Electronic Supplementary
Taguchi method and grey relational analysis to improve in-situ production of FAME from sunflower and Jatrophacurcas kernels with subcritical solvent mixture
SylvianaSutantoa,*, Alchris Woo Goa,c,*, SuryadiIsmadjib,**and Yi-Hsu Jua,**
a Department of Chemical Engineering, National Taiwan University of Science and Technology, 43, Keelung Rd., Sec. 4, Taipei 106-07, Taiwan
bDepartment of Chemical Engineering, Widya Mandala Surabaya Catholic University, Kalijudan 37, Surabaya 60114, Indonesia
c Department of Chemical Engineering, University of San Carlos – Talamban Campus, Nasipit, Talamban, Cebu City, 6000, Philippines
*The first two author have equal contributions to this work
** Corresponding author
** Corresponding author +886-2-27376612
ABSTRACT
This study investigates the possibility of employing in situ (trans)esterification (ISTE) under subcritical condition (200 to 250°C) of solvent mixture (methanol + acetic acid)with high solid loading and low solvent to solid ratio (SSR). Taguchi method together with grey relational analysis was used to optimize both FAME yield and productivity. It was found that temperature reaction time and SSR were factors which contributed the most in obtaining high FAME yields. In addition to the above mentioned factors the addition of acetic acid also significantly improved the productivity. From the optimization method employed the following conditions: 250°C; 8.8 MPa; 3 to 7 mL/g SSR; 10% acetic acid were found to provide high FAME yield and productivity. Confirmatory test resulted in FAME yield of 87.5 to 92.7% for sunflowerkernels and 88.2 to 97.22% for Jatrophacurcas L. kernels and productivity up to 37.5 kg/m3/h can be obtained with good repeatability. Furthermore, the process developed in this study can tolerate moisture and free fatty acid content up to 25%. The direct application of the method using whole kernels was also investigated.
Keywords:biodiesel, Jatrophacurcas, subcritical methanol, subcritical acetic acid,sunflower oil, (trans)esterification
Table S1. Levels of variables and responses of in-situ transesterification of sunflower kernels
Experiments / Parameters / ResponseAcetic Acid (v/v) / SSR
(mL/g) / Time
(min) / Temp
(C) / Yield (kg/kg) / Productivity (kg/m3/h)
A / B / C / D / FAME (%) / SD
(%) / S/N db) / FAME / SD / S/N (db)
1,2 / 0 / 3 / 0 / 200 / 2.42 / 0.55 / 7.34 / 3.66 / 0.82 / 10.93
3,4 / 0 / 5 / 30 / 225 / 22.97 / 2.11 / 27.17 / 10.30 / 0.92 / 20.21
5,6 / 0 / 7 / 60 / 250 / 84.87 / 1.35 / 38.57 / 18.46 / 0.21 / 25.33
7,8 / 5 / 3 / 30 / 250 / 65.91 / 6.54 / 36.31 / 39.71 / 3.85 / 31.92
9,10 / 5 / 5 / 60 / 200 / 33.94 / 0.91 / 30.61 / 11.38 / 0.34 / 21.12
11,12 / 5 / 7 / 0 / 225 / 9.89 / 0.49 / 19.89 / 6.10 / 0.29 / 15.70
13,14 / 10 / 3 / 60 / 225 / 62.06 / 1.00 / 35.86 / 28.91 / 0.37 / 29.22
15,16 / 10 / 5 / 0 / 250 / 25.80 / 2.99 / 28.15 / 17.33 / 2.09 / 24.68
17,18 / 10 / 7 / 30 / 200 / 27.96 / 2.62 / 28.87 / 10.62 / 0.97 / 20.47
CT1a / 10 / 3 / 60 / 250 / 87.23 / - / - / 37.54 / - / -
CT2 / 10 / 3 / 60 / 250 / 87.71 / - / - / 37.83 / - / -
CT3 / 10 / 3 / 60 / 250 / 86.34 / - / - / 36.99 / - / -
Average / 10 / 3 / 60 / 250 / 87.09 / 0.70 / 38.79 / 37.46 / 0.42 / 31.47
a CT-confirmation test runs
Table S2. Summary of grey relational analysis forin-situ transesterification of sunflower kernels
Factors/Levels / S/N
Ratio / Normalized response / Difference Sequence / GRG / Ave.
GRG
Yield / Productivity / x1* / x2* / D1 / D2 / y1 / y2 / y
A1 / 24.36 / 18.82 / 0.3566 / 0.1684 / 0.6434 / 0.8316 / 0.4373 / 0.3755 / 0.4064
A2 / 28.94 / 22.91 / 0.6330 / 0.5690 / 0.3670 / 0.4310 / 0.5767 / 0.5371 / 0.5569
A3 / 30.96 / 24.79 / 0.7551 / 0.7534 / 0.2449 / 0.2466 / 0.6712 / 0.6697 / 0.6705
B1 / 26.50 / 24.02 / 0.4860 / 0.6782 / 0.5140 / 0.3218 / 0.4931 / 0.6084 / 0.5508
B2 / 28.64 / 22.00 / 0.6152 / 0.4799 / 0.3848 / 0.5201 / 0.5651 / 0.4902 / 0.5276
B3 / 29.11 / 20.50 / 0.6436 / 0.3327 / 0.3564 / 0.6673 / 0.5838 / 0.4283 / 0.5061
C1 / 18.46 / 17.10 / 0.0000 / 0.0000 / 1.0000 / 1.0000 / 0.3333 / 0.3333 / 0.3333
C2 / 30.79 / 24.20 / 0.7447 / 0.6953 / 0.2553 / 0.3047 / 0.6620 / 0.6214 / 0.6417
C3 / 35.01 / 25.22 / 1.0000 / 0.7955 / 0.0000 / 0.2045 / 1.0000 / 0.7097 / 0.8549
D1 / 22.27 / 17.51 / 0.2305 / 0.0396 / 0.7695 / 0.9604 / 0.3939 / 0.3424 / 0.3681
D2 / 27.64 / 21.71 / 0.5545 / 0.4513 / 0.4455 / 0.5487 / 0.5288 / 0.4768 / 0.5028
D3 / 34.34 / 27.31 / 0.9597 / 1.0000 / 0.0403 / 0.0000 / 0.9253 / 1.0000 / 0.9627
Min / 18.46 / 17.10 / 0.00 / 0.00 / Average / 0.5735
Max / 35.01 / 27.31 / 1.00 / 1.00
Range / 16.56 / 10.20 / 1.00 / 1.00
Table S3.Product yields for in situ transesterification of ground SF kernels in subcritical methanol acetic acid mixture at 250°C for 1 h.
Kernel(g) / SSR
(mL/g) / AA
(%) / FAME / FFA / MG / DG / TG / Total
21 / 7 / - / 85.84 / 3.94 / 5.46 / 2.91 / 3.12 / 101.27
21 / 7 / 6.25 / 92.75 / 5.01 / 4.71 / 2.23 / 1.62 / 106.33
42 / 3 / 6.25 / 83.79 / 14.00 / 2.70 / 0.58 / 0.67 / 101.75
Table S4.In situ (trans)esterification of JC kernels in sub or supercritical solvents.
Space Loading (mL/g)a / SSR(mL/g) / Co-solvent/
Catalyst / Utilized Reactor Volume (mL) / Temp (C)/
Pressure
(MPa) / Time (h) / Extraction efficiency (%)d / Yield (%)e
(Kernel)f / Ref
54.0 / 5 / - / ~52 (11.56%)b / 300/9.5 / 0.5 (2.0)c / ~109 / 99.67 (66.6) / [11]
30.0 / 5.9 / 2.0 MPa CO2 / ~108
(24%) / 300/20.0 / 0.21 (1.71) / 101.21 / ~92 (~61.5) / [12]
12.0 / 7 / 25% AA / ~165
(68.8 %) / 250/10.0 / 1.0 (1.75) / 99.29 / 88.24 (51.14) / [13]
12.0 / 7 / 25% AA /
2.0 MPa CO2 / ~165
(68.8 %) / 250/21.0 / 1.0 (1.75) / 99.29 / 94.43 (54.72) / [13]
6.0 / 3 / 10% AA / ~165
(68.8 %) / 250/9.0 / 1.0 (1.75) / 99.86 / 88.15 (47.97) / This study
12.0 / 7 / 10% AA / ~165
(68.8 %) / 250/10.5 / 1.0 (1.75) / ~109 / 97.22 (52.90) / This
study
aSpace loading = effective reactor volume in mLper gram dry solid
b Fraction of the reactor volume utilized.
c Total reaction time including time required to heat up reactor to the desired temperature
d Mass fraction (%) of hexane soluble product based on hexane extractable lipid
e Mass fraction (%) of FAME based on hexane extractable lipid
f Mass fraction (%) of FAME based on dry kernel
Table S5.Applicability of ISTE to whole SF kernels(Single experimental runs are carried out in this part of the study as a preliminary for utilizing whole kernels).
Run / Space Loading (mL/g)a / SSR(mL/g) / Co-solvent/
Catalyst / Water
(mL/g)b / Utilized Reactor Volume (mL) / Temp (°C) /
Pressure
(MPa) / Time (h) / Extraction efficiency (%)e / FAME Yield (%)f
1 / 6.0 / 3 / 10% AA / 0.3 / ~181
(75 %)c / 250/9.5 / 1.0 (1.75)d / 72.82 / 58.05
2 / 6.0 / 3 / 25% AA / 0.3 / ~181
(75 %) / 250/8.5 / 1.0 (1.71) / 71.49 / 54.52
3 / 12.0 / 7 / 10% AA / 0.3 / ~171
(71 %) / 250/11.5 / 1.0 (1.75) / 89.72 / 69.52
4 / 12.0 / 7 / 10% AA /
2.0 MPa CO2 / 0.3 / ~171
(71 %) / 250/21.5 / 1.0 (1.75) / 87.98 / 71.93
5 / 12.0 / 7 / 10% AA / 0.3 / ~171
(71 %) / 250/11.5 / 1.5 (2.25) / 88.73 / 75.79
6 / 12.0 / 7 / 25% AA / 0.3 / ~171
(71 %) / 250/10.5 / 1.0 (1.75) / 96.92 / 77.68
7 / 12.0 / 7 / 25% AA / 1.0 / ~185
(77 %) / 250/14.5 / 1.0 (1.75) / 92.33 / 73.02
8 / 12.0 / 7 / 25% AA / 0 / ~165
(68.8 %) / 250/9.0 / 1.0 (1.75) / 89.20 / 72.80
9 / 12.0 / 7 / 25% AA / 0.3 / ~171
(71 %) / 250/10.5 / 1.5 (2.25) / 93.73 / 78.06
aSpace loading = reactor effective volume/dry solid weight
b Volume of water added per gram of kernel
c Fraction of reactor volume utilized.
d Total reaction time including time to heat up the reactor
e Mass fraction (%) of hexane soluble product based on hexane extractable lipid
f Mass fraction (%) of FAME based on hexane extractable lipid
Table S6.Extraction efficiency of ISTE process.
Extraction Efficiency /Product Recovery / Yield
FAME / FFA / MG / DG / TG / Total
Ground kernels: SSR = 3 mL/g; AA = 10%; T = 250 °C; t = 1 h
ISTE / 87.47% / 13.98% / 4.13% / 0.42% / 0.00% / 106.00%
Solid residue / 0.34% / 0.51% / 0.90% / 0.17% / 0.13% / 1.47%
Whole kernels: SSR = 3 mL/g; AA = 10%; T = 250 °C; t = 1 h
ISTE / 58.05% / 12.07% / 2.38% / 0.14% / 0.17% / 72.82%
Solid residue / 17.25% / 3.65% / 0.76% / 0.05% / 0.05% / 21.76%
Whole kernels: SSR = 7 mL/g; AA = 10%; T = 250 °C; t = 1.5 h
ISTE / 75.79% / 9.68% / 2.67% / 0.40% / 0.20% / 88.73%
Solid residue / 10.32% / 1.47% / 0.47% / 0.05% / 0.03% / 12.34%
Whole kernels: SSR = 7 mL/g; AA = 25%; T = 250 °C; t = 1.5 h
ISTE / 78.06% / 12.86% / 2.47% / 0.34% / 0.00% / 93.73%
Solid residue / 7.97% / 1.34% / 0.43% / 0.05% / 0.02% / 9.81%
Table S7.Comparison of different in situ (trans)esterification of SF kernels.
Transesterification process / Lipid Content (%) / SSR(mL/g) / Temp (C)/
Pressure
(MPa) / Time (h) / Yield (%)b
(Kernel)c / Conversion (%) / Ref
Acid Catalyzed
(100 % H2SO4 base on oil wt.)
Stirred / 55.6 / 6.1 / 64.5/0.1 / 1.0 / 90.7 (~50.4) / - / [21]
Base Catalyzed
(2.0% NaOH base on oil wt.)
Ultrasound assisted (24 kHz) / - / 10 / 60/0.1 / 0.33 / - / ~95 / [6]
Base Catalyzed
(2.9% NaOH base on oil wt.)
Stirred (150 rpm)
DEM as Co-solvent
(0.57 DEM:MeOH) / ~46 / 7.6 / 20/0.1 / 0.22 / 94.1 (~45) / ~94 / [22]
Subcritical Solvent (Methanol+Acetic Acid)
6.25% AA in the Solvent / 45.2 / 7 / 250/9.0 / 1.0 (1.75)a / 92.8 (41.9) / 100.8 / This study
Subcritical Solvent (Methanol+Acetic Acid)
10% AA in the Solvent / 45.2 / 3 / 250/9.0 / 1.0 (1.75) / 87.1 (39.5) / 94.6 / This study
a Total reaction time including the time to heat up the reactor
b Mass fraction (%) of FAME based on the hexane extractable lipid
c Mass fraction (%) of FAME based on the dry kernel