Preparation of Aluminum-Oleic Acid Nano-Composite

Materials Transactions, Vol. 52, No. 6 (2011) pp. 1222 to 1227 #2011 The Japan Institute of Metals

Preparation of Aluminum-Oleic Acid Nano-Composite for Application to Electrode for Si Solar Cells Hye Moon Lee* and Jung-Yeul Yun Functional Materials Division, Korea Institute of Materials Science, 531 Changwondaero, Changwon, Gyeongnam, 641-831, Korea Aluminum-oleic acid composite nanoparticles with a mean diameter of 85 nm were successfully prepared by means of a wet chemical process. The Al/oleic acid molar ratio has effects the thickness of the oleic acid layer on Al nanoparticles. Al electrodes can be formed by firing an Al nanoparticle paste film at 600 C, and the firing temperature is about 300 C lower than that required for micrometer-sized Al particles. The electrode formed from commercial Al nanoaparticles is not electrically conducted because of the oxide layer Al nanoparticles; however, the film from the prepared Al nanoparticles for an Al/oleic acid molar ratio of 1 : 0:05 has a minimum specific resistance of 5.6 mcm. [doi:10.2320/matertrans.M2010409] (Received December 2, 2010; Accepted March 10, 2011; Published May 25, 2011) Keywords: aluminum, organic, composite, nanoparticles, electrode, wet chemical process

1.

Introduction

Thick aluminum (Al) films are widely used as the backside contact in single-crystalline or polycrystalline Si solar cells. Al electrodes are commonly fabricated by firing a printed layer formed with a screen printing method; the electrodes are formed by using an Al paste containing Al powder, and organic and inorganic binders.1–3) Micrometer-sized Al particles have been widely used in the fabrication of Al electrodes for Si solar cells. The process of fabricating the Al electrode has three steps: preparing the Al paste with Al particles and organic and inorganic binders, forming Al paste film via screen printing, and firing the film at a temperature equal to or higher than 900 C. Because Al has a relatively low melting point (660 C), a temperature equal to or lower than 660 C is theoretically enough to fire the Al paste film. However, micrometer-sized Al particles contain oxide layers with a thickness of about 10 nm, so the firing temperature for the Al electrode preparation should be equal to or higher than 900 C. Even though the firing process in a high temperature region effectively enhances the electrical properties of the Al electrode, the high firing temperature causes thermal defects, such as a bowing effect and cracks between the Al film and the Si substrate. The thermal defects seriously hinder efforts to improve Si solar cell efficiency. Enhancements to Si solar cell efficiency have been widely researched. However, the focus of most studies was on the improvement of Si solar cell efficiency not by decreasing the thermal defects but by modifying the electrode structure or enhancing the electrode density. Nanoparticles with a size range of a few 100 nm manifest unprecedented chemical and physical behavior.4) Some of the unique properties of nanoparticles include enhanced magnetic moment and a decreased melting temperature.5–7) The use of nanometer-sized Al particles in the fabrication of electrodes is expected to minimize the thermal defects on account of the lower firing temperature. However, Al *Corresponding

author, E-mail: [email protected]

nanoparticles are more reactive than micrometer-sized particles toward oxidation; hence the oxide layer forms rapidly as soon as the Al nanoparticles are exposed to air. Even though the oxide layer is around 2 nm to 5 nm thick, it significantly affects the application of the Al nanoparticles with the oxide layer to an Al conducting film for a Si solar cell. As mentioned earlier, the oxide layer raises the required firing temperature, so that the thermal defects created by the high firing temperature are not removed. In order to decrease the thermal defects, the application of Al nanoparticles without oxide layer to fabrication of Al conducting films is necessary. Organic passivation on the surface of Al nanoparticles is very effective in the prevention of particle oxidation. Al nanoparticles with organic layers have been prepared by various processes. Fernando et al.8) applied a thermal decomposition of alane (AlH3 ) to the preparation of Al nanoparticles with oleic acid layers. Chung et al.9) and Jouet et al.10) synthesized Al nanoparticles with alkyl-substituted epoxide and perfluoroalkyl carboxylic acid layers, respectively. However, all of these Al particles with organic layers were for solid fuel, such as rocket propellant; they have not been yet applied to electrodes for Si solar cells. Thus, in this study we tried to apply the Al nanoparticles with organic layers to the fabrication of an Al electrode for a Si solar cell. The objective of this work is to prepare Al nanoparticles with an oleic acid layer and to apply them to the fabrication of an Al conducting film at a low firing temperature. A wet chemical process was used for the synthesis of the Al nanoaprticles with oleic acid layers. We then characterized the synthesized particles by means of X-ray diffraction (XRD), transmission electron microscopy (TEM), Fourier transform infrared specra (FT-IR), and thermo gravimetric (TG) for our investigations of the crystal structure, morphology, formation- and amount of organic layer on the Al nanoparticles, respectively. The Al electrodes were formed by firing the screen printed film of Al nanoparticles with oleic acid layers. We also investigated the structure and electrical properties of the Al electrodes.

Preparation of Aluminum-Oleic Acid Nano-Composite for Application to Electrode for Si Solar Cells

(b)

Al (JCPDS 04-0787)

20

30

40

50

60

(311) (222)

(220)

(200)

Intensity [a.u.]

(111)

(a)

1223

70

80

200 nm

90

2 θ [ °] Fig. 1 (a) XRD patterns and (b) TEM images of Al nanoparticles produced by a wet chemical process.

2.

Experimental

2.1 Preparation of aluminum nanoparticles The Al nanoparticles were prepared by means of a wet chemical process. Aluminum chloride (AlCl3 ) was used as a precursor for preparing the Al nanoparticles, and lithium aluminum hydride (LiAlH4 ; LAH) was used as both a precursor and a reduction agent. The solvent for the reaction of both chemicals was 1,3,5-trimethylbenzene (mesitylene). All the chemical reagents were purchased from Aldrich Chemical and used as received. For a typical reaction of the Al nanoparticle preparation, we added 5 mmol of AlCl3 and 15 mmol of LAH to 125 ml of 1,3,5-trimethylbenzene and then heated the solution to 164 C with magnetic stirring for about 2 h. All these processes were performed in a glove box containing an argon atmosphere to prevent the precursor or reducing agent from reacting with air or moisture. To remove the byproduct of LiCl, we extracted the prepared particles by a centrifuge process (16000 rpm for 20 min) and washed them with EtOH anhydrous. 2.2

Organic passivation on the surface of aluminum nanoparticles The oleic acid, purchased from Aldrich Chemical, was used as an organic substance for the surface passivation of the Al nanoparticles. For the organic passivation, we added the oleic acid to the hexane anhydrous with the prepared Al nanoparticles and stirred the mixed suspension for approximately 24 h. The Al/oleic acid molar ratio was varied from 1 : 0:025 to 1 : 0:2 for the purpose of investigating how the quantity of oleic acid added to the prepared Al nanoparticles affects the thickness of the organic passivation layer on the surface of the particles. 2.3

Preparation of Al conducting film with aluminumoleic acid composite nanoparticles To prepare an Al conducting film, we used the following three-step process: first, we prepared the Al paste; we then printed the Al paste film on a Si substrate; and finally, we fired the Al paste film. The Al paste was prepared by mixing 0.1 g of the prepared Al-oleic acid particles with 0.15 g of an organic vehicle binder. The Al paste film was formed by

screen printing the Al paste on a Si substrate. The printed Si substrate was annealed at four different heating conditions; first the printed substrate was dried at 120 C for 30 min to remove the solvent; it was then fired at 350 C for 30 min at a heating rate of 10 C/min to remove the organic vehicle; next it was fired at 600 C for 10 min to sinter the Al particles; and, finally, it was cooled down to room temperature in the furnace. The experiments for firing the screen-printed substrate were performed in an atmosphere of argon gas with 20% H2 . 2.4 Characterizations The crystal structure of the prepared Al nanoparticles was analyzed with an X-ray diffractometer (XRD, Rigaku, D/Max 2200) with Cu Ka radiation. The morphology of the particles was examined with a high-resolution transfer electron microscopy (HR-TEM; JEOL, JEOL-2100F). A fourier transform infrared spectroscopy (FTIR; Varian, 640IR) and a thermo-gravimetric (TG; Perkin Elmer, TGA 7) analyses were performed to investigate whether the oleic acid layer was well organized on the surface of Al nanoparticles. The surface structures of the Al conducting films were observed with a scanning electron microscopy (SEM, JEOL, JSM-6060). A four-point probe method (CMT-SR 1000N, Advanced Instrument Technology) was used to measure the specific resistance of the Al conducting film. 3.

Results and Discussion

The particles synthesized by a wet chemical process are illustrated in an XRD pattern in Fig. 1(a) and in a TEM image in Fig. 1(b). The XRD pattern reveals that the prepared particles are a face-centered-cubic aluminum metal. The microscopic investigation of the prepared particles reveals aggregates of small Al crystallites and nonaggregated polyhedral Al crystallites. These results are in good agreement with the research results of Haber and Buhro11) except for the particle size. Their particles had an average diameter of about 200 nm. In contrast, as shown in Fig. 1(b), the Al particles prepared in our study have an average diameter of about 85 nm. This discrepancy is due to the fact that the aging time of our Al particle preparation (2 h) was much shorter

1224

H. M. Lee and J.-Y. Yun

Mass Increase [%]

1625 1559 1459

2915 2859

Transmittance [a.u.]

50

1:0.2 1:0.1 1:0.05 1:0.025

40 30 20

649°C

10 190°C

3500

3000

464°C

0 -10 -20 0

4000

900°C

2500

2000

1500

1000

50

100

150

200

250

300

350

Time [min.]

-1

Wavenumbers (cm ) Fig. 2 FTIR spectra of oleic acid (top) and Al nanoparticles with an oleic acid layer (bottom).

than the aging time used by Haber and Buhro (16 h). The aging time for particle preparation has a significant effect on particle (crystal) growth in a wet chemical process. This observation confirms the results of Wiley et al.,12) who readily controlled the metal particle size by varying the aging time. The formation of an organic layer on the surface of Al nanoparticles is critical for preventing particle surface oxidation. We tried to investigate the formation and thickness of the oleic acid layer on Al nanoparticles via TEM observation, but images were unclear due to the evaporation of the organic layer in a vacuum condition. Thus, we applied FTIR and TG analyses to confirm the formation and quantity of the oleic acid layer on the Al particles, respectively. Figure 2 shows the FTIR spectra of the oleic acid (top) and the Al nanoparticles with the oleic acid layer. The FTIR analysis clearly confirms the presence of the reacted oleic acid; it shows the same C-H stretch pattern before and after the reaction (at around 2900 cm 1 ); however, it also shows the loss of the C=O peak at about 1700 cm 1 and the formation of the bands for bonding between the metal surface and the carboxylic acids at about 1450 cm 1 and 1540 cm 1 . These results suggest that the wet chemical process used in this study permits the production of Al nanoparticles with an oleic acid layer. Figure 3 shows the TG curves for Al-oleic acid composite nanoparticles with the Al/oleic acid molar ratio. The mass changes for all the samples can be divided into five regions. Initially, all the samples lose about 3.9% of their weight due to the volatilization of surface-adsorbed species, such as organic impurities and water; this weight loss is essentially completed before the samples reach 190 C. In the following region (where there is a temperature range of 190 C to 464 C), the weights abruptly decrease for all the samples. According to Lee et al.,13) organics, such as oleic acid and oleyl amine, on the surface of the metal nanoparticles start to evaporate at a temperature range of about 100 C to 200 C, and the evaporation is terminated at about 500 C. These are very similar to our result that the secondary weight loss

Fig. 3 Thermogravimetric analysis curves of Al nanoparticles with Al/oleic acid molar ratios.

occurred at the temperature range of about 190 C to 464 C, indicating that the weight loss occurred at the secondary region was caused by the evaporation of oleic acid layer on Al nanoparticles. In addition, the weight loss that occurs at the secondary region increases as Al/oleic acid molar ratio decreases; specially, the weight losses are 8.5%, 9.4%, 15.6%, and 17.1% for samples with Al/oleic acid molar ratios of 1 : 0:025, 1 : 0:05, 1 : 0:1, and 1 : 0:2, respectively. These values indicate that the thickness of the oleic acid layer on the surface of Al nanoparticles can be adjusted by controlling the molar ratio of the Al/oleic acid. The weight gain is caused by the aluminum oxidation. Because the TG analysis was performed to investigate whether the oleic acid layer is well organized on the particle surface, detailed explanation of the weight gain is outside the scope of this study. Figures 4 and 5 show the structures of Al conducting films formed from the commercial and prepared Al nanoparticles. The electrode fabricated from the commercial Al particles exhibits a relatively dense microstructure as well as a low porosity as shown in Fig. 4(a), implying that the particles are connected each other and thus electrons are likely to migrate from one particle to another. However the conductive behavior was not detected in this sample and it is assumably related with the formation of oxide layer at Al surface. A bright field TEM image of Fig. 4(b), indicative of the clear difference in contrast at outermost area of Al particle, shows that the thin layer is formed around the commercial Al particle. To confirm whether this thin layer is an oxide film which can deteriorate the electrical property, energy filtered (EF) TEM analysis selecting oxygen K edge was carried out and its result is described in Fig. 4(c), which clearly shows that the thin layer formed around Al particles (Fig. 4(b)) includes the oxygen atoms to some degree. Though EFTEM image cannot demonstrate the quantitative information of the elemental species, Figs. 4(b) and (c) clearly show that the observed thin layer is oxide layer and therefore it must give rise to the non-conducting behavior in this prepared electrode. While the oxide layer formed on micrometeric Al particles is incapable of preventing the migration of electrons, the thin layer on nanosized Al particles can play a


(a)

Oxide layer ( ~8nm)

(b)

1225

(c)

Al

5 μm

50 nm

50 nm

Fig. 4 (a) SEM image of electrode prepared by commercial Al particles, (b) bright field TEM image, and (c) oxygen K-edge filtered TEM image of as-received commercial Al particles.

(a)

(b)

5 μm

5 μm

(c)

(d)

5 μm

5 μm

Fig. 5 SEM images of conducting films formed from Al nanoparticles with an oleic acid layer; the Al/oleic acid molar ratios are (a) 1 : 0:025, (b) 1 : 0:05, (c) 1 : 0:1, (d) 1 : 0:2.

significant role to block the migration of electrons due to the relatively high volumetric occupation in contrast to the oxide film on the micrometric Al partices. Therefore the film formed from commercial Al nanoparticles in Fig. 4(b) seems not to be electrically conductive. Figure 5 shows surface images of electrodes prepared from the synthesized Al particles. All the electrodes have a lower dense structure and a higher pore volume than the electrode formed from the commercial particles. We can surmise that the lower density and higher pore volume are caused by the oleic acid layer on the surface of the Al nanoparticles. As shown in Fig. 6, Al conducting film is formed by a few processes; (a) the formation of Al paste film via a screen printing process, (b) the removal of organic binder and some oleic acid by heating the Al paste film, and (c) the formation of Al conducting film by removing the oleic acid and sintering the Al particles. In

process (a), an Al paste film is densely packed with Al particles that have an oleic acid layer. Evaporation of the organic binder and some of the oleic acid during heat treatment of the Al paste film causes the densely packed Al composite particles to be divided into many clusters. Thus, as shown in Fig. 6((c)-1), pores are generated among the clusters and reside in the Al conducting film that is formed after perfect evaporation of the organic materials and sintering of the Al particles without an oleic acid layer. The SEM images of the Al conducting films clearly show that the molar ratio of the Al/oleic acid is proportional to the pore volume. This behavior is in contrast with a common feature in which the concentration of organic materials mixed with metal particles is proportional to the pore volume of a metal film formed after specific heat treatment. We subjected the film formed from the paste with an Al/oleic acid ratio of

1226

H. M. Lee and J.-Y. Yun pore

Oleic acid Al

Organic binder

Evaporation of Oleic acid + Al particle sintering

Evaporation of oleic acid & organic binder

(c)-1

Al Heating at 200 - 400 °C

Al2O3

Heating at 600 °C

(a)

pore

(b) (c)-2

Fig. 6 Process of Al electrode formation from Al paste consisting of Al nanoparticles with an oleic acid layer and an organic vehicle binder: (a) Al paste film formed by a screen printing process, (b) Al film consisting of Al nanoparticles with an oleic acid layer after the evaporation of the organic vehicle binder and some of the oleic acid layer, and (c) Al conducting films formed by evaporating the organic vehicle binder and the oleic acid layer and sintering of the Al particles ((c)-1) film formed from Al paste with a high molar ratio of Al/oleic acid; ((c)-2) film formed from Al paste with a low molar ratio of Al/oleic acid).

10000

Specifice resistance [mΩ cm]

(311) (222)

(440)

(311)

(400)

(220)

Intensity [a.u.]

(200)

(111)

Al (JCPDS 04-0787) Al2O3 (JCPDS 10-0425)

1000

100

10

1:0.025

20

30

40

50

60

70

80

90

2 θ [ °] Fig. 7 XRD pattern of Al conducting film formed after heat treatment at 600 C and in an atmosphere of pure 20% H2 and 80% Ar.

1 : 0:1 XRD analysis to explain the uncommon results. Figure 7 shows the XRD pattern of the film after heat treatment at 600 C and in an atmosphere of pure 20% H2 and 80% Ar. The XRD pattern clearly shows patterns of aluminum and aluminum oxide (-Al2 O3 ). This result indicates, as shown in Fig. 6((c)-2), that aluminum oxidation occurs during the heat treatment, and that the film’s pore volume decreases due to an increase in the particle volume. More advanced studies are necessary for a clear investigation of aluminum oxidation in non-oxygen surrounding. However, given that the pore volume decreases as the oleic acid concentration in the Al paste increases, we can surmise that the oleic acid significantly affects the oxidation process. Figure 8 shows the electrical resistances of the Al conducting films prepared from Al paste with Al/oleic acid molar ratios of 1 : 0:025, 1 : 0:05, 1 : 0:1, 1 : 0:2, and 1 : 1. Clearly, the electrical resistances decrease as the molar ratio of the Al/oleic acid increases. The conducting film that forms from Al paste with an Al/oleic acid molar ratio of 20 (1 : 0:05) has a specific resistance of 5.6 mcm, whereas the

1:0.05

1:0.1

1:0.2

1:1

Al/oleic acid molar ratio [ - ] Fig. 8 Specific resistances of Al conducting films formed from Al-oleic acid composite nanoparticles for an Al/oleic acid molar ratio in the range of 1 : 0:025 to 1 : 1.

conducting film that forms from the Al paste with an Al/oleic acid molar ratio of 10 (1 : 0:1) has a specific resistance value that is greater than 100 mcm. As mentioned earlier, the increase of oleic acid affects the formation of aluminum oxides. Because aluminum oxides form on the surface of the Al particles, electrons are incapable of passing along the conducting films. 4.

Conclusions

Al nanoparticles with an oleic acid layer were chemically prepared by a wet chemical process and used in the preparation of a conducting film at a low firing temperature for the prevention of thermal defects in the rear electrode of a Si solar cell. We successfully used the wet chemical process to prepare the Al nanoparticles with an oleic acid layer, and we adjusted the thickness of the oleic acid layer by controlling the Al/oleic acid molar ratio. Al electrodes can be formed by firing the Al nanoparticle paste film at 600 C, which is about 300 C lower than the firing temperature of


micrometer-sized Al particles. In contrast to the film formed from commercial Al nanoparticles, the film we manufactured from the prepared Al nanoparticles with an oleic acid layer was electrically conductive. However, the conducting films show some high specific resistances due to either the low film density or the Al2 O3 generated during the electrode firing process. Thus, further studies are needed on the formation of conducting film made of Al-organic composite nanoparticles so that we can improve improve the film’s electrical properties and apply it to the rear electrode of a Si solar cell. Acknowledgement This study was supported by a grant (M2009010025) from the Fundamental R&D Program for Core Technology of Materials funded by the Ministry of Knowledge Economy (MKE), Republic of Korea. REFERENCES 1) K. Sopian, N. Amin, N. Asim and S. H. Zaidi: Eur. J. Sci. Res. 24 (2008) 365–372.

1227

2) C. H. Lin, S. Y. Tsai, S. P. Hsu and M. H. Hsieh: Solar Energy Mater. Solar Cells 92 (2008) 986–991. 3) S. Kim, S. Sridharan, C. Khadilkar and A. Shaikh: IEEE 31st Photovoltaic Specialists Conference and Exhibition, (January 3–7, 2005, Orlando, Florida, USA). 4) W. I. Lenggoro, H. M. Lee and K. Okuyama: J. Coll. Interf. Sci. 303 (2006) 124–130. 5) J. S. Chen, C. Y. Tan, S. Y. Chow, B. Liu and G. M. Chow: J. Appl. Phys. 98 (2005) 064306. 6) Z. R. Dai, S. Sun and Z. L. Wang: Nano Lett. 1 (2001) 443–447. 7) K. W. Edmonds, C. Binnis, S. H. Baker, M. J. Maher, S. C. Thornton, O. Tjernberg and N. B. Brookes: J. Magn. Magn. Mater. 231 (2001) 113–119. 8) K. A. S. Fernando, M. J. Smith, B. A. Harruff, W. K. Lewis, E. A. Guliants and C. E. Bunker: J. Phys. Chem. C 113 (2009) 500–503. 9) S. W. Chung, E. A. Guliants, C. E. Bunker, D. W. Hammerstroem, Y. Deng, M. A. Burgers, P. A. Jelliss and S. W. Buckner: Langmuir 25 (2009) 8883–8887. 10) R. J. Jouet, A. D. Warren, D. M. Rosenberg, V. J. Bellitto, K. Park and M. R. Zachariah: Chem. Mater. 17 (225) 2987–2996. 11) J. A. Haber and W. E. Buhro: J. Am. Chem. Soc. 120 (1998) 10847– 10855. 12) B. Wiley, T. Herricks, Y. G. Sun and Y. N. Xia: Nano Lett. 4 (2004) 1733–1739. 13) H. M. Lee, S. G. Kim, I. Matsui, T. Iwaki, F. Iskandar, I. W. Lenggoro and K. Okuyama: J. Magn. Magn. Mater. 313 (2007) 62–68.