here as a reference material. Both porous BN and h-BN exhibit a strong Raman
peak at ~1370 cm-1, corresponding to the counter-phase B-N E2g vibration mode within the
BN structure (Supplementary Figure 2).45, 46 We observed the two characteristic IR bands of
BN at ~1380 cm-1 (in-plane B-N transverse stretching) and ~800 cm-1 (out-of-plane B-N-B
bending) in both samples,47 while no B-O bands, typically observed at ~1000 cm-1, are noted
(Figure 2a).28, 48 Yet, elemental mapping conducted using SEM EDX points to the presence
of oxygen homogeneously distributed through the structure (Figure 2b). To gain further
insight into the relative atomic surface composition and chemical states of the elements, we
collected high resolution core level spectra through XPS. The analysis indicates a significant
proportion (10 at. % ± 1 at. %) of oxygen in porous BN (Figure 2c) with a peak at 533.1 eV
(Supplementary Figure 3). We attribute this peak to boron oxynitride (B-Ox-Ny) species,
which stems from the in-plane substitution of oxygen atoms into the BN lattice, as we have
described in a prior study.29, 49 Oxygen atoms were observed on h-BN and are most likely
related to edge hydroxyl groups (Figure 2c). The formation of BN is confirmed through the
fitted core level spectra of the B 1s and N 1s and show the presence of B-N bonds (191.0 eV
for B 1s and 398.5 eV for N 1s, Supplementary Figure 3).50, 51 We also observe shake-up
satellite peaks the B 1s and N 1s core level spectra for porous BN (Supplementary Figure 3),
which provides evidence for the formation of a sp2-hybridised hexagonal BN phase.52-54 The
6

Page 7 of 26
Journal of Materials Chemistry A
carbon content in the sample was comparatively much lower (~ 3 at. % ± 1 at. %)DaOnI: d10.i1s03V9ie/wCA9rTtiAcl0e2O7n9li3nAe
linked to the presence of adventitious carbon impurities, i.e. not part of the material. Oxygen
atoms can enhance CO2 adsorption55, 56 and can also lead to reduced band gap, as
indicated by both computational and experimental work.28, 57, 58 Both aspects are important in
the context of CO2 capture and photoconversion.
Next, we probed the optoelectronic properties of porous BN through UV-vis diffuse
reflectance spectroscopy. The absorption spectrum (Figure 2d) of the material clearly
indicates a red-shift when moving from h-BN to porous BN, with the absorption edge moving
from 215 nm to 290 nm. From the Tauc plot of the transformed Kubelka-Munk function
against photon energy (Supplementary Figure 5a), the sample exhibits a UV-range band gap
of 4.21 eV ± 0.05 eV, significantly lower than that of h-BN (~5.51 eV ± 0.02 eV)
(Supplementary Figure 5b). These results point to the creation of a semiconducting porous
BN material. At this stage, we cannot infer whether the semiconducting property is related to
the presence of functional groups, their specific location on the BN nanosheets or the
introduction of porosity.
We next collected time-resolved photoluminescence spectra to characterise the excited state
properties of our porous BN. We compare the results to the behaviour of another metal-free
photocatalyst, whose layered structure is similar to that of h-BN: graphitic carbon nitride (g-
C3N4). For both materials, photoluminescence decay components are extracted by global
analysis of the time-resolved spectra, where each component is characterised by a certain
photoluminescence spectrum and decay (details described in Supplementary Figure 6). For
g-C3N4, a single component with a spectrum shown in Figure 2e is found, following power
law decay kinetics ∝ t-1.6. This observation is in good agreement with a study by Godin et al.
(2017) where power law decay kinetics ∝ t-1.5 were observed over a wide range of time
scales in g-C3N4.59 Such power law kinetics are indicative of recombination via trapping-
detrapping steps, and this component can thus be assigned to shallowly trapped charges
(charges trapped at energies close to the band edge).59, 60 In contrast, two components are
7

Journal of Materials Chemistry A
Page 8 of 26
View Article Online
found for porous BN (Figure 2f). An exponentially decaying component (time constant τ = 1.5
DOI: 10.1039/C9TA02793A
ns) dominates at early times, which can be assigned to excitons (bound electron-hole pairs)
based on the exponential decay and the relatively narrow photoluminescence spectrum
close to the band gap transition. At later times, a second component following the same
power law decay kinetics ∝ t-1.6 as in g-C3N4 dominates, analogously assigned to shallowly
trapped charges. These results demonstrate that BN exhibits longer exciton lifetimes than g-
C3N4 and that shallowly trapped charges follow the same decay kinetics in porous BN and g-
C3N4. We note that charge yields, which could not be compared for the used samples, might
provide further insights into photocatalytic activities.
The presence of oxygen atoms in the interior structure of porous BN, arising from the
substitution of nitrogen atoms, results in unpaired electrons upon in-plane bonding with
boron atoms. This in turn yields associated paramagnetic properties of the porous BN
catalyst, which we probed through room temperature electron paramagnetic resonance
(EPR) measurements and compared to h-BN as a reference. We detected a strong EPR
signal at a magnetic field of approximately 342 mT with a calculated g-value of 2.0032
(Figure 2g). This value is in good agreement with literature values.28, 35 In comparison, no
resonance peak was observed for h-BN. This paramagnetic signature of the material
indicates the presence of unpaired free electrons in porous BN, which stems from the
isolated, conjugated –OB3 bonding pattern.28 Functional groups, such as edge hydroxyl
groups, or other bonding patterns such as –O2B are not capable of generating unpaired
electrons, and therefore would not contribute towards paramagnetism. Hence, this confirms
the presence of interior-substituted oxygen atoms in porous BN. The resonance signal at a
g-value of 2.0032 for porous BN in Figure 2g exhibits a symmetric signal, which
demonstrates an isotropic g of a uniform electron species. An approximate 30% increase in
the EPR peak intensity for porous BN was observed when the sample was irradiated using
UV-vis light, as shown in Figure 2h. Previous work has attributed the contribution to the
generation of electron-hole pairs.35 This would correspond to the short-lived photoinduced
8

Page 9 of 26
Journal of Materials Chemistry A
View Article Online
excitons, which were previously assigned to the exponential decay component ofDtOhIe: 10.1039/C9TA02793A
photoluminescence spectra in Figure 2f. However, the signal enhancement could also be
due to the release of the trapped charges under illumination, therefore time-resolved
measurements would be required to discriminate between these two interpretations. These
findings nonetheless indicate porous BN is able to function as a semiconductor upon
excitation.
9

Journal of Materials Chemistry A
Page 10 of 26
View Article Online
DOI: 10.1039/C9TA02793A
10

Page 11 of 26
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C9TA02793A
Figure 2 I Chemical, electronic, photophysical and paramagnetic characterisation of
porous BN as compared to h-BN and g-C3N4. (a) FT-IR spectra of as-synthesized porous
BN, with key characteristic bands highlighted, and h-BN as a reference, (b) SEM image,
EDX spectrum and elemental mapping for porous BN showing the homogeneous distribution
of the elements throughout the structure, (c) Relative atomic composition of porous BN
obtained through XPS analysis, (d) Absorption spectrum of the porous BN sample compared
to that of h-BN, (e, f) Component spectra extracted by global fitting of time-resolved
photoluminescence spectra for graphitic carbon nitride (g-C3N4) and porous BN, obtained
upon excitation at 282 nm. Models used for the global fits are described in the Supporting
Information, (g) Room temperature EPR spectra for porous BN compared to that of h-BN
with the g-value indicated, (h) Comparison of the EPR spectra for porous BN under dark and
UV-vis light irradiation conditions (300 W Xe arc lamp, 25 °C).
Prior to photocatalytic testing, we measured the CO2 adsorption capacity of the material at
the reaction conditions (25 oC, 0-1 bara) and compared to ‘reference’ materials, i.e. TiO2
P25, h-BN and carbon nitride. We have selected these reference materials for the following
reasons. TiO2 P25 is a benchmark commercial material in the field, which enables
comparison with other groups’ work, despite differences in experimental set-ups. Being a
commercial material, P25 acts as an internal reference and its use for comparison is in part
of the guidance in the field.61 Graphitic carbon nitride is the closest material in terms of
chemistry and structure to porous BN and a well-known visible light photocatalyst. Some of
the photophysics properties and behaviour of porous BN are likely to bear resemblances
with those observed in carbon nitride. h-BN is the most well-known BN form due to its long
standing use in industry as a lubricant. It’s the crystalline and insulating nature may have
acted as barriers for researchers to explore the semiconducting behaviour of other forms of
BN. We have also included results from a metal organic framework (MOF, see
Supplementary Figure 8 as a representative of a porous polymeric material. The chemistry
and photophysics are very however very different between BN and MOF.
11

Journal of Materials Chemistry A
Page 12 of 26
View Article Online
DOI: 10.1039/C9TA02793A
Unsurprisingly, porous BN adsorbs significantly more CO2 than the reference materials
(Figure 3a), most likely due to its higher porosity (Supplementary Tables 1 and 2). Other
groups have already reported the use of porous BN as an adsorbent for CO2, with similar
capacities reported.41, 55, 62, 63 Next, we tested porous BN and the reference materials for the
gas phase photoreduction of CO2 under UV-vis and pure visible irradiation, without doping or
the presence of a co-catalyst. For all materials, after an irradiation time of 5 hours, the only
product observed is CO. The CO2 adsorption capacities at 25 °C and specific CO evolution
rates of all of the materials investigated in this study are tabulated in Table 1. Under these
conditions, porous BN drove the photoreduction of CO2 with an average specific CO
production rate of 1.16 mol g-1 h-1 (Figure 3b). Under UV-vis irradiation, porous BN exhibits
μ
a higher specific CO production rate than those for the reference materials. We attribute the
significant increase in CO production going from h-BN to porous BN to the increase in
surface area and porosity as well as the greater light harvesting efficiency. We note that the
surface area of g-C3N4 is much lower than that of porous BN. Should its surface area be
increased, a higher performance might be observed. Porous BN is surpassed by g-C3N4 and
P25 using visible irradiation, which is to be expected given the lower band gaps of these
material (2.72 eV and 3.20 eV, respectively). Such low band gap may be attainable in BN
after significant O doping as suggested in the literature28, 57, 58. Interestingly both TiO2 and
porous BN exhibit visible light activity despite their relatively band gaps. The visible light
activity of P25 has been observed by other research groups worldwide and is attributed to
the presence of shallow trap states and mixed-phase (rutile-anatase) nanoclusters that
extends the photoactivity of P25 TiO2 to the visible region.64-66 For porous BN, the origin of
this visible light activity is still unclear. At present, we speculate that the presence of trap
states, namely the shallow trapped state shown in Figure 2f, is responsible for the visible
light activity. To investigate this further, we tested porous BN for CO2 photoreduction under
the same setup using a 550 nm long pass visible filter. No CO or other gaseous products
휆
were detected for > 550 nm (Figure 3b and Supplementary Table 2). As porous BN
휆
evolves CO in the visible region for 400 < < 550 nm, over which the shallow trapped
12

Page 13 of 26
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C9TA02793A
charge is present, we postulate that this contributes to the visible light activity of the material.
Further experimental studies are planned in a follow up to this study to identify the nature of
these trap states. These trap states could originate from N vacancies or O substituting N
atoms in porous BN. We note that porous BN can also photoreduce CO2 to CO with a 100%
selectivity using H2O rather than H2 as the sacrificial agent (Figure 3b).
We confirmed the origin of CO through a number of control tests: absence of catalyst, dark
mode, N2/H2 atmosphere (Supplementary Table 2). All supported the fact that the source of
the CO evolved stems from the photocatalytic reduction of CO2. To further substantiate this
claim, we carried out isotopic experiments using 13CO2. Aside from 13CO2 being used as the
feed gas in tandem with H2 as the sacrificial agent, we maintained the same catalyst
preparation method and reaction conditions from the original system. The gas chromatogram
corresponding to the 13CO peak and the mass spectra for the photocatalytic 13CO2 reduction
system are presented in Supplementary Figure 8. The results point to the formation of 13CO
and provides direct evidence that the origin of the evolved CO is indeed from the reduction
of CO2 itself.
The kinetic study (Figure 3c) points to a constant CO evolution rate over a 6 hour testing
period, which shows that the catalyst was able to maintain its photoactivity throughout the
course of the reaction. We tested further the recyclability of the catalyst over 3 cycles under
UV-vis irradiation (Figure 3d). We observed a 14% decrease in CO production after the third
cycle (Figure 3d).
Further, we checked the stability of the material through XRD pattern of the used catalyst
and observed a slight increase in crystallinity after CO2 photoreduction (Supplementary
Figure 9). This finding suggests that amorphous domains of the material degraded upon
reaction. This is supported by N2 sorption tests, which showed a 25% decrease in porosity
after reaction using H2 or H2O. Notably though, the porous BN sample still maintained a high
surface area and porosity of 1196 m2 g-1 (1190 m2 g-1 for H2O) and 1.01 cm3 g-1 (0.94 cm3 g-1
for H2O, Supplementary Table 1). We did not observe any distinct changes in the FT-IR
13

Journal of Materials Chemistry A
Page 14 of 26
View Article Online
DOI: 10.1039/C9TA02793A
spectrum of the sample after undergoing a CO2 photoreduction cycle (Supplementary Figure
9). Similarly, we did not observe any changes in the sample morphology (Supplementary
Figure 1) or the relative atomic composition of the sample after undergoing a CO2
photoreduction cycle (Supplementary Figure 4).
To further demonstrate that porous BN can facilitate the gas phase photoreduction of CO2 as
a photocatalyst, we resolved the band structure of the material, specifically the positions of
the valence band (VB) and conduction band (CB) on the absolute energy scale. Importantly,
despite numerous reports of the band gap, which only gives the relative positions of the CB
and VB, the band positions for either porous BN or h-BN on the absolute scale are yet to be
reported in the literature. We first conducted valence band XPS measurements on porous
BN to determine the relative distance between the Fermi level (EF) and the onset of the
valence band, also known as the valence band offset. The average valence band offset was
determined to be 3.05 eV ± 0.02 eV (Figure 3e). High resolution core level spectra (B 1s, N
1s, C 1s, and O 1s) were simultaneously measured along with the valence band spectrum
for porous BN and no effects of charging were observed. Aligning the complete band
structure of porous BN on the absolute energy scale requires knowledge of the absolute
position of the Fermi level (EF). This can be obtained by conducting work function
measurements to measure the secondary electron cut-off through photoelectron
spectroscopy. The work function of the material (Φ) gives the distance between Fermi level
of the material (EF) and the absolute vacuum level (0 eV). We note that work function
measurements conducted through photoelectron spectroscopy are extremely sensitive to
subtle changes in experimental conditions and necessitate the preparation of a well-
controlled, homogeneous surface. To negate the effects of surface roughness, a known
mass of porous BN was pelletized to form a thin homogeneous surface. Further, to account
for potential variations across the surface of the material, the work function was measured at
four different locations and an average was taken (Supplementary Figure 10). The average
work function (Φ) for porous BN was determined to be 2.90 eV ± 0.30 eV. Therefore, as the
14

Page 15 of 26
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C9TA02793A
absolute vacuum level is at 0 eV (the datum for the absolute energy scale), it follows that the
Fermi level of the sample lies at 2.90 eV ± 0.30 eV. Using the valence band offset of 3.05 ±
0.02 eV in Figure 3e), the valence band must lie at 5.95 eV ± 0.32 eV on the absolute
energy scale. Hence, if the band gap of the material (EG) is 4.21 eV ± 0.05 eV
(Supplementary Figure 5), the conduction band is positioned at 1.74 eV ± 0.37 eV on the
absolute energy scale. To align the absolute energy scale vs. vacuum with the redox
potential scale vs. the standard hydrogen electrode (SHE), we used the fact that 4.44 eV on
the former corresponds to 0.00 V on the latter at 298 K.67, 68 In doing so, we can align the
known redox potentials for CO2/CO (-0.53 V) and H+/H2 (0.00 V) reported in the literature on
the absolute energy scale.69 Combining this data with the values calculated from the valence
band and work function measurements yields the full band structure for porous BN shown in
Figure 3f. A detailed derivation and error propagation analysis is presented in the
supplementary information. The key conclusions from this are two-fold. Firstly, the CB and
VB of porous BN do indeed straddle the redox potentials for CO2/CO and H+/H2,
respectively, in the photoredox system with a reduction overpotential (ηCO2/CO) of 2.16 ± 0.37
eV. This confirms that the band structure of porous BN are aligned to provide a sufficient
thermodynamic driving force to facilitate the reduction of CO2 to CO.
In addition, the separation of the Fermi level (EF) to the CB is 1.16 eV ± 0.34 eV, whilst the
distance from the former to the VB is 3.05 eV ± 0.02 eV (Figures 3e and 3f). As the Fermi
level lies closer to the CB than the VB, porous BN behaves as an n-type semiconductor.
Therefore, electrons are the major charge carriers in the photocatalyst, which is ideally
suited to fulfilling the aim of CO2 conversion through gas phase photoreduction. To
conclude, the analysis conducted herein shows that the band structure of porous BN
enables it to be an ideal catalyst to facilitate gas phase CO2 photoreduction and further
substantiates the claim that the CO evolved is indeed due to the action of the porous BN
catalyst itself.
15

Journal of Materials Chemistry A
Page 16 of 26
View Article Online
DOI: 10.1039/C9TA02793A
Figure 3 I Photocatalytic testing of the porous BN sample. (a) CO2 adsorption isotherms
collected at 25 °C for the materials investigated in this study (note the isotherm for h-BN is
directly overlapped by that of g-C3N4), (b) Specific production rate of CO for porous BN,
compared to TiO2 P25, h-BN and carbon nitride (g-C3N4) as reference materials, under UV-
16

Page 17 of 26
Journal of Materials Chemistry A
View Article Online
vis and pure visible light irradiation and H2 or H2O as sacrificial agents (300 W XeDaOrI:c10l.a10m39p/C,9TA02793A
9.5 cm distance from catalyst to lamp source, 5 hours, 25 °C). Intensity at the catalyst
-2
-2
휆
surface was 53800 W m without the UV-filter and 53100 W m with the UV filter ( < 400
nm), (c) Photocatalytic CO evolution from porous BN as a function of time, (d) CO evolution
from porous BN repeated over three 5 hour photocatalytic cycles, (e) Valence band XPS
spectrum for porous BN to ascertain the distance between the Fermi level and valence band
maximum, (f) Band structure for porous BN on the absolute energy scale vs. vacuum with
the conduction and valence bands, Fermi level, and redox potentials for CO2/CO, H+/H2 and
O2/H2O illustrated.
Table 1. CO2 adsorption capacities (25 °C) and specific CO evolution rates of the materials
investigated in this study.
Sample
CO2 adsorption
capacity
CO evolution rate
(UV-Vis)
CO evolution rate
(Visible)
(mmol g-1 STP)
(흀
( mol g-1 h-1)
> 400 nm)
μ
( mol g-1 h-1)
μ
Porous BN
P25
1.22
0.21
0.07
0.03
±
±
±
±
±
±
±
1.17
1.08
0.45
0.01
0.02
0.01
0.04
0.01
0.20
0.46
0.33
0.02
0.04
0.03
g-C3N4
h-BN
n.d
n.d – none detected
In summary, we investigate a new material platform based on porous BN material that is
able to facilitate combined CO2 capture and photoreduction in a simple gas/solid phase set-
up under ambient conditions, without the use of co-catalysts. Unlike its crystalline
counterpart, porous BN exhibits semiconducting behaviour and can photoreduce CO2 to CO
under both UV-vis and visible irradiations, with evolution rates higher than the benchmark in
the field, i.e. P25 TiO2. Porous BN exhibits a steady CO evolution rate over a single cycle
and maintained about 85% of its initial performance over multiple catalytic reactions. We
present, for the first time, the complete band structure of porous BN to show that the material
behaves as an n-type semiconductor, with the conduction and valence bands straddling the
redox potentials for CO2/CO and H+/H2. Given the chemical and structural tunability of this
material, there is considerable scope to expand the photocatalyst functionality to the visible-
17

Journal of Materials Chemistry A
Page 18 of 26
View Article Online
DOI: 10.1039/C9TA02793A
light region, and increase its rate performance. This would constitute a major stride in the
field of solar fuels synthesis.
Methods
Synthesis of porous BN. In a typical synthesis, boric acid (H3BO3, ACS reagent, 99.0 %,
Sigma-Aldrich), urea (CH4N2O, molecular biology grade, Sigma-Aldrich), and melamine
(C3H6N6, ACS reagent, 99.0 %, Sigma-Aldrich), with a 1:1 molar ratio of boric acid to melamine
and 1:5 molar ratio of boric acid to urea were mechanically mixed together and finely ground
for 5 minutes in an agate mortar to form a homogeneous powder. The finely ground powder
was subsequently transferred to an alumina boat crucible, which was placed in a horizontal
tubular furnace. The sample was initially degassed at ambient temperature for 3 hours under
an inert nitrogen atmosphere (flow rate of 250 cm3 min-1). Once the degas was complete, the
nitrogen flow rate was decreased to 50 cm3 min-1, and the sample was heated from ambient
temperature to 1050 °C with a ramp rate of 10 °C min-1. This steady-state temperature was
maintained for 3.5 hours, after which the samples were allowed to naturally cool to room
temperature, whilst maintaining the same nitrogen flow rate. Upon completion of the synthesis,
a porous white powder was obtained. The reference materials, h-BN and P25 TiO2 were
industrially synthesised and obtained as follows: h-BN (Tres BN, max. particle size 74 nm)
and P25 TiO2 (Sigma Aldrich, >99.5%, 21 nm primary particle size).
Synthesis of g-C3N4. In a typical synthesis, 12 g of melamine (C3H6N6, ACS reagent, 99.0 %,
Sigma-Aldrich) were weighed and transferred to an alumina crucible, which was placed in a
muffle furnace. The sample was heated from ambient temperature to 560 °C with a ramp rate
of 5 °C min-1. This steady-state temperature was maintained for 4 hours, after which the
samples were allowed to naturally cool to room temperature. Upon completion of the
synthesis, a yellow solid product was obtained, which was subsequently ground in an agate
mortar to form a fine, homogeneous powder.
18

Page 19 of 26
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C9TA02793A
Characterization. For FT-IR analysis, the samples were first ground to a powder using an
agate mortar. Subsequently, the spectra were obtained in the range of 500 – 4000 cm-1 using
a Perkin-Elmer Spectrum 100 FT-IR spectrometer equipped with an attenuated total
reflectance (ATR) accessory. X-ray photoelectron spectroscopy (XPS) to determine the
relative elemental composition of the samples, the chemical states of the elements, the
valence band off-set and secondary electron cut-off was conducted using a Thermo Scientific
+
훼
K-Alpha X-ray Photoelectron Spectrometer equipped with a MXR3 Al K monochromated X-
휈
ray source (h = 1486.6 eV). The samples were initially ground and mounted onto an XPS
sample holder using a small rectangular piece of conductive carbon tape. The X-ray gun power
was set to 72 W (6 mA and 12 kV). Survey scans were acquired using 200 eV pass energy,
0.5 eV step size and 100 ms (50 ms x 2 scans) dwell times. All of the high resolution core level
spectra (B 1s, N 1s, C 1s, and O 1s) were obtained using a 20 eV pass energy and 0.1 eV
step size. The valence band spectra were obtained using a 15 eV pass energy and 0.05 eV
step size. The results were analysed using the Thermo Avantage data analysis program.
Powder X-ray diffraction (XRD) was performed using a PANalytical X’Pert Pro X-ray
diffractometer in reflection-transmission mode with a spinning stage (2 revolutions/second).
An anode voltage of 40 kV and emission current of 20 mA were chosen as the operating
훼
휆
conditions using a monochromatic Cu-K radiation source ( = 1.54178 Å). The X’Celerator
silicon strip detector was used in the diffractometer. Nitrogen and CO2 sorption isotherms were
measured using a porosity and surface area analyser (Micrometrics 3 Flex) at -196 °C and
25 °C, respectively. The samples were initially degassed overnight at 120 °C at approximately
0.2 mbar pressure. Subsequently, prior to the sorption isotherm measurement, the samples
were degassed in-situ for 4 hours at 120 °C. The equivalent specific surface areas of the
samples were determined using the Brunauer-Emmett-Teller (BET) method.70 The total pore
volume was ascertained from the volume of N2 adsorbed at a relative pressure (P/P0) of 0.97.
The micropore volume was determined using the Dubnin-Radushkevich model. Transmission
19

Journal of Materials Chemistry A
Page 20 of 26
Electron Microscopy (TEM) images were collected on a JEOL 2100Plus with 20V0iewkAVrticle Online
DOI: 10.1039/C9TA02793A
acceleration voltage. The samples were previously dispersed in ethanol and deposited on a
carbon holey grid. SEM-EDX analysis was performed on an Auriga Zeiss microscope at 5 keV
for both imaging and EDX analysis. The powder samples were compressed on a carbon tape
and coated with 10 nm of gold.
Photocatalytic reduction of CO2. A gas/solid photoreactor setup was assembled to
conduct the photocatalytic reduction of CO2. The as-synthesised, finely ground porous BN
photocatalysts were deposited on a circular metal disc with a fixed area of 9.6 cm2. To do so,
30 mg of porous BN powder was added to 1.5 mL of ethanol and stirred rapidly to form a
homogeneous suspension, which was drop casted onto the sample holder. Research grade
(99.999%) CO2 and H2 (99.9995%, Peak Scientific PH200 hydrogen generator) were flowed
at controlled rates using mass flow controllers (Omega Engineering, 0–50 mL/min). Firstly,
the photoreactor (35 cm3) was vacuumed and replenished with a gas mixture of CO2 and
H2 (1.5 vol/vol ratio) ten times. The same gas mixture of CO2 and H2 was subsequently
passed over the catalyst bed in the photoreactor for 15 residence times before it was sealed
at 1.25 bara and irradiated for 5 hours. For experiments in which water vapour was used as
the sacrificial reductant, this was generated from a saturator at room temperature (20 °C)
providing a CO2/H2O ratio of around 40. Similarly, the photoreactor (35 cm3) was vacuumed
and replenished with a gas mixture of CO2 and H2O ten times. A xenon arc lamp (300 W,
λ > 325 nm, LOT Quantum Design), equipped with a water filter was used as the irradiation
source with the distance from the lamp to the sample being 9.5 cm and the intensity at the
catalyst surface measured to be 53800 W m-2. A long pass UV filter (λ < 400 nm) (LOT
Quantum Design) and a long pass visible filter (λ < 550 nm) (Thor Labs) were used to
conduct the tests under visible light irradiation and the lamp emission spectra, both with and
without the UV filters, are provided in Supplementary Figure 7. The evolved gases were
detected by a gas chromatograph (GC, Agilent Technologies) with HayeSep and molecular
sieve columns in series, and thermal conductivity (TCD) and flame ionisation (FID)
20

Page 21 of 26
Journal of Materials Chemistry A
View Article Online
detectors. During the recyclability tests to assess the photostability of the materiaDlsO,I:t1h0e.1039/C9TA02793A
aforementioned process was repeated after each 5-hour irradiation cycle without opening
the photoreactor. In addition, isotopic tracing experiments were performed with 13CO2 (BOC,
>98% atom 13CO2 compared to 12CO2, >99%).
Time-resolved photoluminescence. Time-resolved emission experiments were conducted
using a DeltaFlex time-resolved single photon counting system (Horiba) equipped with a
PPD-650 detection module (Horiba) and a 282 nm JI-IBH NanoLED (1 ns pulse width,
Horiba) excitation source. The acquired time-resolved emission spectra were analysed using
a custom-built global analysis routine by fitting each dataset to a model (described in further
detail in the Supporting Information), including convolution with the measured instrument
response function.
Electron paramagnetic resonance (EPR) spectroscopy. EPR experiments were acquired
using a Bruker Elexsys E580 CW EPR spectrometer operating at X-band frequencies (9–10
GHz/0.3 T), equipped with a Bruker ER4118-X MD5 resonator. The microwave frequencies
used were 9.61 GHz with microwave power 2mW and in the detection sequence a 100kHz
modulation with 2G amplitude was used. Light measurements of the BN sample was carried
out under direct illumination of a 300W Xe arc lamp. All spectra were recorded at room
temperature in air atmosphere in 4mm EPR suprasil tubes.
Work function measurements. The work function of porous BN was determined by
measuring the secondary electron cut-off in the low kinetic energy region using a Thermo
+
훼
Scientific K-Alpha X-ray Photoelectron Spectrometer equipped with a MXR3 Al K
휈
monochromated X-ray source (h = 1486.6 eV). A known mass (100 mg) of sample was
pelletized to form a thin, homogeneous pellet, which was clipped to the sample holder. The
sample holder contained a clean gold standard sample, which was used as a reference
material to ensure correct calibration. A sample bias of -30 V was applied to the samples
using the ion gun and the cut-off spectra were obtained using a pass energy of 10 eV. To
21

Journal of Materials Chemistry A
Page 22 of 26
account for potential variations across the surface of the material, the work function wasView Article Online
DOI: 10.1039/C9TA02793A
measured at four different locations and an average was taken.
Acknowledgements
The authors would like to thank S. Marchesini for her help in producing the TEM images, to
A. Crake for his technical input in the photocatalytic testing and isotopic testing, to A. Hankin
for her help in measuring the lamp emission spectra, to L. Malms and F. Xie for their help in
obtaining the UV-Vis diffuse-reflectance measurements. The authors would also like to
acknowledge the funding from EPSRC through the Doctoral Partnership fund (1855454),
which made this research possible, as well as the EPSRC equipment funding for SPIN-Lab
(EP/P030548/1). AR acknowledges support from Imperial College for her Imperial College
Research Fellowship. We also acknowledge support from the Department of Chemical
Engineering at Imperial College London.
Declaration
The authors declare that there are no competing financial interests.
References
1.
G. Centi; E. A. Quadrelli; S. Perathoner. Energy & Environmental Science, 2013, 6,
1711-1731.
2.
3.
T. Inoue; A. Fujishima; S. Konishi; K. Honda. Nature, 1979, 277, 637-638.
A. Crake; K. C. Christoforidis; A. Kafizas; S. Zafeiratos; C. Petit. Applied Catalysis B:
Environmental, 2017, 210, 131-140.
4. T. W. Woolerton; S. Sheard; E. Reisner; E. Pierce; S. W. Ragsdale; F. A. Armstrong.
Journal of the American Chemical Society, 2010, 132, 2132-2133.
5. T. W. Woolerton; S. Sheard; E. Pierce; S. W. Ragsdale; F. A. Armstrong. Energy &
Environmental Science, 2011, 4, 2393-2399.
22

Page 23 of 26
Journal of Materials Chemistry A
View Article Online
6.
J. Nunez; A. Víctor; P. Jana; J. M. Coronado; D. P. Serrano. Catalysis Today, 2013,
DOI: 10.1039/C9TA02793A
209, 21-27.
7.
F. Yoshitomi; K. Sekizawa; K. Maeda; O. Ishitani. ACS Applied Materials &
Interfaces, 2015, 7, 13092-13097.
8.
N. Zhang; S. Ouyang; T. Kako; J. Ye. Chemical Communications, 2012, 48, 1269-
1271.
9.
A. Crake. Materials Science and Technology, 2017, 33, 1737-1749.
R. Li; W. Zhang; K. Zhou. Advanced Materials, 2018, 1705512.
C. Wang; R. L. Thompson; P. Ohodnicki; J. Baltrus; C. Matranga. Journal of
10.
11.
Materials Chemistry, 2011, 21, 13452-13457.
12. F. Liao; Z. Zeng; C. Eley; Q. Lu; X. Hong; S. C. E. Tsang. Angewandte Chemie
International Edition, 2012, 51, 5832-5836.
13. S. Zhou; Y. Liu; J. Li; Y. Wang; G. Jiang; Z. Zhao; D. Wang; A. Duan; J. Liu; Y. Wei.
Applied Catalysis B: Environmental, 2014, 158, 20-29.
14.
J. Lin; Z. Pan; X. Wang. ACS Sustainable Chemistry & Engineering, 2013, 2, 353-
358.
15.
J. Qin; S. Wang; H. Ren; Y. Hou; X. Wang. Applied Catalysis B: Environmental,
2015, 179, 1-8.
16.
25.
17.
T. Ohno; N. Murakami; T. Koyanagi; Y. Yang. Journal of CO2 Utilization, 2014, 6, 17-
X. Wang; K. Maeda; A. Thomas; K. Takanabe; G. Xin; J. M. Carlsson; K. Domen; M.
Antonietti. Nature Materials, 2009, 8, 76.
18.
19.
P. Niu; M. Qiao; Y. Li; L. Huang; T. Zhai. Nano Energy, 2018, 44, 73-81.
Q. Liang; Z. Li; X. Yu; Z. H. Huang; F. Kang; Q. H. Yang. Advanced Materials, 2015,
27, 4634-4639.
20.
D. J. Martin; K. Qiu; S. A. Shevlin; A. D. Handoko; X. Chen; Z. Guo; J. Tang.
Angewandte Chemie International Edition, 2014, 53, 9240-9245.
21.
22.
M. Shen; L. Zhang; J. Shi. Nanotechnology, 2018, 29, 412001.
W. Tu; Y. Xu; J. Wang; B. Zhang; T. Zhou; S. Yin; S. Wu; C. Li; Y. Huang; Y. Zhou.
ACS Sustainable Chemistry & Engineering, 2017, 5, 7260-7268.
23. Z. Sun; J. M. T. A. Fischer; Q. Li; J. Hu; Q. Tang; H. Wang; Z. Wu; M. Hankel; D. J.
Searles; L. Wang. Applied Catalysis B: Environmental, 2017, 216, 146-155.
24. J. Liu; S. Wen; Y. Hou; F. Zuo; G. J. Beran; P. Feng. Angewandte Chemie, 2013,
125, 3323-3327.
25.
M. Wang; M. Li; L. Xu; L. Wang; Z. Ju; G. Li; Y. Qian. Catalysis Science &
Technology, 2011, 1, 1159-1165.
26.
X. Li; J. Zhao; J. Yang. Scientific Reports, 2013, 3, 1858.
23

Journal of Materials Chemistry A
Page 24 of 26
View Article Online
DOI: 10.1039/C9TA02793A
27.
of Colloid and Interface Science, 2017, 508, 121-128.
28. Q. Weng; D. G. Kvashnin; X. Wang; O. Cretu; Y. Yang; M. Zhou; C. Zhang; D. M.
Tang; P. B. Sorokin; Y. Bando. Advanced Materials, 2017, 29.
29. S. Marchesini; A. Regoutz; D. Payne; C. Petit. Microporous and Mesoporous
Materials, 2017, 243, 154-163.
30. J. Kim; J. Han; M. Seo; S. Kang; D. Kim; J. Ihm. Journal of Materials Chemistry A,
2013, 1, 1014-1017.
J. Pang; Y. Chao; H. Chang; H. Li; J. Xiong; M. He; Q. Zhang; H. Li; W. Zhu. Journal
31.
W. Lei; D. Portehault; R. Dimova; M. Antonietti. Journal of the American Chemical
Society, 2011, 133, 7121-7127.
32.
Balicas. Nature Materials, 2010, 9, 430.
33. J. Lu; K. Zhang; X. F. Liu; H. Zhang; T. C. Sum; A. H. C. Neto; K. P. Loh. Nature
Communications, 2013, 4, 2681.
34. L.-W. Yin; Y. Bando; D. Golberg; A. Gloter; M.-S. Li; X. Yuan; T. Sekiguchi. Journal
of the American Chemical Society, 2005, 127, 16354-16355.
35. C. Huang; C. Chen; M. Zhang; L. Lin; X. Ye; S. Lin; M. Antonietti; X. Wang. Nature
Communications, 2015, 6, 7698.
L. Ci; L. Song; C. Jin; D. Jariwala; D. Wu; Y. Li; A. Srivastava; Z. Wang; K. Storr; L.
36.
M. Zhou; S. Wang; P. Yang; C. Huang; X. Wang. ACS Catalysis, 2018, 8, 4928-
4936.
37.
J. Li; N. Lei; H. Hao; J. Zhou. Chemical Physics Letters, 2017, 672, 99-104.
38.
K. Sivaprakash; M. Induja; P. G. Priya. Materials Research Bulletin, 2018, 100, 313-
321.
39.
M. Florent; T. J. Bandosz. Journal of Materials Chemistry A, 2018, 6, 3510-3521.
S. Alkoy; C. Toy; T. Gönül; A. Tekin. Journal of the European Ceramic Society, 1997,
40.
17, 1415-1422.
41.
S. Marchesini; C. M. McGilvery; J. Bailey; C. Petit. ACS Nano, 2017, 11, 10003-
10011.
42.
Y. Lin; J. W. Connell. Nanoscale, 2012, 4, 6908-6939.
43.
X.-F. Jiang; Q. Weng; X.-B. Wang; X. Li; J. Zhang; D. Golberg; Y. Bando. Journal of
Materials Science & Technology, 2015, 31, 589-598.
44. O. O. Kurakevych; V. L. Solozhenko. Acta Crystallographica Section C: Crystal
Structure Communications, 2007, 63, i80-i82.
45. Z. Yan; J. Lin; X. Yuan; T. Song; C. Yu; Z. Liu; X. He; J. Liang; C. Tang; Y. Huang.
Scientific Reports, 2017, 7, 3297.
46. M. Jedrzejczak-Silicka; M. Trukawka; M. Dudziak; K. Piotrowska; E. Mijowska.
Nanomaterials, 2018, 8, 605.
24

Page 25 of 26
Journal of Materials Chemistry A
View Article Online
DOI: 10.1039/C9TA02793A
47.
48.
49.
R. Geick; C. Perry; G. Rupprecht. Physical Review, 1966, 146, 543.
R. Shankar; S. Marchesini; C. Petit. The Journal of Physical Chemistry C, 2018.
X. Gouin; P. Grange; L. Bois; P. L'Haridon; Y. Laurent. Journal of Alloys and
Compounds, 1995, 224, 22-28.
50. G. Ciofani; G. G. Genchi; I. Liakos; A. Athanassiou; D. Dinucci; F. Chiellini; V. Mattoli.
Journal of Colloid and Interface Science, 2012, 374, 308-314.
51. V. Cholet; L. Vandenbulcke; J. Rouan; P. Baillif; R. Erre. Journal of Materials
Science, 1994, 29, 1417-1435.
52.
53.
54.
J. Riviere; Y. Pacaud; M. Cahoreau. Thin Solid Films, 1993, 227, 44-53.
K. Park; D. Lee; K. Kim; D. Moon. Applied Physics Letters, 1997, 70, 315-317.
A. S. Nazarov; V. N. Demin; E. D. Grayfer; A. I. Bulavchenko; A. T. Arymbaeva; H. J.
Shin; J. Y. Choi; V. E. Fedorov. Chemistry–An Asian Journal, 2012, 7, 554-560.
55.
56.
Y. Chen; J. Wang; Y. Chen; D. Liu; S. Huang; W. Lei. Nanoscale, 2018.
W. Lei; H. Zhang; Y. Wu; B. Zhang; D. Liu; S. Qin; Z. Liu; L. Liu; Y. Ma; Y. Chen.
Nano Energy, 2014, 6, 219-224.
57. L. A. Silva; S. C. Guerini; V. Lemos. IEEE Transactions on Nanotechnology, 2006, 5,
517-522.
58.
59.
J. Wu; W. Zhang. Solid State Communications, 2009, 149, 486-490.
R. Godin; Y. Wang; M. A. Zwijnenburg; J. Tang; J. R. Durrant. Journal of the
American Chemical Society, 2017, 139, 5216-5224.
60. F. M. Pesci; G. Wang; D. R. Klug; Y. Li; A. J. Cowan. The Journal of Physical
Chemistry C, 2013, 117, 25837-25844.
61.
62.
M. Marszewski; S. Cao; J. Yu; M. Jaroniec. Materials Horizons, 2015, 2, 261-278.
Q. Sun; Z. Li; D. J. Searles; Y. Chen; G. Lu; A. Du. Journal of the American Chemical
Society, 2013, 135, 8246-8253.
63. S. Chen; P. Li; S. Xu; X. Pan; Q. Fu; X. Bao. Journal of Materials Chemistry A, 2018,
6, 1832-1839.
64.
D. C. Hurum; A. G. Agrios; K. A. Gray; T. Rajh; M. C. Thurnauer. The Journal of
Physical Chemistry B, 2003, 107, 4545-4549.
65.
66.
B. Choudhury; M. Dey; A. Choudhury. Applied Nanoscience, 2014, 4, 499-506.
Y. Lei; L. Zhang; G. Meng; G. Li; X. Zhang; C. Liang; W. Chen; S. Wang. Applied
Physics Letters, 2001, 78, 1125-1127.
67. S. Koizumi; C. Nebel; M. Nesladek, Physics and applications of CVD diamond. John
Wiley & Sons: 2008.
68.
J. O. M. Bockris; B. E. Conway; R. E. White, Modern aspects of electrochemistry.
Springer Science & Business Media: 2012; Vol. 22.
25

Journal of Materials Chemistry A
Page 26 of 26
View Article Online
DOI: 10.1039/C9TA02793A
69.
J. K. Stolarczyk; S. Bhattacharyya; L. Polavarapu; J. Feldmann. ACS Catalysis,
2018, 8, 3602-3635.
70.
S. Brunauer; P. H. Emmett; E. Teller. Journal of the American Chemical Society,
1938, 60, 309-319.
26