Comparative Evaluation of Thermochemical Liquefaction and

ARTICLE pubs.acs.org/EF

Comparative Evaluation of Thermochemical Liquefaction and Pyrolysis for Bio-Oil Production from Microalgae Umakanta Jena and K. C. Das* Biorefining and Carbon Cycling Program, Department of Biological and Agricultural Engineering, The University of Georgia, Athens, GA 30602, United States

bS Supporting Information ABSTRACT: Bio-oil is the liquid product of thermochemical liquefaction or pyrolysis of biomass. Thermochemical liquefaction (TCL) is a low temperature (250
350 °C) and high pressure (5
20 MPa) process particularly suited for high moisture feedstocks, whereas pyrolysis is accomplished at moderate to high temperatures (400
600 °C) and atmospheric pressure and requires drying of the feedstock. In this paper, we present experimental results that provide a critical comparison of TCL and slow pyrolysis processes for producing bio-oil from algae. TCL experiments were performed in a 1.8-L Parr reactor using algae slurry (80% moisture) and pyrolysis runs were carried out in an 8-L mild steel cubical reactor, using dried algal powder as received (∼4% moisture). Yields and composition of bio-oil, char, gases, and aqueous phase were evaluated and compared for TCL and pyrolysis. TCL resulted in higher bio-oil yields (∼41%), lower char yields (∼6.3%), and lower energy consumption ratio compared to pyrolysis, which resulted in 23
29% bio-oil, and 28
40% solids yields. Bio-oil obtained from TCL was found to have higher energy density and superior fuel properties such as thermal and storage stabilities, compared to pyrolysis bio-oil.

1. INTRODUCTION Pyrolysis and thermochemical liquefaction (TCL) are two important thermochemical conversion processes that convert the organic constituents in biomass into a liquid fuel commonly referred to as “bio-oil” or “biocrude”.1
3 Bio-oil has an upper edge over solid fuels and syngas because of its higher energy density,4 and it is easier to transport and store than gaseous products. Bio-oil has the potential to be used as a fuel oil substitute and can be used in engine, turbine, and burner applications.5,6 They can also be upgraded to superior quality biofuels using current technologies such as hydrodexygenation, catalytic cracking, emulsification, and steam reforming.3,7 In addition to the above qualities, bio-oil is biodegradable and CO2/green house gas (GHG) neutral;8 hence, sustainable production of bio-oil has been of significant research interest for last two decades. In both pyrolysis and TCL, the organic constituents of biomass breakdown into smaller components in excess hydrogen (H+), produced either from the biomass or from water, that further undergo a series of depolymerization and repolymerization reactions to form liquids in an oxygen free environment.1,2 The major differences between TCL and pyrolysis are as follows: (1) TCL occurs at 5
20 MPa (725
2900 psi) pressure and low temperature (250
350 °C) whereas pyrolysis proceeds at near atmospheric pressure (0.1
0.5 MPa, i.e. 14.5
72.5 psi) and 400
600 °C temperature.1,3 (2) TCL can convert high moisture biomass (>80% moisture) without water evaporation and using the excess water in biomass as a highly reactant medium. This eliminates the need for drying, whereas pyrolysis requires complete drying.1,2 (3) TCL results in a higher yield of bio-oil whereas pyrolysis, in most cases (except fast pyrolysis), leads to the production of higher quantities of solid char than other products.1
3 r 2011 American Chemical Society

Microalgae are among the aquatic biomass feedstocks that are considered to be one of the best sources of liquid fuels. They can accumulate lipids that can be converted into biofuels and have drawn significant attention for research and business ventures.9 Algae have higher photoconversion efficiency (in the range 3
8%, compared to 0.5% for terrestrial plants), higher oil productivity (can accumulate 10
70% lipids), the ability to grow in water rather than land, and the ability to sequester CO2 from the atmosphere.10,11 In the literature, various conversion technologies including thermochemical conversions such as pyrolysis and TCL have been suggested for production of biofuel oil and gas from algae.12,13 Slow pyrolysis is the heating of biomass at slow heating rates (5
80 °C min
1)14 and longer residence times (5
30 min)8 compared to high heating rate (1000 °C min
1) and short residence times (10
20 s) in case of fast pyrolysis.13 Although pyrolysis has been extensively reported for a wide variety of lignocellulosic biomass including different species of wood,8 there is not much information about laboratory scale pyrolysis of algal biomass except the ones on fast pyrolysis of microalgae Chlorella protothecoides and Microcystis aeruginosa15 and the slow pyrolysis of Chlorella protothecoides.16 Bio-oil yields of 17.5% and 23.7% were reported from the fast pyrolysis of C. protothecoides and M. aeruginosa, respectively. The bio-oil yield from the slow pyrolysis of C. protothecoides was temperature dependent and was 5.7
55.3%. The fuel properties of bio-oils from algae in the above studies were found superior to that obtained from wood. TCL of microalgae biomass has been widely reported to result in higher bio-oil yields.17
19 Bio-oil yields vary from species to species and with TCL operating Received: September 12, 2011 Revised: October 17, 2011 Published: October 17, 2011 5472

dx.doi.org/10.1021/ef201373m | Energy Fuels 2011, 25, 5472–5482

Energy & Fuels

ARTICLE

conditions and have been reported as 22
43% for Nanochloropsis sp.17 and 18
46% for Spirulina platensis.19 Although an overall comparison of the fuel properties of bio-oil from TCL and pyrolysis were reviewed,1,3 there has not been a single study reported on a side-by-side comparative evaluation of yields and Table 1. Composition of Raw Algae Biomass Samples (as Received Basis wt %) proximate analysis moisture volatiles

4.54 ( 0.32% 79.14 ( 0.20%

ashes

6.56 ( 0.24%

fixed carbon

15.24 ( 0.41%

HHV (MJ kg
1)

20.52 ( 0.23%

element composition C

46.16 ( 0.79%

H

7.14 ( 0.20%

N S

10.56 ( 0.14% 0.74 ( 0.21%

O (by difference)

35.44 ( 0.31%

chemical content carbohydrates

30.21 ( 0.26%

lipids

13.30 ( 1.50%

proteins

48.36 ( 0.50%

ashes

6.56 ( 0.24%

fuel properties of bio-oil obtained from TCL and pyrolysis on a single feedstock. Considering the ongoing research attention centered on liquid fuel production from biomass and rapid developments in the fields of pyrolysis and TCL, further information is needed. In the present study, we provide a comparative assessment of the bio-oil yield and qualities from a microalgal feedstock using slow pyrolysis and TCL. To maintain uniform residence times for TCL and pyrolysis, slow pyrolysis was adopted in this study. The goals of the present study were to (1) compare the process yields and product distribution from TCL and pyrolysis of S. platensis and (2) make a detailed evaluation of bio-oil fuel properties and storage characteristics. Other co-products such as solid char, gases, and the aqueous phase are also evaluated and reported in this paper.

2. EXPERIMENTAL SECTION 2.1. Feedstock. Microalga S. platensis was shipped vacuum packaged in dry powder form (4
6% moisture content) from Earthrise Nutritional, LLC (Calipatria, CA) and was stored in a dry cool ventilated place until further use. Initial testing including ultimate, proximate, and biochemical composition analyses was done following the standard laboratory procedures described in section2.3 and presented in Table 1. 2.2. TCL and Pyrolysis Processes and Methods. Slow pyrolysis and TCL experiments were conducted in Bioconversion Laboratory at The University of Georgia, Athens, using batch reactors (Figure 1). For TCL runs, a slurry with 20% solids was prepared by mixing 150 g of algae (dry basis) with 600 g of deionized water and was

Figure 1. Experimental setup for (a) thermochemical liquefaction and (b) pyrolysis runs. 5473

dx.doi.org/10.1021/ef201373m |Energy Fuels 2011, 25, 5472–5482

Energy & Fuels processed in a 1.8-L stirred reactor (Parr Instruments Co., Moline, PA) (Figure 1a). The reactor was purged thoroughly and pressurized to 2 MPa (∼290 psi) with nitrogen gas. TCL runs were completed in two steps: (i) heating the reactor and contents to 350 °C (corresponding water pressure was 20.6 MPa (∼3000 psi)) with an electrical heater at a heating rate of ∼3.3 °C min
1 and (ii) providing an isothermal reaction step, where the temperature was held at 350 °C ((1 °C) for 60 min. Separation of bio-oil, solids (char), and aqueous phase was accomplished by a series of filtration and acetone washing, as detailed in our previous study.20 Gaseous products were collected in a Tedlar sample bag for further analysis before the reactor was cooled. Pyrolysis of algae was performed at two different temperatures, 350 and 500 °C, in a nitrogen atmosphere using an 8-L unstirred mild steel reactor with internal dimensions of 0.20  0.20  0.20 m. The reactor had two ports each having an 0.013-m internal diameter (i.d.) for the introduction of nitrogen gas and the removal of evolved gases and vapors and arrangements for a thermocouple for measuring biomass temperature and for measuring nitrogen flow during the experiment (Figure 1b). Pyrolysis runs were started by placing the reactor in a programmable muffle furnace. In a typical run, 500 g of dry algae sample was loaded into the reactor and sealed airtight. A ceramic mesh was placed on the surface of the biomass to prevent biomass powders from flowing out along with the gas. Pyrolysis runs were completed in two steps: (i) a nonisothermal step of heating the sample at ∼3.5 °C min
1 and ∼7 °C min
1 for the reaction temperatures of 350 °C and 500 °C, respectively, and (ii) an isothermal reaction step, where the reaction was held at the desired temperature for 60 min. The nitrogen gas flow rate was maintained at 0.25 L min
1. Following the 60 min reaction time, the furnace was turned off and the reactor was allowed to cool to room temperature. The vapors released during the process were condensed in a series of stainless steel traps that were placed in an ice bath. Gas samples were collected at the vent at the end of pyrolysis runs. The total liquids collected in the condensers were combined and weighed. Condensed liquids were separated into two phases (oily phase in the top called bio-oil and aqueous phase called “water-solubles”) using gravity separation in a separatory funnel. Solid residues remaining in the reactor were collected and weighed, which gave the solid char yield. Bio-oil and water-solubles obtained from different runs of pyrolysis and TCL experiments were stored in firmly closed glass bottles in darkness at 4 °C for further analysis. Treatments for TCL and pyrolysis are represented as TCL350, Pyro350, and Pyro500 for the rest of the paper. 2.3. Analysis Methods. Yields of different products were determined from the ratio of the weight of the respective product to the initial weight of biomass and were expressed as percentage yields. Bio-oil yields were reported on a moisture free basis. One-way analysis of variance (ANOVA) was performed for bio-oil yields. Significant differences between the treatments are reported at α = 0.05 level of significance. Yield data were analyzed from five experimental runs for all the treatments. Methods used to measure different physicochemical properties and composition of bio-oil, gaseous products, water-solubles, and solid char are described as follows. Elemental C, H, N, S, and O were analyzed by ASTM D-5291 and D-3176 methods, using a LECO brand (model CHNS-932) elemental analyzer. The analyzer was calibrated using sulfamethazine (C, 51.78%; H, 5.07%; N, 20.13%; and S, 11.52%) as the standard material. Moisture, volatiles, ash, and fixed carbon of solid and bio-oil samples were measured using a LECO TGA-701 proximate analyzer (Leco Corp., St. Joseph, MI) following the ASTM D-5142 method. Carbohydrate contents of algal biomass were analyzed by using the DuBois method;21 protein contents were approximated by multiplying elemental N concentrations by a factor of 4.58,21 while lipid contents were analyzed by a gravimetric method described in detail elsewhere.19 The concentrations of metals in algal biomass and bio-oil samples were analyzed using an

ARTICLE

inductively coupled plasma (argon) spectrometer (ICP) equipped with a mass spectrometer (MS) detector system.22 Higher heating values (HHV) of algae, bio-oil, and char samples were measured using an isoperibol bomb calorimeter (Parr, model 1351) following the ASTM D-5865 and D-4809 methods. The bomb calorimeter was calibrated using benzoic acid as the standard material. Dynamic viscosities of biooil samples were determined using a Brookfield DV-I+ viscometer with an UL/YZ spindle adapter. The method used was a modified version of ASTM D-2983, using higher temperatures than the standard because of the very high viscosity of bio-oil at low temperature. A circulating bath using ethylene glycol as coolant was used to maintain constant temperatures from 10 to 60 °C with an accuracy of (1 °C. Viscosity measurements were replicated three times and showed an average value with precision of (3.5%. Most samples were analyzed for viscosity at 40 °C and 60 °C. However, one stability analysis method involved measuring viscosity at 40 °C before and after an accelerated aging procedure. Densities of bio-oil samples were obtained from the specific gravity data. Specific gravity (g mL
1) was determined gravimetrically by laboratory methods using a 2-mL Gay-Lussac pycnometer. The pH of bio-oil and solid samples was measured with a portable digital pH meter (AP62, Thermo Fisher Scientific, Accumet AP 62). For the measurement of the pH of solid chars, finely ground samples were suspended in deionized water using a 1:100 (w/w) ratio.23 Samples were thoroughly mixed in a vibratory shaker and allowed to equilibrate for 1 h prior to measuring the pH. Gas chromatography (GC) was used to identify key chemical compounds in bio-oil samples, water-solubles, and gaseous products. GC-MS analyses were performed by a Hewlett-Packard (Model HP6890) gas chromatograph using a Hewlett-Packard mass spectrometer (model HP-5973) with a mass selective detector and a 30 m length  0.25 mm i.d. HP-5 MS column. A sample size of 1 μL was injected at 230 °C inlet temperature, 280 °C detector temperature, and at a He flow rate of 1 mL min
1. Oven temperature was maintained at 40 °C for 2.5 min followed by a ramp at 8 °C min
1 to 250 °C and held for 5 min. The mass spectrometer scan range was from 15 to 500 mass units, and compounds were identified using the mass spectral library (NIST 98). For bio-oil analysis, samples were prepared by diluting bio-oil to 2.5% with acetone (v/v). For analysis of water-solubles, samples were prepared by adding 10% of water-soluble products to a solvent mixture made of acetone and methanol at a 1:1 (v/v) ratio. Gaseous products were analyzed in a portable micro-GC (Agilent Technologies G2858A), having a thermal conductivity detector (TCD), a molecular sieve column (5A PLOT) of 10 m  0.32 mm size, and using He as the carrier gas. The GC was calibrated using a refinery gas calibration mix (Agilent part no. 5184
3543). Hydrocarbons in the gas samples were analyzed by the Hewlett-Packard GC-MS, using 150 μL of sample at a He flow rate 1.51 mL min
1 and holding it for 1 min at 200 °C oven temperature with initial oven temperature 50 °C (held for 2.5 min). The mass spectrometer scan range was from 15 to 500 mass units for analyzing the gas samples. FT-IR of the bio-oil and solid char samples was performed using a Varian model Scimitar 2000 (Palo Alto, CA) to identify the structural groups. All samples were analyzed in triplicate in the range 3600
600 cm
1. Solid samples were analyzed using KBr as transparent pellets. The stability of algal bio-oil samples was assessed by analyzing their storage behavior, thermal stability, and oxidative stability. Storage behavior was determined by measuring dynamic viscosity at regular time intervals until 120 days and, then, taking a final measurement at the 270th day of storing the bio-oil at room temperature (21 °C). The thermal stability of bio-oil was determined by measuring viscosity changes during an accelerated aging procedure.24 In this procedure, preweighed bio-oil samples were heated in a forced air oven at 80 °C for 24 h. Then, the samples were allowed to cool at room temperature and were weighed again to assess any weight change due to loss of volatiles. 5474

dx.doi.org/10.1021/ef201373m |Energy Fuels 2011, 25, 5472–5482

Energy & Fuels

ARTICLE

Viscosities of the samples were measured at 40 °C prior to and after aging (heating in the oven). Changes in viscosity were used as indicators of thermal stability. Oxidative stability of bio-oil was determined by measuring oxidation onset temperature (OOT). OOT is the temperature at which oxidation (combustion) begins with higher OOT, indicating greater stability. The oxidation onset test for the algal bio-oil samples was performed using the ASTM E-2009 method. The OOTs were measured using a differential scanning calorimeter (DSC), by placing approximately 3 mg of sample into the measuring cell of the DSC and heating the sample from 25 to 350 °C at 10 °C min
1 in an ultrahigh purity oxygen atmosphere (50 mL min
1). The OOT was the temperature at which combustion began (marked by the exothermic peak) and was determined as the point of intersection obtained by extending the tangent line from the peak and the baseline of the DSC curves, as described by Roger et al.24 The water-solubles were analyzed for formates, acetates, ethanol, and propionates by a high performance liquid chromatograph (LC-20 AT, Shimadzu Corp., U.S.A.), using an RID-10A refractive index detector and a 7.8  300 mm Corezel 64-H transgenomic analytical column. About 5 μL of the centrifuged and filtered samples were injected into the column using the LC-20 AT Shimadzu autoinjector. The samples were analyzed at an eluent (4 mN H2SO4) flow of 0.6 mL min
1 for 30 min retention time at 1000 psi pressure and 60 °C internal oven temperature. Calculations of Energy and Mass Balance. Energy recovery (ER) and energy consumption ratio (ECR) are generally used to express the process energy balance.25 ER was calculated from the yield data and expressed as percentage of the total energy in bio-oil as ER ¼

HHV of bio-oil  mass of bio-oil  100 HHV of raw feedstock  mass of raw feedstock

ð1Þ

The ECR was estimated from the experimental results and was defined as ECR ¼

energy required in TCL or pyrolysis process available energy of bio-oil produced

ð2Þ

Calculations of ECR for TCL and pyrolysis processes were done under the assumptions that algal biomass was available from the cultivation system at 80% initial moisture before thermochemical conversion. This is based on our desire to compare the conversion processes only, and the energy inputs in algae cultivation and harvesting would be the same in both cases. Also, calculations were done based on the production of 1 kg of bio-oil from the required amount of algal slurry in all cases. Energy required for TCL was calculated according to the method of Minowa et al.,26 as follows: ETCL ¼ WCpw ðT
20Þ þ ð1
WÞCps ðT
20Þ

ð3Þ

and, ECRTCL ¼

WCpw ðT
20Þ þ ð1
WÞCps ðT
20Þ rYH0

ð4Þ

where W is the moisture content, Cpw and Cps are the average specific heats of water and dry solid, T is the process temperature, r is the efficiency of available combustion energy, Y is the bio-oil yield, and H0 is the HHV of algal biomass. Cpw, Cps, and r were taken as 4.18 kJ kg
1 K
1, 4.18 kJ kg
1 K
1, and 0.6, respectively.26 The energy required for pyrolysis was calculated by adding the energy required for evaporating excess water (product of mass of water and latent heat of vaporization, 2260 kJ kg
1) to achieve dry algae biomass (at 5% moisture content) to eq 3 and was defined as Epyro ¼ ðMw Lw Þ þ WCpw ðT
20Þ þ ð1
WÞCps ðT
20Þ

ð5Þ

and, ECR pyro ¼

ðMw Lw Þ þ WCpw ðT
20Þ þ ð1
WÞCps ðT
20Þ rYH0 ð6Þ

where Mw and Lw are mass of water to be evaporated and latent heat of evaporation for water, respectively. Mass balances in TCL and pyrolysis were expressed in two ways: (1) carbon and hydrogen recovery (CHR), and (2) nitrogen disposition in products. CHR was calculated from the elemental composition of bio-oil and raw feedstock and was defined as CHR ð%Þ ¼

weight of C and H in bio-oil  100 weight of C and H in the raw feedstock

ð7Þ

Disposition of nitrogen in various products was defined as follows: N-disposition ð%Þ ¼

weight of N in products fraction  100 weight of N in the raw feedstock

ð8Þ

3. RESULTS AND DISCUSSION 3.1. Product Yield and Distribution in TCL and Pyrolysis Processes. Product yields obtained in TCL and pyrolysis

experiments are shown in Figure 2. The ANOVA test on the bio-oil yields showed that, for all the treatments (TCL and pyrolysis), yields were significantly different from each other (p < 0.0001, F = 123.30). TCL resulted in 40.7% bio-oil and 6.7% solids yield compared to 23.8
28.2% bio-oil and 25.6
39.7% solids yield for pyrolysis. As expected, pyrolysis performed at a lower temperature (350 °C) resulted in a higher yield of solids (39.7%), a lower yield of bio-oil (23.8%), and a lower yield of non-condensable gases (19.2%), compared to 25.6% solids yield, 28.5% bio-oil yield, and 28.0% gaseous yield for pyrolysis performed at the higher temperature of 500 °C. Generally, the bio-oil yield from pyrolysis is known as a function of pyrolysis temperature and heating rate.3,8 Pyrolysis of wood at low temperature (9), in contrast to the acidic pH (2
3.8) of bio-oils obtained from most lignocellulosic biomass.8 The alkaline pH could be due to the higher protein content (49.2%) of the algal biomass (Table 1). Generally, biomass with a protein constituent, such as poultry litter, were also reported to yield bio-oil with higher pH.27 The density of algal bio-oil obtained from TCL (0.97 kg L
1) was comparatively lower than that obtained from pyrolysis (1.05
1.20 kg L
1) (Table 2). The results obtained from the pyrolysis in our study were comparable to the results reported in the literature for bio-oil from the fast pyrolysis of microalgae.15 The density of algal bio-oil in our study was less than that of wood bio-oil (1.20 kg L
1) and was close to the density of fossil oil (0.75
1.00 kg L
1). The viscosity of bio-oil in our study was measured at two temperatures, 40 and 60 °C. Viscosity values for bio-oil samples obtained from TCL were 189.8 cP and 51.2 cP at 40 and 60 °C, respectively, compared to viscosity values of 100.7 cP and 34.3 cP for bio-oil obtained from pyrolysis at 350 °C (Pyro350) and 79.2 cP and 23.1 cP for bio-oil obtained at 500 °C (Pyro500), respectively (Table 2). In general, viscosity of algal bio-oil in our study was reported to be higher than the bio-oil obtained from pyrolysis of wood and was higher than the viscosity of petroleum crude (23.0 cP) and was in good agreement with earlier studies on fast pyrolysis of C. protothecoides.15 Viscosity is one of the key properties for use of bio-oil as a fossil fuel substitute. Higher viscosity of bio-oil will require more robust fuel filters and injectors for engine use and hence will need suitable upgrading. The elemental C, H, N, S, and O analyses (Table 2) show that bio-oil from TCL350 had lower nitrogen content (6.30%) than that from pyrolysis (7.13
10.71%). High nitrogen contents in algal bio-oil are due to the biomass composition, dominated by the presence of chlorophyll and proteins (Miao et al., 2004).15 Bio-oil from high temperature pyrolysis (Pyro500) had an

Table 2. Physical Properties, Ultimate Analysis, Inorganic Elements of Algal Bio-Oil Samples, and Energy and Mass Balance in TCL and Pyrolysis Processes conversion processes properties

a

TCL350

Pyro350b

Pyro500

color

black

reddish brown

reddish brown

odor

smoky

acrid smoky

acrid smoky

pH

9.60

9.35

9.52

F,c kg L
1

0.97

1.20

1.05

51.20 189.80

34.30 100.67

23.10 79.20

C, %

73.73

67.52

74.66

H, %

8.90

9.82

10.57

N, %

6.30

10.71

7.13

S, %

0.90

0.45

0.81

O, %

10.17

11.34

6.81

H/C ratio

1.44

1.73

1.68

O/C ratio HHV,d MJ/kg

0.10 34.21

0.13 29.30

0.06 33.62

viscosity, cP at 60 °C at 40 °C

Inorganics in Ash, mg kg
1 Na

14.6

14.6

14.0

Mg

69.3

11.3

2.3

Al

60.1

58.8

10.7

Si

115

54.8

15.5

P

249

63.2

39.6

Ca

116

35.4

7.6

Fe Ni

848 72.1

135 6.4

180 24.1

ER,e %

67.9

33.9

46.7

ECRf

0.70

2.11

1.56

CHR,g %

71.7

39.3

51.3

Energy and Mass Balance

TCL: thermochemical liquefaction. b Pyro: pyrolysis. c F: density. d HHV: higher heating values. e ER: energy recovery. f ECR: net energy ratio. g CHR: carbon and hydrogen recovery. a

5476

dx.doi.org/10.1021/ef201373m |Energy Fuels 2011, 25, 5472–5482

Energy & Fuels

ARTICLE

Table 3. Main Components of the Algal Bio-Oils Obtained from TCL and Pyrolysis Processes, as Identified by GC-MS area, % RT,a min

a

compd

TCL350b

Pyro350c

Pyro500

0.92

1.06

0.63

0.59

3.38

pentanenitrile

3.77

4-penten-2-one, 4-methyl-

3.93

pyrrole

0.77

1.09

4.05

toluene

2.00

6.78

11.90

4.76

3-penten-2-one, 4-methyl-

2.24

0.59

5.75

butanenitrile, 3-methyl-

1.61

2.01

6.16

ethylbenzene

2.36

1.01

2.41

6.85 9.55

styrene 2(1H)-pyridinone, 4-hydroxy-1,6-dimethyl-

1.86

0.46 26.94

1.57 13.98

9.81

phenol

1.83

0.50

0.91

10.52

benzene, (2-methylpropyl)-

0.70

0.41

0.41

11.76

4-piperidinone, 2,2,6,6-tetramethyl-

14.20

benzenepropanenitrile

0.93

15.25

1H-indole, 6-methyl-

1.67

2.24

18.32

pentadecane

1.14

1.03

1.7965

19.80 21.22

hexadecane heptadecane

7.78

0.45 5.65

0.7147 8.17

23.00

2-hexadecene, 3,7,11,15-tetramethyl-, [R-[R*,R*-(E)]]-

2.23

1.00

1.78

23.62

E-6-octadecen-1-ol acetate

1.72

2.99

23.89

hexadecanenitrile

6.64

7.45

24.17

pentadecanoic acid, 14-methyl-, methyl ester

24.72

hexadecanoic acid

26.19

9,12-octadecadienoic acid (Z,Z)-, methyl ester

1.67

0.87

27.25 29.24

hexadecanamide 2(1H)-naphthalenone, octahydro-4a-methyl-7-(1-methylethyl)-

6.95 5.19

12.23 2.12

17.43 2.69

31.77

pyrrolidine, 1-(1-oxopentadecyl)-

2.61

0.96

3.34

0.75

0.37

1.44

18.14

RT: retention time. b TCL: thermochemical liquefaction. c Pyro: pyrolysis.

elemental composition similar to that obtained from TCL350, and it was superior to pyrolysis performed at low temperature (Pyro350). Generally, bio-oil obtained from low temperature pyrolysis run was characterized by higher elemental O, leading to lower energy content (29.30 MJ kg
1) than that from pyrolysis at 500 °C (33.62 MJ kg
1) and TCL performed at 350 °C (34.21 MJ kg
1). The elemental composition and energy content of bio-oil obtained from S. platensis, in our study, were similar to that of the bio-oil obtained from fast pyrolysis of heterotrophic C. protothecoids.15 The energy content of bio-oil from C. protothecoids was reported as 30
41 MJ mg
1. The O/C ratio of algal bio-oil in our study was lower and the H/C ratio was higher, compared to that of bio-oil obtained from pyrolysis of wood and other lignocellulosic feedstocks.8 The sulfur content of algal bio-oil was less than 1% in all cases, compared to 0.05
5.0% in fossil oil. Major inorganic elements analyzed for algal bio-oil were Ni, Fe, Ca, P, Si, Al, Mg, and Na and are presented in Table 2. In general, algal bio-oil reported higher amounts of inorganic elements compared to that of wood pyrolysis oil as reported by Mohan et al., 2006.8 Bio-oil obtained from TCL had higher amounts of inorganic elements than that from pyrolysis, which was likely because TCL is a high pressure process and leads to more intense reactions. Also, the liquids and solids generated in TCL were allowed to remain in the pressurized reactor (Figure 1a) until complete cooling before separation of the products; therefore,

greater amounts of inorganic elements could have leached from solids ending up in the liquids/bio-oil. In contrast, in pyrolysis, the bio-oil vapor was collected in a set of condensers, leaving the solids inside the main reactor (Figure 1b), as the reaction further proceeded. The results of GC-MS analysis reveal that algal bio-oil is an extremely complex mixture of numerous compounds. Among the compounds identified in Table 3, the main components of the bio-oil included aromatic hydrocarbons, N-heterocyclics, amides, amines, carboxylic acids, esters, ketones, and straight chain hydrocarbons. A higher abundance of heterocyclic N compounds in all bio-oil samples is attributed to the chlorophyll and protein content of the biomass. In general, bio-oils generated from pyrolysis, especially at 350 °C, were characterized by an abundance of higher percentages of nitrogenous compounds and aromatic heterocycles, as compared to those from TCL, which was characterized by higher abundance of straight chain compounds, indicating that the TCL bio-oil would be easier to upgrade. Algal bio-oil generally reported fewer oxygenated compounds than bio-oils obtained from lignocellulosic biomass. Different kinds of biomass take different reaction pathways in pyrolysis, resulting in complexity and differences in bio-oil composition. FT-IR of the bio-oil samples presents normalized absorbance plotted against wavelengths (Figure 3). Bio-oils obtained from TCL and pyrolysis have similar spectra, suggesting the presence 5477

dx.doi.org/10.1021/ef201373m |Energy Fuels 2011, 25, 5472–5482

Energy & Fuels

ARTICLE

Figure 3. Infrared spectra of bio-oil samples obtained from TCL and pyrolysis processes.

Table 4. Stability Analysis of Bio-Oils Obtained from TCL and Pyrolysis Processes OOTa

viscosity increase treatments

%

rank

°C

rank

b

TCL350

72.88 ( 1.74

1

190.87 ( 3.23

1

Pyro350c

232.67 ( 1.52

3

121.22 ( 3.23

3

Pyro500

118.36 ( 6.37

2

155.41 ( 3.49

2

a

OOT: onset oxidation temperature. b TCL: thermochemical liquefaction. c Pyro: pyrolysis.

of the same functional groups. However, samples from Pyro350 and Pyro500 had higher peaks at certain bandwidths, compared to those of TCL350, suggesting a higher abundance of particular compounds. The spectra in the wavelength 3150
3500 cm
1 are assigned to O
H functional groups and represent hydroxy groups. A distinct band at about 3300 cm
1 for bio-oil from pyrolysis runs corresponds to N
H functional groups and represents a higher abundance of nitrogenous compounds than that of TCL bio-oil. Band widths of 2850
3100 cm
1 represent C—H stretching vibrations for alkyl derivatives. Lower peaks in bio-oil from TCL350 at 1670 cm
1 (CdO functional groups) represent less abundance of carboxylic acids, esters, or aryl ketones and polar components (CdC) in the bandwidth 1450
1670 cm
1 than that of pyrolysis bio-oils. The stretching band at 1165 cm
1 in all samples corresponds to functional groups CH3, CH2, and CH. 3.3. Bio-Oil Stability Assessment. Table 4 presents the results of thermal and oxidative stability of bio-oils from TCL and pyrolysis of algae. There was no significant change in weight of samples during the accelerated aging process, as confirmed by weight measurements before and after aging (data not shown). Bio-oil from TCL showed the best thermal stability characteristics, followed by bio-oil from Pyro500 and Pyro350. Bio-oil from Pyro350 showed the worst thermal stability characteristics in accelerated aging and had a 232.7% increase in viscosity within

24 h (Table 4). OOT evaluated for oxidative stability of bio-oils from TCL and pyrolysis showed that TCL bio-oil had the highest OOT (190.8 °C) and ranked first compared to 121.2 °C for biooil from Pyro350 and 155.4 °C for bio-oil from Pyro500. The higher OOT of the TCL bio-oil indicates better stability relative to bio-oils from pyrolysis processes at 350 and 500 °C. The pattern of change in the viscosities of the bio-oil samples stored over 270 days and measured at 60 °C is shown in Figure 4. The change in the viscosities of bio-oils during storage is also known as aging and could be due to the polymerization and condensation reactions of the constituents to form carbonyls, ethers, and esters.28 In the initial 50 days, the viscosity change was rapid for all bio-oil samples and was characterized by a steep slope. However, TCL bio-oil showed better storage stability over the pyrolytic bio-oil and was characterized by the plateauing of this curve at the end of 90 days, whereas viscosities of bio-oils from Pyro350 and Pyro500 continued to increase until 120 days. In summary, bio-oil generated from TCL had better stability characteristics than pyrolysis oils, and the bio-oil obtained at 350 °C had the least stability and led to gum formation, as described earlier in 3.2. 3.4. Gaseous Products Analysis. More than 90% of the gaseous products obtained from TCL and pyrolysis runs could be identified by GC. Gaseous products were analyzed for determination of volumetric abundance of N2, H2, CO, CO2, and other hydrocarbons (HC) in the range C1
C4 (CH4, C2H6, C3H8, C4H10, and C3H4). Nitrogen was the dominant gaseous component for both TCL and pyrolysis, as both processes were performed in nitrogen atmosphere. Gases from TCL had 89% N2, 4
6% CO2, 0.2
0.3% H2, a negligible amount of CO, and 1.0
1.5% HC gases, whereas gases from pyrolysis had 75
83% N2, 7.0
8.5% CO2, 0.37
0.73% H2, 1.2
2.5% CO, and 8.2
10.5% HC gases. Gases from Pyro350 were characterized by higher percentages of N2 and CO2, and by lower percentages of CO, H2, and HC than that of higher temperature pyrolysis (Pyro500). Higher hydrocarbons in gases (higher than C5) are presented in Table 5. The majority of the higher hydrocarbons in the gases consisted of N-heterocycles, substituted furans, 5478

dx.doi.org/10.1021/ef201373m |Energy Fuels 2011, 25, 5472–5482

Energy & Fuels

ARTICLE

Figure 4. Storage stability characteristics (viscosity variation) of algal bio-oil obtained from TCL and pyrolysis processes.

Table 5. Compositions of Hydrocarbons in the Gaseous Products Identified by GC-MS

Table 6. HPLC Analysis of Water-Solubles in the Aqueous Phase Obtained from TCL and Pyrolysis Processes concentration, g L
1

area, % RTa 1.37

a

TCL 350b

compd cyclopentene

1.43

propanenitrile

1.50 1.56

2-butanone furan, 2-methyl-

1.84

cyclopentene, 4-methyl-

1.92

1-pentanamine

1.95

benzene

2.17

1-heptene

2.26

heptane

2.28

heptane

2.83 3.18

1H-pyrrole, 1-ethyltoluene

3.82

styrene

4.74

ethylbenzene

4.88

p-xylene

0.95

5.31

1-butylpyrrolidine

2.47

5.91

2-pyrrolidinone, 1-propyl-

RT: retention time. pyrolysis.

10.81 2.76

Pyro350c

Pyro500

treatments

0.32

1.20

TCL350b

5.45

5.38

c

Pyro350

9.28 ( 0.32

16.16 ( 0.07

BDL

2.84 ( 0.03

9.13 22.64

12.26 4.31

Pyro500

10.67 ( 0.88

29.12 ( 0.09

BDL

2.14 ( 0.04

1.49 3.62

10.63 6.85

1.45

0.35

2.44

1.44 1.90

28.63

55.00

3.42 3.85

b

2.99

1.97

TCL: thermochemical liquefaction.

c

2.03 ( 0.00

acetate 3.36 ( 0.00

BDL: below detection limit. Pyro: pyrolysis.

b

ethanol 0.80 ( 0.00

propionate BDLa

TCL: thermochemical liquefaction.

4.47 6.76

1.42

a

formate

c

Pyro:

toluene, styrene, xylene, benzene and substituted benzenes, as well as aliphatic C-chains. The abundance of aromatic ring compounds in gases from pyrolysis runs was higher than that of TCL runs (Table 5). 3.5. Characteristics of Water-Soluble Products. Watersolubles generated from TCL were pale yellow to light brown in color and were characterized by a light smoky odor, whereas water-solubles from pyrolysis were deep brown in color and had a strong smell. HPLC analysis of the water-solubles from both TCL and pyrolysis indicate the presence of formate, acetate, ethanol, and propionate. Water-solubles from TCL had lower abundance of formate and acetate (2.03 g L
1, and 3.36 g L
1, respectively) compared to Pyro350 (9.28 g L
1 and 16.16 g L
1) and Pyro500 (10.67 g L
1 and 29.12 g L
1), respectively

(Table 6). These results suggest that both TCL and pyrolysis processes proceeded via hydrolysis, decarboxylation, and deoxygenation reactions. This argument is further supported by the GC-MS identification of organic compounds in the watersolubles (Figure 5) that are characterized by the presence of substituted acetate compounds such as acetic acids and acetamides. Also, the identification of several amides and N-heterocycles suggests that both TCL and pyrolysis of algae proceeds through the scission of peptide linkages in protein and the subsequent condensation/repolyemrization of amide polymers. 3.6. Properties of Solid Char. Table 7 shows various properties of solids obtained from TCL and pyrolysis processes. Solid char obtained from pyrolysis had higher percentages of volatiles, higher fixed carbon, and lower ash content, indicating lower organic conversion than in TCL. Solid char obtained from pyrolysis had higher energy values, 23.77 MJ kg
1 and 26.12 MJ kg
1, compared to 10.98 MJ kg
1 for the char from TCL. This confirms that most of the chemical energy in the original biomass was transformed into valuable products such as liquids and gases in TCL, whereas in pyrolysis most of them remained unconverted in the char. Between the two pyrolysis treatments, pyrolysis performed at 350 °C (Pyro350) and 3.5 °C min
1 produced a more energy dense char, with higher volatiles and lower ash contents, compared to pyrolysis at 500 °C (Table 7). Solid char obtained from TCL had almost neutral pH (6.77) compared to the alkaline pH values of 11.06 and 11.69 for chars obtained from pyrolysis at 350 and 500 °C, respectively. The FTIR spectra of the solid chars obtained from TCL350 and Pyro350 5479

dx.doi.org/10.1021/ef201373m |Energy Fuels 2011, 25, 5472–5482

Energy & Fuels

ARTICLE

Figure 5. GC-MS analysis of aqueous phase; (a) from TCL, and (b) from pyrolysis performed at 350 °C. The labeled peaks are (1) acetic acid, (2) pyrazine, methyl-, (3) acetamide, N-methyl, (4) acetamide, N-methyl, (5) 2-propanamine, N-(1-methylpropylidene)-, (6) 2-pyrrolidine, 1-methyl, (7) 2-pyrrolidinone, (8) piperdine, 1-butyl-.

Table 7. Analysis of Solid Char Obtained from Algae in TCL and Pyrolysis Processes param.

TCL350

Pyro350

Pyro500

VM,a % (db) FC,b % (db)

10.73 0.23

36.67 43.43

16.58 57.29

ash, % (db)

88.42

18.61

25.98

C, %

11.85

59.24

47.80

H, %

2.62

4.83

2.78

N, %

2.07

10.96

9.51

S, %

0.89

0.91

0.84

HHV, MJ/kg

7.98

26.12

23.77

pH

6.77

11.06

11.69

a

VM: volatile matter; db: dry basis. b Fixed carbon (FC) is calculated by difference: FC = 100
(ash + VM).

in Figure 6 were similar, indicating the presence of same functional groups. A distinct peak at band 1033 cm
1 was due to the C
O stretch for the O
CH3 and C
OH, and the peak

at wavelength 2920 cm
1 was assigned to the aliphatic CH compounds. However, solids obtained from the Pyro500 process showed a distinctly broader band in the wavelength range 2860
3000 cm
1 that was attributed to the CH bending vibration. Solid char obtained from pyrolysis of biomass is also named as “biochar” and has been a major area of research interest, as biochar is reportedly known to enhance the soil fertility, thereby increasing the crop productivity. The application of biochar helps in ameliorating soil conditions for crop production by increasing soil porosity, available soil
water content, organic-C, soil pH, available P, cation exchange capacity (CEC), exchangeable K, and Ca.23 Biochar alkalinity is a key factor in controlling the liming effect on acidic soils and can be a useful for the amendment of acidic soils. Solid chars obtained in our study had similar a pH and carbon content to that of biochar obtained from other lignocellulosic feedstocks.29 3.7. Energy and Mass Balance. ER and ECR were calculated for TCL and pyrolysis using eqs 1
6 and are presented in Table 2. (Detailed calculations for ECR are given in the Supporting Information). About 67.9% of the energy in the 5480

dx.doi.org/10.1021/ef201373m |Energy Fuels 2011, 25, 5472–5482

Energy & Fuels

ARTICLE

Figure 6. Infrared spectra of solids char samples obtained from TCL and pyrolysis processes.

Figure 7. Nitrogen disposition in TCL and pyrolysis (expressed as percentage weight distribution of initial nitrogen in dry algae biomass).

original biomass was recovered in bio-oil in TCL350, whereas pyrolysis processes reported low ER values of 33.9% for Pyro350 and 46.7% for Pyro500. TCL reported an ECR value of 0.70, whereas ECR values for Pyro350 and Pyro500 were 2.11 and 1.56, respectively. The ECR value of TCL (