Isotope Effects in Plasmonic Photosynthesis

Article abstract of DOI:10.1002/anie.202011805

The photoexcitation of plasmonic nanoparticles has been shown to drive multistep, multicarrier transformations, such as the conversion of CO₂ into hydrocarbons. But for such plasmon-driven chemistry to be precisely understood and modeled, the critical photoinitiation step in the reaction cascade must be identified. We meet this goal by measuring H/D and ¹²C/¹³C kinetic isotope effects (KIEs) in plasmonic photosynthesis. In particular, we found that the substitution of H₂O with D₂O slows hydrocarbon production by a factor of 5–8. This primary H/D KIE leads to the inference that hole-driven scission of the O?H bond in H₂O is a critical, limiting step in plasmonic photosynthesis. This study advances mechanistic understanding of light-driven chemical reactions on plasmonic nanoparticles.

Full text of DOI:10.1002/anie.202011805

Angewandte
Communications
Chemie
How to cite:
International Edition: doi.org/10.1002/anie.202011805
German Edition: doi.org/10.1002/ange.202011805
CO2 Reduction
Isotope Effects in Plasmonic Photosynthesis
Sungju Yu and Prashant K. Jain*
Abstract: The photoexcitation of plasmonic nanoparticles has
been shown to drive multistep, multicarrier transformations,
such as the conversion of CO2 into hydrocarbons. But for such
plasmon-driven chemistry to be precisely understood and
modeled, the critical photoinitiation step in the reaction
cascade must be identified. We meet this goal by measuring
H/D and 12C/13C kinetic isotope effects (KIEs) in plasmonic
photosynthesis. In particular, we found that the substitution of
H2O with D2O slows hydrocarbon production by a factor of 5–
8. This primary H/D KIE leads to the inference that hole-
free energy. The mechanism by which plasmonic excitation
boosts this process is therefore worthy of elucidation. This
task is rendered difficult by the complexity of plasmonic
photosynthesis: multiple electron (eꢀ) and proton (H+)
=
ꢀ
transfer, C O bond cleavage, C H bond formation, and
ꢀ
C C coupling steps are involved in the conversion. In
addition to the reduction side processes, there is an oxidation
half-reaction involving splitting of H2O. Which of these
elementary steps is rate determining in the plasmon-excita-
tion-driven conversion of CO2? To answer this question, we
study the effect of carbon-13 (13C) substitution in CO2 and
deuterium (D) substitution in H2O on the kinetics of the
photoreaction. The measured kinetic isotope effects (KIEs)
ꢀ
driven scission of the O H bond in H2O is a critical, limiting
step in plasmonic photosynthesis. This study advances mech-
anistic understanding of light-driven chemical reactions on
plasmonic nanoparticles.
ꢀ
are consistent with a dissociated O H bond of H2O in the
transition state (TS). The crucial insight we draw is that light-
driven, hole (h+)-mediated splitting of H2O, which supplies
H+ and eꢀ for hydrocarbon synthesis, is the critical, limiting
step in plasmonic photosynthesis.
M
etal nanostructures have become prime platforms for
concentrating and harvesting visible light.[1] The plasmon
resonances exhibited by these materials impart them with
strong, tunable light absorption. Recent years have seen
a rising number of case studies where the photoexcitation of
plasmon resonances of coinage metal nanoparticles (NPs)
induces chemical reactions.[2] Processes ranging from metal
ion reduction to bond dissociation have been triggered by
plasmonic excitation. These phenomena are enabling the
conversion of light into chemical energy. For instance, light
excitation drives the conversion of CO2 into energy-rich
hydrocarbons on plasmonic Au NPs in the presence of
1-ethyl-3-methylimidazolium tetrafluoroborate (EMIM–
BF4).[3] The CO2 reduction reaction (CO2RR), while of
central importance in renewable energy and CO2 valoriza-
tion,[4] is kinetically challenging[5] and requires the input of
For the KIE investigation, we employed the photoreac-
tion system described in our previous studies.[3] A photo-
catalyst film of approximately 12 nm Au NPs was immersed in
CO2-saturated water containing 5 mol% EMIM–BF4 as
a promoter. The film was irradiated with a continuous-wave
(CW) 532 nm laser with an intensity of 1 Wcmꢀ2. The laser
excites the localized surface plasmon resonances (LSPRs) of
the Au NPs, which have a LSPR band centered at approx-
imately 520 nm (Figure S1). Under plasmonic excitation, CO2
is typically reduced to C1, C2, and C3 hydrocarbons, and H2O is
oxidized to H2O2. It must be noted that a large fraction of the
light absorbed by the Au NPs is dissipated as heat, which
results in an increase in the temperature of the reaction
medium to about 488C under constant 532 nm laser excitation
of an intensity of 1 Wcmꢀ2 [3a,b]
However, such photothermal
.
[*] Dr. S. Yu, Prof. P. K. Jain
heating of the reaction medium is not responsible for the
observed CO2RR activity. In the absence of light excitation,
setting the reaction medium to a temperature of 508C[3a] or
even to a temperature as high as 708C,[3b] while keeping all
other reaction conditions the same, did not lead to any
observable hydrocarbon production. These control studies
show that the CO2RR activity observed under plasmonic
excitation is photochemically driven and not simply induced
by photothermal heating.
Turnover frequencies (TOFs) of production of individual
hydrocarbons by plasmonic photosynthesis were measured in
three different reaction mixtures (Figure 1): 1) 12CO2-satu-
rated H2O, 2) 13CO2-saturated H2O, and 3) 12CO2-saturated
D2O. It must be noted that the EMIM–BF4 was not of the
deuterated form. Therefore, the reaction mixture is not 100%
deuterated; rather it consists of a minor fraction of non-
deuterated H+ originating from the exchangeable H+ of the
imidazolium salt. The methodology of TOF measurements is
Department of Chemistry, University of Illinois at Urbana-Champaign
Urbana, Illinois 61801 (USA)
E-mail: jain@illinois.edu
Prof. P. K. Jain
Materials Research Laboratory
University of Illinois at Urbana-Champaign
Urbana, Illinois 61801 (USA)
and
Department of Physics, University of Illinois at Urbana-Champaign
Urbana, Illinois 61801 (USA),
and
Beckman Institute for Advanced Science and Technology
University of Illinois at Urbana-Champaign
Urbana, Illinois 61801 (USA)
Dr. S. Yu
Present address: Department of Energy Systems Research
Department of Chemistry, Ajou University
Suwon 16499 (Republic of Korea)
Supporting information and the ORCID identification number(s) for
the author(s) of this article can be found under:
https://doi.org/10.1002/anie.202011805.
Angew. Chem. Int. Ed. 2020, 59, 1 – 5
ꢀ 2020 Wiley-VCH GmbH
1
These are not the final page numbers!

Angewandte
Communications
Chemie
Figure 2. Selectivity (%) for each hydrocarbon product formed in
plasmon-excitation-driven CO2RR on Au NPs in 5 mol% EMIM–BF4 for
three different isotopologue combinations: 12CO2 +H2O (left),
13CO2 +H2O (middle), and 12CO2 +D2O (right). The photoreaction was
conducted under 532 nm laser excitation with an incident light
intensity of 1 Wcmꢀ2. The height of each data bar represents the
average of the selectivity (%) measured for a hydrocarbon product in
three independent, identical trials; the error bar represents the SD of
these measurements.
Figure 1. TOFs of hydrocarbons produced in plasmon-excitation-driven
CO2RR on Au NPs in 5 mol% EMIM–BF4 for three different isotopo-
logue combinations: 12CO2 +H2O (left), 13CO2 +H2O (middle), and
12CO2 +D2O (right). The photoreaction was conducted under 532 nm
laser excitation with an incident light intensity of 1 Wcmꢀ2. Each data
bar shows the average of TOF values measured in three independent,
identical trials; the error bar represents the standard deviation (SD) of
these measurements.
described in detail in the Supporting Information, and the
results are presented in Figures S2–S7.
In the reaction in 12CO2-saturated H2O, CH4, C2H2, C2H4,
C3H6 and C3H8 are produced. The product selectivity followed
the order: C1 > C2 > C3. The C2+ selectivity is close to 50%
(Figure 2). With the heavier 13CO2 isotopologue, the observed
products and the order of selectivity are similar. However, the
TOFs are marginally lower. Although the change in the TOF
is not significant for the C1 product, the C2 and C3 production
show a statistically significant decrease in TOF in 13CO2-
saturated H2O. In fact, it appears that the longer the carbon
chain, the more significantly is its production rate affected by
13C-substitution. There is a slight decrease in the C2+
selectivity and an associated increase in the C1 selectivity in
13CO2-saturated H2O, as compared to 12CO2-saturated H2O.
The isotope effects of D-substitution in H2O are more
dramatic (Figure 1 and Table S1). In 12CO2-saturated D2O, the
kinetics of hydrocarbon production are considerably sup-
pressed as compared to those in 12CO2-saturated H2O. The
TOFs of C1 and C2 production are approximately sixfold
lower in D2O; moreover, no C3 products are formed and the
product profile is dominated by the C1 hydrocarbon, CH4.
Plasmonically generated eꢀ are harvested in the form of
hydrocarbons produced by CO2RR. The effect of the isotopic
substitution can also be characterized in terms of the rate of
eꢀ harvesting (Figure 3). As compared to the normal CO2
isotopologue, the eꢀ harvesting rates achieved with the 13C
isotopologue are lower by approximately 7% for C1, 14% for
C2, and 25% for C3 products; the overall rate of eꢀ harvesting
is lower by approximately 15%. The eꢀ harvesting rates
achieved in D2O are lower than those in H2O by approx-
imately 80% for C1, 82% for C2, and 100% for C3 products;
the overall eꢀ harvesting rate is lower by approximately 87%.
Figure 3. Rate of harvesting of eꢀ, d[eꢀ]/dt, in plasmon-excitation-
driven CO2RR on Au NPs in 5 mol% EMIM–BF4 for three different
isotopologue combinations: 12CO2 +H2O (left), 13CO2 +H2O (middle),
and 12CO2 +D2O (right). The photoreaction was conducted under
532 nm laser excitation with an incident light intensity of 1 Wcmꢀ2
.
The height of each data bar represents the average of d[eꢀ]/dt
measured for a hydrocarbon product in three independent, identical
trials; the error bar represents the SD of these measurements. The
method for determination of d[eꢀ]/dt is described in the Supporting
Information.
Table S1 lists the KIEs in terms of each product and the
net eꢀ harvesting rate. When the 95% confidence interval
(average ꢁ 1.96 ꢀ standard error (SE))[6] is taken into account,
the 12C/13C KIEs are not significantly different from 1 for C1
and C2 products, while the 12C/13C KIEs for the C3 products
2
www.angewandte.org
ꢀ 2020 Wiley-VCH GmbH
Angew. Chem. Int. Ed. 2020, 59, 1 – 5
These are not the final page numbers!

Angewandte
Communications
Chemie
are greater than 1 in a statistically significant manner. This
stretching modes of H2O, respectively. Each term, 1/2hv, is
the ZPVE of the respective mode. At ordinary temperatures,
excited vibrational states are not populated significantly; only
the ground zero-point vibrational state is important. From
Equation (3), we get:
12
13
=
C/ C KIE may be reflective of the involvement of C O
bond dissociation in the activation of CO2 for conversion to
13
=
a hydrocarbon. With C-substitution, the C O bond has an
approximately 5% larger reduced mass and consequently an
approximately 2.5% lower zero-point vibrational energy
ꢀꢂ
ꢃ
ꢂ
ꢃ
ꢁ
1
2
=
DG°D ꢀ DG°H
¼
hnH O;s ꢀ hnD
þ hnH O;as ꢀ hnD
ꢀ ðhnOH ꢀ hnOD
Þ
(ZPVE). The activation barrier for C O bond dissociation
2 O;s
2O;as
2
2
is therefore larger for the 13C isotopologue, which would lead
ð4Þ
to a marginal reduction in the hydrocarbon production rate.
=
However, this putative role of C O bond dissociation cannot
ꢀ
For a vibrating bond A B treated as a harmonic oscillator,
its stretching frequency is given by:
be definitively concluded due to the weak statistical signifi-
cance of the 12C/13C KIEs.
rffiffiffi
The H/D KIE is statistically significant (Table S1), allow-
ing more definitive mechanistic insights to be drawn from its
analysis. If C H bond formation, which is an elementary step
in hydrocarbon synthesis, were rate limiting, an inverse H/D
KIE would have been observed. Instead, we observe a normal
H/D KIE with a value of 5–8. This magnitude is indicative of
a primary KIE, attributable to the dissociation of a bond
1
2p
a
m
ð5Þ
n ¼
ꢀ
Here a is the spring constant, which is same for the H- and
D-isotopologues. m is the reduced mass of the diatomic
fragment, AB, which is given by:
mAmB
mA þ mB
ꢀ
involving a H atom in the TS. The dissociation of the O H
m ¼
ð6Þ
ð7Þ
bond of H2O, an elementary step in plasmonic photosynthesis,
is a prime candidate. In fact, a KIE of a similar magnitude has
been observed in H2O oxidation and dissociation.[7]
in terms of the atomic masses, mA and mB. Thus, we get:
The premise of a photoinduced h+-mediated cleavage of
rffiffiffiffiffiffiffiffiꢄ
rffiffiffiffiffiffiffiffiꢅ
1
a
mOH
the O H bond in H2O (i.e., H2O + h ! COH + H+) as the
+
nOH ꢀ nOD
¼
1 ꢀ
ꢀ
2p mOH
mOD
critical step in plasmonic photosynthesis reproduces the
measured H/D KIE. In Eyring rate theory, the H/D KIE is
expressed as the ratio of the apparent rate constant for the
reaction involving the H-isotopologue, kapp,H, to that of the
which is equal to 0.272vOH, when the values of the reduced
masses are substituted. Thus, vOD = 0.728vOH, which is fairly
close to the estimate: vOD = 0.737vOH obtained by using
[8]
heavier D-isotopologue, kapp,D
:
known experimental values for the mode frequencies:
v
OH
=
3574 cmꢀ1 and
rewritten as:
v
OD = 2635 cmꢀ1. Finally, Equation (4) is
expðꢀDG°H=kbTÞ
kapp;H
kapp;D
ð1Þ
¼
expðꢀDG°D=kbTÞ
ꢂ
ꢃ
0:272
2
DG°D ꢀ DG°H
¼
hnH O;s þ hnH O;as ꢀ hnOH
ð8Þ
where DG°H is the free energy of activation for the rate-
limiting step involving the H isotopologue and DG°D is the
analogous quantity for the D-isotopologue. kb is the Boltz-
mann constant and T is the temperature. Equation (1) can be
expressed as:
2
2
From the known values of nH O;s, nH O;as, and nOH, which
2
2
are 3280 cmꢀ1, 3490 cmꢀ1, and 3574 cmꢀ1, respectively,[8,9] and
Equations (2) and (8), we calculate the H/D KIE. We obtain
an estimate of 7.02 at 321 K or 488C, the temperature of the
bulk solution in the photoreaction. When we account for the
isotope effect of D-substitution on the acidity of the reaction
medium (Supporting Note), we theoretically estimate the
apparent H/D KIE to be 8.42.
ꢀ
ꢁ
kapp;H
kapp;D
¼ exp ðDG°D ꢀ DGH°Þ=kbT
ð2Þ
In general, DG° is dominated by the difference in the
vibrational energies of the TS and the reactant state (RS). If
This simple estimate does not include contributions from
H+ tunneling, although these lead to much larger KIEs[10], or
other phenomena such as H+-coupled eꢀ transfer. Secondly,
ꢀ
we assume that the O H bond is fully broken in the TS, i.e.,
a late TS, the vibrational energy of the TS is solely contributed
ꢀ
by the O H stretching mode of the COH radical formed by
ꢀ
the O H bond in the TS may not be fully broken; or, in the
bond dissociation. The vibrational energy of RS is contributed
TS, the H+ (formed by H2O dissociation) may be partially
bonded with a CO2ꢀC anion radical (formed by 1 eꢀ reduction
of CO2) eventually generating a HOCO radical.[11] In such
cases, the measured H/D KIE will be smaller than the value
calculated here.
ꢀ
by the symmetric and antisymmetric O H stretching modes
of H2O. We neglect the low frequency bending mode that has
a much smaller contribution to the DG°. Thus, DGH° is given
by:
Notwithstanding the approximations made, the corre-
spondence between the magnitudes of the calculated and
measured H/D KIEs (Table S1) indicates that h+-mediated
1
2
1
2
1
2
DG°H
¼
hnOH
ꢀ
hnH O;s ꢀ hnH
ð3Þ
2 O;as
2
ꢀ
ꢀ
where h is the Planck constant. vOH is the frequency of the O
dissociation of the O H bond in H2O is the limiting step in
H stretching mode of the COH radical. nH O;s and nH O;as are the
plasmonic photosynthesis. This elementary step supplies eꢀ
and H+ for activation and hydrogenation of CO2. These
2
2
ꢀ
frequencies of the O H symmetric and antisymmetric
Angew. Chem. Int. Ed. 2020, 59, 1 – 5
ꢀ 2020 Wiley-VCH GmbH
www.angewandte.org
3
These are not the final page numbers!

Angewandte
Communications
Chemie
ꢀ
subsequent CO2 hydrogenation or C H bond formation steps
appear to be kinetically downstream. The production of
longer hydrocarbons has a higher reaction order in the
concentration of eꢀ and H+ formed from H2O dissociation;[3a]
this is likely reason for the stronger KIE measured for the
longer hydrocarbons.
Keywords: catalysis · CO2 reduction · hot electrons ·
localized surface plasmon resonance · nanoparticles
[1] S. Yu, V. Mohan, P. K. Jain, MRS Bull. 2020, 45, 43 – 48.
[2] a) R. Jin, Y. Cao, C. A. Mirkin, K. L. Kelly, G. C. Schatz, J. G.
Zheng, Science 2001, 294, 1901; b) M. Maillard, P. Huang, L.
Brus, Nano Lett. 2003, 3, 1611 – 1615; c) X. Wu, P. L. Redmond,
H. Liu, Y. Chen, M. Steigerwald, L. Brus, J. Am. Chem. Soc.
2008, 130, 9500 – 9506; d) P. Christopher, H. Xin, S. Linic, Nat.
Chem. 2011, 3, 467 – 472; e) P. Christopher, H. Xin, A. Marimu-
thu, S. Linic, Nat. Mater. 2012, 11, 1044 – 1050; f) S. Mukherjee, F.
Libisch, N. Large, O. Neumann, L. V. Brown, J. Cheng, J. B.
Lassiter, E. A. Carter, P. Nordlander, N. J. Halas, Nano Lett.
2013, 13, 240 – 247; g) S. Mukherjee, L. Zhou, A. M. Goodman,
N. Large, C. Ayala-Orozco, Y. Zhang, P. Nordlander, N. J. Halas,
J. Am. Chem. Soc. 2014, 136, 64 – 67; h) K. M. Choi, D. Kim, B.
Rungtaweevoranit, C. A. Trickett, J. T. D. Barmanbek, A. S.
Alshammari, P. Yang, O. M. Yaghi, J. Am. Chem. Soc. 2017, 139,
356 – 362; i) A. Naldoni, T. Montini, F. Malara, M. M. Mrꢁz, A.
Beltram, T. Virgili, C. L. Boldrini, M. Marelli, I. Romero-OcaÇa,
J. J. Delgado, V. Dal Santo, P. Fornasiero, ACS Catal. 2017, 7,
1270 – 1278; j) L. Zhou, D. F. Swearer, C. Zhang, H. Robatjazi,
H. Zhao, L. Henderson, L. Dong, P. Christopher, E. A. Carter, P.
Nordlander, N. J. Halas, Science 2018, 362, 69; k) O. A. Baturina,
A. Epshteyn, B. S. Simpkins, N. Bhattarai, T. H. Brintlinger,
E. Y. Santiago, A. O. Govorov, J. Electrochem. Soc. 2019, 166,
H485 – H493; l) J. S. DuChene, G. Tagliabue, A. J. Welch, W.-H.
Cheng, H. A. Atwater, Nano Lett. 2018, 18, 2545 – 2550; m) H.
Gao, C. Liu, H. E. Jeong, P. Yang, ACS Nano 2012, 6, 234 – 240;
n) M. T. Sheldon, J. van de Groep, A. M. Brown, A. Polman,
H. A. Atwater, Science 2014, 346, 828; o) S. Yu, A. J. Wilson, J.
Heo, P. K. Jain, Nano Lett. 2018, 18, 2189 – 2194.
[3] a) S. Yu, P. K. Jain, Nat. Commun. 2019, 10, 2022; b) S. Yu, P. K.
Jain, ACS Energy Lett. 2019, 4, 2295 – 2300; c) S. Yu, P. K. Jain,
Angew. Chem. Int. Ed. 2020, 59, 2085 – 2088; Angew. Chem.
2020, 132, 2101 – 2104.
[4] S. Yu, A. J. Wilson, G. Kumari, X. Zhang, P. K. Jain, ACS Energy
Lett. 2017, 2, 2058 – 2070.
[5] a) F. Calle-Vallejo, M. T. M. Koper, Angew. Chem. Int. Ed. 2013,
52, 7282 – 7285; Angew. Chem. 2013, 125, 7423 – 7426; b) X. Nie,
M. R. Esopi, M. J. Janik, A. Asthagiri, Angew. Chem. Int. Ed.
2013, 52, 2459 – 2462; Angew. Chem. 2013, 125, 2519 – 2522; c) K.
Jiang, R. B. Sandberg, A. J. Akey, X. Liu, D. C. Bell, J. K.
Nørskov, K. Chan, H. Wang, Nat. Catal. 2018, 1, 111 – 119.
[6] K. Francis, A. Kohen, Perspect. Sci. 2014, 1, 110 – 120.
[7] a) E. D. German, M. Sheintuch, J. Phys. Chem. C 2010, 114,
3089 – 3097; b) A. Klauss, M. Haumann, H. Dau, Proc. Natl.
Acad. Sci. USA 2012, 109, 16035.
The inference that h+-mediated oxidative scission of the
O H bond of H2O is the critical step in plasmonic photosyn-
ꢀ
thesis is further supported by the dependence of the CO2RR
activity on the excitation wavelength (Figure S8a–d). It is
known from previous work[2o,12] that interband excitation of
Au NPs induces redox chemical activity relatively more
efficiently than on-resonant excitation of the LSPR band.
Interband excitation produces energetic d-band h+, which are
required for the thermodynamically challenging water oxida-
tion half-reaction. On the other hand, only those LSPR
excitations that relax via interband damping produce ener-
getic d-band holes. As a result, interband excitation is
expected to drive plasmonic photosynthesis relatively more
efficiently than LSPR excitation of the Au NPs. To test this
expectation, we measured the CO2RR activity in a 12CO2-
saturated solution of 5 mol% EMIM–BF4 in H2O under CW
light excitation of a wavelength of 488 nm, which is more
effective at exciting interband transitions in Au than on-
resonant LSPR excitation at a wavelength of 532 nm. All
other conditions, including the incident light intensity, were
identical to the photoreactions performed under 532 nm
excitation. While similar CO2RR activity and products are
observed under 488 nm excitation, the TOFs are higher under
488 nm excitation (Figure S8a) as compared to those under
532 nm excitation (Figure S2) for all hydrocarbon products.
This is despite the lower absorbance of the Au NP photo-
catalyst at 488 nm excitation. The net rate of eꢀ harvesting is
1.6 ꢀ higher under 488 nm excitation (Figure S8c) than that at
532 nm excitation (Figure 3). Taken together, the light-driven
CO2RR activity is higher under interband (488 nm) excitation
as compared to that under LSPR (532 nm) excitation. We can
infer that interband excitation, which produces d-band holes
more efficiently, is more effective at driving h+-mediated
oxidative scission of H2O, the rate-limiting step in plasmonic
photosynthesis.
In conclusion, the plasmon-excitation-driven CO2RR on
Au NPs with an ionic-liquid promoter shows a primary H/D
KIE. We learn that the conversion of CO2 to hydrocarbons is
limited by light-driven dissociation of H2O.
[8] J. F. Ogilvie, Nature 1964, 204, 572 – 572.
[9] G. E. Walrafen, E. Pugh, J. Solution Chem. 2004, 33, 81 – 97.
[10] Y. Cha, C. J. Murray, J. P. Klinman, Science 1989, 243, 1325.
[11] G. Kumari, X. Zhang, D. Devasia, J. Heo, P. K. Jain, ACS Nano
2018, 12, 8330 – 8340.
[12] a) Y. Kim, D. Dumett Torres, P. K. Jain, Nano Lett. 2016, 16,
3399 – 3407; b) Y. Kim, J. G. Smith, P. K. Jain, Nat. Chem. 2018,
10, 763 – 769; c) A. J. Wilson, V. Mohan, P. K. Jain, J. Phys.
Chem. C 2019, 123, 29360 – 29369.
Acknowledgements
Funding for this work was provided by the Energy &
Biosciences Institute (EBI) through the EBI-Shell program.
We thank Varun Mohan, who compared pH and pD values of
EMIM–BF4.
Conflict of interest
Manuscript received: August 28, 2020
Accepted manuscript online: September 8, 2020
Version of record online: && &&, &&&&
The authors declare no conflict of interest.
4
www.angewandte.org
ꢀ 2020 Wiley-VCH GmbH
Angew. Chem. Int. Ed. 2020, 59, 1 – 5
These are not the final page numbers!

Angewandte
Communications
Chemie
Communications
CO2 Reduction
In plasmon-excitation-driven photosyn-
thesis on Au nanoparticles, the primary
step is the hole-mediated splitting of
water, which supplies protons and elec-
trons for carbon dioxide conversion. This
mechanistic insight into a complex, mul-
tistep transformation was obtained from
measurements of kinetic isotope effects
in the reaction (see picture).
S. Yu, P. K. Jain*
&&&& — &&&&
Isotope Effects in Plasmonic
Photosynthesis
Angew. Chem. Int. Ed. 2020, 59, 1 – 5
ꢀ 2020 Wiley-VCH GmbH
www.angewandte.org
5
These are not the final page numbers!