Introduction

Nitrous oxide (N2O) is one of the greenhouse gases that causes global warming. It has 300 times of the heat-trapping ability in the atmosphere more than carbon dioxide (CO2) compared to the same amount within 100 years [1,2,3]. N2O comes from both man-made and natural sources. Examples of the former are petrochemical industries (adipic acid, nitric acid plant, agriculture (use of nitrogen fertilizer), combustion processes of fossil fuels and biomass; while the latter’s are volcano emission and animal wastes. [1, 4,5,6,7]. The current industrial nitrous oxide abatement in Thailand is nitrous oxide decomposition [8]. The nitrous oxide is converted to nitrogen (N2) and oxygen (O2) at 600–800 °C, where the non-toxic gaseous product could be released to the environment. A part from being environmentally friendly, this method is costly because it requires a lot of external energy and does not give any valuable products.

Nitrous oxide can be utilized as oxygenated feedstock in chemical and energy production (i.e. fuel cells, production of syngas, carbonyl compounds, methanol, dimethyl ether, formaldehyde and so on) [9,10,11,12,13]. This work focuses on syngas production, as it is not only an energy source but also an importance feedstock in which it further conversion can give other valuable chemicals such as synthetic fuel, methanol, dimethyl ether, acetic acid etc. [9,10,11,12,13,14]. The nitrous oxide will be used as an oxidant providing a synergy of N2O utilization and syngas production simultaneously. Transition metals (Co, Fe, and Ni) [15,16,17,18,19,20], precious metals (Pd, Pt, Re, Rh) [11,12,13], and metal oxide compound alumina (Al2O3), lanthanum cobaltite (LaCoO3), lanthanum cobalt iron oxide (LaCoFeO3), lanthanum ferrite (LaFeO3), lanthanum manganite (LaMnO3), ruthenium oxide (RuO2), yttria-stabilized zirconia (YSZ) [15, 16, 18, 21,22,23,24,25,26,27].

The methane partial oxidation by nitrous oxide was studied widely in a packed-bed reactor [11]. Since both methane partial oxidation and nitrous oxide decomposition are exothermic reaction, therefore, a large amount of heat could be created causing temperature runaway. The conventional packed-bed reactor also requires a separation unit leading to a high investment cost. Membrane reactors have attracted significant attention due to its ability to function as a reactor and separator at the same time. However, the methane partial oxidation using nitrous oxide has not yet been researched in a membrane reactor. The membrane reactor could eliminate the complexity of the separation process and to ease complication of the reaction, narrow down the product distribution, increase selectivity, reduce energy consumption [12, 28,29,30,31,32,33]. This work studied methane partial oxidation and nitrous oxide decomposition which supposed to occur at the feed side and permeate side of the membrane reactor, respectively. Thermodynamics is useful for process assessment and optimal condition prediction, thus in this work, chemical equilibrium of the reactions were estimated using Aspen Plus program. The reactions were experimented using 10%NiO–La0.3Sr0.7Co0.7Fe0.3O3−δ (10%NiO–LSCF) to investigate actual reactions and their surface mechanisms. Ni and LSCF3773 was chosen according to its economically acceptable yield and its oxygen storage capacity, respectively [34,35,36,37].

Methodology

Possible reactions

The possible reactions in the membrane reactor are tabulated in Table 1. Table 2 represents relevant mathematical formulas for the changes of the standard Gibbs free energy calculation.

Table 1 The possible reactions in a membrane reactor
Full size table
Table 2 Equations for calculation of the changes of the standard Gibbs free energy
Full size table

At the feed side, nitrous oxide was decomposed on the surface of the membrane giving gaseous oxygen and nitrogen (1). Nitrous oxide could react with oxygen giving nitrogen dioxide as the undesired product (2). Oxygen was assumed to flow through the membrane to the permeable side and partially oxidize methane giving synthesis gas (4) and other by-products via associated reactions (5–11) [38,39,40]. The chemical equilibrium properties were calculated using Aspen Plus program. Changes of the standard Gibbs free energy were calculated based on heat capacity of each reactions [41, 42] using mathematical formulas presented in Fig. 5, where i denotes species of substance, γi is the stoichiometric coefficient of species i, ∆Cp is heat capacity. ∆rH, ∆rS, and ∆rG are the different of enthalpy, entropy, and Gibbs free energy of reaction, respectively. T is temperature. A, B, C, D, and E are characteristic constants [43,44,45].

Catalyst preparation and characterization

The aqueous solution of lanthanum (III) nitrate hexahydrate (La(NO3)3·H2O), strontium nitrate (Sr(NO3)2), cobalt Nitrate (Co(NO3)2·6H2O), and iron(III) nitrate (Fe(NO3)3·9H2O) were added stoichiometrically and stirred until homogeneous. EDTA-NH4OH, citric acid and ethylene glycol was added drop-wise while stirred continuously. The pH of the solution was kept at 6 to 7 using NH4OH. The resulting solution was evaporated at 80–90 °C. The residual muddy dark red gel was left in an oven overnight at 200 °C. The sample was calcined in moving air at 1100 °C for 12 h. giving LSCF3773 powder. The powder was mixed into Ni(NO3)2·6H2O solution. The same procedure of drying, evaporation and calcination were applied. The crystallinity of the catalysts were characterized using X-ray diffraction technique (XRD, Rigaku TTRAX III using Cu Kα radiation, λ = 1.5418 Å in a range of 30° < 2θ < 80° with a scanning step of 0.01°). The reducibility of the catalyst was studied by H2-TPR technique (H2 temperature-programmed reduction). Interaction between CH4 and catalyst surface was investigated by CH4-TPSR technique (CH4 temperature-programmed surface reaction). The catalysts was packed in a lab-scaled tubular reactor. 100 ml/min of 5%H2/Ar (for H2-TPR) or 5%CH4/Ar (for CH4-TPSR) was introduced into the system while temperature was increased at 5 °C/min from 25 to 950 °C. The consumption of the reactant and the formation of the product gases were measured by real-time quadrupole mass spectrometer gas analyzer, based on molecular weights for gases CH4 (16), CO2 (44), CO (28), H2 (2), and Ar (40).

Experimental set-up

The schematic diagram of the rig was presented in Fig. 1. 2 g of 10%NiO–LSCF3773 was packed into a quartz tubular reactor with an inside diameter of 10 mm. The reactor was placed in the middle of an electrical furnace (Chavachote, Thailand) with 30 cm heating zone. The catalyst was reduced in 40%H2 in Ar at 500 °C for 2 h before the reaction for all experiments. For nitrous oxide decomposition, total flow rate of 100 ml/min of N2O (99.95%, company) balanced in Ar (99.995%, BIG) at 1:1 ratio was fed into the system using mass flow controllers (Brooks instrument flow, 0–220 ml/min). A mixture of 50 ml/min of CH4 (40%CH4/Ar, company), 25 ml/min of O2 (40%O2/Ar, company), and 25 ml/min of Ar was introduced into the reactor for methane partial oxidation at 2:1 molar ratio of CH4 to O2. All the experiments were isothermally operated under atmospheric pressure at temperature range of 500–1000 °C using type K thermocouple and PID temperature controller. Gas compositions were analyzed using gas chromatography method (Shimadzu GC-2014ATP).

Fig. 1
figure 1

Rig’s diagram for nitrous oxide decomposition and methane partial oxidation

Full size image

Conversions percentages were estimated using the following equations;

$${X_i} = \left( {\frac{{{n_{i,in}} - {n_{i,out}}}}{{{n_{i,in}}}}} \right) \times 100\%$$
(22)

Here i denotes to N2O, CH4, and O2 and ni is mole or mole flow rate of species i.

Result and discussion

Characterization techniques

X-ray diffraction

Fig. 2 shows diffractograms of (a) NiO, (b) LSCF3773, and (c) 10%NiO–LSCF3773. The XRD pattern of 10%NiO–LSCF3773 has perovskite structure and NiO as major and minor crystalline phase, respectively. The Ni-based LSCF catalyst shown the mixed XRD diffraction peaks at 2θ = 23.20, 33.04, 37.32, 40.74, 43.36, 47.42, 53.94, 58.96, 62.92, 69.22, 75.46, 78.84 and 80.16, the diffractograms also confirmed the presence of NiO (ICDD card No.03-065-5745; 37.32, 43.36, 62.92, 75.46, and 80.16 corresponding to (111), (200), (220), (311), and (222) planes) with a cubic crystal structure. However, there was no peak shift in XRD pattern when more doped Ni content on LSCF, which noticed that the introduction of Ni loading did not alter the crystal structure of prepared-catalyst.

Fig. 2
figure 2

The XRD patterns; (a) NiO, (b) LSCF, and (c) 10%NiO–LSCF

Full size image

H2 Temperature-programmed reduction (H2-TPR)

Reduction property of the synthesized NiO-based LSCF catalyst were evaluated using H2-TPR. H2 consumption profiles of the catalysts were compared and shown in Fig. 3. The TPR profile of pure NiO showed two main reduction peaks centered at 390 and 500 °C, indicating reduction of Ni3+ ions to Ni2+ ions and reduction of Ni2+ ions to Ni0 ions, respectively [46,47,48,49,50]. For LSCF, the lowest temperature of reduction at 270 °C was attributed to surface oxygen reduction [51], while bulk oxygen reduction took placed during higher temperatures at 505, 625, and 735 °C. The bulk oxygen reduction indicate three stages of reduction (1) reduction of Co3+/Fe4+, Fe5+ ions to Co2+/Fe3+ ions at about 505 °C, (2) reduction of Co2+/Fe3+ ions to Co0/Fe2+ ions at 625 °C, and (3) reduction of Co0/Fe2+ ions to Co0/Fe0 ions at 735 °C [51,52,53]. The reduction temperatures of 10%NiO–LSCF showed more similarity to that of LSCF rather than of NiO. This is because the NiO level was not high enough to dominate the crystal structure [46, 54,55,56]. The H2 consumption could be interpreted as oxygen vacancy/oxygen lattice/oxygen storage capacity/oxygen deficiency, which was 11,188 µmol/gsolid, 5754 µmol/gsolid, and 6442 µmol/gsolid, for NiO, LSCF3773 and 10%NiO–LSCF3773, respectively. The 10%NiO was in situ reduced before each run at 500 °C, achieving Ni0 ions which acts as an active site for the partial oxidation of methane.

Fig. 3
figure 3

H2-TP–R profiles of synthesized catalysts; (a) NiO, (b) LSCF, and (c) 10%NiO–LSCF

Full size image

CH4 Temperature-programmed surface reaction (CH4-TPR)

CH4 dissociation on the surface of 10%NiO-LSCF was characterized by CH4-TPR technique, shown in Fig. 4. Four regions of temperature range can be considered according to its mechanisms.

Fig. 4
figure 4

CH4-TPSR profiles of 10%NiO–LSCF catalyst

Full size image

No CH4 consumption was detected at temperature lower than 450 °C, in this region I.

During temperature from 450 to 600 °C, referred to as region II in Fig. 4, CH4 dissociation and combustion were presumably occurred on Ni active site (Eq. 23), and adsorbed oxygen on the surface \(\left( {O_{\left( {ads} \right)}^*} \right)\) (Eq. 24) via Eley–Rideal mechanism,.

$$C{H_{4,(g)}} + Ni_{(s)}^* \to 2{H_{2,(g)}} + C \times Ni_{(ads)}^*$$
(23)
$$C{H_{4,(g)}} + 4O_{(ads)}^* \to C{O_{2,(g)}} + 2{H_2}{O_{(g)}} + {*_{(s)}}$$
(24)

A slight CH4 consumption and rapid H2 formation were correspondingly observed at the beginning of this region. Meanwhile, adsorbed oxygen on LSCF’s surface was reduced by CH4, giving H2O and CO2 as gaseous products. During the process, the oxygen vacancies was achieved (Eq. 24). These oxygen vacancies are to be re-oxidized by gaseous N2O during the N2O decomposition.

During temperature from 600 to 900 °C, shown as region III in Fig. 4, CH4 was significantly consumed while all the gaseous products, H2, CO, CO2 and H2O were found. Two main redox mechanisms were proposed as below:

  1. (1)

    CH4 partial oxidation/combustion occurred on the oxygen surface of LSCF, giving H2, H2O, CO, CO2 and oxygen vacancies as products, shown below as Eq. 25.

    $$C{{H}_{4,(g)}}+O_{(ads)}^{*}\to \left[ \begin{matrix} {{H}_{2,(g)}} \\ {{H}_{2}}{{O}_{(g)}} \\\end{matrix} \right]+\left[ \begin{matrix} C{{O}_{(g)}} \\ C{{O}_{2,(g)}} \\\end{matrix} \right]+{{*}_{(s)}}$$
    (25)
  2. (2)

    Langmuir–Hinshelwood surface reaction between the coked Ni site \(\left( {{{(C \times N{i^*})}_{\left( {ads} \right)}}} \right)\) (obtained from Eq. 23) and surface oxygen \(\left( {O_{\left( {ads} \right)}^*} \right)\) was occurred, producing gaseous products of CO and CO2. The oxygen vacancy \(\left( {_{\left( s \right)}^*} \right)\) was obtained as the result of coke combustion, between the adsorbed coke \(\left( {{{(C \cdot N{i^*})}_{\left( {ads} \right)}}} \right)\) and the lattice oxygen \(\left( {O_{\left( {ads} \right)}^*} \right)\), represented by Eq.  26, where the coked-Ni site \(\left( {{{(C \times N{i^*})}_{\left( {ads} \right)}}} \right)\) was regenerated back to its original form of active Ni \(\;\left( {Ni_{\left( s \right)}^*} \right)\). This regenerated Ni \(\;\left( {Ni_{\left( s \right)}^*} \right)\) could react cyclically with CH4 giving more H2 and coked-Ni \(\left( {{{(C \times N{i^*})}_{\left( {ads} \right)}}} \right)\), according to Eq. 23. The coked-Ni \(\left( {{{(C \times N{i^*})}_{\left( {ads} \right)}}} \right)\) would consequently reduce the LSCF surface by reacting with lattice oxygen \(\left( {O_{\left( {ads} \right)}^*} \right)\). The oxygen vacancy \(\left( {_{\left( s \right)}^*} \right)\) was then achieved via Eq. 26.

    $$C\times Ni_{(ads)}^{*}+O_{(ads)}^{*}\to \left[ \begin{matrix} C{{O}_{(g)}} \\ C{{O}_{2,(g)}} \\\end{matrix} \right]+Ni_{(s)}^{*}+{{*}_{(s)}}$$
    (26)
  3. (3)

    In the 4th region where the temperature was higher than 900 °C, CH4 concentration returned back to its initial level due to the absence of the adsorbed oxygen on the surface of LSCF before started to slightly decrease again by CH4 cracking over the coked-Ni active site. This resulted in an increase in H2 formation, where the CO, CO2 and H2O concentration, more or less, remain unchanged. However, H2O, CO, and CO2 re-adsorption could be suggested.

$$C{H_{4,(g)}} + C \times Ni_{(ads)}^* \to 2C \times Ni_{(ads)}^* + 2{H_{2,(g)}}$$
(27)

The results conclude that CH4 combustion was dominant when the temperature was lower than 600 °C. The desired CH4 partial oxidation occurred at temperature between 600 and 900 °C, while the CH4 cracking took part at 900 °C and beyond.

Chemical equilibrium estimation and effect of operating temperature

Changes of the standard Gibbs free energy \(\left( {{\Delta_r}{G^0}} \right)\;\) of the possible reactions listed in Table 1 were calculated using formulas in Table 2 shown below in Fig. 5. The selected temperature for the study was ranged from 500 to 1000 °C.

Fig. 5
figure 5

Changes of the standard Gibbs free energy of each reactions

Full size image

The result showed that the NO2 formation (reaction 2) was theoretically not possible for all temperature range while CH4 cracking could occur at temperature higher than 700 °C, approximately. From 500 to 600 °C, combustion of methane would be evidenced (reactions 4, 10, 11). Methane partial oxidation (reaction 4) would occur during temperature from 700 to 1000 °C. Coke formation could be deposited on the surface of the catalyst at temperature higher than 700 °C.

Figs. 6 and 7 present conversions of reactants and products of N2O decomposition and CH4 partial oxidation, which supposed to occurred at the feed side and permeate side of the membrane reactor, respectively.

Fig. 6
figure 6

a Conversion of the reactant (N2O) as a function of temperature. Mole fraction product of N2O decomposition using b no catalyst (homogeneous) and c 10%NiO–LSCF catalyst compared with equilibrium mole fraction using Aspen Plus program (line) run WHSV = 50 ml/g/min at 1 atm from 500 to 1000 °C

Full size image
Fig. 7
figure 7

a The temperature dependence of methane conversion. Mole fraction product at the permeate side of the membrane process using b no catalyst (homogeneous) and c 10%NiO–LSCF catalyst compared with equilibrium mole fraction using Aspen Plus program (line) run is WHSV = 50 ml/g min at 1 atm from 500 to 1000 °C

Full size image

Fig. 6a shows the highest N2O conversion of 68.71 was in the non-catalytic reaction and complete N2O conversion (99.13%) when using 10%NiO–LSCF as a catalyst. The conversion was close to the theoretical when using the 10%NiO–LSCF catalyst.

Fig. 6b represents non-catalytic nitrous oxide decomposition. The highest N2O conversion was obtained at 1000 °C giving N2 and O2 of 56.39 and 17.93 mol%, respectively, although both products began to be detectable at 800 °C. The productivities were increased with increasing temperature. The ratio of N2 to O2 was around 2.3–2.4 to 1 which disagreed with theoretical ratio from stoichiometry (2 to 1). There was no NO2 detected for all temperature range. In addition, NO was not measured in the experiment, although NO formation as an intermediate could be suggested via N2O oxidation/decomposition.

Fig. 6c shows that complete N2O conversion started to be achievable at 600 °C, giving 66.59 mol% of N2 and 32.78mol% of O2. Ratio of N2 to O2 was 2:1 during temperature from 600 to 800 °C, agreeing with its stoichiometry theoretically. At temperature higher than 800 C, the ratio was found to increase by the influence of oxygen inhibition, explained in the later part. The following reaction pathway could be suggested.

N2O was adsorbed and dissociated on the oxygen vacancy, achieved by CH4 partial oxidation via reaction (23)–(26), giving N2 and surface oxygen \(\left( {O_{\left( {ads} \right)}^*} \right)\), represented by Eq. 28. The resulting surface oxygen \(\left( {O_{\left( {ads} \right)}^*} \right)\), was reacted further with gaseous N2O via Eley–Rideal surface mechanism via Eq. (29), where more N2 and adsorbed oxygen molecule \(\left( {O_{2\;\left( {ads} \right)}^*} \right)\) were obtained. Afterwards, the adsorbed oxygen molecule \(\left( {O_{2\;\left( {ads} \right)}^*} \right)\) could be; (i) desorbed giving gaseous O2 while leaving the oxygen lattice vacant again \(\left( {_{\left( s \right)}^*} \right)\), demonstrated in Eq. 30, (ii) react with oxygen vacancy \(\left( {_{\left( s \right)}^*} \right)\) via Langmuir–Hinshelwood surface reaction, providing more surface oxygen \(\left( {O_{\left( {ads} \right)}^*} \right)\), shown in Eq. 31.

$${N_2}{O_{(g)}} + {*_{(s)}}\xrightarrow{{{k_1}}}{N_{2,(g)}} + O_{(ads)}^*$$
(28)
$${N_2}{O_{(g)}} + O_{(ads)}^*\xrightarrow{{{k_2}}}{N_{2,(g)}} + O_{2(ads)}^*$$
(29)
$$O_{2(ads)}^*\xrightarrow{{{k_3}\,}}{O_{2,(g)}} + {*_{(s)}}$$
(30)
$$O_{2(ads)}^* + {*_{(s)}}\xrightarrow{{{k_4}}}2O_{(ads)}^*$$
(31)

However, the gaseous oxygen produced via Eq. 30 could be re-adsorbed on the oxygen vacancy \(\left( {_{\left( s \right)}^*} \right)\), represented by Eq. 32. The surface oxygen was obtained.

$${O_{2,(g)}} + 2{*_{(s)}}\xrightarrow{{{k_5}}}2O_{(ads)}^*$$
(32)

Oxygen inhibition, where full surface coverage was achieved by oxygen re-adsorption, could occur if k1, k4 and/or k5 is larger than k2 and/or k3. This would be evident by N2 to O2 ratio higher than 2. However, as the ratio of N2 to O2 was approximately 2 to, this implies that k1, k4 and/or k5 is less than k2 and/or k3, meaning that reactions 28, 31 and/or 32 is the rate limiting step. At temperature from 900 to 1000 °C, the unwanted NO2 formation was utilized the surface oxygen, allowing Eq. 32 to go more forward. This leads to the decrease of O2, resulted in the higher ratio of N2 to O2. Mechanism of NO2 formation was proposed below, where the N2O adsorption/dissociation (reaction 28) occurred, then followed by the below Eley–Rideal surface reactions.

$${N_2}{O_{(g)}} + O_{(ads)}^*\xrightarrow{{{k_6}}}2N{O_{(g)}} + {*_{(s)}}$$
(33)
$$N{O_{(g)}} + O_{(ads)}^*\xrightarrow{{{k_7}}}N{O_{2,(g)}} + {*_{(s)}}$$
(34)

Fig. 6 (line) shows the equilibrium molar fraction of N2 and O2 at 0.67 and 0.33, calculated using Aspen Plus program, for all temperatures. NO2 was not found. The results agreed with the calculation reported in Fig. 5, where reaction 1 was spontaneous and reaction 2 was thermodynamically limited. The catalytic conversion of N2O and its productivities (Fig. 6c) were found to approach the equilibrium approximation as shown in (Fig. 6 (line)) especially during temperature from 500 to 900 °C.

Fig. 7 shows CH4 partial oxidation using no catalyst (b), in the presence of 10%NiO-LSCF (c), and (line) equilibrium concentration calculated using Aspen Plus program. From Fig. 7a, it can be seen that maximum CH4 conversion in non-catalyst, achieved at 1000 °C, was 53.51. CH4 consumption began to be observable at 800 °C, giving conversions of 2.87. In using 10%NiO–LSCF catalyst, complete conversion of CH4 was obtained at 1000 °C, which gives conversion that are close to the theoretical.

From Fig. 7b, less than 3mol% fraction of H2, CO2, H2O, C2H4, and C2H6 were produced at this temperature. Reverse water gas shift reaction (RWGS), which is the combination of reaction (10) and 11, was evident as H2O production was noticed. The Boudouard reaction, the combination of reactions 4, 9, and 10, was not observed, as there was no CO2 formation, therefore, coke formation did not occurred at this temperature range. On the other hand, CO2 was detected at higher temperature ranging from 900 to 1000 °C, suggesting the possibility of the Boudouard reaction, therefore, coke formation was also predicted. In addition, H2, CO, CO2, and H2O molar fraction were increased with increasing temperature. This indicated that complete combustion, reverse water gas shift and the Boudouard reaction were presumably taken place. However, C2 + was increased at temperature from 800 to 900 °C, then decreased again at 900 to 1000 °C. This implied that reactions 5 and 7 were driven backward at temperature lower than 900 °C, and moved forward at temperature higher than 900 °C.

From Fig. 7c, H2O and CO2 were detected at temperature lower than 700 °C and found to decrease with increasing temperature. Small amount (less than 1mol%) of H2O and CO2 were found at temperature higher than 700 °C. On the other hand, H2 and CO were increased with increasing temperature. The reaction temperatures of the CH4 partial oxidation, CH4 combustion and CH4 cracking shown in Fig. 7c were correspondingly agreed with those shown in Fig. 4. The catalytic experimental result was found to approach the equilibrium composition, shown in Fig. 7c, and corresponded with calculation of Gibbs free energy changes tabulated in Fig. 5. The negative value of ∆Go of reaction 4 became larger with increasing temperature, while that of reactions 10 and 11 became smaller. This means that reaction 4 was more spontaneous at higher temperature leading to higher yield of H2 and CO, while reaction 10 and 11 were more thermodynamically limited resulting in less amount of CO2 and H2O production. At temperature higher than 700 °C, the possible mechanism was proposed as follows;

The surface oxygen or lattice oxygen \(\left( {O_{\left( {ads} \right)}^*} \right)\), produced from reaction (28), assumed to diffuse through the membrane from the feed side and reacted with CH4, giving adsorbed CO \(\left( {CO_{\left( {ads} \right)}^*} \right)\) and gaseous H2 as shown in Eq. 35. The adsorbed CO \(\left( {CO_{\left( {ads} \right)}^*} \right)\) was afterwards desorbed to gaseous CO and left the active site \(\left( {_{\left( s \right)}^*} \right)\) free again, represented by Eq. 36:

$$C{H_{4,(g)}} + O_{(ads)}^*\xrightarrow{{{k_8}}}CO_{(ads)}^* + 2{H_{2,(g)}}$$
(35)
$$CO_{(ads)}^*\xrightarrow{{{k_9}}}C{O_{(g)}} + {*_{(s)}}$$
(36)

CO2 and H2O splitting were occurred on the oxygen vacancy \(\left( {_{\left( s \right)}^*} \right)\) as shown in Eq. (37) and (38), producing CO and H2 as gaseous products. On the other hand, the oxygen vacancy \(\left( {_{\left( s \right)}^*} \right)\) was simultaneously re-oxidized by CO2 and H2O becoming the adsorbed oxygen \(\left( {O_{\left( {ads} \right)}^*} \right)\);

$$C{O_{2,(g)}} + {*_{(s)}}\xrightarrow{{{k_{10}}}}C{O_{(g)}} + O_{(ads)}^*$$
(37)
$${H_2}{O_{(g)}} + {*_{(s)}}\xrightarrow{{{k_{11}}}}{H_{2,\;(g)}} + O_{(ads)}^*$$
(38)

From Fig. 7c, no significant amount of C2 + and C3 + were found for all temperature range, indicating the occurrence of reaction (5), (6), (7), and (8) where C2 + and C3 + were reacted with oxygen giving H2 and CO. The result agreed with the standard Gibbs free energy changes, calculated and shown in Fig. 5. From Fig. 5, the standard Gibbs free energy changes of reaction (9) was positive at temperature from 500 to 600 °C, and became negative at temperature higher than 600 °C. This means that coke formation via reaction (9) was not occurred at 500 to 600 °C, but instead was likely to occur at temperature higher than 600 °C.

Figure 8 compared syngas productivities and its ratio between obtained from catalytic process using 10%NiO–LSCF, and from the chemical equilibrium calculation. The experimental result was found to be in good agreement with the calculation. Syngas ratio was approaching 2 at temperature from 700 to 1000 °C.

Fig. 8
figure 8

Experimental catalytic performance using 10%NiO–LSCF compared with equilibrium composition

Full size image

Conclusion

10%NiO–LSCF was a promising catalyst for high purity syngas production via methane partial oxidation using nitrous oxide as an oxidant in a membrane reactor. Surface mechanisms of the reactions, at both feed side and permeate side were proposed, based on the experimental results. The catalytic experimental result was in good agreement with the chemical equilibrium properties, calculated using Aspen Plus. The optimal operating temperature for the reaction carried out in a membrane reactor is 800 °C under atmosphere pressure. At this temperature, NO2 formation was not occurred, and full conversion of N2O decomposition at the feed side, and methane partial oxidation at the permeate side were achieved. Heavier hydrocarbons, i.e. C2 + and C3 + , should not be found. High purity syngas could be obtained at the ratio of H2 to CO equal to 2.