Combination of steam explosion and TEMPO-mediated oxidation as pretreatments to produce nanofibril of cellulose from Posidonia oceanica bleached fibres

Abstract

Posidonia oceanica is the dominant sea grass in the Mediterranean Sea. This biomass has great potential for use as a novel lignocellulosic material on an industrial scale. In this work, an innovative approach was applied to produce cellulose nanofibril (CNF) from Posidonia. First, fibres were isolated by a delignification-bleaching process, followed by refining and TEMPO-mediated oxidation to facilitate their further microfibrillation. Cellulose nanofibril suspensions were then produced by steam explosion or grinding (Masuko supermasscolloider). Next, CNF gel-like suspensions were characterized by several techniques such as morphological analysis (Morfi, optical microscopy, atomic force microscopy, transmission electron microscopy) and turbidity measurements. Nanopapers, prepared by filtration, were used to perform tensile tests. Finally, the efficiency of the combination of TEMPO-mediated oxidation and steam explosion was discussed. Obtained results show that the steam explosion process allows the production of CNF with a width between 4 and 10 nm and properties close to those obtained by the grinding process.

Full Article

Combination of Steam Explosion and TEMPO-mediated Oxidation as Pretreatments to Produce Nanofibril of Cellulose from Posidonia oceanica Bleached Fibres

Malek Khadraoui,^a,b Ramzi Khiari,^a,c,d,* Nicolas Brosse,^e Latifa Bergaoui,^b and Evelyne Mauret ^a,*

Posidonia oceanica is the dominant sea grass in the Mediterranean Sea. This biomass has great potential for use as a novel lignocellulosic material on an industrial scale. In this work, an innovative approach was applied to produce cellulose nanofibril (CNF) from Posidonia. First, fibres were isolated by a delignification-bleaching process, followed by refining and TEMPO-mediated oxidation to facilitate their further microfibrillation. Cellulose nanofibril suspensions were then produced by steam explosion or grinding (Masuko supermasscolloider). Next, CNF gel-like suspensions were characterized by several techniques such as morphological analysis (Morfi, optical microscopy, atomic force microscopy, transmission electron microscopy) and turbidity measurements. Nanopapers, prepared by filtration, were used to perform tensile tests. Finally, the efficiency of the combination of TEMPO-mediated oxidation and steam explosion was discussed. Obtained results show that the steam explosion process allows the production of CNF with a width between 4 and 10 nm and properties close to those obtained by the grinding process.

DOI: 10.15376/biores.17.2.2933-2958

Keywords: Posidonia oceanica; Cellulose nanofibril; Steam explosion; Grinding; TEMPO-mediated oxidation; Quality index

Contact information: a: Univ. Grenoble Alpes, CNRS, Grenoble INP (Institute of Engineering Univ. Grenoble Alpes), LGP2, F-38000 Grenoble, France; b: Carthage University, National Institute of Applied Sciences and Technology, EcoChimie Laboratory, Tunis, Tunisia; c: University of Monastir, Faculty of Sciences of Monastir, Laboratory of Environmental Chemistry and Clean Process (LCE2P- LR21ES04), 5019 Monastir, Tunisia; d: Higher Institute of Technological Studies of Ksar Hellal, Department of Textile, Ksar Hellal, Tunisia; e: Université de Lorraine – Faculté des Sciences et Technologies, Laboratoire D’Etudes et de Recherche sur le Matériau Bois, 54506, Vandœuvre-Lès-Nancy Cedex, France;

*Corresponding authors: khiari_ramzi2000@yahoo.fr; Evelyne.mauret@grenoble-inp.fr; Univ. Grenoble Alpes, CNRS, Grenoble INP (Institute of Engineering Univ. Grenoble Alpes), LGP2, F-38000 Grenoble, France

GRAPHICAL ABSTRACT

INTRODUCTION

Cellulosic fibres are used in various fields, such as pulp and paper industries, and for several applications in packaging, automotive industry, building construction, etc. (Belgacem and Gandini 2011; Khiari and Belgacem 2017).

Of all the available wide variety of lignocellulosic sources, wood is the most commonly used in the world (Klemm et al. 2005). However, in forest-deficient areas, the utilization of other types of fibres has been developed. Currently, great attention is focused on the exploitation of residues and agricultural crops as alternative sources of cellulosic fibres: kenaf (Villar et al. 2009; Valdebenito et al. 2017), hemp bast (Puangsin et al. 2012), flax (Cao et al. 2007; Alila et al. 2013), wheat straw (Josset et al. 2014), sisal (Siqueira et al. 2010), rice straw (Jiang and Hsieh 2013), bamboo residue (Liu et al. 2010), bagasse (Bhattacharya et al. 2008; Afra et al. 2013; Grande et al. 2018), date palm tree (Benhamou et al. 2014), alfa (Marrakchi et al. 2011), corn cobs and husk (Shogren et al. 2011), sugar beet pulp (Habibi and Vignon 2008; Li et al. 2014), banana rachis and culinary banana peel (Zuluaga et al. 2009; Khawas and Deka 2016), and Posidonia oceanica (Bettaieb et al. 2015).

This study deals with the investigation of fragments of Posidonia (Posidonia oceanica), an abundant marine plant in the Mediterranean Sea. The debris from this marine biomass are released by the sea in the form of balls and/or leaves and are accumulated in large quantities on the coasts during each summer season, leading to regular beach cleanings. Therefore, the valorisation of this waste as a source of fibres can be of interest.

Some studies have already been conducted on the delignification of Posidonia to produce cellulose, cellulose derivatives (Aguir and M’Henni 2006; Khiari et al. 2010), or composites (Rammou et al. 2021). Other researchers suggest the exploitation of Posidonia waste as an adsorbent material for dyes (Ncibi et al. 2006, 2007; Guezguez et al. 2009; Dural et al. 2011; Douissa et al. 2013). These materials have also been studied to produce a renewable and biodegradable nanomaterial called nanocellulose (Bettaieb et al. 2015). Since the 1980s, cellulose nanofibril (CNF, also called nanofibrillated cellulose) has been produced, traditionally by the application of specific mechanical treatments, such as microfluidization (Zimmermann et al. 2004), homogenization (Dinand et al. 1999; Dufresne et al. 2000), grinding (Abe et al. 2007; Afra et al. 2013; Nair et al. 2014), or cryo-cruching (Dufresne et al. 1997; Chakraborty et al. 2005; Alemdar and Sain 2008). Although these processes allow for the preparation of CNF exhibiting good properties, a major problem in all cases, regardless of the processes applied, is the high energy requirement (Spence et al. 2011). This is why great attention has been paid to the development of pretreatments leading to a reduction of the energy cost by facilitating the mechanical microfibrillation.

The most common pretreatments are biological and chemical: enzyme-assisted hydrolysis (Henriksson et al. 2007; Pääkkö et al. 2007), TEMPO-mediated oxidation (Saito et al. 2006), carboxymethylation (Heinze and Koschella 2005), and/or solvent-assisted pretreatments (Sirviö et al. 2015). In addition, other studies have been oriented towards the use of extrusion with the objective of increasing the concentration of the CNF gel-like suspensions (Ho et al. 2015; Rol et al. 2017).

Steam explosion (SE), which is considered to be inexpensive and clean, has also attracted several researchers for the production of CNF (Cherian et al. 2010; Deepa et al. 2011; Diop et al. 2015; Bansal et al. 2016; Phinichka and Kaenthong 2018). This is a thermomechanical process that allows the fragmentation and/or break-up of the lignocellulosic material by the combination of several factors, namely temperature in the reactor and high shear forces resulting from the sudden pressure drop when the treated material is blown into the final tank. The steam explosion process was firstly introduced for wood disintegration in the manufacture of panels (Mason 1926). Moreover, to produce fermentable sugars and alcohol from wood, Babcock and Kenvii (1932) used this process as a pretreatment of the raw material. In severe conditions, such as long residence time in the reactor or with the use of chemicals, this process leads to the depolymerisation of some lignin and the hydrolysis of hemicelluloses to elementary sugars and water-soluble oligomers (Fernández-Bolaños et al. 2001; Sun et al. 2005). Recently, this thermo-mechanical process has gained attention again for the separation of cellulose fibres from different lignocellulosic sources (Kaushik and Singh 2011; Bansal et al. 2016; Saelee et al. 2016; Luo et al. 2018). At the same time, this treatment was proposed for producing CNF from different lignocellulosic materials (Table 1).

Table 1. Various Processing Conditions for Fibres and CNF Production Using Steam Explosion Apparatus or Modified Autoclaves System

Various processing conditions of steam explosion were used to separate and/or defibrillate the fibres. However, several authors have worked with modified autoclave systems that do not allow applying high steam pressure, constituting a real limitation. Subsequently, the depressurization stage is not well controlled and its intensity not sufficient. In the literature, it was referred that steam explosion is composed by two steps: steam cracking and explosive decompression. Steam cracking consists in the penetration of water into the material in the form of steam under high pressure and at a high temperature. During the second stage called “explosive decompression”, the wet saturated lignocellulosic material “bursts” as a result of the sudden release of pressure in the reactor. It was shown that the pressure difference between the inside of the reactor and atmospheric pressure is proportional to the intensity of the shear forces applied to the biomass. So, taking advantage of this effect, the authors decided to use steam explosion for microfibrillation.

This paper proposes to study the steam explosion process SE to produce CNF from Posidonia oceanica (balls and leaves). A steam explosion pilot with a well-controlled decompressive stage was used. It was coupled to a conventional chemical treatment, i.e., TEMPO-mediated oxidation. Then, the obtained CNF was characterized and compared to TEMPO-oxidized CNF produced by conventional grinding. The objective is to understand the impact of the steam explosion on CNF morphologies (optical microscopy, atomic force microscopy, transmission electron microscopy, turbidity, and nanosized fraction) and mechanical properties (Young’s modulus). Finally, the results were compared with available data from the literature dealing with wood, non-wood, and agricultural crop-based sources. The novelty of this study was to use steam explosion for the first time as the main process for microfibrillation of delignified and bleached Posidonia fibres (combined or not to a chemical pretreatment). This is could be considered as a short-time process with less reduced environmental impact. The difference with the previous studies is that the explosion conditions were different mostly realized in autoclaves, not in steam explosion pilots with a well-controlled decompressive stage.

EXPERIMENTAL

Materials and Chemical Products

Posidonia oceanica balls (POB) and leaves (POL) were used as a starting raw material to produce CNF fibres. These wastes were collected in Monastir, Tunisia in December 2020. After a washing step, to eliminate sand and contaminants, the material was dried under ambient conditions for one month (average relative humidity: 71%; average temperature: 26 °C). All materials and chemicals were used as received from the suppliers without further purification.

Delignification and Bleaching of Posidonia oceanica Fibres

Bleached fibres from Posidonia balls (F-POB) and leaves (F-POL) were obtained after two operations, namely, delignification of the raw material and bleaching of the fibrous suspension. Delignification was carried out using soda in the presence of a small quantity of anthraquinone. This method has been commonly used for pulping non-wood plants (Antunes et al. 2000). More precisely, the delignification of the Posidonia was performed according to the procedure described by Khiari et al. (2010). For this purpose, 60 g (dry material) of raw material was treated at 160 °C with an alkali charge (sodium hydroxide) of 20% (w/w, based on dry material) and a liquor to solid ratio fixed at 10. Additionally, 0.1% (w/w, dry material) of anthraquinone was added as a catalyst. Pulping was carried out in electrically-heated rotating autoclaves (one-litre volume) for 120 min. Afterwards, the pulp was washed and filtered using a nylon mesh (100 μm) and finally was stored in a freezer.

After delignification, the bleaching step was completed with sodium chlorite (NaClO₂) in a buffer medium. This treatment mainly consisted of two phases: a sodium chlorite bleaching phase followed by an alkaline extraction. For the bleaching phase, 50 g of the delignified pulp were treated in a heat-sealable bag with buffer solution (pH <4.7) prepared from sodium acetate trihydrate and acetic acid. Then 3.95 g of NaClO₂ dissolved in 100 mL of deionized water were added and the bag was filled with water to 1 kg. Then, it was kept in a water bath for 90 min at 80 ° C and shaken manually every 5 min. At the end, sodium sulfite was added to neutralize the chlorine dioxide, and then the pulp was filtered with a nylon sieve of 40 μm mesh size and washed with water.

An alkaline extraction is then performed to facilitate lignin extraction. In a heat-sealable bag, the treated pulp is introduced into 400 mL of a sodium hydroxide solution (3% w/w). The bag was filled up to 1 kg with deionized water, then closed, kept in a water bath at 70 °C for 60 min and stirred every 5 min. Afterwards, the pulp was filtered with a nylon sieve of 40 μm mesh size and washed with water. Those two steps were repeated 5 times.

CNF Production

Fibres refining

Using the PFI mill, as per ISO 5264-2 (2011) standard, 300 mL of an aqueous suspension containing 30 g of dry bleached pulp (10% w/w) were refined until a drainage index of 82 °SR (Schopper Riegler degree), which corresponds to 6000 and 4000 revolutions of the mill for F-POB and F-POL, respectively.

TEMPO-mediated oxidation of fibres

Chemical modification of the fibres by TEMPO (2,2,6,6-tetramethylpiperidine-1-oxide radical)-oxidation was performed to selectively convert the primary hydroxyl of cellulose to carboxylate functions (Saito et al. 2006; Isogai et al. 2011). The bleached and refined pulp at a concentration of 1% (w/w) was treated according to the protocol proposed by Saito et al. (2006). The TEMPO/NaBr/NaClO system was used at 25 °C with 0.1/1/6.22 mmol of the reactants per gram of cellulose (mmol/g), respectively. The ensuing TEMPO-oxidized cellulose fibres were filtered and washed using deionized water until the filtrate conductivity reached a value below 5 μS cm^-1.

Microfibrillation by steam explosion or grinding

Steam explosion experiments were conducted at LERMAB, Nancy, France. These experiments involved three main steps. First, 100 g of dry fibres were placed in a 4.8-L reactor that once closed, was saturated with steam until the target couple of temperature/pressure (170 °C/8 bar) was reached. These conditions were kept for 30 s. The release regulator was then opened, and the material was subjected to a sudden pressure drop (depressurization) during its rapid ejection from the reactor to be finally recovered in the blow tank. These conditions were selected after a preliminary study. In particular, the low residence time was chosen to limit the degradation of cellulose and hemicelluloses.

Fig. 1. Schematic diagram of the proposed pathways to produce CNF from POB and POL

In this study, the main objective of the steam explosion was to take advantage of the sudden drop of pressure and to assess if a possible effect on the fibres (expected breakage of the cell wall) may be sufficient to produce CNF from TEMPO-oxidized fibres.

Grinding was also done using the ultrafine friction grinder Supermasscolloider (model MKZA6-2, disk model MKG-C 80, Masuko Sangyo Co., Ltd., Saitama, Japan), available at LGP2, Grenoble, France. The ultrafine grinder was used at 1500 rpm to treat the fibre suspension. The nominal gaps between the rotating and the static stones were -5 for 10 passes and -10 for 15 passes. Figure 1 presents the proposed routes leading to the preparation of CNF using the different selected processes (cooking, bleaching, TEMPO oxidation, steam explosion, or grinding). Table 2 summarizes all sample labels.

Table 2. Abbreviations of the Different Samples

Characterisation

Several methods and techniques were used to characterize the obtained fibres as well as the produced CNF.

Carboxylic acid content

Conductometric titrations were used to determine the carboxylate content of the TEMPO-oxidized fibres (Besbes et al. 2011). To this purpose, TEMPO-oxidized fibres (0.5 g dry matter) were suspended in 10 mL of HCl (0.1 mol L^-1) in such a way that the pH reached 3. Afterwards, the suspension was titrated with NaOH (0.01 mol L^-1). Titration curves were used to calculate the carboxyl content according to the following Eq. 1,

(1)

where C_NaOH is the concentration of NaOH (mmol L^-1), V₁ is the equivalent volume of NaOH solution added to titrate the strong acid corresponding to the excess of HCl (L), V₂ is the equivalent volume of NaOH solution added to titrate the weak acid corresponding to the carboxyl content (L), and w is the dry weight of sample (g).

Morphological properties

The morphological properties of the prepared samples were investigated. To determine the size distribution and fines content of the suspensions, a MorFi LB-01 fibre analyser (Techpap, Gières, France) was used. Suspensions were diluted to 0.3 g L^-1. The fibres/fines limit were set at 200 μm in length, and the analysis time was set at 5 min. The measurements were triplicated.

Optical microscopy was performed in transmission mode, using an Axio Imager M1 optical microscope (Carl Zeiss, Munich, Germany). For each prepared CNF gel-like suspension (at a concentration of 0.5 wt%) at least 10 images were taken.

Atomic force microscopy is a method for visualizing the topography of the sample surface or for characterizing the physical properties of materials with a resolution between 0.1 and 10 nm (under the best conditions for microscopic observation). Atomic force microscopic (DI, Veeco, Instrumentation Group, Plainview, NY, USA) observations were thus performed by using CNF gel-like suspensions diluted at 10^-4% (w/w) with deionized water. A droplet of the prepared sample was deposited on a mica disc and then allowed to dry under ambient conditions for 24 h.

Transmission electron microscopic analysis was performed using a JEOL 200CX transmission electron microscope at 80 kV. The images were obtained after depositing approximately 0.5 μL of the diluted suspension onto a carbon-coated 300-mesh copper grid.

Fourier transform infrared spectroscopy (FTIR) and crystallinity index (CI)

Attenuated total reflectance-Fourier transform infrared (ATR-FTIR) was used to characterize the prepared papers from refined pulp and TEMPO-oxidized pulp (prepared by filtration under vacuum: see material and method, section Quality index).
The spectra were obtained using a spectrophotometer (Spectrum 65 FT-IR, Perkin-Elmer, Waltham, MA, USA). It covered the spectral range from 600 to 4000 cm^-1. The acquisition conditions of this analysis were 16 scans and 2 cm^-1 resolution.

The crystalline structure of the prepared samples was studied using an X-ray diffractometer (XRD) (D8-Advance, Bruker, Billerica, MA, USA) at room temperature with a monochromatic CuKα radiation source (λ = 0.154 nm). Experiments were performed in step-scan mode with a 2θ angle ranging from 5° to 60° with a step of 0.04 and a scanning time of 5.0 min. The CI was calculated using deconvolution method (amorphous subtraction method).

Quality index

A quality index (Desmaisons et al. 2017) was used to facilitate the monitoring of nanocellulose production at laboratory and industrial scales. The simplified quality index (QI^*) was based on four parameters, namely nanosized fraction, turbidity, average microparticle size (µm²) of CNF gel-like suspensions, and Young’s modulus of nanopapers. The QI^* is expressed according to Eq. 2,

QI ^*= 0.3 × x₁ – 0.03 × x₂ – 0.072 × x₃² + 2.54 × x₃ – 5.34 Ln(x₄) + 58.62 (2)

where x₁ represents the nanosized fraction (%), x₂ is the turbidity (NTU), x₃ is the Young’s modulus (GPa), and x₄ is the average microparticle size (µm²).

The nanosized fraction in CNF gel-like suspensions was measured according to Naderi et al. (2015). This method allows determining the quantity of particles at the nanometric scale in the CNF gel-like suspensions by a gravimetric method. At first, the CNF gel-like suspensions were diluted to 0.02% (w/w) and centrifuged at 1000 g for 15 min using a centrifuge (Sigma 3-18 KS, Sigma Laborzentrifugen GmbH, Osterode am Harz, Germany). The concentrations in the supernatant before (C_bc) and after centrifugation (C_ac) were used to calculate the nanosized fraction according to Eq. 3,

(3)

where the concentration is expressed in percentage (w/w).

A turbidimeter (AL250T-IR, Aqualytic, Dortmund, Germany) was used to measure the turbidity of the diluted CNF suspensions (0.1% w/w). Turbidity is related to the size, number, and refractive index of the visible suspended particles. The presence of nanoparticles that are not detectable makes the value of the turbidity close to zero. The average of the turbidity values was obtained from the measurement of five different samples.

The CNF gel-like suspensions were diluted to 0.5 wt%, and the resulting samples were observed with an optical microscope (Axio Imager M1, Carl Zeiss, Munich, Germany). Images were taken at 20x magnification and analysed with Fiji software (ImageJ, Java 1.8.0_172, Madison, WI, USA). They were converted to 8-byte type and thresholded before assessing the average length of the visible particles remaining in the suspension after steam explosion or grinding. At least five pictures were taken, and the most representative ones were used.

Nanopapers with basis weight of approximately 60 g m^-2 were finally produced using a standard sheet former (Rapid-Köthen, Noviprofibre, Eybens, France) as per ISO 5269-2 (2004) standard and equipped with a nylon sieve of 1 μm mesh size (Buisine, Clermont de l’Oise, France). The suspensions were filtered under vacuum, and the obtained wet mats were dried for 12 min at 90 °C. For each CNF suspension, five nanopapers were produced and stored at 23 °C and 50% relative humidity (RH) for at least 48 h before further characterisation. The thickness of the nanopapers was measured using a micrometer (Adamel Lhomargy, Sms-Labo, Saint-Baldoph, France) at different positions on the nanopaper, and an average value was used for further calculation (m).

Finally, the Young’s modulus of the nanopapers were measured. The samples (width of 15 mm and length between jaws of 100 mm) were tested using a vertical extensometer an (Instron 5965, Instron, Norwood, MA, USA) with a capacity of 5 kN and a tensile velocity of 10 mm/min.

RESULTS AND DISCUSSION

Fibre Properties

Carboxylic acid content

As previously explained, bleached and refined fibres, R-FOB and R-FOL, were submitted to a TEMPO-mediated oxidation (TEMPO/NaBr/NaClO).

Table 3. Carboxylic Content of Various Biomasses Treated under Different Conditions by TEMPO-mediated Oxidation

This reaction was intended to modify the cellulose and to convert the primary alcohols of carbons from glucose rings to carboxylic acid cycles. Under the tested conditions, the treated samples were produced with important carboxylic contents, which are 0.97 mmol g^-1 and 1.07 mmol g^-1 for T-RFOB and T-RFOL, respectively.

The present results agree with previous studies (Bettaieb et al. 2015). Table 3 summarizes the carboxylic acid content of various biomasses treated under different conditions by TEMPO-mediated oxidation, showing that the amounts of acidic groups ranged from 0.3 to 1.75 mmol g^-1. This range depended on many factors, especially the chemical composition of the starting materials, the different treatments undergone by the fibres during the bleaching step, and finally the quantities of TEMPO/NaBr/NaClO reagents. However, it is worth noting that 0.3 mmol g^-1 of ionic groups seemed to be a minimal level to perceive an effect on the microfibrillation behavior (Besbes et al. 2011).

Morphological properties

Fibre morphology was assessed for the unrefined and refined pulps as well as for TEMPO-oxidized fibres by using SEM analysis and MorFi Lab. From SEM observations, it appears that the length of Posidonia oceanica fibres was remarkably lower than that observed for other common plants. MorFi analysis confirmed this observation: the average length and width as well as the percentage in fine elements are summarized in Table 4.

Table 4. Morphological Properties and Crystallinity of Bleached F-POB and F-POL, Refined R-FOB and R-FOL, and TEMPO-oxidized Fibres T-RFOB and T-RFOL