Mechanism Investigation of Solvent Effect on Selective Decomposition of Formic Acid

Abstract

The selective decomposition of formic acid to hydrogen gas represents a highly promising strategy for sustainable energy production. The influence of solvent effects on the selective decomposition of formic acid into H₂ and CO₂ or H₂O and CO was investigated. A variety of solvents, including polar protic solvents (e.g., water, ethanol, methanol), polar aprotic solvents (e.g., tetrahydrofuran, dimethyl sulfoxide), and ionic liquids, were employed in conjunction with a 5 wt% Pd/C catalyst. The yield of formic acid decomposition and the turnover number (TON) were found to be dependent on the choice of solvent. To elucidate the solvent effects, classical solvent parameters and Kamlet–Taft solvatochromic parameters were studied. The study revealed correlations between the TON and the solubility of hydrogen, Kamlet–Taft parameters (acidity, basicity, and polarity/dipolarity), hydrogen bond donor (HBD) capability, and hydrogen bond acceptor (HBA) capacity. The solvent identity was found to play a dominant role in both the polarity/dipolarity and the catalytic mechanism of formic acid decomposition.

1. Introduction

Hydrogen gas, as a clean and sustainable fuel [1,2], can be produced from various feedstocks and processes [3], including natural gas reforming, electrolysis of water, and biomass-derived formic acid (FA) decomposition [4,5,6]. Formic acid is considered one of the most promising candidates for hydrogen storage due to its high hydrogen capacity (4.4 wt%) [7,8], liquid state at room temperature, non-toxicity, and its presence as a major by-product in lignocellulosic biomass hydrolysis degradation reactions [9,10,11].

Numerous studies have reported the use of various noble metal and non-noble metal catalysts (e.g., Pd, Ru, Pt, Ir, and Ni based catalysts) for the selective decomposition of formic acid into CO₂ and H₂ [12,13,14,15,16]. The decomposition of formic acid occurs via two main pathways: (1) the formate route, where the O-H bond in formic acid is activated first to form HCOO* and H* (* indicates chemisorbed species), followed by cleavage of the H-COO* bond to produce CO₂ [17], and (2) the carboxyl route, where the C-H bond in formic acid is cleaved to form COOH* and H*, followed by cleavage of the O-H bond in COOH* to produce CO₂ (Scheme 1) [18,19]. The density functional theory (DFT) calculations have shown that the formation of COOH* is more challenging than the formation of HCOO*, while the decomposition of HCOO* into CO₂ is more difficult than that of COOH* [20]. This implies distinct rate-limiting steps for each pathway: the decomposition of HCOO is rate-limiting in the formate route, whereas the O-H bond breaking in COOH is rate-limiting in the carboxyl route. In addition, the COOH^* can also go through cleaving the C-O bond cleavage, leading to water molecular formation (carboxyl route, Scheme 1) [21,22]. The reaction pathways and product yields are influenced by various parameters, such as reaction temperature, but can be controlled using appropriate catalysts and solvents [23,24]. Since the 1970s, researchers have investigated the reverse reaction of CO₂ to formic acid using a series of metal complex catalysts (e.g., Pd-, Ru-, and Ni-based catalysts) [25,26,27]. Among these catalysts, Pd supported on carbon (Pd/C) presents high selectivity for H₂ production, excellent activity under mild conditions, and practical applicability.

Therefore, it is crucial to find a suitable catalyst and solvent for the formic acid selectivity decomposition into hydrogen gas [28,29]. Although several theoretical studies have explored formic acid decomposition on different metal surfaces [30], the specific roles of the commonly used solvents in the decomposition of formic acid have not yet been fully understood [31,32]. Further experimental techniques are needed to develop models correlating product yields and guiding solvent selection. As chemical transformations occur at the solid–liquid interface, solvents are generally considered to affect the adsorption energies between the catalyst and reactant and to modify the active sites of the catalyst.

In this study, to systematically dissect the solvent effects, we investigated the selective decomposition of formic acid using Pd/C in solvents selected from three distinct classes based on their physicochemical properties and relevance to the reaction. (1) Polar protic solvents (water, ethanol): These solvents possess strong hydrogen bond donor (HBD) capabilities, which can interact with the carboxyl group of formic acid to modulate its dissociation equilibrium—critical for understanding how hydrogen bonding influences intermediate formation (e.g., HCOO* or COOH*). Their relevance is heightened as they are common in biomass-derived formic acid systems. (2) Polar aprotic solvents (THF, DMSO): These solvents lack HBD ability but exhibit high dipolarity/polarizability, allowing us to decouple the effects of hydrogen bonding from pure dipolar interactions. This is key to verifying the role of solvent polarity in reaction kinetics, which is subsequently analyzed via Kamlet–Taft parameters. (3) Ionic liquids ([BMIM]Cl, [BMIM][Tf₂N], [BMIM][OAc]): These imidazolium-based ionic liquids were chosen for their structural tunability (via anions: Cl⁻, [Tf₂N]⁻, [OAc]⁻) and strong hydrogen bond acceptor (HBA) capabilities. The varying anions induce significant differences in HBA strength and polarity, enabling a probe into how HBA capacity affects formic acid stabilization or inhibition—an aspect underexplored in prior studies. It was found that the rate of formic acid decomposition was higher in polar solvents such as ethanol and tetrahydrofuran than in ionic liquids. The role of solvents in formic acid decomposition was studied through comparative experiments and Kamlet–Taft solvatochromic parameters. Attenuated total reflection (ATR)-FTIR and ¹H NMR spectroscopy were employed to quantify the interactions between formic acid and solvents, thereby elucidating the solvent effects on the decomposition of formic acid into H₂ and CO₂.

2. Results and Discussion

2.1. Formic Acid Decomposition in Different Solvent Systems

To investigate the solvent effects, we selected the decomposition of formic acid catalyzed by 5 wt% Pd/C as a model reaction due to its well-documented high activity and selectivity for formic acid decomposition into H₂ and CO₂ in various solvents (protic solvents: water, ethanol; aprotic polar solvents: DMSO, THF; or ionic liquids: [BMIM]Cl, [BMIM][Tf₂N], and [BMIM][OAc]) (Figure 1). Pd-based catalysts, particularly Pd/C, offer advantages such as excellent dispersion, stability, and efficiency in dehydrogenation reactions, making them a preferred choice for heterogeneous catalytic systems. out. When using ionic liquids such as [BMIM]Cl or [BMIM][OAc], the formic acid conversion yield increased gradually over the first hour of reaction time (Figure 1). The results showed that formic acid conversions were generally higher in polar and dipolar aprotic solvents compared to ionic liquids. However, further prolonging the reaction time led to only a slight increase in conversion. This behavior can be attributed to the decomposition rate of formic acid being greater than the reduction rate of CO₂, causing the formic acid conversion yield to increase initially. The yield of formic acid conversion reached its maximum when the formation rate of formic acid equaled its reduction rate. Afterward, formic acid conversion began to decrease, as the CO₂ reduction rate became greater than the formic acid composition rate. It is important to note that the decrease in conversion over time may also involve catalyst deactivation (e.g., via CO adsorption or poisoning) alongside these equilibrium effects.

Subsequently, a series of experiments on formic acid decomposition was carried out in water with different amounts of DMSO (Figure 2a). The TON was used to show the influence of Kamlet–Taft parameters (acidity α, basicity β, and dipolarity/polarizability π*) of aqueous solutions with different DMSO concentrations on the chemical equilibrium and formic acid dissociation (Figure 2a). Pure water, which exhibits the highest acidity (α), demonstrated the best performance, likely due to its ability to shift the equilibrium towards formic acid decomposition (Figure 2a). However, the performance of other solvent mixtures did not strictly follow the trend of their Kamlet–Taft acidity parameters. Although there was a noticeable difference in TON between pure water and a 20 mol% DMSO aqueous solution, the difference was not statistically substantial. This suggests that the Kamlet–Taft acidity parameters alone cannot fully account for the formic acid decomposition rate. It appears that multiple solvent parameters collectively influence formic acid decomposition.

Additionally, the effect of ethanol addition was explored by varying the ethanol content in water (Figure 2b). Compared to reactions conducted without a solvent, the formic acid conversion yield increased with the addition of pure water. The formic acid conversion increased with the addition of ethanol when the ethanol was increased to 2.1 mL and 2.4 mL (ethanol–water = 7:3). However, increasing the ethanol amount to 2.7 mL led to a slight decrease in conversion. In pure solvent systems (ethanol or water), the formic acid conversion increased gradually over time. In contrast, in ethanol–water mixture systems, the conversion yield rapidly rose to 40% within the first 2 h and then remained relatively stable with further prolonged reaction time. The higher formic acid conversion observed in polar and dipolar aprotic solvents is possibly attributed to their ability to stabilize the rate-limiting HCOO* intermediate through strong solvation of the surface adsorbed formate, thereby lowering the effective activation barrier for C-H cleavage [33].

In addition to the interaction between solvents and formic acid molecules, the direct interaction of solvents with the surface of Pd/C catalysts may also affect catalytic activity. Although the catalyst after use was not directly characterized in this study, the observed differences in activity, particularly the lower TON in ionic liquids, suggest the possibility of such interactions. For example, anions (e.g., Cl⁻) in ionic liquids or polar solvent molecules (e.g., S=O groups in DMSO) may competitively adsorb to Pd active sites, hindering the adsorption and activation of formic acid [15,32]. In addition, solvents may affect the surface oxidation state or electron density of Pd nanoparticles, thereby altering their catalytic properties. Therefore, the observed solvent effect should be the result of the combined action of solvent–reactant and solvent–catalyst.

2.2. Correlation of TON with Reaction Solvent Parameters

As indicated in the results above, the kinds of solvents played a vital role in the formic acid decomposition. To better understand the solvent effects on formic acid decomposition, the logTON was correlated with various solvents parameters, including classical measures of the solvatochromic scales of hydrogen bond acceptor (HBA) capability, hydrogen bond donor (HBD) capability, dipolarity/polarity, Reichardt’ dye $E_T^N$, and dielectric constant (ε) (Figure 3). As shown in Figure 3, a clear correlation was observed between solvent polarity and formic acid decomposition, while a poor correlation was found between solvent acidity (or basicity, dielectric constant, Reichardt’ dye $E_T^N$) and the logTON. The strong correlation with solvent polarity/dipolarity suggests that lower polarity environments enhance formic acid decomposition by reducing solvation effects, which favor reactant destabilization and intermediate formation. In contrast, HBA and HBD parameters show weak correlations because the specific mode of hydrogen bond interaction (whether it is a single facilitator hydrogen bond or a stable multiple hydrogen bond network) is more critical than the effect of its absolute strength on the reaction rate. Therefore, polarity primarily governs electron transfer and adsorption processes. While solvent acidity and basicity have limited influence, this implies that while polarity is a dominant factor, other parameters also play roles in influencing the reaction kinetics.

The second-order rate constant and conversion were correlated with the Kamlet–Taft dipolarity/polarizability, acidity, and basicity in ethanol (1)–water (2) mixtures as a function of mole fraction of water (x₂) (Figure 4). With the addition of water into ethanol (x₂ < 0.2), an increase in rate constant and conversion (Figure 4a) is possibly caused due to increasing the dipolarity/polarizability and acidity values (Figure 4b,c). However, with a further increase in water amount, the rate constant and conversion tended to decrease.

The evolution of the Kamlet–Taft parameters (Figure 4b–d) provides a physicochemical rationale for this non-monotonic activity trend. The initial enhancement in activity at low water fractions (x₂ < 0.2) correlates well with a sharp increase in the dipolarity/polarizability (π*), which is expected to stabilize the charged transition states involved in the formate or carboxyl pathways. Concurrently, the increasing solvent acidity (α, Figure 4c) may also contribute by facilitating O-H bond cleavage in formic acid. However, as the water content increases beyond x₂ ≈ 0.2, the activity declines despite continued, albeit slower, rises in π* and α. This suggests that the beneficial effects of high polarity are eventually counteracted by inhibitory factors. These likely include the formation of a strong, stabilizing hydrogen bond network around the formic acid molecule (reflected in the high α value of water) and a potential shift in the reaction equilibrium towards formic acid formation due to the excess water, which is a product of the dehydration side-pathway. The basicity (β, Figure d) shows a monotonic decrease with increasing water content, indicating that the role of hydrogen bond acceptance diminishes as the protic character of the mixture increases. In summary, the optimal catalytic performance in ethanol–water mixtures are achieved at a composition that balances high polarity for transition state stabilization against the inhibitory effects of excessive hydrogen bond donation from water.

2.3. Estimation of the Hydrogen Bond of Formic Acid and Solvents

To elucidate the effects of solvent polarity on formic acid decomposition, the ATR-FTIR spectra and NMR chemical shifts in formic acid in various solvents at different concentrations were systematically measured and compared. The ATR-FTIR spectra of formic acid in different solvent mixtures were recorded across the entire concentration range (Figure 5). The results show that the C=O stretching vibration of formic acid exhibits a significant blue shift in ionic liquids such as [BMIM][Tf₂N] and in water as the solvent concentration increases (Figure 5a). In contrast, a slight red shift in the C=O stretching vibration of formic acid is observed in DMSO, [BMIM]Cl, and [BMIM][OAc] with increasing solvent concentration. These shifts in the wavenumber of the C=O stretching vibration indicate that [BMIM][Tf₂N] and water have a stronger hydrogen bonding ability with the C=O group in formic acid compared to other solvents like DMSO, [BMIM]Cl, and [BMIM][OAc] (Figure 5a), while solvents such as DMSO are weaker. Traditionally, the red shift of C=O telescopic vibrations is often attributed to the interaction of the solvent as a hydrogen bond donor with carbonyl oxygen atoms, weakening the C=O bond. However, the blue shift phenomenon we observed reveals a more complex pattern of competitive interactions: in ionic liquids (e.g., [BMIM][Tf₂N]) and water, solvent molecules preferentially form strong hydrogen bonds with the O-H hydrogen atoms of formic acid as hydrogen bond acceptors, rather than the traditional interaction with carbonyl oxygen atoms. This interaction causes electrons to be transferred from the solvent to the σ antibond orbitals of the formic acid O-H bonds, making the O-H bonds more prone to breaking, thereby promoting formic acid breakdown. Conversely, in DMSO and some ionic liquids, red shifting is consistent with the traditional perception that solvents interact with C=O as hydrogen bond donors, which may stabilize formic acid molecules and inhibit decomposition. The carboxyl groups of formic acid gradually shift to higher wavenumbers (blue shift) with increasing solvent mole fraction. This effect is more pronounced in formic acid–ethanol, formic acid–THF, formic acid–GVL, and formic acid–water systems than in formic acid–ionic liquid systems. This suggests that the hydrogen bonding interactions between formic acid and these solvents are stronger, leading to significant changes in the vibrational frequencies. When formic acid decomposition is carried out in ionic liquids (e.g., [BMIM][OAc] or [BMIM][Tf₂N]), the anion stretching vibrations (νₛ(COO⁻) at around 1380 cm⁻¹ and δ(COO⁻) at 640 cm⁻¹) exhibit a blue shift with increasing formic acid mole fraction (Figure S3). Additionally, the S=O bending vibrations at around 1200 cm⁻¹, 1100 cm⁻¹, and 1000 cm⁻¹ show a blue shift for [BMIM][Tf₂N] (Figure S3). These observations suggest that in the [BMIM][Tf₂N]–formic acid and [BMIM][OAc]–formic acid solvent systems, not only the C-H hydrogen bonds but also the anions of the ionic liquids stabilize formic acid through the formation of multiple intermolecular hydrogen bonds.

¹H NMR measurements of pure formic acid, various solvents, and a series of formic acid–solvent mixtures were conducted at 25 °C. The changes in the chemical shifts in the individual hydrogen atoms along with the mole fraction of solvents were evaluated (Figure 6). The dependences of the chemical shifts (∆δ) of hydrogen atoms in the carboxyl group on the mole fraction of solvents are presented in Figure 6.

The ∆δ values for formic acid–ionic liquid mixtures are positive (Figure 6a), indicating a downfield shift (blue shift) of the hydrogen atoms in the carboxyl group with increasing ionic liquid concentration. This suggests that the interaction between formic acid and ionic liquids is stronger, leading to a more stable formic acid molecule and inhibiting its decomposition. This is consistent with the FTIR blue shift phenomenon: the solvent interacts with O-H as a hydrogen bond acceptor, reducing the density of electrons around the carboxyl hydrogen and making the O-H bond more prone to breakage. As shown in Figure 6b, the ∆δ values of hydrogen atoms in formic acid–polar solvent and formic acid–dipolar aprotic solvent mixtures are negative. This indicates that the hydrogen atoms in the carboxyl group of formic acid experience an upfield shift (red shift) with increasing concentrations of polar or dipolar aprotic solvents (e.g., ethanol, THF). This suggests that these solvents form hydrogen bonds with formic acid, stabilizing the carboxyl group and promoting the decomposition process.

Based on the spectroscopic analysis, the observed chemical shift changes can be attributed to hydrogen bonding interactions between formic acid and the solvents (Figure 6). The upfield shift (red shift) in the hydrogen atoms of the carboxyl group indicates the formation of hydrogen bonds between formic acid and the solvent. This interaction likely involves electron transfer from the solvent (an electron acceptor) to formic acid (an electron donor), stabilizing the carboxyl group and promoting the release of the H-atom in the CO-H bond. Bond. The downfield shift (blue shift) suggests strong interactions between formic acid and the ionic liquid, stabilizing the formic acid molecule and inhibiting its decomposition. This different interaction mode directly affects the adsorption configuration and decomposition path of formic acid on the surface of Pd/C catalysts. In blue shift systems (e.g., ionic liquids, water), strong interactions between solvents and formic acid O-H may lead to the adsorption of formic acid molecules on the catalyst surface in a more “parallel” manner, favoring O-H bond breaking (formate path). In red shift systems (such as DMSO), the interaction of the solvent with C=O may affect the adsorption of formic acid on the surface of the catalyst through carbonyl oxygen. The total effect of the complex polyhydrogen bond network with formic acid O-H (as hydrogen bond acceptor) and C=O (as hydrogen bond donor) in ionic liquids and water in some cases manifests as blue shifting of C=O bonds, while having a net effect on the stability and reactivity of formic acid molecules, which is consistent with the lower decomposition rate results we observed in ionic liquids. The hydrogen atoms in the carboxyl group of formic acid form hydrogen bonds with the oxygen atoms in the anion groups of the ionic liquid (e.g., [Tf₂N]⁻). This interaction involves electron transfer from the ionic liquid (an electron acceptor) to formic acid (an electron donor), promoting the release of the H-atom in the CO-H bond. Then, the oxygen atoms in the carboxyl group of formic acid form hydrogen bonds with the ν(CH) groups of the imidazolium ring in the ionic liquid. This interaction involves electron transfer from formic acid (an electron donor) to the ionic liquid (an electron acceptor), stabilizing the formic acid molecule and inhibiting the release of the H-atom in the CO-H bond.

It is important to note that while our NMR and ATR-FTIR data provide strong indirect evidence for the interactions leading to intermediate formation (e.g., chemical shifts and vibrational frequency shifts indicating hydrogen bonding that precedes bond cleavage), the direct detection of transient intermediates (e.g., HCOO*or COOH*) was challenging due to their short lifetimes under the experimental conditions. Nevertheless, the consistent spectroscopic changes observed support the proposed bond breaking steps in the reaction mechanism.

In the multi-step decomposition of formic acid, the solvent exerts a dual influence. On one hand, these solvents promote the cleavage of either the C-H bond (in COOH*) or the O-H bond (in HCOO*) through hydrogen bonding interactions. The electron transfer from the solvent to formic acid weakens the CO-H bond, facilitating its release and enhancing the decomposition rate. The strong interactions between formic acid and ionic liquids involve multiple hydrogen bonds. While some interactions promote the release of the H-atom, others stabilize the formic acid molecule, inhibiting its decomposition. The overall effect is a net stabilization of formic acid, leading to a lower decomposition rate.

2.4. Mechanism for Formic Acid Decomposition

Based on the characterization using FTIR and ¹H NMR, a mechanism for solvent-promoted formic acid decomposition is proposed. We chose to use ethanol as a representative solvent to explain the mechanism of the solvent promoting the decomposition of formic acid. Ethanol acts as a hydrogen bond acceptor, facilitating proton transfer and stabilizing transition states. In the formate route, ethanol promotes O-H bond cleavage, while in the carboxyl route, it aids C-H activation. According to the FTIR and NMR results, the interaction mode between solvent and formic acid determines the choice of decomposition path: in blue shifted systems (such as water or ionic liquid), the strong hydrogen bond interaction between solvent and O-H preferentially promotes O-H bond breakage, favoring the formate pathway. In red shift systems (such as DMSO), the interaction between solvent and C=O may stabilize the carboxyl intermediate and affect the rate of the carboxyl pathway. Ethanol is a common polar proton solvent that exhibits high conversion rates and TON in formic acid decomposition, and its hydrogen bond interaction with formic acid is representative. We will focus on the analysis of spectroscopic data (e.g., NMR chemical shifts and FTIR spectra) of ethanol and formic acid as a basis for discussing solvent effects. For example, in Figure 1, ethanol (red pentagonal line) shows a higher rate of formic acid decomposition, which provides an experimental basis for subsequent mechanism analysis. The process begins with the preferential adsorption of formic acid onto the catalyst surface (Pt/C) via the carbonyl oxygen bond. The FTIR-ATR and NMR results suggest that formic acid orients itself toward a bridge site with the solvent through its hydroxyl group hydrogen, which may weaken the interaction between formic acid and the Pd/C catalyst. Following the adsorption step, the dehydrogenation reaction commences with rotation steps that break either the H-O bond or the C-H bond, forming adsorbed formate or carboxyl intermediates, respectively. Combined with the above characterization results and previous DFT results [34], the interaction between formic acid and the solvent can promote the reorientation of formic acid from a perpendicular trans-configuration to a nearly parallel cis-configuration (Figure 7), facilitating subsequent bond breaking. On the formate route, breaking the O-H bond in formic acid is exothermic on the Pd/C catalyst. Once the O-H bond in formic acid is broken, the next step is to break the C-H bond in the forming formate intermediate HCOO*. The interaction with polar aprotic solvents promotes the breaking of the oxygen-Pd bond, leading to the formation of monodentate HCOO and CO₂.

In the carboxyl route, hydrogen bonding complexes form between the carboxyl group and the solvent. This interaction involves a partial transfer of electron density from the O-H bond orbital of formic acid (the hydrogen bond donor, HBD) to the solvent molecule (the hydrogen bond acceptor, HBA). This electron density redistribution weakens the O-H bond, which is consistent with the observed upfield NMR shift (red shift) and promotes the breaking of the O-H bond. Additionally, the high gas solubility in ionic liquids shifts the equilibrium towards formic acid formation. Both water and ionic liquids interact not only with the hydrogen atoms in formic acid but also with the oxygen atoms (Figure 7).

3. Materials and Methods

3.1. Materials

The Pd/C (5 wt%), formic acid (98%), ethanol (99.5%), tetrahydrofuran (99.9%, THF) for HPLC, and ultrapure water for LC/MS were purchased from Macklin and used without further purification. The chloroform-d (99.8% with 0.03% V/V tetramethylsilane, CDCl₃) was obtained from Macklin. Dipolar solvent dimethyl sulfoxide (DMSO, 99.7%), γ-valerolactone (GVL, 99%), ionic liquid [BMIM][OAc] (≥95%), [BMIM]Cl (≥98%), and [BMIM][Tf₂N] (≥98%) were purchased from Sigma-Aldrich Corporation. All chemicals were used as received without further treatment.

3.2. General Procedure for Reaction Experiments

The decomposition of formic acid was carried out in a custom-designed 15 mL stainless-steel autoclave with pressure indicator reactor. The reactor was charged with 0.01 g of Pd/C, 3 mL solvent, and 0.05 mL formic acid. The reactor was flushed with N₂ gas ten times to remove the air and heated up with a magnetic stirrer at the desired reaction temperature. After the desired reaction time, the reactor was cooled to room temperature. Formic acid conversion was analyzed with a high-performance liquid chromatography (HPLC) equipped with UV/Refractive index detector and HPLC-RI SH 1011 column. The column oven temperature was 60 °C, and 0.5 mM aqueous solution of sulfuric acid was used as the mobile phase at a flow rate of 1 mL/min. The gas chromatography was performed on a Shimadzu (GC-2014) instrument with a TCD detector (100 °C) and SHINCARBON ST 50/80 mesh column (50 °C). The carrier gas was N₂ with 30 mL/min rate. The GC-TCD analysis results of the sample show that hydrogen gas and carbon dioxide were the products (Figure S1).

3.3. Spectroscopy Techniques

3.3.1. Measurement of Kamlet–Taft Solvatochromic Parameters

Binary mixed solvent systems (DMSO–FA, water–FA, or THF–FA) were prepared by mass using a microbalance (Mettler Toledo, Model AX 504, Greifensee, Switzerland) with an uncertainty of ±1 × 10⁻⁴ g at the corresponding mass ratios of 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, and 0.9 at room temperature, and the concentration of indicator in the mixed solvent systems ranged from 3 to 5 × 10⁻⁵ mol/dm. All UV-Vis spectra of the solution were measured with the UV-Vis spectrophotometer (Jasco, model V-530) at a resolution of 0.2 nm using a 1 cm path length quartz cell contained in an encapsulating case that was held at a constant temperature of 20 ± 0.1 °C using a temperature controller (Jasco, ETC 505, Tokyo, Japan). All UV-vis spectra of the solution were repeated three times. The maximum adsorption wavenumber (ν_mixture) of the solute in the mixtures was evaluated by numerically smoothing the raw data using the first-order derivative of adsorption and origin software (Origin 2023).

3.3.2. ATR-Infrared Spectroscopy

A series of formic acid–solvents mixtures were prepared by weight with mass using a microbalance (Sartorius, MSE 5245-000-DU). All infrared spectra of the solutions were measured with a resolution of 0.07 cm⁻¹ using an FTIR spectrophotometer (Jasco, model 6300, Tokyo, Japan) equipped with an ATR single reflection attachment (Jasco, ATR PRO 450-S) without temperature controller (room temperature 24 ± 1 °C, Jasco, Tokyo, Japan) from 600 to 4000 cm⁻¹. The angle of incidence was 45° with the Ge crystal.

H NMR Measurements

The ¹H NMR spectra of samples were obtained on a Bruker/AV-400 NMR spectrometer (Bruker, Billerica, Massachusetts MA, USA) at 25 °C with tetramethylsilane (TMS) as an internal standard in CDCl₃. The chemical shifts variation (∆δ) is defined as the difference between the chemical shifts in the mixture and those of the pure chemicals.

3.4. Calculations of Formic Acid Conversion and TON

The formic acid (FA) conversion (%) was determined as the ratio of the reduced FA concentration to the initial FA concentration.

$$Theformicacidconversion(\%)={\displaystyle \frac{{FA}_0-{FA}_t}{{FA}_0}}\times 100\%$$

Here, FA₀ means the initial concentration of formic acid and FA_t is the concentration of formic acid measured at time t.

TON is defined as the moles of formic acid converted per mole of active metal (Pd) in the catalyst, reflecting the catalytic efficiency. The calculation formula is as follows:

$$TON={\displaystyle \frac{n_{converted,HCOOH}}{n_{Pdincatalyst}}}$$

where n_{converted, HCOOH} means moles of formic acid converted and n_{Pd in catalyst} means moles of Pd in the 5 wt% Pd/C catalyst.

4. Conclusions

In conclusion, the effects of solvents in heterogeneous catalysis were investigated for the selective decomposition of formic acid using 5 wt.% Pd/C in various solvents, including water, organic solvents, and ionic liquids. Through spectroscopic and theoretical analyses, the role of solvents in formic acid decomposition in the presence of Pd/C was elucidated. The conversion yield of formic acid and the TON were found to be significantly influenced by the solvent, with higher yields and TON values generally observed in solvents with greater dipolarity/polarizability. The TON for formic acid decomposition into CO₂ and H₂ can be optimized by selecting solvents based on their dipolarity/polarizability. For instance, ethanol–water mixtures were found to enhance formic acid decomposition, demonstrating that such mixed solvent systems can effectively promote the reaction. The interaction between formic acid and the solvent plays a dual role, whereby hydrogen bonding between the solvent and formic acid can inhibit the adsorption of formic acid onto the catalyst surface and facilitate the decomposition of formic acid by stabilizing carboxyl/formate intermediates. Most importantly, this study found that solvents function through specific hydrogen bonding patterns: in polar solvents, a single solvent–hydrogen formate bond (solvent as HBA) weakens the O-H bond through electron transfer, promoting decomposition; in ionic liquids, the multiple hydrogen bond networks formed by anions and cations and formic acid have a “locking” effect on molecules, stabilizing formic acid molecules as a whole and inhibiting their decomposition. Simultaneously, we cannot rule out the direct interaction between solvents and the Pd/C catalyst surface, such as competitive adsorption or electronic modification, which may also contribute to the observed activity trends, especially in the case of ionic liquids. The choice of solvent is crucial for maximizing the efficiency of formic acid decomposition in heterogeneous catalysis. Solvents with high dipolarity/polarizability, as well as mixed solvent systems like ethanol–water, offer promising avenues for enhancing reaction performance. This study provides a foundation for improving efficiency in practical applications. Future research should focus on exploring additional solvent combinations and employing in situ surface characterization techniques to decouple the solvent–reactant and solvent–catalyst interactions.