Cyclic Voltammetric and Chronoamperometric Deposition of Cds

Cyclic Voltammetric and Chronoamperometric Deposition of CdS

Yongkuk Kim, Jaegoo Jung, Seunghun Kim and Won-Seok Chae+

Department of Chemistry, Daejin University, Pocheon, 487-711 Gyounggi, Korea

CdS was electrochemically deposited on ITO using cyclic voltammetry (CV) and chronoamperometry (CA) in aqueous solution. To fabricate the CdS thin films, a potential ranging from ¹0.3 to ¹1.2 V was applied for CV and a constant potential of ¹1.1 V was applied for CA (Ag«AgCl). The deposited CdS thin films were characterized in detail using XRD, AFM and SEM. UV/Vis-transmittance was used to determine the band-gap energy (Eg) of the films. The dopant densities and flat band-potentials (Efb) were determined using Mott-Schottky plots. The CdS film obtained by CV deposition was observed to be rougher and sparser than that obtained by CA deposition. This was attributed to the alternating reductionoxidation, corresponding to depositiondissolution of CdS. The carrier density of the CdS film deposited by CV was measured to be as much as 17%. [doi:10.2320/matertrans.M2013125]

(Received March 28, 2013; Accepted May 10, 2013; Published June 21, 2013) Keywords: cadmium sulﬁde, electrodeposition, cyclic voltammetry, chronoamperometry

2+ 1. Introduction Na2S2O3·5H2O (99% Shinyo, Japan), as a source of Cd and S2¹ respectively, and the pH of solutions was 7.4. Thin films of CdS have been used as a window1,2) or a Transparent conducting glass (ITO, Samsung Corning, sensitizer3) for solar cells for many years. However, the use Korea) and Pt mesh (99.99% Sigma-Aldrich) were used as of thin CdS is often accompanied by reduced quantum the working electrode (deposition substrate) and the counter efficiency of the cell, which is attributed to pinholes or other electrode, respectively. A reference electrode, Ag«AgCl defects. The interesting of a thin CdS film is therefore (saturated KCl) was encased in a bridge tube connected to required in order to achieve high-efficiency solar cell. the solution by a Vycor tip. All the potentials were measured Physical and chemical deposition processes have gained against the reference electrode. For the CV, the potential scan increasing interest for this purpose. These include a wide was started at ¹0.3 V (Ag«AgCl), reversed at ¹1.2 V, and variety of methods such as sputtering,4) chemical bath then terminated at ¹0.3 V, with 20 cycles of repetition at a deposition,5) chemical vapor deposition6,7) or closed space scan rate of 0.1 V/s. For the CA, a constant potential of sublimation8) that requires a low vacuum and high temper- ¹1.1 V was applied for 900 s. ature.9) Electrodeposition, an extremely convenient method Electrochemical measurements and impedance studies for for the production of chalcogenide semiconductors, has been Mott-Schottky analysis were carried out using an EG&G used to fabricate CdS thin films.1014) This approach offers PARC M 283 potentiostat and FRA 1053 frequency analyzer several advantages over conventional methods, including low (Ametek, TN, USA). Impedance spectra and photocurrent cost, large surface area coverage, low-temperature growth and measurements of CdS electrodeposited were obtained in direct control of film thickness, morphology and composition. 0.10 mol/L LiClO4 as an electrolyte, with the CdS film on Various electrodeposition-methods have been reported ITO used as the working electrode, after being thoroughly for the preparation of CdS films from nonaqueous1517) or rinsed with distilled water to remove any residual precursors. aqueous solutions.18,19) The most commonly used approach The light source used for measuring the photocurrent was a involves cathodic reduction to form a thin film from an 150 W W-lamp (KLS-150, Dongwoo Optron, Korea). X-ray aqueous solution of Cd2+ ion along with a source of S. The diffraction (XRD) patterns were measured with a Rigaku electrodeposition of such films on conductive substrates is X-ray diffractometer (Tokyo, Japan) using a Cu K¡ (40 kV, well suited for the manufacture of film solar cells. In this 30 mA) radiation. The 2ª range from 20 to 70° was scanned paper, we used a three-electrode cell to compare the films at 3° min¹1. The surface morphology of the films was of CdS electrodeposited by cyclic voltammetry (CV) and observed using an atomic force microscope (AFM; Park chronoamperometry (CA) onto In-doped SnO2 (ITO) coated Systems XE-120, Korea) and a scanning electron microscope glass substrates from aqueous solution. The films were (SEM; Philips XL-30 ESEM, USA). Optical studies were structurally, optically and electrochemically characterized, performed on the films deposited on ITO substrates using and the photoelectrochemical behavior was assessed using a Hewlett-Packard 8453 UVVis spectrophotometer (Palo current-potential measurements under dark and illuminated Alto, CA, USA). conditions. 3. Results and Discussions 2. Experimental Details The electrodeposition of CdS was achieved using both CV The electrochemical deposition of CdS was carried out by and CA, as shown in Figs. 1 and 2, respectively. For the CV, ¹1 CV and CA from a solution containing 0.010 mol·L CdSO4 the well-defined cathodic peak seen in the 1st cycle is the (99.0% Sigma-Aldrich, St. Louis, MO) and 0.35 mol·L¹1 reduction peak of Cd2+ to Cd0 at ¹1.14 V, followed by the current looping in the reversal sweep direction. This type of +Corresponding author, E-mail: [email protected] hysteresis loop, which is marked with arrows in Fig. 1, is a 1468 Y. Kim, J. Jung, S. Kim and W.-S. Chae

5 2.5 1.0 1st (a) (b)

-2 4 2nd 2.0

-2 0.8 rd th 3 3 cm 4 · mA·cm 1.5 0.6 2 ) / mA I / 2

I Experimental max I

Instantaneous / I 1.0 Progressive ( 1 0.4

0 0.5 0.2 Current Density, Current Density, -1 st

Current Density, Current Density, 1 0.0 0.0 0 200 400 600 800 020406080100 -2 Time, t / s t/t -0.4 -0.6 -0.8 -1.0 -1.2 max Potential, E / V vs. Ag|AgCl Fig. 2 CA electrodeposition of CdS on ITO from an aqueous solution of ¹ Fig. 1 CV electrodeposition of CdS on ITO from an aqueous solution of CdSO4 and Na2S2O3·5H2O. Applied potential was 1.1 V. ¹1 CdSO4 and Na2S2O3·5H2O. Scan rate was 0.1 V·s .

the transient was carried out by employing the equations from characteristic feature of a nucleation and growth-deposition Scharifker-Hills model,23) with the data obtained from the process.20,21) The reduction peak was shifted to positive calculations shown in Fig. 2(b). According to the model, potential from the 2nd to the 4th cycle, and the currents there are two limiting nucleation processes: the instantaneous gradually decreased owing to the under potential deposition and progressive. Instantaneous nucleation corresponds to a of CdS and the concentration polarization in double layer, slow growth of nuclei on a small number of active sites, all respectively. Color variations after the 5th cycle were activated at the same time; whereas, progressive nucleation observed on the ITO-surface owing to the buildup of a film corresponds to fast growth of nuclei on many active sites, all with the characteristic yellow color of CdS. The shifting activated during the course of the electroreduction.2325) The of the potential and decreasing of the current indicated a models of the theoretical transients for the instantaneous and growing inactivation of the ITO electrode during the the progressive nucleation are given by eqs. (3) and (4), deposition process. Anodic peaks were observed at ¹0.638 respectively: and ¹0.735 V in the 1st scan, corresponding to the I 2 1:9542 t 2 dissolution of Cd0 and CdS, respectively. The anodic ¼ : ð Þ I t 1 exp 1 2564 t 3 oxidation peak of Cd0 was not present from the 2nd cycle max max + t onwards because the reduction process of Cd2 was restricted max ()"# by the deposited CdS, which is less conductive than ITO. 2 I 2 1:2254 t 2 A negative shift was observed from the 5th cycle between ¼ : ð Þ I t 1 exp 2 3367 t 4 ¹0.95 and ¹1.0 V, which is likely because of the reduction max max 2¹ 2+ 2¹ t of S2O3 ions with Cd to CdS. S2O3 is a source of max 0 elemental S as it can undergo a disproportionation reaction, where Imax and tmax are the coordinates of a CA peak. 12) I2 I2 t t as shown in eq. (1). Figure 2(b) shows the nondimensional max vs. max plots of the data for the CA deposition of CdS film. The S O 2 ! S þ SO 2 ð1Þ 2 3 3 solid and dashed lines are the theoretical transients of the The deposition of CdS could be confirmed by the clear instantaneous and the progressive nucleation, respectively. color change to yellow of the ITO surface. The identification The nucleation and growth processes of CdS can be predicted of CdS films is examined later in the article. The overall from Fig. 2(b). At early stage, the experimental curve well co-deposition was proposed to proceed as shown in fits the curve of the progressive nucleation model by which eq. (2);12,19,22) CdS nucleation occurred on many active sites of ITO surface, þ subsequently, the deposition then changes to an instantaneous Cd2 þ S O 2 þ 2e ! CdS þ SO 2 ð2Þ 2 3 3 nucleation and growth. It could be inferred from these results Equations (1) and (2) were considered as electrochemical- that the CdS film deposited by CA was more dense and chemical type process, in which an initial electrochemical uniform than that by CV-deposition, which is discussed in reduction of Cd2+ to Cd0 is followed by the rate determining further detail below. 0 2¹ 10) reaction of Cd with S2O3 . In CV, the alternating cathodicanodic scan causes the Electrodeposition of CdS was also carried out by CA, reductionoxidation of the CdS film, resulting in repetitive which is a common electrochemical technique used for deposition and dissolution during the forward and reverse electroplating of metals and alloys. Currenttime transients scans, respectively (eq. (2)). It is speculated that this would were observed as the potential is applied (Fig. 2(a)). The cause the formation of a rough, sparse film. sudden change in the current on application of the constant The peaks observed in the XRD spectra in Fig. 3 can be potential was attributed to the double-layer charge between indexed as hexagonal wurtzite-structured CdS, and are highly the ITO surface and the solution with the CdS nucleation and consistent with the standard card (JCPDS 41-1049).26,29) The deposition process occurring immediately after. Analysis of even front CdS films exhibited moderately weaker peaks than Cyclic Voltammetric and Chronoamperometric Deposition of CdS 1469

(a) SnO (In) 2 (002) (101) (a) (100)

(102) (110) (103)

JCPDS 41-1049

(002) (b) (101) Intensity / a.u.

(100)

(110) (102) (103)

20 30 40 50 60 2θ / deg

Fig. 3 (a) and (b) show the XRD spectra of the deposited CdS films obtained by CV and CA deposition, respectively. (b) the underlying ITO thin film, which may be because the CdS films were amorphous. The average particle size of these films was calculated from the half band width of the main XRD peak of plane (101) using the Scherrer equation;2729) K D ¼ ð5Þ T cos ª where D is the peak width in radians (full width at half maximum), K is a constant (0.9), is the X-ray wavelength (0.154 nm), T is the crystallite size of the film and ª is the diffraction angle. The values of T were 28.9 and 30.0 nm for Fig. 4 AFM images of CdS prepared by CV (a) and CA (b) deposition. CdS deposited by CV and CA, respectively. From the XRD data it could be concluded that the CdS films electrochemically deposited using CV and CA were effectively coated CA, however, resulted in constant deposition of CdS under onto the ITO as a hexagonal wurzite structure, and that the reduction conditions, as shown in eq. (2). This enabled particle size was larger in the case of the CA, with a constant continuous particle growth and therefore the formation of cathodic reaction, than the CV, with an alternating anodic larger granules. cathodic potential. The relatively sharp and intense (002) Optical absorption studies of the CdS films were carried peak evident at 26.5 degree in Fig. 3(b) was dissimilar to the out in the wavelength range 2801100 nm. In order to JCPDS 41-1049 (Fig. 3(a)), supporting an extended c-axis, estimate the band-gap energy (Eg) of the CdS, the values of perpendicular direction of ITO-substrate, stacking domain of (¡h¯)2 versus h¯ were plotted, as shown in Fig. 6. The optical the wurtzite lattice in the CA deposited film. band-gap was calculated using the Tauc relation:30) m AFM was subsequently employed in order to observe the ¡h¯ ¼ Aðh¯ EgÞ ð6Þ surface of the deposited CdS films (Fig. 4). For the CV, the 1 CdS appeared rough and relatively thick, with an average fi T thickness of 1.34 µm (Fig. 4(a)). The lm obtained by CA on ¡ ¼ ln ð7Þ the other hand, was a lot smoother and thinner (average d thickness is 0.5 µm) (Fig. 4(b)). These results indicate that where A is a constant, h¯ is the incident photon energy, ¡ is the CdS films deposited by both CV and CA had the same absorption coefficient, Eg is the band-gap energy of the film, crystal structures, but the quality of the film was better when m is equal to 1/2 for the allowed direct transition, T is CA deposition was used. transmittance and d is film thickness. The direct band-gap The surfaces of the two CdS films can be observed in the of CdS films electrodeposited by CV (Fig. 6(a)) and CA SEM images shown in Fig. 5. The surface of CdS deposited (Fig. 6(b)) were determined by extrapolating the straight line by CV consisted of smaller aggregates compared to that to the energy axis, and were found to be 2.59 and 2.71 eV, deposited by CA, which had more defined granules with respectively. These measured values of Eg were larger definite boundaries, even though the particle size was than typical value accepted for polycrystalline CdS film determined to be similar by XRD data. In the cathodic (2.42 eV)31,32) suitable for solar cell applications, and the anodic scan of the CV, CdS was deposited and dissolved possible reason could be regarded that the quantum confine- during the cycling, which may have restricted particle growth ment contributes to the widening of band-gap at small to some extent. The constant cathodic potential applied in the crystallite sizes of CdS film. 1470 Y. Kim, J. Jung, S. Kim and W.-S. Chae

(a) (c)

(b) (d)

Fig. 5 SEM images CdS electrodeposited by CV (a) and (b), and by CA (c) and (d).

5 10 1.8 (a) (b) 1.6 (a) (b) 4 8 1.4 -4 1.2 cm ·

14 1.0 -2

13 3 6 F 10 3 0.8 x / 10 10 2 x x

/ 0.6 ν) 2 / h

2 4 2 ν) α 0.4 C ( h 1/ α

( 0.2 1 2 0.0 -0.5 -0.4 -0.3 -0.2 -0.1 0.0 0.1 Potential, E / V vs. Ag|AgCl 0 0 1.5 2.0 2.5 3.0 3.5 1.5 2.0 2.5 3.0 3.5 Fig. 7 Mott-Schottky plots of 1/C2 as a function of potential for CdS Photon Energy, hν/eV Photon Energy, hν/eV electrodeposited by CV (a) and by CA (b). Fig. 6 Plots of (¡h¯)2 vs. h¯ for determination of band-gap of CdS films electrodeposited by CV (a) and by CA (b). e is the electron charge; E is the applied potential; ND and NA are the donor and acceptor concentration in the film, The Mott-Schottky relation is usually used to describe the respectively; Efb the flat band potential; k the Boltzmann electric properties of films, and is based on the measurement constant; and T is absolute temperature. Figure 7 represents of electrode capacitance as a function of the applied potential. the 1/C2 versus potential plots for CdS films electrodeposited The reciprocal of measured electrode capacity, 1/C, is equal by CV (Fig. 7(a)) and by CA (Fig. 7(b)), with aqueous ¹1 to that of space charge layer, 1/Csc. The relationship between 0.10 mol·L LiClO4 solution as a supporting electrolyte. The C and E for n-type and p-type semiconductors is shown in Mott-Schottky plots of Fig. 7 display positive slopes that 3336) eqs. (8) and (9), respectively: indicate n-type semiconductor behavior. The values of Efb for fi 1 2 kT the CdS lms electrodeposited by the two different methods ¼ E E ð8Þ ¹ C2 ¾¾ A2eN fb e were almost same as 0.45 V. The donor concentration can 0 D be estimated from the slopes. The difference in the values 1 2 kT ¼ E E ð9Þ of the slopes obtained from the Mott-Schottky plot was C2 ¾¾ A2eN fb e 0 A measured to be 17% larger in the CdS film deposited by CV ¹14 in which ¾0 is the vacuum permittivity (8.85 © 10 than by CA deposition. These results suggest that the CdS F·cm¹1); ¾ is the dielectric constant; A is the surface area; film deposited by CV was doped with more ions such as Li+ Cyclic Voltammetric and Chronoamperometric Deposition of CdS 1471

-100 4000 a Photocurrent a b 3500 b

-2 c -80 Dark current d 3000 cm · A μ 2500 -4 -60 a / I cm

b .

2000 2 A -40 μ

1500 / 2 J 1000 -20

Current Density, Current Density, c 500 d 0 0 -0.4 -0.3 -0.2 -0.1 -0.4 -0.3 -0.2 -0.1 0.0 Potential, E / V vs. Ag|AgCl

Fig. 8 Photocurrents of CdS electrodeposited by CV (a) and by CA (b). (c) and (d) are dark conditions.

¹ or ClO4 from the supporting electrolyte solutions, which Eg and Efb, were measured to be almost the same for the two was attributed to its rougher surface morphology. films. However, the dopant density of the CdS deposited by Figure 8 shows the characteristics of the photocurrent CV was found to be greater by 17%, which was attributed to density (Jph) and dark current vs. measured potential for the its increased roughness. The present investigation is expected CdS films deposited by CV and CA. The photocurrent to stimulate interest in the electrodeposition of other binary measurements were carried out in 0.10 mol/L aqueous chalcogenide films, for example, ZnS, ZnSe and CdSe LiClO4 solution under polychromic illumination. The using the CV and CA processes, in which the convenient photo-onset currents of the electrodeposited n-CdS film formulation in solution and the easy controlling morphology started at the anodic scan from the point after Efb was reached of films were accessible. (¹0.45 V), which was estimated in Fig. 7. Figure 8 also 2 shows the dependence of J ph on E, illustrating the functional Acknowledgment relationship predicted by eq. (9):37) 2 This work was supported by the Daejin University Special ND Jph E Efb ¼ ð10Þ Research Grants 2009. 2e¾0¾ ¡J0 ¡ fi where is the light absorption coef cient and J0 is the REFERENCES intensity of the light flux entering the semiconductor. The 2 linear relationship between J ph and E reflects the character- 1) V. K. Singh, P. Chauhan, S. K. Mishra and R. K. Srivastava: Electron. istic dependence of the deletion layer thickness on the Mater. Lett. 8 (2012) 295. potential, and is analogous to the Mott-Schottky dependence 2) S. K. Mishra, R. K. Srivastava, S. G. Prakash, R. S. Yadav and A. C. / 2 Panday: Electron. Mater. Lett. 7 (2011) 31. of 1 C on E. The straight dotted lines in right section of 3) J. S. Hong, J. W. Jeong, W.-S. Chae and K.-J. Kim: Bull. Korean Fig. 8 intersect the potential axis at one point, ¹0.42 V, Chem. Soc. 20 (1999) 597. which almost coincides with the value of Efb measured by the 4) P. K. Ghosh, U. N. Maiti and K. K. Chattopadhyay: Mater. Lett. 60 Mott-Schottky plot. (2006) 2881. 5) H. Zhang, X. Ma, J. Xu and D. Yang: J. Cryst. Growth 263 (2004) 372. 6) H. Uda, H. Yonezawa, Y. Ohtsubo, M. Kosaka and H. Sonomura: Sol. 4. Conclusion Energy Mater. Sol. Cells 75 (2003) 219. 7) S. H. Yoon, S. S. Lee, K. W. Seo and I. W. Shim: Bull. Korean Chem. In this work, CdS films were electrochemically deposited Soc. 27 (2006) 2071. on ITO by CV and CA techniques from an aqueous CdSO4 8) Z. Hu, Y. Yu and H. Hu: Mater. Lett. 64 (2010) 863. 9) N. Romeo, A. Bosio, V. Canevari and A. Podesta: Sol. Energy 77 and Na2S2O3. Both deposited CdS films were identified as fi (2004) 795. having a wurzite hexagonal structure, with the CA lm 10) J. Nishino, S. Chatani, Y. Uotani and Y. Nosaka: J. Electroanal. Chem. displaying a predominant extension of c-axis. From the XRD 473 (1999) 217. data, the particle sizes were found to be similar for the two 11) N. W. Duffy, D. Lane, M. E. Ozsan, L. M. Peter, K. D. Rogers and films; however, the aggregation of particles was more R. L. Wang: Thin Solid Films 361362 (2000) 314. ę significant for the CA. It was found using AFM analysis 12) K. Zar bska and M. Skompska: Electrochim. Acta 56 (2011) 5731. 13) K. R. Murali, S. Kumaresan and J. J. Prince: Mater. Sci. Semi. Proc. 10 that the surface of the CdS electrodeposited by CV was (2007) 56. rougher and thicker compared with that deposited by CA. 14) S. P. Mondal, A. Dhar and S. K. Ray: Mater. Sci. Semi. Proc. 10 (2007) The values for the electrochemical characteristic properties, 185. 1472 Y. Kim, J. Jung, S. Kim and W.-S. Chae

15) S. J. Lade and C. D. Lokhande: Mater. Chem. Phys. 49 (1997) 160. 405 (2010) 919. 16) E. Fatas and P. Herrasti: Electrochim. Acta 33 (1988) 959. 29) I. S. Elashmawi, N. A. Hakeem and M. S. Selim: Mater. Chem. Phys. 17) K. Premaratne, S. N. Akuranthilaka, I. M. Dharmadasa and A. P. 115 (2009) 132. Samantilleka: Renew. Energy 29 (2004) 549. 30) J. Tauc: Amorphous and Liquid Semiconductors, (Plenum Press, New 18) J. Nishino, S. Chatani, Y. Uotani and Y. Nosaka: J. Electroanal. Chem. York, 1974) pp. 159220. 473 (1999) 217. 31) M. Ichimura, F. Goto and E. Arai: J. Appl. Phys. 85 (1999) 7411. 19) S. Dennison: Electrochim. Acta 38 (1993) 2395. 32) V. Popescu, E. M. Pica, I. Pop and R. Grecu: Thin Solid Films 349 20) J. Zhang, M. An and L. Chang: Electrochim. Acta 54 (2009) 2883. (1999) 67. 21) A. I. Inamdar, S. H. Mujawar and P. S. Patil: Int. J. Electrochem. Soc. 2 33) S. Kose, F. Atay, V. Bilgin, I. Akyuz and E. Ketenci: Appl. Surf. Sci. (2007) 797. 256 (2010) 4299. 22) G. Sasikala, R. Dhanasekaran and C. Subramanian: Thin Solid Films 34) J. O’M. Bockris and S. U. M. Khan: Surface Electrochemistry, A 302 (1997) 7176. Molecular Level Approach, (Plenum Press, N.Y., 1993) p. 179. 23) B. Scharifker and G. Hills: Electrochim. Acta 28 (1983) 879. 35) Y. V. Pleskov and Y. Ya Gurevich: Semiconductor Photoelelctrochem- 24) M. Palomar-Pardavé, M. T. Ramírez, I. González, A. Serruya and B. R. istry, (Translation Editor P. N. Bartlett, Consultants Bureau, N.Y., 1986) Scharifker: J. Electrochem. Soc. 143 (1996) 1551. p. 95. 25) D. Grujicic and B. Pesic: Electrochim. Acta 49 (2004) 4719. 36) H. O. Finklea: Semiconductor Electrodes, (Elsevier Science, N.Y., 26) L. Song, H. Wei, H. Xu and J. Zhan: Mater. Res. Bull. 45 (2010) 1396. 1988) p. 27. 27) A. W. Burton, K. Ong, T. Rea and I. Y. Chan: Micro. Meso. Mater. 117 37) Y. V. Pleskov and Y. Ya Gurevich: Semiconductor Photoelelctrochem- (2009) 75. istry, (Translation Editor P. N. Bartlett, Consultants Bureau, N.Y., 1986) 28) R. Seoudi, A. Shabaka, W. H. Eisa, B. Anies and N. M. Farage: Phys. B p. 193.