Carbon nanoparticles as visible-light photocatalysts for efficient CO 2 conversion and beyond

Article abstract of DOI:10.1021/ja200804h

Increasing atmospheric CO₂ levels have generated much concern, driving the ongoing carbon sequestration effort. A compelling CO₂ sequestration option is its photocatalytic conversion to hydrocarbons, for which the use of solar irradiation represents an ultimate solution. Here we report a new strategy of using surface-functionalized small carbon nanoparticles to harvest visible photons for subsequent charge separation on the particle surface in order to drive the efficient photocatalytic process. The aqueous solubility of the catalysts enables photoreduction under more desirable homogeneous reaction conditions. Beyond CO₂ conversion, the nanoscale carbon-based photocatalysts are also useful for the photogeneration of H ₂ from water under similar conditions.

Full text of DOI:10.1021/ja200804h

COMMUNICATION
pubs.acs.org/JACS
Carbon Nanoparticles as Visible-Light Photocatalysts for Efficient CO2
Conversion and Beyond
Li Cao,† Sushant Sahu,† Parambath Anilkumar,† Christopher E. Bunker,*,‡ Juan Xu,† K. A. Shiral Fernando,§
Ping Wang,† Elena A. Guliants,§ Kenneth N. Tackett, II,† and Ya-Ping Sun*,†
†Department of Chemistry and Laboratory for Emerging Materials and Technology, Clemson University, Clemson, South Carolina
29634, United States
‡Propulsion Directorate, Air Force Research Laboratory, Wright-Patterson Air Force Base, Ohio 45433, United States
§Sensors Technology Office, University of Dayton Research Institute, Dayton, Ohio 45469, United States
ABSTRACT: Increasing atmospheric CO2 levels have
generated much concern, driving the ongoing carbon se-
questration effort. A compelling CO2 sequestration option
is its photocatalytic conversion to hydrocarbons, for which
the use of solar irradiation represents an ultimate solution.
Here we report a new strategy of using surface-functiona-
lized small carbon nanoparticles to harvest visible photons
for subsequent charge separation on the particle surface in
order to drive the efficient photocatalytic process. The
aqueous solubility of the catalysts enables photoreduction
under more desirable homogeneous reaction conditions.
Beyond CO2 conversion, the nanoscale carbon-based
photocatalysts are also useful for the photogeneration of
H2 from water under similar conditions.
Figure 1. Aqueous soluble PEG-functionalized carbon nanoparticles
before (left, fluorescent) and after (right, photocatalytic for CO2
reduction in aqueous solution) coating with metal.
emission in the visible spectral region that may also extend into
the near-IR.7,8 The fluorescence emissions are attributed to
radiative combinations of electrons and holes confined on the
carbon nanoparticle surface (Figure 1),7,9,10 phenomenologically
resembling those found in nanoscale semiconductors. In this
study, the carbon particle surface was also doped with gold or
platinum to concentrate the photogenerated electrons to enable
more efficient CO2 reduction.
Carbon powders containing sub-10 nm nanoparticles were
refluxed in aqueous nitric acid (2.6 M) for 12 h, dialyzed in
membrane tubing (pore size equivalent to MW ∼ 1000) against
fresh water, and then centrifuged to retain the supernatant as an
aqueous suspension of small carbon nanoparticles bearing sur-
face carboxylic acid moieties.7 The particles recovered from the
suspension were functionalized with oligomeric poly(ethylene
glycol) diamine (PEG1500N; Figure 1) in a classical amidation
reaction that involved the acylation of the particle-surface
carboxylic acids in neat thionyl chloride as an intermediate
step.7,10 The functionalized carbon nanoparticles were readily
dissolved in water to form stable aqueous solutions of various
concentrations, with solution colors ranging from yellow at low
concentrations to black at very high concentrations. A UVꢀvis
absorption spectrum of the solution used in subsequent photo-
lysis for gold coating is shown in Figure 2.
he significant rise in atmospheric CO2 levels due to the
T
combustion of hydrocarbon fuels has generated much con-
cern. Among various CO2 sequestration options, a compelling
approach is photocatalytic conversion to recycle CO2 back to
hydrocarbon fuels, for which the use of solar irradiation may
represent an ultimate solution. However, there are major chal-
lenges in finding potent photocatalysts for such a task.1ꢀ3
Nanoscale wide-band-gap semiconductors such as titanium
dioxide (TiO2) and cadmium sulfide (CdS) were originally used
and have since been quite popular in CO2 photoreduction and
related photocatalytic reactions,1ꢀ4 but their limitations in terms
of the requirement for UV excitation and generally low conver-
sion efficiencies have also become evident.2,3 Various improve-
ment strategies, including the use of TiO2 nanotubes and dye
sensitization to extend the absorption of the photocatalysts into
the visible region, have been developed.5,6 Here we report a new
approach involving the use of surface-functionalized small carbon
nanoparticles to harvest visible photons for subsequent charge
separation on the particle surface in order to drive the photo-
catalytic process. The aqueous solubility of the catalysts enables
photoreduction under essentially homogeneous reaction condi-
tions. Beyond CO2 conversion, the nanoscale carbon-based
photocatalysts are also useful to the photogeneration of H2 from
water under similar experimental conditions.
The functionalized particles were coated with gold or plati-
num by simple solution-phase photolysis, as these metals are
known in the literature for their photocatalytically enhanced
reduction reactions.11ꢀ13 The photoexcited carbon particles
acted as electron donors to reduce the gold or platinum salt
It is now known that small carbon nanoparticles can be
covalently functionalized to give them strong absorption and
Received: January 26, 2011
Published: March 14, 2011
r
2011 American Chemical Society
4754
dx.doi.org/10.1021/ja200804h J. Am. Chem. Soc. 2011, 133, 4754–4757
|

Journal of the American Chemical Society
COMMUNICATION
Figure 3. TEM images (inset: high-resolution) of the functionalized
carbon nanoparticles with gold coating.
Figure 2. Absorption spectra of the functionalized carbon nanoparticles
in aqueous solution before (dashed line) and after (solid line) coating
with gold.
(HAuCl4 or H2PtCl6, respectively) to the corresponding metal
on the particle surface, similar to what was observed in the
photoreduction of aqueous silver cation into silver metal.9 The
photolysis conditions were such that only the functionalized
particles were allowed to get photons (405ꢀ720 nm for gold and
425ꢀ720 nm for platinum, to exclude not only the metal salts but
also any residual impurities in the salt samples from photo-
excitation), thus eliminating the possibility of any “free” (i.e., not
associated with carbon nanoparticles) gold or platinum nano-
particles. For gold, the progressive photoreductive coating of the
carbon particles was monitored by the emergence and rapid
increase of the surface plasmon absorption of the coated nano-
scale gold (Figure 2, corresponding to a gold-to-core carbon
molar ratio of ∼1:38). No gold formation was found in repeated
control experiments in the absence of the functionalized carbon
nanoparticles, as expected.
Transmission electron microscopy (TEM) imaging, which
took advantage of the high electron density and crystalline fringes
of the coated metal, confirmed the gold coating (Figure 3), as did
results from energy-dispersive X-ray (EDX) analyses. The aver-
age diameter of the gold-coated particles was ∼9 nm according to
statistical analysis of multiple TEM images, and the size distribu-
tion standard deviation was ∼3 nm.
Figure 4. Absorption spectra of the functionalized carbon nanoparticles
in aqueous solution before (solid line) and after (dashed line) platinum
coating. Inset: TEM images of the coated particles.
For platinum coating of the functionalized carbon nanoparti-
cles, the convenient monitoring of a plasmon absorption was not
possible (such an absorption was absent in the concerned
wavelength region; see Figure 4), so the coating was again con-
firmed by TEM imaging (Figure 4) and the associated EDX
results.
photoproduct in the aqueous solution post-photoirradiation was
harvested by simple distillation, from which a new aqueous
solution containing formic acid was collected and then basified
to pH 12 with aqueous NaOH to convert the acid to sodium
formate. Upon the evaporation of water, the salt was analyzed by
1H and 13C NMR spectroscopy to confirm and quantify the
amount of formate using sodium acetate as an internal standard
(Figure 5).
The formation of formic acid from CO2 photoconversion was
confirmed by 13C-labeling experiments using aqueous NaH13-
CO3 (at the saturated CO2-equivalent concentration) as the feed
under neutral and acidic pH conditions (mimicking aqueous
CO2). The 13C-labeled formic acid thus obtained from the
same photocatalytic reduction was readily detected by 13C
NMR spectroscopy with only a few hundred scans (vs >20 000
scans required for regular formic acid from aqueous CO2 or
NaHCO3).
The functionalized carbon nanoparticles with gold or plati-
num coating were used as photocatalysts for the reduction of
CO2. In a typical experiment, an aqueous solution of the gold-
coated particles (corresponding to the UVꢀvis absorption
spectrum in Figure 1) was added to an optical cell in the
photolysis setup, after which the solution was purged with
CO2 gas toward saturation at ambient temperature. The resulting
solution in the optical cell was photoirradiated with visible light
(425ꢀ720 nm) for 5 h. As known in the literature, the photo-
reduction of CO2 generally yields formic acid as a significant
product.1ꢀ3 In this study, the formic acid formed as a volatile
4755
dx.doi.org/10.1021/ja200804h |J. Am. Chem. Soc. 2011, 133, 4754–4757

Journal of the American Chemical Society
COMMUNICATION
Figure 6. Schematic diagram of the setup used for the detection of
photocatalytically generated H2.
the obviously important advantage of strong visible absorption,
other distinctive features of the nanoscale carbon-based photo-
catalysts include their aqueous solubility for the photoconversion
under homogeneous reaction conditions and the confinement of
the photoinduced charge separation to the particle surface, thus
facilitating more efficient electron harvesting by the coated gold
or platinum as the cocatalyst in the CO2 conversion.
Figure 5. 1H NMR spectra of (a) HCOONa harvested from the CO2
photocatalytic reduction and (b) CH3COONa as an internal standard
in D2O.
Beyond CO2 photoreduction, the same functionalized carbon
nanoparticles with gold or platinum coating could be used as
photocatalysts for H2 generation from water.4,14,15 Experimen-
tally, the same aqueous solution of the gold- or platinum-coated
nanoparticles in the absence of CO2 was irradiated under the
same conditions as those described above. The top of the sealed
optical cell for the photoirradiation was connected to a H2
indicator containing an aqueous CuSO4 solution (bright blue
color) (Figure 6). H2 generation in the photocatalytic reaction of
water was detected by a color change and precipitation in the H2
indicator (due to reduction of Cu2þ to Cuþ and Cu by H2 at
30 °C). A more quantitative determination of the amount of
photogenerated H2 from water and the associated reaction
quantum yields is in progress. According to literature results,
there are overlaps in terms of the configuration in photocatalysts
for CO2 reduction and water splitting.2ꢀ4,12,14 In fact, one of the
mechanistic explanations for CO2 reduction in aqueous solution
is based on a two-step process, namely, initial photocatalytic
generation of atomic H followed by reduction of CO2.12
In summary, surface-modified small carbon nanoparticles
represent a highly promising new platform for visible-light
photocatalysis of CO2 conversion and beyond. The photocata-
lysts are stable in the photoreduction reactions. The photocon-
version efficiencies likely benefit from the solubility of the
catalysts (which allows for homogeneous reaction conditions
in solution), surface confinement of the photoinduced charge
separation and trapping, and the straightforward doping of metal
cocatalysts. There is obviously more room for improvement in
the further development of these carbon-based photocatalysts.
The quantification of the photoproduct was used to estimate
quantum yields for photocatalytic reactions of CO2 under the
specific experimental conditions.6 For the reduction to formic
acid only, the estimated quantum yield was ∼0.3%, which was
likely lower than the overall quantum yield for the CO2 conver-
sion (because of conversion to other photoproducts such as
methanol;13 see Figure 1). Nevertheless, even for the production
of formic acid alone, the observed quantum yield is already at the
higher end of the available literature values for reactions with a
variety of photocatalysts under many different conditions.3 As a
more direct comparison, the use of suspended TiO2 nanoparti-
cles (Degussa P25) as photocatalysts with UV light irradiation
(through a 350 nm cutoff filter) in the same photolysis setup
resulted in quantum yields that were an order of magnitude lower
(also for formic acid only).
For the platinum-coated photocatalysts in the CO2 conversion
reaction under otherwise identical experimental conditions, the
results were generally similar to those with the gold-coated ones
described above, though more quantitative comparisons, includ-
ing those for effects of varying amounts of metal coating, are still
being pursued.
Mechanistically, photoexcitation of the surface-passivated small
carbon nanoparticles likely results in charge separation to form
surface-confined electrons and holes (Figure 1),9,10 on which
significant experimental evidence has already become available.
For example, the bright fluorescence emissions, which can be
attributed to radiative recombinations on the particle surface,
could be quenched by both electron donors and acceptors in an
equally efficient fashion.9 The emissions were also essentially
diminished by the gold or platinum coating, as the coated metal
was designed to soak up the surface-confined electrons, disrupt-
ing the radiative recombinations. Therefore, the functionalized
carbon nanoparticles apparently served the function of harvest-
ing visible photons to drive the photoreduction process, with the
particle surface defects facilitating the charge separation by
trapping the separated electrons and holes, phenomenologically
similar to what occurs for metal-coated semiconductor nanopar-
ticles (e.g., platinum-coated CdS or TiO2).11ꢀ13 In addition to
’ AUTHOR INFORMATION
Corresponding Author
christopher.bunker@wpafb.af.mil; syaping@clemson.edu
’ ACKNOWLEDGMENT
This work was made possible by support from the Air Force
Research Laboratory through the Nanoenergetics Program.
Funding from NSF (Y.-P.S.) and ACS-PRF (Y.-P.S.) for partial
support of this work is also acknowledged. L.C. was supported by
4756
dx.doi.org/10.1021/ja200804h |J. Am. Chem. Soc. 2011, 133, 4754–4757

Journal of the American Chemical Society
COMMUNICATION
a Susan G. Komen for the Cure Postdoctoral Fellowship. J.X. was
on leave from Changzhou University in Changzhou, China, with
a visiting scholarship provided by the National Natural Science
Foundation of China and the Abroad Scholarship of Jiangsu
Province. P.W. was a participant in the Palmetto Academy, an
education-training program managed by the South Carolina
Space Grant Consortium.
’ REFERENCES
(1) Inoue, T.; Fujishima, A.; Konishi, S.; Honda, K. Nature 1979,
277, 637.
(2) Usubharatana, P.; McMartin, D.; Veawab, A.; Tontiwachwuthikul,
P. Ind. Eng. Chem. Res. 2006, 45, 2558.
(3) Roy, S. C.; Varghese, O. K.; Paulose, M.; Grimes, C. A. ACS Nano
2010, 4, 1259.
(4) (a) Kudo, A.; Miseki, Y. Chem. Soc. Rev. 2009, 38, 253. (b) Chen,
X.; Shen, S.; Guo, L.; Mao, S. S. Chem. Rev. 2010, 110, 6503.
(5) Varghese, O. K.; Paulose, M.; LaTempa, T. J.; Grimes, C. A.
Nano Lett. 2009, 9, 731.
(6) (a) Maeda, K.; Eguchi, M.; Youngblood, W. J.; Mallouk, T. E.
Chem. Mater. 2008, 20, 6770. (b) Maeda, K.; Eguchi, M.; Lee, S.-H. A.;
Youngblood, W. J.; Hata, H.; Mallouk, T. E. J. Phys. Chem. C 2009,
131, 7962.
(7) Sun, Y.-P.; Zhou, B.; Lin, Y.; Wang, W.; Fernando, K. A. S.;
Pathak, P.; Harruff, B. A.; Wang, X.; Wang, H.; Luo, P. G.; Yang, H.;
Chen, B.; Veca, L. M.; Xie, S.-Y. J. Am. Chem. Soc. 2006, 128, 7756.
(8) (a) Bourlinos, A. B.; Stassinopoulos, A.; Anglos, D.; Zboril, R.;
Karakassides, M.; Giannelis, E. P. Small 2008, 4, 455. (b) Peng, H.;
Travas-Sejdic, J. Chem. Mater. 2009, 21, 5563. (c) Jiang, H. Q.; Chen, F.;
Kagally, M. G.; Denes, F. S. Langmuir 2010, 26, 1991.
(9) Wang, X.; Cao, L.; Lu, F.; Meziani, M. J.; Li, H.; Qi, G.; Zhou, B.;
Harruff, B. A.; Sun, Y.-P. Chem. Commun. 2009, 3774.
(10) Wang, X.; Cao, L.; Yang, S.-T.; Lu, F.; Meziani, M. J.; Tian, L.;
Sun, K. W.; Bloodgood, M. A.; Sun, Y.-P. Angew Chem., Int. Ed. 2010,
49, 5310.
(11) (a) Coren, Z.; Willner, I. J. Phys. Chem. 1990, 94, 3784. (b)
Zhang, Q. H.; Han, W. D.; Hong, Y. J.; Yu, J. G. Catal. Today 2009,
148, 335.
(12) Fox, M. A.; Dulay, M. T. Chem. Rev. 1993, 93, 341.
(13) Pathak, P.; Meziani, M. J.; Castillo, L.; Sun, Y.-P. Green Chem.
2005, 7, 667.
(14) Yan, H.; Yang, J.; Ma, G.; Wu, G.; Zong, X.; Lei, Z.; Shi, J.; Li, C.
J. Catal. 2008, 254, 198.
(15) Maeda, K.; Domen, K. J. Phys. Chem. Lett. 2010, 1, 2655.
4757
dx.doi.org/10.1021/ja200804h |J. Am. Chem. Soc. 2011, 133, 4754–4757