Catalytic properties of monometallic and bimetallic palladium and rhodium nanoparticles obtained in reverse micellar systems

1 Introduction

In recent years, the interest in the study of nanoparticles (NPs) was significantly increased. This is due not only to the unique properties of NPs, but that they have opened up new promising opportunities of nanostructured metal particle synthesis for creating nanocomposite materials with desired properties. Many parameters and properties of nanosized particles such as size, shape, particle time stability, the possibility to impregnate them on different supports, the physicochemical properties, including adsorption and catalytic properties – all of them depend on the method of producing and conditions for the NP synthesis. Therefore, a great attention is paid to the choice of method of NP synthesis.

The size effects in catalysis are attracting increasing attention to researchers all over the world, so the problem of NP obtaining with the given size is very relevant and rather complicated. Further development and intensification of catalytic processes are impossible without the using of nano-effects. Nanocomposite catalytic systems based only not on the monometals, but on binary or ternary systems of different metals, are the important way to improve the catalysts.

Catalysts based on platinum group metals are the most active in catalytic processes associated with activation of molecular hydrogen. The catalytic reaction of homomolecular hydrogen isotope exchange studied in this paper is a structure-sensitive reaction, so it can be used to find out the size dependence of the catalytic properties of metal NPs in numerous reactions proceeding with hydrogen involvement.

This paper is devoted to the synthesis and study of promising catalysts based on metal NPs: palladium and rhodium – for processes occurring with hydrogen involvement.

The main feature of chemical properties of metal NPs is their high reactivity, so they have very short lifetime. The NPs aggregate and react easily with different chemical compounds. Therefore, the great attention is paid to their stability in developing methods of NP synthesis. One of these methods to obtain highly stable metal NPs is synthesis in reverse micelles.

A micelle can be represented as a microreactor in which NPs are formed. A micelle shell creates certain barriers to the growth of these aggregates, allowing to produce particles with small size. The size of produced NPs can be influenced directly by varying the size of the water pool, which is characterized by the ratio ω=[H₂O]/[surfactant] [1–4]. The micelle shell holds back the formed NP aggregation, that is why the particles are stored in solution without losing its high specific surface area and the particular properties, which are characteristic of nanosized state of matter in liquid phase for a long time.

2 Materials and methods

Reverse micelles are ternary system aqueous solution of a salt/surfactant/nonpolar solvent. The aqueous solutions 0.04 m RhCl₃·xH₂O (99.9 mass.%, Aldrich) and 0.015 m PdCl₂ (99.9 mass.%, Aldrich) were taken as metal salts, AOT (98 mass.%, Aldrich) was used as a surfactant, isooctane (99.95 mass.%, Referense-1, COMPONENT-REAKTIV) was used as the dispersion medium. Water with a resistivity equal to 18 MOhm·cm (Millipore) was used in the synthesis. Concentration RhCl₃ was 0.015 m in synthesis of bimetallic particles. AOT concentration was 0.15 m for all syntheses.

Two methods of metal ion reduction: a radiation-chemical (RadChem) [5] and chemical (Chem) [6] were used in the NPs synthesis. In both methods, reduction of the metal ions to atoms occurs in the water pool of reverse micelles with further aggregation and nanostructure formation.

Radiation-chemical reduction method involves generating the transient particles having reducing properties under the influence of ionizing radiation. The principle of this method is as follows. As the reaction constant of molecular oxygen with a solvated electron is k_II=(1.9±0.2) ×10¹⁰m^-1·s^-1, and reaction constants eaq- with metal atoms are of the same order of magnitude [7, 8], the dissolved oxygen must be removed from the reverse micellar system. For this, the previously prepared in glass ampoules reverse micellar solutions are deaerated by vacuum or by bubbling of inert gas (helium, argon) through the solution. Then, the ampoules with the solution are sealed and irradiated by γ rays (⁶⁰Co). The absorbed doses of these samples were 15 kGy and 30 kGy. Short-lived intermediate particles with reducing (eaq-, and ·H) and oxidizing properties (·OH) are formed by γ-radiolysis of water:

Isopropyl alcohol or acetone is added to the solution before the start of the synthesis in order to suppress a particle with oxidizing properties (·OH radicals). Interacting of the alcohol molecule with the ·OH radical leads to the formation of hydroxyisopropyl radical, which possesses high reducing potential as well as eaq- does. This provides increasing yield of reduced metal ions. Produced anomalous valence metal ions are unstable, so their rapid aggregation occurs, leading to the formation of small clusters followed by the formation of larger nanoparticles.

A second method for the NP synthesis is different in that the reducing agent is a chemical compound of polyphenol nature – quercetin (C₁₅H₁₀O₇·2H₂O, Merck), which has a high reduction potential. The concentration of quercetin in the reverse micellar solution is 150 μm. In this case, the reduction reaction of the metal ions takes place under aerobic conditions in contrast to the first method. Furthermore, formation of the metal aggregates does not occur in the absence of oxygen, as a ternary complex of quercetin, oxygen, and an ion of metal nQrδ⁺···mO₂δ^–···pMe^z+, which leads to the reduction of the metal ion, is not formed [9].

The bimetallic palladium and rhodium nanoparticles were produced by chemical reduction. Solutions with the ratio ω=5.0 were prepared for the study. Synthesis of bimetallic NPs was performed in three ways. The first way is to reduce joint rhodium and palladium ions in the reverse micellar solution, which is carried out by simultaneous loading of aqueous solutions of metals in reverse micellar solution in equal amounts (Rh-Pd BMNPs). The second and third methods are synthesis of NPs of core/shell type: Rh/Pd and Pd/Rh. In this case, first, the monometallic nanoparticles were prepared in reverse micellar solution with the ratio ω equal to one half of the final value, then an aqueous solution of second metal salts added thereto. Its aqueous volume was such that the final reverse micellar solution with a predetermined value ratio ω was formed by mixing. Thus, the final solutions in all three methods contained two metals in molar ratio of 1:1.

Synthesis process and the evolution of NPs in reverse micellar solutions were monitored by spectrophotometry in UV-VIS region on instrument Hitachi U-3010. Quartz cells with an optical path length of 1 mm were used.

The synthesized NPs are adsorbed onto a support γ-Al₂O₃ (Trylistnik^™, Redkinsky Catalyst Plant) having a specific surface area equal to 200–220 m²/g. Adsorption of NPs also was monitored spectrophotometrically in the UV-VIS region. Adsorption had been performed for 90 min after which the sample was removed from the solution. This absorption time is caused by that AOT adsorption proceeds more rapidly than the adsorption of the NPs, which is undesirable in the preparation of catalysts. Next, the catalyst was allowed to dry completely in air atmosphere. After that, the catalyst granules were washed three times, first with hexane (ACS, COMPONENT-REAKTIV), then acetone (ACS, COMPONENT-REAKTIV). Preparation of the catalyst was completed by drying in a muffle furnace at 500 K during 2 h.

Sizes of NPs obtained in reverse micellar solutions with different ratio ω were determined by AFM on devices EnviroScope (Veeco) and MultiMode (Bruker) in tapping mode in the parameter Z (height) with an accuracy of ±0.1 nm. Silicon cantilevers NSG-01 (NT-MDT) were used. In the preparation procedure for AFM, the sample of NPs was applied in reverse micellar solution on atomically smooth mica surface, followed by drying and removal of the AOT. All measurements on the instrument EnviroScope were performed under nitrogen atmosphere.

Studies of adsorption properties and catalytic activity of nanocomposites were carried out in glass high vacuum equipment (Figure 1). Operation of the equipment is as follows. Gases (H₂ and D₂) previously passed through a purification system consisting of zeolite column and a Petryanov filter (FP™) (1) disposed in the liquid nitrogen trap (77 K), and a palladium membrane (2) at a temperature of 600 K. After purification, all the gases are stored in special containers of this equipment, from which they are dispensed into the total volume in a predetermined ratio and at the desired pressure.

Figure 1

Scheme of the equipment for the catalytic studies: 1 – Petryanov filter (FP^™), 2 – palladium membrane, 3 – backing vacuum (rotary vane pump) and oil diffusion pumps, 4 – two glass McLeod gauge designed for two pressure ranges, 5 – gallium gate, 6 – reactor with a vacuum jacket, 7 – conductivity cells, 8 – liquid nitrogen trap; dotted – entire reaction volume, which is heated at vacuum pumping up studies.

The reaction volume of the equipment, in which investigations are carried out, comprises a reactor with a vacuum jacket (6), into which the analyzed catalyst is placed, and the conductivity cells (7) included in a Wheatstone bridge and analyzed composition of the reaction mixture during the course of the experiment on the magnitude of the current caused by changing the thermal conductivity of the mixture. The reactor is divided from other parts of the equipment by a gallium gate (5) (eutectic Ga-In-Sn (59.6%-26%-14.4% mass.%) with a vapor pressure of ∼10^-8 Torr at 800 K) and a liquid nitrogen trap (8). Before analyzing, the reactor (6) was cut off from the equipment, a catalyst sample with a mass of 0.2–1.5 g was placed in it, and the reactor was soldered to the glass equipment.

Before the studies of adsorptive and catalytic activity, the catalyst was placed in the reactor of the equipment, and the entire reaction volume (dotted in Figure 1) is heated at vacuum pumping for 6 h at 600 K to remove residual organic impurities.

Adsorption of hydrogen on the catalysts was determined by volumetric method, where the increment of adsorption is measured at each step, and then, it is summarizes with the adsorbed gas volume from the previous stages of measurement. The active surface area of the catalyst was calculated from the adsorption data. The calculation assumed that the adsorption of hydrogen is completely dissociative, and each hydrogen atom occupies a single metal atom:

where n_m is the amount of adsorbed hydrogen in the monolayer, mol; N_a is Avogadro constant, mol^-1; and χ_Me is the area of the site occupied by a hydrogen atom, m². It is an accepted value equal to πr², where r is the radius of the metal atom, m. Specific active surface area S_m was calculated per gram of catalyst (m²/g).

Two hydrogen isotope exchange reactions were used to test the catalytic properties of the metal NPs. The first is the isotope exchange in molecular hydrogen (3):

The second is the ortho-para conversion of protium (4):

The reactions were studied in the temperature range from 77 K to 300 K.

The kinetics of the reactions were investigated under static conditions (without circulation of gas) at the reaction mixture pressure of 0.5 Torr. The effect of diffusion is excluded at such pressure, and the true kinetic relationships can be observed. Analysis of the gas mixture was continuously carried out on the thermal conductivity. The feed gas was an equimolar mixture of H₂ and D₂ (1:1 by volume) to reaction (3), and a mixture of ortho- and para-hydrogen (3:1 by volume) equilibrated at 600 K to the reaction (4). Before protium and deuterium exchange was studied at 77 K and 110 K, the ortho-para conversion of hydrogen previously was carried out, after which the deuterium was added to obtain an equimolar mixture.

The specific catalytic activity K_s was defined as being the first-order rate constant k_o per area of the metal surface S, including the number of hydrogen molecules N_T in a reaction volume at the given temperature:

where C_o, C_τ, C_∞ are the concentration of HD molecules at the initial moment, at a time interval τ, and at equilibrium, respectively. Change in the current caused by a change in the resistance in arms of a Wheatstone bridge was proportional change in the concentration of formed HD molecules.

The values of activation energy E (kJ/mol) and pre-exponential factor B (molecule·cm^-2·c^-1) were determined based on the temperature dependence of К_s by the Arrhenius equation (6):

As far as К_s is an absolute rate of reaction, the factor B includes the surface concentration of the reacting molecules.

The activation energy E was calculated from the slope of the line (α) on the graph Arrhenius dependence (7):

The pre-exponential factor B is defined as the point of intersection of this line with the y-axis.

In addition, the salient point demarcating the low- and high-temperature regions on the graphs of the catalytic activity was determined.

3 Results and discussion

3.1 Synthesis

3.1.1 Palladium nanoparticles

Palladium NPs were obtained by radiation-chemical reduction from six reverse micellar solutions with the ratio: ω₁=1, ω₂=2, ω₃=3, ω₄=4, ω₅=5, and ω₆=8. Absorbed dose was 30 kGy. Another two solutions with the ratio of ω₁=2 and ω₂=5 were prepared to determine the effect of radiation dose on the reduction of palladium ions. In this case, the absorbed dose was reduced twice and was 15 kGy. Absorbed dose rate was the same in both cases.

Figure 2 shows the optical absorption spectra of solutions of palladium with absorbed dose of 30 kGy. It is seen that the optical absorption intensity of Pd NPs increases with increasing ω from 1 to 5 at wavelengths of ∼220 nm, ∼267 nm, and ∼325 nm. It is worth noting that these peaks are composed of several optical absorption peaks belonging to the particles with different sizes and shapes. The intensity of absorption in the long wavelength region increases with increasing ω that indicates an increase in the amount of synthesized NPs with larger sizes. But in the transition from ω=5.0 to ω=8.0, decrease in the optical absorption intensity as well as noticeable reduction of the proportion of large particles are observed, which is why there is a shift of the spectral peaks to ∼251 nm and ∼317 nm.

Figure 2

Absorption spectra of Pd^RadChem NPs (Dose=30 kGy) in reversed micellar solutions at different values of ω_i.

Figure 3 shows a comparison of the intensities of the optical absorption of Pd NPs with different absorbed doses. The formation of NPs is essentially not observed for ω=2 at a dose of 15 kGy., Here we can distinguish two absorption peaks at wavelengths of 281 nm and 317 nm. In both cases, the absorption intensity increases ∼1.5 times, when increasing the dose twofold. The proportion of large particles is reduced at low doses, as evidenced by the shift of the optical absorption peaks to shorter wavelengths, but the concentration of nanoparticles is low.

Figure 3

Absorption spectra of Pd^RadChem NPs (Dose=15 kGy and 30 kGy) in reversed micellar solutions at different values: ω₁=2.0 and ω₂=5.0.

Pd NPs obtained in reverse micelles were adsorbed on γ-Al₂O₃. As an example, Figure 4 shows the absorption spectra of the initial micellar solution ω=5.0 and the spectra of the solution after 70 min location of γ-Al₂O₃ granules in the solution. Here, there is a distinct process of a more rapid absorption of large NPs compared to the small NPs, which is why there is a shift of spectra to shorter wavelengths. Analyzing the ratio of the absorption intensity at λ∼267 nm and ∼325 nm in the adsorption process for all samples, we can see that this ratio is constant, despite the shift of the peaks, which can indicate the belonging of these two peaks to the same particles (having the identical size and shape).

Figure 4

Absorption spectra of Pd^RadChem NPs (Dose=30 kGy, ω=5.0.) in reversed micellar solutions at different moments, t, min: 0 – upper curve, 70 – lower curve.

3.1.2 Rhodium nanoparticles

The two solutions prepared by radiation-chemical way, with the ratio ω₁=1 and ω₂=5, and the four solutions prepared by the chemical way with the ratio ω₁=1.0, ω₂=5.0, ω₃=8.0, and ω₄=15.0, were taken for studying NPs.

The solutions of rhodium NPs (Figure 5) obtained by the radiation-chemical method have the absorption peaks at λ∼260 nm, λ∼300 nm, and λ∼391 nm. The solutions of NPs (Figure 6) produced by the chemical way have several other absorption peaks. The difference may be due to the influence on the optical properties of the NP environment. It can be seen that the Rh^Chem NPs with ω=1.0 have absorption peaks at λ∼233 nm, λ∼292 nm, λ∼421 nm. When the ratio ω increases to ω=8 and ω=15, the last absorption peak shifts to ∼455 nm, and the peak at ∼292 nm is still present, but its intensity increases. The peak in the long-wave region is undecidable. The quercetin absorption intensity (λ_max∼375 nm) with increasing ratio ω regularly decreased. Quercetin spectrum corresponding to the spectrum of the solution, based on which the nanoparticles are synthesized by the addition of the water solution with metal ions in it, is given for comparison of the NP spectra. For the preparation of catalyst systems based on the obtained nanoparticles, its adsorption on γ-Al₂O₃ was performed during 90 min.

Figure 5

Absorption spectra of Rh^RadChem NPs (Dose=15 kGy) in reversed micellar solutions at ω₁=1.0 (lower curve) and ω₂=5.0 (upper curve).

Figure 6

Absorption spectra of Rh^Chem NPs in reversed micellar solutions at different ω: ω₁=1.0, ω₂=8.0, ω₃=15. Quercetin spectrum corresponds to the spectrum of the solution from which the nanoparticles are synthesized by adding metal salt to it.

3.1.3 Bimetallic nanoparticles

The complex time dependence of the optical absorption intensity was observed in the synthesis of BMNPs of Rh_core/Pd_shell type. Let us pay attention to the process of rhodium postreduction in reverse micellar system, as indicated by the increase in absorbance intensity at λ∼425 nm (Figure 7) at the initial period of the synthesis (up to 150 min). Its further decrease is due to covering up the rhodium particle by palladium shell. The intensity peak at ∼294 nm increases during the observed process. It is possible that it characterizes the change of surface structure of the rhodium kernel, but also, in this case, the rhodium-palladium boundary. The entire behavior of the absorption spectrum in the wide wavelength region is difficult to interpret at the moment.

Figure 7

The optical absorption spectra of Rh/Pd^Chem NPs in reversed micellar solutions (ω=5.0) at different moments. Quercetin spectrum is given for comparison of NPs spectra.

The spectrum of Pd_core/Rh_shellNPs type with ω=5.0 (Figure 8) has the lower intensity of the band with λ∼292 nm, due to the rhodium presence as the shell and its small amount as monometallic particles. This also explains the intensity of this band for the bimetallic particles, such as Rh-Pd substitution alloy. More detailed interpretation of the optical absorption spectra of transition metal NPs requires further study.

Figure 8

The optical absorption spectra of bimetallic Pd_core/Rh_shell, Rh_core/Pd_shellNPs^Chem, and Rh-Pd substitution alloy in reversed micellar solutions (ω=5.0). Quercetin as a reducing agent is given for comparison of NPs spectra.

3.2 AFM measurements

The results of atomic force microscopy are presented in Table 1. Examples of AFM images and differential particle size distributions are shown in Figure 9. These data show that the method of the NP synthesis in reverse micelles allows obtaining sufficiently small particles. It is important that the differential particle size distribution is generally monomodal and close to the Gaussian type.

Table 1

Metal particle sizes.

Metal NP	ω	Mean particles size (nm)
PdRadChem (30 kGy)	1.0	1.4
	2.0	2.3
	3.0	3.3
	4.0	5
	5.0	6.5
	8.0	2.1
RhRadChem	1.0	1
	5.0	1–3
RhChem	1.0	2
	5.0	3.1
	8.0	1.4
	15.0	1.3
RhChem (800°C)	5.0	2.9a
Rh/PdChem	5.0	0.7
Pd/RhChem	5.0	1.6
Rh-PdChem	5.0	1.7

^aValue calculated from the adsorption data.

Figure 9

AFM images and corresponding particle size distributions of NPs: Pd NPs (ω=2.0); Rh NPs (ω=8.0); Pd_core/Rh_shellNPs (ω=5.0).

The solutions with ω>8.0 have a particle size smaller than a solution with 5.0. Maybe the NP formation depends on the water pool size in reverse micelles (for instance, there are several crystal seeds), and that it causes the phenomenon. Also worth noting is that the larger micelles are less stable, causing over time the precipitation. In addition, it can be concluded on the basis of previously studied behavior of reverse micellar systems with other metals that the stability of micelles decreases with increasing charge of metal ions used for the synthesis of the corresponding nanoparticle.

3.3 Adsorption measurements of hydrogen

The typical hydrogen adsorption isotherm obtained for the Rh/γ-Al₂O₃ω=1.0 sample is shown in Figure 10. The isotherm has a quite delineated plateau, which is taken as a monolayer of chemisorbed hydrogen. Repeated adsorption of hydrogen (R), carried out after vacuuming of hydrogen from the catalyst surface at 77 K, completely coincided with the primary isotherm of (A), i.e., adsorbed hydrogen is easily and fully removed from the catalyst surface. This means that there is only a loosely bound form of hydrogen on the catalyst [10, 11].

Figure 10

Hydrogen adsorption isotherm at 77 K for Rh/γ-Al₂O₃ nanocomposite on the base Rh NPs (ω=1.0). A – primary isotherm, R – repeated isotherm obtained after pumping of hydrogen (at 77 K) adsorbed in the primary adsorption.

The isotherms obtained for all the studied sample catalysts have a similar appearance. The designed specific active surface areas are given in Table 2.

Table 2

Characteristics of the catalysts.

Sample	Salient point (К)	Sm (cm2/g)	High-temperature region		Low-temperature region
			EaHT (kJ/mol)	BHT, molecule/(cm2·c)	EaLT (kJ/mol)	BLT, molecule/(cm2·c)
RhCh/γ-Al2O3ω=1	150	430±60	8.24	1.53×1017	∼0.3	2.51×1014
RhCh/γ-Al2O3ω=5	150	1180±140	5.11	3.52×1016	∼0.3	6.32×1014
RhCh/γ-Al2O3ω=8	140	1090±110	6.77	2.11×1016	∼0.3	3.98×1013
RhCh/γ-Al2O3 (800°C) ω=8	155	575±60	8.54	2.77×1017	∼0.3	6.31×1014
PdRadCh/γ-Al2O3ω=1	180	880±120	9.21	4.9×1016	0.7	1.86×1014
PdRadCh/γ-Al2O3ω=5	180	1360±220	8.47	1.17×1017	0.9	1.20×1015
PdRadCh/γ-Al2O3ω=8	170	285±40	9.14	1.51×1017	0.5	2.63×1014
Rh-Pd/γ-Al2O3ω=5	170	980±40	10.03	7.24×1016	∼0	7.76×1013
Pdcore/Rhshellω=5	160	800±70	8.5	3.29×1016	<0.3	8.61×1013
Rhcore/Pdshellω=5	160	930±80	10.3	5.32×1017	<0.3	2.29×1014

Before the installation into the reactor, the Rh/Al₂O₃ (800°C) ω=8 nanocomposite was baked in the muffle furnace up to 800°C. The specific active surface becomes equal to 575±60 cm²/g based on the results of hydrogen adsorption. This value is less than two times the initial value (1100±110 cm²/g) of the active surface of the Rh/Al₂O₃ω=8 sample. Most likely, this is due to the fact that there was the agglomeration of the nanoparticles at high temperature, thereby reducing their surface. The estimated value of the mean size of the formed particles is presented in Table 1.

3.4 Catalysis

3.4.1 Homomolecular isotope exchange of hydrogen

The activation energy for low-temperature E^LT and high-temperature E^HT regions, the values of pre-exponential factors B^LT and B^HT for the same regions calculated according to equations (6 and 7) are presented in Table 2. The table also presents the temperatures conventionally differentiating high- and low-temperature regions – the salient point on the Arrhenius plot (the logarithm of the specific catalytic activity of the inverse temperature). These temperature ranges are related to certain mechanisms of the isotope exchange reaction (3), and the salient point indicates the change of one mechanism to another. The reaction proceeds virtually without activation energy on the Eley mechanism at low temperatures, it can occur either according to Rideal mechanism or, more likely, according to Bonhoeffer-Farkas mechanism [12] at high temperatures.

Calculated from the experimental kinetics of hydrogen isotope exchange reaction, values of specific catalytic activity (К_s) are shown in Figures 11A and B as a function of the logarithm of the K_s average value on the reciprocal temperature. The dependences LgK_s=f(1/T) for samples of deposited metal NPs with different sizes are plotted on each graph.

Figure 11

(A) Comparison of the catalytic properties of Rh NPs. (B) Comparison of the catalytic properties of Pd NPs.

The experimental data (Figure 11A) suggest that the activity of rhodium NPs increases with particle size increasing. The Rh NPs of 3.1 nm (ω=5.0) have the most the specific catalytic activity, the NPs of 1.4 nm (ω=8.0) have the least specific catalytic activity. The most distinct difference in the catalytic properties is observed in the low-temperature region of the reaction, where the activation energy is close to zero. The activity of this catalysts differs by 16-fold at 77 K. The dependence of the specific catalytic activity (К_s) of the NPs size (d, nm) is described by the following equation:

(8)lgКs77К=1.4188 lnd+13.168 (8)

The part of the prepared Rh/Al₂O₃ nanocomposite sample (ω=8.0) was calcinated in the air atmosphere at 800°C that led to the great increase in its catalytic activity. Thus, the specific catalytic activity of Rh/Al₂O₃ω=8.0 (800°C) became equal to the value of the activity of Rh/Al₂O₃ω=5.0. Probably, the catalytic activity increase is connected with particle aggregation at a high temperature. This assumption can be proven by the fact that the melting point of the metal decreases with the size of its particles. For Rh NPs, the temperature should be in the region of 600°C that is below the calcination temperature of this sample. As a result, the particle sizes increased to the same size for the Rh/Al₂O₃ω=5.0 sample. It can be confirmed by the calculation of the expected size of particles by measuring the area of the active surface after the high-temperature processing. The calculated particle size is 2.9 nm, which coincides with the size of NPs in Rh/Al₂O₃ω=5.0. The activity of the heated sample is identical to the activity of Rh/Al₂O₃ω=5.0, which confirms our assumption.

The NPs (ω=5.0) having the diameter equal to 6.5 nm, showed the greatest activity among the Pd NPs in the hydrogen isotope exchange reaction The lowest activity was demonstrated by Pd NPs (ω=1.0) with the particle diameter of 1.4 nm. The differences in values of activity range from five to 10 times depending on the temperature. The differences in the values of the catalytic activity of the samples (Figure 11B) are observed in the entire temperature range, and most clearly expressed at low temperatures. The dependence of the catalytic activity (К_s) of Pd NPs at 77 K on the particle size (d) is described by the equation:

(9)lgКs=0.3242 lnd+13.781 (9)

The Rh_core/Pd_shellNPs (ω=5.0) showed the greatest catalytic activity among all studied BMNPs (Figure 12). The two other catalytic systems showed almost the same catalytic activity throughout the temperature range from 77 K to 300 K. Comparing the systems with their monometallic particles, one can speak about the manifestation of synergetic effect for the Rh_core/Pd_shell system. In this case, its activity at low-temperature and at far high-temperature regions is three to five times higher than that of monometallic species (1.4 nm). The spectral characteristics of this system are also very different from the other two. The Pd_core/Rh_shell and Rh-Pd spectra are similar. These two systems exhibit the same catalytic activity as the monometallic particles (Figure 12B). The obtained composition and structure of BMNPs require further study.

Figure 12

(A) Comparison of the catalytic properties of BMNPs. (B) Comparison of the catalytic properties of the investigated NPs at 77 K.

3.4.2 Ortho-para protium conversion

As for the reaction of the ortho-para conversion of protium (4), it follows the same laws as the hydrogen isotope exchange reaction. Because of the lower energy of the activation transition complex of this reaction, it usually proceeds at high rates. This is explained by the kinetic isotope effect. Thus, at low temperatures (from 77 K to 150 K), the reaction of ortho-para protium conversion proceeds by a chemical mechanism, similar to the mechanism of the hydrogen-deuterium exchange named by the Eley mechanism.

The rate of ortho-para conversion for Rh NPs is comparable to the isotopic exchange. The rate of ortho-para conversion is four to five times higher than the rate of isotope exchange for Pd NPs, for Rh_core/Pd_shell – is 1.1 times higher, for Pd_core/Rh_shell – is three times higher, for Rh-Pd – is three times higher. Here, we see the correlation of values of the ratio for Rh NPs with Rh_core/Pd_shellNPs, as well as of the values of the ratio for Pd NPs with Pd_core/Rh_shell and Rh-Pd NPs. Hence, it is clear that palladium plays a decisive role in the catalytic activity of Pd_core/Rh_shell and Rh-Pd systems. It is possible that its presence in the reverse micellar system affects the behavior of Rh ions, in particular, on reducing Rh ions to atoms. This fact requires further study.

4 Conclusion

This work shows that the synthesis of nanoparticles in reverse micellar system leads to the formation of small particles (0.7–6.5 nm) with a sufficiently narrow size distribution (∼±1.5 nm).

The size of the formed particles increases with increasing value of ω, but in systems with higher ω, the formation of small particles was observed, which may be associated with the occurrence of several stable crystallization centers in one micelle.

Optical absorption spectra of NPs are difficult to interpret accurately, but provide information on the nature of the produced particles, consistent with data obtained by other methods. The ratio of particle with different sizes in a micellar system changes with ω increasing.

The dominant role of palladium in the BMNP formation was found out. The BMNPs of Rh_core/Pd_shell type, showing a synergistic effect in the catalytic properties of reactions involving molecular hydrogen, were obtained. Their activity was more than four times higher than the activity of Rh and Pd NPs.

The dependence of the catalytic activity for Rh and Pd NPs on their sizes was obtained. The catalytic activity of the NPs increases with their diameters in the studied range of sizes. The phenomenon of aggregation (sintering) of Rh NPs at a temperature of 800°C was observed.

The ortho-para protium conversion is faster than the reaction of hydrogen isotope exchange because of its thermodynamic properties. The values found for the BMNP activity helped to clarify their structure and identify the role of palladium in their formation.