Magnificent CdS three-dimensional nanostructure arrays: the synthesis of a novel nanostructure family for nanotechnology

Abstract

Magnificent CdS three-dimensional nanostructure arrays, composed of caky plating made up of Cd micro-platelet arrays which are surrounded by an outer layer of well-aligned CdS nanowire/pillar arrays, were successfully prepared through a combination of electroplating and subsequent solvothermal reaction on Cd-coated copper substrates. The nucleation and growth behavior of CdS nanowire/pillars were qualitatively analyzed in terms of kinetics. The results demonstrated that the microstructure of the products shows great dependence on the synthetic conditions, including reactant concentration, growth temperature and time, and the morphological characteristics of the substrate. The coalescence growth of adjacent CdS nanowire/pillars during the growth stage was firstly demonstrated. In addition, the formation mechanism of caky Cd plating was well discussed and also a hydrogen bubbles-assisted growth mechanism based on electrochemistry was proposed to explain the unique plating. The relationship between the microstructure of the products and the synthetic conditions is beneficial for modifying the shape of CdS nanostructures. The as-prepared CdS three-dimensional nanostructure arrays are advantageous for their large surface area, highly ordered structure and conductive growth substrates, and thus may have great potential in many advanced material areas, especially in the field of solar cells.

---

Introduction

With the rapid development of characterization there has been an enhanced capability of microscopic observation, and much research has focused on the fields of nanomaterials and nanostructures over the past decades. In the big family of nanomaterials, one-dimensional (1D) semiconductor nanostructures such as nanowires, nanorods, and nanotubes are of particular interest to researchers owing to their unique geometry and intriguing properties and accordingly potential applications in many advanced nanodevices.^1–9 In particular, the synthesis of well-aligned 1D nanostructures is a subject of increasing interest in that they are expected to enhance the performance of various nanodevices. As an important II–VI group semiconductor compound, 1D CdS crystallites (band gap energy of 2.42 eV at 300 K) have proved to be promising candidate materials in many fields such as solar cells, field-effect transistors, lasers, light emitting diodes, and field emission devices.^10–14 Recently, the successful demonstration of excellent photovoltaic properties of 1D CdS nanopillars used in CdS/CdTe thin film solar cells again verifies the promising research value of aligned 1D CdS nanostructures.¹⁵ Thus far, numerous methods have been developed for fabricating ordered 1D CdS nanostructures. Among the various reported methods in the literature, the typical thermal evaporation techniques based on well-known vapor–liquid–solid (VLS) mechanism produce high-quality 1D CdS nanostructures at elevated temperatures.^14,16 However, these processes often require a catalyst, consume large amounts of energy and are not applicable for large scale production. As an alternate method, anodic aluminium oxide (AAO) films with tunable diameter and homogeneous distribution of pores are ideal templates for preparing well-aligned 1D CdS nanostructuresvia limited growth of the template.^17–21 The template-assisted methods, however, are commonly accompanied by the electrochemical deposition, leading to a series of undesirable consequences that the resulting products are usually in low crystallinity and they are mostly polycrystalline and moreover, there exist unavoidable damages to the products after removing the hard templates.

Compared to the methods mentioned above, solution chemical routes to ordered 1D CdS nanostructures are attractive for their observable advantages including low growth temperatures and good potential for scale up which is strongly required for practical applications.^22,23 In this respect, our lab reported the in situgrowth of well-aligned CdS nanowire arrays (NWs) on Cd foils, where the Cd foils served the double functions of growth substrate and Cd source.²⁴ Later, a series of reports have utilized this synthetic system for fabricating CdS NWs.^25–27 For instance, Wang et al. have adopted our methods to fabricate CdS NWs and also the relevant piezoelectric properties were characterized out of single CdS nanowire-based piezoelectric nanogenerators.²⁸ Together with the inherent merits of solution approach, the distinct advantage of this approach is that this method allows large number of NWs to grow directly on the electrode, thus avoiding the reconstruction of electrical contact to the NWs in nanodevice fabrication.^29,30 For example, if used in the field of solar cells, the cadmium base that supports the NWs will facilitate the electrical contact to collect the photogenerated charge carriers. However, most of the current investigations to date focus on the alignment of NWs, rather than on the synthetic system itself. Therefore, determining how the system works is still a necessary concern which will guide the rational synthesis of CdS NWs and thus further benefit their applications.

The microstructure control is another concern in nanomaterial research as the properties of the nanomaterials show close dependence on their crystal size, morphology, aspect ratio and orientation in comparison with their bulk counterparts owing to the low dimensionality and the resultant highly probable quantum size effect.^31–33 In general, there are two mainstream methods in this regard. One is the proper control of the synthetic conditions from the viewpoint of thermodynamics or kinetics.^34–40 The method has wide applications almost in all cases, but requires a deeper understanding of the synthetic systems. An alternative and more facile approach is to alter the characteristics of the substrate, including the crystal structure, orientation and the morphology. This method has been successfully demonstrated in a series of III–V or II–VI group semiconductor compounds, such as GaN, ZnO, etc.^41–43

On the basis of the preceding concerns, we focused on the preparation of novel CdS three-dimensional nanostructure arrays with quasi-textured surface where the Cd source, the caky Cd plating of micro-platelet (MP) arrays, was formed on conductive substrates through electroplating methods. With this unique structure, the synthetic system was well investigated and the nucleation and growth behavior of CdS is discussed. Together with the rational control of electroplating methods, CdS three-dimensional nanostructure arrays with various morphologies were obtained by altering the reaction conditions. Moreover, the electroplating technique was carefully analyzed in terms of electrochemical theory so that the electroplating technique may be extended to the synthesis of other metals with such unique plating morphology, which may pave the way for the fabrication of a new kind of nanostructure family.

Experimental

As electroplating techniques can be applied to various conductive substrates, herein we use copper foils for demonstration. The detailed experimental process is described as follows: first, a copper foil (with dimensions of 40 × 20 × 0.1 mm³, pretreated by sonication in dilute hydrochloric acid, acetone, and successively distilled water to remove oxide layer and any surface oil contaminant) was used as the working electrode for the deposition of Cd. The electrolytic solution consists of 0.23 mol L⁻¹CdSO₄, 0.24 mol L⁻¹(NH₄)₂SO₄, 0.04 mol L⁻¹Al₂(SO₄)₃ and a certain amount of some given additives. The electrodeposition of Cd was performed at ∼25 °C in a two-electrode configuration with a copper foil as the cathode and a Cd foil as the anode, under a constant cathode potential of −1.0 V for 30 min. After being washed sufficiently by deionized water to remove any surface residual ions, the Cd-coated copper foil was loaded into a Teflon-lined stainless steel autoclave (50 mL) containing ∼40 mL ethylenediamine solution and 0.2 mol L⁻¹thiosemicarbazide. Then, the autoclave was sealed and heated to 180 °C and subsequently kept at this temperature for two hours until the reaction was complete. After the autoclave was air cooled down, the foil covered with yellow products was taken out from the solution and washed with both deionized water and ethanol sufficiently, and successively dried under vacuum at 80 °C for 4 h before characterization.

Morphology and structure characterization of the products were carried out by using X-ray diffraction (XRD, D/max-γB), field-emission scanning electron microcopy (FESEM, SIRION 200, FEI), and high-resolution transmission electron microscopy (HRTEM, JEM-2010, JEOL) with selected-area electron diffraction (SAED). Composition analysis was performed using energy-dispersive X-ray spectroscopy (EDS, OXFORD, attached to the TEM).

Results and discussion

Morphologies and structure

The general morphologies of the as-prepared Cd plating are shown in Fig. 1(a). The electrodeposition was performed under a constant cathode potential of −1.0 V for 30 min in an alternating current (ac) mode. It was found that a layer of caky Cd plating composed of uniform Cd MP arrays with a diameter of about 20 μm was formed on the copper foil. Fig. 1(b) shows a high-resolution SEM image of the deposited Cd MP. Clear growth trace (as marked by the dotted lines) of the Cd MP shown in Fig. 1(b) suggests that the growth of the deposited Cd complies to spiral dislocation growth. In the experimental stage, the electroplating of Cd had been conducted on some other conductive substrates such as Mo, Ni, and ITO. No obvious difference in the morphology of Cd plating was observed in the same experimental conditions, demonstrating the universality of the electrodeposition growth of Cd plating on other conductive substrates. The TEM image of one slice of a Cd MP is shown in Fig. 1(e), the size of which is comparable to that shown in Fig. 1(a). The hexagon-shaped diffraction spots shown in the inset of Fig. 1(f) were obtained out of the Cd slice, confirming the single crystal characteristic of the Cd MP. In addition, the hexagonal diffraction spots can be indexed as the [001] zone axis of hexagonal Cd phase. In this case, the Cd MP would grow along the close packed plane of (002), leading to the exposed upper surface of the Cd MP of (002) with the lowest specific surface energy in hexagonal close packed structure. The high-resolution transmission electron microscopy (HRTEM) image taken from the slice of the Cd MP is shown in Fig. 1(f), which revealed the close packed Cd atoms, thereby further proving our conclusion. This growth behavior complies with the Curie–Wulff principle requiring the resultant crystal to have the minimum surface energy, which is applied to crystal growth behavior concerning the liquid–solid phase transition in equilibrium. The single-crystal characteristic of the Cd MP may also be understood in terms of the growth behavior of spiral dislocation. Spiral dislocation, also referred to as screw dislocation or helix dislocation, is classified as a sort of line defect. During the crystal growth stage, the presence of spiral dislocation will offer a trihedral nook benefiting the filling of new growth units, rendering the new growth units to stack in a layered mode. The macroscopical growth trace with step-shaped appearance as observed in Fig. 1(b), arises from the different growth rates which depend on the distance from the dislocation line. The detailed growth mechanism of the caky plating will be discussed later. Fig. 1(c) shows the morphologies of the resultant magnificent CdS three-dimensional nanostructure arrays in situ grown on the surface of the Cd MP arrays with reaction temperature and time of 180 °C and 2 h, respectively. Obviously, an outer layer of well-aligned CdS nanowire arrays was observed covering the surface of the Cd MP arrays and the as-obtained products show a quasi-textured surface. Compared to the Cd MP, the as-formed products exhibit an obvious drop in diameter but an increase in thickness. Understandably, the former could be ascribed to the dissolution of Cd while the latter arises from the vertical growth of CdS with respect to the Cd MP. A feature of the as-formed individual CdS three-dimensional nanostructure is shown in Fig. 1(d), which clearly shows that well-aligned CdS nanowire arrays grew perpendicularly to the Cd MP. Owing to the coalescent growth as discussed below, the diameters of CdS nanowires possess a wide distribution range from about 30 to 300 nm, whereas the lengths are relatively homogenous, about 1μm.

Fig. 1 FE-SEM images of the arrays of the electrodeposited Cd MP under a constant cathode potential of −1.0 V for 30 min: (a) low-resolution and (b) high-resolution. FE-SEM images of the products after reaction at 180 °C for 2 h with a molar concentration of 0.2 mol L⁻¹ of thiosemicarbazide: (c) the low-resolution and (d) the high-resolution. Also shown are the TEM image (e), the HRTEM image (f) and the relevant SAED patterns (inset) of one slice of the Cd MP.

The X-ray diffraction patterns of the caky Cd plating before and after the growth of CdS nanowire arrays are presented in Fig. 2. All diffraction patterns can be attributed to hexagonal wurtzite structure CdS (JCPDS No. 41-1049), while all other patterns arise from the deposited Cd (JCPDS No. 05-0674). Meanwhile, the strong and narrow (002) diffraction peak shows that CdS nanowires have a preferential growth direction along [001]. The appearance of strong Cd peaks signifies that there is still some Cd left after the reaction and thus an intermediate layer of Cd exists between the copper foil and the CdS which keeps the copper foil from erosion and therefore avoids the contamination of any Cu impurity, which agrees well with the preceding XRD patterns.

Fig. 2 XRD patterns of the caky Cd plating (the lower) and the resultant CdS three-dimensional nanostructure arrays (the upper).

Fig. 3(a) shows a transmission electron micrograph of the as-prepared CdS nanowire. The step-like top of the nanowire suggests that the nanowire is unnecessarily composed of a single nanowire, and several adjacent nanowires may grow together during the growth stage. Shown in Fig. 3(b) is the HRTEM image of the nanowires. The HRTEM image further confirms our assumption and shows the coalescent growth of CdS nanowires, which has been observed in nanostructured ZnO.^44,45 A feature of the coalesced boundary (as marked by a circle in Fig. 3(b)) was shown in Fig. 3(c), which reveals that the nanowires are parallel to each other and grow preferentially along [001]. The selected area electron diffraction (SAED) patterns shown in Fig. 3(d) suggest that the resulting CdS nanowire is not a perfect single crystal. The elongated diffraction spots along [001] arise from the existence of stack fault. The primary diffraction spots marked by solid lines can be attributed to the hexagonal wurtzite structure CdS whereas the other set of diffraction spots marked by dotted lines arise from the associated face-centered cubic structure cells caused by the changing of stacking order from the hexagonal structure of CdS. We may understand the coalescent mechanism in terms of the nucleation and successive growth of CdS nanowires. As discussed above, the exposed plane of the Cd MP was confirmed to be the closed packed plane of (002). The exposed (002) plane of Cd MP together with the almost identical growing conditions as a result of the small size of the Cd MP will benefit the uniform nucleation of CdS with the same orientation within the surface of one Cd MP. On the basis of the large lattice mismatch of up to 39% between the Cd substrate and the resultant CdS nanowires (the crystal parameter a of Cd and CdS are 0.2979 nm and 0.4141 nm, respectively), epitaxial growth of CdS nanowires from the Cd MP may not be favored. In this case, in the initial nucleation stage of CdS, S atoms may stack on the Cd MP in hexagonal close packed manner and Cd atoms will fill half of the resulting tetrahedral interstices formed by the hexagonal close packed S atoms, which can explain the growth direction of the nanowires along [001] and their vertical growth with respect to the growth substrate of the Cd MP. The successive coalescent growth behavior can be understood in terms of thermodynamics, as the surface energy (and thus the energy of the system) would show a drop via coalescent growth. A schematic diagram for the coalescent growth manner is shown in Fig. 3(e). Therefore, both hexagonal structure Cd MP and CdS nanowires will keep their [001] orientation parallel to each other. In addition, as shown in Fig. 3(f) the attached EDS quantitative analysis indicates that the atomic ratio of Cd to S is 49.32 : 50.68, close to the stoichiometry of CdS.

Fig. 3 TEM image (a) and HRTEM images (b) and (c) of single CdS NW prepared at 180 °C for 2 h with a molar concentration of thiosemicarbazide of 0.2 mol L⁻¹. (d) The corresponding SAED patterns indicating hexagonal structure with cubic stacking faults. (e) Schematic diagram for the coalescence growth of CdS NWs. (f) The corresponding EDS spectrum.

Nucleation and growth behavior of CdS

To investigate how the synthetic system works, a series of experiments were performed. First of all, the effect of reactant concentration was investigated by altering the amount of thiosemicarbazide. To assure the validity of the comparison, the copper foils used were pre-coated by Cd under the same electroplating conditions (a cathode potential of −1.0 V, an electroplating time of 30 min and a working mode of ac), which will guarantee the constant amount of Cd source. Together with Fig. 1(c), Fig. 4 shows the general morphological trend with the variable concentration of thiosemicarbazide. Under the lower concentration, i.e., 0.05 mol L⁻¹, micro-scale pillar arrays were obtained as shown in Fig. 4(a) and (b). The diameter of the pillars is about 400 nm as shown in the inset of Fig. 4(b), whereas the length is about 250 nm, corresponding to an aspect ratio of less than 1 : 1. As the concentration of thiosemicarbazide increases, the diameter exhibits an obvious decrease whereas the length shows a dramatic increase. When the concentration was finally increased to 0.3 mol L⁻¹, the diameter is about 100 nm and the concomitant diameter is about 1 μm, leading to an aspect ratio of about 10 : 1. In addition, though it is hard to accurately measure the density of CdS nanowires on the individual Cd platelet, still we can get the conclusion that the density also possesses a rising trend with the rising concentration of thiosemicarbazide as in all cases the nanowires grow densely and compactly. Thus far, many reports have demonstrated that the increased less reactive precursor often benefit the preferential growth of crystal with high aspect ratio.^36,46

Fig. 4 FE-SEM images of the arrays of CdS three-dimensional nanostructure prepared at 180 °C for 2 h but with different molar concentration of thiosemicarbazide: (a) 0.05 mol L⁻¹, (c) 0.1 mol L⁻¹, (e) 0.3 mol L⁻¹. FE-SEM images of (b), (d) and (f) are the magnified images corresponding to (a), (c) and (e), respectively.

The dramatic increase in both the density and aspect ratio of CdS nanowire arrays with increasing reactant concentration required a deeper understanding of the synthetic system. First of all, the evident difference in aspect ratio confirms that the growth of CdS is a typical kinetics controlled process, rather than a thermodynamics controlled process. Therefore, we may find our answer from the viewpoint of kinetics of crystal nucleation and growth. In general, the solution-based crystal growth contains two processes, i.e. nucleation and subsequent crystal growth. For the growth of CdS, heterogeneous nucleation is more energetically favored because of the declined nucleation barrier compared with that of homogeneous nucleation. As for the nucleation process, the nucleation velocity, I_V, can be expressed as follows: I_V = PD, where P is mainly determined by nucleation barrier and D is influenced by diffusion activation energy. On the other hand, supercooling and supersaturation are the two main impetuses driving the solution-based crystal growth. For our synthetic system, as the products are mainly formed during the reaction stage rather than in the cooling period, it is therefore strongly believed that the growth velocity is affected by the supersaturation which manifests itself in the form of diffusion. With increasing reactant concentration, factor D will be enhanced greatly which will accelerate the nucleation velocity. Meanwhile, the diffusion from solution to the surface of the products will be promoted compared to the inverted process, thus promoting the crystal growth to a large extent. In addition, it has been reported that high monomer concentration benefits the growth of the long axis, namely, the preferential growth.³¹ In the case of our synthetic system, a large quantity of reactant will correspondingly lead to high monomer concentration which will in turn help the preferential growth along the long axis. The combined effect of enhanced nucleation velocity and accelerated growth velocity contributes to CdS nanowire arrays with high density and increased aspect ratio. To support our assumption, however, we still need to rule out another possibility. If at low concentration of thiosemicarbazide crystal growth along the short axis is more favored, the surface of Cd platelet will be covered much quicker. The decreased exposed Cd surface will not only lead to decreased nucleation sites (thus restraining the heterogeneous nucleation velocity of CdS) but further decrease the concentration of Cd source owing to the weakened dissolution of Cd, which is unfavorable for the preferential growth along the long axis. In this case, CdS nanopillars with a small aspect ratio will be formed at low reactant concentration. To determine whether this possibility exists, control experiments were also performed for a short reaction time. Fig. 5(a) and (b) show the morphologies of the products synthesized at 180 °C with a reaction time of 0.5 h. We can see that at lower concentration, i.e. 0.05 mol L⁻¹, as shown in Fig. 5(a) only a small number of CdS nanowires were formed. In contrast, the number was greatly increased at a concentration of 0.3 mol L⁻¹, as shown in Fig. 5(b). The evident increase in the density of CdS nanowires demonstrates the great influence of reactant concentration on the nucleation velocity and further confirms our previous assumption.

Fig. 5 FE-SEM images of the products prepared at 180 °C for 0.5 h but with different concentration of thiosemicarbazide: (a) 0.05 mol L⁻¹, (b) 0.3 mol L⁻¹. FE-SEM images of the products prepared at 160 °C: (c) 2 h and (e) 4 h, and 200 °C: (d) 2 h and (f) 4 h.

Then, the effect of growth temperature on the products was also investigated. Together with Fig. 1(d), Fig. 5(c) and (d) show the morphologies of the products prepared at different temperatures but with an equal part of thiosemicarbazide. Although there is no obvious change in the aspect ratio of the obtained CdS NWs, the length of the NWs, however, increases evidently as the temperature arises. The lengths of the NWs synthesized at 160 °C, 180 °C and 200 °C were 0.2 μm, 1 μm and more than 2μm, respectively. The distinct increase in the length of the NWs could be attributed to the increased reactant concentration of Cd, resulting from quicker dissolution of Cd at elevated temperature. As temperature increases, the dissolution of Cd will be considerably enhanced. With the enhanced dissolution velocity, the Cd source will be greatly increased which will promote the rapid growth of CdS NWs in terms of diffusion kinetics. This can explain the difference of more than one order of magnitude in the length of the NWs. Although it is hard to testify the correctness of our assumption, we may get some valuable information from the synthetic reactions carried out for a longer period but at different temperatures. Fig. 5(e) and (f) reveal the morphologies of the products synthesized for 4 h but under different temperatures. It can be seen that the Cd MP dissolved completely at 200 °C as shown in Fig. 5(f), whereas some platelet-like structure still exists at 160 °C as shown in Fig. 5(e), indicating the accelerated dissolution of Cd at elevated temperatures. Moreover, the coalescent growth of CdS nanowires was more clearly revealed and bulk CdS crystals composed of numerous nanowires were formed for such long reaction periods.

Furthermore, the dependence of the alignment of CdS NWs on the surface characteristic of the Cd layer was also studied. In experimenting, it was found that the morphology of the substrate has great influence on the alignment of the NWs. When the rectifier works in a mode of direct current (dc), it is hard to find the trail of the spiral dislocation growth of Cd grain although caky Cd plating was also obtained. Compared to the spiral dislocation growth as-mentioned above, the variation in the surface morphology should arise from the characteristics of the employed current. In a dc mode, Cd²⁺ will be continuously reduced and deposited at the cathode. In an ac mode, however, the deposition and electrolysis of Cd will occur alternately at the cathode with the alternation of the current direction, which will weaken the growth rate of Cd to a large extent. Thus, the dc mode will offer a high deposition rate of Cd compared to that of the ac mode. The high deposition rate, however, may distort the ordered growth of Cd atoms in the form of spiral dislocation growth, whereas the decreased deposition rate at the ac mode may benefit the ordered growth. On this occasion, it is easy to understand the morphological differences of Cd platelets at different current characteristics. With the fast-grown caky Cd plating prepared in a dc mode, the solvothermal reaction was carried out also. Fig. 6(a) and (b) demonstrate the SEM images of the fast-grown caky Cd plating and the relevant products synthesized at 180 °C for 2 h with a thiosemicarbazide concentration of 0.2 mol L⁻¹. In contrast to the above-mentioned CdS NWs grown on the Cd platelet prepared in an ac mode, it is apparent that the alignment declined greatly and the orientation appears to be comparatively random. The difference in the alignment should arise from the only variable parameter, i.e. the morphological characteristics of the substrate. To further understand this morphology dependence, the synthetic reactions were also applied to plate-free Cd plating. Fig. 6(c) and (d) show the SEM images of the plate-free Cd plating and the relevant products, respectively. The platelet-free Cd plating was obtained under a cathode potential of −0.6 V for 30 min in an ac mode. Large Cd grains with a diameter of several microns and clear crystal boundary were observed from Fig. 6(c). After a reaction time of 1 h, a layer of CdS nanorods were grown on the surface of the substrate shown in Fig. 6(d). It is worth noting that the as-obtained CdS nanorods show good alignment only in small domains as shown in the circles, the size of which is comparable to that of the Cd grain marked by a dashed circle in Fig. 6(c). The small-domain alignment can be ascribed to the local uniform of the Cd surface. Significantly, it can be found that all nanorods grow perpendicularly to the substrate, which favors our understanding on the preceding difference in the alignment of the resulting nanowires. For the formation of CdS NWs, two typical processes occur in sequence. At first, heterogeneous nucleation of CdS nucleus will occur at the surface of the Cd surface. Then, preferential growth along the long axis leads to the formation of CdS NWs. The Cd MP prepared in an ac mode is favorable for the uniform nucleation and subsequent growth of the as-formed crystal nucleus with the same growth direction, thus resulting in highly ordered NWs. On the other hand, if the substrate is quite coarse at microscale, as the nanowires tend to grow perpendicularly to the substrate, the growth direction will be random and thereby less ordered NWs will be correspondingly formed owing to the emanant growth. When the reaction time was prolonged to 2 h, the alignment disappears completely and CdS NWs were randomly distributed on the substrate due to the formation of an interpenetrating network composed of CdS NWs from different domains.

Fig. 6 FE-SEM images of the fast-grown caky Cd plating under a dc mode before (a) and after solvothermal reaction (b) with reaction temperature, time and thiosemicarbazide concentration of 180 °C, 2 h, and 0.2 mol L⁻¹, respectively. FE-SEM images of the smooth Cd plating grown under an ac mode before (c) and after solvothermal reaction (d) with reaction temperature, time and thiosemicarbazide concentration of 180 °C, 1 h, and 0.2 mol L⁻¹, respectively.

Growth mechanism of caky Cd plating and controlled synthesis of CdS three-dimensional nanostructure arrays

While working on the electrodeposition synthesis of the caky plating, we had observed an increase in the density of the Cd MP with the applied cathode potential shifting to the more negative side and the caky structure appears only when the cathode potential is in a given range from ∼−0.8 V to ∼−1.0 V. If the cathode potential is too high, i.e. above −0.6 V, only a layer of Cd plating with a smooth surface will be obtained. By contrast, if the cathode potential surpasses −1.0 V, taking −1.2 V for example, a large amount of bubbles will be formed on the cathodal surface and subsequently precipitate out of the solution, leading to the formation of porous Cd plating accordingly. Thus by controlling the cathode potential, we should be able to fabricate CdS three-dimensional nanostructure arrays with controlled density. Fig. 1(c) and Fig. 7(a) and (b) show the density evolution of the Cd MP as a function of the applied cathode potential.

Fig. 7 FE-SEM images of the arrays of CdS three-dimensional nanostructure prepared at 180 °C for 2 h with a molar concentration of thiosemicarbazide of 0.2 mol L⁻¹ but with caky Cd plating substrates prepared at different cathode potentials: (a) −0.9 V, (b) −0.8 V. (c) The schematic diagram for the formation mechanism of caky Cd plating.

With the results discussed above, it is apparent that a high correlation exists between the formation mechanism of the caky plating and the applied cathode potential. In general, electroplating mainly involves two concurrent oxidation–reduction reactions, i.e.oxidation reaction at the anode and reduction reaction at the cathode. As for our electroplating system, two reduction reactions may occur simultaneously at the cathode as follows.

Cd²⁺ + 2e⁻ = Cd (1)

2H⁺ + 2e⁻ = H₂ (2)

As depicted by eqn (1), Cd²⁺ will be reduced and deposited on the cathode, forming a strong metallic bond with the metal substrate. On the other hand, together with the deposition of Cd, a side reaction, the precipitation of hydrogen, shown in eqn (2), may also occur which is not favorable for the deposition of Cd. Taking into account the normal electrode potential of Cd²⁺/Cd and H⁺/H₂ of −0.403 V and 0 V, respectively, we may easily draw the conclusion that the precipitation of hydrogen is more energetically favored and thus unavoidable. There is, however, no visual gas precipitating from the solution when the cathode potential is above −1.0 V. This can be explained from the viewpoint of the existence of the overpotential of H⁺/H₂. The H⁺/H₂ usually has a relatively large overpotential towards the negative side at Cd or Zn electrode whereas the overpotential of metals such as Cd is usually very small and could be neglected, rendering Cd²⁺ to be more easily reduced than H⁺. Meanwhile, only at a cathode potential of −1.2 V will the precipitation of obvious gas occur from the solution, which strongly proves the existence of the overpotential. On the basis of the above facts, we may find some clues for the formation mechanism of the caky plating. As the precipitation of hydrogen is a gradual process, there may exist the precipitation of hydrogen but the process may be relatively weak when the cathode potential ranges from −0.8 V to −1.0 V. Due to the weak precipitation of hydrogen, the precipitated hydrogen will form small bubbles loosely locating at the surface of the Cd platelet which will in turn lead to an asymmetric distribution of the concentration of Cd²⁺ at the surface of the cathode, namely, a low concentration at those bubble-covered domains and a high concentration at other areas. Owing to the spatial shielding effect of the hydrogen bubbles only the domains without bubbles can ensure the effective and continuous deposition of Cd. Moreover, it is worth noting that as discussed above the growth of Cd crystal complies with spiral dislocation growth which favors the growth of Cd grains with layered structure. Together with the spiral dislocation growth characteristic of Cd, hydrogen bubbles may serve as the template, just as the AAO template, for the formation of the caky plating due to their spatial block effect. A schematic growth mechanism for Cd MP arrays is shown in Fig. 7. It should be noted that the real case might be more complicated as the surface of the substrate is not uniform at microscale. As a result, hydrogen bubbles are more likely to be formed at the raised surface of the substrate because of the point discharge, resulting in an asymmetric distribution of the bubbles. This may explain the declined structural regularity of the Cd MP arrays in real case.

Conclusions

In summary, novel and unique CdS three-dimensional nanostructure arrays have been successfully synthesized on the Cd-coated copper foils viasolvothermal reaction combined with the introduction of electroplating. The results demonstrated that the growth of CdS nanowire/pillars in the synthetic system is a typical kinetics-controlled process and the nucleation and growth of CdS follow conventional diffusion kinetics. The aspect ratio and density of the nanowires can be tuned by altering the reactant concentration and reaction temperature. The alignment of the nanowires depends greatly on the morphological characteristics of the growth substrate. We believe our work should provide an insightful view into the synthetic system which will guide the rational synthesis of CdS nanowire arrays. Furthermore, for the first time the electroplating technique was introduced to obtain the plating layer and tune the morphological feature of the substrate, thus tuning the microstructure of the products. The electroplating methodology was well discussed and a model based on both the electrochemical theory and the experimental results was proposed to understand the formation mechanism of the caky plating structure, which favors the extension of the electroplating technique to the preparation of some other metals with such unique structure. Consequently, the synthesis strategy may shed new light on the synthesis of a new nanomaterial family with such novel nanostructure.

Acknowledgements

The work was financially supported by the National High Technology Research and Development Program of China (No. 2007AA03Z301), the Natural Science Foundations of China (No. 20771032, No. 60806028) and Anhui Province (070414200), and the National Key Basic Research Program of China (No. 2007CB936001 and No. 2009CB623703).

References

1. Y. Xia, P. Yang, Y. Sun, Y. Wu, B. Mayers, B. Gates, Y. Yin, F. Kim and H. Yan, Adv. Mater., 2003, 15, 353.
2. Y. Tak, S. Hong, J. Lee and K. Yong, J. Mater. Chem., 2009, 19, 5945.
3. Y. Liu, J. Mao, P. Jiang, Z. Xu, Hongyan Yuan and D. Xiao, CrystEngComm, 2009, 11, 2285.
4. J. Chen, B. J. Wiley and Y. Xia, Langmuir, 2007, 23, 4120.
5. Y. Yoon, K. Park, J. Heo, J. Park, S. Nahm and K. Choi, J. Mater. Chem., 2010, 20, 2386.
6. D. Wu, Z. Huang, G. Yin, Y. Yao, X. Liao, D. Han, X. Huang and J. Gu, CrystEngComm, 2010, 12, 192.
7. H. Unalan, P. Hiralal, D. Kuo, B. Parekh, G. Amaratunga and M. Chhowalla, J. Mater. Chem., 2008, 18, 5909.
8. Y. Jiao, H. J. Zhu, X. F. Wang, L. Shi, Y. Liu, L. M. Peng and Q. Li, CrystEngComm, 2010, 12, 940.
9. W. Lu and C. M. Lieber, J. Phys. D: Appl. Phys., 2006, 39, R387.
10. H. Lee, H. C. Leventis, S.-J. Moon, P. Chen, S. Ito, S. A. Haque, T. Torres, F. Nüesch, T. Geiger, S. M. Zakeeruddin, M. Grätzel and M. K. Nazeeruddin, Adv. Funct. Mater., 2009, 19, 2735.
11. P. Wu, R. Ma, C. Liu, T. Sun, Y. Ye and L. Dai, J. Mater. Chem., 2009, 19, 2125.
12. A. B. Greytak, C. J. Barrelet, Y. Li and C. M. Lieber, Appl. Phys. Lett., 2005, 87, 151103.
13. S. K. Mandal, A. B. Maity, J. Dutta, R. Pal, S. Chaudhuri and A. K. Pal, Phys. Status Solidi A, 1997, 163, 433.
14. A. Pan, R. Liu, Q. Yang, Y. Zhu, G. Yang, B. Zou and K. Chen, J. Phys. Chem. B, 2005, 109, 24268.
15. Z. Fan, H. Razavi, J.-w. Do, Aimee Moriwaki, O. Ergen, Y.-L. Chueh, P. Leu, J. C. Ho, T. Takahashi, L. A. Reichertz, S. Neale, K. Yu, Ming Wu, J. Ager and A. Javey, Nat. Mater., 2009, 8, 648.
16. H. Pan, C. K. Poh, Y. Zhu, G. Xing, K. C. Chin, Y. P. Feng, J. Lin, C. H. Sow, W. Ji and A. T. S. Wee, J. Phys. Chem. C, 2008, 112, 11227.
17. G. Cao and D. Liu, Adv. Colloid Interface Sci., 2008, 136, 45.
18. W. Yang, Z. Wu, Z. Lu, X. Yang and L. Song, Microelectron. Eng., 2006, 83, 1971.
19. D. Routkevitch, T. Bigioni, M. Moskovits and J. M. Xu, J. Phys. Chem., 1996, 100, 14037.
20. Y. Liang, C. Zhen, D. Zou and D. Xu, J. Am. Chem. Soc., 2004, 126, 16338.
21. Y. Li, D. Xu, Q. Zhang, D. Chen, F. Huang, Y. Xu, G. Guo and Z. Gu, Chem. Mater., 1999, 11, 3433.
22. G. Hua, Y. Tian, L. Yin and L. Zhang, Cryst. Growth Des., 2009, 9, 4653.
23. X. Feng, K. Shankar, O. K. Varghese, M. Paulose, T. J. Latempa and C. A. Grimes, Nano Lett., 2008, 8, 3781.
24. B. L. Cao, Y. Jiang, C. Wang, W. H. Wang, L. Z. Wang, M. Niu, W. J. Zhang, Y. Q. Li and S. T. Lee, Adv. Funct. Mater., 2007, 17, 1501.
25. F. Chen, R. Zhou, L. Yang, M. Shi, G. Wu, M. Wang and H. Chen, J. Phys. Chem. C, 2008, 112, 13457.
26. A. Datta, P. G. Chavan, F. J. Sheini, M. A. More, D. S. Joag and A. Patra, Cryst. Growth Des., 2009, 9, 4157.
27. S. Biswas, S. Kar, S. Santra, Y. Jompol, M. Arif and S. I. Khondaker, J. Phys. Chem. C, 2009, 113, 3617.
28. Y. F. Lin, J. H. Song, Y. Ding, S. Y. Liu and Z. L. Wang, Appl. Phys. Lett., 2008, 92, 022105.
29. C. J. Novotny, E. T. Yu and P. K. L. Yu, Nano Lett., 2008, 8, 775.
30. W. Zhang and S. Yang, Acc. Chem. Res., 2009, 42, 1617.
31. V. H. Grassian, J. Phys. Chem. C, 2008, 112, 18303.
32. T. Zhai, X. Fang, Y. Bando, Q. Liao, X. Xu, H. Zeng, Y. Ma, J. Yao and D. Golberg, ACS Nano, 2009, 3, 949.
33. W. Shi, R. W. Hughes, S. J. Denholme and D. H. Gregory, CrystEngComm, 2010, 12, 641.
34. A. Bhatt, A. Mechler, L. Martin and A. Bond, J. Mater. Chem., 2007, 17, 2241.
35. H. Li, D. Chen, L. Li, F. Tang, L. Zhang and J. Ren, CrystEngComm, 2010, 12, 1127.
36. Z. A. Peng and X. Peng, J. Am. Chem. Soc., 2001, 123, 1389.
37. P. Christian and P. O'Brien, J. Mater. Chem., 2008, 18, 1689.
38. D. Moore and Z. Wang, J. Mater. Chem., 2006, 16, 3898.
39. J. Zhang, J. Shi, J. Tan, X. Wang and M. Gong, CrystEngComm, 2010, 12, 1079.
40. J. J. Carvajal and J. C. Rojo, Cryst. Growth Des., 2009, 9, 320.
41. S. D. Hersee, X. Sun and X. Wang, Nano Lett., 2006, 6, 1808.
42. Z. Gu, M. P. Paranthaman, J. Xu and Z. W. Pan, ACS Nano, 2009, 3, 273.
43. Z. L. Wang, ACS Nano, 2008, 2, 1987.
44. C.-C. Lin, S.-Y. Chen and S.-Y. Cheng, J. Cryst. Growth, 2005, 283, 141.
45. T.-Y. Liu, H.-C. Liao, C.-C. Lin, S.-H. Hu and S.-Y. Chen, Langmuir, 2006, 22, 5804.
46. L. Shi, Y. Xu and Q. Li, J. Phys. Chem. C, 2009, 113, 1795.

---