Morphological evolution of ZnO nanorod arrays induced by a pH-buffering effect during electrochemical deposition

Abstract

We report a new parameter of ‘pH-buffering action’ in electrodeposition of ZnO nanorod arrays from dilute Zn(NO₃)₂ solutions; the buffering action improves deposition efficiency and changes the structural morphology of the resulting ZnO nanorods.

---

Zinc oxide (ZnO) is a unique oxide semiconductor with a wide band gap energy of 3.3 eV, a large exciton binding energy of 59 meV and high electron mobility, thereby making transparent n-type conductivity and room-temperature UV emission possible. Since the development of a method for cathodic electrodeposition of ZnO in 1996 by Izaki et al.¹ and Lincot et al.,² extensive efforts have been devoted to the preparation of morphology-controlled ZnO films. By modifying electrodeposition parameters such as the Zn²⁺ precursor concentration, applied potential/current, bath temperature and the nature of additives used, various ZnO structures, including nanorods (nanowires, nanopillars),³ nanocauliflowers,⁴ nanoplates⁵ and epitaxial films,⁶ have been deposited onto conductive substrates in mild aqueous solutions (60–90 °C) under ambient atmosphere. Among them, an array of nanorods is one of the most important structure because of its ability to serve as fundamental 1D architecture with applications in solar cells,⁷ light-emitting diodes⁸ and chemical sensors.⁹ Regarding electrodeposition of these 1D ZnO arrays, most previous research has been mainly focused on the effect of Zn²⁺ concentration, [Zn²⁺], on the morphology of the resulting arrays, revealing that the use of very dilute Zn(NO₃)₂ or ZnCl₂ aqueous solutions, [Zn²⁺] ≤ 0.5 mM (M = mol L⁻¹) is required to avoid coalescence between nanorods during deposition.^3a–f However, such low precursor concentrations limit the efficiency of deposition and/or the controllability of the nanorod diameter (see Fig. S1 in ESI†). Here, we report a new strategy that involves a pH-buffering effect to resolve both these problems.

A plausible mechanism for ZnO electrodeposition has been proposed as follows:¹⁰

O₂ + 2H₂O + 4e⁻ → 4OH⁻ (1-1)

NO₃⁻ + H₂O + 2e⁻ → NO₂⁻ + 2OH⁻ (1-2)

Zn²⁺ + 2OH⁻ → Zn(OH)₂ (2)

Zn(OH)₂ → ZnO + H₂O (3)

The cathodic current applied on a substrate reduces dissolved O₂ and/or NO₃⁻ to OH⁻, thereby causing a local pH increase in the vicinity of the substrate; this process is followed by the deposition of Zn(OH)₂ and dehydration to form crystalline ZnO. According to the described mechanism, the applied current determines the generation rate of OH⁻, whereas the formation rates of Zn(OH)₂ and ZnO are affected by the concentrations of OH⁻ and Zn²⁺ and the bath temperature, respectively. Thus, an increase in pH (i.e. generation of OH⁻ on the substrate) is a key factor in the ZnO electrodeposition, and an appropriate pH-buffering agent is expected to decisively influence the manner of ZnO growth.

To demonstrate the effect of pH buffering on the morphology of ZnO nanorods electrodeposited from dilute [Zn²⁺], we conducted two-electrode galvanostatic electrodeposition from 0.5 mM Zn(NO₃)₂–0.1 M NaNO₃-based aqueous solutions (150 mL) using an F-doped SnO₂-coated glass (FTO, deposition area of 1.5 cm², Asahi Glass, ∼10 Ω sq⁻¹) substrate. Galvanostatic mode was chosen to maintain a constant OH⁻ generation rate and avoid contamination from the reference electrode during electrolysis. The current density was determined from our previous linear sweep voltammetry measurements to be 0.2 mA cm⁻², which is almost minimal within a practicable range.¹¹ The added NaNO₃ serves two functions: as a source of sufficient NO₃⁻ and as a supporting electrolyte.^3e,11 A high-purity Zn bar was used as a counter electrode to compensate the [Zn²⁺] consumed during the deposition of ZnO. NH₄NO₃ was used as a pH-buffering agent because the pH of the as-prepared 0.5 mM Zn(NO₃)₂–0.1 M NaNO₃ solution was 5.7 and the theoretical pH boundary between 0.5 mM Zn²⁺ and Zn(OH)₂ at 25 °C is ∼7.6 according to the corresponding Pourbaix diagram (see Fig. S2 in ESI†),¹² indicating that pH-buffering action is required in the weak-acid region. Because NH₄NO₃ is the salt of a strong acid and weak base, NH₄NO₃ solutions are weakly acidic (pH = 5–6) and NH₄⁺ can act as a OH⁻ scavenger:

NH₄⁺ + OH⁻ ↔ NH₃ + H₂O, pK_a = 9.3. (4)

To verify this behaviour, we performed pH titrations on three different solutions whose initial pH levels were adjusted to pH = 3.0 with HNO₃: (i) 0.1 M NaNO₃, (ii) 0.5 mM Zn(NO₃)₂–0.1 M NaNO₃ and (iii) 10 mM NH₄NO₃–0.5 mM Zn(NO₃)₂–0.1 M NaNO₃. The solutions were titrated with 10 mM NaOH solution. The titration of the simple NaNO₃ solution (i) resulted in a typical pH titration curve (Fig. 1a), whereas that of the solution (ii) containing Zn(NO₃)₂ resulted in a curve with a transient plateau at a pH range of 7.5–8.2 (Fig. 1b), corresponding to the formation of Zn(OH)₂ (eqn (2)). This result agrees with the pH value (∼7.6) estimated from the Pourbaix diagram. In contrast, the curve for the solution (iii) containing NH₄NO₃ exhibited a continuous plateau (Fig. 1c), and the curve shifted slightly to the right compared with that for the solution without NH₄NO₃ (Fig. 1 inset). This result indicates that the reaction of OH⁻ with NH₄⁺ (eqn (4)) occurred before the reaction with Zn²⁺ and that the pH gradually reached ∼7.5, i.e. the onset of Zn(OH)₂ formation. Notably, because the complexation of NH₄⁺ with Zn²⁺ to form [Zn(NH₃)₄]²⁺ (log K = 9.65)¹³ is thermodynamically unfavourable in such dilute solutions, the theoretical pH boundary between Zn²⁺ and Zn(OH)₂ remains ∼7.6. Thus, NH₄NO₃ was proven to be an effective pH-buffering agent in the ZnO electrodeposition solution.

Fig. 1 pH titration curves of (a) 0.1 M NaNO₃, (b) 0.5 mM Zn(NO₃)₂–0.1 M NaNO₃ and (c) 10 mM NH₄NO₃–0.5 mM Zn(NO₃)₂–0.1 M NaNO₃ aqueous solutions. All solutions were initially adjusted to pH = 3.0 with HNO₃ and subsequently titrated with 10 mM NaOH at room temperature. The inset presents an expanded view of the onset of the plateau in curves (b) and (c).

Fig. 2 shows field-emission scanning electron microscopy (FESEM) images of ZnO electrodeposited on a bare FTO substrate from as-prepared 0.5 mM Zn(NO₃)₂–0.1 M NaNO₃ aqueous solutions (75 °C) with varying concentrations (0–20 mM) of NH₄NO₃ while the other electrodeposition conditions were fixed. For all the samples, well-faceted columnar ZnO nanorods were observed and the samples' X-ray diffraction patterns indicated a hexagonal wurtzite structure (see Fig. S3 in ESI†). The morphology, including the diameter, length and growth density, of the obtained ZnO nanorods varied considerably depending on [NH₄NO₃]. The variations of the morphology are plotted as a function of the added NH₄NO₃ in Fig. 3. The diameter and length of nanorods increased from 0.14 to 0.88 μm and from 1.0 to 4.3 μm, respectively, and the growth density per unit substrate area decreased from 16.8 to 0.3 μm⁻² with increasing [NH₄NO₃]. We calculated the volume of ZnO deposited per unit substrate area under the assumption that the nanorods are cylindrical; the results are shown in Fig. 3d, where the current efficiency was also estimated by Faraday's laws of electrolysis.¹⁴ The current efficiency was increased from 17 to 60–70% by the addition of NH₄NO₃.

Fig. 2 Cross-sectional (left) and top-view (right) FESEM images of ZnO nanorod arrays electrodeposited on FTO substrates at a current density of 0.2 mA cm⁻² with a total electric charge of 2.0 C cm⁻² from 0.5 mM Zn(NO₃)₂–0.1 M NaNO₃ aqueous solutions at 75 °C with and without NH₄NO₃: (a and b) 0, (c and d) 1.0, (e and f) 5.0, (g and h) 10 and (i and j) 20 mM. All scale bars are 1 μm.

Fig. 3 Morphological variation of ZnO nanorod arrays as a function of NH₄NO₃ concentration. Data in (a–c) are statistical averages evaluated from FESEM images including Fig. 2; the results in (d) were calculated from these averages.

Thus, the effects of NH₄NO₃ on the ZnO growth manner include the following two points: (i) improvement of the current efficiency and (ii) change in the nanorod morphology. Although the former exhibits little dependence on [NH₄NO₃] while the latter depends on it strongly, these results can be explained consistently by the pH-buffering action (eqn (4)). As previously mentioned, OH⁻ is generated on the substrate at a constant rate defined by the applied current density (0.2 mA cm⁻²) and the total amount of OH⁻ is defined by the electric charge (2.0 C cm⁻²). If [Zn²⁺] is sufficiently high and diffusion of Zn²⁺ to the substrate is equivalent to or greater than the OH⁻ generation rate, the current efficiency should be nearly 100%. In fact, the current efficiency was ∼87% when ZnO was electrodeposited from a 50 mM Zn(NO₃)₂–0.1 M NaNO₃ solution (see Fig. S1 in ESI†). The observed low current efficiency of 17% for the 0.5 mM Zn²⁺ solution without NH₄NO₃ indicates that diffusion of Zn²⁺ to the substrate is the rate-determining step and that surplus (unreacted) OH⁻ is generated on the substrate. The surplus OH⁻ will diffuse to the bulk of the solution and induce undesirable precipitation of ZnO some distance from the substrate, thereby preventing the supply of Zn²⁺ to the substrate during electrodeposition. The OH⁻ scavenger, NH₄⁺, can suppress the diffusion of OH⁻ and assist the smooth supply of Zn²⁺, resulting in a current efficiency 3.6–4.2 times greater. In fact, the solution after ZnO electrodeposition without NH₄NO₃ exhibited a slight white turbidity due to the precipitated ZnO, whereas that with NH₄NO₃ remained clear.

Thus, the threshold of this effect depends not on the NH₄NO₃ concentration, but on the amount of NH₄⁺ in a solution relative to the amount of surplus OH⁻ generated through the electrodeposition. The precise amount of surplus OH⁻ is unclear; however, under the present electrodeposition conditions, the amount of NH₄⁺ in 150 mL of 1.0 mM NH₄NO₃ solution is sufficiently large (4.8 equiv.) compared with the total OH⁻ generated when an electric charge of 3.0 C was passed (total charge density = 2.0 C cm⁻², deposition area = 1.5 cm²). However, because the presence of NH₄NO₃ can suppress the diffusion of surplus OH⁻ but cannot prevent generation of surplus OH⁻ itself (the surplus OH⁻ is attributed to a Zn²⁺ diffusion-limited condition), the current efficiency is lower by ∼20% than that observed in the case of a sufficiently high [Zn²⁺].

Regarding the second point concerning (ii) a change in the nanorod morphology, this change is due to the effect of pH-buffering on the ZnO nucleation step. Because ZnO nanorods grow exclusively in the longitudinal direction (vertical to the substrate) in the case of a Zn²⁺ diffusion-limited condition, the nanorod diameter remains almost constant during deposition.^3a,d,e,15 Thus, the resulting diameter and growth density are defined predominantly at the initial nucleation step. According to the results of previous studies, lower nucleation rates tend to result in larger and sparser ZnO grains, and vice versa.^3e,16 Because the pH-buffering action can delay the initial nucleation rate, the diameter and growth density increased and decreased, respectively, with increasing [NH₄NO₃] (Fig. 3a and c). After nucleation, the nanorods grew to the volume defined by the total electric charge and current efficiency, determining the length of the nanorods. In fact, when the electric charge was increased from 2.0 to 3.4 C cm⁻², the length of ZnO increased from ∼2.0 to ∼3.1 μm, along with a slight increase in diameter, as shown in Fig. 4.

Fig. 4 FESEM image of ZnO nanorod arrays electrodeposited on an FTO substrates at a current density of 0.2 mA cm⁻² with a total electric charge of 3.4 C cm⁻² from a 5.0 mM NH₄NO₃–0.5 mM Zn(NO₃)₂–0.1 M NaNO₃ aqueous solution at 75 °C. Scale bar is 1 μm.

The increase in the pH-buffering action with increasing [NH₄NO₃] indicates that the amount of OH⁻ that reacts with NH₄⁺ before the formation of ZnO increases; i.e. the amount of OH⁻ that can react with Zn²⁺ decreases with increasing [NH₄NO₃]. This action is reflected in Fig. 3d, where the ZnO volume (current efficiency) decreases gradually at NH₄NO₃ concentrations greater than 5.0 mM. Furthermore, in the presence of excessive NH₄NO₃ (50 mM), no deposition was observed because of a strong pH-buffering action (its titration curve is presented in Fig. S4 in ESI†), where pH in the vicinity of the substrate will not reach the ZnO-nucleation level.

In summary, we demonstrated that NH₄NO₃ acts as a pH-buffering agent in the electrodeposition of ZnO from dilute Zn(NO₃)₂ solutions, thus improving the current efficiency and morphological controllability of the ZnO nanorod arrays. This study offers a new parameter of ‘pH-buffering action’ in addition to known parameters, such as Zn²⁺ concentration and applied potential/current, for readily preparing separated ZnO nanorods vertically aligned on FTO substrates.

Acknowledgements

This work was supported by a grant-in-aid from the Ministry of Education, Science, Sports and Technology (MEXT), Japan.

Notes and references

1. M. Izaki and T. Omi, Appl. Phys. Lett., 1996, 68, 2439.
2. S. Peulon and D. Lincot, Adv. Mater., 1996, 8, 166–170.
3. (a) J. Elias, R. Tena-Zaera and C. Lévy-Clément, J. Electroanal. Chem., 2008, 621, 171–177; (b) R. Tena-Zaera, J. Elias, G. Wang and C. Lévy-Clément, J. Phys. Chem. C, 2007, 111, 16706–16711; (c) O. Lupan, T. Pauporté, B. Viana, I. Tiginyanu, V. Ursaki and R. Cortés, ACS Appl. Mater. Interfaces, 2010, 2, 2083–2090; (d) T. Pauporté, G. Bataille, L. Joulaud and F. J. Vermersch, J. Phys. Chem. C, 2009, 114, 194–202; (e) M. R. Khajavi, D. J. Blackwood, G. Cabanero and R. Tena-Zaera, Electrochim. Acta, 2012, 69, 181–189; (f) S.-J. Lee, S. K. Park, C. R. Park, J. Y. Lee, J. Park and Y. R. Do, J. Phys. Chem. C, 2007, 111, 11793–11801; (g) M. Izaki, T. Shinagawa and H. Takahashi, J. Phys. D: Appl. Phys., 2006, 39, 1481–1484.
4. (a) M. Watanabe, H. Aritomo, I. Yamaguchi, T. Shinagawa, T. Tamai, A. Tasaka and M. Izaki, Chem. Lett., 2007, 36, 680–681; (b) M. Izaki, M. Watanabe, H. Aritomo, I. Yamaguchi, S. Asahina, T. Shinagawa, M. Chigane, M. Inaba and A. Tasaka, Cryst. Growth Des., 2008, 8, 1418–1421.
5. D. Pradhan and K. T. Leung, Langmuir, 2008, 24, 9707–9716.
6. (a) R. Liu, A. A. Vertegel, E. W. Bohannan, T. A. Sorenson and J. A. Switzer, Chem. Mater., 2001, 13, 508–512; (b) T. Pauporté and D. Lincot, Appl. Phys. Lett., 1999, 75, 3817; (c) M. Izaki, S. Watase and H. Takahashi, Adv. Mater., 2003, 15, 2000–2002; (d) M. Izaki, S. Watase and H. Takahashi, Appl. Phys. Lett., 2003, 83, 4930.
7. (a) J. Ajuria, I. Etxebarria, E. Azaceta, R. Tena-Zaera, N. Fernandez-Montcada, E. Palomares and R. Pacios, Phys. Chem. Chem. Phys., 2011, 13, 20871–20876; (b) K. P. Musselman, A. Wisnet, D. C. Iza, H. C. Hesse, C. Scheu, J. L. MacManus-Driscoll and L. Schmidt-Mende, Adv. Mater., 2010, 22, E254–E258; (c) R. Könenkamp, A. Nadarajah and R. C. Word, Appl. Phys. Lett., 2009, 95, 053103; (d) T. Le Bahers, F. Labat, T. Pauporté, P. P. Lain and I. Ciofini, J. Am. Chem. Soc., 2011.
8. (a) A. Nadarajah, C. Robert, J. Meiss and R. Könenkamp, Nano Lett., 2008, 8, 534–537; (b) R. Könenkamp, R. Word and M. Godinez, Nano Lett., 2005, 5, 2005–2008; (c) O. Lupan, T. Pauporté, T. Le Bahers, B. Viana and I. Ciofini, Adv. Funct. Mater., 2011; (d) O. Lupan, T. Pauporté and B. Viana, Adv. Mater., 2010, 22, 3298–3302.
9. (a) D. Pradhan, F. Niroui and K. Leung, ACS Appl. Mater. Interfaces, 2010, 2, 2409–2412; (b) Y.-J. Li, K.-M. Li, C.-Y. Wang, C.-I. Kuo and L.-J. Chen, Sens. Actuators, B, 2012, 161, 734–739.
10. (a) A. Goux, T. Pauporté, J. Chivot and D. Lincot, Electrochim. Acta, 2005, 50, 2239–2248; (b) S. Peulon and D. Lincot, J. Electrochem. Soc., 1998, 145, 864; (c) M. Izaki and T. Omi, J. Electrochem. Soc., 1996, 143, 53.
11. T. Shinagawa, M. Chigane, K. Murase and M. Izaki, J. Phys. Chem. C, 2012, 116, 15925–15931.
12. M. Pourbaix, Atlas of electrochemical equilibria in aqueous solutions, NACE, Houston, TX, 1974.
13. R. Smith and A. Martell, Critical Stability Constants Vol. 4, Plenum, New York, 1976.
14. Current efficiency is defined by n(obtained)/n(theoretical), where n (mol) is the amount of the substance. The value of n(obtained) was calculated from Vd/M, where V (cm³) is volume, d (g cm⁻³) is density and M (g mol⁻¹) is the formula weight for the object substance. The value of n (theoretical) was calculated according to Faraday's laws of electrolysis: n = C/(zF), where C (C) is electric charge, z is the number of electrons involved in the reaction and F (C mol⁻¹) is the Faraday constant.
15. T. Shinagawa, S. Watase and M. Izaki, Cryst. Growth Des., 2011, 11, 5533–5539.
16. (a) J. Elias, R. Tena-Zaera and C. Lévy-Clément, Thin Solid Films, 2007, 515, 8553–8557; (b) T. Shinagawa, K. Shibata, O. Shimomura, M. Chigane, R. Nomura and M. Izaki, J. Mater. Chem. C, 2014, 2, 2908–2917.

---

Footnote

† Electronic supplementary information (ESI) available: Experimental details, FESEM images, potential-pH diagram, XRD patterns and pH titration curves. See DOI: 10.1039/c4ra04342a

---