Innovative physically-assisted soda fractionation of rapeseed hulls for better recovery of biopolymers

Abstract

A better knowledge of the effect of non-conventional pretreatment technologies, which can avoid the use of high temperatures and detrimental solvents, is necessary. Thus, physical pretreatment (ultrasounds (US), microwaves (MW), high voltage electrical discharges (HVED) and pulsed electric fields (PEF)) were applied in order to evaluate their effects on acid insoluble residue removal and enzymatic hydrolysis yields from rapeseed hulls. The subsequent chemical treatment consisted of adding 0.3 mol L⁻¹ of hydroxide sodium and maintaining the suspension at 60 °C for 2 h. The results showed that applying physical pretreatments resulted in an increasing yield of acid insoluble residue by 5% (PEF), 6% (MW), 8% (HVED), 12% (US) respectively, in comparison with the chemical treatment. The pulps isolated by physical-assisted extraction, showed a higher enzymatic digestibility. The results obtained in the present study confirm that coupling pretreatments constitute a promising alternative for the valorization of this by-product.

---

1. Introduction

Over 72 million tons of rapeseed were cultivated worldwide in 2013, primarily dedicated to oil production for food purposes and for biodiesel. The content of hulls in rapeseed varies from 10.5 to 20% of the seed weight and 20 to 30% of the defatted meal, on a dry weight basis. The use of rapeseed coproducts has been limited even for animal feed due to the occurrence within the product matrix of a number of antinutritional constituents. There are relatively little reports focusing on the use of rapeseed residues. Most of them deal with thermal applications like pyrolysis,¹ extraction of pectins or condensed tannins with protein precipitating capacity.² In fact, the poor overall quality of the rapeseed hulls is mainly due to the presence of a hard, black and resistant layer with a high resistance against the biological and/or chemical destroying medium. This layer, called phytomelanin, acts as a protective layer of the seed, which protects the developing embryo from external invasion of insects and pests. The chemical composition of rapeseed hulls black layer is still not clear. It is a non lignified material but the absence of lignin has never been clearly demonstrated, till now. On the other hand, rapeseed hulls have a relatively high sugar content that makes it a potential interesting raw material for second-generation ethanol production.

The production of ethanol from rapeseed hulls would contribute to the biorefinery development based on this crop. However, this feedstock poses many challenges for pretreatment processes because of the presence of the phytomelanin layer and its high phenolics content which complicates the sugars recovery and the action of the biocatalysts. Utilization of rapeseed hulls for the production of bioethanol has to the best of our knowledge never been previously reported.

The goal of any pretreatment technology is the removal of structural and compositional impediments to hydrolysis to enhance the rate of enzyme hydrolysis and increase yields of fermentable sugars from cellulose or hemicelluloses.³ There are a number of key features for the effective pretreatment of recalcitrant biomass. The fractionation process should have a low capital and operational cost. It should result in the effective recovery of biomass components in separate fractions. For this purpose, chemical and/or physical pretreatments are investigated.

Several review articles provide a general overview of the current pretreatment approaches.^4–6 Steam Explosion (SE) is the most commonly used pretreatment of biomass and uses both physical and chemical strategies to break the structure of the material through an hydrothermal treatment. However, SE generates some toxic derivatives which can inhibit the successive hydrolysis and fermentation steps,⁷ it has also been revealed to be less effective than some other treatments.⁸ Ammonia fiber explosion (AFEX) approach is similar to steam explosion: biomass is exposed to liquid ammonia (anhydrous or concentrated, >70%) under high temperature and pressure and then the pressure is quickly released. Ammonia loading and also residence time have the highest impact on the economics of the process. The acid pretreatments are the most effective methods used for dissolving hemicellulose and recovering most of the cellulose.⁹ Though, the drawbacks are the high equipment cost and side reactions releasing toxic molecules such as furfural and 5-HMF that are inhibitory to fermentation step. Alkaline pretreatment has been extensively studied for the pretreatment of biomass, especially annual crops and agricultural co-products. It has been shown to disrupt the phenolic structure of the biomass, increasing the accessibility of enzymes to cellulose and hemicelluloses. Some advantages of this process is high pulp production rates and yields, use of atmosphere pressure and low temperatures,¹⁰ and the possibility of processing any type of biomass, woods and alternative species. However, this type of fractionation requires long residence times and high concentrations of base. Another considerable drawback is the conversion of alkali into irrecoverable salts and/or their incorporation into the biomass during the fractionation process,³ which makes it a challenging issue for alkaline approach.

In order to benefit from chemical effect of alkali treatment while using smaller quantities of reagents, coupling with innovative physical treatments is proposed in this work. Physical pretreatment technologies include ultrasounds, microwaves, pulsed electric fields and high voltage electrical discharges. Ultrasounds irradiations are widely used in food industry because it increases reaction rates. This process catalyzes depolymerization of biopolymers, emulsification, tanning of vegetables, extraction of oils from almond,¹¹ ginger,¹² and woad seeds.¹³ Besides, applications of ultrasound irradiation processes were reported in municipal wastewater pretreatment to disrupt flocks,¹⁴ biodiesel production from micro-algae,¹⁵ and pretreatment for bioethanol production from cassava chips.¹⁶ Microwave heating is a proper route to accomplish the disruption of recalcitrant structures in the biomass. Microwave-assisted extraction (MAE) is a key, sustainable technology in achieving the objectives of green chemistry. It has been rapidly developed as one of the hot-spot techniques for isolating interesting high added-value compounds from solid samples. Pulsed electric field (PEF) technology may cause lethal damage to cells or induce sublethal stress by transient permeabilization of cell membranes and electrophoretic movement of charged species between cellular compartments.¹⁷ Different aspects of PEF application for disintegration of soft cellular tissues were intensively discussed in literature.^17,18 High voltage electrical discharges (HVED) induce the electrical breakdown in water.¹⁹ The electrical discharge results in the generation of hot and localized plasmas that emit high-intensity UV light, produce shock waves, and generate hydroxyl radicals during water photodissociation. Electrical discharges have been studied as a means for cell disruption in biochemistry, biology, medicine and drug delivery. It also has been applied for the sterilization of milk or fruit juice,²⁰ and extraction from grape seeds.²¹

The novelty of this study is to combine alkaline treatment with physical pretreatment for rapeseed hulls fractionation. Based on the specific biopolymers content which refers to the main fractions: acid insoluble residue and sugars and on recalcitrance of rapeseed hulls, a combinative fractionation process was carried out involving a physical treatment as a first step followed by soda treatment as a second step. Besides, the effect on the pulp composition and structure after different treatments was also studied with additional information obtained from scanning electron microscope.

2. Materials and methods

2.1 Biological material

The rapeseed hulls were provided by the company CREOL (Pessac, France) and were obtained from Denis D50 separator. The air-dried rapeseed residues were stored at room temperature until use. The dry matter content of provided hulls is 86% (w/w).

2.2 Chemicals and enzymes

The enzyme cocktail (Viscozyme L) was supplied by Sigma-Aldrich (Steinheim, Germany) and presents a specific activity of 100 fungal beta-glucanase units (FBG) mL⁻¹, in which 1 FBG is the amount of enzyme required under the standard conditions (30 °C, pH = 5.0 and 30 min of reaction time) that can hydrolyze barley β-glucan to reducing carbohydrates, with a reducing power corresponding to 1 μmol glucose min⁻¹. Anhydrous sodium acetate (99%) was purchased from Fluka Sigma-Aldrich (Steinheim, Germany), sodium hydroxide and sulfuric acid (72%) from Fisher Scientific (Illkirch, France).

2.3 Pretreatment processes

2.3.1 Ultrasounds pretreatment. The ultrasound treatment chamber (Fig. 1) is a 1 L narrownecked glass flask containing a titanium ultrasound probe (H14 Hielscher GmbH, Germany), with a length of 100 mm and a diameter of 14 mm, connected to the ultrasonic generator (Hielscher GmbH, Stuttgar, Germany). The ultrasound generator can provide a maximal power of 400 W and a maximal frequency of 24 kHz. In this study, the power and frequency regulators were fixed to 400 W and 12 kHz respectively. A mixture containing 43.5 g of rapeseed hulls, 435 g of water (60 °C) and 5.7 g NaOH (0.3 mol L⁻¹ of NaOH) was introduced in the treatment chamber. The temperature was controlled and maintained constant (60 ± 2 °C) by using a water bath.

Fig. 1 Experimental set-up.

The specific ultrasounds energy input E (kJ/Kg) was calculated as follows:

where t_US is the effective treatment duration (s), m is the mass of suspension of rape-seed hulls (kg) and P_US is the generator power (400 W).

In this study, two treatment energies were compared: 150 kJ kg⁻¹ and 1500 kJ kg⁻¹. The corresponding total extraction times were 3 min and 30 min respectively.

2.3.2 Microwave treatment. Microwave pretreatment experiments were performed in a Milestone reactor (NEOS-GR, Milestone, Italy). This apparatus is a multimode microwave reactor 2.45 GHz with a maximum delivered power of 900 W. Experiments were performed at atmospheric pressure.

A suspension containing 43.5 g of rapeseed hulls, 435 g of water (60 °C) and 5.7 g NaOH (0.3 mol L⁻¹ of NaOH) was introduced in the vessel and placed into the microwave reactor and the treatment program with the desired time–power settings was chosen. The treatment temperature was maintained constant (60 ± 2 °C).

The specific microwave energy input E (kJ/Kg) was calculated as follows:

where t_MW is the effective treatment duration (s), m is the mass of suspension of rapeseed hulls (kg) and P_MW is the generator power (400 W).

In this study, two treatment energies were compared: 150 kJ kg⁻¹ and 1500 kJ kg⁻¹. The corresponding total extraction times were 3 min and 30 min respectively.

2.3.3 PEF treatment. The apparatus is composed of a pulsed high voltage power supply (Tomsk Polytechnic University, Tomsk, Russia) and a treatment chamber with a 1 L capacity. The chamber was equipped with two parallel disc electrodes (11 cm of diameter). The pulse duration was of 10 μs and the frequency was 0.5 Hz. The generator can provide exponential decay pulses with a maximum voltage of 40 kV and a maximum current of 10 kA. The distance between electrodes was fixed at 3 cm which corresponded to an electric field strength of 13 kV cm⁻¹. A more detailed description of this equipment is presented in literature.²¹ The specific energy input E (kJ kg⁻¹) was obtained from eqn (3):where W_PEF is the pulse energy (kJ per pulse), and m is the mass of the suspension of rapeseed hulls (kg). W_PEF is determined from eqn (4).where U is the voltage (V) and I is the current (A).

For the pre-treatments, 58 g of hulls, 7.5 g NaOH and 580 g of distilled water (60 °C) were added to the chamber.

Treatment energy applied to the seeds was 150 kJ kg⁻¹. The PEF treatment consisted of applying up to n_PEF = 599 pulses. Thus the time of PEF application (t_PEF, s) was calculated as the product of the average pulse width (t_i, s) and the number of pulses (n_PEF):

t_PEF = n_PEF × t_i (5)

where t_i is the average pulse width and was about 10 μs.

However, the total time of solid–liquid contact during the HVED treatment was much longer and corresponds to:

where f is the pulse frequency.

In this study, the effective PEF application time and the corresponding total solid–liquid contact time were 6 ms and 30 min respectively.

2.3.4 HVED treatment. The same generator as that used for PEF was used for HVED experiments but the treatment chamber was different. A 1 L treatment chamber (inner diameter = 10 cm, wall thickness = 2.5 cm) was equipped with needle-plate geometry electrodes. The diameters of stainless steel needle and the grounded disk electrodes were 10 and 35 mm respectively. The distance between the electrodes was 0.5 cm. A positive pulse voltage was applied to the needle electrode. The peak pulse voltage (U) was 40 kV. Rapeseed hulls (58 g) were introduced between the electrodes. 480 g water at 60 °C and 7.5 g NaOH were added to the solid. The liquid-to-solid ratio (w/w) was 10.

The electrical discharges were generated by electrical breakdown in liquid. The energy input of HVED treatment was calculated as shown in eqn (7) and (8). The specific treatment energy input was fixed by using 599 discharges with a frequency f of 0.5 Hz.

The specific energy input E (kJ kg⁻¹) was obtained from eqn (7):

where W_HVED is the discharge energy (kJ per discharge), m is the suspension mass (kg).

W_HVED is determined from eqn (8).

where U is the voltage (V) and I is the current strength (A).

Treatment energy applied to the seeds was 150 kJ kg⁻¹. The time of HVED application (t_HVED, s) was calculated as the product of the average pulse width (t_i, s) and the number of pulses (n_HVED):

t_HVED = n_HVED × t_i (9)

where t_i is the average pulse width and was about 10 μs.

However, the total time of solid–liquid contact during the HVED treatment was longer and corresponds to:

where f is the pulse frequency.

In this study, the effective HVED application time and the corresponding total solid–liquid contact time were 6 ms and 30 min respectively.

All operating conditions of physical treatments are summarized in Table 1.

Table 1 Physical treatments conditions

Technology	Dry material (g)	Solvent (g)	Treatment conditions	Total duration (min)	Energy (kJ kg^−1)	Specific energy (kJ kg^−1 DM)	Temperature (°C)
MW	43.5	435	(400 W, 30 min)	30	1500	150	60 ± 2
US	43.5	435	(400 W, 30 min)	30	1500	150	60 ± 2
PEF	58	580	(20 kV cm^−1, 6 ms)	30	150	15	60 ± 2
HVED	58	580	(40 kV, 6 ms)	30	150	15	60 ± 2

---

2.3.5 Soda: alkaline extraction. Soda extraction was carried out in a 500 mL glass beaker. 50 g of rapeseed hulls were used for experiments. The liquid to solid ratio was fixed at 10. The solvent was composed of 430 g of water and 5.6 g of NaOH (0.3 mol L⁻¹ of NaOH). The soda extraction was carried out at 60 °C under agitation (750 rpm). For control experiments (without physical pretreatment), the soda extraction was performed for 2 h. For treated samples by PEF, HVED, US or MW, the duration of soda extraction was fixed so that the total extraction (physical treatment + soda) time corresponded to 2 h. After extraction, the mixture was filtered. The solid residue (pulp) was washed with 1 L distilled water. The pulp was then dried at 60 °C for weight determination and characterization. The liquid (black liquor) was collected and stored at 4 °C until analysis.

2.4 Analysis

2.4.1 Analysis of the raw material. Moisture content was determined using infrared dryer (Scaltec, Germany). Results were expressed on dry matter (DM) basis. Ash content was determined by calcination at 525 °C for 5 h. Extractives were determined by extracting untreated hulls (2 g) with hexane (200 mL), using a Soxhlet apparatus (volume = 250 mL) for 6 h (about 36 extraction cycles). The content of nitrogen was analyzed using a ThermoFinnigan model Eager 300 analyzer. The percentage of proteins was calculated as N(%) × 6.25.²² Pectins were specifically extracted with hydrochloric acid (pH 1.5) at 85 °C for 1 h. Total soluble polyphenols were extracted from the samples (4 g) with 100 mL of solvent (water/acetone/acetic acid, 29.5/70/0.5) for 2 h at 50 °C under stirring. The total polyphenols contents were measured by the Folin–Ciocalteu method, based on a colorimetric oxidation/reduction reaction of phenols.

2.4.2 pH determination. The pH of the suspension was determined by using a pHmeter (Ecoscan pH5, Eutech Instruments, France) at 20 °C.

2.4.3 Non saccharidic fraction quantification.
Acid soluble phenolics. The acid-soluble phenolics (ASP) content was determined from absorbance values at 205 nm according to the laboratory analytical procedure (LAP) provided by the National Renewable Energy Laboratory (NREL/TP-510-42618).where ε is the extinction coefficient of lignin at 205 nm L g⁻¹ cm⁻¹; V_t is the total volume of sample (L); DM_i is the initial dry matter (g).
Acid insoluble residue. 175 mg of recovered pulps were hydrolyzed with 1.5 mL of sulfuric acid (72%) for 1 h at 30 °C and autoclaved for 1 h at 121 °C after being diluted to 3% sulfuric acid with the addition of water. The autoclaved samples were filtered, and the dried residue was weighed to obtain the acid insoluble residue content.

2.4.4 Non saccharidic fraction characterization. All solid state nuclear magnetic resonance (NMR) spectroscopy experiments were performed on a Bruker Avance-400 spectrometer. Liquid state ¹³C NMR spectra were obtained with a Bruker Avance 300 NMR spectrometer. Quantitative NMR spectra were acquired using an inverse-gated decoupling (Waltz-16) pulse sequence to avoid nuclear. The samples (150 mg) were dissolved in DMSO-d₆ (0.40 mL) with slight heating and stirring with a micro stir bar. Spectral analyses were performed using Bruker software. Approximatively 30–40 mg of dried acid insoluble residue was placed in a glass thermal desorption tube with ¼ in. outer diameter and 3.5 in. long from Supelco, Inc. and end capped with glass wool. Thermal treatment was performed in a Turbomatrix 300 Thermal Desorber system from Perkin Elmer, USA. The conditions included a prepurge for 1 min at room temperature using helium carrier gas at 1 mL min⁻¹ followed by heating at 210, 230 and 280 °C for different durations in the glass thermal desorption tube of the thermodesorber using carrier gas at 20 mL min⁻¹ and splitless trapping at −30 °C. The trap was desorbed at 300 °C at 1 mL min⁻¹ through a transfer line at 300 °C with a split flow of 40 mL min⁻¹. The volatile degradation products were analyzed by GC-MS after thermodesorption. GC-MS analysis was performed on a Clarus 500 GC gas chromatograph (Perkin-Elmer) equipped with a 5% diphenyl/95% dimethyl polysiloxane fused-silica capillary column (J&WScientific DB-5, 30 m × 0.25 mm × 0.25 m) and controlled by Turbomass (v5.4.2) software. Ionization was achieved under the electron impact method (70 eVionization energy). The components were identified on the basis of comparison of their mass spectrum with the NIST Library 2005 through the NIST MS Search 2.0. Identification was considered as relevant for the match and reverse match coefficient values above 900.

2.4.5 Monosaccharide quantification. Separation and quantification of neutral sugars were performed using a Dionex ICS-3000 system consisting of a gradient pump, an autosampler, an electrochemical detector with a gold working electrode, an Ag/AgCl reference electrode and Chromeleon version 6.8 (Dionex Corp., USA). A Carbopac PA1 (250 mm, Dionex) column with a guard column (4 × 50 mm, Dionex) was used as a stationary phase using isocratic conditions with 1 mM sodium hydroxide as eluent. Eluents were prepared by dilution of a 46–48% NaOH solution (PA S/4930/05 Fisher Scientific) in ultrapure water. All eluents were degassed before use by flushing with helium for 20 min; subsequently they were kept under constant helium pressure (eluent degassing module, Dionex). After each run, the column was washed for 10 min with 200 mM NaOH and reequilibrated for 15 min with the starting conditions. Samples were injected through a 25 μL full loop and separations were performed at 35 °C at a rate of 4 mL min⁻¹. The pulse sequence for pulsed amperometric detection consisted of potentials of +100 mV (0–200 ms), +100 mV integration (200–400 ms), 2000 mV (410–420 ms), +600 mV (430 ms), and 100 mV (440–500 ms).²³

2.4.6 Enzymatic saccharification. The enzyme-catalyzed hydrolysis of the different lignocellulosic substrates (control or pretreated) was carried out in stirred flasks. In a typical hydrolysis reaction, 400 mg of rapeseed hulls samples or the corresponding biomass amount recovered after their pretreatment were added to 18 mL of acetate buffer (50 mM, pH 4.5) and incubated for 2 h (50 °C; 175 rpm). After this preincubation step, hydrolysis was initiated by adding 2 mL of 10 mg mL⁻¹ viscozyme preparation. The final concentrations of lignocellulosic substrate and enzyme in the reaction medium (20 mL) were 2% w/v and 1 mg mL⁻¹, respectively. The hydrolysis reaction was stopped after 24 h by incubating the withdrawn sample at 90 °C for 20 min to deactivate the enzyme.²⁴ Then, the sample was diluted in ultrapure water and filtered (0.45 mm) prior to quantify its fermentable sugars content by DNS method. The yields of the substrate conversion into reducing sugars were expressed as follows:where HS is the hydrolysed sugars quantified with DNS method (g); M_p is the pulps mass treated with enzymes (g).

2.4.7 Total reducing sugars. The amount of total reducing sugars (TRS) in the product sample was measured using the DNS method. 0.5 mL of supernatant from the product sample was mixed with 1 mL of DNS (3,5-dinitrosalicylic acid) reagent. The resulting solution was heated in boiling water for 5 min. The absorbance of the mixture was measured at 540 nm using a UV-vis spectrophotometer. The TRS concentration was calculated by using a calibration curve of the standard glucose solution in accordance with the standard laboratory analytical procedures provided by the National Renewable Energy Laboratory (NREL).

2.4.8 Severity correlation. The severity correlation which describes the severity of the pretreatment as a function of total treatment duration (t_tot, min), temperature (T, °C) and final pH of the black liquor, was calculated using the formula stated below:²⁵where CS is the severity factor, T is the process temperature (°C), T_ref = 100 °C, 14.75 is the activation energy value (J mol⁻¹) in the conditions where process kinetics are of first order and obey to Arrhenius law; and pH is the black liquor pH.

2.4.9 Scanning electron microscopy. The morphology and organization of control and pretreated pulps samples particles was investigated by scanning electron microscopy (SEM). An environmental high-resolution electron scanning microscope QUANTA 250 FEG (FEI Company) was used in low-vacuum mode.

3. Results and discussion

3.1 Composition of raw material

3.1.1 General composition. Table 2 shows the composition of the raw material which is in good agreement with other values found in the literature for this material.²⁵ Rapeseed hulls have a relatively high proportion of acid insoluble residue (47.41 ± 0.58%). Holocelluloses accounted for 28.5% of the whole raw material (% dry weight, w/w) including 12.01 ± 0.37% of glucans, 5.67 ± 0.50% of arabinans, 3.57 ± 0.13% of galactans and 0.98 ± 0.08% of mannans.

Table 2 Raw material composition (%)

Component	Composition (% dry weight, w/w)
Holocelluloses	28.46 ± 1.71
Acid insoluble residue	47.41 ± 0.58
Soluble phenolics	1.44 ± 0.05
Extractives	9.25 ± 0.38
Proteins	13.38 ± 0.53
Ashes	3.75 ± 0.08

---

3.1.2 Composition of the acid insoluble fraction. The composition of the non saccharidic fraction of rapeseed hulls has not been the subject of many works in the literature. The absence of lignin was not clearly reported, till now. The solid residue recovered after acid hydrolysis (acid insoluble residue (AIR), usually called Klason lignin) was analyzed by solid state ¹³C NMR (results not shown). The spectrum revealed the presence of sugars and phenols and important signals at 20–40 ppm and 170–175 ppm were assigned to fatty acids.

The acid insoluble residue (AIR) was also analyzed by thermodesorption coupled to GC-MS, which is a robust and reliable method for identification and quantification of volatile degradation products formed during mild pyrolysis. The main identified compounds are given in the Table 3 with their retention time and relative concentration. A high content in fatty acids and/or fatty acid derivatives and significant amount of phenolic compounds (cresol and catechols) were observed. The phenolic compounds identified are given in the Table 3. Interestingly, it was previously described that phytomelanin layers from sunflower seeds and other seeds are catechol-type phytomelanin and that catechol melanin may be fairly widespread in plant kingdom. The absence of intense signal at 57 ppm in the NMR spectrum associated to the methoxy group of lignin as well as the absence of methoxylated aromatic compounds in Pyr-GC-MS seems to confirm that rapeseed hulls constitute a non lignified feedstock.

Table 3 Retention time (RT) and relative concentration (RC) of identified compounds in the acid insoluble fraction with Pyr-GC-MS analysis

Component	RT, min	RC, %
Oleic acid	39.94	5.10
Hexadecanoic acid	40.01	3.60
Octadecanoic acid	40.22	2.80
Stigmastan-3,5-diene	54.45	2.00
Catechol	17.84	1.34
9-Hexadecenoic acid	35.88	1.10
13-Octadecenoic acid	41.51	1.10
p-Cresol	14.06	0.80
Campesterol	52.88	0.80
Phenol-3-phenoxy	30.85	0.70
Linoleic acid	41.16	0.70
Phenol	10.79	0.60
8-Heptadecen	30.28	0.60
Dimethylphenol	16.38	0.20
4-Methyl-benzen-1,2-diol	20.58	0.20

---

3.2 Pretreatment

Taking into account the chemical composition of the non-saccharidic fraction of rapeseed hulls (mainly phenolics and fatty acids), pretreatment in a basic water-based medium appears to be suitable for its extraction. Moreover, combinative processes involving a physical pretreatment step have also been experimented in this study to reduce recalcitrance of rapeseed hulls. The goal is to alter cell wall structural features so that the polysaccharide fractions (mainly cellulose) locked in the phytomelanin layer of plant cell walls can become more accessible and amenable to enzymatic hydrolysis. Four physical pretreatments (ultrasounds (US), microwaves (MW), high voltage electrical discharges (HVED) and pulsed electric fields (PEF)) were applied as a first step followed by a chemical treatment with sodium hydroxide sodium.

3.2.1 Effect of physical pretreatments on acid insoluble residue recovery. Preliminary tests of acid insoluble residue extraction were conducted by applying all physical treatments at equivalent energy (150 kJ kg⁻¹) and the results are shown in Table 4. As illustrated, applying a very low energy input with MW and US is without interest for this study as there is nearly no difference between untreated and treated samples.

Table 4 Effect of the combined treatment: soda (control), PEF + soda, HVED + soda, MW + soda and US + soda on acid insoluble residue (AIR) removal and pulp yield (soda conditions: 60 °C, 2 h, 0.3 mol L⁻¹, physical treatment condition: 150 kJ kg⁻¹)

	Soda	MW + soda	US + soda	PEF + soda	HVED + soda
AIR removal (%)	30.66 ± 0.60	31.44 ± 1.18	31.36 ± 0.36	34.95 ± 1.95	37.78 ± 1.28
Pulps yield (%)	76.27 ± 2.65	77.82 ± 0.83	76.33 ± 0.28	73.36 ± 1.51	72.47 ± 2.07

---

To achieve the objectives of this study, the physical methods must be employed as pretreatment and thus having a limited duration, but assuring efficiency. For that reason, the same duration of 30 minutes was chosen for all treatments. It corresponded to an equivalent energy input of 150 kJ kg⁻¹ for electrical treatments (HVED, PEF) and 1500 kJ kg⁻¹ for ultrasounds and microwave treatments.

Fig. 2 presents the effect of soda and combinative pretreatments on precipitated acid insoluble residue amount, residual acid insoluble residue (AIR) in pulp, acid insoluble residue removal and pulps yield. AIR is recovered from black liquor by precipitation with sulfuric acid (20%). Results show that all studied physical treatments improved the acid insoluble residue recovery as compared to the control (soda treatment alone). The present study reports removal of 35%, 36%, 38% and 42% of the acid insoluble residue (AIR) present in the native matrix of rapeseed hulls, treated with pulsed electric fields, microwaves, high voltage electrical discharges and ultrasounds respectively which represents an increase by 5%, 6%, 8% and 12% respectively compared to the control. The ultrasounds pretreatment allowed thus the highest recovery of acid insoluble residue. The yield of total solid after washing (pulps) was conversely proportional to acid insoluble residue removal as illustrated in Fig. 2d. Dilute NaOH treatment of cellulosic biomass causes swelling, resulting in an increase in internal surface area, a decrease in the degree of polymerization, a decrease in crystallinity, separation of structural linkages between acid insoluble residue and carbohydrates, and disruption of the acid insoluble residue structure. The additional effect observed for ultrasounds assisted pretreatment in the considered conditions could be due to the more effective cellulosic structure disrupting by explosion of cavitation bubbles induced by ultrasounds. These results are in accordance with the work of Sasmal et al., 2012 (ref. 26), for the ultrasounds assisted lime pretreatment of lignocellulosic biomass toward bioethanol production. Other studied physical pretreatments have also shown effectiveness in the extraction of AIR and recovery of sugars and this thanks to (1) the induced molecules vibration,²⁷ (2) the arc formation and the subsequent shock wave propagation¹⁹ and (3) the cell membrane permeabilisation,²⁸ which are the main effects of microwaves, high voltage electrical discharges and pulsed electric field respectively.

Fig. 2 Effect of the combined treatment: soda (control), PEF + soda, HVED + soda, MW + soda and US + soda on (a), acid insoluble residue precipitated recovery (b), residual acid insoluble residue in pulps (c), acid insoluble residue removal yield and (d) recovered pulps yield (soda conditions: 60 °C, 2 h, 0.3 mol L⁻¹, physical treatment condition: 30 min of total solid–liquid contact time (150 kJ kg⁻¹ for PEF and HVED, 1500 kJ kg⁻¹ for US and MW)).

Fig. 3 presents the energy consumption of HVED, PEF, MW and US for acid insoluble residue extraction. HVED was more effective in terms of energy input to achieve a higher acid insoluble residue removal, while MW treatment required the highest energy input to recover acid insoluble residue in the black liquor. These results can be explained by the physical cell disruption mechanisms involved in each technology. HVED has been demonstrated to induce fragmentation of the particles due to the propagation of shock waves and explosion of cavitation bubbles, therefore facilitating the extraction of soluble biomolecules.¹⁹ Moreover, some hydrodynamic modifications can occur after HVED treatment.¹⁹ Note that the energy input applied for HVED and PEF (150 kJ kg⁻¹) and for MW and US (1500 kJ kg⁻¹) is slightly higher than that usually applied in the case of biomolecules extraction such as polyphenols and proteins.²⁹ The higher energy input required for acid insoluble residue recovery is due to this interaction with cellulose and hemicelluloses which is thus less accessible.

Fig. 3 Energy consumption of HVED, PEF, MW and US treatments for acid insoluble residue recovery (soda conditions: 60 °C, 2 h, 0.3 mol L⁻¹, physical treatment condition: 30 min of total solid–liquid contact time (150 kJ kg⁻¹ for PEF and HVED, 1500 kJ kg⁻¹ for US and MW)).

Fig. 4 illustrates the recovery yield of the acid insoluble residue (AIR) as a function of treatment severity for the combined treatment: soda (control), PEF + soda, HVED + soda, MW + soda and US + soda. It appears that an increase in combined physical/soda treatment severity resulted in the improvement of the acid insoluble residue (AIR) removal. However, control experiment (soda without physical pretreatment) showed a severity value lower than that of US + soda but giving a lower delignification yield by 12%.

Fig. 4 Effect of the severity of the combined treatment: soda (control), PEF + soda, HVED + soda, MW + soda and US + soda on acid insoluble residue removal efficiency (soda conditions: 60 °C, 2 h, 0.3 mol L⁻¹, physical treatment condition: 30 min of total solid–liquid contact time (150 kJ kg⁻¹ for PEF and HVED, 1500 kJ kg⁻¹ for US and MW)).

3.2.2 Effect of physical pretreatments on sugars recovery and fibers properties. Table 5 shows the yields of cellulose and hemicelluloses in recovered pulps for the combined treatment: soda (control), PEF + soda, HVED + soda, MW + soda and US + soda. The results indicate a high recovery of hemicelluloses and cellulose irrespective of applied physical treatment. This is consistent with the typical polymeric solids composition after biomass fractionation as a function of reaction pH characteristic of each pretreatment proposed by Carvalheiro et al., 2008.⁵ Applied pretreatments render sugars mainly in the recovered pulps which can be upgraded by farther treatment like enzyme hydrolysis for bioethanol production.

Table 5 Effect of the combined treatment: soda (control), PEF + soda, HVED + soda, MW + soda and US + soda on sugar content of recovered pulps (soda conditions: 60 °C, 2 h, 0.3 mol L⁻¹, physical treatment condition: 30 min of total solid–liquid contact time (150 kJ kg⁻¹ for PEF and HVED, 1500 kJ kg⁻¹ for US and MW))

	Raw material	Soda	MW + soda	US + soda	PEF + soda	HVED + soda
Cellulose (%)	12.01 ± 0.37	12.19 ± 1.08	11.83 ± 0.90	12.28 ± 0.13	10.86 ± 0.36	10.66 ± 0.08
Hemicelluloses (%)	15.47 ± 0.67	14.15 ± 1.30	12.61 ± 1.16	13.98 ± 0.05	13.92 ± 0.46	13.83 ± 0.14

---

Table 6 presents the effect of the combined treatment on the evolution of pulps yield and the content of acid insoluble residue (AIR), acid soluble phenolics (ASP), total reducing sugars (TRS) in the obtained black liquor. While a fraction of the acid insoluble residue has been precipitated during conventional acid hydrolysis of pulps, acid-soluble residue was quantified in the black liquor as phenolic compounds of low molar mass. These phenolics are detrimental to fermentation by microorganisms.³⁰ The AIR recovery in the black liquor shows a particular efficiency of physically assisted pretreatments in the recovery of a less contaminated sugar fraction. Sugars yields recovered in black liquor are nearly the same irrespective of the applied physical treatment. This is in agreement with previous pulps characterization (Table 5). It has also been reported that ultrasound assisted alkali pretreatment process not only eliminates acid insoluble residue but also improves the quality and recovery of pretreated lignocellulosic biomass.²⁶

Table 6 Effect of the combined treatment: soda (control), PEF + soda, HVED + soda, MW + soda and US + soda on the removal pulp yield and the content of acid insoluble residue (AIR), acid soluble phenolics (ASP), total reducing sugars (TRS) in the obtained black liquor (soda conditions: 60 °C, 2 h, 0.3 mol L⁻¹, physical treatment condition: 30 min of total solid–liquid contact time (150 kJ kg⁻¹ for PEF and HVED, 1500 kJ kg⁻¹ for US and MW))

%		Soda	US + soda	MW + soda	HVED + soda	PEF + soda
Pulps		76.27 ± 2.65	67.26 ± 2.77	70.00 ± 4.80	72.47 ± 2.07	73.36 ± 1.51
Black liquor	AIR	14.54 ± 0.29	19.97 ± 0.77	16.90 ± 0.68	17.91 ± 0.61	16.57 ± 0.93
	ASP	2.13 ± 0.14	2.45 ± 0.36	2.18 ± 0.44	1.69 ± 0.41	1.60 ± 0.38
	TRS	2.00 ± 0.30	1.71 ± 0.02	1.49 ± 0.15	1.24 ± 0.06	1.81 ± 0.07

---

Fig. 5 compares the recovered samples from the combined treatment (soda, PEF + soda, HVED + soda, US + soda, MW + soda) based on their enzymatic saccharification performances after 24 h. The enzymatic digestibility of physically-pretreated rapeseed hulls at 60 °C was higher than the one obtained for the control (soda). It has been shown in previous studies,³¹ that microwave-assisted alkali treatment is an efficient way to improve the enzymatic digestibility of switchgrass. For similar operating conditions but a temperature of 70 °C and a longer alkali treatment (2 h) a recovery of 36.4 g/100 g biomass was observed after alkali microwave assisted treatment of switchgrass.³¹ It has also been concluded by Abeywickrama et al., 2012 (ref. 32) that parallel microwave technology could be used for rapid optimization of biomass pretreatment.

Fig. 5 Effect of the combined treatment (soda (control), PEF + soda, HVED + soda, US + soda, MW + soda) on sugar recovery after 24 hours of enzyme hydrolysis (soda conditions: 60 °C, 2 h, 0.3 mol L⁻¹, physical treatment condition: 30 min of total solid–liquid contact time (150 kJ kg⁻¹ for PEF and HVED, 1500 kJ kg⁻¹ for US)).

Fig. 6 presents SEM images used to complete the analysis on the effect of the physical pretreatment processes on the samples structure. The control sample exhibited a low porosity and highly tied cells (Fig. 6B). In contrast, the pulps obtained after physical treatments appeared to be more distorted (Fig. 6C–F) because of the increase of the external surface area and the porosity. Sample treated with US demonstrated a higher porosity allowing to be more susceptible to enzymatic attack. In fact this treatment broke rigid structure of rapeseed hulls due to partial removal of hemicellulose and phytomelanin layer as revealed by yields results (Fig. 2).

Fig. 6 SEM photographs of raw material (A), control (soda) (B) and pretreated pulps samples: PEF + soda (C); MW + soda (D); HVED + soda (E) and US + soda (F) (magnification ×100).

4. Conclusion

This work is an innovative approach to study combined treatments effect on pulps and acid insoluble residue generated aiming at an integral valorization of highly recalcitrant by-products. The above investigation revealed that physical-assisted soda promoted the extraction of the phytomelanin layer of rapeseed hulls and enhanced the enzymatic hydrolysis of recovered pulps. The results particularly highlighted the effectiveness of ultrasounds for biocompounds recovery and enzyme hydrolysis. However high voltage electrical discharges followed closely the effectiveness observed for ultrasounds but is less energy-consuming. However, more studies are required to elucidate the effects of these technologies on other raw materials delignification.

Acknowledgements

This work was performed in partnership with the SAS PIVERT. Within the frame of the French Institute for the Energy Transition (Institut pour la Transition Energétique (ITE)) P.I.V.E.R.T. (http://www.institut-pivert.com) selected as an Investment for the Future (“Investissements d’Avenir”). This work was supported as part of the Investments for the Future by the French Government under reference ANR-001-01.

References

1. M. Naczk, R. Amarowicz, R. Zadernowski and F. Shahidi, Food Chem., 2001, 73, 467–471.
2. F. Karaosmanoğlu, E. Tetik and E. Göllü, Fuel Process. Technol., 1999, 59, 1–12.
3. N. Mosier, C. Wyman, B. Dale, R. Elander, Y. Y. Lee, M. Holtzapple and M. Ladisch, Bioresour. Technol., 2005, 96, 673–686.
4. D. Chiaramonti, M. Prussi, S. Ferrero, L. Oriani, P. Ottonello, P. Torre and F. Cherchi, Biomass Bioenergy, 2012, 46, 25–35.
5. F. Carvalheiro, L. C. Duarte and F. M. Gírio, J. Sci. Ind. Res., 2008, 67, 849–864.
6. A. T. W. M. Hendriks and G. Zeeman, Bioresour. Technol., 2009, 100, 10–18.
7. P. Kumar, D. M. Barrett, M. J. Delwiche and P. Stroeve, Ind. Eng. Chem. Res., 2009, 48, 3713–3729.
8. X. Chen, Y. Gao, L. Wang, H. Chen and N. Yan, ChemPlusChem, 2015, 80, 1565–1572.
9. M. Balat, Energy Convers. Manage., 2011, 52, 858–875.
10. X. Zhao, F. Peng, K. Cheng and D. Liu, Enzyme Microb. Technol., 2009, 44, 17–23.
11. Q.-A. Zhang, Z.-Q. Zhang, X.-F. Yue, X.-H. Fan, T. Li and S.-F. Chen, Food Chem., 2009, 116, 513–518.
12. S. Balachandran, S. E. Kentish, R. Mawson and M. Ashokkumar, Ultrason. Sonochem., 2006, 13, 471–479.
13. M. Romdhane, Chem. Eng. J., 2002, 87, 11–19.
14. J. H. Gibson, H. Hon, R. Farnood, I. G. Droppo and P. Seto, Water Res., 2009, 43, 2251–2259.
15. R. Halim, M. K. Danquah and P. A. Webley, Biotechnol. Adv., 2012, 30, 709–732.
16. S. Nitayavardhana, P. Shrestha, M. L. Rasmussen, B. P. Lamsal, J. (Hans) van Leeuwen and S. K. Khanal, Bioresour. Technol., 2010, 101, 2741–2747.
17. E. Vorobiev and N. Lebovka, Food Eng. Rev., 2010, 2, 95–108.
18. S. Töpfl, Design, 2006, 194.
19. N. Boussetta, E. Soichi, J.-L. Lanoisellé and E. Vorobiev, Ind. Crops Prod., 2014, 52, 347–353.
20. H. Zuckerman, Y. E. Krasik and J. Felsteiner, Innovative Food Sci. Emerging Technol., 2002, 3, 329–336.
21. N. Boussetta, E. Vorobiev, L. H. Le, A. Cordin-Falcimaigne and J.-L. Lanoisellé, LWT--Food Sci. Technol., 2012, 46, 127–134.
22. M. N. Mohamad Ibrahim, N. Zakaria, C. S. Sipaut, O. Sulaiman and R. Hashim, Carbohydr. Polym., 2011, 86, 112–119.
23. M. Brahim, F. Gambier and N. Brosse, Ind. Crops Prod., 2014, 52, 18–22.
24. T. Auxenfans, S. Buchoux, K. Djellab, C. Avondo, E. Husson and C. Sarazin, Carbohydr. Polym., 2012, 90, 805–813.
25. M. Pedersen and A. S. Meyer, New Biotechnol., 2010, 27, 739–750.
26. S. Sasmal, V. V. Goud and K. Mohanty, Energy Fuels, 2012, 26, 3777–3784.
27. U. Zimmermann, F. Riemann and G. Pilwat, Biochim. Biophys. Acta, Biomembr., 1976, 436, 460–474.
28. J. R. Sarkis, N. Boussetta, C. Blouet, I. C. Tessaro, L. D. F. Marczak and E. Vorobiev, Innovative Food Sci. Emerging Technol., 2015, 29, 170–177.
29. Z. Hu and Z. Wen, Biochem. Eng. J., 2008, 38, 369–378.
30. S. Yasuda, K. Fukushima and A. Kakehi, J. Wood Sci., 2001, 47, 69–72.
31. H. Zhenhu and W. Zhiyou, Biochem. Eng. J., 2008, 38, 369–378.
32. C. J. Abeywickrama, Y. P. Timilsena, S. K. Rakshit, L. Chrusciel and N. Brosse, Ind. Eng. Chem. Res., 2013, 52(4), 1691–1697.

---