Electrolysis of Liquefied Biomass for Sustainable Hydrogen and Organic Compound Production: A Biorefinery Approach

Abstract

Liquefaction is an effective thermochemical process for converting lignocellulosic biomass into bio-oil, a hydrocarbon-rich resource. This study explores liquefied biomass electrolysis as a novel method to promote the electrocracking of organic molecules into value-added compounds while simultaneously producing hydrogen (H₂). Key innovations include utilizing water from the liquefaction process as an electrolyte component to minimize industrial waste and incorporating carbon dioxide (CO₂) into the process to enhance decarbonization efforts and generate valuable byproducts. The electrolysis process was optimized by adding 2 M KOH, and voltammetric methods were employed to analyze the resulting emulsion. The experimental conditions, such as the temperature, anode material, current type, applied cell voltage, and CO₂ bubbling, were systematically evaluated. Direct current electrolysis at 70 °C using nickel electrodes produced 55 mL of H₂ gas with the highest Faradaic (43%) and energetic (39%) efficiency. On the other hand, pulsed electrolysis at room temperature generated a higher H₂ gas volume (102 mL) but was less efficient, showing 30% Faradaic and 11% energetic efficiency. FTIR analysis revealed no significant functional group changes in the electrolyte post-electrolysis. Additionally, the solid deposits formed at the anode had an ash content of 36%. This work demonstrates that electrocracking bio-oil is a clean, sustainable approach to H₂ production and the synthesis of valuable organic compounds, offering significant potential for biorefinery applications.

1. Introduction

The term “biomass” primarily refers to organic matter derived from plants. Biomass is produced by green plants that transform sunlight into plant material through photosynthesis [1]. Lignocellulosic biomass is the most abundant plant material and, therefore, widely available in high quantities and at a low cost [2]. However, when biomass is used as a raw material, it is crucial to implement reforestation programs because it is not an inexhaustible energy source [1]. Biomass meets various energy needs, such as electricity generation, domestic heating, vehicle fuel, and industrial heating. Despite these applications, biomass is primarily used for direct combustion [3].

The biomass liquefaction process has been introduced as an effective thermochemical method to convert lignocellulosic biomass into liquid products using a solvent [4]. A key goal of liquefaction is to reduce the oxygen content in biomass, which constitutes 40–50 wt.% of its composition [5]. The resulting bio-oil has a complex composition containing various oxygenated compounds, including acetic acid, ethanol, carboxylic acids, phenols, aldehydes, ketones, alcohols, and esters [3]. Therefore, bio-oil can serve as a biofuel and can also be used to produce bio-polyols, antioxidants, and other chemicals derived from hydrocarbons and lignin [6].

Hydrogen (H₂) is considered a clean fuel for the future, as its combustion generates only water. Moreover, it has an energy density approximately 2.75 times higher than hydrocarbon fuels like gasoline, offering low pollution and versatile applications [3,7]. Water electrolysis is one of the most promising technologies for producing H₂ [7]. This process induces the dissociation of water into H₂ and oxygen (O₂) by applying an electric current [8]. Water electrolysis has been extensively studied due to its potential to enable scalable H₂ production, particularly when coupled with renewable energy sources. In this case, the resulting H₂ is classified as green H₂ [9]. Additionally, among existing carbon dioxide (CO₂) capture technologies, the electrochemical reduction of CO₂ has recently garnered the most interest [10]. This process can capture CO₂ while producing value-added products such as carbon monoxide (CO), formic acid (HCOOH), formaldehyde (HCHO), and methane (CH₄) [11]. When water electrolysis is combined with CO₂ reduction, the resulting co-electrolysis process typically produces syngas, a versatile fuel comprising CO and H₂ that is widely used for energy generation and chemical synthesis [12].

Despite the apparent simplicity of the water electrolysis process, several challenges are delaying the large-scale production of green H₂. These include the large overpotentials required for operating the electrolyzer, thereby increasing electricity costs, or concerns regarding the use of freshwater, which directly competes with agricultural and human consumption. To overcome these limitations, researchers are focusing on new approaches, such as seawater and biomass electrolysis. The former is an exciting alternative, as it leverages abundant seawater and the large-scale deployment of offshore renewable energy; however, current commercial electrolyzers require the prior desalination of seawater, increasing the process costs and creating environmental issues concerning the disposal of concentrated brine [13].

To avoid desalination, researchers are exploring direct seawater electrolysis as a cost-effective method for green H₂ production. A significant challenge remains in developing durable, earth-abundant electrocatalysts that withstand corrosive conditions. Yu et al. have addressed this by fabricating self-supported nickel–iron phosphosulfide (NiFeSP) nanotube array electrocatalysts with excellent activity and durability for hydrogen and oxygen evolution reactions (HER and OER) in simulated alkaline seawater. Additionally, the NiFeSP electrode exhibited promising activity during urea oxidation as an alternative anodic process during hybrid seawater electrolysis [14].

The oxidation of organic matter at the anode allows the replacement of the kinetically hindered OER and the “corrosive” chlorine evolution reaction as the anodic process, with the additional advantage of reducing the voltage required for electrolysis [14,15]. For this reason, researchers are investigating hybrid seawater electrolysis as an innovative and promising approach for green H₂ production, as its more thermodynamically favorable anodic oxidation reaction reduces the electrolysis energy demand while enabling the co-production of valuable chemicals or the elimination of environmental pollutants [15].

Accordingly, a key motivation for studying biomass electrolysis is that the oxidation of oxygenated organic species from biomass can reduce the electrolysis cell voltage below 1.23 V. Various organic molecules can be utilized in the anodic process, enabling the direct valorization of agricultural and forestry waste (lignocellulosic materials) and by-products from biofuel industries [16]. Recently, Umer et al. carried out FeCl₃-mediated biomass electrolysis using different biomass feedstocks (i.e., glucose, starch, lignin, and cellulose) to produce H₂ at ambient temperature using a cell voltage as low as 1.20 V. It was demonstrated that the process could produce H₂ with a higher yield rate and lower electricity consumption than conventional water electrolysis [17].

The term “electrocracking” has also been introduced as a novel concept linking H₂ production with the electrolysis of liquefied biomass. This process involves decomposing organic compounds in the liquefied biomass into lighter hydrocarbons by passing an electric current between two electrodes [18]. Building on this concept, the present study integrates electrocracking with a patented biomass liquefaction technology to explore sustainable pathways for H₂ and organic compound production. Specifically, it investigates the electrolysis of biomass liquefied through an industrial liquefaction process, incorporating water recovered from that process and subsequent CO₂ capture.

The liquefied biomass was supplied by the Portuguese cement group CMP–Cimentos Maceira e Pataias (Secil), which holds a patent for its biomass liquefaction process, named ENERGREEN [19]. The central objective of this pilot-scale facility is to produce a clean and sustainable biofuel with a high calorific value for use in cement kilns, effectively replacing fossil fuels. The feedstock used was Eucalyptus globulus sawdust. Through acid liquefaction at 160 °C and atmospheric pressure, and by using 2-ethyl-hexanol as the solvent and p-toluenesulfonic acid as the catalyst, the process yields liquefied biomass, polyols, condensates, extraction water, and sugars during the reaction. The final step involves separating the liquefied material into aqueous and organic phases. The aqueous phase is collected in two stages: first, through the condensation of vapors produced during the liquefaction reaction, yielding condensed water, and second, by the liquid–liquid extraction of the biofuel, which also generates an aqueous fraction.

After extracting the sugars, the liquefied biomass was subjected to electrolysis trials, wherein operating parameters such as the temperature, anode material, current type, cell voltage, and the effect of CO₂ sparging were evaluated. Furthermore, fundamental electrochemical analyses, including cyclic voltammetry (CV) and linear scan voltammetry (LSV), were conducted to characterize the redox behavior and stability of the samples under electrolytic conditions. By examining these parameters and performing detailed electrochemical analyses, the present work aims to advance the understanding of biomass electrolysis as a viable pathway for sustainable H₂ and organic compound production, contributing to the development of integrated biorefinery systems.

2. Materials and Methods

2.1. Electrolyte Preparation and Characterization

The liquefied biomass used in this study was supplied directly by CMP–Cimentos Maceira e Pataias (Secil, Portugal) cement group, which produced it from Eucalyptus globulus sawdust using their patented process [19]. Previous studies have demonstrated that the ionic conductivity of liquefied biomass solutions can be significantly enhanced by preparing an alkaline emulsion suitable for electrochemical analysis [20]. Based on these findings, this study utilized an emulsion of Eucalyptus globulus sawdust liquefied biomass with 2 M KOH aqueous solution (50 vol.%). Tests were also performed using waters from the two distinct phases of the liquefaction process (condensed water and two-phase water). The phases of the two-phase water system were named the light-colored phase and the dark phase.

For both electrolytes, the values of conductivity ($\sigma$) and pH were registered. Additionally, the density ($\rho$) of the emulsion (liquefied biomass + 2 M KOH) was measured, and a humidity content test was conducted using a VWR (IT 1400108) balance. A HANNA Instruments (pH20) device was used for conductivity and pH measurements, and a pycnometer was used for the density measurement.

2.2. Electrochemical Measurements

Cyclic voltammetry (CV) and linear scan voltammetry (LSV) measurements were performed using an ALS 2325 bipotentiostat in a 60 mL electrolytic cell with a thermostatic water jacket for temperature control. A typical three-electrode arrangement was used, comprising a platinum (Pt) disk (A = 0.1256 cm²) as the working electrode, a Pt mesh (A = 50 cm²) as the counter electrode, and a saturated calomel electrode (SCE) as the reference electrode. Unless otherwise noted, all potentials are given alongside the SCE reference. For the hydrogen evolution reaction (HER) studies conducted via LSV, two working electrodes were utilized: a Pt electrode with an area of 1 cm² and a nickel (Ni) electrode (A = 22.4 cm²). In CV experiments, the potential was scanned between the open-circuit potential (OCP) and 0.6 V at scan rates ranging from 5 to 1000 mV s^{− 1}. Measurements were conducted at room temperature (RT = 17 °C) and 65 °C. For LSV studies focusing on the HER, the potential was scanned at 1 mV s⁻¹ from the OCP to −2 V, with tests performed at 25 °C and 70 °C.

2.3. Electrolysis Tests

The electrolysis tests were performed in a small-scale laboratory electrolyzer (300 mL) using PAR 273A computer-controlled potentiostat/galvanostat (Princeton Applied Research, Inc., Oak Ridge, TN, USA) and the associated PowerSUITE package to control the experiments. Gas quantification was performed using a Ritter MilliGascounter (MGC-1 V3.4 PMMQA, Ritter, Bochum, Germany) apparatus. Most electrolysis tests used identical Ni plates (A = 22.4 cm²) as the cathode and anode. Additional tests were carried out with a Ni plate as the cathode and a TiO₂-RuO₂ alloy anode.

Besides the regular direct current (DC) electrolysis tests, other tests included using an AMI-7001S anion-exchange membrane (Membranes International Inc., Ringwood, NJ, USA), bubbling CO₂ into the electrolyte during electrolysis, and applying alternating current (AC) waves. All these tests used Ni electrodes. Specifically, electrolysis using AC explored two types of wave shapes: half-sinusoidal AC waves and pulsating waves. The wave shape and electrolysis parameters were verified on a TDS 1012B oscilloscope (Tektronix, Beaverton, OR, USA). For the first type of wave, an alternating power supply (230/12 V, 1 A) was used with a diode (ultra-fast rectifier BYV32) to rectify the sinusoidal wave. A microcontroller based on an Atmega328 processor was used for pulse generation, and a code in C language was programmed to control the pulse frequency and duty cycle. The cell voltage applied in these studies was 1.5 V, 2.4 V, and 3 V; temperatures of 25, 50, 60, and 70 °C were tested.

3. Results and Discussion

3.1. Electrolyte Characterization

Table 1. Values of conductivity, pH, density, and humidity for the emulsion prepared and the waters from the liquefaction process.

	$\mathsf{\sigma }$, mS cm−1	pH	$\mathsf{\rho }$, g dm−3	Humidity, %
Emulsion	115	13.8	1123	78
Condensed water	8.73 × 10−3	2.51	−	−
Light-colored phase	0.01 × 10−3	1.53	−	−
Dark phase	11.2	2.09	−	−

lists the conductivity, pH, density, and humidity data measured for the used solutions.

3.2. Electrochemical Studies

3.2.1. Cyclic Voltammetry

Figure 1, Figure 2, Figure 3 and Figure 4 present the CVs on the 50/50 emulsion of liquefied biomass + 2 M KOH run with a Pt electrode at 17 and 65 °C, respectively. Figure 1 shows two oxidation peaks, a₁ at −0.2 V and a₂ at 0.28 V, and one reduction peak, c₁ at −0.45 V. It is also notable that peak a₂ becomes ill-defined as the scan rate increases.

Several hypotheses could explain the presence of these three peaks: (i) the oxidation of two different species, one of which is later reduced; (ii) a two-step oxidation of the same species, which is subsequently reduced; or (iii) the oxidation of one species that undergoes a chemical reaction, generating a different species that is oxidized (corresponding to peak a₂) and then reduced. Regardless of these possibilities, a more detailed investigation of peak c₁ is necessary, as it may be independent of the oxidation peaks, as further elaborated below. To analyze the effect of the scan rate on the system, Figure 2 presents three CVs recorded at different scan rates (5, 50, and 200 mV s⁻¹).

Figure 2A shows that, at a low scan rate (5 mV s⁻¹), in addition to the two oxidation peaks (a₁ and a₂), there is a third oxidation peak (c₂ at −0.6 V) in the backscan. This third anodic peak could result from a diffusional phenomenon. Specifically, after the second anodic peak (a₂), the low scan rate allows sufficient time for the oxidized species on the electrode’s surface to diffuse into the bulk solution. Consequently, new species susceptible to oxidation can reach the electrode’s surface. The disappearance of peak c₂ and the appearance of peak c₁ at higher scan rates (Figure 2B,C) can be explained conversely: when increasing the scan rate, after the second oxidation peak (a₂), the oxidized species lacks sufficient time to diffuse into the bulk solution, undergoing reduction as more negative potentials are reached.

Figure 2. CVs of a 50/50 emulsion of liquefied biomass + 2 M KOH using a Pt working electrode at room temperature (17 °C) and applying scan rates of (A) 5 mV s⁻¹, (B) 50 mV s⁻¹, and (C) 200 mV s⁻¹.

Figure 2. CVs of a 50/50 emulsion of liquefied biomass + 2 M KOH using a Pt working electrode at room temperature (17 °C) and applying scan rates of (A) 5 mV s⁻¹, (B) 50 mV s⁻¹, and (C) 200 mV s⁻¹.

Going back to Figure 1, the presence of peak c₁ can indicate a possible reversible reaction. In other words, peak c₁ can correspond to the inverse process (i.e., a reduction) of one of the anodic peaks, a₁ or a₂. According to Pacheco et al., when a process is electrochemically reversible, the anodic peak potential does not change with the scan rate [21]. Thus, it would be intuitive to say that the oxidation process in peak a₂ is reversible since its potential is always the same, unlike peak a₁. However, Elgrishi et al. explain that reversibility depends on the scan rate, i.e., for sufficiently high scan rates, all processes can appear as irreversible [22]. Therefore, at low scan rates (Figure 2A), the reduction peak (c₁) is not present; thus, it cannot be concluded that the oxidation process from peak a₂ is reversible and corresponds to peak c₁.

To disprove the hypothesis that peak a₁ is reversible, CVs at different scan rates were run until an upper potential limit of 0.1 V was reached, not reaching the potential necessary for the process represented by peak a₂ to occur (Figure 3).

Figure 3. CVs of a 50/50 emulsion of liquefied biomass + 2 M KOH using a Pt working electrode and applying scan rates from 0.5 to 1000 mV s⁻¹ at room temperature (17 °C).

Figure 3. CVs of a 50/50 emulsion of liquefied biomass + 2 M KOH using a Pt working electrode and applying scan rates from 0.5 to 1000 mV s⁻¹ at room temperature (17 °C).

Figure 3 clearly shows peak a₁ at −0.2 V; however, peak c₁ is not present. This fact reinforces the assumption that peak c₁ is independent of the oxidation of the species represented in peak a₁. As there is no clear indication of the reversibility of any of the peaks and their dependency on the scan rate, it was decided that the electrochemical equation for irreversible processes would be applied to calculate the relevant kinetic parameters. These uncertainties are expected because irreversibility is more prevalent than reversibility in some redox pairs, including organics [23]. Furthermore, Espinoza et al. claim that oxidative or reductive degradation, dimerization, reaction with solvent, fast mass transport, and other processes that diminish the quantity of the species produced on the electrode’s surface have a strong influence on the CVs, causing asymmetries and even the complete elimination of the anodic and cathodic peaks [23].

The oxidation and reduction peak’s current dependency on the scan rate for irreversible processes is given by Equations (1) and (2), respectively,

$${\mathrm{i}}_{\mathrm{p}}=0.496\sqrt{(1-\mathsf{\alpha }){\mathrm{n}}_{\mathrm{a}}} \times \mathrm{n}\mathrm{F}\mathrm{A}\mathrm{C}\sqrt{\left(\mathrm{F}\mathrm{D}\mathsf{\nu }\right)/\left(\mathrm{R}\mathrm{T}\right)}$$

(1)

$${\mathrm{i}}_{\mathrm{p}}=0.496\sqrt{\mathsf{\alpha }{\mathrm{n}}_{\mathrm{a}}}\times \mathrm{n}\mathrm{F}\mathrm{A}\mathrm{C}\sqrt{\left(\mathrm{F}\mathrm{D}\mathsf{\nu }\right)/\left(\mathrm{R}\mathrm{T}\right)}$$

(2)

where i_p is the peak current, n is the number of exchanged electrons, F is Faraday’s constant (96,485 C mol⁻¹), A is the Pt disk working electrode’s area (0.1256 cm²), C is the concentration of electrochemically active species in the solution (3.34 × 10⁻⁵ mol cm⁻³), D is the diffusion coefficient of that species (6.98 × 10⁻⁹ cm² s⁻¹), α is the charge transfer coefficient, n_a is the number of electrons involved in the rate-determining step (being 1 the most likely value), $\mathsf{\nu }$ is the scan rate in V s⁻¹, R is the universal gas constant (8.314 J K⁻¹ mol⁻¹), and T is the temperature in K. Equations (1) and (2) were used to calculate the number of exchanged electrons, n, in the processes represented by peaks a₁, a₂, and c₁, following the determination of the charge transfer coefficient, α. The α coefficient is usually interpreted as being a measure of the position of the transition state of the reaction; in this case, the reduction reaction, and the oxidation reaction, this coefficient is defined as 1 − α. When the α value is near 1, this implies the transition state is “product-like”, and similarly, a value close to zero indicates a “reactant-like” transition state [24]. Thus, Equations (3) and (4) were used to calculate the anodic and cathodic charge transfer coefficients, respectively, for an irreversible redox process [25,26]. The results obtained for n and α at room temperature are shown in

Table 2. Kinetic parameters of the redox processes in the bio-oil emulsion using a Pt electrode at room temperature and 65 °C.

Temperature (°C)	Kinetic Parameters	Peak a1	Peak a2	Peak c1
RT (17 °C)	α	0.61	0.63	0.96
	n	3.1	4.7	2.0
65 °C	α	0.57	−	0.93
	n	4.6	−	1.0

.

$${\mathrm{E}}_{\mathrm{p}}=\left(\mathrm{R}\mathrm{T}\right)/\left(2\right(1-\mathsf{\alpha }\left){\mathrm{n}}_{\mathrm{a}}\mathrm{F}\right)\times \ln \left(\mathsf{\nu }\right)+\mathrm{b}$$

(3)

$${\mathrm{E}}_{\mathrm{p}}=-\left(\mathrm{R}\mathrm{T}\right)/\left(2\mathsf{\alpha }{\mathrm{n}}_{\mathrm{a}}\mathrm{F}\right)\times \ln \left(\mathsf{\nu }\right)+\mathrm{b}$$

(4)

To study the effect of temperature, CVs were also run at 65 °C (Figure 4). Only one anodic peak (a₁), at −0.1 V, can be seen in Figure 4, with a cathodic peak (c₁) at ca. −0.4 V. The presence of a single anodic peak is probably due to the coalescence of peaks a₁ and a₂, shown in Figure 1. The peaks’ coalescence should be due to the similarity of oxidation peak potentials at higher temperatures. Since the anodic peak in Figure 4 has a similar potential to peak a₁ in Figure 1, it is plausible that this peak is not influenced by temperature. Considering this hypothesis, it is possible to claim that peak a₂ from Figure 1 shifts to more negative potentials, towards peak a₁. This shift in peak a₂ indicates that the oxidation of the species corresponding to peak a₂ becomes easier at higher temperatures.

Figure 4. CVs of a 50/50 emulsion of liquefied biomass + 2 M KOH using a Pt working electrode and scan rates of 5 to 1000 mV s⁻¹ at 65 °C.

Figure 4. CVs of a 50/50 emulsion of liquefied biomass + 2 M KOH using a Pt working electrode and scan rates of 5 to 1000 mV s⁻¹ at 65 °C.

Equations (1)–(4) were also used to calculate the kinetic parameters (n and α) from the data in Figure 4 for the processes denoted as a₁ and c₁ in Figure 5 (Table 2). To better illustrate the influence of the scan rate on the system at 65 °C, CVs at 5, 50, and 200 mV s⁻¹ are shown in Figure 5.

The system at 65 °C (Figure 5) exhibits a similar trend to the one observed at 17 °C (Figure 2), with the main difference being the disappearance of peak a₂ and the broadening of peak a₁. This can be clearly seen in Figure 6, which shows CVs at two different scan rates to further compare the differences between the results obtained at room temperature (17 °C) and 65 °C. Furthermore, Figure 6 shows that the temperature increase also significantly increases the current density, demonstrating that it is beneficial to carry out liquified biomass electrolysis at temperatures higher than room temperature.

Lastly, Figure 7 shows CVs when using two different scan rates (10 and 50 mV s⁻¹) for the 50/50 emulsion of liquefied biomass + 2 M KOH before and after a 3 h electrolysis test. It shows that an increase in the current density is recorded with the emulsion after electrolysis and a slight shift in the oxidation peaks towards more positive potentials.

The observed current density increase agrees with previous studies using liquefied biomass from cork [27]. This behavior can be explained by the fact that some of the organic compounds present in the liquified biomass adsorb during electrolysis into the electrode’s surface; hence, the solution resulting from the electrolysis process will have a lower organic load, thus increasing the emulsion’s conductivity.

The second event, i.e., the shift in the oxidation peaks towards more positive potentials, was also expected, as the organic components of the emulsion are more oxidized after electrolysis. Thus, a higher potential is necessary to oxidize the remaining species. Furthermore, Figure 7 shows a third oxidation peak in the backscan of the CV of the emulsion after electrolysis. This peak could be due to the increased ionic conductivity of the solution, which indicates an increase in the diffusion coefficient, thereby increasing the transportation rate of new species susceptible to oxidation to the electrode surface.

3.2.2. Hydrogen Evolution Reaction Study

The HER was studied through LSV (Figure 8) using Pt and Ni electrodes, and then the Tafel equation (Equation (5)) was applied to the experimental data,

$$\mathsf{\eta }=\mathrm{b}\times \log \left(\mathrm{j}\right)+\mathrm{a}=\left(2.3\mathrm{R}\mathrm{T}\right)/\left(\mathsf{\alpha }\mathrm{F}\right)\times \log \left(-\mathrm{j}\right)+\left(2.3\mathrm{R}\mathrm{T}\right)/\left(\mathsf{\alpha }\mathrm{F}\right)\times \log \left(-{\mathrm{j}}_0\right)$$

(5)

where η is the overpotential, j is the current density, and j₀ is the exchange current density, which is defined as the current density that flows equally in equilibrium and in both ways, resulting in a zero net current and a zero net reaction rate [28]. The LSV plots recorded with Pt and Ni electrodes at 25 and 70 °C are presented in Figure 8, with the corresponding Tafel plots shown in Figure 9. The kinetic parameters obtained when applying Tafel analysis to these data are listed in

Table 3. Kinetic parameters obtained from Tafel analysis using Pt and Ni working electrodes.

		25 °C	70 °C
Pt	b (mV dec−1)	284	319
	α	0.21	0.21
	j0 (mA cm−2)	6.59 × 10−2	1.58
Ni	b (mV dec−1)	180	515
	α	0.33	0.11
	j0 (mA cm−2)	6.30 × 10−2	3.14

.

Table 3 shows that an increase in temperature leads to an increase in both the Tafel slope (b) and the j₀ value. Although it would be desirable to see an increase in j₀ together with a decrease in the Tafel slope, the observed behavior is not uncommon. Moreover, Ghatak et al. reported the Tafel slopes obtained with Pt electrodes at 25 °C during HER in alkaline electrolyte and in eucalyptus black liquor at 113 and 135 mV dec⁻¹, and j₀ values of 1.08 and 9.86 mA cm⁻², respectively [29]. Comparing these j₀ values to the ones obtained in the present work, it can be concluded that the HER in these conditions (25 °C and Pt electrode) in a 50/50 emulsion of liquefied biomass + 2 M KOH is kinetically less favorable than in alkaline water or in eucalyptus black liquor, as the herein obtained j₀ value is significantly lower (6.59 × 10⁻² mA cm⁻²). However, the j₀ value significantly increases (1.58 mA cm⁻²) when increasing the temperature to 70 °C, reaching a value slightly above that of conventional alkaline water electrolysis [29]. This has also been verified for the j₀ values obtained when using the Ni electrode. According to Shinagawa et al., the Tafel slopes using Ni electrodes typically vary between 200 and 1000 mV dec⁻¹ due to the formation of Ni hydrides that limit the HER [30]. The Tafel slopes obtained herein with the Ni electrode at both temperatures fall within this range. Furthermore, the Tafel slopes for all cases are higher than 120 mV dec⁻¹, meaning that Volmer’s step limits HER in this system [29,31].

3.3. Electrolysis Studies

This section of the work explores the electrolysis of the bio-oil alkaline emulsion using two different current modes, namely DC and AC (half-sinusoidal and pulsed waves) electrolysis.

3.3.1. Direct Current Electrolysis

Preliminary electrolysis tests using chronoamperometry were carried out to evaluate the most influencing variables during bio-oil alkaline emulsion electrolysis (Figure 10). These parameters were the applied cell voltage (Figure 10A), the composition of the anode material (Figure 10B), and the temperature (Figure 10C).

Figure 10A shows that increasing the cell voltage by 0.9 V, from 1.5 to 2.4 V, increases the current density from ca. 0 to 0.6 mA cm⁻² in the stationary state. Although working at low overpotentials is usually an advantage in terms of energetic costs, these results suggest that applying a cell voltage of 2.4 V significantly benefits bio-oil alkaline emulsion electrolysis, maintaining reasonable currents even after 180 min. Figure 10B shows that Ni anodes are considerably better than TiO₂-RuO₂ anodes for this electrolysis; using Ni as both the cathode and anode generated higher current densities, making it the preferred electrode material. Figure 10C shows the influence of temperature on bio-oil alkaline emulsion electrolysis using Ni electrodes. Notably, when the temperature is increased from 25 to 50 °C, the current density in the stationary state almost doubles, rising from 2 to 7 mA cm⁻². Therefore, running electrolysis at temperatures higher than 25 °C is clearly beneficial.

Finally, a polarization study was conducted to evaluate the optimum electrolysis cell operating potential, considering that one goal is producing H₂. Figure 11, together with the previous LSV studies (Figure 8), confirms that for cathode polarization to reach −1.5 V and to obtain reasonable currents for the HER, it would be necessary to apply at least 2 V to the bio-oil electrolysis cell.

Subsequently, tests with Ni electrodes were conducted to optimize the electrolysis process. Results from Figure 10 and additional experiments revealed that at 2.4 V and 25 °C, no gas production occurred. Therefore, the cell potential was increased to 3 V, and the temperature was raised to 70 °C, as shown in Figure 12.

Figure 12 shows that gas production follows the same behavior as the current density, i.e., gas is produced (55 mL) during the first 80 min until the current density reaches stabilization; when the current density values stabilize, the gas production stops. To perform a quantitative analysis of these results, the Faradaic and energetic efficiencies were calculated considering that all the gas produced is H₂ (

Table 4. Electrolysis performance with Ni electrodes at 3 V for 3 h, with direct current at 70 °C, and applying pulses (300 Hz) at room temperature: theoretical H₂ production, experimentally produced gas, Faradaic efficiency, and energetic efficiency.

Test	Theoretical H2/mL	Produced Gas/mL	${\mathsf{\eta }}_{\mathbf{F}\mathbf{a}\mathbf{r}\mathbf{a}\mathbf{d}\mathbf{a}\mathbf{i}\mathbf{c}}$/%	${\mathsf{\eta }}_{\mathbf{e}\mathbf{n}\mathbf{e}\mathbf{r}\mathbf{g}\mathbf{e}\mathbf{t}\mathbf{i}\mathbf{c}}$/%
DC	127	55	43	39
Pulsed	343	102	30	11

). The Faradaic efficiency was determined as the ratio of the experimentally measured H₂ volume to the theoretical H₂ volume; the latter was calculated by using the H₂ mass determined from Faraday’s laws (Equations (6) and (7)) and converting it to the H₂ volume,

$$1\mathrm{st} \mathrm{Faraday}’\mathrm{s} \mathrm{law} \mathrm{Q}=\mathrm{i}\times \mathrm{t}$$

(6)

$$2\mathrm{nd} \mathrm{Faraday}’\mathrm{s} \mathrm{law} {\mathrm{m}}_{{\mathrm{H}}_2}=(\mathrm{Q}\times \mathrm{M})/(\mathrm{F}\times \mathrm{n})$$

(7)

where Q is the charge in A s, i is the current in A, t is the time in s, ${\mathrm{m}}_{{\mathrm{H}}_2}$ is the produced H₂ mass in g, M is the molar mass of H₂ (2.02 g mol⁻¹), and n is the number of exchanged electrons (in this case, n = 2). The energetic efficiency was determined using Equation (8),

$${\mathsf{\eta }}_{\mathrm{e}\mathrm{n}\mathrm{e}\mathrm{r}\mathrm{g}\mathrm{y}}=\left({\mathrm{n}}_{{\dot{\dot{\mathrm{H}}}}_2}\times \mathrm{H}\mathrm{H}{\mathrm{V}}_{{\mathrm{H}}_2}\right)/{\dot{\mathrm{W}}}_{\mathrm{e}}$$

(8)

where ${\mathrm{n}}_{{\dot{\dot{\mathrm{H}}}}_2}$ is the produced H₂ flow (mol s⁻¹), HHV_H2 is the H₂ higher heating value (286.6 kJ mol⁻¹), and ${\dot{\mathrm{W}}}_{\mathrm{e}}$ is the energy given to the system (kW). Direct current electrolysis (Figure 12) exhibited the highest efficiency; the calculated Faradaic and energetic efficiencies were 43 and 39%, respectively. Despite the process’s innovative character, the obtained energetic efficiency is far from DOE’s target of 74% for alkaline electrolyzer stacks [32].

Electrolysis at 3 V and 60 °C using Ni electrodes and an anion-exchange membrane (AMI-7001S) was also conducted. This test revealed that gas production occurred only on the cathode side, indicating that the gas is H₂. This finding is supported by a study conducted by Caravaca et al. [33], suggesting that organic matter oxidation, specifically lignin oxidation, occurs at a lower potential than water oxidation. Therefore, O₂ production does not occur [33]. This electrolysis test primarily confirmed the production of gas, as it showed no further improvement in current density and posed a risk of membrane fouling.

One of this study’s goals was to investigate the incorporation of CO₂ into electrolysis; therefore, additional electrolysis tests were run at 3 V using Ni electrodes and bubbling CO₂. Figure 13 presents the data obtained at room temperature and 50 °C, demonstrating that under these conditions, not only are the currents low with the CO₂ bubbling, but the effect of temperature on current density is negligible, as the currents stabilize at similar values.

3.3.2. Alternating Current Electrolysis

Liu et al. claim that using a fluctuating current power supply can disturb the diffusion layer [34]; for this reason, some tests using this current mode were carried out. A half-sinusoidal wave was first used in these electrolysis tests to avoid electrode polarity inversion and ensure that ions moved only in one direction. So, the potential varied between 0 and a maximum of 3 V (in this case). It should be noted that in this study, the applied potential corresponds to the root mean square (RMS) value. The results showed that while the current density was higher, it was not significantly different from that generated by the continuous current tests, suggesting that electrolysis may still be limited by diffusion.

Due to the lack of a significant improvement with the half-sinusoidal AC wave, a switch to pulsed electrolysis was made. Pulsed electrolysis is known to effectively prevent the formation of the diffusion layer by interrupting the current before the layer fully develops, thereby reducing cell overpotentials [34]. Additionally, Cheng et al. reported that pulsed electrolysis can significantly enhance H₂ production efficiency at low loads, specifically by optimizing the pulse current magnitude and duty cycle [35].

Here, the pulse characteristics used were 3 V, 300 Hz, and a peak-to-peak value of 4 V. To better compare pulsed electrolysis with other current modes, Figure 14 presents a series of electrolysis tests conducted under various operating conditions to determine the optimal parameters.

Figure 14 shows that pulsed electrolysis not only produces more stable current densities but also generates significantly higher currents compared to all other electrolysis tests despite being conducted at 25 °C. Still, the amount of gas produced was the minimum amount detectable by the gas-measuring device (3.2 mL).

Another test involved bubbling CO₂ during pulsed electrolysis. Figure 15 compares the system’s performance with and without CO₂ bubbling. The results show that electrolysis without CO₂ bubbling yields more stable current densities. Although higher currents were observed during the first 75 min of CO₂ bubbling, they subsequently decreased. For this reason, tests with CO₂ bubbling were abandoned at this point. The poor performance of the CO₂ bubbling and its minimal impact on electrolysis current densities, observed in both continuous current and pulsed electrolysis experiments, are mostly related to the low electrocatalytic activity of the used electrode material (i.e., nickel) for the CO₂ reduction reaction and suboptimal cathode design. The results suggest that future studies on co-electrolysis should incorporate a well-designed gas diffusion electrode that provides the required pathways for CO₂ to effectively reach the catalyst layer, ensuring a uniform flow of gas and efficient reaction kinetics at the catalyst surface.

To address the low gas production observed in the pulsed electrolysis experiment shown in Figure 14, the peak-to-peak voltage was increased to 7 V while keeping the same RMS value. Furthermore, to prevent the pulse shape from deforming over time, the circuit was modified to trigger a discharge of the diffusion layer when the potential reached 0 V. The results from these adjustments are presented in Figure 16 and Table 4.

The conditions for pulsed electrolysis shown in Figure 16 generated higher current densities, stabilizing at 10 mA cm⁻² and producing a significantly higher amount of gas (102 mL). However, compared to direct current electrolysis tests, the calculated Faradaic and energetic efficiency were lower, at 30% and 11%, respectively. Although the electrolysis results data in Table 4 are promising, optimized operating conditions are still needed to increase the Faradaic and energetic efficiency.

3.4. Electrolysis Using Waters from the Liquefaction Process

To reuse water from the liquefaction process, bio-oil emulsions were prepared using this water source, and their performance on electrolysis was evaluated. These emulsions contained 50 vol.% of liquefied biomass in water, i.e., the condensed water and dark phase from the liquefaction process. The light-colored water phase exhibited no electrochemical activity due to its extremely low conductivity (0.01 × 10⁻³, Table 1) and was therefore excluded from further tests.

Since the conductivity of the emulsions prepared with the bio-oil and each of the condensed water and dark phases from the liquefaction process was very low, 1 mL of 50 wt.% H₂SO₄ was added to increase this parameter. Figure 17 compares the results of the 3 h DC electrolysis tests using the two water phases from the liquefaction process to those of the previous alkaline emulsion (liquefied biomass + 2 M KOH).

It is clear from Figure 17 that the electrolysis run with waters from the liquefaction process led to significantly higher current densities than the alkaline emulsion. In fact, for the first hour, the current densities using the emulsion with the condensed water led to more than double the current density of the alkaline emulsion. However, after stabilization, the current values were similar to those obtained with the dark phase emulsion. Despite the promising electrolysis results obtained with the waters from the liquefaction process, these tests led to a higher Ni anode mass loss than the alkaline emulsion at the end of the 3 h test. This effect is attributed to a more corrosive acidic environment, decreasing the Ni electrode stability. Additionally, gas production should be measured to calculate the Faradaic and energetic efficiencies and evaluate whether this approach, using the waters from the liquefaction process, is economically worth it.

4. Conclusions

The main goals of this study were to optimize the electrolysis of liquefied biomass to produce H₂, produce value-added compounds through the oxidation of organic matter, and capture CO₂ in the electrolysis medium. Since little information exists about the electrochemical activity of liquefied biomass (and consequently about the prepared emulsions), preliminary electrochemical studies, including CV and LSV, were conducted.

From CV, it was verified that the organic matter in the liquefied biomass is susceptible to redox reactions. Comparing CVs from the bio-oil emulsion before and after electrolysis, one can notice an increase in the current density, a deviation of peaks towards more positive potentials, and the appearance of a third anodic peak. The first observation indicates that this is due to the loss of some organic compounds that adsorb into the electrodes; the second indicates that the emulsion after electrolysis is more oxidized; and the third is due to an increase in the emulsion’s conductivity. The HER study in the bio-oil emulsion, using a Pt electrode, demonstrated that the reaction is more kinetically hindered than in alkaline water or black liquor.

Preliminary electrolysis tests with direct current showed that a potential higher than 2 V is required to attain reasonable current densities. Additionally, Ni electrodes perform better than TiO₂-RuO₂ alloy electrodes, and a temperature higher than room temperature is beneficial for electrolysis. The best result from the direct current electrolysis tests involved applying 3 V and 70 °C with Ni electrodes, leading to a 39% energetic efficiency and a 43% Faradaic efficiency. Although the electrolysis results are promising, the energetic efficiency of the process is still far from the ultimate (i.e., by 2031) DOE target for liquid alkaline electrolyzer stacks, which has been set as 74%.

The electrolysis test that produced a higher gas volume was carried out with a pulsed current of 300 Hz at 3 V and room temperature; this test had 11% energetic efficiency and 30% Faradaic efficiency. Additional electrolysis tests involving the bubbling of CO₂ into the bio-oil emulsion showed no significant improvement in the recorded current densities, both in the direct current and pulsed electrolysis modes. On the other hand, incorporating water from the liquefaction process showed higher current densities during electrolysis; however, an increased Ni electrode mass loss was verified.

This study demonstrates a promising approach to producing H₂ without O₂ contamination, addressing a critical challenge in H₂ production. However, the further optimization of operating conditions is required to enhance efficiency and scalability. Additionally, more in-depth investigations are necessary to identify the added-value compounds generated during organic matter oxidation and to further evaluate whether CO₂ bubbling could facilitate the formation of other valuable gases (e.g., CO, CH₄).