Several reactions and intermediate steps are involved in the recovery of H₂ from biomass oxygenates with the inherent removal of CO₂. These reactions are incorporated into the models for exploring thermodynamic limits for enhanced H₂ recovery with carbon removal via mineralization. Reaction 2 represents the deconstruction of biomass oxygenates to produce CO₂ and H₂. This reaction represents the overall pathway for biomass oxygenate refining without CO₂ capture. Biomass oxygenate reforming:C x H y O z ( aq ) + ( 2 x - z ) H 2 O ( 1 ) → xCO 2 ( g ) + ( 2 x - z + 0 . 5 y ) H 2 ( g ) (2) Standard heat of reaction ( ΔH 500 0 ) = 108.6^a, 285.2^b, 268.1^c, 195.5^d, 388.7^e, 220.3^f kJ mol − 1

a, b, c, d, e, f corresponds to methanol, ethanol, glycerol, ethylene glycol, acetone, and acetic acid, respectively.

Reaction 2 involves several intermediate steps as described below. The first step in this pathway is the preferential formation of CO through bio-oxygenate decomposition as shown in Eq. 3a for C:O molar ratio of 1:1, and hydrolysis as shown in Eq. 3b for C:O molar ratios greater than 1:1. The CO produced then undergoes an exothermic water gas shift reaction (WGSR) to produce H₂ and CO₂. Biomass oxygenate decomposition:C x H y O z ( aq ) → xCO ( g ) + ( 0 . 5 y ) H 2 ( g ) (3a) C x H y O z ( aq ) + ( x - z ) H 2 O → xCO ( g ) + ( x - z + 0 . 5 y ) H 2 ( g ) (3b) ΔH 500 0 =   127 . 1 g , 322 . 2 b , 323 . 6 c , 232 . 5 h , 444 . 2 e , 257 , 2 i       kJmol − 1 Water - Gas Shift Reaction:CO ( g ) + H 2 O ( l ) → H 2 ( g ) + CO 2 ( g ) (4) ΔH 500 0 = - 18 . 5   kJmol − 1

Competing reactions that result in CH₄ formation as opposed to H₂ formation need to be incorporated (see reactions 5 and 6). Compared to WGS reactions, methanation reactions are more exothermic which make them favorable at low temperatures. Methanation reactions:CO ( g ) + 3 H 2 ( g ) → CH 4 ( g ) + H 2 O ( l ) (5) CO 2   +   4 H 2 → CH 4 +   2 H 2 O ( l ) (6)

ΔH 500 0 = −235.6  kJmol − 1 and ΔH 500 0 = −217.1  kJmol − 1 for Eqs 5, 6 respectively.

Furthermore, bio-derived oxygenates and their intermediate products can undergo dehydration, cracking, decomposition, and reduction to produce more undesired products such as CH₄, CO, solid C, and C₂H₄, as shown by reactions 7–11 below. Ethanol dehydration: C 2 H 6 O ( l ) → C 2 H 4 ( g ) +  H 2 O ( l ) (7) ΔH 500 0 =   58 . 4       kJmol − 1 Boudouard Reaction:  2 CO ( g ) → CO 2 ( g ) + C ( s ) (8) ΔH 500 0 =   - 173 . 8         kJmol − 1 Methane Cracking: CH 4 ( g ) → 2 H 2 ( g )   +  C ( s ) (9) ΔH 500 0 =   80 . 3                 kJmol − 1 Carbon monoxide reduction: CO ( g )   +  H 2 ( g ) → H 2 O ( l )   +  C ( s ) (10) ΔH 500 0 =   - 155 . 3     kJmol − 1 Carbon dioxide reduction: CO 2 ( g )   +   2 H 2 ( g ) →   2 H 2 O ( g )   +  C ( s ) (11) ΔH 500 0 =   - 136 . 8       kJmol − 1

Bio-oxygenate reforming reactions can be endothermic. Therefore, coupling these with exothermic carbon mineralization reactions (Eq. 12) enhances the overall spontaneity of these reactions. The overall exothermicity resulting from the coupling of these reaction is shown in Eq. 1, and the heat of reactions at 500 K for methanol, ethanol, glycerol, ethylene glycol, acetone, and acetic acid is presented in Table 2. Carbonate formation reaction:CO 2 ( g ) +  CaO ( s ) → CaCO 3 ( s ) (12) ΔH 500 0 =   - 174 . 3     kJmol − 1

The equilibrium compositions of various reacting processes are determined using the Gibbs energy minimization method. A more detailed description of this methodology was described by Lima et al. (Lima da Silva et al., 2009; Lima Da Silva and Müller, 2011). The advantage of this approach is that it addresses systems with potential for simultaneous reactions. The objective function shown in Eq. 14 is minimized subject to the elemental balance constraint in Eq. 15, G RT = { ∑ i=1 N n i [ G 0 i RT + ln ( z i P ) ] } + 1 RT ∑ j = 1 N n j G o j (14) ∑ j = 1 N n i α ik + ∑ j = 1 N n j α jk = b k   K = 1 … … M (15) Where G, G o j , T, R, z i , P, n j , α jk , and b k represents the total Gibbs free energy of the system, Gibbs free energy of species i at its standard state, temperature of the system, universal gas constant, mole fraction of species i in the gas phase, pressure of the system, number of atoms of kth component present in each molecules of species i, and the total number of atomic masses of kth component in the system, respectively.

Minimizing the objective function of the system indicates that a mixture of chemical species approaches an equilibrium state. These equilibrium calculations were performed and simulated using ASPEN^® Plus V8.6 process simulation software. The Gibbs reactor (see Figure 2A,B) was modelled as an Isothermal equilibrium reactor, which calculates the equilibrium composition of product species using the Gibbs energy minimization method at various operating conditions, keeping our reaction temperature constant, while accounting for the heat of reactions. The Peng-Robinson (Peng-Rob) equations of state were used to model the property parameter (Peng and Robinson, 1976). This property method has been widely used for processes containing oxygenates, water, and combustion gases (Fateen et al., 2013; Ali et al., 2014; Jimmy et al., 2017).

Prior to been fed into the reactor, biomass oxygenate, and water is mixed, pressurized, and heated using a mixer, pump, and heater, respectively. In cases where the effect of alkalinity was investigated, the alkaline source was fed directly into the reactor at its specified temperature and pressure. It was found that the temperature of the input feed stream does not affect the thermodynamic results, because fixing the reactor parameters produces the equilibrium effect determined by the set temperature and pressure. However, the configuration was still used to ensure proper representation of the real batch reformer, as it would account for the energy required to heat the feedstock to the desired temperature and cool the products before analysis. Using this setup, H₂ evolution with varying biomass oxygenate-to-water ratios in the range of 5–95%, a low temperature range of 150–325°C, and varying alkaline sources, was studied at 50 bar. Prior studies have shown that methanation reactions can also be effectively suppressed using catalysts (Davda et al., 2005; Coronado et al., 2016). We also investigated the influence of 1) methanation and 2) suppressed methanation. Both cases were also investigated with and without CO₂ capture via carbon mineralization. This approach provides the optimal and sub-optimal scenarios that are dependent on the activity of the catalysts.

The equilibrium conversions of biomass oxygenates to products and associated product yields were defined as follows: % Conversion of Biomass Oxygenates ( BO )   = BO in −  BO out BO in  *    100 (17) % H 2   yield  = [ Moles of H 2   produced  Moles of biomass oxygenate fed * Stoichometric Ratio  ( SR ) ]  *    100 (18) % CO ,  CO 2   and alkanes yield   = Moles of C in compond produced  Moles of C atoms in BO fed   *    100 (19)

Water plays a synergistic role in enhanced H₂ recovery with inherent carbon removal from biomass oxygenates. Water facilitates the dissolution of calcium oxide to release calcium ions that can readily react with the evolved CO₂ to produce inorganic carbonates, as opposed to steam. The enhanced ion transport in gas-liquid-solid environments as opposed to in gas-solid environments favors accelerated formation of carbonates at lower temperatures as opposed to sorption-enhanced stream methane reforming that occurs at higher temperatures (Wang et al., 2008; Lindén et al., 2011). Further, the need for additional energy input in the form of latent heat of vaporization is avoided by maintaining water in the aqueous phase. To harness these advantages and resolve the compositions of the product stream, simulations were performed at 227°C, N₂ pressure of 50 bar, with and without stoichiometric amounts of CaO as defined by reaction 1 and 13. H₂, CH₄, and CO₂ yields with and without CaO are shown in Figures 3–8 for methanol, ethanol, glycerol, ethylene glycol, acetone, and acetic acid, respectively.

In the absence of CaO for CO₂ capture, the yields of H₂ and CH₄ remain unaffected by an increase in the concentration of methanol and ethanol up to 40 wt% (See Figures 3A,C, 4A,C). Further increases in the concentrations of methanol and ethanol led to an increase in CH₄ yield and a simultaneous decrease in the production of hydrogen. Decrease in H₂ yield is accompanied by a decrease in the production of gaseous CO₂ gas, which can be attributed to the limited amount of H₂O available to undergo a water gas shift reaction as shown in Eq. 4. Limited water concentration cause a thermodynamic shift to favor CH₄ production via CO methanation reaction at equilibrium. When CaO was introduced to capture CO₂, H₂ concentrations decreased after methanol and ethanol concentrations exceeded 20 wt% (See Figures 3B,D, 4B,D). This change is attributed to the use of water for the water gas shift reaction and the dissolution of CaO to produce calcium carbonate.

The hypothesis that CaO significantly enhances H₂ yields and suppresses CH₄ yields was shown to be true since H₂ yields exceeding 75% were achieved around 20 wt% for all the biomass oxygenates reported in this study (Figures 3B–8B). In the absence of CaO, less than 2% yield of H₂ was noted (Figures 3A–8A). However, methane yields either remained unchanged at low biomass oxygenate concentrations or increased in most cases except for acetone. At acetone concentrations of 80 wt% or higher, H₂ yields approached zero and methane yields decreased significantly. Also, the concentration of unreacted acetone in the product stream rapidly increased indicating low conversions of acetone (19–81%) at concentrations of 80wt% or higher, elsewhere the conversion was 100%. The observed anomaly could imply a different reaction mechanism caused by inadequate amounts of water leading to acetone decomposition to produce unstable radicals described as (CH₃COCH₃ → 2CH₃ ^* + CO) (Spence and Wild, 1936; Smith and Hinshelwood, 1944). Further, the absence of H₂O or H₂ for the CO produced during this reaction to undergo WGS reactions or methanation reaction as described in reactions 4) and 5) shifts the equilibrium back to acetone production at room temperature, leading to low conversions. With all the biomass oxygenates investigated, non-monotonic relationships between the concentrations of biomass oxygenates and the yields of H₂ and CH₄ were noted.

It was interesting to note a significantly lower yield of H₂ at biomass oxygenate concentrations exceeding 60% for all the biomass oxygenates reported in this study which are methanol, ethanol, glycerol, ethylene glycol, acetone, and acetic acid. However, the trends in the concentrations of H₂ and CH₄ at biomass oxygenate concentrations below 60 wt% differ. The effectiveness of using stoichiometric amounts of CaO is evident from the negligibly low concentrations of CO₂ in the gas phase to the order of 10^–12 mol of CO₂ per mole of biomass oxygenate fed after CO₂ capture (Figures 3D–8D). At biomass oxygenate concentrations of 20 wt% or less, near complete conversions of CaO to CaCO₃ are noted (Figures 3D–9D). In addition to aiding the separation of CO₂, the heat generated during carbon mineralization aids the water gas shift reaction for enhanced H₂ conversion.

Temperature has a significant effect on coupled multiphase chemical interactions involved in enhanced H₂ recovery with inherent CO₂ removal from biomass oxygenate precursors. Increase in temperature favors product and reactant formation in endothermic and exothermic reactions, respectively. Further, the solubility of calcium carbonate decreases with temperature which favors the removal of CO₂. Aqueous biomass oxygenate decomposition is an endothermic reaction (see Reaction 3) and higher temperatures are expected to aid the decomposition behavior. In contrast, the conversion of CO and H₂ to CH₄ is exothermic (and higher temperatures may shift the equilibrium towards the reactants, which favors H₂ yield. Coupling the exothermic water-gas-shift reaction (see Reaction 4) with the carbon mineralization reaction removes CO₂ from the product stream and pushes the equilibrium to the products. This coupling of reactions counterbalances the effect of temperature which is to push the equilibrium towards the products. To investigate the hypothesis that the reaction temperature has a significant effect on enhanced H₂ formation with inherent carbon removal, the gas compositions were determined at temperatures in the range of 150°C–325°C, while all other conditions such as the biomass oxygenate composition and the N₂ pressure were held constant at 50 wt% and 50 bar.

In the absence of a catalyst and CaO for CO₂ capture, H₂ yields are less than 1% for all the biomass oxygenates and CH₄ yields exceed 49% in the temperature range of 150°C–325°C (Figures 9A–14A). H₂ and CO₂ yields increase monotonically with temperature (Figures 9A,C–14A,C). The yields of CH₄ decrease monotonically with temperature at these conditions (Figures 9A–14A). In the presence of CaO for CO₂ capture, H₂ yields of 46–65% or higher are achieved at temperatures exceeding 220°C. At these conditions, CH₄ yields are 30% or lower at temperatures exceeding 220°C (Figures 9B–14B). These data suggest that alkaline environments (bearing CaO in this case) favor alkaline hydrolysis that contributes to the cleavage of C-C bonds as opposed to acid hydrolysis that results in C-O cleavage which promotes alkane formation. (Davda et al., 2005). Further, CO yields to the order of 10^–3% and 10^–9% were noted in the absence and presence of CaO, suggesting that the removal of CO₂ from the gas phase accelerates the conversion of CO to CO₂ (See Supplementary Figure S2A–F).

Unlike the cases without CaO, where we obtained a constant increase in H₂ yield, an asymptotic pattern was observed in cases with in-situ CO₂ capture for temperatures greater than 225°C. Furthermore, despite the low amounts of CO being produced, there was a sharp increase CO formation with increasing temperature between 225°C and 320°C. These observations imply that biomass oxygenate decomposition is aided at higher temperatures to produce CO. However, the low concentrations of CO and higher temperatures do not favor product formation in the exothermic water-gas-shift and methanation reactions. These factors contribute to the asymptotic behavior of H₂ above 225°C which also corresponds to the near complete uptake of CO₂ to produce CaCO₃ (Figures 9D–14D). Negligible concentrations of CO₂ to the order 10^–7–10^–10% were found in the product gas stream for all cases of biomass oxygenates studied after CO₂ capture using CaO (See Figures 9D–14D).

The key limiting factor in accelerating H₂ recovery from biomass oxygenates is the formation of CH₄. CH₄ is produced from CO or CO₂ and H₂ as building blocks via exothermic reaction pathways (see Reactions 5 and 6). One approach to limit CH₄ formation is to use catalysts that favor C-C bond cleavage and the water gas shift (WGS) reaction over C-O bond cleavage and methanation reactions (Davda et al., 2003; Davda et al., 2005; Coronado et al., 2016). Some examples of these catalysts include rare earth metals supported catalysts such as Pt/SiO_2, Pd/Al₂O₃, and Ru/C. Further, catalysts that have a basic character or operate in alkaline environments such as earth abundant Ni catalysts have also been shown to be effective in aiding H₂ evolution (Stonor et al., 2017b).

In this study, we build on these observations to investigate the influence of methanation and its suppression on H₂, CH₄, CO, CO₂, and CaCO₃ yields. In the suppressed methanation case, the reactions associated with CH₄ formation from CO and CO₂ precursors are not included in the model. The studies are conducted at 227°C, N₂ pressure of 50 bar, and biomass oxygenate concentrations of 20 wt%, with and without CO₂ capture. For cases with CO₂ capture, we used a stoichiometric amounts of CaO as defined by (reaction 13). In this study, we consider four scenarios for hydrothermal biomass oxygenate reactions: 1) with methanation, 2) without methanation, 3) alkaline treatment with methanation, and 4) alkaline treatment without methanation reactions (Figure 15).

In the scenario with methanation and without a catalyst, CH₄ and CO₂ compositions are dominant and H₂ yields are less than 0.3%. In the scenario where methanation is suppressed by using a catalyst, H₂ and CO₂ yields are dominant, while CO formation is suppressed and there is no CH₄ observed. For alkaline treatment with methanation, more than 70% conversion of H₂ and CaCO₃, and substantial CH₄ yields are noted. No CO formation is noted in this case. For alkaline treatment without methanation, H₂ and CaCO₃ are the dominant phases. These results demonstrate that strategies to harness biomass oxygenates for H₂ recovery with CO₂ removal need to incorporate catalysts to suppress methane formation and alkaline sources for in-situ CO₂ capture.

Analysis of the product streams provided further insights into the reactivities of various biomass oxygenates. The reactivities of the functional groups proceed in the descending order of OH > COOH > COR. This trend was similar with experimental results obtained by Fu and co-workers (Fu et al., 2020) who investigated the steam reforming of bio-oil and its derivatives and reported the highest H₂ yield from reforming of ethylene glycol, followed by acetic acid, and acetone. Additionally, the separation of CO₂ from the gas phase in the alkaline environment greatly limits methanation kinetically at low and moderate temperatures (less than 227°C) and enhances H₂ production. The lower temperatures are potentially beneficial to catalyst life, and they significantly reduce cost. Furthermore, trace amounts of CO(10⁻¹² %) were evident in all cases studied, implying that these processes are effective for generating hydrogen with CO content within the 20 ppm limit for fuel cell applications.