Synthesis of Palladium Nanoparticles and Palladium/Spherical

Carbon Composite Particles in the Solid–Liquid System

of Palladium Oxide–Alcohol by Microwave Irradiation

Yoshihiro Sekiguchi*, Yamato Hayashi and Hirotsugu Takizawa

Department of Applied Chemistry, Graduate School of Engineering, Tohoku University, Sendai 980-8579, Japan

Palladium nanoparticles were synthesized in the solid–liquid system of palladium oxide–alcohol by microwave irradiation. They werecompared with those produced by a conventional heating method. We also used various alcohol solvents and compared the products obtained.The products contained particles that had diameters of several nanometers. In these measurements, microwave heating produced smallerparticles than conventional heating because it provided homogeneous and direct heating. Additionally, Pd/spherical carbon (SC) compositeparticles could be prepared by the same method. For microwave heating, SC particles can support palladium particles without calcination, whichis due to selective heating by microwaves. [doi:10.2320/matertrans.M2010429]

(Received December 20, 2010; Accepted February 15, 2011; Published May 1, 2011)

Keywords: palladium, alcohol, nanoparticles, microwave, solid-liquid system

1. Introduction

The ratio of surface atoms to inner atoms generallyincreases with decreasing particle diameter.1) Quantum sizeeffects appear when particle diameters are of the order ofnanometers. Noble metal nanoparticles such as platinumand palladium catalyze oxidation of carbon monoxide andhydrocarbons2) as well as hydrogenation–dehydrogenationof organic compounds.3,4) Consequently, they are used inmany applications including automobile catalysts5) andchemical synthesis. They reduce the load on the environmentand enhance productivity. Hence, synthesis of noble metalnanoparticles has been intensively researched. Variousmethods have been developed for synthesizing noble metalnanoparticles, including gas evaporation,6) metal salt reduc-tion,7) and metal complex decomposition. However, most ofthese methods have problems associated with them such asthe use of expensive vacuum chambers, toxic inorganic metalsalts, and expensive organometallic compounds.8) Thus,cheaper methods with lower environmental loads are re-quired.

Over the last few decades, microwave heating has receivedmuch interest for synthesizing nanoparticles. Tu and Liusynthesized Pt, Ir, Rh, Pd, Au, and Ru nanoparticles inaqueous methanol solution or ethylene glycol from noblemetal chlorides by microwave heating.9) They point out thatmicrowave heating has the advantages of being rapid andhomogeneous. Therefore, large energy reductions are ex-pected because synthesis is completed rapidly.10) Further-more, microwave heating achieves fast and homogeneousnucleation, making it possible to achieve narrower particlediameter distributions than conventional heating.11)

Ishikawa et al. heated platinum oxide in aqueous ethanolsolution by microwaves and prepared ca. 20 nm platinumnanoparticles without using capping or dispersing agents.12)

In this method, the metal source is an oxide, which is lesstoxic and is cheaper than other starting materials. Moreover,

it does not emit toxic anions. In addition, alcohols areinexpensive and low toxicity organic materials.

In the present study, we synthesized palladium nano-particles by irradiating a solid–liquid system consisting ofpalladium oxide–alcohol with microwaves. We investigatedthe dependence on the alcohol used in this synthesis.Furthermore, we examined carbon-supported catalysts byadding spherical glassy carbon and preparing compositeparticles.

2. Experimental

2.1 Compounds and reactorsPalladium oxide (II) (PdO, 99.9%, Kojundo Chemical Lab.

Co., Ltd.) was used as the starting material. Ethanol (EtOH,99.5%, Wako Pure Chemical Industries, Ltd.), 1-propanol(1-PrOH, 99.5%, Wako Pure Chemical Industries, Ltd.), 2-propanol (2-PrOH, 99.7%, Wako Pure Chemical Industries,Ltd.), 1-butanol (1-BtOH, 99%, Wako Pure ChemicalIndustries, Ltd.) were used as solvents. Spherical glassycarbon (SC; Engineered Carbons Inc.) was used as thesupport. A 2.45 GHz microwave reactor (� Reactor, ShikokuInstrumentation Co., Ltd.) was used for microwave heating.In this experiment, a fluorescence optic-fiber temperaturesensor (FL-2000, Anritsu Meter Co., Ltd.) was used tomeasure the temperature. A hot stirrer (DP-2S, Iuchi SeieidoCo., Ltd.) was used for conventional heating and K-typethermocouples were used to measure the temperature in thisexperiment.

2.2 Preparation of starting suspension0.1 g PdO powder and zirconia balls were added to 100 ml

alcohol solvent and wet ball milled for 24 h. A blacksuspension was obtained that turned dark brown. To preparePd/SC composite particles, SC was added after ball millingand it was dispersed by ultrasonic agitation.

2.3 Synthesis of palladium nanoparticlesThe starting suspension and a magnetic stirrer were placed*Graduate Student, Tohoku University

Materials Transactions, Vol. 52, No. 5 (2011) pp. 1048 to 1052#2011 The Japan Institute of Metals

in a 300 ml three-necked flask. The flask was placed in amicrowave reactor with a Dimroth condenser (see Fig. 1).Microwave irradiation or conventional heating was per-formed for 10–300 min and the suspension was then refluxedat the boiling point of each solvent.

2.4 Characterization of productsThe obtained suspension was dropped onto Au-coated

glass substrates and observed by scanning electron microsco-py (SEM; LEO1420, LEO Electron Microscopy Ltd.). Thesuspension was also dropped into a polypropylene cell andthe average particle diameter was measured by dynamic lightscattering (DLS).

In addition, the samples were dropped onto carbon-coatedCu microgrids and observed by field-emission transmissionelectron microscopy (FE-TEM; HF-2000, Hitachi High-Technologies Co.). Air-dried samples were characterized bypowder X-ray diffraction (XRD; Cu K�, RINT-2000PC,Rigaku Co.).

3. Results and Discussion

3.1 Dependence on reaction timeFigure 2 shows XRD patterns of samples synthesized by

microwave heating using 1-PrOH as the solvent. The PdOpeak intensities decrease with time and face-centered cubicmetallic Pd was obtained at 180 min. Table 1 shows thecrystallite diameter calculated by the Debye–Scherrer equa-tion13) for the (111) peak in these patterns. Additionally,Fig. 3 shows a FE-TEM image and an electron diffractionpattern of the products obtained after heating for 300 min.The electron diffraction pattern reveals that the sample hasbeen reduced to metal Pd. The FE-TEM image shows that theprimary particles have diameters in the range 5–10 nm. Thisresult is supported by the crystallite diameters and theDebye–Scherrer rings in the electron diffraction pattern.14)

3.2 Dependence on microwave powerFigure 4 shows XRD patterns of samples synthesized by

300 or 700 W microwave heating for 120 min in 1-PrOH andFig. 5 shows temperature profiles for the first 2 min. In theseprofiles, achieving temperatures were over 103�C, which ishigher than original boiling point of 1-PrOH (97�C). In caseof polar solvents such as alcohols, solvents achieve highertemperature than conventional boiling point under micro-wave irradiation, without stirring.15) This is called super-

heating,16) and similar phenomenon was observed in theseexperiments despite of stirring. At boiling point, solventneeded only about 100 W experimentally. So it is hard forheat of vaporization to remove redundant heat of strongmicrowave.

The XRD pattern in Fig. 4 for 700 W heating reveals thatPdO was completely reduced, whereas that obtained for300 W heating for the same reaction temperature indicatesthe presence of some residual starting material. This differ-ence is attributed to different heating rates. The solvent willreach the reaction temperature faster when a higher heatingpower is used, but the time to reach the reaction temperaturediffers by only 1 min for 300 and 700 W heating. As

Fig. 1 Experimental setup for microwave heating.

Fig. 2 XRD patterns of (a) starting material and samples synthesized by

300 W microwave heating in 1-PrOH for (b) 60, (c) 120, (d) 180, (e) 240,

and (f) 300 min.

Table 1 Crystallite diameters of samples synthesized by 300 W microwave

heating in 1-PrOH for various times.

Time, t/min 60 180 300

Crystallite diameter, dC/nm 5.9 8.6 10

Fig. 3 TEM image and electron diffraction of sample synthesized by

microwave heating in 1-PrOH for 300 min.

Synthesis of Palladium Nanoparticles and Palladium/Spherical Carbon Composite Particles 1049

mentioned above, some researchers have reported thatmicrowave heating accelerates reactions.9–11) For example,Komarneni et al. synthesized ceramic powders by a micro-wave-hydrothermal process and found that this methodenhanced the crystallization kinetics of the ceramics by oneor two orders of magnitude.17) In addition, in the present case,microwave irradiation is likely to not only maintain thesolvent at boiling point but also promote reduction. In otherwords, the reaction is promoted by using a higher power eventhough the solvent temperature remains the same. Bogdolet al. reported selectivity oxidation of alcohols intocorresponding carbonyl compounds by solid oxidant(Magtrieve�) under microwave irradiation.18) In the experi-ments, oxidant was heated up to about 360�C by microwave.In the present work, the amount of solvent is far major thansolid state (PdO) and we did not observed extreme thermalgradient. But it is possible that PdO absorbed microwaveand heated up selectively, then thermal gradient promotesreduction on solid-liquid interfaces. Table 2 shows thecrystallite diameters obtained by the above-mentioned

method and the average diameters obtained by DLS. In thistable, the crystallite and average diameters differ greatly.This suggests that secondary particles (with average diam-eters of a few hundreds of nanometers) contain many primaryparticles (with crystallite diameters of several nanometers).

3.3 Dependence on solventFigure 6 shows the reaction times of samples synthesized

in various solvents by 300 W microwave heating. Except forEtOH, all the alcohols reduced the raw materials to metalpalladium. EtOH was unable to completely reduce the rawmaterials even after a reaction time of 8 h.

Table 3 shows the reduction times; they increase in theorder 2-PrOH < 1-BtOH < 1-PrOH < EtOH (see Fig. 6).

2-PrOH is the only secondary alcohol in these solvents; itis oxidized to ketone (acetone) by palladium oxide.

PdOþ CH3CH(OH)CH3! Pdþ CH3COCH3 þ H2O

�Gr ¼ �102 kJ/mol19{21Þ ð1Þ

On the other hand, when primary alcohols are oxidized,they become aldehydes, and finally, carboxylic acids. Forexample, in EtOH,

PdOþ CH3CH2OH! Pdþ CH3CHOþ H2O

�Gr ¼ �81:4 kJ/mol19{21Þ ð2ÞPdOþ CH3CHO! Pdþ CH3COOH

�Gr ¼ �159:27 kJ/mol19{21Þ ð3ÞAs Table 3 shows, the primary alcohols have similar

standard Gibbs energies of formation, so that their �Gr

(eq. (2)) are also similar.20,22) For the oxidation of alcohol

Fig. 4 XRD patterns of (a) starting material and samples synthesized by

microwave heating in 1-PrOH for 120 min at (b) 300 and (c) 700 W.

Fig. 5 Temperature profiles for the first 2 min during microwave heating at

(a) 300 and (b) 700 W.

Table 2 Crystallite and average diameters of samples synthesized by

microwave heating in 1-PrOH for 180 min.

Power, P/W 300 700

Crystallite diameter, dC/nm 8.6 8.4

Average diameter, dA/nm 291 199

Fig. 6 XRD patterns of (a) starting material and samples synthesized by

microwave heating in (b) in EtOH at 78�C for 480 min, (c) in 1-PrOH at

97�C for 180 min, (d) in 2-PrOH at 84�C for 10 min, and (e) in n-BtOH at

117�C for 40 min.

1050 Y. Sekiguchi, Y. Hayashi and H. Takizawa

(the first step of this reaction), 2-PrOH has a lower �Gr

(eq. (1)). It is thus thermodynamically favored, so that 2-PrOH has the highest reaction rate. In the case of primaryalcohols, �Gr for the reaction between PdO and aldehydes(eq. (3)) (the second step of the reaction) are lower than thosebetween PdO and 2-PrOH (eq. (1)). Thus, even though thealdehydes have high reductivities, the production of alde-hydes is slow because of slow reaction between PdO andprimary alcohols (eq. (2)), which is the first step of thereaction. Consequently, �Gr is high caused by the lowaldehyde concentrations, as expressed by:

�Gr ¼ �Gr� þ RT lnQ;

where Q is the reaction ratio. Additionally, many palladiumcatalysts oxidize primary alcohol to aldehydes but notaldehydes to carboxylic acids.23) Therefore, aldehydes isnot likely to reduce PdO in this system. For these reasons, thereaction rates were lower for a primary alcohol than for asecondary alcohol.

Moreover, the boiling points of the primary alcoholsdecrease with decreasing carbon chain length as 1-BtOH(117�C) > 1-PrOH (97�C) > EtOH (78�C). The reactionsin this experiment were performed at boiling point. Thereaction rate generally increases with increasing reactiontemperature, as predicted by the Arrhenius equation (k ¼A expð�Ea=RTÞ). Therefore, reduction is promoted by usingan alcohol with a high boiling point (i.e., with a long alkylchain). In these experiments, we did not attempt to reducePdO at the same temperature (e.g., 70�C) using differentprimary alcohols, but we expect that there will be no notabledifferences in their reaction rates because they have similar�Gr.

Table 4 shows the reaction times, the crystallite diameters,and the average particle diameters of the products. In thistable, the crystallite and average diameters increase in theorder 2-PrOH < 1-PrOH < n-BtOH (i.e., the same order astheir boiling points). The average diameter (of the secondaryparticles) generally increases at higher temperatures sinceparticles in a colloidal dispersion system tend to aggregate athigher temperatures. A similar phenomenon is considered tobe occurred in this experiment.

3.4 Comparison with conventional heatingFigure 7 shows XRD patterns of products synthesized by

conventional heating at the boiling point (97�C). The reactionproceeded in the similar way as for microwave heating.Table 5 shows the crystallite and average diameters of theproducts. These results reveal that conventional heatingproduces primary and secondary particles with largerdiameters than microwave heating.

Microwave heating heats molecules in the solvent homo-geneously11) and directly.12) For example, Tu et al. synthe-sized noble metal colloids by microwave heating and oil-bathheating.9) They found that microwave heating producesparticles with smaller average sizes and narrower sizedistributions than oil-bath heating because its heats homoge-neously and directly. In contrast, in conventional heating, thebottom of the flask is heated by a hot plate whose temperatureexceeds the boiling point of the solvent (about 200�C)causing particles to aggregate at the bottom of the flask.Microwave heating has the advantage that it suppressesaggregation.

3.5 Palladium/spherical glassy carbon composite par-ticles

Figure 8 shows SEM images of composite particlesprepared by microwave and conventional heating in threedifferent solvents. These images suggest that palladiumnanoparticles are supported on SC particles. More palladiumparticles are supported on SC prepared by microwave heatingthan by conventional heating, especially in n-BtOH.

Table 3 Standard Gibbs energies of formation of alcohols and aldehydes.

Number of Standard Gibbs energy of formation, �Gf�/kJ mol�1

carbon atoms Alcohol Aldehyde

2 174.7816Þ 128.1216Þ

3 170.617Þ 127.919Þ

4 162.517Þ 119.019Þ

Table 4 Reaction times, boiling point, reaction temperatures, crystallite

diameters, and average diameters of samples synthesized by 300 W

microwave heating in three solvents.

Solvent 1-propanol 2-propanol 1-butanol

Reaction time, t/min 180 10 40

Boiling point, T/�C 97 84 117

Reaction temperature, T/�C 103 85 120

Crystallite diameter, dC/nm 8.6 8.0 8.8

Average diameter, dA/nm 291 201 1090

Fig. 7 XRD patterns of (a) starting material and samples synthesized by

conventional heating in 1-PrOH (b) for 60 (c) and 180 min.

Table 5 Crystallite and average diameters of samples synthesized by

microwave and conventional heating in 1-PrOH for 180 min.

Method Microwave Conventional

Crystallite diameter, dC/nm 8.6 9.2

Average diameter, dA/nm 291 427

Synthesis of Palladium Nanoparticles and Palladium/Spherical Carbon Composite Particles 1051

It is generally difficult to load metal particles onto carbon,especially without calcination. Few palladium particles aresupported on SC synthesized by conventional heating. Incontrast, many more composite particles synthesized bymicrowave heating can load palladium onto SC. Sincecarbonaceous materials generally absorb microwaves and areself-heated by Joule heat loss, SC could be expected to belocally heated. In this experiment, palladium particles arepossible to adhere to SC without additional heat treatment.For these reasons, microwave heating has many advantagesfor preparing supported catalysts.

4. Conclusion

We prepared palladium nanoparticles by microwaveirradiation of the solid–liquid system, palladium oxide–alcohol. The primary particles of the products are severalnanometers in diameter. We were able to synthesizepalladium nanoparticles inexpensively and with a lowenvironmental load. Increasing the microwave power pro-motes reduction of PdO. Secondary alcohols are more activethan primary alcohols due to their higher reactivities. Inaddition, longer primary alcohols are more active thanshorter primary alcohols due to their higher boiling points. Inthis experiment, microwave heating produced smaller par-

ticles than conventional heating. Moreover, we synthesizedcomposite particles by adding SC. It is possible to preparecomposite particles in a single step by microwave heating.Therefore, microwave heating is expected to be used tosynthesize supporting catalysts in the future. Microwaveheating can prepare both fine Pd particles and Pd/SCcomposite particles easier than conventional methods be-cause it provides direct, rapid, and local heating.

Acknowledgement

This study was supported by Suzuki Foundation.

REFERENCES

1) W. P. Halperin: Rev. Mod. Phys. 58 (1986) 533–606.

2) M. Skoglundh, H. Johansson, L. Lowendahl, K. Jansson, L. Dahl and B.

Hirschauer: Appl. Catal. B 7 (1995) 299–319.

3) G. C. Bond: Catalysis by metals, (Academic Press, London, 1962)

pp. 395–436.

4) P. Tetenyi and K. Schachter: Acta. Chim. Acad. Sci. Hung. 56 (1968)

141–152.

5) P. Granger, C. Dujardin, J.-F. Paul and G. Leclercq: J. Mol. Catal. A:

Chem. 228 (2005) 241–253.

6) R. Uyeda: J. Cryst. Growth 24 (1974) 69–75.

7) J. M. Thomas: Pure Appl. Chem. 60 (1988) 1517–1528.

8) Y. Hayashi, H. Takizawa, M. Inoue, K. Niihara and K. Suganuma:

IEEE Trans. Electron. Packag. Manuf. 28 (2005) 338–343.

9) W. Tu and H. Liu: J. Mater. Chem. 10 (2000) 2207–2211.

10) S. Komarneni, D. Li, B. Newalkar, H. Katsuki and A. S. Bhalla:

Langmuir 18 (2002) 5959–5962.

11) T. Yamamoto, Y. Wada, T. Sakata, H. Mori, M. Goto, S. Hibino and S.

Yanagida: Chem. Lett. 33 (2004) 158–159.

12) D. Ishikawa, Y. Hayashi and H. Takizawa: J. Nanosci. Nanotechnol. 8

(2008) 4482–4487.

13) P. Scherrer: Nachr. Ges. Wiss. Gottingen 26 (1918) 98–100.

14) M. Brust, J. Fink, D. Bethell, D. J. Schiffrin and C. Kiely: J. Chem. Soc.

Chem. Commun. (1995) 1655–1656.

15) R. N. Gedye, F. E. Smith and K. C. Westaway: Can. J. Chem. 66 (1988)

17–20.

16) D. R. Baghurst and D. M. P. Mingos: J. Chem. Soc. Chem. Commun.

(1992) 674–677.

17) S. Komarneni, R. Roy and Q. H. Li: Mater. Res. Bull. 27 (1992) 1393–

1405.

18) D. Bogdal, M. Lukasiewics, J. Pielichowski, A. Miciak and Sz.

Bednarz: Tetrahedron 59 (2003) 649–653.

19) D. D. Wagman, W. H. Evans, V. B. Parker, R. H. Schumm, I. Halow,

S. M. Bailey, K. L. Churney and R. L. Nuttall: J. Phys. Chem. Ref. Data

11 (1982) Suppl. 2.

20) R. C. Wilhoit and B. J. Zwolinski: J. Phys. Chem. Ref. Data 2 (1973)

Suppl. 1.

21) J. Nell and H. St. C. O’Neill: Geochim. Cosmochim. Acta 60 (1996)

2487–2493.

22) E. Buckley and J. D. Cox: Trans. Faraday Soc. 63 (1967) 895–901.

23) M. B. Smith and J. March: March’s Advanced Organic Chemistry:

Reactions, Mechanisms, and Structure, 6th Edition, (John Wiley &

Sons, Inc., New Jersey, 2007) pp. 1703–1869.

Fig. 8 SEM images of samples synthesized with spherical glassy carbon

(a) (b) in 1-PrOH for 180 min, (c) (d) in 2-PrOH for 10 min, (e) (f) in n-

BtOH for 40 min; (a) (c) (e) are synthesized by conventional heating and

(b) (d) (f) are synthesized by microwave heating.

1052 Y. Sekiguchi, Y. Hayashi and H. Takizawa