Visible Light-driven Dehydrogenation of Benzylamine under Liberation of H2

Article abstract of DOI:10.1002/cctc.202000329

The visible light-driven transformation of chemical compounds in combination with the liberation of H₂ is highly attractive. Herein, we report on a photocatalyst that allows the acceptorless dehydrogenation of benzylamine. Upon light absorption, free charge carriers are generated and used for the concerted imine formation and liberation of H₂. Our photocatalyst consists of CdS as a light harvesting semiconductor supported on colloidal metal-organic framework crystallites. The decoration with co-catalytic nickel nanoparticles promotes hydrogen evolution and, in addition, stabilizes the CdS component under irradiation.

Full text of DOI:10.1002/cctc.202000329

The European Society Journal for Catalysis
Supported by
Accepted Article
Title: Visible Light-Driven Dehydrogenation of Benzylamine under
Liberation of H2
Authors: Rhett Kempe, Mara Klarner, Sebastian Hammon, Sebastian
Feulner, Stephan Kümmel, and Lothar Kador
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemCatChem 10.1002/cctc.202000329
Link to VoR: https://doi.org/10.1002/cctc.202000329
01/2020

10.1002/cctc.202000329
ChemCatChem
FULL PAPER
Visible Light-Driven Dehydrogenation of Benzylamine under
Liberation of H2
Mara Klarner[a], Sebastian Hammon[b], Sebastian Feulner[c], Stephan Kümmel[b], Lothar Kador[c] and
Rhett Kempe*[a]
[a]
M. Klarner, Prof. Dr. R. Kempe
Anorganische Chemie II
University of Bayreuth
Universitätsstraße 30, 95440 Bayreuth, Germany
E-mail: kempe@uni-bayreuth.de
S. Hammon, Prof. Dr. S. Kümmel
Theoretische Physik IV
[b]
[c]
University of Bayreuth
Universitätsstraße 30, 95447 Bayreuth, Germany
S. Feulner, Prof. Dr. Lothar Kador
Institute of Physics
University of Bayreuth
Universitätsstraße 30, 95447 Bayreuth, Germany
Supporting information for this article is given via a link at the end of the document.
Abstract: The visible light-driven transformation of chemical
compounds in combination with the liberation of H2 is highly
attractive. Herein, we report on a photocatalyst that allows the
The reaction proceeds additive-free, without requiring an
electron acceptor, and involves amine oxidation towards an
imine in combination with liberation of one equivalent of H2. The
acceptorless dehydrogenation of benzylamine. Upon light absorption, synthesis of non-symmetric imines is possible by using a second,
free charge carriers are generated and used for the concerted imine
formation and liberation of H2. Our photocatalyst consists of CdS as
a light harvesting semiconductor supported on colloidal metal-
organic framework crystallites. The decoration with co-catalytic
nickel nanoparticles promotes hydrogen evolution and, in addition,
stabilizes the CdS component under irradiation.
non-benzylic amine that is not dehydrogenated by the
photocatalyst. To accomplish amine splitting, we developed a
novel metal-organic framework-based photocatalyst system
(Ni/CdS@MIL-101).
Introduction
The solar-driven upgrading of organic compounds is a promising
and sustainable way to produce value-added products.[1] The
simultaneous liberation of H2 during such upgrading processes
generates an additional, highly attractive, byproduct. Meanwhile.
a well investigated approach is the photocatalytic acceptorless
dehydrogenation of alcohols (alcohol splitting) to yield carbonyl
compounds (Scheme 1, top).[2] In a similar way, amines may be
transformed into imines and H2 (Scheme 1, bottom). The
photocatalytic amine dehydrogenation described so far requires
sacrificial agents mostly molecular oxygen (aerobic amine
oxidation). The visible light-mediated aerobic amine oxidation is
Scheme 1. State of the art in visible light-mediated photocatalytic
dehydrogenation of alcohols with liberation of H2 (alcohol splitting) and the
acceptorless dehydrogenation of benzylamine (amine splitting) introduced
here.
catalyzed by semiconductor materials including CdS,[3] graphitic
carbon nitride,[4] Nb2O5,[5] WS2,[6] WO3
and bismuth
[7]
oxyhalides.[8] Zhao and coworkers developed a Ni/CdS catalyst
system for the visible light-driven H2 generation from water in
combination with the coupling of amines.[9] Furthermore,
Ni/CdS@MIL-101 is composed of three components. a) The
metal organic framework (MOF) known as MIL-101(Cr) is used
as the photocatalytically inactive support material that acts as a
molecular sponge permitting the semiconductor and catalyst
synthesis and determines the overall size of the photocatalyst
for efficient recycling.[14] b) CdS is the visible light absorbing
semiconductor material. c) Nickel nanoparticles are an efficient
co-catalyst for H2 evolution and enhance the stability of CdS
under photocatalytic conditions since the oxidation of sulfide to
sulfate is prevented. Fluorescence lifetime measurements
heterojunction
photocatalysts,[10]
plasmonic
composite
catalysts[11] and MOF-based photocatalysts[12] are described to
yield homocoupled imines via aerobic amine oxidation. The
transformation of two different amines into a non-symmetric
imine is more challenging, requiring high temperatures and
oxygen pressure.[13]
Herein, we report on the photocatalytic and visible light-driven
acceptorless dehydrogenation of benzylamine (amine splitting).
1
This article is protected by copyright. All rights reserved.

10.1002/cctc.202000329
ChemCatChem
FULL PAPER
indicate a faster electron transfer from the conduction band of
CdS to the Ni particles in comparison to nanoparticles of noble
metals such as Pd or Pt. Thereby, the probability of charge
recombination within the semiconductor is reduced. In addition,
DFT calculations suggest a stronger substrate binding of the
amine to the Ni than to Pd nanoparticles.
temperature by applying static vacuum and subsequently
reducing to Ni0 at 90 °C and 10 bar H2.[18] We modified
CdS@MIL-101 with different amounts of nickel, 1 wt.%, 5 wt.%,
and
10 wt.%.
(η3-Allyl)(η5-cyclopentadienyl)-palladium(II)
[Pd(C3H5)(C5H5)] was used to generate 5 wt.% Pd0 nanoparticles
by sublimation at 32 °C in dynamic vacuum and reduction at
70 °C and 70 bar H2.[19] The modification with 5 wt.% Pt0
proceeds via the gas phase loading of trimethyl-
(methylcyclopentadienyl)-palladium(IV) [Me3Pt(CH3-C5H4)] into
the pores of CdS@MIL-101 at 37 °C in static vacuum, followed
by a reduction step at 80 °C and 50 bar H2.[20] TEM analysis
Results and Discussion
The photocatalytically inactive MOF MIL-101 with its
(Scheme 2; Supporting Information, Figure S3) indicated
a
characteristic
pore
structure
(Supporting
Information,
Figure S2a) was shown to have a surface area of 2700 m2/g.[15]
The MIL-101 crystallite size, between 100 and 300 nm, is
desirable for the generation of active and reusable core-shell
catalysts. The MIL-101 crystallites are small enough to exhibit a
proper outer surface area for the modification with photoactive
components.[16] This is combined with the easy separation of the
colloidal photocatalyst, since the crystallites are, on the other
hand, large enough to ensure recyclability by centrifugation. The
highly porous and size optimized support material can be
selectively loaded with precursor molecules permitting the
catalyst synthesis. The outer surface of single MIL-101
crystallites is decorated with visible light-absorbing CdS particles
by a simple solvothermal route. Cadmium acetate is infiltrated
into MIL-101 in dimethyl sulfoxide as sulfur source and
crystallizes as cubic CdS under solvothermal conditions at
180 °C during 10 h to yield CdS@MIL-101.[17] Transmission
electron microscopy revealed the arrangement of CdS particles
with an average size between 20 and 30 nm on the structure
determining core MIL-101 (Scheme 2; Supporting Information,
Figure S4). The gas phase infiltration of volatile metal-organic
precursors into the porous CdS@MIL-101 composite material
allows for the generation of metal nanoparticles, denoted as
M/CdS@MIL-101. For the modification with nickel nanoparticles,
bis(cyclopentadienyl)nickel(II) (Ni(C5H5)2) was infiltrated at room
homogeneous distribution of metallic nanoparticles smaller than
2 nm which are located on the CdS particles, forming an
interface with the visible light-absorbing semiconducting material.
The metal content of M5CdS@MIL-101 catalysts (theoretically
5 wt.%) was examined by inductively coupled plasma optical
emission spectroscopy (ICP-OES) and found to be 5.1 wt.% Ni,
5.2 wt.% Pd, and 4.9 wt.% Pt, respectively (Supporting
Information, Table S1). The calculated weight percentage of
CdS is 45 wt.% for each catalyst material. The final
M/CdS@MIL-101 core-shell composite showed the original
octahedral shape determined by the MIL-101. The metal
nanoparticles are assumed to serve as an electron reservoir,
since a directed electron transfer from the conduction band of
the semiconductor CdS across the semiconductor/nanoparticle
interface has been postulated.[21] Scheme 2 presents the general
photocatalytic concept of M/CdS@MIL-101: An electron-hole
pair is generated upon visible light excitation of CdS. The
electron is transferred from the conduction band (CB) of CdS to
a metal particle by the built-in electric field, thereby reducing the
possibility of charge recombination. Subsequently, spatially
separated redox reactions can be catalyzed. Electrons reduce
protons for molecular hydrogen generation at the co-catalytic
nanoparticles and, simultaneously, holes in the CdS valence
band are consumed by oxidizing substrates.
Scheme 2. Synthesis of the photocatalyst M/CdS@MIL-101. The decoration with CdS semiconductor particles proceeds via a solvothermal modification. The
metal precursor (Ni: Ni(C5H5)2; Pd: Pd(C3H5)(C5H5); Pt: Me3Pt(CH3-C5H4)) is infiltrated into MIL-101 (green) by gas phase deposition and reduced by hydrogen
treatment. In the semiconductor CdS (yellow) an electron is promoted from the valance band (VB) to the conduction band (CB) by the absorption of visible light.
After the directed electron transfer toward the metal particles (grey), spatially separated redox reactions can be catalysed.
For
determining
the
absorbance
characteristics
of
compared to bare MIL-101 absorbing around 600 and 450 nm,
an increased absorption was observed for the dark yellow
CdS@MIL-101 (Supporting Information, Figure S4a). Upon
Ni5CdS@MIL-101, we performed diffuse-reflectance ultraviolet-
visible spectroscopy (DRS) between 350 nm and 700 nm. As
2
This article is protected by copyright. All rights reserved.

10.1002/cctc.202000329
ChemCatChem
FULL PAPER
generating the photoactive compound, the characteristic
absorbance of MIL-101 is extinguished. Photons of wavelengths
headspace via GC-TCD (Figure 1b). Switching the light on and
off verified the release of H2 only under visible light illumination
(Supporting Information, Figure S10). The reusability of
Ni5CdS@MIL-101 photocatalyst was investigated by the
acceptorless dehydrogenation of benzylamine. We performed
five consecutive runs without a remarkable loss of activity
(Supporting Information, Figure S11).
smaller
than
510 nm
are
predominantly
absorbed,
corresponding to the band gap of the semiconductor CdS. With
the Munk-Kubelka equation, the optical band gap of CdS was
determined as 2.44 eV, in agreement with the literature value for
cubic CdS (Supporting Information, Figure S4b). The
photocatalyst Ni5CdS@MIL-101 exhibited nearly constant
absorbance throughout the visible range. We examined the
surface area and the pore size distribution of the different
compounds during synthesis of the photocatalyst by nitrogen
physisorption measurements. The modification of porous
MIL-101 with CdS results in a significant decrease of the initial
surface area by 53 %. Further modification with Ni nanoparticles
leads only to a minor additional decrease of the surface area to
1040 m2/g. X-ray powder diffractometry (PXRD) reveals the
exclusive formation of cubic CdS particles (Supporting
Information, Figure S8a) showing reflections at 26.5°, 43.9°, and
51.9° (2θ). The reflections from 2° to 20° (2θ) are assigned to
the preserved MIL-101 core. X-ray photoelectron spectroscopy
(XPS) indicated the formation of metallic Ni nanoparticles
(Supporting Information, Figure S5). The major peak within the
S2p region at a binding energy of 161.5 eV is assigned to
metallic sulfide (CdS). Traces of metal sulfate were identified at
a binding energy of 169 eV; they are ascribed to surface
oxidation of CdS due to the handling in air. The XPS survey
shows characteristic elemental signals for Cr, O, and C of the
MIL-101 support material and additional signals for Cd, S, and
Ni, which form the photoactive shell. The homogeneous
distribution of Ni particles on the CdS@MIL-101 material was
confirmed by energy dispersive X-ray (EDX) elemental mapping
(Supporting Information, Figure S6). Furthermore, we performed
Table 1. Photocatalytic acceptorless dehydrogenation of benzylamine.[a]
Catalyst
Yield [%] [b]
Yield H2 [%] [c]
1
2
3
4
Ni5CdS@MIL-101
CdS@MIL-101
Without hv
99
62
0
98
59
0
Without catalyst
0
0
[a] 1 mmol benzylamine, 1.5 mL MeCN, 10 h, 5 mg catalyst, rt, Ar, 470 nm
blue LED (50 W). [b] Determined by GC using n-dodecane as an internal
standard. [c] Quantified by GC-TDC using methane as an internal standard.
We also developed an approach to access non-symmetric
imines using a second, non-benzylic amine which is not
dehydrogenated by the photocatalyst under the given conditions.
Figure 1c describes the cross-coupling of benzylamine and
cyclohexylamine via the dehydrogenation of benzylamine and
subsequent reversible transimination to yield the heterocoupled
N-cyclohexyl-1-phenylmethanimine
(Figure 1c,
d).
This
exchange is a non-photocatalytic equilibrium reaction, in which
one equivalent of benzylamine is liberated and fed back into the
dehydrogenation cycle. Under visible light illumination, the
equilibrium is shifted towards the non-symmetric imine,
guaranteeing its predominant formation (Supporting Information,
Table S5).
Fourier-transform
infrared
spectroscopy
(FT-IR)
of
Ni5CdS@MIL-101 and the intermediate materials during
synthesis (Supporting Information, Figure S7a). Due to the
surface modification with the light harvesting CdS, MIL-101
signals are reduced between 700 and 2000 cm-1. A total weight
loss of 42 % was observed for the Ni5CdS@MIL-101
photocatalyst as compared to 72 % for MIL-101 in
thermogravimetric analysis (TGA) (Supporting Information,
Figure S7b). By covering the MOF core with the photocatalytic
active shell, the temperature stability of the system is enhanced.
Next, we performed the photocatalytic acceptorless
dehydrogenation of benzylamine via liberation of molecular H2.
Thereby, the intermediate aldimine reacts with a second
equivalent of benzylamine in a condensation reaction to yield the
homocoupled
product
N-benzyl-1-phenylmethanimine
(Figure 1a). The elimination of gaseous ammonia was verified
by gas chromatography (thermal conductive detector, GC-TCD).
The flat band potential of CdS is sufficiently positive (VB 1.7 V
vs NHE)[22] to thermodynamically allow for the oxidation of
benzylamine (standard reduction potential 0.9 V vs NHE).[23] The
photocatalytic amine dehydrogenation by Ni5CdS@MIL-101 was
performed under an inert-gas atmosphere at room temperature
without the use of any additives or acceptor molecules. In the
presence of co-catalytic Ni nanoparticles, the dehydrogenation
of benzylamine proceeds with quantitative yield, whereas the
neat CdS@MIL-101 leads to the formation of the tertiary amine
Figure 1. (a) Acceptorless dehydrogenation of benzylamine. (b) H2 liberation
from benzylamine (1 mmol) is observed under visible light illumination and
quantified by GC-TCD. (c) Non-symmetric imine formation by cross-coupling
benzylamine and cyclohexylamine. (d) Kinetic study (the colour code refers to
(c)).
as
a byproduct (Table 1, Entry 1, 2). We confirmed the
We synthesized NixCdS@MIL-101 catalyst systems with
different Ni contents (x equal to 1, 5, 10 wt.% Ni) to investigate
equimolar liberation of molecular H2 by analyzing the reaction
3
This article is protected by copyright. All rights reserved.

10.1002/cctc.202000329
ChemCatChem
FULL PAPER
the influence of the co-catalytic nanoparticles on the
dehydrogenative cross-coupling. To convert 1 mmol of
substrates under optimized reaction conditions, 5 mg
Ni5CdS@MIL-101 were suspended in 1.5 ml ethanol and
illumination no conversion of the substrates takes place (Table 2,
Entries 10, 11). Interestingly, a significantly reduced yield is
observed in ambient air (Table 2, Entry 12).
illuminated with
Information, Figure S1, Table S3 and S4).
a
50 W blue LED (470 nm) (Supporting
Next, we studied the stability of the semiconducting CdS
component of Ni5CdS@MIL-101 under photocatalytic conditions.
The corrosion of CdS is a frequently discussed phenomenon
and a main challenge in the use of such photocatalysts.[24]
Recently, the benzylamine mediated oxidation of sulfide was
reported by DiMeglio and co-workers during the non-aqueous
dehydrogenation of benzylamine with O2 as a sacrificial agent.[25]
Table 2. Photocatalytic dehydrogenation of benzylamine and cross-coupling
with cyclohexylamine. Comparison of different Ni contents, different metal
nanoparticles and different reaction conditions.[a]
The
standard
reduction
potential
of
benzylamine
thermodynamically allows for the oxidation of CdS via the
formation of the amine-radical cation. In a typical photocatalytic
set-up, the photocatalysts Ni5CdS@MIL-101 and CdS@MIL-101
were illuminated in the presence of benzylamine and ethanol as
solvent in an inert gas atmosphere.
Catalyst
Metal
Yield 1
[%] [c]
Yield 2
[%] [c, d]
[wt %] [b]
1
2
3
4
Ni1CdS@MIL-101
Ni5CdS@MIL-101
Ni10CdS@MIL-101
Ni@CdS[f]
1
5.1[e]
10
45
74
17
18
12
7
37
4
5
5
6
CdS@MIL-101
Ni@MIL-101
MIL-101
-
42
0
17
0
5
7
-
0
0
8
Pd5CdS@MIL-101
Pt5CdS@MIL-101
Without hν
5.2[e]
4.9[e]
5.1[e]
-
8
34
51
0
9
7
10
11
12
0
Without catalyst
Ambient air
0
0
5.1[e]
11
23
[a] 1 mmol benzylamine, 1 mmol cyclohexylamine, 1.5 mL EtOH,
5 mg catalyst, 470 nm blue LED (50 W), 20 h, rt, Ar. [b] Co-catalytic metal
nanoparticles (Ni, Pd, Pt). [c] Determined by GC using n-dodecane as an
internal standard. [d] Referred to 0.5 mmol. [e] Determined by ICP-OES
analysis. [f] Commercial CdS (Alfa Aesar).
CdS@MIL-101 loaded with 5 wt.% Ni showed the best catalytic
performance yielding 74 % heterocoupled product 1 and 7 %
homocoupled product 2 (Table 2, Entry 2). With increasing Ni
content, the nanoparticles covering the CdS crystals reduce the
light absorbance and, thereby, the generation of free charge
carriers resulting in 17 % of 1 and 37 % of 2 (Table 2, Entry 3).
With 1 wt.% Ni, the photocatalytic activity is similar to neat
CdS@MIL-101 (Table 2, Entry 1). The MIL-101 support is
Figure 2. XPS studies of Ni5CdS@MIL-101 and CdS@MIL-101 demonstrating
the enhanced photocatalytic stability of CdS in the presence of Ni
nanoparticles. Ni5CdS@MIL-101 does not show significant degradation of the
surface of CdS [(3), (4)], whereas the CdS in CdS@MIL-101 is oxidized under
photocatalytic conditions [(1), (2)]. For details see text.
The pre-catalytic CdS@MIL-101 exhibits one major sulfur signal
2
at P3/2 161.5 eV (86 %) corresponding to metal sulfide and a
2
essential, since Ni-modified commercial CdS shows
a
small signal for metal sulfate at P3/2 168.7 eV (14 %) (Figure 2,
significantly lower activity (Table 2, Entry 4). Also, the
unmodified CdS@MIL-101 shows a lower catalytic activity
(Table 2, Entry 5) underlining the importance of Ni nanoparticles
which was already observed for the formation of the
homocoupled product N-benzyl-1-phenylmethanimine 2. We
found no photocatalytic activity without the light-harvesting CdS
component in both Ni@MIL-101 and the neat support MIL-101
(Table 2, Entries 6, 7). The modification with the noble metals
Pd and Pt as co-catalysts leads mainly to the formation of the
homo-coupled product 2 (Table 2, Entries 8, 9). This fact
highlights the superior performance of co-catalytic non-noble Ni
nanoparticles. The composite material Ni5CdS@MIL-101
catalyzes the cross-coupling of amines as shown by several
control experiments: Without the catalyst and without visible light
(1) and (2)). Post-catalytic PXRD analysis indicates the
formation of oxidized sulfur species due to the presence of
several reflections besides those of cubic CdS (Supporting
Information, Figure S8). XPS analysis in the S2p region confirms
an increase in surface sulfate to 51 %. The modification with Ni
nanoparticles improves the stability of the CdS component since
no corrosion effects are observed in PXRD. The ratio of CdS
(94 %) to CdSO4 (6 %) is constant for pre- and post-catalytic
Ni5CdS@MIL-101 (Figure 2, (3) and (4)). We performed several
control experiments without either visible light illumination, or the
presence of benzylamine or solvent and, in addition, in dry
acetonitrile. PXRD analysis of post-catalytic Ni5CdS@MIL-101
materials showed no evidence for CdS corrosion in all these
cases, which supports the idea of benzylamine-mediated sulfide
4
This article is protected by copyright. All rights reserved.

10.1002/cctc.202000329
ChemCatChem
FULL PAPER
oxidation in an oxygen-containing solvent (Supporting
Information, Figure S8). The observed stability of CdS is
decisive for the catalytic recyclability of Ni5CdS@MIL-101.
Pt5CdS@MIL-101 photocatalysts revealed that the supposed
stabilizing effect of metal nanoparticles on the CdS
semiconductor occurs also with these noble metals.
Further
experiments
on
Pd5CdS@MIL-101
and
Figure 3. Fluorescence lifetime studies of MIL-101 supported catalysts confirming the directed electron transfer from excited CdS to Ni nanoparticles. Polar-plot
representations of the data for NixCdS@MIL-101 and M5CdS@MIL-101: (a) neat CdS, (b) 1 wt.% Ni, (c) 5 wt.% Ni, (d) 10 wt.% Ni, (e) 5 wt.% Pd and (f) 5 wt.% Pt.
The fluorescence lifetime of CdS decreases with increasing Ni content (right end point of the straight line on the semi-circle). It also depends on the metal
nanoparticle (Ni, Pd, Pt; 5 wt.%) indicating an influence of the latter on the electron transfer efficiency. Data points correspond to diffraction-limited spots on the
sample within an area of size 10 × 10 μm, the red dot indicating their centre of mass. The fluorescence intensity of each spot is color-coded. The modulation
frequency of the laser was 110 MHz. For further details see text.
The directed electron transfer from photo-excited CdS across
the semiconductor-metal interface to Ni nanoparticles was
investigated with fluorescence lifetime imaging microscopy in
the frequency domain (FD-FLIM). A custom-built apparatus was
used with a semiconductor laser at 487 nm and modulation
frequencies between 110 and 155 MHz (adapted from Ref. [26]).
The combination of the semiconductor material CdS with
metallic particles opens up an additional decay channel of
charge carriers in the conduction band of CdS. By absorbing
visible light, an electron-hole pair is generated in the CdS
component which can recombine after an intrinsic lifetime by
emitting a photon. FD-FLIM data are conveniently plotted in the
so-called polar-plot representation, where the quadrature
component of the normalized fluorescence signal is plotted
versus the component which is in phase with the excitation.
the MIL-101. The position of the data points on the straight line
represents the relative contributions of the two lifetime
components to the luminescence signal. With increasing Ni
content, the CdS lifetime gradually shortens from 115 ps
(CdS@MIL-101)
to
97 ps
(Ni1CdS@MIL-101),
67 ps
(Ni5CdS@MIL-101) and 11 ps (Ni10CdS@MIL-101), respectively,
verifying the directed electron transfer (Figure 3a-d). The relative
variation of these lifetimes is reliable; the absolute numbers are
subject to an uncertainty of about a factor two to three, however,
given the comparatively low modulation frequencies. In
accordance with the observed photocatalytic activity, the
efficiency of the charge transfer increases with the Ni content.
The charge carrier separation over two different catalyst
components leads to the enhanced photocatalytic activity of
Ni/CdS@MIL-101 in redox reactions. Comparing the 3d metal Ni
with the noble metals Pd and Pt, the lifetime of neat CdS is most
strongly affected by neighboring Ni particles, indicating faster
electron transfer to Ni particles than to either Pd or Pt (Figure 3c,
e, f). Comparing catalyst systems with 5 wt.% of the three
metals, the lifetime of neat CdS is reduced from 115 ps to
108 ps (Pd5CdS@MIL-101), 92 ps (Pt5CdS@MIL-101) and
67 ps (Ni5CdS@MIL-101). This result underlines the superior
performance of Ni particles in the discussed photocatalytic
reactions.
Single-exponential decays correspond to data on
a
characteristic semi-circle with radius 0.5 around the point (0.5; 0)
in this plot. Data points within the semi-circle, on the other hand,
represent fluorescence decays featuring more than one lifetime
component and can be linearly decomposed in the complex
plane.[27] The photoluminescence lifetime of neat CdS supported
on MIL-101 is determined by drawing a straight line through the
data points in the polar plot and extrapolating it to the
intersections with the semicircle (Figure 3a). Two lifetime
components are extracted which are independent of the
modulation frequency (Supporting Information, Figure S9). The
shorter lifetime is interpreted as the luminescence lifetime of
CdS, whereas the longer one is ascribed to the luminescence of
In addition to a faster electron transfer from CdS to Ni in
comparison to Pd or Pt particles, another reason for the superior
photocatalytic performance of Ni/CdS@MIL-101 may be the
5
This article is protected by copyright. All rights reserved.

10.1002/cctc.202000329
ChemCatChem
FULL PAPER
interaction or binding of benzylamine with the metal
nanoparticles. As a straightforward test we computed the
binding energy of benzylamine to the metal nanoparticles. While
this type of calculation can be performed using density functional
theory (DFT) in principle, it requires careful considerations in
practice: The structure of metal particles in general is difficult to
determine due to the existence of many isomers of similar
energy,[28] yet it may influence binding energies. Furthermore,
common exchange-correlation approximations may reach their
accuracy limits for d-electron systems.[29] For these reasons, we
focused on 13-atom clusters of Ni and Pd as two representative
which is not dehydrogenated by the photocatalyst under the
given conditions. The MOF-supported colloidal photocatalyst is
composed of the visible light-absorbing semiconductor CdS and
Ni nanoparticles. We confirmed the directed electron transfer
from the conduction band of CdS to metallic nanoparticles (Ni,
Pd and Pt) and observed a faster electron transfer to the
abundant non-noble metal Ni. The co-catalytic Ni (as well as Pd
and Pt) also enhances the stability of CdS against photo-
corrosion in the presence of benzylamine. In addition, DFT
calculations reveal a stronger binding of benzylamine to Ni than
to Pd clusters.
test cases. Both form similar and stable cluster geometries, such
30]
as the icosahedron.[28,
Our calculations suggest that the
icosahedral structure is arguably the most relevant one for Ni13
and Pd13 at room temperature in solution (Supporting Information,
Theoretical Procedure). Therefore, we investigated the binding
energy of one benzylamine molecule to one 13-atom metal
cluster (Ni13 or Pd13) by computing the electronic and geometric
structure using TURBOMOLE[31] (Supporting Information,
Theoretical Procedure). We checked for possible limitations of
the predictive power of the DFT calculations due to the presence
of d-electrons by computing the binding energies with different
exchange-correlation functionals.[32] In particular, the importance
of localization and self-interaction was investigated by using
functionals with different amounts of exact exchange.
Acknowledgements
This work was supported by the German Research Foundation
(DFG SFB 840, B1). The authors also acknowledge the support
of the Bavarian Polymer Institute (University of Bayreuth,
KeyLab Electron and Optical Microscopy, KeyLab Theory and
Simulation) and the initiative Solar Technologies go Hybrid. We
thank Florian Puchtler (XRD) and Prof. Dr. Sven Hüttner (XPS).
Furthermore, we thank the Elite Network Bavaria and the DAAD
(Bayreuth-Melbourne Colloid/Polymer Network) for financial and
other support.
Keywords: hydrogen generation • MOF • nickel • amines •
photocatalysis
[1]
[2]
G. Han, Y.-H. Jin, R. A. Burgess, N. E. Dickenson, X.-M. Cao, Y. Sun, J.
Am. Chem. Soc. 2017, 139, 15584-15587.
a) T. P. A. Ruberu, N. C. Nelson, I. I. Slowing, J. Vela, J. Phys. Chem.
Lett. 2012, 3, 2798-2802; b) Z. Chai, T.-T. Zeng, Q. Li, L.-Q. Lu, W.-J.
Xiao, D. Xu, J. Am. Chem. Soc. 2016, 138, 10128-10131; c) D. Jiang, X.
Chen, Z. Zhang, L. Zhang, Y. Wang, Z. Sun, R. M. Irfan, P. Du, J. Catal.
2018, 357, 147-153; d) S. Meng, X. Ye, J. Zhang, X. Fu, S. Chen, J.
Catal. 2018, 367, 159-170; e) D. Tilgner, M. Klarner, S. Hammon, M.
Friedrich, A. Verch, N. de Jonge, S. Kümmel, R. Kempe, Aust. J. Chem.
2019, 72, 842-847; f) T. Mitkina, C. Stanglmair, W. Setzer, M. Gruber,
H. Kisch, B. Konig, Org. Biomol. Chem. 2012, 10, 3556-3561; g) K.
Imamura, H. Tsukahara, K. Hamamichi, N. Seto, K. Hasimoto, H.
Kominami, Appl. Catal. A General 2013, 450, 28-33.
Figure 4. Lowest energy geometries obtained from Born-Oppenheimer DFT
molecular dynamics simulations in which a benzylamine molecule binds to a
13-atom Pd icosahedral particle (a) and a 13-atom Ni icosahedral particle (b).
[3]
[4]
a) W. Zhao, C. Liu, L. Cao, X. Yin, H. Xu, B. Zhang, RSC Adv. 2013, 3,
22944-22948; b) R. Wu, S. Wang, Y. Zhou, J. Long, F. Dong, W. Zhang,
ACS Appl. Nano Mater. 2019, 2, 6818-6827.
We found the general trend that benzylamine binds stronger to
Ni13 than to Pd13 by several hundred meV (Supporting
Information, Table S8). The lowest energy geometries are
depicted in Figure 4. This result correlates with a previous study
on the binding energy of benzyl alcohol to Ni13 and Pd13.[17a]
Furthermore, we calculated the binding energy for the
dehydrogenated intermediate (C7H7N). This intermediate also
shows a higher binding energy to Ni13 than to Pd13, yet overall
slightly lower than benzylamine (Supporting Information,
Theoretical Procedure).
a) F. Su, S. C. Mathew, L. Möhlmann, M. Antonietti, X. Wang, S.
Blechert, Angew. Chem. Int. Ed. 2011, 50, 657-660; b) A. Kumar, P.
Kumar, C. Joshi, S. Ponnada, A. K. Pathak, A. Ali, B. Sreedhar, S. L.
Jain, Green Chem. 2016, 18, 2514-2521.
[5]
[6]
S. Furukawa, Y. Ohno, T. Shishido, K. Teramura, T. Tanaka, ACS
Catal. 2011, 1, 1150-1153.
a) F. Raza, J. H. Park, H.-R. Lee, H.-I. Kim, S.-J. Jeon, J.-H. Kim, ACS
Catal. 2016, 6, 2754-2759; b) H. Liang, B.-Q. Zhang, J.-M. Song,
ChemCatChem 2019, 11, 6288-6294.
[7]
N. Zhang, X. Li, H. Ye, S. Chen, H. Ju, D. Liu, Y. Lin, W. Ye, C. Wang,
Q. Xu, J. Zhu, L. Song, J. Jiang, Y. Xiong, J. Am. Chem. Soc. 2016,
138, 8928-8935.
Conclusion
In conclusion, we have reported the first example of
[8]
[9]
A. Han, H. Zhang G.-K. Chuah, S. Jaenicke, Appl. Catal. B Environ.
2017, 219, 269-275.
photocatalytic.
visible
light-driven
acceptorless
amine
dehydrogenation (amine splitting). The noble metal-free
photocatalyst Ni/CdS@MIL-101 oxidizes benzylamine with the
liberation of one equivalent of H2 most efficiently. The synthesis
of non-symmetric imines is possible by adding a second amine
W. Yu, D. Zhang, X. Guo, C. Song, Z. Zhao, Catal. Sci Technol. 2018,
8, 5148-5154.
[10] a) S. Samanta, S. Khilari, D. Pradhan, R. Srivastava, ACS Sustainable
Chem. Eng. 2017, 5, 2562-2577; b) K. Zhang, H. Su, H. Wang, J.
6
This article is protected by copyright. All rights reserved.

10.1002/cctc.202000329
ChemCatChem
FULL PAPER
Zhang, S. Zhao, W. Lei, X. Wei, X. Li, J. Chen, Adv. Sci. 2018, 5,
1800062.
[30] a) G. L. Gutsev, C. W. Weatherford, K. G. Belay, B. R. Ramachandran,
P. Jena, J. Chem. Phys. 2013, 138, 164303; b) J. P. Chou, C. R. Hsing,
C. M. Wei, C. Cheng, C. M. Chang, J. Phys.: Condens. Matter 2013, 25,
125305; c) A. M. Köster, P. Calaminici, E. Orgaz, D. R. Roy, J. U.
Reveles, S. N. Khanna, J. Am. Chem. Soc. 2011, 133, 12192-12196; d)
B. Fresch, H.-G. Boyen, F. Remacle, Nanoscale 2012, 4, 4138–4147.
[11] a) S. Naya, K. Kimura, H. Tada, ACS Catal. 2013, 3, 10-13; b) S.
Sarina, H. Zhu, E. Jaatinen, Q. Xiao, H. Liu, J. Jia, C. Chen, J. Zhao, J.
Am. Chem. Soc. 2013, 135, 5793-5801; c) H. Chen, C. Liu, M. Wang, C.
Zhang, N. Luo, Y. Wang, H. Abroshan, G. Li, F. Wang, ACS Catal.
2017, 7, 3632-3638.
[31] TURBMOLE V7.3 2018, (University of Karlsruhe and Forschungs-
zentrum Karlsruhe GmbH, Karlsruhe, Germany)
[12] a) D. Sun, L. Ye, Z. Li, Appl. Catal. B Environ. 2015, 164, 428-432; b) H.
Liu, C. Xu, D. Li, H.-L. Jiang, Angew. Chem. Int. Ed. 2018, 57, 5379-
5383.
[32] a) J. P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett. 1996, 77,
3865-3868; b) J. P. Perdew, K. Burke, M. Ernzerhof, Phys. Rev. Lett.
1997, 78, 1396; c) J. P. Perdew, M. Ernzerhof, J. Chem. Phys. 1996,
105, 9982; d) C. Adamo, V. Barone J. Chem. Phys. 1999, 110, 6158; e)
A. D. Becke, J. Chem. Phys. 1993, 98, 5648-5652; f) P.J. Stephens, F.
J. Devlin, C. F. Chabalowski, M. J. Frisch, J. Phys. Chem. 1994, 98,
11623-11627; g) J. Tao, J. P. Perdew, V. N. Staroverov, G. E. Scuseria,
Phys. Rev. Lett. 2003, 91, 146401; h) V. N. Staroverov, G. E. Scuseria,
J. Tao, J. P. Perdew, J. Chem. Phys. 2003, 119, 12129–12137; i) V. N.
Staroverov, G. E. Scuseria, J. Tao, J. P. Perdew, J. Chem. Phys. 2004,
121, 11507–11507; j) J. Sun; A. Ruzsinszky; J. P. Perdew. , Phys. Rev.
Lett. 2015, 115, 036402.
[13] a) A. Grirrane, A. Corma, H. Garcia, J. Catal. 2009, 264, 138-144; b) L.
Liu, S. Zhang, X. Fu, C.-H. Yan, Chem. Commun. 2011, 47, 10148-
10150; c) X. Qiu, C. Len, R. Luque, Y. Li, ChemSusChem 2014, 7,
1684-1688; d) M. Largeron, M.-B. Fleury, Chem. Eur. J. 2015, 21,
3815-3820.
[14] a) D. Tilgner, M. Friedrich, J. Hermannsdörfer, R. Kempe,
ChemCatChem, 2015, 7, 3916-3922; b) D. Tilgner, R. Kempe, Chem.
Eur J. 2017, 23, 3184-3190; c) D. Tilgner, M. Friedrich, A. Verch, N. de
Jonge, R. Kempe, ChemPhotoChem 2018, 2, 349-352; d) D. Tilgner, M.
Klarner, S. Hammon, M. Friedrich, A. Verch, N. de Jonge, S. Kümmel,
R. Kempe, Aust. J. Chem. 2019, 72, 842-847.
[15] G. Ferey, C. Mellot-Draznieks, C. Serre, F. Millange, J. Dutour, S.
Surble, I. Margiolaki, Science 2005, 309, 2040-2042.
[16] J. Hermannsdörfer, M. Friedrich, R. Kempe, Chem. Eur. J. 2013, 19,
13652-13657.
[17] a) D. Tilgner, M. Klarner, S. Hammon, M. Friedrich, A. Verch, N. de
Jonge, S. Kümmel, R. Kempe, Aust. J. Chem. 2019, 72, 842-847; b) Y.
Wang, Y. Zhang, Z. Jiang, G, Jiang, Z. Zhao, Q. Wu, Y. Liu, Q. Xu, A.
Duan, C. Xu, Appl. Catal. B Environ. 2016, 185, 307-314; c) J. He, Z.
Yan, J. Wang, J. Xie, L. Jiang, Y. Shi, F. Yuan, F. Yu, Y. Sun, Chem.
Commun. 2013, 49, 6761-6763. d) D. Ding, Z. Jiang, J. Jin, J. Li, D. Ji,
Y. Zhang, L. Zan, J. Catal. 2019, 375, 21-31.
[18] J. Hermannsdörfer, M. Friedrich, N. Miyajima, R. Albuquerque, S.
Kümmel, R. Kempe, Angew. Chem. Int. Ed. 2012, 51, 11473-11477.
[19] J. Hermannsdörfer, R. Kempe, Chem. Eur. J. 2011, 17, 8071-8077.
[20] S. Proch, J. Hermannsdörfer, R. Kempe, C. Kern, A. Jess, L. Seyfarth,
J. Senker, Chem. Eur. J. 2008, 14, 8204-8212.
[21] a) T. Simon, N. Bouchonville, M. J. Berr, A. Vaneski, A. Adrovic, D.
Volbers, R. Wyrwich, M. Döblinger, A. S. Susha, A. L. Rogach, F.
Jäckel, J. K. Stolarczyk, J. Feldmann, Nat. Mater. 2014, 13, 1013-1018;
b) Y. Xu, R. Xu, Appl. Surf. Sci. 2015, 351, 779-793; c) S. Cao, C.-J-
Wang, X.-J. Lv, Y. Chen, W.-F. Fu, Appl. Catal. B. Environ 2015, 162,
381-391; d) Z. Chai, T.-T. Zeng, Q. Li, L.-Q. Lu, W.-J. Xiao, D. Xu, J.
Am. Chem. Soc. 2016, 138, 10128-10131; e) G. Han, Y.-H. Jin, R. A.
Burgess, N. E. Dickenson, X.-M. Cao, Y. Sun, J. Am. Chem. Soc. 2017,
139, 15584-15587.
[22] A. J. Bard, M.S. Wrigthon, J. Electrochem. Sci. Technol. 1977, 124,
1706-1710.
[23] K. Ohkubo, T. Nanjo, S. Fukuzumi, Bull. Chem. Soc. Jpn. 2006, 79,
1489-1500.
[24] a) A. Henglein, Ber. Bunsenges. Phys. Chem. 1982, 86, 301-305; b) D.
Meissner, C. Brenndorf, R. Memming, Appl. Surf. Sci. 1987, 27, 423-
436; c) Y. H. Hsieh, C. P. Huang, Colloids Surf. 1991, 53, 275-295.
[25] J. L. DiMeglio, B. M. Bartlett, Chem. Mater. 2017, 29, 7579-7586.
[26] S. Zahner, L. Kador, K. R. Allakhverdiev, E. Yu. Salaev, M. F.
Huseyinoğlu, J. Appl. Phys. 2014, 115, 043504.
[27]
a) G. I. Redford, R. M. Clegg, J. Fluoresc. 2005, 15, 805-815; b) M. A.
Digman, V. R. Caiolfa, M. Zamai, E. Gratton, Biophys. J. 2008, 94, L14-
L16.
[28] L. Leppert, R. Kempe, S. Kümmel, Phys. Chem. Chem. Phys. 2015, 17,
26140-26148.
[29]
T. Schmidt, S. Kümmel, Computation 2016, 4, 33-1 – 33-15.
7
This article is protected by copyright. All rights reserved.

10.1002/cctc.202000329
ChemCatChem
FULL PAPER
Entry for the Table of Contents
The photocatalyst Ni/CdS@MIL-101 (MIL-101 green, CdS yellow, Ni grey) dehydrogenates benzylamine via the liberation of H2
(amine splitting). A faster electron transfer from CdS to Ni and a stronger benzylamine binding at the Ni clusters make the earth
abundant metal Ni superior in comparison to Pd and Pt.
8
This article is protected by copyright. All rights reserved.