Low-molecular-weight carbon nitrides for solar hydrogen evolution

Article abstract of DOI:10.1021/ja511802c

This work focuses on the control of the polymerization process for melon ( graphitic carbon nitride ), with the aim of improving its photocatalytic activity intrinsically. We demonstrate here that reduction of the synthesis temperature leads to a mixture of the monomer melem and its higher condensates. We show that this mixture can be separated and provide evidence that the higher condensates are isolated oligomers of melem. On evaluating their photocatalytic activity for hydrogen evolution, the oligomers were found to be the most active species, having up to twice the activity of the monomer/oligomer mixture of the as-synthesized material, which in turn has 3 times the activity of the polymer melon, the literature benchmark. These results highlight the role of defects , i.e., chain terminations, in increasing the catalytic activity of carbon nitrides and at the same time point to the ample potential of intrinsically improving the photocatalytic activity of carbon nitride , especially through the selective synthesis of the active phase. (Chemical Equation Presented).

Full text of DOI:10.1021/ja511802c

Subscriber access provided by Yale University Library
Article
Low molecular-weight carbon nitrides for solar hydrogen evolution
Bettina V. Lotsch
J. Am. Chem. Soc., Just Accepted Manuscript • DOI: 10.1021/ja511802c • Publication Date (Web): 24 Dec 2014
Downloaded from http://pubs.acs.org on December 26, 2014
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of the American Chemical Society is published by the American Chemical
Society. 1155 Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Journal of the American Chemical Society
This document is confidential and is proprietary to the American Chemical Society and its authors. Do not
copy or disclose without written permission. If you have received this item in error, notify the sender and
delete all copies.
Low molecular-weight carbon nitrides for solar hydrogen
evolution
Journal: Journal of the American Chemical Society
Manuscript ID: ja-2014-11802c.R1
Manuscript Type: Article
Date Submitted by the Author: 18-Dec-2014
Complete List of Authors: Lau, Vincent; Max Planck Institute for Solid State Research,
Mesch, Maria; University of Bayreuth, Department for Inorganic Chemistry
I
Duppel, Viola; Max Planck Institute for Solid State Research,
Blum, Volker; Duke University, Department of Mechanical Engineering and
Materials Science
Senker, Jürgen; University of Bayreuth, Department for Inorganic
Chemistry I
Lotsch, Bettina; Max Planck Institute for Solid State Research,
ACS Paragon Plus Environment

Page 1 of 11
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
Low molecular-weight carbon nitrides for solar hydrogen evo-
lution
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Vincent Wing-hei Lau†§, Maria B. Mesch‡, Viola Duppel†§, Volker Blumǁ, Jürgen Senker‡, Bet-
tina V. Lotsch†§&*
AUTHOR ADDRESS
† Max Planck Institute for Solid State Research, Heisenbergstraße 1, 70569 Stuttgart, Germany
‡ Department of Inorganic Chemistry III, University of Bayreuth, Universitätsstraße 30, 95447 Bayreuth, Germany
§ Department of Chemistry, University of Munich, Butenandtstraße 5-13, 81377 Munich, Germany
Department of Mechanical Engineering and Materials Science and Center for Materials Genomics, Duke Universi-
ty, Durham, North Carolina 27708, United States
& Nanosystems Initiative Munich (NIM) and Center for Nanoscience, Schellingstraße 4, 80799 Munich, Germany
KEYWORDS Carbon nitride, oligomer, structure-activity relationship, hydrogen evolution, photocatalysis
ABSTRACT: This work focuses on the control of the polymerisation process for melon (“graphitic carbon nitride”), with
the aim of improving its photocatalytic activity intrinsically. We demonstrate here that reduction of the synthesis tem-
perature leads to a mixture of the monomer melem and its higher condensates. We show that this mixture can be sepa-
rated and provide evidence that the higher condensates are isolated oligomers of melem. On evaluating their photocata-
lytic activity for hydrogen evolution, the oligomers were found to be the most active species, having up to twice the activi-
ty of the monomer/oligomer mixture of the as-synthesized material, which in turn has three times the activity of the pol-
ymer melon, the literature benchmark. These results highlight the role of “defects”, i.e. chain terminations, in increasing
the catalytic activity of carbon nitrides and at the same time points to the ample potential of intrinsically improving the
photocatalytic activity of “carbon nitride”, especially through the selective synthesis of the active phase.
INTRODUCTION
One new and promising class of photocatalysts is the
so-called “graphitic carbon nitride”, which are polymers
based on the condensation products of melem (2,5,8-
triamino-tri-s-triazine, Scheme S1).6-8 Note that the term
“graphitic carbon nitride” or “g-C3N4” is used only to keep
consistency with the literature; it is in fact a misnomer
considering that these materials do not possess the stoi-
chiometric ratio C3N4 but always contain a sizeable
amount of hydrogen (≈ 2 wt%), consistent with the com-
position of Liebig’s melon, C6N9H3, a 1D carbon nitride
polymer. Originally inspired by the “harder than dia-
mond” postulate,9 this family of materials were thought to
be a precursor to the predicted β-C3N4 and have been
heavily investigated for their synthetic pathway and struc-
ture.10-16 The main advantages of these materials as photo-
catalysts are that they can be facilely synthesized by heat
treatment of inexpensive precursors (e.g. melamine or
dicyandiamide), are active under visible light irradiation,
and chemically stable. Although their photocatalytic ac-
tivity is low compared to the record holders mentioned
above, there are still significant avenues for performance
enhancement. Some strategies for improving their activity
include texturization or morphology control to increase
the surface area,17-22 chemical modification to red-shift the
Hydrogen production by photocatalytic water splitting
is a promising technology for the storage and direct con-
version of solar into chemical energy. The chemical ener-
gy can then be extracted using, for example, fuel cells
with high efficiency, yielding only water as the by-
product. This cycle forms the basis of the energy para-
digm known as the “Hydrogen Economy”, a sustainable
and environmentally benign alternative to the current
practice based on the combustion of fossil fuels.1 Follow-
ing the seminal publication of Honda and Fujishima on
TiO2-photocatalyzed water splitting four decades ago,2
many photocatalytic materials have been identified, of
3
which the most active ones include La-doped NaTaO3
and the solid solution Ga0.88Zn0.12N0.88O0.12.4 Yet, none of
these photocatalysts satisfies the requirements for com-
mercial application, as they often require complex syn-
thetic processes or are composed of rare and expensive
elements, making their preparation uneconomical. Fur-
thermore, many of these materials, especially the oxides,
have low photocatalytic activity under visible light, the
main constituent of solar irradiation at sea level.5
ACS Paragon Plus Environment

Journal of the American Chemical Society
absorption edge,23-26 and band alignment by coupling with
other semiconductors, photocatalysts, and co-catalysts to
improve the electron-hole separation.27-34
Page 2 of 11
1
2
3
4
5
6
7
8
9
Despite the growing number of publications exploring
these strategies, the literature has focused almost entirely
on the polymer melon as the photocatalyst and over-
looked the rich diversity of other related carbon nitride
materials with different degrees of polymerization. For
example, a change in synthesis conditions (temperature,
pressure, atmosphere or synthesis medium) can yield a
variety of phases, including oligomers and polymers of
triazine, adduct phases of melem, coordination and hy-
drogen-bonded compounds.15,35-38 Recent results have
shown that not only are triazine polymers active for pho-
tocatalytic hydrogen and oxygen evolution, but also
chemical modifications can enhance their activity to val-
ues even exceeding that of the heptazine polymer, mel-
on.39,40
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Another sparsely explored topic concerns the funda-
mental properties of these carbon nitride materials, par-
ticularly the nature of the active species, its mechanistic
operation, and strategies to intrinsically improve their
activity, rather than extrinsically by red-shifting the ab-
sorption through doping and increasing the surface area
through morphology control. Since highly crystalline
melem and melon, obtainable only by synthesis in closed
ampules under an autogenous pressure of ammonia
(which induces partial de- and repolymerization, thereby
leading to defect healing14), have negligible photocatalytic
activity, the currently accepted postulate is that “surface
terminations and defects seem to be the real active sites”.7
A more detailed understanding of its catalytic aspects can
provide the guiding principles for bottom-up design and
synthesis of carbon nitride photocatalysts, particularly
when many synthetic modifications are now available for
the heptazine skeleton.41
Figure 1 a) Photocatalytic hydrogen evolution of the as-
synthesized melem oligomer, compared with melon and b)
extended photocatalytic test of the as-synthesized melem
oligomer, both in pH 7 phosphate buffer system (PBS).
After every overnight run, the reactor headspace was evac-
uated and purged with argon. Methanol (500 µL) was add-
ed on the 52nd and 118th hour after the overnight run.
Thus, to gain further insights into the nature of the ac-
tive sites and identify the criteria necessary for photocata-
lytic activity in the carbon nitride family, we direct our
focus away from the polymer. Specifically, we investigate
here the synthesis and characterization of a carbon ni-
tride material with lower molecular weight than the pol-
ymer, and how its structural features affect its perfor-
mance for photocatalytic hydrogen evolution when com-
pared to the polymer.
The product obtained at the reduced synthesis temper-
ature of 450 °C for 12 hours of calcination at 5 °C min-1
ramp, tentatively notated as “as-synthesized melem oli-
gomer”, exhibited a photocatalytic activity more than
twice that of melon in acidic, neutral or alkaline condi-
tions under AM 1.5 conditions (Figure 1 and Figure S2).
Here, the benchmark material melon was synthesized at
550 °C for 12 hours in argon, corresponding to the typical
literature protocols;43 crystalline melon was not used as
we have verified its negligible activity. The photocatalytic
activity of the as-synthesized melem oligomer was sus-
tained for over 120 h without any decrease, demonstrating
that the catalyst stability is comparable to melon. As will
be shown and discussed in more details below, the as-
synthesized melem oligomer absorbs less in the visible
spectrum than melon, indicating that its better activity is
not due to collection of more photons, but rather either
RESULTS & DISCUSSIONS
In our methodology, we selected synthetic conditions
that favor the condensation of melamine into melem and
higher condensation products, but not completely to
melon (temperatures between 380 and 490 °C based on
results from temperature-programmed XRD and thermo-
gravimetric analysis).15,42 We also employed an open ves-
sel rather than a closed ampule, such that the ammono-
lytic depolymerization did not occur, thus favoring the
kinetically rather than thermodynamically driven for-
mation of low condensed melem oligomers, in line with
the increased formation of chain terminations.
ACS Paragon Plus Environment

Page 3 of 11
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 2 a) XRD patterns, b) ATR-IR and c) diffuse reflectance UV-Vis spectra of the melem oligomer and melon in com-
parison with that of crystalline melem. The tick marks in the XRD show the prominent peaks of melem from the litera-
ture.
due to improved charge separation or to interfacial charge
transfer.44,45 We consider improved charge separation as
an unlikely (albeit not ruled out) explanation for the bet-
ter activity, since the photo-excited charge carriers in
carbon nitrides have short lifetime and appear to be con-
fined to the individual heptazine unit without diffusing
along the polymer or oligomer.46,47 The latter explanation
appears more likely, and suggests that the melem oligo-
mer contains features that accelerate the oxidation of
methanol or reduction of hydrogen via the platinum cen-
ters. These encouraging photocatalysis results thus moti-
vated us to determine the identity of this melem oligo-
mer, with the aim of elucidating the properties or criteria
for superior photocatalytic activity in carbon nitride ma-
terials.
possibly 1316 cm-1 have been shown to be characteristic of
the C–NH–C unit in melam, and thus assigned to bridging
secondary amine units.51 We also note that the UV-Vis
absorption for melem oligomer is between that of melem
and melon. Assuming that melon behaves like a partially
π-conjugated heptazine polymer,6 this result suggests that
the aromatic system is larger than the monomer but
smaller than the polymer. We however caution that re-
cent optical investigations question this organic semicon-
ductor model, with results from photoluminescence de-
cay showing instead that the electronic structure of
g-C3N4 can rather be described as that of a “quasi-
monomer”, that is, a low degree of delocalization between
the individual heptazine units.47,52 A separate explanation
independent of the semiconductor model is that the
melem oligomer contains an amount of reacted primary
amines between that of the monomer and polymer, since
derivatization of the amines on melem can also red-shift
its absorption.48
The XRD pattern of the as-synthesized melem oligo-
mer, depicted in Figure 2a, is similar to melem but shows
various small shifts in the reflection positions, indicating
that the crystalline portion of the oligomer sample con-
tains a melem-related structure that has been observed
and employed for synthesizing melem derivatives in the
literature.48 TEM analysis of the heat-treated melem oli-
gomer (Figure S4) showed the sample to have a sheet-like
morphology, with lateral dimensions in the micron scale.
Electron diffraction analysis of this as-synthesized melem
oligomer sample shows the presence of melem, an exam-
ple given in the center image of Figure S4. An additional,
melem-related phase was also observed by electron dif-
fraction (Figure S4, right) with slightly increased d-
spacing for both 002 and 011, consistent with the melem-
related phase in the XRD mentioned above. While carbon
nitrides have been synthesized under similar conditions
before, e.g. Guo et al. performed the synthesis under iden-
tical conditions except in air,49,50 we found that synthesis
in air worsened the reproducibility of the XRD pattern
and also reduced the photocatalytic activity (which may
hint at the negative effect of oxygenated or oxidized func-
tionalities).
Lastly, the C:N atomic ratio of the as-synthesized
melem oligomer sample lies between that of melem and
the theoretical C3N4, indicating an increased degree of
condensation compared to melem. Combined with the
XRD and TEM analyses, these results suggest that this
melem oligomer sample contained a mixture of mono-
meric, crystalline heptazine monomers (i.e. melem) and
more highly condensed but less crystalline products.
While we do not rule out the possibility of forming tria-
zine oligomers or oligomers between melem and triazine,
we believe their presence to be improbable on thermody-
namic considerations. For example, melam (triazine-
triazine dimer, Scheme S1) is a short-lived intermediate in
the synthesis of melem,15 while polymers of triazine are
not expected to form unless under ionothermal condi-
tions.39
Characterization of the spent as-synthesized catalyst
after the photocatalytic reaction showed the formation of
a melem hydrate phase,38 as evident in the XRD patterns,
but virtually no change in the IR spectra, from both the
8 h and the 129 h experiments (Figure S5 and S6 respec-
tively). The hydrate phase is likely formed from the re-
crystallization of the monomers present in the sample
under the aqueous conditions of the photocatalytic reac-
tion, rather than depolymerization of the higher conden-
Despite the fact that only the monomer is observed by
crystallographic techniques in the as-synthesized melem
oligomer sample, results from ATR-IR and diffuse reflec-
tance UV-Vis spectroscopy (Figure 2b and 2c), as well as
elemental analyses (Table S1) are consistent with those
expected of higher condensation products of melem. In
the IR spectra, the absorptions at 1206 and 1235 cm-1 and
ACS Paragon Plus Environment

Journal of the American Chemical Society
sation species, since the IR absorption bands correspond-
Page 4 of 11
ered (mg)
175 (48%)
1
2
ing to the bridging secondary amines are still present.
White
DMSO soluble
HRTEM analyses of the spent catalyst (Figure S7)
showed the formation of spherical platinum particles of
diameter 1–2 nm on the catalyst surface as found previ-
ously.43 X-ray photoelectron spectroscopy (XPS) of the
spent catalyst shows the Pt 4f 7/2 peak (Figure S8, top
right) to have a binding energy (B.E.) of around 71.3 eV,
which is assigned to Pt0. This thus confirms the formation
of metallic platinum, although HRTEM and XPS naturally
do not allow us to determine the molecular binding sites
of the metal particles to the light harvester. Integrity of
the melem oligomer was also corroborated from spectra
in the C 1s (Figure S8, bottom left) and N 1s regions (Fig-
ure S8, bottom right). The C 1s peak at 288.1 eV has been
assigned to a C–N–C coordination, while the N 1s peaks at
398.6 eV, 399.4 eV and 401.2 eV have been assigned to the
C–N–C group, the tertiary nitrogen N–(C)3 group, and the
hydrogen-bearing amine group, respectively.53 The larger
intensity of the peak at 401.2 eV, compared to literature
spectra, can be attributed to the presence of the monomer
melem in the sample, which is in line with our observa-
tions outlined above.54 No change is discernable when
comparing the C 1s and N 1s spectra of this material be-
fore and after the photocatalytic reaction, other than the
absence of platinum.
3
4
5
6
7
8
9
Fraction RCF
60000
57 (16%)
132 (35%)
Off-white/light yel-
low
Fraction RCF
10000
Light yellow
In the DMSO soluble fraction, the most intense reflec-
tions in the XRD coincide with the (011) and the (212)
reflections of melem hydrate, the water likely originating
from residual moisture in the DMSO. Thus, this fraction
can be assigned to the hydrate phase, indicating that
melem from the as-synthesized sample was reorganized
into the hydrate phase during the separation procedure,
whilst keeping the molecular structure and integrity of
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
the triamino-heptazine.
A small amount of sulfur
(~1 wt%) was detected only in this fraction (but not the
other two) by elemental analysis using energy dispersive
x-ray spectroscopy (EDX), which we attribute to the mi-
nute presence of a DMSO-solvate of melem.55 Its IR spec-
trum shows two broad bands at 1596 and 1442 cm-1, again
coinciding with those of melem, while the lack of absorp-
tion at around 1231 cm-1 indicates the absence of the oli-
gomeric species in this fraction. Its absorption onset in
the UV-Vis spectra, which is below 400 nm, suggests an
underivatized monomeric structure. The 13C NMR spec-
trum exhibits two signal groups at 164 ppm and at
155 ppm which can be assigned to the amine-bearing and
the remaining heptazine carbons, respectively. In melem,
both signals exhibit a fine structure, due to different hy-
drogen bonding environments.10 Owing to the weak crys-
tallinity, the signals for the present material are not as
resolved and slightly shifted to lower ppm values similar
to melem hydrate.38 Three peaks in the 15N can be as-
signed thus: the broad signal in the range of -190 to -
210 ppm corresponds to the peripheral nitrogens in the
heptazine ring, -229 to -240 ppm to the center nitrogen of
heptazine, and -265 to -285 ppm to the amine. The values
of these shifts coincide with those of melem or melem
hydrate.10,38 Lastly, the mass spectroscopy showed peaks
at 219 and 420 a.m.u. (Figure S9), corresponding to the
protonated forms of the monomer and the dimer respec-
tively, but nothing of higher mass. It is possible that the
dimer is somewhat soluble in the DMSO, or may be pre-
sent as a contaminant in minute amount below the detec-
tion limit of NMR (≈ 5%). From this characterization, we
can conclude that the DMSO-soluble fraction is predomi-
nantly melem, illustrating the separation of the monomer
from the higher condensates.
As we are interested in elucidating the structure-
property-activity relationship in carbon nitride photo-
catalysts, it is thus necessary to separate these higher
condensation species in the sample from the monomer,
which has not been reported to possess photocatalytic
activity. Since DMSO can solubilize melem13 but does not
affect the polymer chemically, we therefore can remove
the majority of the monomer from the bulk melem oli-
gomer sample and isolate the oligomer, which will have a
lower solubility in DMSO. The insoluble component was
further separated by differential centrifugation at RCF
(relative centrifugal force) values of 10000 and 60000.
Descriptions of the fractions are summarized in Table 1
and their characterization by XRD, UV-Vis, IR, MALDI-
13
15
TOF-MS, and solid-state C and N magic angle spinning
(MAS) NMR spectroscopy are shown in Figure 3. MALDI-
TOF spectra of the fractions at lower mass ranges are
shown in Figure S9. A summary of the BET surface areas,
photocatalytic hydrogen evolution rate and quantum
efficiencies are given in Table S2.
Table. 1 Fractions obtained by differential centrifuga-
tion of a DMSO suspension containing melem oli-
gomer (400 mg in 100 mL).
Amount recov-
Appearance
ACS Paragon Plus Environment

Page 5 of 11
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 3 a) XRD patterns, b) ATR-IR, c) diffuse reflectance UV-Vis, d) MALDI-TOF at mass range >850 a.m.u. in positive
mode, e) C and f) N solid state CP-MAS-NMR spectra, and g) the postulated average structures of the three fractions
from differential centrifugation.
13
15
The other two fractions showed similarities in their IR
spectra with that of melon, the noteworthy feature being
the bridging amine absorption at 1231 cm-1, while their
UV-Vis absorption onsets are in the visible region. Com-
pared with melon, their XRD patterns exhibit a relatively
sharp reflection at 12.8° 2θ and broad reflections in the
range 25-29° 2θ suggesting long-range order to some ex-
tent across the plane and largely turbostratic disorder in
the stacking direction. These two fractions are however
differentiated by their NMR spectra. In the 13C spectrum,
the RCF 60000 fraction is characterized by the same
shoulder at 154 ppm as the DMSO soluble fraction, sug-
gesting a more melem-like carbon environment. None-
theless, both these fractions are unequivocally not the
monomer, as evidenced by the peak at -245 ppm in the 15N
spectrum, which coincides with the secondary amine (NH
group) in melon. We also note the expansion of the signal
of the peripheral heptazine nitrogens towards higher ppm
values, being characteristic for melon;14 this peak is more
intense for the RCF 10000 than the 60000 fraction. Lastly,
we emphasize that the shift of the central heptazine ni-
trogen provides information regarding the extent of the
polymerization, since this peak shifts from -234 ppm for
melem to -225 ppm for the polymer.10,14 Although this
peak is broad for both fractions, a comparison of the max-
ima revealed that the shift is at -226 ppm for the RCF
10000 fraction, whereas it is at -228 ppm for the RCF
60000 fraction, thus suggesting that this latter fraction is
less melon-like than the former. An additional qualitative
(but by no means quantitative, due to cross polarization
condition) measure of polymerization is to compare the
integral ratios of the 15NH2:15NH peaks in the different
spectra. In this case, this ratio is around 3.3 and 2.4 for the
RCF 60000 and 10000 fractions, respectively, indicating
that the former has more unreacted primary amine
groups, i.e. less degree of polymerization. As the ratio
depends on the chain length (Figure S10), the size of the
oligomer can thus be approximated. Therefore, the less
dense fraction (RCF 60000) with a ratio of 3.3 will mainly
consist of dimers with an average length of 2.3 melem
units and only a small amount of higher condensates,
whereas the denser fraction (RCF 10000) is shifted to-
wards higher condensates, mostly trimers with an average
length of 3.2 melem units. The postulated structures of
ACS Paragon Plus Environment

Journal of the American Chemical Society
these two fractions as linear polymers are shown in Fig-
Page 6 of 11
1
2
3
4
5
6
7
8
ure 3g. Note that as these fractions are inevitably mix-
tures, the oligomer lengths shown are average structures
that are consistent with the NMR results. While we do
not rule out the possibility of non-linear structures (e.g.
dendrimer or cyclic structures35), no signal corresponding
to a tertiary amine was detected from the 15N NMR spec-
trum. We therefore surmise that non-linear structures
involving a tertiary nitrogen are either absent or present
only in negligible amounts. Regarding the possible pres-
ence of triazine moieties, we note that the spectral signa-
ture of the 13C and 15N 1D spectra clearly fit better to a
heptazine-based oligomer, rather than a triazine-based
one.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
As spectral overlap could mask the minute presence of
such triazine-species, we have performed matrix-assisted
laser desorption/ionization time of flight (MALDI-TOF)
mass spectrometry to corroborate the NMR results. Both
samples exhibit peaks at mass corresponding to the pro-
tonated monomer (219 a.m.u.), dimer (420 a.m.u.), trimer
(621 a.m.u.), and further increments of 201 a.m.u. corre-
sponding to every additional monomer up to the nonam-
er (1627 a.m.u.) can be clearly resolved, while no triazine-
containing oligomer was observed. These signals are un-
likely to be fragments of the polymer since the spectrum
for melon does not contain these peaks. In contrast, in
analyzing melon at the high mass range, a series of peaks
with mass separated by 24 a.m.u. was observed. Note that
this separation is also observed for the RCF 10000 frac-
tion, and is assigned to the C2 fragment resulting from the
high laser power required to generate sufficient signal
intensity. While we acknowledge previous MALDI-TOF
analysis of heptazine polymers showing mass intervals of
201 a.m.u.,56 we emphasize that the spectra in this work
cleanly show the molecular peaks at the exact masses
expected for the oligomer, rather than the fragments.
Although MALDI-TOF was unable to differentiate be-
tween these two fractions in terms of their oligomer
length, it nonetheless verifies our hypothesis that these
are indeed isolated oligomers of melem.
Figure 4 TEM image of the fraction obtained at RCF 60000.
Figure 5 Photocatalytic activity under visible light irradiation (λ >
420 nm) of the fractions in PBS (18 mL of 0.1 M solution at pH
7) containing methanol (2 mL) and dihydrogen hexachloroplatiꢀ
nate (5 µL of 8 wt% aqueous solution). Catalyst loading in all
cases was 20 mg.
The thermal properties of these fractions and evolved
volatiles were analyzed by thermogravimetric analysis
coupled with mass spectroscopy under an argon atmos-
phere (Figure S11). Melon has one mass loss step, begin-
ning at around 520 °C accompanied by evolution of HCN,
before complete decomposition takes place at 620–700 °C
with the evolution of C2N2 and small amounts of NH3. The
DMSO soluble fraction – identified above to be melem –
undergoes two mass loss steps, the first beginning at
around 350 °C with evolution of ammonia and small
amounts of HCN, which we attribute to the condensation
polymerization of the monomer. The second step is simi-
lar to melon decomposition, with HCN and C2N2 evolu-
tion onset at around 520 °C and 600 °C, respectively. Both
RCF 10000 and 60000 fractions have an identical mass
loss profile with a single step mass loss at a temperature
between that of the monomer and the polymer, con-
sistent with the fact that these fractions are intermediates
between the monomer and the polymer. The condensa-
tion reaction, as in melem, was not detected, and can be
rationalized by the negligible presence of the monomer as
precursor.
The RCF 60000 fraction has a layered structure similar
to the plate-like morphology observed in crystalline parti-
cles of melon,14,20 as shown in its SEM (Figure S12) and
TEM (Figure 4), with lateral dimensions in the micron
regime. These fractions were tested for photocatalytic
hydrogen evolution at pH 7 under irradiation from visible
light (>420 nm) and the results are shown in Figure 5. The
DMSO-soluble fraction, which we have assigned to
melem, exhibited negligible activity as expected, since its
absorption onset is in the UV region. The fractions col-
lected at RCF 10000 and 60000, which together constitute
ACS Paragon Plus Environment

Page 7 of 11
Journal of the American Chemical Society
about 50% of the total weight of the original sample, ex-
hibited a hydrogen evolution rate of 4.8 and 5.4 µmol h-1
1
2
respectively, with an apparent quantum efficiency (AQE)
estimate at 400±25 nm of 0.087% and 0.1%; these hydro-
gen evolution rates outperformed the as-synthesized
sample by 1.66 and 1.71 times, respectively, and outper-
formed the polymer (0.9 µmol h-1, AQE 0.015% at
400±25 nm) by 5.9 and 6.1 times, respectively. The BET
surface areas of these two active fractions are no more
than three times that of the polymer (see Table S2), nor
does their morphology differ significantly from melon
(other than having small particle size in general as a re-
sult of the differential centrifugation). Solvent interaction
and dispersibility of the polymer and the RCF 10000 frac-
tion are also quite similar, as gauged by zeta potential
measurements (Figure S13), which show that the polymer
and this fraction has zeta potentials of -18 mV and -28 mV
at pH 6.6, and isoelectric points of pH 3.3 and pH 2.7,
respectively. The values for the polymer are in broad
agreement with literature values.27 Thus, while textural
properties, solvent effects and mass transport may have
an effect on the photocatalytic performances, the vast
activity differences of the fractions indicates that these
attributes are not the dominating contributors, but point
towards other intrinsic factors as the principal explana-
tion for the large outperformances of the oligomeric frac-
tions, i.e. increased activity per surface area. The separa-
tion procedure using DMSO does not affect the heptazine
polymer chemically, thereby ruling this factor out as the
source of the increased intrinsic activity. We then consid-
er the common explanations for intrinsic increase in ac-
tivity in photocatalysis, which are: 1) improvement in light
absorption, 2) lengthened lifetime of photo-generated
charges and/or 3) interfacial charge transfer rate. We can
rule out the first factor immediately, as these active frac-
tions absorb less in the visible region than the polymer,
based on their diffuse reflectance UV-Vis spectra. These
active fractions are also poorly crystalline from the XRD, a
factor usually associated with enhanced electron-hole
recombination and impeded charge percolation.57 Fur-
thermore, the carrier lifetime and mobility of heptazine
polymers are considered to be disfavored for photocataly-
sis, due to strong carrier localization.47,58-60 The life-
time/mobility interplay may be varied slightly by chang-
ing the stacking separation, with the distance directly
correlated with carrier lifetime but inversely correlated to
mobility.47 This relation however may not be applicable
for oligomers due to differences in crystal structure and
co-facial distances in the aromatic units.47 We thus con-
sider the second factor – increased activity due to more
favorable carrier dynamics – an improbable explanation
for the large difference in activity (even after normalizing
the oligomers’ activities to surface area; see below).
Therefore, the remaining explanation, that the high activ-
ity of the as-synthesized melem oligomer is attributable
to faster extraction of the photo-generated charges (i.e.
faster kinetics for interfacial charge transfer), is consid-
ered the likeliest explanation. This step is typically the
slowest and thus the bottleneck in most photocatalytic
systems. Faster transfer kinetics can result from the cata-
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 6 Highest occupied and lowest unoccupied orbitals
of the monomer to the tetramer.
lyst having larger number of exposed reactive sites, which
may be a surface area effect. However, when the activities
are normalized to the BET surface area, we see that the
activity of the RCF 60000 fraction at 11 µmol h-1 m-2 is
more than twice the activity of the polymer
(5.3 µmol h-1 m-2), thus illustrating that the surface area
cannot be the principal factor in explaining this outper-
formance.
To understand the high activity of the oligomers, we
have undertaken calculations of the HOMO/LUMO or-
bitals of the mono- and oligomers and their variation with
conformation, employing van der Waals-corrected densi-
ty functional theory61,62 for the structure and an optimized
hybrid functional63 to validate orbital energy trends, using
the FHI-aims all-electron code64,65 (orbitals represented in
Figure 6 and Table S4; energies summarized in Table S3;
conformational variations in Table S4 and additional
discussions in the supporting information). The computa-
tional trends are corroborated with photoluminescence
(PL) experiments (Figure S14). Consistent with a previous
study on the correlation of band gap with degree of
polymerization,52 the energy gap was found to expand
with shorter oligomer length. Current Kohn-Sham density
functional approximations cannot be expected to give
absolute band gap energies, but the trend of the band gap
and HOMO/LUMO energies is consistent across different
functionals. For the optimized functional (PBE61 function-
al), the gas-phase electronic gap increases from 7.83
(2.68) eV for the tetramer, to 7.93 (2.77) eV for the trimer
and 8.15 (2.95) eV and 8.98 (3.50) eV for the di- and mon-
omer, respectively. From the PL spectra, both the oligo-
meric fractions and the polymer possess an emission state
at about 375 nm, but differ for the emission in the visible
region. The polymer has an emission at 455 nm, while the
spectra for the oligomers have a shoulder in the main
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 11
emission peak at around 425 nm. The energy gap for these
metal organic framework75).These amines allow either
1
2
3
4
5
6
7
8
emissions correspond to 2.7 eV for the polymer and 2.9 eV
for the oligomers. While the absolute agreement with
DFT-PBE is fortuitous, again the trend with length is in
good agreement with the computational results. This shift
in energy level is due to the LUMO shifting to more posi-
tive energy while the HOMO shifts much less with oligo-
mer length, implying that the photoelectrons in the
shorter oligomer are thermodynamically more reducing
than the longer ones (i.e. more driving force for electron
transfer), while the photo-holes have largely invariant
oxidizing ability. In relating this LUMO shift to the activi-
ty, we find that the absorption edge from the UV-Vis
spectra for the oligomeric fractions (440 nm, 2.8 eV) is
blue-shifted by about 20 nm (0.1 eV) from the polymer
(460 nm, 2.7 eV). Likewise for the PL, the emission is
blue-shifted by 30 nm (0.2 eV) going from the polymer to
the oligomer. In other words, the oligomer fraction has
over fivefold increase in hydrogen evolution rate over
melon (Figure 5; twofold on a per surface area basis) with
only about 100-200 mV potential change. This variation is
large, considering that semiconductor photocatalysts
exhibit far less than a twofold variation in activity for a
similar shift in band potential resulting from quantum
confinement or by varying the composition of solid solu-
tion.66-69 Thus, in addition to the slightly higher reducing
ability of the photoelectron due to the LUMO shift, we
also considered factors related to kinetics, such as the
location and distribution of the charge transfer sites. We
find that for all oligomers the location of the HOMO,
which is likely associated with the oxidation sites where
the photo-holes are quenched after interaction with
methanol as the sacrificial electron donor, are localized
exclusively on the tertiary heptazine nitrogen atoms for
the small oligomers (Figure 6 and Table S4).6,58 Since
these are the positions where direct hydrogen bonding
interactions to methanol is most likely to occur, the
HOMO location is therefore “ideal” for all oligomers.
However, for all oligomers calculated, the LUMO, which
typically is used as an approximation for the reducing
sites,70 are delocalized across the carbon and nitrogen
atoms of the heptazine, including the central nitrogen
and the bridging secondary amine, as well as localized on
the primary amines. The LUMO is not equally spread
across the multiple heptazine units of the oligomer, i.e.,
not entirely delocalized; even stronger localization effects
are observed when conformational variations are ac-
counted for (see Table S4), consistent with the “quasi-
monomer” description.47 An explanation consistent with
our empirical results would be to consider the primary
and/or secondary amines to be the primary reducing sites,
as they are not only more prevalent in the oligomeric than
the polymeric case, but better exposed on account of the
stacking disorder and conformational flexibility for the
less condensed oligomer, as observed in the broad XRD
better coordination to the platinum centers for efficient
electron transfer (e.g. via inner sphere transfer mecha-
nism), or for amine-mediated proton-coupled electron
transfer,76 or even as the actual site for recombination of
the hydrogen atoms to molecular hydrogen.77
These findings therefore lend evidence to the often-
invoked relevance of “defects”, which are commonly con-
sidered to increase the photocatalytic activity of carbon
nitrides, as the nature of such defects can in fact be asso-
ciated with polymer terminations and increased exposure
of the peripheral functional groups (e.g. –NH2). It will
therefore be key to explore whether the most NH2-rich
species - melem and its derivatives - possess photocatalyt-
ic activity, and how the modification of the heptazine
skeleton affects the optoelectronic and photocatalytic
properties. From a computational approach, it will be
instructive to investigate conformer dependency, as ob-
served in the preliminary calculations for the tetramer
(Table S4), where symmetry-breaking and hydrogen
bonding have been shown to shift both the
HOMO/LUMO energy and location, and thus may offer
alternative strategies for activity increase through altered
photogenerated charge carrier location and separation.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The high activity of our experimentally resolved oligo-
mer fractions highlights the potential of active site engi-
neering as an alternative strategy to improve the intrinsic
photocatalytic activity of carbon nitrides. The hydrogen
evolution rates of these oligomers are quite competitive
with values in the literature (after normalizing for catalyst
loading and surface area), which are typically in the range
of 1–4 µmol H2 h-1 m-2 for g-C3N4 that is composited with
other semiconductors.27,30,31,50 Note that direct comparison
is difficult on account of differences in reaction parame-
ters (electron donor used, reactor configuration, and
whether the catalyst loading is limited by mass transfer78).
Nonetheless, we expect the activity of the oligomer can be
further enhanced on a per weight basis provided that the
active phase can be selectively synthesized. We note that
these two “oligomeric” fractions constitute approximately
52 wt% of the as-synthesized sample, with about half of
the melem remaining unreacted even after 12 h at 450 °C.
We foresee that the yield and polymerization degree are
far from optimized, and a simple fine-tuning of the syn-
thesis parameters may already provide ample scope for
further improvement in the photocatalytic activity of
these heptazine-based polymers.
CONCLUSIONS
We have experimentally demonstrated that the length
of the melon polymer chain can be reduced by decreasing
the synthesis temperature, yielding a hydrogen evolution
photocatalyst with three times the activity of the polymer
under AM 1.5 condition, despite absorbing less in the
visible region. Through a separation procedure and de-
tailed characterization, this photocatalyst was shown to
be a mixture of the monomer and oligomers of melem.
The former has negligible activity, while the latter has
reflections. These amine groups would also be the pre-
2-
ferred site for the in-situ reduction of Pt(0) from PtCl6
,
given the numerous literature precedence on the encap-
sulation of platinum particles within photo- and electro-
catalytic amine-bearing matrices (e.g. dendrimer71-74 and
ACS Paragon Plus Environment

Page 9 of 11
Journal of the American Chemical Society
nearly twice the activity of the as-synthesized material on
a per weight basis, resulting in a total increase in activity
AUTHOR INFORMATION
1
2
3
4
5
6
7
8
Corresponding Author
of six times the polymer. Moreover, the photocatalytic
activity also appeared to be inversely correlated with the
oligomer length, consistent with the periphery of the
heptazine moiety being involved in the catalytic reaction,
which has been corroborated by DFT calculations. The
successful isolation of the active component(s) demon-
strated in this work provides a platform for future investi-
gations into the catalytic properties of carbon nitride
photocatalysts. Furthermore, these results illustrate a
strategy based on active site engineering for intrinsically
improving carbon nitride photocatalysts, thus highlight-
ing the need for developing synthetic protocols selective
to the active species.
*Email: b.lotsch@fkf.mpg.de
Funding Sources
VWL gratefully acknowledges a postdoctoral scholarship
from the Max Planck Society. This work was supported by
the Deutsche Forschungsgemeinschaft (projects LO1801/1-1,
SE1417/5-1), the Max Planck Society, the cluster of excellence
Nanosystems Initiative Munich (NIM), the Fonds der Chem-
ischen Industrie (FCI) and the Center for Nanoscience
(CeNS). This work is dedicated to Prof. Arndt Simon on the
occasion of his 75th birthday.
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
ACKNOWLEDGMENT
EXPERIMENTAL SECTION
The authors would like to thank Dr Stephan Rauschenbach
for assistance in the preparation and analysis of the samples
for MALDI-TOF measurements. The authors also thank Ms
Marie-Luise Schreiber for the elemental analyses, Dr Ko-
numa for the XPS analyses, and Mr Daniel Weber for the
TGA-MS experiments.
Detailed descriptions of the characterization tech-
niques are provided in the supplementary information.
A sample tentatively notated as “melem oligomer” was
prepared as follows. Melamine (1.00 g, Carl Roth, >99%)
was loaded into a loosely covered alumina boat and was
placed into a quartz tube furnace. After complete removal
of air by an extended argon purge, the furnace was heated
under a slow argon flow to 450 °C for 12 h at a 5 °C min-1
ramp, then allowed to cool to ambient temperature at an
unrestricted rate. The resulting light yellow solid
(390 mg) was then thoroughly ground with an agate mor-
tar and pestle prior to characterization and photocatalytic
evaluation. The crystallinity of the product can be im-
proved by further heat treatment at 300 °C for 24 h in
argon.
REFERENCES
(1)
Rand, D. A. J.; Dell, R. M. Hydrogen Energy: Challenges
and Prospects; RSC Publishing: Cambridge, 2008.
(2)
(3)
Fujishima, A.; Honda, K. Nature 1972, 238, 37
Kato, H.; Asakura, K.; Kudo, A. J. Am. Chem. Soc. 2003,
125, 3082
(4)
Lu, D.; Takata, T.; Saito, N.; Inoue, Y.; Domen, K. Na-
ture 2006, 440, 295
(5)
(6)
Maeda, K.; Domen, K. J. Phys. Chem. C 2007, 111, 7851
Wang, X.; Maeda, K.; Thomas, A.; Takanabe, K.; Xin,
For comparison, a sample of melon was prepared in an
identical fashion, except that melamine was heated to
550 °C.
G.; Carlsson, J. M.; Domen, K.; Antonietti, M. Nat. Mater. 2009,
8, 76
(7)
Wang, Y.; Wang, X.; Antonietti, M. Angew. Chem. Int.
Ed. 2012, 51, 68
Separation of the melem oligomer into different frac-
tions was carried out by stirring 400 mg of the ground
solid in DMSO (100 mL) for one week. The resulting sus-
pension was centrifuged at a relative centrifugal force
(RCF) of 10000. The precipitate (notated as “fraction RCF
10000”) was washed with acetone and dried at 100 °C. The
supernatant was centrifuged again at a RCF of 60000 and
the precipitate (fraction RCF 60000) was washed and
dried as in the previous fraction. For the supernatant, the
DMSO was completely evaporated off at 70 °C in vacuo
and the resulting white powder was subjected to the same
acetone wash and drying procedure as described above
for the two other fractions.
(8)
(9)
(10)
Cao, S.; Yu, J. J. Phys. Chem. Lett. 2014, 5, 2101
Liu, A. Y.; Cohen, M. L. Science 1989, 245, 841
Jürgens, B.; Irran, E.; Senker, J.; Kroll, P.; Müller, H.;
Schnick, W. J. Am. Chem. Soc. 2003, 125, 10288
(11)
(12)
(13)
Lotsch, B. V.; Schnick, W. Chem. Mater. 2005, 17, 3976
Lotsch, B. V.; Schnick, W. Chem. Mater. 2006, 18, 1891
Sattler, A.; Schnick, W. Z. Anorg. Allg. Chem. 2006,
632, 238
(14)
Lotsch, B. V.; Döblinger, M.; Sehnert, J.; Seyfarth, L.;
Senker, J.; Oeckler, O.; Schnick, W. Chem. Eur. J. 2007, 13, 4969
(15)
(16)
Lotsch, B. V.; Schnick, W. Chem. Eur. J. 2007, 13, 4956
Seyfarth, L.; Seyfarth, J.; Lotsch, B. V.; Schnick, W.;
Senker, J. Phys. Chem. Chem. Phys. 2010, 12, 2227
(17) Chen, X.; Jun, Y.-S.; Takanabe, K.; Maeda, K.; Domen,
K.; Fu, X.; Antonietti, M.; Wang, X. Chem. Mater. 2009, 21, 4093
(18) Wang, X.; Maeda, K.; Chen, X.; Takanabe, K.; Domen,
K.; Hou, Y.; Fu, X.; Antonietti, M. J. Am. Chem. Soc. 2009, 131,
1680
ASSOCIATED CONTENT
Supporting Information. Details of the experimental section
and theoretical calculations, additional characterizations
(XPS, elemental analyses, surface areas, TGA-MS, SEM and
PL) of the melem oligomer and of the isolated species from
differential centrifugation are provided in the supporting
information. This material is available free of charge via the
Internet at http://pubs.acs.org.
(19)
Li, X.-H.; Zhang, J.; Chen, X.; Fischer, A.; Thomas, A.;
Antonietti, M.; Wang, X. Chem. Mater. 2011, 23, 4344
(20) Niu, P.; Zhang, L.; Liu, G.; Cheng, H.-M. Adv. Func.
Mater. 2012, 22, 2763
(21)
Wu, G.; Thind, S. S.; Wen, J.; Yan, K.; Chen, A. Appl.
Catal. B 2013, 142-143, 590
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 10 of 11
(22) Yang, S.; Gong, Y.; Zhang, J.; Zhan, L.; Ma, L.; Fang, Z.;
(51)
Komatsu, T. J. Mater. Chem. 2001, 11, 799
1
2
3
4
5
6
7
8
Vajtai, R.; Wang, X.; Ajayan, P. M. Adv. Mater. 2013, 25, 2452
(23) Liu, G.; Niu, P.; Sun, C.; Smith, S. C.; Chen, Z.; Lu, G.
Q. M.; Cheng, H.-M. J. Am. Chem. Soc. 2010, 132, 11642
(24) Zhang, J.; Sun, J.; Maeda, K.; Domen, K.; Liu, P.; Anto-
nietti, M.; Fu, X.; Wang, X. Energy Environ. Sci. 2011, 4, 675
(25) Zhang, G.; Zhang, J.; Zhang, M.; Wang, X. J. Mater.
Chem. 2012, 22, 8083
(26) Zhang, J.; Zhang, G.; Chen, X.; Lin, S.; Möhlmann, L.;
Dołęga, G.; Lipner, G.; Antonietti, M.; Blechert, S.; Wang, X.
Angew. Chem. Int. Ed. 2012, 51, 3183
(27) Di, Y.; Wang, X.; Thomas, A.; Antonietti, M. Chem-
CatChem 2010, 2, 834
(28) Ding, Z.; Chen, X.; Antonietti, M.; Wang, X. ChemSus-
Chem 2011, 4, 274
(29) Xiang, Q.; Yu, J.; Jaroniec, M. J. Phys. Chem. C 2011, 115,
7355
(30) Ge, L.; Zuo, F.; Liu, J.; Ma, Q.; Wang, C.; Sun, D.; Bar-
tels, L.; Feng, P. J. Phys. Chem. C 2012, 116, 13708
(52) Tyborski, T.; Merschjann, C.; Orthmann, S.; Yang, F.;
Lux-Steiner, M.-C.; Schedel-Niedrig, T. J. Phys. Condens. Matter
2012, 24, 162201
(53) Thomas, A.; Fischer, A.; Goettmann, F.; Antonietti, M.;
Müller, J.-O.; Schlögl, R.; Carlsson, J. M. J. Mater. Chem. 2008, 18,
4893
(54) Ishii, A.; Habu, K.; Kishi, S.; Ohtsu, H.; Komatsu, T.;
Osaka, K.; Kato, K.; Kimura, S.; Takata, M.; Hasegawa, M.; Shi-
gesato, Y. Photochem. Photobiol. Sci. 2007, 6, 804
(55) Lotsch, B. V. PhD Thesis, Ludwig Maximilians Univer-
sität, München, 2006.
(56) Komatsu, T. Macromol. Chem. Phys. 2001, 202, 19
(57) Xing, J.; Fang, W. Q.; Zhao, H. J.; Yang, H. G. Chem.
Asian J. 2012, 7, 642
(58) Huda, M. N.; Turner, J. A. J. Appl. Phys. 2010, 107,
123703
(59) Meek, G. A.; Baczewski, A. D.; Little, D. J.; Levine, B. G.
J. Phys. Chem. C 2014, 118, 4023
(60) Reshak, A. H.; Khan, S. A.; Auluck, S. RSC Adv. 2014, 4,
6957
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(31)
Kang, H. W.; Lim, S. N.; Song, D.; Park, S. B. Intl. J. Hy-
drogen Energy 2012, 37, 11602
(32) Sun, L.; Zhao, X.; Jia, C.-J.; Zhou, Y.; Cheng, X.; Li, P.;
Liu, L.; Fan, W. J. Mater. Chem. 2012, 22, 23428
(33) Cao, S.-W.; Yuan, Y.-P.; Barber, J.; Loo, S. C. J.; Xue, C.
Appl. Surf. Sci. 2014, 319, 344
(34) Chen, J.; Shen, S.; Guo, P.; Wang, M.; Wu, P.; Wang, X.;
Guo, L. Appl. Catal. B 2014, 152-153, 335
(35) Döblinger, M.; Lotsch, B. V.; Wack, J.; Thun, J.; Senker,
J.; Schnick, W. Chem. Commun. 2009, 1541
(36) Sattler, A.; Pagano, S.; Zeuner, M.; Zurawski, A.;
Gunzelmann, D.; Senker, J.; Müller-Buschbaum, K.; Schnick, W.
Chem. Eur. J. 2009, 15, 13161
(37) Wirnhier, E.; Döblinger, M.; Gunzelmann, D.; Senker,
J.; Lotsch, B. V.; Schnick, W. Chem. Eur. J. 2011, 17, 3213
(38) Makowski, S. J.; Köstler, P.; Schnick, W. Chem. Eur. J.
2012, 18, 3248
(39) Ham, Y.; Maeda, K.; Cha, D.; Takanabe, K.; Domen, K.
Chem. Asian J. 2013, 8, 218
(40) Schwinghammer, K.; Tuffy, B.; Mesch, M. B.; Wirnhier,
E.; Martineau, C.; Taulelle, F.; Schnick, W.; Senker, J.; Lotsch, B.
V. Angew. Chem. Int. Ed. 2013, 52, 2435
(61)
Perdew, J. P.; Burke, K.; Ernzerhof, M. Phys. Rev. Lett.
1996, 77, 3865
(62) Tkatchenko, A.; Scheffler, M. Phys. Rev. Lett. 2009, 102,
073005
(63) Atalla, V.; Yoon, M.; Caruso, F.; Rinke, P.; Scheffler, M.
Phys. Rev. B 2013, 88, 165122
(64) Blum, V.; Gehrke, R.; Hanke, F.; Havu, P.; Havu, V.;
Ren, X.; Reuter, K.; Scheffler, M. Comput. Phys. Commun. 2009,
180, 2175
(65) Ren, X.; Rinke, P.; Blum, V.; Wieferink, J.; Tkatchenko,
A.; Sanfilippo, A.; Reuter, K.; Scheffler, M. New J. Phys. 2012, 14,
053020
(66) Lee, H.-S.; Woo, C.-S.; Youn, B.-K.; Kim, S.-Y.; Oh, S.-
T.; Sung, Y.-E.; Lee, H.-I. Top. Catal. 2005, 35, 255
(67) Datta, A.; Priyam, A.; Bhattacharyya, S. N.; Mukherjea,
K. K.; Saha, A. J. Colloid Interface Sci. 2008, 322, 128
(68) Luo, W.; Li, Z.; Jiang, X.; Yu, T.; Liu, L.; Chen, X.; Ye, J.;
Zou, Z. Phys. Chem. Chem. Phys. 2008, 10, 6717
(69) Zhang, G.; Monllor-Satoca, D.; Choi, W. Catal. Sci.
Technol. 2013, 3, 1790
(41)
Schwarzer, A.; Saplinova, T.; Kroke, E. Coord. Chem.
(70) Ceroni, P.; Balzani, V. In The Exploration of Supramo-
lecular Systems and Nanostructures by Photochemical Tech-
niques; Ceroni, P., Ed.; Springer: Dordrecht, 2012.
Rev. 2013, 257, 2032
(42) Bojdys, Michael J. PhD Thesis, Max-Planck Institut für
Kolloid- und Grenzflächenforschung, Potsdam, 2009.
(43) Maeda, K.; Wang, X.; Nishihara, Y.; Lu, D.; Antonietti,
M.; Domen, K. J. Phys. Chem. C 2009, 113, 4940
(44) Sabio, E. M.; Chamousis, R. L.; Browning, N. D.; Oster-
loh, F. E. J. Phys. Chem. C 2012, 116, 3161
(45) Osterloh, F. E. Chem. Soc. Rev. 2013, 42, 2294
(46) Yang, F.; Kuznietsov, V.; Lublow, M.; Merschjann, C.;
Steigert, A.; Klaer, J.; Thomas, A.; Schedel-Niedrig, T. J. Mater.
Chem. A 2013, 1, 6407
(47) Merschjann, C.; Tyborski, T.; Orthmann, S.; Yang, F.;
Schwarzburg, K.; Lublow, M.; Lux-Steiner, M.-C.; Schedel-
Niedrig, T. Phys. Rev. B 2013, 87, 205204
(48) Schwarzer, A.; Böhme, U.; Kroke, E. Chem. Eur. J. 2012,
18, 12052
(71)
Ye, H.; Scott, R. W. J.; Crooks, R. M. Langmuir 2004,
20, 2915
(72) Esumi, K.; Suzuki, A.; Yamahira, A.; Torigoe, K. Lang-
muir 2000, 16, 2604
(73) Yang, L.; Luo, Y.; Jia, X.; Ji, Y.; You, L.; Zhou, Q.; Wei,
Y. J. Phys. Chem. B 2004, 108, 1176
(74) Zhao, M.; Crooks, R. M. Adv. Mater. 1999, 11, 217
(75) Wen, M.; Mori, K.; Kamegawa, T.; Yamashita, H. Chem.
Commun. 2014, 50, 11645
(76) DuBois, M. R.; DuBois, D. L. Chem. Soc. Rev. 2009, 38,
62
(77) Joo, J. B.; Dillon, R.; Lee, I.; Yin, Y.; Bardeen, C. J.; Zae-
ra, F. Proc. Natl. Acad. Sci. U.S.A. 2014, 111, 7942
(78) Maschmeyer, T.; Che, M. Angew. Chem. Int. Ed. 2010,
49, 1536
(49) Guo, Y.; Chu, S.; Yan, S.; Wang, Y.; Zou, Z. Chem.
Commun. 2010, 46, 7325
(50) Yuan, Y.-P.; Xu, W.-T.; Yin, L.-S.; Cao, S.-W.; Liao, Y.-
S.; Tng, Y.-Q.; Xue, C. Intl. J. Hydrogen Energy 2013, 38, 13159
ACS Paragon Plus Environment

Page 11 of 11
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
11
ACS Paragon Plus Environment