Chromium(0) nanoparticles as effective catalyst for the conversion of glucose into 5-hydroxymethylfurfural

Article abstract of DOI:10.1002/cssc.201200795

It's nano: Small and uniform chromium nanoparticles, either preformed or generated in situ, effectively catalyze the conversion of glucose into 5-hydroxymethyl furfural. The results compare favorably with those achieved by using a catalyst system based on divalent CrCl₂ in ionic liquids (ILs). In addition, the chromium nanoparticles are found in the CrCl ₂/IL system, and the implications of their presence in that system is investigated. Copyright

Full text of DOI:10.1002/cssc.201200795

CHEMSUSCHEM
COMMUNICATIONS
DOI: 10.1002/cssc.201200795
Chromium(0) Nanoparticles as Effective Catalyst for the
Conversion of Glucose into 5-Hydroxymethylfurfural
Jianghua He, Yuetao Zhang, and Eugene Y.-X. Chen*[a]
In 2007, a seminal work by Zhang et al. revealed that glucose
can be converted into 5-hydroxymethylfurfural (HMF) in good
yields by using a catalyst system based on divalent CrCl2 in
ionic liquids (ILs), particularly 1-ethyl-3-methylimidazolium
chloride ([EMIM]Cl).[1] This work is significant because (i) glu-
cose, derived from hydrolysis of cellulosic biomass, is a renewa-
ble source of HMF and a more desirable feedstock than the
food-based fructose previously employed for HMF production;
and (ii) the main product of the process, HMF, is a versatile, key
biorefining building block for sustainable chemicals, materials,
and liquid fuels.[2] Currently major efforts are directed at devel-
oping economically viable pathways to convert nonfood bio-
mass into biofuels and/or feedstock chemicals,[3] and the glu-
cose/cellulose-to-HMF conversion process holds great potential
to meet this goal. However, the challenge has been how to ad-
vance the catalyst system so that it becomes technologically
and economically competitive, compared to petroleum refiner-
ies. Since the discovery of the CrCl2/IL catalyst system, a large
number of effective metal catalyst systems have been reported
for this common goal of glucose/cellulose-to-HMF conversion
in ILs; they are mostly halide,[4] but also a few alkoxide,[5] com-
plexes of metals, but all at +n oxidation states and, to the best
of our knowledge, no catalyst system based on metal nanoparti-
cles (M-NPs) has been reported for this process.
systems use ILs as solvent, the IL solvent can also function as
the stabilizer for M0-NPs.[8]
In this Communication, we present several lines of evidence
that suggest that Cr0-NPs, either preformed or generated
in situ, serve as the true catalyst for the conversion of glucose
to HMF in ILs employing Cr(CO)6 (Scheme 1). Also reported is
Scheme 1. General scheme for the conversion of glucose into HMF by Cr0-
NPs.
evidence for the presence of Cr-NPs even in glucose conver-
sions that employ divalent CrCl2 in ILs or in DMF, although de-
tailed investigations indicated that the dominant catalysis in
the CrCl2/IL system is not brought about by these Cr-NPs.
Under the standard conditions employed in this study (i.e.,
0.50 g [EMIM]Cl, 10 mol% precatalyst relative to glucose, T=
1208C), the CrCl2-based benchmark catalyst system gave
a HMF yield of 45% in 6 h or 50% in 24 h (Table S1).[9] In com-
parison, the HMF yield from glucose conversion using Cr(CO)6
was 49% in 6 h or 41% in 24 h, indicating that Cr(CO)6 was
more effective than CrCl2 in the shorter-time run (6 h). Howev-
er, other group 6 metal (Mo and W) carbonyl complexes were
much less effective (Table S1). A benzene derivative prepared
by replacing three carbonyl ligands in Cr(CO)6 with one h6-ben-
zene ligand, giving (h6-C6H6)Cr(CO)3, proved a poor precatalyst
for the glucose conversion, affording HMF in low yields of
ꢀ14% under various reaction conditions (Table S1). Under the
current conditions, group 8 iron carbonyl complexes, Fe(CO)5
and Fe2(CO)9, were not capable of catalyzing this conversion,
nor were carbonyl complexes of groups 7 and 9, Mn2(CO)10
and Co2(CO)8. Focusing on the best-performing carbonyl com-
plex of this series, Cr(CO)6, we investigated the effects of reac-
tion temperature and time, catalyst loading, IL structure, and
additives on the glucose conversion. Figure S3 summarizes the
effects of reaction temperature and time on the HMF yield.
When the conversion was carried out at T=1008C, the HMF
yield increased gradually and reached a maximum of 47%
after 42 h, but the yield dropped sharply to only 4.5% after
48 h. On the other hand, at T=1208C the reaction reached the
We reasoned that the currently unexplored zero-valent
metal carbonyl complexes could be especially attractive for the
glucose-to-HMF conversion in ILs for several reasons. (i) They
are known to be excellent precursors to well-defined M0-NPs
under microwave irradiation[6a] or thermal decomposition (ther-
molysis)[6b] conditions in ILs, thereby offering an interesting op-
portunity to explore new nanocatalysis for this process.
(ii) They are typically less expensive than the corresponding
halide complexes (e.g., the cost of the air-stable, zero-valent
Cr(CO)6 is about half that of the air-sensitive, divalent CrCl2).
(iii) Because small M0-NPs are considered strong Lewis acids,
which is the main reason they typically require stabilization,
with stabilizers ranging from donor polymers or solvents to
anions,[7] they could promote efficient catalysis for the glucose-
to-HMF conversion following the general catalysis scheme pro-
posed for other Lewis acids such as CrClx (x=2,3), that is, acid-
catalyzed isomerization of glucose to fructose followed by de-
hydration to HMF.[1,4] (iv) As homogenous biomass conversion
[a] Dr. J. He, Dr. Y. Zhang, Prof. Dr. E. Y.-X. Chen
Department of Chemistry, Colorado State University
Fort Collins, CO 80523-1872 (USA)
Fax: (+1) 970-491-1801
E-mail: eugene.chen@colostate.edu
Supporting Information for this article is available on the WWW under
http://dx.doi.org/10.1002/cssc.201200795.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2013, 6, 61 – 64 61

CHEMSUSCHEM
COMMUNICATIONS
www.chemsuschem.org
maximum yield of ca. 50% much more quickly (6 h), and main-
tained this yield level during up to 12 h of reaction, after
which the yield began to decline (41% after 24 h). A much
lower HMF yield achieved for the reaction after extended time
is likely due to gradual degradation of HMF at high tempera-
tures.[4f,5a]
ternative hypothesis that the formation of Cr0-NPs was induced
by the TEM beam. Specifically, a sample of Cr(CO)6 in a room-
temperature IL, 1-butyl-3-methylimidazolium tetrafluoroborate
([BMIM]BF4), which received neither microwave irradiation nor
thermolysis pretreatment, was imaged using TEM conditions
identical to those used to image the NPs shown in Figure 1.
We did not observe NPs under these conditions. When the
TEM electron beam was focused on a small area with increased
beam current density, some much larger NPs were formed
(34.5Æ6.6 nm; Figure S7). Notably, the smaller NPs formed
under the real reaction conditions (i.e., the NPs shown in
Figure 1) did not change during such focusing steps.
A critical question is whether or not the Cr0-NPs formed
during the reaction are responsible for the catalysis. To address
this question, we preformed Cr0-NPs via microwave irradiation
of a dispersion of Cr(CO)6 in [BMIM]BF4.[6,9] TEM images shown
in Figure S6 confirmed the formation of uniform NPs (3.6Æ
0.7 nm). Two experimental results that followed are significant.
(i) Glucose conversion by using these preformed Cr0-NPs gave
a noticeably higher HMF yield (54%) than that achieved by di-
rectly employing Cr(CO)6 (49%) under the current standard
conditions. (ii) Catalysis using the preformed NPs reached the
maximum HMF yield much sooner than the 6 h period typically
observed for the reaction when employing the molecular pre-
cursor Cr(CO)6. For example, after only 1 h, Cr0-NPs derived
from thermolysis by heating Cr(CO)6 in DMF (T=1208C for
6 h), [EMIM]Cl (T=1208C for 6 h), [T=EMIM]Cl (2308C for
12 h), and [T=BMIM]BF4 (2308C for 12 h) achieved a maximum
HMF yield of 43%, 50%, 48%, and 49%, respectively, during
glucose conversion carried out in [EMIM]Cl at T=1208C. Over-
all, the above results based on the preformed Cr0-NPs are con-
sistent with the assertion that the NPs are responsible for the
catalysis.
Several interesting initial observations on the Cr(CO)6 system
revealed its unique catalytic behavior. Firstly, unlike for the
CrCl2 system, the HMF yield achieved by Cr(CO)6 was insensi-
tive to the precatalyst loading. Specifically, using the condi-
tions that gave the best HMF yield (T=1208C, t=6–12 h), we
varied the Cr(CO)6 loading from 10 mol% to 1.25 mol% and
observed a nearly constant HMF yield of 50Æ3% (Figure S4).
Secondly, replacing the IL solvent [EMIM]Cl by polar organic
solvents such as DMSO and DMF prohibited the catalysis,
which is in drastic contrast to the CrCl2 system, which gave
comparable yields of HMF in [EMIM]Cl and the above polar or-
ganic solvents.[4i] Thirdly, the addition of one equivalent (rela-
tive to Cr) of an N-heterocyclic carbene (NHC) ligand, 1,3-
bis(2,4,6-trimethylphenyl)imidazol-2-ylidene (IMes), completely
shut down the catalysis. This result is again in sharp contrast
to the CrCl2 system, for which an enhanced HFM yield upon
addition of an NHC ligand has been reported,[4l] but is consis-
tent with catalysis by M0-NPs considering the NHC ligand as
a potent M0-NP poison.[10] Finally, performing the conversion in
the common ammonium salt Et4NCl, which is known to be an
effective stabilizer of M-NPs,[11] was as effective as [EMIM]Cl, re-
quiring either a longer reaction time (45% yield at T=1208C
for 24 h) or higher temperature (49% yield at T=1308C for
9 h).
Because the results described above are consistent with the
catalysis behavior of M0-NPs, we sought direct evidence for the
presence of Cr0-NPs in the glucose conversion system that em-
ployed Cr(CO)6 as the source of catalyst. Specifically, we ana-
lyzed samples withdrawn from the solution at the end of the
conversion reactions by transmission electron microscopy
(TEM).[9] Indeed, TEM images of these post-reaction samples
clearly revealed the presence of small, monodisperse Cr0-NPs
with an average size of NP(100) of 2.3Æ0.4 nm (Figure 1).
Cautioned by a report that some M-NPs can be formed from
metal salts in ILs under accelerated electron-beam irradia-
tion,[12] we carried out a control experiment to examine the al-
To further confirm that the Cr0-NPs truly serve as catalyst, we
performed quantitative poisoning experiments on the glucose-
to-HMF catalysis in the presence of varied amounts of 1,10-
phenathroline (PHEN). Amongst several common poisons in-
vestigated (e.g., Hg, CS2), PHEN has been identified as the
most suitable poison for distinguishing between a M-NP cata-
lyst and a ligand-stabilized molecular or sub-nanometer cluster
catalyst in
a catalytic system operating in conditions
ꢁ1008C.[13] Figure 2 plots the relative HMF yield achieved by
the preformed Cr0-NPs via microwave irradiation versus the
number of equivalents of PHEN (relative to Cr) subsequently
added to the reaction mixture. Figure 2 clearly shows that
a sub-stoichiometric amount of PHEN drastically suppressed
the catalytic activity; the linear, extrapolated portion of the
data yielded an intercept of 0.28, which can be considered as
the calculated equivalent of PHEN (per metal) needed to halt
the catalysis.[13] A similar poisoning experiment performed on
a system with in situ generated Cr0-NPs from thermolysis
showed that the catalytic activity diminished to the back-
ground level (1–4% HMF yield in [EMIM]Cl, without any cata-
lyst) in the presence of a fractional equivalent of PHEN,[9] and
the linear, extrapolated portion of the data yielded an intercept
of 0.20 (Figure S11), comparable to that of the microwave irra-
diation route. Overall, the above quantitative poisoning experi-
Figure 1. TEM images of Cr0-NPs obtained at the end of the glucose conver-
sion reaction in [EMIM]Cl, employing Cr(CO)6 at 1208C for 6 h. The average
size of the NPs(100) is 2.3Æ0.4 nm.
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2013, 6, 61 – 64 62

CHEMSUSCHEM
COMMUNICATIONS
www.chemsuschem.org
sponding metallate complex, but not the Cr-NPs present in the
system.
In conclusion, the zero-valent Cr(CO)6-based catalyst system
proved more effective for the conversion of glucose to HMF
under current conditions ([EMIM]Cl, T=1208C) than the diva-
lent CrCl2-based benchmark catalyst system, especially at low
catalyst loadings. Evidence from different sources, obtained by
in situ, ex situ, and quantitative poisoning experiments, dem-
onstrate for the first time that small, uniform Cr0-NPs, either
preformed via microwave irradiation (3.6Æ0.7 nm) or generat-
ed in situ via thermolysis (2.3Æ0.4) during the reaction, are
active species responsible for the observed catalysis when
using Cr(CO)6 as the precatalyst. Being strong Lewis acids, such
small Cr0-NPs promote efficient catalysis, presumably following
the general mechanism proposed for other Lewis acids such as
CrClx. In view of some of the favorable characteristics of the
Cr(CO)6-derived NP catalyst system, including the relatively low
cost and air-stability of the precatalyst as well as its ability to
maintaining high efficiency levels at low catalyst loadings, the
results reported herein should stimulate future efforts to devel-
op more effective M-NP catalysts for glucose- or related bio-
mass-conversion processes.
Figure 2. Plot of the relative HMF yield vs equiv of the poison PHEN added
to the conversion system in [EMIM]Cl (1208C, 6 h) by Cr0-NPs derived from
microwave irradiation.
ments showed that a sub-stoichiometric amount of PHEN
(<0.3 equiv for both Cr-NP systems) can halt the catalysis, pre-
senting very strong evidence for NP catalysis by the Cr(CO)6/IL
system.
To probe the possibility that Cr-NPs could also be present in
the glucose conversion system employing CrCl2, we studied
the CrCl2 system in a similar manner as we did the Cr(CO)6
system. Intriguingly, TEM images of post-reaction samples
taken from the glucose conversion employing CrCl2 in
[EMIM]Cl or in DMF at T=1208C for 6 h also revealed the pres-
ence of small, uniform NPs, with the average size of NPs(100)
being 3.4Æ0.6 nm for reactions in [EMIM]Cl (Figure S8) or
3.0Æ0.4 nm for reactions in DMF (Figure S9).[14] A control ex-
periment on a sample of CrCl2 in [BMIM]BF4 that was not sub-
ject to the reaction conditions revealed no formation of NPs al-
though, with increased beam current density, some much
larger NPs (24.7Æ7.9 nm) were formed (Figure S10). Worth
noting again is that the small NPs formed under the reaction
conditions did not change upon such treatment. To determine
if the catalysis demonstrated by the CrCl2-based system was
brought about by the presence of small NPs in the glucose
conversion system, or by molecular complexes derived from
CrCl2 in the IL or DMF, we again performed quantitative poi-
soning experiments using PHEN. Behaving differently from the
Cr(CO)6 system, a super-stoichiometric amount of PHEN added
to a preheated solution of CrCl2 at T=1208C for 3 h in
[EMIM]Cl (to generate NPs or other types of active species) was
required to substantially suppress the catalytic activity of the
system. The linear, extrapolated portion of the data yielded an
intercept of 5.0 (Figure S12), which represents the calculated
equivalents of PHEN needed to halt the catalysis; remarkably,
this value is identical to the reported 5 equiv of 2,2’-bipyridine
needed to shut down catalysis by CrCl2 in [EMIM]Cl.[1] The
same poisoning experiment performed on the CrCl2 system in
DMF, which gave a HMF yield comparable to the conversion in
[EMIM]Cl under the current conditions, yielded an intercept of
2.5 (Figure S13). Overall, the above results indicate that Cr-NPs
and molecular active species co-exist in the CrCl2-based cata-
lyst system, but the quantitative poisoning experiment data
seemed indicative of dominant catalysis by CrClx or the corre-
Experimental Section
The Supporting Information contains experimental details, as well
as Tables (3), and Figures (13)referred to in the text.
Acknowledgements
This work was supported by the US Department of Energy–Office
of Basic Energy Sciences, grant DE-FG02-10ER16193.
Keywords: 5-hydroxymethylfurfural
·
ionic
liquids
·
nanocatalysis · nanoparticles · renewable resources
[1] H. Zhao, J. E. Holladay, H. Brown, Z. C. Zhang, Science 2007, 316, 1597–
1600.
[2] Selected recent Reviews: a) A. A. Rosatella, S. P. Simeonov, R. F. M. Frade,
C. A. M. Afonso, Green Chem. 2011, 13, 754–793; b) O. O. James, S.
Maity, L. A. Usman, K. O. Ajanaku, O. O. Ajani, T. O. Siyanbola, S. Sahu, R.
Chaubey, Energy Environ. Sci. 2010, 3, 1833–1850; c) Top Value Added
Chemicals from Biomass (Eds.: T. Werpy, G. Petersen), U.S. Department
of Energy (DOE) report: DOE/GO-102004-1992, 2004.
[3] Selected recent Reviews and examples: a) W. Yang, A. Sen, Chem-
SusChem 2011, 4, 349–352; b) M. J. Climent, A. Corma, S. Iborra, Green
Chem. 2011, 13, 520–540; c) D. M. Alonso, J. Q. Bond, J. A. Dumesic,
Green Chem. 2010, 12, 1493–1513; d) W. Yang, A. Sen, ChemSusChem
2010, 3, 597–603; e) Breaking the Chemical and Engineering Barriers to
Lignocellulosic Biofuels: Next Generation Hydrocarbon Biorefineries (Ed.:
G. W. Huber), U.S. National Science Foundation, Washington, 2008; f) A.
Corma, S. Iborra, A. Velty, Chem. Rev. 2007, 107, 2411–2502; g) G. W.
Huber, A. Corma, Angew. Chem. 2007, 119, 7320–7338; Angew. Chem.
Int. Ed. 2007, 46, 7184–7201; h) J. N. Chheda, G. W. Huber, J. A. Dume-
sic, Angew. Chem. 2007, 119, 7298–7318; Angew. Chem. Int. Ed. 2007,
46, 7164–7183; i) G. W. Huber, S. Iborra, A. Corma, Chem. Rev. 2006, 106,
4044–4098.
[4] Selected recent examples: a) Z. Zhang, Q. Wang, H. Xie, W. Liu, Z. K.
Zhao, ChemSusChem 2011, 4, 131–138; b) B. Kim, J. Jeong, D. Lee, S.
Kim, H.-J. Yoon, Y.-S. Lee, J. K. Cho, Green Chem. 2011, 13, 1503–1506;
c) Y. Zhang, E. A. Pidko, E. J. M. Hensen, Chem. Eur. J. 2011, 17, 5281–
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2013, 6, 61 – 64 63

CHEMSUSCHEM
COMMUNICATIONS
www.chemsuschem.org
5288; d) Y. Zhang, H. Du, X. Qian, E. Y.-X. Chen, Energy Fuels 2010, 24,
2410–2417; e) E. A. Pidko, V. Degirmenci, R. A. van Santen, E. J. M.
Hensen, Angew. Chem. 2010, 122, 2584–2588; Angew. Chem. Int. Ed.
2010, 49, 2530–2534; f) T. Stꢁhlberg, M. G. Sørensen, A. Riisager, Green
Chem. 2010, 12, 321–325; g) F. Tao, H. Song, L. Chou, ChemSusChem
2010, 3, 1298–1303; h) X. Qi, M. Watanabe, T. M. Aida, R. L. Smith, Jr.,
ChemSusChem 2010, 3, 1071–1077; i) J. B. Binder, R. T. Raines, J. Am.
Chem. Soc. 2009, 131, 1979–1985; j) Y. Su, H. M. Brown, X. Huang, X.-D.
Zhou, J. E. Amonette, Z. C. Zhang, Appl. Catal. A 2009, 361, 117–122;
k) S. Hu, Z. Zhang, J. Song, Y. Zhou, B. Han, Green Chem. 2009, 11,
1746–1749; l) G. Yong, Y. Zhang, J. Y. Ying, Angew. Chem. 2008, 120,
9485–9488; Angew. Chem. Int. Ed. 2008, 47, 9345–9348.
[8] C. Vollmer, C. Janiak, Coord. Chem. Rev. 2011, 255, 2039–2057.
[9] See the Supporting Information for details.
[10] C. M. Hagen, L. Vieille-Petit, G. Laurenczy, G. Sꢂss-Fink, R. G. Finke, Orga-
nometallics 2005, 24, 1819–1831.
[11] D. Astruc, F. Lu, J. R. Aranzaes, Angew. Chem. 2005, 117, 8062–8083;
Angew. Chem. Int. Ed. 2005, 44, 7852–7872.
[12] T. Tsuda, T. Sakamoto, Y. Nishimura, S. Seino, A. Imanishi, S. Kuwabata,
Chem. Commun. 2012, 48, 1925–1927.
[13] E. Bayram, J. C. Linehan, J. Fulton, J. A. S. Roberts, N. K. Szymczak, T. D.
Smurthwaite, S. ꢃzkar, M. Balasubramanian, R. G. Finke, J. Am. Chem.
Soc. 2011, 133, 18889–18902.
[14] It is not clear at present how a small amount of CrII is reduced to Cr0
under the current reaction conditions (in IL or DMF, 1208C, 6 h), but the
presence of glucose (a reducing sugar) is critical for Cr0-NP formation.
[5] a) D. Liu, E. Y.-X. Chen, Appl. Catal. A 2012, 435–436, 78–85; b) X. Qi, M.
Watanabe, T. M. Aida, R. L. Smith, Bioresour. Technol. 2012, 109, 224–
228.
[6] a) C. Vollmer, E. Redel, K. Abu-Shandi, R. Thomann, H. Manyar, C. Harda-
cre, C. Janiak, Chem. Eur. J. 2010, 16, 3849–3858; b) E. Redel, R. Tho-
mann, C. Janiak, Chem. Commun. 2008, 1789–1791.
Received: October 21, 2012
[7] L. Starkey-Ott, R. G. Finke, Coord. Chem. Rev. 2007, 251, 1075–1100.
Published online on December 7, 2012
ꢀ 2013 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim
ChemSusChem 2013, 6, 61 – 64 64