Hydrazine-Assisted CO₂ Capture and TiO₂ Photoinduced Reactivity for Artificial Photosynthesis-Inspired Hydrogen Evolution

Abstract

A TiO₂/hydrazine system was investigated as a proof-of-concept platform for coupling chemical CO₂ capture with light-driven H₂ evolution under UV irradiation. Hydrazine served as the CO₂ capture agent, leading to the formation of carbamate-type intermediates, while TiO₂ acted as the photoresponsive solid. FT-IR, UV-Vis, and mass spectrometry analyses supported carbamate formation after CO₂ uptake and confirmed H₂ generation during irradiation, reaching a maximum of 33.2 μmol under the conditions evaluated. Deuterated experiments showed no detectable HD or D₂, indicating that H₂ evolution predominantly proceeded via hydrazine dehydrogenation rather than direct water splitting. On the basis of the available spectroscopic evidence, a tentative pathway involving carbamate intermediates and nitrogen-containing oxidation products is proposed. However, key control experiments required to confirm a strictly photocatalytic origin of H₂ evolution were not performed in the present exploratory study. Therefore, the observed behavior is more appropriately interpreted as preliminary photoinduced reactivity in a TiO₂/hydrazine/CO₂ system rather than definitive proof of a fully established photocatalytic mechanism. Overall, the results establish a preliminary proof of concept, while the limitations related to control experiments, product identification, quantification, and reproducibility are recognized.

1. Introduction

In a world increasingly affected by global warming, the energy sector faces the big challenge of identifying sources that are not only competitive but also guarantee sustainability and security. Actual energy systems depend on fossil fuels such as oil, natural gas and coal, whose combustion produces large quantities of carbon dioxide (CO₂) and other greenhouse gases into the atmosphere. It is estimated that around 75% of global CO₂ emissions originate from the energy sector, making it a principal agent of climate change [1,2]. This situation has motivated the development of efficient and innovative energy technologies, including the valorization of CO₂ to produce value-added products or synthetic fuels through processes like Artificial Photosynthesis, thus contributing to the decarbonization of the atmosphere.

Inspired by the natural process in plants, Artificial Photosynthesis uses light-induced reactions to capture CO₂ and convert it into solar fuels, such as green hydrogen and hydrocarbons [3,4]. Natural photosynthesis begins with light absorption, leading to energy transfers and chemical reactions that produce ATP and NADPH, which are then utilized for CO₂ fixation [5,6]. Artificial photosynthesis seeks to mimic this by employing semiconductor materials as replacements for living cells. As shown in Figure 1, UV irradiation of semiconductor materials such as titanium dioxide (TiO₂) can generate electron-hole pairs that may drive redox reactions relevant to artificial photosynthesis [7,8]. However, simulating this process is complex due to significant thermodynamic and quantum barriers, as both CO₂ (ΔG⁰ = 394.4 kJ·mol⁻¹) and H₂O (ΔG⁰ = 237.2 kJ·mol⁻¹) are highly stable molecules requiring substantial energy input for activation [9,10].

To address these obstacles, this work proposes the use of titanium dioxide (TiO₂), an abundant and low-cost photocatalyst that absorbs UV light, combined with hydrazine (N₂H₄), a strong reducing agent that acts as a chemical capturer of CO₂ to form carbamates, thus helping to overcome thermodynamic barriers [11,12]. From a renewable energy perspective, this approach is driven by the fact that synthetic hydrocarbons diversify the current energy portfolio, reducing dependence on fossil fuels. The key characteristic of synthetic fuels produced this way is the use of solar energy, an endless and clean energy source, creating a CO₂ consumption–production cycle that would reduce the carbon footprint in the atmosphere while enhancing energy security [13,14].

The main objective of this work is to explore a light-driven artificial-photosynthesis-inspired system combining water, hydrazine as a CO₂ capture agent, and TiO₂ as a UV-responsive solid. To achieve this, the specific objectives were to (i) construct and optimize a UV-LED photoreactor for TiO₂ activation, (ii) evaluate the capture of CO₂ by hydrazine to form carbamates, (iii) characterize the gaseous and liquid products formed upon UV irradiation using FT-IR and mass spectrometry, and (iv) propose a reasonable and explicitly tentative reaction pathway consistent with the available evidence. Given the exploratory nature of this study and the absence of critical control experiments, the goal is not to claim a fully validated photocatalytic solar-fuel process, but rather to examine whether coupling chemical CO₂ capture with UV-induced reactivity in the presence of TiO₂ can generate detectable hydrogen and chemically transformed intermediates.

2. Results

2.1. CO₂ Capture and Carbamate Formation

Controlled introduction of CO₂ into aqueous hydrazine produced carbamate-type species, as indicated by FT-IR analysis (Figure 2). Compared with the spectrum of the hydrazine solution prior to gas injection, the post-capture spectra showed new bands attributable to carbonyl-containing and carbamate-related functionalities. In particular, absorptions near 1680 cm⁻¹ were assigned to C=O stretching vibrations, while features in the 1400–1000 cm⁻¹ region were consistent with coupled C–O and C–N stretching modes. At the same time, the broad band in the 3700–3000 cm⁻¹ region, associated with overlapping N–H and O–H stretching vibrations, decreased in intensity, suggesting the participation of hydrazine functional groups in CO₂ capture.

The intensity of the carbamate-related bands increased with increasing CO₂ loading from 0.5 g to 2.0 g, indicating progressive formation of CO₂-derived intermediates in solution. These observations are consistent with the formation of hydrazine-derived carbamates in aqueous medium.

2.2. Mass Spectrometry Analysis of Gaseous Products (Batches 1, 2, and 3)

Mass spectrometry analysis of the gas phase showed the presence of molecular hydrogen in the irradiated reaction systems. A signal at m/z = 2 was consistently detected after irradiation, consistent with H₂ formation. In the deuterated experiments, no significant signals corresponding to HD (m/z = 3) or D₂ (m/z = 4) were observed, suggesting that the evolved hydrogen did not originate predominantly from the deuterated solvent. Figure 3 summarizes the comparative gas-phase response obtained under the different irradiation conditions.

Hydrogen quantification was estimated from the integrated mass spectrometric response assigned to m/z = 2 using the instrument response under a constant analytical configuration, and the resulting values are therefore presented as semi-quantitative estimates rather than as absolute gas-chromatographic determinations. Because the present study did not include external calibration with certified H₂ standards across the full concentration range, these values should be interpreted primarily as comparative measurements between experimental batches operated under the same analytical method.

Under this approach, the maximum estimated H₂ production was 33.2 μmol in Batch 1 (

Table 1. Quantification of hydrogen production for different experimental batches.

Batch	Reaction Conditions	H2 Produced (μmol)	H2 Yield (%)
1	N2H4·H2O, CO2, TiO2	33.2	0.0151
2	N2H4·H2O, CO2, TiO2, (NADP+, D2O in some tests)	18.2	0.0083
3	N2H4·H2O, CO2, TiO2, (NADP+, D2O in some tests), 380 nm LED	31.99	0.0146

). Batch 2 produced 18.2 μmol, while Batch 3 produced 31.99 μmol. In this work, hydrogen yield was estimated as the percentage ratio of experimentally estimated moles of H₂ to the theoretical maximum moles of H₂ obtainable from the initial hydrazine content, assuming complete dehydrogenation under the tested reaction conditions. On this basis, the highest calculated yield was 0.0151% for Batch 1, followed by 0.0083% for Batch 2 and 0.0146% for Batch 3.

Although these values indicate measurable H₂ evolution under irradiation, the yields are very low and should be interpreted only as preliminary proof-of-concept results. In addition, because the experimental batches differed simultaneously in several variables, including irradiation source, reaction time, reactor geometry, and the presence or absence of NADP⁺ and D₂O, direct causal comparison among batches must be made with caution. Most importantly, because essential control experiments were not conducted, the gas evolution observed here cannot yet be assigned unambiguously to a purely photocatalytic pathway.

2.3. FT-IR Analysis of Liquid Products (Batches 2 and 3)

FT-IR spectra of the post-reaction liquid samples revealed the persistence of carbamate-related bands together with additional features suggesting chemical transformation during irradiation (Figure 4). In Batch 2, broad absorptions in the 3700–3000 cm⁻¹ region were assigned to overlapping O–H and N–H stretching vibrations. Signals in the 3000–2800 cm⁻¹ interval suggested the presence of C–H-containing species, possibly associated with partially transformed organic products. The region between 1700 and 1550 cm⁻¹ contained bands compatible with carbonyl-containing species, while the 1550–1450 cm⁻¹ interval included contributions that may arise from overlapping C–N stretching and N–H bending modes. The complex 1200–1000 cm⁻¹ region likely contained overlapping contributions from C–O, C–N, N–N, and, where applicable, phosphorus-containing groups associated with NADP⁺. The 800–500 cm⁻¹ interval contained the expected Ti–O–Ti lattice vibrations from TiO₂.

In Batch 3, the spectra displayed similar features but with improved band intensity and spectral definition. Sample 5 showed an additional band in the 2250–2350 cm⁻¹ region. At this stage, this feature can only be assigned tentatively to strongly unsaturated nitrogen-containing species such as nitrile-like or carbodiimide-like functionalities; however, this interpretation is not conclusive from FT-IR evidence alone and would require confirmation by complementary structural methods.

Overall, the FT-IR data support the persistence of CO₂-derived nitrogen-containing intermediates after irradiation, but specific product assignments should be regarded as provisional because of band overlap and the lack of orthogonal molecular identification techniques.

2.4. UV-Vis Analysis of CO₂ Capture Samples

UV-Vis spectroscopy was employed as a complementary tool to examine the post-reaction liquid samples obtained at different CO₂ loadings. As presented in Figure 5, all samples showed a similar spectral profile, with a strong absorption maximum at approximately 228–230 nm, broad features in the 240–300 nm range, and a sharp decrease in absorbance above 300 nm. The intense band at 228–230 nm is consistent with π→π^∗ transitions of carbamate-related species, while the broader absorption between 240 and 300 nm may be attributed to overlapping n→π^∗ transitions of carbonyl- and nitrogen-containing products formed during CO₂ capture. Considering the reaction chemistry reported for hydrazine-based systems, these absorptions are more appropriately interpreted as the combined response of mono-carbamate and di-carbamate species in equilibrium rather than as signatures of isolated N–N or N–C bonds. The 2.0 g sample exhibited slightly more pronounced absorption in the 230–280 nm region, suggesting a higher relative abundance of UV-active intermediates, although the strong overlap of bands prevents an unambiguous assignment.

3. Discussion

3.1. Interpretation of the CO₂ Capture Step

The FT-IR data in Figure 2 indicate that hydrazine reacts effectively with CO₂ in aqueous medium to produce carbamate-type species, in agreement with previous reports on hydrazine-based CO₂ capture chemistry [15,16]. The appearance of carbonyl-related and C–N/C–O bands after gas loading supports the interpretation that the stable CO₂ molecule becomes chemically activated through carbamate formation.

This chemical activation step is likely central to the present system. Rather than relying exclusively on direct photocatalytic reduction of dissolved CO₂, the system first transforms CO₂ into a chemically bound intermediate that may then participate in subsequent light-driven reactions or interact with the TiO₂ surface.

3.2. Light-Induced Reactivity and H₂ Production

The production of H₂ under UV irradiation (Figure 3, Table 1) demonstrates that the TiO₂/hydrazine/CO₂ system is chemically reactive under illuminated conditions. However, in the absence of critical control experiments—such as TiO₂ + hydrazine without CO₂, hydrazine + CO₂ without TiO₂, and dark controls—the present data do not yet allow an unambiguous attribution of H₂ evolution exclusively to photocatalysis at the TiO₂ surface.

Within this limitation, the irradiated TiO₂-containing system is consistent with a photoinduced process in which UV excitation of TiO₂ may contribute to charge generation and redox chemistry, while hydrazine likely serves as the principal hydrogen source. The absence of HD or D₂ (m/z = 3 or 4) in the deuterated experiments strongly suggests that the evolved H₂ does not originate predominantly from solvent splitting, but rather from hydrazine dehydrogenation. This interpretation is also consistent with the presence of nitrogen-containing bands in the FT-IR spectra after reaction [17,18].

Accordingly, the current evidence supports the statement that H₂ evolution occurs under UV irradiation in the presence of TiO₂ and hydrazine, but the extent to which TiO₂-mediated photocatalysis, direct photochemistry, thermally assisted decomposition, or coupled pathways contribute to the observed response remains to be established experimentally [19,20].

3.3. Influence of Experimental Parameters on H₂ Yield

The differences in estimated H₂ production between batches can be associated with several experimental factors, although the exploratory design of the study prevents strict variable-by-variable interpretation. Batch 2 included NADP⁺ in some samples and showed lower H₂ yield compared to Batch 1. NADP⁺ is known to act as an electron mediator in natural photosynthesis [21], but in the present system its role remains uncertain. Its presence may have introduced competing redox pathways or contributed to the formation of additional phosphorus-containing and nitrogen-containing species, as suggested by the more complex FT-IR response in the 1200–1000 cm⁻¹ region. This interpretation, however, remains tentative.

The addition of D₂O in Batches 2 and 3 may also have influenced reaction kinetics through isotopic effects on hydrogen bonding and proton-transfer steps. At the same time, the absence of HD or D₂ supports the conclusion that the detected H₂ originated mainly from hydrazine rather than from the solvent.

A major operational difference existed between Batch 2 (UV-A lamp, 18 mW cm⁻²) and Batch 3 (380 nm LED array, 156 mW cm⁻²). The higher m/z = 2 response observed for Batch 3 suggests that stronger and more wavelength-matched irradiation enhanced the light-driven behavior of the system. Because 380 nm irradiation is close to the anatase TiO₂ excitation threshold, this condition may favor more efficient photoexcitation of TiO₂ [22,23]. Nevertheless, since the batches also differed in other aspects, this trend should be interpreted as suggestive rather than mechanistically definitive.

3.4. Tentative Reaction Pathway

Based on the available spectroscopic evidence, a tentative reaction pathway may be proposed, while explicitly recognizing that it is not yet experimentally validated in mechanistic detail. As shown in Scheme 1 (I), hydrazine first reacts with CO₂ in solution to form carbamate-type intermediates [15,24,25]. Upon UV irradiation in the presence of TiO₂, the system undergoes light-induced chemical transformation. One plausible interpretation is that hydrazine or hydrazine-derived species undergo oxidation/dehydrogenation steps, while reduction equivalents contribute to H₂ evolution.

The deuterated experiments suggest that hydrogen evolution is associated primarily with hydrazine rather than solvent-derived water splitting. In parallel, the persistence of carbamate-related bands and the appearance of additional nitrogen-containing signals in the FT-IR spectra indicate that some of the captured CO₂-derived species remain in solution and may participate in secondary transformations.

The resulting liquid-phase products have not yet been identified conclusively. Therefore, species such as nitrile-like or carbodiimide-like products should be considered only as tentative assignments consistent with selected FT-IR features, particularly the band in the 2250–2350 cm⁻¹ region. Overall, Scheme 1 should be understood as a working hypothesis intended to organize the observed spectroscopic trends.

3.5. Sustainability Considerations and Future Work

Although this work is inspired by concepts from artificial photosynthesis and CO₂ utilization, the use of hydrazine introduces clear environmental, health, and process-sustainability limitations. Hydrazine is toxic and hazardous, and its production is itself energy-intensive. Therefore, the present system should not be interpreted as a green or scalable energy technology in its current form. Rather, it serves as a laboratory model for exploring whether chemical CO₂ capture and light-induced H₂ evolution can be coupled in a single reaction framework [26,27].

Future work should focus first on the critical control experiments required to validate the origin of H₂ evolution, including: (i) TiO₂ + hydrazine under irradiation without CO₂, (ii) hydrazine + CO₂ under irradiation without TiO₂, and (iii) dark controls for the complete reaction mixture. In addition, future studies should include statistically supported replicates, external calibration for gas quantification, structural identification of liquid-phase products by complementary techniques, and mechanistic validation through photoelectrochemical and computational methods, including DFT.

A broader evaluation of energy balance, environmental impact, and process viability would also be essential before any technological implications could be considered.

4. Materials and Methods

4.1. Materials

Hydrazine monohydrate (N₂H₄·H₂O, 64–65%, reagent grade 98%), titanium dioxide (TiO₂, anatase powder, 99.8% trace metals basis), Deuterium oxide (D₂O, 99.9 atom %), and nicotinamide adenine dinucleotide phosphate (NADP⁺, disodium salt, approximately 98%) were purchased from Sigma-Aldrich (St. Louis, MO, USA). Carbon dioxide (CO₂, Research Grade UN 1013) and nitrogen (N₂, Research Grade UN 1066) were supplied by INFRA (Mexico City, Mexico). All reagents were used as received without further purification.

4.2. Photoreactor Design

A custom photoreactor was designed and assembled for the UV light-induced experiments. Two illumination configurations were used.

In the first configuration, a UV-A lamp emitting in the 380–400 nm range with an irradiance of 18 mW cm⁻² was employed. This setup was used for Batch 1 and Batch 2 experiments.

In the second configuration, implemented for Batch 3, the reactor was upgraded with an array of 200 UV LEDs centered at 380 nm. The LEDs were mounted on phenolic boards and provided an irradiance of 156 mW cm⁻² at the sample surface. This wavelength was selected to closely match the anatase TiO₂ band-gap excitation region.

A cooling fan (YOUNUON, model YN-Server-Cooling Fans, 12 V, 0.45 A; YOUNUON, Wuxi, China) was used to maintain thermal stability during prolonged irradiation periods. Batch 1 reactions were conducted in a 1 mL quartz cuvette supplied by YiXing Purshee Optical Elements Co., Ltd. (Yixing, China), whereas Batches 2 and 3 were carried out in hermetically sealed 25–50 mL borosilicate flasks (manufactured in Mexico City, Mexico) fitted with septum caps from KSTEK (Suzhou, China) to allow gas accumulation and sampling [24,25].

4.3. Experimental Procedure

4.3.1. Batch 1: CO₂ Capture and Photoinduced Conversion

Three samples containing 5.5 g of aqueous hydrazine were prepared [15]. CO₂ was introduced into each sample while the mass increase was monitored using an analytical balance (Mettler Toledo, MS105DU; Switzerland). Final CO₂ loads of 0.5 g, 1.4 g, and 2.0 g were obtained. The resulting solutions were then transferred to the reaction vessel, and 2.0 mg of TiO₂ was added. The suspensions were irradiated under the UV-A lamp configuration for 24 h with continuous magnetic stirring.

4.3.2. Batch 2: Photoinduced Conversion with UV-A Irradiation

Batch 2 experiments were scaled using the most favorable condition identified in Batch 1, corresponding to 2.0 g of CO₂ per 5.5 g of hydrazine solution. The required quantities of hydrazine, captured CO₂, and TiO₂ were placed in sealed borosilicate flasks. In selected experiments, D₂O (1 mL) and/or NADP⁺ were added. Before irradiation, the reaction system was purged with N₂ for 5 min to remove atmospheric oxygen. The flasks were then irradiated for two weeks under the UV-A lamp configuration with magnetic stirring.

4.3.3. Batch 3: Photoinduced Conversion Under 380 nm LED Irradiation

Batch 3 experiments followed the same general protocol as Batch 2 but were performed using the optimized 380 nm LED array instead of the UV-A lamp. Four tests were conducted under two-week irradiation conditions, varying the presence of NADP⁺, D₂O, and TiO₂ loading as defined in the experimental design.

Methodological note: The present study was exploratory and did not include a full factorial design. Therefore, the experimental batches should not be interpreted as isolated variable-by-variable comparisons.

4.4. Product Analysis

4.4.1. Gas Phase Analysis by Mass Spectrometry

After irradiation, gaseous products were displaced with N₂ and collected in 200 mL Tedlar gas-sampling bags. Gas analysis was performed using a mass spectrometer (DSMS HIDEN ANALYTICAL BALZERS TCP 015; WA, Hiden Analytical, Warrington, England) operating in residual gas analysis mode. The instrument was set to an emission current of 1000 μA, electron energy of 60 eV, and a scanned mass range of 0.4–50 amu with a step size of 0.05 amu. Gaseous species were identified according to their characteristic m/z values, with particular attention to H₂, HD, D₂, CO₂, O₂, and water-related fragments.

For the purposes of this exploratory study, the m/z = 2 response was used as an indicator of H₂ formation and as the basis for semi-quantitative estimation of hydrogen production under a fixed analytical setup. Because a full external calibration with certified H₂ standards was not performed, the absolute H₂ amounts reported should be interpreted as method-dependent estimates suitable primarily for internal comparison among experiments performed under the same instrumental conditions.

4.4.2. Liquid Phase Analysis by FT-IR Spectroscopy

After each experiment, the liquid phase was recovered and filtered through a 0.45 μm PES membrane to remove suspended TiO₂ particles. FT-IR spectra were recorded using a FT-IR spectrometer (PerkinElmer Spectrum GX; Shelton, CT, USA) in transmission mode over the range 4000–400 cm⁻¹ with a resolution of 4 cm⁻¹. Background subtraction was performed using a clean reference spectrum. Band assignments were made by comparison with literature data for carbamates, hydrazine derivatives, and related nitrogen-containing compounds.

4.4.3. UV-Vis Spectroscopy

UV-Vis spectra of selected liquid products were recorded as complementary evidence of chemical changes after CO₂ capture and photoinduced conversion. The post-reaction samples were analyzed in the wavelength range of 190–400 nm. Spectral profiles were compared for solutions prepared with different CO₂ loadings (0.5, 1.4, and 2.0 g) in order to identify variations in the UV absorption region associated with the formation of carbamate-type and other nitrogen-containing species.

5. Conclusions

A hydrazine-assisted CO₂ capture system combined with TiO₂ under UV irradiation was evaluated as an artificial-photosynthesis-inspired platform for light-induced hydrogen evolution. FT-IR results demonstrated that hydrazine efficiently captures CO₂ and forms carbamate-type intermediates in aqueous medium. Under irradiation, the system generated molecular hydrogen, confirming chemical reactivity in the illuminated TiO₂/hydrazine/CO₂ system under the tested conditions.

The maximum estimated H₂ production observed was 33.2 μmol, and the highest calculated H₂ yield was 0.0151% relative to the theoretical hydrogen obtainable from the initial hydrazine content. Deuterated experiments showed no detectable HD or D₂, indicating that hydrogen evolution originated predominantly from hydrazine dehydrogenation rather than direct water splitting. UV-Vis and FT-IR analyses further supported the presence of nitrogen-containing and carbamate-related species in the liquid phase after reaction.

At the same time, the present work has important limitations. No carbon-based synthetic fuel was identified or quantified, the experimental design was not fully controlled across all variables, and key control experiments were not included. Therefore, the study should be interpreted as a preliminary proof of concept of photoinduced reactivity, rather than as a demonstration of a fully validated photocatalytic solar-fuel process.

Future work should prioritize the corresponding dark and component control experiments, improve reproducibility and statistical treatment, strengthen gas quantification, identify liquid-phase products by complementary analytical methods, validate mechanisms through DFT and photoelectrochemical studies, and replace hydrazine with safer and more sustainable capture/reducing agents.