Etched heavy ion tracks in polycarbonate as template for copper nanowires

Etched heavy ion tracks in polycarbonate as template for copper nanowires

Nuclear Instruments and Methods in Physics Research B 185 (2001) 192±197 www.elsevier.com/locate/nimb Etched heavy ion tracks in polycarbonate as te...

340KB Sizes 1 Downloads 33 Views

Nuclear Instruments and Methods in Physics Research B 185 (2001) 192±197

www.elsevier.com/locate/nimb

Etched heavy ion tracks in polycarbonate as template for copper nanowires M.E. Toimil Molares a,*, J. Br otz b, V. Buschmann b, D. Dobrev a, R. Neumann a, R. Scholz c, I.U. Schuchert a, C. Trautmann a, J. Vetter a a

b

Gesellschaft fur Schwerionenforschung ± GSI, Planckstr. 1, D-64291 Darmstadt, Germany FG Strukturforschung, FB 21, TU- Darmstadt, Petersenstr. 23, D-64287 Darmstadt, Germany c Max-Planck-Institut fur Mikrostrukturphysik, Weinberg 2, D-06120 Halle, Germany

Abstract 30 and 40 lm thick polycarbonate foils were irradiated with Au197 and Pb208 ions of kinetic energy 1±2 GeV with ¯uences between 106 and 109 ions/cm2 . The latent tracks generated by the heavy ions were chemically etched providing membranes with cylindrical pores of diameters between 30 and 200 nm. These membranes have been used as templates for the creation of metallic nanowires of very high aspect ratio. A thin metal ®lm deposited on one side of the membrane acted as cathode in the two-electrode electrochemical cell, while a copper cone served as anode. The wires were grown potentiostatically. The electrochemical process was monitored by registering chronoamperometric curves for di€erent cathode overvoltages, temperatures and concentrations of the electrolyte. Under suitable conditions, single-crystalline needles were produced. The morphology and crystallinity of the copper nanowires were studied by means of scanning electron microscopy (SEM), transmission electron microscopy (TEM) and X-ray di€raction. Ó 2001 Elsevier Science B.V. All rights reserved.

1. Introduction Recently, increasing interest has focused on the investigation of nanoscale materials. They have potential applications in many di€erent ®elds such as magnetism [1±3], optics [4±6] and electronics [7]. Tremendous in¯uence on nanotechnologies is expected due to new physical properties arising from the small dimensions. In particular, nanowires have been produced and investigated by many

*

Corresponding author. E-mail address: [email protected] (M.E. Toimil Molares).

di€erent techniques. In this paper, we report the production of copper nanowires using the template method [8,9]. In comparison with lithographic techniques (electron- or X-ray lithography) the template method allows us to create very thin wires (as small as a few tens of nanometers) with aspect ratios (length over diameter) as high as 1000. Various types of structures have been synthesized by di€erent groups such as tips or tubes of conducting polymers [10,11], metallic needles [12±15] and multilayered wires [1,2]. As templates, most commonly porous alumina, mica and, as used here, polymeric ion track membranes are employed. For the formation of suitable ion track

0168-583X/01/$ - see front matter Ó 2001 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 8 - 5 8 3 X ( 0 1 ) 0 0 7 5 5 - 8

M.E. Toimil Molares et al. / Nucl. Instr. and Meth. in Phys. Res. B 185 (2001) 192±197

membranes, the etch rate along the track (Vt ) should be much larger than the etch rate of the undamaged bulk material (Vb ). The ratio Vt =Vb determines the resulting geometry of the pores [16]. For a given set of etching parameters, this ratio mainly depends on the damage density in the track, which is directly related to the stopping power of the projectiles in the polymer …dE=dx† [17]. In particular, polycarbonate and polyethyleneterephthalate have been used in the membrane technology because of a high Vt =Vb ratio from several hundreds to 1000 leading to extremely cylindrical pores. We present experiments of copper deposition in polycarbonate membranes with pore diameters between 30 and 200 nm. The resulting nanowires are investigated regarding morphology and crystallinity by means of scanning electron microscopy (SEM), transmission electron microscopy (TEM), and X-ray di€raction. 2. Experiments and results 2.1. Production of polymer template 30 and 40 lm thick polycarbonate foils (Makrofol N, Bayer Leverkusen) were irradiated at the linear accelerator (UNILAC) of GSI with heavy ions (Au197 and Pb208 ) of 11.4 MeV per nucleon at normal incidence. Under these conditions, the penetration range of the ions in polycarbonate was larger than the thickness of the foils and the dE=dx of the ions was well above the threshold required for homogeneous etching [16]. Depending on the intended pore diameter, ¯uences between 106 and 109 ions/cm2 were applied. The irradiated polycarbonate foils were chemically etched at 50°C in a 6 N NaOH solution containing 10% methanol and 1% surfactant. The resulting pores are cylindrical, their diameters increasing linearly with the time of etching [18]. In this work we created pores with diameters between 30 and 200 nm. 2.2. Electrochemical deposition of copper Prior to the electrochemical deposition process, a thin gold layer (thickness  100 nm) was sput-

193

tered onto one side of the polymer membrane and subsequently reinforced galvanostatically with copper. This metal layer provides a stable substrate for the growth of needles. Simultaneously, the metallized side of the membrane served as cathode in our two-electrode electrochemical cell [14]. As anode, we used a cone-shaped copper electrode containing a small amount of phosphor. Due to this geometry, the electrochemical deposition is homogeneous over the whole membrane surface. The electrolyte consisted of an aqueous solution of 238 g=l CuSO4  5H2 O and 21 g/l sulfuric acid. High concentration of CuSO4 is important to provide a suciently large number of copper ions inside the pores during the galvanic deposition process. Sulfuric acid increases the conductivity of the solution and lowers the cathode overvoltage [19]. Electrodeposition was performed potentiostatically at temperatures between room temperature and 70 °C. By applying low voltages …jlj < 120 mV†, side reactions such as hydrogen evolution were avoided. During the deposition process, we recorded the electrical current as a function of time. A typical chronoamperometric curve is presented in Fig. 1. In general, four di€erent zones can be distinguished. The sharp current increase at the beginning of the process (I) is ascribed to the charge of the double layer and the creation of the di€usion layer. During the growth of the needles in the pores, the current remains nearly constant (II). As

Fig. 1. Current versus time curve for the deposition of copper nanowires with diameter 50 nm at 50 °C and overvoltage l ˆ 120 mV. The di€erent deposition phases (I±IV) are represented schematically.

194

M.E. Toimil Molares et al. / Nucl. Instr. and Meth. in Phys. Res. B 185 (2001) 192±197

Fig. 2. SEM images of (a) single- and (b) poly-crystalline caps, grown after the copper deposition has reached the surface of the polymeric membrane.

soon as the wires reach the upper surface, caps are formed and the current increases (III). Finally caps are growing to macroscopic sizes at an almost constant rate (IV). The deposition process was stopped either during phase II (to characterize the needles) or in phase IV (to observe the caps). In the following, we will describe di€erent observations concerning morphology and crystallinity of the copper needles, whereas a detailed study of the various deposition parameters (electrolyte, current density, temperature) in¯uencing the crystalline structure of the metal has been published elsewhere [20]. 2.3. Microscopic analysis: morphology and crystallinity When the deposition process is stopped during phase III (cf. Fig. 1), a ®rst indication for the crystalline needle structure is given by the shape of the caps, grown on the membrane surface. Fig. 2 displays two typical cases: faceted caps are related to single-crystalline needles (Fig. 2(a)), whereas caps with round shape are formed if the needles are poly-crystalline (Fig. 2(b)). For the characterization of the needles by means of SEM and TEM, the electrodeposition process was stopped during phase II. Copper wires, freestanding on the substrate, were obtained by dissolving the polycarbonate matrix in dichloromethane. We used selected-area electron di€raction (SAED), and dark- and bright-®eld imaging.

A TEM image of both a single- and a polycrystalline wire together with their corresponding SAED patterns is shown in Fig. 3. In the case of Fig. 3(a), the wire was deposited at 50 °C and a voltage of l ˆ 50 mV. The needle possesses cylindrical geometry with a constant diameter and a smooth contour over the entire length. The contour of the poly-crystalline wire (Fig. 3(b)) is clearly rougher. In this case, the deposition took place at room temperature and at much higher voltages. Fig. 4 shows a TEM image of a wire with diameter 60 nm, which was deposited at 50 °C by applying a voltage l ˆ 45 mV. The high-magni®cation image, exhibiting the undisturbed lattice with atomic resolution, con®rms the single-crystallinity of the needle. Moreover, no defects or inclusions of a second phase are observed. Note that the diameter is constant over the full length. Our smallest copper wire produced up to now has a diameter of 30 nm and an aspect ratio of 1000 (Fig. 5). 2.4. X-ray di€raction: texture In order to con®rm the single-crystalline quality and to investigate possible textures of the wires, Xray di€raction was performed on needles of diameter 60 nm, still embedded in the polycarbonate membrane (see deposition conditions of Fig. 3). A STOE four-circle di€ractometer with graphite  monochromatized CoKa radiation (k ˆ 1:7902 A) was employed. The di€ractogram of a singlecrystalline sample shows an extremely strong

M.E. Toimil Molares et al. / Nucl. Instr. and Meth. in Phys. Res. B 185 (2001) 192±197

195

Fig. 3. TEM images of (a) single- and (b) poly-crystalline wires. The insets show SAED patterns, respectively con®rming the single- and polycrystalline structure.

Fig. 4. Sequence of TEM pictures of a copper nanowire of diameter 60 nm, showing its homogeneous cylindrical shape. The highresolution image con®rms the single-crystallinity.

196

M.E. Toimil Molares et al. / Nucl. Instr. and Meth. in Phys. Res. B 185 (2001) 192±197

Fig. 5. SEM image of a copper nanowire with aspect ratio 1000 (length 30 lm and diameter 30 nm) showing a smooth and homogeneous contour.

(2 2 0) re¯ection (Fig. 6(a)). The ratio of the measured intensities of the (2 2 0) and the (1 1 1) re¯ections is by a factor of 18 larger than expected for a standard powder sample, which indicates a strong h1 1 0i texture. The rocking curve of the (2 2 0) copper re¯ection is given in Fig. 6(b). While the (1 1 1) and the (1 0 0) re¯ection planes did not show any preferred orientation, we found a strong texture for the (1 1 0) re¯ection planes. The maximum at x ˆ 44° gives evidence for strongly preferred orientation of the di€racting planes parallel to the substrate surface (i.e. perpendicular to the wires). Most striking is the narrow peak width (FWHM ˆ 0.6°) indicating both a very high degree of orientation of the crystalline planes in the copper wires and also of the wires with respect to the membrane. 3. Discussion

Fig. 6. (a) h±2h di€ractogram measured on 60 nm diameter copper nanowires. (b) Rocking curve on the (2 2 0) re¯ection, centered at x ˆ 44°. The FWHM is 0.6°.

As the shape of the wires directly re¯ects the geometry of the pores in the polycarbonate, there is clear evidence that under the employed irradiation and etching conditions, the pores are well aligned and have cylindrical geometry. It should be emphasized that needles with cigarlike shapes as discussed by several authors [21,22] were never observed. We assume that such an e€ect is linked either to special properties of the polymer or caused by additives to the etchant solution (e.g. surfactants) [23]. The surface smoothness of the wires depends on several factors such as the quality of the polymer and the etching process. Our observations strongly indicate that also the conditions during electrodeposition play an important role. The presence of air bubbles in the pores, hydrogen evolution during depositions at too high overvoltages, or bad quality of the deposited metal at too high current densities result in rather large roughness of the wire surface. Single-crystalline wires, grown at higher temperatures and lower voltages [20], possess very homogeneous and smooth surfaces. Concerning the single-crystallinity of the wires, the X-ray di€raction analysis con®rms the results obtained by TEM. In addition, it gives evidence for a good alignment of the pores perpendicular to

M.E. Toimil Molares et al. / Nucl. Instr. and Meth. in Phys. Res. B 185 (2001) 192±197

the membrane surface. This property of our templates can be of great importance if homogeneous needle growth is required. Problems due to large angular distributions with tilt angles of more than 30° in commercially available membranes have been reported so far [24]. Polycarbonate membranes, as presented in this paper, provide suitable templates for basic research and for applications where a high degree of parallelism is of essential importance. Topics of interest are the study of magnetic and electrical properties, and the creation of magneto-resistive devices, photonic elements and ®eld emission arrays, to mention a few. In conclusion, it should be emphasized that polycarbonate foils irradiated with energetic heavy ions are most suitable for the production of templates with cylindrical pores. In such membranes, we have grown copper wires with length 30 lm and diameters as small as 30 nm. Under appropriate conditions, even single-crystalline wires exhibiting strong h1 1 0i texture were generated. Acknowledgements The authors would like to thank E. Hoehberger and L. Pescini for their help with SEM measurements. References [1] L. Piraux, J.M. George, J.F. Despres, C. Leroy, E. Ferain, R. Legras, K. Ounadjela, A. Fert, Appl. Phys. Lett. 65 (1994) 2484. [2] K. Liu, K. Nagodawithana, P.C. Searson, C.L. Chien, Phys. Rev. B (1995) 7381.

197

[3] V. Scarani, B. Doudin, J.-P. Ansermet, J. Magn. Mater. 205 (1999) 248. [4] M. Saito, T. Kano, T. Seki, M. Miyagi, Infrared Phys. Technol. 35 (1994) 709. [5] C.A. Foss Jr., M.J. Tierney, C.R. Martin, J. Phys. Chem. 96 (1992) 9001. [6] C.A. Foss Jr., G.L. Hornyak, J.A. Stocker, C.R. Martin, Adv. Mater. 5 (1993) 135. [7] S. Krongelb, L.T. Romankiw, J.A. Tornello, IBM J. Res. Develop. 42 (1998) 575. [8] J. Vetter, R. Spohr, Nucl. Instr. and Meth. B 79 (1993) 691. [9] C.R. Martin, Science 266 (1993) 1961. [10] V.P. Menon, J. Lei, C.R. Martin, Chem. Mater. 8 (1996) 2382. [11] S. Demoustier-Champagne, E. Ferain, C. Jerome, R. Jerome, R. Legras, Eur. Polym. J. 34 (1998) 1767. [12] D. Dobrev, J. Vetter, N. Angert, R. Neumann, Electrochim. Acta 45 (2000) 3117. [13] C.J. Brumlik, V.P. Menon, C.R. Martin, J. Mater. Res. 9 (1994) 1174. [14] D. Dobrev, J. Vetter, N. Angert, Nucl. Instr. and Meth. B 149 (1999) 207. [15] J. Vetter, R. Spohr, Nucl. Instr. and Meth. B 79 (1993) 691. [16] C. Trautmann, S. Bou€ard, R. Spohr, Nucl. Instr. and Meth. B 116 (1996) 429. [17] C. Trautmann, Nucl. Instr. and Meth. B 105 (1995) 81. [18] R.L. Fleischer, P.B. Price, R.M. Walker, Nuclear Tracks in Solids, Principles and Applications, University of California Press, Berkeley, 1975. [19] Gmelins Handbuch der Anorganischen Chemie, Copper, Vol. 60, Verlag Chemie, Weinheim, 1955. [20] M.E. Toimil Molares, V. Buschmann, D. Dobrev, R. Neumann, R. Scholz, I.U. Schuchert, J. Vetter, Adv. Mater. 13 (2001) 62. [21] C. Sch onenberger, B.M.I. van der Zande, L.G.J. Fokkink, M. Henny, C. Schmid, M. Kr uger, A. Bachtold, R. Huber, H. Birk, U. Staufer, J. Phys. Chem. B 101 (1997) 5407. [22] E. Ferain, R. Legras, Nucl. Instr. and Meth. B 131 (1997) 97. [23] P. Apel, JINR, Dubna, Russia, Private communication. [24] V.P. Menon, C.R. Martin, Anal. Chem. 67 (1995) 1920.