ISSN 00231584, Kinetics and Catalysis, 2013, Vol. 54, No. 5, pp. 586–596. © Pleiades Publishing, Ltd., 2013.
The Properties of Pd/Au Bimetallic Colloidal Catalysts Stabilized by Chitosan and Prepared by Simultaneous and Stepwise Chemical Reduction of the Precursor Ions1 M. Adlima * and M. A. Bakarb a
FKIP Program Studi Pend Kimia Universitas Syiah Kuala Darussalam Banda Aceh 23111, Provinsi Aceh, Indonesia b School of Chemical Science, University Sains Malaysia, 11800, Minden, Penang, Malaysia *email:
[email protected];
[email protected] Received June 28, 2011
Abstract—Bimetallic Pd/Au nanoparticle catalysts were prepared with chitosan as a stabilizer. The prepara tion procedure included mixing or stepwise adding palladium and gold ions in various molar ratios followed by simultaneous or stepwise reduction using either methanol or sodium borohydride (nb) as reducing agents. TEM and UV–Vis characterization showed that the particle size of bimetallic Chi–Pd/Au prepared by simultaneous reduction was smaller than that of the samples prepared by stepwise reduction methods. The particle size varied in the 1 to 24 nm range at all Pd/Au molar ratios of bimetallic compositions. Sodium boro hydride was the most effective reducing agent for the preparation of bimetallic Chi–PdcoreAushell by the step wise reduction. The catalytic activities of Chi–Pd/Au prepared by either simultaneous or stepwise reductions were generally higher than those of the respective monometallic systems whereas the most active catalysts were prepared by the simultaneous reduction. Shielding the palladium metal colloid with gold sol led to the decrease in catalytic activity. The turnover frequencies (TOFs) for Chi–Pd/Au–me in catalytic hydrogena tion of 1octene were as high as 20.855 and 89.336 for monometallic and bimetallic catalysts respectively. TOFs for Chi–Pd/Au–nb were in the region between 2.978 and 87.429. The core–shell and alloy formation of the bimetallic Chi–Pd/Au were inferred from the particle size measurements and evaluation of catalytic activity. DOI: 10.1134/S0023158413050017 1
Chitosan or poly[β(1–4)2amino2deoxyD glucose] is a natural polymer that is biodegradable, non toxic and has numerous applications in industrial and manufacturing processes [1]. It has been used as a pro tecting agent for colloidal catalyst preparations [2, 3].
Activity, the metal synergic catalytic effect, selectiv ity, resistance for deactivation and stability of bimetallic nanoparticles catalysts were described in many reports [4, 5]. The morphology of bimetallic nanoparticles also has a significant importance for designing electronic devices [6, 7]. Most studies on the bimetallic catalysts have been focused on the relationship between catalytic activities and the ratio of the metals in bimetallic cata lysts clusters prepared by simultaneous reduction. Bimetallic coreshell structure of Pd–Au colloids has been prepared with various techniques, i.e. irradiation, calorimetric, sonochemical methods and using syn thetic polymer or dendrimers [8–12]. It is interesting to compare catalytic properties of the colloidal alloy and the coreshell Pd/Au colloid in which Pd metal colloid is partially covered by Au sol in chitosan matrix. This comparison can help to elucidate a specific assembling of nanosized particles. The present work describes the 1
The article is published in the original.
effect of the way of reduction on the particle size and catalytic activities of chitosan–Pd/Au catalysts. Reduction procedures included simultaneous or step wise reduction of goldpalladium ions followed by pro gressive covering palladium sol with gold colloid stabi lized by chitosan. EXPERIMENTAL Materials and Equipment Sodium borohydride (95%, “Reidel de Haen,” Ger many), 1octene (99.99%) and chitosan of medium molecular weight (∼400000, “Fluka,” Switzerland), PdCl2 (99.99%, “Merck,” USA); HAuCl4 ⋅ 3H2O (99.5%, “Sigma,” USA) and methanol (ACS certified grade, “Systerm,” Malaysia) were purchased and used without further purification. Philips CM 12 transmission electron microscope was used to obtain TEM micrographs for the determi nation of particle size and particlesize distribution. The particle diameters were measured using a computer software “analysis Docu 2.11” (“GmbH,” Germany, 1986–1997). The average particle size and particlesize distribution were obtained from at least 300 particles. The results of calculations verified the earlier report
586
THE PROPERTIES OF Pd/Au BIMETALLIC COLLOIDAL CATALYSTS
[13]. A powder Xray diffractometer, Siemens model D5000 (CuKα), was used to determine particle crystal linity. Sample preparations for powder XRD character ization were prepared in a solid by pouring the chito san–stabilized colloidal metal solution in acetone to obtain gel–like precipitates. The precipitate was washed several times with acetone–water mixture (1 : 1). The precipitate was smeared on a piece of silicon wafer (1 × 1 cm) and dried in vacuum prior to characteriza tion. Preparations of Colloidal Catalysts Bimetallic Pd/Au metal colloid stabilized by chito san is designated as Chi–Pd/Au. Methanol and sodium borohydride used as reducing agents are denoted by let ters “me” and “nb”, correspondingly. The bimetallic samples prepared by simultaneous reduction of the monometallic species taken at various molar ratios and employing various reducing agents are signified as Chi– Pd1Au1.5–me and Chi–Au1Pd1.5–nb. The bimetallic samples prepared by the stepwise metal reduction are labeled as Chi–Pd1coreAu1shell–nb or Chi– Au1corePd1.5shell–me. The subscript (numerical num ber) such as 1, 1.5 stands for molar ratios of each monometallic species. The molar fraction of the first metal in the bimetallic composition was held con stant (2.8 × 10–5 mole) and is indicated by the numerical index “1.” The molar fraction of the sec ond metal was related to the fraction of first metal as l. The designation Chi–Pd1Au1.5–nb means, for instance, that the first metal is Pd and the mole con tent of Au is 1.5 times that of Pd. Monometallic Catalysts Samples Chi–Pd–me, Chi–Pd–nb, Chi–Au–me and Chi–Au–nb were pre pared according to earlier reports [3]. The amount of chitosan in each experiment was 60 mg and total vol ume of the solution was 45 mL. The concentrations of monometallic component were 1.4 × 10–5, 2.8 × 10–5 and 4.2 × 10–5 mole. Typically, 10 mL of solution con taining 6 mg/mL of chitosan was diluted with 6 mL of a 1.5% (by volume) solution of aqueous acetic acid and 22.5 mL of methanol. Using a micropipette, a 1.24 mL (or 2.8 × 10–5 mole) of palladium chloride stock solu tion was added into the solution. The palladium was reduced either by reflux or by adding 5 mL of aqueous NaBH4, the amount of which was in a 25fold molar excess to the metal. The pH of the solution was adjusted by adding a few drops of concentrated acetic acid and water to maintain the pH at around pH 4 and the total volume of 45 mL. Bimetallic catalysts prepared by simultaneous reduc tion. Similar to monometallic preparative procedures and according to an earlier report [14], bimetallic Chi– Pd/Au was prepared by holding the molar content of one metal constant, while varying the content of other metals. For instance, Chi–Pd1Au1.5 was prepared by diluting 10 mL of 6 mg/mL of chitosan in 5 mL of a 1.5% (by volume) solution of aqueous acetic acid and KINETICS AND CATALYSIS
Vol. 54
No. 5
2013
587
22.5 mL of methanol. A 1.24 mL (or 2.8 × 10–5 mole) of palladium chloride stock solution was quickly mixed with 0.42 mL (or 4.2 × 10–5 mole) of gold stock solu tion and added to the chitosan solution. The bimetallic solution was reduced either by reflux for 5 h or by add ing 5 mL of NaBH4, the amount of which was in a 25fold molar excess to the amount of bimetallic ions. The pH solution was adjusted by adding a few drops of concentrated acetic acid and water to maintain the pH at around pH 4 and with the total volume of 45 mL. Bimetallic catalysts prepared by stepwise reduction. Stepwise preparations are similar to simultaneous reduction method except that the first metal was reduced prior to addition and reduction of the second metal. Typically, Chi–Pd1coreAu1.5shell was prepared by diluting 10 mL solution of chitosan (6 mg/mL) in 5 mL of a 1.5% (by volume) solution of aqueous acetic acid and 22.5 mL of methanol. First, 1.24 mL (or 2.8 ×10–5 mole) of palladium chloride stock solution was added into the chitosan solution. The palladium was reduced either by reflux for 5 h or by adding sodium borohydride (0.0224 g in 2.5 mL). For reflux method the Chi–Pd colloid was cooled down before addition of gold metal. First, 0.42 mL (or 4.2 × 10–5 mole) of gold stock solution was added to the chitosan solution–palladium colloid and reduced by reflux or addition of sodium borohydride (0.0357 g in 2.5 mL). The solution pH was adjusted at pH 4 by adding a few drops of concentrated acetic acid and water to maintain a total volume of 45 mL. Catalytic Test on Hydrogenation of 1Octene Hydrogenations of 1octene either with monome tallic or bimetallic Pd/Au catalysts were performed at atmospheric pressure and 30°C in a closed 50 mL glass reactor as described in a previous study [2]. Exactly 5.3 mL solution of the colloidal catalyst was diluted with methanol (44.7 mL) and fed into the reactor. Hydrogen gas was fed several times to eliminate air and then the catalyst was activated for 60 min with vigorous stirring. The reaction was started with the injection of 0.5 mL of 1octene (3.1 × 10–3 mole). The hydrogen consumption was monitored with a graduated gas burette. The reaction mixtures were sampled at differ ent intervals and the samples containing catalyst were separated with a membrane filter prior to gas chroma tography analysis (Hitachi G3000) with FID at 20 m long carbowax capillary column (HP20 M) main tained at 50°C. RESULTS AND DISCUSSION Spectrophotometry Studies UV–Vis spectra of monometallic Chi–Pd and Chi– Au were recorded and the observed spectrum of Chi– Au–me has a sharp maximum band at 525 nm as shown in Fig. 1. The broader bands of colloidal gold were observed in the spectra of the Chi–Au–nb samples. The
588
ADLIM, BAKAR
Absorbance, a. u. 1.2 1.0
5
0.8
4
0.6 3 2
0.4 0.2 0 400
1 500
600
700 Wavelength, nm
Fig. 1. UV–Vis spectra of colloidal Pd/Au samples: 1— Chi–Pd–me, 2—colloidal mixture of Chi–Pd–me and Chi–Au–me, 3—Chi–Pd1Au1–nb, 4—Chi–Pd1Au1– me, 5—Chi–Au–me.
spectra of colloidal Chi–Pd–me and Chi–Pd–nb cat alysts, however, exhibit no maximum in wavelength region of 450–700 nm. Figure 1 shows also the spectra of bimetallic samples. The spectrum of Chi–Pd1Au1– me shows a broader band compared to that of Chi–Au– me, while the band of Chi–Pd1Au1–nb is much broader than that of Chi–Pd1Au1–me. The Chi– Pd1Au1–me sample containing particles formed as crystallites exibits a broad peak at 525 nm. At the same time, the Chi–Pd1Au1–nb catalyst composed of spher ical particles as shown in the TEM images, has a very broad band at lower wavelength of 475 nm. The differ ences in UV–Vis spectral profile correspond to the size differences in mono and bimetallic colloids. Colloidal solutions that show a broader band often contain smaller particles and the solutions that exhibit a sharper band usually contain larger particles [15]. The relation between shape of particles and the UV–Vis spectra was recorded in literature. Metal colloids containing non spherical particles often absorb at a higher wavelength than those having spherical particles [16]. The differ ences between spectra registered for bimetallic and monometallic systems have been interpreted as due to the formation of bimetallic species and a possible for mation of alloy for the catalytic system prepared by simultaneous reduction [17]. D’Souza and coworkers used UV–Vis studies to confirm the formation of Pd/Ag alloy [18]. Given these spectral changes and reported studies, it is rational to admit the formation of bimetallic Pd/Au species and a possible formation Pd/Au alloy. UV–Vis characteristics of Pd/Au alloy have been reported earlier [19]. Bimetallic Chi–Pd/Au catalysts prepared by step wise reduction method. The differences in the pattern of UV–Vis spectrum and shifting absorbance peak of the bimetallic Chi–Pd/Au were also observed for the sam ples prepared by the stepwise reduction method. For
example the Chi–Au1corePd1.5shell–me catalyst shows broader bands in the spectrum compared to those of the Chi–Au1corePd1shell–me and Chi–Au1corePd0.5shell– me samples. The difficulties in fabricating a bimetallic system with a less noble metal as the core with a more noble metal as the shell (Chi–PdcoreAushell) using a chemical reduction has been addressed by other authors [12]. However, if the reducing agent for the metal was strong enough such as sodium borohydride, the formation of Chi–PdcoreAushell is feasible as reported for FecoreAushell preparations in reverse micelle reaction using sodium borohydride as reducing agent [20]. Accordingly, shielding palladium metal with gold metal using methanol as the reducing agent might not be a completed process since gold (III) ions had been initially reduced by palladium metal and the palladium metal was reoxidized back to palladium ions prior to reduction with methanol. Therefore, using methanol as a reducing agent was an inventive development to pre pare Chi–PdcoreAushell. Substituting methanol by sodium borohydride as the reducing agent preceded preparation of Chi–PdcoreAushell. Particle Size and ParticleSize Distribution Monometallic Chi–Pd and Chi–Au. The effect of reducing agent on the particle size and the particle–size distribution in Chi–Pd, Chi–Au and the bimetallic Chi–Pd/Au colloidal samples was inferred from the TEM micrographs as summarized in Table 1. In the samples reduced in refluxing methanol, the diameter of particles in the monometallic Chi–Au–me colloid var ied from 13.2 ± 4.1 nm to 9.4 ± 3.8 nm and that of the Chi–Pd–me were in the 2.6 ± 0.5 nm to 2.3 ± 0.4 nm range. The particles were dispersed and Chi–Au–me gave a broader particle–size distribution than Chi– Pd–me. Relative standard deviation (σ/d) for Chi– Au–me and Chi–Pd–me samples was around 0.3 and 0.2 respectively. However, if sodium borohydride was used as a reducing agent, Chi–Pd–nb and Chi–Au–nb have nearly equal particle size and particle–size distri bution and all particles were dispersed. The particles Chi–Pd–nb were from 1.9 ± 0.3 nm to 2.6 ± 0.5 nm in size, and the mean diameter of Chi–Au–nb particles was between 2.1 ± 0.4 nm and 2.6 ± 0.5 nm. The rep resentative TEM images of monometallic Chi–Pd and Chi–Au prepared with both reducing agents were shown in Fig. 2. It appears that the particles were dis persed except for the Chi–Pd–me sample. Some Chi– Au–me particles take the shape of crystallites. Bimetallic Chi–Pd/Au. The particle dimensions of bimetallic Chi–Pd/Au catalysts prepared using metha nol and sodium borohydride as reducing agents and reduced either by simultaneous or stepwise reduction are shown in Table 1. (i) Methanol as reducing agent (Chi–Pd/Au–me samples). The bimetallic Chi–Pd/Au–me samples pre pared by simultaneous reduction have a particle size in KINETICS AND CATALYSIS
Vol. 54
No. 5
2013
THE PROPERTIES OF Pd/Au BIMETALLIC COLLOIDAL CATALYSTS
589
Table 1. Particle size and particlesize distribution of monometallic Chi–Pd, Chi–Au and bimetallic Chi–Pd/Au nano particles prepared with methanol or sodium borohydride as reducing agents Molar ratios*
Stabilizers/reducing Type of metal composition and agents reduction method Chitosan/MeOH, reflux
d ± σ, nm
σ/d
2.6 ± 0.5
0.2
dispersed
0 0 1.5 1.0 0.5
2.3 ± 0.4 2.5 ± 0.5 13.2 ± 4.1 11.9 ± 3.6 9.4 ± 3.8
0.2 0.2 0.3 0.3 0.4
less dispersed dispersed less dispersed '' dispersed
1.0 1 1 1.5 0.5
1.5 1 0.5 1 1
8.5 ± 4.1 5.3 ± 3.7 3.3 ± 1.8 3.2 ± 0.7 5.1 ± 1.8
0.5 0.7 0.5 0.2 0.3
less dispersed dispersed '' '' ''
1 1 1
1.5 1 0.5
8.8 ± 2.9 9.6 ± 2.7 7.5 ± 1.9
0.3 0.3 0.2
dispersed '' less dispersed
1.5 1 0.5
1 1 1
13.8 ± 4.2 10.7 ± 3.2 24.1 ± 5.1
0.3 0.3 0.2
dispersed less dispersed ''
monometallic catalysts
1.5 1 0.5 0 0 0
0 0 0 1.5 1.0 0.5
2.0 ± 0.3 2.6 ± 0.5 1.9 ± 0.3 2.4 ± 0.5 2.1 ± 0.4 2.6 ± 0.5
0.1 0.2 0.2 0.2 0.2 0.2
dispersed '' '' '' '' ''
bimetallic catalysts prepared by simultaneous reduction
1 1 1 1.5 0.5
1.5 1 0.5 1 1
2.4 ± 0.7 1.9 ± 0.3 2.1 ± 0.8 2.4 ± 0.7 2.0 ± 0.6
0.3 0.2 0.4 0.3 0.3
dispersed '' '' '' ''
bimetallic catalysts Pdcore prepared by step wise reduction
1 1 1
1.5 1 0.5
2.2 ± 0.6 1.9 ± 0.3 2.1 ± 1.2
0.3 0.2 0.6
dispersed '' ''
1.5 1 0.5
1 1 1
2.0 ± 0.3 2.1 ± 0.3 2.2 ± 0.4
0.1 0.1 0.2
dispersed '' ''
Pd
Au
1.5
0
1 0.5 0 0 0 bimetallic catalysts prepared by simultaneous reduction
bimetallic catalysts Pdcore prepared by step wise reduction
monometallic catalysts
Aucore
Chitosan/NaBH4
Aucore
Distribution
Molar ratio index meaning “1” = 2.8 × 10–5 mol.
the 3.3 to 8.5 nm range with particles smaller compared to those in monometallic Chi–Au–me samples but slightly larger than particles in Chi–Pd–me catalysts (Table 1). The size of bimetallic particles increases with increasing gold content in the bimetallic composition. KINETICS AND CATALYSIS
Vol. 54
No. 5
2013
A similar trend in the particle size was also observed for the samples prepared by the stepwise–metal reduction in which the fraction of gold in the Chi– Pd/Au clusters significantly affects the diameter of the bimetallic particles. With Pd as a core and Au as
590
ADLIM, BAKAR
Distribution, %
Chi–Pd–me 10 : 1, d = (2.3 ± 0.4) nm
20 0 1.0– 1.6– 2.1– 2.6– 3.1– 3.6– 1.5 2.0 2.5 3.0 3.5 4.0 Diameter of particles, nm
Pd–me 20 nm
Distribution, %
(а)
40
60
Chi–Pd–nb 1 : 10, d = (1.9 ± 0.4) nm
40 20 0 1.0–1.5
(b)
20 nm
Distribution, %
50
Chi–Au–me 10 : 1, d = (11.9 ± 3.6) nm
40 30 20 10 0
Au–me, 1:10 (c)
1.6–2.0 2.1–2.6 2.6–3.0 Diameter of particles, nm
6–8 9–11 12–14 15–17 18–20 21–23 Particle diameter, nm
20 nm
Distribution, %
Chi–Au–nb 10 : 1, d = (2.6 ± 0.5) nm
Au, 1:10 nb (d)
20 nm
40 30 20 10 0 1.5– 2.1– 2.6– 3.1– 3.6– 4.1– 2.0 2.5 3.0 3.5 4.0 5.0 Particle diameter, nm
Fig. 2. TEM and histogram of particlesize distribution of monometallic Chi–Pd prepared by using: a—methanol (me), b—sodium borohydride (nb), and Chi–Au prepared by using: c—methanol and d—sodium borohydride as reducing agents. All scale bars are 20 nm except for the Chi–Au–me TEM which is 50 nm. KINETICS AND CATALYSIS
Vol. 54
No. 5
2013
THE PROPERTIES OF Pd/Au BIMETALLIC COLLOIDAL CATALYSTS
Distribution, %
60
591
Chi–Au1Pd1–me, d = (5.3 ± 3.7) nm
40 20 0 1–3 4–6 7–9 10–1213–1516–18 16–18 Particle diameter, nm
(а)
Au1Pb1
50 nm
Distribution, %
40 Chi–PdcoreAushell–me, d = (9.6 ± 2.7) nm 30 20 10 0 3–5 PdsAu1 50 nm
(b)
Distribution, %
40
6–8
9–11 12–14 15–17 Particle diameter, nm
Chi–AucorePdshell–me, d = (10.7 ± 3.2) nm
30 20 10 0
AusPd1 50 nm
(c)
4–6 7–9 10–1213–1516–1819–21 22–24 Particle diameter, nm
Fig. 3. Representative TEM and particlesize distribution histograms of bimetallic Chi–Pd/Au–me prepared with simultaneous and stepwise reduction; a—Chi–Au1 Pd1–me (simultaneous reduction), b—Chi–Pd1core Au1shell–me, c—Chi–Au1core Pd1shell– me. Arrows indicate particles with geometrically regular shape. The scale bars inside TEM images are 50 nm.
a shell (Chi–Pdcore Aushell–me samples) the particle dimensions were smaller than those of bimetallic com positions with gold as a core (Chi–AucorePdshell–me). Chi–AucorePdshell–me particles were less dispersed as shown in Fig. 3. Some of these particles were shaped as crystallites like triangles and octahedrons as shown in KINETICS AND CATALYSIS
Vol. 54
No. 5
2013
Fig. 3c. This observation is consistent with the size properties of their respective monometallic systems, in which Chi–Pd–me particles were larger and less dis persed, compared to Chi–Au–nb particles. The smaller particle size of the Chi–PdcoreAushell– me samples compared to the Chi–AucorePdshell–me cat
592
ADLIM, BAKAR (a) 38.150
Chi–Pd1coreAu1shell–me
44.476 64.221 77.627 115.565 111.283
81.928
0 20
40
60
80 (b)
100
120
136.941
140 2θ
38.150
Chi–Au1corePd1shell–me
44.223 64.721 77.880 112.117
81.675
0 20
40
60
80
100
120
140 2θ
(c)
38.150
Chi–Au1Pd1–me
44.476 54.468 78.133 111.012 115.434
138.989
0 20
40
60
80
100
120
140 2θ
reduced to gold metal and palladium metal was oxi dized to palladium ions, which subsequently mixed with the remaining gold ions. This ionic mixture was later reduced by methanol. Therefore, in this stepwise reduction method fragments of Chi–PdcoreAushell–me might be formed to contribute to the formation of smaller particles as observed for the particle size trend in Table 1 (entry “MeOH (reflux)”). Given the trend of particle size data, it is suggested that particle structure of bimetallic Chi–Pd/Au prepared by simultaneous reduction was different from those prepared by stepwise preparations, where simultaneous reduction gave smaller and bimetallic Chi–Pd/Au–me sample with particles dispersed. From UV–Vis spectra alteration and with allowance for the size of the bimetallic particles it can be con cluded that the core–shell structure and alloying were possibly formed since those properties have been recog nized as bimetallic characteristic in earlier report [20]. (ii) Sodium borohydrate as reducing agent (Chi– Pd/Au–nb samples). In stepwise reduction with sodium borohydride as a reducing agent, the size of Chi–PdcoreAushell–nb particles was similar to the parti cle size of Chi–AucorePdshell–nb (around 2 nm). The particle size of Chi–Pd/Au remained about 2 nm regardless of preparation technique used. Since the par ticles were very small, the size alteration was hardly rec ognizable. The bimetallic colloidal nanoparticles were dispersed with uniform shape (spherical) and a very narrow size distribution. The particle dimensions of mono and bimetallic Pd/Au catalysts stabilized by chitosan especially those of samples prepared using sodium borohydride as a reducing agent are generally smaller than the particle size of samples made with synthetic polymers as stabi lizers [14].
(d) 38.905
Chi–Pd1coreAu1shell–nb
65.733
78.639 117.103
0 20
40
60
80
100
120
140 2θ
Fig. 4. XRD patterns of bimetallic Chi–Pd/Au prepared by simultaneous and stepwise reduction using sodium borohy dride or methanol (refluxing) as reducing agents. Peaks marked with arrows belong to palladium diffraction.
alysts is probably due to the effect of electrochemical reduction that occurred during formation of Chi– PdcoreAushell–me colloid. In mixture of palladium metal and gold ions, a certain amount of gold ions was
Xray Diffraction Study Chi–Pd/Au Bimetallic Colloids The relative intensity of XRD peaks for gold is higher than that for palladium [7]. In addition, it has been also recorded that difficulties in the analysis of the corresponding XRD patterns are caused by the pres ence of very small particles leading to diffuse reflec tions [18]. XRD analysis of Chi–Pd/Au bimetallic colloidal catalysts cast into film was carried out in the 1° ≤ 2θ ≤ 150° range. All samples regardless of reduc tion technique exhibited very strong gold reflection and very weak peaks of palladium as shown in Fig. 4 and tabulated in Table 2. The XRD patterns of Chi– Pd1core Au1shell–nb showed broader reflections than those of Chi–Pd/Au–me (Fig. 4d). This suggests that crystallinity of all Chi–Pd/Au–me samples was higher than that of Chi–Pd1coreAu1shell–nb. The particles shaped as crystallites were also observed on the TEM images of the corresponding monometallic Chi–Au– me and bimetallic Chi–Pd/Au–me samples (Figs. 2 and 3). KINETICS AND CATALYSIS
Vol. 54
No. 5
2013
THE PROPERTIES OF Pd/Au BIMETALLIC COLLOIDAL CATALYSTS
593
Table 2. Tabulated 2θ value of peaks taken from diffractograms of Chi–Pd/Au. Standard/background Pd 40.0 47.0 68.0 – 82.0 86.4 105.0 109.5 119.5 124.4
Chi–Pd1core Au1shell–me
Chi–Au1core Pd1shell–me
Chi–Au1Pd1–me
Chi–Pd1core Au1shell–nb
Au
Pd
Au
Pd
Au
Pd
Au
Pd
Au
38.4 44.4 64.5 77.7 81.8 98.2 111.0 115.5 137.0 –
40.0 * – – – – – 109.5 119.5 –
38.15 44.5 64.5 78.1 – – 111.0 116.0 137.0 –
40.0 – – – – – – 109.5 119.5 –
38.15 44.5 64.5 78.1 – – 111.0 116.0 137.0 –
40.0 – – – – – – 109.5 119.5 –
38.15 44.5 64.5 78.1 – – 111.0 116.0 137.0 –
40.0 – – – – – – 109.5 119.5 –
38.15 44.5 64.5 78.1 – – 111.0 116.0 137.0 –
* There is no peak in the given 2q angle number.
Catalytic Activity Catalytic activity of Chi–Pd/Au and the respective monometallic samples was evaluated from the catalytic hydrogenation of 1octene at temperature of 29°C and atmospheric pressure. The activity was defined in terms of mole percentages of 1octene conversions. 1Octene was catalytically hydrogenated to octane and isomer ized into 2octene and 3octene as observed in all experiments. Isomerization of 1octene usually accom panies catalytic hydrogenation as reported also by other authors [21, 22]. However cis–trans ratio was not used to characterize the activity. The catalytic activity related to mole of metal is represented in terms of turnover frequency (TOF). TOF was calculated on the basis of molar amount of product (moles of 1octene converted to a product) per mole of catalyst per hour. The molar amount of the catalyst used in the TOF calculations is the number of moles of the monometallic component, which was held constant (2.8 × 10–5 mole) in bimetallic composition. For example, total molar amount of metals in bimetallic composition, such as Chi–Pd1Au1.5, Chi–Pd1Au0.5 and Chi–Pd1coreAu1.5shell catalysts, was varied but molar content of palladium was held constant (2.8 × 10–5 mole). Chi–Pd/Au–me. Table 3 shows data characterizing the hydrogenation of 1octene with mono and bime tallic Chi–Pd/Au–me samples. In the presence of the Chi–Pd–me catalyst, the 1octene conversion increased from 11.19 up to 22.20%, while the TOF increased from 11.216 to 20.855 with increasing pal ladium content. Catalytic activities of Chi–Au–me samples were much lower than those of Chi–Pd–me with the 1octene conversion level as low as 1.88– 5.71 mol %. The bimetallic Chi–Au/Pd prepared with simultaneous reduction gave TOF much higher than TOFs calculated by summation of catalytic activities of KINETICS AND CATALYSIS
Vol. 54
No. 5
2013
individual Chi–Pd–me and Chi–Au–me samples. When palladium dominated the composition of the bimetallic Chi–Pd/Au–me catalytic activity was higher. Generally, the catalytic activities of Chi– Pd/Au–me samples prepared by stepwise reduction were lower than those made by simultaneous reduction but higher than those of the respective monometallic catalysts. An improvement in catalytic properties of Chi–Pd/Au–me prepared by simultaneous reduction can be related to alloying of Pd/Au, because it is known that Pd/Au alloys as catalysts are superior to respective monometallic samples [23]. The possibility of alloying Pd/Au was supported by the particle size measure ments and UV–Vis data outlined above. Since palla dium is more catalytically active than gold, it can be expected that a bimetallic catalyst prepared by cover ing palladium clusters as the core with gold metal as the shell, Chi–PdcoreAushell–me, has a lower catalytic activity than Chi–AucorePdshell–me. However, as shown in Table 2, the activity of the Chi–PdcoreAushell– me sample was generally higher than that of the Chi– AucorePdshell–me catalyst. The enhancement of activity observed on going from Chi–PdcoreAushell–me to Chi– AucorePdshell–me tends to support the effect of elec trochemical reduction of gold by palladium metal. Attempt to shield Pd metal with gold (Chi– PdcoreAushell–me) by stepwise reduction and refluxing methanol still led to a possible formation of Pd/Au alloy which contributed to higher catalytic activities. An interesting property was observed when both metals having equal composition (molar ratio of 1 : 1) either prepared with simultaneous or stepwise reduction as shown in Table 2. The highest TOF was obtained at a ratio of palladium to gold 1 : 1 regardless of reducing techniques. An exception, however, was observed where TOF of Chi–Au1Pd1.5–me (89.336) was higher than Chi–Pd1Au1–me (87.176) and the catalytic selectivity
594
ADLIM, BAKAR
Table 3. Catalytic activity of Chi–Pd, Chi–Au and the bimetallic Chi–Pd/Au nanoparticles prepared with methanol as a reducing agent Molar ratios Metal system
Conversion, mol %
Substrate uncon verted, mol %
Product, mol %
1octene
octane 2octene 3octene
TOFs, c–1
Pd
Au
Monometallic catalysts
1.50 1.25 1.00 0.75 0.50 0 0 0 0 0
0 0 0 0 0 1.50 1.25 1.00 0.75 0.50
22.20 19.90 17.19 14.94 11.94 5.71 4.28 2.68 1.88 0
77.80 80.10 82.81 85.06 88.06 94.29 95.72 97.32 98.12 100.00
9.97 8.92 8.24 7.38 6.60 3.26 2.35 1.51 1.02 0
9.21 7.75 6.70 5.21 3.94 1.81 1.59 0.80 0.74 0
3.02 3.23 2.25 2.35 1.40 0.64 0.34 0.37 0.12 0
20.855 18.694 16.148 14.035 11.216 5.364 4.021 2.518 1.766 0
Bimetallic catalysts prepared by simultaneous reduction
1 1 1 1 1 1.50 1.25 0.75 0.50
1.50 1.25 1.00 0.75 0.50 1 1 1 1
74.31 73.82 92.80 87.27 79.06 95.10 88.63 86.05 84.88
25.69 26.18 7.20 12.73 20.94 4.90 11.37 13.95 15.12
34.69 36.22 58.11 38.14 33.03 52.79 50.21 51.27 57.44
27.21 26.37 27.63 35.02 33.18 31.11 28.17 23.45 20.09
12.41 11.23 7.06 14.11 12.85 11.20 10.25 11.33 7.35
69.806 69.346 87.176 81.981 74.268 89.336 83.258 80.835 79.736
Bimetallic catalysts Pdcore prepared by step wise reduction
1 1 1 1 1
1.50 1.25 1.00 0.75 0.50
41.52 66.30 30.86 36.47 61.37
41.52 66.30 30.86 36.47 61.37
29.65 21.00 36.08 37.21 15.75
20.52 5.30 23.93 17.22 16.99
8.31 7.40 9.13 9.10 5.89
54.936 31.658 64.950 59.680 36.289
1.50 1.25 1.00 0.75 0.50
1 1 1 1 1
58.15 56.51 32.42 83.99 88.03
58.15 56.51 32.42 83.99 88.03
17.32 17.99 33.19 7.13 5.22
18.27 18.29 25.05 5.32 5.26
6.26 7.21 9.34 3.56 1.49
39.314 40.854 63.484 15.040 11.245
Aucore
to octane was 62.62 and 50.01% respectively. Accordingly, in stepwise reduction preparation, Chi–Pd1coreAu1–shell–me and Chi–Au1corePd1shell– me demonstrated the highest catalytic activities among the series. This finding is in agreement with other studies previously reported [16]. Chi–Pd/Au–nb. Catalytic activities of Chi–Pd–nb, Chi–Au–nb and the bimetallic Chi–Pd/Au–nb in the hydrogenation of 1–octene are collected in Table 4. Among the samples prepared by a simultaneous reduc tion using sodium borohydride as a reducing agent, there is a trend of changing catalytic activities similar to that recorded for bimetallic compositions made with methanol as a reducing agent. The bimetallic catalysts show higher catalytic activities than the monometallic
catalysts. The catalytic activities of Chi–Pd/Au increased with increasing palladium content in the bimetallic formulations. Contrary to the trend observed for bimetallic cata lysts prepared with methanol as a reluctant, the Chi– PdcoreAushell–nb catalysts were less active than the Chi– AucorePdshell–nb samples. These data confirmed that electrochemical reduction of palladium ions on gold metal did not occur if sodium borohydride was used as a reducing agent. As predicted, covering palladium clusters with gold metal led to a decrease in the catalytic activities of the coreshell catalyst. When palladium was covered partially with gold metal, the TOF was much higher than TOF of bimetallic samples prepared by fully covered the Pd metal with gold sol. This is consistent KINETICS AND CATALYSIS
Vol. 54
No. 5
2013
THE PROPERTIES OF Pd/Au BIMETALLIC COLLOIDAL CATALYSTS
595
Table 4. Catalytic activity of Chi–Pd, Chi–Au and Chi–Pd/Au colloidal nanoparticles prepared with sodium borohydride as a reducing agent Molar ratio Metal system
Unconverted Conversion, substrate, mol % mol % 1octene
Pd
Au
Monometallic catalysts
1.5 1 0.5 0 0 0
0 0 0 1.5 1 0.5
70.13 62.66 56.64 6.67 5.26 3.17
Bimetallic catalysts prepared by simulta neous reduction
1 1 1 1.5 0.5
1.5 1 0.5 1 1
Bimetallic Pdcore catalysts prepared by stepwise reduction Au core
1 1 1 1.5 1 0.5
TOFs, c–1
octane
2octene
3octene
29.87 37.34 43.36 93.33 94.74 96.83
32.21 29.05 27.95 3.79 2.93 1.15
27.67 22.24 20.17 1.64 1.92 1.69
10.25 11.37 8.52 1.24 0.41 0.33
65.880 58.862 53.207 6.266 4.941 2.978
74.72 93.19 84.79 93.07 71.30
25.28 6.81 15.21 6.93 28.70
46.00 56.63 37.88 60.65 37.22
20.28 29.04 36.54 26.70 26.34
8.44 7.52 10.37 5.72 7.74
70.192 87.542 79.651 87.429 66.979
1.5 1 0.5
55.34 20.49 6.45
55.34 20.49 6.45
19.42 36.19 57.32
19.51 33.11 28.33
5.73 10.21 7.90
41.953 74.691 87.880
1 1 1
7.64 11.48 26.30
7.64 11.48 26.30
52.01 45.26 33.09
31.03 33.21 29.99
9.32 10.05 10.62
86.762 83.155 69.233
with the sequence of catalytic activities for Chi– PdcoreAushell–nb samples Chi–Pd1core Au1.5shell–nb < Chi–Pd1core Au1shel–nb < Chi–Pd1coreAu0.5shel–nb. For the samples prepared by depositing palladium metal on gold clusters, Chi–AucorePdshell–nb, an oppo site general trend is observed Chi–Au1corePd1.5shell–nb > Chi–Au1corePd1shell–nb > Chi–Au1corePd0.5shell–nb. The Chi–AucorePdshell–nb catalysts were more active than the monometallic palladium catalyst. The evidence presented above shows that Chi– AucorePdshell catalysts containing palladium covered with gold gave higher catalytic activity compared to the monometallic palladium catalyst Chi–Pd. This conclu sion applies to all preparations regardless of the type of reducing agent and reducing technique. Also, as a rule, Chi–PdcoreAushell samples show higher catalytic activity than Chi–Pd catalysts. An enhancement of catalytic activity for Chi–AucorePdshell can be accounted for by the electronic effect that gold exerts on the surface of palladium atoms. An improvement to catalytic activity found for the catalyst Chi–PdcoreAushell could be a result of other type of synergic effects between palladium and gold, which still are far from being fairly understood [24, 25]. Since the catalyst precursors were reduced by using sodium borohydride and adjusted to pH 4, the metal boride and boric acid might be formed and these species remained in catalysts. Earlier reports confirm the for KINETICS AND CATALYSIS
Product, mol %
Vol. 54
No. 5
2013
mation of metal boride in nickel boride catalysts [26, 27]. It was also reported previously that boron signifi cantly affects the catalytic activity of ruthenium colloi dal catalyst for the hydrogenation of o–chloronitroben zene [28]. Based on the current study, however, increas ing amount of boron apparently has no considerable impact on the catalytic activity in the hydrogenation of 1octene. As shown in Table 4 (entry, “bimetallic simul taneous reduction”) bimetallic catalysts with Pd/Au molar ratios of 1 : 1.5 and 1.5 : 1 have similar amounts of sodium borohydride but have different catalytic activity. A similar case described in the entry “bimetallic system stepwise reduction” (Table 4) illustrates how at a similar amount of NaBH4 the samples prepared with Pd/Au molar ratios of 1 : 1.5 and 1.5 : 1 show quite dif ferent catalytic activities. The presence of the second metal in bimetallic com position, the nature of the reducing agent and the tech nique of reduction of Chi–Pd/Au affect the particle sizes and the catalytic activities of corresponding bime tallic catalysts. The particle size of bimetallic Chi– Pd/Au prepared with simultaneous reduction was smaller than particle dimensions of those prepared with stepwise reduction methods. Chi–Pd/Au samples pre pared with sodium borohydride as a reducing agent are composed of smaller particles compared to the catalysts made by the reduction with methanol. The catalytic activities of Chi–Pd/Au catalysts prepared with either simultaneous or stepwise reductions were generally
596
ADLIM, BAKAR
higher than those shown by the respective monometal lic systems. Bimetallic catalysts prepared with simulta neous reduction exhibited much higher catalytic activi ties compared with those prepared by the stepwise reduction. Covering the palladium metal colloid with gold metal sol in chitosan matrix decreases the catalytic activity but this trend was hardly discernable among Pdcore–Aushell samples prepared with methanol as the reducing agent. The coreshell and alloy formation of bimetallic Chi–Pd/Au could be inferred from the results of particle size determination and evaluation of the catalytic activities. REFERENCES 1. Kumar, M.N.V.R., React. Funct. Polym., 2000, vol. 47, p. 1. 2. Adlim, M., Bakar, M.A., Liew, K.Y., and Ismail, J., J. Mol. Catal. A: Chem., 2004, vol. 212, p. 141. 3. Adlim, M. and Bakar, M.A., Indonesian J. Chem., 2008, vol. 8, p. 320. 4. Verdier, S., Didillon, B., Morin, S., Jumas, J.C., Four cade, J.O., and Uzio, D., J. Catal., 2003, vol. 218, p. 280. 5. Tonetto, G.M. and Damiani, D.E., J. Mol. Catal. A: Chem., 2003, vol. 202, p. 289. 6. Jun, Y.W., Choi, J.S., and Cheon, J., Angew. Chem. Int. Ed., 2006, vol. 45, p. 3414. 7. Park, J., Joo, J., Soon, G.K., Jang, Y., and Hyeon, T., Angew. Chem. Int. Ed., 2007, vol. 46, p. 4630. 8. Xiong, S. and Li, Z., An, Y., Ma, R., and Shi, L., J. Col loid Interface Sci., 2010, vol. 350, p. 260. 9. Taguchi, N., Iwase, A., Maeda, N., Kojima, T., Tanigu chi, R., Okuda, S., Akita, T., Abe, T., Kambara, T., Ryuto, H., and Hori, F., Radiat. Phys. Chem., 2009, vol. 78, p. 1049. 10. Akita, T., Hiroki, T., Tanaka, S., Kojima, T., Kohya ma, M., Iwase, A., and Hori, F., Catal. Today, 2008, vol. 131, p. 90.
11. Toshima, N., Kanemaru, M., Shiraishi, Y., and Koga, Y., J. Phys. Chem. B, 2005, vol. 109, p. 16326. 12. Knecht, M.R., Weir, M.G., Frenkel, A.I., and Crooks, R.M., Chem. Mater., 2008, vol. 20, p. 1019. 13. Ishizuki, N., Torigoe, K., Esumi, K., and Meguro, K., Colloids Surf., 1991, vol. 55, p. 15. 14. Cardenas, T.G. and Segura, D.R., Mater. Res. Bull., 2000, vol. 35, p. 1369. 15. Riahi, G., Guillemot, D., PolissetThfoin, M., Kho dadadi, A.A., and Fraissard, J., Catal. Today, 2002, vol. 72, p. 115. 16. Baia, L., Baia, M., Kiefer, W., Popp, J., and Simon, S., Chem. Phys., 2006, vol. 327, p. 63. 17. Bond, G.C. and Rawle, A.F., J. Mol. Catal. A: Chem., 1996, vol. 109, p. 261. 18. D’Souza, L., Bera, P., and Sampath, S., J. Colloid Inter face Sci., 2002, vol. 246, p. 92. 19. Mandal, M., Kundu, S., Ghosh, S.K., and Pal, T., J. Photochem. Photobiol., A, 2004, vol. 167, p. 17. 20. Zhou, W.L., Carpenter, E.E., Lin, J., Kumbhar, A., Sims, J., and O’Connor, C.J., Eur. Phys. J. D, 2001, vol. 16, p. 289. 21. Sederman, A.J., Mantle, M.D., Dunckley, C.P., Huang, Z., and Gladden, L.F., Catal. Lett., 2005, vol. 103, p. 1. 22. Smits, H.A., Moulijn, J.A., Glasz, W.Ch., and Stan kiewicz, A., React. Kinet. Catal. Lett., 1997, vol. 60, p. 351. 23. Bonarowska, M., Pielaszek, J., Pintar, W., and Kar pinski, Z., J. Catal., 2000, vol. 195, p. 304. 24. Rodriguez, J.A., Surf. Sci., 1994, vol. 318, p. 253. 25. Storm, J.R.M., Lambert, R.M., Memmel, N., Ons gaard, J., and Taglauer, E., Surf. Sci., 1999, vol. 436, p. 259. 26. Glavee, G.N., Klabunde, K.J., Sorensen, C.M., and Hadjapanayis, G.C., Langmuir, 1992, vol. 8, p. 771. 27. Nishimura, S., Handbook of Heterogeneous Catalytic Hydrogenation for Organic Synthesis, New York: Wiley, 2001. 28. Yan, X., Liu, M., Liu, H., and Liew, K.Y., J. Mol. Catal. A: Chem., 2001, vol. 169, p. 225.
KINETICS AND CATALYSIS
Vol. 54
No. 5
2013
