1. Introduction
Chinese medicine (CM), which is primarily composed of plant-based herbs (> 87%) [
1], [
2], [
3], has been extensively used in China and southeast Asia for its potential in treating various diseases, such as cancer, liver diseases, atherosclerosis, and corona virus disease 2019 (COVID-19) [
1], [
4], [
5]. As of 2020, China’s cultivation area for natural medicinal biological resources exceeded 5806 km
2 [
6]. However, the excessive production and processing of CM have led to the disposal of large amounts of CM residues (CMRs), particularly in the post-pandemic era, as the discharge of CMRs has continued to increase [
7], [
8]. Statistics indicate that the annual discharge of CMRs in China is estimated to reach 60-70 million tonnes [
9]. Conventional methods for handling CMRs, such as landfilling, incineration, and stacking, can result in resource waste and secondary environmental pollution [
10], [
11]. Therefore, there is an immediate need to devise solid waste management strategies for the disposal of CMRs in order to reduce their environmental footprint.
Given that CMRs are classified as ordinary solid waste and are devoid of harmful substances, there is a growing emphasis on the re-utilization of CMRs, including for biofuel production, biogas fermentation and ethanol production [
12], [
13]. However, CMRs contain a relatively high proportion of cellulose (up to 36%) [
11], [
14], which has been the subject of limited research. Notably, the cellulose in CMRs serves as an accessible source characterized by high-quality and high-value components. In general, cellulose is lightweight with variable dimensions (i.e., micro/nanoscale) and distinct morphologies, along with a substantial surface area, making cellulose an ideal reinforcement material for green material fabrication and biocomposite production [
15], [
16]. However, eliminating the recalcitrant components of lignin and hemicellulose in CMRs poses a significant challenge for the extraction of cellulose.
Various approaches, such as mechanical disintegration, acid hydrolysis, and enzymatic treatment, have been employed to extract cellulose from natural waste [
15], [
17]. However, the aforementioned methods present specific challenges and deficiencies, such as yielding low-quality results, requiring specialized equipment, being time-consuming, and so forth [
18], [
19], [
20]. These factors limit their practical application for cellulose extraction. Currently, the oxidation method is the most commonly used approach [
21]. In this method, hydrogen peroxide (H
2O
2) or ammonium persulfate (APS) are widely employed to facilitate the individualization of fibers; however, H
2O
2 and APS are nonselective and have limited effectiveness for complex solid wastes [
22], [
23]. Therefore, there is pressing demand for a simple and highly efficient method to selectively extract cellulose from CMRs.
In this study, we present a facile and efficient thermally activated peroxymonosulfate (PMS) oxidation protocol, referred to herein as “one-step PMS oxidation,” for selectively extracting high-quality microcrystalline cellulose (MCC) from CMRs. We elucidate the preferential reactions of the active sites in amorphous polymers (i.e., lignin, hemicellulose, and the amorphous regions of cellulose) with PMS that contribute to the selective extraction of MCC by means of quenching experiments, electron paramagnetic resonance (EPR) trapping, and density functional theory (DFT) calculations. The length and width distribution, thermal stability, and crystallinity of cellulose are also evaluated. Furthermore, we examine the practical performance of this method in diverse CMRs and other cellulose-based materials (i.e., tobacco, jute, and straw) to determine its application potential. We also thoroughly evaluate the cost of cellulose extraction using techno-economic analysis (TEA) in terms of capital, operating, and material costs. Overall, this work not only demonstrates an effective method for extracting cellulose from various CMRs but also provides a platform for treating other cellulose-based natural wastes.
2. Materials and methods
2.1. Chemicals and reagents
PMS (2KHSO5·KHSO4·K2SO4, ≥ 42% KHSO5 basis), sodium chlorite (NaClO2), methanol (MeOH), tert-butanol (TBA), 5,5-dimethyl-1-pyrroline-N-oxide (DMPO), 9,10-diphenanthraquinone dye (DPA), furfuryl alcohol (FFA), and L-histidine were purchased from Aladdin (China). Sodium hypochlorite (NaClO, available chlorine ≥ 30%) and 2,2,6,6-tetramethylpiperidine (TEMP) were purchased from Macklin (China). APS and H2O2 (30% w/w) were purchased from Sinopharm Group (China). Peracetic acid (PAA, ∼15%) was purchased from ANNJET (China). A Milli-Q ultrapure water purification system (Millipore, USA; resistivity > 18.2 MΩ·cm) was used to produce deionized water. Radix Paeoniae Rubra (RPR), tobacco, Astragalus, Acorus, Suhuang (containing Ephedra, Schisandra chinensis, Perilla frutescens, etc.), Achyranthes, Shaxi (containing licorice, Akebia stem, Pyrrosia lingua, Cynanchum paniculatum, etc.), and Weisu were provided by the Nanchang Traditional Chinese Medicine Science and Technology Innovation City (China). Jute and straw were obtained from the Jiangsu province (China). Unless otherwise specified, all data and characterizations were based on RPR.
2.2. Extraction approach of cellulose from CMRs
The obtained solid CMRs were dried in a 60 °C oven for 12 h. Subsequently, a crushing device was used to break the sample into powder (60 mesh). Next, the CMRs (1 g) were introduced into 100 mL of 0.3 mol·L−1 PMS solution. The mixture was then reacted at the designated temperature (T = 40, 60, and 80 °C) with vigorous stirring for 10 h to yield a cellulose suspension. Subsequently, the solution was allowed to settle, decanted, and centrifuged at about 4400g (g = 9.8 m·s−2) for 10 min. Approximately 250 mL of water was then added and mixed thoroughly for 5 min, which was followed by settling, decanting, and centrifuging. This process was repeated five times until the solution reached a pH of approximately 7. Finally, the resulting cellulose suspension was crushed in an ultrasonic homogenizer (∼1080 W; JY98-IIID Ningbo Scientz Biotechnology Co., Ltd., China) for 20 min to obtain cellulose with a smaller particle size.
2.3. Analytical methods for assessing the content of cellulose, hemicellulose, and lignin
The contents of cellulose, hemicellulose, and lignin in the CMRs were investigated using a paradigmatic detergent-based method [
24] with a crude fiber determination instrument (ANKOM220, USA). The fundamental principle is as follows: the sample is boiled with a neutral detergent, and the remaining residue consists of neutral detergent fibers (e.g., cellulose, hemicellulose, lignin, and ash). After treatment with an acidic detergent, the remaining residue comprises acidic detergent fibers (e.g., cellulose, lignin, and ash). After treatment with 72% sulfuric acid, the residue from the acidic detergent fibers comprises lignin and ash. The background contents for different pristine and treated CMRs are shown in Tables S1 and S2 in Appendix A.
2.4. Characterization
Multiple morphological changes in the cellulose during the PMS treatment process were observed by scanning electron microscopy (SEM; ZEISS GeminiSEM 300, Germany). Fourier-transform infrared (FTIR) spectroscopy was conducted on a Thermo Scientific Nicolet iS20 instrument (USA) to examine changes in the functional groups of the cellulose. The crystal structure and crystallinity index (CRI) of the MCC were examined through X-ray diffraction (XRD; Rigaku Miniflex 600, Japan) using Cu Kα radiation. Atomic force microscopy (AFM; Bruker Dimension Icon, Germany) was employed to investigate the morphology and dimensions of the samples. A thermal stability analysis of the MCC sample was conducted using a Perkin-Elmer STA 6000 thermogravimetric analysis (TGA, USA) instrument. 1H nuclear magnetic resonance (1H NMR), 13C nuclear magnetic resonance (13C NMR) spectra and two-dimensional heteronuclear single-quantum coherence (2D-HSQC) NMR analysis were employed to monitor the structure of the CMRs on a Bruker AVANCE NEO 400 M spectrometer (Germany). 13C cross-polarization magic angle spinning solid-state NMR was used to investigate the microstructure of the obtained MCC, which was conducted on a Bruker Avance III 600 M instrument (Germany). The EPR technique was employed to trap reactive species (HO·, $\mathrm{SO}_{\dot{4}}^{-}$, 1O2) using a Bruker EPR 5000 spectrometer (Germany). More details of the characterizations and experimental process are provided in Texts S1-S3 in Appendix A. Details of the TEA assessment can be found in Supplementary Texts S4 in Appendix A.
2.5. DFT calculations
To address computational constraints, we selected model compounds, including 2-phenoxy-1-phenylethanol to represent lignin [
25], [
26], xylobiose for hemicellulose [
27], and cellobiose for cellulose [
28]. These choices were informed by the 2D-HSQC results (Fig. S1 and Text S1). All calculations were conducted using the DMoL
3 module within Materials Studio 2019 software (USA). The optimization of ground-state geometries for lignin, hemicellulose, and cellulose model compounds, as well as the total energy calculations, were based on the Becke’s 1988 exchange functional with the correlation functional by Lee, Yang, and Parr (BLYP) [
29], [
30] within the generalized gradient approximation. More specifically, equilibrium geometries and vibrational frequencies for these systems were computed using the BLYP exchange and correlation functions in conjunction with the double numerical polarized (3.5) basis set. Simultaneously, natural population analysis of the atomic charge, highest occupied molecular orbitals (HOMOs), lowest unoccupied molecular orbitals (LUMOs), and Fukui index (f-) were performed through the same method.
3. Results and discussion
3.1. Proof of concept for the selective extraction of MCC via one-step PMS oxidation
RPR, which is distinguished by its historical significance, widespread utilization, and high cellulose content [
31], [
32], was chosen as the model substance to evaluate the cellulose extraction efficiency. Various oxidants (i.e., PMS, NaClO, APS, NaClO
2, PAA, and H
2O
2) were applied and thermally activated at different temperatures (i.e., 40, 60, and 80 °C) to assess the viability of cellulose extraction from RPR, as detailed in Tables S3-S5 in Appendix A. The PMS-targeted setup (80 °C, PMS concentration 0.3 mol·L
-1) demonstrated remarkable efficacy in MCC extraction (∼95%), surpassing the other oxidants (
Fig. 1(a)) and alternative conditions (Figs. S2 and S3 in Appendix A). Conversely, NaClO, APS, and H
2O
2 exhibited relatively low cellulose extraction efficiencies (< 60%) under identical conditions. Even widely used pulping and bleaching agents (i.e., NaClO
2 and PAA) were less effective than PMS under thermal activation. SEM images revealed that the initially compact structure of the RPR underwent notable opening and disintegration following PMS oxidation treatment (Fig. S4 in Appendix A). FTIR spectra showed a gradual decrease in the intensity of the characteristic peaks at 1460, 1030, 860, and 1240 cm
−1 associated with lignin (C-H bending in CH
2 and CH
3, C-O stretching, and C-H out-of-plane bending) [
33], [
34] and hemicellulose (C-O stretching in carboxylic acid) [
35] after the PMS oxidation treatment (Fig. S5 in Appendix A). Moreover, the results from two-dimensional infrared (2D-IR) spectroscopy demonstrated that the intensities of the characteristic peaks
Φ (1380, 1460) and
Φ (1226, 1730) were positive, while
Φ (808, 1660) was negative (
Fig. 1(b)). Conversely, in the asynchronous spectrum,
Φ (1380, 1460) exhibited a negative value, while
Φ (1226, 1730) and
Φ (808, 1660) both exhibited positive values (
Fig. 1(c)). These spectral variations provided additional evidence that the removal of the relevant peak of lignin preceded the removal of hemicellulose, as detailed in Table S6 in Appendix A.
The cellulose extraction efficiency reached equilibrium after a 10 h reaction, with the complete elimination of lignin and only about 5% hemicellulose remaining (
Fig. 1(d)). This result implies that PMS can effectively promote the
in situ extraction of cellulose by removing lignin and hemicellulose. We subsequently applied this method to various commonly used CMRs, including
Astragalus,
Acorus, Suhuang,
Achyranthes, Shaxi, and Weisu, whether they consisted of single or mixed components. In each case, we consistently achieved an average cellulose yield exceeding 75% (
Fig. 1(e) and Table S2), underscoring the practicality and applicability of our approach for cellulose extraction from a diverse range of CMRs.
3.2. Features of the extracted MCC from one-step PMS oxidation
It has been reported that the fibers of MCC consist of strongly hydrogen-bonded nanofibers, and that individualized nanofiber bundles are present on the surface of the MCC [
36]. As the extracted cellulose initially comprised micro-sized fibers with diameters ranging from 1-20 μm, we subsequently applied ultrasonic treatment to facilitate the formation of nanocrystalline cellulose (NCC). The as-prepared NCC nanoparticles displayed a needle-like shape with widths ranging from 17 to 39 nm and lengths spanning from 192 to 1499 nm, as characterized by AFM and ImageJ software (
Fig. 2(a); Fig. S6 in Appendix A). These dimensions make them particularly attractive for applications in the fields of engineering and medicine [
15], [
37].
The degree of the CRI increased with oxidation processing (
Fig. 2(b); Table S7 in Appendix A), which can be attributed to the reduction and removal of non-cellulosic components and amorphous regions. Furthermore, the XRD pattern of the MCC revealed prominent peaks at 16.5°, 22.6°, and 34.5°, corresponding to the (110), (200), and (040) crystal planes, respectively [
23], [
38]. These features closely resemble the diffraction pattern of cellulose I, suggesting that the cellulose structure remained identical after PMS oxidation treatment, with the majority of its crystalline regions preserved. Cellulose I is known for its superior mechanical properties [
39], [
40], as depicted in Fig. S7 in Appendix A, which make it a valuable enhancer for improving the mechanical characteristics of composite materials.
The solid-state NMR spectra of the MCC confirmed the presence of highly pure and crystalline cellulose I, consistent with the XRD analysis. Notably, the signal at chemical shift (
δ) = 175 confirmed the presence of carboxylic acid groups (
Fig. 2(c)). This observation supports the idea that the oxidation process also targets RPR amorphous regions, leading to the formation of MCC-COOH [
23]. Carboxylated MCC has various and widespread applications, offering active sites for surface modification and the immobilization of proteins/enzymes and serving as a template for nanoparticle synthesis [
41]. Furthermore, given the necessity of thermal stability in material processing, we conducted a detailed investigation into the thermostability of the MCC using thermogravimetric (TG) and derivative thermogravimetric (DTG) curves. The results revealed that the MCC exhibited remarkable stability, with negligible mass loss (< 5%) occurring at approximately 220-250 °C and no observable pyrolysis occurring at temperatures lower than 150 °C (
Fig. 2(d)). The findings suggested that the gradual removal of lignin and hemicellulose contributed to the improved thermal stability of the MCC [
42], as hemicellulose, lignin, and pectin tend to decompose at lower temperatures than cellulose. These results indicate that the MCC possesses applicable thermal stability, especially given that most thermoplastic polymers are processed within the temperature range of 150-250 °C [
34].
3.3. Mechanistic insight into the selective extraction of MCC via one-step PMS oxidation
Scavenging experiments using MeOH and TBA did not show apparent retarding effects (Fig. S8 in Appendix A). These results imply that HO
·, $\mathrm{SO}_{\dot{4}}^{-}$ were not responsible for the cellulose extraction. Upon the addition of
L-histidine, the cellulose extraction rate decreased, and lignin and hemicellulose removal decreased (Fig. S9 in Appendix A). As reported in a previous study, in some cases, it is inappropriate to use L-histidine as a quencher to confirm the role of
1O
2 due to its rapid reaction with PMS [
43]. We further used DMPO as a trapping agent [
44]. EPR spectroscopy revealed distinct signals of DMPO-HO
· and DMPO-$\mathrm{SO}_{\dot{4}}^{-}$ in the thermally activated APS and H
2O
2 systems. However, the thermally activated PMS system exhibited no observable HO
· or $\mathrm{SO}_{\dot{4}}^{-}$ signals (
Fig. 3(a)). This finding indicated that the radicals generated in the NaClO, APS, and H
2O
2 oxidation systems are ineffective at removing lignin, hemicellulose, and the amorphous regions of cellulose—a result that differs from conventional cellulose extraction mechanisms that rely on HO
· and $\mathrm{SO}_{\dot{4}}^{-}$ radicals [
23], [
45]. Using TEMP spin-trapping EPR spectra [
44], a high-intensity peak of TEMP-
1O
2 was observed in the system. Notably, the high-intensity
1O
2 signal remained consistently stable throughout the entire oxidation process (
Fig. 3(b)). This included a remarkably intense signal within the first 2 h of the reaction, which agrees with the findings from the DPA experiments (
Fig. 3(c)). As the reaction proceeded, the intensity of the absorption peak at approximately 378 nm gradually decreased, indicating that
1O
2 was continuously present [
46]. Hence, we propose that
1O
2 is present during the one-step PMS oxidation process; however, it appears to be unserviceable for cellulose extraction.
In this context, we further monitored the PMS decomposition to elucidate the underlying mechanism. As depicted in
Fig. 3(d), it is difficult for PMS to decompose without RPR. However, introducing RPR into the system leads to significant PMS decomposition and RPR decolorization. This result suggests that specific components in RPR react strongly with PMS, rapidly consuming PMS through direct reactions [
43]. It has been reported that electrophilic peroxide readily attacks various electron-rich sites, such as aromatic rings and olefinic sites, in lignin [
43], [
47], [
48]. Attenuated total reflectance (ATR)-FTIR spectra confirmed electron transfer in the PMS-RPR interaction (Fig. S10 in Appendix A). The addition of RPR led to a notable reduction in the intensity of the 886 cm
-1 peak (attributed to the O-O stretching vibrations of PMS), signifying PMS decomposition. Furthermore, the S-O band (1059 cm
-1) associated with HSO
5- broadened with RPR, indicating the occurrence of electron transfer [
49], [
50]. Interestingly, the SO
42- vibration (1108 cm
-1) shifted to 1117 cm
-1, reflecting increased electron density—more specifically, the transfer of electrons from RPR to PMS [
51], [
52]. In conclusion, these results suggest that electrophilic PMS can selectively target specific components of RPR, such as lignin and hemicellulose. This finding further elucidates why the cellulose extraction efficiency of NaClO
2 and PAA is surpassed only by that of PMS, as research indicates that NaClO
2 and PAA can selectively remove lignin [
48], [
53].
Quantum chemistry calculations were utilized to further elucidate the mechanism involved. According to frontier molecular orbital theory, electrophilic reactions are likely to occur in electron-rich (HOMO) regions. The order of the HOMOs was 2-phenoxy-1-phenylethanol (
Fig. 3(e-i)) > xylobiose (
Fig. 3(e-ii)) > cellobiose (
Fig. 3(e-iii) (Table S8 in Appendix A). This finding indicated that lignin and hemicellulose are relatively prone to electron escape, making them susceptible to electrophilic attacks triggered by PMS [
54]. The electrostatic potential (ESP) results indicated that 2-phenoxy-1-phenylethanol has more electron-rich regions than xylobiose and cellobiose. In addition, cellobiose and xylobiose exhibited significant negative ESP areas [
55]. However, it was not possible to precisely describe the gain or loss of electrons at each site on the model molecules during the reaction. The condensed
f- was further calculated as a representation of the electrophilic susceptibility (Tables S9-S11 in Appendix A). Notably, O9 (
f- = 0.129) and C10-15 (
f- = 0.018-0.075) in lignin; O12 (
f- = 0.105), O13 (
f- = 0.104), and O15 (
f- = 0.057) in hemicellulose; and O9 (
f- = 0.160) and O11 (
f- = 0.112) in cellulose were found to have comparatively high Fukui values. This indicated that the β-O-4 bond, the aromatic ring in lignin, and the O sites in xylobiose and cellobiose are highly reactive. These findings are consistent with the reported susceptibility of phenolic compounds and vanillin to electrophilic attacks by peroxides [
48], [
53], [
56]. As more lignin is removed, there is partial degradation of the polysaccharide fraction, possibly due to the removal of lignin-carbohydrate complexes [
57], [
58]. The amorphous regions of cellulose are prone to oxidative cleavage. However, in crystalline domains, they are less susceptible to attack by PMS [
53]. Overall, this provides proof-of-concept that one-step PMS oxidation can selectively target lignin, hemicellulose, and amorphous cellulose domains, leading to the efficient extraction of MCC.
3.4. TEA evaluation of the one-step PMS oxidation process and its practical applicability
Notably, the obtained MCC can be readily transformed into NCC through an ultrasonication process. Despite the existence of an efficient protocol for renewable NCC extraction, the industrial scale-up necessitates careful consideration of both cost and environmental impact. The TEA of NCC, which encompasses material, operational, and capital costs, was 2.04 USD·g
−1 (
Fig. 4(a); Text S4 in Appendix A), which falls below the reported high cost for NCC slurry preparation (5 USD·g
−1) [
13]. This result emphasizes that the one-step direct PMS oxidation process is not only cost-effective but also holds economic promise for NCC extraction. It is important to note that the primary cost lies in PMS consumption, which requires the management of sulfur-containing wastewater (Table S12 in Appendix A) and addressing backend treatment. This aspect may require further attention when implementing the approach in practical fieldwork. The second most significant cost is that of electricity, which could be reduced with advancements in renewable electricity generation technologies [
57].
In addition to its relatively low cost, this study introduces a remarkably scalable method for cellulose extraction from various cellulose-based sources, including tobacco, jute, and straw (
Fig. 4(b); Table S13 in Appendix A). The utilization of the one-step PMS oxidation technique enables the efficient retrieval of cellulose from these plant materials, achieving an average extraction rate of approximately 80%. These findings suggest that this highly effective cellulose extraction process extends beyond CMRs to encompass natural residues such as tobacco, jute and straw, demonstrating its practical versatility and applicability.
4. Conclusions
In summary, the facile and efficient one-step PMS oxidation method was successfully shown to extract MCC, demonstrating highly selective cellulose extraction from a diverse array of CMRs. This finding underscores this method’s potential for overcoming the recalcitrance of lignin and hemicellulose to oxidation. The resultant high-quality MCC holds significant promise for a myriad of applications, including nanofillers, textiles, materials synthesis, and drug delivery. EPR characterization and theoretical calculations revealed that thermally activated PMS primarily targets the β-O-4 bond, the aromatic ring in lignin, specific O sites in hemicellulose, and the amorphous region of cellulose, thereby facilitating the selective extraction of MCC. Experimental investigations conducted on tobacco, jute, and straw validated the real-world applicability of the one-step PMS oxidation method, confirming its high scalability. This approach also successfully addresses the challenge of extracting NCC with notable cost effectiveness. In summary, the one-step PMS oxidation method effectively tackles a key challenge in converting solid waste into a valuable resource. Its notable potential for onsite applications is particularly promising for addressing cellulose-based natural waste.
Acknowledgments
This work was financially supported by the National Key Research and Development Program of China (2022YFC3205300).
Compliance with ethics guidelines
Jinjing Huang, Xia Liu, Kaixing Fu, Shengyun Yang, Shiqing Zhou, and Jinming Luo declare that they have no conflicts of interest or financial conflicts to disclose.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.org/10.1016/j.eng.2024.03.008.
PDF
(1618KB)
