Highly efficient and N-bromosuccinimide-mediated conversion of carbohydrates to 5-hydroxymethylfurfural under mild conditions

Article abstract of DOI:10.1007/s11164-012-0837-x

Highly efficient and selective conversion of different carbohydrates to 5-hydroxymethylfurfural (HMF) has been successfully performed with N-bromosuccinimide (NBS) as a promoter. In the presence of single NBS, a 64.2 % yield of HMF from fructose was obtained in N-methylpyrrolidone for 2 h. The effects of time, temperature and reaction media are discussed. It was concluded that the preliminary bromination of substrate could improve the generation of HMF compared to the direct dehydration process. Moreover, the HMF yield could be elevated to 79.6 and 82.3 % when FeCl₃ and SnCl₄ were used as the additives, respectively. Furthermore, the addition of CrCl ₃ facilitated the conversion pathway from glucose, sucrose, inulin, or cellulose to HMF. A 57.3, 68.2, 62.4, or 6.1 % yield of HMF was, respectively, obtained in the presence of CrCl₃ and NBS under mild conditions, which will therefore generate a promising application strategy for biomass transformation.

Full text of DOI:10.1007/s11164-012-0837-x

Res Chem Intermed
DOI 10.1007/s11164-012-0837-x
Highly efficient and N-bromosuccinimide-mediated
conversion of carbohydrates to
5
-hydroxymethylfurfural under mild conditions
Guo Tian
Yongtao Yan
•
Xinli Tong
•
•
Yanhua Wang
•
Song Xue
Received: 10 August 2012 / Accepted: 25 September 2012
Ó Springer Science+Business Media Dordrecht 2012
Abstract Highly efficient and selective conversion of different carbohydrates to
5
-hydroxymethylfurfural (HMF) has been successfully performed with N-bromo-
succinimide (NBS) as a promoter. In the presence of single NBS, a 64.2 % yield of
HMF from fructose was obtained in N-methylpyrrolidone for 2 h. The effects of
time, temperature and reaction media are discussed. It was concluded that the
preliminary bromination of substrate could improve the generation of HMF com-
pared to the direct dehydration process. Moreover, the HMF yield could be elevated
to 79.6 and 82.3 % when FeCl and SnCl were used as the additives, respectively.
3
4
Furthermore, the addition of CrCl facilitated the conversion pathway from glucose,
3
sucrose, inulin, or cellulose to HMF. A 57.3, 68.2, 62.4, or 6.1 % yield of HMF was,
respectively, obtained in the presence of CrCl and NBS under mild conditions,
3
which will therefore generate a promising application strategy for biomass
transformation.
Keywords Dehydration Á Carbohydrates Á N-Bromosuccinimide Á
5
-Hydroxymethylfurfural Á Biomass conversion
Electronic supplementary material The online version of this article
doi:10.1007/s11164-012-0837-x) contains supplementary material, which is available to authorized
users.
(
G. Tian Á X. Tong (&) Á Y. Wang Á Y. Yan Á S. Xue (&)
School of Chemistry and Chemical Engineering, Tianjin University of Technology,
Tianjin 300384, People’s Republic of China
e-mail: tongxinli@tju.edu.cn; tongxli@sohu.com
S. Xue
e-mail: xuesong@ustc.edu.cn
123

G. Tian et al.
Introduction
Nowadays, producing fuels and chemicals from renewable sources instead of fossil
resources has become a most attractive research area in the chemistry field. Biomass
feedstocks are considered as promising alternatives because of their abundance,
renewability and worldwide distribution [1–3]. Particular attention has been paid to
the conversion of different carbohydrates to value-added chemicals, in which
5
-hydroxymethylfurfural (HMF), a valuable intermediate used in the production of
fine chemicals, pharmaceuticals, and thermoresistant polymers, and produced from
the dehydration of carbohydrate, is very interesting and significant [4].
In the last decade, fructose, sucrose and inulin have been employed as the
reactants for the preparation of HMF using sustainable methods [5]. For the fructose
dehydration, general protonic acids or Lewis acids can be employed as the catalysts
[
6, 7]. Antal et al. [8, 9] reported the dehydration of D-fructose with H SO as a
2 4
catalyst in subcritical water at 250 °C, and gained a yield of HMF as high as 53 %.
Moreover, Bicker et al. [10] found that the yield of HMF reached 78 % at 180 °C in
the presence of H SO when supercritical acetone–water mixture was employed as a
2
4
reaction medium. Furthermore, Roman-Leshkov et al. [11] reported a two-phase
reactor system with HCl as the catalyst. In this work, the D-fructose was dehydrated
in the aqueous phase with HCl as a catalyst. Otherwise, DMSO and poly(1-vinyl-2-
pyrrolidinone) were added to suppress the undesired side reactions. The product
5
-HMF was continuously extracted into an organic methylisobutylketone (MIBK)
phase modified with 2-butanol to enhance the partitioning from the reactive aqueous
solution. As a result, it was found that 80 % HMF selectivity at a 90 % conversion
was achieved for 10 wt% D-fructose solution. In addition, a continuous microre-
action process has been set up based on the HCl-catalyzed dehydration of fructose in
pure aqueous solution in order to improve the ‘‘green’’ synthesis of HMF [12]. As a
result, the process conditions were deliberately shifted to high temperatures and
pressures (185 °C, 17 bar) in only 1 min, and the product HMF was obtained with
5
4 % yield at 71 % fructose conversion. On the other hand, Chheda et al. [13] found
that a good selectivity for HMF at high conversion from polysaccharides, such as
sucrose, starch, cellobiose, and xylan, can be obtained using mineral acid (HCl,
H SO or H PO ) as the catalyst in a biphasic reactor system, which was composed
2
4
3
4
of the reactive aqueous phase modified with DMSO and the organic extracting
phase consisting of a 7:3 (w/w) MIBK–2-butanol mixture. Also, Ilgen et al. [14]
reported that 25 and 57 % yields of HMF were obtained, respectively, from inulin
and sucrose using p-toluenesulfonic acid as catalyst in melt systems consisting of
choline chloride (ChCl) and up to 50 wt% of carbohydrates.
Recently, the development of novel and efficient catalytic systems has become a
hot topic for the synthesis of HMF from the dehydration of carbohydrates [15]. For
example, strong acid cation exchange resins [16–18], H-form zeolites [19, 20],
metal chlorides [21–24], and supported heteropolyacids [25] have been employed
for the dehydration of fructose to HMF. In addition, acidic ionic liquids have also
been used as catalysts or solvents in the dehydration of carbohydrates [26–30].
Neutral ionic liquids can promote the dehydration reaction due to the enhanced
dissolution of catalysts and substrates; for example, Moreau et al. [26] reported that
1
23

N-bromosuccinimide-mediated conversion of carbohydrates
the ILs [bmim]PF and [bmim]BF are suitable reaction media for the dehydration
6
4
of D-fructose with Amberlyst-15 as a catalyst. Moreover, Zhao et al. [27] found that,
in the IL [emim]Cl, chromium(II) chloride (CrCl ) could efficiently catalyze the
2
dehydration of D-fructose, leading to a yield of HMF close to 70 %. In addition,
several special ILs, such as 1-[H]-3-methylimidazolium chloride and choline
chloride/citric acid, are also employed as both solvent and catalyst for the
generation of HMF from fructose [28]. Recently, we have studied the dehydration of
sugar with acidic ionic liquid. It was found that N-methyl-2-pyrrolidonium-based
and N-methyl-morpholinium-based ionic liquids showed high catalytic activity for
the dehydration of fructose or sucrose in the organic solvents [29, 30]. Most of these
systems suffer from the need for relatively high reaction temperatures and poor
yields or low product selectivities. For instance, the poor yield arose from the
decomposition and polymerization of HMF after it was produced, and the main by-
products included levulinic acid, formic acid and humin among other components
[5, 31, 32]. The production of HMF in a high yield and by a cost-effective technique
is still a challenge.
It is well known that N-bromosuccinimide (NBS) can mediate the oxidation and
dehydrogenation of alcohols in most organic solvents [33–35], while the promotion
of NBS on the selective dehydration of carbohydrates is still vacant. Previous
investigations have indicated that the existence of bromide is helpful for the
selective conversion of sugar owing to its nucleophilic ability [23, 36]. In this work,
we have focused on using a cheap, easy-to-use and nucleophilic NBS instead of
corrosive inorganic acid as the catalyst for direct dehydration of fructose to HMF.
Moreover, the mediation of NBS on conversion of polysaccharides in the presence
of metal chlorides is also investigated and discussed.
Results and discussion
The dehydration of carbohydrates to 5-hydroxymethylfurfural
To start with, the catalytic performance of different N-substituted succinimides was
investigated in the dehydration of fructose (Scheme 1), where a 10.0 mol% catalyst
was employed for a catalytic reaction for 1 h at 90 °C, and the corresponding
dehydration results are summarized in Table 1.
Scheme 1 The dehydration of D-fructose with different N-substituted succinimides
123

G. Tian et al.
Table 1 Synthesis of HMF from carbohydrate with different catalysts
a
b
Entry
Reactant
Catalyst
Time (h)
Yield (%)
1
2
3
4
5
6
7
8
9
Fructose
Fructose
Fructose
Fructose
Fructose
Fructose
Fructose
Fructose
Glucose
Sucrose
Fruc. ? Gl.
Inulin
no
1
1
1
1
2
1
2
2
1
1
1
1
0.8
56.1
36.2
31.6
1.0
N-Bromosuccinimide
N-Chlorosuccinimide
N-Iodosuccinimide
N-Hydroxysuccinimide
N-Chlorophthalimide
N-Hydroxyphthalimide
Phthalimide
31.0
0.6
0.9
N-Bromosuccinimide
N-Bromosuccinimide
N-Bromosuccinimide
N-Bromosuccinimide
3.2
10
11
12
32.4
27.9
24.1
c
Reaction conditions: 1.0 g substrate, 10.0 mol% catalyst, in 12 mL NMP, reaction temperature 90 °C
a
The molecular structure of catalysts is shown in the following:
a
b
O
O
N X
N-Z
O
O
(X=Br, Cl, I , or OH group)
(Z=Cl, OH or H)
b
c
The results are obtained by HPLC analysis
The mass ratio of fructose and glucose is 1:1
It can be seen that the dehydration of fructose occurs only slightly with N-
methylpyrrolidone (NMP) as the solvent in absence of any catalyst, in which only
.8 % yield of HMF is obtained (entry 1), while, in the presence of 10.0 mol% NBS,
0
the yield of HMF can be elevated to 56.1 % (entry 2). Moreover, the catalytic
performances of other N-substituted succinimides were also investigated. It was
found that N-chlorosuccinimide (NCS) and N-iodosuccinimide (NIS) exhibit
comparative activities, and 36.2 or 31.6 % yield of HMF is obtained with NCS
or NIS, respectively (entries 3 and 4). In order to reveal further the promotion effect
of these organic molecules, N-hydroxysuccinimide (NHC) and N-chlorophthalimide
were also chosen as the catalysts for the fructose dehydration. It was found that only
1
.0 % yield of HMF is obtained with NHC for 5 h, and 31.0 % yield of HMF is
obtained with N-chlorophthalimide for 2 h (entries 5 and 6), which showed that
halide-substituted imides are more favorable in the dehydration. Furthermore, the
experimental data with N-hydroxyphthalimide and phthalimide as the promoters
also validate this point, in which only 0.6 and 0.9 % yield of HMF were obtained
(entries 7 and 8). In addition, it was also seen that NBS is more efficient than NCS
and NIS under similar conditions, which indicated that the bromo group probably
1
23

N-bromosuccinimide-mediated conversion of carbohydrates
has an important role in the catalytic reaction. Also, it was concluded that the
preliminary bromination of substrate could improve the generation of HMF
compared to the direct dehydration process.
Moreover, several representative carbohydrates were further studied as the
reactants for HMF production. It was found that only 3.2 % of HMF is obtained for
the conversion of glucose in the presence of NBS, which indicates that the catalytic
efficiency of NBS is closely related to the reaction substrate (entry 9). On the other
hand, the HMF yields of 32.4, 27.9 and 24.1 % were obtained when sucrose,
fructose ? glucose (1:1), and inulin, respectively, are employed as reactants (entries
1
0–12). Thus, it can be concluded that NBS can promote the conversion of
numerous carbohydrates to HMF, and that the ketose is more easily dehydrated than
aldose in this reaction system.
Effect of reaction time in the fructose dehydration
The effect of reaction time in the dehydration of fructose with NBS is outlined in
Fig. 1. In the dehydration of fructose, the yield of HMF increases gradually before
6
6
0 min, and the yield of HMF remains almost unchanged during the period of
0–120 min, in which 62.2 % yield of HMF was obtained for 120 min at 90 °C.
However, the HMF yield began to decline when the reaction time was further
prolonged, which shows that the conversion of HMF to by-product is probably more
rapid than the generation of HMF after 120 min.
Effect of temperature in the fructose dehydration
The effect of temperature in the dehydration of fructose with NBS as a catalyst is
presented in Fig. 2. It is seen that the yield of HMF increases from 40 to 100 °C,
and the yield decreases after 100 °C, which is probably attributed to the occurrence
Fig. 1 Effect of reaction time in the dehydration of fructose with NBS as the catalyst (reaction
conditions: 1.0 g reactant, 10.0 mol% catalyst, in 12 mL solvent, reaction temperature 90 °C)
123

G. Tian et al.
Fig. 2 Effect of reaction temperature in the dehydration of fructose with NBS as the catalyst. (reaction
conditions: 1.0 g reactant, 10.0 mol% catalyst, in 12 mL solvent, 2.0 h)
Table 2 The conversion of
carbohydrate to HMF with NBS
in different solvents
a
Entry
Substrate
Solvent
Yield of HMF (%)
1
2
3
4
5
6
7
8
Fructose
Fructose
Fructose
Fructose
Fructose
Sucrose
Inulin
EtOH
DMF
8.9
29.3
7.6
CH
DMSO
CCl
3
CN
Reaction conditions: 1.0 g
47.5
1.9
reaction substrate, 10.0 mol%
NBS catalyst, in 12 mL solvent,
reaction time 2 h, temperature
4
DMSO
DMSO
DMF
31.0
22.5
14.8
1
a
00 °C
The results are obtained by
HPLC analysis
Sucrose
of side reactions. Here, the highest yield of HMF is 64.2 % at 100 °C for 2 h with
NBS as the catalyst in the dehydration reaction (Fig. 2).
Effect of solvents in the fructose dehydration
Different solvents, including polar [ethanol, N,N-dimethyl formamide (DMF),
acetonitrile, DMSO] and non-polar one (CCl ), were employed in the dehydration
4
of fructose (Table 2). As a result, a 47.5 % yield or a 29.3 % yield of HMF with
DMSO or DMF, respectively, as a solvent is superior to the 7.6 % yield or 8.9 %
yield with acetonitrile or ethanol, respectively, as a solvent. Otherwise, due to the
limited solubility of fructose, the dehydration reaction is scarcely performed when
carbon tetrachloride is used as a solvent. So, it is concluded that polar and aprotic
solvent is more suitable for this reaction system, which can be attributed to its
better solubility and differentiating effect. In addition, DMSO is also a favorable
1
23

N-bromosuccinimide-mediated conversion of carbohydrates
Table 3 Dehydration of various carbohydrate to HMF with NBS and metal chlorides
a
Entry
Substrate
Catalyst
Solvent
Yield (%)
1
2
3
4
5
6
7
8
Fructose
Fructose
Glucose
Glucose
Glucose
Glucose
Sucrose
Inulin
FeCl
3
? NBS
? NBS
NMP
79.6
82.3
2.7
SnCl
NBS
SnCl
SnCl
4
NMP
DMSO
DMSO
DMSO
DMSO
NMP
4
2
? NBS
? NBS
? NBS
? NBS
? NBS
? NBS
? NBS
? NBS
? NBS
29.9
10.5
57.3
40.7
47.2
4.9
CrCl
3
SnCl
SnCl
SnCl
4
4
4
NMP
b
9
Cellulose
Sucrose
Inulin
NMP
10
11
12
CrCl
CrCl
CrCl
3
3
3
DMSO
DMSO
DMSO
68.2
62.4
6.1
b
Cellulose
Reaction conditions: 1.0 g reaction substrate, 10.0 mol% NBS catalyst, in 12 mL solvent, reaction time
2
h, temperature 100 °C
The results are obtained by HPLC analysis
a
b
The reaction is performed at 140 °C for 3 h
solvent for the dehydration of fructose to HMF due to its ability to prevent the
formation of by-products such as levulinic acid and humin. Furthermore, a 31.0 or
1
4.8 % yield of HMF was obtained in the presence of 10.0 mol% NBS from
sucrose when DMSO or DMF, respectively, is used as solvent. For the conversion
of inulin to HMF, a yield of 22.5 % is obtained, which is very near to that obtained
in NMP solvent.
Dehydration of carbohydrates with NBS and metal chlorides
In the following, the combination of NBS and metal chlorides were employed as
catalyst system for the conversion of different carbohydrates to HMF. As
indicated in Table 3, the yield of HMF is increased to 79.6 and 82.3 % when
FeCl and SnCl are employed as the additives, respectively (entries 1 and 2).
3
3
For the conversion of glucose, only 2.7 % of HMF is obtained with NBS as the
catalyst in DMSO solvent. However, 29.9, 10.5 or 57.3 % yields of HMF are
respectively obtained in the presence of SnCl –NBS, SnCl –NBS or CrCl –NBS
4
2
3
catalytic system (entries 3–5). On the other hand, the yields of HMF are 40.7,
7.2 and 4.9 % for the conversion of sucrose, inulin and cellulose, respectively,
in the presence of NBS and SnCl (entries 6–8). Yields of 68.2, 62.4 or 6.1 %
4
4
HMF are respectively obtained for conversion of sucrose, inulin, or cellulose in
the presence of NBS and CrCl in DMSO solvent (entries 9–12). These results
3
showed that the addition of metal Cr facilitates the conversion pathway
from glucose, sucrose, inulin, respectively, or cellulose to HMF in this catalyst
system.
123

G. Tian et al.
Experimental
Reagents and equipment
Fructose, glucose, sucrose, inulin, cellulose, lactose, NBS, NCS, NIS, NHC,
anhydrous ethanol, acetonitrile, N-chlorophthalimide, N-hydroxyphthalimide,
phthalimide, FeCl , SnCl , SnCl , CrCl , NaHCO , acetic acetate, and Na SO
4
3
4
2
3
3
2
were analytical grade and used without further purification. The solvent including
NMP, ethanol, acetonitrile, DMF, DMSO and CCl was rectified before being used.
4
The NMR spectra were recorded on an INOVA 500MHZ spectrometer. The
qualitative analysis was performed using an Agilent 6890/5973 GC–MS, and the
quantitative analysis was performed by the Agilent 2100 HPLC with UV and
Refractive Index detectors.
Reaction conditions for the dehydration of different carbohydrates
Typical procedure for the dehydration of carbohydrate is as follows: 1.0 g substrate,
NBS (10.0 mol%) and 12 mL NMP solution are charged in order into a 100-mL
flask equipped with magnetic stirring and a condenser, The atmosphere inside is
replaced with nitrogen before the flask is airproofed. Under stirring, the flask is
preheated to the certain temperature with an oil bath and then kept there for the
necessary time. After the reaction, the obtained mixture is decanted into a
volumetric flask with anhydrous ethanol as diluter, and then analyzed by HPLC with
UV and Refractive Index detectors.
Typical separation procedure for HMF
After the reaction, the mixture is transferred into a flask, and then a saturated
aqueous solution of NaHCO is added. The obtained liquid mixture is stirred with a
3
magnetic stirrer overnight and extracted three times with acetic acetate, and then the
organic phase is collected. After being dried with anhydrous Na SO , the organic
2
4
layer is distilled under reduced pressure to obtain the pure HMF as a main product.
The purity is more than 98 % from HPLC analysis.
Conclusion
In summary, the efficient and catalytic synthesis of HMF from different
carbohydrates has been successfully performed with NBS as the promoter under
mild conditions. A 64.2 % yield of HMF is obtained from fructose in the presence
of only NBS for 2 h at 100 °C. The effects of the reaction temperature and reaction
time in the fructose dehydration are examined. Moreover, the addition of CrCl
3
facilitated the conversion pathway from glucose, sucrose, inulin, or cellulose to
HMF. This efficient catalyst system containing NBS will therefore generate a
promising application strategy for biomass transformation.
1
23

N-bromosuccinimide-mediated conversion of carbohydrates
Acknowledgment We thank the financial support of the Natural Science Foundation of China (Project
Number 21003093).
References
1
2
3
4
5
6
7
8
9
. G.W. Huber, S. Iborra, A. Corma, Chem. Rev. 106, 4044–4098 (2006)
. J.N. Chheda, G.W. Huber, J.A. Dumesic, Angew. Chem. Int. Ed. 46, 7164–7183 (2007)
. G.W. Huber, A. Corma, Angew. Chem. Int. Ed. 46, 7184–7201 (2007)
. X. Tong, Y. Ma, Y. Li, Appl. Catal. A Gen. 385, 1–13 (2010)
. A. Corma, S. Iborra, A. Velty, Chem. Rev. 107, 2411–2502 (2007)
. C.J. Moye, Rev. Pure Appl. Chem. 14, 161–170 (1964)
. M.S. Feather, J.F. Harris, Adv. Carbohydr. Chem. Biochem. 28, 161–224 (1973)
. M.J. Antal, W.S.L. Mok, Carbohydr. Res. 199, 91–109 (1990)
. M.J. Antal, W.S. Mok, in Research in Thermochemistry and Biomass Conversion (Elsevier, New
York, 1988) pp. 464–472
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
0. M. Bicker, J. Hirth, H. Vogel, Green Chem. 5, 280–284 (2003)
1. Y. Roman-Leshkov, J.N. Chheda, J.A. Dumesic, Science 312, 1933–1937 (2006)
2. T. Tuercke, S. Panic, S. Loebbecke, Chem. Eng. Technol. 32, 1815–1822 (2009)
3. J.N. Chheda, Y. Roman-Leshkov, J.A. Dumesic, Green Chem. 9, 342–350 (2007)
4. F. Ilgen, D. Ott, D. Kralisch, C. Reil, A. Palmberger, B. K o¨ nig, Green Chem. 11, 1948–1954 (2009)
5. A.A.M. Lapis, L.F. de Oliveira, B.A.D. Neto, J. Dupont, ChemSusChem 1, 759–762 (2008)
6. L. Rigal, A. Gaset, Biomass 8, 267–276 (1985)
7. X. Qi, M. Watanabe, T.M. Aida, R.L. Smith Jr, Green Chem. 10, 799–805 (2008)
8. X. Qi, M. Watanabe, T.M. Aida, R.L. Smith Jr, Ind. Eng. Chem. Res. 47, 9234–9239 (2008)
9. C. Moreau, R. Durand, C. Pourcheron, S. Razigade, Ind. Crops Prod. 3, 85–90 (1994)
0. P. Rivalier, J. Duhamet, C. Moreau, R. Durand, Catal. Today 24, 165–171 (1995)
1. Z. Zhang, Q. Wang, H. Xie, W. Liu, Z.K. Zhao, ChemSusChem 4, 131–138 (2011)
2. G. Yong, Y.J. Zhang, Y. Ying, Angew. Chem. Int. Ed. 47, 9345–9348 (2008)
3. X. Tong, M. Li, N. Yan, Y. Ma, P.J. Dyson, Y. Li, Catal. Today 175, 524–527 (2011)
4. C. Wang, L. Fu, X. Tong, Q. Yang, W. Zhang, Carbohydr. Res. 347, 182–185 (2012)
5. M.A. Schwegler, P. Vinke, M. van der Eijk, H. van Bekkum, Appl. Catal. A Gen. 80, 41–57 (1992)
6. C. Lansalot-Matras, C. Moreau, Catal. Commun. 4, 517–520 (2003)
7. H. Zhao, J.E. Holladay, H. Brown, Z.C. Zhang, Science 316, 1597–1600 (2007)
8. C. Moreau, A. Finiels, L. Vanoye, J. Mol. Catal. A Chem. 253, 165–169 (2006)
9. X. Tong, Y. Li, ChemSusChem 3, 350–355 (2010)
0. X. Tong, Y. Ma, Y. Li, Carbohydr. Res. 345, 1698–1701 (2010)
1. B.F.M. Kuster, Starch/St a¨ rke 42, 314–321 (1990)
2. J. Lewkowski, ARKIVOC (i), 17–54 (2001)
3. N.S. Srinivasan, Tetrahedron 30, 419–425 (1974)
4. R. Filler, Chem. Rev. 63, 21–43 (1963)
5. K. Surendra, N.S. Krishnaveni, K.R. Rao, Tetrahedron Lett. 46, 4111–4113 (2005)
6. J.B. Binder, R.T. Raines, J. Am. Chem. Soc. 131, 1979–1985 (2009)
123