On-demand hydrogen production from formic acid by light-active dinuclear iridium catalysts

Article abstract of DOI:10.1039/d0cc00704h

Light-active dinuclear iridium pentahydride complexes catalyze the decomposition of formic acid to generate H₂ by irradiation (λ =395 nm) under ambient temperature and base-free conditions. The catalyst activity is sensitive to light producing

Full text of DOI:10.1039/d0cc00704h

ChemComm
View Article Online
View Journal
COMMUNICATION
On-demand hydrogen production from formic
acid by light-active dinuclear iridium catalysts†
Cite this:DOI: 10.1039/d0cc00704h
Yuki Sofue, Kotohiro Nomura
and Akiko Inagaki
*
Received 27th January 2020,
Accepted 18th March 2020
DOI: 10.1039/d0cc00704h
rsc.li/chemcomm
Light-active dinuclear iridium pentahydride complexes catalyze the generated hydrogen, the production of an on-demand hydro-
decomposition of formic acid to generate H2 by irradiation (k =395 nm) gen supply through appropriate catalysis control is a matter of
under ambient temperature and base-free conditions. The catalyst special importance. However, control of the catalyst reactivity
activity is sensitive to light producing H2 under light irradiation, but by external stimuli is difficult, and understandably, switching
with no reaction being observed in the absence of light or when the of the reaction remains an unexplored area.11,12 Previously, we
light is switched off, thereby demonstrating the clear ON/OFF have reported the syntheses and photocatalytic activities of
switching ability of this system. Importantly, the dinuclear structure multinuclear metal catalysts composed of visible-light absorbing
of the catalyst is sufficiently stable to be maintained under the units.13–18 Building on this previous work, we herein report our
catalytic conditions employed herein.
study into on-demand hydrogen production through the catalytic
decomposition of FA by light-active catalysts using light as an
As a result of the continuous rise in the global energy demand, external stimulus.
in addition to the negative impact of fossil fuel consumption on
climate change, the discovery and development of alternative
energy resources and more sustainable technologies has received
growing attention.1–3 More specifically, the efficient utilization of
renewable resources such as biomass, wind, and sunlight has
received considerable attention in recent years, and as a result,
(1)
new technologies for energy storage are in strong demand. For
Following our report into the syntheses of novel light-active
example, fuel cells have shown promise for the generation of dinuclear iridium pentahydride complexes,19 we additionally
electricity from hydrogen for powering vehicles and/or power synthesized catalysts 1d and 1e bearing 3,5-dimethylphenyl and
units.4 In terms of the safety concerns associated with the use of 4-dimethylamino-3,5-dimethylpheny substituents, respectively,
compressed hydrogen gas, which is highly flammable and on the phosphorous atom of the BINAP ligand (ESI,† Fig. S1–S7).
difficult to store, formic acid (FA) has been accepted as a Both complexes showed similar spectroscopic features to the
suitable liquid hydrogen carrier.5,6 FA has a high hydrogen previously described catalysts in the 1H NMR and 31P NMR spectra,
content (4.4 wt% hydrogen; 5.22 MJ kgꢀ1), and it presents many in addition to their UV-vis absorption spectra (Fig. S8, ESI†).
advantageous physical and chemical properties, such as being a
The catalytic FA decomposition reaction was conducted in a
low toxic liquid (bp = 101 1C), which allows its facile transportation, Pyrex 5f NMR tube containing a MeOH solution of the dinuclear
and a feasible hydrogenation/dehydrogenation cycle that can be iridium catalyst (1a–1e, 3 mol%, Fig. 1) at ambient temperature
achieved in the presence of suitable catalysts.
under light irradiation (3 W LED lamp, l = 395 nm) (eqn (1)). All
To date, a wide variety of homogeneous and heterogeneous catalysts were active in the catalytic reaction, and the obtained
catalysts for the selective generation of hydrogen (with low results are summarized in Table 1, whereby each value is an
CO contents) through the catalytic decomposition of FA have average of 3 experiments. The reaction was found to proceed
been reported.7–12 In terms of the efficient utilization of the under light irradiation, but was sluggish under the corresponding
dark conditions, thereby confirming a sharp ON/OFF switching
behavior (entries 1 and 2). Catalyst 1d, bearing the 3,5-dimethyl-
Minami-Osawa, Hachioji city, 192-0397, Tokyo, Japan. E-mail: ainagaki@tmu.ac.jp
phenyl substituent, was most active in this reaction presenting a
† Electronic supplementary information (ESI) available: NMR spectra and structural
contrasting trend to the activity of alkene hydrogenation19
details. CCDC 1980047. For ESI and crystallographic data in CIF or other electronic
format see DOI: 10.1039/d0cc00704h
(entries 1, 3–6). In order to check the wavelength dependence,
Chem. Commun.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Communication
ChemComm
Table 2 Solvent effect in the catalytic decomposition of FA by 1d and 1e
(X = BF4) in a 7.8 mL Pyrex glass tubea
Entry
Time/h
Cat.
Solvent
H2b/mL
FA conv.c/%
TON
1
2
24
24
24
24
8
1d
1d
1d
1d
1d
1e
MeOH
H2O
HCOOH
CH3NO2
CH3NO2
H2O
3.06
0.89
0.72
6.44
3.75
3.57
12
3
3
28
96
19
125
37
28
280
102
184
3
4d
5de
6
24
a
Catalytic FA (1000 eq. vs. catalyst) decomposition conducted in a 7.8 mL
Pyrex screw glass tubes under a N2 atmosphere with 395 nm irradiation
b
and with stirring unless otherwise noted. H2 volume was quantified by
c
TCD-GC. Each value is an average of 3 experiments. Calcd based on the
% of (mol H2 production)/(mol starting FA). d A 14.2 mL Pyrex screw glass
e
tube was used. 100 eq. vs. catalyst was used.
is sufficiently high to conduct the reaction iteratively. Upon
investigation of the solvent effect, nitromethane exhibited a
higher reactivity compared to other solvents such as methanol,
H2O, and neat FA (Table 2). It was also found that the catalysts
can promote FA decomposition under the base-free conditions
in water or neat FA. Catalyst 1e can present a higher activity in
H2O than 1d, likely due to the greater solubility of 1e.
Fig. 1 Light active dinuclear iridium complexes 1a–1e (X = BF4).
Table 1 Catalytic FA decomposition activities of 1a–1e (X = BF4)a
As mentioned above, it is notable that the sharp ON/OFF
switching behavior was observed by simply turning on and off
the LED lamp. To confirm the switching behavior of the system,
a solution of 1d in MeOH (0.1 mol% vs. FA) was exposed to
light, and dark conditions of 12 and 6 h intervals, respectively
(Fig. 2). After each interval, the volume of H2 gas produced was
quantified, and the solution was degassed using the freeze–
thaw–pump method to initialize the system. In addition to the
ON/OFF switching behavior, it was obvious that the catalyst
maintained its activity after each dark period, and with similar
reaction rates being observed upon commencing irradiation
once again. These results imply that on-demand H2 gas production
could be possible using this system, thereby indicating its potential
to contribute to an efficient and energy-saving supply of H2 for
fuel cells.
Entry
Catalyst
H2b/mL
FA conversionc/%
TON
1
2
3
4
5
6
7
8
1a
1a, dark
1b
1c
1d
1e
0.23
0.01
0.27
0.17
0.39
0.23
0.24
0.18
27
0
9.3
0
32
21
47
33
29
22
10.1
6.4
16.2
11.3
9.1
7.1
1d, 365 nm
1d, 425 nm
a
Catalytic FA decomposition conducted in MeOH-d4 in 5f NMR tubes
under a N2 atmosphere with 395 nm irradiation (3 W LED lamp) unless
b
otherwise noted. H2 volume was quantified by TCD-GC. Each value is an
c
average of 3 experiments. Calcd based on the % of (mol H2 production)/
(mol starting FA).
1
When the reaction of 1a was followed by H NMR spectro-
the reaction was conducted under different wavelength irradia-
tion (entries 7 and 8, 3 W LED lamp, l = 365 nm and 425 nm,
respectively). Irradiation under 365 nm resulted in the lower
activity (entries 7 vs. 5) than that obtained under 395 nm due to
the decomposition of the catalyst although e at 365 nm is much
higher than that of 395 nm (e365 = 12621 vs. e395 = 4803 for 1d).20
Irradiation under 425 nm also gave lower activity (entries 8 vs. 5)
according to the less absorption coefficient of 1d at 425 nm
(e425 = 2702, e395 = 4803).
scopy, two intermediate species Aa and Ba (intermediates A and
B formed from 1a, respectively, Fig. 3) were observed in the
These reactions were also scaled up in a glass tube (Table 2,
7.8- or 14.2 mL volume) with stirring to give higher turnover
numbers (TONs). Although these TONs are lower than those of
the reported catalysts taking part in thermal reactions,7,21
catalysts 1a–1e are sufficiently stable to continue generating
H2 even after 24 h. The reaction by 1a in MeOH over 24 h, the
reaction mixture was collected, dried, and the residual solid was
analyzed by 1H NMR spectroscopy to give signals originating
from the corresponding monoformate complex (intermediate A,
vide infra). This result confirmed that the stability of the catalyst
Fig. 2 Total amount of H2 gas formed through ON/OFF switching of the
light source by a MeOH solution of 1d containing FA (1000 eq.).
Chem. Commun.
This journal is © The Royal Society of Chemistry 2020

View Article Online
ChemComm
Communication
of that of the parent complex 1a. Although the reason for this
difference is currently unclear, it is possible that the lower
absorption coefficient of Aa to 1a at 395 nm may affect the
catalyst reactivity (Fig. S15, ESI†). More specifically, when com-
paring the UV-vis absorption spectra of 1a, Aa, and Ba, the
absorption coefficients of the intermediates at 395 nm were
found to be less than 1/2 of that of the parent complex (i.e.,
e395 Mꢀ1 cmꢀ1 = 1884 (Aa), 1845 (Ba), 4328 (1a)).
(2)
(3)
Fig. 3 Time course of the NMR spectral progression in the reaction of 1a
with FA (30 eq.) in THF-d8 under irr. at 395 nm.
When the isolated intermediate Ba was independently exposed to
irradiation, slight conversion to Aa was observed under irradiation
(18% conversion after 6 h of irradiation). These results indicate the
presence of an equilibrium between A and B, which slightly
shifts toward A upon irradiation. Based on the above observations,
a plausible reaction mechanism was proposed, as depicted in
Scheme 1. Starting from complex 1, reaction with 1 eq. of FA gives
the monoformate complex A with H2 gas discharge (steps (i) and
(ii)).23–25 A second reaction with FA gives the bisformate complex B
as the resting state. CO2 elimination from monoformate complex A
via C regenerates the starting catalyst to complete the catalytic
cycle (steps (iii) and (iv)). However, whether the CO2 elimination
process takes place via a dinuclear process or through b-H
elimination on the single Ir center, remains unclear at this
moment.26–28 Since complex 1 and intermediate A did not react
with FA without irradiation, it is apparent that direct protonation
reaction mixture under irradiation at 395 nm (A : B = 0.17 : 0.83
after 7.0 h). In the reactions catalyzed by the other catalysts, i.e.,
1b–1e, only the corresponding intermediate A, but not B, was
observed. It should be noted that the relationship between the
A to B ratio and the catalytic activity is not clear at this moment.
Furthermore, examination of the ESI-MS spectrum of the reaction
mixture after the photocatalysis showed signals corresponding
to the monoformate (m/z 1791) and bisformate (m/z 1835) com-
plexes together with the corresponding mono-oxidized species
(Fig. S9, ESI†). Intermediate species Aa and Ba were independently
synthesized in a quantitative reaction between FA and 1a (eqn (2)
and (3)).22 Indeed, the isolated species presented identical spectro-
scopic features to those observed in the NMR spectra during the
reaction. However, the 13C NMR spectrum of Aa displayed broad
signals even at ꢀ30 1C, and so the corresponding formate complex
with ligand c (Ac) was measured (Fig. S10–S13, ESI†), and this
presented a characteristic carbon signal corresponding to the
m-k,k0-OCHO formate ligand at 175 ppm. The NMR data of Ac
are presented in the ESI,† (Fig. S10–S13).
A single crystal of Bb suitable for analysis by X-ray diffraction
study was then obtained from toluene at ꢀ30 1C, and its molecular
structure was determined. The ORTEP diagram of the cationic unit
of Bb is shown in Fig. S14 (ESI†), where a single BF4 anion and two
toluene molecules per Ir2 unit were present. This complex
possessed m-k,k0-OCHO formate ligands between the two iridium
centers, and the Ir–Ir bond distance was lengthened from
2.5174(7) Å in 1b to 3.0733(7) Å in Bb. This lengthening of the
Ir–Ir bond may affect the photophysical properties of the complex
as a photocatalyst.
The independently synthesized Aa (eqn (2) and (3)) was then
employed as a catalyst under the conditions outlined in Table 1,
and the reaction proceeded similarly under irradiated conditions,
although the reactivity was about 1/3 (0.07 mL (Aa) vs. 0.23 mL (1a))
Scheme 1 Plausible reaction mechanism for the catalytic decomposition
of FA. Steps requiring light irradiation are indicated in red.
Chem. Commun.
This journal is © The Royal Society of Chemistry 2020

View Article Online
Communication
ChemComm
L. E. Manzer, T. J. Marks, K. Morokuma, K. M. Nicholas, R. Periana,
L. Que, J. Rostrup-Nielson, W. M. H. Sachtler, L. D. Schmidt, A. Sen,
G. A. Somorjai, P. C. Stair, B. R. Stults and W. Tumas, Chem. Rev.,
2001, 101, 953–996.
of the hydride ligand to generate H2 does not occur in this
system.29,30 Most importantly, light is required to drive the processes
of (i) and (iii), (iv).
4 N. Armaroli and V. Balzani, ChemSusChem, 2011, 4, 21–36.
In summary, we herein demonstrated the ON/OFF switching
behavior of light-active dinuclear iridium hydride complexes
in the catalytic decomposition of formic acid (FA), resulting in
on-demand H2 production using light as an external stimulus.
Among the various catalysts examined, 1d bearing a 3,5-dimethyl-
phenyl substituent on the phosphorous atom of the BINAP ligand
exhibited the highest catalytic activity under the base-free con-
ditions employed herein. In addition, the catalysts were stable
enough to conduct the reaction without deterioration even after
continuous switching cycles for 54 h. The highest activity was
observed in nitromethane solution, and the catalytic process
also proceeded in water or a neat FA solution. Furthermore,
careful analysis allowed observation of the mono- and bisformate
intermediates, and these were independently synthesized from the
quantitative reactions. The isolated intermediates exhibited catalytic
activity under irradiated conditions, thereby confirming their
contribution to the catalytic cycle. These results are expected to
contribute to the use of FA as a liquid hydrogen carrier for
energy storage applications.
´
5 F. Joo, ChemSusChem, 2008, 1, 805–808.
6 T. C. Johnson, D. J. Morris and M. Wills, Chem. Soc. Rev., 2010, 39,
81–88.
7 K. Sordakis, C. Tang, L. K. Vogt, H. Junge, P. J. Dyson, M. Beller and
G. Laurenczy, Chem. Rev., 2018, 118, 372–433.
8 W.-H. Wang, Y. Himeda, J. T. Muckerman, G. F. Manbeck and
E. Fujita, Chem. Rev., 2015, 115, 12936–12973.
9 G. Laurenczy and P. J. Dyson, J. Braz. Chem. Soc., 2014, 25, 2157–2163.
10 M. Iglesias and L. A. Oro, Eur. J. Inorg. Chem., 2018, 2125–2138.
¨
11 A. Boddien, F. Gartner, R. Jackstell, H. Junge, A. Spannenberg,
W. Baumann, R. Ludwig and M. Beller, Angew. Chem., Int. Ed.,
2010, 49, 8993–8996.
¨
12 A. Boddien, B. Loges, F. Gartner, C. Torborg, K. Fumino, H. Junge,
R. Ludwig and M. Beller, J. Am. Chem. Soc., 2010, 132, 8924–8934.
13 S. Shitaya, K. Nomura and A. Inagaki, Chem. Commun., 2019, 55,
5087–5090.
14 S. Shitaya, K. Nomura and A. Inagaki, Dalton Trans., 2018, 47,
12046–12050.
15 Y. Matsusaka, S. Shitaya, K. Nomura and A. Inagaki, Inorg. Chem.,
2017, 56, 1027–1030.
16 S. Kikuchi, K. Saito, M. Akita and A. Inagaki, Organometallics, 2018,
37, 359–366.
17 H. Nitadori, T. Takahashi, A. Inagaki and M. Akita, Inorg. Chem.,
2012, 51, 51–62.
18 S. Phungsripheng, M. Akita and A. Inagaki, Inorg. Chem., 2017, 56,
12996–13006.
19 Y. Sofue, K. Nomura and A. Inagaki, Organometallics, 2019, 38,
2408–2411.
This work was supported by Cooperative Research Program
of ‘‘Network Joint Research Center for Materials and Devices’’
and JSPS KAKENHI (B) Grant Number 16H0412100.
20 Noticeable decomposition of the catalst 1a in CD3OD solution was
confirmed by irradition by 365 nm LED lamp (77% decomposition
(365 nm) vs. 3% decomposition (395 nm) under identical condition).
21 A. Matsunami, S. Kuwata and Y. Kayaki, ACS Catal., 2017, 7,
4479–4484.
22 Other complexes having different substituents can be synthesized in
the same procedure.
Conflicts of interest
There are no conflicts to declare.
23 R. L. Sweany, J. Am. Chem. Soc., 1981, 103, 2410–2412.
24 A. J. Esswein, A. S. Veige and D. G. Nocera, J. Am. Chem. Soc., 2005,
127, 16641–16651.
Notes and references
1 A. M. Appel, J. E. Bercaw, A. B. Bocarsly, H. Dobbek, D. L. DuBois,
M. Dupuis, J. G. Ferry, E. Fujita, R. Hille, P. J. A. Kenis, C. A. Kerfeld, 25 R. H. Morris, H. C. Foley, T. S. Targos and G. L. Geoffroy, J. Am.
R. H. Morris, C. H. F. Peden, A. R. Portis, S. W. Ragsdale, T. B. Chem. Soc., 1981, 103, 7337–7339.
Rauchfuss, J. N. H. Reek, L. C. Seefeldt, R. K. Thauer and G. L. 26 Y. Gao, J. K. Kuncheria, H. A. Jenkins, R. J. Puddephatt and G. P. A.
Waldrop, Chem. Rev., 2013, 113, 6621–6658. Yap, J. Chem. Soc., Dalton Trans., 2000, 3212–3217, DOI: 10.1039/B004234J.
2 M. Aresta, A. Dibenedetto and A. Angelini, Chem. Rev., 2014, 114, 27 J. J. A. Celaje, Z. Lu, E. A. Kedzie, N. J. Terrile, J. N. Lo and T. J.
1709–1742. Williams, Nat. Commun., 2016, 7, 11308.
3 H. Arakawa, M. Aresta, J. N. Armor, M. A. Barteau, E. J. Beckman, 28 M. Iguchi, H. Zhong, Y. Himeda and H. Kawanami, Chem. – Eur. J.,
A. T. Bell, J. E. Bercaw, C. Creutz, E. Dinjus, D. A. Dixon, K. Domen, 2017, 23, 17017–17021.
D. L. DuBois, J. Eckert, E. Fujita, D. H. Gibson, W. A. Goddard, 29 J. Li, J. Li, D. Zhang and C. Liu, ACS Catal., 2016, 6, 4746–4754.
D. W. Goodman, J. Keller, G. J. Kubas, H. H. Kung, J. E. Lyons, 30 Z. Treigerman and Y. Sasson, ChemistrySelect, 2017, 2, 5816–5823.
Chem. Commun.
This journal is © The Royal Society of Chemistry 2020