Improved microbial oil production from oil palm empty fruit bunch by Mucor plumbeus

Fuel 194 (2017) 180–187

Contents lists available at ScienceDirect

Fuel journal homepage: www.elsevier.com/locate/fuel

Full Length Article

Improved microbial oil production from oil palm empty fruit bunch by Mucor plumbeus Farah B. Ahmad a,⇑, Zhanying Zhang a, William O.S. Doherty a, Valentino S.J. Te’o b, Ian M. O’Hara a a b

Centre for Tropical Crops and Biocommodities, Queensland University of Technology, Brisbane, Australia School of Earth, Environmental and Biological Sciences, Faculty of Science and Engineering, Queensland University of Technology, Brisbane, Australia

a r t i c l e

i n f o

Article history: Received 1 June 2016 Received in revised form 16 November 2016 Accepted 5 January 2017

Keywords: Biodiesel Empty fruit bunch Lignocellulose Lipid Microbial oil Palm oil

a b s t r a c t This study investigated the effect of cultivation parameters on microbial oil production from hydrolysate of oil palm empty fruit bunch (EFB) using fungus Mucor plumbeus. The parameters selected for evaluation were sugar concentration (30–100 g/L), yeast extract concentration (0–13.3%, g yeast extract/g sugar), pH (5–7) and spore concentration (4.3–6.3, log spore number/mL medium). Response surface methodology was used to optimise the cultivation conditions which were based on the oil concentration and oil yield. Sugar concentration was the most influential parameter that affected oil concentration. However, the cultivation at high sugar concentration (
100 g/L) also resulted in ethanol accumulation. The optimum condition for oil yield was found at 30 g/L sugar, 0 g/L yeast extract and pH 5.0. Cultivation in 1 L bioreactor under the optimum conditions resulted in
1.8-fold increase in oil yield compared to the shake-flask cultivation. Microbial oil produced from EFB hydrolysate has the potential to be used as the feedstock for biodiesel production from non-food feedstock, with cheaper cost of biodiesel production in comparison to glucose-derived microbial oil. Ó 2017 Published by Elsevier Ltd.

1. Introduction Lignocellulosic biomass is an attractive feedstock for microbial oil production due to its high availability and low price. Microbial oil or lipid can be produced through biochemical conversion of lignocellulosic biomass, which typically involves pretreatment and enzymatic hydrolysis of lignocellulosic biomass for breaking down polysaccharides to hexoses and pentoses, that can be utilised by oleaginous microorganisms for oil production. Oils in the form of triacylglycerides (TAG) can be used as feedstocks for the production of the second generation biodiesel through transesterification process. In comparison to the first generation biodiesel derived from plant oils, the production of biodiesel from microbial oil has several advantages, including low requirement of land-use and labour, as well as higher oil productivity [1,2]. Oleaginous filamentous fungi are promising candidates for microbial oil production from lignocellulosic biomass due to their capacity to grow on a broad range of carbon substrates (e.g., gluAbbreviations: HMF, 5-hydroxymethylfurfural; C/N, carbon-to-nitrogen ratio; DO, dissolved oxygen; EFB, empty fruit bunch; EH, enzymatic hydrolysate; RMC, raw material cost; RSM, Response surface methodology. ⇑ Corresponding author. E-mail address: [email protected] (F.B. Ahmad). http://dx.doi.org/10.1016/j.fuel.2017.01.013 0016-2361/Ó 2017 Published by Elsevier Ltd.

cose, xylose, glycerol, etc.) [3,4], and their ability to tolerate low concentrations of growth inhibitors (e.g., furfural and 5hydroxymethylfurfural (HMF)) resulting from chemical pretreatment of lignocellulosic biomass [5]. The morphology of filamentous fungi, either in pellet or filamentous hyphal form, allows simple filtration technique for down-steam processing [6,7]. Cultivation conditions play important roles in oil production. Carbon-to-nitrogen (C/N) ratio is possibly the most important factor as oleaginous microorganisms accumulate oil under limitingnitrogen conditions [8,9]. Other cultivation conditions that have influence on microbial growth and oil production are cultivation pH [10], and inoculum concentration. pH of the cultivation medium may affect microbial cells’ membrane permeability [11]. Spore inoculum concentration of fungi was reported to have an effect on fungal morphology and metabolic activity [12], which subsequently could influence oil accumulation. Despite the effect of a variety of cultivation conditions on oil production, studies with a systematic approach to process optimisation of oil production from lignocellulosic biomass by filamentous fungi are limited. Oil palm empty fruit bunch (EFB), is the lignocellulosic byproduct of palm oil processing and makes up the highest percentage of wastes generated in palm oil mills. In this study, the cultivation of the filamentous fungus, Mucor plumbeus, on EFB

F.B. Ahmad et al. / Fuel 194 (2017) 180–187

hydrolysates for microbial oil production was optimised. M. plumbeus was shown to be the best candidate for oil production from EFB [13]. EFB hydrolysate was prepared through dilute acid pretreatment and enzymatic hydrolysis. Response surface methodology (RSM) was used to optimise the oil production by assessing the impact of the parameters of cultivation (sugar concentration, yeast extract concentration, spores concentration, and pH) on the oil concentration (g/L) and oil yield (mg oil per g sugars consumed). Oil yield is an important response parameter for the optimisation as it measures the efficiency of converting the carbon substrates (i.e., an operating cost) to product (i.e., a revenue) [3]. Subsequently, microbial oil production was scaled up into a bioreactor system to investigate the effect of reactor operation on microbial biomass and oil production. Studies of cultivation in the bioreactor systems are important for assessing the feasibility and the economics of progressing to industrial scales (e.g., >1000 L). However, there are limited studies on microbial oil production from lignocellulosic hydrolysates in the bioreactor systems. Microbial oil from EFB could be a promising source for biodiesel production. The use of cheap feedstock like EFB could potentially improve the economics of microbial oil production, which is one of the challenges for commercialising biodiesel production.

2. Material and methods 2.1. Materials Oil palm EFB was provided by KKS East Mill, Sime Darby Plantation Sdn. Bhd, Malaysia. Air-dried EFB consisted of 34.0% glucan, 17.2% xylan, 29.6% lignin, 7.5% moisture, 6.5% ash, 14.2% water extractive and 6.3% ethanol extractive based on compositional analysis procedure developed by National Renewable Energy Laboratory [14,15]. Mucor plumbeus (FRR no.: 2412) strain was purchased from FRR Culture Collection (Australia). The spores of fungal strain were maintained on potato dextrose agar (PDA) at 4 °C [2]. 2.2. EFB hydrolysate preparation EFB was pretreated at 170 °C with 0.8 wt% sulfuric acid and a solid/liquid ratio of 1:6 in 7.5 L Parr reactor (Model 4554, Parr Instrument Company, USA). The stirring speed was 100 rpm and the reaction time was 15 min. Following pretreatment, the liquid fraction and the solid residue were separated by filtration using Whatman filter paper (Grade 1, Whatman, England). The solid residue was washed twice with tap water. Enzymatic hydrolysis of the washed EFB solid residue was performed at a glucan loading of 7% (w/w) with a cellulase dosage of 20 FPU/g glucan (AcceleraseTM 1500, Batch no: 4901298419). The pH of the mixture was adjusted to 5.0 and the mixture was then placed in a shaking incubator (OM15, Ratek, Australia) for 72 h at 50 °C and 150 rpm. At the end of enzymatic hydrolysis, the liquid fraction of enzymatic hydrolysis was separated by centrifugation. The supernatant was labelled as EFB enzymatic hydrolysate (EH). EH was concentrated using a rotary evaporator (Rotavapor, BUCHI, UK) at 60 °C. The concentrated EH consisted of 118.52 g/L glucose, 9.55 g/L xylose and 1.00 g/L arabinose. 2.3. Optimisation in shake flasks A response surface methodology (RSM) with face-centred central composite design (CCD) (Supplementary Data A.1) was applied for the cultivation of M. plumbeus on EFB enzymatic hydrolysate

181

(EH) by varying the parameters of cultivation (independent variables) which were: 1. Sugar concentration of EH (g/L) (X1) 2. Relative concentration of yeast extract to sugars in EH (%, g yeast extract/g sugar in EH) (X2) 3. Spore concentration (log spores number/mL medium) (X3) 4. Initial pH (X4) The response factors (dependant variables) for optimisation were oil concentration (g/L) (Y1) and oil yield (mg/g) (Y2). The coded and actual values of each variable and its levels (1 for low value and 1 for high value) for this experimental design are shown in Table 1. The levels were determined based on the preliminary study (unpublished data), where the selection of the pH level was on the basis that M. plumbeus showed no capacity to grow at pH 4 and poor growth above pH 8. For this optimisation study, a total of twenty-six experimental runs were conducted in random, which consisted of the combination of sixteen factorial points, six axial points and a centre point with five replicates. Design of experiments, mathematical modelling and optimisation of process parameters were performed using the Design Expert software (Stat-Ease Inc., USA). Three dimensional surface plots were drawn using MATLAB R2009a (The MathWorks, Inc., USA). Each response variable was fitted to a quadratic model to correlate the response variable to the independent variables. Analysis of variance (ANOVA) was evaluated through statistical analysis of the model. The statistical significance of the model terms was assessed using the p-value approach. For cultivation experiments, sugar concentrations, yeast extract and spore inoculum concentration as well as pH were used according to Table 1. The cultivation media were prepared by supplementing EH with the same nutrients compositions used in previous study (0.4 g/L MgSO47H2O, 2 g/L KH2PO4, 3 mg/L MnSO4H2O and 0.1 mg/L CuSO45H2O) [13]. The cultivation was performed with 30 mL working volume in 250 mL Erlenmeyer flasks at 28 °C and 200 rpm on an OM15 orbital shaking incubator (Ratek, Australia) for seven days. Fungal biomasses were harvested by vacuum filtration and washing, followed by biomass freeze-drying to constant weight [3]. A validation experiment was performed using the experimental conditions of the optimised parameters cultivation. The cultivation of control was performed using the media preparation based on repeatedly used cultivation methods [2,16]. The nutrients supplementation used in the control was the same as the present study. Yeast extract concentration for the control cultivation was 1.5 g/L and the initial pH was 5.5. 2.4. Cultivation in bioreactor The cultivation of M. plumbeus on EH was scaled up and performed in a 1 L bioreactor (New BrunswickTM BioFloÒ/CelliGenÒ 115 Fermentor and Bioreactor, Eppendorf AG, Germany). The cultivation was carried out at 28 °C with an initial agitation speed of 200 rpm and aeration rate of 1 vvm (volume air per working volume per minute). A cascade control for dissolved oxygen (DO) regulation was applied to keep the DO level higher than 20% by automated increment of agitation speed and aeration rate. The cultivation in the bioreactor was carried out using EH at the same concentrations of sugar (30 g/L) and yeast extract (0 g/L), and initial pH (pH 5.0) of the optimised conditions, and supplemented with the same nutrients formulation of 0.4 g/L MgSO47H2O, 2 g/L KH2PO4, 3 mg/L MnSO4H2O and 0.1 mg/L CuSO45H2O [13]. Samples from the cultivation medium were observed under an Olympus BX41TF microscope (Japan) with an Olympus DP11 Microscope Digital Camera system (Japan).

182

F.B. Ahmad et al. / Fuel 194 (2017) 180–187

Table 1 The coded and actual values of each variable and its levels for the experimental design. Independent variable

Sugar concentration Relative concentration of yeast extract Spore concentration Initial pH

Sugar %YE Spore pH

Unit

Symbol

g/L %, g yeast extract/g sugar log spores number/mL medium

X1 X2 X3 X4

The medium for bioreactor cultivation were inoculated with 10% (v/v) preculture medium. The preculture medium was prepared using EH diluted to
15 g/L sugars, 1 g/L yeast extract and spore concentration of 6.3 log spores/mL medium (2  106 spore/ mL). The cultivation of preculture medium was conducted at 28 °C on an OM15 orbital shaking incubator (Ratek, Australia) for three days.

2.5. Oil extraction and analyses Microbial oil was extracted from the biomass based on the procedure described in Ahmad et al. [13]. All results for oil content are reported on a dry weight (DW) basis. The oil yield (mg/g) was calculated by dividing the oil concentration (mg/L) with the total glucose and xylose consumed (g). For the determination of the fatty acids composition of oils, microbial oils were derivatised to fatty acid methyl esters (FAME) based on a previous study [17]. FAME analysis was carried out using a gas chromatography-mass spectrometry (GC–MS) system (Shimadzu GCMS-TQ8040, Japan) equipped with an RtxÒ-2330 column (60 m long  0.25 mm I.D.  0.2 lm film thickness) (Restek, USA), based on the GC–MS method described in Ahmad et al. [13]. Sugar concentrations (cellobiose, glucose, xylose and arabinose), organic acids (formic acid, acetic acid and levulinic acid), furans (furfural and 5-hydroxymethylfurfural (HMF)) and ethanol were analysed using high-performance liquid chromatography (HPLC) by a Waters HPLC system and a refractive index (RI) detector (Waters 410, US) [18]. The column used for analysing sugars was a SP810 carbohydrate column (300 mm  8.0 mm, Shodex, Japan) that was set at 85 °C with water as mobile phase at a flow rate of 0.5 mL/min [18]. Organic acids, furans and ethanol were analysed using an Aminex HPX-87H column (300 mm  8.0 mm, Bio-Rad, US), that was set at 65 °C with 5 mM H2SO4 as the mobile phase at a flow rate of 0.6 mL/min [18]. The analysis of total nitrogen was conducted using TOC-VCSH (Shimadzu Corporation, Japan) with TNM-1 (TN (Total Nitrogen) Unit) (Shimadzu Corporation, Japan) [13]. Carbon to nitrogen (C/ N) ratio (mol/mol) was calculated as follows,

½Glucose and Xylose and Acetic Acid  0:4 ðmol=molÞ C=N ¼ ½TN in yeast extract þ ½TN in hydrolysate

ð1Þ

where [Glucose and Xylose and Acetic acid] was the total concentration of glucose and xylose and acetic acid (g/L). For [TN in yeast extract] and [TN in hydrolysate], TN refers to total nitrogen (g/L). Nitrogen composition of yeast extract was 8% (w/w) [13]. The oil yield (mg/g) was determined by dividing the oil concentration (mg/L) with the total glucose and xylose consumed (g).

Code level 1

0

1

30 0 4.3 5.0

65 6.7 5.3 6.0

100 13.3 6.3 7.0

3. Results and discussion 3.1. Optimisation of cultivation conditions by response surface methodology (RSM) 3.1.1. Chemical compositions of EFB hydrolysates The concentrations of acetic acid and 5-hydroxymethylfurfural (HMF) of EFB enzymatic hydrolysate (EH) at different sugar concentrations are shown in Table 2. Acetic acid, HMF and furfural were potential microbial growth inhibitors that typically present in lignocellulosic hydrolysates. Furfural was not detected in the hydrolysates. HMF was below the detection level in EH30 and EH65. The concentration of HMF was considered to be too low in EH100 to have a negative impact on microbial growth [5]. The concentrated EH (undiluted) consists of 0.89 g/L of nitrogen. Nitrogen may have originated from the nitrogen-rich commercial hydrolytic enzymes [19]. 3.1.2. Mathematical modelling and statistical analysis of optimisation study The optimisation experiment on the cultivation of M. plumbeus on EH was performed based on the experimental design as described in Table 1. The analysis of variance (ANOVA) of the model indicated that the model of oil concentration and oil yield are significant. Further statistical analysis was presented in Supplementary Data A. The analysis demonstrated that the mathematical models are reliable for simulating the oil concentration and oil yield for the cultivation of M. plumbeus on EFB hydrolysate. The results of the experimental and predicted responses, and the standard deviation from the simulation using the mathematical models are shown in Table 3. Further evaluation on the predicted results from the mathematical models was presented in the Supplementary Data A.6. The plots of residuals versus predicted responses displayed random pattern, and not outward-opening funnel pattern which is the undesirable pattern. Therefore, this analysis demonstrated that the mathematical model is reliable and the assumptions are satisfied [20]. The surface plots simulated from the mathematical models are shown in Supplementary Data B for the oil concentration and Supplementary Data C for the oil yield. 3.2. Effects of sugar concentration, yeast extract concentration, spore concentration and pH on microbial oil production Sugar concentration was selected as one of the parameters for optimisation in order to assess the influence of carbon concentration on the oil production as C/N ratio is one of the key factors affecting oil production. The surface analysis indicated that sugar concentration (X1) was significant (p-value < 0.0050) to the oil concentration, with positive linear coefficient (Supplementary Data A.3), which is in accordance to higher sugar concentration (carbon) leads to a higher C/N ratio and subsequently higher oil accumulation by oleaginous microorganisms [8]. The synergy of sugar concentration with yeast extract concentration (X1X2) was significant to the oil yield, where from the surface plots of oil yield, lowering the yeast extract concentration supplemented to the

183

F.B. Ahmad et al. / Fuel 194 (2017) 180–187 Table 2 Concentrations of sugars, organic acids and furan in enzymatic hydrolysates (EHs). Sugar concentration (g/L) of EH

Medium label

Glucose (g/L)

Xylose (g/L)

Formic acid (g/L)

Acetic acid (g/L)

HMF (g/L)

30 65 100

EH30 EH65 EH100

28.90 57.09 95.06

2.59 5.00 8.23

n.d. 0.39 0.51

1.63 2.51 3.56

n.d. n.d. 0.06

n.d. – not detected.

Table 3 The experimental and predicted responses from the simulation of the mathematical models for cultivating M. plumbeus on enzymatic hydrolysates (EHs) at different concentrations of sugars (X1), yeast extract (X2) and spores (X3), and pH (X4), as well as standard deviation (SD). Run

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26

Independent variables

Oil concentration (g/L) (Y1)

Oil yield (mg/g) (Y2)

Sugar (X1)

%YE (X2)

Spore (X3)

pH (X4)

Experi-mental value

Predicted value

SD

Experi-mental value

Predicted value

SD

0 1 1 1 1 1 0 0 0 1 1 1 0 1 1 1 1 1 1 0 1 1 1 0 0 0

1 1 1 1 1 0 0 0 0 1 1 1 0 1 1 1 1 0 1 0 1 1 1 0 0 0

0 1 1 1 1 0 0 0 1 1 1 1 0 1 1 1 1 0 1 0 1 1 1 0 0 1

0 1 1 1 1 0 0 1 0 1 1 1 0 1 1 1 1 0 1 0 1 1 1 0 0 0

1.82 1.49 1.23 4.26 8.08 5.46 2.17 2.16 2.91 0.67 3.77 2.39 2.46 7.56 1.49 1.40 1.10 1.02 1.48 1.90 2.54 3.46 1.69 2.16 2.08 2.40

2.01 0.63 1.54 3.72 7.49 5.05 2.05 2.35 2.55 0.92 3.13 2.60 2.05 7.60 1.77 0.38 1.09 1.57 1.91 2.05 3.68 4.14 1.82 2.05 2.05 2.96

0.14 0.61 0.22 0.38 0.42 0.29 0.09 0.14 0.26 0.18 0.45 0.15 0.29 0.03 0.20 0.72 0.01 0.39 0.30 0.11 0.81 0.49 0.09 0.08 0.03 0.40

36 60 43 59 85 54 38 37 53 6 37 86 43 74 48 45 37 33 51 34 54 35 59 38 37 42

39 56 50 58 78 47 36 40 47 5 36 84 36 77 50 37 36 42 53 36 59 41 53 36 36 51

2.01 2.56 5.48 0.45 4.39 4.89 1.35 2.01 3.92 1.05 0.03 1.52 4.86 2.02 1.47 6.06 0.55 6.53 1.40 1.97 4.02 4.09 3.90 1.20 0.22 5.93

media would lead to higher oil yield at any concentration of sugars (Supplementary Data C.1). The yeast extract concentration was shown to be significant to both oil concentration and oil yield. The binary interaction of yeast extract and spore concentration (X2X3) was significant to the oil concentration and oil yield. The surface plots (Supplementary Data B.4 and C.3) showed that at the maximum yeast extract concentration, increasing spores inoculum concentration could lower the oil production. This is possibly because higher yeast extract concentration provides surplus-nitrogen condition, where this condition was known to stimulate cell proliferation [21]. Therefore, a higher spore inoculation to surplus-nitrogen medium may have caused a rapid formation of fungal biomass in the culture, which could lead to poor oxygen-mass transfer and subsequently low oil production. This study also revealed that fungi could grow even without yeast extract for optimum oil concentration (Supplementary Data B.1, B.4 and B.5). From the response surface analysis, pH (X4) and the binary interaction of pH and sugar concentration (X1X4) were shown to be significant to the oil concentration. Based on the surface plots of the interaction between pH and sugar concentration for both oil concentration (Supplementary Data B.3) and oil yield (Supplementary Data C.2), it can be concluded that the impact of pH on oil concentration or oil yield varied according to the level of sugar concentration. At the minimal sugar concentration (30 g/L), a higher oil yield could be achieved by decreasing the value of pH (Supplementary Data C.2). However, a total opposite correlation

applied at maximum sugar concentration of 100 g/L, where decreasing pH values would decrease the oil yield. The correlation between pH and sugar concentration might be influenced by acetic acid concentration. EH100 contains 3.56 g/L acetic acid, which is three times higher than acetic acid concentration in EH30. The pKa value of acetic acid is 4.75, which means that at pH 4.75, the concentration of the un-dissociated and dissociated forms of acetic acid in the cultivation media were equal [22]. Acetic acid in undissociated form was found to be inhibitory to microbial growth [23]. When the medium was adjusted to pH 7.0, there would be 99% of acetic acid dissociated into acetate anions [23], which reduces the inhibitory effect of acetic acid on microorganisms. Therefore, pH 7.0 was more favourable for oil production on EH100.

3.3. Substrates consumption, fatty acids composition and other metabolites accumulation The impact of different cultivation parameters on microbial growth was analysed through the sugar consumption behaviour of M. plumbeus on EFB enzymatic hydrolysates (EHs) (Fig. 1 for glucose and Supplementary Data D for xylose). From the experimental run performed on EH30 and EH100 (Fig. 1b and c), the cultivations with the maximum yeast extract and spore concentration resulted in complete consumption of glucose within 48 h (EH30) and 74 h (EH100), regardless of the initial pH of the media. The cultivation with yeast extract concentration had a faster consumption rate of

184

F.B. Ahmad et al. / Fuel 194 (2017) 180–187

Glucose concentration (g/L)

(A)

35

min YE, min spore, pH5

30

max YE, min spore, pH5

25

min YE, max spore, pH5

20

max YE, max spore, pH5

15

min YE, min spore, pH6 max YE, min spore, pH7

10

min YE, max spore, pH7

5 0

max YE, max spore, pH7 0

24

48

72

96

120

144

168

mean YE, mean spore, pH6

Time (h) Glucose concentration (g/L)

(B)

70

min YE, mean spore, pH5

60

mean YE, min spore, pH5

50 40

mean YE, max spore, pH6

30 20

mean YE, mean spore, pH7

10 0

0

24

48

72

96

120

144

168

mean YE, mean spore, pH6

Time (h)

Glucose concentration (g/L)

(C)

min YE, min spore, pH5

120

max YE, min spore, pH5

100

min YE, max spore, pH5

80

max YE, max spore, pH5

60

min YE, min spore, pH7

40

min YE,max spore, pH7

20 0

max YE, max spore, pH7

0

24

48

72

96

Time (h)

120

144

168

mean YE, mean spore, pH6

Fig. 1. The consumption of glucose from the cultivation on EFB enzymatic hydrolysate (EH) diluted to 30 g/L sugars (EH30) (A), 65 g/L sugars (EH65) (B) and 100 g/L sugars (EH100) (C).

sugars in comparison to the cultivation on EH30 and EH65 with no yeast extract supplementation (Fig. 1a and b). The trend of sugar consumption was shown to be correlated to the concentration of yeast extract (nitrogen concentration). The cultivation with maximum yeast extract concentration had the fastest consumption of glucose and xylose. As nutrient (e.g., nitrogen) depletion was shown to negatively impact cell proliferation and carbon substrates uptake rates [24], the opposite may promotes carbon substrates consumption. Even though the cultivation with maximal yeast extract supplementation could potentially reduce the number of days of cultivation due to fast sugar consumption, the oil yields were low in comparison to the cultivation with no additional yeast extract (Supplementary Data C.1, C.3 and C.4). The overall rate of glucose consumption of all cultivation was higher than the consumption rate for xylose. In this study, for all cultivations, M. plumbeus exhibited sequential sugar assimilation where xylose consumption began after the majority of glucose was consumed in the media. The ratio of the amount of glucose to xylose in the media was 12:1. The sequential sugar assimilation was common for media that contained a higher proportion of glucose than xylose, where the assimilation pattern could be due to catabolite repression by glucose or allosteric competition for sugar transporters [25].

The analysis of fatty acids composition showed that the microbial oils consist of palmitic (C16:0), stearic (C18:0), oleic (C18:1) and linoleic (C18:2) acid, which was similar to fatty acid compositions reported for the cultivation of other oleaginous microorganism [2,13,26]. Ethanol was detected in some of the cultivation runs (Supplementary Data E). Ethanol accumulation in the cultivation with higher sugar concentrations showed that M. plumbeus did not possess strict aerobic metabolism. Some Mucor species (e.g., Mucor indicus) were utilised for ethanol production from acid hydrolysates of rice straw and spruce forest residues [27]. In this study, ethanol accumulation was prevalent in the cultivation with high sugar concentration (
100 g/L). High carbon substrates loading caused a rapid growth of biomass that could lead to an increase in the viscosity of the culture and further reduce the efficiency of oxygen-mass transfer inside the fungal biomass. This phenomenon could result in carbon assimilation under limiting-oxygen conditions which then led to the production of ethanol. Ethanol was accumulated most likely at the expense of carbon-to-oil conversion efficiency, where cultivation with ethanol accumulation of more than 20 g/L obtained low oil yields (5 mg/g for Run #10 and 36 mg/g for Run #11) (Table 3). The cultivation on EH100 also resulted in higher ethanol yields than oil yields (Supplementary Data E). Even though ethanol is a valuable co-product, the ethanol

F.B. Ahmad et al. / Fuel 194 (2017) 180–187

yields were too low in comparison to the theoretical ethanol yield (514 mg/g glucose), therefore the process of ethanol recovery could lead to unfavourable economics for the overall cost of production. 3.4. Validating optimisation parameters for oil production from EFB hydrolysates (EH) by M. plumbeus The main criteria for finding optimum parameters for oil production are high oil yield and high oil concentration. The cultivation at the maximum sugar concentration (100 g/L) and pH 7.0 with no additional yeast extract was predicted to give the highest oil concentrations (7.6 0 g/L), with predicted oil yield at
77 mg/g (Table 3). The cultivation with a minimum sugar concentration and maximum spore concentration at pH 5.0 without additional yeast extract was predicted to result in the highest oil yield at 84 mg/g (Table 3). However, the cultivation run at the maximum sugar concentration led to the accumulation of more than 5 g/L ethanol with ethanol yields of more than 50 mg per g sugars consumed (Supplementary Data E). The cultivation at 100 g/L also resulted in rapid formation of biomass in high volume which was a potential challenge for the cultivation in bioreactor. Therefore, the parameters for optimisation were selected at 30 g/L sugar concentration, pH 5.0 with spore concentration at 6.3 log spores number/mL medium, without additional yeast extract. The optimised parameters were predicted to result in an oil concentration of 2.60 g/L and 84 mg/ g oil yield. The validation experiment resulted in the production of 11.6 g/L biomass and 2.67 g/L oil (oil content of 23.1%), with an oil yield of 94 mg/g. The relative amounts of sugars consumed from EH after seven days of cultivation were 86% of glucose and 38% of xylose, respectively. The analysis on the fatty acids of microbial oil from the optimised cultivation revealed that the oil consists of 18.9% palmitic, 31.0% stearic, 34.7% oleic and 15.5% linoleic acids, which is similar to fatty acid compositions of M. plumbeus grown in EFB hydrolysates in our previous study [13]. The cultivation of the control, performed on EH at the optimised sugar concentration (30 g/L), resulted in the production of 13.2 g/L biomass and 2.13 g/L oil with 16.2% oil content and 52 mg/g oil yield. The resulting oil concentration and oil yield from the control were lower than those of the optimised cultivation. However, both glucose and xylose in the cultivation medium of the control were consumed completely by the end of cultivation. The cultivation medium of the control contains a higher nitrogen content and lower C/N ratio than the optimised culture. The higher nitrogen content may have contributed to a better consumption of the carbon sources in the control, as sufficient nitrogen supply promotes cell proliferation. Even with high sugar consumption rates in the control, the efficiency of carbon substrates conversion into oil was lower than that in the optimised cultivation. It is noted that higher sugar consumption rates did not necessarily lead to a higher oil yield, which provided further evidence on the impact of nitrogen concentration on oil production as discussed in Section 3.3. 3.5. Microbial oil production in bioreactor The cultivation of M. plumbeus on EH was subsequently performed in a bioreactor, based on the optimised conditions, to investigate the impact of the bioreactor system on oil production as well as the feasibility of process scale-up. Supplementary Data F.1 shows gradual consumption of sugars throughout the cultivation. The pH increased slightly at the beginning of the cultivation, but dropped gradually after 24 h, and reached pH 5.0 at the end of the cultivation. The DO level was reduced sharply within the first 24 h, as glucose was consumed rapidly at this phase of the cultivation (Supplementary Data F.1). Due to DO control in the bioreactor system, the DO level was maintained at 20% by increasing the agi-

185

tation speed from its initial speed at 200 rpm (minimum setpoint) to 335 rpm (maximum setpoint) by 41 h. The agitation speed was reduced afterward to the original speed at 200 rpm, corresponding to the raising DO level and remained at 200 rpm until the end of cultivation. For the first two days of the cultivation, fungal biomass only grew in pellets. The transition in fungal morphology in the bioreactor became apparent after three days of cultivation with the appearance of dispersed mycelia mixed with fungal pellets (Supplementary Data G). The transition in morphology, from pelletonly biomass to mixture of pellet and dispersed mycelia, was most likely attributed to the high agitation speed. Filamentous fungus Mortierella isabellina in dispersed mycelial form was reported to obtain higher oil yield and oil content in comparison to fungi in pellet form [7]. This is because mycelia aggregates possibly had better oxygen and nutrient intake due to lower biomass density and smaller radius [7]. The results of batch cultivations in the bioreactor were shown in Table 4. In comparison to the shake-flask cultivation, the growth of M. plumbeus on EH in a bioreactor showed an improvement in the oil concentration and oil yield. Even at a similar C/N ratio, the cultivation in the bioreactor resulted in much higher oil concentration and oil yield. The enhanced microbial oil production was possibly attributed to enhanced oxygen supply because of proper agitation and aeration in the culture. The transition in fungal morphology was one of the possible factors of the increased oil production. The sugar consumption of the cultivation in the bioreactor was similar to the shake-flask cultivation under a similar C/N ratio (Supplementary Data F.2), as discussed previously that the nitrogen content of media might play role in sugar consumption. Table 4 also showed the comparison between the bioreactor cultivation of M. plumbeus on EH to other bioreactor cultivation of oleaginous yeasts and fungi on various lignocellulosic hydrolysates. The oil concentration (5.3 g/L) and oil yield (168 mg/g) from the bioreactor cultivation on EH were comparable to the results of bioreactor cultivations from other studies such as microbial oil production from corn stover by M. isabellina (6.9 g/L oil) and from rice straw by Rhodotorula glutinis (
7.3 g/L oil) [26,28]. The results of oil concentration produced from corn stover by Rhodotorula graminis was much higher than the oil concentration obtained in the present study [29], possibly due to higher sugar concentration in corn stover hydrolysate. However, the oil yield in the present study was higher than that obtained from the cultivation on corn stover hydrolysate by R. graminis (89 mg/g) [29]. The study of cultivation of R. glutinis on rice straw hydrolysates in an airlift bioreactor by Yen et al. obtained a higher oil concentration than the present study [28], possibly due to higher C/N ratio as the oil concentration of R. glutinis was close to the results of the cultivation M. isabellina at C/N ratio of 91 [26]. Even though the oil concentration obtained by Yen et al. was higher than the present study [28], the oil yield was comparable to the present study. The outcome of the bioreactor cultivation study demonstrated the potential of scaling-up of M. plumbeus for large-scale oil production from EFB hydrolysates without the addition of nitrogen such as yeast extract. 3.6. Biodiesel production from EFB The fatty acids composition of oil from the bioreactor cultivation on EFB hydrolysate were 15.6% palmitic, 15.9% stearic, 21.1% oleic and 47.4% linoleic acid, which was different to the oil produced from the shake-flask cultivation. The fatty acids compositions were further used to evaluate the fuel properties of microbial oil via empirical calculation. The fuel properties were analysed based on the assessment of fatty acid methyl ester (FAME) as described in Ahmad et al. [13]. The results showed that

186

F.B. Ahmad et al. / Fuel 194 (2017) 180–187

Table 4 Results of different cultivation performed in the present study and the comparison with other batch cultivation of oleaginous yeasts and fungi from the literature. Feedstock (hydrolysate)

Reactor type

Strains

Glucose (g/L)

Xylose (g/L)

C/N ratio

Oil (g/L)

Oil yield (mg/g)

Reference

EFB residue EFB residue Corn stover residue Corncob stover Rice straw

Shake-flask Stirred-tank bioreactor Stirred-tank bioreactor Stirred-tank bioreactor Airlift bioreactor

Mucor plumbeus Mucor plumbeus Mortierella isabellina Rhodotorula graminis Rhodotorula glutinis

31.2 33.6 28.6 126.0 23.9

3.7 2.8 16.1 87.1 6.1

62 65 91 n/a n/a

2.7 5.3 6.9 16.3 7.3a

94 168 147 89a 170

The present study The present study [26] [29] [28]

n/a – Not available. a Data was not provided, therefore the value was estimated.

the oil has a cetane number of 57.11, iodine value of 99.95 and kinematic viscosity of 4.34 mm2/s, which are within the limit set by the European standard for FAME of EN 14214 (>51 for cetane number, <120 for iodine value and 3.5–5.0 mm2/s for kinematic viscosity). Therefore, microbial oil from EFB was a promising source for good quality biodiesel production. The production cost of biodiesel production from EFB was calculated based on Koutinas et al. for biodiesel production from pure glucose through microbial cultivation, oil extraction and transesterification (Table 5) [30]. For the production of 10,000 t biodiesel from EFB, based on the oil yield of 0.168 g/g (oil/sugars) and the conversion rate of oil to biodiesel of 90%, it was estimated that 66,138 t of sugars from EFB was required. The cost of sugars production from EFB was estimated to be at $256/t based on the selling price of diluted sugars from lignocellulosic biomass [2,31]. The raw material cost (RMC) for biodiesel production was calculated as the total cost for cultivation (carbon substrates and yeast extract), extraction and transesterification (Table 5a). RMC for the production of 10,000 t/year biodiesel from glucose, was evaluated to be at $21.23 million [30]. RMC for EFB biodiesel was estimated to be lower at $17.84 million, due to the use of cheaper lignocellulosic sugars and no additional yeast extract in the cultivation. Table 5b presents the components of the production cost based on the calculation of the cost of manufacturing for glucose biodiesel [30]. Assuming that the utilities cost and operating labour cost

for the cultivation of EFB sugars for the production of microbial oil were similar to the production of microbial oil from glucose, the cost of EFB biodiesel production was estimated to be $5472/t biodiesel. The cost of production for EFB biodiesel was estimated to be 7.3% lower than that for glucose biodiesel. The cost of production could be reduced with an improved oil yield through optimising cultivation in bioreactor, as shown in Table 5c, where the cost for EFB biodiesel at the theoretical oil yield (0.32 g oil/g glucose [8]) (theoretical EFB feedstock) was 24% lower than that for glucose biodiesel. The cost of biodiesel production from microbial oil is still higher than vegetable oil-derived biodiesel at $1318/t [32]. However, the production of the first generation biodiesel from vegetable oils can cause adverse impacts on global food security. The global food price from 2010 was forecasted to climb up by 47% in 2040, based on annual growth rate of firstgeneration biofuels production at 2.7% (as predicted by U.S. EIA) [33]. The analysis of 2013 global consumption of the firstgeneration biodiesel revealed that the calorie content from the oil crops used for biodiesel production could be equivalent to the calorie requirement to feed 70 million people [34]. In addition, there is limited global agricultural land to support the current demand from food and fuel markets [34]. The production of microbial oil-derived biodiesel could be economically viable through implementation of an energy policy which limits the use of foodbased feedstock, technology advancement for reducing the cost

Table 5 The comparison of raw materials cost (RMC) (A), production cost (B) and cost of biodiesel (C) for the production of biodiesel from EFB and glucose based on Koutinas et al. for the production of 10,000 t/year biodiesel [30]. Process

Raw material

(A) Cultivation RMC on pure glucose

Glucose Yeast extract Total Cultivation RMC on EFB EFB sugar Yeast extract Total Extraction & transesterification RMC (hexane, methanol, NaOH and HCl) a RMC for glucose RMC for EFBa

a b c d

Quantity (t/y)

Unit cost ($/t)

Total cost ($ million/y)

42,081 4370

400 800

66,138 0

256 0

16.833 3.495 20.328 16.931 0 16.931 0.905 21.233 17.836

Cost item

Total cost ($ million/y)

(B) Fixed capital investment (FCI) Operating labour cost (OLC) Utility cost (UC) Production cost for glucoseb,c Production cost for EFBb,c

73.65 1.005 7.563 58.895 54.718

Feedstock

Cost of biodiesel ($/t)

(C) Glucose EFB Theoretical EFB (at theoretical oil yield)d

5900 5472 4483

RMC = Cultivation RMC + Extraction & transesterification RMC. Assuming Wastewater treatment cost (WTC) = 0. Operating cost = 0.28FCI + 2.73OLC + 1.23 (RMC + UC + WTC). Theoretical FB sugars = 34,722 t/year, oil yield = 0.32 g/g glucose.

F.B. Ahmad et al. / Fuel 194 (2017) 180–187

of microbial cultivation (e.g., consolidated bioprocessing) and biorefinery integration. 4. Conclusions The optimisation of microbial oil production from EFB hydrolysate by Mucor plumbeus was performed using response surface methodology by evaluating the impact of cultivation parameters on oil concentration and oil yield. The analysis showed that increasing sugar concentration resulted in an increased oil concentration, and decreasing yeast extract concentration led to an increased oil yield. However, the cultivation at high sugar concentration (
100 g/L) resulted in ethanol production, which subsequently led to lower oil yields. The optimum conditions for microbial oil production were identified and utilised for the cultivation in 1 L bioreactor, which resulted in an increase of the oil yield in comparison to the shake-flask cultivation. The outcome of this study demonstrated the potential of scaling-up of M. plumbeus for large-scale microbial oil production from EFB hydrolysates without the addition of nitrogen source. The oils produced from EFB have the potential to be used for good quality biodiesel production, with cheaper cost of production than glucose derived oil. Acknowledgement The authors acknowledge Ministry of Higher Education, Malaysia for the postgraduate scholarship of Farah B. Ahmad. The authors thank Mr Mohan Sivasubramaniam from KKS East Mill Sdn. Bhd., Malaysia for the supply of EFB. The authors also thank the QUT Banyo Pilot Plant and the QUT Central Analytical Research Facility for its support on sample preparation and analyses. Appendix A. Supplementary material Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.fuel.2017.01.013. References [1] Liang M-H, Jiang J-G. Advancing oleaginous microorganisms to produce lipid via metabolic engineering technology. Prog Lipid Res 2013;52:395–408. [2] Zheng Y, Yu X, Zeng J, Chen S. Feasibility of filamentous fungi for biofuel production using hydrolysate from dilute sulfuric acid pretreatment of wheat straw. Biotechnol Biofuels 2012;5:1–10. [3] Ahmad FB, Zhang Z, Doherty WOS, O’Hara IM. A multi-criteria analysis approach for ranking and selection of microorganisms for the production of oils for biodiesel production. Bioresour Technol 2015;190:264–73. [4] Huang C, Chen X-F, Xiong L, Chen X-D, Ma L-L, Chen Y. Single cell oil production from low-cost substrates: the possibility and potential of its industrialization. Biotechnol Adv 2013;31:129–39. [5] Ruan Z, Hollinshead W, Isaguirre C, Tang YJ, Liao W, Liu Y. Effects of inhibitory compounds in lignocellulosic hydrolysates on Mortierella isabellina growth and carbon utilization. Bioresour Technol 2015;183:18–24. [6] Xia C, Zhang J, Zhang W, Hu B. A new cultivation method for microbial oil production: cell pelletization and lipid accumulation by Mucor circinelloides. Biotechnol Biofuels 2011;4:1–10. [7] Gao D, Zeng J, Yu X, Dong T, Chen S. Improved lipid accumulation by morphology engineering of oleaginous fungus Mortierella isabellina. Biotechnol Bioeng 2014;111:1758–66. [8] Papanikolaou S, Aggelis G. Lipids of oleaginous yeasts. Part II: Technology and potential applications. Eur J Lipid Sci Technol 2011;113:1052–73.

187

[9] Ratledge C. Fatty acid biosynthesis in microorganisms being used for single cell oil production. Biochimie 2004;86:807–15. [10] Venkata Subhash G, Venkata Mohan S. Lipid accumulation for biodiesel production by oleaginous fungus Aspergillus awamori: influence of critical factors. Fuel 2014;116:509–15. [11] Chew TL, Bhatia S. Catalytic processes towards the production of biofuels in a palm oil and oil palm biomass-based biorefinery. Bioresour Technol 2008;99:7911–22. [12] Gao D, Zeng J, Zheng Y, Yu X, Chen S. Microbial lipid production from xylose by Mortierella isabellina. Bioresour Technol 2013;133:315–21. [13] Ahmad FB, Zhang Z, Doherty WOS, O’Hara IM. Evaluation of oil production from oil palm empty fruit bunch by oleaginous micro-organisms. Biofuels Bioprod Biorefin 2016;10:378–92. [14] Sluiter A, Hames B, Ruiz R, Scarlata C, Sluiter J, Templeton D, et al. Determination of structural carbohydrates and lignin in biomass. In: Laboratory analytical procedure. Golden, Colorado: National Renewable Energy Laboratory; 2011. [15] Sluiter A, Ruiz R, Scarlata C, Sluiter J, Templeton D. Determination of extractives in biomass. In: Laboratory analytical procedure. Golden, Colorado: National Renewable Energy Laboratory; 2005. [16] Yu X, Zheng Y, Dorgan KM, Chen S. Oil production by oleaginous yeasts using the hydrolysate from pretreatment of wheat straw with dilute sulfuric acid. Bioresour Technol 2011;102:6134–40. [17] Mulbry W, Kondrad S, Buyer J, Luthria D. Optimization of an oil extraction process for algae from the treatment of manure effluent. J Am Oil Chem Soc 2009;86:909–15. [18] Zhang Z, Wong HH, Albertson PL, Doherty WOS, O’Hara IM. Laboratory and pilot scale pretreatment of sugarcane bagasse by acidified aqueous glycerol solutions. Bioresour Technol 2013;138:14–21. [19] Jin M, Slininger PJ, Dien BS, Waghmode S, Moser BR, Orjuela A, et al. Microbial lipid-based lignocellulosic biorefinery: feasibility and challenges. Trends Biotechnol 2015;33:43–54. [20] Montgomery DC. Design and analysis of experiments. Hoboken (NJ): John Wiley & Sons; 2005. [21] Gong Z, Shen H, Yang X, Wang Q, Xie H, Zhao ZK. Lipid production from corn stover by the oleaginous yeast Cryptococcus curvatus. Biotechnol Biofuels 2014;7:158. [22] Palmqvist E, Hahn-Hägerdal B. Fermentation of lignocellulosic hydrolysates. II: inhibitors and mechanisms of inhibition. Bioresour Technol 2000;74:25–33. [23] Gong Z, Shen H, Zhou W, Wang Y, Yang X, Zhao ZK. Efficient conversion of acetate into lipids by the oleaginous yeast Cryptococcus curvatus. Biotechnol Biofuels 2015;8:189. [24] Cescut J, Fillaudeau L, Molina-Jouve C, Uribelarrea J-L. Carbon accumulation in Rhodotorula glutinis induced by nitrogen limitation. Biotechnol Biofuels 2014;7:1–11. [25] Hu C, Wu S, Wang Q, Jin G, Shen H, Zhao ZK. Simultaneous utilization of glucose and xylose for lipid production by Trichosporon cutaneum. Biotechnol Biofuels 2011;4:1–8. [26] Ruan Z, Zanotti M, Archer S, Liao W, Liu Y. Oleaginous fungal lipid fermentation on combined acid- and alkali-pretreated corn stover hydrolysate for advanced biofuel production. Bioresour Technol 2014;163:12–7. [27] Karimi K, Zamani A. Mucor indicus: biology and industrial application perspectives: a review. Biotechnol Adv 2013;31:466–81. [28] Yen H-W, Chang J-T, Chang J-S. The growth of oleaginous Rhodotorula glutinis in an internal-loop airlift bioreactor by using mixture substrates of rice straw hydrolysate and crude glycerol. Biomass Bioenergy 2015;80:38–43. [29] Galafassi S, Cucchetti D, Pizza F, Franzosi G, Bianchi D, Compagno C. Lipid production for second generation biodiesel by the oleaginous yeast Rhodotorula graminis. Bioresour Technol 2012;111:398–403. [30] Koutinas AA, Chatzifragkou A, Kopsahelis N, Papanikolaou S, Kookos IK. Design and techno-economic evaluation of microbial oil production as a renewable resource for biodiesel and oleochemical production. Fuel 2014;116:566–77. [31] Humbird D, Davis R, Tao L, Kinchin C, Hsu D, Aden A, et al. Process design and economics for biochemical conversion of lignocellulosic biomass to ethanol: dilute-acid pretreatment and enzymatic hydrolysis of corn stover. Golden (CO): National Renewable Energy Laboratory (NREL); 2011. [32] Apostolakou AA, Kookos IK, Marazioti C, Angelopoulos KC. Techno-economic analysis of a biodiesel production process from vegetable oils. Fuel Process Technol 2009;90:1023–31. [33] To H, Grafton RQ. Oil prices, biofuels production and food security: past trends and future challenges. Food Secur 2015;7:323–36. [34] Rulli MC, Bellomi D, Cazzoli A, De Carolis G, D’Odorico P. The water-land-food nexus of first-generation biofuels. Sci Rep 2016;6.