Bioresource Technology

Bioresource Technology https://doi.org/10.1016/j.biortech.2019.121633

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com

Short Communication Evaluation of a wet processing strategy for mixed phumdi biomass conversion to bioethanol

Anoop Puthiyamadam ⁠a⁠, ⁠2, Velayudhanpillai Prasannakumari Adarsh ⁠a⁠, ⁠2, Kiran Kumar Mallapureddy ⁠a, Anil Mathew ⁠a⁠, ⁠1, Jitendra Kumar ⁠b, Sudhakara Reddy Yenumala ⁠b, Thallada Bhaskar ⁠b, Ummalyama Sabeela Beevi ⁠c, Dinabandhu Sahoo ⁠c, Rajeev K Sukumaran ⁠a⁠, ⁎⁠ a Biofuels and Biorefineries Section, Microbial Processes and Technology Division (MPTD), CSIR-National Institute for Interdisciplinary Science and Technology, Thiruvananthapuram 695019,India b Biomass Conversion Area (BCA), Materials Resource Efficiency Division (MRED), CSIR-Indian Institute of Petroleum (IIP), of Scientific and Innovative Research (AcSIR), Dehradun 248005, India c Institute of Bioresources and Sustainable Development, Takyelpat, Imphal 795001, India

ARTICLE INFO ABSTRACT

Keywords: Biorefineries typically use dry feedstock due to technical and logistic issues, but in unique cases where climatic Wet processing conditions are unfavorable and where the biomass has to be processed without a holding time, wet processing Biomass might be advantageous. The present study evaluated the possibility of using the fresh (non-dried) mixed biomass Phumdi harvested from Phumdis; which are floating vegetation unique to Loktak lake in Manipur, India, for bioethanol Loktak production. Pretreatment with dilute alkali (1.5% at 120 °C for 60 min) resulted in 36% lignin removal and an Bioethanol enhancement of cellulose content to 48% from 37%, and enzymatic hydrolysis released 25 g/L glucose. Fermen- Lignocellulose tation of the hydrolysates was highly efficient at 95% attained in 36 h and 80% in just 12 h. The newwetpro- cessing strategy could help in value addition of mixed phumdi biomass.

1. Introduction again has led to a disposal problem and a solution to the perpetual issue would be value addition the biomass so that it is converted quickly to a Loktak Lake in Manipur, India is home to a unique ecosystem called product in demand, so that there is local income generation along with the Phumdis which are floating mats of vegetation, soil and organic disposal of the otherwise waste biomass (Singh, 2015). matter under various stages of decomposition (Singh and Khundrakpam, We had previously described the potential of using biomass from 2011). Anthropogenic activities have led to increased organic content selected invasive species like Paragrass (Brachairia mutica) and wild and nutrient levels- especially phosphorous and nitrogen in the lake, –rice-grass (Zizania latifolia) from the Loktak lake for production of resulting in rapid proliferation of invasive species like paragrass bioethanol (Sahoo et al., 2018, 2017). These studies involved use of a (Brachairia mutica) in the phumdis (Sahoo et al., 2017). Also there is single variety of plant in dry form as the feedstock for bioethanol pro- an increased proliferation of aquatic weeds like Water hyacinth and duction. However, the harvesting of phumdis performed mechanically Salvinia. The changes in vegetation has significantly affected the health yields a mixed vegetation with high water content which is practically of phumdis, leading to rapid disintegration and sinking, further adding difficult remove. A strategy that uses wet (fresh/non-dried) biomass as to the nutrient levels in the lake and increased proliferation of the in- feedstock for conversion to bioethanol or any other value added product vasive plant varieties. The annual biomass production of Loktak lake would be more beneficial in this scenario. Also, the use of an unsegre- is estimated to be ∼2 million tons (Singh, 2015). The lake contributes gated (mixed) biomass as obtained from the harvesting would be highly to the livelihood of a significant population and the local govern- desirable as it is practically difficult to segregate large quantities of wet ment is spending significant amount of money and effort to period- biomass. ically remove the overgrown floating vegetation from the lake. This While is generally accepted that dry biomass possesses several ad- vantages with respect to preventing damage while storing, economics

⁎ Corresponding author at: Microbial Processes and Technology Division, CSIR-National Institute for Interdisciplinary Science and Technology, Thiruvananthapuram 695 019, India. Email address: [email protected] (R.K. Sukumaran) 1 Present Address: Department of Chemical Engineering, Hanyang University, Seoul, Republic of Korea. 2 Both contributed equally to the work. https://doi.org/10.1016/j.biortech.2019.121633 Received 14 April 2019; Received in revised form 7 June 2019; Accepted 9 June 2019 Available online 12 June 2019 0960-8524/ © 2019. A. Puthiyamadam et al. Bioresource Technology xxx (xxxx) xxx-xxx

Table 1 Changes in mixed phumdi biomass composition on pretreatment.

Temp (°C) Acid/Alkali Conc. (%w/w) Acid pretreated Biomass (%) Presoaked and Alkali Pretreated Biomass (%)

Cellulose Hemi-cellulose Lignin Cellulose Hemi-cellulose Lignin

80 0.5 38.7 ± 0.9 19.6 ± 1.1 22.3 ± 1.5 39.4 ± 0.0 22.9 24.0 ± 0.4 1.0 37.9 ± 1.5 18.4 ± 3.3 24.1 ± 0.7 41.0 ± 0.5 29.3 21.4 ± 0.4 1.5 38.5 ± 0.5 18.1 ± 4.0 25.1 ± 0.5 45.5 ± 10.1 25.5 18.6 ± 0.5 100 0.5 37.9 12.6 ± 0.2 24.9 ± 0.1 41.7 ± 0.2 24.9 ± 3.6 21.7 ± 0.9 1.0 40.8 13.8 26.3 ± 0.3 44.0 ± 1.7 26.8 ± 7.0 19.4 ± 1.6 1.5 44.0 12.1 ± 0.3 25.1 ± 0.7 46.2 ± 0.6 26.2 ± 1.4 17.6 ± 0.2 120 0.5 41.1 ± 4.5 20.5 ± 1.4 25.8 ± 0.2 41.5 ± 2.1 26.0 ± 0.3 19.5 ± 4.4 1.0 45.3 ± 0.7 17.5 ± 0.2 25.3 ± 1.9 46.1 ± 0.8 23.2 ± 0.9 17.7 ± 0.2 1.5 48.9 ± 2.0 13.6 ± 0.9 26.4 ± 0.3 48.1 ± 0.5 22.3 ± 0.3 16.0 ± 0.1

Untreated biomass composition (%): Cellulose −36.7 ± 0.9, Hemicellulose – 22.6 ± 1.7, Lignin – 25 ± 0.8.

Fig. 1. Comparison of the composition of alkali pretreated samples processed with or without pre-soaking. Fig. 2. Sugar yields from enzymatic hydrolysis of differently pretreated MPB.

Table 2 Sugar yields on hydrolysis of acid or alkali pretreated MPB. of moisture. Weight, moisture content and bulk densities were measured as per standard protocols. Before pretreatment, the MPD was chopped Glucose Released for Alkali and wet ground in a food blender. The slurry obtained was analyzed for Temp Glucose Released for Acid presoaked and Pretreated Sample dry solids content and used directly as feedstock for pretreatment. (°C) Pretreated Sample (mg/ml) (mg/ml)

24 h 48 h 24 h 48 h 2.2. Pretreatment

0.5 80 3.2 ± 0.60 4.3 ± 0.01 8.6 ± 0.60 12.0 ± 0.50 Dilute acid (H SO and alkaline (NaOH) pretreatments were car- 100 3.6 ± 0.30 5.3 ± 0.30 13.4 ± 1.20 13.3 ± 1.00 ⁠2 ⁠4) 120 5.7 ± 1.00 7.8 ± 0.90 14.3 ± 1.49 17.7 ± 2.16 ried out using modified methods previously optimized at CSIR-NIIST 1.0 80 4.3 ± 0.10 5.7 ± 0.40 11.0 ± 0.10 13.4 ± 3.00 for other feedstock (Kuttiraja et al, 2013; Christopher et al, 2017). Op- 100 5.81 ± 1.00 5.9 ± 0.30 16.5 ± 1.80 17.6 ± 0.80 timization of pretreatment was carried out by varying the acid/alkali 120 11.5 ± 1.00 15.4 ± 1.00 13.1 ± 1.40 19.8 ± 2.44 loading from 0.5 to 1.5% (w/w) at three different temperatures – 80 °C, 1.5 80 4.9 ± 1.00 7.2 ± 0.20 11.0 ± 2.00 12.9 ± 0.70 100 8.4 ± 1.00 11.7 ± 0.80 19.9 ± 4.20 15.2 ± 1.10 100 °C and 120 °C. Biomass loading was constant at 15% w/w. After pre- 120 14.6 19.3 20.1 ± 2.14 25.1 ± 2.00 treatment, the slurry was adjusted to a pH of 6–6.5 and filtered using a nylon sieve. The pretreated biomass was either used directly for enzymatic hydrolysis or air dried and used later. For evaluation of presoak- of transportation/logistics and handling (Ewanick and Bura, 2011), the ing as a pre-pretreatment strategy, fresh MPD was soaked in 0.25% w/ context of harvested phumdi biomass presents a different challenge and v NaOH solution overnight followed by pretreatment at different alkali it seems that wet processing may be of advantage in this unique con- loadings and temperature as above. text. As could be expected, there is very less information on wet/fresh processing of lignocellulosic biomass for bioethanol production due to 2.3. Hydrolysis obvious reasons. In this study, a wet processing strategy was evaluated for processing of the mixed wet phumdi biomass for bioethanol produc- Hydrolysis was carried out in 100 ml screw-capped conical flasks at tion. 50 °C and at 200 rpm agitation in a water bath shaker. The total reac- tion volume was kept constant at 20 ml, and the biomass loading was 2. Materials and methods 10% (w/w). Enzymatic hydrolysis was carried out using a commercial acid cellulase (Zytex India Pvt. Ltd, Mumbai, India), used at a loading 2.1. Biomass collection and properties of 15 FPU/g. The commercial acid cellulase had a total cellulase activity of 75.7 FPUs/ml, endoglucanase (carboxy methyl cellulase/CM- Mixed phumdi biomass (MPD) was collected fresh from Loktak lake Case) activity of 2003 IU/ml, xylanase activity of 889 IU/ml, beta –glu in Manipur, India and transported to laboratory ensuring minima loss

2 A. Puthiyamadam et al. Bioresource Technology xxx (xxxx) xxx-xxx cosidase activity of 56 IU/ml and had optimal pH and temperature of 3.2. Pretreatment of mixed phumdi biomass and hydrolysis 4.8 and 50 °C respectively. Samples were taken at regular intervals from 0−48 h and sugar concentrations were estimated. Both dilute acid and alkali pretreatment was evaluated for the pretreatment of MPB. For alkali pretreatment, an overnight pre-soaking in 2.4. Fermentation 0.25% NaOH was performed so as to improve penetration of alkali into the biomass, as it has been reported earlier that such a pre-soaking treat- Ethanol fermentation was carried out in 10 ml screw capped glass ment enhances the efficiency of pretreatment (Park and Kim, 2012). vials with a working volume of 5 ml for a period of 48 h. Hydrolysate Composition analyses of the pretreated biomass was performed to de- was concentrated to 5% w/v glucose using a rotary vacuum evaporator. termine the changes in cellulose, hemicellulose and lignin which dic- The inoculum used was 4% (wet w/v) of a novel strain of yeast – Sac- tates the hydrolytic efficiency. In general, acid pretreatment resulted in charomyces cerevisiae – RRP-03 N and fermentation was carried out un- lowering of hemicellulose content with increase in acid concentration der static conditions at a temperature of 30 ± 2 °C. Samples were taken and temperature, and in the case of alkali pretreatment, increase in its at regular intervals and analyzed for sugar and ethanol concentration. concentration and temperature resulted in lowering of lignin content. In both cases, there was a corresponding increase in cellulose concentra- 2.5. Analytical methods tion (Table 1). Highest removal of lignin was achieved with 1.5% Alkali concentration and 120 °C pretreatment temperature. 2.5.1. Composition analysis A comparative study was made for pretreatment of MPB samples The composition of the biomass, before and after pretreatment, was that were presoaked with 0.25% NaOH before pretreatment and sam- determined as per the protocols specified by the National Renewable ples that were pretreated without pre-soaking. It was observed that the Energy Laboratory (NREL) (Sluiter et al., 2012). lignin removal efficiency and the increase in cellulose concentration was very similar in both cases (Fig. 1). 2.5.2. Sugar estimation The efficiency of pretreatment was further evaluated by checking the The concentration of sugars was determined by HPLC (Shimadzu susceptibility to enzymatic hydrolysis. The differently pretreated MPB Prominence UFLC) using a Rezex® RPM Monosaccharide Pb column was subjected to hydrolysis using 15 FPU/g of commercial acid cellu- (8% cross-linked) (300 × 7.8 mm) (Phenomenex, India) and RI detec- lase. Glucose yields were monitored by HPLC. Sugar yield was enhanced tor following previously optimized protocol (Christopher et al., 2016). by both an increase in concentration of the chemical agent (acid/alkali) The column temperature was maintained at 80 °C and the flow rate was and the temperature. Highest glucose yield (25 mg/ml)) was obtained 0.6 ml/min. De-gassed and deionized distilled water was used as mobile in presoaked and alkali pretreated biomass, for the highest alkali con- phase. centration and temperature tried (Table 2). A comparison of the sugar yields between enzymatically hydrolyzed untreated mixed phumdi bio- 2.5.3. Ethanol estimation mass (UT), acid pretreated (1.5% and 120 °C) MPB (Ac-T), Alkali pre- The ethanol concentration was measured using a Gas Chromato- soaked and alkali pretreated (1.5% and 120 °C) MPB (Al-PS-T) and al- graph (Shimadzu GC2014) fitted with RTX-1701 30m capillary column kali pretreated (1.5% and 120 °C) MPB (Al-T) indicated that alkali pre- (Restek- USA) and equipped with FID detector. In the analysis, the col- treated MPB whether presoaked or not, was hydrolyzed better than the umn temperature was maintained at 150 °C with a holding period of acid pretreated biomass (Fig. 2). Compared to acid hydrolysis as a pre- 10 min. The injection volume was 1 µl and was done in the split injection treatment strategy, NaOH pretreatment has been proposed to improve ratio: 1:20. The injector temperature was 160 °C and the mobile phase enzymatic digestion due to higher delignification efficiency by alkali used was nitrogen gas at a linear velocity of 30 cm/sec and a column (Ioelovich and Morag, 2012). It has also been reported that a mild tem- flow rate of 1 ml/min. perature NaOH treatment induces significant solubilization of lignin and xylan and this treatment prior to a pretreatment scheme designed to 3. Results and discussion improve enzymatic hydrolysis would improve the saccharification efficiency (Liu et al., 2014). However, in this case alkali pre-soaking before 3.1. Biomass properties pretreatment did not have much influence on the sugar yields compared to pretreatment without presoaking (Fig. 2). This might be because of The mixed phumdi biomass mainly consisted of the leaves and stems the fact that the biomass was already wet and allowed penetration of of Para grass (Brachiaria mutica) while it also contained smaller amounts alkali at the same level as with a pre-soaked sample. Since presoaking of the areal portions of other plants like tall reed (Phragmatis karka) wild would require extended time of operation and more infrastructure for rice grass (Zizania latifolia), some water hyacinth (Eichornia crassipes) the plants when operated at higher scales, further studies were carried and minor quantities of other unidentified plants. The submerged por- out on the MPD biomass without presoaking before pretreatment. tion of biomass was dark and mainly contained roots and soil. Composi- tion analysis of the aerial portion and roots of the biomass revealed high lignin content (>30%) for the roots. For unit quantity (1 kg) wet weight 3.3. Fermentation of alkali pretreated mixed phumdi biomass of the aerial portion of biomass, the dry weight was 0.15 kg while it was about 0.18 kg for the root portion. Mean bulk density of the aer- Fermentation was carried out with a yeast strain Saccharomyces cere- ial phumdi biomass was 0.107 kg/L while this was not determined for visiae RRP-03N which is known to grow and produce ethanol from the root portion as it contained considerable amount of soil and silt biomass hydrolysates even when they are not de-toxified (Christopher unless washed. The average moisture content for aerial and root por- et al., 2017). Pretreatment process often introduces sugar degradation tions were 75.5% and 71.8% respectively. Since the lignin content was products and the degradation products from lignin, which are not com- very high and because of the presence of high amount of silt and soil, pletely removed from the pretreated biomass, especially when a final roots were not used for further studies. The mixed wet biomass used for washing step is not included. The pretreatment step in present study the study was the aerial portion of the mixed phumdi biomass (MPB) did not include a washing step deliberately and post processing after containing a representation of the predominant plant varieties. Com- pretreatment was only a neutralization and solid liquid separation. This position analysis of the MPB indicated that it contained 36.7 ± 0.91% while affecting the enzymatic hydrolysis, did not pose much of prob- cellulose, 22.6 ± 1.68% hemicellulose, 25.03 ± 0.75% lignin and about lem for the fermentation, as the yeast could almost completely utilize 4.98 ± 0.07% Ash. glucose in 36 h and more than 81% of glucose was consumed in the

3 A. Puthiyamadam et al. Bioresource Technology xxx (xxxx) xxx-xxx first 12 h itself. By 24 h, 94.5% of glucose was utilized and ∼99% glu- IIP and IBSD through DBT to RKS with Project Number IBSD/A1/ cose was consumed in 36 h. The ethanol production, starting from a glu- P(PH-2)/11. cose concentration of 5.6% w/v was 2.75, 3.07 and 3.28% v/v at 12, 24 and 36 h respectively. References The ethanol yield was 95% of the theoretical maximum at 36 h, and even at 12 h of fermentation, the ethanol yield was ∼80% of the theo- Christopher, M., Anusree, M., Mathew, A.K., Nampoothiri, K.M., Sukumaran, R.K., Pandey, A., 2016. Detoxification of acidic biorefinery waste liquor for production of high value retical maximum indicating that the yeast was not affected by any pos- amino acid. Bioresour. Technol. 213, 270–275. https://doi.org/10.1016/j.biortech. sible inhibitory compounds present in the hydrolysate. We had previ- 2016.03.054. ously performed fermentation of Paragrass biomass hydrolysate using Christopher, M., Mathew, A.K., Kiran Kumar, M., Pandey, A., Sukumaran, R.K., 2017. A biorefinery-based approach for the production of ethanol from enzymatically hy- the same strain, where also the unique property of tolerance to poten- drolysed cotton stalks. Bioresour. Technol. 242, 178–183. https://doi.org/10.1016/j. tial inhibitors of fermentation was observed (Sahoo et al., 2017). Appar- biortech.2017.03.190. ently, the Saccharomyces cerevisiae strain RRP-03N could be considered Ewanick, S., Bura, R., 2011. The effect of biomass moisture content on bioethanol yields as a highly potent yeast for performing ethanol fermentation from bio- from steam pretreated switchgrass and sugarcane bagasse. Bioresour. Technol. 102, 2651–2658. https://doi.org/10.1016/J.BIORTECH.2010.10.117. mass hydrolysates where the fermentation inhibitors have very low or Ioelovich, M., Morag, E., 2012. Study of enzymatic hydrolysis of mild pretreated lignocel- no effect on the organism performance. lulosic biomasses. BioResources 7, 1040–1052. https://doi.org/10.15376/biores.7.1. 1040-1052. 4. Conclusions Liu, T., Williams, D.L., Pattathil, S., Li, M., Hahn, M.G., Hodge, D.B., 2014. Coupling alkaline pre-extraction with alkaline-oxidative post-treatment of corn stover to enhance enzymatic hydrolysis and fermentability. Biotechnol. Biofuels 7, 48. https://doi.org/ A wet processing strategy was evaluated for pretreatment of mixed 10.1186/1754-6834-7-48. biomass from Phumdis, the unique floating vegetation in Manipur’s Lok- Park, Y.C., Kim, J.S., 2012. Comparison of various alkaline pretreatment methods of lignocellulosic biomass. Energy 47, 31–35. https://doi.org/10.1016/J.ENERGY.2012.08. tak lake. Alkaline pretreated biomass yielded ∼25 g/L glucose on enzy- 010. matic saccharification. Direct processing of mixed wet biomass isbeing Sahoo, D., Ummalyma, S.B., Okram, A.K., Pandey, A., Sankar, M., Sukumaran, R.K., 2018. reported for the first time to the best of our knowledge. Though wetpro- Effect of dilute acid pretreatment of wild rice grass (Zizania latifolia) from Loktak Lake for enzymatic hydrolysis. Bioresour. Technol. 253, 252–255. https://doi.org/10. cessing may be advantageous where drying is not a feasible option. Also 1016/j.biortech.2018.01.048. the process did not generate much inhibitors evidenced by the >92% of Sahoo, D., Ummalyma, S.B., Okram, A.K., Sukumaran, R.K., George, E., Pandey, A., 2017. the theoretical efficiency in fermentation. These indicate the potential Potential of Brachiaria mutica (Para grass) for bioethanol production from Loktak Lake. Bioresour. Technol. 242, 133–138. https://doi.org/10.1016/J.BIORTECH.2017. to use mixed phumdi biomass in fresh form for any value addition 03.047. Singh,W. R, 2015. Composting of Floating Biomass (Phumdi and Salvinia natans) of Loktak Acknowledgements Lake (Manipur, India).Ph.D Thesis, Submitted to IIT Guwahati. TH-1481_09610471. Singh, A.L., Khundrakpam, M.L., 2011. Phumdi proliferation: a case study of Loktak lake. Manipur. Water Environ. J. 25, 99–105. https://doi.org/10.1111/j.1747-6593.2009. The authors would like to thank the Institute of Bioresources and 00197.x. Sustainable Development (IBSD), Imphal, Manipur for the financial as- Sluiter, A., Hames, B., Ruiz, R., Scarlata, C., Sluiter, J., Templeton, D., Crocker, D., 2012. sistance in the form of collaborative project between CSIR-NIIST-CSIR- NREL/TP-510-42618 analytical procedure - determination of structural carbohydrates and lignin in biomass. Lab. Anal. Proced. 17, NREL/TP-510-42618.