- Email: [email protected]
Direct catalytic production of sorbitol from waste cellulosic materials
Accepted Manuscript Direct catalytic production of sorbitol from waste cellulosic materials Lucília Sousa Ribeiro, José J. de Melo Órfão, Manuel Fernando Ribeiro Pereira PII: DOI: Reference:
S0960-8524(17)30115-3 http://dx.doi.org/10.1016/j.biortech.2017.02.008 BITE 17571
To appear in:
Bioresource Technology
Received Date: Revised Date: Accepted Date:
5 December 2016 2 February 2017 3 February 2017
Please cite this article as: Ribeiro, L.S., de Melo Órfão, J.J., Pereira, M.F.R., Direct catalytic production of sorbitol from waste cellulosic materials, Bioresource Technology (2017), doi: http://dx.doi.org/10.1016/j.biortech. 2017.02.008
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
Direct catalytic production of sorbitol from waste
2
cellulosic materials
3
Lucília Sousa Ribeiro, José J. de Melo Órfão, Manuel Fernando Ribeiro Pereira*
4 5
Laboratório de Processos de Separação e Reação - Laboratório de Catálise e
6
Materiais (LSRE-LCM), Departamento de Engenharia Química, Faculdade de
7
Engenharia, Universidade do Porto, Rua Dr. Roberto Frias, 4200-465 Porto, Portugal
8 9
[email protected], [email protected], [email protected]*
10 11
*Corresponding author:
12
Manuel Fernando Ribeiro Pereira
13
Tel. +351 225 081 468
14
Fax: +351 225 081 449
15 16
Abstract
17
Cotton wool, cotton textile, tissue paper and printing paper, all potential waste cellulosic
18
materials, were directly converted to sorbitol using a Ru/CNT catalyst in the presence
19
of H2 and using only water as solvent, without any acids. Conversions up to 38% were
20
attained for the raw substrates, with sorbitol yields below 10%. Ball-milling of the
21
materials disrupted their crystallinity, allowing reaching 100% conversion of cotton
22
wool, cotton textile and tissue paper after 4 h, with sorbitol yields around 50%. Mix-
23
milling these materials with the catalyst greatly enhanced their conversion rate, and the
24
materials were efficiently converted to sorbitol with a yield around 50% in 2 h. However,
25
ball- and mix-milled printing paper presented a conversion of only 50% after 5 h, with
1
26
sorbitol yields of 7%. Amounts of sorbitol of 0.525, 0.511 and 0.559 g could be
27
obtained from 1 g of cotton wool, cotton textile and tissue paper, respectively.
28
Keywords: biomass conversion; cellulosic materials; cotton; paper; sorbitol
29
Abbreviations:
30
CNT – carbon nanotubes
31
DTG – differential thermogravimetry
32
HPLC – high performance liquid chromatography
33
RI – refractive index
34
TG – thermogravimetry
35
TOC – total organic carbon
36
XRD – X-ray diffraction
2
37
1. Introduction
38
Concerns about global warming have motivated the search for alternative renewable
39
resources, leading to a significant increase in research activities directed towards their
40
use (Byun and Han, 2016a, 2016b; Deng et al., 2015; Kobayashi et al., 2014).
41
Currently, about 10% of the world’s primary energy is biomass, which is mostly used to
42
generate power and heat (Li et al., 2015). Lignocellulose is the most abundant and less
43
expensive type of biomass on earth, therefore being a promising feedstock for the
44
production of renewable energy, especially biofuels, and chemicals (Deng et al., 2015;
45
Feng et al., 2016). Unlike corn and stalk, lignocellulose is inedible for humans, and its
46
use will not impose a negative impact on food supplies since it does not compete with
47
food production (Li et al., 2015; Yamaguchi et al., 2016). Lignocellulose consists of 35-
48
50% cellulose (a polymer of D-glucose), 25-30% hemicellulose (a polymer of C5 and C6
49
sugars) and 15-30% lignin (Deng et al., 2015). The hydrolytic hydrogenation of these
50
sugar polymers produces sugar alcohols, which are used as low-calorie and low-
51
cariogenic sweeteners and sugar substitutes for diabetics, as humectants in cosmetic
52
and pharmaceutical products, in paper and tobacco and as precursors to plastics
53
(Kobayashi et al., 2014; Kusserow et al., 2003; Mäki-Arvela et al., 2011; Rao et al.,
54
2016). Sorbitol is one of the most important sugar alcohols and can serve as platform
55
chemical for the synthesis of various value-added chemicals such as glycerol, glycols,
56
lactic acid, isosorbide, 1,4-sorbitan and L-sorbose (Deng et al., 2015). The annual
57
production of sorbitol has already reached 6.5×105 ton (Kobayashi et al., 2014).
58
Therefore, the catalytic hydrolytic hydrogenation of biomass and its components in the
59
presence of hydrogen has greatly attracted attention for the establishment of a
60
sustainable society, because it circumvents metastable glucose and allows high
61
selectivity to high-value chemicals or fuels (Zhao et al., 2015).
62
The catalytic conversion of cellulose to sorbitol using supported metal catalysts in the
63
presence of hydrogen has already been reported by some research groups (Deng et
3
64
al., 2009; Ding et al., 2010; Fukuoka and Dhepe, 2006; Kobayashi et al., 2011; Luo et
65
al., 2007; Ribeiro et al., 2015b; Romero et al., 2016; Van de Vyver et al., 2010; Van de
66
Vyver et al., 2012; Yang et al., 2012). Carbon nanotubes, which have been gaining
67
increasing attention as supports in heterogeneous catalysis, have already shown to be
68
the most effective support for the direct conversion of cellulose to sorbitol (Deng et al.,
69
2009; Ribeiro et al., 2017; Wang et al., 2012). Also, Ru catalysts have shown to be the
70
most effective in the direct conversion of cellulose into polyols (Deng et al., 2009; Guo
71
et al., 2014; Ribeiro et al., 2015a; Wu et al., 2012). However, pure cellulose has to be
72
obtained from lignocellulose, normally using strong acid/base catalysts to completely
73
remove the hemicellulose and lignin components. Thus, the direct conversion of the
74
individual components in woody biomass to valuable chemicals is of great importance
75
to open new possibilities of using biomass (Yamaguchi et al., 2014). Palkovits et al.
76
reported the conversion of spruce chips using supported metal catalysts combined with
77
sulfuric/phosphoric acid with a yield of 55% (based on cellulose) of sugar alcohols
78
(sorbitol, sorbitan, isosorbide) (Palkovits et al., 2010). However, the use of acids has
79
problems associated with corrosion of the reactor or neutralization processes for the
80
removal of the acid, and so should be avoided. Beeck et al. also studied the conversion
81
of several lignocellulosic biomass into sorbitan and isosorbide (yield of 15% based on
82
carbon) using a combination of supported Ru catalyst and heteropoly acids (Beeck et
83
al., 2013). Nevertheless, the acid catalysts were still required so far for the conversion
84
of lignocellulosic materials to sugar alcohols. Guha et al. were successful in producing
85
arabitol from sugar beet fiber (yield of 83% based on hemicellulose) (Guha et al.,
86
2011), but the direct catalytic conversion of cellulose in lignocellulosic biomass
87
remained a challenge. Recently, Yamaguchi et al. directly converted cellulose and
88
hemicellulose in wood chips using 4%Pt/BP2000 in water without the use of any acids,
89
achieving a 94% conversion and 62% yield of sugar alcohols (Yamaguchi et al., 2014).
90
Using tungsten-based catalysts, Li et al. managed to convert raw woody biomass
91
(poplar, basswood, ashtree, beech, xylosma, bagasse, pine and yate) into ethylene 4
92
glycol and other diols with a total yield up to 75.6% (based on the amount of cellulose
93
and hemicellulose) and into monophenols with a yield of 46.5% (based on lignin) (Li et
94
al., 2015). More recently, Yamaguchi et al. also reported the direct conversion of
95
lignocellulosic biomass (Japanese cedar, eucalyptus, bagasse, empty fruit bunch and
96
rice straw) into sugar alcohols (sorbitol, mannitol, galactitol, xylitol, arabitol) using
97
supported Pt and Ru-Pt catalysts in the presence of hydrogen in water without any
98
acids, obtaining an amount of sugar alcohols up to 0.551 g from 1 g of milled bagasse
99
(Yamaguchi et al., 2016).
100
Although these works have already focused on the conversion of woody biomass, such
101
as forestry wastes, agricultural residues and crops, to the best of our knowledge the
102
direct conversion of materials that are mainly composed of cellulose and also
103
considered as residues, such as paper or cotton, has not yet been reported. So, this
104
work will focus on the one-pot catalytic conversion of waste cellulosic materials into
105
high added-value chemicals, especially sorbitol, in the presence of a Ru catalyst
106
supported on multi-walled carbon nanotubes, using only water as solvent under H2
107
pressure. The performance of the metal catalyst will be examined in the conversion of
108
cellulosic materials that are normally considered as residues, such as printing paper
109
(white or recycled), tissue paper, cotton wool and cotton textile. Furthermore, the
110
process used in this work can be considered environmentally friendly, since only uses
111
water as solvent and does not require the use of any acids for the reaction neither for
112
the pre-treatment of the substrates or catalyst. Additionally, the effect of ball-milling the
113
substrates or mix-milling them with the catalyst will also be considered.
114 115 116
2. Materials and Methods 2.1. Materials
5
117
A cotton textile sample was supplied by Arcotêxteis (Portugal), with the following
118
features: 100% cotton prepared for dyeing (warp: 3726 threads, weft: 52 threads),
119
previously desized and bleached. Cotton wool (Continente), printing paper (Navigator
120
Universal, 80 g·m2) and tissue paper (Renova) were acquired from Continente and
121
recycled printing paper (Staples 100% recycled, 80 g·m2) from Staples. Microcrystalline
122
cellulose, sorbitol (98%) and the metal precursor ruthenium (III) chloride (RuCl 3 99.9%,
123
Ru 38%) were provided by Alfa Aesar. Nanocyl-3100 multi-walled carbon nanotubes
124
were obtained from Nanocyl and sulphuric acid (> 95%) was supplied by Fisher
125
Chemical. TiO2 P25 was obtained from Degussa. Ultrapure water with a conductivity of
126
18.2 µS·cm-1 was obtained in a Milli-Q Millipore System and used for the preparation of
127
the solutions.
128 129
2.2. Preparation procedures
130
The different materials were ball-milled in a 10 cm3 ceramic pot with two zirconium
131
oxide balls (12 mm of diameter) using a laboratory ball mill (Retsch Mixer Mill MM200)
132
for 4 h at a frequency of 20 vibrations/s.
133
The different materials were also ball-milled together with the catalyst in the same
134
conditions, by introducing both catalyst and substrate in the same ceramic pot.
135
A 0.4 wt% ruthenium catalyst was prepared by the incipient wetness impregnation of
136
commercial multi-walled carbon nanotubes (CNT) with an aqueous solution of the
137
metallic precursor (RuCl3). This metal loading has shown to be an optimum for the
138
transformation of cellulose into sorbitol under the present conditions (Ribeiro et al.,
139
2015a). After impregnation, the resulting material was dried overnight in an oven at 110
140
ºC. Finally, the catalyst was heat treated under nitrogen flow for 3 h at 250 ºC and
141
subsequently reduced under hydrogen flow for 3 h at 250 ºC. The sample was denoted
142
as Ru/CNT.
143
Further details can be found elsewhere (Ribeiro et al., 2017).
144 6
145
2.3. Characterization
146
X-ray diffraction (XRD) patterns were recorded by a Phillips X’Pert MPD diffractometer
147
(Cu-Kα = 0.15406 nm), where the diffracted intensity of Cu-Kα radiation was measured
148
in the 10-100º range of 2θ. Elemental analysis was performed on an EA1108 CHNS-O
149
elemental analyser from Carlo Erba Instruments. Thermogravimetric (TG) analysis was
150
carried out under nitrogen and air using a STA 409 PC/4/H Luxx Netzsch thermal
151
analyser. The samples were heated from 50 to 800 ºC at a 10 ºC·min-1 heating rate.
152 153
2.4. Reaction
154
Cotton wool, cotton textile, printing paper, recycled printing paper, tissue paper and
155
commercial microcrystralline cellulose were used as substrates. With exception of
156
cellulose, these materials were previously cut into small shaped pieces (about 1 cm
157
side) before use, so as to minimize voids inside the reactor. Each material was used
158
without any pre-treatment (besides cutting and ball-milling).
159
The hydrolytic hydrogenation experiments were performed in a 1000 mL stainless steel
160
reactor (Parr Instruments), which was loaded with 750 mg of substrate, 300 mg of
161
Ru/CNT and 300 mL of water. The reactor was then flushed three times with nitrogen
162
to remove ambient air and subsequently heated to 205 ºC at 150 rpm. After achieving
163
the desired temperature, the reaction was initiated by switching from inert gas to H2 (50
164
bar) and stopped after 5 h. Samples (1 mL) were periodically withdrawn for analysis by
165
high performance liquid chromatography (HPLC) and total organic carbon (TOC). The
166
chromatograph was equipped with a refractive index (RI) detector and the products
167
were separated in an ion exclusion Alltech OA-1000 column (300 × 6.5 mm), using a
168
0.005 mol·L-1 H2SO4 mobile phase as eluent at a 0.5 mL·min-1 flow rate and an
169
injection volume of 30 µL. The yield ( ) of each product i was determined according to
170
the following equation: (%) =
moles of product formed (measured from HPLC) × 100% (1) moles of substrate initially present 7
171
TOC data was obtained with a Shimadzu TOC 5000-A and the conversion (!) of each
172
substrate determined using the equation: ! (%) =
moles of total organic carbon in the resultant liquid × 100% (2) moles of carbon in the substrate charged to the reactor
173
For calculations, it was considered that the materials used are practically only made up
174
cellulose, except of paper (printing and recycled). For the paper samples, about 15%
175
are inorganic impurities (see Section 3.1), and so we considered that the remaining
176
85% was cellulose.
177
Typical error in the catalytic experiments was within ± 3%.
178 179 180
3. Results and Discussion 3.1. Characterization of materials
181
In order to facilitate the contact with the catalyst, the crystalline structure of cellulose
182
and all the materials has to be considered. In this work, ball-milling was used for
183
disrupting their crystal structures. The XRD peak height method is widely used to
184
determine the crystallinity index, allowing fast comparison of the original and ball-milled
185
materials (Park et al., 2010; Yabushita et al., 2014). The crystallinity index (CrI) is
186
calculated according to the following equation: CrI (%) =
'(() − '+, × 100% (3) '(()
187
where '(() is the maximum intensity of the (002) lattice diffraction (2θ = 22.6º) and '+,
188
is the intensity diffraction at 2θ = 18º. '(() represents both crystalline and amorphous
189
parts whereas '+, represents amorphous parts only.
190
The XRD patterns of the different materials used before and after ball-milling are
191
shown in supplementary material. In every unmilled sample it is possible to observe
192
two strong diffraction peaks at 2θ of 15.0º and 22.6º, which are characteristic
193
diffractions of the (101) and (002) crystalline planes, respectively. After ball-milling, the
194
intensity of the diffraction peaks evidently decreased and the amorphous peak at 2θ 8
195
around 18-20º increased, indicating a decrease in the crystallinity index of the ball-
196
milled samples. In the case of cellulose, ball-milling allowed decreasing the degree of
197
crystallinity from 92 to 23% (Ribeiro et al., 2015b). Clearly, the transformation of the
198
original crystalline materials into amorphous materials indicates that ball-milling has
199
weakened the hydrogen bond networks within the crystalline materials. In general, the
200
XRD patterns of the materials are not very different from that of the microcrystalline
201
cellulose, since they are mainly composed of cellulose. Printing paper (recycled or not)
202
presents some extra peaks at 2θ higher than 25-30º, which may be attributed to the
203
additives used during its manufacturing, like TiO2 that is detected as a mixture of
204
crystals including anatase and rutile phases (Cheng et al., 2014; Yadav et al., 2012).
205
These peaks are not found on tissue paper, since this kind of paper is mainly
206
composed of cellulose (> 90%) on the opposite to printing paper that can contain
207
around 15% of additives, and so only containing around 85% of cellulose. Also, cotton
208
wool and cotton textile do not differ much between themselves, presenting similar
209
crystalline structures.
210
The photographs of the cellulosic materials before and after ball-milling are presented
211
in supplementary material. After the ball-milling treatment every material turns
212
completely into powder. Also, the printing paper sample, which was originally white,
213
acquires a light blue colour after ball-milling.
214
The elemental analysis of the different materials was performed and the results are
215
shown in Table 1. The analysis of cellulose revealed carbon, hydrogen and oxygen
216
contents of about 44%, 7% and 49%, respectively, which are compatible to its
217
molecular formula, (C6H10O5)n. The contents of C, H and O of printing paper and
218
recycled printing paper are practically identical. Both paper samples presented a
219
slightly lower content of carbon and hydrogen and larger content of oxygen, when
220
compared to cellulose; this slightly modified composition of printing paper samples can
221
be related to the presence of additives used for the manufacturing of paper, as
222
mentioned above. Tissue paper, cotton wool and cotton textile present similar C, H and 9
223
O contents, which are not very different from those of cellulose, which is in accordance
224
to the literature (Moltó et al., 2006) and to the fact that these materials are mainly
225
composed of cellulose. So, in general, the results are quite similar for all the materials
226
and do not differ much from those of cellulose.
227
Table 1
228
The mass loss history of the cellulosic materials under different atmospheres was
229
tracked as shown in Figure 1. Only one sharp mass loss between 260 and 360 ºC
230
(over 80% of total mass) was found for cotton wool, cotton textile and tissue paper
231
samples under inert atmosphere (N2) due to pyrolysis processes. Under oxidative
232
atmosphere (air), a second mass loss was detected at higher temperatures (400-500
233
ºC) for these three samples, due to the oxidation of the char. Also, the temperature
234
range that corresponds to the first mass loss did not vary with the surrounding
235
environment. The results are similar to those of microcrystalline cellulose (Ribeiro et
236
al., 2015b) and are consistent with those reported in literature (Moltó et al., 2006; Órfão
237
et al., 1999; Shen et al., 2013; Sivasangar et al., 2013). The decomposition of cotton
238
wool, cotton textile and tissue paper in air was practically completed at about 500 ºC,
239
while cellulose was completely decomposed at about 580 ºC. Couhert et al. reported
240
that the decomposition of pure components (such as cellulose) differs from real
241
materials due to interferences of other components (Couhert et al., 2009). The solid
242
residues of cotton wool, cotton textile and tissue paper samples were 2%, 5% and 2%,
243
respectively. The TG curves of these materials were very similar among them and in
244
comparison with cellulose; nevertheless, cellulose and cotton wool start to decompose
245
at slightly higher temperatures than tissue paper and cotton textile, indicating that the
246
cellulose of cotton wool has a slightly different structure than cellulose from tissue
247
paper and cotton textile.
248
Both printing paper samples presented an extra peak under inert atmosphere around
249
700 ºC, which maintained under oxidative atmosphere. Moreover, these two samples
250
only presented a total weight loss around 85% under air, which could be attributed to 10
251
the presence of inorganic compounds that are added during paper manufacturing.
252
Accordingly, these two paper samples present a lower cellulose content than cotton
253
wool, cotton textile and tissue paper. This difference on the content of cellulose
254
between the samples was considered for calculations, as mentioned in Section 2.4.
255
Figure 1
256
The catalyst has been extensively characterized and the results were reported
257
elsewhere (Ribeiro et al., 2017).
258 259
3.2. Conversion of cellulosic materials
260
We have previously reported the direct conversion of cellulose into sugar alcohols
261
using Ru/CNT as catalyst (Ribeiro et al., 2017). A total yield of sorbitol of 51% was
262
obtained after 5 h of reaction in the conversion of ball-milled cellulose with 50 bar H2 at
263
205 ºC. The yield of sorbitol was further increased to 61% in just 1 h of reaction by mix-
264
milling cellulose with the catalyst. In addition, the catalyst presented excellent stability
265
during the current reaction conditions and could be reused up to at least four
266
successive runs with practically no loss in activity and selectivity or metal leaching to
267
solution.
268
In the present work, we applied the previous direct conversion method to obtain sugar
269
alcohols from cellulosic materials, such as cotton (wool and textile) and paper (tissue,
270
printing and recycled) in the presence of Ru/CNT. Figure 2 shows the evolution of the
271
conversion of the different untreated materials in comparison to that of cellulose. The
272
conversion after 5 h varied between 18 and 38%, the lowest conversion obtained with
273
cotton wool and the highest with tissue paper and microcrystalline cellulose. The
274
conversions obtained were very small, which is explained by the rigid crystalline
275
structure of the cellulosic materials (see Section 3.1) that difficults the hydrolysis of the
276
materials. Accordingly, the conversion of the unmilled materials only afforded yields of
11
277
sorbitol up to 8%. Therefore, these results had proven that the pre-treatment of the
278
substrates is a pre-requisite for achieving high conversions of cellulosic materials.
279
Figure 2
280
Then, in order to increase the conversion, the materials were ball-milled prior to
281
reaction. A huge difference was observed for the conversion of the materials, as
282
depicted in Figure 2a. A 100% conversion of cotton wool, cotton textile and tissue
283
paper was achieved after just 4 h of reaction, which was even higher than that of
284
cellulose (83.5%). The ball-milling of the materials greatly increased the conversion
285
due to the decrease of cellulose crystallinity. In Section 3.1 it was observed that the
286
(002) diffraction peak at 22.6º was broadened by the ball-milling treatment of the
287
cellulosic samples, indicating that the cellulose crystallinity in cotton wool, cotton textile
288
and tissue paper was decreased. The conversion of the printing paper samples after 5
289
h of reaction only presented an increase from about 25-33% to 50% with the ball-
290
milling pre-treatment. The lowest conversion obtained in comparison to the remaining
291
materials can be explained by the presence of inorganic impurities and additives that
292
were used upon the manufacturing of paper (e.g. TiO2). These substances could inhibit
293
the conversion of printing paper under the conditions used in the present work.
294
Yields of sorbitol of 47.6, 45.0 and 45.3% were attained after 5 h of reaction from the
295
conversion of cotton wool, cotton textile and tissue paper, respectively. Although the
296
yields of sorbitol from the direct conversion of pure cellulose were higher during most of
297
the reaction time, the yield of sorbitol achieved at the end of the reaction (50.8% after 5
298
h) was very close to that obtained from the conversion of the waste materials. It is also
299
possible to observe from Figure 2b that the yields of sorbitol obtained during the
300
reaction are very similar for the three samples (cotton wool, cotton textile and tissue
301
paper), and especially between the two cotton samples. Therefore, the decrease of the
302
materials crystallinity has a direct effect on the improvement of the conversion and
303
production of sorbitol. Besides sorbitol, C2-C6 sugar alcohols such as xylitol, glycerol, 12
304
glucose, ethylene glycol (EG) and propylene glycol (PG) could also be obtained directly
305
from the cellulosic materials (Table 2). The conversion of cotton wool, cotton textile and
306
tissue paper increased with increasing time and consequently the water-soluble
307
products also increased, such as glucose, xylitol and glycerol. A total amount of sugar
308
alcohols of 0.803, 0.703 and 0.733 g were obtained from 1 g of cotton wool, cotton
309
textile and tissue paper after 5 h of reaction, respectively, where the amounts of
310
sorbitol, xylitol and glycerol were between 0.509-0.534 g, 0.030-0.101 g and 0.055-
311
0.070 g, respectively (Table 2). Tissue paper and cotton wool also afforded EG and PG
312
with amounts of EG of 0.055 and 0.059 g and of 0.012 and 0.025 g of PG after 5 h of
313
reaction, respectively.
314
Table 2
315
In addition to the low conversion achieved, the printing paper samples were not
316
significantly converted into sorbitol (yields around 7% after 5 h of reaction) (Figure 2b)
317
neither into the other sugar alcohols. The amount of total sugar alcohols (0.194
318
g/gprinting-paper and 0.213 g/grecycled-paper after 5 h of reaction) was considerably lower than
319
that directly obtained from tissue paper, cotton wool and cotton textile (Table 2). The
320
reason could be related to the presence of impurities that inhibit the conversion to
321
sorbitol and other sugar alcohols.
322
There are many substances that are added to paper during its manufacturing. In order
323
to test if the presence of those compounds can affect the reaction, we selected one of
324
the most important known additives (TiO2) and added it to the reaction mixture along
325
with cellulose. The reaction was performed under the same conditions and the results
326
obtained are presented in Figure 3. The evolution of the yield of sorbitol was not very
327
different up to 1.5 h of reaction. After that time, a continuous decrease was observed
328
during the remaining reaction time. The yield of sorbitol achieved after 5 h of reaction
329
decreased from 50.8 to 22.4% due to the presence of TiO2 on the reaction mixture.
330
Therefore, we can conclude that the presence of TiO2 was unfavourable for the 13
331
formation of sorbitol directly from cellulose. Since there are many other substances that
332
are added during paper manufacturing and each of them could affect the conversion,
333
as well as their combination, an intensive study needs to be carried out in order to
334
better understand the effect that each added component could have on the
335
transformation of printing paper samples.
336
Figure 3
337
Finally, in this work we also tested the effect of mix-milling the substrates with the
338
catalyst. Such approach had already proven to allow achieving higher yields of sugar
339
alcohols in less reaction time, by greatly increasing the conversion rates (Kobayashi et
340
al., 2013; Komanoya et al., 2014; Ribeiro et al., 2015b; Ribeiro et al., 2017). A
341
conversion of 100% was achieved for tissue paper, cotton wool and cotton textile in just
342
1, 1.5 and 2 h of reaction, respectively (Figure 2a). Once more, the conversion of
343
printing paper samples did not went further than about 50% after 5 h of reaction.
344
Therefore, with exception of the two printing paper samples, the rate of conversion of
345
the materials was greatly enhanced by the mix-milling pre-treatment. The catalyst and
346
the different materials (tissue paper, cotton wool and cotton textile) were also
347
separately ball-milled and tested under the same conditions to understand the effect of
348
ball-milling (supplementary material). The mentioned increase of the conversion rate
349
was not observed when the catalyst and the substrates were separately ball-milled,
350
confirming that the enhancement of the performance was not due to the catalyst ball-
351
milling, but to the good physical contact between the catalyst and the substrates, which
352
resulted from the mix-milling.
353
Yields of sorbitol around 50% were attained in just 2 h of reaction from the conversion
354
of cotton wool, cotton textile and tissue paper (Figure 2b), corresponding to amounts of
355
sorbitol over 0.5 g/gsubstrate (see Table 2). Although no significant increase was observed
356
for the yields of sorbitol at the end of the reaction (i.e. 5 h) with the mix-milling
357
comparatively to the ball-milling, they were reached in less than half of the time. The 14
358
mix-milling pre-treatment of cotton wool, cotton textile and tissue paper enhanced the
359
total sugar alcohol yield from 0.372 to 0.674 g, from 0.353 to 0.709 g and from 0.607 to
360
0.792 g after 2 h of reaction, respectively (Table 2). The amounts of sorbitol were also
361
drastically increased from 0.183 to 0.525 g for cotton wool, from 0.217 to 0.511 g for
362
cotton textile and from 0.338 to 0.559 g for tissue paper, but the amount of other
363
products did not change so significantly. Tissue paper also afforded an amount of
364
xylitol of 0.112 g after 2 h of reaction. The mix-milling had once again no effect on the
365
yield of sorbitol obtained from the printing paper samples.
366
As a final point, with (ball- or mix-milling) or without pre-treatment, the two printing
367
paper samples presented similar performances indicating that the fact that the sample
368
is recycled or not might not influence the reaction.
369
The results presented in this work show that we have succeeded in producing sugar
370
alcohols, especially sorbitol, from cellulosic materials using a Ru/CNT catalyst in water
371
without any acid catalysts, pre-treatment or reaction medium. Moreover, the catalyst
372
used can be easily recovered by filtration from the reaction mixture, allowing its
373
reusability.
374 375
4. Conclusions
376
Waste cellulosic materials were converted into sorbitol using a Ru/CNT catalyst, and
377
conversions close to 40% were obtained after 5 h, with sorbitol yields below 10%. Ball-
378
milling of cotton textile, cotton wool and tissue paper resulted in 100% conversion of
379
these materials and sorbitol yields around 50% after 5 h. Mix-milling these materials
380
with Ru/CNT greatly enhanced their rate of conversion and sorbitol yields of 50% were
381
reached in 2 h. However, printing paper only had a conversion up to 50% after 5 h, with
382
sorbitol yields around 7%. Moreover, TiO2 had a negative impact on the production of
383
sorbitol.
15
384 385
Acknowledgements
386
This work was financially supported by project POCI-01-0145-FEDER-006984 –
387
Associate Laboratory LSRE-LCM funded by FEDER through COMPETE2020 -
388
Programa Operacional Competitividade e Internacionalização (POCI) – and by national
389
funds through FCT - Fundação para a Ciência e a Tecnologia. L.S. Ribeiro
390
acknowledges her Ph.D. scholarship (SFRH/BD/86580/2012) from FCT.
391 392
Appendix A. Supplementary data
393 394
References
395
1.
Beeck, B.O.d., Geboers, J., Vyver, S.V.d., Lishout, J.V., Snelders, J., Huijgen,
396
W.J.J., Courtin, C.M., Jacobs, P.A., Sels, B.F. 2013. Conversion of
397
(Ligno)Cellulose Feeds to Isosorbide with Heteropoly Acids and Ru on Carbon.
398
ChemSusChem, 6, 199-208.
399
2.
Byun, J., Han, J. 2016a. Catalytic production of biofuels (butene oligomers) and
400
biochemicals (tetrahydrofurfuryl alcohol) from corn stover. Bioresource Technol,
401
211, 360-366.
402
3.
Byun, J., Han, J. 2016b. Process synthesis and analysis for catalytic conversion
403
of lignocellulosic biomass to fuels: Separate conversion of cellulose and
404
hemicellulose using 2-sec-butylphenol (SBP) solvent. Appl Energ, 171, 483-
405
490.
406
4.
Cheng, L., Liu, H., Cui, Y., Xue, N., Ding, W. 2014. Direct conversion of corn
407
cob to formic and acetic acids over nano oxide catalysts. J Energ Chem, 23, 43-
408
49.
409
5.
Couhert, C., Commandre, J.-M., Salvador, S. 2009. Is it possible to predict gas
410
yields of any biomass after rapid pyrolysis at high temperature from its
411
composition in cellulose, hemicellulose and lignin? Fuel, 88, 408-417.
412
6.
Deng, W., Tan, X., Fang, W., Zhang, Q., Wang, Y. 2009. Conversion of
413
Cellulose into Sorbitol over Carbon Nanotube-Supported Ruthenium Catalyst.
414
Catal Lett, 133, 167-174.
16
415
7.
Deng, W., Zhang, H., Xue, L., Zhang, Q., Wang, Y. 2015. Selective activation of
416
the C–O bonds in lignocellulosic biomass for the efficient production of
417
chemicals. Chinese J Catal, 36, 1440-1460.
418
8.
Ding, L.N., Wang, A.Q., Zheng, M.Y., Zhang, T. 2010. Selective transformation
419
of cellulose into sorbitol by using a bifunctional nickel phosphide catalyst.
420
ChemSusChem, 3, 818-821.
421
9.
Feng, Y., Li, G., Li, X., Zhu, N., Xiao, B., Li, J., Wang, Y. 2016. Enhancement of
422
biomass conversion in catalytic fast pyrolysis by microwave-assisted formic acid
423
pretreatment. Bioresource Technol, 214, 520-527.
424
10.
425 426
Fukuoka, A., Dhepe, P. 2006. Catalytic conversion of cellulose into sugar alcohols. Angew Chem Int Edit, 45, 5161-5163.
11.
Guha, S.K., Kobayashi, H., Hara, K., Kikuchi, H., Aritsuka, T., Fukuoka, A.
427
2011. Hydrogenolysis of sugar beet fiber by supported metal catalyst. Catal
428
Commun, 12, 980-983.
429
12.
Guo, X., Wang, X., Guan, J., Chen, X., Qin, Z., Mu, X., Xian, M. 2014. Selective
430
hydrogenation of D-glucose to D-sorbitol over Ru/ZSM-5 catalysts. Chinese J
431
Catal, 35, 733-740.
432
13.
Kobayashi, H., Ito, Y., Komanoya, T., Hosaka, Y., Dhepe, P.L., Kasai, K., Hara,
433
K., Fukuoka, A. 2011. Synthesis of sugar alcohols by hydrolytic hydrogenation
434
of cellulose over supported metal catalysts. Green Chem, 13, 326-333.
435
14.
Kobayashi, H., Yabushita, M., Komanoya, T., Hara, K., Fujita, I., Fukuoka, A.
436
2013. High-Yielding One-Pot Synthesis of Glucose from Cellulose Using Simple
437
Activated Carbons and Trace Hydrochloric Acid. ACS Catalysis, 3, 581-587.
438
15.
Kobayashi, H., Yamakoshi, Y., Hosaka, Y., Yabushita, M., Fukuoka, A. 2014.
439
Production of sugar alcohols from real biomass by supported platinum catalyst.
440
Catal Today, 226, 204-209.
441
16.
Komanoya, T., Kobayashi, H., Hara, K., Chun, W.-J., Fukuoka, A. 2014. Kinetic
442
Study of Catalytic Conversion of Cellulose to Sugar Alcohols under Low-
443
Pressure Hydrogen. ChemCatChem, 6, 230-236.
444
17.
445 446
Kusserow, B., Schimpf, S., Claus, P. 2003. Hydrogenation of Glucose to Sorbitol over Nickel and Ruthenium Catalysts. Adv Synth Catal, 345, 289-299.
18.
Li, C., Zhao, X., Wang, A., Huber, G.W., Zhang, T. 2015. Catalytic
447
Transformation of Lignin for the Production of Chemicals and Fuels. Chem Rev,
448
115, 11559-11624.
449
19.
Luo, C., Wang, S., Liu, H. 2007. Cellulose conversion into polyols catalysed by
450
reversible formed acids and supported ruthenium clusters in hot water. Angew
451
Chem Int Edit, 46, 7636-7639. 17
452
20.
Mäki-Arvela, P.i., Salmi, T., Holmbom, B., Willför, S., Murzin, D.Y. 2011.
453
Synthesis of Sugars by Hydrolysis of Hemicelluloses - A Review. Chem Rev,
454
111, 5638-5666.
455
21.
Moltó, J., Font, R., Conesa, J.A., Martín-Gullón, I. 2006. Thermogravimetric
456
analysis during the decomposition of cotton fabrics in an inert and air
457
environment. J Anal Appl Pyrolysis, 76, 124-131.
458
22.
Órfão, J.J.M., Antunes, F.J.A., Figueiredo, J.L. 1999. Pyrolysis kinetics of
459
lignocellulosic materials—three independent reactions model. Fuel, 78, 349-
460
358.
461
23.
Palkovits, R., Tajvidi, K., Procelewska, J., Rinaldi, R., Ruppert, A. 2010.
462
Hydrogenolysis of cellulose combining mineral acids and hydrogenation
463
catalysts. Green Chem, 12, 972-978.
464
24.
Park, S., Baker, J., Himmel, M., Parilla, P., Johnson, D. 2010. Cellulose
465
crystallinity index: measurement techniques and their impact on interpreting
466
cellulase performance. Biotechnol Biofuels, 3, 1-10.
467
25.
468 469
Rao, L.V., Goli, J.K., Gentela, J., Koti, S. 2016. Bioconversion of Lignocellulosic biomass to xylitol: An overview. Bioresource Technol, 213, 299-310.
26.
Ribeiro, L.S., Órfão, J.J.M., Pereira, M.F.R. 2015a. Comparative study of
470
different catalysts for the direct conversion of cellulose to sorbitol. Green
471
Process Synth, 4, 71-78.
472
27.
473 474
Ribeiro, L.S., Órfão, J.J.M., Pereira, M.F.R. 2015b. Enhanced direct production of sorbitol by cellulose ball-milling. Green Chem 17, 2973-2980.
28.
Ribeiro, L.S., Delgado, J.J., Órfão, J.J.M., Pereira, M.F.R. 2017. Direct
475
conversion of cellulose to sorbitol over ruthenium catalysts: Influence of the
476
support. Catal Today, 279, 244-251.
477
29.
Romero, A., Alonso, E., Sastre, Á., Nieto-Márquez, A. 2016. Conversion of
478
biomass into sorbitol: Cellulose hydrolysis on MCM-48 and d-Glucose
479
hydrogenation on Ru/MCM-48. Micropor Mesopor Mat, 224, 1-8.
480
30.
Shen, D., Ye, J., Xiao, R., Zhang, H. 2013. TG-MS analysis for thermal
481
decomposition of cellulose under different atmospheres. Carbohyd Polym, 98,
482
514-521.
483
31.
Sivasangar, S., Taufiq-Yap, Y.H., Zainal, Z., Kitagawa, K. 2013. Thermal
484
behavior of lignocellulosic materials under aerobic/anaerobic environments. Int
485
J Hydrogen Energ, 38, 16011-16019.
486 487
32.
Van de Vyver, S., Geboers, J., Dusselier, M., Schepers, H., Vosch, T., Zhang, L., Van Tendeloo, G., Jacobs, P.A., Sels, B.F. 2010. Selective Bifunctional
18
488
Catalytic Conversion of Cellulose over Reshaped Ni Particles at the Tip of
489
Carbon Nanofibers. ChemSusChem, 3, 698-701.
490
33.
Van de Vyver, S., Geboers, J., Schutyser, W., Dusselier, M., Eloy, P., Dornez,
491
E., Seo, J.W., Courtin, C.M., Gaigneaux, E.M., Jacobs, P.A., Sels, B.F. 2012.
492
Tuning the Acid/Metal Balance of Carbon Nanofiber-Supported Nickel Catalysts
493
for Hydrolytic Hydrogenation of Cellulose. ChemSusChem, 5, 1549-1558.
494
34.
Wang, H., Zhu, L., Peng, S., Peng, F., Yu, H., Yang, J. 2012. High efficient
495
conversion of cellulose to polyols with Ru/CNTs as catalyst. Renew Energ, 37,
496
192-196.
497
35.
Wu, Z., Ge, S., Ren, C., Zhang, M., Yip, A., Xu, C. 2012. Selective conversion
498
of cellulose into bulk chemicals over Brønsted acid-promoted ruthenium
499
catalyst: one-pot vs. sequential process. Green Chem, 14, 3336-3343.
500
36.
501 502
Yabushita, M., Kobayashi, H., Fukuoka, A. 2014. Catalytic transformation of cellulose into platform chemicals. Appl Catal B-Environ, 145, 1-9.
37.
Yadav, M., Mishra, D.K., Hwang, J.-S. 2012. Catalytic hydrogenation of xylose
503
to xylitol using ruthenium catalyst on NiO modified TiO2 support. Appl Catal A-
504
Gen, 425–426, 110-116.
505
38.
Yamaguchi, A., Sato, O., Mimura, N., Hirosaki, Y., Kobayashi, H., Fukuoka, A.,
506
Shirai, M. 2014. Direct production of sugar alcohols from wood chips using
507
supported platinum catalysts in water. Catal Commun, 54, 22-26.
508
39.
509 510
Yamaguchi, A., Sato, O., Mimura, N., Shirai, M. 2016. Catalytic production of sugar alcohols from lignocellulosic biomass. Catal Today, 265, 199-202.
40.
Yang, P., Kobayashi, H., Hara, K., Fukuoka, A. 2012. Phase change of nickel
511
phosphide catalysts in the conversion of cellulose into sorbitol. ChemSusChem,
512
5, 920-926.
513 514
41.
Zhao, X., Xu, J., Wang, A., Zhang, T. 2015. Porous carbon in catalytic transformation of cellulose. Chin J Catal, 36, 1419-1427.
515
19
516
Figure Captions
517 518 519 520
Figure 1 – TG (on the left) and DTG (on the right) curves of cellulosic materials under a) air and b) nitrogen. Figure 2 – a) Conversion and b) yield of sorbitol of the different unmilled (A), ball-milled
521
(B) and mix-milled (C) materials. Reaction conditions: 300 mL water, 750 mg
522
substrate, 300 mg Ru/CNT, 205 ºC, 50 bar H2, 150 rpm.
523
Figure 3 – Effect of TiO2 on the production of sorbitol from cellulose. Reaction
524
conditions: 300 mL water, 750 mg ball-milled cellulose (+ 150 mg TiO2), 300 mg
525
Ru/CNT, 205 ºC, 50 bar H2, 150 rpm.
526
20
527 528
Tables Table 1 – Elemental analysis results for the different materials. Sample
C (at.%)
H (at.%)
O (at.%)
Cellulose
44.2 ± 0.1
6.9 ± 0.1
48.9 ± 0.3
Cotton wool
41.7 ± 0.8
5.7 ± 0.1
52.6 ± 0.8
Cotton textile
42.9 ± 0.5
6.4 ± 0.1
50.7 ± 0.3
Printing paper
37.2 ± 0.3
4.6 ± 0.1
58.2 ± 0.5
Recycled printing paper
38.8 ± 0.3
5.0 ± 0.1
56.2 ± 0.4
Tissue paper
43.7 ± 0.6
5.9 ± 0.3
50.4 ± 0.7
529 530
21
531
Table 2 – Amount of products obtained in the conversion of cellulosic materials after 2 h [and 5 h] of reaction. Reaction
532
conditions: 300 mL water, 750 mg substrate, 300 mg Ru/CNT, 205 ºC, 50 bar H2, 150 rpm.
Material
Cellulose
Cotton wool
Cotton textile
Printing paper
Recycled printing paper
Tissue paper
Pre-treatment
Amount of product (g/gmaterial) Sorbitol
Glucose
Xylitol
Glycerol
EG
PG
Total
Unmilled
0.098 [0.106]
0.000 [0.000]
0.005 [0.017]
0.031 [0.040]
0.000 [0.000]
0.000 [0.000]
0.134 [0.163]
Ball-milled
0.533 [0.571]
0.040 [0.000]
0.037 [0.082]
0.045 [0.052]
0.047 [0.057]
0.000 [0.016]
0.702 [0.778]
Mix-milled
0.782 [0.656]
0.000 [0.000]
0.029 [0.053]
0.038 [0.053]
0.000 [0.000]
0.000 [0.000]
0.849 [0.762]
Unmilled
0.066 [0.073]
0.000 [0.000]
0.000 [0.000]
0.021 [0.028]
0.047 [0.049]
0.000 [0.005]
0.134 [0.155]
Ball-milled
0.183 [0.534]
0.086 [0.085]
0.008 [0.030]
0.038 [0.070]
0.050 [0.059]
0.007 [0.025]
0.372 [0.803]
Mix-milled
0.525 [0.562]
0.093 [0.081]
0.018 [0.040]
0.038 [0.061]
0.000 [0.000]
0.000 [0.000]
0.674 [0.744]
Unmilled
0.064 [0.072]
0.000 [0.000]
0.000 [0.000]
0.023 [0.027]
0.000 [0.000]
0.000 [0.000]
0.087 [0.099]
Ball-milled
0.217 [0.510]
0.090 [0.084]
0.009 [0.054]
0.037 [0.055]
0.000 [0.000]
0.000 [0.000]
0.353 [0.703]
Mix-milled
0.511 [0.505]
0.089 [0.081]
0.015 [0.024]
0.035 [0.048]
0.051 [0.054]
0.008 [0.013]
0.709 [0.725]
Unmilled
0.063 [0.063]
0.000 [0.000]
0.000 [0.000]
0.027 [0.028]
0.053 [0.059]
0.000 [0.013]
0.143 [0.137]
Ball-milled
0.063 [0.064]
0.000 [0.000]
0.003 [0.002]
0.031 [0.029]
0.059 [0.069]
0.016 [0.030]
0.172 [0.194]
Mix-milled
0.065 [0.066]
0.000 [0.000]
0.003 [0.001]
0.029 [0.036]
0.061 [0.073]
0.022 [0.045]
0.180 [0.221]
Unmilled
0.062 [0.062]
0.000 [0.000]
0.001 [0.001]
0.030 [0.038]
0.056 [0.064]
0.015 [0.023]
0.164 [0.188]
Ball-milled
0.064 [0.065]
0.000 [0.000]
0.002 [0.002]
0.042 [0.045]
0.063 [0.071]
0.020 [0.030]
0.191 [0.213]
Mix-milled
0.065 [0.065]
0.000 [0.000]
0.000 [0.001]
0.042 [0.044]
0.064 [0.071]
0.027 [0.042]
0.198 [0.223]
Unmilled
0.083 [0.093]
0.000 [0.000]
0.057 [0.052]
0.029 [0.038]
0.000 [0.000]
0.000 [0.000]
0.169 [0.183]
Ball-milled
0.338 [0.509]
0.082 [0.000]
0.094 [0.101]
0.046 [0.056]
0.047 [0.055]
0.000 [0.012]
0.607 [0.733]
Mix-milled
0.559 [0.517]
0.000 [0.000]
0.112 [0.102]
0.055 [0.056]
0.053 [0.055]
0.013 [0.014]
0.792 [0.744]
22
533
Figures
534 535
Figure 1 – TG (on the left) and DTG (on the right) curves of cellulosic materials under a) air and b) nitrogen. 23
536 537
Figure 2 – a) Conversion and b) yield of sorbitol of the different unmilled (A), ball-milled (B) and mix-milled (C) materials. Reaction conditions:
538
300 mL water, 750 mg substrate, 300 mg Ru/CNT, 205 ºC, 50 bar H2, 150 rpm. 24
539
540 541 542
Figure 3 – Effect of TiO2 on the production of sorbitol from cellulose. Reaction conditions: 300 mL water, 750 mg ball-milled substrate, 150 mg TiO2, 300 mg Ru/CNT, 205 ºC, 50 bar H2, 150 rpm.
543
25
544 545
26
546
Highlights:
547
•
Materials ball-milling disrupted their crystallinity, easing the direct conversion
548
•
Sorbitol was directly obtained from cotton wool, cotton textile and tissue paper
549
•
Mix-milling greatly enhanced the rate of conversion of cellulosic biomass
550
•
Sorbitol amounts higher than 0.5 g could be obtained from 1 g of cellulosic biomass
551
•
Printing paper (white and recycled) was not converted to sugar alcohols
552 553
27
S0960-8524(17)30115-3 http://dx.doi.org/10.1016/j.biortech.2017.02.008 BITE 17571
To appear in:
Bioresource Technology
Received Date: Revised Date: Accepted Date:
5 December 2016 2 February 2017 3 February 2017
Please cite this article as: Ribeiro, L.S., de Melo Órfão, J.J., Pereira, M.F.R., Direct catalytic production of sorbitol from waste cellulosic materials, Bioresource Technology (2017), doi: http://dx.doi.org/10.1016/j.biortech. 2017.02.008
This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
1
Direct catalytic production of sorbitol from waste
2
cellulosic materials
3
Lucília Sousa Ribeiro, José J. de Melo Órfão, Manuel Fernando Ribeiro Pereira*
4 5
Laboratório de Processos de Separação e Reação - Laboratório de Catálise e
6
Materiais (LSRE-LCM), Departamento de Engenharia Química, Faculdade de
7
Engenharia, Universidade do Porto, Rua Dr. Roberto Frias, 4200-465 Porto, Portugal
8 9
[email protected], [email protected], [email protected]*
10 11
*Corresponding author:
12
Manuel Fernando Ribeiro Pereira
13
Tel. +351 225 081 468
14
Fax: +351 225 081 449
15 16
Abstract
17
Cotton wool, cotton textile, tissue paper and printing paper, all potential waste cellulosic
18
materials, were directly converted to sorbitol using a Ru/CNT catalyst in the presence
19
of H2 and using only water as solvent, without any acids. Conversions up to 38% were
20
attained for the raw substrates, with sorbitol yields below 10%. Ball-milling of the
21
materials disrupted their crystallinity, allowing reaching 100% conversion of cotton
22
wool, cotton textile and tissue paper after 4 h, with sorbitol yields around 50%. Mix-
23
milling these materials with the catalyst greatly enhanced their conversion rate, and the
24
materials were efficiently converted to sorbitol with a yield around 50% in 2 h. However,
25
ball- and mix-milled printing paper presented a conversion of only 50% after 5 h, with
1
26
sorbitol yields of 7%. Amounts of sorbitol of 0.525, 0.511 and 0.559 g could be
27
obtained from 1 g of cotton wool, cotton textile and tissue paper, respectively.
28
Keywords: biomass conversion; cellulosic materials; cotton; paper; sorbitol
29
Abbreviations:
30
CNT – carbon nanotubes
31
DTG – differential thermogravimetry
32
HPLC – high performance liquid chromatography
33
RI – refractive index
34
TG – thermogravimetry
35
TOC – total organic carbon
36
XRD – X-ray diffraction
2
37
1. Introduction
38
Concerns about global warming have motivated the search for alternative renewable
39
resources, leading to a significant increase in research activities directed towards their
40
use (Byun and Han, 2016a, 2016b; Deng et al., 2015; Kobayashi et al., 2014).
41
Currently, about 10% of the world’s primary energy is biomass, which is mostly used to
42
generate power and heat (Li et al., 2015). Lignocellulose is the most abundant and less
43
expensive type of biomass on earth, therefore being a promising feedstock for the
44
production of renewable energy, especially biofuels, and chemicals (Deng et al., 2015;
45
Feng et al., 2016). Unlike corn and stalk, lignocellulose is inedible for humans, and its
46
use will not impose a negative impact on food supplies since it does not compete with
47
food production (Li et al., 2015; Yamaguchi et al., 2016). Lignocellulose consists of 35-
48
50% cellulose (a polymer of D-glucose), 25-30% hemicellulose (a polymer of C5 and C6
49
sugars) and 15-30% lignin (Deng et al., 2015). The hydrolytic hydrogenation of these
50
sugar polymers produces sugar alcohols, which are used as low-calorie and low-
51
cariogenic sweeteners and sugar substitutes for diabetics, as humectants in cosmetic
52
and pharmaceutical products, in paper and tobacco and as precursors to plastics
53
(Kobayashi et al., 2014; Kusserow et al., 2003; Mäki-Arvela et al., 2011; Rao et al.,
54
2016). Sorbitol is one of the most important sugar alcohols and can serve as platform
55
chemical for the synthesis of various value-added chemicals such as glycerol, glycols,
56
lactic acid, isosorbide, 1,4-sorbitan and L-sorbose (Deng et al., 2015). The annual
57
production of sorbitol has already reached 6.5×105 ton (Kobayashi et al., 2014).
58
Therefore, the catalytic hydrolytic hydrogenation of biomass and its components in the
59
presence of hydrogen has greatly attracted attention for the establishment of a
60
sustainable society, because it circumvents metastable glucose and allows high
61
selectivity to high-value chemicals or fuels (Zhao et al., 2015).
62
The catalytic conversion of cellulose to sorbitol using supported metal catalysts in the
63
presence of hydrogen has already been reported by some research groups (Deng et
3
64
al., 2009; Ding et al., 2010; Fukuoka and Dhepe, 2006; Kobayashi et al., 2011; Luo et
65
al., 2007; Ribeiro et al., 2015b; Romero et al., 2016; Van de Vyver et al., 2010; Van de
66
Vyver et al., 2012; Yang et al., 2012). Carbon nanotubes, which have been gaining
67
increasing attention as supports in heterogeneous catalysis, have already shown to be
68
the most effective support for the direct conversion of cellulose to sorbitol (Deng et al.,
69
2009; Ribeiro et al., 2017; Wang et al., 2012). Also, Ru catalysts have shown to be the
70
most effective in the direct conversion of cellulose into polyols (Deng et al., 2009; Guo
71
et al., 2014; Ribeiro et al., 2015a; Wu et al., 2012). However, pure cellulose has to be
72
obtained from lignocellulose, normally using strong acid/base catalysts to completely
73
remove the hemicellulose and lignin components. Thus, the direct conversion of the
74
individual components in woody biomass to valuable chemicals is of great importance
75
to open new possibilities of using biomass (Yamaguchi et al., 2014). Palkovits et al.
76
reported the conversion of spruce chips using supported metal catalysts combined with
77
sulfuric/phosphoric acid with a yield of 55% (based on cellulose) of sugar alcohols
78
(sorbitol, sorbitan, isosorbide) (Palkovits et al., 2010). However, the use of acids has
79
problems associated with corrosion of the reactor or neutralization processes for the
80
removal of the acid, and so should be avoided. Beeck et al. also studied the conversion
81
of several lignocellulosic biomass into sorbitan and isosorbide (yield of 15% based on
82
carbon) using a combination of supported Ru catalyst and heteropoly acids (Beeck et
83
al., 2013). Nevertheless, the acid catalysts were still required so far for the conversion
84
of lignocellulosic materials to sugar alcohols. Guha et al. were successful in producing
85
arabitol from sugar beet fiber (yield of 83% based on hemicellulose) (Guha et al.,
86
2011), but the direct catalytic conversion of cellulose in lignocellulosic biomass
87
remained a challenge. Recently, Yamaguchi et al. directly converted cellulose and
88
hemicellulose in wood chips using 4%Pt/BP2000 in water without the use of any acids,
89
achieving a 94% conversion and 62% yield of sugar alcohols (Yamaguchi et al., 2014).
90
Using tungsten-based catalysts, Li et al. managed to convert raw woody biomass
91
(poplar, basswood, ashtree, beech, xylosma, bagasse, pine and yate) into ethylene 4
92
glycol and other diols with a total yield up to 75.6% (based on the amount of cellulose
93
and hemicellulose) and into monophenols with a yield of 46.5% (based on lignin) (Li et
94
al., 2015). More recently, Yamaguchi et al. also reported the direct conversion of
95
lignocellulosic biomass (Japanese cedar, eucalyptus, bagasse, empty fruit bunch and
96
rice straw) into sugar alcohols (sorbitol, mannitol, galactitol, xylitol, arabitol) using
97
supported Pt and Ru-Pt catalysts in the presence of hydrogen in water without any
98
acids, obtaining an amount of sugar alcohols up to 0.551 g from 1 g of milled bagasse
99
(Yamaguchi et al., 2016).
100
Although these works have already focused on the conversion of woody biomass, such
101
as forestry wastes, agricultural residues and crops, to the best of our knowledge the
102
direct conversion of materials that are mainly composed of cellulose and also
103
considered as residues, such as paper or cotton, has not yet been reported. So, this
104
work will focus on the one-pot catalytic conversion of waste cellulosic materials into
105
high added-value chemicals, especially sorbitol, in the presence of a Ru catalyst
106
supported on multi-walled carbon nanotubes, using only water as solvent under H2
107
pressure. The performance of the metal catalyst will be examined in the conversion of
108
cellulosic materials that are normally considered as residues, such as printing paper
109
(white or recycled), tissue paper, cotton wool and cotton textile. Furthermore, the
110
process used in this work can be considered environmentally friendly, since only uses
111
water as solvent and does not require the use of any acids for the reaction neither for
112
the pre-treatment of the substrates or catalyst. Additionally, the effect of ball-milling the
113
substrates or mix-milling them with the catalyst will also be considered.
114 115 116
2. Materials and Methods 2.1. Materials
5
117
A cotton textile sample was supplied by Arcotêxteis (Portugal), with the following
118
features: 100% cotton prepared for dyeing (warp: 3726 threads, weft: 52 threads),
119
previously desized and bleached. Cotton wool (Continente), printing paper (Navigator
120
Universal, 80 g·m2) and tissue paper (Renova) were acquired from Continente and
121
recycled printing paper (Staples 100% recycled, 80 g·m2) from Staples. Microcrystalline
122
cellulose, sorbitol (98%) and the metal precursor ruthenium (III) chloride (RuCl 3 99.9%,
123
Ru 38%) were provided by Alfa Aesar. Nanocyl-3100 multi-walled carbon nanotubes
124
were obtained from Nanocyl and sulphuric acid (> 95%) was supplied by Fisher
125
Chemical. TiO2 P25 was obtained from Degussa. Ultrapure water with a conductivity of
126
18.2 µS·cm-1 was obtained in a Milli-Q Millipore System and used for the preparation of
127
the solutions.
128 129
2.2. Preparation procedures
130
The different materials were ball-milled in a 10 cm3 ceramic pot with two zirconium
131
oxide balls (12 mm of diameter) using a laboratory ball mill (Retsch Mixer Mill MM200)
132
for 4 h at a frequency of 20 vibrations/s.
133
The different materials were also ball-milled together with the catalyst in the same
134
conditions, by introducing both catalyst and substrate in the same ceramic pot.
135
A 0.4 wt% ruthenium catalyst was prepared by the incipient wetness impregnation of
136
commercial multi-walled carbon nanotubes (CNT) with an aqueous solution of the
137
metallic precursor (RuCl3). This metal loading has shown to be an optimum for the
138
transformation of cellulose into sorbitol under the present conditions (Ribeiro et al.,
139
2015a). After impregnation, the resulting material was dried overnight in an oven at 110
140
ºC. Finally, the catalyst was heat treated under nitrogen flow for 3 h at 250 ºC and
141
subsequently reduced under hydrogen flow for 3 h at 250 ºC. The sample was denoted
142
as Ru/CNT.
143
Further details can be found elsewhere (Ribeiro et al., 2017).
144 6
145
2.3. Characterization
146
X-ray diffraction (XRD) patterns were recorded by a Phillips X’Pert MPD diffractometer
147
(Cu-Kα = 0.15406 nm), where the diffracted intensity of Cu-Kα radiation was measured
148
in the 10-100º range of 2θ. Elemental analysis was performed on an EA1108 CHNS-O
149
elemental analyser from Carlo Erba Instruments. Thermogravimetric (TG) analysis was
150
carried out under nitrogen and air using a STA 409 PC/4/H Luxx Netzsch thermal
151
analyser. The samples were heated from 50 to 800 ºC at a 10 ºC·min-1 heating rate.
152 153
2.4. Reaction
154
Cotton wool, cotton textile, printing paper, recycled printing paper, tissue paper and
155
commercial microcrystralline cellulose were used as substrates. With exception of
156
cellulose, these materials were previously cut into small shaped pieces (about 1 cm
157
side) before use, so as to minimize voids inside the reactor. Each material was used
158
without any pre-treatment (besides cutting and ball-milling).
159
The hydrolytic hydrogenation experiments were performed in a 1000 mL stainless steel
160
reactor (Parr Instruments), which was loaded with 750 mg of substrate, 300 mg of
161
Ru/CNT and 300 mL of water. The reactor was then flushed three times with nitrogen
162
to remove ambient air and subsequently heated to 205 ºC at 150 rpm. After achieving
163
the desired temperature, the reaction was initiated by switching from inert gas to H2 (50
164
bar) and stopped after 5 h. Samples (1 mL) were periodically withdrawn for analysis by
165
high performance liquid chromatography (HPLC) and total organic carbon (TOC). The
166
chromatograph was equipped with a refractive index (RI) detector and the products
167
were separated in an ion exclusion Alltech OA-1000 column (300 × 6.5 mm), using a
168
0.005 mol·L-1 H2SO4 mobile phase as eluent at a 0.5 mL·min-1 flow rate and an
169
injection volume of 30 µL. The yield ( ) of each product i was determined according to
170
the following equation: (%) =
moles of product formed (measured from HPLC) × 100% (1) moles of substrate initially present 7
171
TOC data was obtained with a Shimadzu TOC 5000-A and the conversion (!) of each
172
substrate determined using the equation: ! (%) =
moles of total organic carbon in the resultant liquid × 100% (2) moles of carbon in the substrate charged to the reactor
173
For calculations, it was considered that the materials used are practically only made up
174
cellulose, except of paper (printing and recycled). For the paper samples, about 15%
175
are inorganic impurities (see Section 3.1), and so we considered that the remaining
176
85% was cellulose.
177
Typical error in the catalytic experiments was within ± 3%.
178 179 180
3. Results and Discussion 3.1. Characterization of materials
181
In order to facilitate the contact with the catalyst, the crystalline structure of cellulose
182
and all the materials has to be considered. In this work, ball-milling was used for
183
disrupting their crystal structures. The XRD peak height method is widely used to
184
determine the crystallinity index, allowing fast comparison of the original and ball-milled
185
materials (Park et al., 2010; Yabushita et al., 2014). The crystallinity index (CrI) is
186
calculated according to the following equation: CrI (%) =
'(() − '+, × 100% (3) '(()
187
where '(() is the maximum intensity of the (002) lattice diffraction (2θ = 22.6º) and '+,
188
is the intensity diffraction at 2θ = 18º. '(() represents both crystalline and amorphous
189
parts whereas '+, represents amorphous parts only.
190
The XRD patterns of the different materials used before and after ball-milling are
191
shown in supplementary material. In every unmilled sample it is possible to observe
192
two strong diffraction peaks at 2θ of 15.0º and 22.6º, which are characteristic
193
diffractions of the (101) and (002) crystalline planes, respectively. After ball-milling, the
194
intensity of the diffraction peaks evidently decreased and the amorphous peak at 2θ 8
195
around 18-20º increased, indicating a decrease in the crystallinity index of the ball-
196
milled samples. In the case of cellulose, ball-milling allowed decreasing the degree of
197
crystallinity from 92 to 23% (Ribeiro et al., 2015b). Clearly, the transformation of the
198
original crystalline materials into amorphous materials indicates that ball-milling has
199
weakened the hydrogen bond networks within the crystalline materials. In general, the
200
XRD patterns of the materials are not very different from that of the microcrystalline
201
cellulose, since they are mainly composed of cellulose. Printing paper (recycled or not)
202
presents some extra peaks at 2θ higher than 25-30º, which may be attributed to the
203
additives used during its manufacturing, like TiO2 that is detected as a mixture of
204
crystals including anatase and rutile phases (Cheng et al., 2014; Yadav et al., 2012).
205
These peaks are not found on tissue paper, since this kind of paper is mainly
206
composed of cellulose (> 90%) on the opposite to printing paper that can contain
207
around 15% of additives, and so only containing around 85% of cellulose. Also, cotton
208
wool and cotton textile do not differ much between themselves, presenting similar
209
crystalline structures.
210
The photographs of the cellulosic materials before and after ball-milling are presented
211
in supplementary material. After the ball-milling treatment every material turns
212
completely into powder. Also, the printing paper sample, which was originally white,
213
acquires a light blue colour after ball-milling.
214
The elemental analysis of the different materials was performed and the results are
215
shown in Table 1. The analysis of cellulose revealed carbon, hydrogen and oxygen
216
contents of about 44%, 7% and 49%, respectively, which are compatible to its
217
molecular formula, (C6H10O5)n. The contents of C, H and O of printing paper and
218
recycled printing paper are practically identical. Both paper samples presented a
219
slightly lower content of carbon and hydrogen and larger content of oxygen, when
220
compared to cellulose; this slightly modified composition of printing paper samples can
221
be related to the presence of additives used for the manufacturing of paper, as
222
mentioned above. Tissue paper, cotton wool and cotton textile present similar C, H and 9
223
O contents, which are not very different from those of cellulose, which is in accordance
224
to the literature (Moltó et al., 2006) and to the fact that these materials are mainly
225
composed of cellulose. So, in general, the results are quite similar for all the materials
226
and do not differ much from those of cellulose.
227
Table 1
228
The mass loss history of the cellulosic materials under different atmospheres was
229
tracked as shown in Figure 1. Only one sharp mass loss between 260 and 360 ºC
230
(over 80% of total mass) was found for cotton wool, cotton textile and tissue paper
231
samples under inert atmosphere (N2) due to pyrolysis processes. Under oxidative
232
atmosphere (air), a second mass loss was detected at higher temperatures (400-500
233
ºC) for these three samples, due to the oxidation of the char. Also, the temperature
234
range that corresponds to the first mass loss did not vary with the surrounding
235
environment. The results are similar to those of microcrystalline cellulose (Ribeiro et
236
al., 2015b) and are consistent with those reported in literature (Moltó et al., 2006; Órfão
237
et al., 1999; Shen et al., 2013; Sivasangar et al., 2013). The decomposition of cotton
238
wool, cotton textile and tissue paper in air was practically completed at about 500 ºC,
239
while cellulose was completely decomposed at about 580 ºC. Couhert et al. reported
240
that the decomposition of pure components (such as cellulose) differs from real
241
materials due to interferences of other components (Couhert et al., 2009). The solid
242
residues of cotton wool, cotton textile and tissue paper samples were 2%, 5% and 2%,
243
respectively. The TG curves of these materials were very similar among them and in
244
comparison with cellulose; nevertheless, cellulose and cotton wool start to decompose
245
at slightly higher temperatures than tissue paper and cotton textile, indicating that the
246
cellulose of cotton wool has a slightly different structure than cellulose from tissue
247
paper and cotton textile.
248
Both printing paper samples presented an extra peak under inert atmosphere around
249
700 ºC, which maintained under oxidative atmosphere. Moreover, these two samples
250
only presented a total weight loss around 85% under air, which could be attributed to 10
251
the presence of inorganic compounds that are added during paper manufacturing.
252
Accordingly, these two paper samples present a lower cellulose content than cotton
253
wool, cotton textile and tissue paper. This difference on the content of cellulose
254
between the samples was considered for calculations, as mentioned in Section 2.4.
255
Figure 1
256
The catalyst has been extensively characterized and the results were reported
257
elsewhere (Ribeiro et al., 2017).
258 259
3.2. Conversion of cellulosic materials
260
We have previously reported the direct conversion of cellulose into sugar alcohols
261
using Ru/CNT as catalyst (Ribeiro et al., 2017). A total yield of sorbitol of 51% was
262
obtained after 5 h of reaction in the conversion of ball-milled cellulose with 50 bar H2 at
263
205 ºC. The yield of sorbitol was further increased to 61% in just 1 h of reaction by mix-
264
milling cellulose with the catalyst. In addition, the catalyst presented excellent stability
265
during the current reaction conditions and could be reused up to at least four
266
successive runs with practically no loss in activity and selectivity or metal leaching to
267
solution.
268
In the present work, we applied the previous direct conversion method to obtain sugar
269
alcohols from cellulosic materials, such as cotton (wool and textile) and paper (tissue,
270
printing and recycled) in the presence of Ru/CNT. Figure 2 shows the evolution of the
271
conversion of the different untreated materials in comparison to that of cellulose. The
272
conversion after 5 h varied between 18 and 38%, the lowest conversion obtained with
273
cotton wool and the highest with tissue paper and microcrystalline cellulose. The
274
conversions obtained were very small, which is explained by the rigid crystalline
275
structure of the cellulosic materials (see Section 3.1) that difficults the hydrolysis of the
276
materials. Accordingly, the conversion of the unmilled materials only afforded yields of
11
277
sorbitol up to 8%. Therefore, these results had proven that the pre-treatment of the
278
substrates is a pre-requisite for achieving high conversions of cellulosic materials.
279
Figure 2
280
Then, in order to increase the conversion, the materials were ball-milled prior to
281
reaction. A huge difference was observed for the conversion of the materials, as
282
depicted in Figure 2a. A 100% conversion of cotton wool, cotton textile and tissue
283
paper was achieved after just 4 h of reaction, which was even higher than that of
284
cellulose (83.5%). The ball-milling of the materials greatly increased the conversion
285
due to the decrease of cellulose crystallinity. In Section 3.1 it was observed that the
286
(002) diffraction peak at 22.6º was broadened by the ball-milling treatment of the
287
cellulosic samples, indicating that the cellulose crystallinity in cotton wool, cotton textile
288
and tissue paper was decreased. The conversion of the printing paper samples after 5
289
h of reaction only presented an increase from about 25-33% to 50% with the ball-
290
milling pre-treatment. The lowest conversion obtained in comparison to the remaining
291
materials can be explained by the presence of inorganic impurities and additives that
292
were used upon the manufacturing of paper (e.g. TiO2). These substances could inhibit
293
the conversion of printing paper under the conditions used in the present work.
294
Yields of sorbitol of 47.6, 45.0 and 45.3% were attained after 5 h of reaction from the
295
conversion of cotton wool, cotton textile and tissue paper, respectively. Although the
296
yields of sorbitol from the direct conversion of pure cellulose were higher during most of
297
the reaction time, the yield of sorbitol achieved at the end of the reaction (50.8% after 5
298
h) was very close to that obtained from the conversion of the waste materials. It is also
299
possible to observe from Figure 2b that the yields of sorbitol obtained during the
300
reaction are very similar for the three samples (cotton wool, cotton textile and tissue
301
paper), and especially between the two cotton samples. Therefore, the decrease of the
302
materials crystallinity has a direct effect on the improvement of the conversion and
303
production of sorbitol. Besides sorbitol, C2-C6 sugar alcohols such as xylitol, glycerol, 12
304
glucose, ethylene glycol (EG) and propylene glycol (PG) could also be obtained directly
305
from the cellulosic materials (Table 2). The conversion of cotton wool, cotton textile and
306
tissue paper increased with increasing time and consequently the water-soluble
307
products also increased, such as glucose, xylitol and glycerol. A total amount of sugar
308
alcohols of 0.803, 0.703 and 0.733 g were obtained from 1 g of cotton wool, cotton
309
textile and tissue paper after 5 h of reaction, respectively, where the amounts of
310
sorbitol, xylitol and glycerol were between 0.509-0.534 g, 0.030-0.101 g and 0.055-
311
0.070 g, respectively (Table 2). Tissue paper and cotton wool also afforded EG and PG
312
with amounts of EG of 0.055 and 0.059 g and of 0.012 and 0.025 g of PG after 5 h of
313
reaction, respectively.
314
Table 2
315
In addition to the low conversion achieved, the printing paper samples were not
316
significantly converted into sorbitol (yields around 7% after 5 h of reaction) (Figure 2b)
317
neither into the other sugar alcohols. The amount of total sugar alcohols (0.194
318
g/gprinting-paper and 0.213 g/grecycled-paper after 5 h of reaction) was considerably lower than
319
that directly obtained from tissue paper, cotton wool and cotton textile (Table 2). The
320
reason could be related to the presence of impurities that inhibit the conversion to
321
sorbitol and other sugar alcohols.
322
There are many substances that are added to paper during its manufacturing. In order
323
to test if the presence of those compounds can affect the reaction, we selected one of
324
the most important known additives (TiO2) and added it to the reaction mixture along
325
with cellulose. The reaction was performed under the same conditions and the results
326
obtained are presented in Figure 3. The evolution of the yield of sorbitol was not very
327
different up to 1.5 h of reaction. After that time, a continuous decrease was observed
328
during the remaining reaction time. The yield of sorbitol achieved after 5 h of reaction
329
decreased from 50.8 to 22.4% due to the presence of TiO2 on the reaction mixture.
330
Therefore, we can conclude that the presence of TiO2 was unfavourable for the 13
331
formation of sorbitol directly from cellulose. Since there are many other substances that
332
are added during paper manufacturing and each of them could affect the conversion,
333
as well as their combination, an intensive study needs to be carried out in order to
334
better understand the effect that each added component could have on the
335
transformation of printing paper samples.
336
Figure 3
337
Finally, in this work we also tested the effect of mix-milling the substrates with the
338
catalyst. Such approach had already proven to allow achieving higher yields of sugar
339
alcohols in less reaction time, by greatly increasing the conversion rates (Kobayashi et
340
al., 2013; Komanoya et al., 2014; Ribeiro et al., 2015b; Ribeiro et al., 2017). A
341
conversion of 100% was achieved for tissue paper, cotton wool and cotton textile in just
342
1, 1.5 and 2 h of reaction, respectively (Figure 2a). Once more, the conversion of
343
printing paper samples did not went further than about 50% after 5 h of reaction.
344
Therefore, with exception of the two printing paper samples, the rate of conversion of
345
the materials was greatly enhanced by the mix-milling pre-treatment. The catalyst and
346
the different materials (tissue paper, cotton wool and cotton textile) were also
347
separately ball-milled and tested under the same conditions to understand the effect of
348
ball-milling (supplementary material). The mentioned increase of the conversion rate
349
was not observed when the catalyst and the substrates were separately ball-milled,
350
confirming that the enhancement of the performance was not due to the catalyst ball-
351
milling, but to the good physical contact between the catalyst and the substrates, which
352
resulted from the mix-milling.
353
Yields of sorbitol around 50% were attained in just 2 h of reaction from the conversion
354
of cotton wool, cotton textile and tissue paper (Figure 2b), corresponding to amounts of
355
sorbitol over 0.5 g/gsubstrate (see Table 2). Although no significant increase was observed
356
for the yields of sorbitol at the end of the reaction (i.e. 5 h) with the mix-milling
357
comparatively to the ball-milling, they were reached in less than half of the time. The 14
358
mix-milling pre-treatment of cotton wool, cotton textile and tissue paper enhanced the
359
total sugar alcohol yield from 0.372 to 0.674 g, from 0.353 to 0.709 g and from 0.607 to
360
0.792 g after 2 h of reaction, respectively (Table 2). The amounts of sorbitol were also
361
drastically increased from 0.183 to 0.525 g for cotton wool, from 0.217 to 0.511 g for
362
cotton textile and from 0.338 to 0.559 g for tissue paper, but the amount of other
363
products did not change so significantly. Tissue paper also afforded an amount of
364
xylitol of 0.112 g after 2 h of reaction. The mix-milling had once again no effect on the
365
yield of sorbitol obtained from the printing paper samples.
366
As a final point, with (ball- or mix-milling) or without pre-treatment, the two printing
367
paper samples presented similar performances indicating that the fact that the sample
368
is recycled or not might not influence the reaction.
369
The results presented in this work show that we have succeeded in producing sugar
370
alcohols, especially sorbitol, from cellulosic materials using a Ru/CNT catalyst in water
371
without any acid catalysts, pre-treatment or reaction medium. Moreover, the catalyst
372
used can be easily recovered by filtration from the reaction mixture, allowing its
373
reusability.
374 375
4. Conclusions
376
Waste cellulosic materials were converted into sorbitol using a Ru/CNT catalyst, and
377
conversions close to 40% were obtained after 5 h, with sorbitol yields below 10%. Ball-
378
milling of cotton textile, cotton wool and tissue paper resulted in 100% conversion of
379
these materials and sorbitol yields around 50% after 5 h. Mix-milling these materials
380
with Ru/CNT greatly enhanced their rate of conversion and sorbitol yields of 50% were
381
reached in 2 h. However, printing paper only had a conversion up to 50% after 5 h, with
382
sorbitol yields around 7%. Moreover, TiO2 had a negative impact on the production of
383
sorbitol.
15
384 385
Acknowledgements
386
This work was financially supported by project POCI-01-0145-FEDER-006984 –
387
Associate Laboratory LSRE-LCM funded by FEDER through COMPETE2020 -
388
Programa Operacional Competitividade e Internacionalização (POCI) – and by national
389
funds through FCT - Fundação para a Ciência e a Tecnologia. L.S. Ribeiro
390
acknowledges her Ph.D. scholarship (SFRH/BD/86580/2012) from FCT.
391 392
Appendix A. Supplementary data
393 394
References
395
1.
Beeck, B.O.d., Geboers, J., Vyver, S.V.d., Lishout, J.V., Snelders, J., Huijgen,
396
W.J.J., Courtin, C.M., Jacobs, P.A., Sels, B.F. 2013. Conversion of
397
(Ligno)Cellulose Feeds to Isosorbide with Heteropoly Acids and Ru on Carbon.
398
ChemSusChem, 6, 199-208.
399
2.
Byun, J., Han, J. 2016a. Catalytic production of biofuels (butene oligomers) and
400
biochemicals (tetrahydrofurfuryl alcohol) from corn stover. Bioresource Technol,
401
211, 360-366.
402
3.
Byun, J., Han, J. 2016b. Process synthesis and analysis for catalytic conversion
403
of lignocellulosic biomass to fuels: Separate conversion of cellulose and
404
hemicellulose using 2-sec-butylphenol (SBP) solvent. Appl Energ, 171, 483-
405
490.
406
4.
Cheng, L., Liu, H., Cui, Y., Xue, N., Ding, W. 2014. Direct conversion of corn
407
cob to formic and acetic acids over nano oxide catalysts. J Energ Chem, 23, 43-
408
49.
409
5.
Couhert, C., Commandre, J.-M., Salvador, S. 2009. Is it possible to predict gas
410
yields of any biomass after rapid pyrolysis at high temperature from its
411
composition in cellulose, hemicellulose and lignin? Fuel, 88, 408-417.
412
6.
Deng, W., Tan, X., Fang, W., Zhang, Q., Wang, Y. 2009. Conversion of
413
Cellulose into Sorbitol over Carbon Nanotube-Supported Ruthenium Catalyst.
414
Catal Lett, 133, 167-174.
16
415
7.
Deng, W., Zhang, H., Xue, L., Zhang, Q., Wang, Y. 2015. Selective activation of
416
the C–O bonds in lignocellulosic biomass for the efficient production of
417
chemicals. Chinese J Catal, 36, 1440-1460.
418
8.
Ding, L.N., Wang, A.Q., Zheng, M.Y., Zhang, T. 2010. Selective transformation
419
of cellulose into sorbitol by using a bifunctional nickel phosphide catalyst.
420
ChemSusChem, 3, 818-821.
421
9.
Feng, Y., Li, G., Li, X., Zhu, N., Xiao, B., Li, J., Wang, Y. 2016. Enhancement of
422
biomass conversion in catalytic fast pyrolysis by microwave-assisted formic acid
423
pretreatment. Bioresource Technol, 214, 520-527.
424
10.
425 426
Fukuoka, A., Dhepe, P. 2006. Catalytic conversion of cellulose into sugar alcohols. Angew Chem Int Edit, 45, 5161-5163.
11.
Guha, S.K., Kobayashi, H., Hara, K., Kikuchi, H., Aritsuka, T., Fukuoka, A.
427
2011. Hydrogenolysis of sugar beet fiber by supported metal catalyst. Catal
428
Commun, 12, 980-983.
429
12.
Guo, X., Wang, X., Guan, J., Chen, X., Qin, Z., Mu, X., Xian, M. 2014. Selective
430
hydrogenation of D-glucose to D-sorbitol over Ru/ZSM-5 catalysts. Chinese J
431
Catal, 35, 733-740.
432
13.
Kobayashi, H., Ito, Y., Komanoya, T., Hosaka, Y., Dhepe, P.L., Kasai, K., Hara,
433
K., Fukuoka, A. 2011. Synthesis of sugar alcohols by hydrolytic hydrogenation
434
of cellulose over supported metal catalysts. Green Chem, 13, 326-333.
435
14.
Kobayashi, H., Yabushita, M., Komanoya, T., Hara, K., Fujita, I., Fukuoka, A.
436
2013. High-Yielding One-Pot Synthesis of Glucose from Cellulose Using Simple
437
Activated Carbons and Trace Hydrochloric Acid. ACS Catalysis, 3, 581-587.
438
15.
Kobayashi, H., Yamakoshi, Y., Hosaka, Y., Yabushita, M., Fukuoka, A. 2014.
439
Production of sugar alcohols from real biomass by supported platinum catalyst.
440
Catal Today, 226, 204-209.
441
16.
Komanoya, T., Kobayashi, H., Hara, K., Chun, W.-J., Fukuoka, A. 2014. Kinetic
442
Study of Catalytic Conversion of Cellulose to Sugar Alcohols under Low-
443
Pressure Hydrogen. ChemCatChem, 6, 230-236.
444
17.
445 446
Kusserow, B., Schimpf, S., Claus, P. 2003. Hydrogenation of Glucose to Sorbitol over Nickel and Ruthenium Catalysts. Adv Synth Catal, 345, 289-299.
18.
Li, C., Zhao, X., Wang, A., Huber, G.W., Zhang, T. 2015. Catalytic
447
Transformation of Lignin for the Production of Chemicals and Fuels. Chem Rev,
448
115, 11559-11624.
449
19.
Luo, C., Wang, S., Liu, H. 2007. Cellulose conversion into polyols catalysed by
450
reversible formed acids and supported ruthenium clusters in hot water. Angew
451
Chem Int Edit, 46, 7636-7639. 17
452
20.
Mäki-Arvela, P.i., Salmi, T., Holmbom, B., Willför, S., Murzin, D.Y. 2011.
453
Synthesis of Sugars by Hydrolysis of Hemicelluloses - A Review. Chem Rev,
454
111, 5638-5666.
455
21.
Moltó, J., Font, R., Conesa, J.A., Martín-Gullón, I. 2006. Thermogravimetric
456
analysis during the decomposition of cotton fabrics in an inert and air
457
environment. J Anal Appl Pyrolysis, 76, 124-131.
458
22.
Órfão, J.J.M., Antunes, F.J.A., Figueiredo, J.L. 1999. Pyrolysis kinetics of
459
lignocellulosic materials—three independent reactions model. Fuel, 78, 349-
460
358.
461
23.
Palkovits, R., Tajvidi, K., Procelewska, J., Rinaldi, R., Ruppert, A. 2010.
462
Hydrogenolysis of cellulose combining mineral acids and hydrogenation
463
catalysts. Green Chem, 12, 972-978.
464
24.
Park, S., Baker, J., Himmel, M., Parilla, P., Johnson, D. 2010. Cellulose
465
crystallinity index: measurement techniques and their impact on interpreting
466
cellulase performance. Biotechnol Biofuels, 3, 1-10.
467
25.
468 469
Rao, L.V., Goli, J.K., Gentela, J., Koti, S. 2016. Bioconversion of Lignocellulosic biomass to xylitol: An overview. Bioresource Technol, 213, 299-310.
26.
Ribeiro, L.S., Órfão, J.J.M., Pereira, M.F.R. 2015a. Comparative study of
470
different catalysts for the direct conversion of cellulose to sorbitol. Green
471
Process Synth, 4, 71-78.
472
27.
473 474
Ribeiro, L.S., Órfão, J.J.M., Pereira, M.F.R. 2015b. Enhanced direct production of sorbitol by cellulose ball-milling. Green Chem 17, 2973-2980.
28.
Ribeiro, L.S., Delgado, J.J., Órfão, J.J.M., Pereira, M.F.R. 2017. Direct
475
conversion of cellulose to sorbitol over ruthenium catalysts: Influence of the
476
support. Catal Today, 279, 244-251.
477
29.
Romero, A., Alonso, E., Sastre, Á., Nieto-Márquez, A. 2016. Conversion of
478
biomass into sorbitol: Cellulose hydrolysis on MCM-48 and d-Glucose
479
hydrogenation on Ru/MCM-48. Micropor Mesopor Mat, 224, 1-8.
480
30.
Shen, D., Ye, J., Xiao, R., Zhang, H. 2013. TG-MS analysis for thermal
481
decomposition of cellulose under different atmospheres. Carbohyd Polym, 98,
482
514-521.
483
31.
Sivasangar, S., Taufiq-Yap, Y.H., Zainal, Z., Kitagawa, K. 2013. Thermal
484
behavior of lignocellulosic materials under aerobic/anaerobic environments. Int
485
J Hydrogen Energ, 38, 16011-16019.
486 487
32.
Van de Vyver, S., Geboers, J., Dusselier, M., Schepers, H., Vosch, T., Zhang, L., Van Tendeloo, G., Jacobs, P.A., Sels, B.F. 2010. Selective Bifunctional
18
488
Catalytic Conversion of Cellulose over Reshaped Ni Particles at the Tip of
489
Carbon Nanofibers. ChemSusChem, 3, 698-701.
490
33.
Van de Vyver, S., Geboers, J., Schutyser, W., Dusselier, M., Eloy, P., Dornez,
491
E., Seo, J.W., Courtin, C.M., Gaigneaux, E.M., Jacobs, P.A., Sels, B.F. 2012.
492
Tuning the Acid/Metal Balance of Carbon Nanofiber-Supported Nickel Catalysts
493
for Hydrolytic Hydrogenation of Cellulose. ChemSusChem, 5, 1549-1558.
494
34.
Wang, H., Zhu, L., Peng, S., Peng, F., Yu, H., Yang, J. 2012. High efficient
495
conversion of cellulose to polyols with Ru/CNTs as catalyst. Renew Energ, 37,
496
192-196.
497
35.
Wu, Z., Ge, S., Ren, C., Zhang, M., Yip, A., Xu, C. 2012. Selective conversion
498
of cellulose into bulk chemicals over Brønsted acid-promoted ruthenium
499
catalyst: one-pot vs. sequential process. Green Chem, 14, 3336-3343.
500
36.
501 502
Yabushita, M., Kobayashi, H., Fukuoka, A. 2014. Catalytic transformation of cellulose into platform chemicals. Appl Catal B-Environ, 145, 1-9.
37.
Yadav, M., Mishra, D.K., Hwang, J.-S. 2012. Catalytic hydrogenation of xylose
503
to xylitol using ruthenium catalyst on NiO modified TiO2 support. Appl Catal A-
504
Gen, 425–426, 110-116.
505
38.
Yamaguchi, A., Sato, O., Mimura, N., Hirosaki, Y., Kobayashi, H., Fukuoka, A.,
506
Shirai, M. 2014. Direct production of sugar alcohols from wood chips using
507
supported platinum catalysts in water. Catal Commun, 54, 22-26.
508
39.
509 510
Yamaguchi, A., Sato, O., Mimura, N., Shirai, M. 2016. Catalytic production of sugar alcohols from lignocellulosic biomass. Catal Today, 265, 199-202.
40.
Yang, P., Kobayashi, H., Hara, K., Fukuoka, A. 2012. Phase change of nickel
511
phosphide catalysts in the conversion of cellulose into sorbitol. ChemSusChem,
512
5, 920-926.
513 514
41.
Zhao, X., Xu, J., Wang, A., Zhang, T. 2015. Porous carbon in catalytic transformation of cellulose. Chin J Catal, 36, 1419-1427.
515
19
516
Figure Captions
517 518 519 520
Figure 1 – TG (on the left) and DTG (on the right) curves of cellulosic materials under a) air and b) nitrogen. Figure 2 – a) Conversion and b) yield of sorbitol of the different unmilled (A), ball-milled
521
(B) and mix-milled (C) materials. Reaction conditions: 300 mL water, 750 mg
522
substrate, 300 mg Ru/CNT, 205 ºC, 50 bar H2, 150 rpm.
523
Figure 3 – Effect of TiO2 on the production of sorbitol from cellulose. Reaction
524
conditions: 300 mL water, 750 mg ball-milled cellulose (+ 150 mg TiO2), 300 mg
525
Ru/CNT, 205 ºC, 50 bar H2, 150 rpm.
526
20
527 528
Tables Table 1 – Elemental analysis results for the different materials. Sample
C (at.%)
H (at.%)
O (at.%)
Cellulose
44.2 ± 0.1
6.9 ± 0.1
48.9 ± 0.3
Cotton wool
41.7 ± 0.8
5.7 ± 0.1
52.6 ± 0.8
Cotton textile
42.9 ± 0.5
6.4 ± 0.1
50.7 ± 0.3
Printing paper
37.2 ± 0.3
4.6 ± 0.1
58.2 ± 0.5
Recycled printing paper
38.8 ± 0.3
5.0 ± 0.1
56.2 ± 0.4
Tissue paper
43.7 ± 0.6
5.9 ± 0.3
50.4 ± 0.7
529 530
21
531
Table 2 – Amount of products obtained in the conversion of cellulosic materials after 2 h [and 5 h] of reaction. Reaction
532
conditions: 300 mL water, 750 mg substrate, 300 mg Ru/CNT, 205 ºC, 50 bar H2, 150 rpm.
Material
Cellulose
Cotton wool
Cotton textile
Printing paper
Recycled printing paper
Tissue paper
Pre-treatment
Amount of product (g/gmaterial) Sorbitol
Glucose
Xylitol
Glycerol
EG
PG
Total
Unmilled
0.098 [0.106]
0.000 [0.000]
0.005 [0.017]
0.031 [0.040]
0.000 [0.000]
0.000 [0.000]
0.134 [0.163]
Ball-milled
0.533 [0.571]
0.040 [0.000]
0.037 [0.082]
0.045 [0.052]
0.047 [0.057]
0.000 [0.016]
0.702 [0.778]
Mix-milled
0.782 [0.656]
0.000 [0.000]
0.029 [0.053]
0.038 [0.053]
0.000 [0.000]
0.000 [0.000]
0.849 [0.762]
Unmilled
0.066 [0.073]
0.000 [0.000]
0.000 [0.000]
0.021 [0.028]
0.047 [0.049]
0.000 [0.005]
0.134 [0.155]
Ball-milled
0.183 [0.534]
0.086 [0.085]
0.008 [0.030]
0.038 [0.070]
0.050 [0.059]
0.007 [0.025]
0.372 [0.803]
Mix-milled
0.525 [0.562]
0.093 [0.081]
0.018 [0.040]
0.038 [0.061]
0.000 [0.000]
0.000 [0.000]
0.674 [0.744]
Unmilled
0.064 [0.072]
0.000 [0.000]
0.000 [0.000]
0.023 [0.027]
0.000 [0.000]
0.000 [0.000]
0.087 [0.099]
Ball-milled
0.217 [0.510]
0.090 [0.084]
0.009 [0.054]
0.037 [0.055]
0.000 [0.000]
0.000 [0.000]
0.353 [0.703]
Mix-milled
0.511 [0.505]
0.089 [0.081]
0.015 [0.024]
0.035 [0.048]
0.051 [0.054]
0.008 [0.013]
0.709 [0.725]
Unmilled
0.063 [0.063]
0.000 [0.000]
0.000 [0.000]
0.027 [0.028]
0.053 [0.059]
0.000 [0.013]
0.143 [0.137]
Ball-milled
0.063 [0.064]
0.000 [0.000]
0.003 [0.002]
0.031 [0.029]
0.059 [0.069]
0.016 [0.030]
0.172 [0.194]
Mix-milled
0.065 [0.066]
0.000 [0.000]
0.003 [0.001]
0.029 [0.036]
0.061 [0.073]
0.022 [0.045]
0.180 [0.221]
Unmilled
0.062 [0.062]
0.000 [0.000]
0.001 [0.001]
0.030 [0.038]
0.056 [0.064]
0.015 [0.023]
0.164 [0.188]
Ball-milled
0.064 [0.065]
0.000 [0.000]
0.002 [0.002]
0.042 [0.045]
0.063 [0.071]
0.020 [0.030]
0.191 [0.213]
Mix-milled
0.065 [0.065]
0.000 [0.000]
0.000 [0.001]
0.042 [0.044]
0.064 [0.071]
0.027 [0.042]
0.198 [0.223]
Unmilled
0.083 [0.093]
0.000 [0.000]
0.057 [0.052]
0.029 [0.038]
0.000 [0.000]
0.000 [0.000]
0.169 [0.183]
Ball-milled
0.338 [0.509]
0.082 [0.000]
0.094 [0.101]
0.046 [0.056]
0.047 [0.055]
0.000 [0.012]
0.607 [0.733]
Mix-milled
0.559 [0.517]
0.000 [0.000]
0.112 [0.102]
0.055 [0.056]
0.053 [0.055]
0.013 [0.014]
0.792 [0.744]
22
533
Figures
534 535
Figure 1 – TG (on the left) and DTG (on the right) curves of cellulosic materials under a) air and b) nitrogen. 23
536 537
Figure 2 – a) Conversion and b) yield of sorbitol of the different unmilled (A), ball-milled (B) and mix-milled (C) materials. Reaction conditions:
538
300 mL water, 750 mg substrate, 300 mg Ru/CNT, 205 ºC, 50 bar H2, 150 rpm. 24
539
540 541 542
Figure 3 – Effect of TiO2 on the production of sorbitol from cellulose. Reaction conditions: 300 mL water, 750 mg ball-milled substrate, 150 mg TiO2, 300 mg Ru/CNT, 205 ºC, 50 bar H2, 150 rpm.
543
25
544 545
26
546
Highlights:
547
•
Materials ball-milling disrupted their crystallinity, easing the direct conversion
548
•
Sorbitol was directly obtained from cotton wool, cotton textile and tissue paper
549
•
Mix-milling greatly enhanced the rate of conversion of cellulosic biomass
550
•
Sorbitol amounts higher than 0.5 g could be obtained from 1 g of cellulosic biomass
551
•
Printing paper (white and recycled) was not converted to sugar alcohols
552 553
27