Tri-reforming of methane over Ni@SiO2 catalyst

A nickel-silica core@shell catalyst was applied for a methane tri-reforming process in a fixed-bed reactor. To determine the optimal condition of the tri-reforming process for production of syngas appropriate for methanol synthesis the effect of reaction temperature (550e750 C), CH4:H2O molar ratio (1:0e3.0) and CH4:O2 molar ratio (1:0e0.5) in the feedstock was investigated. CH4 conversion rate and H2/CO ratio in the produced syngas were influenced by the feedstock composition. Increasing the amount of steam above the proportion of CH4:H2O 1:0.5 reduced the H2:CO molar ratio in produced syngas to ~1.5. Increasing oxygen partial pressure improved methane conversion to 90% at 750 C. At low ~550 C reaction temperature the tri-reforming process was not effective with low hydrogen production (H2 yield ~20%) and very low <5% CO2 conversion. Increasing reaction temperature increased hydrogen yield to ~85% at 750 C. From all the tested reaction conditions the optimal for tri-reforming over the 11%Ni@SiO2 catalyst was: feed composition with molar ratio CH4:CO2:H2O:O2:He 1:0.5:0.5:0.1:0.4 at T 1⁄4 750 C. The results were explained in the context of characterisation of the catalysts used. The obtained results showed that the tri-reforming process can be applied for production of syngas with composition suitable for methanol synthesis. Copyright © 2014, The Authors. Published by Elsevier Ltd on behalf of Hydrogen Energy Publications, LLC. This is an open access article under the CC BY license (http:// creativecommons.org/licenses/by/3.0/).


Introduction
Methane reforming is a well established industrial process for syngas production. The goal of methane reforming is to achieve high methane conversion with the required H 2 :CO ratio without coke deposition. The process can sometimes be combined with CO 2 conversion and utilisation. However, that process requires high energy input and there is a risk of coke deposition that would deactivate the catalyst [1].
In the novel process of tri-reforming, syngas production and CH 4 conversion is possible without CO 2 separation and with relatively low energy consumption [2]. The fact that it is not necessary to separate CO 2 from methane can reduce the cost of reforming and at the same time can reduce the volume of CO 2 emissions. The tri-reforming process combines the three generally used methane reforming processes into one. In the tri-reforming of methane in a single reactor, the following reactions are coupled: methane steam reforming (1), methane partial oxidation (2) and carbon dioxide reforming of methane (3): CH 4 þCO 2 /2CO þ 2H 2 À DH 0 298 þ247 kJ mol The highly exothermic complete methane oxidation (4) can also occur, which increases energy efficiency: In line with this, the process combines the endothermic reactions of steam (1) and dry (3) reforming with exothermic partial (2) and complete oxidation of methane (4). Addition of oxygen to the reactor can generate heat required by steam and dry reforming of methane and make the reforming process more energy efficient [3]. Halmann and Steinfeld proposed to combine the tri-reforming process with a lime carbonation [4] or with a carbothermic reduction of iron [5] to achieve the fully thermo-neutral process.
In the tri-reforming process, CO 2 is utilized in the methane dry reforming reaction (3). Moreover, flue gas from the combustion processes of power plants can be used as a CO 2 source for tri-reforming [2]. An average flue gas composition contains CO 2 3e16%, O 2 2e13%, H 2 O 6e8%, N 2 75e76% [6].
The tri-reforming process is more energy efficient than steam reforming or dry reforming of methane for production of syngas with H 2 :CO molar ratio in the range of 1.5e2, as required by methanol synthesis [5]. Adjusting the molar ratio of compounds during tri-reforming allows control of the H 2 :CO ratio in the produced gases to the required level.
The tri-reforming process can be used for transformation of low quality, CO 2 -rich natural gas into useful syngas. Trireforming is also a good option for hydrogen production from biogas [7]. In biogas, the molar ratio of CH 4 :CO 2 is usually sufficient for tri-reforming. Furthermore, the tri-reforming process can be used to obtain syngas by upgrading gas from biomass gasification.
During the reforming process of methane, coke formation sometimes occurs during methane cracking (5), the Boudouard reaction (6) and reduction of CO to carbon (7): Applying tri-reforming for reforming of methane can reduce the coke formation problem. During the tri-reforming process coke formation may occur simultaneously with the reduction of formed coke due to the coke oxidation (reaction 8 and reverse of reaction 7) Reaction (5) tends to generate carbon at higher (~750 C) reaction temperature. The thermodynamic equilibrium of reactions (6) and (7) can be shifted to the right side at low (~550 C) reaction temperature. At higher reaction temperature reactions (6) and (7) can be influenced by equilibrium limitations.
So far, the concept of methane tri-reforming has been studied at a laboratory scale and mathematical models of trireforming reactors have been analysed. Minutillo and Perna [6,8] using thermal efficiency calculations suggested optimal operating conditions for tri-reforming of 850 C and a molar ratio between flue gas and methane of 2e3.
The most popular catalysts for tri-reforming of methane are nickel supported by a wide range of different materials (Al 2 O 3 , ZrO 2 CeO 2 etc), similar to steam or to dry reforming. In tri-reforming of methane, there is a risk of re-oxidation of catalyst by oxygen present in the feed. For catalysts resistant to coke formation, re-oxidation can be the main reason of catalyst deactivation [9]. In line with this, not all catalysts applicable for methane steam reforming can be used for trireforming. Solov'ev et al. [3] observed that NiO/Al 2 O 3 catalyst which had almost 100% of methane conversion during steam reforming, gave a methane conversion of only 15% during trireforming.
The mixture of gases produced by the tri-reforming process contains mostly H 2 and CO and can be applied for solid oxide fuel cells (SOFC) and molten carbonate fuel cells (MCFC) for electricity generation [10,11]. The syngas produced during the tri-reforming process usually has a H 2 :CO ratio of 1.5e2 [12]. That proportion makes produced syngas suitable for use in the FischereTropsch process [13], for methanol synthesis [8] or for dimethyl-ether synthesis [14,15].
A variety of compositions of synthesis-gas can be used for the FishereTropsch process. The required H 2 /CO ratio is usually around 1.7e2.15 depending on the catalyst used and operating conditions [16,17]. The maximum output of methanol can sometimes be obtained for the H 2 /CO ratio 2.5 [18]. The ideal required H 2 /CO ratio for synthesizing methanol is 2. However, the reaction conditions and especially the addition of CO 2 can change the required ratio [19,20].
Silica is rarely used as a catalyst support in steam reforming because as the steam pressure is increased silica can become volatile. That would damage the support structure and reduce the surface area of a catalyst prepared by a wet impregnation method. The tested Ni@SiO 2 catalyst had the core@shell structure. In that structure the silica core is completely covered by a nickel shell and that should limit access to the silica surface. In addition, the reduced water partial pressure in the tri-reforming process can decrease the eventual carryover of silica in the form of vapour species. The core@shell structure of the catalyst had the advantage of high utilisation of metal and enhancement of catalyst activity. No significant change to the catalyst structure was observed after 4 h reaction. An alternative solution for the eventual silica vaporisation is to dissolve silica cores by alkaline solution to obtain nickel hollow nanospheres [23].
The aim of this work was to establish the optimum conditions for the tri-reforming process with a nickelesilica core@shell catalyst. The variables studied were chosen according to literature data and according to limitations of the setup used. The Ni@SiO 2 core shell catalyst was selected due to its satisfactory results in methane steam reforming reactions [21]. No reports have been found on characterisation of SiO 2 supported catalyst for methane and CO 2 conversion in the presence of O 2 and H 2 O. In addition, the application of core shell catalysts has not yet been studied for tri-reforming processes. However, some authors [22] tested catalyst prepared by a deposition-precipitation method. In this paper, an evaluation was made of the ability to control the H 2 :CO ratio in produced syngas by adjusting the molar ratio of compounds in i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 2 5 7 8 e1 2 5 8 5 feedstock to obtain synthesis-gas suitable for methanol synthesis.

Catalyst preparation and characterisation
The catalyst was prepared according to the method described previously [21]. Silica spheres were prepared by the St€ ober process and were covered by nickel using the depositionprecipitation method [23]. The St€ ober process combined with the deposition-precipitation of Ni allows reproductive formulation of particles with required size. Silica spheres were covered by three-dimensional Ni film. Core shell structured catalysts, especially those with a thin shell, exhibit enhanced catalytic properties. After drying in air at 105 C overnight, the prepared catalyst was calcined at 700 C for 4 h in air, at a heating rate of 5 C min À1 . Characterisation of freshly prepared catalyst was described previously [21]. It was confirmed that the catalyst had a core@shell structure with silica cores covered by Ni shells [21]. The coated catalyst had a nickel concentration of~11 wt% (11%Ni@SiO 2 ). The particle size measured from SEM micrographs was in the range of 0.7e1.0 mm with an average diameter of~0.9 mm for both the fresh and the spent catalyst. The surface area of the catalyst after calcination was~68.1 m 2 g À1 .
The properties of the spent catalyst were analysed using the same procedures as for the freshly prepared catalyst [21] for better understanding of catalyst activity in the trireforming process. The BET surface area, average pore diameter and total pore volume were determined by N 2 adsorption. The crystallization behaviour of the catalyst was analysed by X-ray powder diffraction (XRD). The surface chemistry of the catalysts were analysed using Fourier transform infrared spectroscopy (IR). The morphological properties of the tested catalyst were measured using a scanning electron microscope. The coke deposition was analysed using temperaturedependent mass change profiles (TGA). Samples of spent catalyst were examined by a temperature programmed oxidation method (TGA-TPO) by heating samples of the used catalyst (10e20 mg) in air at a flow rate of 50 ml min À1 over the temperature range 25e900 C at a rate of 10 C/min. Sample preparation techniques and details of equipment used were described previously [21].

Catalytic reaction
The performance of the 11%Ni@SiO 2 catalyst for tri-reforming of methane was tested in a bench scale, continuous flow, fixed-bed reactor. The stainless steel reactor with inner diameter 5 mm operated at atmospheric pressure was equipped with a gas flow control system and an on-line gas chromatograph (GC). For each test, 0.2 g of the catalyst (as a powder) was diluted with inert CSi (1:9). The catalyst was diluted to avoid gas pressure drop, particles clogging and catalyst bypassing. The reactor was installed into a tubular, electrically heated furnace. Prior to reaction, the catalyst was activated "in situ" at 650 C with hydrogen flow 10 ml min À1 during 1 h. After the reduction process, the reactor was purged by nitrogen (100 ml min À1 , 5 min) then a gas mixture of CH 4 :CO 2 :O 2 and He (carrier gas) was switched into the reactor. The catalyst was tested under a constant flow rate of CH 4 25 ml min À1 at a temperature range of 550e750 C. Steam was generated from deionised water fed to a heater (120 C) by a pump at a flow of 0e0.06 ml min À1 and mixed with the gas mixture directly before the reactor inlet. Helium was used as a carrier gas and as a balance in order to obtain the total gas flow constant. However, the flow rate of the feed varied accordingly to the amount of added steam for those experiments where the influence of steam partial pressure was investigated ( Table 1). The effluent was cooled by passing through an ice trap where eventual liquid products were condensed. The on-line GC was used for quantitative and qualitative analysis of reaction products. CH 4 and CO 2 conversions were calculated from the molar concentrations of CH 4 and CO 2 respectively at the reactor inlet and outlet. For the CO 2 it was a net conversion that included eventual CO 2 generation from reactions (4), (6) and (8) and water-gas shift reaction.
The results of tri-reforming obtained with the 11%Ni@SiO 2 catalyst were compared with commercial catalyst HiFUEL™ R110 from Alfa Aesar at the reaction condition optimal for the 11%Ni@SiO 2 catalyst. Fig. 1 presents results of catalyst activity experiments as a function of reaction temperature. All reactions were carried out separately for each tested reaction temperature. No significant deactivation of the tested catalyst was detected in any of reactions during the first 4 h. Conversion of methane and conversion of CO 2 increased with increasing reaction temperature. Additionally, the H 2 :CO molar ratio in produced syngas decreased at elevated reaction temperature. Increasing reaction temperature from 550 to 750 C resulted in increasing methane conversion from 24% to above 70%. Over the same range of temperatures CO 2 conversion increased more than 10 times from 4% to 52%. Simultaneously, the H 2 :CO molar ratio reduced from 3.7 to 2.6.

Results and discussion
Increasing reaction temperature favoured endothermic reactions like steam (1) and dry (3) reforming of methane. This is particularly the case for dry reforming, as the highly endothermic reaction became more intensive at elevated temperature. As a result, with increasing reaction temperature methane and CO 2 conversion increased and enhanced CO  production was observed. Moreover, the reverse water-gas shift reaction was thermodynamically favoured at high reaction temperature. The reverse water-gas shift reaction consumed part of the hydrogen produced by methane reforming and produced more CO. It was concluded that all added H 2 O was converted at elevated reaction temperature. There were no liquid products after all reactions with the molar ratio of CH 4 :H 2 O 1:0.5 in the feedstock. The residual oxygen and water concentrations were essentially zero. The tri-reforming process at the lowest reaction temperature of 550 C was not effective, with only just above 20% methane conversion (Fig. 2). Conversion of CO 2 was significantly lower than of methane at that temperature. In the trireforming process, CO 2 is not only the reaction substrate but also the reaction product. During the reaction at 550 C, the average CO 2 conversion was below 5%. Steam reforming (1) slowed down and dry reforming (3) probably terminated. Zhou et al. [24] suggested that dry reforming does not proceed at the temperature of the tri-reforming process below 650 C. CO 2 conversion at 550 C slightly decreased with reaction time, probably due to the carbon deposition (Fig. 2). At lower reaction temperature, the thermodynamic equilibrium favoured the forward direction of the water-gas shift reaction resulting in conversion of CO and steam to CO 2 and H 2 , with an accompanying increase of CO 2 and H 2 concentration in produced syngas. Moreover, for a period of time, the phenomenon of negative CO 2 conversion was observed for the reaction at 550 C (Fig. 2). CO 2 could be produced also by the complete (exhaustive) methane oxidation (4) at that reaction temperature. CO production was reduced at low reaction temperature, as it was thought that excess of added water reacted with CO in the water-gas shift reaction under these conditions. Fig. 3 presents results of methane tri-reforming under different proportions of oxygen to methane in the feedstock, during the constant CO 2 and steam partial pressure in the feed. To obtain constant feed flow rate and constant CO 2 and steam partial pressure with increasing O 2 partial pressure accordingly decreased partial pressure of He used as a balance. All reactions were carried out separately for 4 h at each condition with no significant deactivation of the catalyst. It was observed (Fig. 3) that syngas with slightly different H 2 :CO molar ratio was obtained by varying the O 2 partial pressure. Increasing the amount of oxygen added to the tri-reforming process improved methane conversion from 70% to 90%. However, hydrogen yield remained almost unchanged. Conversion of CO 2 decreased from 80% when no oxygen was added to reactor to~55% for reactions with added oxygen.
Increased O 2 partial pressure promotes partial (2) and complete (4) oxidation of methane and that could result in high methane consumption without hydrogen production. As a result, there was less methane available for dry (3) and steam (1) reforming. Addition of O 2 resulted in methane oxidation (4) and CO 2 production which in turn affected the net CO 2 conversion. CO 2 conversion could be also affected by the coke oxidation reaction (8). Oxygen reacts quickly and was assumed to have been completely consumed in the tested temperature range, since there was no oxygen measured in produced gas for    the all tested reaction conditions. Walker at al. [22] suggested that the effect of rapid oxygen consumption occurs as a result of high affinity of O 2 for the catalyst active sites.
The influence of added O 2 on CO 2 conversion is not clear. Solov'ev et al. [3] observed that the influence of oxygen concentration is related to the catalyst structure. They noticed that addition of oxygen with the NiO/Al 2 O 3 catalyst significantly reduced methane reforming, but for NiO/Al 2 O 3 eLa 2 O 3 or NiO/Al 2 O 3 eCeO 2 increased methane conversion and the H 2 :CO ratio, but reduced conversion of CO 2 .
It can be concluded that CH 4 conversion increased with increasing reaction temperature and with increasing oxygen concentration. In line with this observation, as the amount of added oxygen increased, the required reaction temperature to obtain high CH 4 conversion became lower. Moreover, the positive aspect of oxygen addition was to enhance catalyst activity by oxidation of deposited carbon (8).
With increasing the amount of steam added to the process from molar ratio of CH 4 :H 2 O ¼ 0 to 0.5 the CH 4 conversion increased form~30e73% and conversion of CO 2 decreased accordingly from 91 to 56% (Table 1). It can be concluded that in the situation where no water was added to the reaction (auto-thermal reforming) there was no competition for CO 2 to access to catalyst active sites. That resulted in intensive CO 2 conversion, above 90% and relatively small (~30%) CH 4 conversion. The ratio between H 2 and CO in the produced syngas was around 2. That was higher than the stoichiometric value obtained from only the dry reforming reaction (3). The additional hydrogen could be deduced to have come from methane partial oxidation and methane cracking reaction (5). The activity of the methane cracking reaction (5) could also explain the higher coke deposition. It must be pointed out that for the feed composition of The presence of steam promoted the methane steam reforming reaction (1). That led to higher concentration of H 2 in the produced syngas and higher (2.6) H 2 :CO molar ratio. The fact that the H 2 :CO ratio was still below 3 was probably due to the reverse water-gas shift reaction which consumed some of the produced H 2 . According to the literature [22] it is possible that water competes with CO 2 for active sites upon the catalyst and that resulted in reduction in CO 2 conversion. Reaction of CH 4 with steam (1) is thermodynamically favoured over the reaction with CO 2 (3). The coke deposition was very small, only 5 mg g cat À1 .
Methane conversion changed only slightly with further increase in the water partial pressure. The proportion between H 2 and CO in the produced syngas decreased to around 1.5, the value required for methanol synthesis. However, the amount of deposited coke increased significantly ( Table 1). The decrease in H 2 :CO molar ratio was probably due to the reverse water-gas shift reaction. The influence of steam on coke deposition is still not clear. Excessive water was collected by the ice trap where the amount of steam added to reaction was above the ratio CH 4 :H 2 O 1:0.5. The volume of redundant water increased with increasing water partial pressure in the feedstock. Table 2 presents analyses of the spent catalyst after exposure to different reaction conditions. It can be seen that both the reaction temperature and the amount of oxygen added to the reactor influenced the catalyst condition. Increasing reaction temperature from 550 to 750 C reduced the amount of deposited coke from 99 to 5 mg g cat À1 . The higher coke deposits in the reactions at 550 and 650 C were thought to be due to the fact that at lower reaction temperature, exothermic reactions were favoured. At temperature 550e650 C the thermodynamic equilibrium favoured the Boudouard reaction (6) and CO reduction reaction (7) which could result in high coke deposition. Increasing oxygen concentration from 0 to 0.2 M fraction decreased the coke deposition at constant reaction temperature of 750 C ( Table 2). Addition of oxygen in the proportion CH 4 :O 2 0.1 was sufficient to reduce the coke deposition to 5 mg g cat À1 from 24 mg g cat À1 without oxygen, although further increase in oxygen partial pressure did not further influence the coke deposit. It was thought that the effect of deposited coke reduction in the presence of oxygen was due to the carbon oxidation reaction (8). Moreover, higher oxygen concentration could increase temperature in the reactor by promoting oxidation and partial oxidation of methane. Higher temperature could shift thermodynamic equilibrium to the reverse Boudouard reaction (reverse of reaction (6)) and to reduction of coke deposition due to the presence of steam (reverse of reaction (7)). Furthermore, the reaction of oxygen with methane reduced the amount of methane available for eventual methane cracking reaction (5). Referring to Table 2, the surface area of the catalyst changed after reaction if the tri-reforming process was carried out at temperature above 650 C. The surface area decreased from 68.1 m 2 g À1 for the fresh catalyst to~41 m 2 g À1 after reaction at 650 C and to~26.1 m 2 g À1 after reaction at 750 C. Reduction in catalyst surface decreased with higher oxygen concentration in the feed. As the surface area of used catalyst decreased the average pore diameter increased. The decrease in the surface area of the catalyst could be due to the sintering of Ni particles, coke deposition or due to the formation of the clusters made of 2e3 catalyst particles. The eventual reoxidation of Ni also could affect the catalyst surface area. The TGA analyses of spent catalyst showed that the coke formed during the tri-reforming process was oxidized at temperature 585e620 C and that may suggest that the coke was in the graphite carbon form [25].  The surface area of the fresh unused catalyst after calcination was 68.1 m 2 g À1 .
i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 2 5 7 8 e1 2 5 8 5 SEM scans of the catalyst after 4 h reaction were taken to analyse the catalyst morphology (Figs. 4 and 5). The particles of the spent catalyst looked similar to particles of the fresh catalyst described previously [21]. Cracks or breaks in nickel coating were not apparent on catalyst grain structure. SEM micrographs of the catalyst after the tri-reforming process confirmed formation of coke on catalyst surface during the reaction at low temperature. SEM analyses of the catalyst after the reaction at 550 C showed some whisker structures believed to be carbon in the form of whisker carbon or nanotubes (Fig. 4). The number of carbon tubes after the reaction at 750 C was insignificant (Fig. 5), consistent with TPO analyses ( Table 2). The number of clusters made of 2e3 particles was slightly higher after the reaction at 750 C than that at 550 C. Fig. 6 presents results of FTIR analyses of the fresh and the spent catalyst. The characterisation of the freshly prepared catalyst was described previously [21]. The structure of the catalyst did not change significantly after the tri-reforming process. All the bands characteristic for the core@shell nickelesilica catalyst were detected for samples of the spent catalyst after all tested reaction conditions. Bands specific for nickel phyllosilicate (3649, 3629, 711 and 670 cm À1 ) became less intensive after tri-reforming, probably due to Ni sintering and due to the higher (750 C) reaction temperature. The catalyst was reduced at 650 C and 2:1 nickel phyllosilicate can decompose at higher than 650 C temperature [26].
The only difference between the fresh and the spent catalyst was the presence of several small new bands in the region of 1400e1600 and 2825e2890 cm À1 for the catalyst after reaction. The two minor bands at 2825 and 2890 cm À1 could be related to the symmetric and asymmetric stretching vibration of the CeH bonds associated to methylene groups connected to the aromatic rings or aliphatic groups [27]. After the trireforming process the "water" band at 1628 cm À1 from the bending of HeOeH bond had an additional shoulder at 1580 cm À1 . This was a frequency characteristic of aromatic skeletal vibrations. The spent catalyst also showed a band at 1485 cm À1 and 1445 cm À1 . According to Ryczkowski [28] those bands could be related to the aromatic CeC stretching.
Conventional coke usually has an aromatic character. The IR results obtained confirmed the presence of coke. However, from IR results of analysed samples it was difficult to determine unquestionably the form of deposited coke. Because of the small amount of deposited coke, all samples were analysed in situ without catalyst etching. As a result, intensive bands related to the catalyst surface overlapped part of the bands related to deposited coke.
No diffraction peaks were detected in the XRD patterns of the fresh or the spent catalyst. Even after reaction at high 750 C reaction temperature the Ni@SiO 2 catalyst was still in an amorphous form. The eventual Ni crystals were highly dispersed and were below the XRD detection level.
From all the tested reaction conditions the optimal for trireforming over the 11%Ni@SiO 2 catalyst was determined to be the feed composition with molar ratio CH 4    i n t e r n a t i o n a l j o u r n a l o f h y d r o g e n e n e r g y 3 9 ( 2 0 1 4 ) 1 2 5 7 8 e1 2 5 8 5 1:0.5:0.5:0.1:0.4 at T ¼ 750 C. For that reaction condition the H 2 :CO molar ratio was 2.5, being at the edge of acceptance for methanol synthesis. Methane conversion was high (73%) and coke deposition was very low (5 mg g cat À1 ).
Commercial catalyst HiFUEL™ R110 from Alfa Aesar with higher nickel loading 14.5% showed slightly lower CH 4 conversion (71%) and slightly higher CO 2 conversion (65%) at the same reaction condition. The molar ratio of H 2 :CO in syngas was lower 2.1 but coke deposition was much higher 126 mg g cat À1 .

Conclusion
The tested catalyst showed stable activity during the first 4 h of tri-reforming at 750 C. . In the tri-reforming process over the 11%Ni@SiO 2 catalyst above 70% methane conversion was obtained in the tested temperature range 550e750 C. The H 2 :CO molar ratio of~2.5 was obtained from the reaction at 750 C. Syngas with different H 2 :CO molar ratios was obtained by varying the feedstock gas composition. For the tested catalyst, increasing the amount of oxygen in the feedstock improved methane conversion to 90% and reduced coke deposition but did not improve the H 2 :CO ratio in produced syngas. Reduction of steam partial pressure to zero reduced the H 2 :CO molar ratio in produced syngas to value~2 and increased CO 2 conversion to over 90%. However, methane conversion decreased over the same range of compositions. Increasing the amount of added steam above the molar ratio CH 4 :H 2 O 1:0.5 reduced the proportion between H 2 :CO in produced syngas to~1.5 without affecting methane conversion. However, coke deposition increased when more water was added to the tri-reforming process. The surface area of the catalyst decreased after tri-reforming, probably due to the Ni sintering and formation of the clusters made of 2e3 catalyst particles. The Ni/SiO 2 catalyst showed slightly higher CH 4 conversion, lower coke deposition and lower CO 2 conversion compared to the commercial catalyst HiFUEL™ R110 at the optimal for the Ni/SiO 2 catalyst conditions. The results obtained showed that the Ni/SiO 2 catalyst can be considered as a promising catalyst for the tri-reforming process.