Swelling of cellulose stimulates etherification

2.1 Materials

Avicel PH-101 (Sigma) a microcrystalline cellulose with an approximate particle size of 50 µm was used as a model compound for pure cellulose throughout all experiments. The degree of polymerization (DP) for Avicel was calculated to be approximately 360. This DP was derived from the intrinsic viscosity of pure cellulose, following the methodology outlined by Evans and Wallis (1989). The viscosity measurements were done following the ISO standard 5351:2010. Sodium hydroxide (NaOH), acetic acid, and sodium chloroacetic sodium salt were purchased from VWR. All chemicals were of analytical grade and all washing steps were done with deionized water, water retention value (WRV) measurements were done with normal water (<25 mS/cm) according to the ISO standard.

2.2 Preparing swollen cellulose

Swollen cellulose was prepared using the method by (Wang et al. 2014) with some modifications. Briefly, 25 g of Avicel was added under continuous stirring to a 1 L 10 % NaOH solution. The solution was thoroughly stirred overnight in a refrigerator (+6 °C) giving a translucent but light-yellow colloidal suspension. Subsequently, the cellulose was reprecipitated by lowering the pH to 10 with acetic acid and separated from the alkaline solution by vacuum filtration using a glass microfiber filter (VWR) with particle retention of 1.2 µm. The resulting swollen cellulose was repeatedly washed with deionized and filtered until the conductivity of the filtrate was below 25 µS cm⁻¹. The degree of polymerization was calculated to be approximately 250.

2.3 Drying swollen cellulose

After the washing, the swollen cellulose had a dry content of approximately 8 % and was accurately divided into five equal parts. These parts were further dried using methods previously described (von Schreeb et al. 2025), never-dried, air-dried, freeze-dried, glycerol-dried, and acetone-dried. The benefit of these drying methods is that hornification, measured by water retention value (WRV), can be mitigated when freeze drying, drying from glycerol or acetone. The never-dried swollen cellulose was sealed and left in the refrigerator, thus retaining the water. Air drying was done at room temperature. In the case of glycerol drying, excess glycerol was added (1:4 ratio to the dry mass of cellulose) to act as a spacer when drying. Acetone drying was done via two solvent exchanges; first, the swollen cellulose was immersed and mixed in ethanol for 1 h followed by filtration, this step was repeated with acetone.

2.4 Water retention value

Water retention value (WRV) was used to evaluate the degree of hornification for the dried cellulose materials. This method has previously been shown to correlate well with the hornification of pulps (Claramunt et al. 2011; Laivins and Scallan 1993) since it evaluates the swelling behavior both before and after drying. The WRV measurements were conducted following the ISO 23714 (2015) standard, with modifications to adapt the procedure to a smaller scale. For this, 1 g of dried cellulose samples were soaked in 500 mL water for more than 12 h. The samples were thereafter disintegrated to ensure homogeneous suspensions. The suspensions were filtered until the surface water had disappeared using a glass microfiber filter with 1.5 µm particle retention (WVR). Approximately 400 mg of wet sample was transferred to Eppendorf tubes equipped with filter membrane 10 kDa (WVR). The samples were thereafter centrifuged in a MiniSpin plus centrifuge (Eppendorf) with a centrifugal force of 3,000 ± 50 g for 30 min ± 30 s following the ISO standard 23714 (2015). The measurement was carried out in triplicate and the WRV is reported as the ratio between the weight directly after centrifugation (WRV_wet) and after drying at 105 °C for at least 8 h (WRV_dry):

2.5 X-ray diffraction

Before carboxymethylation, the change in crystallinity of all cellulose samples was observed by analyzing the X-ray diffraction (XRD) patterns. The XRD patterns were obtained using a PANalytical X’Pert PRO powder X-ray diffractometer under ambient conditions. The reference Avicel and the freeze-dried sample were analyzed directly, whereas the air-dried and acetone-dried samples were ground using a pestle. The glycerol-dried sample with remaining glycerol was simply pressed into a thin cake before analysis. Scans from 2θ were collected from 5° to 45° with a step of 0.04°, at 45 kV and 40 mA. The crystallinity index (CI) was determined using the Segal peak height method, where it is calculated as the ratio of the height of the 002 peak (I₀₀₂) to the height of the minimum (I_AM) between the 002 and 101 peaks (Segal et al. 1959):

The (I_AM) is typically located around 18° 2θ, however, the swollen reprecipitated material does not have distinct 101 peaks and the minimum value has therefore been estimated around 18° 2θ.

2.6 Carboxymethylation

All dried samples of swollen cellulose were carboxymethylated to asses the reactivity. For this, a reaction media was prepared by dissolving 6.3 g of NaOH in 150 mL EtOH (90 % v/v), once dissolved, 10 g of dry-weight cellulose sample was added under constant stirring. The mixture was added to a pre-heated oil bath and continuously mixed until the reaction mixture reached 60 °C. Subsequently, 8.7 g of chloroacetic sodium salt dissolved in 25 mL EtOH (90 % v/v) was added and the reaction proceeded for 60 min under reflux. The reaction was terminated by the addition of acetic acid. The carboxymethylated samples were filtered under low vacuum and washed once with 180 mL EtOH (90 % v/v) followed by 180 mL MeOH (80 % v/v). Finally, the carboxymethylated cellulose (CMC) was dried under vacuum.

2.7 Fourier transform infrared spectroscopy

The carboxymethylation was confirmed using Fourier transform infrared spectroscopy (FT-IR) and compared with the unmodified dried swollen cellulose samples. The spectra were recorded with a Spectrum 100 FT-IR Spectrometer (Perkin Elmer USA); 16 scans were recorded for each sample. The spectra were analyzed with Perkin Elmer Spectrum software and all samples were baseline and ATR corrected. Spectra were normalized at 1,162 cm⁻¹ corresponding to the C–O–C stretching vibration in the cellulose backbone.

2.8 Titrimetric determination of the degree of substitution

Conductometric titration was used to determine the degree of substitution (DS), following the protocol by (Eyler et al. 1947) with some modifications. Briefly, 0.3 g (dry-weight) CMC was soaked in 15 mL MeOH (70 % v/v) for 10 min. Thereafter, 200 mL of deionized water was added together with 5 mL 0.3 M NaOH. The mixture was titrated with a total of 25 mL of 0.15 M HCl by taking the conductivity reading after each addition of 0.5 mL HCl, allowing time for mixing.

The obtained conductivity values were plotted against the volume of acid and three tangential lines of the curves could be extrapolated to determine the value for V1 and V2 which corresponds to the intersection point from the titration. These intersection points were used to calculate the total carboxyl content per gram sample (A) and the free carboxyl content per gram sample (B):

where V₁ and V₂ are the volume of HCl in mL obtained from the intersection points in the titration curve, V_NaOH is the added 5 mL 0.3 M NaOH, M_HCl is the molarity of the HCl, and m is the mass of the sample in grams.

The degree of substitution (DS) was thereafter calculated as follows;

where 162 is the molecular weight of the anhydroglucose unit of cellulose. 80 is the net increase in molecular weight when a sodium carboxymethyl group is substituted, associated with the total carboxyl content (A). The free carboxyl content is associated with the free carboxyl content and 22 is obtained by subtracting the net increase in molecular weight for each carboxymethyl group substituted.

3.1 Degree of hornification and crystallinity of swollen cellulose

Initially, the degree of hornification was evaluated for the different drying methods. As seen in Figure 2, the never-dried cellulose has the highest WRV, and no hornification has occurred since it is induced by drying and removing water (Kato and Cameron 1999; Sjöstrand et al. 2023). In contrast, the reference Avicel and the air-dried swollen cellulose have the lowest WRV. This result is expected since Avicel is a highly crystalline compound and air-drying induces hornification, which lowers the water-holding capacity.

Figure 2:

The water retention value was used to assess the degree of substitution. The reference Avicel has a low water retention value, due to its crystalline structure. Upon reprecipitation, the cellulose becomes highly swollen and the water retention value drastically increases, as shown for the never-dried sample. The glycerol-dried showed the highest water retention value, indicating no or low hornification. In contrast, air-, acetone-, and freeze-drying induce some hornification, which is reflected in their lower water retention values.

From the XRD measurements, the reference Avicel shows a pattern of cellulose I from the characteristic diffraction peaks of the planes 1–10, 110, and 200, located around 14.5, 16.5, and 22.8° 2θ, respectively (Figure 3). The reference Avicel also has the highest crystallinity index (CI), approximated with the Segal peak height method (Segal et al. 1959) displayed in Table 1. However, the high CI, almost 82 %, is in line with previously reported results for Avicel PH-101 (Park et al. 2010).

Figure 3:

Diffractograms of swollen cellulose dried using different drying methods. The reference Avicel shows characteristics for cellulose I, whereas the swollen cellulose that was dried using different drying methods shows a cellulose II pattern. However, the glycerol-dried sample lacks any peak around 12° 2θ.

In contrast, all swelled and dried samples show XRD patterns of cellulose II, with diffraction peaks around 12.3°, 20.5°, and 22.0° 2θ. The swelled freeze-, air-, and acetone-dried samples lacked distinct I_AM minimum, but the value was approximated around 18° 2θ. Furthermore, the glycerol-dried sample exhibited no peak around 12° 2θ and therefore did not show an I_AM minimum, this value was also approximated around 18° 2θ. This absence may be attributed to the highly swollen and more amorphous structure being better preserved by glycerol, as indicated by the water retention value. However, due to the limitations of the XRD technique used, this cannot be determined with certainty. The calculated CI% is still high for the swollen cellulose (Table 1), even though the diffractograms show a cellulose II pattern. Additionally, there is no significant difference in CI% between glycerol-, freeze-, or air-drying. However, the Segal height method is known for giving high values of crystallinity since it only measures two heights in the diffractogram. In addition, the method is only suggested for comparing relative differences between samples (Park et al. 2010).

The acetone-dried material has the highest CI of the different drying methods and also a low WRV, thus indicating that it might recrystallize during the drying. This finding is rather unexpected since the solvent exchange (acetone) before drying has been reported to mitigate hornification measured by WRV (Hashemzehi et al. 2024). However, although these results indicate that the acetone-dried sample is as hornified as the air-dried material and has the same crystallinity, the appearance of the two materials is completely different. The air-dried material dries into hard stone-like structures, whereas the acetone-dried sample is more powder-like.

Finally, adding glycerol as a spacer inhibits the hornification significantly and the water-holding capacity is significantly higher than for all other drying methods (Figure 2). This result is in line with previous findings (Moser et al. 2018). However, glycerol is still present in the samples which limits the end use for the cellulose. Freeze drying seems to preserve the swollen and less crystalline structure, and thus increase the WRV and still inhibit hornification.

Overall, these results indicate that swelling in cold alkali followed by precipitation creates cellulose II and a more swollen structure. If the material is not dried the water holding capacity is more than five times higher than the starting material, and freeze drying seems to be the most convenient way to preserve this structure for further processing.

3.2 Assessing the reactivity of swollen cellulose

The first set of analyses examined the impact of the drying methods used to prepare the swollen cellulose samples. However, the main aim was to assess the reactivity of these pretreated materials. Since the structure of the reprecipitated cellulose is less crystalline and more swollen we suppose that the OH-groups in cellulose are more accessible, and hence the materials should be more reactive. Hornification and the presence of cellulose II can, however, negatively impact the reactivity of cellulose (Luo and Zhu 2011; Sjöstrand et al. 2023).

Initially, FT-IR spectroscopy was used to confirm the carboxymethylation of the reference sample and the cellulose dried using different methods. The spectra (Figure 4) showed characteristic peaks corresponding to the carboxymethyl group (C=O) around 1,600 cm⁻¹, indicating that derivatization was achieved and the chemical structure of the Avicel was modified accordingly (Ellerbrock and Gerke 2021; Moura et al. 2024).

Figure 4:

FT-IR spectra of carboxymethylated cellulose (CMC) from reference Avicel and swollen Avicel. The swollen Avicel shows the characteristic C=O carboxyl group band at ∼1,600 cm⁻¹, which is absent in untreated reference Avicel. Spectra were analyzed using Perkin Elmer Spectrum software, baseline and ATR corrected, and normalized at 1,162 cm⁻¹.

There is to date no standard method to determine cellulose reactivity, however, Fock’s method has been widely used to assess the reactivity of viscose (Tian et al. 2013). This study used carboxymethylation, and the degree of substitutions (DS) gives a value of the cellulose reactivity. Commercial carboxymethylated cellulose (CMC) DS varies normally between 0.4 and 1.5 depending on the intended application (Heinze and Koschella 2005). A high DS yields a more water-soluble derivative compared to a low DS. The theoretical maximum DS for cellulose is 3.

Mercerization is a common pretreatment to etherification and previous findings have shown that higher NaOH concentration (18–27.5 %) increases the reactivity and thereby the DS. While cellulose II content and temperature have negative impacts on the DS (Almlöf et al. 2012; Ambjörnsson et al. 2013). Since both mercerization and regeneration (dissolution followed by reprecipitation) yield cellulose II (Sixta 2006), the reprecipitation into swelled cellulose is expected to increase the accessibility and reactivity, resulting in high DS.

The DS for all dried cellulose samples and the untreated Avicel reference is listed in Table 2, the average is calculated from two titrations. Freeze-drying gives the highest DS, probably due to the drying method preserving the swollen and more accessible structure, which was also confirmed by XRD and WRV. It was somewhat unexpected that the acetone-dried cellulose had a lower DS than the untreated reference. This result suggests that even though the cellulose structure is more swollen and accessible, its reactivity is still poor due to hornification caused by the drying process. Another factor that could explain the poor reactivity of the acetone sample is the change in crystal structure from cellulose I to II as previously reported by (Nagarajan et al. 2017).

The air-dried material was the most hornified according to both the appearance and WRV and consequently, this material had the lowest DS. For the glycerol-dried material, the spacer remains in the material, and since glycerol has an abundance of alcohol groups the chloroacetic sodium salt might be consumed by the hydroxyl groups in glycerol.

In summary, these reactivity results show that it is possible to produce carboxymethylated cellulose from dried swollen cellulose. When analyzing the relationship between the degree of substitution (DS) and the water retention value (WRV) (Figure 5), it is observed that higher WRV values generally correspond to higher DS. However, this is not the case for the glycerol-dried sample. This deviation is likely attributed to residual glycerol retained within the cellulose matrix, wherein the hydroxyl groups of glycerol potentially compete with the substrate, thereby influencing the substitution process. Swollen cellulose is expected to have higher chemical reactivity, since more reaction sites, i.e., alcohols, are easily available than in a more crystalline form of cellulose. However, when swollen forms of cellulose are air-dried, they will in many cases undergo very strong hornifications (von Schreeb et al. 2025), which lower the reactivity. Thus, the degree of swelling (i.e., available surface, low crystallinity) might be of higher importance for the chemical reactivity of cellulose, than if the cellulose crystals are in the form of cellulose I or cellulose II. This is in line with the views of Wada et al. (2010) who observed that “hydrated cellulose II,” likely representing swollen cellulose, demonstrated high reactivity, whereas “dried cellulose II,” likely corresponding to strongly hornified cellulose, exhibited reduced reactivity.

Figure 5:

Relationship between degree of substitution and water retention value (WRV). Higher WRV gives a higher degree of substitution, except for the glycerol-dried sample where it is likely that the hydroxyl groups in the swelled cellulose matrix compete for the substrate.

3.3 Technical significance

Achieving an extensive swelling of cellulose, as described in this work could be a potential approach to improve cellulose ether production. While drying remains challenging due to strong hornification, our results indicate that these issues can be managed. However, methods like freeze-drying, while effective, are not practical on an industrial scale. Future work should explore scaling up these findings to pilot or mill-scale processes to assess their feasibility in industrial applications.