Effect of Y-Methacryloxypropyltrimethoxysilane (MPS) and Tetraethoxysilane
(TEOS) Towards Preparation of Oil Absorbent Foams from Polyvinyl Alcohol
(PVA) Reinforced with Microfibrillated Cellulose (MFC)

Increasing usage of foams in various industry sectors had causing serious disposal problems once it reaches the end of its life-cycle. Herein, PVA-MFC foam was prepared by freeze-drying using polyvinyl alcohol (PVA) and microfibrillated cellulose (MFC) as a reinforced material from sugarcane bagasse (SCB). In this study, the PVA-MFC foam was chemically silylated with Y-methacryloxypropyltrimethoxysilane (MPS) and tetraethoxysilane (TEOS). The wetting ability and mechanical strength of the silylated 2,20PVA-MFC foam was greatly enhanced compared with unmodified 2,20PVA-MFC foam. The silane chemicals (MPS and TEOS) had been confirmed grafted on 2,20PVA-MFC foam due to the presence of Si-C and Si-O-C stretching vibration as showed in Fourier Transform Infrared (FTIR) spectra and cloud-like coating of porous pore was observed in scanning electron microscopy (SEM) images. The silylated 2,20PVA-MFC foam (MPS and TEOS) exhibited a series of desirable properties such as lower swelling ratio and high absorption capacity of solvents and oils but had low thermal stability in thermogravimetric (TGA) analysis. The characterization of 2,20PVA-MFC foam using TEOS was further investigated. A significant difference in morphology was clearly observed between the unmodified and silylated 2,20PVA-MFC-TEOS foam through field emission scanning electron microscopy (FESEM) images. The X-ray photoelectron (XPS) analysis of silylated 2,20PVA-MFC-TEOS foam confirmed the presence of C, O and trace amount of Si elements. These synthesized 2,20PVA-MFC foam could be a promising material for broad range of polymer foam applications.


Introduction
Polymer foams have attracted considerable attention toward various applications such as thermal and acoustic insulation, energy dissipation, packaging, shock protection and many more [1]. However, these polymer foams are made from non-degradable polymers which are inevitably brought substantive waste and serious environmental pollution [2]. Recently, trend toward fabrication of degradable polymer foams reinforced with renewable resources such as natural fibers is considerable increased due to the demand of ecofriendly and non-toxic material. Hence, the use of natural fibers such as microfibrillated cellulose (MFCs) is considered to be a promising material as an alternative to inorganic fillers in porous composite materials [2]. MFC is also known as nanocelluloses which can be extracted from lignocellulosic materials through chemical and/or mechanical disintegration methods [3]. Typically, it is in the range of 20-40 nm since it usually consists of aggregates of cellulose microfibrils [4]. They are usually used as reinforced materials in composites due to its desirable properties such as lightweight, nontoxic and biodegradable. In this study, the microfibrillated cellulose was extracted from sugarcane bagasse (SCB). SCB is a by-product which is generated after juice extraction from sugarcane mills [5] and have dry weight basis of cellulose approximately 43.8% [6]. It has been extensively applied in various industries such as in pulp and paper production [7], textiles [8] and furniture industry [9]. Due to its intrinsic mechanical and chemical properties, it can be used to build hierarchical structures which can enhance the performance of traditional polymeric materials [10][11][12].
In this respect, the development of 'green foams' from biodegradable polymer reinforced with natural fibres from renewable resources is necessity as an alternative to non-degradable polymer foams. Polyvinyl alcohol (PVA) is consider as one of the polymers that is nontoxic, biodegradable, biocompatible and water soluble synthetic polymer. In presence of suitable microorganisms, PVA is easily susceptible to biodegradation [1,13]. PVA and SCB are highly hydrophilic which rich in pendant hydroxyl groups that act as attachment sites for biomolecules [14] and have low thermal stability that can limit the choice of polymer matrices as well as processing technologies for production of foam composites [15]. Thus, incorporating MFC-SCB into PVA foam will be less attractive as it lacks fine interfacial adhesion, low melting point and poor resistance towards moisture which may lead to fibers agglomeration and cause poor mechanical properties [16]. Therefore, the most feasible solution to this problem is increasing the hydrophobicity of PVA-MFC composite foams by chemically modification using coupling agents such as alkoxysilane.
Alkoxysilane is one example of organosilicon compounds which particularly effective in hydrophobizing hydroxylated substrates [17] and improved degree of cross-linking in the interface region besides offering a perfect bonding [18]. After silanols react with hydroxyl groups of MFC and PVA, they can form stable covalent bonds to the cell wall that are chemisorbed onto the foam surface [18]. Hence, the swelling ratio of the PVA-MFC composite foams (i.e., creation of a crosslinked network) due to covalent bonding between the matrices can be restrained by modification with silane.
PVA foam has poor dimensional stability and mechanical properties, while pure fibrillar cellulose foam has loose characteristics [2]. Therefore, in this study, the microfibrillated cellulose of sugarcane bagasse (MFC-SCB) was incorporated with PVA in order to prepare composite foam and chemically modified with Y-methacryloxypropyltrimethoxysilane (MPS) and tetraethoxysilane (TEOS). The best silane concentration and reaction time in preparing the silylated PVA-MFC composite foams were determined. The characteristics such as morphology, chemical structure, thermal, mechanical and wetting ability of PVA-MFC composite foams before and after treatment were investigated. The swelling ratio and absorption capacity of the unmodified and silylated foams were also compared. The silylated 2,20 PVA-MFC-TEOS 10,2 foams showed a good characterization in terms of morphology, absorption capacity, wetting ability, thermal and mechanical properties. Thus, further examinations of this foam were performed in respects of morphology, chemical compositions and specific surface area using FESEM, XPS and BET surface area, respectively.

Materials
The agricultural residues (SCB) were obtained from Sri Gombak market, Selangor. Sodium hydroxide (NaOH) pellets, 35% hydrogen peroxide (H 2 O 2 ), 95% sulphuric acids (H 2 SO 4 ) and polyvinyl alcohol (PVA) with degree of hydrolysis (87-89%) at molecular weight 50 k were purchased from Sigma-Aldrich. Two silane coupling agent Y-methacryloxypropyltrimethoxysilane (MPS) and tetraethoxysilane (TEOS) were purchased from Merck and Fisher Scientific (Malaysia) respectively. Others reagent such as ethanol (HmbG chemical) and glacial acetic acid (CH 3 COOH) were purchased from Merck (Malaysia). All the reagents are analytical grade and were used as received without further purification.

Extraction of Microfibrillated Cellulose (MFC) from Sugarcane Bagasse (SCB)
Extraction of cellulose from SCB was adapted from Anchor et al. [19] with slightly modification as described by Asem et al. [20]. The SCB was sun-dried and ground-using grinder until it achieved 100 μm size. The SCB powder (25 g) were mixed with 500 ml of 2% (w/v) of NaOH and water bath for 5 h at 80°C. The treated SCB were filtered, then bleached with 500 ml of H 2 O 2 in 1:1 aqueous dilution at 75°C. The samples were further treated with 12% (w/v) of NaOH solution and incubated in water bath at 80°C for 15 min. In order to breakdown the chemical bond of hemicellulose and lignin of SCB completely, the SCB was treated two times using NaOH. The treated SCB sample was vacuum filtered and washed for few times to ensure there were no NaOH and H 2 O 2 residues in the cellulose fibers suspension. The treated SCB samples were then further treated with 1% v/v of H 2 SO 4 at 80°C. The extracted samples were cleaned with distilled water until achieved neutral pH value. The samples were sonicated using probe sonicator (B.Braun Biotech International, Germany) for few hours at amplitude of 70. The MFC-SCB samples were centrifuged for 15 min at 6000 rpm and stored in the refrigerator at 4°C until further used.

Preparation of PVA-MFC Foams
The PVA-MFC foams were prepared with 2 w/w% concentration of MFC-SCB suspension, 20 w/v% concentration of PVA (MW: 50 k). The mixtures were slowly stirred at room temperature for 4 h and frozen at -22°C freezer (Polar 530SV, AS biomedical division, Italy) for 24 h. The frozen samples were subsequently dried for 4 days in ALPHA 1-4 LD Plus freeze-dryer (Christ, Germany) to sublimate the water under vacuum. This method is adapted from Dash et al. with slightly modification [21]. The PVA-MFC foam was denoted as 2,20 PVA-MFC foam, where 2 denotes as concentration of MFC-SCB suspension (w/w%) and 20 as concentration of PVA (w/v%).

Modification of PVA-MFC Foams
The freeze dried PVA-MFC foam was firstly cut with diameter of 1.6 cm and 1 cm of thickness. The coupling agents, MPS and TEOS were gently stirred in ethanol/water mixture in ratio of 80:20 at different concentrations (5 wt%, 10 wt% and 15 wt%). In order to ensure efficient coating of MPS/TEOS on PVA-MFC foam, the mixture was gently stirred at room temperature and pH of the solution was adjusted to 5 using acetic acid as described in previous studies [22]. The PVA-MFC foam was immersed into 1 mL of MPS/TEOS and 1 mL of deionized water at different reaction time (2 h, 4 h and 6 h) in the conical flasks. These flasks were tightly sealed and placed in an oven at 70°C. Then, the silylated sample was taken out and placed in oven for another 24 h at 60°C. The silylated PVA-MFC foams were stored in desiccator until further used. The silylated PVA-MFC foams are denoted as PVA-MFC-MPSx,y foam and PVA-MFC-TEOSx,y foam. The x and y are denoted as silane concentration and reaction time, respectively. The scanning electron microscopy (SEM) analysis of 50 mg unmodified and silylated PVA-MFC foam were determined using JEOL (JSM IT 100, Japan) at an accelerating voltage of 8 kV. The foam samples were fixed on sample holders using carbon pad and sputter-coated with a thin layer of platinum.

Structural Characterization by Fourier Transform Infrared (FTIR) of PVA-MFC Foam
Fourier transform infrared (FTIR) spectra of unmodified and silylated PVA-MFC foams were recorded on IRAffinity-1S FTIR (Shimadzu™ Corporation, Japan) in the transmittance mode in range of 4000-400 cm -1 using KBr pellets with resolution of 4 cm -1 .

Wetting Ability of PVA-MFC Foam
The surface wettability of unmodified and silylated PVA-MFC foams were evaluated by contact angle measurement, using Static Contact Angle (Data Physics Instruments, Germany) combined with high speed camera. At the top surface of the sample, 0.1 μL/s of water droplet were deposited directly using a precision stainless steel tip at room temperature. Three measurements of each sample were taken after contact angle measurement had reached a stable value after 30 s.

Thermogravimetric Analysis (TGA) of PVA-MFC Foam
Thermal stability measurements of unmodified and silylated PVA-MFC foam were carried out using simultaneous thermal analyzer (TG2, STA7300 HITACHI, Japan) at a heating rate of 10°C/min, under argon atmosphere (flow rate about 100 mL/min) from 25°C to 600°C.

Compression Test Analysis of PVA-MFC Foam
Compression testing of foam sample was performed on an electronic universal tensile machine (LR10K Plus, LLYOD Instruments, Germany) with thickness 1 cm and diameter of 1.6 cm. The preload for each sample was 0.5 N with a speed of 1 mm/min to 50% of its original thickness for 10 s. The samples were prepared according to the American Society for Testing and Materials, ASTM D3574 Part C. Five different experiments of unmodified and silylated 2,20 PVA-MFC foams were performed.

Swelling Ratio of PVA-MFC Foam
For swelling ratio (Sr) test, three measurements of each PVA-MFC foam (thickness 1 cm, diameter 1.6 cm) were performed and the average values were calculated. Each foam sample was immersed into distilled water until its weight remained constant. Calculation of PVA-MFC foams swelling behavior were evaluated using Eq. (1) [13]: where, W s is the weight of the sample in the swollen state and W d represent the dried weight of the sample.

Solvent and Oil Absorption Capacities of PVA-MFC Foam
The solvent and oil absorption capacities (Cm) of PVA-MFC foam were determined by dipped the unmodified and silylated PVA-MFC foams (thickness 1 cm and diameter of 1.6 cm) into 20 mL of solvents (phosphate buffered saline, ethanol, dichloromethane, dimethyl sulfoxide) and oil (castor oil, synthetic oil, cooking oil) for 5 s. The average of Cm obtained by measuring three different foam samples and was calculated as follows [17]: where m 0 and m 1 are the weights of the sample before and after absorption, respectively.

Field Emission Scanning Electron Microscopy (FESEM)
The unmodified and silylated 2,20 PVA-MFC-TEOS 10,2 foam were further observed by field emission scanning electron microscopy (FESEM, MERLIN, ZEISS, Germany). The foams samples (10 mg) were mounted on the aluminium studs and sputter-coated with a thin layer of iridium.

X-Ray Photoelectron Spectrometer (XPS)
XPS was performed using Axis Ultra DLD (Kratos, SHIMADZU, Japan) with an aluminum anode (AlKa = 1486.6 eV) operating at 150 W with a background pressure 1-10 torr. The low-resolution survey scans were taken with 50 eV analyzer pass energy and high-resolution spectra was taken with 20 eV analyzer pass energy. The atomic ratio of oxygen-to-carbon (O/C) was calculated from their normalized peak areas as [23]: where I o and I c are the normalized integrated area of the peaks for oxygen and carbon, respectively while SC/ SO is the corrected term for the sensitivity factor. The quantitative data on the surface chemical composition are acquired primarily from the O/C atomic proportion and the relative quantity of each carbon type (C 1 s).
For example, the oxygen-to-carbon (O/C) nuclear ratio was calculated using region sensitivity variables as an original indication of 2,20 PVA-MFC foam surface treatment.

Brunauer-Emmet-Teller (BET) Specific Surface Area
The BET specific surface area of the best silylated PVA-MFC foam (silylated 2,20 PVA-MFC-TEOS 10,2 foam) was determined by nitrogen adsorption using surface area and pore size analyzer (Autosorb-1, Quantachrome, USA). 30 mg of foam sample was dried under a nitrogen gas flow at 105°C for 18 h and measured at -196°C, under a range of relative vapor pressures between 0.05 and 0.2. The BET surface area and pore size were evaluated from the obtained adsorption isotherm.
3 Results and Discussion 3.1 FESEM Analysis of Microfibrillated Cellulose of Sugarcane Bagasse (MFC-SCB) Fig. 1 showed the even surface morphology of sugarcane bagasse microfibrillated cellulose (MFC-SCB) at 90 k magnification. It is clearly showed that the individualized forms of MFC covering the whole area. These fibrillated microcelluloses with high entangled-structure that is several microns in length but 1-100 nm in diameter.

Determination of Appropriate Parameters (Silane Concentration and Reaction Time) of Silane Modification on PVA-MFC Composite Foam
The density and porosity obtained from unmodified 2,20 PVA-MFC foams (Mw: 50 k) were 0.49 g/cm 3 and 68%, respectively. However, the density values of silylated 2,20 PVA-MFC-MPS foams and 2,20 PVA-MFC-TEOS foams were higher compared to unmodified 2,20 PVA-MFC foams. This is because, increased in density values lead to the decrement of porosity values of the silylated 2,20 PVA-MFC foams. From the result obtained, the porosity values of silylated 2,20 PVA-MFC-MPS foams were decreased from 68% to 32% when increasing in silane concentration and reaction time. Meanwhile, silylated 2,20 PVA-MFC-TEOS foams also showed a decreasing of porosity values (68% to 26%), when increasing in silane concentration and reaction time. Previous study reported that a gradual decreased in porosity of both silylated foams (MPS and TEOS) indicated that the void volume fraction progressively decreased within the silylated foams due to the thickening of the cellulosic scaffold after silylation [18]. Moreover, this phenomenon may also due to the number of hydroxyl groups of PVA and cellulose decreased with increasing of silane concentration which resulted an increased in density values of silylated 2,20 PVA-MFC foam [24]. Both porosity of silylated foams (MPS and TEOS) also showed a decreased value when the reaction time was increased. It might be due to longer reaction time in complex interactions between cellulose, PVA compounds and solvents may affect the physical properties of 2,20 PVA-MFC foams [25,26]. Therefore, these two parameters (silane concentration and reaction time) need to control carefully in order to optimize the surface modification of 2,20 PVA-MFC foams. The highest porosity values achieved in this work were 65% ( 2,20 PVA-MFC-MPS foam) and 67% ( 2,20 PVA-MFC-TEOS foam). The appropriate conditions to achieve this porosity value were observed at conditions of 10 wt% of silane concentration and 2 h of reaction time. As can be seen, the modification of foams with silane resulted in a continuous and cloud-like coating unlike unmodified 2,20 PVA-MFC foam. The unmodified 2,20 PVA-MFC foam showed a random threedimensional network structure. This formation might be due to the nucleation and growth of large ice crystals within the network when the MFC are pushed out from its original location during the freezedrying process [27]. When coupling agent (MPS or TEOS) were introduced, the 2,20 PVA-MFC foam  Fig. 4 shows the FTIR spectra of 2,20 PVA-MFC foams before and after silanization ( 2,20 PVA-MFC-MPS 10,2 foams and 2,20 PVA-MFC-TEOS 10,2 foams). As observed, bands at 3340 cm -1 , 2940 cm -1 , 1646 cm -1 , 1429 cm -1 , 1143 cm -1 , 1096 cm -1 and 895 cm -1 were presence in all samples. The band at 3340 cm -1 is attributed to O-H vibration which mainly results from hydrogen bonds in cellulose [25]. Meanwhile, the bands at 2940 cm -1 , 1429 cm -1 and 1143 cm -1 are associated with C-H, C-H 2 and C-O stretching vibrations, respectively. The shoulder band presence at 1096 cm -1 is attributed to C-O stretching, which could be detected in PVA nanocomposites [28]. From the spectra, it was observed that the band at 1594 cm -1 and 895 cm -1 associated to C-NH 2 bending vibration [29] and ß-1,4-glucosidic bond vibration [30], respectively. Further inspection of these spectra reveals the presence of two small peaks around 950 cm -1 and 800 cm -1 in both silylated 2,20 PVA-MFC-MPS 10,2 foams and silylated 2,20 PVA-MFC-TEOS 10,2 foams. These bands were attributed to Si-O stretching vibration and deformation of Si-OH stretching vibrations, respectively [31].

Wetting Ability of Unmodified and Silylated PVA-MFC Foams
2,20 PVA-MFC foam is easily to absorb moisture and swell in water due to high density of the hydroxyl groups present at the surface of PVA and MFC-SCB [27]. Therefore, this hydrophilicity features were modified with silane coupling agents (MPS and TEOS). The resulting wetting properties were evaluated by measuring contact angle of a water droplet (WCA) deposited directly at the top surfaces of the unmodified and silylated 2,20 PVA-MFC foams. Fig. 5a shows a distinct hydrophilic character of unmodified 2,20 PVA-MFC because WCA could not be measured as it was immediately absorbed within the foam structure in less than 1 s. This is due to the amphiphilic character of cellulose [32] and hydrophilic properties of PVA [33,34]. In contrast, the water droplets could stand on the surfaces of the silylated 2,20 PVA-MFC-MPS 10,2 (123.3°± 0.4°) foams and 2,20 PVA-MFC-TEOS 10,2 (99.6°± 0.6°) foams. Both silylated 2,20 PVA-MFC foams kept their original shapes well without significant absorption within 30 s (Figs. 5b and 5c). The difference in contact angle might also be attributed at least in part to the different surface chemistry as reported in previous studies [34][35][36]. TEOS has short pendant group compared to MPS [37], thus extent of silane hydrolysis and its degree of polymerization are known to exert a great influence on the silane adsorption at the foam surface [38]. The improved in WCA measurements of both silylated 2,20 PVA-MFC foams (MPS and TEOS) can be attributed to the formation of a thin and interpenetrated silane coating network on the surface of the  [27]. In this respect, the hydrophobic nature of alkoxy groups on the surface of the silane coatings can influence the WCA measurements of silane treated 2,20 PVA-MFC-MPS 10,2 foams and 2,20 PVA-MFC-TEOS 10,2 foams [27,39].

Thermal Properties of Unmodified and Silylated PVA-MFC Foams
Figs. 6a and 6b show the thermo gravimetric (TG) and derived curves (DTG) of unmodified 2,20 PVA-MFC foams, silylated 2,20 PVA-MFC-MPS 10,2 foams and silylated 2,20 PVA-MFC-TEOS 10,2 foams, respectively. The starting curve was not stable due to different moisture uptake that different sample exhibits. Modification with silane was not only reduced ambient moisture uptake but also acted as barrier preventing inherent available bound water within PVA/MFC foam from easily evaporated (or diffused out) at the beginning of TGA test. This can be seen clearly in Fig. 6 when unmodified sample experienced faster weight loss from ambient up to 100°C). Based on the observation, the initial weight loss in the range of 30-150°C for silylated 2,20 PVA-MFC-MPS 10,2 foams and silylated 2,20 PVA-MFC-TEOS 10,2 foams were attributed to the removal of intermolecular H-bonded water [25,40]. However, without modification with silane, the unmodified 2,20 PVA-MFC foams showed an initial and slow weight loss until around 200°C, which might be attributed to the loss of residual moisture in the MFC as well as the slow degradation of MFC [41]. This phenomenon also occurs due to the evolution of water produced by elimination of hydroxyl side groups [13,42] and formation of some volatile products which yields predominantly macromolecules with a polyene structure in PVA [13,43].
The degradation temperature for unmodified 2,20 PVA-MFC foams was 237°C. Upon treated the foam samples with silane, the degradation temperatures for silylated 2,20 PVA-MFC-MPS 10,2 foams and silylated 2,20 PVA-MFC-TEOS 10,2 foams shifted to lower temperature which is 158°C and 191°C, respectively. This might be due to the agglomeration of silane treated inside the silylated 2,20 PVA-MFC foams. This phenomenon occurred due to the agglomeration of silane treated composites that could result in poor interfacial adhesion with the polymer matrix [43]. Thus, due to poor dispersion of silane treated composite in the polymer matrix, strong van der Waals forces between composites can cause the formation of aggregates in the matrix and the thermal stability of the polymer nanocomposites was reduced. Furthermore, other studies reported that addition of silane can decrease the crystalline, and thereby decreased the thermostability [24].
As seen from DTG curves (Fig. 6b), for silylated 2,20 PVA-MFC-MPS 10,2 foams and silylated 2,20 PVA-MFC-TEOS 10,2 foams, both weight loss started at around 30.3°C, followed by a drastic weight loss at 119°C and 134°C. In contrast, the unmodified 2,20 PVA-MFC foams showed a slow weight loss in range of 201°C-243°C. By comparing both types of silylated foams, a better improvement of thermal stability was observed  10,2 foams. This result could be attributed to the better retardation of PVA thermal decomposition by covalent bonds formed between PVA and silane chemical [27] during silane modification using TEOS.

Compressive Properties of Unmodified and Silylated PVA-MFC Foams
In this study, unmodified 2,20 PVA-MFC foams were easily loose and fragmented while silylated Based on compressive stress-strain curves presented in Fig. 7, all samples exhibited typical characteristics of elasto-plastic foams, showing an elastic behavior at low stress, followed by collapse region and then the stress increased steeply. The deformation mechanism of the foam could be separated into three different phases during compression [44,45]. At the early stage, a linear-elastic region appears where the cell walls start to bend. Then, there is a plateau-like state which corresponds to the buckling of cell walls. When the foam was further compressed to higher strain, its volume decreased and the cell walls began to collapse, which showed a significant increase in the stress. The compressive strength of the foams lies in the order of silylated 2,20 PVA-MFC-TEOS 10,2 foam > silylated 2,20 PVA-MFC-MPS 10,2 foam > unmodified 2,20 PVA-MFC foam in the elastic region of the curves. It showed that the compression strength of the silylated foams were higher than unmodified 2,20 PVA-MFC foams.
The compression stress of unmodified 2,20 PVA-MFC foams was 0.31 MPa at 50% strain. Meanwhile, the compressive strength of silylated 2,20 PVA-MFC-MPS 10,2 foams and silylated 2,20 PVA-MFC-TEOS 10,2 foams were 0.43 MPa and 0.64 MPa, respectively (as shown in Tab. 1). It is clearly showed that the silane treatment could improve the mechanical properties of 2,20 PVA-MFC foams which opposite to the thermal properties results. This result could be attributed to the increase in the degree of hydrogen bonding within the PVA-MFC network after silylation, which incite an increase in the arrangement of networks between PVA and MFC [17]. As observed in Fig. 8, all foams reached an equilibrium swelling state within 24 h. The highest swelling ratio observed in unmodified 2,20 PVA-MFC foams was 10.8%. This result can be ascribed to the hydrophilicity feature of PVA and MFC. The swelling behavior can be explained by taking into account that hydrogen-bonding interactions between PVA and MFC free hydroxyl groups occurs in 2,20 PVA-MFC foams [13]. After 2 h water immersion, silane treated sample exhibited~2.5 times less absorption (2-3% swelling ratio) compared to unmodified sample (10% swelling ratio). The low values of swelling ratios in silylated 2,20 PVA-MFC-MPS 10,2 foams (3.95%) and silylated 2,20 PVA-MFC-TEOS 10,2 foams (2.65%) were observed. This might be due to the silane treatment was successfully carried out to enhance the hydrophobicity of the foam. These differences occurred due to the origin of silylated 2,20 PVA-MFC foams (MPS and TEOS) in specific ionic or hydrogen bonding phenomena [46]. The interaction H-bond networks between PVA and MFC-SCB are occurred due to both PVA and MFC-SCB are hydrophilic [47]. In this respect, the silane chemicals (MPS and TEOS) have removed the H-bond networks. Hence, the ionic charges are no longer active when the surface of the 2,20 PVA-MFC foams are silylated and soaked into water. This phenomenon of hydrophobicity of silylated 2,20 PVA-MFC foams was confirmed by water contact angle analysis.

Solvents and Oils Absorption Capacity of Unmodified and Silylated PVA-MFC Foams
Through the silanization using MPS or TEOS, the 2,20 PVA-MFC foams exhibited decreasing of theirs inherent hydrophilicity features. Hydrophobizing of hydroxylated PVA-MFC by silylating agents (MPS and TEOS), were expected to serve as absorbents for effective clean and collection of organic solvents and oils [48]. As seen in Fig. 9, the silylated 2,20 PVA-MFC foams were kept floating on the water surface, whereas the unmodified 2,20 PVA-MFC foams was sank into water rapidly.  The absorption capacity of silylated 2,20 PVA-MFC-MPS 10,2 foams and silylated 2,20 PVA-MFC-TEOS 10,2 foams for different oils and organic solvents were measured and presented in Fig. 11. The absorption capacity of solvents and oils in silylated 2,20 PVA-MFC-TEOS 10,2 foams were range from 0.41 g/g to 48.59 g/g and 16.42 g/g to 23.7 g/g, respectively. It shows that the highest absorption capacity compared to silylated 2,20 PVA-MFC-MPS 10,2 foams. While, the range of solvents and oils in silylated 2,20 PVA-MFC-MPS 10,2 foams were from 0.78 g/g to 37.8 g/g and 14.56 g/g to 22 g/g, respectively. In general, the absorption capacity of the silylated 2,20 PVA-MFC foams increased with the density of the liquids, as reported for other absorbent materials [49][50][51]. The chemical functionalization by MPS and TEOS could cause the modification of surface chemistry of the pores and led to increment of the oleophilicity, which due to an increase in the capillary pressure of the medium and low pore connectivity foams [52]. As expected lower absorption values obtained for deionized water could be linked to the hydrophobicity of silylated 2,20 PVA-MFC foams.

Characterization of Silylated 2,20 PVA-MFC-TEOS 10,2 Foams
Based on the initial characterization of unmodified and silylated 2,20 PVA-MFC-SCB foams (MPS and TEOS), the silylated 2,20 PVA-MFC-TEOS 10,2 foams showed good characterization in terms of Figure 9: The unmodified and silylated 2,20 PVA-MFC foams were immersed in water morphology, absorption capacity, wetting ability, thermal and mechanical properties. Therefore, it was further examined in terms of morphology, chemical compositions and specific surface area using FESEM, XPS and BET surface area, respectively.  The FESEM micrographs (Figs. 12a and 12b) showed the surface of the prepared foams (unmodified 2,20 PVA-MFC foams and silylated 2,20 PVA-MFC-TEOS 10,2 foams) respectively with magnification of 5 k X. A significant difference in morphology between the unmodified and silylated foams was clearly observed. Fig. 12a show open small pores of unmodified 2,20 PVA-MFC foams. The small pores size were probably due to the removal of ice crystals in the samples during freeze-drying, which can only form very small bubbles in the polymer matrix [53]. As can be seen in the image of unmodified 2,20 PVA-MFC foams, the interior of the pores possessed a continuously three-dimensional (3D) porous structure, formed by randomly entangled MFC fibers and PVA. After 2,20 PVA-MFC foam was silylated with TEOS, there were closed small pores and cloud-like structure observed on 2,20 PVA-MFC-TEOS 10,2 foam surface. The change in chemical compositions after silane modification with TEOS was further characterized by X-ray photoelectron spectrometer (XPS). As shown in Fig. 13, the XPS analysis of the unmodified 2,20 PVA-MFC foams and silylated 2,20 PVA-MFC-TEOS 2,20 foams confirmed the presence of C,O and trace amount of Si elements. The Si 2p peak appeared in the silylated foam demonstrating the successful silane treatment of the 2,20 PVA-MFC-TEOS 10,2 foams. The peak spectrum in Fig. 14 showed the XPS spectrum between 108 and 98 eV of binding energy which indicated the existence of Si 2p peaks as reported in previous study [54]. The peaks at 102.1 eV and 101.5 eV are represented the Si-O-Si bond and C-O-Si bond, respectively [48,54,55]. The subsequent pore structure data was presented in Tab. 2. As can be seen in this table, the surface area and pore volume of silylated 2,20 PVA-MFC-TEOS 10,2 foam were 3.14 m 2 /g and 0.01 cm 3 /g, respectively. A low BET surface area of the silylated foam might be due to the characteristic of silane coupling agent (TEOS) which is easy to be self-polymerization especially on PVA-MFC foam, where it is frequently decreased the specific surface area and obstructed the inner pore structure [25,57].

Conclusion
The successfully grafted of MPS and TEOS on the surface of foams was proven as showed in FTIR results. The WCA of both silylated foams possessed a hydrophobic characteristic than the unmodified 2,20 PVA-MFC foam. In contrast to unmodified 2,20 PVA-MFC foam, both silylated 2,20 PVA-MFC foam show low thermal properties but displayed high compressive strength. The silylated 2,20 PVA-MFC foam also exhibited low swelling ratio and excellent absorption capacity on organic solvents and oils compared to unmodified 2,20 PVA-MFC foam. Silylated 2,20 PVA-MFC foam with TEOS showed good characterization in terms of morphology, absorption capacity, wetting ability, thermal and mechanical properties compared to the 2,20 PVA-MFC foam with MPS. Perhaps this absorbent foam could be alternative materials used for oil-water separation.  Conflicts of Interest: The authors declare that they have no conflicts of interest to report regarding the present study.