Two-Stage Hydrothermal Liquefaction of Sweet Sorghum Biomass

May 15, 2018 - Based on the previous studies,(33−35) homogeneous alkaline catalyst, .... Institute of Standards and Technology (NIST) mass spectral ...
3 downloads 0 Views 609KB Size
Subscriber access provided by Kaohsiung Medical University

Biofuels and Biomass

Two-stage hydrothermal liquefaction of sweet sorghum biomass - Part II: Production of upgraded biocrude oil Yang Yue, James R Kastner, and Sudhagar Mani Energy Fuels, Just Accepted Manuscript • DOI: 10.1021/acs.energyfuels.8b00669 • Publication Date (Web): 15 May 2018 Downloaded from http://pubs.acs.org on May 16, 2018

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

A schematic of two-stage HTL-HDO process for sweet sorghum biomass to produce upgraded biocrude oil 275x113mm (96 x 96 DPI)

ACS Paragon Plus Environment

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

Two-stage hydrothermal liquefaction of sweet sorghum biomass -

2

Part II: Production of upgraded biocrude oil

3

Yang Yue, James R. Kastner and Sudhagar Mani*

4

School of Chemical, Materials and Biomedical Engineering, University of Georgia, Athens, GA

5

*

Corresponding Author: Ph: (706) 542-2358; email: [email protected]

6 7

ABSTRACT: The hydrothermal liquefaction (HTL) followed by hydro-deoxygenation (HDO)

8

of lignin-rich biomass has a potential to improve the yield and the quality of upgraded biocrude

9

oil. In this study, the lignin-rich biomass obtained from the low-temperature HTL (stage 1) of

10

sweet sorghum bagasse was investigated to evaluate the biocrude oil yield and its compositions

11

using catalytic high-temperature HTL-HDO (stage 2) and compared with conventional, whole-

12

stage HTL-HDO process. Biocrude oil yield was achieved up to 38% during a high-temperature

13

HTL, while 42% was obtained after catalytic (5% Ru/C) HDO process. Up to 96% hydrocarbon

14

content of biocrude oil was achieved, which is twice higher than that of conventional whole-

15

stage HTL-HDO process. Aromatic hydrocarbons and long-chain alkanes were also dominant in

16

the upgraded biocrude oil, which collectively accounted for 28% of total hydrocarbons. From

17

raw sweet sorghum bagasse, the hydrocarbons yield from second-stage HTL-HDO process (11%)

18

was substantially increased by 78% compared to that of whole-stage HTL-HDO process (7%).

19

The upgraded biocrude oil could be directly processed or co-processed in the existing refinery to

20

produce drop-in fuels.

21 22

Keywords: High-temperature HTL, lignin, biocrude oil, hydrodexoygenation,

23 24

1 ACS Paragon Plus Environment

Page 2 of 32

Page 3 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

1

1. INTRODUCTION

2

Lignocellulosic biomass is both sustainable and renewable energy resources for biofuel

3

production due to its abundant availability and potential to reduce greenhouse gas emissions.

4

Sweet sorghum, a typical C4 annual energy crop can be grown on an uncompetitive arable land.

5

This energy crop has a short growth cycle, produces high biomass yield, requires low

6

fertilization input and is tolerance to drought condition.1-3 Therefore, it can be an excellent

7

biorefinery feedstock to produce diverse stream of biochemical and biofuels. Sweet sorghum

8

biomass also accumulates sugars in plant stalks, which can be squeezed to extract juices rich in

9

sucrose.4, 5 The sugar content of the juice varies from 9 to 15%, with the total juice yield of 3.6 to

10

15.5 Mg ha-1. The bagasse, the byproduct of stalk after juice extraction represents approximately

11

two-third of the dry biomass, and has limited value due to its low energy content and high stalk

12

moisture content.6 However, the bagasse with high moisture content ranging from 60-80% (wet

13

basis) could be an excellent feedstock for hydrothermal liquefaction (HTL) based biorefinery.

14

Hydrothermal liquefaction (HTL) is a rapidly developing technology to convert wet biomass

15

into biocrude oil using subcritical water as the solvent at moderate temperature (280-370 °C) and

16

high pressure (10-25 M Pa).7 During HTL process, all the solid biomass is completely dissolved

17

into mainly liquid and gaseous products. The liquid product can be further separated into

18

aqueous and oil (biocrude oil) fractions. The unconverted solids and other inorganic minerals in

19

the biomass are called residual solids or char. The HTL biocrude oil yield from lignocellulosic

20

biomass ranged from 30 to 40%;8-12 this yield was impacted by various HTL parameters,

21

including temperature, catalyst, and retention time. It was reported that the typical HTL

22

temperature range was from 300 °C to 350 °C, to obtain a maximum biocrude oil yield using

23

water as a solvent with or without catalysts.13-16 The utilization of an alkaline catalyst in the HTL

2 ACS Paragon Plus Environment

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

process suppresses the formation of solid products, which would enhance the biocrude oil yield.9

2

Karagöze et al. employed 0.94 M K2CO3 as a catalyst to HTL of wood and found that the

3

biocrude oil yield increased from 17.8 to 33.7% and the solid residue yield was reduced from 42

4

to 4%.17, 18 For most lignocellulosic biomass, the reported optimum holding time ranged from 15

5

min to 60 min depending on the reaction temperature (250~375 °C) and heating rate (5-

6

140 °C/min).7, 19, 20 Alkaline catalyst improved the processes of hydrolysis and decomposition

7

and shortened holding time; however, the most important effect of alkaline catalyst was to

8

increase biocrude oil yield.21 It was also reported by Boocock22 and Zhu13 that prolonged holding

9

time suppressed the biocrude oil yield because the cracking reaction of biocrude oil or its

10

precursors promoted the formation of gases and chars through condensation and

11

repolymerization reactions.23 The reduction in oil yield with longer holding time was also

12

confirmed by the National Collaborative Research Infrastructure Strategy (NCRIS, University of

13

Sydney, Australia) through a continuous HTL process, which reported that shorter residence time

14

with a higher heating rate benefited the higher biocrude oil yield.24

15

The obtained biocrude oil through HTL contains a wide-range of highly-oxygenated small

16

molecular compounds as well as undesirable water, which resulted in poor combustion properties

17

due to low heating value, high acidity, high reactivity, and corrosion against combustor.25 Hydro-

18

deoxygenation (HDO) is, therefore, essentially required to improve biocrude oil quality by

19

catalytically removing oxygenated compounds from biocrude oil through hydrodeoxygenation

20

and decarboxylation reactions in the presence of hydrogen and heterologous catalysts at

21

moderate temperature (250-450 °C) and high pressure (75-300 bar) before applying for drop-in

22

fuels.26 The most commonly used HDO catalyst is ruthenium metal catalyst supported with

3 ACS Paragon Plus Environment

Page 4 of 32

Page 5 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

1

carbon (Ru/C), which has high potential to increase biocrude oil yield (52%) and higher degree

2

of HDO (86%) among various transition metal catalysts at 350 °C for 4 h reaction time.27, 28

3

One major issue concerning HTL of lignocellulosic biomass is that the presence of

4

hemicellulose in the biomass limits the biocrude oil yield and further impairs the downstream

5

hydro-deoxygenation (HDO) upgrading process due to high content of relatively small

6

oxygenates.29 Although the hemicellulose in the biomass accounts for approximately 20-40%,

7

but it was reported to contribute only up to 5.3% of biocrude oil yield.30 The majority of the

8

hemicellulose during HTL process were converted into aqueous compounds (up to 42%).30 Liu

9

also reported the cross reactions occurred between hemicellulose derived intermediates and

10

lignin units, further promoted the formation of solid residues through repolymerization

11

reactions.31 However, limited research was conducted on improving the HTL oil yield, while

12

effectively utilizing all fractions of the lignocellulosic biomass into useful products. Therefore,

13

we developed a two-stage HTL-HDO technology to comprehensively utilize lignocellulosic

14

biomass. In the first-stage HTL, hemicellulose in the biomass can be selectively hydrolyzed

15

under low-temperature HTL conditions and the remaining lignin-rich solid are prepared for

16

further conversion into upgraded biocrude oil (Fig. 1). The optimal first-stage HTL conditions

17

for sweet sorghum bagasse and the sugar yields were investigated in our previous study and

18

found that (i) about 94.1% of hemicellulose and 49.2% cellulose were respectively hydrolyzed

19

into C5 and C6 sugars platform with the concentration of 28.52 g/L; (ii) the co-product of lignin-

20

rich solid fraction, an ideal source of phenolics feedstock can be further hydrothermally liquefied

21

into high-quality biocrude oil.32 This article aims to investigate the feasibility of conversion

22

pretreated lignin-rich biomass into high-quality biocrude oil. The main objectives of this paper

23

were to experimentally investigate the high-temperature HTL-HDO of lignin-rich sweet sorghum

4 ACS Paragon Plus Environment

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

bagasse to determine the biocrude oil yield compared with whole-stage HTL-HDO process and

2

to evaluate the fuel characteristics and the compositions of second-stage HTL and HDO

3

upgraded biocrude oil to produce drop-in biofuels.

4

5 6 7 8 9

Fig.1 A schematic of two-stage HTL-HDO process for sweet sorghum biomass to produce upgraded biocrude oil.

10

2. EXPERIMENTAL SECTION

11

2. 1 Materials. Sweet sorghum bagasse used for the whole-stage HTL-HDO was obtained from

12

the experimental research station in Fort Valley State University, Fort Valley, Georgia, USA.

13

Sweet sorghum variety of EJ7281 (Ceres Inc., Thousand Oaks, California) was harvested in mid-

14

October and the stalks were squeezed to extract sugar juices. The remaining solids, sweet

15

sorghum bagasse was size reduced using a knife cutting mill (Retsch SM 2000, Germany) into a

16

fine powder with an average size of 0.3 mm and then oven-dried at 105 °C overnight. The

17

sample was stored in a desiccator for subsequent use. It composed of around 20% hemicellulose,

18

36% cellulose and 18% lignin. The lignin-rich sweet sorghum bagasse used in the second-stage

19

HTL-HDO was prepared from the first-stage HTL process using the optimal condition of 170 °C

5 ACS Paragon Plus Environment

Page 6 of 32

Page 7 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

1

for 90 min based on our previous study.32 The lignin-rich biomass contained about 33% lignin

2

and 43% cellulose with a negligible hemicellulose content. Reagent grade dichloromethane

3

(DCM), magnesium perchlorate, and 5% ruthenium on carbon (Ru/C, particle size of 20 µm)

4

were all purchased from Sigma-Aldrich.

5

2.2 HTL and HDO Experiment. The stainless-steel reactor utilized for HTL and HDO was a

6

Parr 4598 batch micro-stirred reactors system (Parr Instrument Company, Moline, USA) with a

7

maximum temperature of 500 °C and pressure of 5000 psi. The vessel capability was 100 ml. A

8

consumable Parr Grafoil graphite gasket was used to avoid pressurized gas leakage during

9

reaction. A 4-blade impeller (0.81 in dia.) powered by a magnetic stirrer (model no. A1120HC6,

10

1/8 hp variable speed) provided agitation with a maximum speed of 750 rpm. The vessel was

11

heated through a ceramic fiber external jacket (700 watts), and the processes of HTL and HDO

12

reaction were monitored with Parr Specview Software. The sketch of the experimental apparatus

13

is showed as Fig. 2. The average heating rate was 14 °C/min; the average cooling rate was

14

34.5 °C/min in a water bath. 3

He 2

4

Monitor 5 6 1

Micro-stirred reactor 15

6 ACS Paragon Plus Environment

7

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1 2 3

Fig. 2 A schematic graph of experimental micro-stirred reactor. 1. Helium cylinder, 2. Pressure gauge, 3. rotor, 4. Thermocouple, 5. Heating jacket, 6. Gas sample collector, 7. Liquid sample collector.

4

About 5.0 g of dried sweet sorghum bagasse sample was placed into the vessel followed by

5

the addition of 50 ml 0.94 M potassium carbonate solution or water for each HTL and control

6

experiments. Helium was used to purge and displace the air in headspace until 500 psi was

7

reached. The reaction was monitored through Parr Specview Software at a temperature of 350 °C

8

and with an agitation of 300 rpm. After the holding time, the vessel was immediately cooled in a

9

water bath until reaching room temperature. The gas sample was then collected from the

10

headspace through a blowdown needle valve into a sealed-air sample bag. The vessel was

11

unloaded after releasing pressure through the blowdown valve. The liquid and solid products

12

were analyzed or immediately used for HDO experiment. Based on the previous studies,33-35

13

homogeneous alkaline catalyst, K2CO3 was used a catalyst in the present study; and three

14

different holding times (15, 30 and 60 min) were investigated for the whole-stage HTL to study

15

the effects on biocrude oil yield and its quality. The optimum condition of whole-stage HTL

16

process for the maximum biocrude oil yield was used to study the second-stage HTL. In the

17

second-stage HTL, the lignin-rich sweet sorghum bagasse obtained from first-stage HTL

18

conditions (170 °C for 90 min) was used to determine the biocrude oil yield and its compositions.

19

HDO Experiment: Hydro-deoxygenation (HDO) experiment was carried out immediately

20

after HTL experiment in the same Parr reactor. About 1.5 g Ru/C was added as a catalyst into the

21

vessel. The properties of Ru/C catalyst have been reported in our pervious study with the surface

22

area of 714 m2/g, average pore size of 14.4 Å and pore volume of 0.52 cm3/g.36 Hydrogen was

23

used to purge the headspace under a pressure of 50 psi for 30 s three times and finally

24

pressurized until 500 psi was reached. The HDO process was conducted at 350 °C for 4 h with

25

an agitation rate of 500 rpm. After HDO reaction, the gas sample was collected as similar to the 7 ACS Paragon Plus Environment

Page 8 of 32

Page 9 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

1

HTL procedure described previously. After unloading the vessel, the solid and liquid samples

2

were collected and stored in a refrigerator at 4 °C before separation for compositional analysis.

3

Three replicates of HTL and HDO operations were carried out and the average data was reported

4

with standard derivation. All the yields and compositions of HTL products were determined

5

based on mass basis.

6

2.3 Separation and Extraction Procedures. About 40 ml DCM solvent was added into the

7

mixture of HTL or HTL-HDO liquid and solid products to extract biocrude oil (oil phase). The

8

mixture was then filtered with vacuum filtration through Whatmann #4 90mm filter paper. The

9

solid residue was rinsed with an additional 15 ml DCM to collect residual biocrude oil. The

10

rinsing solution was merged with the filtrate. The biocrude oil fraction in liquid products was

11

extracted with the combined DCM and decanted off from aqueous phase using a separatory

12

funnel. To remove residual moisture, 1.0 g anhydrone (magnesium perchlorate) was added into

13

the biocrude oil fraction. This anhydrone-biocrude oil mixture was then filtered again through

14

Whatmann #4 90mm filter paper, and the extracted de-watered biocrude oil fraction was further

15

transferred into a round-bottomed flask for rotary evaporation in a water bath of 36 °C for 60 min

16

under a vacuum of 2 mmHg. The DCM free biocrude oil or upgraded biocrude oil was then

17

collected and weighed before storing in a refrigerator for compositional analysis.

18

2.4 HTL and HDO Products Analysis. Major compounds in aqueous products were

19

determined with HPLC using a Coragel 94 column with the mobile phase of 4 mN sulfuric acid

20

at the flow rate 0.6 ml/min. The oven temperature was kept constantly at 60 °C. 5 µl sample was

21

injected for analysis.

22

The elemental composition of biocrude oil was measured using an elemental analyzer (LECO

23

CHNS 932, LECO Corporation, St. Joseph, MI), based on the procedure of ASTM D3176–89 in

8 ACS Paragon Plus Environment

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

which oxygen content was obtained by difference.37 Moisture content of biocrude oil was

2

measured with Karl Fisher Titration.38

3

The higher heating value (HHV) of biocrude oil could be calculated based on the elemental

4

composition, using the Dulong Formula as

5

(MJ/kg) = 0.3383C + 1.422 (H − O/8) 39

6

The compositions of biocrude oil were analyzed with GC-MS using a HP-5ms Capillary

7

Column (30m x 0.25um x 0.25mm). The GC oven temperature was set to 40 °C initially and

8

hold for 4 min, then ramped at 5 °C /min until 275 °C and held consistently for 5 min. The Mass

9

Spec interface and inlet temperature were set at 280 °C and 260 °C, respectively. Helium was

10

employed as GC carrier gas at the flow rate of 0.8 mL/min. The split ratio was set to 50:1. The

11

electron impact (EI) ionization was set at 70 eV with a mass to charge range from 50 to 600 u.

12

Major compounds identified with GC-MS from biocrude oil and upgraded biocrude oil were

13

quantitatively analyzed with an internal standard method. Neat compounds of toluene, trimethyl

14

benzene, pentadecane, 2-methoxy phenol and 2-cyclopenten-1-one were used to estimate the

15

mass percentage of aromatic hydrocarbons, benzenes, alkanes, phenols and cyclic ketones,

16

respectively. DCM was used as a solvent to mix with the internal standard, 1-hexanol, at a

17

concentration of 1.04 g/L. The standard calibration curves were developed by calculating the

18

quantification factor of the slope obtained from concentration versus GC area of above neat

19

compounds with three replications. For biocrude oil and upgraded biocrude oil compositional

20

analyses, a standard stock solution was prepared immediately before GC-MS operation by

21

adding 10 µl 1-hexanol into 800 µl DCM. After vortexing, 35 µl standard stock solution was

22

mixed with 302 µl biocrude oil or upgraded biocrude oil sample. 1 µl mixed sample solution was

23

injected into GC-MS after filtration with a 0.45 µm filter. The solvent delay was setup for 3 min

9 ACS Paragon Plus Environment

Page 10 of 32

Page 11 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

1

to ignore DCM. The compounds were identified by matching fragmentation patterns with

2

National Institute of Standards and Technology (NIST) mass spectral library. Each sample from

3

HTL and HDO experiments was repeated thrice. The average result was reported with standard

4

deviation.

5 6

3. RESULTS AND DISCUSSION

7

3.1 Whole-Stage vs Second-Stage HTL Experiment

8

3.1.1 HTL Biocrude Oil Yield. The biocrude oil yield increased up to 1.44 times in the presence

9

of K2CO3 catalyst due to decreased aqueous product and solid yields (Table 1). During HTL of

10

lignocellulosic biomass, K2CO3 catalyzed the hydrolysis of lignin and cellulose macromolecules

11

into oligomers and monomers; thus, promoted the following decomposition into a diversity of

12

secondary small-molecular compounds through dehydration, dehydrogenation, deoxygenation,

13

and decarboxylation reactions. Once produced on the basis of their activities, these small-

14

molecular compounds experienced rearrangement reactions of condensation, cyclization and

15

polymerization and formed the majority of biocrude oil.40 In addition, the solid product yield

16

decreased by 58%, while the gas product yield increased by 2.27-fold due to an alkali carbonate

17

effect proposed by Appell.41 This gasification effect was crucial for deoxygenation, but at the

18

cost of carbon loss.13 The yield trends of HTL with alkaline catalysts were similar to that of other

19

studies conducted elsewhere.7, 14

20 21 22 23

10 ACS Paragon Plus Environment

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

1

Page 12 of 32

Table 1. A summary of product yields from various HTL and HDO conditions Whole-stage HTL a Catalyst Holding time (min)

0.94 M K2CO3 15

30

60

Second-stage HTL b

Whole-stage HTL-HDO c

Second-stage HTL-HDO d

N/A

0.94 M K2CO3

5% Ru/C

5% Ru/C

30

30

240

240

Product Yields,% Gas

9.33 (0.67)

36.03 (1.33)

41.30 (1.33)

11.00 (1.17)

36.90 (2.07)

34.76 (1.79)

32.91 (2.25)

Solid

8.03 (0.00)

6.67 (0.67)

5.33 (0.67)

15.89 (1.28)

7.47 (1.60)

4.01 (0.70)

7.42 (1.84)

Biocrudeoil

17.30 (0.67)

25.30 (1.33)

21.33 (0.67)

10.38 (1.48)

37.83 (2.32)

27.81 (2.46)

41.70 (0.67)

Aqueous

65.34 (0.00)

32.00 (0.67)

32.04 (0.00)

62.80 (1.40)

17.83 (3.84)

33.42 (4.95)

17.97 (3.40)

2 3 4 5 6

Note: a, yield of whole-stage HTL was calculated on the basis of dried raw sweet sorghum bagasse; b, yield of second-stage HTL was calculated on the basis of first-stage HTL solid product; c, yield of wholestage HTL-HDO was calculated on the basis of solid and liquid products from whole-stage HTL; d, yield of second-stage HTL-HDO was calculated on the basis of solid and liquid products from second-stage HTL.

7

The effect of holding time on the biocrude oil yield of whole-stage HTL with alkaline catalyst

8

was also studied. Biocrude oil yield increased by 46% from 15 to 30 min residence time and then

9

decreased to 16% from 30 to 60 min. The maximum biocrude oil yield of 25% was obtained at

10

30 min. A prolonged retention time suppressed the biocrude oil yield as reported by Boocock19

11

and Zhu13. The decline in biocrude oil yield with longer time could be due to the condensation

12

and repolymerization reactions within the biocrude oil.23 Although the detailed mechanism of

13

biocrude oil decomposition are still unclear, our study confirmed an increase in gas yield with

14

longer reaction time. The result further confirmed that the use of 0.94 M K2CO3 had prevented

15

the repolymerization reaction between hemicellulose decomposition intermediates and lignin

16

unit derived from raw sweet sorghum bagasse. Therefore, the HTL condition of 350 °C, 30 min

11 ACS Paragon Plus Environment

Page 13 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

1

with 0.94 M K2CO3 was the most suited conditions for maximum biocrude oil yield and thus was

2

used as a basis for comparing the second-stage HTL to produce biocrude oil from lignin-rich

3

sweet sorghum bagasse.

4

The lignin-rich sweet sorghum bagasse had 94% increase in lignin and 18% increase in

5

cellulose content as compared with that of raw sweet sorghum bagasse. Table 1 summarized the

6

HTL products yield for second-stage HTL of lignin-rich feedstock. The biocrude oil yield of

7

second-stage HTL was increased by 49% as compared with that of whole-stage HTL. The

8

second-stage HTL biocrude oil yield of 38% was obtained from hemicellulose-free sweet

9

sorghum bagasse and the yield was higher than reported elsewhere for pine wood,17 birth wood,12

10

sawdust42 and cornstalks10. The second-stage HTL process also reduced the aqueous product

11

yield by 44% compared with that of whole-stage HTL. In addition, it was found that there was

12

no significant solid yield difference between the second-stage HTL and whole-stage HTL, which

13

confirmed that the potential formation of char through cross-reaction and repolymerization

14

between hemicellulose and lignin intermediates was suppressed from whole-stage HTL with the

15

presentence of K2CO3.17, 18

16

Table 2. Major compositions of biocrude-oil from whole-stage HTL Catalytic (K2CO3) whole-stage HTL condition Holding time (min)

15

Compounds

30

60

Mass%

Cyclic ketones

74.57

56.75

58.48

Cyclopentanone, 2,5-dimethyl-

4.23

2.78

4.72

2-Cyclopenten-1-one, 2,3-dimethyl-

5.11

2.51

1.64

2-Cyclopenten-1-one, 2,3,4-trimethyl-

31.57

21.95

15.72

2-Cyclopenten-1-one, 2,3,4,5tetramethyl-

33.66

29.51

36.40

12 ACS Paragon Plus Environment

Energy & Fuels 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 32

Phenolics

18.62

35.83

35.30

Phenol

0.88

1.66

-

Phenol, 2-methyl-

0.00

2.43

0.99

Phenol, 4-methyl-

3.21

5.21

3.06

Phenol, 2,4-dimethyl-

-

-

13.59

Phenol, 3,5-dimethyl-

8.54

10.65

-

Phenol, 2-ethyl-4-methyl-

1.84

1.46

2.16

Phenol, 2-ethyl-5-methyl-

-

3.38

-

Phenol, 4-ethyl-3-methyl-

-

-

4.05

1.29

2.52

-

-

-

3.67

2.86

8.51

7.78

6.81

7.43

6.22

4.61

7.43

5.07

1,3-Dimethyl-1-cyclohexene

0.93

-

-

Toluene

0.65

1.21

0.55

1H-Indene, 2,3-dimethyl-

2.22

1.32

2.24

-

4.05

1.56

0.81

0.85

0.71

2.20

-

1.15

Pentadecane

1.20

-

0.72

Heptadecane

1.00

-

0.43

Phenol, 2-methyl-5-(1-methylethyl)Thymol Other phenolics Hydrocarbons Aromatic hydrocarbons

1H-Indene, 2,3-dihydro-1,6-dimethylOther aromatic hydrocarbons Alkanes

1 2

3.1.2 HTL Fuel Properties. The elemental composition of biocrude oil from HTL experiments

3

is summarized in Table 3. In the presence of 0.94 M K2CO3, about 43% of carbon in the sweet

4

sorghum bagasse was converted into biocrude oil after 30 min, which is 2.45 times more than

5

that of non-catalytic biocrude oil yield. In absence of K2CO3, a major fraction of the carbon was

6

converted into carboxylic acid and ends up in the aqueous fraction of the biocrude oil.

13 ACS Paragon Plus Environment

Page 15 of 32 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Energy & Fuels

1

The HHVs of biocrude oils ranged from 29.10 MJ/kg to 32.88 MJ/kg (Table 3), which almost

2

doubled that of the sweet sorghum bagasse feedstock of 16.30 MJ/kg.

3

Table 3. Fuel properties of biocrude oils obtained from different HTL conditions. Whole-stage HTL

Catalyst

4 5

0.94 M K2CO3

Secondstage HTL

Wholestage HTLHDO

Secondstage HTLHDO

Noncatalyst

0.94 M K2CO3

5% Ru/C

5% Ru/C

Benchmark Petroleum43

Holding time (min)

15

30

60

30

30

240

240

C [%]

72.81 (0.25)

69.62 (0.34)

73.96 (0.41)

69.13 (0.29)

79.30 (0.53)

86.87 (2.12)

88.22 (2.28)

83~86

H [%]

7.46 (0.02)

7.24 (0.01)

7.66 (0.08)

6.86 (0.06)

8.37 (0.06)

9.06 (0.31)

9.11 (0.20)

11~14

N [%]

1.36 (0.06)

1.35 (0.02)

1.30 (0.02)

1.24 (0.02)

0.63 (0.03)

1.28 (0.02)

0.61 (0.03)