Sequential Oxidation on Wood and Its Application in Pb2+ Removal from Contaminated Water

Raw wood was subjected to sequential oxidation to produce 2,3,6-tricarboxycellulose (TCC) nanofibers with a high surficial charge of 1.14 mmol/g in the form of carboxylate groups. Three oxidation steps, including nitro-oxidation, periodate, and sodium chlorite oxidation, were successfully applied to generate TCC nanofibers from raw wood. The morphology of extracted TCC nanofibers measured using TEM and AFM indicated the average length, width, and thickness were in the range of 750 ± 110, 4.5 ± 1.8, and 1.23 nm, respectively. Due to high negative surficial charges on TCC, it was studied for its absorption capabilities against Pb2+ ions. The remediation results indicated that a low concentration of TCC nanofibers (0.02 wt%) was able to remove a wide range of Pb2+ ion impurities from 5–250 ppm with an efficiency between 709–99%, whereby the maximum adsorption capacity (Qm) was 1569 mg/g with R2 0.69531 calculated from Langmuir fitting. It was observed that the high adsorption capacity of TCC nanofibers was due to the collective effect of adsorption and precipitation confirmed by the FTIR and SEM/EDS analysis. The high carboxylate content and fiber morphology of TCC has enabled it as an excellent substrate to remove Pb2+ ions impurities.


Introduction
Cellulose is a naturally occurring, renewable, inexpensive, and environmentally friendly material that exhibits properties that make it a viable biomaterial [1]. It is the most abundant biopolymer on Earth, and approximately 1.5 trillion tons of cellulose biomass are produced on an annual basis [2]. It is comprised of β-(1-4)-linked D-glucose units and are linked at the C1 and C4 positions of the anhydroglucose units (AGUs). The most active groups are the hydroxyl groups (-OH) that appear at the C2, C3, and C6 positions. Oxidation of these hydroxyl groups into carboxyl groups can modify the properties of cellulose and generate an important derivative of cellulose called carboxycellulose [3].
In this study, we have introduced raw wood to produce 6CC using the nitro-oxidation method. In previous studies, the nitro-oxidation method was performed on softwood and grasses [3,8,29] to produce 6CC nanofibers. Nitro-oxidation is a simple, less-chemically oriented, and cost-effective pathway as compared to TEMPO-oxidation to generate 6CC nanofibers. Hence, we have utilized this method to produce highly charged TCC from raw wood. The whole process of producing TCC from raw wood involve sequential oxidation using (i) nitro-oxidation; (ii) periodate oxidation, and (iii) sodium chlorite oxidation.
TCC nanofibers have been used widely in biomedical applications such as in surgical gauzes [30], wound dressing materials [31], hemostatic material [32], drug delivery [33,34], lanthanum recovery [35], and antibacterial and antituberculosis material [36]. Sphericalshape TCC nanoparticles are also discussed in the literature and used to stabilize carbon nanotubes suspension with minimal ultrasonication [37][38][39]. There have been few reports dealing with the successful selective oxidation of TCC nanofibers that have been carried out in a simple and cost-effective manner. It has been previously studied [28] that the adsorption of heavy metal ions by the carboxycellulose nanofibers increases with higher carboxylate concentrations. However, previous studies have only shown high concentrations of carboxylate groups in 6CC nanofibers. This study looks to produce TCC nanofibers, which we have hypothesized will lead to a higher concentration of carboxylate groups than the 6CC nanofibers because, in TCC, all three active hydroxyl groups will randomly transform to carboxylate groups (COO -) or ionic sites. Additionally, in TCC, the original AGU structure will be disturbed due to breakage of C2 and C3 bond that will cause a reduction in crystallinity of nanofibers. Hence, we assume that the increase in carboxylate content and reduction in crystallinity in TCC will increase the adsorption properties of TCC, making it a more effective adsorbent.

Materials and Methods
Raw wood was purchased from Brinkmann (Hickory Wood Chips, New York, NY, USA), which was purchased at local hardware stores. The wood chips were first cut into small pieces (less than 2 cm) and then put into an IKA MF10 basic grinder to be turned into a fine powder (200 µ). The samples were not washed nor subjected to any pretreatments or bleaching processes before the reactions.

Preparation of 6CC Using Nitro-Oxidation
The 6CC fibers were prepared following the procedure described in Sharma et al., 2017 [9]. In accordance with this procedure, 10 g of untreated fibers was placed in a threenecked, round-bottom flask (RBF), and then 140 mL of 60% nitric acid was added into it and allowed to spin via stir bar and magnetic stirrer. After 20 min, 9.6 g of sodium nitrite was added to the mixture and the reaction could take place for 12 h at 50 • C. After 12 h, the reaction was stopped by adding 250 mL of distilled water to the solution. The reaction mixture was then poured into the beaker. The fibers settled at the bottom of the beaker after a few minutes, and the upper portion (liquid) was decanted into a separate beaker, which was then neutralized with sodium bicarbonate. The process of decantation was repeated 3-4 times. Then, the fiber slurry was put into a dialysis bag and kept in distilled water for further purification. The conductivity of the water was continually checked for two weeks until the conductivity was under 5 µS. Once the conductivity reached below 5 µS, preparation of DAC could begin.

Preparation of 2,3-dialdehyde Using Sodium Metaperiodate Oxidation
Following the dialysis of the 6CC fibers, the procedure from Varma, 2002 [22] was followed to produce DAC; however, certain parameters were modified to suit the current study. More specifically, a 100 mL suspension of water containing approximately 1.0 g 6CC fibers was added to a 250 mL glass bottle. The bottle was first wrapped in aluminum foil to prevent the exposure of light, and the lights were turned off to decrease any potential light exposure. Then, 0.5 g of sodium metaperiodate was added to the bottle, and the reaction was continued for 6 h in an oil bath kept at 52 • C with the lights turned off so as not to expose the samples to light, and was continuously stirred via a stir bar. After 6 h, the reaction was quenched by adding 50 mL of ethanol and allowed to stir for another 15 min. The samples were then put into dialysis bags for purification, and the conductivity was monitored. The sample was put through the centrifuge multiple times at 5000 rpm for 15 min until the upper suspension of the solution had a pH of 7.

Preparation of TCC Using Sodium Chlorite Oxidation
Following the production and centrifugation of the DAC fibers, the sample (6-carboxy-2,3-dialdehyde cellulose) was subjected to sodium chlorite oxidation to produce TCC fibers. More specifically, approximately 0.5 g of the sample (in suspension) was placed into a 600 mL beaker. Then, a measured amount of sodium chlorite (0.838 g) and acetic acid (0.278 mL) was dispersed in 50 mL of water and then added slowly, dropwise, into the 6-carboxy-2,3-dialdehyde cellulose suspension. The reaction was carried out for 6 h between 20-25 • C and constant stirring via stir bar. After 6 h, the reaction was quenched with 100 mL of ethanol. The solution was then washed with ethanol and centrifuged four times each at 6000 rpm for 10 min. The scheme for the preparation of TCC is presented in Supplementary Information in Figure S4.
The structural and functional characterization of 6CC and TCC was performed using Fourier transform infrared spectrometer (FTIR) (PerkinElmer Spectrum One instrument, Melville, NY, USA) in attenuated total reflectance (ATR) mode, 13

Preparation of Pb 2+ Solution for Removal Study
A stock solution of 5000 ppm Pb(OAc) 2 (lead (II) acetate) contaminated water was made by dissolving 0.4795 g of Pb(OAc) 2 in 50 mL of H 2 O. A serial dilution was then carried out to obtain solutions with the following concentrations (in ppm): 250, 100, 50, 25, 10, and 5. For the remediation experiments, 5 mL of each concentrated lead (II) acetate solution was mixed with 5 mL of the 6CC nanofibers. The same parameters were followed for remediation experiments with the TCC nanofiber solutions. Each nanofiber solution had a wt% of 0.02.

Static Adsorption Study
The static adsorption study on Pb 2+ and TCC was performed by employing the data obtained from ICP-MS analysis. The data obtained were used to calculate Qe (adsorption capacity adsorbed at equilibrium) and plotted against Ce (equilibrium concentration of adsorbate). The ICP-MS data was evaluated using Langmuir Isotherm model. Langmuir model is based on monolayer adsorption on the active site of adsorbent, Langmuir isotherm model is expressed in Equation (1): where: Ce-equilibrium concentration of adsorbate; Qe-adsorption capacity adsorbed at equilibrium; Qm and K-Langmuir constants which can be calculated from the intercept and slope of the linear plot based on Ce/Qe versus Ce [40]. Figure 1 shows the structure of 6CC and TCC. The structure of 6CC possesses repeated anhydroglucoronic units with carboxylic (COOH) or carboxylate (COONa) groups distributed randomly at C6 carbon. While TCC structure shows repeated anhydroglucoronic units with carboxylic (COOH) or carboxylate (COONa) groups presented arbitrarily at C6, C2, C3 carbons. The FTIR spectra of raw wood in Figure 2i shows several characteristic peaks whereby 3328 cm −1 corresponded to OH stretching and peaks at 2900 cm −1 belonged to C-H symetrical stretching in the cellulose unit; 1515 cm −1 was responsible for C=C aromatic symmetrical stretching in the lignin unit; 1739 cm −1 , 1460 cm −1 , 1240 cm −1, and 810 cm −1 resembled to xylan and glucomannan of hemicellulose moieties. It was observed that the OH stretching peak at 3328 cm −1 in 6CC and TCC became much sharper than that in raw wood. This could be due to the disturbance in the hydrogen bonding from the modification of C6 primary hydroxyl groups to carboxyl groups in anhydroglucose units of the cellulose chain. Notably, the peak at 1597 cm −1 appeared in 6CC, and TCC corresponded to the carboxylate (COONa) groups. The peak was sharper for TCC than 6CC represents the higher carboxyl groups as compared to 6CC. The quantitative determination of carboxyl groups is determined by the conductometric titration, as explained in the next section. Interestingly, the peaks at 1739 cm −1 , 1460 cm −1 , 1240 cm −1, and 810 cm −1 due to hemicellulose and lignin have disappeared. The peak at 1515 cm −1 responsible for C=C aromatic symmetrical stretching in the lignin unit decreased in 6CC and completely disappeared in TCC. The FTIR spectra for each sample are a replica of 20 runs. Conductometric titration curves for 6CC and TCC are presented in Figure S1 in Supplementary Information and were used to calculate the carboxylate content. The carboxylate content observed for 6CC was 0.64 mmol/g and TCC was 1.14 mmol/g. The results clearly indicate that TCC posses higher surfacial charges as compared to 6CC, which will lead to comparatively higher removal of Pb 2+ ions from water.

Structural Characterization of 6CC and TCC
The solid-state 13 C CPMAS NMR spectra of raw wood fibers, 6CC and TCC are presented in Figure 2ii. The area between 60-70 ppm represented C6 carbon of the primary alcohol group in the cellulose chain. Another group of peaks between 70-80 ppm was designated C2, C3, and C5 carbons, while the peaks appeared between 80-95 ppm associated with C4 carbon, and the peaks between 100-110 ppm corresponded to the anomeric carbon C1. The NMR spectra of raw wood fibers displayed discrete peaks at 21 and 56 ppm represent glucoroxylans of hemicellulose. Additionally, the peaks at 153 and 171-172 ppm belonged to C1 carbon of guaiacyl units and C4, C3 carbons of syringyl units of lignin. The loss of distinct lignin and hemicellulose peaks in NMR of 6CC and TCC confirms the removal of hemicellulose and lignin during the oxidation processes. It was already evidenced that nitro-oxidation is very effective in removing lignin, hemicellulose, and other impurities from raw biomass [3].
WAXD patterns obtained from raw wood, 6CC and TCC are illustrated in Figure S2 in Supplementary Information. These patterns depict the presence of cellulose I structures in raw wood, 6CC and TCC, whereby the peaks appeared at 2θ of 16.4 • , 22.5 • , corresponding to the (110), and (200) reflection planes, respectively. The diffraction peak at 2θ of 34.9 • due to (004) plane was present in wood and 6CC but disappeared in the TCC sample. Notably, the diffraction peaks related to the (110) and (102) planes of cellulose I structure did not appear in the WAXD patterns of raw wood, 6CC and TCC [37]. The disappearance of the (110) peak can be attributed to the presence of residual hemicellulose and lignin around the crystal structure of cellulose which disrupts the hydrogen bonding of the plane. The WAXD pattern in TCC showed the broad peaks rather than sharper peaks in WAXD pattern for wood and 6CC. This indicates that the crystallinity in TCC was highly decreased due to the introduction of carboxylate groups at C2, C3, C6 positions.

Morphological Characterization of 6CC and TCC
TEM images of 6CC and TCC obtained from raw wood fiber are shown in Figure 3. These images depict that the 6CC and TCC appeared in namometer size fibrous structure. In total 20 fibers were taken from the images to calculate the average (avg.) length (L) and width (W) of fibers using ImageJ software. It was found that 6CC possessed avg. L of 1044 ± 208 nm and avg. W = 6 ± 1.5 nm. While TCC fibers were in the range of avg. L of 750 ± 110 nm and avg. W of 4.5 + 1.8 nm. AFM images in Figure 4 were used to calculate the thickness of the fibers. In total 5-6 spots on different fibers of AFM image were considered to calculate the average thickness of the fibers. Based on that, the avg. thickness measured for 6CC and TCC was 1.69 and 1.23 nm, respectively.

Metal Ion Remediation of Pb 2+ Contaminated Water by 6CC and TCC
Measured amounts (5 mL) of 6CC and TCC samples were introduced into water solutions that had been contaminated by Pb 2+ metal ions. The 6CC and TCC nanofibers were introduced into solutions containing different concentrations of Pb 2+ .
In Figure 5i, the photographs on the remediation of Pb 2+ ions using 6CC are presented. It was noted that in most sample vials, a white powder could be seen settled at the bottom of the vials. We believe this white powder to be the Pb 2+ ions that have been precipitated out on the introduction of the 6CC. In the samples containing 6CC nanofibers, a gelation effect is hardly seen in the samples containing Pb 2+ ions. This is likely due to when the carboxycellulose has either less or equal carboxylate concentration to the concentration of the metal ions; the carboxycellulose nanofibers cannot fully adsorb the metal ions and allow them to precipitate out. This phenomenon does not occur in the TCC samples, and mostly gelation was seen in the vial. This further indicates that the TCC consist of higher carboxylate content that has allowed the Pb 2+ ions to get adsorbed on its surface more efficiently. To confirm the removal efficiency of 6CC and TCC against Pb 2+ ions, the FTIR of floc was performed as presented in Figure S3 in Supplementary Information. Here, a floc sample is referred to a sample containing adsorbed Pb 2+ impurities on carboxycellulose nanofibers (6CC or TCC). The graph indicates a broad distinct peak of 1594 cm −1 in both 6CC and TCC floc representing the cross-linking of carboxylate groups of nanofibers and Pb 2+ ions, although this phenomenon was not clearly visible in the photographs presenting the remediation of Pb 2+ ions using 6CC (Figure 5i).
The phenomenon of Pb 2+ ion removal by nanofibers (6CC and TCC) in suspension is explained by SEM/EDS methods. SEM images of floc formed on adding Pb 2+ (250 ppm) solution in nanofibers suspension is depicted in Figure 6i,iii. SEM images were taken after drying/coating the samples on Si wafer. As a result, the original morphology of nanofibers disappeared, and the sample appeared as a sheet. The floc of Pb 2+ (250 ppm)@6CC appeared as a smoother surface as compared to Pb 2+ (250 ppm)@TCC because of the longer fiber length in 6CC than TCC. In Figure 6iiii, a thick cloud of Pb 2+ ions on the surface of the nanofibers can be seen, which confirms the precipitation of Pb impurities during the remediation process. The quantitative evidence of the presence of Pb 2+ onto nanofibers surface was supported by EDS spectra, shown in Figure 6ii,iv. In corresponding EDS spectra, the Carbon (C), Oxygen (O), Sodium (Na) peaks represented carboxylate nanofibers, while Pb peak designated Pb 2+ ions. The Pb peak was more prominent in TCC than 6CC, indicating the adsorption of more Pb 2+ ions onto the TCC surface due to high carboxylate content. The data obtained from ICP-MS was used to calculate the parameters to fit into Langmuir isotherm model. Table 1 demonstrates the calculations for measuring Qe (experimental adsorption capacity), Ce (concentration of Pb 2+ used for the experiments). The ideal adsorption capacity for TCC was calculated based on the concentration of Pb 2+ used in the solution to the concentration of TCC used for the experiments. Correspondingly, the adsorption efficiency of 6CC and TCC was calculated at different concentrations of Pb 2+ ions. The removal efficiency of the 6CC and TCC nanofibers was between 77-99% when exposed to Pb 2+ concentrations between 5 ppm and 250 ppm (Figure 7ii). The TCC showed a higher removal efficiency of 99% when the concentration of Pb 2+ ions was 5 ppm, and this decreased to 78% when the Pb 2+ concentration was 250 ppm. As compared to TCC, 6CC showed lower removal efficiency against Pb 2+ ions. This can be explained by the difference in the carboxylate content of 6CC and TCC. TCC exhibited carboxylate content of 1.14 mmol/g, and 6CC showed 0.64 mmol/g.  Based on the Langmuir isotherm model, the coefficient of the LSRL of Ce against Ce/Qe is 6.37316E-4 as the reciprocal of the adsorption capacity. Thus, the adsorption capacity (Qm) obtained for TCC was 1569 mg/g with R 2 0.69531, showing conformity to the Langmuir isotherm model (Figure 7i).
A comparative table showing the maximum adsorption capacity (Qm) of the reported substrate in the literature and TCC is given in Table 2. It was interesting to note that the TCC demonstrated in this study is lower the maximum adsorption capacity of the most effective adsorbent reported to date, i.e., nitro-oxidized cellulose nanofibers (NOCNF) [9]. However, TCC has shown higher adsorption capacity than the reported organic substrate such as modified lignin [41], sago waste [42], thiol-, ethylenediamine-and diethylenediamine-modified cellulose nanofibers [43,44]. Lately, a nanocomposite system containing carboxymethylcellulose (CMC) and polyacrylic acid (PAA) was demonstrated, which exhibited very efficient Pb 2+ ion removal of 900-952 mg/g, which is significantly lower than the TCC reported in this study. There have been several reports discussing the adsorption efficiency of biobased materials [45][46][47].

Conclusions
In this study, TCC with a high negative surface charge (1.14 mmol/g carboxylate content) was prepared using the three-step oxidation process, including nitro-oxidation, periodate oxidation and sodium chlorite oxidation. This is the first study to introduce raw wood in nitro-oxidation to produce 6CC. To obtain the high surface-charged substrate for Pb 2+ ions removal, 6CC was further introduced to sequential oxidation. By doing so, the obtained TCC was of high carboxylate content (1.14 mmol/g) and low crystallinity, which has enabled it as an excellent substrate to remove Pb 2+ ion impurities with a maximum removal efficiency of 1569 mg/g. Based on the evidence gathered from this study, it is proposed that the mechanism in which Pb 2+ ions are removed from solution by TCC nanofibers is via adsorption and precipitation.