Introduction Biomass as a near-zero emission and novel energy source is regarded as an optimal alternative to fossil fuels. Biomass gasification technology is extensively used due to its superior efficiency in energy utilization, environmental congeniality, and sustainability. However, tar as a predominant by-product of this process presents some formidable challenges, i.e., combustion difficulties, potential clogging of boiler flue gas ducts, damage to gas-operating machinery, and health risks. These issues constitute a significant impediment to the industrial deployment of biomass gasification technology. Addressing the efficient processing of tar is pivotal for the extensive utilization of biomass gasification. The strategy of employing catalysts for the catalytic reforming of tar to produce hydrogen is emerging as an effective solution, characterized by lower requisite temperatures and enhanced conversion rates. Biomass-derived carbon with its cost-effectiveness, high specific surface area, and outstanding thermal stability is projected to have a significant potential in the transformation of biomass tar. Nevertheless, the suboptimal metal loading capability in the catalyst preparation from biomass carbon results in a reduced catalytic activity and an inferior hydrogen selectivity during the reforming process. Furthermore, there is a notable deficiency in research pertaining to the tar produced in actual production processes. Consequently, there is an urgent need for comprehensive investigations into the mechanism of hydrogen generation via catalytic reforming using biomass-based carbon catalysts.Methods An apricot shell powder with a granularity of 100-150 mesh was utilized as a feedstock for char production. A biomass char carrier was obtained via carbonizing this material at 350 ℃ for 30 min. Subsequently, the biomass char carrier was impregnated and stirred in a 1 mol/L KOH solution for 3 h, followed by a filtration process, with an impregnation ratio of 1 g of char carrier to 10 mL of KOH solution. The impregnated material was firstly calcined at 300 ℃ for 2 h, and then at 800 ℃ for 1 h 30 min. For post-activation, the biomass char underwent oxidative modification. It was washed to a pH value of 7, impregnated with a 15% hydrogen peroxide solution for 3 h, further washed until the pH value was 7, and then dried the post-filtration. Nickel and cobalt were chosen as metals for loading by impregnation. A 100 mL solution of 1 mol./L Ni(NO3)2 and Co(NO3)2 was prepared, and a mixture of cobalt nitrate and nickel nitrate was configured in a ratio of 1:4 to produce a 100 mL mixed solution. This solution was then impregnated with the prepared char carrier in a water bath at 90 ℃ for 4 h and filtered. After drying for 12 h, the material was calcined at 800 ℃ for 4 h. For the experiment, a cold trap method was employed to collect tar at the outlet of a straw gasifier under atmospheric pressure. The experiment primarily utilized a vertical tube furnace, enabling the catalytic reforming of tar inside a quartz tube. A catalyst bed was placed inside the tube furnace. A basket containing tar was set in the quartz tube in the heating furnace. In the tube furnace, the tar transformed into a gaseous phase and passed through the catalyst bed, undergoing a reforming reaction with the catalysts on the bed. The reaction products entered two conical flasks containing dichloromethane at the lower part of the tube furnace. These flasks were placed in an ice bath pot to collect the residual tar, and after drying, the gases were collected in gas bags for gas chromatographic analysis. The prepared catalyst was characterized by specific surface area analysis, scanning electron microscopy and X-ray diffraction.Results and discussion The XRD patterns indicate that during thermal treatment, nickel forms metal-carbon complex compounds with surface carbon, while cobalt results in cobalt oxide formation on the biomass char surface. The SEM images revealed a porous structure on the Ni/C surface, with metal-carbon complexes causing indentations on the char carrier, evidenced by visible surface metal particles. A distinctive feature of Co/C compared to nickel-based catalysts is the presence of etched grooves on the surface. The EDS analysis showed an even distribution of nickel on the Ni/C surface with a localized enrichment, alongside some surface collapse. The nickel loading on the catalyst surface is 12.01%. Cobalt displays a more uniform distribution on Co/C, indicating a fewer cobalt oxide enrichment sites.The experimental results with and without catalysts for tar catalytic reforming demonstrate that catalyst addition increases a tar reforming efficiency and H2 and CO contents in syngas. Tar reforming conversion rate increases from 36.12% to 76.67%, and H2 and CO contents in syngas increase from 22.76% and 28.80% to 25.62% and 32.78%, respectively, while CO2 content decreases from 27.11% to 17.45%. The results for the temperature impact on tar catalytic reforming reveal that an hydrogen yield initially increases and then decreases having the maximum value at 700 ℃. Tar conversion rate gradually increases from 76.67% to 94.95% with increasing the temperature, finally stabilizing after 800 ℃. An optimal steam/tar ratio (i.e., mSteam/mTar) is 3 based on hydrogen yield, production, and tar conversion rate. Similarly, an optimal tar/catalyst ratio (mTar/mCatalyst) is 2. Comparing single metal catalysts with Ni?Co/C catalyst for tar catalytic reforming, Ni?Co/C composite catalyst has a higher hydrogen yield in the gaseous products (i.e., 72.86%) with reduced production rates of C2H4, CO and CO2.Conclusions The preparation of carbon-based metal catalysts consists of two stages, i.e., the preparation of the char carrier and metal loading. The optimal temperature for preparing apricot shell char carrier was 350 ℃. Through KOH activation, its specific surface area increased from 5.502 m2/g to 1 124.16 m2/g, with the formation of abundant mesopores and micropores. Subsequently, H2O2 oxidation of the activated char carrier effectively enhanced the oxygen-containing acidic groups on its surface. The char carrier, prepared through carbonization, activation, and oxidative modification, exhibited a well-developed mesoporous and microporous structure suitable for efficient metal loading. On the surface of nickel-based catalysts, complex metal-nickel and carbon compounds formed, whereas cobalt- and iron-based catalysts formed metal oxides. The catalyst surfaces showed a good integration of metal with the carbon carrier, with uniform distribution. The carbon-based catalysts with a large specific surface area and a developed mesoporous structure were suitable for tar catalytic reforming. The results indicated that under nickel-based catalyst conditions, a hydrogen yield of 91.52 g H2 per 1 kg of tar could be obtained at the optimal parameters for hydrogen production from tar catalytic reforming (i.e., 800 ℃, mSteam/mTar of 3 and mTar/mCatalyst of 2). Cobalt-based catalysts exhibited a stronger catalytic activity and a higher hydrogen selectivity, compared to nickel-based catalysts. However, nickel- and cobalt-based catalysts both demonstrated a lower catalytic activity, compared to the Ni?Co/C composite catalyst.