Energy & Environmental Science PAPER A plant-like battery: a biodegradable power source ecodesigned for precision agriculture† Cite this: Energy Environ. Sci., 2022, 15, 2900 Marina Navarro-Segarra,a Carles Tortosa,a Carlos Ruiz-Dı́ez,a Denis Desmaële,a Teresa Gea, b Raquel Barrena,b Neus Sabatéac and Juan Pablo Esquivel *ade The natural environment has always been a source of inspiration for the research community. Nature has evolved over thousands of years to create the most complex living systems, with the ability to leverage inner and outside energetic interactions in the most efficient way. This work presents a flow battery profoundly inspired by nature, which mimics the fluid transport in plants to generate electric power. The battery was ecodesigned to meet a life cycle for precision agriculture (PA) applications; from raw material selection to disposability considerations, the battery is conceived to minimize its environmental impact while meeting PA power requirements. The paper-based fluidic system relies on evaporation as the main pumping force to pull the reactants through a pair of porous carbon electrodes where the electrochemical reaction takes place. This naturally occurring transpiration effect enables to significantly expand the operational lifespan of the battery, overcoming the time-limitation of current capillary-based power sources. Most relevant parameters affecting the battery performance, such as evaporation flow and redox species degradation, are thoroughly studied to carry out device optimization. Flow rates and power Received 21st February 2022, outputs comparable to those of capillary-based power sources are achieved. The prototype practicality Accepted 30th May 2022 has been demonstrated by powering a wireless plant-caring device. Standardized biodegradability and DOI: 10.1039/d2ee00597b phytotoxicity assessments show that the battery is harmless to the environment at the end of its operational lifetime. Placing sustainability as the main driver leads to the generation of a disruptive battery rsc.li/ees concept that aims to address societal needs within the planetary environmental boundaries. Broader context Waste electrical and electronic equipment (WEEE) or e-waste is becoming the fastest growing waste stream worldwide. Due to their content of heavy metals and strong electrolytes, batteries are one of the most hazardous components of e-waste. Regrettably, the vast majority of the batteries sold worldwide are neither properly disposed nor recycled after use. The demand for portable batteries is expected to increase significantly in the following years due to the upcoming wave of power hungry Internet-of-Things (IoT) devices. Agriculture is a strategic sector that particularly demands electronic devices to increase the efficiency of food production while minimizing the use of agrochemicals. This demand for portable energy must be fulfilled by power sources with high performance and also they should possess environmental sustainability in terms of their materials, fabrication methods and disposal. This work presents a nature-inspired battery, specifically ecodesigned to follow the lifecycle for applications in precision agriculture. The proposed battery meets the power requirements for wireless sensors and is composed of benign non-toxic materials that undergo biodegradation at the end of its operational life cycle. a Instituto de Microelectrónica de Barcelona, IMB-CNM (CSIC), C/dels Tillers sn, Campus UAB, 08193 Bellaterra Barcelona, Spain. b E-mail: juanpablo.esquivel@csic.es Universitat Autònoma de Barcelona (UAB), 08193 Bellaterra Barcelona, Spain 1. Introduction c Catalan Institution for Research and Advanced Studies (ICREA), Passeig Lluı́s Agriculture needs a new paradigm. With the worldwide population Companys 23, 08010 Barcelona, Spain d BCMaterials, Basque Centre for Materials, Applications and Nanostructures, UPV/ predicted to reach 9 billion inhabitants by 2050,1 it is likely that the EHU Science Park, 48940 Leioa, Spain. overall food demand will continue to grow. To meet this demand, E-mail: juanpablo.esquivel@bcmaterials.net ever-increasing crop yields are required. Nevertheless, it is today e IKERBASQUE, Basque Foundation for Science, 48009 Bilbao, Spain widely recognized that the intensive use of agrochemicals (e.g., † Electronic supplementary information (ESI) available: Battery half-cell voltam- fertilizers, herbicides, pesticides, etc.) that have been utilised so far mograms, photograph of device components, derivation of evaporation rate equation, hygrometric chart, additional figures of redox species characterization to boost food production is not sustainable in the long term. and commercial sensor electronic evaluation. See DOI: https://doi.org/10.1039/ Intensive agriculture based on high inputs of chemicals proves to d2ee00597b be one of the major drivers of soil degradation, pollution of natural 2900 | Energy Environ. Sci., 2022, 15, 2900–2915 This journal is © The Royal Society of Chemistry 2022 Paper Energy & Environmental Science water resources, and biodiversity loss. In this regard, wireless membraneless microfluidic fuel cells (mFCs) or microfluidic redox sensor networks (WSNs) can readily help growers to adapt their flow cells. One of the great advantages of mGCs is that they can practices and shift towards precision agriculture (PA). generate electrical energy using a variety of reactants, including In a nutshell, WSNs for PA are based on a collection of biofuels.12–16 For instance, mFCs exploiting methanol, ethanol, or individual nodes deployed across agricultural lands to monitor glucose as fuel have been reported.17–25View Article Online The possibility of driving View Journal | View Issue a variety of local parameters (e.g., air temperature, soil moisture, low-consumption electronic systems, including wireless sensors, soil PH, salinity, light, etc.). Measurements are then used to was also demonstrated.26 mFCs have thereby been envisioned as derive microclimate conditions and/or detect variability among promising candidates to replace small conventional primary different parcels. This way, instead of treating uniformly (and batteries. Unfortunately, even after almost two decades of intense This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. sometimes unnecessarily) a whole field, growers can better apply research efforts, mFCs have not succeeded in leaving the lab bench the right amount of inputs, at the right place and at the right yet. The problem is that syringe pumps or pressure controllers are Open Access Article. Published on 30 May 2022. Downloaded on 5/11/2023 2:47:37 PM. time. To the best of our knowledge, however, WSNs for PA are required to drive the reactants towards the electrochemical reac- deployed in a limited number of agricultural settings; mainly tive zones inside the mFCs. Such auxiliary external pieces of those whose configuration remains relatively unchanged from equipment are bulky and expensive. Even worse is that they season to season (e.g., vineyards,2,3 olive groves4 or orchards5,6). may consume more electrical energy than the MFCs can actually Comparatively, there is a scarcity of studies reporting the use of produce. To eliminate external pumping systems, the scientific WSNs in cereal crops (e.g., wheat, millet, quinoa, etc.). From our community has come up with mFCs made of paper materials. personal analysis, a plausible main reason that hampers a wider Paper can indeed serve as a low-cost passive pumping mechanism adoption of WSNs in cereal crops is that the nodes may actually because it can absorb and drive liquids through capillary constitute obstacles to be avoided during harvesting. Recollect- action.27,28 Paper-based materials are even more appealing for ing tens or even hundreds of nodes before harvesting, however, PA because they are nature friendly (e.g., cellulose can be made of is not a plausible option. In the future, low-cost nodes that could fibers coming from corn straw, bamboo, bagasse, grass-plants like just be left in the soil indefinitely without negatively impacting reeds, etc.). Nonetheless, although paper offers a very simple and the environment could bring innovative solutions to such limita- elegant solution to bypass the use of external pumps, many paper- tions. Nevertheless, there are several technological challenges to based mFCs still make use of critical raw materials to produce be tackled to turn such an ideal case scenario into reality. One of electricity (e.g., platinum, palladium, platinum-ruthenium alloys, the major hindrances is that current nodes are far from being etc.).15,29 The use of such noble metal catalysts does not align well eco-friendly. In particular, the power source turns out to be a with an ideal circular sustainable approach. Although a paper- major source of potential pollution. At present, most nodes in PA based mFC using enzymes as biocatalysts was reported to be are battery-powered. Commercially available batteries, however, capable of driving a digital clock,30 the well-recognized restricted are loaded with heavy metals and/or toxic chemicals (e.g., stability of enzymes over time appears as a strong limiting factor lithium, cadmium, manganese, copper, zinc, etc.) which prove for targeting real PA applications. Moreover, enzymes are often to be highly detrimental to the environment.7 The use of solar most active solely in a precise range of temperature.31,32 However, panels, wind generators, or hydro-generators connected to unlike in indoor environments, strong variations under ambient irrigation pipes was reported for supplying power to nodes in conditions over days/nights (e.g., temperature, humidity, etc.) PA.8 Nonetheless, such energy harvesting devices cannot be used must be expected in an open field. In 2017, Esquivel et al. reported on their own. Indeed, they solely produce electrical energy spor- PowerPAD3.33 Different from the paper-mFCs previously reported, adically (i.e., if it is windy, if it is not too cloudy, etc.). As a direct the completely all-organic PowerPAD architecture did not involve consequence, they actually serve exclusively as auxiliary rechar- the use of any metals or enzymes. Notably, the authors assessed ging units to prolong the batteries’ lifespan. Since conventional and validated that the battery could be biotically degraded by batteries are still needed, the situation is actually even worsened microorganisms. This represented a significant step forward, from the point-of-view of the environmental impact. For instance, since it means that the battery could be discarded in a field with solar panels also include a variety of hazardous materials.9,10 a minimal environmental impact. Since then, several other exam- Generally speaking, the end-of-life of electrical electronic equip- ples of batteries focusing on biodegradation as the alternative end- ment (i.e., their disposal, dismantling, difficult and limited recy- of-life have been reported.34–36 A limitation of the PowerPAD, cling) has already started to arise alarming societal concerns. The however, was that electricity was produced mainly through a deployment of new digital technologies in PA should not come at diffusion process which limited the overall battery efficiency. More the price of an additional uncontrollable electronic waste burden. recently, the same group reported an enhanced design where the Because batteries are considered as one of the major sources of battery performance was increased by exploiting quasi-steady e-waste, novel sustainable power sources are needed to enable a capillary flows. Despite the gain obtained in performance, the more massive deployment of WSNs in PA.11 operational lifetime remained limited; in both designs, the bat- In the context of PA, new battery architectures featuring teries could generate electric energy only for a few hours. Unfortu- nature-based characteristics appear especially appealing. This nately, this is too restrictive for targeting PA applications. is the case for microfluidic galvanic cells (mGCs). We here refer Herein, we introduce the FlowER: an Evaporation Flow to mGCs as miniature fluidic devices that can produce an Redox battery intended to bring new solutions to some of the electric current by means of a redox reaction, such as challenging requirements imposed by PA applications (Fig. 1). This journal is © The Royal Society of Chemistry 2022 Energy Environ. Sci., 2022, 15, 2900–2915 | 2901 View Article Online Energy & Environmental Science Paper This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Open Access Article. Published on 30 May 2022. Downloaded on 5/11/2023 2:47:37 PM. Fig. 1 FlowER battery concept evolution from the inspiration, found in the liquid transport in plants, to an operational prototype. The technological adaptation of the FlowER battery concept has been performed through the selection of environmentally harmless materials to produce two inlet channels acting as roots, from which the device nurtures with dissolved redox species, just like plants take water and nutrients from their surroundings. The main laminated paper core, analogous to a plant stem, includes two porous carbon current collectors and a salt bridge, indispensable components to create a membraneless galvanic cell. The core ends with two outlet channels that transport the fluid towards the top. A rounded leaf-like absorbent pad exposed to the atmosphere keeps wicking the solution by transpiration pull through the battery. Finally, a 3D printed compostable green casing (i.e., reservoirs + lid + base + paper supports) was used to provide mechanical robustness and verticality. The FlowER battery conception and development approach 2. Results and discussion invokes a change in the current portable battery models. This 2.1 Fundamentals of water transport in plants change pursues a new battery paradigm in which power sources are designed to be paired with the life cycle of the intended All living organisms need to transport fluids to survive and application. Environmental sustainability has been placed as prosper. In the animal kingdom, many species rely on a heart to the core priority to create a disruptive technology that can ensure blood circulation within the body. In contrast, plants are evolve within the environmental boundaries of the planet. In not equipped with analogue contractile organs. Yet some this process, eco-design was used as a fundamental tool, from raw species of trees are able to transport fluids upwards over material selection to disposability considerations – including incredibly long distances (e.g., some sequoia trees can be more manufacturing processes, operational time, and energy capacity – than 100 m high37). Strikingly, they do it against gravity and in to create a prototype that meets PA-lifecycle requirements, a passive way, namely, without dedicating metabolic energy to namely, affordability, extended operational lifespan, and biode- pumping. Instead, they rely on a combination of three inter- gradability. The overall architecture of this new battery is pro- twined passive flow mechanisms. foundly inspired by nature: both its appearance and the way it The plant roots first capture the water found in the operates make it a true ‘‘flower-like’’ battery. The ‘‘roots, stem and surrounding soil. Then, an osmotic-flow mechanism generates petals/leaves’’ are all made of low-cost paper-based materials. a positive pressure that pushes the water collected towards the An absorbent pad, placed on the top of the FlowER, exploits xylem. The xylem is a portion of the sapwood composed evaporation combined with cohesion and capillary forces in a of repeating narrow vessels that run vertically up the plant. way analogous to transpiration pull in plants. Liquid reactants This vascular tissue is made of the hollow walls of dead cell contained in reservoirs can thereby flow continuously through skeletons, and is mainly involved in the ascent of bulk – almost porous carbon electrodes where electrochemical reactions take place. pure – water from the roots up to the leaves.‡ Once the water Notably, the operation time, which usually ranges from solely tens of has reached the xylem conduits, capillarity comes into play. minutes to a few hours maximum in conventional paper-based fuel Indeed, depending upon the species and location of the plant, cells, can be significantly extended. Furthermore, the harmless green the inner diameter of the xylem conduits can range from 10 mm chemicals selected prove to be powerful enough for real-world to ca. 100 mm. Nevertheless, capillary rise varies with the inverse applications. As a proof-of-concept, we demonstrate that the FlowER of the vessel diameter. As discussed by Steudle, even with the can supply power to a commercial wireless monitoring unit dedi- cated to precision horticulture. Finally, the possibility for the new ‡ For a more exhaustive description about plant physiology (e.g., consideration of plant-inspired battery to end its life cycle by returning to nature is the phloem tissue, etc.) the interested reader may consult dedicated review assessed by standardized ecotoxicity and biodegradation tests. articles (e.g., see 81 and references therein). 2902 | Energy Environ. Sci., 2022, 15, 2900–2915 This journal is © The Royal Society of Chemistry 2022 View Article Online Paper Energy & Environmental Science smallest xylem conduits, water can rise not more than 3 m high and connect to two separate outlet-channels of 40.5 mm  when considering solely capillary effects.38 It is the capillarity- 12 mm (defining a geometrical electrode area of 0.25 cm2). The transpiration combination that is actually considered to be the fluidic system ends up in the center of a flower-shaped absor- fundamental mechanism of water transport in large multi- bent pad that absorbs liquids omnidirectionally around its cellular plants. Transpiration (i.e., the conversion of water from 3601, maximizing the evaporation exposed area while keeping liquid to gas) is predominant in the foliage where water a co-laminar flow. Fan-shaped absorbent pads were first molecules evaporate through the multitude of closable micro- proposed by Mendez et al. in 2010, proving that the gradual scopic slit-like pores (stomata) mainly present on the under- increase of the unwetted area induces a quasi-steady flow rate sides of leaves. The sun provides the energy to overcome the in the preceding channel sections.46 More recently, our group This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. latent heat of evaporation of the water molecules. The removal incorporated this fluid control mechanism in a paper-based of the water molecules from the water–air interface (i.e., redox flow battery.47 The quasi-steady flow translated into a Open Access Article. Published on 30 May 2022. Downloaded on 5/11/2023 2:47:37 PM. menisci) cause the interfaces to recede into the pores and stable power delivery over time. When working under evapora- change their shape39 (i.e., the menisci become more concave). tion conditions the absorbent pad will be called evaporation Capillarity acts to restore the initial equilibrium shape of the pad. The paper assembly is fed with redox species solutions menisci, generating a strong pulling force so that water lost already dissolved in a 15 mL volumetric capacity reservoir. during the evaporation process is replaced. In order to create the final prototype, the battery paper struc- Interestingly, several engineering approaches have tried to ture is coupled with a 3D printed compostable casing, which capture the main attributes of the xylem and/or the transpiration provides verticality and mechanical robustness and compen- mechanism occurring in plants to create miniature synthetic sates the thickness of the different paper layer (a photograph of devices capable of pumping fluids without the use of any moving all the device components is shown in Fig. S1, ESI†). mechanical parts.40–44 Importantly, one could envision to use Using the evaporation action to drive the battery’s main flow such miniature pumps in conjunction with mFC architectures is, for paper microfluidics, the next natural step towards auton- in order to partly solve some of the aforementioned problems omy and competitiveness. Making use of a transpiration mecha- (e.g., bypassing the use of conventional bulky syringe pumps). nism instead of capillary forces allows the system to become Nonetheless, advanced microfabrication techniques and/or non- independent of the absorbent volumetric capacity. Hence, the environmentally friendly materials were still mostly needed to operation time is also decoupled from the paper’s total volu- fabricate these devices. As an alternative, porous materials with metric capacity (137 ml for this particular shape system), but pores in the micrometer-size range have been shown to be instead depends on the volume of the reactant reservoirs (15 mL appropriate for efficient capillary pumping of water. This is the in this case). On the other hand, the fluid flow rate changes its case for paper-based materials. Furthermore, in a manner dependence from the absorbent pad material wicking proper- similar to plants, paper-based materials can take advantage of ties to the evaporation-exposed area, becoming constant (under both capillary and evaporation effects.28,45 Because the large certain atmospheric conditions) and easily tunable. Combining deployment of numerous nodes for high resolution measure- these two characteristics, the result is a miniaturized low-cost ments capabilities in PA will require low-cost devices employing and paper-based autonomous mid-term operational battery environmentally responsible materials, we selected paper as the (from hours to days), which can start to compete in operational building block for the fabrication of our FlowER battery. time terms with traditionally commercial ones, thus bringing paper-based microfluidics power suppliers up to the next level. 2.2 Device design and operation principle A evaporation-driven flow battery has been designed with a 2.3 Influence of the evaporation pad area on the evaporation plant-inspired operation principle, indicating that it has two flow rate inlet channels acting as roots, from which the device nurtures Flow rate (Q) is one of the key parameters that determine a flow with already dissolved redox species, just like plants take water battery performance, as it defines the reactants availability for and nutrients from their surroundings. The main laminated electrochemical conversion. In the particular case of the device paper core, analogous to a plant stem, includes two porous herein presented, evaporation-driving forces govern the flow carbon electrodes connected by a U-shape paper that acts as a rate. Even though, the paper structure fills up by capillary salt bridge, indispensable components to create a membrane- forces, once completely full, the FlowER battery takes advan- less microfluidic galvanic cell. The core ends with two outlet tage of the transpiration mechanism naturally occurring on the channels that transport the fluid towards the top of the device. evaporation pad area to pull fresh redox species through the Finally, a rounded leaf-like absorbent pad exposed to atmo- electrodes and clean the reaction products, keeping an active sphere keeps wicking the solution by transpiration pull through flow towards the top. the core (see Fig. 1). The evaporation phenomenon has been widely studied in The prototype has been designed to be able to be planted in paper microfluidics due to the direct influence it may have on the bare ground. In a nutshell, a 110 mm  12 mm U-shaped precision and reproducibility of the devices.48–51 Here, a study piece of paper creates the inlet channels for the solutions and was carried out to investigate how to control and adapt tran- also forms the salt-bridge. Two porous carbon electrodes of spiration pull to the battery requirements. The evaporation pad 5 mm  10 mm create a flow-through electrode configuration area (Apad) is an easily tunable parameter that determines the This journal is © The Royal Society of Chemistry 2022 Energy Environ. Sci., 2022, 15, 2900–2915 | 2903 View Article Online Energy & Environmental Science Paper This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Open Access Article. Published on 30 May 2022. Downloaded on 5/11/2023 2:47:37 PM. Fig. 2 Evaporation flow rate control related to the evaporation area exposed to the air: 2, 1, 0.5 and 0 times the area of one side of the evaporation pad (APad = 19.6 cm2). Linear fit adjustment coefficients: R2A2 = 0.9997, R1A2 = 0.9996, R0.5A2 = 0.9924 and R0A2 = 0.9427. Environmental conditions: HR% A [51, 56]%, T A [28, 30] 1C and P = 1 atm. total liquid–air interface where the evaporation phenomenon environmental conditions on the evaporation flow rate. For this takes place, and hence the evaporation flow rate (Qevap) gener- reason, a simple evaporation flow rate modeling was proposed ated through the fluidic system. Four different APad configura- (the complete mathematical development is presented in the tions were tested (2  APad, 1  APad, 0.5  APad, and 0  APad), ESI†). Eqn (1) shows the dependencies and parameters that by covering different percentages of the evaporation pad area, influence the generated evaporation flow [g s1], 1  APad being the area of one side of the absorbent pad ð1  HR%Þ (19.6 cm2). A blue dye was used to visualize the total volume of Q ¼ ADHsat (1) Dx liquid evaporated from the reservoirs through the fluidic system over time. As shown in Fig. 2, the evaporation phenomenon at where D [m2 s1] is the diffusion coefficient of the liquid, Dx [m] the absorbent pad creates a quasi-steady flow (i.e. a constant represents the distance between 2 points in space, A [m2] is the slope over time) for 8 consecutive hours. Furthermore, the exposed area (i.e. total surface of liquid–air interface), Hsat volume of the liquid evaporated at a given time increases [g m3] is the saturation humidity (i.e., total amount of water proportionally while increasing the APad. Thus, under the same that can be dissolved in the dry air at a given temperature), and environmental conditions, the ratios between the evaporation HR% is the relative humidity (i.e., amount of water dissolved in flow rates (467 mL h1, 231 mL h1, 98 mL h1, and 8 mL h1) and the air in percentage respect to the saturation humidity). Here, the corresponding evaporation exposed areas 2  APad, 1  APad, D is approximated to 25.6  106 m2 s1 (T = 20 1C) and Dx is set 0.5  APad, and 0  APad are equal. as 106 m to normalize the orders of magnitude of experimental The flow rates, reported previously by our group, obtained and theoretical flow rates. with a capillary-driven fluidic system, were 900 mL h1, 540 mL h1 In order to validate the proposed Q.evap modeling, the and 288 mL h1 depending on the outlet channel length,47 experimental evaporation flow rates presented in the previous with operational times of 7, 13 and 25 minutes, respectively. section were compared with the theoretical flow rates estimated Comparing both fluidic systems, it can be concluded that using using eqn (1). The validation was done with the 2  APad evaporation as a driving force, the flow rates generated are within configuration, where A = 35.5 cm2. Ambient temperature and the same order of magnitude than those generated by capillarity. relative humidity values were recorded with a sensor during the Finally, it has to be noticed that, by completely covering the set-up course. Fig. 3a shows the comparison between the absorbed pad area, 0  APad, a negligible evaporation flow was experimental and theoretical flow rates for nine repeats. As it obtained, indicating that no other evaporation sites are created can be observed, both values significantly differ; however, when than the ones desired on the evaporation pad. computing the percentage difference between theoretical and experimental values, a constant k = 44.3  2.7% is obtained. The disagreement between the experimental and theoretical 2.4 Influence of the environmental conditions on the flows, expressed in the k parameter, could be explained due to a evaporation flow modeling discrepancy in the exposed evaporation area. The parameter A As mentioned above, the power delivered by the battery in eqn (1) stands for the total liquid–gas interface; taking into depends on the flow rate. Hence, for the purpose of this work account that in this case a paper-based evaporation pad is it was necessary to fully understand the influence of being used, the geometrical area does not correspond with the 2904 | Energy Environ. Sci., 2022, 15, 2900–2915 This journal is © The Royal Society of Chemistry 2022 View Article Online Paper Energy & Environmental Science system. Once completely full, the evaporation forces occurring at the liquid–air interface of the absorbent pad become pre- dominant over the capillary effect. This is when the flow battery is mainly evaporation-driven and the prototype is ready to be used. In this work, the battery was characterized using a well- known couple of quinone-based organic species, which have been already used as redox couples in power sources.33,47,52 This particular chemistry has already shown to successfully undergo anaerobic degradation and non-toxicity in previous This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Fig. 3 Theoretical vs. experimental evaporation flow rates and the corres- studies.33 As shown in Fig. 1, both half-cell redox reactions ponding ratio, k (%), between each of them. N = 9 repeats. present a proton-dependence; therefore, mixed-media condi- Open Access Article. Published on 30 May 2022. Downloaded on 5/11/2023 2:47:37 PM. tions are used as a strategy to increase the electrochemical cell voltage.53,54 Ultimately, a solution of 0.1 M p-BQ in 0.5 M of evaporation exposed area (see Fig. 3b). Due to the physical oxalic acid is used as the acid catholyte, while 0.1 M H2BQS in space occupied by the paper fibers, the actual liquid–gas inter- 1 M of KOH solution composes the alkaline anolyte. The half- face area is lower than the geometric one. The k ratio compen- cell reaction characterization in a conventional three-electrode sates for this disagreement and other collateral effects of using electrochemical cell can be found in Fig. S2 (ESI†). a specific porous material as an evaporation pad (i.e. adhesion Discrete characterization of the battery was performed for forces generated between the liquid and the paper walls). This 8h by consecutively recording the polarization curves every ratio also allows the calculation of the actual flow rate of the hour. The OCV was in all cases between 0.8 and 0.9 V. As paper-based fluidic system directly from pre-recorded environ- depicted in Fig. 4a, the battery yields a maximum power density mental conditions (T1, HR%, P). The theoretical modeling of the of 1.42 mW cm2 at time 0 h, and then decreases linearly over evaporation flow rate eliminates the need for photographic time until 0.69 mW cm2 at time 8 h. It is important to notice analysis, deepens the environmental conditions’ influence com- that this power density is in the same order of magnitude as prehension, and makes it possible to predict the device perfor- the maximum power density reported in our previous work mance in different climates and environments. Henceforth, all (1.88 mW cm2 at 540 mL h1), using the same fluidic system the prototype characterization experiments were conducted but driven by capillary forces. while simultaneously measuring ambient parameters with a Despite showing a decrease in output power during the first commercial sensor. The recorded environmental data provide 8 h, these results demonstrate that evaporation-driving forces information about the battery working conditions and allow the are able to maintain the co-laminarity of the two reactants computation of the actual working Qevap, strengthening the streams, expanding the operational time of the paper-based prototype control during characterization. Furthermore, it allows battery from minutes to hours. The power density loss causes the prediction of the battery behavior in terms of Qevap, for a are further studied in the section below when assessing redox wide range of ambient conditions. Fig. S3 (ESI†) depicts the species’ stability. hygrometric chart relating the battery theoretical flow rate with Next, continuous operation was characterized by measuring temperatures from 5 1C to 35 1C and relative humidity condi- battery voltage evolution over time under four discharge cur- tions from 0% to 90%. rents. Fig. 4b shows the discharge curves obtained in terms of battery voltage and power density. This characterization under 2.5 Battery operation and optimization continuous operation was used to assess the device efficiency. Once the paper structure was fluidically validated, the configu- The faradaic efficiency (ZF) allows evaluation of the electroactive ration creating the highest flow rate (Qevap = 467 mL for 2  APad) species utilization, quantifying the theoretical charge capacity was selected to further characterize the evaporation flow that is actually being converted into current (see eqn (2)). The battery. This section reports the battery electrochemical char- energy efficiency (ZE) was also evaluated by comparing the acterization and the studies regarding the durability of the power delivered to the energy that ideally would be delivered device’s critical components: redox species reactants, the paper (see eqn (3)). core (i.e., paper channels and electrodes), and the evaporation The fluidic battery is able to deliver 100 mA for 8 h without pad. Isolating and studying each component allows the accurate voltage or power decrease, delivering a total of 0.8 mA h of depiction of the battery performance and set-up strategies to capacity and 13.5% ZF and 9% ZE conversion rates. These values further expand the operational lifetime. are within the average range for the reported paper-based microfluidics electrochemical power sources (1–10% ZF and Operation characterization 1–5% ZE).47 As expected, for higher discharging currents the In order to set up the prototype, 5 mL of the electroactive voltage decreases while efficiencies increase, leading to a 15% species previously dissolved in water-based solutions are placed decrease for 250 mA (with 2 mA h of capacity, 28% ZF and 16% ZE) in the reservoirs. Then, the inlet channels are immersed into and a 50% for 500 mA (4 mA h of total capacity, 71% ZF and 24% the solutions. The paper strips absorb the liquid creating a ZE). The device operating at 500 mA delivered the highest power convective flow driven by capillary forces and filling up the values (0.2 mW to 0.1 mW) during the whole measurement. In This journal is © The Royal Society of Chemistry 2022 Energy Environ. Sci., 2022, 15, 2900–2915 | 2905 View Article Online Energy & Environmental Science Paper Durability of the device’s core To study the durability of the device’s paper core (paper channels and electrodes), the battery was set up for four days and subjected to continuous current demand of 100 mA. Every 24 h the anolyte and catholyte reservoirs were refilled with fresh solutions and the evaporation pad was replaced with a new dry one. Moreover, the battery polarization curve was recorded before and after this daily conditioning. Depicted in Fig. 5 are This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. the obtained discharge curves together with the polarization curves of the battery over the days. The battery shows a continuous power delivery of around 200 mW cm2 for four days. Open Access Article. Published on 30 May 2022. Downloaded on 5/11/2023 2:47:37 PM. A transition regime appears when changing the species and the evaporation pad, in which the capillary flow caused by the new evaporation pad replenishment perturbs the battery per- formance. When analysing the polarization curves, however, it can be observed that the battery performance improves after the daily conditioning during the first three days. This result agrees with the current density peak loss caused by the anodic species degradation, which was obtained from the analysis of the solution in the reservoir every day before replenishment. However, on the fourth day of continuous operation, the battery performance dropped to 0 V. No further improvement was seen from changing the solutions or the evaporation pad, indicating that the device’s core operational lifetime was reached. Redox species stability The anolyte and catholyte degradation and its relation with the total loss of performance of the battery were investigated. The redox species were analyzed via cyclic voltammetry (CV) for 9 days. Three different volumes (5 mL, 10 mL, and 15 mL) were assessed, in order to study the influence of exposed area on the oxygen vs. total solution volume stored. Fig. 6a shows the Fig. 4 Battery characterization for 8 h, (a) discharge polarization curves of correlation between the current density peak and storage time the evaporated driven flow battery, recorded every hour since the start (black for each volume. In the case of the anodic reactant, H2BQS, the line) until the 8th hour (palest line); (b) continuously measured discharge slope decay was palliated when increasing the total solution curves at four different discharge currents. Prototype supplied with 0.1 M H2BQS in 1 M KOH as the anolyte and 0.1 M p-BQ in 0.5 M C2H2O4 as the volume; the signal loss was 100% for 5 mL solution, 47% for catholyte. Average flow rates calculated from eqn (1), and environmental 10 mL, and 20% for 15 mL, showing a direct dependence conditions recorded with a commercial sensor: (a) Qevap = 239 mL h1. between oxygen exposure and H2BQS spontaneous degradation. HR% A [54, 57]%, T A [19, 21] 1C. (b) I = 100 mA; Qevap = 276 mL h1, HR% Since all samples were stored in the same kind of reservoirs, the A [62, 70]%, T A [25, 27] 1C; I = 250 mA; Qevap = 333 mL h1, HR% A [55, 59]%, contact surface area between the reactive and the air was the T A [25, 26] 1C; I = 500 mA; Qevap = 260 mL h1, HR% A [58, 63]%, T A [23, 24] 1C; I = 1000 mA; Qevap = 351 mL h1, HR% A [55, 63]%, same, indicating that the amount of H2BQS oxidized by the air T A [26, 28] 1C (P = 1 atm). was equal for all three samples. Therefore, the different volumes of reagents led to different concentration changes and thus maximum current peaks. With respect to the cathodic species, all three cases, the battery was able to maintain a continuous p-BQ, the decay of the current density peak is independent of the power supply for 8 h. However, when the battery was subjected to storage volume; 94% of the signal loss was measured in all cases. 1000 mA working current (i.e. 4 mA cm2), the operational This result agrees with the naturally occurring reduction of p-BQ working time decays to 30 minutes, yielding a 0.5 mA h capacity, in aqueous media. The reduction product, hydroquinone (HQ), an increase in faradaic efficiency up to 99% ZF but a decrease in spontaneously polymerizes into large poly-hydroquinone chains, energy efficiency to 12% ZE. Higher current densities provoke a losing its electrochemical activity.56 As a consequence of these faster depletion of the reactants at the electrodes, thus the results, the prototype operating volume was increased to 15 mL. galvanic cell evolves from working in ohmic to mass transport With the aim of preventing the redox species degradation, loss regions in which the performance decays nonlinearly with deoxygenation with N2(g) and storage of the solutions were current density.16,55 Hence, a working condition limitation was assessed with CV for 14 days. Fig. S4 (ESI†) shows the current found above 500 mA or 2 mA cm2 current density.47 density peaks obtained. This approach demonstrates to 2906 | Energy Environ. Sci., 2022, 15, 2900–2915 This journal is © The Royal Society of Chemistry 2022 View Article Online Paper Energy & Environmental Science This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Open Access Article. Published on 30 May 2022. Downloaded on 5/11/2023 2:47:37 PM. Fig. 5 Continuous discharge curve of the evaporation-driven flow battery working at 100 mA for 4 days with daily renewal of redox species solutions and evaporation pad. Insets: Battery OCVs vs. time, battery polarization curves before and after evaporation pad and redox species replacement, and anolyte and catholyte degradation study in one day. Average flow rates calculated from eqn (1), and environmental conditions recorded with a commercial sensor: Day 1; Qevap = 103 mL h1, HR% A [60, 81]%, T A [19, 22] 1C; Day 2; Qevap = 184 mL h1, HR% A [59, 64]%, T A [19, 22] 1C; Day 3; Qevap = 215 mL h1, HR% A [57, 66]%, T A [17, 20] 1C; Day 4; Qevap = 157 mL h1, HR% A [65, 73]%, T A [18, 21] 1C (P = 1 atm). with oxygen from air. The chemical stability of organic redox compounds in aqueous media has been widely studied, due to their interest for redox flow batteries. The quinones containing ketone groups, such as p-BQ, generally show instability because their carbon–oxygen double bonds are susceptible to reacting with nucleophiles, such as water, via Michael addition.57,58 Optimized continuous operation prototype Finally, an optimized FlowER battery prototype was assembled and tested under a continuous discharge current of 100 mA. In this case, the 2  APad configuration was selected and a 2 mm of Fig. 6 Battery reactant degradation through time. (a) Effect of time the protective oil layer was added on the top of the 15 mL storage on different solution volumes. (b) Effect of time storage on solution in the battery reservoirs. The battery yielded a power 10 mL volume with or without a top protective layer of sunflower oil. density between 0.25 and 0.2 mW cm1 with averages of 13.3% ZF and 7.7% ZE, over 3 days and a half (see Fig. 7). Perturbations on the power delivered can be attributed to changes in the successfully prevent the degradation of the H2BQS anodic evaporation conditions (i.e. Tamb, P, and HR%) that are reactant, whereas the p-BQ degradation continues regardless reflected in the flow rate. The performance drastically drops of the deoxygenation. Consequently, as a simple and low-cost on the fourth working day. As stated above, the oil protection method to avoid oxygen molecules from reaching the redox layer does not prevent the p-BQ species from degradation; the species, we added a 2 mm thick layer of vegetable oil on the top catholyte and anolyte solutions were tested and a 56% signal of the reactant solutions. The result of species electrochemical decrease was obtained for p-BQ. response over 56 days (8 weeks) is depicted in Fig. 6b; all stored Overall, the flow battery operational time was expanded in the same volume (10 mL) and concentration. from hours to days. Achieving an operation time that considerably As can be seen, the H2BQS current peaks decay less with exceeds those demonstrated in the literature for paper-based time when the protective oil layer is present. Quantitatively, energy conversion devices (usually in the range of minutes).47 only 37% of signal loss was recorded for H2BQS with the oil After the conducted studies, a few situations that could protection, while the bare solution presented a 84% decrease. produce a decrease in the device performance were identified. Regarding p-BQ, no stabilization of the redox species was Depending on the cause, these failure modes can be categor- possible, 50% of degradation is reached in 3 days and 100% ized as effects related to the intrinsic assembly of the battery or of degradation in 28 days for both solutions. Confirming that, to the external events due to the environmental conditions. For as previously stated, its degradation comes from the fact of example, the evaporation process results in an accumulation of being diluted in a water media and not from being in contact solid compounds in two regions, the evaporation pad and the This journal is © The Royal Society of Chemistry 2022 Energy Environ. Sci., 2022, 15, 2900–2915 | 2907 View Article Online Energy & Environmental Science Paper model CR2032) powers the device. As a first prerequisite to drive the sensor’s internal circuitry using the FlowER battery, we need to amplify the voltage generated (initially inferior to 1 V, see for instance Fig. 4). A common strategy used in the literature to increase the output voltage of paper power sources is cell stacking; different studies have reported stacking designs and their effects on the cell output.59–63 In this work, however, a voltage boost converter (VBC) integrated circuit was used to amplify the operating voltage up to 3.1 V (see additional details This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. in ESI†). We favoured this option because the commercial Fig. 7 Continuous power density delivered by the evaporation-driven Flower Caret sensing unit is not an eco-friendly sensing unit; Open Access Article. Published on 30 May 2022. Downloaded on 5/11/2023 2:47:37 PM. flow battery discharged at 100 mA for 4 days. Also depicted is the flow rate evolution. Inset: 4 day degradation of the anolyte and catholyte namely, it is made of conventional electronic components (e.g., solutions. Average flow rates calculated from eqn (1), and environmental printed circuit board, integrated circuits, etc.). Consequently, conditions recorded with a commercial sensor: Day 1; Qevap = 237 mL h1, the additional use of the VBC did not appear critical for HR% A [68, 72]%, T A [26, 27] 1C; Day 2; Qevap = 260 mL h1, HR% A demonstration purposes at this stage of development. [61, 73]%, T A [27, 27] 1C; Day 3; Qevap = 308 mL h1, HR% A [55, 67]%, T A [23, 27] 1C; Day 4; Qevap = 358 mL h1, HR% A [54, 58]%, T A [25, 27] 1C In addition, it turns out that the communication modes of (P = 1 atm for all days). the BLE protocol frequently imposed current peaks up to 5– 6 mA (e.g., see Fig. S6 and S7, ESI†). This means net power consumption peaks up to E18 mW. Comparatively, the max- electrode edges. This effect of saturation of the evaporation imum net power generated by the FlowER battery did not area could be mitigated by introducing a washing buffer, exceed 0.4 mW (see maximum power points in Fig. 4). To whereas the electrode zone sealing could be improved to avoid overcome this difficulty, a 6.8 mF super-capacitor was selected parasitic evaporation. Regarding working conditions, the pro- as a refillable energy reservoir capable of absorbing any totype has demonstrated good performance under typical punctual current/power peaks that could not come directly greenhouse conditions; however, high working temperatures from our evaporation-driven battery. and/or low relative humidity would increase the flow rate, thus First, the capability of the FlowER battery to fully charge the causing early exhaustion of the active materials. In this case, super-capacitor fully discharged was verified. As soon as the the design should be appropriately scaled to the targeted FlowER battery with the VBC was connected, the voltage ambient conditions. Thus providing an adequate volume of generated abruptly dropped from its initial OCV value to active solutions according to the estimated flow rate in the 0.5 V (see A in Fig. 8). The battery output dropped further hydrometric chart (Fig. S3, ESI†). Furthermore, extreme ambi- down to E0.3 V when the VBC started to charge the super- ent conditions, such as heavy rain or flooding, would cause a capacitor (see B). The full charging process took about 15 min, malfunction. To mitigate the detrimental effect that excess with two distinct phases. Indeed, the VBC needed to overcome a water would have on the paper-based prototype, a breathable cold-start phase before it could normally charge the capacitor. impermeable and biodegradable coating (such as natural wax) During the first slower charging rate phase (see B and C), the could be applied onto the paper structure. It is foreseen that current demand was higher. Comparatively, the charging rate these failure modes could be tackled with design optimization became much higher during the last 3 minutes of the process and through a tailoring process addressing specific working (see C and D). In the course of this second phase, the VBC’s conditions or requirements. Furthermore, thanks to the mod- internal circuitry could exploit its advanced non-linear switch- ular design, the FlowER battery could be easily modified to based technique to optimize the voltage across the storage enable evaporation pad and/or electrode replacement or acces- element. As a direct consequence, the current demand sible refilling of the active solutions, which would also result in decreased and the battery voltage re-increased to 0.5 V. At the an extension of its operational lifespan. end of the charging process (i.e., when the super-capacitor voltage reached 3.1 V), the current consumption dropped to 2.6 Driving a wireless smart monitoring unit for precision the nano-microamps range and the FlowER battery voltage horticulture returned swiftly to 0.7 V (see D). After the full charge of the With the aim to test a realistic PA application scenario, a Flower super-capacitor, the Flower Caret sensor was turned on. At this Caret smart monitoring unit was purchased. Flower Caret is moment, the capacitor voltage suddenly dropped from 3.1 to a low-cost device commercialized for horticulture or indoor 2.3 V (see E). This was due to the start-up sequence of Flower gardening. The unit encompasses four sensors for measuring Caret (see Fig. S5, ESI†). Once activated, the sensor entered essential PA parameters. In particular, the unit can concur- into the sleep mode and ran advertising events every 1 s (see rently monitor the soil conductivity, soil humidity, air tempera- Fig. S6, ESI†). These events also include three current peaks ture, and light exposure of a nearby plant. Flower Caret runs reaching 6 mA, but their duration was very short (o1 ms). the Bluetooth Low Power (BLE) protocol for transferring Despite their presence, the battery proved to be powerful the data measured on a smartphone (see Fig. S5–S7, ESI†). enough to recharge the capacitor back to 3.1 V in less than Normally, a primary 3 V cell coin battery (i.e., lithium battery 2 min. The sensor can thereby stay activated and run the 2908 | Energy Environ. Sci., 2022, 15, 2900–2915 This journal is © The Royal Society of Chemistry 2022 View Article Online Paper Energy & Environmental Science This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Open Access Article. Published on 30 May 2022. Downloaded on 5/11/2023 2:47:37 PM. Fig. 8 Proof of concept of the evaporation driven flow battery demonstrating the capability of powering the start-up, passive mode and Bluetooth connection of a horticulture caring device. advertising mode as long as needed. To transfer data, the microbiome (including meso and thermophilic bacteria and Flower Caret was connected to the smartphone. A second fungi). Therein, molecules are first hydrolyzed by extracellular sudden drop in the capacitor voltage occurred (see F). Using enzymes to simpler molecules and later fully mineralized to CO2 this scenario, Fig. 8 shows that the FlowER feasibility battery and H2O by the involved microorganisms. could successfully recharge the capacitor and maintain wireless First, biodegradability was assessed under controlled data transmission to a smartphone. Despite the fact that this aerobic conditions and following a procedure adapted from first proof-of-concept is based on the use of additional external the standard EN 13432.72 Three different types of samples were electronic components (i.e., the VBC and a super capacitor), it analysed: (i) compost alone (used as control); (ii) compost with clearly demonstrates that the FlowER battery is already capable un-used evaporation pads (i.e., cellulose without reactants); of generating enough electrical power to drive off-the-shelf (iii) compost with used evaporation pads (i.e., cellulose with electronic components useful for PA applications. Thanks to reaction products accumulated for 3 days). The dynamic the modularity of its design, the FlowER battery hence appears respiration index (DRI) was computed as an indicator of the as an ideal candidate to supply power to novel approaches in microbial activity evolution through the degradation process green electronics that are being developed for the smart agri- using eqn (4).73,74 The high initial DRI values (Fig. 9a) indicate a culture sector, also pursuing biodegradability as plausible rapid microbial growth derived from the presence of easily electronics end-of-life.64–66 biodegradable carbon. This behaviour is even sharper in the case of the evaporation pad with reaction products. Moreover, 2.7 Aerobic biodegradability and phytotoxicity tests the dynamics up to day 12 support this trend, confirming that The FlowER battery is intended as a sustainable power supply for the samples are acting as a substrate and being consumed by PA applications. For this reason, it is compelling to demonstrate the consortium of microorganisms present in the compost. that the prototype end-of-life is completely aligned with this From day 15 onwards, the DRI curves flatten and there are no purpose. The FlowER battery green casing structure is made of remarkable differences in the microorganisms’ activity for the COMPOST3D, a commercially available material certified as different samples. biodegradable under OK Compost standard (OK Compost Fig. 9b depicts the accumulated oxygen uptake (eqn (5)) by HOME, TÜV Austria). Hence, the tests reported in this section the three samples over 18 consecutive days. The total oxygen analyse the evaporation pad as a representative portion of the consumed was 2.75  0.54 g for the compost, 3.30  0.38 g for paper structure which, furthermore, ends up having the highest the evaporation pads without products, and 4.02  0.6 g for the accumulation of redox species and electrolyte salts. Biodegrada- evaporation pads with accumulated products. Total oxygen tion pathways for standard organic molecules as cellulose or consumed by the microorganisms during the biodegradation simple aromatic compounds, like the battery components, are process allows for the calculation of the percentage of biode- well stablished in the general literature related to composting gradation as compared to the theoretical oxygen demand and bioremediation.67–71 Both natural environment (soil) and (eqn (6)). Standard EN 13432 requires 90% of biodegradation composting matrices are complex ecosystems with a rich in less than 6 months to consider a material as biodegradable. This journal is © The Royal Society of Chemistry 2022 Energy Environ. Sci., 2022, 15, 2900–2915 | 2909 View Article Online Energy & Environmental Science Paper whereas composts with GI values higher than 80% assure the absence of phytotoxic substances.75 As observed, compost con- trol has a positive effect both in IG and RRE. Regarding com- post from aerobic biodegradation of evaporation pads with reaction products, RRE does not show significant differences from that of compost control, whereas the percentage of germination is slightly lower. In any case, the GI is higher than 100% ensuring its beneficial use. However, when the same chemical species used in the battery were added directly to This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. a compost control but not subjected to a biodegradation process, a significant negative effect is observed. This effect is Open Access Article. Published on 30 May 2022. Downloaded on 5/11/2023 2:47:37 PM. especially harmful in RRE, inhibiting radicle growth, which is a more sensitive indicator of toxicity than the percentage of germination.76 The evaporation pad without species (cellulose) used as a control affected more than expected the germination indices. We assume this may be due to the incomplete degra- dation of cellulose during the test. Instead, as mentioned above, the alterations of the cellulose structure during the battery operation and the presence of the reaction products seem to speed up evaporation pad degradation and lead to a more stable, harmless final material. Altogether, the results reported above confirm the FlowER battery’s capability to be aerobically biodegraded according to standardized tests. Furthermore, a nutritious non-toxic com- post was obtained from the biodegradation test, ensuring the harmlessness of the biotic degradation products. 3. Experimental section Chemicals The redox species used for setting up the battery were Fig. 9 Aerobic biodegradability assessment of the FlowER battery, (N = 3). p-benzoquinone (p-BQ) and hydroquinone sulfonic acid potas- (a) Dynamic respiration index (DRI) of evaporation pads with and without sium salt (H2BQS), with oxalic acid and potassium hydroxide the reaction products accumulated over 3 days and compost used as a (KOH) as electrolytes. All of them were purchased from Sigma- blank; and (b) total oxygen consumed by the three samples and level of Aldrich (Sigma Aldrich, St Louis, Missouri, USA) and used as biodegradation achieved at day 18. received, as well as the blue colorant used for the evaporation flow rate characterization, erioglaucine disodium salt. All The levels of biodegradation reached in only 18 days were 66  solutions were prepared in deionized water; 0.1 M p-BQ in 16% and 90  6% for the evaporation pads without and with 0.5 M C2H2O4 and 0.1 M H2BQS in 1 M KOH were used as reactants, respectively. It is noteworthy that the oxygen uptake catholyte and anolyte, respectively. rate of the used evaporation pads was considerably higher than that of the unused. This may be due to physico-chemical Device fabrication changes in the cellulose structure during the battery lifespan The device’s paper-based fluidic system was designed using a or to the presence of the reaction products, both factors CAD program (CorelDRAW, Corel, Ottawa, ON, Canada). The favouring microbial degradation of the evaporation pads. This paper structure was fabricated as follows: due to its low fluidic effect was confirmed when the reactors were opened at day 18 resistance, glass fiber (Standard 14 355 mm thick) is chosen as and non-recognizable parts of the sorbents were found in inlet channels; whereas cellulose (Whatman 1180 mm thick) is evaporation pads with reactants while small fragments from used to fabricate the outlet channels, fluidic connector and un-used pads were observed in the final compost. absorbent/evaporation pad. Whatman 1 material is a well- After the aerobic biodegradation assay, seed germination known cellulosic filter paper, typically used in microfluidics tests were carried out in order to assess the toxicity of the three as an absorbent pad, since it creates a homogeneous capillary generated composts. Germination index (GI), relative root pressure. This material seemed also adequate as the evapora- elongation (RRE), and germination percentage of all composts tion pad because it is one of the thinnest commercially avail- are shown in Table 1. GI values over 100% are considered to able cellulosic filters, a property that allows the reduction of the have a positive effect on germination and root elongation absorbent volume versus its superficial area, resulting in an 2910 | Energy Environ. Sci., 2022, 15, 2900–2915 This journal is © The Royal Society of Chemistry 2022 View Article Online Paper Energy & Environmental Science Table 1 Germination test for compost generated from biodegradation assessments of evaporation pads of the FlowER battery: with and without products in comparison with a compost control and compost with pure reactants added directly. Pictures show the differences between the representative samples analysed. The results have been statistically analysed using the Holm–Sidack pairwise multiple comparison procedure; hence the compost samples have been paired under different letters. Those samples paired under the same letter showed non-significant differences (p o 0.05) (Holm–Sidack Test, Sigmaplot 11) Compost generated from the biodegradability assessment of Germination index (%) Percentage of germination (%) Relative root elongation (%) Control compost 195  75a 100  20a 195  83a Evaporation pads with products 77  44c 81  12b 95  61b Evaporation pads with reaction products 147  46b 86  15b 171  65a This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Control compost with pure reactants 7  5d 52  4c 13  6c Open Access Article. Published on 30 May 2022. Downloaded on 5/11/2023 2:47:37 PM. enhancement of the total liquid–air interface. Both materials and temperature, recorded with a sensor (ALPS Sensor IoT were purchased from GE Healthcare, Pittsburgh, PA, USA. Network Smart Module, Alps Electric Co., Ltd, Osaki, 04, JP). These paper-like materials were cut using a CO2 laser cutter (Mini 24, Epilog Laser, Golden, CO, USA). Porous carbon Electrochemical studies electrodes were cut to size (10  5 mm) from sheets of Toray The effect of storage time on redox species degradation was carbon paper (TGPH-090 E-TEK, 280 mm). For the purpose of characterized via cyclic voltammetry. In the case of different this work, the porous carbon electrodes were thermally pre- storage volumes and deoxygenation studies, the measurements treated to confer them with hydrophilic behaviour. were performed in a three-electrode electrochemical cell. Glassy A compostable and 3D printable filament, COMPOST3D, was carbon electrode (0.071 cm2), platinum electrode, and Ag/AgCl used for building the mechanical structure (B4Plastics, IQ electrode (CH Instruments Inc., TX, USA) were used as the Parklaan 2A, 3650 Dilsen-Stokkem, Belgium), by 3D-printing a working, counter and reference electrodes, respectively. The custom-made model, designed using an open-source software, effect of the sunflower oil protection layer was assessed on Blender (Blender Foundation, Stichting Blender Foundation, carbon SPEs (screen-printed electrodes), which use Ag/AgCl as Buikslotermeerplein 161, 1025 ET Amsterdam, the Netherlands). the reference electrode (fabricated by the authors). The different parts of the prototype were manually assembled The polarization and power curves for battery characteriza- with a water-based bioadhesive, BioTAK S100, a double-sided tion were generated using linear sweep voltammetry technique, PSA formulated to include a high content of renewable materials from premeasured OCV to 0 V at a scan rate of 20 mV s1. The (BioTAK, Sustainable Adhesive Products B.V., Venkelbaan 82, discharging curves were generated using chronopotentiometry NL-2908 KE Capelle aan den IJssel, The Netherlands). technique by holding the battery at different discharging currents. Flow rate characterization All electrochemical measurements were recorded with a DropSens mStat400 bipotentiostat/galvanostat and DropView The experimental evaporation flow rate over time was charac- 8400 Software (DropSens S.L., Asturias, Spain), at a scan rate terized by measuring the decrease of the liquid height in device of 20 mV s1. trays. To enhance image contrast, in these experiments, 100 mM solution of euroglaucine disodium salt dye was used Faradaic and energy efficiencies to simulate and visualize the anolyte and catholyte flow. The images were recorded with a camera (C920, Logitech, Fremont, The faradaic efficiency allows the evaluation of the redox CA, USA) controlled by time-lapse software (Sky Studio Pro, free species utilization, by comparing the total theoretical charge licensed), which saved images at a rate of 1 frame every 20 min. capacity with the one experimentally harvested. It can be The flow rates were calculated from the captured images by computed using the following expression: measuring the height decrease in the trays over time. The Ðt IðtÞdt images were analysed by using ImageJ software (US National ZF ¼ 0 (2) nc0 VF Institutes of Health, Bethesda, Maryland, USA). The theoretical evaporation flow rate over time was char- where I(t) is the current recorded or demanded during the acterized through the environmental conditions of humidity measure, t is the measurement time, n is the number of This journal is © The Royal Society of Chemistry 2022 Energy Environ. Sci., 2022, 15, 2900–2915 | 2911 View Article Online Energy & Environmental Science Paper electrons per mole of reactive specie, c0 stands for the bulk Aerobic biodegradability and phytotoxicity assessment species concentration, V is the total volume that has gone Aerobic biodegradability tests were performed in 0.5 L self- through the electrodes and F is the Faraday constant. made PVC cylindrical reactors (13 cm high  7 cm diameter) On the other hand, the energy efficiency compares the with forced air inlet (at 20.9% O2) and outlet. A constant energy that the battery could theoretically deliver with the flow rate was controlled with a mass flow meter (Mass-Stream power actually delivered under continuous operation. It can D-6311, Bronkhorst, NL) at 15 ml min1.74 The incubator be expressed as follows: (Memmert IF 160, Schwabach, Germany) temperature was set at 58.9 1C to favour the growth of thermophilic microorgan- Ðt This article is licensed under a Creative Commons Attribution 3.0 Unported Licence.  EðtÞdt isms. Compost moisture content was initially adjusted to 54% 0 IðtÞ ZE ¼ (3) and maintained by constant water vapour-saturation of the nc0 VFE inlet air. Three samples were analysed: 60 g of compost from Open Access Article. Published on 30 May 2022. Downloaded on 5/11/2023 2:47:37 PM. industrial plant (Consorci per a la Gestió dels Residus del Vallès where E(t) is the operating voltage and E is the theoretical Oriental, Barcelona, Spain) were used as a control, 60 g of com- Nernst cell voltage. post plus three un-used evaporation pads (0.49 g of cellulose) and 60 g of compost plus three used evaporation pads (1.30  0.05 g of Flower Caret sensor and voltage boost converter cellulose and reaction products accumulated for 3 days). The quantities were set based on proportions proposed by Ruggero A Flower Caret (HHCC-Xiaomi, Beijing, China) wireless plant et al. (2019), slightly reduced due to sample quantity limitations.77 monitoring unit was purchased from an ecommerce platform. The oxygen content at the outlet air flow was monitored The accompanying app was installed on a smartphone running using electrochemical O2–A2 oxygen sensors (Alphasense, UK). Android 11. The sensor was preliminary linked to the app The Dynamic Respiration Index (DRI) was determined following before running experiments. UNI 11184: 2016, To monitor the voltage variations of the FlowER battery   under load, a portable potentiostat/galvanostat was used (Drop- F  O2;in  O2;out P  32  60 DRI ¼  (4) sens mStat400, DropSens S.L., Oviedo, Spain) coupled to the M  100 T  R  %DM Dropview 8400 software installed on a laptop. where DRI (mg O2 h1 g1 DM); F, the air flow rate into the To amplify the initial voltage generated by the FlowER reactor (L min1); O2,in, the percentage of oxygen at the inlet air battery, the evaluation module of an ultra-low-power DC-DC flow (%, mol O2 mol1 O2), in this experiment was set at 20.9; boost converter (BQ25504-EVM, Texas Instruments, Dallas, O2,out, the oxygen concentration at the outlet air flow (%, mol O2 USA) was used in association with a 6.8 mF super-capacitor mol O21); M, the total mass inside the reactor (g); P, atmospheric (BZ05FB682ZSB, AVX). The BQ25504-EVM is equipped with pressure (atm); 32 oxygen molecular weight (%, mol O2 mol1 O2); three terminal blocks: the input source (J1), the charger output 60, conversion factor from minutes to hours; T, ambient tempera- (J5), and the battery connection (J2). The FlowER battery was ture (K); R, ideal gas constant (0.08206 L atm mol1 K1); and connected to J1 after the establishment and stabilization of the %DM, fraction of dry matter (g DM g1).78 evaporation phase (i.e., the evaporation effect became domi- The accumulated oxygen consumed (g O2) depicted in nant over the capillary effect so that the battery was predomi- Fig. 9b is obtained by the integration of DRI and then multi- nantly evaporation-driven). The Flower Caret sensor and the plying by the sample dry mass (g); super-capacitor were connected to J5 and J2, respectively. The ðt functions available on the evaluation board (e.g., maximum DRI dt ¼ gO2  g1 DM (5) power tracking point, under-voltage settings, etc.) were used as 0 provided by the manufacturer. Finally, the biodegradation level (%) was obtained as follows: Electrical signals with a duration ranging from a few milli- Total Oxygen Consumed seconds to several tens of seconds were measured using a %Biodegradability ¼ (6) ThOD digital oscilloscope (DSOX2002A, Agilent Technologies, Santa Clara, USA and Waverunner 44Xi, LeCroy, Chestnut Ridge, USA) The total oxygen consumed (g O2) was calculated by the coupled to a standard probe (P2200, Tektronix). Built-in high- accumulated oxygen uptake of the sample in 18 days while resolution and/or fine modes were used to filter the back- ThOD is the theoretical oxygen demand (g O2). 79 The calculated ground noise and facilitate the visualization of the overall samples’ ThOD are 0.576 g for the evaporation pads without signal shapes. reaction products and 1.408 g for the used evaporation pads. Other electrical signals lasting longer than a few tens of A germination test using radish seeds (Raphanus sativus) was seconds were recorded using a digital multimeter 6.5 digit performed to evaluate compost phytotoxicity.80 A compost resolution (34401A, Hewlett Packard, Palo Alto, USA). The extract (10 mL) in a ratio of 10 : 1 (water volume, in ml, to dry multimeter was connected to a computer using a GPIB-USB- weight, in g) was used to evaluate the germination of 10 radish HS Card (National Instruments, Austin, USA). A block diagram- seeds plated in a Petri dish in triplicate. Germinated seeds and based interface in LabVIEW (National Instruments) was devel- root length were measured after five days in a lighted area at oped to automate the acquisition of the data. room temperature. Seed germination percentage and relative 2912 | Energy Environ. Sci., 2022, 15, 2900–2915 This journal is © The Royal Society of Chemistry 2022 View Article Online Paper Energy & Environmental Science root elongation (RRE) were calculated and expressed as a conceptualization, methodology, validation, resources, writing – percentage about the results of the control using distilled reviewing and editing, supervision, project administration, and water. Germination Index (GI) was calculated considering both funding acquisition. germination percentage and RRE as described by Komilis and Tziouvaras (2009).75 Conflicts of interest There are no conflicts to declare. 4. Conclusions This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. Technological solutions are currently being used to tackle the challenges of digitalized societies and ensure a resilient food Acknowledgements Open Access Article. Published on 30 May 2022. Downloaded on 5/11/2023 2:47:37 PM. system. At the same time, waste electrical and electronic equip- The authors acknowledge financial support from projects ERC- ment is now the fastest growing waste stream worldwide, being CoG SUPERCELL (GA.648518), Ecotronic (RTI2018-102070-B- accumulated in landfills and generating considerable environ- C21, MCIN), BIDEKO (PLEC2021-007801), funded by MCIN/ mental and human health risks. In this context, the FlowER AEI/10.13039/501100011033 and European Union ‘‘NextGener- battery presented herein aims to evince that it is possible to ationEU’’/PRTR’’, and Marie Curie-Skłodowska Individual Fel- redefine technological priorities. By placing sustainability as a lowship under the European Commission’s Horizon 2020 core priority, feasible and efficient solutions can be created while Framework (project 893995,‘‘GREAT-4PA’’). having a neutral or positive environmental impact. Specifically, conceived and eco-designed to follow the lifecycle of agriculture procedures, the FlowER battery mimics plant mechanisms of fluid Notes and references transport to passively move reactants through a paper-based 1 United Nations – Department Economic and Social affairs – fluidic structure. This plant-inspired operation principle enables Population Division – World Population Prospects: High- the creation of a flow-through configuration cell for energy lights, 2019. harvesting. Due to transpiration phenomena, co-laminar flow is 2 A. Matese, F. P. Vaccari, D. Tomasi, S. F. Di Gennaro, maintained and reactants are refreshed on the electrode surface. J. Primicerio, F. Sabatini and S. Guidoni, Sensors, 2013, 13, Thus, the need for external pumps is bypassed, while overcoming 7652–7667. the operation time limitations of capillary-based flow cells. Para- 3 R. Beckwith, D. Teibel and P. Bowen, Proc. IEEE Sens., 2004, meters affecting the evaporation flow rate and battery perfor- 2, 561–564. mance have been studied, leading to an energy autonomy of up to 4 A. Zervopoulos, A. Tsipis, A. G. Alvanou, K. Bezas, four days with the present configuration. Operational time was A. Papamichail, S. Vergis, A. Stylidou, G. Tsoumanis, found to be constrained to the particular selected quinone redox V. Komianos, G. Koufoudakis and K. Oikonomou, Agricul- couple degradation; thus it could be expanded by using more ture, 2020, 10, 1–20. stable redox chemistries. The prototype practicality has been 5 Z. Sai, Y. Fan, T. Yuliang, X. Lei and Z. Yifong, Comput. tested by powering a wireless plant monitoring system, providing Electron. Agric., 2016, 127, 76–86. enough energy for the start-up, passive monitoring, and Bluetooth 6 A. C. Tagarakis, D. Kateris and R. Berruto, Appl. Sci., 2021, communication sequences. Ultimately, the results obtained in 1–13. aerobic biodegradation and germination tests show that the 7 M. A. Parvez Mahmud, N. Huda, S. H. Farjana and C. Lang, depleted FlowER battery could be safely disposed of or composted Batteries, 2019, 5(1), 22. as agricultural waste, resembling the way a plant comes back to 8 R. Morais, S. G. Matos, M. A. Fernandes, A. L. G. Valente, nature at the end of its lifecycle. S. F. S. P. Soares, P. J. S. G. Ferreira and M. J. C. S. Reis, Comput. Electron. Agric., 2008, 64, 120–132. Author contributions 9 Y. Xu, J. Li, Q. Tan, A. L. Peters and C. Yang, Waste Manage., 2018, 75, 450–458. Marina Navarro-Segarra: conceptualization, validation, investiga- 10 B. Bakhiyi, F. Labrèche and J. Zayed, Environ. Int., 2014, 73, tion, data curation, writing – original draft, writing – reviewing 224–234. and editing, visualization, and project administration. Carles 11 A. O. Afolayan, Global Challenges, 2018, 2, 1700090. Tortosa: conceptualization, validation, investigation, data cura- 12 Y. Wang, S. Luo, H. Y. H. Kwok, W. Pan, Y. Zhang, X. Zhao tion, writing – original draft, and visualization. Carlos Ruı́z Dı́ez: and D. Y. C. Leung, Renewable Sustainable Energy Rev., 2021, validation, investigation, data curation, and writing – original 141, 110806. draft. Denis Desmaële: validation, investigation, data curation, 13 E. Kjeang, N. Djilali and D. Sinton, J. Power Sources, 2009, and writing – original draft. Teresa Gea: validation, resources, 186, 353–369. writing – reviewing and editing, and supervision. Raquel Bar- 14 J. W. Lee and E. Kjeang, Biomicrofluidics, 2010, 4, 041301. rena: validation, investigation, data curation, writing – original 15 J. H. Rivera, D. O. Dı́az, D. M. A. Cruz, J. Rodrı́guez- draft, and supervision. Neus Sabaté: conceptualization, valida- Reséndiz, J. M. O. Ramı́rez, A. Dector, D. Dector, tion, resources, and funding acquisition. Juan Pablo Esquivel: R. Galindo and H. E. E. Ponce, Energies, 2020, 13, 1–13. This journal is © The Royal Society of Chemistry 2022 Energy Environ. Sci., 2022, 15, 2900–2915 | 2913 View Article Online Energy & Environmental Science Paper 16 O. A. Ibrahim, M. Navarro-Segarra, P. Sadeghi, N. Sabaté, 39 H. R. Brown, Phys. Perspect., 2013, 15, 320–358. J. P. Esquivel and E. Kjeang, Chem. Rev., 2022, 122(7), 40 T. D. Wheeler and A. D. Stroock, Nature, 2008, 455, 208–212. 7236–7266. 41 H. Kim, K. Kim and S. J. Lee, Sci. Rep., 2016, 6, 1–10. 17 D. Desmaële, T. T. Nguyen-Boisse, L. Renaud and S. Tingry, 42 L. Jingmin, L. Chong, X. Zheng, Z. Kaiping, K. Xue and Microelectron. Eng., 2016, 165, 23–25. W. Liding, PLoS One, 2012, 7(11), DOI: 10.1371/ 18 A. J. Armenta-González, R. Carrera-Cerritos, A. Moreno- journal.pone.0050320. Zuria, L. Álvarez-Contreras, J. Ledesma-Garcı́a, F. M. Cuevas- 43 P. Agrawal, P. S. Gandhi, M. Majumder and P. Kumar, Phys. Muñiz and L. G. Arriaga, Fuel, 2016, 167, 240–247. Rev. Appl., 2019, 12, 1. 19 K. G. Lim and G. T. R. Palmore, Biosens. Bioelectron., 2007, 44 J. M. Li, C. Liu, Z. Xu, K. P. Zhang, X. Ke, C. Y. Li and This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. 22, 941–947. L. D. Wang, Lab Chip, 2011, 11, 2785–2789. 20 D. Desmaële, L. Renaud and S. Tingry, Biomicrofluidics, 45 S. A. Baldwin, S. M. Van Bruggen, J. M. Koelbl, Open Access Article. Published on 30 May 2022. Downloaded on 5/11/2023 2:47:37 PM. 2015, 9(4), DOI: 10.1063/1.4928946. R. Appalabhotla, J. E. Bear and J. M. Haugh, Biomicrofluidics, 21 A. Zebda, L. Renaud, M. Cretin, C. Innocent, F. Pichot, 2021, 15(4), DOI: 10.1063/5.0054764. R. Ferrigno and S. Tingry, J. Power Sources, 2009, 193, 46 S. Mendez, E. M. Fenton, G. R. Gallegos, D. N. Petsev, 602–606. S. S. Sibbett, H. A. Stone, Y. Zhang and G. P. López, Lang- 22 Y. H. Kwok, Y. F. Wang, A. C. H. Tsang and D. Y. C. Leung, muir, 2010, 26, 1380–1385. Appl. Energy, 2018, 217, 258–265. 47 M. Navarro-Segarra, P. P. Alday, D. Garcia, O. A. Ibrahim, 23 A. S. Hollinger, R. J. Maloney, R. S. Jayashree, D. Natarajan, E. Kjeang, N. Sabaté and J. P. Esquivel, ChemSusChem, 2020, L. J. Markoski and P. J. A. Kenis, J. Power Sources, 2010, 195, 201903511. 3523–3528. 48 Z. Liu, J. Hu, Y. Zhao, Z. Qu and F. Xu, Appl. Therm. Eng., 24 Y. Wang and D. Y. C. Leung, Appl. Energy, 2016, 184, 2015, 88, 280–287. 659–669. 49 C. Castro, C. Rosillo and H. Tsutsui, Microfluid. Nanofluid., 25 J. Galindo-De-La-Rosa, A. Álvarez, M. P. Gurrola, J. A. Rodrı́guez- 2017, 21, 1–14. Morales, G. Oza, L. G. Arriaga and J. Ledesma-Garcı́a, ACS 50 K. M. Schilling, A. L. Lepore, J. A. Kurian and Sustainable Chem. Eng., 2020, 8, 10900–10910. A. W. Martinez, Anal. Chem., 2012, 84, 1579–1585. 26 D. Desmaële, L. Renaud and S. Tingry, Sens. Actuators, B, 51 Z. Liu, X. He, J. Han, X. Zhang, F. Li, A. Li, Z. Qu and F. Xu, 2015, 220, 583–589. Microfluid. Nanofluid., 2018, 22, 1–20. 27 E. Fu, S. A. Ramsey, P. Kauffman, B. Lutz and P. Yager, 52 O. A. Ibrahim, P. Alday, N. Sabaté, J. P. Esquivel and Microfluid. Nanofluid., 2011, 10, 29–35. E. Kjeang, J. Electrochem. Soc., 2017, 164, A2448–A2456. 28 W. Batchelor and D. Söderberg, Advances in Pulp and Paper 53 Y. Ding and G. Yu, Angew. Chem., Int. Ed., 2016, 55, 4772–4776. Research, Oxford 2017, Pulp & Paper Fundamental Research 54 Y. Ding, C. Zhang, L. Zhang, Y. Zhou and G. Yu, Chem. Soc. Society, 2017. Rev., 2018, 47, 69–103. 29 J. P. Esquivel, F. J. Del Campo, J. L. Gómez De La Fuente, 55 M. Tanveer and K. Y. Kim, Energies, 2021, 14, 3381. S. Rojas and N. Sabaté, Energy Environ. Sci., 2014, 7, 1744–1749. 56 A. v Razimov, F. T. Bekmashi and B. I. Liogon’kii, Polym. Sci. 30 C. W. Narváez Villarrubia, C. Lau, G. P. M. K. Ciniciato, U.S.S.R., 1975, 17, 3164–3170. S. O. Garcia, S. S. Sibbett, D. N. Petsev, S. Babanova, 57 K. Wedege, E. Dražević, D. Konya and A. Bentien, Sci. Rep., G. Gupta and P. Atanassov, Electrochem. Commun., 2014, 2016, 6, 1–13. 45, 44–47. 58 B. Yang, L. Hoober-Burkhardt, S. Krishnamoorthy, A. Murali, 31 S. C. Barton, J. Gallaway and P. Atanassov, Chem. Rev., 2004, G. K. S. Prakash and S. R. Narayanan, J. Electrochem. Soc., 104, 4867–4886. 2016, 163, A1442–A1449. 32 M. Wallenstein, S. D. Allison, J. Ernakovich, J. M. Steinweg 59 L. Li, Y. He, Q. Xu, T. Liu, S. Bei, K. Zheng, J. Yang, H. Wang and R. Sinsabaugh, in Soil enzymology, ed. G. Shukla and and M. K. H. Leung, Int. J. Energy Res., 2022, 46, 8389–8397. A. Varma, 2010, pp. 245–258. 60 M. Tanveer, A. Ambreen, H. Khan, G. M. Kim and C. W. 33 J. P. Esquivel, P. Alday, O. A. Ibrahim, B. Fernández, Park, Energy Convers. Manage., 2022, 264, 115732. E. Kjeang and N. Sabaté, Adv. Energy Mater., 2017, 7(18), 61 T. S. Copenhaver, K. H. Purohit, K. Domalaon, L. Pham, DOI: 10.1002/aenm.201700275. B. J. Burgess, N. Manorothkul, V. Galvan, S. Sotez, F. A. 34 N. Mittal, A. Ojanguren, M. Niederberger and E. Lizundia, Gomez and J. L. Haan, Electrophoresis, 2015, 36, 1825–1829. Adv. Sci., 2021, 2004814. 62 A. Fraiwan, S. Mukherjee, S. Sundermier, H. S. Lee and 35 E. di Mauro, D. Rho and C. Santato, Nat. Commun., 2021, S. Choi, Biosens. Bioelectron., 2013, 49, 410–414. 12, 3167. 63 L. L. Shen, G. R. Zhang, T. Venter, M. Biesalski and 36 M. H. Lee, J. Lee, S. Jung, D. Kang, M. S. Park, G. D. Cha, B. J. M. Etzold, Electrochim. Acta, 2019, 298, 389–399. K. W. Cho, J. Song, S. Moon, Y. S. Yun, S. J. Kim, Y. W. Lim, 64 M. Irimia-Vladu, Chem. Soc. Rev., 2014, 43, 588–610. D. Kim and K. Kang, Adv. Mater., 2021, 33(10), 2004902. 65 E. di Mauro, D. Rho and C. Santato, Nat. Commun., 2021, 12, 37 G. W. Koch, S. C. Stillet, G. M. Jennings and S. D. Davis, 1–10. Nature, 2004, 428, 851–854. 66 A. Zvezdin, E. di Mauro, D. Rho, C. Santato and M. Khalil, MRS 38 E. Steudle, Nature, 1995, 378, 663–664. Energy Sustainability, 2020, 7(1), DOI: 10.1557/MRE.2020.16. 2914 | Energy Environ. Sci., 2022, 15, 2900–2915 This journal is © The Royal Society of Chemistry 2022 View Article Online Paper Energy & Environmental Science 67 J. Xu, B. Wang, W. H. Zhang, F. J. Zhang, Y. D. Deng, 74 L. Mejias, D. Komilis, T. Gea and A. Sánchez, Waste Manage., Y. Wang, J. J. Gao, Y. S. Tian, R. H. Peng and Q. H. Yao, 2017, 65, 22–28. AMB Express, 2021, 11, 1–8. 75 D. P. Komilis and I. S. Tziouvaras, Waste Manage., 2009, 29, 68 J. Devillers, P. Boule, P. Vasseur, P. Prevot, R. Steiman, 1504–1513. F. Seigle-Murandi, J. L. Benoit-Guyod, M. Nendza, C. Grioni, 76 Y. Luo, J. Liang, G. Zeng, M. Chen, D. Mo, G. Li and D. Dive and P. Chambon, Ecotoxicol. Environ. Saf., 1990, 19, D. Zhang, Waste Manage., 2018, 71, 109–114. 327–354. 77 F. Ruggero, R. Gori and C. Lubello, Waste Manage. Res., 69 F. J. Enguita and A. L. Leitão, BioMed Res. Int., 2013, DOI: 2019, 37, 959–975. 10.1155/2013/542168. 78 UNI 11184:2016, Waste and refuse derived fuels – Determi- This article is licensed under a Creative Commons Attribution 3.0 Unported Licence. 70 J. Pérez, J. Muñoz-Dorado, T. de La Rubia and J. Martı́nez, nation of biological stability by dinamic respirometric Int. Microbiol., 2002, 5, 53–63. index, 2016. Open Access Article. Published on 30 May 2022. Downloaded on 5/11/2023 2:47:37 PM. 71 P. Béguin and J.-P. Aubert, FEMS Microbiol. Rev., 1994, 13, 79 EN ISO 17556:2020, Plastics – Determination of the ultimate 25–58. aerobic biodegradability of plastic materials in soil by 72 BS EN 13432:2000, Packaging-requirements for packaging measuring the oxygen demand in a respirometer or the recoverable through composting and biodegradation: test amount of carbon dioxide evolved. scheme and evaluation criteria for the final acceptance of 80 C. Ballardo, M. del, C. Vargas-Garcı́a, A. Sánchez, R. Barrena packaging. and A. Artola, Waste Manage., 2020, 106, 32–43. 73 R. Barrena Gómez, F. Vázquez Lima and A. Sánchez Ferrer, 81 A. D. Stroock, V. V. Pagay, M. A. Zwieniecki and N. M. Waste Manage. Res., 2006, 24, 37–47. Holbrook, Annu. Rev. Fluid Mech., 2014, 46, 615–642. This journal is © The Royal Society of Chemistry 2022 Energy Environ. Sci., 2022, 15, 2900–2915 | 2915