Platinum nanowires with pronounced texture, controlled crystallite size and excellent growth homogeneity fabricated by optimized pulsed electrodeposition

Abstract

Platinum nanowires with controlled texture and crystallite size were fabricated in nanoporous ion-track etched polycarbonate membranes by electrochemical deposition with different potential pulse sequences. The application of specific potential pulses ranging from −0.5 V to −1.3 V and reverse pulses with +0.2 V allows switching between a light 〈111〉 texture and a pronounced 〈100〉 texture along the nanowire axis. At the same time, the crystallite size determined by XRD was significantly increased from approx. 20 nm to 45 nm, yielding oligocrystalline wires, which are very difficult to obtain with Pt electrodeposition due to its pronounced tendency towards instantaneous nucleation. TEM verified the increase of the calculated average crystallite size of the Pt nanowires. We have been able to prove the necessity of each potential pulse and pulse duration by changing them in a systematical way. Key strategy to achieve large crystallite sizes and a pronounced texture was to reduce oversaturation of Pt adatoms during the reduction step and to preferentially dissolve lattice defects and nucleation sites by anodic removal. Additionally, the homogeneity of the Pt nanowire growth was evaluated for the applied pulse sequences by SEM. The results show that by reducing the deposition current density, the uniformity of the Pt nanowires was strongly enhanced. The proposed rational synthetic strategy allows to optimize the crystallinity, texture and monodispersity of Pt nanowires and is thus of considerable relevance for tailoring the functional properties of these structures.

---

Introduction

Metal nanowires (NWs) are an emerging class of functional nanomaterials which possess numerous interesting and desirable properties such as being precisely defined and showing shape anisotropy, high surface-to-volume ratio, pronounced magnetism,¹ improved yield strength² or high catalytic activity.³ By utilizing these functionalities, NWs have been successfully applied in gas sensing^4,5 and for designing highly durable and active heterogeneous catalysts.⁶ Due to the pronounced dependence of the NW properties on their size,¹ composition⁷ and crystallinity,⁸ synthetic approaches which ensure a high degree of control and variability are of major importance for both academic and applied research. For instance, in the case of improving chemoresistive gas-sensing devices based on metal NWs, it is necessary to adjust their crystallinity, owing to its influence on electrical conductivity.^9,10 Also, other fundamental properties of metal NWs are affected by the NW crystallinity. Moore et al. fabricated single and polycrystalline Bi NWs and observed a strong influence of the NW structure on the thermal conductivity.¹¹ In the case of Co NWs it has been shown that magnetic characteristics strongly depend on the control of shape anisotropy and texture.¹ The special magnetic properties of Ni NWs as a function of the crystallinity have also been varied by Napolskii et al.⁸

The application of NWs in devices gives rise to the need of adjusting the NWs properties by precisely tuning their structure via the growth process. NW arrays in e.g. gas sensing devices are often electrically contacted at each side.^4,5 An inhomogeneous NW growth therefore leads to a reduced amount of available NWs, since some do not have a connection to the contact electrode on each side. The application of NWs as gas sensing elements especially requires the ability to improve the gas sensing sensitivity⁷ and NW homogeneity.^8,12,13 This necessitates the development of manufacturing processes which allow controlling the NWs properties and demands an appropriate choice of material. For instance, Pt NWs are eminently eligible as gas sensing elements due to the high melting point and oxidation stability of Pt which enables the use of Pt NWs in high temperature applications and chemically reactive environment.

There are several methods for producing NWs with advanced reproducibility and shape control. Besides the electron-beam¹⁴ and nanoimprint lithography,^15,16 NWs can be fabricated by the use of vapor deposition¹¹ and oblique angle deposition method.⁵ The fabrication of NWs by electrochemical deposition into hard-pulse anodized AAO templates¹⁷ or into ion-track etched polycarbonate membranes¹⁸ allows an excellent degree of morphological control, while the nanostructure of the wires can be controlled by the electrodeposition parameters. Especially for Pt, which is a very important material for many applications, there is limited knowledge about the control of texture, crystallite size (CS) and growth homogeneity.^1,8,19–23 This is due to the very low critical nucleus size of Pt leading to polycrystalline deposits with small crystallite sizes. The influence of a low critical nucleus size has previously been discussed for Co, Ni and Rh by Tian et al.²³

In this work, Pt NWs were produced by electrochemical deposition into ion-track etched polycarbonate membranes. On the basis that the properties of Pt NWs can be changed by controlling the electrodeposition parameters,^24,25 we systematically investigated different potential pulse sequences for the electrochemical deposition of Pt NWs and examined the resulting textures, CS and homogeneity of NW growth by XRD, TEM and SEM. By choosing appropriate deposition conditions, we were able to significantly increase CS up to locally single crystalline Pt NWs with diameters of about 100 nm. Also nearly maximum 〈100〉 textured Pt NWs were achieved and a considerably improved homogeneity of the Pt NWs growth front was realized. The mechanism of the Pt NW growth is discussed elaborately.

Experimental

Polycarbonate foil (Makrofol N®, Bayer Leverkusen, nominal thickness: 30 μm) was irradiated with ¹⁹⁷Au²⁵⁺ ions (kinetic energy E ≈ 11 MeV per nucleon, fluence ϕ = 5 × 10⁸ ions cm⁻², using a honeycomb slit mask) at the UNILAC linear accelerator of GSI Helmholtzzentrum für Schwerionenforschung GmbH (Darmstadt, Germany). In this step, latent tracks were created within the polycarbonate foil which were selectively etched out with NaOH solution (6 M, 50 °C, 200 s etching time), yielding cylindrical nanopores of about 80–120 nm in diameter. After etching, a contact electrode was produced by sputter depositing a thin Au film on one side of the polycarbonate foil. Subsequently, the Au film was galvanically thickened by a coulometric Cu deposition of 2.6 C cm⁻² at a constant potential of −0.5 V resulting in an approximately 2 μm thick layer.

During electrodeposition of Pt a 2-electrode setup was used where the contact layer on the polycarbonate membrane served as the cathode and a coiled Pt wire was used as the anode. As electrolyte, a commercially available H₂Pt(OH)₆ solution (METAKEM Platinum-OH, Usingen, Germany, Pt content: 10 g l⁻¹) was employed. The Pt NW depositions were performed at an electrolyte temperature of 305 ± 0.5 K with a LabVIEW controlled Keithley 2602 dual-channel device. For variation of the Pt NW crystallinity, texture and growth homogeneity, 2 to 4 different potential pulses were applied as described in the text. To control the deposition of the Pt NWs, current vs. time curves were recorded, and the NW growth was stopped after 1.0 C to avoid the growth of caps on top of template-penetrating NWs.²⁶

After Pt NW deposition, the contact layer was chemically removed (Cu: diluted nitric acid; Au: aqueous solution containing KI and I). The crystallinity of the Pt NWs remaining parallel aligned within the pores of the polycarbonate membrane was investigated using X-ray diffraction (SEIFERT XRD 30003 PTS-3, Cu Kα, λ = 154.13 pm; for the measurement setup, see Fig. 1) and transmission electron microscopy (FEI CM 20 ST, 200 kV). For TEM sample preparation, the Pt NWs were embedded with resin prior to microtome cutting.

Fig. 1 Setup for the XRD measurement of the Pt NWs which are still within the polycarbonate membrane.

In addition, the Pt NWs were investigated with scanning electron microscopy (Philips XL30 FEG) for examining their growth homogeneity by measuring the length of 100 NWs per sample at different sample positions. For these measurements, the template was dissolved in dichloromethane on top of a Si wafer piece, leaving the isolated NWs on the wafer surface.

Results and discussion

Pt NWs were deposited using different pulse sequences, including reverse pulses and varying pulse durations. To examine the function of the applied pulses, a Pt NW sample was subjected to pulses with identical potential, but with elongated pulse duration in order to exclude the effect of capacitive currents. The resulting current vs. time response is depicted in Fig. 2 showing one cathodic pulse (−1.3 V) with Pt reduction, one anodic pulse (+0.2 V) with Pt surface dissolution, and two pulses (−0.5 V and −0.7 V) with no significant faradaic currents.

Fig. 2 Current vs. time response of a Pt NW sample with elongated pulse durations in order to exclude the effect of capacitive currents.

In Fig. 3, the diffraction patterns (a) and the electrodeposition parameters (b) of all performed experiments are summarized. All diffraction patterns clearly show the crystal structure of pure face-centered cubic (fcc) Pt; besides slight reflexes at about 2θ = 38°, owing to a small amount of Au residue due to the contact layer removal, no characteristic reflexes of other phases were detected. The reflexes of a standard Pt powder sample (ICDD pdf file 00-004-0802) are depicted in Fig. 3(a) as vertical lines directly above the 2θ-axis.

Fig. 3 The measured X-ray diffractograms (a) and the corresponding potential pulse sequences for the fabrication of the different Pt NW samples (b).

The Pt NWs of sample (a) in Fig. 3 were deposited using a −1.3 V potential pulse with a pulse duration of 300 ms, followed by a −0.5 V potential pulse with a pulse duration of 1000 ms. These electrodeposition parameters lead to a diffraction pattern showing a slight 〈111〉 texture along the Pt NW axis. The potential pulse sequence for sample (a) was used as a point of origin, because on the one hand the direct current electrodeposition has already been investigated for several fcc metals and resulted always in 〈110〉 textured nanowire growth.^25,27–31 On the other hand, the current density during the −0.5 V pulse of all samples is negligible and shows no faradaic contribution as mentioned above and depicted in Fig. 2. Therefore, the Pt reduction is only maintained by the use of the −1.3 V potential pulses.

For sample (b), an additional anodic pulse with a potential of +0.2 V and a short duration of 20 ms was added after the reduction pulse of the previous procedure which was intended to re-oxidize the Pt surface (see Fig. 2). The parameter set of sample (b) lead to a low 〈100〉 texture. In sample (c), a fourth, moderately cathodic pulse was introduced before the Pt reduction pulse (potential: −0.7 V; duration: 100 ms) which does not contribute to the reduction of Pt. Compared to sample (b), this procedure strengthened the 〈100〉 texture of the Pt NWs. The last investigated samples (d) and (e) were electrodeposited by using the same potential pulses as in sample (c). The only variations were the change of the pulse durations of the −1.3 V and the +0.2 V potential pulse, respectively. For the deposition of sample (d), the pulse duration of the reduction pulse at −1.3 V was reduced to 175 ms. The XRD analysis of sample (d) revealed a pronounced 〈100〉 texture of the Pt NWs along the wire axis. In comparison to sample (d), the electrodeposition of sample (e) was conducted just by changing the pulse duration of the +0.2 V potential pulse from 20 ms to 30 ms, revealing a further increase of the 〈100〉 texture. The summation of the electrodeposition parameters for manufacturing the Pt NWs are listed in Table 1.

Table 1 Potential pulse sequences for the fabrication of the electrodeposited Pt NWs together with the calculated values for the texture coefficients TC₍₁₁₁₎ and TC₍₁₀₀₎, average crystallite size CS, relative standard deviation RSD, average nanowire length L̄ after deposition of 1 C as well as the pore filling ratio PFR and the average current density j_AVG

Sample	Deposition parameters (pulse voltage/pulse duration)	Column 1	Column 2	Column 3	Column 4	Column 5	Column 6	Column 7
		TC(111)	TC(100)	CS [nm]	RSD [%]	L̄ [μm]	PFR [%]	jAVG [mA cm^−2]
(a)	−1.3 V/300 ms, −0.5 V/1000 ms	1.4	0.8	20	35	4, 6	26	7.9
(b)	−1.3 V/300 ms, +0.2 V/20 ms, −0.5 V/1000 ms	0.8	1.4	23	18	5, 8	33	6.5
(c)	−0.7 V/100 ms, −1.3 V/300 ms, +0.2 V/20 ms, −0.5 V/1000 ms	0.6	1.8	32	14	6, 8	39	5.7
(d)	−0.7 V/100 ms, −1.3 V/175 ms, +0.2 V/20 ms, −0.5 V/1000 ms	0.3	3.4	32	13	13	76	0.8
(e)	−0.7 V/100 ms, −1.3 V/175 ms, +0.2 V/30 ms, −0.5 V/1000 ms	0.1	3.8	45	8.8	17	97	0.5

---

The texture along the wire axis of the Pt NWs manufactured with different electrodeposition parameters was evaluated by calculating the texture coefficient (TC_(hikili)) using the following formula (1):³²

Here, I(h_ik_il_i) and I₀(h_ik_il_i) are the intensities of the Pt NW lattice planes (h_ik_il_i) and of a standard Pt sample, respectively. n denominates the total number of reflexes taken into account for the calculation of TC_(hikili). In our case, we used n = 4 because we investigated the reflexes 〈111〉, 〈200〉, 〈220〉 and 〈311〉. TC_(hikili) > 1 indicates a preferred orientation of the (h_ik_il_i) lattice plane perpendicular to the Pt NWs axis. The theoretically accessible maximum value of the calculated texture coefficient is TC_(hikili) = 4. In that case all (h_ik_il_i) lattice planes in the Pt grains are oriented perpendicular to the NWs axis. The results of the texture investigation are listed in column 1 and 2 of Table 1. All investigated samples didn't show significant 〈110〉 or 〈311〉 textures which is the reason for the absence of TC₍₁₁₀₎ and TC₍₃₁₁₎ in Table 1.

Supplementary, the average CS for each sample was calculated using the XRD data and Scherrer's formula (2) with the assumption of spherical crystallites which corresponds to a shape factor k of 0.89. The parameter λ is the used X-ray wavelength for the XRD measurements, FWHM is the full width have maximum value and cos θ the angle of the corresponding reflexes. The calculated values of the CS for each sample are listed in column 3 of Table 1.

In addition to the TC and CS values, the relative standard deviation (RSD) of the NW length was calculated by using formula (3) and a set of 100 wire lengths for each sample (measured by SEM). L_DEV is the standard deviation of the measured length values and L̄ is the average NW length. The calculated RSD and average NW length L̄ of each sample is tabulated in column 4 and 5 of Table 1. The RSD was introduced to evaluate the different electrodeposition parameters in terms of the homogeneity of the NW growth front. The lower the calculated RSD of a Pt NW sample, the more regular was the length of the NWs.

Furthermore, the pore filling ratio PFR was calculated with eqn (4) by using the average length L̄ of the Pt NWs (used in formula (3)) and the calculated theoretical length L_Th of the NWs, which was determined by using Faradays law. The calculation of the PFR with formula (4) was created, because the experiments were stopped at a charge of 1 C for the Pt NWs deposition. Therefore, it was not possible to determine the electrodeposited charge Q_exp, which is experimentally necessary for filling the pores completely and to compare Q_exp with Q_theo, where Q_theo is the theoretical charge calculated by using Faradays law for entirely filling the pores.²⁴ Considering the constant deposited charge for all experiments the calculated PFR (formula (4)) allows an overview for all applied potentials pulse sequences used for Pt electrodeposition in consideration of their ability to fill every available polymer pore completely. The higher the PFR the more homogeneous are all available pores filled with Pt NWs. The results for the calculation of the PFR are included in column 6 of Table 1.

The electrodeposition mechanism of the Pt NWs is based on the adsorption of the hexahydroxoplatinate ions [Pt(OH)₆]²⁻ of the used METAKEM Platinum-OH electrolyte on the electrode surface and the generation of adatoms by a formal 4-electron-reduction per ion. The adatoms either aggregate to form new nuclei or integrate into the crystal lattice of the NW. To switch from a low energetic TC₍₁₁₁₎ texture (sample a) to the thermodynamically less favorable TC₍₁₀₀₎ texture (sample e), we had to optimize the ratio between the faradaic currents for Pt crystallite growth and the currents for the re-oxidation of loosely bound surface atoms and lattice defects. This consideration is approved with the results of Maurer et al.²⁵ They investigated the surface energies necessary for the growth of NWs with 〈111〉, 〈100〉 and 〈110〉 textures dependent on their aspect ratio. They concluded that NWs consisting of fcc metals and grown by direct current electrodeposition conditions reveal a 〈110〉 texture along their wire axis in the case of an aspect ratio h/d higher than 5, where h is the height and d the NWs diameter. A further achievement of Maurer et al. was the 〈100〉 texture for Cu and Au NWs with an aspect ratio above 5 by using alternating current conditions, where the {110} surfaces were preferentially dissolved.

For the growth of Pt NWs, a possibility for the reduction of the faradaic current is given by changing the Pt precursor concentration near the electrode surface. The high concentration of the METAKEM Platinum-OH electrolyte (10 g l⁻¹) is suspected to give rise to an instantaneous Pt nucleation on the NW surface due to a pronounced supersaturation of Pt adatoms produced by [Pt(OH)₆]²⁻ reduction.³³ The change from the instantaneous nucleation to a much lower amount of nuclei on top of the cathode surface and with that to lower faradaic currents during the −1.3 V potential pulses is supposed to be available by reducing the Pt concentration in the vicinity of the cathode surface by applying suitable potential pulses.

Considering the negative charge of the [Pt(OH)₆]²⁻, we used the potential pulses of −0.5 V and −0.7 V to control the Pt concentration at the electrode surface right before the upcoming reduction with a −1.3 V potential pulse. The influence of the −0.5 V potential pulse was supposed to prevent an accumulation of [Pt(OH)₆]²⁻ near the electrode surface. The −0.7 V potential pulse in comparison to the −0.5 V potential pulse was assumed to generate a much stronger repulsion and hence was used to cause a depletion of the [Pt(OH)₆]²⁻ concentration in the vicinity of the electrode surface. During the following −1.3 V potential pulse, only a few adatoms are created because of the preliminary concentration depletion with the −0.7 V potential pulse, resulting in a small number of nuclei arised by the reduction of the adsorbed [Pt(OH)₆]²⁻ to the elemental state. By using the −0.5 V and −0.7 V potential pulses, the reduction process was running with low faradaic currents and hence lead only to a small amount of crystallite growth. Comparing the pulse sequences with differently strong [Pt(OH)₆]²⁻ repulsion just before the Pt reduction pulse, according to the above-mentioned reasoning, lower faradaic currents were found in case of stronger depletion (sample c), (with −0.7 V pulse) than compared to fewer depletion (sample b), (without −0.7 V pulse). The average reduction currents of sample (c) during the −1.3 V potential pulses are 15.9% lower than the corresponding currents of sample (b).

Vice versa, the +0.2 V potential pulse induced an increase in the [Pt(OH)₆]²⁻ concentration at the electrode's surface because of the attraction of negatively charged ions and due to the re-oxidation of loosely bound surface atoms, {110} surfaces, and defects on the crystal surface.^19–21 The latter are supposed to be active sites for nucleation¹⁹ which results in a reduced quantity of growing nuclei during the Pt reduction in case of their preferential dissolution by the use of the +0.2 V potential pulse. The variation of the −1.3 V and +0.2 V potential pulse durations finally was the key to optimize the re-oxidation process and the appropriate concentration of [Pt(OH)₆]²⁻ at the electrode surface. This step enabled a further increase in the 〈100〉 texture of the electrodeposited Pt NWs yielding a maximum value of TC₍₁₀₀₎ = 3.8 for sample (e) (see Table 1, column 2). To our best knowledge the maximum reported TC₍₁₀₀₎ for electrodeposited Pt NWs in templates amounts to 2.35.²⁴

The calculation results of the CS, RSD and PFR together with the resulting average length L̄ of the Pt NWs listed in columns 3 to 6 of Table 1 are illustrated in Fig. 4 and 5 with respect to the calculated average current densities j_AVG. The j_AVG is also included in column 7 of Table 1.

Fig. 4 Calculated values of the relative standard deviation RSD and the crystallite size CS are depicted as a function of calculated average current density j_AVG.

Fig. 5 The calculated pore filling ratio PFR and the average length L̄ of the Pt NWs are shown versus the calculated average current density j_AVG.

Fig. 4 and 5 show that decreasing j_AVG also reduces the RSD while the CS, PFR and L̄ increase. The rise of the CS by diminishing j_AVG has already been observed for Ni.³⁴ The behavior of the RSD, CS, PFR and L̄ can be explained by the application and duration of the −0.7 V potential pulse and +0.2 V reverse potential pulse. Comparison of sample (a) and (b) or of sample (d) and (e) clearly shows that the use of the +0.2 V potential pulse and an increased duration of this pulse results in a decrease of j_AVG and RSD as well as in an increase of CS, PFR and L̄. In the same manner, comparing sample (b) and (c), the influence of the −0.7 V potential pulse leads to a decrease of j_AVG and RSD plus to an increase of CS, PFR and L̄.

It is supposed that the reduced RSD and increased CS, PFR and L̄ values are related to the preferential re-oxidization of surface inhomogeneities during the reverse pulse³⁵ and to the reduced quantity of adatoms due to the −0.7 V pulse, leading to a lower amount of growing crystallites and a more regular Pt NW growth. Increasing the ratio of the pulse durations t_{+0.2 V}/t_{−1.3 V} further minimizes the RSD as well as rises the CS, PFR and L̄ value, as can be seen by comparing sample (c) and (d) as well as (d) and (e). In accordance with the previous reasoning, this is expected since a higher relative duration of the re-oxidation pulse leads to a more extensive removal of irregularities on the NW surface, to a reduced amount of growing crystallites and thus to more equal growth conditions during the reduction pulse.^35,36 A low amount of growing crystallites causes high mean CS, because the average distance between the crystallites is high which results in a more undisturbed growth in every direction before they impinge on each other.³⁷

In addition, the variation of the CS within the Pt NWs, as derived from X-ray diffraction data, was investigated by transmission electron microscopy (TEM). Fig. 6(a) and (b) show representative bright field (BF) images of sample (a) in overview and at a higher magnification, respectively. From Fig. 6(b) the polycrystalline nature of the nanowire is clearly visible.

Fig. 6 (a) Overview bright field (BF) TEM image of a microtome thin section parallel to the nanowire axis and (b) corresponding higher magnification, revealing the polycrystalline nature of sample (a). In (c) and (d) dark field (DF) TEM images of sample (a) and (e) are shown, respectively. The insets reveal the corresponding SAED patterns and the diffraction spot utilized for DF image formation. Clearly, the polycrystalline nature of both samples and the markedly different crystallite size could be determined, being consistent with X-ray diffraction data.

The comparison of Fig. 6(c) and (d), which represent corresponding dark field (DF) images of sample (a) and (e), also reveal the polycrystalline nature of both. While sample (a) shows an average CS of approximately 15 nm, sample (e) was determined to have a CS ranging between 35 and 60 nm. The dark field TEM images confirm the results of the XRD measurement depicted in Fig. 4 (also listed in Table 1) with increasing CS from sample (a) through (e). In accordance with the above-mentioned growth mechanism favoring renucleation, other papers on the template-based electrodeposition of Pt report highly polycrystalline products with small CSs below 10 nm.^24,38,39

Finally the experimental reproducibility was examined by repeating electrodeposition of Pt NWs using the parameter sets of sample (d) and (e) (see Table 1). The maximum differences between the original and repeated samples where 11.9% for CS, 11.4% for RSD as well as no measurable discrepancy for TC₍₁₁₁₎ and TC₍₁₀₀₎. This results clearly underline the great potential of this technique and its applicability for reproducibly fabricating highly textured Pt NWs with increased CS and homogeneous growth fronts.

Conclusion

We have been able to fabricate highly textured Pt NWs via the control of electrodeposition parameters. The NW crystallinity was examined by XRD, and the results were confirmed by TEM measurements. The rational application of specific potential pulse sequences allowed switching between a light 〈111〉 texture and a strong 〈100〉 texture along the NW axis, simultaneously increasing the CS, L̄ and the homogeneity of the NW growth front. We have been able to show the necessity of each potential pulse and pulse duration by changing them in a systematical way.

From a mechanistic point of view, the pulses serve three different purposes: The electrodeposition of Pt, the oxidation of superficial Pt inhomogeneities and the depletion of the ionic Pt precursor close to the electrode surface. By using the −1.3 V reduction pulse with a pulse duration of 300 ms in conjunction with a −0.5 V depletion pulse of 1000 ms duration, a light 〈111〉 texture along the NW axis was obtained. Introducing a second, stronger depletion pulse (−0.7 V, 100 ms duration) before and by adding an anodic pulse (+0.2 V, 20 ms duration) after the reduction pulse, the development of the energetically preferred 〈110〉 NW texture is prevented, leading to a 〈100〉 texture along the NW axis. This texture is further strengthened by decreasing the duration of the reduction pulse and by increasing the duration of the re-oxidation pulse. Key strategy behind all changes in the pulse sequence was to reduce the number of Pt renucleation events, favoring the epitaxial growth of the NWs, thus increasing both the TC and CS. As an additional result of the well-defined nature of the NW growth fronts, the length of the deposited Pt NWs became more homogeneous.

The outlined possibilities to control the crystallinity and regularity of Pt NWs will be of importance for the application-oriented optimization of such nanomaterials.

Acknowledgements

This work has been supported by the Federal Ministry of Education and Research and the VDI/VDE-IT (Project 3-DOING (FKZ 16SV5476)). We thank the members of the Materials Research Department at the GSI Helmholtzzentrum für Schwerionenforschung GmbH (Darmstadt, Germany) for irradiation of the 30 μm thick polycarbonate foil.

References

1. L. Cattaneo, S. Franz, F. Albertini, P. Ranzieri, A. Vicenzo, M. Bestetti and P. L. Cavallotti, Electrochim. Acta, 2012, 85, 57–65.
2. B. Wu, A. Heidelberg and J. J. Boland, Nat. Mater., 2005, 4, 525–529.
3. M. Rauber, I. Alber, S. Müller, R. Neumann, O. Picht, C. Roth, A. Schökel, M. E. Toimil-Molares and W. Ensinger, Nano Lett., 2011, 11, 2304–2310.
4. B. Wang, L. F. Zhu, Y. H. Yang, N. S. Xu and G. W. Yang, J. Phys. Chem. C, 2008, 112, 6643–6647.
5. S. Hwang, H. Kwon, S. Chhajed, J. W. Byon, J. M. Baik, J. Im, S. H. Oh, H. W. Jang, S. J. Yoon and J. K. Kim, Analyst, 2013, 138, 443–450.
6. B. Y. Xia, H. B. Wu, Y. Yan, X. W. Lou and X. Wang, J. Am. Chem. Soc., 2013, 135, 9480–9485.
7. O. Lupan, L. Chow, T. Pauporte, L. K. Ono, B. R. Cuenya and G. Chai, Sens. Actuators, B, 2012, 173, 772–780.
8. K. S. Napolskii, I. V. Roslyakov, A. A. Eliseev, D. I. Petukhov, A. V. Lukashin, S.-F. Chen, C.-P. Liu and G. A. Tsirlina, Electrochim. Acta, 2011, 56, 2378–2384.
9. F. Yang, K. C. Donavan, S.-C. Kung and R. M. Penner, Nano Lett., 2012, 12, 2924–2930.
10. B. J. Murray, E. C. Walter and R. M. Penner, Nano Lett., 2004, 4, 665–670.
11. A. L. Moore, M. T. Pettes, F. Zhou and L. Shi, J. Appl. Phys., 2009, 106, 034310.
12. D. Borissov, S. Isik-Uppenkamp and M. Rohwerder, J. Phys. Chem. C, 2009, 113, 3133–3138.
13. L. Trahey, C. R. Becker and A. M. Stacy, Nano Lett., 2007, 7, 2535–2539.
14. P. Candeloro, A. Carpentiero, S. Cabrini, E. Di Fabrizio, E. Comini, C. Baratto, G. Faglia, G. Sberveglieri and A. Gerardino, Microelectron. Eng., 2005, 78–79, 178–184.
15. A. A. Talin, L. L. Hunter, F. Léonard and B. Rokad, Appl. Phys. Lett., 2006, 89, 153102.
16. J. Yeon, Y. J. Lee, D. E. Yoo, K. J. Yoo, J. S. Kim, J. Lee, J. O. Lee, S. J. Choi, G. W. Yoon, D. W. Lee, G. S. Lee, H. C. Hwang and J. B. Yoon, Nano Lett., 2013, 13, 3978–3984.
17. A. S. Esmaeily, M. Venkatesan, A. S. Razavian and J. M. D. Coey, J. Appl. Phys., 2013, 113, 17A327.
18. M. E. Toimil-Molares, Beilstein J. Nanotechnol., 2012, 3, 860–883.
19. D. Dobrev, J. Vetter, N. Angert and R. Neumann, Appl. Phys. A, 1999, 69, 233–237.
20. D. Dobrev, J. Vetter, N. Angert and R. Neumann, Appl. Phys. A, 2001, 72, 729–733.
21. D. Dobrev, J. Vetter, N. Angert and R. Neumann, Electrochim. Acta, 2000, 45, 3117–3125.
22. G. Riveros, H. Gómez, A. Cortes, R. E. Marotti and E. A. Dalchiele, Appl. Phys. A, 2004, 81, 17–24.
23. M. Tian, J. Wang, J. Kurtz, T. E. Mallouk and M. H. W. Chan, Nano Lett., 2003, 3, 919–923.
24. M. Rauber, J. Brötz, J. Duan, J. Liu, S. Müller, R. Neumann, O. Picht, M. E. Toimil-Molares and W. Ensinger, J. Phys. Chem. C, 2010, 114, 22502–22507.
25. F. Maurer, J. Brötz, S. Karim, M. E. Toimil Molares, C. Trautmann and H. Fuess, Nanotechnology, 2007, 18, 135709.
26. M. E. Toimil Molares, V. Buschmann, D. Dobrev, R. Neumann, R. Scholz, I. U. Schuchert and J. Vetter, Adv. Mater., 2001, 13, 62–65.
27. M. E. Toimil Molares, J. Brötz, V. Buschmann, D. Dobrev, R. Neumann, R. Scholz, I. U. Schuchert, C. Trautmann and J. Vetter, Nucl. Instrum. Methods Phys. Res., Sect. B, 2001, 185, 192–197.
28. H. Pan, H. Sun, C. Poh, Y. Feng and J. Lin, Nanotechnology, 2005, 16, 1559–1564.
29. J. Zhang, X. Wang, X. Peng and L. Zhang, Appl. Phys. A, 2002, 75, 485–488.
30. G. Riveros, S. Green, A. Cortes, H. Gómez, R. E. Marotti and E. A. Dalchiele, Nanotechnology, 2006, 17, 561–570.
31. S. Karim, M. E. Toimil-Molares, F. Maurer, G. Miehe, W. Ensinger, J. Liu, T. W. Cornelius and R. Neumann, Appl. Phys. A, 2006, 84, 403–407.
32. G. B. Harris, Philosophical Magazine Series 7, 1952, vol. 43:336, pp. 113–123.
33. M. Paunovic and M. Schlesinger, in Fundamentals of Electrochemical Deposition, ed. M. Paunovic and M. Schlesinger, Wiley-Interscience, Hoboken, 2nd edn, 2006, vol. 2, p. 115.
34. A. M. Rashidi and A. Amadeh, Surf. Coat. Technol., 2008, 202, 3772–3776.
35. J.-C. Puippe, H. Y. Cheh, D. Landolt, O. Dossenbach, W. Paatsch, B. Sutter, A. Knödler, A. J. Avila, H. Konno and N. M. Osero, in Theory and Practice of Pulse Plating, ed. J.-C. Puippe and F. Leaman, American Electroplaters and Surface Finishers Society, Orlando, 1986, vol. 1, pp. 8–9.
36. E. H. Lyons, in Modern Electroplating, ed. F. A. Lowenheim, Wiley-Interscience, New York, 2nd edn, 1963, vol. 2, p. 37.
37. M. Paunovic and M. Schlesinger, in Fundamentals of Electrochemical Deposition, ed. M. Paunovic and M. Schlesinger, Wiley-Interscience, Hoboken, 2nd edn, 2006, vol. 2, p. 117.
38. C.-H. Cui, H.-H. Li and S.-H. Yu, Chem. Commun., 2010, 46, 940–942.
39. S. M. Choi, J. H. Kim, J. Y. Jung, E. Y. Yoon and W. B. Kim, Electrochim. Acta, 2008, 53, 5804–5811.

---