1 Introduction

One of the main issues in agriculture has always been the massive generation of agricultural waste, which is mainly composed by lignocellulosic fibers from different sources. This is a result of various processing mechanisms, being usually destined for animal feeding or even used as organic fertilizer. While the use of landfills for waste disposal seems to be an acceptable solution, the huge quantity of agricultural waste is leading to their fast overload, making the use of landfills a temporary choice (Berthet et al. 2015).

Reuse is one of the main components of Circular Economy (CE) concept, which consists in the use of a pre-processed material as a raw material for a diversity of processes. These may include energy generation and production of high value added products, both of them being crucial for sustainable development (Wong et al. 2016). Therefore, agricultural waste rich in lignocellulosic fibers (straw, bark, husk, shells, leaves, bagasses) may be of better use if following this concept rather than treating it as a residue. One of the fields that profit from this concept is nanotechnology (García et al. 2016).

The field of nanotechnology has attracted great research interest regarding reinforcing properties when applied to compounds in the form of nanocomposites. Nanotechnology is the study of structures with dimensions of 1–100 nm, such as nanocellulose, which have unique properties when applied in polymers as they naturally present high crystallinity, large surface area, high thermal stability (depending on chosen acid for hydrolysis) and good mechanical properties (Dufresne 2010; Habibi et al. 2010).

Nanocellulose, in turn, is a general term that represents the main products we may obtain from cellulose, which are nanocrystalline cellulose (generated through enzymatic or acid hydrolysis) and cellulose nanofibers (obtained after mechanical defibrillation of cellulosic fibers). While industrial production of nanocrystalline cellulose is a well-known process, characterized by presenting a low energy consumption, easy to handle hydrolysis reaction and high crystallinity of produced nanocrystals, the following review article will focus on processing of lignocellulosic residues for generating nanocrystalline cellulose (Chen et al. 2015).

These nanoscaled compounds have been employed for several applications such as coating (Jabbar et al. 2017), anti-microbial agent (Pal et al. 2017), pharmaceutical excipient (Zainuddin et al. 2017) and strength enhancer (Yin et al. 2016), for instance. They have good potential for the partial replacement of fossil derived polymers because they also improve material resistance when they are incorporated into products. The use of nanocomposites is advantageous due to their low cost, low density and wide variety of cellulosic sources (Chirayil et al. 2014; Rebouillat and Pla 2013). Many reviews have focused solely on production and application of nanocellulose from wood pulp and derivatives while leaving aside agro-industrial sources, which pose as a great potential source for their wide availability and low cost, in addition supplying an alternative to conventional industrial waste treatment (García et al. 2016). This review aims to provide a better characterization of nanocrystals manufacture process from agro-industrial residues as a guide for future studies in the field.

2 Lignocellulosic fibers

In general, fibers are composed of three main structures: cellulose, hemicellulose and lignin. The crystalline portion (cellulose) is found embedded in the amorphous region which is composed of hemicellulose and lignin (Siqueira et al. 2010b). Other compounds may be present in small amounts such as ash, pectins and silica (Casas-Orozco et al. 2015; Johar et al. 2012; Wang et al. 2014a, b). Although the generally proposed model for cellulose presents both crystalline and amorphous domains, it has recently been suggested that its crystalline nature is a result of thermal and mechanical treatment of fibril structure, which leads to reduced accessibility to water, thus improving final crystallinity. Authors suggest that X-ray diffractometry (XRD) behavior is related to an aligned conformational state rather than a crystalline one. As this study verified that unmodified wood cell wall microfibrils are accessible to water, this could indicate that fibrils are not composed of crystals (Wüstenberg 2015). Lignocellulosic components are illustrated in Fig. 1.

Fig. 1
figure 1

Composition of various lignocellulosic residues (USDOE 2016)

Full size image

2.1 Main polymers from lignocellulosic fibers

Cellulose, followed by starch, is the most abundant and renewable polymer resource that is available. It is often found in association with lignin and hemicellulose, although it can be found in its pure form in nature in the seed hairs of cotton plants, for example. Because it is the main product used in the paper industry, it must be isolated from non-cellulosic structures in order to provide its well-known resistant and versatile structure. Cellulose basically consists of α-d-glucopyranose residues linked through β-1,4 bonds. Cellulose chains tend to be very close to each other due to their hydrophilicity, leading to a stable and crystalline structure, which is characteristic of this polymer. Among polymers, cellulose presents the highest tendency to crystallization, which is a positive factor that could be used for better results in relation to structural reinforcement (Wüstenberg 2015).

Despite being found mostly in the cell wall structure of higher plants, cellulose is known to be present in organisms such as algae and certain bacteria. Unlike cellulose extracted from plants, bacterial cellulose is highly crystalline and is often used as a model for polymorph studies (Shen et al. 2015).

In addition to cellulose, which is the target portion from lignocellulosic fibers for obtaining cellulose in nanoscale, hemicellulose and lignin are also found in association in natural fibers. The former mainly consists of xylose, mannose, glucose and traces of other polysaccharides, which may vary depending on the fiber origin. This non-crystalline heteropolysaccharide is readily soluble in alkali solutions, being this treatment the most commonly employed one for cellulose purification (Wang et al. 2015a, b; Yadav and Hicks 2015). Finally, lignin is an amorphous polyphenolic polymer found in plant tissues that works as a structural support in plants. It is characterized by its hydroxyl and methoxyl groups, and its internal structure suggests a highly unsaturated and aromatic character. One of the major challenges in isolating cellulose is removing most of the lignin content without overly affecting cellulose structure. Several studies have attempted to overcome this problem mainly through adapting reaction conditions during processing (Azelee et al. 2014; Hodzic and Shanks 2014; Wang et al. 2015a, b).

3 Nanocellulose crystals

The major compound that can be obtained from purified cellulose is nanocrystalline cellulose (NCC), also known as cellulose nanocrystals (CNCs), or cellulose nanowhiskers (CNWs). These rod-shaped structures can be obtained through mechanical/chemical methods after the isolation of cellulose, generally through acid hydrolysis of pure cellulose. The dimensions of CNCs may vary depending on the source of cellulose and processing conditions used to obtain them (Moriana et al. 2016).

The isolation of stable colloidal suspensions was first described in the 1950s when Ranby and Ribi managed to obtain cellulose crystals from wood and cotton cellulose fibers after sulfuric acid treatment; the structures were approximately 50–60 nm in length and 5–10 nm in width (Rånby and Ribi 1950). Further studies have tried to optimize the obtaining of CNCs and to eliminate amorphous cellulose thereby improving thermal and mechanical behavior of produced nanocellulose.

3.1 Sources of nanocrystals

Although cellulose is commonly found in the cell wall of plants in nature, it is also present in different sources such as agricultural residues, a few species of algae, amoeba and bacteria (Mohammadkazemi et al. 2014; Sundari and Ramesh 2012). Because this review focuses on agro-industrial cellulosic sources, bacterial cellulose will be briefly mentioned for comparison. Some bacteria, such as gram-negative Gluconacetobacter xylinus, naturally synthesize pure cellulose as part of their metabolism through carbon sources. Consequently, it has been widely studied for its biomedical and cosmetic applications as it presents high crystallinity (compared to raw plant cellulose), biodegradability and high pure structure, which avoids the need for cellulose purification, being this a necessary step for obtaining nanocrystals from lignocellulosic samples (Gao et al. 2015; Kiziltas et al. 2015).

Despite presenting good overall properties, bacterial cellulose may not be ideal for the mass production of nanocellulose; wood pulp is the current best source for this process. Bacterial cellulose synthesis is currently only possible using bench scale as it tends to be much more expensive than chemical treatment of cellulose originating from plant sources, as well as the fact it is dependent on cells metabolism and their carbon sources. Viable use of lignocellulosic waste as a source of nanocellulose may be explained by the great difference in the degree of polymerization between sources, which consists of about 13,000–14,000 for plant cellulose and 2000–6000 for bacterial cellulose thereby resulting in higher yield of nanostructures from plant cellulose. These factors may justify the preparation of nanocrystals for the mass production of lignocellulosic sources (Khajavi et al. 2011; Sheykhnazari et al. 2011).

In spite of the fact that wood pulp provides a better yield of nanocrystals, alternative sources may be an interesting alternative for this process. Cellulose content can vary between sources, with cellulose generally representing the majority (in addition to the hemicellulose and lignin fractions) followed by traces of pectin, extracts and ashes (Sun 2010). The use of agro-industrial residues is environmentally friendly because residues can be treated and transformed into a product with high added value. The chemical composition of different lignocellulosic sources is listed in Table 1.

Table 1 Chemical composition of lignocellulosic materials from different sources
Full size table

4 Nanocellulose preparation procedure

The processing of cellulose to nanocrystals can vary depending on different factors, such as temperature, time and concentration of reagents because these factors result in different contents of cellulose, hemicellulose and lignin (Wüstenberg 2015). For example, the obtaining of CNCs from cotton cellulose dispenses alkali and bleaching treatments because it is naturally composed of a high content of pure cellulose (Lu and Hsieh 2010; Morais et al. 2013).

The usual method of processing utilizes alkaline treatment with a strong base such as sodium hydroxide (NaOH) or even potassium hydroxide (KOH). The bleaching process is subsequently performed, which removes most of the lignin content, usually with sodium chlorite (NaClO2) under acidic conditions. Finally, the resulting fibers are further treated with a strong acid such as sulfuric acid (H2SO4), followed by nanocrystals dispersion and the optional drying stage. Although variations in treatments may be performed for the obtaining of nanocellulose from lignocellulosic biomass, which will be cited throughout this article, the general procedure is represented in Fig. 2.

Fig. 2
figure 2

General procedure for processing of lignocellulosic residues to nanocellulose

Full size image

4.1 Pre-treatment

In general, fibers from different sources should be milled or triturated (in order to increase the contact surface and improve the effects of subsequent treatments), and afterwards washed in deionized water to remove soluble impurities that may be present in samples from processing (Abraham et al. 2011; Karimi et al. 2014). Milling to powder is necessary for uniformity of size and also to improve swelling capacity in water (Chen et al. 2011a). Fibers containing excess impurities may lead to an efficiency reduction in subsequent stages. A few studies have also performed fiber de-waxing through treatment with a mixture of benzene/ethanol and toluene/ethanol as washing with deionized water may not be efficient for this purpose. Similarly, treated samples are washed in deionized water and filtered several times to obtain wax-free fibers (Dai et al. 2013; Zimmermann et al. 2010). The fibers may be subsequently submitted to alkali treatment.

4.2 Alkali treatment (mercerization)

This stage is necessary to partially solubilize hemicellulose fraction from fibers in order to expose cellulose crystals for further processing, which enhances crystallinity structure. Lignin and cellulose are barely affected during this stage. Additionally, alkali treatment should remove residual waxes at the pre-treatment stage, as well as pectin, silica ash and natural fats (Jonoobi et al. 2009). The removal of hemicellulose is necessary as decomposition takes place earlier in non-cellulosic components, so if used for reinforcement purposes this stage should not be omitted (Lu and Hsieh 2010). Treatment is generally carried out by using a strong base solution (NaOH or KOH) with varying conditions as presented in Table 2.

Table 2 Mercerization of different lignocellulosic fibers
Full size table

New approaches, such as steam explosion, have shown great results in eliminating lignin fraction while leaving cellulose structure intact and therefore purer. Solubilized components are removed by washing with deionized water and filtered. The conditions employed may vary depending on fiber source and its constitution (Deepa et al. 2011).

4.3 Bleaching (delignification)

Following mercerization, the bleaching process takes place. This treatment is necessary for the removal of components such as residual hemicellulose, but mainly lignin, as this process is also known as delignification. Two treatments are well described in the literature, namely treatment with chlorine compounds under acidic condition (more effective but environmentally harmful), and hydrogen peroxide (H2O2) treatment under alkaline condition (Abraham et al. 2011; Chen et al. 2011a). These and other treatments employed for bleaching are described in Table 3.

Table 3 Bleaching of different lignocellulosic fibers
Full size table

Bleaching is often repeated to provide a higher level of lignin removal; lignin is a very resistant structure turning bleaching necessary for producing a purer cellulose which will be the source of nanocrystals. The removal of components from fibers affects directly their thermal stability and final quality of CNCs because cellulose will be more accessible to acid attack during subsequent treatment for nanocrystals dispersion (Chen et al. 2015; Wang et al. 2014a, b).

4.4 Acid hydrolysis

After obtaining pure cellulose from the previous processing stages, this structure must be further treated to obtain nanocellulose because it still contains amorphous domains that may reduce thermal stability and structural strength. The main technique used is acid hydrolysis, although enzymatic cleavage is also accepted. The use of acids instead of biological hydrolysis presents lower cost and good efficiency for obtaining CNCs, therefore it has been widely used in several studies (Cui et al. 2016; Siqueira et al. 2010d).

It should be noted that the use of different acids may lead to different effects on CNCs. For instance, the sulfate groups of H2SO4 may react with surface hydroxyl groups of cellulose, yielding charged surface esters that promote dispersion of nanocrystals in water. On the other hand, treatment with hydrochloric acid results in reduced dispersion of crystals, leading to their flocculation when in solution (Cudjoe et al. 2017; Roman and Winter 2004).

In addition, recent studies indicate that hydrolysis kinetics through strong acid treatment is greatly dependent on acid concentration where treatment below 58 wt% should lead to cellulose depolymerization and its consequent conversion to soluble sugars thus affecting nanocellulose yield. This work finding should aid for maximizing yield, as it allows to define the desired reaction parameters, as well as to ensure a greater recovery of cellulose solid residue from process which should be of interest for industrial applications when minimizing losses in processing. The same group accomplished CNCs yield of 70% from eucalyptus kraft pulp when defining that, despite rapid depolymerization of cellulose when using mineral acids at high concentrations, CNC properties and reaction may be handled using subtle modifications of process parameter for CNCs formation (Chen et al. 2015; Wang et al. 2014a, b).

Process efficiency is correlated with factors such as the purity of samples because residual non-cellulosic compounds may interfere in hydrolysis, as well as temperature, reaction period and acid concentration where lower temperatures are related to reaction time increase (Wang et al. 2014a, b). The different treatments used for the preparation of nanocrystals are listed in Table 4.

Table 4 Different treatments used to extract nanocrystals from various sources and their dimensions
Full size table

Uncontrolled conditions may lead to either poor cleavage of amorphous domains, resulting in the failure to yield CNCs, or excessive breakage of nanowhiskers, which may reduce their crystallinity (Lu and Hsieh 2010; Roman and Winter 2004). After hydrolysis, the reaction is diluted with an excess of deionized water (usually fivefold to tenfold), with the objective of quenching the reaction. Excess acid is usually removed through a series of centrifugation and washing stages. The precipitate is recovered and is submitted to dialysis until it presents constant pH. Dialysis is also necessary to remove non-reactive sulfate groups, salts and soluble sugars (Henrique et al. 2013; Rosa et al. 2010). Neutralization may substitute this process because an excess of sulfate groups may lead to a decrease in thermal stability, although these same groups also avoids the aggregation of nanowhiskers (Deepa et al. 2011).

4.5 Dispersion of nanocrystals

Homogenization is defined as the final stage after acid hydrolysis; it helps to ensure uniformity of size of obtained CNCs. While optional, this stage may be necessary because it may be difficult to disperse the nanocrystals aggregated from the previous stage into a homogeneous suspension. Of all employed methods, ultra-sonication seems to be the most recommended because it is environmentally friendly, presents high efficiency, low cost, and presents the ability to easily disperse CNCs in suspension (Alemdar and Sain 2008; Lu et al. 2013). Treatment is generally performed under ice bath conditions because otherwise, it would promote desulfation of sulfate groups on CNCs surface (Beck et al. 2010). Increasing the power and sonication length have great influence on CNCs final dimensions and process conditions should be adapted for nanocrystals application (Chen et al. 2011b; Pasquini et al. 2010).

4.6 Drying process

Before being designated as a nanofiller, CNCs may optionally be dried. This procedure allows cost reduction in material transport and also the ease of use in industrial applications. However, drying must be well defined as this process can irreversibly harm the structure of the crystals, stability and also lead to complications for dispersing sample when mishandled (Peng et al. 2012).

As simple oven drying of nanocrystal suspension is known to affects the quality of nanocrystals, as slow water evaporation allows the aggregation of crystals and reduces thermal stability, alternatives have been proposed to overcome these issues and to enable industrial application of nanocrystals (Peng et al. 2012). One of these proposed methods is spray-drying. In spite of this method being low-cost in terms of use and maintenance it may promote re-aggregation due to the high polarity of CNCs, which tend to link through hydrogen bonds. Surface modification is often necessary to overcome this problem when using this technique (Besbes et al. 2011; Peng et al. 2012).

Freeze-drying is the most expensive method presented until now although it shows great potential as it allows fast freezing and water evaporation, without over affecting nanocellulose structure. The surface sulfate groups obtained from sulfuric acid hydrolysis, however, should act to avoid aggregation of nanocellulose to large bundles through irreversible hydrogen bonds formation thus, preventing loss of nanoscale after drying (Peng et al. 2013). However, when this occurs the bonds are not as strong as would happen when using the spray-drying technique, which facilitates redispersion of crystals in solution. The powder easily re-disperses in solvents without any pre-treatment and it is therefore a promising procedure in relation to industrial applications (Peng et al. 2012).

5 Properties

Nanocrystals isolated from lignocellulosic sources show unique properties when applied as nanocomposites. They include the following.

5.1 Low cost and availability

Several studies have attempted to minimize the creation of waste through purification of cellulose from different residues, such as those listed in the previous tables. Although the focus has been on rich cellulosic sources, such as hemp, jute and sisal for instance, some studies have looked at the treatment of unusual sources such as fruit peels, bagasse, husks (hulls), fruit leaves and seeds (Deepa et al. 2011; Kallel et al. 2016; Lamaming et al. 2017; Lu and Hsieh 2012; Lu et al. 2013; Ludueña et al. 2011; Mandal and Chakrabarty 2011; Santos et al. 2013). The wide availability and low cost of all these sources may aid supporting studies for viable fabrication of nanocrystals which will later be discussed in this review.

5.2 Biodegradability

Inherent biodegradability is one of the many properties of nanocrystalline cellulose which can be of use in terms of coating and packaging purposes. Several reports have showed that thermal and mechanical stability of materials may be improved with the aid of nanocomposites in particular biodegradable films (Acharya et al. 2011; Belbekhouche et al. 2011; Swain et al. 2013). For instance, biodegradable films enhanced with nanocrystals present improved properties, such as high tensile strength, elongation potential, thermal and structural stability and oxygen barrier, forming a resistant and yet durable film (Espino-Pérez et al. 2013; Floros et al. 2012). Because biodegradable films often show lower resistance and stability in comparison to synthetic ones, the application of nanocellulose and nanofibers is intended to provide better stability and strength, while also enabling their biodegradation, being defined as green nanocomposites (Mesquita et al. 2012; Minelli et al. 2010; Savadekar and Mhaske 2012).

5.3 Mechanical properties

The interest in use of nanocellulose as a reinforcing agent is exemplified when analyzing Young’s modulus which is a mechanical property designed to estimate the force required to stretch or compress a material. It has been reported that CNCs can present up to 150 GPa for Young’s modulus, which is stronger than steel and similar to Kevlar, although this value may vary depending on the source of cellulose and the type of treatment (Siqueira et al. 2010a). Several studies have reported a great increase in tensile strength when using nanocrystals in comparison with the original fiber state (Sehaqui et al. 2011). This inherent property of CNCs may be used to overcome the mechanical limitations of biodegradable films when they are used as reinforcing agents.

5.4 Thermal resistance

Natural cellulose presents a highly crystalline structure. Acid hydrolysis treatment, which tends to eliminate amorphous components such as residual lignin, hemicellulose, extracts, pectins and imperfect cellulose crystals, leads to a more resistant and stable structure, affecting several properties in CNCs such as their thermal stability. The incorporation of CNCs as nanocomposites should enhance this property because of their inherent characteristic of heat resistance and it also helps to avoid polymeric material decomposition at higher temperatures (Sehaqui et al. 2011; Siqueira et al. 2010a; Sun et al. 2004). However, one should bear in mind that employed treatment may affect this aspect as usual acid hydrolysis with H2SO4 results in formation of surface sulfate groups within cellulose structure which was already reported to drastically reduce its thermal decomposition temperature (Alemdar and Sain 2008; Chen et al. 2016).

6 Future perspectives

In spite of the fact that nanocellulose extraction from wood pulp is well characterized in literature and it generally leads to higher yields of nanocellulose, nevertheless, CNC processing from other sources such as agro-industrial residues should be debated. The reuse and manufacture of high value-added products, a concept extracted from circular economy studies, may become even more important if they show high cellulose content. Commercial process from alternative sources should be based on residues that are rich in cellulose. Cotton, hemp, jute and sisal present higher levels of cellulose than those found in hardwood and softwood and they are very similar to the levels found in corncob, for example (Chen et al. 2015; Morais et al. 2013; Sehaqui et al. 2011; Silvério et al. 2013;).

New approaches should be adopted in order to define less hazardous and environmentally dangerous reagents to replace current acidic treatments, while also providing similar results. An interesting possibility could be the synergic application of shearing forces and hydrolytic reagents for enhancing yield and crystallinity from lignocellulosic biomass, while minimizing the formation of byproducts and reduction of cellulose from simple sugars rather than CNCs. A study has demonstrated that this effect worked using tungstophosphoric acid and sonication simultaneously; this led to a high crystallinity of nanocellulose (88%), while also preventing excessive cellulose degradation. The use of combined techniques with mild conditions represents an environmentally-friendly and cost/time effective approach, while mineral acids not only represent losses in thermal behavior but also being partially responsible for reactors corrosion in industrial scale, which is a constant issue in commercial production (Hamid et al. 2016). Other researchers have also employed cellulose pretreatment with the use of Ionic Liquids (IL), which are defined as salts that are liquid at temperatures lower than 100 °C. These show unique properties such as low vapor pressure, high thermal and electrochemical stability and excellent dissolution performance for most organic and inorganic compounds, including cellulose. These characteristics make IL potential agents for reducing industrial processing of cellulose, while also being environmentally-friendly for its intrinsic properties (Cao et al. 2017; Li et al. 2012).

Another path should be the use of organic acids for obtaining CNCs as harsh treatment employed by strong acids such as sulfuric acid should harm fibers structure to an extent where it is not economically viable if process is not well regulated (Savadekar and Mhaske 2012). Taking aside reduction of thermal stability occurred through H2SO4 treatment of fibrils presence of sulfate groups limit derived crystals application, use of this and others mineral acids tends to be expensive and environmentally harmful as large quantities of this acid (9 kg H2SO4) are necessary for 1 kg of produced CNC, in addition to disposal of residual salt after neutralization (approximately 13 kg Na2SO4 per kg CNC) (Chen et al. 2016). In recent report, it was stated that use of organic acids (maleic, oxalic, ρ-toluene and benzenesulfonic acids) is effective for producing partially hydrolyzed cellulose residue which may be designated for production of nanocellulose after mechanical treatment while also allowing low cost acid recovery after treatment. This study indicated a better performance in comparison to mineral acid treatment as formed crystals presented enhanced dispersion in solution, longer lengths, improved thermal stability and the possibility of recovery of residual cellulose may be recovered from hydrolysis process in order to produce more nanocellulose (Chen et al. 2016).

In addition to organic acids, another alternative should be the use of Lewis acid catalysts for obtaining crystalline nanocellulose. A recent attempt to prepare nanocrystals had focused on ferric chloride (FeCl3) which is a salt already used for esterification, oxidation and condensation reactions. In comparison to conventional generation of nanocrystals out of bamboo pulp, the first study presented an easy approach to obtain short rod-like CNC with length of 100–200 nm and width of 10–20 nm with high crystallinity index (nearly 80%). In addition to similar structure of crystals, the alternative methodology succeeded in preparation of nanocellulose without strong acid use and, consequently, preserving thermal behavior of crystals as this approach avoids hydroxyl surface modification of cellulose (Brito et al. 2012; Lu et al. 2014).

7 Conclusion

Commercial process from agro-industrial waste should benefit from the information presented in this review in order to optimize process where higher extent of acid recovery and generation of nanocellulose variations should be aimed. Modification of conventional hydrolysis should aid to turn it more sustainable and effective thus reducing processing costs and making scale-up possible for agro-industrial residues sources instead of wood pulp.

Having presented the main issues in nanocrystalline cellulose preparation and the recent approaches to overcome them, it is clear that processing from agro-industrial residues should be further discussed. While the choice of a specific source should be dependent on the residue availability and the actual concerns for its conventional disposal/treatment, the generation of nanocellulose may not only result in high value-added products but it may also cooperate through circular economy in the reuse of agro-industrial residues which should represent in economy and reduction of emissions generated from conventional treatment, for example.