- Email: [email protected]
Pyrolysis of corn stalk biomass briquettes in a scaled-up microwave technology
Accepted Manuscript Pyrolysis of corn stalk biomass briquettes in a microwave technology Arshad Adam Salema, Muhammad T. Afzal, Lyes Bennamoun PII: DOI: Reference:
S0960-8524(17)30242-0 http://dx.doi.org/10.1016/j.biortech.2017.02.113 BITE 17685
To appear in:
Bioresource Technology
Received Date: Revised Date: Accepted Date:
19 December 2016 23 February 2017 24 February 2017
Please cite this article as: Salema, A.A., Afzal, M.T., Bennamoun, L., Pyrolysis of corn stalk biomass briquettes in a microwave technology, Bioresource Technology (2017), doi: http://dx.doi.org/10.1016/j.biortech.2017.02.113
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.
Pyrolysis of corn stalk biomass briquettes in a microwave technology
1 2
Arshad Adam Salemaa, Muhammad T. Afzalb*, and Lyes Bennamounb
3 4 a
5 6
Discipline of Mechanical Engineering, School of Engineering, Monash University Malaysia, Jalan Lagoon Selatan, 46150 Bandar Sunway, Selangor, Malaysia.
b
7
Department of Mechanical Engineering, Faculty of Engineering, University of New
8
Brunswick, Fredericton, NB E3B 5A3, Canada.
9 10
Abstract
11
Pyrolysis of corn stalk biomass briquettes was carried out in a developed microwave (MW)
12
reactor supplied with 2.45 GHz frequency using 3kW power generator. MW power and
13
biomass loading were the key parameters investigated in this study. Highest bio-oil,
14
biochar, and gas yield of 19.6 %, 41.1%, and 54.0% was achieved at different process
15
condition. In terms of quality, biochar exhibited good heating value (32 MJ/kg) than bio-oil
16
(2.47 MJ/kg). Bio-oil was also characterised chemically using FTIR and GC-MS method.
17
This work may open new dimension towards development of large-scale MW pyrolysis
18
technology.
19 20
Keywords: Biomass; Briquette; Microwave; Pyrolysis; Product; Yield and quality
21
*
Corresponding author: [email protected]; Tel: +1-506-453-4880; Fax: +1-506-453-5025
1
22 23
1. Introduction
24
Corn is one of the major agricultural crops in the United States and Canada and in
25
several other parts of the world. This agricultural business generates huge amount of corn
26
residues every year in form of corn stalk, corn cobs, and corn stover. For instance, Ontario
27
and Quebec provinces produced about 5.1 and 3.1 million tonnes of corn stalk, respectively
28
between year 2001 and 2005(Bailey, 2006).Presently, corn stalk is either used as soil
29
amendment in the farm itself or utilized as fuel for the production of energy. However,
30
there still remains excess biomass which needs to be disposed of safely. Despite being a
31
waste, one of the potential utilizations of corn biomass materials is the production of fuels
32
and chemicals (Huang et al., 2013). Environmental regulations and serious climatic issues
33
have become major contributing factors for the use of biomass materials as an alternative
34
source of energy and fuels.
35
Despite detailed research work on Microwave Assisted Pyrolysis (MAP) technology
36
(Motasemi and Afzal, 2013; Yin, 2012; Macquarrie et al., 2012), there still remains a clear
37
lack in scale-up of MW technology in processing biomass materials. There are some
38
attempts in scale-up of MW reactor for biomass processing, but no results are yet reported
39
in the open literature which might be due to reason of patent protection or commercial
40
development. Thus, development of large scale MW reactor will surely help in removing
41
the obstacle for industrial applications (Li et al., 2016). Very recently (Pianroj et al., 2016),
42
pyrolysis of oil palm shell was carried out in a scaled up reactor. However, their quartz
43
glass reactor size (10.7 cm in diameter and 14.0 cm in height) was almost similar to that of
44
previous work of Salema and Ani, 2011. Even though, Pianroj et al., 2016 claimed about 2
45
scale-up of MW system, but they processed about 400 g of biomass. There are some
46
exception studies where large amount of biomass was processed (Zhao et al., 2010; Miura
47
et al., 2004), but are limited. Recently, Robinson et al., 2015 reported an interesting result
48
about the effect of biomass sample size on the product yield and quality using MW
49
technology. However, their sample size was in the range of 5 to 20 g only. In most of the
50
research conducted under MW technology, biomass weight was limited in size of grams.
51
Therefore, very little data (Robinson et al., 2015; Salema and Afzal, 2015) is available in
52
the open literature and limited work has been carried out on the effect of biomass sample
53
size. The present study aims to demonstrate the scale-up of the MW reactor system in terms
54
of amount of biomass processed i.e. kilograms.
55
Although several researchers have processed various types of biomass materials and
56
in different forms loose, shredded, grinded (Salema and Ani, 2012a and 2012b; Abubakar
57
et al., 2013), pellets (Robinson et al., 2009; Undri et al., 2015), whole wood log (Miura et
58
al., 2004) and bale (Zhao et al., 2010), but none have conducted MW pyrolysis of biomass
59
briquettes. Certainly, densification (pellet and briquette) increases the density of biomass,
60
thus reducing the volume of biomass which can save the transport cost. Moreover, it can
61
also significantly reduce the reactor volume or size and thereby reduction in the cost.
62
However, the effect of densification on factors such as heat and mass transfer needs to be
63
investigated.
64
Table 1 presents the pyrolysis technologies used to pyrolyse corn stalk biomass. It is
65
clear that most of researchers used conventional technologies to pyrolyze corn stalk
66
biomass, except one (Zhao et al., 2010) and their focus was on the pyrolysis behavior,
67
temperature and weight profile, and little on product yield and no evidence on the product 3
68
quality. Comparatively, MW technology has provided better product quality due to some
69
unique heating mechanism. One of the most important is the direct interaction of the
70
material with the MW. These and other non-thermal effects of MW heating have
71
revolutionized the processing of the biomass under MW. However, biomass is poor
72
absorber of the MW due to low loss dielectric properties. In order to increase the heating
73
rate as well as to induce much faster pyrolysis reactions they are usually mixed with
74
carbonaceous materials.
75
From our past numerical simulation study (Salema and Afzal, 2015), MW power
76
and biomass loading was considered as the key process parameters of the biomass
77
pyrolysis. But how these parameters would influence the product yield and quality in real
78
application? is still a question that needs attention. The objectives of this study was to
79
reveal the development and the potential of rector design, and to pyrolyze the biomass
80
briquettes in a developed MW reactor which was done for the first time, and to study the
81
effects of MW power and biomass loading (size in kilograms) on the product yield and
82
quality. Biochar and bio-oil were subjected to elemental analysis, heating value, and FT-IR
83
in addition to GC-MS analysis of bio-oil.
84 85
2. Materials and methods
86 87
2.1.Materials
88
Corn stalk (CS) biomass was obtained from the neighboring corn farm situated in the
89
Mactaquac region, Fredericton, New Brunswick, Canada. The corn stalk biomass samples
90
were collected from freshly harvested corn crop in the month of October 2015. The samples 4
91
were chopped, shredded, dried in an oven for 6 hours and stored at room condition. A
92
hydraulic briquetting machine with a constant pressure of 100 MPa and holding time of 5 s
93
was used to produce CS briquettes with a size of 0.04 m diameter and 0.02 m length. This
94
study does not take into account the effect of densification parameters. Each CS briquette
95
weighted around 30 g. The CS briquettes were stored in a sealed container to minimize the
96
moisture activity. The proximate and ultimate analysis of CS biomass showed moisture
97
content (6.8 wt.%), volatile matter (76.2 wt.%), fixed carbon (17 wt.%), carbon (46.67
98
wt.%), hydrogen (6.01 wt.%), nitrogen (0.02 wt.%), oxygen (47.28 wt.%), and sulphur
99
(0.02 wt.%). The biochar obtained from the MW pyrolysis of CS briquettes were used as
100
MW absorbers in the present study.
101 102
2.2.Development and assembly of MW pyrolysis system
103
The microwave assisted pyrolysis (MAP) tests were carried out in a custom-built
104
cylindrical stainless steel 309 reactor connected to a 3 kW and 2450 MHz frequency
105
microwave generator system. The MW generator was procured from Muegge GmbH,
106
Reichelsheim, Germany which consisted of MW power supply (Model MX3000DL-
107
152KL, Muegge GmbH), magnetron head (Model MH3000S-250BB, Muegge GmbH)
108
embedded with isolator and MW directional coupler. A complete MW pyrolysis system
109
consisted of MW generator, waveguide, 3 stub-tuner, a stainless steel 309 cylindrical
110
reactor, a nitrogen gas generator, two K-type metallic thermocouples, Pico data logging
111
system, a pilot scale glass condenser and a bio-oil collector as shown in Fig.1.
112 113
The MW power supply and magnetron head was delivered with a water (temperature between 18 and 20 °C) flowing continuously with a flow rate of 4.5 liter per 5
114
minute (L/min) in order to facilitate the cooling process. As a safety feature, the MW
115
generator system will shut OFF automatically if the water flow rate drops below 4 L/min.
116
The same water was also used for isolator to absorb any reflected MW. The magnetron
117
generates the MW that travel through the waveguide and finally to a cylindrical reactor.
118
Any reflected MW gets diverted into a dummy water load (isolator) to avoid damage to the
119
magnetron. A MW directional coupler also embedded in the magnetron head allows to
120
measure any reflected power. The reflected power was much dependent on the applied MW
121
power and it increased with increase in MW power. For 900, 1200, and 1500 W it was
122
about 0%, 2%, and 5 %, respectively. The three stub tuner helps to match the impedance of
123
the waveguide segments to the load. Rectangular waveguide (WR 340) of 13.8 cm × 9.5 cm
124
also acquired from Muegge, Germany was used in this study.
125
A cylindrical stainless steel 309 reactor of 35.5 cm inner diameter and 53.34 cm in
126
height was designed and fabricated to carry out the pyrolysis process. The size of the
127
reactor was determined based on the amount of densified biomass to be processed, say 7 to
128
8 kg. This cylindrical reactor was covered from top and bottom with SS 309 tank dish head.
129
The top tank head was facilitated with 5 openings; two openings were attached with a 60
130
cm long, 6 mm inner diameter and 2 mm thick thermowell and were closed at the bottom
131
for the insertion of K-type thermocouples. The position and the distance of thermocouples
132
are presented in section 3.1. The centered opening was for the entrance of nitrogen gas
133
inside the cylindrical reactor. Any other openings were sealed and closed to avoid any exit
134
of the vapors or MW leakage. The nitrogen gas was supplied by a nitrogen gas generator
135
(VWR Inc., Ontario, Canada, Model CA26000-014). The bottom tank head had a single
136
opening for the exit of the pyrolysis vapor into the condenser. Both the top and bottom tank 6
137
head were sealed and closed tightly with the help of high temperature gaskets and stainless
138
steel clamps. To avoid the pyrolysis vapor to travel through the waveguide and get
139
deposited into the magnetron, mica sheet of 2 mm thick was placed at the waveguide which
140
opened into the SS reactor. The mica sheet is transparent to microwave, but at the same
141
time it restrict any vapor to pass into the waveguide. The height at which the waveguide
142
opened into the reactor was adequate to distribute the MW into the reactor and interact with
143
the biomass sample.
144
During the experiment, the reflected power from the microwave was monitored on
145
the Muegge software, and the temperature readings from the thermocouples were measured
146
and continuously recorded in a real time. The experiments were run under atmospheric
147
pressure. The vapors generated from pyrolysis process were continuously swept out of the
148
reactor with the help of nitrogen gas. The vapors passed through a pilot scale glass
149
condenser (VWR Inc., Ontario, Model no. 6016-141) which condensed into liquid called as
150
bio-oil. The inner diameter of the condenser was 45 cm and height of 100 cm with outer
151
jacket and inner coiled jacket for water circulations. The temperature of the cooling water
152
supplied to condenser was between 8 and 10 °C. Bio-oil was collected in a 2 L glass
153
collector attached at the bottom of the condenser. Any incondensable vapors were purged to
154
the atmosphere through a fume hood.
155 156
2.3.Experimental procedure
157
The effect of process parameters such as MW power (900, 1200 and 1500 W) and
158
biomass loading size (0.5 kg and 1 kg) on the yield and quality of product were
159
investigated. CS briquettes were mixed with 75 g of biochar (MW absorber) and loaded in 7
160
the middle of cylindrical MW reactor. It was ensured that biochar is well distributed with
161
CS briquettes. The mixture of briquettes and biochar were placed on a 0.063 cm thick, 34.6
162
cm in diameter stainless steel 309 distributor plate with 240 holes of 3 mm diameter. The
163
reactor was purged with nitrogen gas of 97.5% purity (measured with the help of oxygen
164
analyzer) and flow rate of 30 L/min before commencing the experiment for about 15
165
minutes, whereas during experiment the flow rate was decreased to 20 L/min. The vapors
166
were removed from the opening of the bottom tank head.
167
The MW power was set using Muegge MW controller software provided by the
168
company. Each batch of experiment was conducted for about 2 h. The pyrolysis
169
temperature was monitored by two shielded and grounded type-K thermocouples connected
170
to 8 channel Pico data logger (Pico Technology, Cambridge shire, U. K.) having high
171
resolution and accuracy, and sampling up to 10 measurements per second. The vapors were
172
observed within 2 to 5 min of the MW start. After the MW was turned off, the pyrolysed
173
product was allowed to cool down to room temperature before opening the reactor. Each
174
experiment was repeated three times, and the final results are presented as an average.
175
The MW leakage (3 to 5 mW/m2) was well below the specified standard (10 mW/m2)
176
during the experiment which was monitored using MW leakage detector (Electron
177
Microscopy Science Model 72083-00). In order to avoid any thermal injury due to high
178
temperature at the surface of the SS reactor and to reduce the heat loss from the reactor, a
179
ceramic wool insulation of 5 cm thick was wrapped around the reactor supported by a steel
180
mesh.
181 182 8
183
2.4.Product yield
184
The weight of the bio-oil and bio-char was calculated in percentage based on the ratio
185
of product collected to the amount of original biomass loaded as follow:
186
Bio-oil yield (%) = × 100
187
Bio-char yield (%) =
188
The yield of gas was calculated based on the difference, i.e. 100 – (bio-oil yield (%) +bio-
189
char yield (%)).
(
× 100
190 191
2.5.Product characterization
192
Fourier transform infrared (FTIR) analysis was conducted to identify the chemical
193
functional groups present in bio-oil and biochar. The FTIR spectra were collected between
194
spectral ranges of 4000–400 cm−1 using a Perkin Elmer Spectrum 2000 FTIR Spectrometer.
195
The chemical analysis of bio-oil was performed using a GC/MS based on USEPA 8260
196
method. The elemental components in bio-oil and biochar were investigated using
197
elemental analyzer. The Higher Heating Value (HHV) was measured using oxygen bomb
198
calorimeter.
199 200
3. Results and discussion
201 202
3.1.Challenges in temperature measurement
203
Temperature measurement was one of the major problems in the design of present
204
reactor system. Two thermowells which are 10.5 cm apart from the centre of the reactor as
9
205
depicted in Fig. 2 were used to accommodate thermocouples. Two K-type thermocouples
206
were placed inside the thermowell and connected to the computer via 8 channel Pico data
207
logger. It should be noted that the thermowells are closed from the bottom.
208
Fig. 2 also present the maximum temperature attained at different MW power and
209
biomass loading. Obviously, the temperature was found to increase with MW power, but
210
interestingly it also increased with biomass loading. Thermocouple T2 showed lower
211
temperature as compared to T1. This might be due to the position of T1 thermocouple in
212
close proximity to biomass sample as compared to T2. Overall, the temperature in this
213
study was much lower (in range of 150 to 250 °C) than typical pyrolysis temperature (500 –
214
700 °C). There could be several reasons behind this i) sluggish response of the K-type
215
thermocouple because of its placing inside the thermowell which means thermocouple was
216
not in direct contact with the sample, ii) therefore, thermocouple might be detecting the
217
gaseous or surrounding environment temperature rather than actual sample temperature, iii)
218
samples were almost positioned in the middle as a result of high MW power density at the
219
centre of the reactor which was observed in our preliminary experiment, iv) the
220
thermocouples were positioned at the side of the reactor while the sample at the middle, v)
221
the thermocouple sensor were shielded by two layers, 2 mm thick stainless steel thermowell
222
and about 1 mm thick thermocouple stainless steel shield which might also lower down the
223
response or sensitivity of thermocouples.
224
Several researchers have raised the issue of monitoring the temperature in a MW
225
technology (Zaini and Kamaruddin, 2013; Salema and Ani, 2011; Salema and Ani, 2012a
226
and b). Some have recommended infra-red or optical fiber sensors, but they suffer from
227
detecting the surface temperature only and are expensive. It was also difficult to design and 10
228
place the infra-red or optical fiber sensor in the present reactor design. Nevertheless, the
229
temperature profile obtained by using metallic K-type thermocouple was pretty smooth and
230
almost free from noise (fluctuations). However, repetition of experiments revealed high
231
standard error in the temperature data as shown in Fig. 2, except for 1500 W power and 1
232
kg sample size. It is highly recommended to position the thermowell in contact with the
233
biomass sample to obtain accurate temperature. Further, the electric field usually gets
234
disturb due to loading of biomass sample. Therefore, due to the above limitation in the
235
temperature data, MW power was used to interpret the results in the present study rather
236
than temperature.
237 238
3.2.Pyrolysis product yield
239
Physically, most of the corn stalk biomass briquettes retained their original shape even
240
after pyrolysis. It is assumed that the external surface of CS briquettes might have heated
241
first and subsequently the heat might have transferred to the inner surface. This is possible
242
due to the presence of microwave absorber (in this case biochar) which surrounds the
243
biomass briquettes. Biochar and other carbon materials are usually used as a strong MW
244
absorber which rapidly heats up to a very high temperature in short time. The mechanism of
245
heating biomass in presence of MW absorber might be quite different as compared to
246
without absorber. This is because the heat was reported (Farag et al., 2012;Miura et al.,
247
2004; Undri et al., 2015; Vongpradubchai and Rattanadecho, 2009) to transfer from inner
248
or core of the biomass pellets and wood log to the outer surface. This mechanism of heat
249
transfer cannot be denied in the present briquettes. The porous and fluffy nature of biomass
250
briquettes might also help to penetrate the MW and create the heat at the core. 11
251
Nevertheless, both heating mechanism (outer to core and core to outer) can co-exists at the
252
same time. This needs further detailed studies. Furthermore, preliminary study revealed that
253
corn stalk in their original form and shredded into small size were difficult to pyrolyze and
254
resulted in poor product yield. However, biomass densified into briquettes provided much
255
better results in terms of product yield and pyrolysis behavior.
256
Certainly, the product yield is sensitive to process conditions as evident from Fig. 3, and
257
it is observed to depend on both MW power and biomass loading. Similar results were
258
reported earlier for CS pyrolysis (Pittman, et al., 2012).The yield of bio-oil, biochar and gas
259
ranged from 13.4 to 19.6 wt.%, 30.9 to 41.1wt.%, and 41.6 to 54.0 wt.%, respectively. The
260
bio-oil yield increased by about 22 % when MW power was increased from 900 to 1500 W
261
at 0.5 kg loading and it was highest (19.6 wt.%) at 1200 W and biomass loading of 1 kg.
262
MW power is one of the significant parameter in determining the product yields (Li et al.,
263
2013; Salema and Ani, 2012a and b; Zhuang et al., 2012).Basically, MW power influences
264
the heating rates(Li et al., 2013) and when they increased the power from 750 W to 1500
265
W, the heating rate increased by 112 %, but when they increase the power from 750 W to
266
2250 W, the heating rate increased by almost 900 %. According to them the change in
267
heating rate also depends on the MW power segments. Therefore, the bio-oil yield did not
268
change much in the present study when MW power was increased from 900 to 1500 W,
269
because this range could fall in the medium-high power segment.
270
Certainly, there is also relation between the pyrolysis product yield and the biomass
271
loading. The effect of biomass loading on the pyrolysis product under MW irradiation is
272
still new and very limited knowledge has been developed in the literature. Very recently,
273
effect of biomass loading on the temperature profile, MW electric field and MW absorption 12
274
energy was studied using numerical simulation (Salema and Afzal, 2015).Another study
275
(Robinson et al., 2015)showed significant contribution of sample size on the product yield
276
and quality using experimental work. The general trend of the product yield was the
277
increase in bio-oil and gas yield and decrease in bio-char yield with increase in MW power.
278
Interestingly, the product yield for half and one kilogram biomass loading were almost
279
similar at 1500 W, while it varied for other powers 900 and 1200 W.
280
A decrease in bio-oil yield is always favored with increase in gas yield. This
281
phenomenon particularly happens when the condensable undergoes secondary cracking
282
reactions on its way towards the condenser or if there is lack in rapid condensation (due to
283
longer residence time in the reactor). Thus, most of the condensable vapor might exit the
284
condenser without forming into liquid or bio-oil product. It is quite clear from Table 2 that
285
the product yield also varies with pyrolysis technologies. There could be several reasons
286
behind such contrasts such as:
287
•
Dissimilar reactor configuration and process conditions
288
•
Heating characteristics in MW (volumetric, rapid and selective) is quite different
289
from conventional (conduction and convection) technologies. Typically, MW has
290
higher heating rate as compared to conventional pyrolysis (Mašek et al., 2013).
291
•
Because of above nature, the temperature profile are also quite different in MW
292
•
The CS biomass load used till date is few grams at lab-scale
293
•
The gas yield generally increases with MW power (Huang et al., 2015)
294
•
Mixing MW absorber to biomass sample can also play a role as a catalyst to
295
define the product yield
13
296
•
Non-uniform heating was observed to some extent in the present reactor due to
297
absence of stirring action. This left some CS briquette without pyrolysis (refer to
298
supplementary material).
299
•
The vapors were observed to condense on the wall of reactor as well as on the
300
distributor plate due to difference in temperature along the height of the reactor.
301
This resulted in condensation of high volatile compound inside the reactor which
302
finally affected the bio-oil yield. The problem of bio-oil deposition is attributed to
303
unique heating nature of the MW called as selective heating, in which only the CS
304
biomass is heated to a very high temperature while the surrounding temperature
305
stills remains comparatively low. Therefore, the vapors condense inside the
306
reactor once they experience a lower surrounding temperature. Even though the
307
technique (Abubakar et al., 2013) of introducing nitrogen from the top of the
308
reactor and collecting the vapor from the bottom was implemented to solve the
309
problem of bio-oil deposition, but was less successful in scaled-up reactor (refer to
310
supplementary material).
311
•
When compared with other MW pyrolysis research work as shown in Table 2, the
312
bio-oil yield obtained in present work was lower. This could be due to lower MW
313
power and higher biomass loading in the present study. The present study was
314
limited to 1500 W power due to avoidance of MW leakage, since further increase
315
in MW power resulted in higher MW leakage (> 10 mW/cm2). However, it is
316
anticipated that bio-oil yield can be increased by increasing the MW power
14
317
beyond 2000 W and modification in reactor design to purge out the heavy phase
318
bio-oil.
319 320
•
Lastly, the type of biomass and its physical and chemical characteristics might also contribute in influencing the product yield
321 322
3.3.Product quality
323
As per the visual observation bio-oil was dark red-brown in color and was found to be
324
less viscous may be due to high amount of water and/or low volatile chemical compounds.
325 326
3.3.1. Elemental compositions
327
The elemental composition (CHNOS) of biochar and bio-oil is shown in Table 3.
328
Biochar showed around 63–74 wt.% of carbon, 1.49–2.91 wt.% of hydrogen, 0.1–0.4 wt.%
329
of nitrogen, and 23 – 35 wt.% of oxygen. The C content in the present biochar samples
330
were higher and nitrogen content was lower than the biochar samples obtained from free
331
fall reactor (Shah et al., 2012) and MW pyrolysis (Borges et al., 2014). The O content was
332
much higher in the present biochar but comparable to other study (Shah et al., 2012). The
333
variation in the elemental properties of biochar from various technologies is expected due
334
to types of feedstock, technology, process conditions and pre-treatment if any. There was
335
little effect of MW power on the elemental composition of biochar. However, the sulfur
336
content was found to increase with biomass loading. Biochar can be used as a soil
337
remediation in agricultural applications or as carbon sequestering agent.
338
Elemental compositions of bio-oil in this study showed a remarkable higher oxygen
339
content (around 80 wt.%) than corn stover bio-oil (around 40 wt.%) (Shah et al., 2012). 15
340
However, the values were close to bio-oil produced from corn stover (Borges et al., 2014)
341
(around 78 wt.%) and wood pellets (Undri et al., 2015) (around 77 wt.%) produced from
342
MW pyrolysis. Conversely, the C content was much lower than conventionally produced
343
bio-oil (Shah et al., 2012), but again almost comparable MW pyrolysis produced bio-oil
344
(Borges et al., 2014). Apparently, the C and O content are far lower (85% low) and higher
345
(100% high), respectively than the typical wood based bio-oil (Mohan et al., 2006). The
346
reason could be due to presence of low volatile or light fraction chemicals in the present
347
bio-oil. Higher O content can also be due to presence of water and number of oxygenated
348
chemical compounds found in bio-oil (Mohan et al., 2006). Amazingly, the hydrogen
349
content in the present bio-oil was higher than typical wood based bio-oil (Mohan et al.,
350
2006), corn stalk bio-oil (Pittman et al., 2012), and comparable to corn stover bio-oil
351
(Borges et al., 2014; Undri et al., 2015). In conclusion, bio-oil with the present quality
352
cannot be used in combustion application due to its high O content and lower C content.
353
Further upgrading of bio-oil will not be feasible method due to expensive process and
354
additional unit operations. It is recommended to improve the reactor and/or process design
355
in order to avoid deposition of heavy volatile chemical compounds inside the reactor and
356
maximize their transfer to condenser.
357
3.3.2. Heating value
358
One of the criteria to select a fuel in a power plant is the heating value. Fig. 4 shows
359
the higher heating value (HHV) of biochar and bio-oil. It should be noted that the heating
360
value of biochar represents only of the pyrolysed CS briquettes, since unpyrolysed CS
361
briquettes was easy to separate. The highest HHV of biochar was 31.70 MJ/kg at 900 W
362
power and 0.5 kg biomass load and lowest was 23.00 MJ/kg at 900 W power and 1 kg 16
363
biomass load. The heating value of biochar was found to decrease slightly with increase in
364
MW power and biomass load. Lower value (23.00 MJ/kg) might be due to unpyrolysed CS
365
briquettes. Overall, heating value of biochar between 23 and 32 MJ/kg strongly indicates
366
it’s potential to be used as fuel for bioenergy applications. Using biochar (a carbonaceous
367
material) to assist the pyrolysis of CS in a MW can assist in achieving higher heating value
368
as reported earlier (Undri et al., 2015).
369
Bio-oil showed poor quality in terms of heating value (1.75 to 2.5 MJ/kg), which
370
might be probably due to presence of upper phase bio-oil. One of the main factors that can
371
contribute to such a low heating value is the water content in the bio-oil. Previous study
372
(Undri et al., 2015)also found very low heating values (6 – 8 MJ/kg)of bio-oil produced
373
from wood pellets. The heating value of conventionally (free fall reactor at 400 °C)
374
produced bio-oil was also found to be lower (around 7 MJ/kg) produced for untreated CS
375
(Pittman et al., 2012).Obviously, the heating value of bio-oil is very much dependent on its
376
elemental composition, chemical compounds, and water content. Undoubtedly, based on
377
the present results of the heating value, bio-oil is not a good candidate for combustion and
378
energy application. The challenge of low heating value can be overcome by condensing
379
heavy fraction into bio-oil.
380
Overall, the specific energy to pyrolyse the CS briquette in the developed MW
381
reactor according to the following equation was about 6.5 MJ/kg for 900 W to 10 MJ/kg for
382
1500 W. =
×
17
383
Where, is the specific energy in J/kg, is the input MW power in J/s, is the time of
384
the sample exposed to the MW in s, and
is the mass of sample processed in kg.
385
3.3.3. FTIR analysis
386
The possible chemical functional groups and its compounds present in the bio-oil
387
obtained from MW pyrolysis of corn stalk briquettes (spectra are provided in the
388
supplementary material). In general, the IR spectra for 0.5 and 1 kg bio-oil at all MW
389
power had similar absorption bands and similar relative intensities. The broad and intense
390
peaks between 3200 and 3600 cm-1 wavelength could be possible due to presence of
391
alcohol or phenolic OH stretch. However, the peak in the region of 3400 cm-1 could also be
392
possibly due to high content of water vapor or H-bonded OH stretching. This indicates that
393
bio-oil contain high fraction of water. A weak broad peak at 2000 cm-1 wavenumber could
394
be because of aromatic substitutions such as C-H or C=C stretch. Simple monosubstituted
395
absorbs due to alkenes C=C stretch can be observed at 1630 cm-1. A sharp peak between
396
1685 and 1725 cm-1 wavenumber could be due to C=O stretch ketones with phenyl group
397
or phenolic compounds. The peak at 1380 cm-1 might be due to CH3 bending absorption of
398
alkanes. Two sharp and medium peaks at 1265 and 1380 cm-1 could be attributed to methyl
399
CH3 bending and primary or secondary alcohol OH in-plane bending, respectively. Lastly,
400
the broad and strong peak between 900 and 400 cm-1 can be attributed to alkyl halide either
401
to C-Cl or C-Br or C-I stretch.
402
There is a clear difference in IR spectra of bio-oil and biochar absorption bands as
403
well as intensities (refer to the supplementary material). The broad and weak peaks between
404
3200 and 3600 cm-1 wavelength could be possible due to presence of alcohol or phenol OH
18
405
stretch or because of moisture content. Two strong peaks between 1600 and 1350 cm-1
406
could be due to presence of nitro compounds in the char such as NO2 stretch or N-H
407
bending from amine group. The broad absorption between 1200 and 1000 cm-1 for 1 kg
408
biomass load could be because of C-O stretch and this band was very weak in case of 0.5
409
kg biomass loading. Few weak peaks around 850 cm-1 may be due to aromatic C-H
410
deformation. These peaks almost disappeared or became very weak in the spectra such of
411
1200W-0.5kg and 900W-1 kg.
412
3.3.4. Chemical analysis of liquid (bio-oil) product
413
Chemically, bio-oil is a mixture of hundreds of chemical compounds such as phenol,
414
ketones, aldehydes, furfural, acetic acid, guaiacols, catecols, syringols, formic acid,
415
alcohols, esters, carboxylic acids and even water. The formation of chemical compounds
416
depends on the type of biomass used in pyrolysis process, its chemical and lignocellulosic
417
structure, pyrolysis process conditions, and condensing parameters. The identification of
418
chemicals in the present bio-oil was limited to certain compounds because the bio-oil
419
obtained was of lower quality and mostly contained water components. Therefore, it was
420
difficult to analyse and fractionate the bio-oil compounds and only some specific chemical
421
compounds are presented in Table 4. Another reason for restricting the type of chemical
422
compound was due to the limitation of GC-MS column and machine specifications.
423
The concentration of chemical compounds in the bio-oil was found to depend on both
424
the MW power and biomass loading. The main phenolic compounds (phenol, guaiacol (2-
425
methoxy phenol), Xylenol (2,6-dimethyl phenol and 2,4-dimethyl phenol)) are mainly the
426
products from lignin pyrolysis. Acetone that basically belongs to ketone group is expected
427
to occur in the aqueous phase of bio-oil and is produced as the by-product in the biomass 19
428
pyrolysis reaction. The concentration of 1,2-Dichloroethane-d4 also commonly known as
429
ethylene dichloride and belonging to organic chemical in the family of alkanes and
430
halocarbons was almost consistent with MW power and biomass loading. Lastly, the
431
concentration of fatty acids (palmitic, stearic, and oleic) varied with both power and
432
loading. These fatty acids are basically found in extractives present in the biomass.
433
Remarkably, the quantity for most of the chemical compounds in Table 4 was independent
434
of the amount of biomass loaded in the rector. This signifies that the quantity as well as the
435
quality of certain chemicals in the bio-oil is less dependent on the biomass loading
436
compared to pyrolysis process conditions (MW power, temperature, condensing
437
temperature and system) and reactor design.
438
Typically, phenol and its derivatives form large fraction in bio-oil of the CS (Lv and
439
Wu, 2012) followed by acetic acid, furfural, ketones, furans and aldehydes. The formation
440
of these chemical compounds is very much dependent on the lignocellulosic content of the
441
biomass. The average concentration of phenols and its compounds was around 60 wt.% as
442
shown in Table 4.This was in agreement with previous studies(Lv et al., 2013; Pittman et
443
al., 2012) who also reported phenol to be the principal component from the lignin pyrolysis.
444
As can be seen from Table 4, the quantity of phenolic compounds produced from CS may
445
not necessarily depend on the biomass loading while keeping the MW power constant. The
446
total phenol content decreased when biomass loading was increased from 0.5 kg to 1 kg.
447
For instance, it was reported (Lv et al., 2013) that increase in temperature favors the
448
formation of specific type of phenolic compound. Moreover, formation of phenolic
449
compounds usually takes place at higher pyrolysis temperature (Mohan et al., 2006). Very
450
limited work has been carried out in the literature about the scale-up of biomass processing 20
451
in MW technology. Thus, additional research work is required to ascertain the effect of
452
scaling-up the MW technology on the quality and quantity of chemical compounds
453
produced from biomass pyrolysis.
454 455
4. Conclusions
456
To our knowledge the pyrolysis of biomass in kilogram (scaled-up reactor) and in form
457
of briquette in MW technology is reported for the first time. The quantity (yield) of biochar,
458
bio-oil, and gas greatly depended on the process condition (MW power and biomass
459
loading). Based on its heating value and elemental composition, biochar is suitable for
460
energy applications, whereas bio-oil produced from CS in this study is neither good for
461
energy application nor for chemical production. Further research work is needed in terms of
462
reactor design and biomass loading to prove the feasibility of system at large-scale
463
production.
464 465
Acknowledgements
466 467
The authors are grateful to the New Brunswick Department of Agriculture, Aquaculture
468
and Fisheries (NBDAAF); and Natural Sciences and Engineering Research Council of
469
Canada (NSERC) for the financial support to this project. First author pursued his
470
postdoctoral studies at the University of New Brunswick, Canada.
471 472 473 21
474
Supplementary materials
475
Fig.A1. FTIR analysis of bio-oil (A and B) and bio-char (C and D) obtained from MW
476
pyrolysis of corn stalk and Table A1 Photos of CS briquettes before and after MW
477
pyrolysis are available on the journal paper website.
478 479 480
References 1. Abubakar, Z., Salema, A.A., and Ani, F.N., 2013. A new technique to pyrolyse
481
biomass in a microwave system: effect of stirrer speed. Bioresour. Technol.128,
482
578-585.
483
2. Bailey S.S., Samson, R., and Lem, C.H., 2006. Biomass Resource Options:
484
Creating a BIOHEAT Supply for the Canadian Greenhouse Industry. Natural
485
Resources Canada.
486
3. Borges, F.C., Du, Z., Xie, Q., Trierweiler, J.O., Cheng, Y., Wan, Y.Q., Liu, Y.H.,
487
Zhu, R.B., Lin, X.Y., Chen, P., and Ruan, R., 2014. Fast microwave assisted
488
pyrolysis of biomass using microwave absorbent. Bioresour. Technol.156, 267-274.
489
4. Cordella, M., Berrueco, C., Santarelli, F., Paterson, N., Kandiyoti, R., Millan, M.,
490
2013. Yields and ageing of the liquids obtained by slow pyrolysis of sorghum,
491
switchgrass and corn stalks. J. Anal. Appl. Pyrol. 104, 316-324.
492
5. Farag, S., Sobhy, A., Akyel, C., Doucet, J., Chaouki, J., 2012. Temperature profile
493
prediction within selected materials heated by microwaves at 2.45GHz. Appl.
494
Therm. Eng. 36, 360-369.
22
495
6. Huang, Y.F., Kuan, W.H., Chang, C.C., Tzou, Y.M., 2013. Catalytic and
496
atmospheric effects on microwave pyrolysis of corn stover. Bioresour. Technol.
497
131, 274-280.
498
7. Li, H., Qu, Y., Yang, Y., Chang, S., and Xu, J., 2016. Microwave irradiation–A
499
green and efficient way to pretreat biomass. Bioresour. Technol.199, 34-41.
500 501 502
8. Li, L., Ma, X., Xu, Q., Hu, Z., 2013. Influence of microwave power, metal oxides and metal salts on the pyrolysis of algae. Bioresour. Technol. 142, 469-474. 9. Liu, X., Zhang, Y., Li, Z., Feng, R., and Zhang, Y., 2014. Characterization of
503
corncob-derived biochar and pyrolysis kinetics in comparison with corn stalk and
504
sawdust. Bioresour. Technol.170, 76-82.
505
10. Liu, S., Xie, Q., Zhang, B., Cheng, Y., Liu, Y., Chen, P., and Ruan, R. 2016. Fast
506
microwave-assisted catalytic co-pyrolysis of corn stover and scum for bio-oil
507
production with CaO and HZSM-5 as the catalyst. Bioresour. Technol. 204, 164-
508
170.
509 510 511
11. Lv, G., Wu, S., 2012. Analytical pyrolysis studies of corn stalk and its three main components by TG-MS and Py-GC/MS. J. Anal. Appl. Pyrol. 97, 11-18. 12. Lv, G., Wu, S., Yang, G., Chen, J., Liu, Y., Kong, F., 2013. Comparative study of
512
pyrolysis behaviors of corn stalk and its three components. J. Anal. Appl. Pyrol.
513
104, 185-193.
514 515
13. Macquarrie, D. J., Clark, J. H., and Fitzpatrick, E., 2012. The microwave pyrolysis of biomass. Biofuels Bioprod. Biorefining6(5), 549-560.
23
516
14. Mašek, O., Budarin, V., Gronnow, M., Crombie, K., Brownsort, P., Fitzpatrick, E.,
517
and Hurst, P., 2013. Microwave and slow pyrolysis biochar—Comparison of
518
physical and functional properties. J. Anal. Appl. Pyrol. 100, 41-48.
519 520 521 522 523 524
15. Miura, M., Kaga, H., Sakurai, A., Kakuchi, T., Takahashi, K. 2004. Rapid pyrolysis of wood block by microwave heating. J. Anal. Appl. Pyrol. 71(1), 187-199. 16. Mohan, D., Pittman, C.U., Steele, P.H., 2006. Pyrolysis of Wood/Biomass for Biooil: A Critical Review. Energy Fuels, 20(3), 848-889. 17. Motasemi, F., and Afzal, M.T., 2013. A review on the microwave-assisted pyrolysis technique. Renew. Sustain. Energy Rev. 28, 317-330.
525
18. Pianroj, Y., Jumrat, S., Werapun, W., Karrila, S., and Tongurai, C., 2016. Scaled-up
526
reactor for microwave induced pyrolysis of oil palm shell. Chem. Eng. Process.:
527
Process Intensification106, 42-49.
528
19. Pittman, C.U., Mohan, D., Eseyin, A., Li, Q., Ingram, L., Hassan, E.B.M., Mitchell,
529
B., Guo, H., Steele, P.H., 2012. Characterization of bio-oils produced from fast
530
pyrolysis of corn stalks in an auger reactor. Energy Fuels 26(6), 3816-3825.
531
20. Robinson, J., Dodds, C., Stavrinides, A., Kingman, S., Katrib, J., Wu, Z., Medrano,
532
J., Overend, R., 2015. Microwave pyrolysis of biomass: control of process
533
parameters for high pyrolysis oil yields and enhanced oil quality. Energy Fuels
534
29(3), 1701-1709.
535 536
21. Robinson, J.P., Kingman, S.W., Barranco, R., Snape, C.E., and Al-Sayegh, H., 2009. Microwave pyrolysis of wood pellets. Ind. Eng. Chem. Res.49 (2), 459-463.
24
537
22. Salema, A.A., Afzal, M.T., 2015. Numerical simulation of heating behaviour in
538
biomass bed and pellets under multimode microwave system. Int. J. Therm. Sci. 91,
539
12-24.
540 541 542 543 544
23. Salema, A.A., and Ani, F.N, 2012a. Microwave-assisted pyrolysis of oil palm shell biomass using an overhead stirrer. J. Anal. Appl. Pyrol.96, 162-172. 24. Salema, A.A., and Ani, F.N., 2011. Microwave induced pyrolysis of oil palm biomass. Bioresour. Technol. 102(3), 3388-3395. 25. Salema, A.A., Ani, F.N., 2012b. Pyrolysis of oil palm empty fruit bunch biomass
545
pellets using multimode microwave irradiation. Bioresour. Technol. 125, 102-107.
546
26. Shah, A., Darr, M.J., Dalluge, D., Medic, D., Webster, K., and Brown, R.C., 2012.
547
Physicochemical properties of bio-oil and biochar produced by fast pyrolysis of
548
stored single-pass corn stover and cobs. Bioresour. Technol. 125, 348-352.
549
27. Undri, A., Abou-Zaid, M., Briens, C., Berruti, F., Rosi, L., Bartoli, M., Frediani,
550
M., Frediani, P., 2015. Bio-oil from pyrolysis of wood pellets using a microwave
551
multimode oven and different microwave absorbers. Fuel 153, 464-482.
552 553 554
28. Uzun, B.B., Sarioğlu, N., 2009. Rapid and catalytic pyrolysis of corn stalks. Fuel Process. Technol. 90(5), 705-716. 29. Vongpradubchai, S., Rattanadecho, P., 2009. The microwave processing of wood
555
using a continuous microwave belt drier. Chem. Eng. Process.: Process
556
Intensification 48(5), 997-1003.
557
30. Wang, T., Ye, X., Yin, J., Lu, Q., Zheng, Z., and Dong, C., 2014. Effects of
558
biopretreatment on pyrolysis behaviors of corn stalk by methanogen. Bioresour.
559
Technol. 164, 416-419. 25
560 561 562 563
31. Yin, C., 2012. Microwave-assisted pyrolysis of biomass for liquid biofuels production. Bioresour. Technol.120, 273-284. 32. Zaini, M.A.A., and Kamaruddin, M.J., 2013. Critical issues in microwave-assisted activated carbon preparation. J. Anal. Appl. Pyrol. 101, 238-241.
564
33. Zhang, B., Zhong, Z., Ding, K., Cao, Y., Liu, Z., 2014. Catalytic upgrading of corn
565
stalk fast pyrolysis vapors with fresh and hydrothermally treated HZSM-5 catalysts
566
using Py-GC/MS. Ind. Eng. Chem. Res. 53(24), 9979-9984.
567
34. Zhao, X., Zhang, J., Song, Z., Liu, H., Li, L., and Ma, C., 2010. Microwave
568
pyrolysis of straw bale and energy balance analysis. J. Anal. Appl. Pyrol. 92(1), 43-
569
49.
570
35. Zhuang, Y., Guo, J., Chen, L., Li, D., Liu, J., Ye, N., 2012. Microwave-assisted
571
direct liquefaction of Ulva prolifera for bio-oil production by acid catalysis.
572
Bioresour. Technol.116, 133-139.
573 574 575
26
576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608
Fig. 1. Schematic diagram of a MW pyrolysis system developed at UNB, Canada
609 610 611 612
27
613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630
Fig. 2. Maximum temperature attained during MW pyrolysis of CS briquette biomass and the
631
schematic diagram of thermocouple design
632 633 634 635 28
636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680
Fig. 3. Product yield obtained from MW pyrolysis of CS briquettes
29
681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702
Fig. 4. Higher heating value (HHV) of (A) biochar and (B) bio-oil
703 30
727 728
Table 1 Pyrolysis of corn stalk using conventional and microwave technology (2009-2016) Reference Technology Method Product Wang et al., 2014 Pyroprobe Methanogen pretreatment Chemicals connected to of corn stalk GC/MS Liu et al., 2014 Lab-scale fixed bed The feedstock were air Biochar reactor was used dried and stored in sealed for slow pyrolysis containers Zhang et al., 2014 Py-GC/MS Corn stalk was driedat 105 Chemicals in bio°C for 24 handgrounded to oil and char 40 mesh size. Catalytic pyrolysis using a HZSM-5 catalyst. Cordella et al., Laboratory-scale Corn stalk samples were Bio-oil and 2013 fixed bed reactor dried overnight at 60 °C biochar and ground to particle sizeless than 1 mm.2–3 g of biomass sample per run was used. Lv et al., 2013 Tubular furnace Corn stalk without leaves Chemicals in bioand Py-GC/MS was groundedto a size of oil and biochar 40–60 mesh (0.28–0.45 mm). Pre-treatment was also used in this study. Corn stalk was also fractionated into cellulose, hemicellulose and lignin.About 350 mg of each samplewere loaded. Pittman, Jr. et al., Auger reactor with Untreated and acid-treated Chemicals in bio2012 feed rate of 1−2.5 corn stalks were used. The oil kg/h and stalks were crushedto a temperature of 400 particlediameter of 0.5−5 to 450 °C. mm and densified under a pressure of 550 psi and166 °C for 4 min. Zhao et al., 2010 Microwave Corn stalk bale was Gas pyrolysis with total weighed and loaded in the microwave input microwave. The size of power of 18 kW corn stalk bale was about 1000mm×600mm×600mm. Uzun and High-speed heated Corn stalks wereair-dried Chemicals in bioSarioğlu, 2009 fixed-bed and ground with average oil
32
tubularreactor
size in range of 0.85
729 730 731 732 733 734 735
33
736 737 738
Table 2 Product yield from CS biomass using different pyrolysis technologies
Technologies
Process conditions
MW pyrolysis
MW power – 900 to 1500 W, radiation time – 2 h, biomass loading – 0.5 and 1 kg, temperature* – 150 to 250 °C Corn stover, Temp - 500 °C, biomass particle size - 1 mm, vacuum degree - 170 mmHg Wood pellets, sample load 211.7 g, MW power - 1.2 to 3.0 kW, Carbon absorber 75 g Corn stover co-pyrolysis, Temp – 450 °C, SiC as absorber, CaO/HZSM-5 catalyst, biomass load – 30 g Temp – 800 °C, residence time – 1 to 2 min Feed rate – 1 to 1.25 kg/h, temp – 400 to 450 °C, residence time – 50 s, pretreatment in 2 wt % aqueous H2SO4 solutions Biomass load – 5 g, residence time – 10 min, temp – 600 °C, used zeolite catalyst Biomass load – 2 to 3 g, temp – 650 °C
MW pyrolysis
MW pyrolysis
MW pyrolysis
Fixed bed tubular furnace Auger reactor
Fixed-bed tubular reactor
739 740 741
Product yield BioBio-oil, Gas, char, wt.% wt.% wt.%
References
17.0
35.0
48.0
This study
64
22
14
Borges et al., 2014
46.1
24.9
29.0
Undri et al., 2015
17.4
71.1
11.5
Liu et al., 2016
30.0
25.0
45.0
(Lv et al., 2013)
35.0
29.0
13.5
(Pittman et al., 2012)
46.0 (with water)
20.0
34.0
(Uzun and Sarioğlu, 2009)
Laboratory(Cordella et scale fixed bed 50.0 30.0 20.0 al., 2013) reactor *Note: Please see section 3.1., for the temperature accuracy. However, actual temperature might be 3 times higher than measured.
742
34
743 744 745 746 747 748 749 750
Table 3 Elemental composition of biochar and bio-oil produced from MW pyrolysis of corn stalk briquette MW power, W Load, kg
900 1200 1500
900 1200 1500
0.5 1 0.5 1 0.5 1 0.5 1 0.5 1 0.5 1
C H N S O wt% wt% wt% wt% wt% Biochar 74.33 1.86 0.35 0.0035 23.45 67.55 2.91 0.40 0.0037 29.13 63.05 1.49 0.14 0.0085 35.31 67.49 2.69 0.36 0.0143 29.43 71.61 1.89 0.35 0.0190 26.13 66.05 2.66 0.48 0.0625 30.75 Bio-oil 7.96 8.80 0.04 0.0016 83.19 9.12 8.46 0.04 0.0018 82.38 7.85 10.43 0.05 0.0017 81.65 9.09 9.05 0.11 0.0030 81.74 8.73 9.79 0.10 0.0019 81.37 8.75 10.13 0.07 0.0023 81.04
751 752 753 754
35
766 767 768 769 770 771
Table 4 GC/MS analysis of bio-oil at different experimental conditions
0.5kg 1kg 900W 1200W 1500W 900W 1200W 1500W Phenol, µg/ml 220 140 190 150 140 170 2,6-dimethyl phenol, µg/ml 2.2 2.0 2.2 2.0 1.8 1.7 2,4-dimethyl phenol, µg/ml 9.5 8.4 11 9.9 10 10 2-methoxy phenol, µg/ml 87 97 78 120 100 100 Acetone, mg/kg 46 170 240 370 360 360 1,2-Dichloroethane-d4, % 96 105 106 108 101 106 C16:0 (Palmitic), mg/kg 24 78 38 27 37 31 C16:1n7 (Palmitoleic), mg/kg < 10 < 10 < 10 < 10 < 10 < 10 C18:0 (Stearic), mg/kg 13 38 25 17 20 24 C18:1n9 (Oleic), mg/kg 17 56 23 24 24 38 C18:2n6 (Linoleic), mg/kg < 10 26 < 10 < 10 11 15 C20:0 (Arachidic), mg/kg < 10 < 10 < 10 < 10 < 10 < 10 C18:3n3 (ALA), mg/kg < 10 < 10 < 10 < 10 < 10 < 10
Chemical names
772 773 774 775 776 777 778
37
791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821
Graphical abstract
39
822 823 824
Highlights •
First time microwave (MW) pyrolysis of biomass briquette was carried out
825
•
Biomass loading was scaled up to kilograms
826
•
HHV of biochar and bio-oil was 32 MJ/kg and 2.5 MJ/kg, respectively
827
•
Pyrolysis product yield dependent on the process parameters, MW power and
828 829
loading •
Reactor design can be further improved to increase the bio-oil quality
830 831
40
S0960-8524(17)30242-0 http://dx.doi.org/10.1016/j.biortech.2017.02.113 BITE 17685
To appear in:
Bioresource Technology
Received Date: Revised Date: Accepted Date:
19 December 2016 23 February 2017 24 February 2017
Please cite this article as: Salema, A.A., Afzal, M.T., Bennamoun, L., Pyrolysis of corn stalk biomass briquettes in a microwave technology, Bioresource Technology (2017), doi: http://dx.doi.org/10.1016/j.biortech.2017.02.113
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.
Pyrolysis of corn stalk biomass briquettes in a microwave technology
1 2
Arshad Adam Salemaa, Muhammad T. Afzalb*, and Lyes Bennamounb
3 4 a
5 6
Discipline of Mechanical Engineering, School of Engineering, Monash University Malaysia, Jalan Lagoon Selatan, 46150 Bandar Sunway, Selangor, Malaysia.
b
7
Department of Mechanical Engineering, Faculty of Engineering, University of New
8
Brunswick, Fredericton, NB E3B 5A3, Canada.
9 10
Abstract
11
Pyrolysis of corn stalk biomass briquettes was carried out in a developed microwave (MW)
12
reactor supplied with 2.45 GHz frequency using 3kW power generator. MW power and
13
biomass loading were the key parameters investigated in this study. Highest bio-oil,
14
biochar, and gas yield of 19.6 %, 41.1%, and 54.0% was achieved at different process
15
condition. In terms of quality, biochar exhibited good heating value (32 MJ/kg) than bio-oil
16
(2.47 MJ/kg). Bio-oil was also characterised chemically using FTIR and GC-MS method.
17
This work may open new dimension towards development of large-scale MW pyrolysis
18
technology.
19 20
Keywords: Biomass; Briquette; Microwave; Pyrolysis; Product; Yield and quality
21
*
Corresponding author: [email protected]; Tel: +1-506-453-4880; Fax: +1-506-453-5025
1
22 23
1. Introduction
24
Corn is one of the major agricultural crops in the United States and Canada and in
25
several other parts of the world. This agricultural business generates huge amount of corn
26
residues every year in form of corn stalk, corn cobs, and corn stover. For instance, Ontario
27
and Quebec provinces produced about 5.1 and 3.1 million tonnes of corn stalk, respectively
28
between year 2001 and 2005(Bailey, 2006).Presently, corn stalk is either used as soil
29
amendment in the farm itself or utilized as fuel for the production of energy. However,
30
there still remains excess biomass which needs to be disposed of safely. Despite being a
31
waste, one of the potential utilizations of corn biomass materials is the production of fuels
32
and chemicals (Huang et al., 2013). Environmental regulations and serious climatic issues
33
have become major contributing factors for the use of biomass materials as an alternative
34
source of energy and fuels.
35
Despite detailed research work on Microwave Assisted Pyrolysis (MAP) technology
36
(Motasemi and Afzal, 2013; Yin, 2012; Macquarrie et al., 2012), there still remains a clear
37
lack in scale-up of MW technology in processing biomass materials. There are some
38
attempts in scale-up of MW reactor for biomass processing, but no results are yet reported
39
in the open literature which might be due to reason of patent protection or commercial
40
development. Thus, development of large scale MW reactor will surely help in removing
41
the obstacle for industrial applications (Li et al., 2016). Very recently (Pianroj et al., 2016),
42
pyrolysis of oil palm shell was carried out in a scaled up reactor. However, their quartz
43
glass reactor size (10.7 cm in diameter and 14.0 cm in height) was almost similar to that of
44
previous work of Salema and Ani, 2011. Even though, Pianroj et al., 2016 claimed about 2
45
scale-up of MW system, but they processed about 400 g of biomass. There are some
46
exception studies where large amount of biomass was processed (Zhao et al., 2010; Miura
47
et al., 2004), but are limited. Recently, Robinson et al., 2015 reported an interesting result
48
about the effect of biomass sample size on the product yield and quality using MW
49
technology. However, their sample size was in the range of 5 to 20 g only. In most of the
50
research conducted under MW technology, biomass weight was limited in size of grams.
51
Therefore, very little data (Robinson et al., 2015; Salema and Afzal, 2015) is available in
52
the open literature and limited work has been carried out on the effect of biomass sample
53
size. The present study aims to demonstrate the scale-up of the MW reactor system in terms
54
of amount of biomass processed i.e. kilograms.
55
Although several researchers have processed various types of biomass materials and
56
in different forms loose, shredded, grinded (Salema and Ani, 2012a and 2012b; Abubakar
57
et al., 2013), pellets (Robinson et al., 2009; Undri et al., 2015), whole wood log (Miura et
58
al., 2004) and bale (Zhao et al., 2010), but none have conducted MW pyrolysis of biomass
59
briquettes. Certainly, densification (pellet and briquette) increases the density of biomass,
60
thus reducing the volume of biomass which can save the transport cost. Moreover, it can
61
also significantly reduce the reactor volume or size and thereby reduction in the cost.
62
However, the effect of densification on factors such as heat and mass transfer needs to be
63
investigated.
64
Table 1 presents the pyrolysis technologies used to pyrolyse corn stalk biomass. It is
65
clear that most of researchers used conventional technologies to pyrolyze corn stalk
66
biomass, except one (Zhao et al., 2010) and their focus was on the pyrolysis behavior,
67
temperature and weight profile, and little on product yield and no evidence on the product 3
68
quality. Comparatively, MW technology has provided better product quality due to some
69
unique heating mechanism. One of the most important is the direct interaction of the
70
material with the MW. These and other non-thermal effects of MW heating have
71
revolutionized the processing of the biomass under MW. However, biomass is poor
72
absorber of the MW due to low loss dielectric properties. In order to increase the heating
73
rate as well as to induce much faster pyrolysis reactions they are usually mixed with
74
carbonaceous materials.
75
From our past numerical simulation study (Salema and Afzal, 2015), MW power
76
and biomass loading was considered as the key process parameters of the biomass
77
pyrolysis. But how these parameters would influence the product yield and quality in real
78
application? is still a question that needs attention. The objectives of this study was to
79
reveal the development and the potential of rector design, and to pyrolyze the biomass
80
briquettes in a developed MW reactor which was done for the first time, and to study the
81
effects of MW power and biomass loading (size in kilograms) on the product yield and
82
quality. Biochar and bio-oil were subjected to elemental analysis, heating value, and FT-IR
83
in addition to GC-MS analysis of bio-oil.
84 85
2. Materials and methods
86 87
2.1.Materials
88
Corn stalk (CS) biomass was obtained from the neighboring corn farm situated in the
89
Mactaquac region, Fredericton, New Brunswick, Canada. The corn stalk biomass samples
90
were collected from freshly harvested corn crop in the month of October 2015. The samples 4
91
were chopped, shredded, dried in an oven for 6 hours and stored at room condition. A
92
hydraulic briquetting machine with a constant pressure of 100 MPa and holding time of 5 s
93
was used to produce CS briquettes with a size of 0.04 m diameter and 0.02 m length. This
94
study does not take into account the effect of densification parameters. Each CS briquette
95
weighted around 30 g. The CS briquettes were stored in a sealed container to minimize the
96
moisture activity. The proximate and ultimate analysis of CS biomass showed moisture
97
content (6.8 wt.%), volatile matter (76.2 wt.%), fixed carbon (17 wt.%), carbon (46.67
98
wt.%), hydrogen (6.01 wt.%), nitrogen (0.02 wt.%), oxygen (47.28 wt.%), and sulphur
99
(0.02 wt.%). The biochar obtained from the MW pyrolysis of CS briquettes were used as
100
MW absorbers in the present study.
101 102
2.2.Development and assembly of MW pyrolysis system
103
The microwave assisted pyrolysis (MAP) tests were carried out in a custom-built
104
cylindrical stainless steel 309 reactor connected to a 3 kW and 2450 MHz frequency
105
microwave generator system. The MW generator was procured from Muegge GmbH,
106
Reichelsheim, Germany which consisted of MW power supply (Model MX3000DL-
107
152KL, Muegge GmbH), magnetron head (Model MH3000S-250BB, Muegge GmbH)
108
embedded with isolator and MW directional coupler. A complete MW pyrolysis system
109
consisted of MW generator, waveguide, 3 stub-tuner, a stainless steel 309 cylindrical
110
reactor, a nitrogen gas generator, two K-type metallic thermocouples, Pico data logging
111
system, a pilot scale glass condenser and a bio-oil collector as shown in Fig.1.
112 113
The MW power supply and magnetron head was delivered with a water (temperature between 18 and 20 °C) flowing continuously with a flow rate of 4.5 liter per 5
114
minute (L/min) in order to facilitate the cooling process. As a safety feature, the MW
115
generator system will shut OFF automatically if the water flow rate drops below 4 L/min.
116
The same water was also used for isolator to absorb any reflected MW. The magnetron
117
generates the MW that travel through the waveguide and finally to a cylindrical reactor.
118
Any reflected MW gets diverted into a dummy water load (isolator) to avoid damage to the
119
magnetron. A MW directional coupler also embedded in the magnetron head allows to
120
measure any reflected power. The reflected power was much dependent on the applied MW
121
power and it increased with increase in MW power. For 900, 1200, and 1500 W it was
122
about 0%, 2%, and 5 %, respectively. The three stub tuner helps to match the impedance of
123
the waveguide segments to the load. Rectangular waveguide (WR 340) of 13.8 cm × 9.5 cm
124
also acquired from Muegge, Germany was used in this study.
125
A cylindrical stainless steel 309 reactor of 35.5 cm inner diameter and 53.34 cm in
126
height was designed and fabricated to carry out the pyrolysis process. The size of the
127
reactor was determined based on the amount of densified biomass to be processed, say 7 to
128
8 kg. This cylindrical reactor was covered from top and bottom with SS 309 tank dish head.
129
The top tank head was facilitated with 5 openings; two openings were attached with a 60
130
cm long, 6 mm inner diameter and 2 mm thick thermowell and were closed at the bottom
131
for the insertion of K-type thermocouples. The position and the distance of thermocouples
132
are presented in section 3.1. The centered opening was for the entrance of nitrogen gas
133
inside the cylindrical reactor. Any other openings were sealed and closed to avoid any exit
134
of the vapors or MW leakage. The nitrogen gas was supplied by a nitrogen gas generator
135
(VWR Inc., Ontario, Canada, Model CA26000-014). The bottom tank head had a single
136
opening for the exit of the pyrolysis vapor into the condenser. Both the top and bottom tank 6
137
head were sealed and closed tightly with the help of high temperature gaskets and stainless
138
steel clamps. To avoid the pyrolysis vapor to travel through the waveguide and get
139
deposited into the magnetron, mica sheet of 2 mm thick was placed at the waveguide which
140
opened into the SS reactor. The mica sheet is transparent to microwave, but at the same
141
time it restrict any vapor to pass into the waveguide. The height at which the waveguide
142
opened into the reactor was adequate to distribute the MW into the reactor and interact with
143
the biomass sample.
144
During the experiment, the reflected power from the microwave was monitored on
145
the Muegge software, and the temperature readings from the thermocouples were measured
146
and continuously recorded in a real time. The experiments were run under atmospheric
147
pressure. The vapors generated from pyrolysis process were continuously swept out of the
148
reactor with the help of nitrogen gas. The vapors passed through a pilot scale glass
149
condenser (VWR Inc., Ontario, Model no. 6016-141) which condensed into liquid called as
150
bio-oil. The inner diameter of the condenser was 45 cm and height of 100 cm with outer
151
jacket and inner coiled jacket for water circulations. The temperature of the cooling water
152
supplied to condenser was between 8 and 10 °C. Bio-oil was collected in a 2 L glass
153
collector attached at the bottom of the condenser. Any incondensable vapors were purged to
154
the atmosphere through a fume hood.
155 156
2.3.Experimental procedure
157
The effect of process parameters such as MW power (900, 1200 and 1500 W) and
158
biomass loading size (0.5 kg and 1 kg) on the yield and quality of product were
159
investigated. CS briquettes were mixed with 75 g of biochar (MW absorber) and loaded in 7
160
the middle of cylindrical MW reactor. It was ensured that biochar is well distributed with
161
CS briquettes. The mixture of briquettes and biochar were placed on a 0.063 cm thick, 34.6
162
cm in diameter stainless steel 309 distributor plate with 240 holes of 3 mm diameter. The
163
reactor was purged with nitrogen gas of 97.5% purity (measured with the help of oxygen
164
analyzer) and flow rate of 30 L/min before commencing the experiment for about 15
165
minutes, whereas during experiment the flow rate was decreased to 20 L/min. The vapors
166
were removed from the opening of the bottom tank head.
167
The MW power was set using Muegge MW controller software provided by the
168
company. Each batch of experiment was conducted for about 2 h. The pyrolysis
169
temperature was monitored by two shielded and grounded type-K thermocouples connected
170
to 8 channel Pico data logger (Pico Technology, Cambridge shire, U. K.) having high
171
resolution and accuracy, and sampling up to 10 measurements per second. The vapors were
172
observed within 2 to 5 min of the MW start. After the MW was turned off, the pyrolysed
173
product was allowed to cool down to room temperature before opening the reactor. Each
174
experiment was repeated three times, and the final results are presented as an average.
175
The MW leakage (3 to 5 mW/m2) was well below the specified standard (10 mW/m2)
176
during the experiment which was monitored using MW leakage detector (Electron
177
Microscopy Science Model 72083-00). In order to avoid any thermal injury due to high
178
temperature at the surface of the SS reactor and to reduce the heat loss from the reactor, a
179
ceramic wool insulation of 5 cm thick was wrapped around the reactor supported by a steel
180
mesh.
181 182 8
183
2.4.Product yield
184
The weight of the bio-oil and bio-char was calculated in percentage based on the ratio
185
of product collected to the amount of original biomass loaded as follow:
186
Bio-oil yield (%) = × 100
187
Bio-char yield (%) =
188
The yield of gas was calculated based on the difference, i.e. 100 – (bio-oil yield (%) +bio-
189
char yield (%)).
(
× 100
190 191
2.5.Product characterization
192
Fourier transform infrared (FTIR) analysis was conducted to identify the chemical
193
functional groups present in bio-oil and biochar. The FTIR spectra were collected between
194
spectral ranges of 4000–400 cm−1 using a Perkin Elmer Spectrum 2000 FTIR Spectrometer.
195
The chemical analysis of bio-oil was performed using a GC/MS based on USEPA 8260
196
method. The elemental components in bio-oil and biochar were investigated using
197
elemental analyzer. The Higher Heating Value (HHV) was measured using oxygen bomb
198
calorimeter.
199 200
3. Results and discussion
201 202
3.1.Challenges in temperature measurement
203
Temperature measurement was one of the major problems in the design of present
204
reactor system. Two thermowells which are 10.5 cm apart from the centre of the reactor as
9
205
depicted in Fig. 2 were used to accommodate thermocouples. Two K-type thermocouples
206
were placed inside the thermowell and connected to the computer via 8 channel Pico data
207
logger. It should be noted that the thermowells are closed from the bottom.
208
Fig. 2 also present the maximum temperature attained at different MW power and
209
biomass loading. Obviously, the temperature was found to increase with MW power, but
210
interestingly it also increased with biomass loading. Thermocouple T2 showed lower
211
temperature as compared to T1. This might be due to the position of T1 thermocouple in
212
close proximity to biomass sample as compared to T2. Overall, the temperature in this
213
study was much lower (in range of 150 to 250 °C) than typical pyrolysis temperature (500 –
214
700 °C). There could be several reasons behind this i) sluggish response of the K-type
215
thermocouple because of its placing inside the thermowell which means thermocouple was
216
not in direct contact with the sample, ii) therefore, thermocouple might be detecting the
217
gaseous or surrounding environment temperature rather than actual sample temperature, iii)
218
samples were almost positioned in the middle as a result of high MW power density at the
219
centre of the reactor which was observed in our preliminary experiment, iv) the
220
thermocouples were positioned at the side of the reactor while the sample at the middle, v)
221
the thermocouple sensor were shielded by two layers, 2 mm thick stainless steel thermowell
222
and about 1 mm thick thermocouple stainless steel shield which might also lower down the
223
response or sensitivity of thermocouples.
224
Several researchers have raised the issue of monitoring the temperature in a MW
225
technology (Zaini and Kamaruddin, 2013; Salema and Ani, 2011; Salema and Ani, 2012a
226
and b). Some have recommended infra-red or optical fiber sensors, but they suffer from
227
detecting the surface temperature only and are expensive. It was also difficult to design and 10
228
place the infra-red or optical fiber sensor in the present reactor design. Nevertheless, the
229
temperature profile obtained by using metallic K-type thermocouple was pretty smooth and
230
almost free from noise (fluctuations). However, repetition of experiments revealed high
231
standard error in the temperature data as shown in Fig. 2, except for 1500 W power and 1
232
kg sample size. It is highly recommended to position the thermowell in contact with the
233
biomass sample to obtain accurate temperature. Further, the electric field usually gets
234
disturb due to loading of biomass sample. Therefore, due to the above limitation in the
235
temperature data, MW power was used to interpret the results in the present study rather
236
than temperature.
237 238
3.2.Pyrolysis product yield
239
Physically, most of the corn stalk biomass briquettes retained their original shape even
240
after pyrolysis. It is assumed that the external surface of CS briquettes might have heated
241
first and subsequently the heat might have transferred to the inner surface. This is possible
242
due to the presence of microwave absorber (in this case biochar) which surrounds the
243
biomass briquettes. Biochar and other carbon materials are usually used as a strong MW
244
absorber which rapidly heats up to a very high temperature in short time. The mechanism of
245
heating biomass in presence of MW absorber might be quite different as compared to
246
without absorber. This is because the heat was reported (Farag et al., 2012;Miura et al.,
247
2004; Undri et al., 2015; Vongpradubchai and Rattanadecho, 2009) to transfer from inner
248
or core of the biomass pellets and wood log to the outer surface. This mechanism of heat
249
transfer cannot be denied in the present briquettes. The porous and fluffy nature of biomass
250
briquettes might also help to penetrate the MW and create the heat at the core. 11
251
Nevertheless, both heating mechanism (outer to core and core to outer) can co-exists at the
252
same time. This needs further detailed studies. Furthermore, preliminary study revealed that
253
corn stalk in their original form and shredded into small size were difficult to pyrolyze and
254
resulted in poor product yield. However, biomass densified into briquettes provided much
255
better results in terms of product yield and pyrolysis behavior.
256
Certainly, the product yield is sensitive to process conditions as evident from Fig. 3, and
257
it is observed to depend on both MW power and biomass loading. Similar results were
258
reported earlier for CS pyrolysis (Pittman, et al., 2012).The yield of bio-oil, biochar and gas
259
ranged from 13.4 to 19.6 wt.%, 30.9 to 41.1wt.%, and 41.6 to 54.0 wt.%, respectively. The
260
bio-oil yield increased by about 22 % when MW power was increased from 900 to 1500 W
261
at 0.5 kg loading and it was highest (19.6 wt.%) at 1200 W and biomass loading of 1 kg.
262
MW power is one of the significant parameter in determining the product yields (Li et al.,
263
2013; Salema and Ani, 2012a and b; Zhuang et al., 2012).Basically, MW power influences
264
the heating rates(Li et al., 2013) and when they increased the power from 750 W to 1500
265
W, the heating rate increased by 112 %, but when they increase the power from 750 W to
266
2250 W, the heating rate increased by almost 900 %. According to them the change in
267
heating rate also depends on the MW power segments. Therefore, the bio-oil yield did not
268
change much in the present study when MW power was increased from 900 to 1500 W,
269
because this range could fall in the medium-high power segment.
270
Certainly, there is also relation between the pyrolysis product yield and the biomass
271
loading. The effect of biomass loading on the pyrolysis product under MW irradiation is
272
still new and very limited knowledge has been developed in the literature. Very recently,
273
effect of biomass loading on the temperature profile, MW electric field and MW absorption 12
274
energy was studied using numerical simulation (Salema and Afzal, 2015).Another study
275
(Robinson et al., 2015)showed significant contribution of sample size on the product yield
276
and quality using experimental work. The general trend of the product yield was the
277
increase in bio-oil and gas yield and decrease in bio-char yield with increase in MW power.
278
Interestingly, the product yield for half and one kilogram biomass loading were almost
279
similar at 1500 W, while it varied for other powers 900 and 1200 W.
280
A decrease in bio-oil yield is always favored with increase in gas yield. This
281
phenomenon particularly happens when the condensable undergoes secondary cracking
282
reactions on its way towards the condenser or if there is lack in rapid condensation (due to
283
longer residence time in the reactor). Thus, most of the condensable vapor might exit the
284
condenser without forming into liquid or bio-oil product. It is quite clear from Table 2 that
285
the product yield also varies with pyrolysis technologies. There could be several reasons
286
behind such contrasts such as:
287
•
Dissimilar reactor configuration and process conditions
288
•
Heating characteristics in MW (volumetric, rapid and selective) is quite different
289
from conventional (conduction and convection) technologies. Typically, MW has
290
higher heating rate as compared to conventional pyrolysis (Mašek et al., 2013).
291
•
Because of above nature, the temperature profile are also quite different in MW
292
•
The CS biomass load used till date is few grams at lab-scale
293
•
The gas yield generally increases with MW power (Huang et al., 2015)
294
•
Mixing MW absorber to biomass sample can also play a role as a catalyst to
295
define the product yield
13
296
•
Non-uniform heating was observed to some extent in the present reactor due to
297
absence of stirring action. This left some CS briquette without pyrolysis (refer to
298
supplementary material).
299
•
The vapors were observed to condense on the wall of reactor as well as on the
300
distributor plate due to difference in temperature along the height of the reactor.
301
This resulted in condensation of high volatile compound inside the reactor which
302
finally affected the bio-oil yield. The problem of bio-oil deposition is attributed to
303
unique heating nature of the MW called as selective heating, in which only the CS
304
biomass is heated to a very high temperature while the surrounding temperature
305
stills remains comparatively low. Therefore, the vapors condense inside the
306
reactor once they experience a lower surrounding temperature. Even though the
307
technique (Abubakar et al., 2013) of introducing nitrogen from the top of the
308
reactor and collecting the vapor from the bottom was implemented to solve the
309
problem of bio-oil deposition, but was less successful in scaled-up reactor (refer to
310
supplementary material).
311
•
When compared with other MW pyrolysis research work as shown in Table 2, the
312
bio-oil yield obtained in present work was lower. This could be due to lower MW
313
power and higher biomass loading in the present study. The present study was
314
limited to 1500 W power due to avoidance of MW leakage, since further increase
315
in MW power resulted in higher MW leakage (> 10 mW/cm2). However, it is
316
anticipated that bio-oil yield can be increased by increasing the MW power
14
317
beyond 2000 W and modification in reactor design to purge out the heavy phase
318
bio-oil.
319 320
•
Lastly, the type of biomass and its physical and chemical characteristics might also contribute in influencing the product yield
321 322
3.3.Product quality
323
As per the visual observation bio-oil was dark red-brown in color and was found to be
324
less viscous may be due to high amount of water and/or low volatile chemical compounds.
325 326
3.3.1. Elemental compositions
327
The elemental composition (CHNOS) of biochar and bio-oil is shown in Table 3.
328
Biochar showed around 63–74 wt.% of carbon, 1.49–2.91 wt.% of hydrogen, 0.1–0.4 wt.%
329
of nitrogen, and 23 – 35 wt.% of oxygen. The C content in the present biochar samples
330
were higher and nitrogen content was lower than the biochar samples obtained from free
331
fall reactor (Shah et al., 2012) and MW pyrolysis (Borges et al., 2014). The O content was
332
much higher in the present biochar but comparable to other study (Shah et al., 2012). The
333
variation in the elemental properties of biochar from various technologies is expected due
334
to types of feedstock, technology, process conditions and pre-treatment if any. There was
335
little effect of MW power on the elemental composition of biochar. However, the sulfur
336
content was found to increase with biomass loading. Biochar can be used as a soil
337
remediation in agricultural applications or as carbon sequestering agent.
338
Elemental compositions of bio-oil in this study showed a remarkable higher oxygen
339
content (around 80 wt.%) than corn stover bio-oil (around 40 wt.%) (Shah et al., 2012). 15
340
However, the values were close to bio-oil produced from corn stover (Borges et al., 2014)
341
(around 78 wt.%) and wood pellets (Undri et al., 2015) (around 77 wt.%) produced from
342
MW pyrolysis. Conversely, the C content was much lower than conventionally produced
343
bio-oil (Shah et al., 2012), but again almost comparable MW pyrolysis produced bio-oil
344
(Borges et al., 2014). Apparently, the C and O content are far lower (85% low) and higher
345
(100% high), respectively than the typical wood based bio-oil (Mohan et al., 2006). The
346
reason could be due to presence of low volatile or light fraction chemicals in the present
347
bio-oil. Higher O content can also be due to presence of water and number of oxygenated
348
chemical compounds found in bio-oil (Mohan et al., 2006). Amazingly, the hydrogen
349
content in the present bio-oil was higher than typical wood based bio-oil (Mohan et al.,
350
2006), corn stalk bio-oil (Pittman et al., 2012), and comparable to corn stover bio-oil
351
(Borges et al., 2014; Undri et al., 2015). In conclusion, bio-oil with the present quality
352
cannot be used in combustion application due to its high O content and lower C content.
353
Further upgrading of bio-oil will not be feasible method due to expensive process and
354
additional unit operations. It is recommended to improve the reactor and/or process design
355
in order to avoid deposition of heavy volatile chemical compounds inside the reactor and
356
maximize their transfer to condenser.
357
3.3.2. Heating value
358
One of the criteria to select a fuel in a power plant is the heating value. Fig. 4 shows
359
the higher heating value (HHV) of biochar and bio-oil. It should be noted that the heating
360
value of biochar represents only of the pyrolysed CS briquettes, since unpyrolysed CS
361
briquettes was easy to separate. The highest HHV of biochar was 31.70 MJ/kg at 900 W
362
power and 0.5 kg biomass load and lowest was 23.00 MJ/kg at 900 W power and 1 kg 16
363
biomass load. The heating value of biochar was found to decrease slightly with increase in
364
MW power and biomass load. Lower value (23.00 MJ/kg) might be due to unpyrolysed CS
365
briquettes. Overall, heating value of biochar between 23 and 32 MJ/kg strongly indicates
366
it’s potential to be used as fuel for bioenergy applications. Using biochar (a carbonaceous
367
material) to assist the pyrolysis of CS in a MW can assist in achieving higher heating value
368
as reported earlier (Undri et al., 2015).
369
Bio-oil showed poor quality in terms of heating value (1.75 to 2.5 MJ/kg), which
370
might be probably due to presence of upper phase bio-oil. One of the main factors that can
371
contribute to such a low heating value is the water content in the bio-oil. Previous study
372
(Undri et al., 2015)also found very low heating values (6 – 8 MJ/kg)of bio-oil produced
373
from wood pellets. The heating value of conventionally (free fall reactor at 400 °C)
374
produced bio-oil was also found to be lower (around 7 MJ/kg) produced for untreated CS
375
(Pittman et al., 2012).Obviously, the heating value of bio-oil is very much dependent on its
376
elemental composition, chemical compounds, and water content. Undoubtedly, based on
377
the present results of the heating value, bio-oil is not a good candidate for combustion and
378
energy application. The challenge of low heating value can be overcome by condensing
379
heavy fraction into bio-oil.
380
Overall, the specific energy to pyrolyse the CS briquette in the developed MW
381
reactor according to the following equation was about 6.5 MJ/kg for 900 W to 10 MJ/kg for
382
1500 W. =
×
17
383
Where, is the specific energy in J/kg, is the input MW power in J/s, is the time of
384
the sample exposed to the MW in s, and
is the mass of sample processed in kg.
385
3.3.3. FTIR analysis
386
The possible chemical functional groups and its compounds present in the bio-oil
387
obtained from MW pyrolysis of corn stalk briquettes (spectra are provided in the
388
supplementary material). In general, the IR spectra for 0.5 and 1 kg bio-oil at all MW
389
power had similar absorption bands and similar relative intensities. The broad and intense
390
peaks between 3200 and 3600 cm-1 wavelength could be possible due to presence of
391
alcohol or phenolic OH stretch. However, the peak in the region of 3400 cm-1 could also be
392
possibly due to high content of water vapor or H-bonded OH stretching. This indicates that
393
bio-oil contain high fraction of water. A weak broad peak at 2000 cm-1 wavenumber could
394
be because of aromatic substitutions such as C-H or C=C stretch. Simple monosubstituted
395
absorbs due to alkenes C=C stretch can be observed at 1630 cm-1. A sharp peak between
396
1685 and 1725 cm-1 wavenumber could be due to C=O stretch ketones with phenyl group
397
or phenolic compounds. The peak at 1380 cm-1 might be due to CH3 bending absorption of
398
alkanes. Two sharp and medium peaks at 1265 and 1380 cm-1 could be attributed to methyl
399
CH3 bending and primary or secondary alcohol OH in-plane bending, respectively. Lastly,
400
the broad and strong peak between 900 and 400 cm-1 can be attributed to alkyl halide either
401
to C-Cl or C-Br or C-I stretch.
402
There is a clear difference in IR spectra of bio-oil and biochar absorption bands as
403
well as intensities (refer to the supplementary material). The broad and weak peaks between
404
3200 and 3600 cm-1 wavelength could be possible due to presence of alcohol or phenol OH
18
405
stretch or because of moisture content. Two strong peaks between 1600 and 1350 cm-1
406
could be due to presence of nitro compounds in the char such as NO2 stretch or N-H
407
bending from amine group. The broad absorption between 1200 and 1000 cm-1 for 1 kg
408
biomass load could be because of C-O stretch and this band was very weak in case of 0.5
409
kg biomass loading. Few weak peaks around 850 cm-1 may be due to aromatic C-H
410
deformation. These peaks almost disappeared or became very weak in the spectra such of
411
1200W-0.5kg and 900W-1 kg.
412
3.3.4. Chemical analysis of liquid (bio-oil) product
413
Chemically, bio-oil is a mixture of hundreds of chemical compounds such as phenol,
414
ketones, aldehydes, furfural, acetic acid, guaiacols, catecols, syringols, formic acid,
415
alcohols, esters, carboxylic acids and even water. The formation of chemical compounds
416
depends on the type of biomass used in pyrolysis process, its chemical and lignocellulosic
417
structure, pyrolysis process conditions, and condensing parameters. The identification of
418
chemicals in the present bio-oil was limited to certain compounds because the bio-oil
419
obtained was of lower quality and mostly contained water components. Therefore, it was
420
difficult to analyse and fractionate the bio-oil compounds and only some specific chemical
421
compounds are presented in Table 4. Another reason for restricting the type of chemical
422
compound was due to the limitation of GC-MS column and machine specifications.
423
The concentration of chemical compounds in the bio-oil was found to depend on both
424
the MW power and biomass loading. The main phenolic compounds (phenol, guaiacol (2-
425
methoxy phenol), Xylenol (2,6-dimethyl phenol and 2,4-dimethyl phenol)) are mainly the
426
products from lignin pyrolysis. Acetone that basically belongs to ketone group is expected
427
to occur in the aqueous phase of bio-oil and is produced as the by-product in the biomass 19
428
pyrolysis reaction. The concentration of 1,2-Dichloroethane-d4 also commonly known as
429
ethylene dichloride and belonging to organic chemical in the family of alkanes and
430
halocarbons was almost consistent with MW power and biomass loading. Lastly, the
431
concentration of fatty acids (palmitic, stearic, and oleic) varied with both power and
432
loading. These fatty acids are basically found in extractives present in the biomass.
433
Remarkably, the quantity for most of the chemical compounds in Table 4 was independent
434
of the amount of biomass loaded in the rector. This signifies that the quantity as well as the
435
quality of certain chemicals in the bio-oil is less dependent on the biomass loading
436
compared to pyrolysis process conditions (MW power, temperature, condensing
437
temperature and system) and reactor design.
438
Typically, phenol and its derivatives form large fraction in bio-oil of the CS (Lv and
439
Wu, 2012) followed by acetic acid, furfural, ketones, furans and aldehydes. The formation
440
of these chemical compounds is very much dependent on the lignocellulosic content of the
441
biomass. The average concentration of phenols and its compounds was around 60 wt.% as
442
shown in Table 4.This was in agreement with previous studies(Lv et al., 2013; Pittman et
443
al., 2012) who also reported phenol to be the principal component from the lignin pyrolysis.
444
As can be seen from Table 4, the quantity of phenolic compounds produced from CS may
445
not necessarily depend on the biomass loading while keeping the MW power constant. The
446
total phenol content decreased when biomass loading was increased from 0.5 kg to 1 kg.
447
For instance, it was reported (Lv et al., 2013) that increase in temperature favors the
448
formation of specific type of phenolic compound. Moreover, formation of phenolic
449
compounds usually takes place at higher pyrolysis temperature (Mohan et al., 2006). Very
450
limited work has been carried out in the literature about the scale-up of biomass processing 20
451
in MW technology. Thus, additional research work is required to ascertain the effect of
452
scaling-up the MW technology on the quality and quantity of chemical compounds
453
produced from biomass pyrolysis.
454 455
4. Conclusions
456
To our knowledge the pyrolysis of biomass in kilogram (scaled-up reactor) and in form
457
of briquette in MW technology is reported for the first time. The quantity (yield) of biochar,
458
bio-oil, and gas greatly depended on the process condition (MW power and biomass
459
loading). Based on its heating value and elemental composition, biochar is suitable for
460
energy applications, whereas bio-oil produced from CS in this study is neither good for
461
energy application nor for chemical production. Further research work is needed in terms of
462
reactor design and biomass loading to prove the feasibility of system at large-scale
463
production.
464 465
Acknowledgements
466 467
The authors are grateful to the New Brunswick Department of Agriculture, Aquaculture
468
and Fisheries (NBDAAF); and Natural Sciences and Engineering Research Council of
469
Canada (NSERC) for the financial support to this project. First author pursued his
470
postdoctoral studies at the University of New Brunswick, Canada.
471 472 473 21
474
Supplementary materials
475
Fig.A1. FTIR analysis of bio-oil (A and B) and bio-char (C and D) obtained from MW
476
pyrolysis of corn stalk and Table A1 Photos of CS briquettes before and after MW
477
pyrolysis are available on the journal paper website.
478 479 480
References 1. Abubakar, Z., Salema, A.A., and Ani, F.N., 2013. A new technique to pyrolyse
481
biomass in a microwave system: effect of stirrer speed. Bioresour. Technol.128,
482
578-585.
483
2. Bailey S.S., Samson, R., and Lem, C.H., 2006. Biomass Resource Options:
484
Creating a BIOHEAT Supply for the Canadian Greenhouse Industry. Natural
485
Resources Canada.
486
3. Borges, F.C., Du, Z., Xie, Q., Trierweiler, J.O., Cheng, Y., Wan, Y.Q., Liu, Y.H.,
487
Zhu, R.B., Lin, X.Y., Chen, P., and Ruan, R., 2014. Fast microwave assisted
488
pyrolysis of biomass using microwave absorbent. Bioresour. Technol.156, 267-274.
489
4. Cordella, M., Berrueco, C., Santarelli, F., Paterson, N., Kandiyoti, R., Millan, M.,
490
2013. Yields and ageing of the liquids obtained by slow pyrolysis of sorghum,
491
switchgrass and corn stalks. J. Anal. Appl. Pyrol. 104, 316-324.
492
5. Farag, S., Sobhy, A., Akyel, C., Doucet, J., Chaouki, J., 2012. Temperature profile
493
prediction within selected materials heated by microwaves at 2.45GHz. Appl.
494
Therm. Eng. 36, 360-369.
22
495
6. Huang, Y.F., Kuan, W.H., Chang, C.C., Tzou, Y.M., 2013. Catalytic and
496
atmospheric effects on microwave pyrolysis of corn stover. Bioresour. Technol.
497
131, 274-280.
498
7. Li, H., Qu, Y., Yang, Y., Chang, S., and Xu, J., 2016. Microwave irradiation–A
499
green and efficient way to pretreat biomass. Bioresour. Technol.199, 34-41.
500 501 502
8. Li, L., Ma, X., Xu, Q., Hu, Z., 2013. Influence of microwave power, metal oxides and metal salts on the pyrolysis of algae. Bioresour. Technol. 142, 469-474. 9. Liu, X., Zhang, Y., Li, Z., Feng, R., and Zhang, Y., 2014. Characterization of
503
corncob-derived biochar and pyrolysis kinetics in comparison with corn stalk and
504
sawdust. Bioresour. Technol.170, 76-82.
505
10. Liu, S., Xie, Q., Zhang, B., Cheng, Y., Liu, Y., Chen, P., and Ruan, R. 2016. Fast
506
microwave-assisted catalytic co-pyrolysis of corn stover and scum for bio-oil
507
production with CaO and HZSM-5 as the catalyst. Bioresour. Technol. 204, 164-
508
170.
509 510 511
11. Lv, G., Wu, S., 2012. Analytical pyrolysis studies of corn stalk and its three main components by TG-MS and Py-GC/MS. J. Anal. Appl. Pyrol. 97, 11-18. 12. Lv, G., Wu, S., Yang, G., Chen, J., Liu, Y., Kong, F., 2013. Comparative study of
512
pyrolysis behaviors of corn stalk and its three components. J. Anal. Appl. Pyrol.
513
104, 185-193.
514 515
13. Macquarrie, D. J., Clark, J. H., and Fitzpatrick, E., 2012. The microwave pyrolysis of biomass. Biofuels Bioprod. Biorefining6(5), 549-560.
23
516
14. Mašek, O., Budarin, V., Gronnow, M., Crombie, K., Brownsort, P., Fitzpatrick, E.,
517
and Hurst, P., 2013. Microwave and slow pyrolysis biochar—Comparison of
518
physical and functional properties. J. Anal. Appl. Pyrol. 100, 41-48.
519 520 521 522 523 524
15. Miura, M., Kaga, H., Sakurai, A., Kakuchi, T., Takahashi, K. 2004. Rapid pyrolysis of wood block by microwave heating. J. Anal. Appl. Pyrol. 71(1), 187-199. 16. Mohan, D., Pittman, C.U., Steele, P.H., 2006. Pyrolysis of Wood/Biomass for Biooil: A Critical Review. Energy Fuels, 20(3), 848-889. 17. Motasemi, F., and Afzal, M.T., 2013. A review on the microwave-assisted pyrolysis technique. Renew. Sustain. Energy Rev. 28, 317-330.
525
18. Pianroj, Y., Jumrat, S., Werapun, W., Karrila, S., and Tongurai, C., 2016. Scaled-up
526
reactor for microwave induced pyrolysis of oil palm shell. Chem. Eng. Process.:
527
Process Intensification106, 42-49.
528
19. Pittman, C.U., Mohan, D., Eseyin, A., Li, Q., Ingram, L., Hassan, E.B.M., Mitchell,
529
B., Guo, H., Steele, P.H., 2012. Characterization of bio-oils produced from fast
530
pyrolysis of corn stalks in an auger reactor. Energy Fuels 26(6), 3816-3825.
531
20. Robinson, J., Dodds, C., Stavrinides, A., Kingman, S., Katrib, J., Wu, Z., Medrano,
532
J., Overend, R., 2015. Microwave pyrolysis of biomass: control of process
533
parameters for high pyrolysis oil yields and enhanced oil quality. Energy Fuels
534
29(3), 1701-1709.
535 536
21. Robinson, J.P., Kingman, S.W., Barranco, R., Snape, C.E., and Al-Sayegh, H., 2009. Microwave pyrolysis of wood pellets. Ind. Eng. Chem. Res.49 (2), 459-463.
24
537
22. Salema, A.A., Afzal, M.T., 2015. Numerical simulation of heating behaviour in
538
biomass bed and pellets under multimode microwave system. Int. J. Therm. Sci. 91,
539
12-24.
540 541 542 543 544
23. Salema, A.A., and Ani, F.N, 2012a. Microwave-assisted pyrolysis of oil palm shell biomass using an overhead stirrer. J. Anal. Appl. Pyrol.96, 162-172. 24. Salema, A.A., and Ani, F.N., 2011. Microwave induced pyrolysis of oil palm biomass. Bioresour. Technol. 102(3), 3388-3395. 25. Salema, A.A., Ani, F.N., 2012b. Pyrolysis of oil palm empty fruit bunch biomass
545
pellets using multimode microwave irradiation. Bioresour. Technol. 125, 102-107.
546
26. Shah, A., Darr, M.J., Dalluge, D., Medic, D., Webster, K., and Brown, R.C., 2012.
547
Physicochemical properties of bio-oil and biochar produced by fast pyrolysis of
548
stored single-pass corn stover and cobs. Bioresour. Technol. 125, 348-352.
549
27. Undri, A., Abou-Zaid, M., Briens, C., Berruti, F., Rosi, L., Bartoli, M., Frediani,
550
M., Frediani, P., 2015. Bio-oil from pyrolysis of wood pellets using a microwave
551
multimode oven and different microwave absorbers. Fuel 153, 464-482.
552 553 554
28. Uzun, B.B., Sarioğlu, N., 2009. Rapid and catalytic pyrolysis of corn stalks. Fuel Process. Technol. 90(5), 705-716. 29. Vongpradubchai, S., Rattanadecho, P., 2009. The microwave processing of wood
555
using a continuous microwave belt drier. Chem. Eng. Process.: Process
556
Intensification 48(5), 997-1003.
557
30. Wang, T., Ye, X., Yin, J., Lu, Q., Zheng, Z., and Dong, C., 2014. Effects of
558
biopretreatment on pyrolysis behaviors of corn stalk by methanogen. Bioresour.
559
Technol. 164, 416-419. 25
560 561 562 563
31. Yin, C., 2012. Microwave-assisted pyrolysis of biomass for liquid biofuels production. Bioresour. Technol.120, 273-284. 32. Zaini, M.A.A., and Kamaruddin, M.J., 2013. Critical issues in microwave-assisted activated carbon preparation. J. Anal. Appl. Pyrol. 101, 238-241.
564
33. Zhang, B., Zhong, Z., Ding, K., Cao, Y., Liu, Z., 2014. Catalytic upgrading of corn
565
stalk fast pyrolysis vapors with fresh and hydrothermally treated HZSM-5 catalysts
566
using Py-GC/MS. Ind. Eng. Chem. Res. 53(24), 9979-9984.
567
34. Zhao, X., Zhang, J., Song, Z., Liu, H., Li, L., and Ma, C., 2010. Microwave
568
pyrolysis of straw bale and energy balance analysis. J. Anal. Appl. Pyrol. 92(1), 43-
569
49.
570
35. Zhuang, Y., Guo, J., Chen, L., Li, D., Liu, J., Ye, N., 2012. Microwave-assisted
571
direct liquefaction of Ulva prolifera for bio-oil production by acid catalysis.
572
Bioresour. Technol.116, 133-139.
573 574 575
26
576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608
Fig. 1. Schematic diagram of a MW pyrolysis system developed at UNB, Canada
609 610 611 612
27
613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630
Fig. 2. Maximum temperature attained during MW pyrolysis of CS briquette biomass and the
631
schematic diagram of thermocouple design
632 633 634 635 28
636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680
Fig. 3. Product yield obtained from MW pyrolysis of CS briquettes
29
681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702
Fig. 4. Higher heating value (HHV) of (A) biochar and (B) bio-oil
703 30
727 728
Table 1 Pyrolysis of corn stalk using conventional and microwave technology (2009-2016) Reference Technology Method Product Wang et al., 2014 Pyroprobe Methanogen pretreatment Chemicals connected to of corn stalk GC/MS Liu et al., 2014 Lab-scale fixed bed The feedstock were air Biochar reactor was used dried and stored in sealed for slow pyrolysis containers Zhang et al., 2014 Py-GC/MS Corn stalk was driedat 105 Chemicals in bio°C for 24 handgrounded to oil and char 40 mesh size. Catalytic pyrolysis using a HZSM-5 catalyst. Cordella et al., Laboratory-scale Corn stalk samples were Bio-oil and 2013 fixed bed reactor dried overnight at 60 °C biochar and ground to particle sizeless than 1 mm.2–3 g of biomass sample per run was used. Lv et al., 2013 Tubular furnace Corn stalk without leaves Chemicals in bioand Py-GC/MS was groundedto a size of oil and biochar 40–60 mesh (0.28–0.45 mm). Pre-treatment was also used in this study. Corn stalk was also fractionated into cellulose, hemicellulose and lignin.About 350 mg of each samplewere loaded. Pittman, Jr. et al., Auger reactor with Untreated and acid-treated Chemicals in bio2012 feed rate of 1−2.5 corn stalks were used. The oil kg/h and stalks were crushedto a temperature of 400 particlediameter of 0.5−5 to 450 °C. mm and densified under a pressure of 550 psi and166 °C for 4 min. Zhao et al., 2010 Microwave Corn stalk bale was Gas pyrolysis with total weighed and loaded in the microwave input microwave. The size of power of 18 kW corn stalk bale was about 1000mm×600mm×600mm. Uzun and High-speed heated Corn stalks wereair-dried Chemicals in bioSarioğlu, 2009 fixed-bed and ground with average oil
32
tubularreactor
size in range of 0.85
729 730 731 732 733 734 735
33
736 737 738
Table 2 Product yield from CS biomass using different pyrolysis technologies
Technologies
Process conditions
MW pyrolysis
MW power – 900 to 1500 W, radiation time – 2 h, biomass loading – 0.5 and 1 kg, temperature* – 150 to 250 °C Corn stover, Temp - 500 °C, biomass particle size - 1 mm, vacuum degree - 170 mmHg Wood pellets, sample load 211.7 g, MW power - 1.2 to 3.0 kW, Carbon absorber 75 g Corn stover co-pyrolysis, Temp – 450 °C, SiC as absorber, CaO/HZSM-5 catalyst, biomass load – 30 g Temp – 800 °C, residence time – 1 to 2 min Feed rate – 1 to 1.25 kg/h, temp – 400 to 450 °C, residence time – 50 s, pretreatment in 2 wt % aqueous H2SO4 solutions Biomass load – 5 g, residence time – 10 min, temp – 600 °C, used zeolite catalyst Biomass load – 2 to 3 g, temp – 650 °C
MW pyrolysis
MW pyrolysis
MW pyrolysis
Fixed bed tubular furnace Auger reactor
Fixed-bed tubular reactor
739 740 741
Product yield BioBio-oil, Gas, char, wt.% wt.% wt.%
References
17.0
35.0
48.0
This study
64
22
14
Borges et al., 2014
46.1
24.9
29.0
Undri et al., 2015
17.4
71.1
11.5
Liu et al., 2016
30.0
25.0
45.0
(Lv et al., 2013)
35.0
29.0
13.5
(Pittman et al., 2012)
46.0 (with water)
20.0
34.0
(Uzun and Sarioğlu, 2009)
Laboratory(Cordella et scale fixed bed 50.0 30.0 20.0 al., 2013) reactor *Note: Please see section 3.1., for the temperature accuracy. However, actual temperature might be 3 times higher than measured.
742
34
743 744 745 746 747 748 749 750
Table 3 Elemental composition of biochar and bio-oil produced from MW pyrolysis of corn stalk briquette MW power, W Load, kg
900 1200 1500
900 1200 1500
0.5 1 0.5 1 0.5 1 0.5 1 0.5 1 0.5 1
C H N S O wt% wt% wt% wt% wt% Biochar 74.33 1.86 0.35 0.0035 23.45 67.55 2.91 0.40 0.0037 29.13 63.05 1.49 0.14 0.0085 35.31 67.49 2.69 0.36 0.0143 29.43 71.61 1.89 0.35 0.0190 26.13 66.05 2.66 0.48 0.0625 30.75 Bio-oil 7.96 8.80 0.04 0.0016 83.19 9.12 8.46 0.04 0.0018 82.38 7.85 10.43 0.05 0.0017 81.65 9.09 9.05 0.11 0.0030 81.74 8.73 9.79 0.10 0.0019 81.37 8.75 10.13 0.07 0.0023 81.04
751 752 753 754
35
766 767 768 769 770 771
Table 4 GC/MS analysis of bio-oil at different experimental conditions
0.5kg 1kg 900W 1200W 1500W 900W 1200W 1500W Phenol, µg/ml 220 140 190 150 140 170 2,6-dimethyl phenol, µg/ml 2.2 2.0 2.2 2.0 1.8 1.7 2,4-dimethyl phenol, µg/ml 9.5 8.4 11 9.9 10 10 2-methoxy phenol, µg/ml 87 97 78 120 100 100 Acetone, mg/kg 46 170 240 370 360 360 1,2-Dichloroethane-d4, % 96 105 106 108 101 106 C16:0 (Palmitic), mg/kg 24 78 38 27 37 31 C16:1n7 (Palmitoleic), mg/kg < 10 < 10 < 10 < 10 < 10 < 10 C18:0 (Stearic), mg/kg 13 38 25 17 20 24 C18:1n9 (Oleic), mg/kg 17 56 23 24 24 38 C18:2n6 (Linoleic), mg/kg < 10 26 < 10 < 10 11 15 C20:0 (Arachidic), mg/kg < 10 < 10 < 10 < 10 < 10 < 10 C18:3n3 (ALA), mg/kg < 10 < 10 < 10 < 10 < 10 < 10
Chemical names
772 773 774 775 776 777 778
37
791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821
Graphical abstract
39
822 823 824
Highlights •
First time microwave (MW) pyrolysis of biomass briquette was carried out
825
•
Biomass loading was scaled up to kilograms
826
•
HHV of biochar and bio-oil was 32 MJ/kg and 2.5 MJ/kg, respectively
827
•
Pyrolysis product yield dependent on the process parameters, MW power and
828 829
loading •
Reactor design can be further improved to increase the bio-oil quality
830 831
40