Assessment of ultrasound-assisted extraction of crambe seed oil for biodiesel synthesis by in situ interesterification

Accepted Manuscript Assessment of ultrasound-assisted extraction of crambe seed oil for biodiesel synthesis by in situ interesterification Gilmar Roberto Tavares, Thainara Bovo Massa, José Eduardo Gonçalves, Camila da Silva, Wanderley Dantas dos Santos PII:

S0960-1481(17)30381-6

DOI:

10.1016/j.renene.2017.04.065

Reference:

RENE 8760

To appear in:

Renewable Energy

Received Date: 16 December 2016 Revised Date:

20 April 2017

Accepted Date: 28 April 2017

Please cite this article as: Tavares GR, Massa TB, Gonçalves JoséEduardo, da Silva C, dos Santos WD, Assessment of ultrasound-assisted extraction of crambe seed oil for biodiesel synthesis by in situ interesterification, Renewable Energy (2017), doi: 10.1016/j.renene.2017.04.065. 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.

ACCEPTED MANUSCRIPT

Assessment of ultrasound-assisted extraction of crambe seed oil for

2

biodiesel synthesis by in situ interesterification

3 4

Gilmar Roberto Tavaresa, Thainara Bovo Massab, José Eduardo Gonçalvesc, Camila da

5

Silvab∗∗ and Wanderley Dantas dos Santosa,d a

6

Departamento de Ciências Agronômicas, Universidade Estadual de Maringá (UEM), Estrada da Paca

7

s/n, Umuarama, PR, 87500-000, Brazil. b

8

RI PT

1

Departamento de Tecnologia, Universidade Estadual de Maringá (UEM), Avenida Ângelo Moreira da

9 10 11

SC

Fonseca 180, Umuarama, PR, 87506-370, Brazil. c

Programa de Mestrado em Tecnologias Limpas e Programa de Mestrado em Promoção da Saúde, Centro Universitário de Maringá, Av. Guedner 1610, Maringa, PR, 87050-900, Brazil.

d

Departamento de Bioquímica, Universidade Estadual de Maringá (UEM), Avenida Colombo 5790,

M AN U

12 13

Maringa, PR, 87020-900, Brazil.

14

Abstract: In this study, the effectiveness of the ultrasound-assisted extraction (UAE)

16

of crambe seed oil was evaluated with the aim of biodiesel synthesis by in situ

17

interesterification. An experimental design was applied, using a mixture of n-hexane

18

and methyl acetate as the solvent, to evaluate the effect of the process variables and

19

determine the conditions that maximize the removal of oil from the seeds. The results

20

indicated that the extraction time and temperature have a greater influence on the oil

21

extraction than the solvent:seed ratio (p<0.05). The extraction carried out at 60 °C

22

using a solvent to seed ratio of 10 (mL g-1) for 90 min provided the maximum oil yield

23

(~37%), representing ~92% of the yield obtained by the conventional method and a

24

20% increase compared with the yield obtained without ultrasound. The fatty acids

25

compositions of oils obtained by UAE and conventional extraction were similar

26

(p>0.05), showing a predominance of erucic, oleic and linoleic acids, which accounted

AC C

EP

TE D

15

∗

To whom correspondence should be addressed. Tel.: (55) 44-3621 9300. Fax: (55) 44-3621 9326. E-mail: [email protected]

10

ACCEPTED MANUSCRIPT for ~86.5% of the oil composition. The water and free fatty acids contents indicate that

2

the oils obtained cannot be processed by the conventional method using alkaline

3

homogeneous catalysis for the production of biodiesel.

4

Keywords: n-hexane, methyl acetate, Crambe abyssinica H., ultrasound.

RI PT

1

5 6

1. Introduction

The production of biodiesel requires the use of raw materials which are of low

8

cost and do not compete with the food sector. Crambe is an option that offers these

9

characteristics. Considered a winter crop, it has a short annual cycle (90-100 days) and a

10

high tolerance to drought [1]. In addition, the culture of crambe requires lower amounts

11

of fertilizer and water compared to other oil crops [2] and it is more resistant to pests

12

and diseases [3]. Seed yields are between 1000 and 1500 kg hectare-1. Crambe can be

13

used for culture rotation with soybeans, corn, wheat, among others and its cultivation

14

can reach four annual cycles [4,5].

M AN U

TE D

15

SC

7

Crambe seeds contain around 30 to 51% of oil [6-8].

Erucic acid is the

predominant component of the fatty acids composition (56-66%), which makes the oil

17

inappropriate for human consumption [9,10] but leads to a high stability and low

18

melting point [4,11].

AC C

19

EP

16

The process traditionally used in biodiesel production involves several different

20

manufacturing steps, including oil extraction and purification, separation and recovery

21

of the solvent and reaction. These processing stages represent 70 to 85% of the total

22

biodiesel production process [12-14]. Aimed at reducing production costs, an in situ

23

process has been proposed, which consists of the reactive extraction of seeds, that is,

11

ACCEPTED MANUSCRIPT 1

concomitant extraction and reaction directly in the oleaginous matrix in a single step

2

[15,16]. In the in situ process, the acyl acceptor acts as a solvent and reagent in the

4

presence of a catalyst, generating fatty acid esters, and the solvent extraction step is no

5

longer required. Among the acyl acceptors that can be used, methyl acetate is an

6

excellent solvent for oil extraction from oilseed [17,18]. Moreover, with the use of this

7

solvent has the interesterification reaction which produces triacetin as coproduct.

8

Triacetin can be added to biodiesel in proportions of up to 10% by weight, without

9

adversely affecting the quality of the fuel [19,20]. Also, it has a higher commercial

10

value than glycerol [19-21], the coproduct generated in the conventional

11

transesterification process with alcohol.

M AN U

SC

RI PT

3

The solvent n-hexane is commonly used for the extraction of vegetable oils and

13

it improves the mass transfer in reaction processes conducted with enzymatic catalysts

14

[22], promoting an increase in the enzyme activity. Thus, n-hexane can be used together

15

with methyl acetate in the in situ process. It is expected to provide high efficiency in the

16

extraction of oil and a high reaction rate for the process is obtained with enzyme

17

catalysts.

EP

TE D

12

Several different techniques can be applied for oil extraction, among which

19

ultrasound-assisted extraction (UAE) is an attractive option since high oil yields are

20

obtained from the vegetable matrix, with short extraction times and a low volume of

21

solvent. These benefits are due to cavitation (the formation, rise and implosion of

22

bubbles during the extraction) generated by the ultrasound, which disrupts the surface of

23

the solid matrix, enhancing mass transfer and accelerating diffusion [23-27]. This

24

method has shown good efficiency in obtaining vegetable oils from various sources [27-

AC C

18

12

ACCEPTED MANUSCRIPT 1

30]; however, ultrasound-assisted extraction does not appear to have been applied to

2

obtain crambe oil. Based on these aspects, the present study was performed to establishment of

4

conditions that maximize crambe seed oil extraction for subsequent in situ

5

interesterification assisted by ultrasound. Experiments were performed in an ultrasound

6

bath evaluating the effect of operating variables on oil yield. The results obtained were

7

compared with those provided by extraction without the application of ultrasonic and by

8

conventional Soxhlet extraction.

SC

RI PT

3

10

2. Materials and Methods

11

2.1. Materials

M AN U

9

The crambe seeds used for the oil extraction were obtained from Fundação MS

13

(cultivar FMS Brilhante) with 3.52±0.05% of humidity. Methyl acetate (Sigma Aldrich,

14

99% purity) and n-hexano (Panreac) were used as solvents. In the characterization of the

15

oil, methyl heptadecanoate (Sigma Aldrich, >99% purity), heptane (Anidrol), ethanol

16

(Anidrol), ethyl ether (Anidrol), sodium hydroxide (Nuclear), phenolphthalein

17

(Nuclear), methanol (BT Baker) and BF3-methanol (Sigma Aldrich) were used.

19 20

EP

AC C

18

TE D

12

2.2. Sample preparation The crambe seeds were subjected to mild grinding in a household blender

21

(Pratic Blender) to break their shells, which were later separated by directing

22

compressed air on the material. The ground seeds were classified according to the

23

standard series of Tyler sieves (Bertel, ASTM) and the fraction retained on a 28-mesh

13

ACCEPTED MANUSCRIPT 1

sieve was used in the experiments. The material was packed in a plastic bag and kept

2

under refrigerator until use.

3 4

2.3. Oil extraction The oil extraction was performed (in duplicate) using the conventional method

6

(Soxhlet) and the process assisted by ultrasound (UAE). In both cases, the oil yield was

7

calculated from Equation 1:

RI PT

5

SC

8

(1)

M AN U

9 10

Conventional extraction (Soxhlet) was performed according to the methodology

12

of the Adolfo Lutz Institute [31], applying an extraction time of 480 min and a seed to

13

solvent ratio of 30:1 (g mL-1). The extraction temperature was kept constant and above

14

the solvent reflux temperature in all runs with the use of an electric heating plate

15

(Logen). The extraction flask was attached to the condenser coupled to a refrigerated

16

bath (model MA 184, Marconi).

EP

TE D

11

The UAE was conducted in a Unique ultrasound bath (Q 5.9/25 A) with indirect

18

contact, power of 165 W and frequency of 25 kHz, using a round bottom flask with a

19

volume of 50 mL. The ultrasonic bath was operated at the desired temperature and

20

maximum power, and the temperature was maintained constant using a thermostated

21

recirculating water bath (Marconi, MA 184) connected to the ultrasonic device. In each

22

extract, 3 g of seeds were added to the flask together with methyl acetate (methyl

23

acetate to oil molar ratio of 12 (whereas oil content was obtained by the conventional

24

method) and n-hexane. The methyl acetate to oil molar ratio was determined

AC C

17

14

ACCEPTED MANUSCRIPT considering the results of Xu et al. [32], Du et al. [33], Jeong and Park [34] and

2

Maddikeri et al. [35], who reported better yields for interesterification under this

3

condition. After the addition of the substrates and enzyme, the flask was connected to a

4

condenser with water recirculation at a temperature of 10 °C (Marconi thermostated

5

bath, model MA 184) and placed in the ultrasonic bath. After the reaction time, the

6

seeds were removed by filtration using quality filter paper with a diameter of 15 cm and

7

weight of 80 g m-2.

RI PT

1

After the extraction period, the solvent remaining in the samples obtained by the

9

two methods was removed of the samples until constant weight, and the samples were

10

stored under refrigeration until analysis. The recovered solvent was reused in the

11

process for 5 cycles, showing no influence (p>0.05) in the oil yield.

M AN U

SC

8

To evaluate the influence of the experimental variables (time, temperature and n-

13

hexane to seed ratio) and determine the optimum levels to maximize the oil yield, a

14

Box-Behnken factorial experimental design with three levels was applied, including

15

four runs at the central point, generated by Statistica® 8.0 software. The levels

16

employed for the independent variables are shown in Table 1.

19

EP

18

Table 1

AC C

17

TE D

12

20

Analysis of variance was used to evaluate the effects of the independent variables on

21

the responses and experimental data were fitted to a second-order polynomial model

22

including the effect of the interaction of linear terms, using Statistica® 8.0 software.

23

The generalized model used is expressed by Equation 2.

15

ACCEPTED MANUSCRIPT (2) where β0, βi, βii and βij are the regression coefficients (β0 = constant term; βi = linear

2

effect; βii = quadratic effect; βij = linear interaction term) and Y is the response variable

3

(oil yield) observed in the experiments. Xi and Xj are the independent variables: time,

4

temperature and n-hexane to seed ratio.

RI PT

1

Based on the results of the experimental design, the extraction kinetics were

6

investigated with the aim of determining the equilibrium yield under these conditions.

7

In addition, a comparison with the extraction performed without the application of

8

ultrasound, in a Marconi orbital shaker (model MA 839a) using agitation of 40 rpm,

9

was carried out.

10

12 13

2.4. Oil Characterization

TE D

11

M AN U

SC

5

The FFA content was determined according to the AOCS method Ca 5a-40 (1990), and the water content was quantified using a Karl Fischer titrator (Orion, AF8). To determine the fatty acids composition, derivatization of the samples was

15

performed according to the AOAC method Ce 2-66 [36] (Walker, 1990). Subsequently,

16

samples were analyzed in a gas chromatograph (Agilent 7890 B) coupled to a mass

17

spectrometer (Agilent 5977 A) equipped with an Agilent HP-5MS capillary column (30

18

m x 0.250 mm x 0.25 µm) applying the following conditions: injection of 1.0 µL in split

19

mode 1:10, initial column temperature of 120 °C, held for 5 min, increasing to 180 °C at

20

a rate of 15 °C min-1 and then to 240 °C at a rate of 5 °C min-1, held for 5 min. The flow

21

of helium as the carrier gas was 1 mL min-1. The temperatures for the ionization source

22

and quadrupole mass analyzer were 230 and 150 °C, respectively. The compounds were

AC C

EP

14

16

ACCEPTED MANUSCRIPT 1

identified through comparison of their mass spectra with the NIST 11.0 library spectra

2

and quantified through analysis using methyl heptadecanoate as the internal standard.

3

2.5 Analysis of data All analysis was performed in duplicate and the data collected were subjected to

5

ANOVA using Excel® 2010 software and Tukey tests (with a 95% confidence

6

interval), to evaluate differences between the results.

3. Results and Discussion

9

3.1. Conventional extraction

M AN U

8

SC

7

RI PT

4

10

Conventional extraction (Soxhlet) resulted in an oil yield of 41.23±0.10%.

12

According to reports in the literature, crambe seeds contain 30-51% of oil by weight [7,

13

37-41]. Among other variables, the difference in the oil contents of crambe seed may be

14

related to the type of fertilizer applied to the soil, as observed in studies by Silva et al.

15

[42] and Soratto et al. [43].

EP

16

TE D

11

3.2. Ultrasound-assisted extraction

18

3.2.1 Effect of experimental variables

19

AC C

17

Table 2 shows the experimental conditions for the UAE of crambe seed oil

20

obtained using a Box-Behnken design, as well as the oil yields obtained under these

21

conditions. The effects of the operational variables analyzed are shown in Table 3.

22 23

Table 2 17

ACCEPTED MANUSCRIPT According to the statistical significance considered (p-value), it can be seen that

2

only the effects of the interactions between time and temperature and between solvent to

3

seed ratio and temperature are not significant (p>0.05). Time, temperature and solvent

4

to seed ratio showed positive linear correlations with the oil yield and thus an increase

5

in the magnitude of these variables results in an increase in the variable response.

6

Table 3

SC

7

RI PT

1

8

The extraction time is the variable with the greatest influence on the extraction

10

of oil by UAE, and the highest yields were obtained within 60 min. Other authors have

11

also observed an increase in oil yield with an ultrasound exposure time of up to 60 min

12

[29,30,44,45]. It can also be seen that higher yields of oil were obtained after 60 min

13

compared with the shorter extraction time of 20 min. This highlights the stages of the

14

extraction process, involving the washing of the particle surface with the solvent, which

15

easily removes accessible oil molecules, followed by the second stage where oil

16

extraction occurs by mass diffusion [44,46].

EP

TE D

M AN U

9

Higher temperatures result in an increase in the vapor pressure of the solutes

18

[47] and a decrease in the viscosity and density of the solvent [48,49]. They also

19

facilitate the disruption of the plant tissue [50]. The combination of these effects favors

20

the oil diffusion into the solvent [46,51]. Furthermore, an increase in the temperature in

21

ultrasound-assisted processes favors the collapse of cavitation bubbles [29], since the

22

vapor pressure has an important influence on the occurrence and intensity of acoustic

23

cavitation [47].

AC C

17

18

ACCEPTED MANUSCRIPT An increase in the temperature provided higher yields in the extraction of

2

crambe seed oil. A similar effect was reported by Hu et al. [52] and Barizão et al. [53]

3

for the ultrasound-assisted extraction of oil from safflower and pomegranate using n-

4

hexane, an increase from 30 to 60 °C favoring the oil extraction. Based on the

5

thermodynamic analysis of data from Cannabis sativa L. oil extraction using n-hexane,

6

Kostić et al. [54] reported negative values for the Gibbs free energy, indicating that the

7

extraction process is spontaneous and that increasing the extraction temperature from 20

8

to 70 °C favors this spontaneity.

SC

RI PT

1

However, for longer extraction times, an increase in the temperature had no

10

effect on the process (p>0.05), since at high temperatures the extraction equilibrium was

11

easily reached.

M AN U

9

The recovery of oil from crambe seed was favored on increasing the solvent to

13

seed molar ratio from 6 to 10. The use of a higher amount of solvent enhances the oil

14

extraction due to the increased concentration gradient between the solid and the liquid

15

[44,47,50,55]. The use of a high solvent to seed ratio could also enhance the diffusion

16

through a reduction in the viscosity.

TE D

12

Increasing the extraction efficiency through the use of larger volumes of solvents

18

has been observed in several studies reported in the literature. As an example, for the

19

extraction of papaya seed oil, Samaran et al. [50] reported an increase in the oil yield on

20

increasing the relative amount of solvent from 6:1 to 10:1 (solvent to seeds). In another

21

study by Gutte et al. [47], an increase in the flaxseed oil yield from 30 to 38% was

22

observed on increasing the solvent to seed ratio from 5:1 to 10:1.

AC C

EP

17

23

The use of high amounts of solvent and long reaction times provided higher oil

24

yields (p<0.05). However, the interaction between these variables and the temperature

25

had no effect (p> 0.05) on the oil yield. This finding could be explained by the fact that 19

ACCEPTED MANUSCRIPT 1

increasing the amount of solvent increases the solubility of oil in solvent, due to a

2

stronger driving force, which promotes mass transfer at elevated temperatures.

3

However, the reaching of equilibrium with excessive amounts of solvent is not feasible

4

[44,56].

6

RI PT

5

3.2.2 Second-order polynomial model

The values for the parameters shown in Table 3 were re-estimated considering

8

those for which the differences did not show statistical significance. From the regression

9

analysis data, it was found that the oil yield and the experimental variables are

11

M AN U

10

SC

7

correlated according to Equation 2 (p<0.05):

Y (%) = 33.70 + 1.52 X 1 + 1.28 X 2 + 1.23 X 3 − 0.51X 12 − 0.52 X 22 − 0.98 X 32 + 0,69 X 1 X 3 (3)

12

The validity of the predictive equation was verified applying the F-test based on

14

the analysis of variance (ANOVA) data. The F-test for regression showed that the

15

equation was able to represent the experimental data for the range of factors

16

investigated, because FCALC>FTAB (calculated from the ANOVA and tabulated data,

17

respectively). The statistical analysis results showed values of 65.73 and 3.5 for FCAL

18

and FTAB, respectively. The value for the correlation coefficient (0.98) and F regression

19

test results showed that the model was able to represent the experimental data in the

20

experimental range assessed. Figure 1 shows the strong correlation between the

21

experimental data and the values predicted by the model.

AC C

EP

TE D

13

22 23

Figure 1

24 20

ACCEPTED MANUSCRIPT 3.2.3 Optimization of extraction conditions From the predictive equation (Equation 3) it was possible to determine the

3

maximum yield which can be obtained within the experimental ranges tested for the

4

variables as shown in Figure 2. As shown in this figure, the predicted maximum yield

5

was 36.42%, obtained after 60 min, at 60 °C and with a seed to solvent ratio of 1:10 (g

6

mL-1). To confirm the predictive ability of the model, experiments were conducted

7

under the predicted conditions, which resulted in 36.35±0.69% of oil. From the

8

application of the Student t-test for averages, it can be observed that the predicted value

9

and the experimental result did not differ statistically (p>0.05).

10

SC

RI PT

2

M AN U

1

Figure 2

11 12

Figure 3 shows the contour plot constructed according to Equation 3, which

14

illustrates the correlation of independent variables with the variable response. Each

15

contour plot is a function of two variables, keeping the third at this central point. Figure

16

3 shows that the oil extraction is favored in the combination of the higher levels of the

17

considered variables. It was observed that the effect of the use of high amount of solvent

18

for long extraction time is more pronounced, as already presented in Table 3.

21

AC C

EP

TE D

13

22

3.2.4. Extraction kinetics

19 20

Figure 3

23

Applying the optimum conditions, a kinetics study on the ultrasound-assisted

24

extraction was carried out and the results are shown in Figure 4. Experimental data 21

ACCEPTED MANUSCRIPT obtained without the application of ultrasound are also reported in this figure. From

2

Figure 4 it can be seen that the oil yield increases with time (p <0.05) up to 90 and 120

3

min for UAE and conventional extraction, respectively. The results clearly show that the

4

application of ultrasound favors the oil extraction (p<0.05), with 37.5% yield in 90 min

5

of extraction compared with 33.4% yield in 120 min for the extraction without

6

ultrasound.

7

Figure 4

SC

8

RI PT

1

9

With the application of ultrasound Sicaire et al. [27], Zhang et al. [28], Lou et al.

11

[29], Mello et al. [30] and Rodrigues et al. [57] obtaining better yields in the extraction

12

of rapeseed, linseed, chickpea, chia and macauba pulp oils. The UAE allowed the target

13

components to dissolve faster in the solvent, thereby boosting the yield obtained in a

14

shorter time. This effect is observed because the ultrasonic wave disrupts the cell walls,

15

allowing a larger area of contact between the solvent and the materials [27,28].

TE D

M AN U

10

The maximum oil yield obtained applying UAE was 37.5±0.6% in 90 min,

17

which is lower than the yield obtained by the conventional method (41.23±0.1%).

18

However, it is noteworthy that the Soxhlet extraction time was 480 min and it was

19

carried out at the boiling temperature of n-hexane (68 °C) using a solvent to seed ratio

20

of 30 (mL g-1). To obtain greater removal of the oil from crambe seeds, future work can

21

be conducted by increasing the range of study variables in the UAE.

AC C

EP

16

22

The experimental conditions for crambe oil extraction have been minimally

23

explored in the literature. However, the results of this study can be compared to those

24

reported by Santos et al. [41] who used subcritical propane as the solvent. The authors

22

ACCEPTED MANUSCRIPT 1

report a maximum yield of 32.8% at 80 °C and 16 Mpa with 80 min of extraction and a

2

solvent to seed ratio of ~9 (mL g-1).

3

3.3. Oil Characterization

Table 3 shows the fatty acids composition and the water and free fatty acids

5

contents for the extracts obtained by Soxhlet and with ultrasound under of the maximum

6

yield conditions. The oils obtained presented water and free fatty acids (FFA) contents

7

above the limits for the application of the conventional method used for biodiesel

8

production, that is, homogeneous alkaline catalysis [58]. The oils obtained with the two

9

techniques showed similar water contents, but the oil obtained by the conventional

10

method had a higher FFA content. This method was carried out at the boiling

11

temperature of the solvent (68 ° C) and with a long extraction time (8 h), and these

12

conditions may favor the formation of FFA.

M AN U

SC

RI PT

4

13

Table 3

TE D

14 15

The extraction methods used showed no significant difference (p> 0.05) in terms

17

of the fatty acids composition, with erucic acid being the major fatty acid, followed by

18

oleic and linoleic acids. The fatty acids composition found is similar to that reported by

19

Santos et al. [41], Lalas et al. [59], and Zanetti et al. [60] who also observed the

20

predominance of erucic acid (63.77, 59.4 and 52.43%) followed by oleic (15.07, 20.17

21

and 16.84%) and linoleic (7.52, 13.06 and 9.11%) acids.

AC C

EP

16

22 23

4. Conclusions

23

ACCEPTED MANUSCRIPT The results of this study demonstrated the efficiency of UAE in the obtainment

2

of crambe oil, using a mixture of methyl acetate and n-hexane as the solvent. It was

3

found that in this technique an increase in the variables time, temperature and

4

solvent:seed ratio (within the experimental range tested) favors the removal of oil from

5

the seeds, the maximum oil yield being obtained at 60 °C with an extraction time of 90

6

min and solvent to seed ratio of 10. The UAE method provided ~92% of the oil yield

7

obtained with the conventional extraction method and an increase of ~ 20% in the oil

8

yield when compared to the process without ultrasound. The major fatty acids were

9

erucic, oleic and linoleic acids and the levels of these fatty acids were not influenced by

10

the extraction method; however, UAE provided oil with a lower FFA content. Data

11

obtained in this study can be used for the conduction of in situ interesterification using a

12

enzimatic catalyst, for example.

13

16

TE D

15

Acknowledgments

The authors are grateful to MS Foundation for the crambe oil donation and CNPq for financial support.

17

EP

14

M AN U

SC

RI PT

1

References

19

[1] T.R.B. Silva, A.C.S. Reis, C.D.G. Maciel, Ind. Crop. Prod. 39 (2012) 135–138.

20

[2] A.S. Carlsson, Biochimie. 91 (2009) 665-670.

21

[3] Pitol, C., Broch, D.L., Roscoe, R. (2010). Tecnologia e produção: crambe.

22

Maracaju: Editora Fundação MS. 60.

AC C

18

24

ACCEPTED MANUSCRIPT [4] Fundação MS, 2008. Crambe (Crambe abyssinica) – cultivar FMS Brilhante: uma

2

boa alternativa para produção de biodiesel. Boletim informativo. Disponível em:

3

http://www.fundacaoms.org.br/produtos/crambe. Access in: 20 sept (2016).

4

[5] Plein, G.S., Favaro, S.P., Souza, A.D.V., Souza, C.F.T., Santos, G.P., Miyahia,

5

M.A.M., Roscoe, R. (2010). Caracterização da fração lipídica em sementes de crambe

6

armazenadas com e sem casca. In: IV Congresso Brasileiro de Mamona e I Simpósio

7

Internacional de Oleaginosas Energéticas. 1, 1812-1816.

8

[6] O.H. Viana, A. Borsoi, R.F. Santos, K. Sanderson, J. Food Agric. Environ. 11

9

(2013) 721-723.

SC

RI PT

1

[7] B.K. Singh, M. Bala, P.K. Rai, Natl. Acad. Sci. Lett. 37 (2014) 219-226.

11

[8] A.G. Brandão, T.R.B. Silva, L.A.V. Henrique, J.S. Santos, F.M. Gonçalves, D.S.

12

Kohatsu, A.C. Gonçalves Jr, Afr. J. Agric. Res. 9 (2014) 927-930.

13

[9] W.T. Wazilewski, R.A. Bariccatti, G.I. Martins, D. Secco, S.M.N. Souza, H.A.

14

Rosa, L.I. Chaves, J. Food Agric. Environ. 43 (2013) 207-212.

15

[10] A.M. Maciel, C.C. Ming, A.P.B. Ribeiro, R.C. Silva, L.A. Gioielli, L.A.G.

16

Gonçalves, J. Am. Oil Chem. Soc. 91 (2014) 111-123.

17

[11] J.M. Vargas-Lopez, D. Wiesenborna, K. Tostensona, L. Cihacekc, J. Am. Oil

18

Chem. Soc. 76 (1999) 801–809.

19

[12] O.K. Choi, J.S. Song, D.K. Cha, J.W. Lee, Bioresource Technol. 166 (2014) 51–

20

56.

21

[13] F. Gunawan, A. Kurniawan, I. Gunawan, Y.H. Ju, A. Ayucitra, F.E. Soetaredjo, S.

22

Ismadji, Biomass Bioenerg. 96 (2014) 28-38.

23

[14] F. Zhu, L. Zhao, H. Jiang, Z. Zhang, Y. Xiong, J. Qi, Energy & Fuels 28 (2014)

24

5277−5283.

AC C

EP

TE D

M AN U

10

25

ACCEPTED MANUSCRIPT [15] K.G. Georgogianni, M.G. Kontominasa, P.J. Pomonisa, D. Avlonitisb, V. Gergisc,

2

Fuel Process. Technol. 89 (2008) 503–509.

3

[16] P. Hidalgo, G. Ciudad, M. Mittelbach, R. Navia, Fuel 153 (2015) 544–551.

4

[17] E. Su, W. Xu, K. Gao, Y. Zheng, D. Wei, J. Mol. Catal. B-Enzy. 48 (2007) 28-32.

5

[18] H. Wu, Y. Liu, J. Zhang, G. Li, Bioresource Technol. 174 (2014) 182–189.

6

[19] A. Go, Y. Lee, Y.H. Kim, S. Park, J. Choi, J. Lee, S.O. Han, S.W. Kim, C. Park,

7

Enzyme Microb. Tech. 53 (2013) 154-158.

8

[20] E.F. Aransiola, T.V. Ojumu, O.O. Oyekola, T.F. Madzimbamuto, D.I.O. Ikhu-

9

Omoregbe, Biomass and Bioenerg. 61 (2014) 276-297.

SC

RI PT

1

[21] A. Casas, M.J. Ramos, A. Pérez, Chem. Eng. J. 171 (2011) 1324-1332.

11

[22] L. Batistella, M.K. Ustra, A. Richetti, S.B.C. Pergher, H. Treichel, J.V. Oliveira, L.

12

Lerin, D. Oliveira, Bioproc. Biosyst. Eng. 35 (2012) 351–358.

13

[23] M.D. Esclapez, J.V. García-Pérez, A. Mulet, J.A. Cárcel, Food Eng. 3 (2011) 108–

14

120.

15

[24] V.G. Yachmenev, E.J. Blanchard, A.H. Lambert, Ultrasonics 42 (2004) 87–91.

16

[25] D. Yu, L. Tian, H. Wu, S. Wang, Y. Wang, D. Ma, X. Fang, Processes

17

Biochemistry. 45 (2010) 519–525.

18

[26] V.B. Veljkovic´, J.M. Avramovic, O.S. Stamenkovic, Renew. Sust. Energ. Rev. 16

19

(2012) 1193–1209.

20

[27] A.G. Sicairea, M.A. Viana, F. Fineb, P. Carréc, S. Tostaind, F. Chemat, Ultrason.

21

Sonochem. 31 (2016) 319–329.

22

[28] Z.S. Zhang, L.J. Wang, D. Li, S.S. Jiao, X.D. Chen, Z.H. Mao, Ind. Crop. Prod. 62

23

(2008) 192–198.

24

[29] Z. Lou, H. Wang, M. Zhang, Z. Wang, J. Food Eng. 98 (2010) 13-18.

AC C

EP

TE D

M AN U

10

26

ACCEPTED MANUSCRIPT [30] Mello, B.T.F., Garcia, V.A.S., Silva, C.(2016). Ultrasound-assisted extraction of

2

oil from chia (Salvia Hispânica L.) Seeds: optimization extraction and fatty acid profile.

3

J. Food Process Eng., DOI: 10.1111/jfpe.12298.

4

[31] Adolfo Lutz Institute. Métodos físico-químicos para análise de alimentos. 4. Ed.

5

Brasília: Anvisa (2008).

6

[32] Y. Xu, W. Du, D. Liu, J. Zeng, Biotechnol. Lett. 25 (2003) 1239–1241.

7

[33] W. Du, Y. Xu, D. Liu, J. Zeng, J. J. Mol. Catal. B-Enzym. 30 (2004) 125–129.

8

[34] G.T. Jeong, D.H. Park, Appl. Biochem. Biotech. 161 (2010) 195–208.

9

[35] G.L. Maddikeri, A.B. Pandit, P.R. Gogate, Fuel Process. Technol. 116 (2013) 241–

SC

RI PT

1

249.

11

[36] American Oil Chemists’ Society – AOCS (1990). Official methods and

12

recommended practices (Method AOCS Ca 5a-40), United States: Editora Champaign.

13

[37] S.L. Falasca, N. Flores, M.C. Lamas, S.M. Carballo, A. Anschau, Int. J. Hydrogen

14

Energ. 35 (2010) 5808–5812.

15

[38] S.Y. No, Renew. Sust. Energ. Rev. 15 (2011) 131-149.

16

[39] A.E. Atabani, A.S. Silitonga, H.C. Ong, T.M.I. Mahlia, H.H. Masjuki, I.A.

17

Badruddin, H. Fayaz, Sust. Energ. Rev. 18 (2013) 211-245.

18

[40] H.A. Rosa, W.T. Wazilewski, D. Secco, L.I. Chaves, G. Veloso, S.N.M. Souza,

19

M.J. Silva, Renew. Sust. Energ. Rev. 38 (2014) 651-655.

20

[41] K.A. Santos, R.A. Bariccatti, L. Cardozo-Filho, R. Schneider, F. Palú, C. Silva,

21

E.A. Silva, J. Supercrit. Fluid. 104 (2015) 54-61.

22

[42] T.R.B. Silva, R.F. Lavagnolli, A. Nolla, J. Food Agric. Environ. 9 (2011) 264-267.

23

[43] R.P. Soratto, G.D. Souza-Schlick, A.M. Fernandes, E.F.C. Souza, Rev. Bras.

24

Ciênc. Solo 37 (2013) 658-666.

AC C

EP

TE D

M AN U

10

27

ACCEPTED MANUSCRIPT [44] I.T. Stanisavljević, M.L. Lazić, V.B. Veljković, Ultrason. Sonochem. 14 (2007)

2

646-652.

3

[45] Q.A. Zhang, Z.Q. Zhang, X.F. Yue, X.H. Fan, T. Li, S.F. Chen, Food Chem. 116

4

(2009) 513-518.

5

[46] T.A. Toda, M.M. Sawada, C.E.C. Rodrigues, Food. Bioprod. Process. 98 (2016) 1-

6

10.

7

[47] K.B. Gutte, A.K. Sahoo, R.C. Ranveer, Oilseeds & Fats Crop and Lipids. 22

8

(2015) 2-7.

9

[48] S. Hemwimol, P. Pavasant, A. Shotipruk, Ultrason. Sonochem. 13 (2006) 543-548.

10

[49] Ramandi, N.F., Ghassempour, A., Najafi, N.M., Ghasemi, E.(2012). Optimization

11

of ultrasonic assisted extraction of fatty acids from Borago Officinalis L. flower by

12

central

13

10.1016/j.arabjc.2012.06.009.

14

[50] S. Samaram, H. Mirhosseini, C.P. Tan, H.M. Ghazali, S. Bordbar, A. Serjouie,

15

Food Chem. 172 (2015) 7-17.

16

[51] R.C.A. Amarante, P.M. Oliveira, F.K. Schwantes, J.A. Moron-Villarreyes, Ind.

17

Eng. Chem. Res. 53 (2014) 6824−6829.

18

[52] A.J. Hu, Q.Q. Feng, J. Zheng, X.H. Hu, C. Wu, C.Y. Liu, J. Food Process Eng. 35

19

(2012) 278-294.

20

[53] É.O. Barizão, J.S. Boeing, A.C. Martins, J.V. Visentainer, V.C. Almeida, Food

21

Anal. Method. 8 (2015) 2392–2400.

22

[54] M.D. Kostić, N.M. Joković, O.S. Stamenković, K.M. Rajković, P.S. Milić, V.B.

23

Veljković, Ind. Crop. Prod. 52 (2014) 679-686.

24

[55] Y. Tian, Z. Xuc, B. Zheng, Y.M. Lo, Ultrason. Sonochem. 20 (2013) 202–208.

design.

Arabian

Journal

of

Chemistry.

DOI:

AC C

EP

TE D

composite

M AN U

SC

RI PT

1

28

ACCEPTED MANUSCRIPT [56] M.D. Kostić, N.M. Joković, O.S. Stamenković, K.M. Rajković, P.S. Milić, V.B.

2

Veljković, Ind. Crop. Prod. 48 (2013) 133-143.

3

[57] G.M. Rodrigues, B.T.F. Mello, V. Garcia, C. Silva, J. Food Process. Eng. (2016),

4

DOI - 10.1111/jfpe.12530.

5

[58] F. Ma, M. Hanna, Bioresource Technol. 70 (1999) 1-15.

6

[59] S. Lalas, O. Gortzi, V. Athanasiadis, E. Dourtoglou, V. Dourtoglou, J. Am. Oil

7

Chem. Soc. 89 (2012) 2253–2258.

8

[60] F. Zanetti, D. Scordia, T. Vamerali, V. Copani, C. Dal Cortivo, G. Mosca, Ind.

9

Crop. Prod. 90 (2016) 152–160.

SC

AC C

EP

TE D

M AN U

10

RI PT

1

29

ACCEPTED MANUSCRIPT 1

Table 1. Factors and actual and coded levels used in the Box-Behnken design for UAE

2

extraction. Levels Low (-1)

Central (0)

High (1)

(X1) time (min)

30

45

60

(X2) temperature (°C)

40

50

60

8

10

(X3) solvent to seed ratio (mL g-1)

6

3

AC C

EP

TE D

M AN U

SC

4

RI PT

Factors

30

ACCEPTED MANUSCRIPT 1

Table 2. Experimental conditions applied and oil yield obtained in experiment to assess

2

the effects of the operating variables using a Box-Behnken design. X11

X21

X31

Y2(%)

1

-1

-1

0

30.20

2

1

-1

0

32.58

3

-1

1

0

31.95

4

1

1

0

5

-1

0

-1

6

1

0

-1

7

-1

0

1

8

1

0

9

0

-1

10

0

11

0

12

0

13

0

14

0

15

30.26

31.76

SC

31.26

35.56

-1

29.61

M AN U

1

-1

32.25

-1

1

32.18

1

1

34.74

0

0

33.91

0

0

33.58

0

0

0

33.93

0

0

0

33.39

EP

As in Table 1; Oil yield obtained by Equation 1.

AC C

3

2

35.94

1

TE D

16 1

RI PT

Run

31

ACCEPTED MANUSCRIPT 1

Table 3. Model coefficients and p-value of the model for the extraction of crambe seed oil by

2

UAE.

Statistical

error d

significanc

0.075

yes

p-value b

Coefficient c

Mean/Interaction

32.35

<0.001

32.35

X1 (L)

3.04

<0.001

1.52

X1 x X1

0.51

0.030

0.25

X2 (L)

2.57

<0.001

1.28

X2 x X2

0.52

0.028

0.26

0.065

yes

X3 (L)

2.46

<0.001

1.23

0.092

yes

X3 x X3

0.98

X1 x X2

0.80

X1 x X3

1.40

X2 x X3

-0.04

0.065

yes

0.092

yes

SC

yes

0.49

0.065

yes

0.054

0.40

0.131

no

0.012

0.70

0.131

yes

0.888

-0.02

0.131

no

TE D

Effect of the independent variable on the dependent variable; bstatistical significance p<0.05’; ccoefficients of

second-order polynomial model (Equation 3); derror associated with the estimated coefficient.

EP

5

a

0.092

0.004

AC C

3 4

RI PT

Effect a

M AN U

Standard

Variables

32

ACCEPTED MANUSCRIPT 1

Table 4. Fatty acids composition in crambe oil obtained using different extraction

2

methods.

3

Extraction method

Palmitic

2.16a±0.03

2.20a±0.01

Palmitoleic

0.08a±0.01

Stearic

1.06a±0.03

Oleic

19.77a±0.80

Linoleic

6.65a±0.05

Linolenic

1.44a±0.04

1.39a±0.06

Arachidonic

1.41a±0.15

1.46a±0.02

Gadoleic

4.59a±0.30

4.15a±0.10

0.90a±0.01

0.92a±0.01

2.76a±0.05

2.96a±0.03

59.18a±0.80

59.33a±0.09

1.89a±0.02

1.17b±0.18

0.34a±0.1

0.29a±0.15

Eicosadienoic Behenic Erucic Free fatty acids content (wt%)

0.09a±0.01 1.11a±0.07

19.67a±0.40 6.75a±0.14

Results in g 100 g-1 of oil. Means followed by same lowercase letters (comparison method) did not differ statistically (p>0.05).

EP

6

1

AC C

4 5

TE D

Water content (wt%)

RI PT

UAE

SC

Fatty acid1

Conventional extraction

M AN U

Property

33

ACCEPTED MANUSCRIPT List of Figure Captions

2

Figure 1. Correlation between the experimental and predicted oil yield (according to

3

second-order polynomial model - Equation 3) for UAE of crambe oil.

4

Figure 2. Optimization of operational variables, obtained from the Equation 3, for UAE

5

of crambe oil: X1 – time; X2 – temperature; X3: seed to solvent ratio.

6

Figure 3. Contour plot for oil yield as a function of: a) time and temperature; b) time

7

and seed to solvent ratio; c) seed to solvent ratio and temperature (all obtained for the

8

central condition).

9

Figure 4. Kinetics of extraction of crambe seed oil obtained at 60 °C with seed to

SC

solvent ratio of 1:10 (g mL-1).

AC C

EP

TE D

11

M AN U

10

RI PT

1

34

ACCEPTED MANUSCRIPT 37 36 35

RI PT

Predicted Values

34 33 32

SC

31

29 28 28

29

30

TE D EP

Figure 1

AC C

4

32

33

Observed Values

1 2

3

31

M AN U

30

35

34

35

36

37

ACCEPTED MANUSCRIPT X1

X2

X3

39.00

28.00

1

1

Levels of variables

1

Figure 2

-1

M AN U

2

-1

SC

-1

RI PT

Oil yield (%)

36.42

AC C

EP

TE D

3

36

1

ACCEPTED MANUSCRIPT 1.0

RI PT

X2

0.5

0.0

(a)

> 35 < 35 < 34 < 33 < 32 < 31 < 30

-1.0 -1.0

-0.5

M AN U

SC

-0.5

0.0

0.5

1.0

X1

1 2

EP

0.0

AC C

X3

0.5

TE D

1.0

(b)

-0.5

-1.0 -1.0

3 4

-0.5

0.0

X1

37

0.5

1.0

> 35 < 35 < 34 < 33 < 32 < 31 < 30

ACCEPTED MANUSCRIPT 1.0

RI PT

0.0

SC

X2

0.5

-1.0 -1.0

-0.5

TE D EP

Figure 3

AC C

4

0.0

X3

1 2

3

M AN U

-0.5

38

0.5

(c)

1.0

> < < < < <

34 34 33 32 31 30

ACCEPTED MANUSCRIPT

40

RI PT

Oil yield (%)

30

20

SC

10

0 0

30

M AN U

without ultrasound UAE

60

90

Time (min)

EP

TE D

Figure 4

AC C

1 2

39

120

150

ACCEPTED MANUSCRIPT •

UAE of crambe seed oil to biodiesel synthesis by in situ interesterification.

•

Extraction time and temperature have a greater influence on the oil extraction.

•

Maximum oil yield was obtained at 60 °C, 90 min and 10 mL of solvent per g of seed.

RI PT

Fatty acids compositions showing a predominance of erucic and oleic acids.

AC C

EP

TE D

M AN U

SC

•

ACCEPTED MANUSCRIPT Answers to the Comments Reviewer 1: After going through the revised manuscript, the revisions submitted by the authors are acceptable, thus I recommend to the Editor to consider this manuscript for publication

RI PT

in Renewable Energy. 1. In conclusion section, a sentence on the very importance of this work and its futuristic application might help the readers about the importance of this work. Answer: The reviewer's request was included in the revised manuscript.

Reviewer 2:

M AN U

and effort spent to improve our research report.

SC

We would like to thank the comments made by the Reviewer, as we recognize the time

This article presented interesting outputs for extraction of oil from crambe seed. However, the following points should be addressed before being accepted for

TE D

publication in "Renewable Energy":

1. In Table 3, use X1×X1 or (X1)^2 or other forms rather than X1(Q) to express the quadratic effect. Insert a column in Table 3 to state if the variable is significant or not.

EP

Answer: As requested by the reviewer the modifications have been made.

AC C

2. In Table 3, how was the Effect (column 2) was calculated (the authors did not reply to this comment)

Answer: The information was added in the revised version of the manuscript (Page 15, line 22).

3. According to Fig. 1, what are the coefficient of determination (r2) and adjusted (r2) values Answer: The coefficient of determination value is presented in the Page 20 – line 18.

ACCEPTED MANUSCRIPT 4. Based on which hypothesis does the optimization in Fig. 2 was performed? Is it interactive response surface methodology, or other? Kindly see Fig. 14 in the article: Bioremediation of red azo dye from aqueous solutions by Aspergillus niger strain isolated from textile wastewater. Journal of Environmental Chemical Engineering 5 (1), 547-554. 2017.

RI PT

Answer: Figure 2 presents the optimal values for the variables within the study range and was conducted using interactive response surface methodology.

SC

5. According to Fig. 2 and Fig. 3, the maximum response occurs at highest X1 (60 min), X2 (60C), and X3 (10). May be the response will still increase with increasing the

M AN U

inputs, or may be a peak occurs. This means that the investigated inputs have no satisfactory ranges. Hence, the authors should mention that future studies are required at wider ranges for input parameters.

Answer: The information was added in the revised version of the manuscript (Page 22, lines 20 and 21).

TE D

We would like to thank the comments made by the Reviewer, as we recognize the time and effort spent to improve our research report. Associate Editor's Comments:

EP

1. The Reviewers consider that the quality of this paper has been much improved. They have some minor comments on some further revisions that would further improve the

AC C

quality of the paper. They suggest that further experimental work could improve the data, so the paper could include suggestions for further work. Answer: All corrections requested by reviewers were included in the revised version of the manuscript.

We would like to thank the comments made by the Editor, as we recognize the time and effort spent to improve our research report.