Processing of CdTe thin films by intense pulsed light in the presence of Cd[Cl.sub.2].
Keywords Intense pulsed light, CdTe, Cd[Cl.sub.2]
Thin films of cadmium telluride (CdTe) for photovoltaic (PV) applications are typically subjected to a post-deposition treatment involving high temperatures in order to enhance the crystallinity and morphology leading to more favorable optical and electrical properties. This often includes the use of halogens such as chlorine (Cl) and fluorine (F) during the heat treatment to improve the crystal size and orientation. (1,2) In particular, cadmium chloride (Cd[Cl.sub.2]) has been used for several decades and can be applied onto deposited CdTe thin films using a simple solution technique and annealed at temperatures of approximately 400[degrees]C.
Recent reviews have indicated that CdTe thin films grown at low temperatures undergo three stages during heat treatment. (3-5) The first stage occurs during low temperature growth of CdTe which typically produces highly orientated (111) films with small grains. These films generally produce solar cells with low performance, with efficiencies less than 12%. As the annealing temperature of the film approaches approximately 380[degrees]C, there is a structural transformation where the predominant (111) orientation drops off leaving a random structure with all three orientations (111), (220), and (311) present. This second region typically produces solar cells with very low performance due to gaps and defects existing between the large grains. As the annealing temperature increases to approximately 450[degrees]C, large grains of the preferential (111) orientation are slowly reestablished. In this third region, high efficiencies are typically realized. Many applications of the thermal process have been studied including vapor phase at low pressures, prolonged thermal exposure, and rapid thermal annealing with optimization of grain size and (111) orientation being the goal.
Thermal annealing is a significant step and techniques that can speed the process would be advantageous to the industry. It has been recently shown that the intense pulsed light (IPL) process could drastically reduce the time required to increase the grain size and preferentially orient the grains in the (111) plane for electrodeposited CdTe thin films. (6) The IPL technique can anneal photosensitive thin films by delivering high-energy light in a very short duration over a large processing area. The method is analogous to additive manufacturing techniques that utilize lasers to locally sinter powders; however, IPL uses a broader spectrum of light and the processing area is far larger and thus is simpler to apply toward economically feasible manufacturing. Currently, IPL has primarily been employed to sinter conducting Ag, Au, and Cu nanoparticle films on to low temperature substrates (e.g. polymers) in air for flexible electronics. (7-11) However, little work has been done with semiconducting materials. Colorado, et al. (12) applied IPL sintering to cubic CdS deposited by CBD. IPL has also been applied to the sintering of CuInGa[Se.sub.2] NPs. (13,14) Our group has explored the IPL process with both CdS and CdTe deposited using electrochemical techniques. (6,15)
The research presented here provides experimental evidence of the transition from the first to second stage using IPL and Cd[Cl.sub.2] treatments. In this study, the impact of the specific energy density (ED) and total energy input of the IPL process in modifying the morphology of electrodeposited CdTe is investigated. The IPL process is then combined with the Cd[Cl.sub.2] treatment. Films are studied using X-ray diffraction, ultraviolet, visible to infrared (UV-Vis-IR) spectroscopy, and scanning electron microscopy (SEM).
Materials and methods
CdTe films were fabricated by 2-electrode cathodic electrochemical deposition on to TEC-8 FTO glass substrates (Hartford Glass Co. Inc.). The FTO was used as the cathode and a graphite rod was used as the anode. The 6.0 x 4.0 [cm.sup.2] substrates were placed in a Teflon vessel holding an 1400 ml aqueous solution of 0.5 M CdS04 (Sigma-Aldrich). The pH of the solution was decreased to 1.44 using 1.0 M [H.sub.2]S[O.sub.4] (Sigma-Aldrich). The deposition was carried out at 85[degrees]C, and the films were deposited for 2 h. A growth voltage ([V.sub.g]) of 1.526 V was applied using a Keithley 2400 source-meter. The deposition current was maintained at ~170 [micro]A[cm.sup.-2], by adjusting the stirring rate and adding Te[O.sub.2] dissolved in 1.0 M [H.sub.2]S[O.sub.4]. The deposited film thickness was approximately 400 nm as measured using a Tencor AlphaStep 500 contact profilometer.
IPL processing of CdTe was carried out in air using a Sinteron 2000 (Xenon Corporation). The system delivers rapid pulses of light with wavelengths ranging from the UV to IR region. In this case, CdTe was treated with pulses lasting 1.0 ms. The system was set so that lamp cycled between its "ON" and "OFF" conditions. The lamp cycled on, pulsing twice for 1.0 s, and off for 1.0 s. The ED of each pulse was 21.6 or 25.9 J [cm.sup.-2], and the number of pulses was adjusted between 20 and 120 (for a total process time of between 20 and 120 s). The total energy input during the IPL process is the product of the number and the ED of the pulses. Total energy inputs of 431, 862, 1294, 1725, 2157, and 2588 J [cm.sup.2] were applied to the CdTe films using 20, 40, 60, 80, 100, and 120 pulses, respectively, with an ED of 21.6 J [cm.sup.2]. Total energy inputs of 414, 880, 1294, 1708, 2123, and 2588 J [cm.sup.-2] were applied to the CdTe films using 16, 34, 50, 66, 82, and 100 pulses, respectively, with an ED of 25.9 J [cm.sup.2].
CdTe was also IPL treated after the deposition of a thin Cd[Cl.sub.2] film. In this case, Cd[Cl.sub.2] was spin coated on to the surface of the as-deposited film from an aqueous solution of 1.0 M Cd[Cl.sub.2] (Sigma-Aldrich). The films were then IPL processed as described above using ED of 8.63, 12.94, 17.26, 21.57, and 25.88 J [cm.sup.2] and 100 pulses, for a total processing time of 100 s. Thus total energy input of 860, 1290, 1730, 2160, and 2590 J [cm.sup.2] was used.
The materials' crystallinity and phases were studied using a Bruker AXS D8 X-ray Diffractometer. The equipment was operated with X-ray source of [CuK.sub.[alpha]] ([lambda] = 0.1548 nm), a position sensitive detector (PCD), a scan speed of 0.5 s/step and step size of 0.02[degrees]. XRD patterns were measured using the [theta]-2[theta] method in the 2[theta] range of 20[degrees]-60[degrees].
An FEI Nova NanoSEM 600 was used to study the morphology of the surfaces with an accelerating voltage of 15 kV and a working distance of 5-6 mm. A thin layer of gold was sputtered on the surface of the samples before being studied by the SEM to avoid charging effects. Optical microscopy images of the films after IPL treatment were carried out using a ZEISS Axio Imager A2 m. The high contrast images were taken in the differential interference contrast mode.
Results and discussion
IPL and the number of pulses (without Cd[Cl.sub.2])
IPL irradiates the surface of the film with pulses of incoherent light from a Xenon lamp. In the semiconductor system, photons with energy greater than or equal to the material's optical bandgap are absorbed, exciting electrons from the valence to the conduction band. Upon relaxation, phonons are produced in the crystal lattice creating a rise in the lattice temperature. The excitation and relaxation of the electrons are considered to be an instantaneous mechanism and can thermally process the material in ambient conditions, without oxidation. Varying the ED of the pulses and the number of pulses yields control of the temperature rise in the film. The ED of the IPL system can be adjusted to deliver up to 25.9 J [cm.sup.2], at a rate of 1 pulse per second. In this study, the ED and number of pulses are regulated where the total energy input from the IPL is the product of the ED and number of pulses. (ED designates the energy intensity per area of the strobe and the total energy input is the cumulative ED delivered during the process).
Initial studies on IPL of CdTe thin films focused on the effect of the specific ED delivered over 100 pulses. This total energy input (product of number of pulses and ED) was then used to bracket the optimal process. It was found that applying 100 pulses with an ED of 21.6 J [cm.sup.2] produced the most crystalline films, while 100 pulses at ED of 25.9 J [cm.sup.2] created significant morphological changes. The process was bracketed using between 90 and 110 pulses of ED of 21.6 J [cm.sup.2], which resulted in very similar crystalline films. These results suggest that a total energy input between 1950 J [cm.sup.2] and 2350 J [cm.sup.2] produced the optimal CdTe films at 400 nm of thickness. (6) The absorption of light is going to change through the layer according to the absorption coefficient. This will create a linear gradient of temperature throughout the film. (15)
In order to further investigate the IPL process, the total energy input was varied by changing the number of pulses using an ED of 21.6 J [cm.sup.2] and 25.9 J [cm.sup.2]. Figure 1(i) shows the change in the [(111).sub.IPL]/[(111).sub.as]. deposited ratio vs the total energy input. The films treated with the 21.6 J [cm.sup.2] pulses show that at least 1725 J [cm.sup.2] (80 pulses) was needed to produce a ratio greater than 1. Below this input, a slight decrease in the ratio was observed, indicating that the cooling rate was faster than the recrystallization rate. Optimum recrystallization occurred at 100 pulses. When the ED was increased to 25.9 J [cm.sup.2], only 34 pulses (880 J [cm.sup.2]) was required for the [(111).sub.IPL./[(111).sub.as-dePosited] to exceed 1.5. Increasing the number of pulses beyond this caused only a slight improvement in the film's crystallinity. The two curves cross at about 2000 J [cm.sup.2]; however, the large reduction in the (111) peak is not seen when an ED of 25.9 J [cm.sup.2] is used. The changes in the (220) and (311) peaks follow the same general trend [Figs, 1(iii), (iv)]. Figure 1(ii) shows the change in the crystallite size with respect to the total energy input during the IPL process at the two ED's. The crystallite size for the films treated with 21.6 J [cm.sup.2] appears to be larger than for the films treated with 25.9 J [cm.sup.-2] pulses, suggesting that solid state diffusion of the particles is occurring to create larger crystallites at the lower energy inputs.
In summary, when CdTe films are IPL treated without Cd[Cl.sub.2], grains continue to grow with (111) preferred orientation and then all three peaks reduce in intensity. This indicates the degrading of crystallinity, most probably due to loss of Cd from the CdTe film.
Figures 2 and 3 show the effects of the number of pulses on the surface morphology of the films when using pulse energy densities of 21.6 and 25.9 J [cm.sup.2], respectively. Figures 2(v) and 2(vi) indicate a drastic change of morphology, and could cause reduction of preferential orientation of the grains. This structural transition seems to occur at the melting of the material in grain boundaries. Figures 3(v) and 3(vi) also show similar changes. After this change, grains are much larger and appear to have turned into random orientations at melting of the grain boundaries. This random nature reduces the (111) preferred orientation and leads to an increase in the (220) and (311) peaks as shown in the XRD in Fig. 1.
From both the XRD and SEM results, it is clear that the ED of the pulses is critical during film formation. Melting and coalescence on the surface of the films becomes apparent only after 60 pulses or 16 pulses with an ED of 21.6 and 25.9 J [cm.sup.2] were used, respectively. This 20% increase in pulse energy yields a temperature rise that is sufficient to increase the crystallinity and begin melting in only 16 pulses. The extremely fast kinetics of the IPL process allows the sample to be treated in air without any observable oxidation. Figure 3 also shows that although the entire film is irradiated with light, melting begins in discrete areas. These areas appear to have been made up from very small nano sized particles. As the number of pulses was increased, the temperature rise was sufficient to show grain growth and melting throughout the film.
The IPL processing of CdTe thin films has demonstrated the processes ability to produce grains as large as 1 [micro]m, in less than 2 min without the use of Cd[Cl.sub.2]. In conventional thermal processes, the material must be kept at temperatures as high as 400[degrees]C for approximately 30 min to achieve similar grain growth. (16)
IPL treatment with Cd[Cl.sub.2]
The use of halogen containing compounds in the thermal processing of CdTe has become prevalent in the development of CdTe based PV devices. Although alternatives such as Freon gases (e.g., dichlorofluoromethane) have been used, Cd[Cl.sub.2] still remains a commonly exploited chemical to produce high efficiency solar cells. While the entire benefits of Cd[Cl.sub.2] are still unknown, it is clear that Cd[Cl.sub.2] promotes the growth and recrystallization of CdTe. One of the most common and simplest methods to apply Cd[Cl.sub.2] before heat treatment is to apply a saturated solution of Cd[Cl.sub.2] on the surface. This results in a thick coating of Cd[Cl.sub.2] on the surface. As discussed previously, the efficiency of the IPL process is based on the materials ability to absorb light. Cd[Cl.sub.2] is known to absorb UV light, therefore much of the energy required to heat the CdTe film will be absorbed by this layer. (17) This will result in the sintering of the Cd[Cl.sub.2], reducing the thermal treatment in the CdTe layer. Therefore, to counteract this effect, a thin coat of Cd[Cl.sub.2] was spin coated on to the surface from an aqueous 1 M solution prior to IPL.
Figure 4(i) shows the effect of the total energy input during the IPL process on the [(111).sub.IPL]/[(111).sub.as-deposited] ratio. The Cd[Cl.sub.2] treated samples display an enhanced trend to the untreated samples with the largest improvement in crystallinity being obtained using pulses with an ED of 21.6 J [cm.sup.2].
Cd[Cl.sub.2] is believed to work as a "fluxing" agent on the surface of the particles, leading to a reduction in the processing temperature needed to induce physical changes. In the IPL process, the pulse ED was not shifted to lower values to promote maximum recrystallization when Cd[Cl.sub.2] was used. However, it was observed that while a pulse ED of 8.6 J [cm.sup.-2] induced disorder in the untreated films, the Cd[Cl.sub.2] caused the [(111).sub.IPL]/ [(111).sub.as-deposited] ratio to remain above 1. This indicates the catalytic effect of Cd[Cl.sub.2] in the re-crystallization process. Cd[Cl.sub.2] treatment has been known to induce a loss of orientation in the (111) crystalline plane and induce random orientation showing other increased CdTe related peaks. The CdTe films treated with Cd[Cl.sub.2] retain their preferential orientation to the (111) crystalline plane until a 100 pulses with an ED of 21.6 J [cm.sup.2] was used. After this point, the collapse of (111) peak is shown by XRD results [see Fig. 4(i)]. Figure 4(ii) shows the effect of energy input on the crystallite size for the Cd[Cl.sub.2] treated samples. A maximum crystallite size of 58 nm was found for the films treated with a pulse ED of 21.6 J [cm.sup.2].
Interestingly, the changes in the (220) and (311) orientations are drastically different when Cd[Cl.sub.2] treatment is used [Figs. 4(iii), (iv)]. Without Cd[Cl.sub.2] on the surface, loss of Cd from CdTe takes place, resulting in a reduction to the intensity of (111) and (220) peaks and minimal change to (311) peaks. The presence of Cd[Cl.sub.2] on the surface prevents loss of Cd from CdTe layer, and acts as a fluxing agent or catalytic source, thus reorientation occurs. In the presence of Cd[Cl.sub.2], rapid growth of (220) and (311) peaks occur when (111) peak starts to collapse. At this point, the grains lose their preferred orientation and show a random nature. The grain boundaries will have entered a melt phase, resulting in a movement of the CdTe grains to a random structure as the film cools down. Grains can also coalesce easily across the liquid margins forming larger grains. This reorientation of large grains should lead to a higher surface roughness.
This is evident in Fig. 5 showing the SEM topographical changes of the CdTe thin films exposed to Cd[Cl.sub.2] after 100 pulses of ED 17.6, 21.6, and 25.9 J [cm.sup.-2]. These films produced particles similar in shape, but larger in size to the uncoated films; however, these films appear to be rougher. This structural reorganization can often precede a rise in grain size followed by a reorientation to the preferential (111) as more energy is delivered to the thin film.
Figure 6 shows optical microscope images of the uncoated and Cd[Cl.sub.2]-treated films. When 100 pulses of light with an ED of 12.9 J [cm.sup.-2] were applied, no structural changes were observed. Upon application of pulses with an ED of 25.9 J [cm.sup.-2], significant macroscopic structures were observable in the films treated with Cd[Cl.sub.2]. The structures in the film appear to have a centralized point, which propagates out to make islands greater than 200 nm in width [Fig. 6(vi)]. This phenomenon, termed "explosive crystallization" has been observed in the laser and IPL treatment of amorphous silicon. (19-20) Exposure to the irradiation source causes the amorphous region to melt and recrystallize. If the latent heat released during crystallization is large enough, adjacent amorphous regions will also melt and recrystallize. In this manner, crystallization in the film accelerates until an autocatalytic process occurs. The propagation front will cease when the rate of heat dissipation (influenced by the thermal conductivity of the material and irradiation time) exceeds the latent heat released during crystallization process. (21) In the CdTe films, it would appear that these explosive fronts originate from areas of densely coated Cd[Cl.sub.2] on the surface. However, this requires further investigation to confirm. Increasing the ED of the pulses further, resulted in the propagation waves being less observable. However, the large discrete islands remained. In addition, delamination of the films appears to have been observed.
During thermal processing, recrystallization will initially proceed, followed by particle growth. However, recrystallization is inextricably linked to both the particle size and stress in the film. This is why CdTe films grown at high temperatures (e.g., CSS) do not always demonstrate changes to their large grains after treatment. The low temperature growth of electrochemical deposition on the other hand produces small particles under stress, which are susceptible to thermal treatment. In the IPL process, pulses with low energy densities (<17.3 J [cm.sup.-2]) generate sufficient energy for the recrystallization process to begin. Under these conditions, particle growth is observed, where the smaller particles undergo solid state surface diffusion. Melting in the films can be achieved by either increasing the number of pulses or the ED of the pulse. Increasing the surface temperature by varying the number of pulses can be an inefficient method as the temperature rise is in direct competition with the cooling rate. The cooling rate is a function of the thermal conductivity of the material, the interval time between pulses and pulse duration. Inducing a sufficient temperature rise by increasing the ED of the pulses is much simpler, as melting will occur during the pulses.
This work identifies and visually observes a structural transition taking place in electrodeposited CdTe layers throughout IPL treatment. During IPL treatment without Cd[Cl.sub.2], grains grow gradually keeping the (111) preferred orientation. At a threshold total energy input, variant on the ED, XRD peaks reduce in intensity indicating deterioration of crystallinity most probably due to loss of Cd from the CdTe film. In the presence of Cd[Cl.sub.2], however, the situation is different. At early stages of IPL treatment, crystallinity improves, keeping the (111) preferred orientation. At a total energy input threshold, the layer becomes randomly oriented, but crystal grains grow continuously. These grains are randomly oriented showing all three CdTe peaks with comparable intensities.
Abbreviations IPL Intense pulsed light CdTe Cadmium telluride Cl Chlorine F Fluorine Cd[Cl.sub.2] Cadmium chloride SEM Scanning electron microscope XRD X-ray diffraction
R. Dharmadasa, B. W. Lavery, T.Druffel ([mail])
Conn Center for Renewable Energy Research, University of Louisville, Louisville, KY 40292, USA
I. M. Dharmadasa
Materials and Engineering Research Institute, Sheffield
Hallam University, Sheffield S1 1WB, UK
Acknowledgment The authors would like to acknowledge the Conn Center for Renewable Energy Research at the University of Louisville for their financial support.
(1.) Romeo, A, Buecheler, S, Giarola, M, Mariotto, G, Tiwari, A, Romeo, N, Bosio, A, Mazzamuto, S, "Study of CSS-and HVE-CdTe by different recrystallization processes." Thin Solid Films, 517 (7) 2132-2135 (2009)
(2.) Paudel, NR, Young, M, Roland, PJ, Ellingson, RJ, Yan, Y, Compaan, AD, "Post-deposition processing options for high-efficiency sputtered CdS/CdTe solar cells." J. Appl. Phys., 115 (6) 064502 (2014)
(3.) Dharmadasa, I, "Review of the Cd[Cl.sub.2] treatment used in CdS/CdTe thin film solar cell development and new evidence towards improved understanding." Coatings, 4 (2) 282-307 (2014)
(4.) Kim, M, Lee, J, Lee, S, Sohn, S, "Study of CdTe/CdS heterostructure by Cd[Cl.sub.2] heat treatment via in situ high temperature XRD." Sol. Energy Mater. Sol. Cells, 109 209-214 (2013)
(5.) Durose, K, Edwards, P, Halliday, D, "Materials aspects of CdTe/CdS solar cells." J. Cryst. Growth, 197 (3) 733-742 (1999)
(6.) Dharmadasa, R, Lavery, B, Dharmadasa, IM, Druffel, T, "Intense pulsed light treatment of cadmium telluride nanoparticle-based thin films." ACS Appl. Mater. Inter., 6 (7) 5034-5040 (2014)
(7.) Chung, WH, Hwang, HJ, Lee, SH, Kim, HS, "In situ monitoring of a flash light sintering process using silver nanoink for producing flexible electronics." Nanotechnology, 24 (3) 035202 (2012)
(8.) Hwang, HJ, Chung, WH, Kim, HS, "In situ monitoring of flash-light sintering of copper nanoparticle ink for printed electronics." Nanotechnology, 23 (48) 485205 (2012)
(9.) Kim, HS, Dhage, SR, Shim, DE, Hahn, HT, "Intense pulsed light sintering of copper nanoink for printed electronics." Appl. Phys. A: Mater. Sci. Process., 97 (4) 791-798 (2009)
(10.) Han, WS, Hong, JM, Kim, HS, Song, YW, "Multi-pulsed white light sintering of printed Cu nanoinks." Nanotechnology, 22 (39) 395705 (2011)
(11.) Dharmadasa, R, Jha, M, Amos, DA, Druffel, T, "Room temperature synthesis of a copper ink for the intense pulsed light sintering of conductive copper films." ACS Appl. Mater. Inter., 5 (24) 13227-13234 (2013)
(12.) Colorado, HA, Dhage, SR, Hahn, HT, "Thermo chemical stability of cadmium sulfide nanoparticles under intense pulsed light irradiation and high temperatures." Mater. Sci. Eng. B: Adv. Fund. Solid-State Mater., 176 (15) 1161-1168 (2011)
(13.) Dhage, SR, Hahn, HT, "Rapid treatment of CIGS particles by intense pulsed light." J. Phys. Chem. Solids, 71 (10) 1480-1483 (2010)
(14.) Dhage, SR, Kim, HS, Hahn, HT, "Cu(In, Ga)[Se.sub.2] thin film preparation from a Cu(In, Ga) metallic alloy and Se nanoparticles by an intense pulsed light technique." J. Electron. Mater., 40 (2) 122-126 (2011)
(15.) Dharmadasa, R, Dharmadasa, IM, Druffel, T, "Intense pulsed light sintering of electrodeposited CdS thin films." Advan. Eng. Mater. (2014). doi:10.1002/adem.201400008
(16.) Moutinho, H, Hasoon, F, Abulfotuh, F, Kazmerski, L, "Investigation of polycrystalline CdTe thin films deposited by physical vapor deposition, close-spaced sublimation, and sputtering." J. Vac. Sci. Technol. A, 13 (6) 2877-2883 (1995)
(17.) Nakagawa, H, Matsumoto, H, "Optical absorption and luminescence of the I-Center in Cd[Cl.sub.2]." J. Phys. Soc. Japan, 36 (6) 1577-1582 (1974)
(18.) Moutinho, H, Al-Jassim, M, Levi, D, Dippo, P, Kazmerski, L, "Effects of Cd[Cl.sub.2] treatment on the recrystallization and electro-optical properties of CdTe thin films." J. Vac. Sci. Technol. A, 16 (3) 1251-1257 (1998)
(19.) Geiler, HD, Glaser, E, Gotz, G, Wagner, M, "Explosive crystallization in silicon." J. Appl. Phys., 59 (9) 3091-3099 (1986)
(20.) Ohdaira, K, Fujiwara, T, Endo, Y, Nishizaki, S, Matsumura, H, "Explosive crystallization of amorphous silicon films by flash lamp annealing." J. Appl. Phys., 106 (4) 044907 (2009)
(21.) Pore, V, Ritala, M, Leskela, M, Saukkonen, T, Jam, M, "Explosive crystallization in atomic layer deposited mixed titanium oxides." Cryst. Growth Des., 9 (7) 2974-2978 (2009)
Please note: Some tables or figures were omitted from this article.
|Printer friendly Cite/link Email Feedback|
|Title Annotation:||cadmium telluride|
|Author:||Dharmadasa, R.; Lavery, Brandon W.; Dharmadasa, I.M.; Druffel, T.|
|Publication:||Journal of Coatings Technology and Research|
|Date:||Sep 1, 2015|
|Previous Article:||Direct thickness measurement of doctor-bladed liquid film on gravure roll surface.|
|Next Article:||Numerical simulation of the dip-coating process with wall effects on the coating film thickness.|