Examining Acid Formation During the Selective Dehydration of Fructose to 5-Hydroxymethylfurfural in Dimethyl Sulfoxide and Water

Article abstract of DOI:10.1002/cssc.201803013

Sustainable conversion of biomass, including fructose dehydration to 5-hydroxymethylfurfural (HMF), remains a challenge. Fructose can be selectively dehydrated to HMF in dimethyl sulfoxide (DMSO) without addition of an acid catalyst. The role of DMSO is examined starting with either fructose or HMF in DMSO/water. With increasing DMSO content, it is observed that fructose conversion, HMF selectivity, and post-reaction solution acidity increase. Although DMSO degradation to sulfuric acid is a potential source of acidity and reactivity, a barium chloride precipitation test demonstrates that sulfate ions are not detectable after reaction, suggesting that DMSO is stable during reaction at 120 °C and 150 °C with oxygen present. Instead, the majority of the acidic species produced are formic acid, levulinic acid, and humins. These acids have a minimal effect on fructose conversion in DMSO. These results suggest that DMSO promotes fructose conversion mainly through solvation effects and not as an origin of acid catalysis. For HMF stabilization, the optimal molar fraction of DMSO in water is 0.20–0.43. Overall, these results indicate that DMSO can promote fructose dehydration to HMF at 120 °C.

Full text of DOI:10.1002/cssc.201803013

Accepted Article
Title: Examining Acid Formation During the Selective Dehydration of
Fructose to 5-Hydroxymethylfurfural in DMSO and Water
Authors: Mariah Rose Whitaker, Aamena Parulkar, Pinaki Ranadive,
Rutuja Joshi, and Nicholas Anthony Brunelli
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: ChemSusChem 10.1002/cssc.201803013
Link to VoR: http://dx.doi.org/10.1002/cssc.201803013
A Journal of
www.chemsuschem.org

10.1002/cssc.201803013
ChemSusChem
1
2
Title: Examining Acid Formation During the Selective Dehydration of Fructose to 5-
Hydroxymethylfurfural in DMSO and Water
3
Authors: Mariah R. Whitaker[a], Dr. Aamena Parulkar[a], Pinaki Ranadive[a], Rutuja Joshi[a], and
Prof. Nicholas A. Brunelli*[a]
4
5
Author address: [a] The Ohio State University, William G. Lowrie Department of Chemical and
Biomolecular Engineering, 151 W. Woodruff Ave., Columbus, OH 43210, Fax: 614-292-3769,
Homepage: catalysts.osu.edu
6
7
8
*Corresponding author - Email: brunelli.2@osu.edu; Twitter: OSUChemEProfBru
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
Abstract:
Sustainable conversion of biomass remains a challenge, including fructose dehydration to
5-hydroxymethylfurfural (HMF). Fructose can be selectively dehydrated to HMF in dimethyl
sulfoxide (DMSO) without addition of an acid catalyst. The role of DMSO is examined starting with
either fructose or HMF in DMSO/water. With increasing DMSO content, it is observed that fructose
conversion, HMF selectivity, and post-reaction solution acidity increase. While DMSO
degradation to sulfuric acid is a potential source of acidity and reactivity, a barium chloride
precipitation test demonstrates that sulfate ions are not detectable after reaction, suggesting that
DMSO is stable during reaction at 120°C and 150°C with oxygen present. Instead, the majority of
the acidic species produced are formic acid, levulinic acid, and humins. These acids have a
minimal effect on fructose conversion in DMSO. These results suggest that DMSO promotes
fructose conversion mainly through solvation effects and not as an origin of acid catalysis. For
HMF stabilization, the optimal molar fraction of DMSO in water is 0.20 to 0.43. Overall, these
results indicate that DMSO can promote fructose dehydration to HMF at 120°C.
24
1
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
25
Table of Contents Image and Text:
26
27
DMSO stability under reaction conditions relevant for fructose dehydration to HMF are
investigated. Post-reaction mixtures are acidic because of formation of formic and levulinic acids
and acidic humins, but sulfuric acid is not detected in these mixtures. An optimal fraction of DMSO
exists where a minimum of HMF conversion is observed for this reaction.
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
Keywords: Biomass conversion, DMSO stability, Fructose dehydration, Homogeneous catalysis,
5-hydroxymethylfurfural
Introduction
Biomass is an abundant and under-utilized carbon source that has great potential to
decrease dependence on fossil fuel by providing an alternative and renewable carbon feedstock
for chemical production. Indeed, promising research has shown that biomass can be used to
produce fuels,[1,2] chemicals,[3–6] and polymers.[2] One important pathway for biomass conversion
involves converting lignocellulose through several intermediates such as glucose and fructose
before producing 5-hydroxymethylfurfural (HMF).[1] HMF was identified by the US Department of
Energy[1,7] as a key platform chemical that can be used as a precursor to produce several valuable
compounds in a few synthetic steps.[6] The current challenge remains achieving high selectivity
for HMF in an efficient, scalable, and economically viable process.[6,8,9] Research has pursued
2
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
many different directions for increasing HMF selectivity, including using either a homogeneous or
heterogeneous acid catalyst in an organic solvent[9–12] or ionic liquid,[13–16] as a co-solvent or
biphasic reaction system.[17,18] These methods result in higher HMF selectivity with varying
degrees of success. Of these, using dimethyl sulfoxide (DMSO) as a co-solvent is an effective
method to obtain high HMF selectivity.[6] However, it is well known that a key limitation of using
DMSO as a homogeneous solvent is its high boiling point, which makes HMF separation
difficult.[12] While including DMSO in the reaction mixture has clear benefits in terms of selectivity,
considerable debate persists concerning the role of DMSO for the dehydration of fructose to HMF,
particularly in the solvent’s ability to significantly increase the reaction rate even in absence of
acid catalyst.[6] Some researchers attribute this observed fructose reactivity in DMSO to DMSO
degradation to sulfuric acid,[19–21] which can be the catalytically active species. Providing additional
insights on the role of DMSO would facilitate the development and improvement of engineering
solutions, such as heterogeneous catalyst design that has the potential to overcome the
limitations of DMSO and lead to more sustainable catalytic material design.
The overall mechanism for the acid catalyzed fructose dehydration involves the
sequential removal of three water molecules, as shown in Scheme 1. In pure water, the
dehydration of fructose results in low selectivity for HMF that is typically in the 30-50% range for
fructose conversions of 50-70%.[22,23] Reaction mixtures including DMSO have frequently been
reported to achieve selectivity values of 90% or higher.[6] While a complete picture of the roles of
DMSO is not currently known, previous research has determined the presence of DMSO has at
least three roles in the reaction. First, DMSO is known to alter the isomer distribution of fructose
from the distribution that is observed in water.[24,25] While fructose exists in five different isomers
(α-pyranose, β-pyranose, α-furanose, β-furanose, and open-chain), NMR measurements of
fructose in DMSO demonstrated that the b-furanose isomer is present in highest concentration
70
as compared to in water where the b-pyranose isomer is dominant.[10,24,26] Since HMF is thought
3
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
to be produced from the furanose form of fructose,[27–29] the increased furanose population in
DMSO contributes to the increased yield of HMF in DMSO as compared to that in water.[10]
Christopher et al developed an equation that was able to accurately predict initial HMF selectivity
strictly from the distribution of fructose isomers obtained from in situ NMR, indicating that the
isomer distribution is the major factor affecting HMF selectivity.[21]
Second, the presence of DMSO has been shown to alter the intermediates observed for
the dehydration reaction. NMR measurements have elucidated several intermediates in DMSO
that are distinct from intermediates observed in water.[10,19] In situ 13C NMR studies of fructose
dehydration in DMSO without added catalyst have identified cyclic intermediates that indicate the
furan ring remains intact while undergoing dehydration.[10,30] These intermediates are thought to
be the origin of the higher selectivity observed in DMSO compared to water. Interestingly, these
results are obtained in the absence of an acid catalyst, and these intermediates associated with
the selective pathway are still observed in water and DMSO solvent mixtures, providing some
scope for tuning the reaction conditions while still maintaining high HMF selectivity.
The final generally accepted benefit is that DMSO stabilizes HMF in solution. Using
attenuated total reflectance FTIR spectroscopy, Tsilomelekis et al demonstrated that DMSO binds
more tightly to HMF than water.[31] The coordination of DMSO to HMF was in agreement with
Density Functional Theory (DFT) calculations. The DFT calculations further demonstrated that
the LUMO energy of HMF increases when solvated by DMSO, which makes HMF less susceptible
to nucleophilic attack and results in stabilizing HMF.[31,32] Interestingly, molecular simulations
suggest DMSO preferentially solvates the carbonyl of HMF, which is thought to prevent the
rehydration of HMF to organic acids.[33,34] Overall, these results demonstrate that DMSO has many
potential roles in increasing the selective formation of HMF.
Although the high selectivity for HMF demonstrates that DMSO provides several benefits
in promoting the selective dehydration of fructose, several perspectives exist on the role of DMSO
in promoting the reactivity for fructose dehydration in the absence of acid catalyst under aerated
4
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
97
98
conditions.[19–21,35,36] Several hypotheses have been proposed by researchers to explain this
increased reactivity in DMSO, including DMSO stabilization of transition states that reduces the
reaction energy barrier and increases the rate of reaction,[10,37] DMSO acting as a catalyst by
initiating hydroxyl dehydration from fructose,[25,30] or DMSO degrading in situ into sulfuric acid,
which is capable of catalyzing the reaction.[21,38–40] These different proposed hypotheses have
important implications for how to design selective catalysts. If DMSO is stabilizing transition states
or acting as a catalyst, an efficient catalyst would potentially consist of sulfoxide species that
would provide the necessary interactions to promote selective dehydration. This could eliminate
the need to add homogeneous acidic species to the process. Eliminating acid species could be
beneficial in the event that biomass processing of glucose to fructose to HMF is performed in a
cascade manner using immobilized amines to catalyze the isomerization of glucose[41] in water,
which is also the solvent of choice for upstream processes. Indeed, both catalysts could be
included in the same reactor to overcome equilibrium conversion limitations for glucose
isomerization, eliminating the need to use moving bed chromatography to purify fructose before
the dehydration reaction.[42] Alternatively, if DMSO degrades in situ to form sulfuric acid, this would
complicate process design and increase cost. First, DMSO would have to be constantly replaced,
adding to the overall costs.[43] Second, DMSO degradation would introduce acidic species that
would need to be neutralized to avoid equipment corrosion.
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
Previous work has studied DMSO decomposition.[40,44,45] These studies have been used
as a basis to suggest DMSO degradation is occurring during fructose dehydration in DMSO under
aerated conditions.[20,21] However, the DMSO degradation work[40,44] was conducted at high
temperatures (>180°C), under an enriched O2 atmosphere, or in presence of added radical
initiators like AIBN or peroxides. These conditions are more extreme than conditions typically
utilized for fructose dehydration. Thus, a critical evaluation of DMSO stability under fructose
dehydration conditions is needed. A sensitive test for DMSO degradation was identified by
Santosusso and Swern.[40] While studying the thermal decomposition of DMSO under reflux
5
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
conditions in oxygen enriched atmosphere, they confirmed the presence of sulfuric acid in the
reaction solution through addition of barium chloride to precipitate sulfate ions in the form of
barium sulfate. Barium sulfate is insoluble and can be identified to be present by XRD and FTIR
analysis, making it a sensitive test for sulfate ions.[40] This result indicated that dissolved oxygen
could promote the generation of in situ acid that could catalyze the reaction. Currently, similar
observations have not been made for fructose dehydration in DMSO under air.
This work seeks to provide additional insights on the role of DMSO through examining
reaction mixtures with different amounts of DMSO in water to determine the effect of solvent
composition on fructose conversion, HMF selectivity, and acid formation. The post-reaction
mixtures are also analyzed using titration experiments to determine how the composition impacts
formation of acidic species. Further work uses a barium chloride precipitation test in conjunction
with titration and HPLC analysis to elucidate the nature of acidic species formed during the
reaction. Finally, experiments are presented that test the stability of HMF in different DMSO and
water solutions. Through the combination of experiments, an optimal fraction of DMSO in water
is elucidated that balances HMF selectivity and DMSO utilization.
138
Results and Discussion
139
140
141
142
143
144
145
146
Effect of DMSO concentration on fructose conversion, HMF selectivity, and solution acidity
Initial work examines how the solvent composition affects the conversion of fructose and
selectivity to HMF. A comparative study is done between two cases: (a) in the absence of acid
catalyst and (b) in the presence of 10 mol% Amberlyst-15, a commercially available sulfonic acid
heterogeneous catalyst. The results in Figure 1a show that conversion and selectivity increase
with increasing DMSO concentration for both cases. Similarly, Jia et al investigated baseline
fructose conversion and HMF selectivity in co-solvents of varying DMSO fractions in absence of
acid catalyst.[36] However, this previous work only preformed these baseline reactions for 1 hour
6
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
of reaction time, and therefore, claimed that a DMSO composition less than 0.70 moles of DMSO
per total moles of solvent (D/S) is sufficient to deactivate the dehydration.[36] Our results for non-
catalytic reactions performed for 24 hours indicate that even in presence of significant amounts
of water – D/S molar compositions less than 0.70 – the dehydration reaction is still active in the
absence of catalyst. In comparison, adding 10 mol% Amberlyst-15 increases the conversion and
selectivity relative to the non-catalytic case, as shown in Figure 1b. Indeed, Amberlyst-15 results
in higher fructose conversion and HMF selectivity at lower DMSO concentrations less than a D/S
of 0.43, indicating that a heterogeneous catalyst can improve the reaction without the need of a
highly DMSO concentrated co-solvent.
Given that previous work has demonstrated DMSO degradation to strong acids for
dehydration reactions,[40] it is possible that the observed conversion in the absence of catalyst
could be caused by the generation of acidic species during the reaction. The results presented in
Figure 1 support that acidic species are beneficial for this reaction. Acid formation is investigated
by base titration experiments and HPLC analysis of non-catalytic reaction solutions quenched
after 24 hours of reaction time. Several reaction solutions are examined at DMSO concentrations
of 0, 0.04, 0.20, and 1 D/S. The titration results depicted in Figure 2 show that as DMSO
concentration increases, the amount of acidic species produced increases. The inflection point of
the titration curve, labeled by the black data points in Figure 2, indicates the point at which the
acid is neutralized by the base. Considering the concentration and volume of base added as well
as the amount to which the solution is diluted, the concentration of acid present in solution can be
calculated and is listed in Table 1. The total acid concentration is determined to monotonically
increase with increasing D/S, ranging from 0.01 to 0.06 mmol H+ per gram solution for D/S ratios
of 0 to 1, respectively. It is known that organic acids are commonly generated in the reaction from
HMF rehydration,[6] however, it has not been explicitly verified if DMSO degradation products such
as sulfuric acid contribute to solution acidity. If sulfuric acid is present, it could potentially be
7
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
responsible for the high activity observed in DMSO solvent for the fructose dehydration reaction
in absence of catalyst.
Identifying acidic species formed in non-catalytic reactions in presence of DMSO
The origin of these acidic species could potentially be associated with DMSO degradation
into sulfuric acid, as has been observed for dehydration reactions in DMSO at high temperatures
in the presence of air.[40] Thus, the following experiments are done to investigate if sulfuric acid is
being produced in situ at conventional reaction conditions for the fructose dehydration reaction.
For this work, it is important to note that 24 hours is well beyond the reaction time required to
obtain complete conversion of fructose in DMSO/water co-solvents, increasing the likelihood for
HMF rehydration to organic acids. The 24-hour reaction time is chosen to monitor any possible
evidence of DMSO degradation, as it would be important for a commercial process that reuses
DMSO. An initial test examines the stability of DMSO at 120°C without added fructose. After
subjecting the DMSO/water mixture to reaction conditions in presence of air, the resultant
solutions are quenched in an ice bath before titrating with base. The titration results for
DMSO/water co-solvents in Figure S1a show that the solvent solutions remain neutral after 24
hours of heating at 120°C without fructose in solution. Additional experiments are performed that
involve heating the sample for 24 hours at 150 and 180°C, as shown in Figure S1b. These
mixtures remain neutral. Only after heating DMSO to 190°C for 48 hours does the solution
become acidic, which is consistent with previous literature.[40] Overall, these results indicate that
DMSO does not thermally decompose to measurable quantities of acidic species at reaction
temperatures up to 180°C in presence of air.
Acidic species do arise in the presence of fructose and could still be associated with
DMSO degradation. To determine if DMSO degradation occurs in the presence of air and
fructose, previous research has performed the fructose dehydration in DMSO in presence of base,
such as sodium bicarbonate, which would neutralize strong acidic products from DMSO
decomposition and ultimately inhibit the reaction.[19] However, the presence of base does not
8
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
accurately capture the reaction environment investigated because base is not present in our
baseline dehydration tests. In addition, the presence of base would introduce hydroxide species,
which affects the fructose dehydration rate and product yield.[11] In addition, base is able to
catalyze the fructose isomerization reaction to glucose.[41,46,47] As an alternative, the post-reaction
mixtures are analyzed using a barium chloride precipitation test that is specific for sulfate ions in
solution. While DMSO degradation can form methane sulfonic acid for which this test is not
sensitive, previous work has demonstrated that methane sulfonic acid forms along with sulfuric
acid.[40] This precipitation test is a way to test for sulfuric acid, a DMSO decomposition product,
post-reaction without changing the initial reaction environment. The precipitation of barium sulfate,
a white crystalline solid, is a positive test for sulfate ions and would confirm the presence of sulfuric
acid, a byproduct of DMSO degradation at high temperatures. For pure water, the limit of detection
(LOD) for the precipitation of barium sulfate is found to be 0.002 mmol per gram solution, a factor
of five lower than the acid concentration measured via titration. After 24 hours of reaction time,
barium chloride salt is added to non-catalytic reaction mixtures heated at 120°C for 24 hours with
the following fractional DMSO content: 0, 0.04, 0.20, 0.43, 1.00 D/S. No solid precipitates out of
solution for 0, 0.04, or 0.20 D/S reaction solutions. However, a brown solid precipitate is obtained
for 0.43, and 1 D/S reaction mixtures. This solid is analyzed using XRD and is determined to be
a non-crystalline substance, as depicted in Figure 3. No peak characteristic of barium sulfate is
present in the spectrum. A control test is performed to determine the sensitivity of the XRD
instrument for low concentrations of barium sulfate. One gram of a 0.002 mmol/g sulfuric acid
solution is intentionally added to 1 gram of a 1 D/S reaction mixture quenched after 24 hours of
reaction at 120°C. After adding barium chloride salt to the resultant mixture, the precipitated solid
is analyzed with XRD. The crystalline diffraction peaks observed are characteristic of barium
sulfate, demonstrating that the XRD analysis is able to detect low concentrations of barium sulfate
that is precipitated from a solution containing a low concentration of sulfate ions. Thus, the barium
chloride precipitation test remains sensitive at low concentrations of sulfuric acid even in the
9
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
224
225
226
227
presence of a complex mixture of fructose, HMF, and other reaction byproducts. Additional tests
using a polymer containing functional groups that are basic in nature are described in the SI.
These adsorption tests were negative for the presence of either sulfuric acid or methane sulfonic
acid.
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
As further evidence that barium sulfate is not present, the precipitates are analyzed using
FTIR since BaSO4 has an intense S=O peak. As a reference, the FTIR spectra are shown of the
brown solids produced from reaction mixtures of 0.43 and 1 D/S concentrations without the
addition of BaCl2 (shown in Figure 4). Additionally, humins produced in the absence of DMSO
(D/S = 0) are also analyzed. Since reactions in water result in low fructose conversion, the humins
produced in pure water are derived from reaction mixtures starting from HMF. The spectrum for
humins produced in pure water is similar to humins derived from fructose in DMSO and water
mixtures, including several peaks in the range of 1000 – 1200 cm-1. These spectra of humins are
similar to previous reports that analyzed humins formed during the dehydration reaction.[34,48–52]
After establishing baseline spectra for humins, the FTIR spectrum for BaSO4 precipitated from
0.1 M sulfuric acid solution is analyzed. The BaSO4 precipitated from the sulfuric acid solution
contained a very intense sulfoxide peaks at 1042 and 600 cm-1 that have similar intensity and can
serve as indicators of BaSO4. With these as standards, the reaction mixture with D/S = 1
containing fructose is mixed with BaCl2, and the resultant precipitate is analyzed with FTIR. While
the spectrum changes in the range of 1000 – 1200 cm-1, the spectrum lacks the characteristic
peak at 600 cm-1 that would be indicative of BaSO4 and the presence of sulfuric acid. This analysis
remains sensitive to sulfuric acid since the intentional addition of sulfuric acid to the mixture results
in clear peaks near 1042 cm-1 and 600 cm-1 that have the expected ratio of intensities to indicate
BaSO4 is present. Since the S=O peak for BaSO4 is strong, trace amounts of BaSO4 would be
observed in the spectrum of the collected solid. Therefore, it is concluded that sulfuric acid is not
10
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
being generated in situ from DMSO degradation, indicating that sulfoxide groups are stable under
conventional reactions conditions for biomass conversion reactions.
The barium chloride precipitation test as well as FTIR analysis of resulting precipitate is
repeated for non-catalytic reactions performed in a DMSO/water co-solvent of 0.20 D/S fractional
DMSO content and in pure DMSO at 150°C, a temperature commonly used for fructose
dehydration.[19,21] Even for reaction mixtures subjected to this higher temperature, the barium
chloride test does not result in barium sulfate, indicating that DMSO remains stable at 150°C for
24 hours as well. XRD analysis shown in Figure S3a of the precipitated solid obtained from 0.20
and 1 D/S reaction solutions find the precipitate to be amorphous, indicating no sulfuric acid
production. IR peaks associated with barium sulfate are also absent from FTIR spectra in Figure
S3b of the solids precipitated from reaction solutions heated at 150°C. These IR spectra also
resemble the spectra of humins that is reported in literature. However, the intensity of the 1670
cm-1 peak relative to other peaks in the same spectrum is larger than what is observed in IR
spectra of humins synthesized at 120°C. Previous literature has suggested that the 1670 cm-1
peak corresponds to the C=O of HMF species incorporated into the humin structure.[52] The larger
relative peak intensity for humins formed at 150°C suggests that a greater concentration of HMF
is being incorporated into humins formed at 150°C than in humins formed at 120°C.
Since it is shown that sulfuric acid is not being detected from reaction mixtures subjected
to conventional reaction conditions, the nature of the acid species is investigated by comparing
organic acid yield as obtained by HPLC analysis to the overall acid yield as calculated from
titration experiments discussed previously. The results of this analysis are listed in Table 1. For
0, 0.04, and 1 D/S reaction mixtures, the total concentration of organic acids is comparable to the
concentrations obtained by titration. Therefore, organic acids are the predominant acids produced
during these non-catalytic reactions. HPLC analysis of DMSO/water reaction solutions starting
with HMF as a reactant confirms that the organic acids are derived from HMF rehydration as
expected. Interestingly, while monitoring HMF conversion at various fractional DMSO contents in
11
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
water for 24 hours at 120°C, it is found that low fractional content of DMSO is sufficient to prevent
further HMF conversion to byproducts in absence and in presence of acid catalyst. Indeed, high
HMF conversion is observed at low DMSO concentrations in water. However, as shown in Figure
5, HMF is not significantly converted after 24 hours at 120°C for D/S ≥ 0.11 in non-catalytic
reactions solutions. Even in presence of 10 mol% Amberlyst-15, an optimal D/S ratio of 0.20 to
0.43 provides sufficient HMF stabilization, as shown in Figure 5. Therefore, a DMSO
concentration of 0.20 to 0.43 D/S could potentially be a target range for DMSO or sulfoxide
incorporation needed to obtain optimal HMF selectivity in a reaction system or catalyst design.
Limiting the amount of DMSO used as a co-solvent or designing a heterogeneous catalyst with
sulfoxide species that exhibit the same beneficial solvation interactions is crucial to improving the
sustainability of the process while maintaining high HMF yields.
Comparing the organic acid yield and overall acid yield as listed in Table 1, a significant
discrepancy exists between the amount of acid titrated and the amount of organic acid quantified
by HPLC for reactions in a co-solvent of 0.20 fractional DMSO content in water, suggesting
another acidic species exists in solution. Another potential product in solution to consider is
humins, whose structure and functionality has been the subject of recent research.[33,34,48,49,51–54]
Research investigating the structure of insoluble humins found that its IR spectrum had several
peaks in the carbonyl region, which are consistent with carboxylic acid functional groups.[49,51,53]
Some proposed humin formation pathways have suggested that insoluble and soluble fractions
of the humins structure could potentially incorporate levulinic acid.[48,54] Therefore, it is suggested
that soluble and insoluble humins formed during the reaction are also being titrated in solution,
thus making up for the discrepancy between acid yield and organic acid yield for 0.20 D/S reaction
solutions. A more detailed discussion about evidence supporting the acidity of humins can be
found in the Supplementary Information.
To determine if the insoluble humins are acidic, solids isolated from 0.43 and 1.00 D/S
reaction mixtures are suspended in 10 mL of 2 M NaCl solution to perform a cation exchange with
12
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
surface acidic groups to release acidic protons into solution in the form of HCl. The solids are
filtered out of the salt solution and the filtrate is titrated with 7.5 mM NaOH using phenolphthalein
as an indicator. For both humins synthesized in pure DMSO and in 0.43 D/S DMSO/water co-
solvents, about 1 mL of titrant was needed to neutralize the acid, indicating that the insoluble
humins indeed have acidic functional groups incorporated in their structure. This result is
consistent with previous research, which has suggested that insoluble and soluble fractions of the
humins structure could potentially incorporate levulinic acid.[48,54]
Reactivity of acidic species
Experiments are done to determine if the organic acids formic acid and levulinic acid
contribute to the fructose conversion observed in non-catalytic reactions in DMSO. The reactions
are performed in presence of 10 mol% of either formic acid or levulinic acid, which is comparable
to the maximum organic acid formed in non-catalytic reactions according to HPLC analysis. The
results in Figure 6 show that in both water and DMSO after 1 hour of reaction time neither organic
acid significantly impacts fructose conversion or HMF selectivity as compared to the non-catalytic
reaction. However, in presence of 0.002 mmol sulfuric acid per gram solution in both water and
DMSO, conversion and selectivity are significantly affected. Thus, if sulfuric acid is being
generated in non-catalytic reactions in DMSO, even at low concentrations, a comparable
conversion should be observed. Since levulinic acid does not impact fructose conversion in either
water or DMSO, this suggests that soluble and insoluble humins would not be reactive for the
dehydration reaction although contribute to solution acidity.
Solvent reactivity under de-aerated conditions
Finally, another common perspective about the origin of the activity of DMSO is that radical
species generated under aerated conditions and high temperature cause DMSO degradation to
strong acid.[39,40,45] Previous research provides evidence for this claim by de-aerating DMSO and
performing the reaction under nitrogen or evacuated conditions, which significantly decreases
HMF yield.[20,21] A set of experiments are designed to determine the effect of oxygen on non-
13
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
catalytic fructose dehydration and acid production in pure DMSO by de-aerating the solvent via
freeze/thaw cycles and preforming the reaction under a nitrogen or evacuated atmosphere. The
results are summarized in Table 2. It is found that in absence of air and dissolved oxygen, fructose
conversion and acid production decreases, as shown in Table 2 and Figure 7, respectively.
Interestingly, performing the reaction under nitrogen results in 74% fructose conversion and 43%
HMF selectivity and significantly less acid production. The source of the acidity generated under
nitrogen is consistent with the formation of levulinic acid and formic acid at concentrations of
0.0005 M. This concentration is below the sensitivity limit of the HPLC for these compounds.
Experiments detailed earlier in this study above distinctly show that the only acids generated in
solution under aerated conditions are organic acid byproducts from HMF rehydration and humins,
not DMSO degradation products such as sulfuric acid. Thus, oxygen seems to have a role in
forming organic acids. Under aerated conditions, a higher HMF selectivity of 78% is obtained at
a higher fructose conversion of 100%.
Ultimately, the tests discussed above provide experimental evidence that no strong acid
products from DMSO decomposition are produced in reaction solutions subjected to conventional
biomass conversion conditions. Additionally, organic acids produced during the reaction – formic
acid, levulinic acid, and humins – have minimal impact on fructose conversion. These insights
suggest that the role of DMSO in promoting the fructose dehydration reaction in aerated DMSO
in absence of catalyst is not the generation of strongly acidic species such as sulfuric acid. This
is an important insight that suggests that DMSO stably promotes the fructose dehydration reaction
mainly by solvation effects and does not degrade to strong acid at temperatures less than 150°C.
Conclusions
The dehydration of fructose to HMF is investigated in mixtures of DMSO and water to
determine the effect of fractional DMSO content in a DMSO/water co-solvent on fructose
conversion, HMF selectivity, and acid formation. HPLC analysis of these non-catalytic reactions
showed that fructose conversion and HMF selectivity increase with increasing DMSO composition.
14
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
Titration results showed that DMSO/water co-solvents remain neutral after heating at 120°C for
24 hours. However, titration of reaction solutions with fructose quenched at 24 hours reveal that
the reaction solution becomes more acidic with increasing DMSO concentration. A barium
chloride precipitation test along with XRD and FTIR analysis reveal that sulfuric acid is not
produced from DMSO degradation in reaction mixtures starting from fructose. Therefore, the only
acidic species that are produced are formic acid, levulinic acid, and soluble and insoluble humins.
Fructose dehydration reactions done in presence of either levulinic acid or formic acid in DMSO
show that organic acid byproducts produced from HMF are not the predominant contributors to
the fructose conversion observed in DMSO in absence of acid catalyst. This experimental
evidence supports that the dehydration of fructose to HMF is promoted mainly by DMSO solvation
effects and not by catalysis of acidic species generated in solution. Indeed, the experimental
evidence suggest that sulfuric acid is not formed at the reaction conditions tested. Additional
experiments test the stability of HMF in DMSO/water co-solvents of varying DMSO content. An
optimal range of 0.20 to 0.43 molar fraction of DMSO per total moles of solvent is found to stabilize
HMF in solution while minimizing the amount of DMSO needed in the solvent. Overall, this work
demonstrates that sulfoxide species are stable under aerated reaction conditions for 24-hour
batch reactions, suggesting that the primary role of DMSO is promoting the dehydration via
solvation effects not through degradation into strongly acidic species.
370
371
Experimental Section
Chemicals
372
373
374
375
All chemicals are used as received, including: D-fructose (99%, Alfa Aesar); 5-
hydroxymethylfurfural (98%, Ark Pharm Inc.); dimethyl sulfoxide (DMSO; ³ 99.5%, GC, plant cell
culture tested, Sigma Aldrich); barium chloride (BaCl2; ACS grade, BDH VWR analytical);
concentrated sulfuric acid (Fischer); Amberlyst-15 H wet ion exchange resin beads (Alfa Aesar);
15
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
376
377
and de-ionized (DI) water is obtained from a tap connected to house supply that produces Milli-Q
quality water.
378
Fructose dehydration and HMF conversion reactions
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
A stock solution of 0.5 mmol fructose or HMF per gram solution is prepared by weighing
out water and DMSO accordingly to achieve the desired D/S molar ratio (defined as moles DMSO
per total moles of solvent (DMSO and water)), as listed in Table S1. All reactions are performed
for 24 hours at 120°C and an agitation rate of 420 RPM in a 15 mL thick walled glass pressure
tube sealed with a Teflon cap containing a back-sealing O-ring. The reaction is performed either
without acid catalyst or in presence of 10 mol% of Amberlyst-15 (titrated acid loading of 3 mmol/g).
Before starting the reaction, the pressure tube with the cap and a stir bar inside is weighed. Two
grams of the prepared stock solution are added to the pressure tube before immersing the sealed
tube in a silicone oil bath heated to 120°C just enough so that the level of the oil bath is just above
the liquid level of the reaction solution in the tube. After 24 hours, the tube is removed from the
oil bath and is immersed in an ice bath to quench the reaction. After the pressure tube has cooled
to room temperature, the tube with the cap on is weighed to demonstrate that no mass is lost.
The reaction solution is diluted by a factor of three in the tube with DI water before the tube is
capped and shaken vigorously. The diluted solution is filtered with a 0.22 µm syringe filter, and
then 1.5 grams of the filtrate is placed in a vial and analyzed with High Performance Liquid
Chromatography (HPLC). A Waters Sugar Pak 1 column equipped with the corresponding guard
column is used to separate components at 70°C using a flow rate of 0.6 mL/min of DI water with
an injection volume of 10 µL. Fructose and HMF concentrations are determined by using
Refractive Index and UV-Vis detectors, respectively. For reactions performed under either a
nitrogen or an evacuated atmosphere, a similar procedure is followed except using a Schlenk
tube as the reaction vessel. The reaction solutions are first degassed by three freeze/thaw degas
cycles, and then the Schlenk tube is either backfilled with nitrogen or left under vacuum before
16
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
401
402
403
immersing in an oil bath heated to 120°C for 24 hours. To quantify levulinic acid and formic acid,
a Waters Ion Exclusion column is used to separate components at 70°C at a flow rate of 0.2
mL/min of 1 mM 1-octanesulfonic acid using an injection volume of 10 µL.
404
Quantification of acid via titration
405
406
407
408
409
410
411
412
413
After the reaction mixture is quenched, it is diluted by a factor of 10 with DI water. The
diluted solution is placed in a 120 mL beaker containing a magnetic stir bar and a calibrated
Mettler Toledo pH probe. After recording the starting pH, the solution is titrated with 7.5 mM NaOH.
The titrant is added dropwise while recording the change in pH after every few drops. As the
mixture reaches a pH of 5, the pH is recorded after addition of 1-2 drops to help capture the
inflection point. After pH 9, the number of drops added between pH recordings is increased. The
titration is ended once a pH of 11 is reached. The pH versus volume curves are generated from
this data. The curves are fit with a sigmoid function to calculate the inflection point that
corresponds to the concentration of acid in the reaction mixture.
414
Barium chloride precipitation test for sulfate ions
415
416
417
418
419
420
421
422
423
After the reaction mixture is quenched, barium chloride (24 mg) salt is added to the
reaction mixtures with D/S of 0, 0.04, 0.20, 0.43, and 1.00. For 0.43 and 1.00 D/S solutions, water
(1 g) is added to dissolve the salt completely. If solid precipitates, the solution is diluted to 20
grams and is then centrifuged. The water is decanted and a suspension of the solid in water is
dried overnight in the hood on an XRD sample holder. The solid is then characterized by powder
X-ray Diffraction using a Bruker D8 X-ray diffractometer (40 kV, 40 mA) to determine if the solid
that precipitates is barium sulfate. The solid is also characterized by ATR-FTIR using a Thermo
Scientific Nicolet iS50 FT-IR with a Smart iTX accessory to determine if any trace barium sulfate
is present.
17
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
424
Humins titration
425
426
427
428
429
430
431
432
433
The humins are precipitated by diluting the reaction mixture 10 times with 0.4 M HCl
solution.[52] The solution is then centrifuged for 5 minutes at 9000 rpm before carefully removing
the supernatant with a pipette leaving the humins behind. To wash the humins, they are
suspended in DI water again, and the centrifugation and supernatant removal process is repeated.
After each wash, the pH of the supernatant is taken. When the pH of the supernatant is above 4,
the suspension is place in a tared vial and dried in the oven. The dry humins are then suspended
in 10 g of 2 M NaCl to exchange the acidic protons for one hour. The suspension is filtered to
remove the humins, and the filtrate is titrated with 7.5 mM NaOH solution with phenolphthalein as
an indicator.
434
435
436
Acknowledgements
437
438
439
We gratefully acknowledge the National Science Foundation (NSF CBET 1605037 and
1653587) and the Ohio State University Institute of Materials Research (OSU IMR FG0211) for
their financial support.
440
441
References
442
443
444
445
[1]
[2]
[3]
M. J. Climent, A. Corma, S. Iborra, Green Chem. 2014, 16, 516–547.
T. Wang, M. W. Nolte, B. H. Shanks, Green Chem. 2014, 16, 548–572.
D. M. Alonso, J. Q. Bond, J. A. Dumesic, J. N. Chheda, G. W. Huber, M. J. Climent, A. Corma, S. Iborra, M. E. Domine, P.
J. Luggren, et al., Catal. Today 2011, 147, 7164–7183.
446
447
[4]
D. S. Park, K. E. Joseph, M. Koehle, C. Krumm, L. Ren, J. N. Damen, M. H. Shete, H. S. Lee, X. Zuo, B. Lee, et al., ACS
Cent. Sci. 2016, 2, 820–824.
448
449
450
[5]
[6]
J. J. Pacheco, M. E. Davis, Proc. Natl. Acad. Sci. U. S. A. 2014, 111, 8363–7.
R.-J. van Putten, J. C. van der Waal, E. de Jong, C. B. Rasrendra, H. J. Heeres, J. G. De Vries, Chem. Rev. 2013, 113,
1499–1597.
451
452
[7]
[8]
J. J. Bozell, G. R. Petersen, Green Chem. 2010, 12, 539–554.
A. I. Torres, P. Daoutidis, M. Tsapatsis, Energy Environ. Sci. 2010, 3, 1560.
18
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
[9]
L. Shuai, J. S. Luterbacher, ChemSusChem 2016, 9, 133–155.
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
[34]
[35]
[36]
[37]
[38]
[39]
[40]
[41]
[42]
H. Kimura, M. Nakahara, N. Matubayasi, J. Phys. Chem. A 2013, 117, 2102.
B. F. M. Kuster, Carbohydr. Res. 1977, 54, 177–183.
D. W. Brown, A. J. Floyd, R. G. Kinsman, Y. Roshan-ali, J. Chem. Tech. Biotechnol. 1982, 32, 920–924.
C. Lansalot-Matras, C. Moreau, Catal. Commun. 2003, 4, 517–520.
C. Moreau, A. Finiels, L. Vanoye, J. Mol. Catal. A Chem. 2006, 253, 165–169.
J. Y. G. Chan, Y. Zhang, ChemSusChem 2009, 2, 731–734.
T. Stahlberg, W. Fu, J. M. Woodley, A. Riisager, ChemSusChem 2011, 4, 451–458.
Y. Roman-Leshkov, J. N. Chheda, J. A. Dumesic, Science 2006, 312, 1933–1937.
Y. J. Pagán-Torres, T. Wang, J. M. R. Gallo, B. H. Shanks, J. A. Dumesic, ACS Catal. 2012, 2, 930–934.
G. R. Akien, L. Qi, I. T. Horváth, Chem. Commun. 2012, 48, 5850.
S. Despax, C. Maurer, B. Estrine, J. Le Bras, N. Hoffmann, S. Marinkovic, J. Muzart, Catal. Commun. 2014, 51, 5–9.
G. S. Svenningsen, R. Kumar, C. E. Wyman, P. Christopher, ACS Catal. 2018, 8, 5591–5600.
S. P. Teong, G. Yi, Y. Zhang, Green Chem. 2014, 16, 2015.
B. F. M. Kuster, Starch-Stärke 1990, 42, 314–321.
S. Frieder W. Lichtenthaler and Ronninger, J. Chem. Soc. Perkin Trans. 1990, 2, 1489–1497.
L. K. Ren, L. F. Zhu, T. Qi, J. Q. Tang, H. Q. Yang, C. W. Hu, ACS Catal. 2017, 7, 2199–2212.
M. S. Feather, J. F. Harris, Adv. Carbohydr. Chem. Biochem. 1973, 28, 161–224.
V. Nikolakis, S. H. Mushrif, B. Herbert, K. S. Booksh, D. G. Vlachos, J. Phys. Chem. B 2012, 116, 11274–11283.
P. Dais, A. Perlin, Carbohydr. Res. 1987, 169, 159–169.
P. Dais, A. S. Perlin, Carbohydr. Res. 1985, 136, 215–223.
A. S. Amarasekara, L. D. Williams, C. C. Ebede, Carbohydr. Res. 2008, 343, 3021–3024.
G. Tsilomelekis, T. R. Josephson, V. Nikolakis, S. Caratzoulas, ChemSusChem 2014, 7, 117–126.
S. H. Mushrif, S. Caratzoulas, D. G. Vlachos, Phys. Chem. Chem. Phys. 2012, 14, 2637.
Horvat Jaroslav, Klaic Branimir, Metelko Biserka, Sunjic Vitomir., Croat. Chem. Acta 1986, 59, 429–438.
S. K. R. Patil, J. Heltzel, C. R. F. Lund, Energy & Fuels 2012, 26, 5281–5293.
R. M. Munavu, R. M. Musau, Biomass 1987, 13, 67–74.
S. Jia, Z. Xu, Z. C. Zhang, Chem. Eng. J. 2014, 254, 333–339.
M. A. Mellmer, C. Sanpitakseree, B. Demir, P. Bai, K. Ma, M. Neurock, J. A. Dumesic, Nat. Catal. 2018, 1, 199–207.
J. Zhang, A. Das, R. S. Assary, L. A. Curtiss, E. Weitz, Appl. Catal. B Environ. 2016, 181, 874–887.
D. L. Head, C. G. McCarty, Tetrahedron Lett. 1973, 1405–1408.
T. M. Santosusso, D. Swern, J. Org. Chem. 1976, 41, 2762–2768.
N. Deshpande, L. Pattanaik, M. R. Whitaker, C.-T. Yang, L. Lin, N. A. Brunelli, J. Catal. 2017, 353, 205–210.
486
487
N. Rajabbeigi, A. I. Torres, C. M. Lew, B. Elyassi, L. Ren, Z. Wang, H. Je Cho, W. Fan, P. Daoutidis, M. Tsapatsis, Chem.
Eng. Sci. 2014, 116, 235–242.
488
489
490
491
492
493
494
495
[43]
[44]
[45]
[46]
[47]
[48]
[49]
[50]
[51]
F. K. Kazi, A. D. Patel, J. C. Serrano-Ruiz, J. A. Dumesic, R. P. Anex, Chem. Eng. J. 2011, 169, 329–338.
D. L. Head, C. G. Mccarty, Tetrahedron Lett. 1973, 16, 1405–1408.
C. Chao-Tung, Y. Shou-Jen, Tetrahedron Lett. 1969, 10, 3855–3856.
J. M. Carraher, C. N. Fleitman, J.-P. Tessonnier, ACS Catal. 2015, 5, 3162–31733.
C. Liu, J. M. Carraher, J. L. Swedberg, C. R. Herndon, C. N. Fleitman, J.-P. Tessonnier, ACS Catal. 2014, 4, 4295–4298.
Z. Cheng, J. Everhart, G. Tsilomelekis, V. Nikolakis, B. Saha, D. Vlachos, Green Chem. 2018, DOI 10.1039/C7GC03054A.
S. Reiche, N. Kowalew, R. Schlögl, ChemPhysChem 2014, 16, 579–587.
I. V. Sumerskii, S. M. Krutov, M. Y. Zarubin, Russ. J. Appl. Chem. 2010, 83, 320–327.
496
497
I. van Zandvoort, E. J. Koers, M. Weingarth, P. C. A. Bruijnincx, M. Baldus, B. M. Weckhuysen, Green Chem. 2015, 17,
4383–4392.
498
[52]
G. Tsilomelekis, M. J. Orella, Z. Lin, Z. Cheng, W. Zheng, V. Nikolakis, D. G. Vlachos, Green Chem. 2016, 18, 1983–1993.
19
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
499
500
501
502
[53]
[54]
I. Van Zandvoort, Y. Wang, C. B. Rasrendra, E. R. H. van Eck, P. C. A. Bruijnincx, H. J. Heeres, B. M. Weckhuysen,
ChemSusChem 2013, 6, 1745–1758.
C. R. Heltzel, J., Patil, S. K., Lund, Reaction Pathways and Mechanisms in Thermocatalytic Biomass Conversion II,
Springer, 2016.
503
504
20
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
505
Captions for Figures and Schemes
506
507
508
Scheme 1. Acid catalyzed mechanism for fructose dehydration to HMF involving the sequential
loss of three water molecules.[19]
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
Figure 1. Comparison of the fructose conversion and HMF selectivity as a function of the mole
fraction of DMSO per moles of solvent (D/S) concentration in (a) the absence of catalyst and (b)
the presence of 10 mol% Amberlyst-15. All reactions are performed for 24 hours at 120°C and
an agitation rate of 420 RPM starting with a fructose concentration 0.5 mmol/g solution. The
error bars represent the standard deviation from multiple reaction samples.
Figure 2. Comparison of the titration curves (pH versus volume titrant volume) for reaction
mixtures (diluted by a factor of 10) that had no catalyst added in DMSO/water co-solvents of
D/S of 0, 0.04, 0.20, 1. The titrant is a 7.5 mM NaOH solution that is added dropwise to 15
grams of the diluted reaction mixture.
Figure 3. Comparison of XRD spectra of solids precipitated from reaction solutions of 0.43 and
1.00 D/S to a BaSO4 product obtained from the reaction of barium chloride and 0.1 M sulfuric
acid. The green spectrum corresponds to the solid precipitated from a reaction solution of 1.00
D/S with 0.002 M sulfuric acid upon addition of barium chloride.
Figure 4. Comparison of ATR-FTIR spectra of humins precipitated by diluting reaction solutions
with 0.4 M HCl (D/S = 0.43 and D/S = 1), humins produced from HMF in water (D/S = 0), humins
precipitated by diluting the reaction solution with 0.1 M barium chloride, humins precipitated by
addition of barium chloride to a reaction mixture spiked with sulfuric acid (control test), and barium
sulfate synthesized by adding 0.1 M barium chloride to a 0.1 M solution of sulfuric acid.
Figure 5. Comparison of the HMF conversion as a function of the mole fraction of DMSO per
moles of solvent (D/S) concentration in (a) the absence of catalyst and (b) the presence of 10
mol% Amberlyst-15. All reactions are performed for 24 hours at 120°C and an agitation rate of
420 RPM starting with a fructose concentration 0.5 mmol/g solution. The error bars represent
the standard deviation from multiple reaction samples.
Figure 5. Comparison of A. fructose conversion and B. HMF selectivity obtained without added
acid catalyst and with the addition of 10 mol% formic acid (FA), levulinic acid (LA), or sulfuric
acid (SA) in water (red) or DMSO (blue) for 24 and 1 hours of reaction time respectively. All
reactions are performed at 120°C and an agitation rate of 420 RPM starting at a fructose
concentration 0.5 mmol/g solution. The error bars represent the standard deviation from multiple
reaction samples.
Figure 6. Comparison of A. fructose conversion and B. HMF selectivity obtained without added
acid catalyst and with the addition of 10 mol% formic acid (FA), levulinic acid (LA), or sulfuric
acid (SA) in water (red) or DMSO (blue) for 24 and 1 hours of reaction time respectively. All
reactions are performed at 120°C and an agitation rate of 420 RPM starting at a fructose
concentration 0.5 mmol/g solution. The error bars represent the standard deviation from multiple
reaction samples.
Figure 7. pH versus titrant volume curves for non-catalytic reaction mixtures in pure DMSO
quenched at 24 hours for different atmospheric conditions: under air, nitrogen, and vacuum. All
21
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
554
555
556
557
reactions are performed in a Schlenk tube for 24 hours at 120°C and an agitation rate of 420
RPM starting at a fructose concentration 0.5 mmol/g solution.
22
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
558
559
560
Tables:
Table 1. Comparison of the Quantification of acidic species starting from fructose by titration
D/S
Fructose
Conversion
[%]a
HMF
Yield
[%][a]
Levulinic and
Formic Acid
Concentration
[mmol/g] [b]
0.01
Acid
Concentration
from Titration
[mmol/g] [c]
0.01
0.00
0.04
0.20
1.00
7
20
53
100
1
7
30
0.02
0.02
0.05
0.03
0.05
0.06
77
561
562
563
564
565
566
[a] Obtained from HPLC analysis of non-catalytic reaction mixtures starting with fructose.
[b] Combined concentration of levulinic and formic acid from HPLC analysis of non-catalytic
reaction mixtures starting from fructose.
[c] Obtained from titration of non-catalytic reactions starting from fructose.
Table 2. Fructose dehydration in pure DMSO under inert atmosphere
Atmosphere
Fructose Conversion
HMF
HMF Yield
[%]
Selectivity [%]
[%]
Vacuum
Nitrogen
Air
1
74
100
0
43
78
0
32
77
567
568
23
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
569
O
O
OH
H
HO
HO
HO
HO
O
OH
OH
H+
-H2
O
H+
-H2
O
H+
-H2
O
HO
HO
O
O
O
HO
HO
HO
570
571
572
573
574
Scheme 1. Acid catalyzed mechanism for fructose dehydration to HMF involving the sequential
loss of three water molecules.[19]
24
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
575
576
577
578
579
580
581
582
583
Figure 1. Comparison of the fructose conversion and HMF selectivity as a function of the mole
fraction of DMSO per moles of solvent (D/S) concentration in (a) the absence of catalyst and (b)
the presence of 10 mol% Amberlyst-15. All reactions are performed for 24 hours at 120°C and
an agitation rate of 420 RPM starting with a fructose concentration 0.5 mmol/g solution. The
error bars represent the standard deviation from multiple reaction samples.
25
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
584
585
586
587
588
589
590
591
Figure 2. Comparison of the titration curves (pH versus volume titrant volume) for reaction
mixtures (diluted by a factor of 10) that had no catalyst added in DMSO/water co-solvents of
D/S of 0, 0.04, 0.20, 1. The titrant is a 7.5 mM NaOH solution that is added dropwise to 15
grams of the diluted reaction mixture.
26
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
592
593
594
595
596
597
Figure 3. Comparison of XRD spectra of solids precipitated from reaction solutions of 0.43 and
1.00 D/S to a BaSO4 product obtained from the reaction of barium chloride and 0.1 M sulfuric
acid. The green spectrum corresponds to the solid precipitated from a reaction solution of 1.00
D/S with 0.002 M sulfuric acid upon addition of barium chloride.
27
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
598
599
600
601
602
603
604
Figure 4. Comparison of ATR-FTIR spectra of humins precipitated by diluting reaction solutions
with 0.4 M HCl (D/S = 0.43 and D/S = 1), humins produced from HMF in water (D/S = 0), humins
precipitated by diluting the reaction solution with 0.1 M barium chloride, humins precipitated by
addition of barium chloride to a reaction mixture spiked with sulfuric acid (control test), and barium
sulfate synthesized by adding 0.1 M barium chloride to a 0.1 M solution of sulfuric acid.
28
This article is protected by copyright. All rights reserved.

10.1002/cssc.201803013
ChemSusChem
605
606
607
608
609
610
611
Figure 5. Comparison of the HMF conversion as a function of the mole fraction of DMSO per
moles of solvent (D/S) concentration in (a) the absence of catalyst and (b) the presence of 10
mol% Amberlyst-15. All reactions are performed for 24 hours at 120°C and an agitation rate of
420 RPM starting with a fructose concentration 0.5 mmol/g solution. The error bars represent
the standard deviation from multiple reaction samples.
29
This article is protected by copyright. All rights reserved.