Synergistic effect of Zn with Ni on ZSM-5 as propane aromatization catalyst: effect of temperature and feed flowrate

This work shows how propane was catalytically converted to aromatic compounds over Zn-Ni/HZSM-5 to investigate the synergistic role of nickel as a zinc stabilizer and promoter in propane aromatization. Co-impregnation method was employed for Zn-Ni/ZSM-5 synthesis fixed 2 wt% of zinc and 1, 2 and 3 wt% of nickel. Modified catalysts’ were fully characterized. Catalysts crystallinity, structure and microporosity were retained from the analysis results. The propane aromatization process was conducted at 540 °C, 1200 ml/g-h gas hourly space velocity and atmospheric pressure. The presence of Ni with Zn improved catalytic performance for all Ni loading. Aromatic selectivity and propane conversion were improved, with the best performance to be Zn-Ni/ZSM-5 with 2 wt% Zn, 2 wt% Ni on ZSM-5 averaging 88 and 60% for twelve hours’ time on stream. Aromatic selectivity on Zn-Ni/ZSM-5 is eight times better than the parent HZSM-5 and one and half times better than the Zn/ZSM-5 catalyst. The electronic interaction of zinc and nickel resulting from equality of oxidation state of +1 and +2 and metal-binding energy change synergistically improved the catalytic performance of the bimetallic Zn-Ni/ZSM-5 over the HZSM-5 and Zn/ZSM-5. Flowrate increase from 6 to 35 ml min−1 decreased propane conversion from 80–40% with increased aromatic selectivity from 55–80%. Temperature increase from 500–580 °C favours both propane conversion and aromatic selectivity increase from 50–68 and 55–92% respectively. The metallic interactions from H2-TPR and XPS analysis revealed a big improvement in propane conversion, aromatic selectivity and product distribution.


Introduction
The transformation of natural gas, liquefied petroleum gases and other lower alkanes to aromatic compounds (BTEX) are receiving uncommon attention in research [1][2][3]. Currently, steam cracking, fluid catalytic cracking (FCC) and naphtha reforming are industrial processes for producting aromatic compounds as feedstocks for petrochemicals though they are insufficient [4,5]. Hence, light alkanes aromatization over zeolitic-based catalysts is gaining ground for producting aromatic compounds. Aromatization being an energy-intensive (endothermic) process, requires a high reaction temperature to achieve meaningful alkane conversion. The first stage in propane aromatization is protolytic cracking and protolytic dehydrogenation to alkenes which further undergo oligomerization-cracking reactions leading to higher hydrocarbon C 4 -C 10 alkenes. The higher alkenes transfer hydrogen to other alkenes to form corresponding dienes and alkanes. The dienes undergo cyclization (intramolecular oligomerization) to form cyclo-alkenes which are further transformed into cyclic dienes and which through further hydrogen transfer yield aromatics [6][7][8][9][10].
The unique pore size and acidic properties of HZSM-5 makes it suitable for propane aromatization. Nevertheless, the process always records low aromatic yield and selectivity because of catalyst deactivation and enhanced cracking reaction leading to the formation of lower alkanes, predominantly methane and ethane [11].
This work focuses on co-impregnating nickel metal with nearly same properties with zinc on HZSM-5 to enhance zinc stability, catalytic activity and overall aromatic selectivity. Comparative performance tests on propane aromatization, selectivity towards aromatics, effect of temperature and feed flow rate change on conversion, aromatic selectivity and product distribution were also studied.

Catalyst characterization
Regular FTIR, FTIR-pyridine spectra measurement, XRD, BET and textural properties, SEM, TEM, H 2 -TPR and XPS, were all used to characterize the catalysts using the same techniques in our previous studies [27,28,35,36].

Catalytic activity test
Propane aromatization was carried out using the procedures in our previous studies in stainless-steel fixed bed reactor with propane (C 3 ) conversion and product selectivity calculated using the same equations [27][28][29].
Nitrogen-sorption isotherms of ZSM-5 catalysts are shown in Fig. 6. All catalysts displayed Type-I isotherms without separate hysteresis loops [35,36]. This is a unique property of materials with micropores and lesser portions of mesopores. Table 1 shows textural properties obtained from of ZSM-5 catalysts. HZSM-5 displayed highest the porosity proven by high surface area and pore volume [48]. Specific surface area and total pore width slightly reduced with metal weight increase on HZSM-5 due to coverage of zeolite surface. Zn-Ni/ZSM-5 microporous surface area and volume were slightly larger than that of Zn/ZSM-5 because Zn and Ni metal species probably not been located as a single oxide phase, but co-existed on the surface of catalyst. This caused the internal surface area to increase. Decreased surface area inferred that the Zn and Ni species incorporated by impregnation were located both on the micropores and mesopores or on the outer surfaces of ZSM-5 crystals and thereby decreasing the catalyst surface area as compared the parent HZSM-5. Modification of HZSM-5 with metal resulted in increased microporous volume (V micro ) signifying that the metals mostly deposited on the external surfaces and partially blocked the catalyst pores. Because of high dispersion of Zn into the zeolite channels, the metal modified had micropore volumes higher than the HZSM-5. The presence of Zn inhibited the migration and dispersion of Ni since conditions. This occurrence suggests most impregnated Zn existed on the external surfaces of ZSM-5 thereby suppressing Ni dispersions into the zeolite pore channels and thus leaving the Ni-unoccupied Bronsted acid sites in the zeolite channels [27,[48][49][50][51][52]. TPR is a suitable method for assessing the strength of metal-modified ZSM-5 interactions and how reducible d metal oxides. Figure 7 presents the hydrogen-TPR analysis for the ZSM-5 catalysts. Sharp reduction peaks are shown at 420 and 600 °C are representing partial reduction of zinc present from [Zn-O-Zn] 2+ to ZnOH + on the ZSM-5 surface when compared with the parent HZSM-5. Zn-Ni/ ZSM-5 showed a reduction that could also be attributed to the presence of nickel from Ni 2+ to Ni + at 540 °C as against the reported peak for Ni/ZSM-5 at 350 and 490 °C [27,28,53,54]. The sharp increase in reduction peak on the addition of Ni was as result of both Zn and Ni reduction to metal form on the surface of the catalyst which ascribed to the hydrogen release approving the intrinsic ability of Ni to intensely absorb hydrogen inside the interstitial cavities of the metal lattice [55]. The strong interaction between zinc and nickel most probably led to the shift in the reduction peak of the bimetallic catalyst when compared to the single metal counterpart. Thus, a more stable Zn-Ni/ZSM-5 catalyst has been formed. This could relate to the electronic interaction between Zn and Ni. Thus, the rearrangement of charge moved d-electrons from Zn into the interface region between Zn and Ni species [56]. Analysis of this result could indicate that some Ni species on the micropore interacted with Zn species in the microporous channels of the catalyst [49].
XPS was used to investigate the chemical state of the ZSM-5 catalysts. Zinc was 2 + both in bimetallic and monometallic catalysts by XPS analysis. The binding energy of the 2p 3/2 and 2p 1/2 Zn 2+ are 1025 and 1048 eV respectively as shown in Fig. 8. This increased as nickel was coimpregnated with zinc due to interaction between the metals to 1030 and 1054 eV which is believed to have aided in strengthening and improving zinc stability [27,28,[57][58][59]. The binding energy of nickel on bimetallic catalyst were 862 and 880 eV for 2p 3/2 and 2p 1/2 , as against the reported values of 853 and 872 eV for nickel in Ni/ZSM-5, this suggests the presence of a new interface created on the catalyst surface via Zn-Ni interaction as also shown by hydrogen-TPR [10,60].

Catalysts performance test
Light alkane aromatization over bifunctional ZSM-5 involves a series of heterogeneous reactions. Bronsted acid sites had been reported to be responsible for oligomerization, cyclization, and cracking of olefins while Lewis acid sites principally aid the dehydrogenation of alkanes and cycloalkanes and recombination of H + to form H 2 in the hydride transfer steps [7,10,[27][28][29]61]. Figures 9 and 10 show the catalytic conversion of propane and aromatic selectivity over HZSM-5, Zn/ ZSM-5 and Zn-Ni/HZSM-5 (2 wt% of Zn, 1-3 wt% of Ni) catalysts. It could be observed that similar conversions were recorded for HZSM-5 and Zn/ZSM-5 as reported in our previous works [27][28][29]. The Bronsted acid site of HZSM-5 is responsible for propane activation, oligomerization and aromatization [7,27]. Propane conversion on Zn/ZSM-5 was low but increased from 1 to 7 h' time one stream because of rich zinc on HZSM-5. Zn/ZSM-5 propane conversion was behaving like the parent HZSM-5 from 8-12 h' time on stream which depicted the loss of zinc active site as evidenced in aromatic products distribution and undesired products as reported in our previous work [27][28][29]. Activity test on Zn-Ni/ZSM-5 (1-3 wt% Ni) showed improved and stable propane conversion from average of 45% for parent HZSM5 and Zn/ZSM-5 to 60, 70 and 75% for Zn-Ni/ZSM-5 with nickel 2, 1 and 3% wt. respectively. The presence of Ni aided propane conversion. The equal metal loading of Zn and Ni promoted proper distribution of Zn over the surface of HZSM-5 thus enhancing its activity. Figure 10 shows the aromatic selectivity of the ZSM-5 catalysts. HZSM-5as control 1 in this work had been reported to have low aromatic selectivity because of cracking of propane to C 1 -C 3 and oligomerized alkanes [27][28][29].
Introduction of 2 wt% Zn as control 2 to HZSM-5 by wetimpregnation method increased the aromatic selectivity. This was as a result of recombination of hydrides when zinc metal was used for dehydrogenation in converting alkenes or alkanes intermediates to aromatic hydrocarbons [7, 27-29, 43, 62-65]. Using Ni as second metal (Zn-Ni/ZSM-5) not only improved selectivity but stabilized and sustained the activity and selectivity for the 12 h time on stream investigated. Zn-Ni/ZSM-5 (2 wt% Zn, 2 wt%) displayed highest aromatic selectivity because of maximum synergistic interactions between zinc and nickel as shown in the XPS results and hydrogen-TPR analyses [10,[53][54][55].
The product distributions of propane conversions on ZSM-5 catalysts are clearly shown in Figs. 11, 12, 13, 14 and 15. HZSM-5 had been reported from previous work to have low aromatic selectivity because of fast beta-scission side reactions producing C 1 -C 3 and oligomerized alkanes (hexane, cyclohexane and heptane) as observed in Fig. 11. The introduction of zinc species on HZSM-5 was also reported to have increased aromatic selectivity by recombining hydrogen specie and inhibited beta-scission side reactions that usually lead to undesirable products [28].  The equal weight percentage of Zn with Ni (2 wt%) on HZSM-5 ( Fig. 13) showed similar aromatic compound distribution to Zn/ZSM-5 only that C 9+ increased with increased TOS due to Zn instability.

Effect of temperature and feed flow rate on propane aromatization over Zn-Ni/ZSM-5
Propane aromatization catalyst of 2 wt% Zn with 2 wt% Ni on ZSM-5 gave the best aromatic selectivity, sustained product distribution and lowest C 9+ . This formed the basis for choosing the catalyst for further performance test of effect of flow rate and temperature change on propane conversion, aromatic yield. Selectivity and product distribution. Figure 15 shows the effect of temperature on propane conversion, aromatic yield and selectivity. Change in temperature plays a vital role as aromatization is an endothermic process. An increase in temperature from 500-580 °C favours endothermic reaction with high production of methane and other C 1 -C 2 gases as a result of propane cracking to lower alkanes [65]. There was a steady increase in activity and aromatic selectivity as displayed in Fig. 15, resulting from the improved propane conversion through stable Zn-Ni/ ZSM-5 catalyst preventing cracking and reduction in aromatic formation. Temperature increase also aided cracking of C 9+ to lower aromatics which were clearly seen in toluene, benzene, o and m-xylene increase in Fig. 16. It could be observed that temperature increase favours dehydrogenation, oligomerization and cyclization [42]. Figures 17 and 18 show the effect of increase in feed flow rate on catalytic performance and product distribution respectively. Propane conversion from Fig. 17 reduced as feed flow rate increased from 6 to 35 ml min −1 because propane had a shorter time of contact with the surface of the catalyst for conversion. As the feed flow rate increased, aromatic selectivity increased as the C 9+ build-up time of higher aromatics was limited. Figure 18 showed the product distribution as C 1 -C 3 gases reduced because the contact time was short to favour cracking or any form of secondary reactions. Benzene was relatively stable among the aromatic compounds. The effect of flow rate increase was seen in toluene increase resulting from a decrease in C 9+ because of limited build-up time for higher aromatics formation, thus an overall increase in aromatic selectivity.

Conclusion
Zn and Ni species were co-impregnated on HZSM-5 to synthesize Zn-Ni/ZSM-5 bimetallic catalysts for propane conversion to aromatics in a fixed-bed reactor. Characterization analysis and performance test showed intermetallic interactions had strong influence on activity, aromatic selectivity and product distribution. Nickel was found to aid zinc stability on ZSM-5 with sustained and improved propane conversion and aromatic selectivity during the twelve hours' time on stream (TOS). This increased and sustained aromatic selectivity minimized formation of lighter alkanes. Zn-Ni/ ZSM-5 developed catalysts are highly selective towards toluene and ethylbenzene among other aromatic compounds. Conversion of propane decreased with an increase in flowrate while the selectivity towards aromatics increased. Both   corresponding author and catalyst testing while all co-authors reviewed the manuscripts severally. The optimal operation of the fixed bed reactor, accompanying on-stream GC and other equipment was ensured in getting accurate research results by AAY. BM EBJ hugely contributed to linking catalyst performance to characterizations thereby giving its performance meaning. ABO contributed greatly in the catalyst synthesis, characterisation and discussions.

Declarations
Conflict of interest The authors of this paper declare that we have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper. Products distribution (%) Flow rate (ml/min) Benzene Toluene EB M-xylene P-xylene O-xylene C 1 -C 2 C 6 -C 7 C 9+