Sustainable Approaches to Incorporate Plant-Based Biomaterials in Power Generation

: Biomass-derived materials have traditionally been used to generate electrical energy through the combustion of their organic components. However, within the past few years, certain common biomass compounds, especially plant-based products such as cellulose and lignin, have drawn attention in the energy ﬁeld due to their wide availability, low cost, and chemical versatility. In the case of cellulose, the combination of crystalline and amorphous domains, along with the high surface area and abundance of hydroxyl groups, has allowed for its application in multiple devices to harvest energy from the environment. However, to date, there are no reviews focusing on the different approaches that have been developed to implement these sustainable materials in the generation of renewable energies and the desirable material properties for these applications. This manuscript reviews alternative ways that have been developed to exploit biomass compounds in power generation, especially cellulose and lignin. Three different types of energy harvesting are discussed: mechanical, osmotic, and thermal energy. In the case of mechanical energy, the application of plant-derived materials in piezoelectric and triboelectric generators is described. In both cases, approaches where the biomass material has an active role in power generation instead of acting as a mechanical support are reported. For osmotic energy, the performance of inverse electrodialysis systems and the use of plant-derived materials, including the chemical modiﬁcations carried out to allow for their use for energy generation, was reviewed. Finally, for thermal energy generation, the reported work on biopolymer-based devices that work using thermoelectricity has been summarised. In each case, the latest advances in the ﬁeld from the materials science perspective and the reported performance were described. Hybrid approaches involving the combination of biomass materials with other components have also been considered and compared with the performance obtained using biopolymers alone. Current limitations and opportunities are, ﬁnally, discussed to offer an overview of the current landscape and indicate future directions of the ﬁeld.


Introduction
Energy generation from biomass represents a promising alternative to the use of fossil fuels and it will play a pivotal role in achieving the net-zero carbon target by 2050. Biomass refers to the organic material that is obtained from living organisms. The most common route for exploiting the energy contained in biomass is through the combustion of its components. This type of energy is also commonly referred to as "bioenergy". Bioenergy is considered carbon neutral, since its use does not, in principle, lead to the accumulation of net carbon in the atmosphere, since biomass is generated through the fixation of CO 2 by plants. Moreover, the natural decomposition of agricultural or forestry residues releases CO 2 , even if they are not used for energy generation [1]. Consequently, this energy source has the potential to reduce the use of underground fossil fuels. Although within the past few years new models and policies have been developed to take into account the

Materials and Methods
Natural cellulose can be found as a mixture of its amorphous and crystalline states. To date, six forms of crystalline cellulose have been described, namely cellulose I, II, IIII, IIIII, IVI, and IVII [36]. The crystallinity of cellulose and the high presence of hydroxyl groups provides cellulose with a strong electron-donating ability and high number of dipoles, which have been exploited for energy harvesting through tribo-and piezoelectricity [19]. The most common form of natural cellulose is cellulose I. Cellulose II, also called regenerated cellulose, is formed upon modification of the structure of cellulose I using different solvents, including basic solutions containing NaOH [37]. In plants, cellulose is predominantly found as cellulose Iβ, whilst cellulose Iα is obtained from less complex organisms such as bacteria. Both crystalline cellulose structures present a parallel chain arrangement but differ in their lattice systems. Cellulose Iα is a one-chain triclinic system, while Iβ presents a two-chain monoclinic structure [36].
The highly hydroxylated composition of cellulose allows for its easy modification and functionalisation. Surface modification reactions such as oxidation [38,39] or acetylation [40], and surface functionalisation using click chemistry [41], among others, have been carried out in cellulose polymers to improve their chemical properties towards specific applications. These modifications can be performed using green chemistry routes, while preserving its sustainability and biodegradability. As such, there has been a great interest in the incorporation of cellulose and its derivatives into multiple applications such as

Materials and Methods
Natural cellulose can be found as a mixture of its amorphous and crystalline states. To date, six forms of crystalline cellulose have been described, namely cellulose I, II, III I , III II , IV I , and IV II [36]. The crystallinity of cellulose and the high presence of hydroxyl groups provides cellulose with a strong electron-donating ability and high number of dipoles, which have been exploited for energy harvesting through tribo-and piezoelectricity [19]. The most common form of natural cellulose is cellulose I. Cellulose II, also called regenerated cellulose, is formed upon modification of the structure of cellulose I using different solvents, including basic solutions containing NaOH [37]. In plants, cellulose is predominantly found as cellulose Iβ, whilst cellulose Iα is obtained from less complex organisms such as bacteria. Both crystalline cellulose structures present a parallel chain arrangement but differ in their lattice systems. Cellulose Iα is a one-chain triclinic system, while Iβ presents a two-chain monoclinic structure [36].
The highly hydroxylated composition of cellulose allows for its easy modification and functionalisation. Surface modification reactions such as oxidation [38,39] or acetylation [40], and surface functionalisation using click chemistry [41], among others, have been carried out in cellulose polymers to improve their chemical properties towards specific applications. These modifications can be performed using green chemistry routes, while preserving its sustainability and biodegradability. As such, there has been a great interest in the incorporation of cellulose and its derivatives into multiple applications such as structural materials [42][43][44], sensing [45][46][47], or smart textiles [48,49], among others. In particular, its use in energy harvesting systems has attracted rising interest in the last few years, given the high versatility of cellulose, which can be easily functionalised and chemically modified to tailor its properties. As such, it represents a renewable and low-cost alternative to current materials employed in energy harvesting.
An alternative material derived from plant biomass is lignin. Lignin is an amorphous macromolecule, which shows a different composition depending on the plant source [50]. In general terms, lignin is formed by three different phenylpropane monomers: p-coumaryl alcohol, coniferyl alcohol, and sinapyl alcohol, with variable ratios. However, despite being the second most widespread biopolymer in nature, the applications of lignin in power generation are limited. This fact can be attributed to its poor electrical performance and lower processability compared to cellulose, given the insolubility in inert solvents and chemical stability of lignin [51]. Consequently, the market size for lignin is considerably lower compared to cellulose, being $730 million in 2014 [52], and it is expected to reach $1.6 billion by 2025 [53]. In comparison, the market size for cellulose was 1.72 billion in 2014 and it is expected to reach $7.4 billion by 2025 only within the Asia-Pacific region [54,55], showing a higher volume and growth. However, lignin shows good mechanical resilience and transparency, which could be exploited within biomedical applications such as electronic skins [56] and structural [57] applications.
This review discusses the current landscape of biomass-derived materials, especially cellulose and lignin, in renewable energy generation. Three major areas are discussed: mechanical, osmotic, and thermal energy. Specifically, we focused on approaches where these biomass materials can be used as an active component instead of a mechanical support. In each case, the properties of the biomass materials that are exploited to generate electricity from a fundamental chemical level and how they relate with the observed performance are described. Hybrid materials where multiple chemicals and nanostructured compounds have been combined are also described and their performance is detailed. The review of these materials and their applications allowed for an objective comparison between the performance of different energy sources that could be used to identify potential areas for improvement and opportunities in this field.

Piezoelectric Generators
Mechanical energy is one of the most ubiquitously available forms of energy in the environment. As such, it represents a large source of renewable power. The piezoelectric effect was first described in quartz crystals by the Curie brothers in 1880. Since then, multiple materials, such as barium titanate, tourmaline, and Rochelle, have been discovered and tested [58]. Piezoelectric generators can produce a change in the electrical polarisation of the active materials as a consequence of an induced mechanical stress. This phenomenon is called the direct piezoelectric effect (Figure 2a). Piezoelectric materials can also deform when they are subjected to an electrical field, defined as the reversed piezoelectric effect (Figure 2b). The piezoelectric effect takes place within materials with no centre of symmetry on their crystal structure. In these materials, a mechanical strain causes an asymmetric shift on the ionic charges in the crystal structure, which generates a voltage. To generate piezoelectricity, the active material must be enclosed between two separate electrodes and, under the application of a mechanical input such as a vibration or pressure, the device generates electricity due to the formation of a voltage difference between the electrodes, allowing for a harvesting of mechanical energy (Figure 2c). Cellulose nanocrystals can produce piezoelectricity due to their particular structure, presenting a noncentrosymmetry, with two different hydrogen bonding networks [60,61]. This effect was first reported in 1954 by Fukada [62], who observed the piezoelectric effects on the annual rings of wood. Since then, multiple devices have been designed to exploit this effect using cellulose-based materials. This material has been employed either as a piezoelectric generator or in the form of a nanocomposite, with nanoparticles embedded.
The power yield of the devices based on natural cellulose alone tends to be low compared to commonly used organic materials in piezoelectric generators such as polylactic acid [63] or poly(vinylidene difluoride) (PVDF) [64]. In the case of PVDF, a power output of 112.8 µW has been reported by Song et al. [65], being similar to current ceramic-based harvesters. However, the poor yield of cellulose in comparison is reflected by a low piezoelectric coefficient, which is related to the strength of the piezoelectric effect on the material and indicates the polarisation on the electrodes obtained upon subjecting the material to a mechanical stress. The use of vertically aligned cellulose nanocrystals has been demonstrated to enhance this piezoelectricity generation [66]. The increase in energy generation is a consequence of the large dipole moment of cellulose within the cellulose chain Cellulose nanocrystals can produce piezoelectricity due to their particular structure, presenting a noncentrosymmetry, with two different hydrogen bonding networks [60,61]. This effect was first reported in 1954 by Fukada [62], who observed the piezoelectric effects on the annual rings of wood. Since then, multiple devices have been designed to exploit this effect using cellulose-based materials. This material has been employed either as a piezoelectric generator or in the form of a nanocomposite, with nanoparticles embedded.
The power yield of the devices based on natural cellulose alone tends to be low compared to commonly used organic materials in piezoelectric generators such as polylactic acid [63] or poly(vinylidene difluoride) (PVDF) [64]. In the case of PVDF, a power output of 112.8 µW has been reported by Song et al. [65], being similar to current ceramic-based harvesters. However, the poor yield of cellulose in comparison is reflected by a low piezoelectric coefficient, which is related to the strength of the piezoelectric effect on the material and indicates the polarisation on the electrodes obtained upon subjecting the material to a mechanical stress. The use of vertically aligned cellulose nanocrystals has been demonstrated to enhance this piezoelectricity generation [66]. The increase in energy generation is a consequence of the large dipole moment of cellulose within the cellulose chain direction [66]. Vertically aligned CNC films achieved a piezoelectric coefficient of 19.3 ± 2.9 pC/N, which is similar to the one observed in PVDF, in the range of 20-30 pC/N [67]. This piezoelectric coefficient was 50 times higher than the longitudinal piezoelectric coefficient observed in wood cellulose fibres, which has been reported to be 0.4 pC/N [68]. However, the formation of vertically aligned cellulose nanocrystals required the use of a DC voltage of 5 kV, reducing the scalability of the fabrication process industrially.
The piezoelectric coefficient of cellulose can be further increased to up to 210 pm/V when used in the form of ultrathin films of aligned nanocrystals [69]. This value is similar to the one obtained by piezoelectric metal oxides and can be obtained by fabricating the films using a uniform electric field onto a mica substrate. Despite the high piezoelectric coefficient achievable by cellulose nanocrystals, especially by aligned nanocrystals, the expensive fabrication and processing involved make it an impracticable technology within a commercial setup. As such, alternative materials that require a lower degree of chemical processing are desired to reduce the production costs and improve the eco-friendliness of the devices.
Unmodified natural cellulose can be used as a power generator, taking advantage of the naturally occurring nanocrystals in plant tissues. As an example, fresh fruits such as pomelo can be used as the piezoelectric material, with a power yield of 12 µW cm −2 [70], which could represent a promising approach for the generation of energy from biowastes. However, the power conversion of this approach was relatively low when compared to the current commercially available devices. In the case of natural wood, the high presence of lignin and its low elastic compressibility lead to a poor piezoelectric output. To circumnavigate these challenges, Sun et al. [71] tested natural wood that had been partially digested using G. applanatum. These fungi could partially remove the lignin and hemicellulose molecules, and altered the structure of cellulose, improving its compressibility (Figure 3a,b). As a consequence, the electrical output of wood increased by 55 times compared to the natural untreated version.
Some attempts have been made to improve the energy generation of cellulose by mixing it with other polymers and organosilicon compounds. Ram et al. [72] developed a piezoelectric generator based on 5 wt% cellulose nanocrystals embedded inside a nylon 11 matrix and using glycerol as a plasticiser to increase the flexibility of the system, which could be easily folded (Figure 3c,d). The final device showed a power output of 500 µW cm −3 when subjected to mechanical impacts. The power yield of polymer-mixed devices could be further improved by the incorporation of polydimethylsiloxane (PDMS) with cellulose nanofibrils. Zheng et al. [59] reported a piezoelectric generator using porous cellulose nanofibrils and coated with PDMS, generating a power density of 6.3 mW/cm 3 ( Figure 3e). However, the synthesis of some piezoelectric polymers for the fabrication of these mixed approaches requires the use of precursors from nonrenewable sources, especially in the case of petrochemical polymers [73], and it tends to be energy intensive [74]. Although some efforts have been made within the last few years to improve the sustainable production of certain polymers, including PVDF [75] and nylon 11 [76,77], these approaches are not being used industrially yet.
One of the most common approaches in the development of cellulose-based piezoelectric generators is the use of nanocomposites, where one or more nanomaterials are incorporated. Cellulose nanofibres are typically employed in this area, given their high surface area and good mechanical compressibility (Figure 3f,g). This approach enables a fast and low-cost way to increase the piezoelectric performance of cellulose-based compounds. Materials such as multiwalled carbon nanotubes (CNTs) or BaTiO 3 have successfully been used in combination with cellulose to produce high-yield piezoelectric generators. Alam et al. [78] developed a nanocomposite-based piezoelectric generator based on cellulose microfibres and PDMS, incorporating MWCNTbs as the conductive filler ( Figure 3h). The final device showed a power density of 9.0 µW/cm 3 , enough to power up 22 LEDs and an LCD screen. In addition, BaTiO 3 has been incorporated, generating in the form of nanocomposites [79], reaching a power density as high as 8.41 µW/cm 3 , reported by Choi et al. [79], using nanocellulose and 40 wt% BaTiO 3 (Figure 3i,j).
Solids 2023, 4, FOR PEER REVIEW 7 nanocomposites [79], reaching a power density as high as 8.41 µW/cm 3 , reported by Choi et al. [79], using nanocellulose and 40 wt% BaTiO3 (Figure 3i,j). The power output of piezoelectric generators can be further increased by combination with triboelectricity, enabling a conversion of mechanical energy into electricity. This concept was applied by Shi et al. [80], who developed a cellulose/BaTiO3 aerogel, increasing the power of the final device from 11.8 µW, when only piezoelectricity was harvested, to up to 85 µW. Thus, the harvesting of mechanical energy through piezoelectricity represents a promising alternative for the powering of devices. However, current piezoelectric generators are limited due to their relatively low power generation and the necessity for a mechanical pressure. As such, new approaches have been developed to enhance this generation of power. A summary of recent approaches in power generation using biomass-derived materials is shown in Table 1. The power output of piezoelectric generators can be further increased by combination with triboelectricity, enabling a conversion of mechanical energy into electricity. This concept was applied by Shi et al. [80], who developed a cellulose/BaTiO 3 aerogel, increasing the power of the final device from 11.8 µW, when only piezoelectricity was harvested, to up to 85 µW. Thus, the harvesting of mechanical energy through piezoelectricity represents a promising alternative for the powering of devices. However, current piezoelectric generators are limited due to their relatively low power generation and the necessity for a mechanical pressure. As such, new approaches have been developed to enhance this generation of power. A summary of recent approaches in power generation using biomass-derived materials is shown in Table 1.

Triboelectric Nanogenerators
The concept of triboelectric generators was first reported by Prof. Zhonglin Wang in 2012 [85]. These devices can transform mechanical energy into electricity, even in small amounts of movement. The discovery of this form of energy opened up a new field in renewable energy harvesting and within 12 months of their discovery, the power output of the devices had been improved by 5 orders of magnitude [86].
Power generation through triboelectricity takes place by an electron exchange from an electron-donating material to a charge-trapping layer. As such, this method can combine the energy generated through contact electrification and electrostatic induction and is dependent on the total surface area [87]. When two materials with different surface potentials are put into contact, there is an electron movement that generates an electrical output (Figure 4a) [88]. Common strategies for increasing the power output of triboelectric generators focus on increasing the charges on the active materials or the surface contact between the electrodes. Materials such as modified fluorinated ethylene propylene through ionised gas injection [89] and plasma-treated PDMS [90] have been employed with high output yields. However, these chemical treatments are energy intensive and their scalability for industrial purposes is low.
The incorporation of cellulose as an active material in energy generation through piezoelectricity or triboelectric nanogenerators represents a promising approach to sustainable power generators. As detailed in the previous section, cellulose presents a piezoelectric behaviour due to its crystalline structure. In addition, among all the available biomass-derived compounds, cellulose represents the most widespread material applied to triboelectric generators. Cellulose materials are positively charged in triboelectric terms due to the high amount of oxygen atoms within their molecular structure [91]. This chemical composition allows for the loss of electrons without requiring a high energy. This property of pure cellulose is reflected in a high frictional contact charge transfer, in the range of −130 µC m −2 , being higher than the average value for common polymers [19].
The first cellulose-based triboelectric generator was reported by Yao et al. in 2016. The cellulose was first oxidised with tetramethylpiperidine-1-oxy and paired with fluorinated ethylene propylene as the negative triboelectric layer (Figure 4b,c) [92]. The differences in the polarity between both the treated cellulose and the fluorinated ethylene propylene allowed for its use without the need for any additional material. In addition, an energy harvesting device combining the tribo-and piezoelectric capabilities of cellulose has been reported, with a power output as high as 10.6 µW·cm −2 for the triboelectric generation and 1.21 µW·cm −2 for the piezoelectric side [82]. This piezo-and triboelectric hybrid system could also be used as a pressure sensor with a low detection limit of 0.2 N cm −2 . This device combined a wood-derived nitrocellulose by modification using nitric acid and a piezoelectric nanocomposite containing BaTiO 3 /MWCNTbs with bacterial cellulose. However, the power output of power generators can be further improved by focusing on strategies to increase the triboelectric output.
The performance of triboelectric generators is directly proportional to the charge density of the active surfaces and the surface area [93]. As such, one of the main approaches in the development of these devices is the incorporation of different nanoparticles that can enhance the surface area in a cellulose matrix, making a nanocomposite. Materials such as ZnO have been used in the form of biocompatible triboelectric generators, showing a power density of 42 mW m −2 (Figure 4d,e) [94]. In addition, BaTiO 3 particles have been incorporated into cellulose nanofibres to enhance the surface roughness and dielectric permittivity of the triboelectric generators, increasing the power density by 2 orders of magnitude, with a reported value of 4.8 W m −2 [95]. The performance of cellulosic materials can be further increased through the functionalisation of the biopolymer chains with electron-donating groups such as aminated siloxanes [96]. In the case of cellulose aerogels, Zhang et al. [97] reported a method for the synthesis of this material that led to a surface area of 221 m 2 g −1 and a maximum power density of 127 mW m −2 when used as a triboelectric generator using polytetrafluoroethylene (PTFE) as the back electrode (Figure 4f,g) [95].
Another biomass-derived material that has shown potential for its incorporation into triboelectric generators is lignin. Despite its poor electrical conductivity, the electron-transfer capabilities of lignin could be exploited by placing this material in contact with chargetrapping films such as Kapton [98]. Given the high stiffness and insolubility of lignin, this material had to be combined with starch, and led to a power density of 173.5 nW cm −2 . This value resulted in a lower value than the ones observed in cellulose, which explains the scarce amount of approaches exploring lignin as an energy material within the literature.
Although the applications of biomass-derived materials, especially cellulose, are well established within the literature, the overexploitation of this resource could enhance deforestation rates [99]. Consequently, the market for alternative sources for cellulose generation, such as bacterial cellulose, have experienced a high growth within the last few years [100]. However, there is still great interest in the exploitation of whole plants for energy harvesting, which could improve the sustainability of the devices.
As mentioned, power generation through triboelectricity is dependent on the surface area of the employed electrodes [87]. As such, to achieve a good performance, the materials have to be processed through lithography [101] or electrospinning [102]. These methods increase the production costs, limiting their commercialisation. Some plant organs, such as leaves and petals, present a large surface area due to their structure, with a hierarchical porosity and high roughness. In the case of leaves, the presence of cuticle and stoma cells can increase the surface area by up to 170 m 2 g −1 for tobacco plants [103]. This surface area value is similar to and even larger than other engineered materials already incorporated as triboelectric generators such as cellulose hydrogels [97,104,105]. As such, plant leaves have been employed as templates for the fabrication of high surface area electrodes for the generation of triboelectricity using PDMS [106]. PDMS elastomer can be used as a negative friction layer in triboelectric generators and it can be patterned using laser ablation to enhance the surface area. Using this method, Lee et al. [107] developed a triboelectric nanogenerator able to produce 7.69 W m −2 . Moreover, by using micropatterned PDMS with incorporated silver nanowires and in contact with human skin, Sun et al. [106] reported a power conversion of 56 V and 3.1 µA (Figure 4h,i). This method could lead to a performance almost 60 times higher than similar devices fabricated using laser ablation [108], saving costs and reducing complexity in the manufacturing.
incorporated as triboelectric generators such as cellulose hydrogels [97,104,105]. As such, plant leaves have been employed as templates for the fabrication of high surface area electrodes for the generation of triboelectricity using PDMS [106]. PDMS elastomer can be used as a negative friction layer in triboelectric generators and it can be patterned using laser ablation to enhance the surface area. Using this method, Lee et al. [107] developed a triboelectric nanogenerator able to produce 7.69 W m −2 . Moreover, by using micropatterned PDMS with incorporated silver nanowires and in contact with human skin, Sun et al. [106] reported a power conversion of 56 V and 3.1 µA (Figure 4h,i). This method could lead to a performance almost 60 times higher than similar devices fabricated using laser ablation [108], saving costs and reducing complexity in the manufacturing.  Unmodified plant leaves have a low electron affinity, which can also be exploited to generate a current when it is placed into contact with materials with higher electron affinities such as poly-methyl methacrylate (PMMA). Jie et al. [110] compared the performance of different leaves from multiple tree species and obtained a maximum output power of 45 mW m −2 in the case of Hosta leaves. This power output resulted in higher yields than in some of the nanocomposite-based approaches reported. The energy yield of these plants could be greatly improved by the modification of leaf powder with poly-L-lysine (PLL) [111]. By applying this method, a maximum of 17.9 mW was obtained, which was enough to power a group of 868 LEDs. Thus, the use of whole plants in energy generation has been proven to be a promising alternative for the development of triboelectric generators with a high yield. However, the performance of this approach still cannot meet the high power output of cellulose nanofibres and commercially available devices. A summary of the reported values for cellulose-based triboelectric generators is provided in Table 2.

Osmotic Energy Harvesting with Cellulose
Whilst the generation of electricity through mechanical energy has been shown to be an efficient way to harvest electrical energy, especially for the powering of wearable devices through human motion, this approach cannot meet the current energetic needs that would enable a reduction in non-renewable energy sources. Thus, alternative renewable forms of energy are required. In recent years, osmotic energy has been developed as an alternative form of power generation. This type of energy is collected from the differences in concentration between two electrolyte solutions. Given the high abundancy of naturally occurring mixtures of fresh and salt water on Earth, especially at the estuaries, it has been estimated that, potentially, up to 2 trillion watts from the environment could be harvested using this technology [114].
The most common method used to generate electrical power through salinity changes is reverse electrodialysis. The first demonstration of this type of energy was reported by Pattle in 1955, using a stack of 94 polyethylene and polystyrene membranes, which generated an external power of 15 mW [115]. Since then, the power output of reverse electrodialysis has greatly improved, with up to 67 W m −2 achieved by a single membrane by employing nanoporous and atomically thin carbon-based membranes [116]. Typical electrodialysis systems employ a stack of alternating cation and anion-selective membranes (Figure 5a). As such, the performance of the devices is directly proportional to the properties of the ion-exchange membranes [117]. A highly concentrated salt solution is then used in contact with these ion membranes, separated from a low-concentrated water solution in contact with the membranes as well. The differences in potential between the low-concentrated solutions due to the diffusion of cations and anions, which is driven by the electrochemical gradient with the salt-concentrated solution, generate a voltage [118]. A first pilot plant ex-ploiting this technology has already been implemented in the Netherlands by the REDStack company, being able to generate several watts per square metre of membrane [119].
Recent discoveries in the field of ion-exchange membranes have allowed for the development of high-performance reverse osmotic systems. However, the further improvement in reverse electrodialysis technologies relies on the discovery of affordable and sustainable materials that can be employed for energy generation. One of the main limitations of this technology, that hinders its full incorporation at an industrial scale, has been the relatively low power efficiency compared to other renewable energy sources. This power efficiency is directly related to the selectivity [114], the charge density, and the conductivity [120] of the ionic-exchange membranes. To improve the power conversion of current osmotic systems, nanoporous materials, such as MXenes [121], boron nitride [122], or graphene [123], have been employed. The porosity and permeation paths of these nanomaterials are located at the sub-nanometre range, allowing for a good selectivity on the cations [124]. However, the diffusion speed of water molecules and ions within such confined compartments is reduced, compromising the performance of the system [119]. In addition, the manufacturing costs of these materials is high, reducing the possibility of scale up.
The incorporation of cellulose-based materials has been proven to be a promising alternative for the fabrication of porous membranes for osmotic energy generation. Specifically, the use of cellulose nanofibrils has demonstrated a high yield performance compared to traditional systems due to its high porosity [125] and the presence of highly polar hydroxyl groups at the surface of cellulose. These properties provide cellulose with a high ion conductivity, which can be further enhanced through chemical functionalisation, such as oxidation [126]. Consequently, cellulose has been employed as a sustainable alternative for ion-exchange membranes in fuel cell applications, among others [127].
Cellulose can be used either in its pure form, after chemical modification [128], or in the form of hybrid materials, with the incorporation of other nanostructured compounds. In the case of hybrid materials, the combination of graphene oxide nanoplatelets and cellulose nanofibrils in the form of assembled layers has been shown to improve the performance of current osmotic energy systems by enlarging the nanochannels while showing a large space charge [124]. These properties can decrease the transfer energy of cations and maintain a good selectivity. Wu et al. [124] implemented this principle for the elaboration of a RED system that could generate up to 4.19 mW m −2 using a graphene oxide nanocomposite, generating a layered structure (Figure 5b). However, although in recent years, the feasible synthesis of graphene oxide using environmentally friendly methods has been reported [129][130][131], the main approach used commercially due to cost-effectiveness is based on modified Hummers methods [132][133][134]. These methods involve an exfoliation step, where strong oxidants and concentrated sulphuric acid are used. Thus, to achieve a true renewable and environmentally free generation of electricity using this technology, green materials are needed. Cellulose can also be used as a modifier for standard ion-exchange membranes, increasing the charge density of the films [135]. In this case, cellulose acetate was used as a modifier of the standard PVC-based cation membrane. The incorporation of this cellulose modification could reduce the electrical resistance of the films, from 6.77 up to 3.10 Ω cm 2 whilst increasing the charge density by 3 mEq g −1 . However, the efficiency of this modification has not been tested yet within a practical device.
The use of cellulose-based nanoporous materials alone has been explored by Wu et al. [136], who reported an efficiency of 0.23 W m −2 . To achieve this efficiency, the cellulose nanofibrils had to be functionalised to acquire the negative and positive charges needed for the separation of cations and anions, respectively (Figure 5c,d). In the case of positively charged films, the membranes were modified with quaternary amine groups using etherification. On the contrary, to charge the membranes negatively, carboxyl groups were introduced through 2,2,6,6tetramethylpiperidine-1-oxyl (TEMPO) oxidation. This approach led to an output power density of 0.23 W m −2 , showing promise for an industrially scalable and low-cost device.
One alternative to replace the nanoporous materials is to take advantage of the naturally occurring porosity in wood and cellulose nanofibres. Wu et al. [117] employed a similar strategy to negatively and positively charge wood sheets using TEMPO and etherification, respectively (Figure 5e,f). However, in this case, the power output of the system was considerably lower than the one reported using cellulose nanofibrils alone, with 5.14 mW m −2 [117].
The use of cellulose-based nanoporous materials alone has been explored by Wu et al. [136], who reported an efficiency of 0.23 W m −2 . To achieve this efficiency, the cellulose nanofibrils had to be functionalised to acquire the negative and positive charges needed for the separation of cations and anions, respectively (Figure 5c,d). In the case of positively charged films, the membranes were modified with quaternary amine groups using etherification. On the contrary, to charge the membranes negatively, carboxyl groups were introduced through 2,2,6,6-tetramethylpiperidine-1-oxyl (TEMPO) oxidation. This approach led to an output power density of 0.23 W m −2 , showing promise for an industrially scalable and low-cost device.
One alternative to replace the nanoporous materials is to take advantage of the naturally occurring porosity in wood and cellulose nanofibres. Wu et al. [117] employed a similar strategy to negatively and positively charge wood sheets using TEMPO and etherification, respectively (Figure 5e,f). However, in this case, the power output of the system was considerably lower than the one reported using cellulose nanofibrils alone, with 5.14 mW m −2 [117]. Despite the low power output achieved in the case of lignin-based devices, which could be a consequence of the presence of less ionically conductive materials within wood such as lignin and hemicellulose, the price of each ion-exchange membrane was in the Despite the low power output achieved in the case of lignin-based devices, which could be a consequence of the presence of less ionically conductive materials within wood such as lignin and hemicellulose, the price of each ion-exchange membrane was in the range of $10, which is considerably lower than the commercially available ones (about $350). However, the performance was significantly lower than the one obtained by the cellulose-based approaches. The reported values for lignin-and cellulose-based osmotic energy generators are shown on Table 3. Table 3.

Thermoelectric Energy Harvesting with Cellulose
Thermoelectricity can make use of temperature gradients to generate an electric voltage. This technology is based on the Seebeck effect and has been proven to be a useful tool in the development of self-powered electronics, which can work with the residual heat from the human body [138,139]. In addition, thermoelectricity can be employed to recover energy from waste heat sources such as internal combustion engines [140] or electrical ovens [141], among others.
The energy conversion of thermoelectric materials is indicated by the Seebeck coefficient, which is defined as the ratio between the electromotive force and the temperature difference. However, the performance of thermoelectric materials is commonly described in terms of the "Figure of merit", which includes the contributions from the electrical conductivity of the device, the thermal conductivity, and the Seebeck coefficient (Equation (1)), and it is related to the energy conversion efficiency of the device.
where σ represents the electrical conductivity, S is the Seebeck coefficient, k is the thermal conductivity, and T is the temperature. From this equation, it can be deduced that materials with a high electrical conductivity and Seebeck coefficient, and low thermal conductivity, are desired. However, since these variables are strongly correlated, a single improvement in one of these is usually compensated for by the rest of the parameters, making it difficult to enhance the value of zT [142]. One of the most common elements found in thermoelectric devices is tellurium. The most widely employed materials in conventional devices are telluride-based semiconductor materials such as Bi 2 Te 3 and PdTe.
Recently, Bi 2 Te 3 thin films were used to develop flexible thermoelectric devices, achieving a power density of 2.1 mW cm −2 and a power factor as high as~200 µW m −1 K −2 [143]. These materials present a zT of about 1 and are commonly used commercially given their good thermoelectric properties. As such, it can be incorporated inside bacterial cellulose nanofibres, showing a high Seebeck coefficient, in the range of 135 µV K −1 , and a power conversion of 25.5 µW m −1 K −2 [144]. In addition, cellulose nanofibre/Bi 2 Te 3 nanocomposites have been applied to wearable devices, increasing the performance by doping the Bi 2 Te 3 with selenium and antimony [145]. However, Te is a relatively scarce element, which has propelled research into Te-free devices [146]. In addition, the synthesis of current thermoelectric materials, especially Bi 2 Te 3 , is carried out by a hydrothermal reaction [147][148][149], where the precursors need to be subjected to high temperatures and pressure, and requires toxic-reducing agents such as NaBH 4 and N 2 H 4 . Thus, the development of alternative materials with thermoelectric properties that can be synthetised using green and eco-friendly processes represents a crucial aspect in the research of thermoelectric materials. Plant-derived biomass materials such as cellulose or lignin do not show any thermoelectric properties under standard conditions. However, cellulose chains can be oxidised to improve ionic conductivity and ions can be infiltrated [150]. This approach can generate a voltage due to the movement of ions inside the cellulose fibres when they are subjected to a temperature gradient. Li et al. [150] demonstrated this concept using natural cellulose that had been treated to remove hemicellulose and lignin, making cellulose II membranes. This cellulose was oxidised using 2,2,6,6-tetramethylpiperidine-1-oxyl to increase the negative charge density of the fibres, and Na + ions were infiltrated using NaOH. This device showed a thermal gradient ratio of 24 mV K -1 and a power factor of 1150 µW m −1 K -2 . However, the most common approach in the development of cellulose-based thermoelectric materials is their use as a matrix for the fabrication of nanocomposites.
One of the most common materials to provide thermoelectricity to cellulose-based nanocomposites are CNTs. This material has been reported to have a high Seebeck coefficient, which is dependent on its structure. Up to 160 µV K −1 at room temperature have been measured for single-walled CNTs and 80 µV K −1 in the case of multiwalled CNTs [151]. Thus, it shows promise for the development of highly efficient and flexible devices. CNTs can be incorporated into cellulose aerogels, either in the form of single-or multiwalled CNTs. As expected, both the Seebeck coefficient and power factor resulted in higher yields in the case of the single-walled CNTs, with 1.1 µW m −1 K −2 and 47.2 µV K −1 , respectively (Figure 6a) [152]. The fabrication process of cellulose-based thermoelectric generators was further simplified by Abol-Fotouh et al. [153], who used a solution containing 0-12 wt% of single-walled CNTs in direct contact with cellulose-making bacteria (K. xylinus), leading to a higher power factor compared to the previous approach, with 20 µW m −1 K −2 , and a similar Seebeck coefficient (30 µV K −1 ). However, the morphology obtained through this approach did not show any CNT bundle which led to a difference in the performance (Figure 6b,c). The properties of these nanocomposites can also be tailored by using heteroatom-doped CNTs [154]. Such doping can lead to negative Seebeck coefficients in the range of −20 µV K −1 when nitrogen was used as a dopant and positive coefficients (~25 µV K −1 ) when boron was used. Additional plant biomass-derived materials such as lignin can also be used in this field. Culebras et al. [155] developed CNT yarns doped with lignin produced from lignocellulosic waste. The use of lignin increased the Seebeck coefficient from~50 µV K −1 up to 98.9 µV K −1 and the power factor from~20 µW m −1 K −2 up to 132.2 µW m −1 K −2 . Flexible thermoelectric generators have additionally been designed by using natural cellulose, extracted from flax, and combining it with graphene and Bi 2 Te 3 . The final material presented a layered structure that maximised the contact area ( Figure 6d). However, the power factor of this material was relatively low when compared to the cellulose-graphene and cellulose-Bi 2 Te 3 nanocomposites alone, showing a power factor of 6.4 × 10 −2 µW m K −2 and a Seebeck coefficient in the range of 20-30 µV K −1 at 52 • C [156]. Thus, the use of biopolymer-based materials represents an emerging material in the field of thermoelectric generators. The performance of the different reported thermoelectric materials based on biomass-derived materials is detailed on Table 4.  Thus, the use of biopolymer-based materials represents an emerging material in the field of thermoelectric generators. The performance of the different reported thermoelectric materials based on biomass-derived materials is detailed on Table 4.

Current Limitations and Opportunities
As observed, despite the high performance of cellulose nanofibres within most of the energy generation applications, they need to be extracted from natural sources. As such, one of the main limitations of biomass-derived materials is the need for processing prior to their use. This is a consequence of the presence of undesired materials on the natural sources such as lignin and hemicellulose and, in some cases, the higher surface area obtained when using cellulose nanofibres. Thus, within all the reported approaches, the utilisation of pure cellulose outperforms unmodified lignocellulose. An opportunity in this field would be the development of simpler and more sustainable chemical routes to purify cellulose from its natural source and chemically modify it with desired functional groups. In addition, the genetic modification of existing plant resources to produce varieties that produce less lignin/hemicellulose could represent an alternative to increase the power output whilst further reducing the fabrication costs.
Despite all the current efforts to use biomass-derived materials as substituents for traditional compounds in power generation, their performance is still considerably lower than some of the recently reported devices. This fact is more prominent within thermoelectric systems, given the low number of reported approaches employing cellulose or lignin. Consequently, the use of hybrid approaches involving cellulose and different nanomaterials still shows the highest performance compared to cellulose alone. As such, another potential opportunity in the improvement in energy generation using biomass-derived materials would be the synthesis of more sustainable fillings that could be used to improve the performance of cellulose. This includes the development of chemical routes to efficiently transform biomass materials into high-performance nanomaterials such as carbon nanotubes and graphene.
Future developments in the field of power generation using biomass-derived materials will focus on the development of high throughput cellulose sources that require minimal processing to achieve a good performance. An example of a potential source is bacterial cellulose which, as observed, presents good physical-chemical properties given the high presence of nanofibrils that enhance the surface area of materials. However, this method for cellulose generation is slow and requires a high energy to maintain bacterial colonies.
In addition to the chemical modification of cellulose biopolymers, the advantage of using vertically aligned nanocrystals has been described. This material can be used to achieve a high piezoelectric coefficient, leading to an enhanced power generation. Further developments in the manufacturing of thin nanocrystalline cellulose films could be incorporated into miniaturised and wearable sensors that will make use of the power generation for battery-less health monitoring. In this context, cellulose films could be used both as a power source for low-power electronics and as sensing materials. The piezoelectric and triboelectric properties of cellulose could be exploited to determine hand movement or muscle contractions, among other applications, while thermoelectric generators could be used as temperature sensors. Moreover, the mechanical versatility of cellulose thermoelectrics could be used in the harvesting of waste heat directly from the generating source. This could enable the recovery of power from small electronic parts, such as electric wires, directly, boosting the efficiency.
Finally, the unprecedented development in materials science and engineering in the last few years will eventually allow for the design of more efficient energy harvesting devices, which would take full advantage of the properties of biomass-derived compounds.

Conclusions and Remarks
In summary, this review describes the latest developments in the use of biomassderived polymers, such as cellulose or lignin, to be applied to the generation of renewable energies. These materials have not only been chosen due to their low cost and sustainability, but also due to their good performance when employed in energy generation.
In the case of cellulose, the complex chemical structure, comprising both amorphous and crystalline domains, allows for the tailoring of its chemical properties to enhance the power yield. Its chemical structure can also be modified to improve its conductivity or charge density. In particular, the crystalline domains in cellulose can generate piezoelectricity due to its noncentrosymmetry, which has been exploited to develop energy harvesting devices. However, so far, the highest yield in these devices based on biomass materials has been reported in the case of nanocomposite materials. Despite the enhanced output achieved by nanocomposite-based piezoelectric generators, there is great interest in the use of biomass materials from low-cost sources to improve the possibility of scale up and manufacturability within a commercial setup. This form of power generation, using cellulose only, is higher when aligned cellulose nanocrystals are employed. However, an electrical output can also be obtained from natural cellulose obtained from certain plant structures, such as pomelo fruit or wood, greatly reducing the time and energy required for the fabrication of power generators.
The development of triboelectric generators has expanded the possibilities for the utilisation of biomass materials in energy harvesting. Cellulose presents a relatively high charge density due to the presence of hydroxyl groups within its surface. In addition, this material shows a high surface area when used in the form of nanofibres. As such, in general terms, the reported power generation rates in the case of triboelectric-based devices employing cellulose tend to be higher than piezoelectric generators. However, the versatility of cellulose molecules has allowed for the design of hybrid devices able to harvest energy employing piezoelectricity and triboelectricity sources.
Another potential field where the use of biomass-derived materials could be incorporated is osmotic energy generation. This method represents a more optimal route to obtain energy from the environment, since the mixture between water sources containing different salinities occurs naturally at estuaries. The possibility of chemically modifying cellulose plays a pivotal role in this technology, since positively and negatively charged membranes can be synthetised through quaternisation and oxidation, respectively. Both natural lignocellulose and cellulose nanofibres have been incorporated in this field. However, the highest yield was achieved by using graphene oxide-based nanocomposites. Among the studied methods in this manuscript, osmotic energy represents the most promising approach to replace traditional carbon-based energy generators, given the high-power output, and the possibility of scaling up the size of the generators.
Contrary to the approaches previously mentioned, the use of biomass materials in thermoelectricity has not been extensively studied and the number of described approaches in the literature is scarce. The use of pure cellulose as an active material by the oxidation of cellulose nanofibres and infiltration of ions has been demonstrated, with a high-power factor and Seebeck coefficient. However, the most popular approach involves the incorporation of carbon nanotubes to form complex nanocomposites. Thus, there is a need for the development of simple and low-cost high-yield thermoelectric generators based on biomass materials exclusively.