Extremely low catalyst loading for electroless deposition on a non-conductive surface by a treatment for reduced graphene oxide

Jing Zhan a, Desmond C. L. Tan a, Swaek Prakitritanon b, Ming Lin c and Hirotaka Sato *a
aSchool of Mechanical & Aerospace Engineering, Nanyang Technological University, 50 Nanyang Avenue, Singapore 639798, Singapore. E-mail: hirosato@ntu.edu.sg
bOkuno-Auromex Co., Ltd Bangoo Industrial Estate, 392 Moo 4, Soi Pattana 3, Sukhumvit Saikao Road, Praksa Sub-district, Muang, Samut Prakan Province 10280, Thailand
cInstitute of Materials Research and Engineering, A*STAR, 2 Fusionopolis Way, Innovis, #08-03, Singapore 138634, Singapore

Received 15th May 2017 , Accepted 6th July 2017

First published on 6th July 2017

Enhancement of electroless deposition using reduced-graphene-oxide (rGO) coating for a conductive surface on the non-conductive substrate is proposed and demonstrated. We confirmed that the rGO treatment enhanced the Pd catalytic activity and achieved a sufficient rate in electroless deposition with extremely low Pd loading (3.94 × 10−2 μg cm−2).

Since its invention by Brenner and Riddell in 1947,1 electroless plating has been widely used as a deposition method in various industries for coating. Compared with electrodeposition, electroless deposition can reduce metallic ions to their metallic state in the presence of a chemical reducing agent in solution. No additional power source is required and no external electrodes are present for the deposition; the electrons are provided by the reducing agent through a chemical reaction. In electroless plating, Pd is the most commonly used catalyst.2,3 The Pd catalyst deposited on a substrate is a cost barrier to the widespread use of electroless deposition because the market price of Pd is considerably high.

Dispersed Pd nanoparticles with a 50 nm diameter and 20% coverage of the substrate are necessary for initiating complete and homogenous metallisation onto a non-conductive substrate, such as glass, plastic, and ceramic.4 In a non-conductive substrate after the catalysation treatment, oxidation of the reducing agent and electron discharge occur locally in the nanoparticle catalysts. Thus, the metal ions receive the electrons and deposit locally near the Pd nanoparticle domains, as illustrated in Fig. 1A. Therefore, to uniformly metallise a non-conductive surface, Pd nanoparticles must be densely distributed and loaded in large quantity, in the order of a few μg cm−2, and this makes electroless deposition expensive.5

image file: c7cc03773b-f1.tif
Fig. 1 Electroless deposition models on (A) non-conductive substrate, (B) conductive substrate and (C) reduced graphene (rGO) coated non-conductive substrate. Electrons discharged from reducing agent (R) are confined within the Pd nanoparticle catalysts domains; electroless deposition takes place locally at the non-conductive substrate (A) and form continuous surface due to electron transfer for the conductive substrate (C). Once rGO is coated on the non-conductive substrate (C), the deposition model as in (B) would be adopted and a high deposition rate is expected even with a low Pd loading.

For conductive substrates after catalysation treatment, electrons discharged from the oxidation of reducing species are transferred away or pushed over the conductive substrate, resulting in complete and homogeneous deposition on the surface, as shown in Fig. 1B.6 In fact, as demonstrated in the electroless deposition process on a conductive surface (Fig. S1, ESI), Ni deposition occurred on the entire copper sheet even though only several small circles were catalysed with Pd. Thus, if a non-conductive substrate can be facilely functionalised with a conductive material for electron transfer, electroless deposition can occur efficiently throughout the surface even with low Pd loading or treatment with a sparsely deposited Pd catalyst (Fig. 1C).

Coating or depositing graphene would make a plastic or ceramic surface conductive because graphene is superconductive; this property of graphene is attributed to its unique two-dimensional (2D) structure formed by the layer of sp2 hybridised carbon atoms.7 However, graphene is not suitable for solution processes because of its hydrophobicity; graphene coagulates and precipitates in water. However, graphene oxide (GO) is hydrophilic and dissolves or disperses well in water. Thus, we can coat GO onto a solid substrate through immersion in a GO solution. The surface of such a GO-coated substrate can be made conductive by treating it with a reducing agent solution to reduce the GO to reduced graphene oxide (rGO).8

Herein, we hypothesised that an rGO coating on a non-conductive surface may induce electroless deposition at a regular deposition rate even with very small Pd loading (Fig. 1C). In the electroless deposition process, we introduced an rGO pre-treatment (reduction of GO to rGO) before the Pd catalysation step. We demonstrated that the metal deposition is feasible at a sufficient deposition rate even with an extremely low Pd loading of 3.94 × 10−2 μg cm−2.

A 0.2 mg ml−1 GO solution and PdCl2 solution with an extremely low concentration of 30 μM (3.2 ppm) were used for the Pd catalysation. To reduce GO to rGO or Pd2+ to a Pd metal, a NaH2PO2 solution (0.2 M) was used. See the ESI for the details of the sample preparation. In brief, a non-conductive substrate (SiO2 coated quartz crystal microbalance (QCM) substrate) was firstly dipped into the GO solution for 1 min followed by air-blow of the substrate and dipping into the NaH2PO2 solution. This treatment to coat rGO was repeated five times. Then, the rGO functionalized substrate was Pd-catalysed by a sequential treatment (dipping in the PdCl2 solution, air-blow, then NaH2PO2 solution for 2 cycles). The sample prepared by these steps is denoted by ‘3.2 ppm Pd & rGO’ as in Table 1. We also prepared two control samples. One was prepared by using the same PdCl2 (3.2 ppm) and another was prepared by high concentrated PdCl2 (160 ppm, this concentration is conventionally used level in the Pd catalysation for electroless plating), and rGO functionalization was skipped for these two control samples. These two control samples are denoted by ‘3.2 ppm Pd’, ‘160 ppm Pd’, respectively. Expectedly, Pd loadings were very low for the 3.2 ppm Pd & rGO and the 3.2 ppm Pd while 160 ppm Pd had a high Pd loading (see Table 1). The Pd loadings from the low PdCl2 concentration were less than 2% of the one from the high PdCl2. These catalysed samples were dipped into a Ni electroless deposition bath to evaluate the catalytic activity (mass activity deposition amount per Pd loading) by QCM.9,10

Table 1 Pd loadings and mass activities of the samples prepared under different PdCl2 concentrations
Sample name 160 ppm Pd 3.2 ppm Pd 3.2 ppm Pd & rGO
PdCl2 solution concentration (μM) 1.5 × 103 3.0 × 10 3.0 × 10
Resultant Pd loading (μg cm−2) 2.12 2.81 × 10−2 3.94 × 10−2
Loading ratio to 160 ppm Pd (%) 100 1.33 1.86
Mass activity at t = 10 s 1.41 47.5

We confirmed that GO was reduced by the NaH2PO2 treatment to rGO. The results of an X-ray diffraction (XRD) analysis (Fig. 2A) show that the GO peak at 9.73°, which corresponds to an interlayer spacing of 0.9083 nm, disappears after the reduction treatment. A broad peak appears at 26.22° and is attributed to typical graphitic (001) reflection associated with an interlayer spacing of 0.3 nm, as observed in the XRD profile of regular graphene or rGO.11 The lower shift observed in NaH2PO2-treated GO (26.22°) as compared with pure graphene (26.5°) implies that phosphorus released via the decomposition (oxidation) of NaH2PO2 was incorporated into the rGO structure, resulting in the expansion of the graphene lattice.12

image file: c7cc03773b-f2.tif
Fig. 2 (A) XRD patterns and (B) Raman spectra of GO before and after the NaH2PO2 treatment. On the sample after NaH2PO2 treatment, the GO peak disappears and another peak arises at a similar position of the graphene peak.

The reduction of GO to rGO by NaH2PO2 was also confirmed from the Raman spectra of GO before and after the treatment (Fig. 2B). The ratio of the peak intensities attributed to the D band at ∼1350 cm−1 and the G band at ∼1600 cm−1, ID/IG, exhibits a significant increase from 0.886 (before treatment) to 1.112 (after the treatment). This increase indicates that the defects and degree of disorder in the GO sheet were modified to those of regular graphitic sp2 domains as a result of the reduction of GO by NaH2PO2.13

From the TEM observations for the catalysed samples, we confirmed that a small amount of Pd particles was deposited on the samples of 3.2 ppm Pd (Fig. 3B) and 3.2 ppm Pd & rGO (Fig. 3C) compared to 160 ppm Pd (Fig. 3A), which agrees with the trend in the Pd loading measured by ICP MS (the third row of Table 1). The 160 ppm Pd sample had relatively dense Pd particles and they coagulated to form large particles (50–100 nm, Fig. 3A).14 Coagulation was also seen in the 3.2 ppm Pd sample as well (Fig. 3B). For the Pd deposition, electrons discharged from the oxidation of hypophosphite reduce a Pd ion to a Pd metal. Since both 160 ppm Pd and 3.2 ppm Pd samples do not have an rGO layer, when hypophosphite was oxidized, the discharged electrons did not widely travel and the Pd deposition occurred locally. On the other hand, for the 3.2 ppm Pd & rGO, the majority of the particles were small (∼5 nm) and well dispersed. This should be due to electrons travelling widely over the rGO layers covered on the 3.2 ppm Pd & rGO sample, similarly discussed in Fig. 1C.

image file: c7cc03773b-f3.tif
Fig. 3 TEM images of (A) 160 ppm Pd, (B) 3.2 ppm Pd and (C) 3.2 ppm Pd & rGO. See Table 1 for the denotations of these samples.

Applying rGO-coated surface-enhanced electroless deposition even with an extremely low Pd loading achieves a deposition rate comparable with that achieved when regular Pd loading is used. For comparison, we employed the Pd catalysation process using a conventionally employed PdCl2 concentration level, i.e., 1.5 mM (‘160 ppm Pd’ sample in Table 1), for conventional Ni electroless deposition, which resulted in a regular Pd loading (2.12 μg cm−2) and Ni deposition rate, as shown in Fig. 4.15 Once the PdCl2 concentration was decreased to 30 μM (3.2 ppm Pd sample), the Pd loading decreased to 2.81 × 10−2 μg cm−2 and, as expected, the deposition rate drastically decreased and no Ni deposition was observed. The QCM result of the reduced Pd concentration experiment was similar to that of the surface with no applied catalysts (“blank samples” in Fig. 4). In principle, electroless Ni deposition does not occur without catalysation. Thus, no Ni was deposited in the electroless deposition process when 30 μM of PdCl2 solution was used.

image file: c7cc03773b-f4.tif
Fig. 4 Deposition amount in Ni electroless deposition on a SiO2-coated QCM electrode after different catalysation processes: 160 ppm Pd, 3.2 ppm Pd, 3.2 ppm Pd & rGO and no catalysation (blank).

With rGO pre-treatment (3.2 ppm Pd & rGO), the Pd loading was still extremely small (3.94 × 10−2 μg cm−2) for 30 μM of the PdCl2 solution; however, the deposition rate was notably competitive to that of 160 ppm Pd without rGO (Fig. 4). The low Pd loading and high deposition rate indicate that rGO is effective in enhancing the catalysation process for electroless deposition and that a low amount of Pd nanoparticle catalysts can be used. Quantitatively, the mass activity, the amount of Ni deposited in μg normalised by Pd loading in μg,10 was 47.5 at a deposition time (t) of 10 s; this value is 34 times higher than that obtained when 160 ppm Pd (the mass activity was 1.41 at t = 10 s) was used, as shown in Fig. 5.

image file: c7cc03773b-f5.tif
Fig. 5 Mass activity, Ni deposit normalised by Pd loadings of 160 ppm Pd and 3.2 ppm Pd & rGO.

Similarly, the enhancement of Ni deposition was demonstrated using another non-conductive substrate, polyimide. For 3.2 ppm Pd & rGO (the sample at the right in Movie S1, ESI), Ni deposited over the catalysed area can be clearly observed. In contrast, for 3.2 ppm Pd without rGO (the sample at the left in Movie S1, ESI), the appearance of the polyimide substrate did not change, indicating insufficient Ni deposition without rGO. The rGO coating of a non-conductive surface can clearly induce electroless deposition at regular deposition rates even with extremely small Pd loading.

Overall, this study verified our hypothesis that the introduction of rGO as a conductive coating on non-conductive surfaces enhances electroless deposition and results in high deposition rates even with an extremely low Pd loading. The rGO pre-treatment proposed and demonstrated in this study will improve the cost-effectiveness of industrial metallisation processes because Pd pre-deposition has been employed in all electroless deposition processes over the past 50 years.

This work was supported by the Nanyang Assistant Professorship (NAP, M4080740), Singapore Ministry of Education (M4020194, MOE2013-T2-2-049) and Okuno-Auromex Co. Ltd (M4061665). The authors appreciate Ms Koh Joo Luang, Ms Yong Mei Yoke and Mr Leong Kwok Phui at School of Mechanical & Aerospace Engineering, NTU, for their continuous effort to set up and maintain the excellent experimental environment.

Notes and references

  1. E. D. Daniel, C. D. Mee and M. H. Clark, Magnetic recording: the first 100 years, John Wiley & Sons, 1999 Search PubMed.
  2. J. Severin, R. Hokke and H. Van der Wel, J. Electrochem. Soc., 1993, 140, 682–687 CrossRef CAS.
  3. L. B. Li, Y. Ma, G. X. Gao, H. Wang, X. C. Yang, J. C. Xie and W. T. Wang, Colloids Surf., A, 2015, 477, 42–48 CrossRef CAS.
  4. S. L. Brandow, W. J. Dressick, C. R. K. Marrian, G. M. Chow and J. M. Calvert, J. Electrochem. Soc., 1995, 142, 2233–2243 CrossRef CAS.
  5. T. Homma, K. Naito, M. Takai, T. Osaka, Y. Yamazaki and T. Namikawa, J. Electrochem. Soc., 1991, 138, 1269–1274 CrossRef CAS.
  6. B. J. Hwang and S. H. Lin, J. Electrochem. Soc., 1995, 142, 3749–3754 CrossRef CAS.
  7. K. S. Novoselov, A. K. Geim, S. V. Morozov, D. Jiang, Y. Zhang, S. V. Dubonos, I. V. Grigorieva and A. A. Firsov, Science, 2004, 306, 666–669 CrossRef CAS PubMed.
  8. J. Liu, L. Hua, S. Li and M. Yu, Appl. Surf. Sci., 2015, 327, 241–245 CrossRef CAS.
  9. H. Ashassi-Sorkhabi, A. Mirmohseni and H. Harrafi, Electrochim. Acta, 2005, 50, 5526–5532 CrossRef CAS.
  10. C. L. Lee, Y. L. Tsai and C. W. Chen, Electrochim. Acta, 2013, 104, 185–190 CrossRef CAS.
  11. P. Tang, G. Hu, Y. Gao, W. Li, S. Yao, Z. Liu and D. Ma, Sci. Rep., 2014, 4, 5901 CrossRef CAS PubMed.
  12. X. Wang, W. Y. Xing, B. Yu, X. M. Feng, L. Song and Y. Hu, J. Mater. Chem. C, 2013, 1, 690–694 RSC.
  13. R. Vidano, D. Fischbach, L. Willis and T. Loehr, Solid State Commun., 1981, 39, 341–344 CrossRef CAS.
  14. M. M. Demir, M. A. Gulgun, Y. Z. Menceloglu, B. Erman, S. S. Abramchuk, E. E. Makhaeva, A. R. Khokhlov, V. G. Matveeva and M. G. Sulman, Macromolecules, 2004, 37, 1787–1792 CrossRef CAS.
  15. C. L. Lee, L. C. Kuo, Y. C. Huang and Y. W. Yen, Electrochem. Commun., 2006, 8, 697–702 CrossRef CAS.


Electronic supplementary information (ESI) available: Details of experiments and additional characterization results. See DOI: 10.1039/c7cc03773b

This journal is © The Royal Society of Chemistry 2017