Reduction of CO2 captured in basic solutions with biomass as reducing agent and metallic catalysts

Reduction of CO 2 captured in basic solutions with biomass as reducing agent and metallic catalysts† Maira I. Chinchilla, ab ´ Angel Mart´ ın, ab J. McGregor, c Fidel A. Mato ab and Mar´ ıa D. Bermejo * ab CO 2 capture and utilization technologies can make an important contribution to the decarbonization of industry. However, capture processes entail significant economic and energy costs, mainly associated with the purification, compression and transport of CO 2 . These costs would be reduced if captured CO 2 could be transformed in situ into useful products, avoiding purification, compression and transport costs. This work presents a hydrothermal process in which CO 2 absorbed in aqueous solutions as bicarbonate is reduced with biomass waste to give formic acid as a joint product of the biomass and CO 2 transformation, and acetic and lactic acids as byproducts from the decomposition of the biomass. Several biomass materials are applied as reductants: softwood, sugarcane bagasse, sugar beet, cork, pine needles, vermicompost and pure cellulose as reference material. Moreover, different catalysts are tested to improve conversion yield: Pd(5%)/C and Pd(10%)/C, Ru(5%)/C and activated carbon. The best results (18% formic acid yield) are obtained using pure cellulose as biomass and Pd(5%)/C catalyst. The next best results are obtained with the biomasses with the highest cellulose content, such as wood (11%) and sugarcane bagasse (9%). Experiments performed with labelled H 13 CO 3 − as carbon source at 300 °C using the Pd(5%)/C catalyst demonstrate that over 70% of the produced formic acid is formed from the inorganic bicarbonate carbon source. These high yields of conversion using renewable biomass as reductant can contribute to improve the technical and economic feasibility of CO 2 capture technology. Sustainability spotlight Currently there is great effort in reducing CO 2 emissions, but the cost of capturing and storage is very high. Developing technologies for producing useful chemicals and fuels from CO 2 will contribute to reduce their cost. In the present work a technology for simultaneously transforming CO 2 and biomass waste in formic acid and other organic waste. The main advantage of this technology is that can be directly coupled with absorption of CO 2 with NaOH and the effluent of the absorber can be taken directly to the reactor, avoiding purication, compression and transport of CO 2 that are the most expensive and energy consuming. It is aligned with several UN SDG, but more clearly with the 13th: climate actions. 1. Introduction Nowadays many efforts are being made to limit the increase in global temperature. In the last three decades, the emission of greenhouse gases has increased by 43%, with 65% of this being emissions of CO 2 . To reduce the emission rates of CO 2 into the atmosphere, research and innovation is being targeted at the development of new technologies promoting the conversion and utilization of this compound. 1–5 The utilization of CO 2 has aroused great interest since this compound can be used as raw material for manufacturing products of high value for industry such as formic acid, ethylene, ethanol, etc., 3,6–9 thus providing an economic revenue to compensate the capture costs and, in a longer term, a sustainable alternative to the current fossil-fuel based organic chemical industry. CO 2 can be directly used in carbonatation processes, or it can be converted to other compounds by means of electrochemical reduction, photoreduction or reduction with H 2 , among other alternatives. However, the high thermodynamic stability of the CO 2 molecule must be overcome to obtain high conversion rates. 10–14 Hydrothermal reduction has been considered as an alternative for the conversion of CO 2 because of the peculiar properties of water at high temperature and pressure, such as low dielectric constant, density and viscosity. Water at temperatures a Research Institute on Bioeconomy –BioEcoUVa, PressTech Group, Universidad de Valladolid, Dr Mergelina s/n., Valladolid, 47011, Spain. E-mail: mariadolores. [email protected] b Department of Chemical Engineering and Environmental Technology, Universidad de Valladolid, Dr Mergelina s/n., Valladolid, 47011, Spain c School of Chemical, Materials and Biological Engineering, University of Sheffield, Sheffield S1 3JD, UK †Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d4su00440j Cite this: RSC Sustainability,2025,3, 822 Received 4th August 2024 Accepted 16th November 2024 DOI: 10.1039/d4su00440j rsc.li/rscsus 822 |RSC Sustainability,2025,3,822–835 © 2025 The Author(s). Published by the Royal Society of Chemistry RSC Sustainability PAPER Open Access Article. Published on 21 November 2024. Downloaded on 2/17/2025 8:51:24 AM. This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. View Article Online View Journal | View Issue near 300 °C and pressures high enough to maintain it in the liquid state is characterized by a high ionic product and therefore a high concentration of H + and HO − ions, which can promote acid/base catalysed reactions. The hydrothermal reduction of CO 2 has been particularly successful in forming formic acid when starting with CO 2 dissolved in slightly basic solutions as bicarbonates or carbamates, such as those formed when CO 2 is captured by amines or ammonia. These processes have the advantage that the wellknown absorption CO 2 capture process can be associated with an in situ reaction process in which the feed of the reactor is the effluent of the absorption column, without the intermediate separation steps and their associated costs. The reaction can be performed using metals such as Al, Fe or Zn as reductants 14–17 or directly with gaseous H 2 . 13,18 Yields of formic acid can be higher than 60% in residence times of 1–2h using bicarbonate at temperatures between 250 °C and 300 °C. In combination with catalysts and using gaseous H 2 as reductant, reaction temperature can be reduced to temperatures near 200 °C (ref. 14,17) or even to 120 °C in the case of ammonium carbamate or CO 2 captured by amines. 13,18 Among the catalysts studied, Pd/C catalysts have presented an especially good performance. 13,14 Jin et al. were the rst to prove that CO 2 captured as bicarbonate could also be reduced to formic acid using organic substances with alcohol groups such as glycerine or isopropanol. 19–21 This opened the possibility to simultaneously transform CO 2 and biomass, as it is well known that lignocellulosic biomass can be dissolved and hydrolysed in hydrothermal media to form sugars such as glucose and other subproducts with alcohol groups. 22–27 In 2018, And´ erez et al. showed that a number of compounds derived from the hydrothermal decomposition of lignocellulosic biomass could react with bicarbonate at 300 °C, pressures above 20 MPa to ensure liquid phase conditions and a residence time of 2 h to produce formic acid. 28 Further studies made with glucose as model organic reductant showed that part of the formic acid proceeded from inorganic CO 2 and the other was originated by glucose decomposition, and that this combined reaction increased the selectivity of the transformation of glucose to formic acid. 29 In 2022, Chinchilla et al. 30 showed the effect of several catalysts in the production of formic acid in solutions of sodium bicarbonate. Yields of formic, lactic and acetic acid were higher using sodium bicarbonate as inorganic carbon source than with ammonium carbamate, and the global yield was not improved or only slightly improved with the catalysts tested. However, with Pd and Ru supported in carbon, the proportion of formic acid coming from inorganic CO 2 increased. The possibility to perform the reduction of CO 2 with biomass residues allows to valorise lignocellulosic material that is commonly considered as residue. To convert these residues into products of high added value promoting the reutilization of biomass material and the reduction of CO 2 emissions into the atmosphere is becoming of great interest. Some authors are implementing microalgae residues from biodiesel extraction or molecules derived from proteins to be utilized as reducing agents for the conversion CO 2 captured. 31,32 And´ erez et al. 33 presented results of the simultaneous conversion of CO 2 and biomass (pine needles and sugarcane bagasse) at temperatures up to 275 °C, resulting in low yields of formic, lactic and acetic acid. Nevertheless, there is possibility to improve the yields of reaction of formic by adding catalyst that promote the hydrogenation of the CO 2 captured. Also, in the previous work the origin of the formic acid was not investigated. Thus, to advance in this investigation, a systematic study encompassing a wider range of biomass reductants and catalysts and assessing the origin of the organic products (either the inorganic CO 2 source or biomass decomposition) is required. In the present work, the conversion of sodium bicarbonate at 300 °C using different biomass reductants (microcrystalline cellulose, sowood, pine needles, sugarcane bagasse, vermicompost and sugar beet and cork waste) and catalysts (granular and powder activated carbon, Pd(5%)/C and Pd(10%)/C and Ru(5%)/C) is studied. The origin of formic acid produced in reactions at 200 °C and 300 °C is evaluated performing experiments with labelled H 13 CO 3 − as inorganic carbon source, with Pd(5%)/C as catalyst and the different biomasses tested. This research aims at assessing the best properties of biomass to act as CO 2 reductant in the hydrothermal process, as well as the best combinations of reductant and catalyst. 2. Material and methods 2.1. Materials All solutions were prepared with deionized water (conductivity 2 mScm −1 ). Sodium bicarbonate (99.7%), sodium bicarbonate 13C (98.5%), acetic acid (99.7%), lactic acid (88%), formic acid (97.5%), galacturonic acid (99%), D-glucose (99%), glycolic acid (99%), glycerin (99.5%), formaldehyde (37%), ethyleneglycol (99.8%), 1–2 propanediol (99.5%), acrylic acid (99%), ethanol (99.8%), acetone (99.5%), 5-HMF (99%), microcrystalline cellulose, lignin (99%), furfural (99%) and sulfuric acid (96%) were purchased from Sigma-Aldrich. Activated carbon and metal supported catalysts (Pd(5%)/C, Pd(10%)/C) were acquired from Sigma-Aldrich, Ru(5%)/C (50% wet paste) was purchased from Strem Chemicals. Xylose (99%), arabinose (99%) and glucose (99%) were used (Sigma-Aldrich) as standards for measuring sugars carbohydrates and lignin in biomass. Biomass materials used as reducing agents were obtained from different sources: pine needles and sugarcane bagasse (kindly provided by prof. James McGregor, University of Shef- eld), sugar beet (kindly provided by AB Azucarera Iberia S.L.U.), vermicompost (elaborated and kindly provided by Eyup Yildirir from Usak University, Turkey) and sowood chips (purchased from VitakraS.L.). 2.2. Preparation of biomass All biomasses were grounded and sieved in a mesh of 100 mm. The siing obtained was used to carry out the experiments. 2.3. Humidity The moisture content of all samples was measured with a Sartorius Model MAT 160. 80 °C was set as the drying temperature. © 2025 The Author(s). Published by the Royal Society of Chemistry RSC Sustainability,2025,3,822–835 | 823 Paper RSC Sustainability Open Access Article. Published on 21 November 2024. Downloaded on 2/17/2025 8:51:24 AM. This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. View Article Online 2.4. Batch hydrothermal reactions Solutions of sodium bicarbonate (SB) 0.5 M and SB 13 C 0.5 M were prepared gravimetrically in deionized water. 4.5 cm 3 of this solution was charged in each high-pressure batch reactor. In experiments, additionally 0.1 g of catalyst and 0.1 g biomass were charged in the reactor. Mass ratio catalyst to biomass was always 1 : 1, except in the experiments in which the inuence of the amount of biomass was studied in which amounts of 1, 0.05 and 0.025 g of biomass were used, but the amount of catalyst was kept constant in 0.1 g. The reactors used consist of stainless-steel high-pressure tubing of 1 2 00 diameter and 10 cm 3 of inner volume, closed with high pressure caps in the end. 14,28,30 A thermocouple was installed in the reactor to measure the inner temperature of the system. For safety reasons and to avoid overpressure during heating, reactors were lled up with liquids only up to 45% of their volume. The nal pressure is thus the vapor pressure of the water of the solution, plus the pressure generated by the gases released. Thus, a fraction of reactor volume not lled with liquid is necessary to avoid overpressure due to this possible gas release. Pressure was not measured inside of the reactor. The pressure inside the reactors was estimated as the saturation pressure of water at the working temperature. Once lled and closed, the reactors were introduced in a horizontal position in a custom-made uidized bed oven already preheated at the operational temperature. This oven consists of a bed of alumina uidized by air and containing electrical resistances. It can work at temperatures as high as 400 °C, and the uidization allows a rapid and uniform heating. Experiments were carried out at 200 °C, 250 °C and 300 °C. In experiments performed in a reactor equipped with a thermocouple it has been determined that the interior of the reactor reaches the operation temperature in 6–7 minutes. The reactors are kept inside the oven for 2 h and then, the reaction was stopped by quenching the reactor in cold water at 10 °C. Once cooled, the reactors were opened. No appreciable amount of gases was collected, thus the initial assumption that the inner pressure of the reactor is the vapor pressure of water is correct. All experiments were repeated at least twice. The uncertainty in the yields was estimated with the standard deviation of replications. The yield of the desired products was calculated by dividing the molar concentration of the product by the initial concentration of SB or SB 13 C as indicated in eqn (1). YProduct ¼CProduct CNaHCO3;initial 100% (1) Although only part of the formic acid comes from bicarbonate and the lactic and acetic acid come entirely from biomass, the yields have been expressed with respect to bicarbonate to facilitate comparison. 2.5. Analysis of solid samples of biomass 2.5.1. Total solids content. To determine the total solid content of biomass samples, 2 g of solids supported in an aluminum pan were placed in a convection oven at 105 °C for 24 hours. The weight of solids was recorded before and aer the treatment. The percentages of total solids (%TS) were calculated with eqn (2), Min all equations represent mass weighted in grams. %TS ¼MDry panþdry sampleMDry pan MSample as received100% (2) 2.5.2. Structural carbohydrates and insoluble lignin. 300 mg of the biomass sample previously dried at 105 °C were used. Each of the weighted samples were put in pressure tubes. Then, 4.92 g of H 2 SO 4 72% were added into each tube for each sample. These tubes were closed and then placed into a water bath at 30 °C for 1 hour. Aer this, 84 mL of water were added into each tube, and then they were put in a silicon bath at 120 °C for 1 h. The same procedure was carried out to analyze the standards. 60 mg of xylose, arabinose and glucose were used as standards. Once hydrolysis of samples took place, the solid and liquid phase were separated by ltration. 5–10 mL of the liquid phase were taken to neutralization to pH 6–7 with CaCO 3 .2mL of the liquid phase were then taken and ltered with 0.22 mm nylon syringe lter and later analyzed in HPLC. The solid phase was collected and placed in crucibles and located in the furnace at 575 °C. Samples were treated for 24 hours. Samples were weighed before and aer the treatment. The ash content was calculated as indicated in eqn (3). %Ash ¼MDry panþashMDry pan MDry panþdry sampleMDry pan100% (3) The percentage of acid insoluble lignin (%AIL) was calculated as shown in eqn (4). %AIL ¼MSample as received MDry panðMCrucibleþAsh MCrucibleÞ MDry panþdry sampleMDry pan 100% (4) The percentage of structural carbohydrates with ve and six carbon atoms aer the hydrolysis of biomass was calculated as shown in eqn (5) and (6), respectively. %C5 ¼ MArabinose from HPLCþMXilose from HPLC MDry panþdry sampleMDry pan!100% (5) %C6 ¼MGlucose from HPLC MDry panþdry sampleMDry pan100% (6) 2.6. Analysis of solid samples of catalysts 2.6.1. X-ray diffraction test (XRD). X-ray diffraction tests for different catalysts were carried out with a Bruker D8 Discover A25 diffractometer, equipped with a 3 kW generator and a 2.2 kW copper ceramic tube type FFF. The detector used was the LynxEye 40 kV 30 mA. 824 |RSC Sustainability,2025,3,822–835 © 2025 The Author(s). Published by the Royal Society of Chemistry RSC Sustainability Paper Open Access Article. Published on 21 November 2024. Downloaded on 2/17/2025 8:51:24 AM. This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. View Article Online 2.6.2. Scanning electron microscope (SEM) images. SEM images were taken in order to observe the external structure of the material. A Hitachi FlexSEM1000 equipment was used. Samples were previously prepared by covering the surface with a 10 nm layer of gold by using a Balzers SCD004 gold evaporator operated at 30 mA. The samples were in the evaporator for 40 seconds. BSEs (Backscattered electrons) for SEM images was carried out at 15 kV and a low vacuum of 30 Pa in a “FlexSEM1000”of the HITACHI brand. 2.6.3. Atomic spectroscopy. To evaluate the reutilization of the metal supported catalysts, the concentration of the metal in the catalyst was measured with atomic spectroscopy by ICPOES. 50–100 mg of pulverized sample was digested with 4 mL HNO 3 and 1 mL HF. Samples were digested in an Ethos Sel digester. The process was carried out in two steps: a rst step at 25 min, 250 °C, and 1500 W; followed by a second digestion at 10 min, 250 °C, and 1500 W. Subsequently, 350 mg of H 3 BO 3 is added and carried again in Ethos Sel digester under the following conditions: rst at 10 min, 160 °C, and 1500 W; and then a second digestion for 5 min, 160 °C, and 1500 W. These last stages allow the total liberation of F − of the solution to avoid the formation of insoluble uorides and corrosion of the quartz parts of the ICP used. Finally, aer digestion, the sample is diluted with MQ water and ltered with a 0.45 mmlter. To measure concentration the digested sample was introduced in a Microwave-ULTRAWAVE (Milestone) and 725-ES ICPOptical Emission Spectrometer (Agilent). 2.6.4. BET surface. The surface area of the catalysts was determined with N 2 adsorption–desorption isotherms in an Autosorb IQ (Quantachrome Instruments) at 1 bar. 2.6.5. Transmission electron microscope (TEM) images. TEM images were taken with a JEOL JEM-FS2200 HRP equipment featured with a 200 kV Field Emission TEM microscope. With accelerating voltages: 80, 100, 120, 160 and 200 kV, and high resolution HRP polar part (spot: 0.23 nm, line: 0.1 nm). 2.7. Analysis of liquid samples Liquid samples coming from the hydrothermal reaction were collected and ltered with 0.22 mm nylon lter to be furtherly analyzed by HPLC or NMR. The gas phase was not analyzed, as in all experiments no gas or a very small amount of gas was produced, and it was not possible to collect it. 2.7.1. High pressure liquid chromatography (HPLC). All liquid samples were analyzed in an HPLC Waters® (separation module Aliance e2695) with RI detector (Waters, module 2414) using a Phenomex Rezex ROA Organic Acid H + column. Temperatures of column and detector were 40 °C and 30 °C, respectively. As mobile phase, a 0.5 mL min −1 ow of 0.25 mM H 2 SO 4 was used. For structural carbohydrates analysis, a Shodex® column was used. Temperatures of column and detector were 60 °C and 30 °C, respectively. As mobile phase, a 0.8 mL min −1 ow of 0.25 mM H 2 SO 4 was used. 2.7.2. Nuclear magnetic resonance (NMR). The samples obtained with SB 13 C were analyzed with a 500 MHz Agilent instrument equipped with OneNMR probe. The acquisition parameters for 13 C NMR spectra were: 25 °C, 70 s relaxation delay between transients, 45° pulse width, spectral width of 31 250 Hz, a total of 16 transients and 1.048 s acquisition time. The inverse gated decoupling technique to suppress the nuclear Overhauser effect (NOE) was used to obtain quantitative measurements. The acquisition parameters for 1 H NMR spectra were: 25 °C, 70 s relaxation delay between transients, 90° pulse width, spectral width of 8012.8 Hz, a total of 4 transients and 2.044 s acquisition time. The PRESAT sequence was used in order to suppress the strong signal of water. 1 H and 13 C NMR chemical shis(d) were reported in parts per million (ppm) and referenced to tetramethylsilane (TMS). For all the samples, a capillary tube lled with chloroform was inserted into the NMR tubes to assure lock conditions. 3. Results 3.1. Biomass characterization In Table 1, the measured properties of the biomass samples before the reaction are reported. Table 2 presents the results of the elemental analysis of these samples. 3.2. Catalysts characterization In Fig. 1, the SEM images of the physical structure of the catalysts is presented. It can be observed that C granular (Fig. 1(a)), C powder (Fig. 1(b)), Pd(5%)/C (Fig. 1(c)), and Pd(10%)/C (Fig. 1(d)) presented an inner tubular structure. On the other hand, Ru(5%)/C (Fig. 1(e)) presents an amorphous structure. In Fig. 2 XRD patterns for all the catalysts before reaction experiments are presented. In Fig. 2(a), the typical XRD pattern of activated carbon in powder form presenting two broad diffraction peaks around 24°–26° and 42°–47° are observed. 34 All catalysts are carbon-based materials, thus these two wide peaks are repeated in all the XRD patterns of all studied materials. XRD Table 1 Properties of biomass before reaction Biomass % Humidity % Total solid content % Ash % Acid insoluble lignin % C5 % C6 Sowood 4.8 95.2 0.6 30 18 33 Sugarcane bagasse 6.6 93.4 1.1 25 23 31 Sugar beet 8.1 91.9 1.8 36 16 23 Cork 3.7 96.3 0.1 75 6 4 Pine needles 6.3 93.7 0.1 44 10 16 Vermicompost 4.6 95.4 38.7 29 2 2 © 2025 The Author(s). Published by the Royal Society of Chemistry RSC Sustainability,2025,3,822–835 | 825 Paper RSC Sustainability Open Access Article. Published on 21 November 2024. Downloaded on 2/17/2025 8:51:24 AM. This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. View Article Online pattern for C granular (Fig. 2(b)) presents several narrow peaks that correspond to crystalline structures that are identied as rests of calcite and quartz that can come from the origin source of this type of carbon. Pd(5%)/C and Pd(10%)/C (Fig. 2(c) and (d)) catalysts presented two narrow peaks around 40° and 47° that corresponded to the face-centered cubic structure of Pd supported. 34 For Ru(5%)/C narrow peaks around 44° correspond to Ru diffraction planes, the other narrow peaks corresponded to rests of quartz remaining in the carbon support. 35,36 In Table 3 the diameter and BET surface of the materials is presented. The C granular presented the lower surface area in comparison with the rest of the materials. 3.3. Inuence of the catalyst in the simultaneous reaction of biomass and CO 2 The main products obtained in all reactions are formic acid (FA), acetic acid (AA), lactic acid (LA). Glyceraldehyde, glycolaldehyde, formaldehyde, ethylene glycol, acetone, pyruvaldehyde, galacturonic acid, 5-HMF, among others, were detected to a lesser extent. All these components are typical products derived from biomass decomposition in hydrothermal media. 22,23 The experiments to determine the inuence of the different catalysts for the different biomass were carried out at a temperature of 300 °C with 0.1 g of catalyst and 0.1 g of biomass and 2 h of reaction time. The 2 h residence time was selected based on previous literature results which suggest that shorter reaction times are less effective in enhancing formic acid production. 19,20,28–30,33 The FA yields obtained under these conditions are shown in Fig. 3(a) (go to Section 1.1, 2.1, and 3.1 of the ESI to see the tabulated data†). It is observed that, using Pd(5%)/C catalyst, for most samples of biomass, the global FA yield was higher than without catalyst, in contrast to the results observed by Chinchilla et al. 30 in experiments with glucose as model organic reductant at 200 and 250 °C. In the cases of granular and powder carbon, the yields of formic acid are of the same order than without catalyst and in the case of Ru(5%)/C they are much lower. The best results were obtained with Pd(5%)/C catalyst, with yields that were even higher than those using Pd(10%)/C. The highest yield of FA obtained was 18%. This yield was obtained using pure cellulose as a biomass derivative (black circles), and with the Pd(5%)/C catalyst; followed by a 16% yield also obtained with cellulose and Pd(10%)/C (black circles). With the sowood biomass, a maximum FA yield of 12% and 10% was obtained using Pd(5)%/C and Pd(10%)/C (navy blue cross), respectively. The yield of formic acid is lower for cork (less than 6%) and pine needles (less than 7%). If we compare it with the composition of the different biomass presented in Table 3, it is observed that samples of biomass rich in cellulose presented higher yields of Table 2 Elemental analysis of biomass before reaction Biomass Element H/C O/C%N %C %H %O Sowood 0.00 48 6.1 43.3 1.5 0.7 Sugarcane bagasse 3.4 44.9 5.6 42.2 1.5 0.7 Sugar beet 0.2 46.9 5.7 41.5 1.4 0.7 Cork 0.4 60.5 7.6 27.5 1.5 0.3 Pine needles 2.0 48.1 5.9 37.7 1.5 0.6 Vermicompost 1.8 24.1 2.6 20.8 1.3 0.6 Fig. 1 SEM of catalyst before reaction: C granular (a), C powder (b), Pd(5%)/C (c), Pd(10%)/C (d), Ru(5%)/C (e). Fig. 2 XRD patterns for catalysts before use: C powder (a), C granular (b), Pd(10%)/C (c), Pd(5%)/C (d), Ru(5%)/C (e). Table 3 Physical properties of catalysts before reaction Catalyst Average diameter BET surface (m 2 g −1 ) C granular 2 mm 75 C powder 55 mm 650 Pd(5%)/C 45 mm 948 Pd(10%)/C 45 mm 880 Ru(5%)/C 50 mm 844 826 |RSC Sustainability,2025,3,822–835 © 2025 The Author(s). Published by the Royal Society of Chemistry RSC Sustainability Paper Open Access Article. Published on 21 November 2024. Downloaded on 2/17/2025 8:51:24 AM. This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. View Article Online conversion to formic acid. The biomass with the lowest yield of FA was vermicompost (less than 2% and not improved by using any catalyst), probably due to the high amount of ashes and the low amounts of C5 and C6 compounds (see Table 2). Theresultsoftheyieldsofaceticacidobtainedinthe experiments are presented in Fig. 3(b). It is observed that, in most cases, the yield of acetic acid improved with the use of catalysts and in general presented similar values for the different biomass samples, with the best results again being achieved with Pd(5%)/C. It is observed that the sugar beet produced lower acetic acid yields with powder and granular C catalysts. Cork presented lower values for Pd and Ru catalysts and sowood presented lower yields of acetic acid with Ru(5%)/C catalyst. The biomasses with which the highest amounts of acetic acid were obtained are sugarcane bagasse (4–6%), sugar beet (3.5–5%) and pine needles (3–4%). The Fig. 3 Effect of different type of catalysts in the yields of formic acid (a); acetic acid (b) and lactic acid (c) in the reaction of HCO 3 − with biomass. Conditions of reaction: mass ratio catalyst : biomass 1 : 1, 0.1 g of catalyst, 0.5 M NaHCO 3 , 45% filling volume of reactor, 300 °C, 2 h. © 2025 The Author(s). Published by the Royal Society of Chemistry RSC Sustainability,2025,3,822–835 | 827 Paper RSC Sustainability Open Access Article. Published on 21 November 2024. Downloaded on 2/17/2025 8:51:24 AM. This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. View Article Online biomass that produced the lowest yield of acetic acid was, in general, cellulose (2%). This can be explained because acetic acid is produced from the hydrolysis and oxidation of the sugars generated by the hydrolysis of hemicellulose and cellulose, but, additionally, the dissolution of hemicelluloses releases acetyl groups that are part of the hemicellulose structure. 37,38 Furthermore, the release of acid increments the hydrolysis reaction rate due to an autocatalytic effect. 39 Pd/C and Ru/C are some of the heterogeneous catalysts most used for hydrogenation applications. The high surface area (around 850 to 950 m 2 g, see Table 3) than allows palladium and ruthenium to disperse uniformly on the solid support, this provide high activity to the catalyst, which can explain the high yields of FA obtained when Pd(5%)/C is used in the reaction. However, Ru/C decompose formic acid and hydrogenate other compounds such as levulinic acid in acid media, 36 this decomposition can be promoting the lower yields of formic acid obtained in the reaction at 300 °C with this catalyst. Fig. 4 Effect of variation of temperature in the yields of formic acid (a); acetic acid (b) and lactic acid (c) in the reaction of HCO 3 − with biomass. Conditions of reaction: mass ratio Pd(5%)/C : biomass 1 : 1, 0.1 g of catalyst, 0.5 M NaHCO 3 , 45% filling volume of reactor, at 200 °C, 250 °C, 300 °C, 2 h. 828 |RSC Sustainability,2025,3,822–835 © 2025 The Author(s). Published by the Royal Society of Chemistry RSC Sustainability Paper Open Access Article. Published on 21 November 2024. Downloaded on 2/17/2025 8:51:24 AM. This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. View Article Online Regarding C powder and C granular, these materials have low BET surface (75 and 650, respectively, see Table 3) which reduce the activity of the material. Besides the low BET surface, it is noted that the addition of the metal in the carbon support enhances the activity of the carbon materials. Fig. 3(c) shows the lactic acid yields. It can be observed that the highest lactic acid yields were obtained when using granular C as catalyst and sugarcane bagasse (15%) and beetroot (16%) as biomass. It is observed that the LA yield slightly increase when using granular C in biomasses rich in sugar such as beet (from 14 to 16%) and sugarcane bagasse (from 9 to 15%). This is possibly due to the presence of saccharose (a dimer made of glucose and fructose), which decomposes faster than structural carbohydrate, and fructose released, that is rapidly converted to lactic acid promoted by the slightly basic media. 40 In the case of vermicompost and cork it is observed that the LA yield did not change signicantly when using adifferent catalyst. Fig. 5 Effect of different amount of biomass in the yields of formic acid (a); acetic acid (b) and lactic acid (c) in the reaction of HCO 3 − with biomass. Conditions of reaction: mass ratio Pd(5%)/C : biomass 1 : 1, 2 : 1, and 4 : 1, 0.1 g of catalyst, 0.5 M NaHCO 3 , 45% filling volume of reactor, 300 °C, 2 h. © 2025 The Author(s). Published by the Royal Society of Chemistry RSC Sustainability,2025,3,822–835 | 829 Paper RSC Sustainability Open Access Article. Published on 21 November 2024. Downloaded on 2/17/2025 8:51:24 AM. This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. View Article Online 3.4. Inuence of the temperature in the simultaneous reaction of biomass and CO 2 The experiments to determine the inuence of the reaction temperature for the different biomass were carried out using Pd(5%)/C as catalyst, as this gave the highest yield of formic acid, as reported in Section 3.2. The aim was to determine if the use of the catalyst could make possible the operation at lower reaction temperatures. Temperatures of 200, 250 and 300 °C were considered. 0.1 g of catalyst and 0.1 g of biomass were used and the reaction time was xed again in 2 h (go to Section 1.3, 2.3 and 3.3 of the ESI to see the tabulated data†). The yields of formic acid at different temperatures are shown in Fig. 4(a). It can be observed that the that the higher the reaction temperature, the higher the yield of formic acid obtained. For example, using pure cellulose, which is the biomass derivative that produces the highest FA yield, a yield of FA of 18% at 300 °C is obtained, a value that drops down to almost 4% at 200 °C. In the case of sowood, the yield falls from 11% at 300 °C to 1% at 200 °C. Similar behavior was observed in experiments with other biomass reductants. Fig. 4(b) shows that in most cases the AA yield increases slightly by increasing temperature. In the case of pine needles, the yield of AA remained approximately constant in the temperature range tested. In the case of sugar beet residues and sugarcane bagasse, it was observed that the AA yield was higher when increasing temperature from 200 °C to 300 °C (increasing from 3 to 5% and from 3 to 4% respectively). The same result was observed with cellulose, sowood, cork, and vermicompost (increasing from 0 to 2%, 1 to 6%, 2 to 3%, 1 to 2%). With respect to the yield to lactic acid (Fig. 4(c)), it was observed in most cases that this yield increased by increasing temperature, except when using sugar beet, with which the maximum yield of lactic acid was achieved at 250 °C. This increase was considerable in the cases of sugarcane bagasse and pine needles (from 2 to 9% and from 1 to 6.5%, respectively). With vermicompost, cellulose, cork and sowood, the LA yield increased slightly by increasing temperature (from 0 to 1.5%, from 3 to 4%, from 1.5 to 3%, and from 1 to 4%, respectively). 3.5. Inuence of the amount of biomass in the simultaneous reaction with CO 2 The effect of the amount of biomass added in the experiment, for a constant amount of bicarbonate was studied with 0.1 g Pd(%5)/C as catalyst at 300 °C, using 0.025, 0.05 or 0.1 g of biomass, at reaction times of 2 h. The yields of formic, lactic and acetic acids obtained in these conditions are shown in Fig. 3 (go to Section 1.2, 2.2, and 3.2 of the ESI to see the tabulated data†). Fig. 5(a) shows that, for cellulose, pine needles and cork, the yield of formic acid increased with the amount of biomass added. For example, with 0.1 g of cellulose, a yield of 18% was produced, while with 0.025 g the yield was 7%. In the case of beet and sugarcane bagasse, the FA yield increased slightly when using 0.05 g of biomass instead of 0.1 g (from 9.3% and 9.5%, to 8.5 and 9.1%, respectively). For vermicompost, the highest yield (4%) was obtained when a smaller amount of biomass of 0.025 g was added in the reactor. And´ erez et al. 29 found that the yield of the formic acid decreases when increasing the biomass (glucose), nevertheless the same authors 33 found this effect less important when using as biomass sugar cane bagasse in amounts similar to that used in this work at 250 °C. Fig. 5(b) shows that, using Pd(5%)/C as a catalyst, in the case of cellulose, sowood, pine needles, and vermicompost, the AA yield increases slightly (from 0 to 2%, 1 to 2%, 1 to 3%, and 0 to 2%) as the amount of biomass increases. In the case of beet residues, sugarcane bagasse, and cork, a higher production of AA is observed with the increase in the amount of biomass (from 1 to 5%, from 1 to 6% and from 1 to 4% respectively). In Fig. 5(c), it is observed that when using Pd(5%)/C as a catalyst, the LA yield increased as the amount of biomass increased in the experiments. The increment was higher when using pine needles, sugarcane bagasse and sugar beet (from 0 to 9%, from 1 to 7.5% and from 1 to 6.5% respectively). The increment in the LA yield of cork and cellulose was lower (from 0 to 4% and from 0 to 2.5%), while it was negligible for vermicompost samples (from 0 to 1.5%). 3.6. Origin of the formic acid at different temperatures and type of biomass According to previous studies, 29,30 FA can be formed via two routes, from the oxidation of biomass compounds or from the reduction of sodium bicarbonate (SB) (CO 2 source). NMR spectroscopy was used to determine how much formic acid comes from each species. The experiments were carried out with NaH 13 CO 3 (SB 13 C), which is an isotope of SB. The NMR carbon spectra in a typical experiment is shown in Fig. 6 (go to Section 4 of the ESI to see the NMR spectra†). In the 13 C NMR spectrum only the compounds derived from the reduction of CO 2 captured as SB 13 C are observed. The products observed were acetic acid (d=183) and formic acid (d =173). The peak at d=163 corresponds to SB 13 C that did not react and the peak at d=127 corresponds to 13 CO 2 dissolved in the sample. The peak at d=183, corresponding to acetic acid, does not indicate that acetic acid is a product of reaction of Fig. 6 NMR carbon spectra. Conditions of reaction: mass ratio Pd(5%)/ C : cellulose 1 : 1, 0.1 g of catalyst, 0.5 M NaH 13 CO 3 , 45% filling volume of reactor, 300 °C, 2 h. 830 |RSC Sustainability,2025,3,822–835 © 2025 The Author(s). Published by the Royal Society of Chemistry RSC Sustainability Paper Open Access Article. Published on 21 November 2024. Downloaded on 2/17/2025 8:51:24 AM. This article is licensed under a Creative Commons Attribution-NonCommercial 3.0 Unported Licence. View Article Online