Effect of carriers on physico-chemical properties and activity of Pd nano-catalyst in n-hexane isomerization

As seen in figure 1, XRD patterns of catalysts Pd/HZSM-5 and Pd/HY are the same as HZSM-5 and HY, respectively. Particle size of carrier can be calculated by the following equation [9]:

$$\begin{equation} d = \frac{6}{{\rho S_{{\rm{BET}}} }}, \end{equation} \tag{ 1 }$$

where ρ (g cm⁻³) is the density of carrier (ρ of HZSM-5 is 0.45 g cm⁻³, of HY is 0.48 g cm⁻³ and of γ-Al₂O₃ is 0.92 g m⁻³), S_BET (m² g⁻¹) is the specific surface area.

Zoom In Zoom Out Reset image size

In figure 2 one can see rectangular cubic crystallites of zeolites with dimensions 200–260 and 400–600 nm, respectively, for catalysts Pd/HZSM-5 and Pd/HY. As follows from table 1, for catalyst Pd/HZSM-5 the calcination temperature did not influence remarkably physico-chemical characteristics of zeolite phase; values of both the quantities d_Zeol and d varied in ranges of 31.5–32.2 and 41.9–44.7 nm, respectively. Also, the changes in values of specific area were not significant. Nevertheless, it is notable that with increase of calcination temperature from 400 °C to 550 °C the dispersity of Pd improved (increased from 6.29 to 28.19%) and the value of Pd cluster size reduced from 18.4 to 4.1 nm. Pd cluster size (d_Pd) calculated by HPC and measured by TEM are relatively close; on Pd/HZSM-5-500, d_Pd is 5 nm by HPC and 7.36 nm by TEM (figure 3). Catalyst Pd/HY calcined at 550 °C is characterized by a higher value of surface area. On this catalyst the determined values by HPC of Pd cluster size and Pd dispersity are 7.3 nm and 15.95%, respectively. Thus, compared to Pd/HZSM-5, catalyst Pd/HY possesses higher value of surface area but is characterized by a worse dispersity of supported metal. The reason may be included in wider pore size and weaker acidity of faujasite-type zeolite that should lead to a weaker interaction between metal and carrier than in the case of Pd on HZSM-5. In study [10] on catalysts 0.88 wt% Pd supported on ZrO₂ and WO₃-promoted ZrO₂, the values 2.93 nm for quantity d_Pd and 4.8 and 3.8 for quantity γ_Pd, respectively, were observed.

Zoom In Zoom Out Reset image size

Zoom In Zoom Out Reset image size

As seen in EDS images (figure 4), Pd is distributed on catalyst surface fairly evenly. The average values of element distribution on catalyst surface is given in table 1. For catalysts Pd/HZSM-5 the values of Si/Al on surface are fairly high; after calcination at 400 °C the value of this ratio was about 15, but calcination at 550 °C made the ratio Si/Al obtain a value of about 18. For catalyst Pd/HY-550 the value of ratio Si/Al was only about 3.

Zoom In Zoom Out Reset image size

According to [11], for both Pd/Al and Au–Pd/Al samples, the presence of TPR peaks at about 81 °C indicates the reduction of PdO species interacting with the alumina surface. In addition, the TPR profile of the Au–Pd/Al sample shows a peak at 31 °C, indicative of the reduction of bulk PdO. The negative peak of H₂ consumption at 84 °C is attributed to H₂-desorption from the decomposition of a bulk palladium hydride formed through H-diffusion within the Pd crystallites [12]. So, TPR diagrams (figure 5(a)) of all Pd/zeolite catalysts had only one peak with T_max = 65–80 °C, which characterizes the reduction of PdO species interacting with the carrier surface. As follows from table 2, reduction extent of PdO is increasing with calcination temperature of NH₄ZSM-5. Samples Pd/HZSM-5–500 and Pd/HZSM-5–550 have approximately the same and the highest value of reduction extent, and this value is higher than that of Pd/HY catalyst.

Zoom In Zoom Out Reset image size

Results of acidity determination indicate that catalysts Pd/HZSM-5 possess higher total acidity compared to catalyst Pd/HY (∼ 33 mmol NH₃ compared to 25.4 mmol NH₃ per 100 g catalyst). Both catalysts Pd/HZSM-5-500 and Pd/HY-550 are characterized by closed values of strong acidity. However, in values of medium acidity catalyst Pd/HY-550 is characterized only by figure of 8.9 mmol per 100 g catalyst, then catalyst Pd/HZSM-5-550-18.4 mmol per 100 g catalyst.

On all the catalysts a common phenomenon can be observed: when reaction temperature increased the conversion of n-hexane increased but the selectivity in isohexane decreased, so for each catalyst the yield of main product must obtain maximal value at a certain temperature. At pressure 0.1 MPa, for catalysts Pd/HZSM-5 optimal temperatures of the reaction were observed in the range 250–275 °C and for catalyst Pd/HY-550 optimal temperature was 350 °C.

Table 3 shows activity and selectivity data of the studied catalysts at their optimal temperatures at 0.1 MPa. The reaction products comprise unreacted n-hexane, isomers of isohexane, such as 2,3-dimethyl butane (2,3- DMB), 2-methyl pentane (2-MP), 3-methyl pentane (3-MP) and products of cracking.

Calcination temperature of (NH₄)ZSM-5 significantly affected the catalytic activity of Pd/HZSM-5. Among the considered catalysts, Pd/HZSM-5-400 and Pd/HZSM-5-450 are characterized by the lower activity, selectivity and isohexane yield. This can be explained by their lower reduction extent. At pressure 0.1 MPa, both samples Pd/HZSM-5-500 and Pd/HZSM-5-550 expressed approximately equally high efficiency in isohexane production probably due to their high reduction extent. As seen in table 3, the first sample expressed higher activity but lower selectivity compared to the second one. Two catalysts gave the same yield of isohexane (about 50%). Considering their acidity (table 2), one can see that the first sample possesses a greater number of medium acidic centers but fewer strong acidic centers than the second one; the values of total acidic centers on both the catalysts are identical. This fact indicates that acidic centers on carrier surface must play their role in activity and selectivity of catalysts for the given reaction. Besides, the ratio of 2,3-DMB: (2-MP + 3-MP) observed on sample Pd/HZSM-5-500 was the highest in comparison with that on other Pd/HZSM-5 catalysts. This is one of the reasons, leading to the highest RON value of the liquid product obtained on this catalyst. The cracking composition was C₃–C₅ hydrocarbons, in which the proportion of C₃ was preferable. It means that the cracked hydrocarbon was broken at the center of the skeleton.

Compared to catalyst Pd/HY, catalysts Pd/HZSM-5 gave higher activity but much lower ratio of two-branched/one-branched isomers [2,3-DMB: (2-MP + 3-MP)]. This should be understandable, because catalyst Pd/HY is characterized by lower acidity, bigger cluster dimension and worse dispersity of Pd and lower reducibility, but much wider pore size than Pd/HZSM-5. As indicated above, pore size of zeolite HY is up to 1.2 nm, and pore size of HZSM-5 is less than 0.6 nm, while diameters calculated by Lennard–Johns for n-C₆H₁₄ is 0.43 nm, for 2-MP is 0.50 nm and for 2,2-DMB is 0.62 nm.

The analysis of results on XRD (figure 1(b), line 1) and SEM images (figure 2(d)) indicates that aluminum oxide exists in amorphous phase like fine loose cotton with particle size in the range 33–40 nm. XRD patterns of catalysts supported on mixed carriers Pd/Al-HZSM-5 and Pd/Al-HY (figure 1(b)) are similar to those of Pd catalysts supported on pure zeolites (figure 1(a)). Characteristic peaks of zeolite HZSM-5 (at 2θ = 7.9°, 9°, 14.8°, 15.6°, 16°; 23.3°, 23.9°, 24.4°, 29.3°, 30.1° degrees etc) and of zeolite HY (at 2θ = 6.5°, 10.5°, 12°, 16°, 19°, 21°, 24°, 27.5°, 32°) also appeared in XRD patterns of catalysts on mixed carriers but with weaker intensities. Also, the ratio Al₂O₃:zeolite does not influence the characteristics of XRD patterns. Besides, the SEM image of catalyst on mixed carrier (figures 2(e) and (f)) is similar to that of catalyst on pure zeolite (figures 2(a)–(c)). In figures 2(e) and (f) one can see rectangular cubic crystallites of zeolites with dimensions 120–300 nm and 300–500 nm respectively for catalysts Pd/Al-HZSM-5 and Pd/Al-HY on loose alumina. Thus, from analysis of the obtained results it should be concluded that the structure of zeolites HZSM-5 and HY in mixed carriers was not subject to change.

It is interesting to note that, according to EDS data (figures 4(c) and (d) and table 4), for catalysts supported on mixed carriers the values of ratio Si/Al was reduced; in several areas atomic composition of aluminum even exceeds that of silicon. It is possible to propose that on catalyst surface the interaction between aluminum oxide and zeolite is able to form different microphases, although, as confirmed by XRD data, the structure of zeolite was not subject to change.

Lower surface area of γ-Al₂O₃ compared to zeolites resulted in smaller S_BET values of the catalysts on mixed carrier (table 4). Like pure zeolite carriers, mixed carriers are characterized by the same crystallites sizes of zeolite. Among catalyst samples in table 4, catalyst Pd on γ-Al₂O₃ possesses the highest value of Pd particle size and the lowest value of metal dispersity. It is noticeable that characteristics of Pd distribution on (HY + γ-Al₂O₃) carriers were better than those on (HZSM-5 + γ-Al₂O₃). Moreover, while on the first type carriers the Pd dispersion improved with zeolite content, on the second type carriers this quantity changed in opposite direction with zeolite content.

TPR diagrams of catalyst carried on Al₂O₃ and mixed carriers had only one peak with T_max = 70–80 °C characterizing the reduction of PdO species interacting with the carrier surface (figure 5(b)). It should be noted that mixing of aluminum oxide to zeolite HY made the reduction extent of catalyst increase from 30% up to ∼ 34–42%, depending on the ratio Al₂O₃:HY. Also, on addition of Al₂O₃ to HZSM-5, this quantity reduced to be lower than that of catalysts Pd/Al-HY. This should be understandable, because catalyst Pd/Al-HZSM-5 is characterized by bigger cluster dimension and lower dispersity of Pd (table 4).

From results in table 5 one can see that catalyst Pd/Al is characterized by a very low acidity, much lower compared to catalysts Pd/HY and Pd/HZSM-5 (table 2). Generally, the acidity of catalyst on a mixed carrier is between the acidity of catalyst supported on aluminum oxide and the acidity of catalyst supported on zeolite and acidity is increasing with zeolite content. The acidity of Pd/Al-HZSM-5 (1:1) is equal to only half in total and one fourth in medium acidity of catalyst Pd/HZSM-5. Among catalysts supported on mixed carrier Al₂O₃ + HY, sample Pd/Al-HY (2.5:1) is characterized by a lowest acidity; its value is four times higher compared to that of Pd/Al and one third compared to Pd/HY. On this catalyst the value of medium acidity is three times higher, but the strong acidity is only 1.5 times higher than on Pd/Al. The total quantity of strong and medium acidity of catalyst Pd/Al-HY(2.5:1) is equal to one eighth of that on catalyst Pd/HY. Thus, the obtained results indicate that mixed carrier is able not only to produce catalyst with suitable acidity but also to control crystallites size and dispersion of the supported metal.

Activity and selectivity of Pd catalysts supported on mixed carriers are presented in table 6.

The reaction was carried out at 'optimal temperature' for each catalyst and pressure of 0.1 MPa. Data in table 6 indicate that, as a rule, catalyst Pd/Al is characterized by the lowest values of n-hexane conversion and isohexane yield and the highest optimal temperature compared to other catalysts. However, this catalyst expressed also the lowest cracking selectivity due to the lowest acidity. One can put the activity order of catalysts supported on single carriers as follows: Pd/HZSM-5-500 > Pd/HY-550 > Pd/Al. The order of optimal reaction temperatures for these catalysts is in the opposite direction. Among these catalysts, the highest values of conversion and main product yield were observed on Pd/HZSM-5, the highest isohexane selectivity belongs to Pd/Al, and Pd/HY gave the highest proportion of two-branched isomers. These results can be explained by the structure and properties of carriers as shown and interpreted above. One can notice a feature included in the distribution of cracking products on Pd/Al and on other catalysts. If on Pd/Al the content of C₄ and C₅ are predominant in products of cracking, on the rest of the catalysts, proportions of hydrocarbons C₃:C₄:C₅ did not vary significantly.

The common trend in activity variation for catalysts supported on mixed carriers is increasing with zeolite content, reaching a maximum at a certain proportion of zeolite and then going down. It should be considered that optimal compositions for this kind of catalysts are as follows: Al₂O₃:HY = 2.5:1 and Al₂O₃:HZSM-5 = 1:1. This could be explained by the fact that in these catalysts the ratio between amount of metallic centers and acidic centers is reaching optimal value. Naturally, when the proportion of zeolite is growing, cracking selectivity increases and selectivity on isohexane reduces.

Catalyst with optimal composition Pd/Al-HY(2.5:1) gave higher values of n-hexane conversion, isomerization selectivity and isohexane yield compared to Pd/HY, while catalyst Pd/Al-HZSM-5(1:1) expressed lower activity compared to catalyst Pd/HZSM-5. This fact can be explained as follows: alumina in mixed carriers reduced the acidity of the obtained catalysts, but alumina created opposite effects for palladium properties on two types of catalysts. As seen above, on catalysts supported on alumina plus HY zeolite the effect is improvement of Pd dispersion (decrease of particle size, increase of dispersity) and reductibility, while on catalysts supported on alumina plus HZSM-5 zeolite, the effect is the reverse. In other words, addition of alumina to zeolite HY made the physico-chemical properties of catalysts change toward being favorable for isomerization reaction, while addition of alumina to zeolite HZSM-5 made these properties become worse for the given reaction. Since the addition of alumina to zeolites leads to decrease of catalyst acidity, it is understandable that herewith the stability of catalysts supported on mixed carriers should be better than that on catalysts supported on zeolites alone. At the given conditions the lifetime of catalyst Pd/Al-HY (2.5:1) was 23.7 h, while the lifetime of Pd/HY was only 1.25 h. The lifetime of Pd/Al-HZSM-5 (1:1) also was longer than that of Pd/HZSM-5 (1.5 h compared to 1.0 h).

In order to improve the activity, selectivity and stability of catalysts, the reaction pressure was moved up to 0.7 MPa. Table 7 shows the results of experiments carried out on three chosen as representative catalysts at their optimal temperatures and at two values of reaction pressure: 0.1 and 0.7 MPa.

As seen in table 7, at 0.7 MPa, all three catalysts gave higher values of conversion, selectivity and isohexane yield than those obtained at atmospheric pressure. On catalysts Pd/HY and Pd/HZSM-5-500 the optimal reaction temperature even decreased 50 and 25 °C, correspondingly. Also at pressure of 0.7 MPa one can observe remarkable reductions in cracking selectivity of all the catalysts and herewith significant improvements of their lifetimes. RON values of liquid products, obtained at 0.7 MPa on all the catalysts were higher compared to the case when the reaction proceeded at atmospheric pressure. Thus, assuming all experimental results one can conclude that among the studied catalysts, sample 0.8 wt% Pd/HZSM-5-500 has been shown to have the best activity, selectivity and stability in n-hexane isomerization at 0.7 MPa. The only drawback of this catalyst is low proportion of two-branched isomers in reaction products.