Abstract
Low-cost production of nanocellulose from diverse lignocellulosic feedstocks has become an important topic for developing sustainable nanomaterials. The available feedstocks include both woody and non-woody plants, where the latter are relatively underutilized. Interestingly, the porous structure and low lignin content in most non-woody plants, such as agricultural residues and natural fibers, also makes them ideal sources for lower energy nanocellulose production using simpler methods than those required to process woody plants. To enhance the goal of circularity, this review first provides an overview of the nanocellulose conversion from cellulose and then comprehensively discusses the use of non-woody feedstocks for nanocellulose production. Specifically, the availability of suitable non-woody feedstocks and the use of low-cost processes for pulping and cellulose oxidation treatments, including alkaline, solvent pulping, and nitrogen-oxidation treatments, are discussed. The information in this review can lead to new opportunities to achieve greater sustainability in biobased economies. Additionally, demonstrations of nanocellulose-based water purification technologies using agricultural residues derived remediation materials are highlighted at the end of this review.
Download PDF
Full Article
Nanocellulose Preparation from Diverse Plant Feedstocks, Processes, and Chemical Treatments: A Review Emphasizing Non-woods
Rasel Das,a Tom Lindström,b,* Madani Khan,a Mahdi Rezaei,a and Benjamin S. Hsiao a,*
Low-cost production of nanocellulose from diverse lignocellulosic feedstocks has become an important topic for developing sustainable nanomaterials. The available feedstocks include both woody and non-woody plants, where the latter are relatively underutilized. Interestingly, the porous structure and low lignin content in most non-woody plants, such as agricultural residues and natural fibers, also makes them ideal sources for lower energy nanocellulose production using simpler methods than those required to process woody plants. To enhance the goal of circularity, this review first provides an overview of the nanocellulose conversion from cellulose and then comprehensively discusses the use of non-woody feedstocks for nanocellulose production. Specifically, the availability of suitable non-woody feedstocks and the use of low-cost processes for pulping and cellulose oxidation treatments, including alkaline, solvent pulping, and nitrogen-oxidation treatments, are discussed. The information in this review can lead to new opportunities to achieve greater sustainability in biobased economies. Additionally, demonstrations of nanocellulose-based water purification technologies using agricultural residues derived remediation materials are highlighted at the end of this review.
DOI: 10.15376/biores.19.1.Das
Keywords: Agricultural residues; Natural fibers; Sustainable feedstocks; Pulping; Cellulose oxidation; Nanocellulose
Contact information: a: Department of Chemistry, Stony Brook University, Stony Brook, NY 11794-3400, USA; b: KTH Royal Institute of Technology, Stockholm 100 44, Sweden;
* Corresponding authors: toml@kth.se; benjamin.hsiao@stonybrook.edu
INTRODUCTION
Cellulose is the most abundant polymer in the world, consisting of several hundred to many thousands of connected D-glucose units in the backbone structure, and it is commonly known as an essential compositional element in plants and algae. Recently, the entity of nanocellulose (cross-sectional dimension in between 2 and 10 nm), often considered as the building block of the plant cell wall, has been recognized as a sustainable nanomaterial for many emerging applications. Interest in this material is heightened due to its high surface area, abundant functionalities, and scaffolding capability, while maintaining the unique and environmentally friendly properties of cellulose. Using top-down preparation approaches, two types of nanocellulose can be produced from lignocellulosic feedstocks: cellulose nanofibers (CNF) and cellulose nanocrystals (CNC). CNFs are longer, pliable, and less crystalline, and they can intertwine among themselves. Alternatively, CNCs are shorter, rigid, with higher crystallinity. CNCs often appear as rod-shaped nanoparticles with lower aspect ratios. The choice between CNF and CNC depends on the requirements of the targeted application. Additionally, bacterial nanocellulose (BNC) can be produced by bacteria using sugar solutions as feedstocks through the bottom-up approach. The details of structures and properties in these different nanocelluloses have been comprehensively discussed in recent reviews (e.g., Das et al. 2022; Lv et al. 2023).
In a chosen defibrillation process, the choice of feedstock can play a vital role in determining the required consumptions of energy, chemicals, and water to produce nanocellulose. Typically, non-woody materials with looser structure need less energy (and chemical treatments) for nanocellulose production, where relatively simple pulping/ defibrillation methods can be adopted. Often these materials, such as agricultural residues and natural fibers, are considered less valuable than woody materials for conventional applications in construction and packaging. As a result, these non-woody biomasses, although truly abundant, are relatively underutilized around the globe. For example, China currently produces over 56 million tons of agricultural waste per annum (Kamel et al. 2020), and the United States is projected to produce 320 million tons of primary crop residues by 2030, including corn stover, wheat straw, barley, sorghum, and oats residues. (US-DOE 2011). Such residues are commonly used either as soil amendments, animal feeds, or for biofuel production. In the last decades, there have been extensive efforts to explore the usage of non-woody fiber residues for ethanol production. However, these efforts have been gravely challenged by the low oil prices; thus, the economics are simply not viable. Several research groups have also investigated the concept of integrated production of nanocellulose and biofuel, but the economics are still not clear (Menon and Rao 2012; Zhu et al. 2011). As agricultural residues are truly abundant and low-cost, and their improper disposal can also cause grave environmental consequences, there is an urgent need to explore simple ways to upcycle these materials and create new economic opportunities.
Agronomic benefits of agricultural residues and natural fibers over woody materials vary from species to species, where the most visible benefits of these non-woody plants include shorter growth cycles and higher biomass yield. Some non-woody plants, such as jute and sorghum, also exhibit lower recalcitrance for biorefinery than wood due to their relatively low lignin content. In other words, these plants require less energy to remove lignin and hemicellulose components, which is the primary step for nanocellulose production. There are, however, other factors that must be considered for valorization of non-woody feedstocks, such as collection, transportation, decortication, drying process, and retting, which are all essential economic and technological issues for agro-based and non-woody fiber manufacturing. Firstly, as the handling, storage and preparation of these materials have not received the same attention as that for woody materials, one must consider their potentially serious logistical problems. For example, the long-distance cost to transport agricultural residues can be prohibitively high. Thus, in large-scale production of fruits, such as palm oil fruit, citrus, and coconut, it may be a viable approach to process the waste on site, to reduce costly transportation logistics. Secondly, the morphology and chemical composition of non-woody plants and their specific breed often depends on harvesting time and geographic location. Thus, their pulp properties change seasonally. This variability of fiber properties constitutes a major challenge to runnability and efficiency of continuous large operations. To continue, the handling of agricultural residues and natural fibers is often labor-intensive, especially for leaf and fruit fibers, due to hand-stripping, while natural fibers require decorticating equipment, which is commonly used for many stem-fiber crops. Moisture content control and storage conditions are also critical factors, with an extensive number of books and papers dealing with such issues (Young, 1997a,b). Thirdly, most non-woody plants contain silicate compounds, which may remain with the fibers in acidic processes, but dissolve in alkali, which will complicate the recovery operations, while high ash content which is particularly prevalent in the family of grasses (Poaceae), poses its own set of challenges. In summary, the handling, preparation, storage, and transportation logistics of utilizing agricultural residues, including the variability of the feedstock and the issue of the high ash content in some of these materials, can create difficulties for large scale modern facilities (such as greenfield kraft pulp mills), and require substantial investment cost and a large minimum production capacity (usually in the range of 500 to 1000 kton/year) for economical operation. Hence, the challenge is to find a suitable and economically feasible treatment method to process non-woody materials, such as agricultural resides and natural fibers, which will be discussed in detail in this review.
The major aim of this review is to improve the circularity of bioeconomy by converting low-valued non-woody feedstocks to high-valued nanocellulose. However, considering the diverse interests of potential readers, the first part of this review summarizes the conversion of non-woody lignocellulosic biomass into nanocellulose using different production methods along with the introduction of some essential subjects related to the field. After that, this review contains an overview of agricultural residues and natural fibers, focusing on (i) the availability of these non-woody feedstocks to produce high quality nanocellulose, especially CNF, (ii) the composition of various biomasses, and (iii) low-cost processes that have been demonstrated for the manufacture of nanocellulose and the viability of their upscaling. Various pulping processes, their advantages and disadvantages, and the development of a unique zero-waste nitro-oxidation process (NOP) are reviewed and discussed. Perspectives on the use of small mills for possible adoption of some of these technologies are also discussed. As there is not yet a market for nanocellulose in the order of million tons/year due to its relatively high cost, this article also argues that the development of viable nanocellulose manufacturing facilities using agricultural residues (or agro-waste) should be small-scale operations, unless operated as a biorefinery.
Conversion of Cellulose into Nanocelluloses
In the past few decades, a plethora of methods have been demonstrated to prepare nanocellulose from lignocellulosic plants (Table 1) and bacterial materials. Many publications and reviews have covered these methods and the corresponding structure-property relationships in resulting nanocellulose. These methods include chemical treatments, biological treatments, and mechanical treatments. This section will summarize some highlights of these methods and the distinct physicochemical properties of resulting nanomaterials.
Chemical treatments
Methods using suitable chemical treatments to attain nanocellulose from lignocellulosic feedstocks can vary depending on the starting feedstocks as they contain different contents of cellulose, lignin, hemicellulose, pectin, and other organic compounds. Typically, a pretreatment method can be used to remove the non-desired hetero-polymers and organic substances (Habibi 2014). The lignin component can be removed using both acid and bases. However, the efficiency of the removal and the type of lignin removed may vary depending on the concentration of acid and base, types of acids and bases, temperature, additive used, size of the starting material, the amount present, and the composition of both biomass and lignin (Klemm et al. 2011; Baruah et al. 2018; Mankar et al. 2021).
Table 1. Acidic Hydrolysis Treatments of Agricultural Residues and Natural Fibers for Nanocellulose Production with Diverse Structural Dimensions
Acid and base treatments of the biomass can also reduce the cellulose component to a lower degree of polymerization, since both acid and base can hydrolyze the glycosidic bond of the cellulose chains. Other chemical pretreatments include the use of organic solvent mixtures, ammonia fiber explosion, co-solvent enhanced lignocellulosic fractionation, supercritical fluids, deep eutectic solvents, and TEMPO-oxidation (Mankar et al. 2021) The most common industrial pretreatment process is kraft pulping, using a mixture of sodium sulfide and sodium hydroxide to produce nearly pure cellulose. The details of these chemical treatments are extensively discussed in the later parts of this review.
Pretreated cellulose materials can be further chemically treated to produce nanocellulose. For example, acid hydrolysis (using hydrochloric acid or sulfuric acid at relatively high concentrations) of pretreated cellulose materials can remove residual lignin and hemicellulose as well as some amorphous cellulose and yield highly crystalline CNC, which have also been called cellulose “whiskers”. CNC has a rod-like shape with a diverse range of cross-sectional dimensions between 5 and 70 nm and a length typically around 100 nm (Cheng et al. 2017; Kusmono et al. 2020; Pawcenis et al. 2022; Wulandari et al. 2016; Yu et al. 2013). Table 1 summarizes the effects of some common acid treatments on the structure of CNF and CNC using different biomass feedstocks.
In addition to acid hydrolysis using relatively strong acids, some chemical methods using milder conditions also have been demonstrated. For example, Ji et al. (2019) demonstrated a relatively green method using citric acid (60 to 80%) for 0.5 to 4 h, followed by 20 min of sonication to produce both CNC and CNF from bleached sugarcane bagasse. The nanocellulose products created by this method included CNC with a cross-sectional dimension between 20 and 30 nm and a length between 250 and 450 nm, as well as CNF with a larger cross-sectional dimension between 30 and 60 nm and a length between 500 and 1000 nm (Ji et al. 2019).
Another popular chemical treatment for nanocellulose production is TEMPO-mediated oxidation of cellulose via the use of the 2,2,6,6-tetramethylpiperidine 1-oxyl radical (TEMPO) agent, sodium bromide, and sodium hypochlorite at pH = 10. The crystallinity, degree of polymerization, degree of oxidation, and aspect ratio of the resulting cellulose nanoscale fibers can be controlled by varying the amount of sodium hypochlorite and the number of homogenization cycles (Zhou et al. 2018). The creation of carboxylate groups due to TEMPO reaction can lead to anionic charges on the cellulose surface, which can facilitate the defibrillation process, leading to a small cross-sectional dimension between 5 and 10 nm and a length between 100 nm and a few micrometers, depending on the experimental conditions (Saito et al. 2007). In the TEMPO reaction, the radical TEMPO-catalyst is first oxidized by sodium hypochlorite to generate the N-oxoammonium-type structure, where the oxoammonium compound can subsequently oxidize the cellulose surface (Isogai et al. 2018).
There are other radical reaction methods used to produce nanocellulose from cellulose pulps. For example, in the Fenton-oxidation approach, a transitional metal with a low oxidation state can be used to react with hydrogen peroxide and generate hydroxyl radicals (Shokri and Fard 2022). In such a reaction, iron is the most common type of metal employed, as it is naturally abundant and possesses the oxidation state of +2 and +3. Fan et al. (2019) demonstrated that microcrystalline cellulose can be converted into CNC using 30% H2O2 and FeSO4 at 60°C. The resulting nanocellulose had a cross-sectional dimension between 19 and 23 nm, a length between 92 and140 nm, and the degree of carboxylation as high as 2.2 mmol/g.
While Fenton oxidation uses metal to catalyze the generation of hydroxyl radicals, oxidants such as ammonium persulfate (APS) can also break down and create highly reactive sulfate radicals at moderately elevated temperatures (60 to 80 °C). These sulfate radicals are capable of simultaneously pulping and oxidizing cellulose (Leung et al. 2011) In one study, when Sansevieria trifasciata, commonly known as “snake plant” or “in-law’s tongue”, was treated with 1.5 M APS at 70 °C for 16 h, the reaction resulted in nanocellulose with an average width of 7.4 ± 2.7 nm, an average length of 156.4 ± 54.6 nm, and a crystallinity index of 87.4% (Indirasetyo and Kusmono 2022). Another study showed that the use of N,N,N’,N’-tetramethylethylenediamine along with APS could increase the yield of carboxylated CNC (up to 62.5%) for cotton pulp, where the crystallinity index could rise to 93% (Liu et al. 2020).
Ionic liquids are different classes of salts that consist of cations and anions, and they exist in the liquid state at room temperature. Certain ionic liquids can dissolve cellulose and hydrolyze cellulose. Haron et al. (2021) published a review on the use and effects of different ionic liquids for nanocellulose production. Their properties and efficiencies and sustainability issues were highlighted. With the appropriate usage, ionic liquid can be used to produce nanocellulose from cellulose, while recycled ionic liquids can be used multiple times without losing much of their efficiency. For nanocellulose production, some demonstrated ionic liquids have included [Bmim][Cl] (1-butyl-3-methylimidazolium chloride) [Bmim] [OAc] (1-butyl-3-methylimidazolium acetate), [Bmim] [HSO4] (1-butyl-3-methylimidazolium hydrogen sulfate), [Emim][Cl] (1-Ethyl-3-methylimidazolium chloride), and [Emim][OAc] (1-ethyl-3-methylimidazolium acetate). The nanocellulose treated with ionic liquids typically exhbited a cross-sectional dimension between 15 and 20 nm (Haron et al. 2021).
Work in the authors’ lab demonstrated that when raw lignocellulosic feedstock (e.g., untreated jute biomass) was treated with a mixture of nitric acid and sodium nitrite, both pulping and cellulose oxidization processes could simultaneously occur. This reaction is termed the nitro-oxidation process, and it is capable of producing carboxylated nanocellulose from any lignocellulosic feedstock. The resulting nanocellulose typically possesses an average cross-sectional dimension of about 5 nm and a length ranging from 200 to 2500 nm (Sharma et al. 2018b,c).
Biological treatments
Biological treatments using fungi or enzymes to remove lignin and hemicellulose from lignocellulosic biomass are often viewed as environmentally friendly and inexpensive methods to produce nearly pure cellulose. Notably, biological treatments require significantly less energy than that of mechanical treatments. The most common biological pretreatment to remove lignin and hemicellulose from raw feedstock is with fungi. Two types of fungi are typically used – hydrolytic and lignolytic. Hydrolytic microorganisms are responsible for degradation of polysaccharides (i.e., cellulose and hemicellulose), whereas lignolytic microorganisms are responsible for degrading lignin. For pulping purposes, white-rot, brown-rot, and soft-rot fungi are typical classes of microorganisms utilized. In the case of enzymes, species such as laccases, lignin peroxidase, manganese peroxidase, and versatile peroxidase are also efficient in removing lignin. Extensive discussions of biological pretreatment have been made by Baruah et al. (2018) and Østby et al. (2020). Once the pure cellulose is produced, various enzyme strains can be further used to produce different-sized cellulose. The final product, purity, aspect ratio, and water content can be finely tuned by using different strains of bacteria and pH medium. Some notable examples are as follows.
Lytic polysaccharide monooxygenases (LPMOs) are mononuclear copper enzymes that can cleave the glycosidic bond of cellulose by inserting oxygen at the C1 and/or C4 positions and oxidizing the reducing end. It has been demonstrated that the addition of LPMO in pretreated flax pulps, cotton linters, softwood (birch), and hardwood pulps could directly produce CNC and CNF (Karnaouri et al. 2022). Endoglucanase enzyme was also found to be able to hydrolyze the glycosidic bond in cellulose, especially in the amorphous region, and when added to pretreated flax and hemp fibers, endoglucanase enzyme can yield rod-like nanofibrils with cross-sectional dimension of about 10 nm and length about 200 nm ( Xu et al. 2013; Wang et al. 2016). Many commercially available enzymes are used in combination with chemical and/or mechanical treatments to produce CNC and CNF from agricultural residues. Some examples include Fibercare® enzyme for treating wheat straw, curauá fibers, and bleached eucalyptus kraft pulp; Pectinase PL AmanoTM enzyme for treating orange peel; OptimashTM VR enzyme for treating soybean straw; and cellulase enzyme (from Sigma) for treating bagasse pulp (Michelin et al. 2020). While enzymatic hydrolysis can generally produce nanoscale fibers with good transparency and high specific surface area, for large-scale operations, mechanical treatments are often supplemented for nanocellulose production (Aguado et al. 2022).
Fig. 1. TEM images of CNCs prepared from different feedstocks using different treatment conditions. These images show some structural variations but similar CNC characteristics. (a) Photographs of sulfuric acid hydrolyzed CNCs suspension (1.0 wt%) (inset: a 10.0 wt% CNC hydrogel), and (b) corresponding TEM image of CNC derived from wood pulp. (c)-(l) TEM images of CNCs derived from different feedstocks using different treatment conditions (Azizi Samir et al. 2005 [Reprinted with permission from Biomacromolecules 6(2), 612-626, American Chemical Society]; Cheng et al. 2022 [Permission from SpringerNature, Reprinted from “Comparative study on properties of nanocellulose derived from sustainable biomass resources,” Cellulose 29(13)]; Habibi et al. 2008 [Used with permission of Royal Society of Chemistry from “Bionanocomposites based on poly (ε-caprolactone)-grafted cellulose nanocrystals by ring-opening polymerization,” Journal of Materials Chemistry 18(41)]; Lin et al. 2022 [Reprinted from Front. Chem. 10:922437, CC BY]; Pornbencha et al. 2023 [Reprinted from RSC, CC BY-NC 3.0 Deed]; Roman and Winter 2004 [Reprinted with permission from Biomacromolecules 5(5), 1671-1677, ACS]; Siqueira et al. 2010 [Reprinted with permission from Langmuir 26(1), 402-411, ACS]; Usov et al. 2015 [Reprinted from Nature Communications 6, 7564, CCBY]; Zainuddin et al. 2017 [Used with permission of Elsevier from Carbohydrate Polymers 163, 261-269]; Zhao et al. 2019 [Used with permission of SpringerNature, CC-BY-NC-ND).
Mechanical treatments
Mechanical treatments are frequently used to pulverize raw biomass such as agricultural residues before and after the applications of other physical or chemical procedures to produce nanocellulose. Common mechanical treatments include grinding, milling, extrusion, high-pressure homogenization, cryocrushing, and sonication. The descriptions of different mechanical methods for nanocellulose production can be found in several reviews (Habibi 2014; Islam et al. 2014; Jonoobi et al. 2015; Patil et al. 2022; Sanchez-Salvador et al. 2022a). While mechanical treatments alone cannot generate pristine nanocellulose from raw biomass, they can be used in tandem with other chemical or enzymatic treatments to increase the yield and crystallinity of nanocellulose, which will be discussed later. However, mechanical treatments require additional cost due to the energy consumption that varies with the structure of the feedstock. Therefore, one should carefully select an appropriate biomass feedstock for producing cost-effective nanocellulose for practical applications.
Serra-Parareda et al. (2021) reported that the nature of raw feedstocks can also impact the resulting nanocellulose structure and morphology. Transmission electron microscopy (TEM) images of CNC and CNF derived from various feedstocks using different treatment methods are illustrated in Figs. 1 and 2, respectively. From these images it is apparent that the size and shape of nanocellulose are very different due to the variations in feedstock materials and treatment conditions.
Fig. 2. TEM images of CNFs prepared from different feedstocks using different treatment conditions. The images show some structural differences but similar CNF characteristics. (a) Photographs of TEMPO oxidized 2.0 wt% CNF suspension (inset: 10.0 wt% CNF hydrogel) and (b) corresponding TEM image of CNF derived from bleached softwood kraft pulp. (c)-(l) TEM images of CNFs derived from different feedstocks using different treatment conditions (Alemdar and Sain 2008 [Reprinted with permission of Elsevier]; Usov et al. 2015 [Reprinted from Nature Communications 6, 7564, CCBY]; Sharma et al. 2017b, 2018a [Reprinted from Biomacromolecules, Copyright 2017, American Chemical Society, and ACS Sustainable Chemistry and Engineering, Copyright 2018, ACS]; Chen et al. 2019 [With permission from SpringerNature, CC-BY-NC-ND]; Chen et al. 2022 [Reprinted from Nanomaterials, CC BY]; Kumari et al. 2019 [With permission from SpringerNature, CC-BY-NC-ND]; Narkpiban et al. 2019; Zhan et al. 2020 [Used with permission of RSC, from Water Research & Technology 6(11), 3080-3090]; Carvajal-Barriga et al. 2022, CC BY, Permission granted by authors).
Although the combination of the microscopic technique (e.g., TEM) and image analysis tool can provide information about the nanofiber width, it is relatively difficult to determine the nanofiber length because of the fiber entanglement and the vague existence of nanofiber ends. Furthermore, as the CNF suspension becomes inhomogeneous, the images can contain both individual and aggregates (bundles) of nanofibers. In this case, the fiber width can vary along the fiber length, unless they are sufficiently delaminated (Balea et al. 2020). In this regard, CNC materials are relatively easier to characterize. In their review, Balea et al. (2021) examined the challenges related to nanocellulose characterization and attempted to address the question of why the market has yet to realize its full potential.
Overview of Agricultural Residues and Natural Fibers for Nanocellulose Production
There have been extensive investigations of nanocellulose production from agricultural residues and natural non-woody fibers. As shown in Fig. 3, a great deal of publications can be found that focus on the characterizations and applications of CNCs and CNFs derived from non-woody feedstocks (including algae). Unfortunately, the amount of published work does not correlate with their economic viability. Because of the relatively high crystallinity (typically around 50%), large surface area (nearly 100 m2/g) and effective fiber network formation capability (Rocha et al. 2018), CNF has secured its position as emerging sustainable nanomaterials for electronic and biomedical applications (amounts to 31% and 29% of the total publications, respectively).
Fig. 3. Top CNF research fields for top 18 non-woody biomass feedstocks (Pennells et al. 2020, with permission from SpringerNature, CC-BY-NC-ND)
In Table 2, the chemical composition (i.e., cellulose, hemicellulose, and lignin) and annual production yield of some agricultural residues are tabulated. In Table 3 the data are also summarized for natural plant fibers and some agricultural fibers, from which nanocellulose can be extracted, albeit not at commercial scale to our knowledge.
Table 2. Chemical Compositions and Annual Production Yields of Varying Agricultural Residues (Tye et al. 2016)
Table 3. Chemical Compositions and Annual Production Yields of Varying Natural Plant Fibers (mainly from non-woody plants)