Effect of Dissolution Time on the Development of All-Cellulose Composites Using the NaOH/Urea Solvent System

: Innovative and sustainable all-cellulose composites (ACCs) can be obtained by partial dissolution of cellulosic ﬁbers and regeneration of the dissolved fraction. Among cellulose solvents, sodium hydroxide/urea solutions are recognized as promising low-environmental impact systems. In this work, ﬁlter paper (FP) was dissolved with a 7 wt% NaOH/12 wt% urea aqueous solution, kept at − 18 ◦ C for different time intervals, regenerated with distilled water and ﬁnally dried under different conditions. The developed ﬁlms were characterized in terms of morphology, porosity, optical properties, crystalline structure, hydration and mechanical properties. The porosity of the composites decreased with dissolution time due to the progressive ﬁlling of voids as the cellulosic ﬁbers’ surface skin layer was dissolved and regenerated. Samples treated for 4 h showed the minimum values of porosity and opacity, high hydration and a substantial change from cellulose I to cellulose II. Hot pressing during drying led to relevant improvements in ACCs stiffness and strength values.


Introduction
Single-polymer composites are a special kind of composites in which the reinforcement and the matrix have the same chemical identity [1]. This concept was first introduced by Capiati and Porter who obtained single HDPE-polymer composites by melt processing [2]. Since the same polymer is used for both the reinforcement and the matrix, excellent compatibility between components is expected and composite recycling is more straightforward [3].
Cellulose is the most abundant biopolymer in nature, and it is widely recognized for its outstanding mechanical properties which have triggered its use as a reinforcement of different polymeric matrices. Although cellulose is insoluble in most common solvents, dissolution in certain low-environmental-impact solvent systems such as ionic liquids and cold alkaline solutions has proved possible [4]. In particular, NaOH/urea aqueous solutions are effective in dissolving cellulose at low temperatures (less than 0 • C), at concentrations in the range of 7 to 10 wt% for NaOH and up to 24 wt% for urea [5,6]. The role of urea in the solvent system is crucial. While under certain conditions NaOH solutions are able to dissolve low-DP cellulose [4,7], the addition of urea enhances the performance by interacting with OHthrough hydrogen bonds, stabilizing Na + cations and promoting the formation of inclusion complexes cellulose-NaOH-urea-H 2 O [8]. Although it is claimed that there is no direct interaction between urea and cellulose, the formation of inclusion

Materials
Composites were prepared using qualitative slow-filtering filter paper (Grade 103, Hangzhou Fuyang Xinxing Co., Zhejiang, China) in the form of cotton cellulose circular samples. Paper discs with 110 mm of diameter, 80 ± 4 g m −2 of grammage, 10-15 µm maximum pore diameter and an ash content lower than 0.15% were employed (data provided by the supplier). Analytical grade NaOH and urea (Anedra, Research AG, Buenos Aires, Argentina) were used in the solvent system. The degree of polymerization (DP) of the filter paper was 505, as determined by the ISO 5351 method [15] using Kes and Christensen parameters [16]. This value falls within the adequate DP range (i.e., 500-900) for cellulose dissolution in the 7 wt% NaOH/12 wt% urea solvent system used [5].

Preparation of All-Cellulose Composites
Filter paper sheets were pretreated first with distilled water (20 min) and then with a 3 wt% NaOH aqueous solution (20 min) at room temperature [13]. Once the time of the pretreatment was over, the liquid was discarded and 13.6 g of the 7 wt% NaOH/12 wt% urea aqueous solution was incorporated. Samples were immediately placed at −18 • C for 1, 2, 3 and 4 h. After these periods, samples were removed from the freezer and kept at room temperature, and the solvent in excess was discarded and replaced by distilled water. This process was repeated several times to wash the samples until a neutral pH was reached. The water holding capacity (WHC) was measured in wet samples according to Equation (1), after the careful removal of excess water using a cloth without applying any pressure [17]; w wet and w o are the weights of the treated sample and the initial dry weight of the filter paper, respectively.
Finally, the samples from the different treatment times were dried in hot air (50-55 • C) until they achieved a constant weight. In addition, aiming to evaluate the benefits derived from drying under hot pressing (HP), some samples treated with 7 wt% NaOH/12 wt% urea aqueous solution for 4 h were dried under pressure (105 • C, 100 kPa, 30 min) in a heated Carver hydraulic press (model 3925, United States). Both drying methods required prior extensive sample washing in order to avoid alkaline degradation during heating (usually evidenced as a color change in samples from white to yellow or brown).

Determination of Porosity and Size Reduction
ACC porosity was determined using Equation (2) [18], and the density value of pure cellulose (1592 kg m −3 ) given by Mwaikambo & Ansell [19]. The density of the sample (ρ sample ) was calculated using Equation (3), where m is the weight of the ACC (±0.01 g), and L is the thickness (n = 12 ± 0.01 mm). A is the area measured using photographs (±0.05 mm) taken as detailed later on (Section 2.4) and processed with ImageJ (v1.53e, National Institutes of Health, Bethesda, MD, USA) [20].
Sample size reduction (%) was determined from the images of the composites according to Equation (4). A o corresponds to the area of the filter paper before treatment and A is the area of the composite after drying.

Scanning Electron Microscopy (SEM) Analysis
The surface of the samples was observed by scanning electron microscopy (Quanta 200, FEI Co., Hillsboro, OR, USA) at 20 kV. Samples were coated with a thin layer of gold before SEM observation. The diameter of the fibers in filter paper was measured from the SEM images using ImageJ software and reported as a histogram.

Brunauer-Emmett-Teller (BET) Surface Area Determination
The specific surface area of previously degassed samples (10 µm Hg, 0.5 h) was determined by nitrogen adsorption-desorption measurements using liquid nitrogen (77.5 K) in a Micromeritics ASAP 2020 equipment. The weight of each sample assayed was approximately 1 g and the complete measurement interval was 120 min.

X-ray Diffraction (XRD) Analysis
Samples were analyzed in an X-ray diffractometer (D/Max-C, Rigaku) operated in reflection mode with a Cu Kα radiation source filtered by nickel, a potential difference of 40 kV and a current intensity of 50 mA. The range of 2θ analyzed was 8.00-50.00 • with a step of 0.02 • .

Optical Characterization of All-Cellulose Composites
The opacity of the composites was analyzed by taking photographs of the samples under controlled conditions. A photographic black chamber was built following the procedure reported by Yam and Papadakis with some modifications [21]. This chamber allows taking photographs under controlled conditions with a defined illumination. LED lights with a color temperature of 6 500 K (12 W, 450 lumens, LT5, Interelec S.A., Argentina) to accomplish the CIELab 1976 standard were placed inside the chamber on the inner surface of the roof, separated by 175 mm. Samples were placed at the same distance from the roof.
Samples were placed over black (L*: 6.65 ± 0.28, a*: −3.81 ± 0.44 and b*: 0.19 ± 0.74) and white (L*: 88.21 ± 0.05, a*: 2.16 ± 0.03 and b*: −4.54 ± 0.04) backgrounds and photographed according to the procedure of Riquelme et al. [22] to determine their opacity. A digital Camera (Sony IMX 230, Kanagawa, Japan) with 21 MP resolution was used. Photographs were taken in the sRGB color space using the following conditions: f/2.0 mm (aperture), ISO 100, cloudy-daylight white balance (according to the lighting used) and total exposure value equal to 11. Photographs were analyzed with ImageJ software, and were converted to CIELab 1976 space using the incorporated function for D65 illumination. In this way, histograms for L* values were obtained. The mode of population for each parameter was taken and the opacity was calculated from Equation (5).
Sample luminosity (L*) was determined with a hand-held colorimeter (CR 400, Konica-Minolta), calibrated using a pattern provided by the manufacturer (L*: 84.0, a*: 0.3160 and b*: 0.3220). Samples were placed over black and white backgrounds (the same used in photographs) and L* was measured at five random points. Opacity was also obtained with the colorimeter using Equation (5). Opacity values from both methods used were compared.

Water Vapor Sorption of All-Cellulose Composites
ACCs previously dried in silica gel for 7 days were placed in sorption containers of 7 liters and kept at 22 • C and 90% r.h. (relative humidity), generated by a saturated solution of BaCl 2 . Samples were weighed (±0.0001 g) at different times to obtain the kinetics of water vapor sorption. Experiments were performed in triplicate. The water content h (in g H 2 O per g of dry mass, d. m.) as a function of time t was fitted using a first-order kinetics biexponential function that considers two rates for the water uptake [23], as shown in Equation (6).
where h 0 is the water content at the beginning (h 0 = 0), ∆h 1 and ∆h 2 are the water contents related to the contribution of processes 1 and 2, respectively, and τ 1 and τ 2 are time constants for the water uptake of each process. The water content at equilibrium h ∞ was calculated as h ∞ = ∆h 1 + ∆h 2 , and the mean time constant τ for the full water uptake process was obtained as τ = (∆h 1 /h ∞ ) τ 1 + (∆h 1 /h ∞ ) τ 2 , where ∆h 1 /h ∞ and ∆h 2 /h ∞ are the fractions of the contribution to the total hydration of processes 1 and 2, respectively. In addition, sorption isotherms were determined by weighing samples periodically in an analytical balance (±10 −4 g) until they reached the equilibrium at selected r.h. Samples of ACCs previously dried in silica gel for 7 days (≈0% r.h.) were placed in containers of 1.5 liters and equilibrated at different r.h. (11%, 33%, 43%, 57%, 75% and 90%) by exposing them to saturated solutions of LiCl, MgCl 2 , K 2 CO 3 , NaBr, NaCl and BaCl 2 at 22 • C, respectively [24,25]. The water content, or hydration h, given in units of g of water per g of dry matter (d. m.), was evaluated as a function of a w (equilibrium r.h./100). Experiments were performed three times. Isotherms were fitted using the Guggenheim-Anderson-DeBoer (GAB) model [26], according to Equation (7): where N is a parameter related to the primary binding sites of water molecules at the monolayer (g of water per g of d. m.), c is a parameter related to the difference of chemical potential in water molecules strongly (monolayer) and weakly (upper layers) bound to the sample and k is associated to the difference of enthalpy of the pure liquid water and water bound to the upper layers of hydration in the sample [27].

Mechanical Properties of All-Cellulose Composites
Uniaxial tensile tests were performed in an Instron dynamometer (model 5985) at 1 mm min −1 at room temperature. Samples were obtained according to ASTM D1708-13 [28] with 22 mm of effective length and 5 mm of width. Nominal stress(σ)-strain(ε) curves were obtained and Young's modulus (E), tensile strength (σ UTS ) and strain at break (ε b ) values were determined from these curves. At least 6 samples were tested for each system, and the average values and their deviations were reported.

Statistical Analysis
Infostat software (v2020I, Grupo Infostat, Argentina, South America), R (v4.0.2, Cran Software) and Jamovi (v2.3, The jamovi project) were used to perform the analysis of variance (ANOVA) and Games-Howell contrast tests. Table 1 shows the porosity values determined for the ACCs produced. A strong decrease in sample porosity was observed with an increase in treatment time, with average porosity values of 3-4% for the highest treatment time assayed, irrespective of the drying conditions used. In the partial dissolution method employed in this work to produce the ACCs, when the dissolved surface skin layer of the cellulosic fibers in FP is regenerated to constitute the matrix of the composite, voids are progressively filled, and porosity is reduced. A similar behavior was observed by Adak and Mukhodpadhyay in ACCs made from lyocell fabrics and 1-butyl-3-methyl imidazolium chloride ([C 4 mim] + Cl − ). Porosity was reduced from almost 40% to less than 5% after 4 h of treatment [29]. Using the same method, other authors observed an increase in the apparent density, from 465 kg m −3 to 1032 kg m −3 after the treatment of wood pulp with 7 wt% NaOH/12 wt% urea for 4 h [10]. Hot pressing of the sample pretreated for 4 h did not induce noticeable changes in the final porosity of the corresponding ACC. Simultaneously, with an increase in treatment time a sustained size reduction of the ACCs films was observed (Table 1), which reached a value close to 80% after 4 h of treatment, irrespective of the drying conditions used. In the same way, the results of BET specific surface area measurement also evidenced a drastic reduction of the accessible area of solid surface per unit mass in samples treated for 4 h, with respect to the filter paper (FP: 1.04 ± 0.03 m 2 g −1 , 4 h: 0.17 ± 0.03 and 4 h HP: 0.18 ± 0.02-83% of reduction). Similar specific surface area values (0.10 m 2 g −1 , 50% of reduction) were reported by Khakalo et al. for all-wood composites, after treatment of delignified wood with the ionic liquid (IL) 1-ethyl-3-methylimidazolium acetate for 30 min [30]. The reduction in the specific surface area of the ACCs is in accordance with the progressive coverage of FP voids already mentioned.

Morphology
In Figure 1 SEM micrographs of filter paper and ACCs are shown. In the case of the original filter paper, randomly arranged microfibers with an average diameter value of 13.3 ± 4.2 µm are shown (the corresponding histogram is also included in Figure 1). As dissolution time increased, the progressive coverage of the voids of FP structure was observed. Indeed, in the micrographs corresponding to the samples of 4 h of treatment, the original pores are no longer distinguishable.
Polysaccharides 2023, 4, FOR PEER REVIEW 6 Overall, the results collected in Table 1 and micrographs shown in Figure 1 show the determining effect of dissolution time on the sample architecture, with a remarkable increase in compactness as longer treatment times were used.

Water Holding Capacity
The water holding capacity (WHC) of FP increased drastically upon treatment with the NaOH/urea system, which may be attributed to the entry of the solvent and the subsequent opening of the paper structure. As shown in Figure 2, all samples studied had a Overall, the results collected in Table 1 and micrographs shown in Figure 1 show the determining effect of dissolution time on the sample architecture, with a remarkable increase in compactness as longer treatment times were used.

Water Holding Capacity
The water holding capacity (WHC) of FP increased drastically upon treatment with the NaOH/urea system, which may be attributed to the entry of the solvent and the subsequent opening of the paper structure. As shown in Figure 2, all samples studied had a WHC higher than that of the original filter paper (up to seven times). The maximum WHC value was reached after 1 h of treatment, whereas longer periods resulted in a progressive reduction in the sample WHC. Longer dissolution times promote larger amounts of dissolved and then regenerated cellulose, which may lead to a decrease in WHC due to the promotion of cellulose-cellulose interactions. Massive regeneration has been claimed to lead to an expulsion of the water trapped between chains and the reduction of total size [31], due to the removal of the Na + and the decrease in the osmotic pressure inside the hydrogel [32].
Polysaccharides 2023, 4, FOR PEER REVIEW 7 WHC higher than that of the original filter paper (up to seven times). The maximum WHC value was reached after 1 h of treatment, whereas longer periods resulted in a progressive reduction in the sample WHC. Longer dissolution times promote larger amounts of dissolved and then regenerated cellulose, which may lead to a decrease in WHC due to the promotion of cellulose-cellulose interactions. Massive regeneration has been claimed to lead to an expulsion of the water trapped between chains and the reduction of total size [31], due to the removal of the Na + and the decrease in the osmotic pressure inside the hydrogel [32].  Figure 3 shows the X-ray diffractograms of the filter paper and ACC samples. The filter paper diffractogram showed peaks characteristic of cellulose I (i.e., 2θ: 14.7°, 16.4°, 22.6°) [33], as previously reported for similar filter paper samples [13,34]. With the progression of treatment with the NaOH/urea solution, the X-ray diffractograms of ACCs significantly changed, and signals characteristic of cellulose II (12.4°, 20.1°, 21.9°) [33,35] appeared with varying intensity. Under the assayed conditions, a notorious change took place between 2 h and 3 h of treatment; in the samples treated for 4 h, only peaks typical of cellulose II were observed, suggesting an allomorphic transition from cellulose I to cellulose II [33,36,37]. Other authors have also reported a transition from cellulose I to cellulose II for other cellulose sources regenerated from a NaOH/urea solution [38]. While cellulose I is packed in parallel form, cellulose II is organized in an antiparallel form, making the change irreversible [39,40]. Cellulose II is characterized by a complex network of hydrogen bonds that results in a crystalline form that is more difficult to dissolve and more stable than cellulose I [41]. Some authors proposed that the mobility in the paracrystalline and amorphous zones could promote the change of the crystalline form in the whole sample [13].  Figure 3 shows the X-ray diffractograms of the filter paper and ACC samples. The filter paper diffractogram showed peaks characteristic of cellulose I (i.e., 2θ: 14.7 • , 16.4 • , 22.6 • ) [33], as previously reported for similar filter paper samples [13,34]. With the progression of treatment with the NaOH/urea solution, the X-ray diffractograms of ACCs significantly changed, and signals characteristic of cellulose II (12.4 • , 20.1 • , 21.9 • ) [33,35] appeared with varying intensity. Under the assayed conditions, a notorious change took place between 2 h and 3 h of treatment; in the samples treated for 4 h, only peaks typical of cellulose II were observed, suggesting an allomorphic transition from cellulose I to cellulose II [33,36,37]. Other authors have also reported a transition from cellulose I to cellulose II for other cellulose sources regenerated from a NaOH/urea solution [38]. While cellulose I is packed in parallel form, cellulose II is organized in an antiparallel form, making the change irreversible [39,40]. Cellulose II is characterized by a complex network of hydrogen bonds that results in a crystalline form that is more difficult to dissolve and more stable than cellulose I [41]. Some authors proposed that the mobility in the paracrystalline and amorphous zones could promote the change of the crystalline form in the whole sample [13].

Opacity
The solubilization of cellulose and its subsequent regeneration may result in visual changes associated with the evolution of the treatment [42]. Table 2 shows the progress in the opacity of the ACCs measured using two different devices. The results showed that while filter paper and samples of 1 h and 2 h of treatment show similar opacity values, close to 0.95, a noticeable reduction in sample opacity took place at 3 h (Photographs method) and 4 h of treatment. This behavior has been previously attributed to the change of cellulose I to cellulose II, as well as to the covering of voids by the regenerated cellulose which acts as the matrix of the composite [43]. Significant reductions in ACC opacity and/or increments in ACC transmittance have been previously reported by other authors using filter paper and other solvent systems [13,14,44].

Opacity
The solubilization of cellulose and its subsequent regeneration may result in visual changes associated with the evolution of the treatment [42]. Table 2 shows the progress in the opacity of the ACCs measured using two different devices. The results showed that while filter paper and samples of 1 h and 2 h of treatment show similar opacity values, close to 0.95, a noticeable reduction in sample opacity took place at 3 h (Photographs method) and 4 h of treatment. This behavior has been previously attributed to the change of cellulose I to cellulose II, as well as to the covering of voids by the regenerated cellulose which acts as the matrix of the composite [43]. Significant reductions in ACC opacity and/or increments in ACC transmittance have been previously reported by other authors using filter paper and other solvent systems [13,14,44].

Water Vapor Sorption
The kinetics of water vapor sorption displayed in Figure 4a illustrate the effect of the NaOH/urea treatment time. Hydration increased with treatment time, with the most relevant changes observed after 3 h of treatment. The highest hydration was found in samples treated for 4 h, with a total hydration value exceeding that of filter paper by 50%.

h HP
0.76 ± 0.06 0.74 ± 0.03 *This lower-cost device allows analysis of the entire surface of the sample which could be particularly relevant in samples of irregular and/or non-flat surfaces [45] such as certain ACCs.

Water Vapor Sorption
The kinetics of water vapor sorption displayed in Figure 4a illustrate the effect of the NaOH/urea treatment time. Hydration increased with treatment time, with the most relevant changes observed after 3 h of treatment. The highest hydration was found in samples treated for 4 h, with a total hydration value exceeding that of filter paper by 50%.
Sorption kinetics were fitted with a bi-exponential model considering two hydration processes, one of them faster (process 1) and the other slower (process 2) ( Table 3). The relevance of the faster process in the total hydration (h∞) increased with the treatment time, and therefore the growth in h∞ was mainly due to the increase in Δh1. For the two processes, time constants 1 and 2 increased, and consequently the mean time of the total hydration was augmented with the treatment time. Regarding the 4 h hot pressed sample, the contributions of processes 1 and 2 were modified with respect to those of the 4 h sample and parameters 1 y 2 were reduced. The described differences in parameters related to kinetic experiments are probably due to the modification of the internal capillary and micropore network, resulting from hot pressing and/or varying the NaOH/urea treatment time. This could be due to the higher amount of amorphous cellulose and coverage of the pores, because of the formation of the matrix and the composite. Amorphous cellulose is completely permeable to water molecules, while crystalline cellulose is not; water molecules can only access the surface areas of the crystalline zones [46].  (6)) and sorption isotherms were fitted with the GAB (Equation (7)) model. Figure 4b reveal an important hydration in the monolayer. The value of the parameter N, related to the hydration in the monolayer, decreased in samples treated for 1 and 2 h; then, in those treated for 3 and 4 h, values increased until they reached the value measured for filter paper. On the other hand, the value of c was higher in samples treated for 1 h with respect to that of filter paper, and then in treatments applied for 2 h or longer time this value decreased. The opening of the filter paper structure induced by the treatment with NaOH/urea for 1 h might have favored the increase in the value of c, but the amount of regenerated cellulose produced in treatments applied for longer times might have contributed to increasing cellulose-cellulose interactions and decreasing the affinity of cellulose functional groups directly bound to water. The k parameters showed a minimal difference between the samples and their values (appreciably lower than 1) indicate the formation of a non-massive multilayer.  (6)) and sorption isotherms were fitted with the GAB (Equation (7)) model. Sorption kinetics were fitted with a bi-exponential model considering two hydration processes, one of them faster (process 1) and the other slower (process 2) ( Table 3). The relevance of the faster process in the total hydration (h ∞ ) increased with the treatment time, and therefore the growth in h ∞ was mainly due to the increase in ∆h 1 . For the two processes, time constants τ 1 and τ 2 increased, and consequently the mean time τ of the total hydration was augmented with the treatment time. Regarding the 4 h hot pressed sample, the contributions of processes 1 and 2 were modified with respect to those of the 4 h sample and parameters τ 1 y τ 2 were reduced. The described differences in parameters related to kinetic experiments are probably due to the modification of the internal capillary and micropore network, resulting from hot pressing and/or varying the NaOH/urea treatment time. This could be due to the higher amount of amorphous cellulose and coverage of the pores, because of the formation of the matrix and the composite. Amorphous cellulose is completely permeable to water molecules, while crystalline cellulose is not; water molecules can only access the surface areas of the crystalline zones [46]. Table 3. Hydration kinetics and GAB parameters fitted for the water sorption kinetics curves and water sorption isotherms of Figure 4a,b. The reported values of the statistical parameter R 2 indicate the goodness of fitting. Water sorption isotherms displayed in Figure 4b reveal an important hydration in the monolayer. The value of the parameter N, related to the hydration in the monolayer, decreased in samples treated for 1 and 2 h; then, in those treated for 3 and 4 h, values increased until they reached the value measured for filter paper. On the other hand, the value of c was higher in samples treated for 1 h with respect to that of filter paper, and then in treatments applied for 2 h or longer time this value decreased. The opening of the filter paper structure induced by the treatment with NaOH/urea for 1 h might have favored the increase in the value of c, but the amount of regenerated cellulose produced in treatments applied for longer times might have contributed to increasing cellulose-cellulose interactions and decreasing the affinity of cellulose functional groups directly bound to water. The k parameters showed a minimal difference between the samples and their values (appreciably lower than 1) indicate the formation of a non-massive multilayer.

Sample Hydration Kinetics Parameters Water Sorption Isotherms Parameters
Regarding the isotherms of the 4 h HP sample, as compared with those of the other samples, some differences in their shape were observed, particularly for a w < 0.6 which corresponds to the hydration water directly bound to the polymeric matrix forming the monolayer. The convexity or concavity of isotherms in that region is linked to the value of parameter c of Equation (7), which is related to the force of the water binding to the monolayer. As parameter c increases, the isotherms at a w < 0.6 become more concave. Particularly, the isotherm of the 4 h HP sample presented the smallest value of c and exhibited more convexity in that region. Hot temperature and pressure drying produces an exhaustive dehydration that results in a greater number of direct interactions between the polar (hydrophilic) groups of the cellulose. Some of these interacting groups then become unavailable for hydration, reducing the number of primary hydration sites.

Mechanical Properties
As shown in the previous sections, among all studied samples, and irrespective of the drying treatment applied, samples treated with NaOH/urea for 4 h exhibited the lowest values of porosity and opacity and the highest conversion from cellulose I to cellulose II. Hence, these ACCs were those chosen to be mechanically characterized. Table 4 summarizes the results obtained for the samples assayed in tension mode (FP, 4 h, 4 h HP). While the 4 h sample exhibited a lower value of stiffness and similar strength value than filter paper, the 4 h hot pressed sample presented significantly higher values of Young's modulus and strength (Table 4). In fact, the stiffness and strength of the 4 h HP ACC were respectively 65% and 163% higher than those of the original filter paper. On the other hand, strain at break values were quite similar for all samples assayed. In spite of the significant void coverage and the highest compactness of the 4 h sample, higher values of tensile properties with respect to those of the filter paper were not found. However, improvements in stiffness and strength upon hot pressing of the 4 h sample were attained, which may be attributed to the exhaustive dehydration resulting from the hot temperature and pressure, and the subsequent greater number of direct interactions between the polar groups of the cellulose.
The results herein obtained were generally in the same range of values reported in the literature for other ACCs prepared from filter paper treated with alkaline solvents [12,13]. Particularly, in comparison with the values obtained by Duchemin et al. using shorter treatment times [13], slightly higher Young's modulus and quite similar tensile strength and strain at break values were observed (HP samples). Wei et al. reached Young's modulus values of 4.8 GPa in ACCs made with cotton linters and processed at −10 • C with NaOH/urea solvent for 5 min [12]. However, it is worth noting that mechanical properties of ACCs developed with alkaline solvents are generally lower than those for composites prepared with DMAc/LiCl [47]. For example, in comparison with the results of Wei et al., Nishino and Arimoto obtained significantly higher values of stiffness and tensile strength using filter paper and DMAc/LiCl to prepare ACCs [14].

Conclusions
ACCs with promising properties were obtained by the dissolution of filter paper with 7 wt% NaOH/12 wt% urea and regeneration in distilled water. Results showed that the treatment time with the NaOH/urea solvent system played a determinant role in the resulting ACC characteristics, as evidenced from the morphology, porosity, crystalline structure, opacity, water vapor sorption and mechanical property analyses performed. In the ACCs obtained from the longest treatment time assayed (4 h), a substantial change from cellulose I to cellulose II was achieved, which was accompanied by the lowest porosity and opacity, and the highest values of total hydration. In addition, hot pressing of these samples was required to significantly improve their mechanical properties. Indeed, hot pressing of 4 h ACCs resulted in materials with stiffness and strength values 65% and 163% higher than those of the original filter paper.
The mechanical and optical properties of ACCs obtained by partial dissolution of filter paper using a relatively low-environmental impact solvent system such as NaOH/urea aqueous solution, suggest them as promising candidates for the development of sustainable biobased and biodegradable composite materials.