Energy Transfer Dynamics of Formate Decomposition on Cu(110)

Article abstract of DOI:10.1002/anie.201611342

Energy transfer dynamics of formate (HCOO_a) decomposition on a Cu(110) surface has been studied by measuring the angle-resolved intensity and translational energy distributions of CO₂ emitted from the surface in a steady-state reaction of HCOOH and O₂. The angular distribution of CO₂ shows a sharp collimation with the direction perpendicular to the surface, as represented by cosⁿθ (n=6). The mean translational energy of CO₂ is measured to be as low as 100 meV and is independent of the surface temperature (T_s). These results clearly indicate that the decomposition of formate is a thermal non-equilibrium process in which a large amount of energy released by the decomposition reaction of formate is transformed into the internal energies of CO₂ molecules. The thermal non-equilibrium features observed in the dynamics of formate decomposition support the proposed Eley–Rideal (ER)-type mechanism for formate synthesis on copper catalysts.

Full text of DOI:10.1002/anie.201611342

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Angewandte
Communications
Chemie
International Edition: DOI: 10.1002/anie.201611342
German Edition: DOI: 10.1002/ange.201611342
Interfacial Reaction Dynamics
Hot Paper
Energy Transfer Dynamics of Formate Decomposition on Cu(110)
Jiamei Quan, Takahiro Kondo, Guichang Wang, and Junji Nakamura*
Abstract: Energy transfer dynamics of formate (HCOOa)
decomposition on a Cu(110) surface has been studied by
measuring the angle-resolved intensity and translational energy
distributions of CO2 emitted from the surface in a steady-state
reaction of HCOOH and O2. The angular distribution of CO2
shows a sharp collimation with the direction perpendicular to
the surface, as represented by cosnq (n = 6). The mean trans-
lational energy of CO2 is measured to be as low as 100 meVand
is independent of the surface temperature (Ts). These results
clearly indicate that the decomposition of formate is a thermal
non-equilibrium process in which a large amount of energy
released by the decomposition reaction of formate is trans-
formed into the internal energies of CO2 molecules. The
thermal non-equilibrium features observed in the dynamics of
formate decomposition support the proposed Eley–Rideal
(ER)-type mechanism for formate synthesis on copper
catalysts.
distributions of the CO2 produced from the decomposition
of formate during the steady-state reaction of HCOOH with
O2 on Cu(110).
Formate (HCOOa) is an important intermediate in the
synthesis of methanol, and is formed on copper surfaces
during the initial elementary CO2 hydrogenation step,
namely: CO2 + 1/2H2!HCOOa.[8–10] Formate synthesis is
unique in terms of reaction mechanism and dynamics, as
well as its structure-insensitive kinetics. We have proposed an
ER-type mechanism for formate synthesis, based on exper-
imental kinetic analysis and theoretical calculations,[11,12] in
which CO2 directly reacts with the adsorbed hydrogen atoms
(Ha) on the copper surface without the involvement of
trapping states and adsorption precursors. This, in turn,
implies that the formate production rate can only be
enhanced by controlling the energy of the CO2 molecules in
terms of a thermal non-equilibrium character. The dynamics
of formate decomposition on copper catalysts is thus
interesting as the reverse reaction of formate synthesis:
HCOOa!CO2 + Ha.
E
nergy transfer and bond rupture/formation are two impor-
tant events in a chemical reaction.[1,2] Consequently, the
reaction mechanism must be examined with both aspects in
mind. For gas-phase bimolecular reactions, energy transfer
processes have been extensively examined by applying
crossed molecular beam techniques, where detailed dynam-
ical parameters (such as total momentum, energy, and angular
momentum, before and after the collision events) can
describe the partitioning of energy into each available mode
of the products.[3,4] On the other hand, characterizing the
energy-transfer processes in gas–surface chemical reactions is
challenging because of the large number of degrees of
freedom available to the surface atoms. To analyze a gas–
surface chemical reaction, angle-resolved (AR) analysis of
the desorbed products from the surface is one of the direct
methods used because it relates to the transition state (TS)
structure of the reaction.[5–7] In this communication, we report
the angle-resolved intensity and translational energy
It is well-known that there are two types of formate on
copper surfaces: monodentate and bidentate formate.[13–16]
We have reported the energy diagram for formate synthesis
from CO2 and H2 in which CO2 is converted to bidentate
formate via monodentate formate.[12] The adsorption energies
of bidentate formate and monodentate formate on Cu(110)
were calculated as ꢀ1.10 eVand ꢀ0.32 eV with respect to the
energy for CO2 + Ha, respectively. That is, bidentate formate
is much more stable compared to monodentate formate. The
energy diagram explains the experimentally measured acti-
vation energies and the structure dependence for formate
synthesis/decomposition well; thus, we believe the decom-
position mechanism proceeds by transformation of bidentate
formate into monodentate formate, followed by decomposi-
tion into CO2 and Ha. Therefore, we consider monodentate
formate an important intermediate for determining the
dynamics of formate decomposition.
Initially, we investigated the structural changes near the
TS of monodentate formate decomposition on Cu(110),
which were calculated by density functional theory (DFT;
Figure 1). The energy diagram profile starts from a mono-
dentate formate/Cu(110) surface as an initial state (IS).
The barrier for the decomposition of formate is calculated to
be 0.50 eV, during which the monodentate formate is rotated
[*] Prof. T. Kondo, Prof. J. Nakamura
Faculty of Pure and Applied Sciences, Tsukuba Research Center for
Interdisciplinary Materials Science (TIMS) & Center for Integrated
Research in Fundamental Science and Engineering (CiRfSE)
University of Tsukuba
Tsukuba, Ibaraki 305-8573 (Japan)
E-mail: nakamura@ims.tsukuba.ac.jp
J. Quan
ꢀ
to form a Cu H bond and is followed by the successive
Graduate school of Pure and Applied Sciences, University of Tsukuba
Tsukuba, Ibaraki 305-8573 (Japan)
ꢀ
ꢀ
cleavage of the Cu O and C H bonds of the formate group.
The unique point is that nascent CO2 has already been formed
at the TS after the cleavage of two bonds. The structural
changes (TS!3!4!5!FS; FS = final state) in Figure 1
provide useful information to discuss the dynamics of formate
decomposition. From the TS to the FS, the O-C-O angle
changes: 154.798!168.608!175.688!176.528!179.298. The
Prof. G. Wang
Department of Chemistry, Key Laboratory of Advanced Energy
Materials Chemistry (Ministry of Education) and Collaborative
Innovation Center of Chemical Science and Engineering (Tianjin)
Nankai University
Tianjin 300071 (P.R. China)
Angew. Chem. Int. Ed. 2017, 56, 1 – 6
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
1
These are not the final page numbers!

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Angewandte
Communications
Chemie
formate on Cu(110). The obtained results provide insights
into energy dissipation because the desorbing CO2 molecules
may interact with the substrate during formate decomposition
on Cu(110).
The formation of 13CO2 during the steady-state reaction of
H13COOH with O2 on Cu(110) was monitored by angle-
resolved intensity measurements using a mass spectrometer.
H13COOH and O2 were continuously introduced and main-
tained at P(H13COOH) = 0.7 ꢁ 10ꢀ7 torr and P(O2) = 1.15 ꢁ
10ꢀ7 torr, respectively. Figure 2(a) shows the Ts dependence
of the AR 13CO2 mass signals (m/z = 45) at q = 08 with heating
and cooling of the Cu(110) sample. The production of 13CO2
was negligible below 400 K; however, it started to rise
significantly above 400 K and reached the maximum level at
452 K, after which it started decreasing gradually. In both
heating and cooling modes, the CO2 signals were observed to
be almost identical, indicating that the reaction of formic acid
with oxygen reached the steady state.
The kinetics of formate decomposition on copper surfaces
have been studied extensively, and it is well-known that this
decomposition takes place at 410–450 K to form CO2 and
H2.[20,21] The formation of CO2 above 400 K in Figure 2(a) is
thus attributed to the decomposition of formate. As shown in
Figure 2(b), the adsorbed formate structure is in a bidentate
form, that is, two oxygen atoms are bound to the two nearest
neighboring surface copper atoms; this has clearly been
Figure 1. Energetics of the monodentate formate decomposition
process (mono-HCOOa!CO2 +Ha) on Cu(110). The insets show the
ꢀ ꢀ
structural changes with bond lengths (ꢀ) and O C O angles (8).
The energy of monodentate formate adsorbed on Cu(110) is taken to
be the reference state (zero relative energy). Key: initial state (IS),
transition state (TS), final state (FS).
154.798 angle corresponds to a highly excited
bending mode, which is 0.44 eV higher than CO2
with an angle of 1808.[17] The angle changes thus
suggest that the energy gain of 0.71 eV from the
TS to the FS in Figure 1 is mainly transformed
into the vibrational energy of CO2. It is also noted
that the distance between CO2 and the copper
surface changes from 2.10 ꢀ!2.14 ꢀ!2.69 ꢀ for
the structures 4!5!FS without significant
changes in the angle of CO2 and Ha/Cu, while
the energy difference between 4 and FS is 0.50 eV.
This energy difference is thus partly ascribed to
the Pauli repulsion between CO2 and the copper
surface. It is suggested that the desorbing CO2
molecule receives some repulsive forces, leading
to an increase in translational energy and/or
rotational energy of CO2. Accordingly, the struc-
tural changes from the TS to the FS suggest that
the energy gain of 0.71 eV is converted to the
internal energy (vibration, rotation, and trans-
lation) of CO2.
Figure 2. a) AR 13CO2 signal, as a function of Ts, along the direction normal (q=08)
to the Cu(110) surface under steady-state reaction conditions. The pressures of
H13COOH and O2 are 0.7ꢁ10ꢀ7 torr and 1.15ꢁ10ꢀ7 torr, respectively. The Ts was first
increased from 325 K to 704 K (red circles) and then decreased to 400 K (blue
circles) as indicted by arrows. b) Depiction of bidentate formate species on Cu(110).
AR distribution of 13CO2 with a polar angle along the [001] direction of Cu(110)
under steady-state reaction conditions. c, d) 13CO2 AR distributions at Ts values of
443 K and 456 K, respectively. Black dashed curves depict the distributions of cosnq
(n=6) obtained by fitting analysis.
In the case of formate decomposition on
platinum foil and Pd(111), infrared emission
spectroscopy showed that the vibrational temperature (Tv)
and rotational temperature (Tr) of the product CO2 were
much higher than Ts (surface temperature).[18,19] Therefore, to
understand the energy-transfer dynamics of the cleavage step
of the last bond (in the TS region), the direct measurement of
translational energy is indispensable. Thus, we measured the
angular intensity and translational energy distributions of the
CO2 desorbed from the surface during the decomposition of
identified previously.[22] It is also known that formic acid
reacts readily with adsorbed oxygen to form formate and
water, even at room temperature.[21,23–25] The signals
corresponding to the masses of 13C16O (m/z = 29) and
13C16O16OH (m/z = 46) are negligible. The elemental steps
for the overall reaction of formic acid with oxygen in
Figure 2a are therefore described by Equations 1–6 (see
subsequent text).[25]
2
www.angewandte.org
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2017, 56, 1 – 6
These are not the final page numbers!

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Angewandte
Communications
Chemie
O2 2 O
ð1Þ
!
a
H13COOH þ Oa H13COO þ OH
ð2Þ
ð3Þ
ð4Þ
ð5Þ
!
a
a
2 OHa H O þ O
!
2
a
H13COOH þ OHa H13COO þ H O
!
a
2
H13COOa 13CO2 þ Ha
!
2 Ha
H
2
ð6Þ
!
Steady-state coverage of intermediates on
Cu(110) is determined by the kinetics of elemen-
tary steps (1)–(6). The elementary steps of (1)–(3)
and (6) are fast at room temperature, while
Equation (5) is slow as the rate-determining step
taking place above 400 K.[25] Although the kinet-
ics of step (4) have not been measured to date, it is
inferred that the major intermediate is formate in
the steady-state reaction above 400 K based on
the kinetic balance. The steady-state coverage of
Figure 3. a) Deconvoluted TOF spectra of 13CO2 (circles) and least squares fits
(lines) along the surface normal (q=08) direction of Cu(110) under steady-state
reaction condition for varying Ts values. The pressure of H13COOH is 1ꢁ10ꢀ7 torr
and O2 is 1ꢁ10ꢀ7 torr. b) The mean translational energy <Et > of 13CO2 derived
from the TOF spectra in (a).
Oa, OHa, and Ha should be small. We consider in the present
reaction conditions that formate decomposition proceeds on
an adsorbate (Oa, OHa, Ha)-free Cu(110) surface.
as 100 meV, independent of reaction temperatures between
450 K and 700 K. The < Et > of 100 meV corresponds to the
CO2 temperature (T) of 580 K by assuming < Et >= 2kT
(k: Boltzmannꢂs constant), which is lower than the Ts of 700 K
in Figure 3. This indicates the presence of a cooling mecha-
Note that the decomposition of formate in Equation (5) is
the reverse reaction of formate synthesis by the hydrogena-
tion of CO2. On the basis of microreversibility, the dynamics
of formate decomposition are interesting, especially in
relation to the proposed ER-type mechanism for formate
synthesis on copper catalysts.[11,12] Hence, we examined the
reaction dynamics of formate decomposition by measuring
the angular intensity and translational energy distributions of
CO2 as functions of desorption angle and Ts. Figure 2(c) and
2(d) show the AR-measured distributions of 13CO2 along the
[001] direction of Cu(110) at 443 K and 456 K, and
P(H13COOH) = P(O2) = 0.7 ꢁ 10ꢀ7 torr. The angular distribu-
tions of desorbing 13CO2 at both temperatures show a sharp
collimation with cosnq (n = 6) along the surface normal
direction. These sharp distributions indicate that the desorp-
tion of 13CO2 molecules by the thermal decomposition of
formate is a thermal non-equilibrium process. This contrasts
with the desorption of fully thermally accommodated
adsorbed species on the surface, which shows a cosine
distribution.[26,27] Between 443 K and 456 K, the AR distribu-
tion was observed to be independent of Ts, suggesting that the
CO2 just before desorption at the TS for formate decom-
position is not energetically coupled with the Cu(110) surface.
The origin of the observed sharp distribution perpendicular to
the surface can be explained by the Pauli repulsion between
CO2 and the copper surface with sp-band character at the
Fermi level,[28] as suggested by the present DFT calculations
in Figure 1. If the CO2 is energetically coupled with the
surface, the AR distribution should be broader at higher Ts
values.
ꢀ
nism near the TS. As described in Figure 1, the Cu O and
ꢀ
C H bonds of formate are cleaved near the TS while the
energy was used for climbing the decomposition reaction
barrier or breaking the bonds. The endothermic process
should cool the TS. The nascent CO2 at the TS is thus cold,
which then experiences the Pauli repulsion from Ha/Cu. The
product CO2 consequently desorbs from the surface with
translational energy only acquired as a result of the energy
released from the TS, which is independent of Ts, as shown by
Figure 3(b). We note here that the mean translational energy
of 100 meV is much smaller than the DFT-calculated 0.71 eV
energy difference between the TS and the FS. Thus, this
0.71 eV energy can be considered to be distributed to the
vibrational energy and the rotational energy too, as described
in the discussion of Figure 1.
The measured thermal non-equilibrium results clearly
indicate that nascent CO2 desorbs without trapping its
adsorption state after passing through the TS during the
decomposition of formate. That is, no significant energy
transfer takes place between CO2 and the surface around the
TS for formate decomposition. From the viewpoint of energy
transfer, this is in contrast to what has been observed in
associative desorption reactions at surfaces reported in the
literature. For example, the translational energy of CO2
formed by the oxidation of CO on Pd(133) increases linearly
with increasing Ts under steady-state reaction conditions and
follows the relationship: < Et >= 0.24 eV + 2kTs.[29] The
difference in Ts dependence of < Et > between the present
study and the literature is probably caused by a difference in
TS structures. In the present formate decomposition case, the
nascent CO2 at the TS does not bond with the copper surface.
That is considered to be the origin of the thermal non-
equilibrium dynamics of formate decomposition.
Figure 3(a) shows a series of time-of-flight (TOF) data for
13CO2 measured during the steady-state reactions of
H13COOH with O2 at various reaction temperatures. As
shown in Figure 3(b), mean translational energies < Et > of
13CO2 (see Supporting Information) were found to be as low
Angew. Chem. Int. Ed. 2017, 56, 1 – 6
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
www.angewandte.org
3
These are not the final page numbers!

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Angewandte
Communications
Chemie
translation) of desorbing CO2. It is also suggested
that the internal energy conversion of desorbing
CO2 takes place by Pauli repulsion and van der
Waals forces.[36] These thermal non-equilibrium
features support the proposed ER-type mecha-
nism of formate synthesis by copper catalysts,
which is an important elemental step in the
synthesis of methanol by the hydrogenation of
CO2.
Acknowledgements
Funding support from the Advanced Catalytic
Transformation Program for Carbon Utilization
(ACT-C) of the Japan Science and Technology
Agency (JST) is gratefully acknowledged.
Figure 4. a) Deconvoluted TOF spectra of 13CO2 (circles) and least squares fits
(dashed lines) along the [110] direction of Cu(110) at 470 K under steady-state
reaction conditions as a function of desorption polar angle q. The pressure of
H13COOH is 1ꢁ10ꢀ7 torr and O2 is 1ꢁ10ꢀ7 torr. b) The mean translational energy
<Et > of 13CO2 derived from the TOF spectra in (a).
¯
Conflict of interest
Finally, the polar angle dependence of TOF was measured
for desorbed 13CO2 from the surface for the steady-state
The authors declare no conflict of interest.
reaction of H13COOH with O2. Figure 4(a) shows the TOF
results for angles of 08, 158, 308, 458, 578, and 608 at a reaction
temperature of 470 K. The < Et > of 13CO2 was then plotted as
a function of the polar angle of CO2 desorption along the
Keywords: energy transfer · formate decomposition ·
surface reaction dynamics · thermal non-equilibrium ·
translational energy
¯
[110] direction, as shown in Figure 4(b). It was found that
< Et > decreases with increasing polar angle. Decreases in
mean translational energy at higher polar angles have also
been reported for CO2 desorption during CO oxidation
reactions on Rh(111)[30–32] and Pt(111).[33] This decrease in
< Et > cannot be explained by the one-dimensional van
Willigen model.[32–35] However, the angle dependence can be
explained by considering Pauli repulsion and van der Waals
interactions. With increasing desorption polar angle, CO2
should interact with the copper surface for longer times
through Pauli repulsion and then by exerting van der Waals
forces. These synergistic forces result in the transformation of
the translational energy of CO2 into internal energy because
unequal electron densities distributed around carbon and two
oxygen atoms in the molecular orbitals interact differently
with the surface by Pauli repulsion and van der Waals
interactions. The importance of van der Waals interactions in
the final energy states of CO2 desorbed from formate
decomposition on Cu(110) was also shown by recent molec-
ular dynamics simulation methods.[36]
[1] G. Ertl, in Advances in Catalysis, Vol. 45 (Eds.: B. C. Gates, H.
Knozinger), Academic Press, San Diego, 2000, pp. 1 – 69.
[2] A. M. Wodtke, D. Matsiev, D. J. Auerbach, Prog. Surf. Sci. 2008,
83, 167 – 214.
[3] J. C. Polanyi, Acc. Chem. Res. 1972, 5, 161 – 168.
[4] Y. T. Lee, Angew. Chem. Int. Ed. Engl. 1987, 26, 939 – 951;
Angew. Chem. 1987, 99, 967 – 980.
[5] G. Comsa, R. David, Surf. Sci. Rep. 1985, 5, 145 – 198.
[6] T. Matsushima, Prog. Surf. Sci. 2007, 82, 435 – 477.
[7] T. Matsushima, K. Shobatake, J. Mol. Catal. A 2010, 315, 135 –
147.
[8] M. Behrens, F. Studt, I. Kasatkin, S. Kꢃhl, M. Hꢄvecker, F.
Abild-Pedersen, S. Zander, F. Girgsdies, P. Kurr, B.-L. Kniep, M.
Tovar, R. W. Fischer, J. K. Nørskov, R. Schlçgl, Science 2012,
336, 893 – 897.
[9] J. Nakamura, Y. Choi, T. Fujitani, Top. Catal. 2003, 22, 277 – 285.
[10] Y. Yang, J. Evans, J. A. Rodriguez, M. G. White, P. Liu, Phys.
Chem. Chem. Phys. 2010, 12, 9909 – 9917.
[11] H. Nakano, I. Nakamura, T. Fujitani, J. Nakamura, J. Phys.
Chem. B 2001, 105, 1355 – 1365.
[12] G. C. Wang, Y. Morikawa, T. Matsumoto, J. Nakamura, J. Phys.
Chem. B 2006, 110, 9 – 11.
[13] Y. Yang, C. A. Mims, R. S. Disselkamp, J.-H. Kwak, C. H. F.
Peden, C. T. Campbell, J. Phys. Chem. C 2010, 114, 17205 –
17211.
In summary, we found that 13CO2 desorption during
formate decomposition in the steady-state reaction of
H13COOH with O2 on Cu(110) is a thermal non-equilibrium
process, in which the translational energy (as low as
ꢁ 100 meV) of 13CO2 is independent of Ts. DFT calculations
of formate decomposition on Cu(110) reveal that CO2 has
[14] G. J. Millar, C. H. Rochester, J. Chem. Soc. Faraday Trans. 1991,
87, 1491 – 1496.
ꢀ
ꢀ
been formed at the TS after cleavages of the Cu O and C H
bonds of adsorbed formate. This suggests that the TS is cooled
down because of the endothermic bond ruptures and that CO2
and Ha/Cu(110) are thermally decoupled at the TS. The
energy difference between TS and FS is thus expected to be
converted to the internal energy (vibration, rotation, and
[15] M. Bowker, S. Haq, R. Holroyd, P. M. Parlett, S. Poulston, N.
Richardson, J. Chem. Soc. Faraday Trans. 1996, 92, 4683 – 4686.
[16] E. Iglesia, M. Boudart, J. Phys. Chem. 1986, 90, 5272 – 5274.
[17] B. Mondal, J. Song, F. Neese, S. Ye, Curr. Opin. Chem. Biol. 2015,
25, 103 – 109.
4
www.angewandte.org
ꢀ 2017 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
Angew. Chem. Int. Ed. 2017, 56, 1 – 6
These are not the final page numbers!