CdTe is of considerable technological importance for use in semiconductor-based radiation detection applications, such as medical imaging and homeland security fields. The fabricated CdTe radiation detectors can be operated at room-temperature due to the favorable band gap (1.44 eV) and have been researched by various groups [1–7].
High resistivity is required for any semiconductor material to be used in realizing high-performance radiation detectors. Because high bias voltage is needed to improve the charge collection efficiency of the detection devices, resistivity of the material used should be high enough, close to a semi-insulating state, since electronic noise of the material originating from the leakage current increases with larger amounts of free charge carriers within the semiconductor . However, it has been reported that undoped CdTe bulk crystal grown from the melt usually has p-type characteristics due to the acceptor-like nature of the inherent point defects including Cd-vacancies . These point defects in the crystals can also behave as traps for increased leakage currents . Also, another point defects such as Te inclusions have been known to adversely affect the charge transport properties, energy resolution, and uniformity of the fabricated radiation detectors [11–13]. Especially, Szeles et al.  discussed that the region of degraded charge transport near Te inclusions extends well beyond the volume of the inclusions themselves possibly due to the field of dislocations and Te precipitates surrounding Te inclusions. In this connection, Cl doping for annihilating all these point defects (i.e. Cd-vacancies and Te inclusions) can be an effective way to obtain semi-insulating (above 109 Ω·cm) high-quality CdTe material [2, 8, 9, 13, 14].
Thus, in this study, we investigated the effects of Cl doping on the microstructural, electrical, and optical properties of CdTe ingots grown by using the low-pressure Bridgman (LPB) method. Samples from three different locations within each ingot (i.e. top, middle, bottom) were also compared. It was shown that the samples from the middle part of the ingot with Cl doping concentrations of 9.94 × 1019 cm−3 and 1.99 × 1020 cm−3 exhibited better crystalline qualities compared to other samples. It was also shown that the average size and surface coverage of Te inclusions decreased with increased Cl concentrations. Resistivity of the samples were increased by 5 to 6 orders of magnitude for Cl-doped samples. Optical properties of the samples measured by spectroscopic ellipsometry (SE) were also reported and discussed.
2.1 Sample preparation
CdTe raw materials of 99.9999 % purity were purchased from 5N Plus Co. and were used as-received. The powders were injected into 1-inch quartz ampoules which were sealed at a background pressure of ∼ 1.33 × 10−4 Pa. We put different amounts of Cl dopant into four different ampoules. In other words, one was undoped, and the other three were doped with Cl at the concentrations of 4.97 × 1019 cm−3, 9.94 × 1019 cm−3 and 1.99 × 1020 cm−3, respectively. The sealed ampoules were then loaded into a LPB furnace. For all four growth runs, 6-zone furnace moved at a speed of 0.93 mm/h and the thermal gradient at the growth zone was approximately 10 °C/cm. Fig. 1 shows the four grown ingots used in this study. The ingots were then sliced perpendicular to the growth direction, by using a diamond wire saw. The samples from each part of the ingots (also designated in Fig. 1) are herein referred to as ’top’, ’middle’ and ’bottom’ samples, respectively. Then, these samples were lapped with 800-, 2400- and 4000-grit SiC papers followed by polishing with different Al2O3 powders of 1 µm, 0.3 µm and 0.05 µm diameters. To remove the mechanical damages at the sample surfaces during the processes, all the samples were chemically etched by dipping in the 2 % Br-methanol solutions. Finally, Au electrodes were deposited onto both sides of the samples by using the AuCl3 solutions.
Microstructural perfections of the samples were investigated using an XRD (X-ray diffraction) system (D/Max 4200, Rigaku) within the 2 theta range of 20° to 80°. Te inclusions within the samples were observed by infrared (IR) mapping system with a Nikon infrared camera. To characterize the optical properties of the samples, spectroscopic ellipsometry (SE, J.A. Woollam Co., photon energy: from 1.1 eV to 6.5 eV) measurements at room temperature (RT) were performed. Finally, electrical measurements were performed by using a Keithley 4200 semiconductor characterization system.
3 Results and discussion
Fig. 2 shows XRD measurement results for the samples from four ingots with different Cl doping concentrations (undoped, 4.97 × 1019 cm−3, 9.94 × 1019 cm−3 and 1.99 × 1020 cm−3) chosen from top, middle and bottom sections, respectively. Regardless of doping concentrations, samples from the middle section of the ingots exhibited better crystallinity, showing very sharp (1 1 1) peak. It is also shown that the middle samples with Cl doping concentrations of 9.94 × 1019 cm−3 and 1.99 × 1020 cm−3 exhibited slightly better crystal quality with reduced unwanted small peaks (e.g. (5 1 1) peak).
Fig. 3 shows typical IR microscope images of Te inclusions in the middle samples with different Cl doping concentrations (4.97 × 1019 cm−3, 9.94 × 1019 cm−3 and 1.99 × 1020 cm−3). Also, histograms of the Te inclusion size analyzed from several IR microscope images (including Fig. 3) for the three samples are shown in Fig. 4. It is noted from Fig. 3 and Fig. 4 that the average size of the Te inclusions decreased with increasing Cl doping concentrations. For example, the samples with different Cl doping concentrations exhibited average sizes of 19.53 µm (4.97 × 1019 cm−3), 9.31 µm (9.94 × 1019 cm−3) and 3.92 µm (1.99 × 1020 cm−3). In addition, the observed average surface coverage of Te inclusions in the CdTe sample decreased with increasing Cl doping concentrations from 3.90 % (4.97 × 1019 cm−3), 1.86 % (9.94 × 1019 cm−3) to 0.47 % (1.99 × 1020 cm−3). Similar result was also obtained by Jayatirtha et al. . However, it should be mentioned that the exact mechanisms underlying the participation of chlorine in the improvement of CdTe morphology are not clearly understood, what is also mentioned in the literature  and references therein, and the subject is currently under further investigations.
Fig. 5a shows a typical dielectric response of undoped and doped (1.99 × 1020 cm−3) samples within the photon energy range of 1.1 eV to 6.5 eV. From the data in Fig. 5a, extinction coefficient (k) and absorption coefficient (α = 4πk/λ, where λ is wavelength in cm) can be obtained. Because CdTe is a direct-transition semiconductor, its optical bandgap can be obtained by extrapolation of the linear part of the α2 vs. photon energy (hν, in eV) curves toα = 0 , as shown in Fig. 5b. It is noted that doped CdTe exhibits slightly larger bandgap, by 8 meV compared to that of undoped sample, which means smaller Urbach tails below/above the conduction/valence band edges . Fig. 5c shows the second derivative spectra  of the ϵ1 curves (from Fig. 5a) near the bandgaps. Gamma values of the peaks  were obtained to be 0.169056 eV for the undoped sample, and 0.153282 eV for the doped sample. From Fig. 5b and Fig. 5c, the doped CdTe sample can be regarded to have better crystalline qualities.
Fig. 6a shows average resistivities of CdTe samples as a function of Cl doping concentration. It is shown that the resistivities have abruptly increased by almost 5 orders of magnitude with an addition of small amounts of Cl dopant (4.97 × 1019 cm−3). Furthermore, with further increase in the dopant concentration, the resistivities were only slightly increased to about 1010 Ω·cm as it is also summarized in Table 1. Similar resistivity saturation phenomenon was also observed by Shin et al. . Fig. 6b shows variation of average resistivities of the samples as a function of different locations within the ingots. It is shown that the samples from the middle part of the ingots exhibit slightly higher resistivities compared to the other parts (i.e. top or bottom samples), as has been also summarized in Table 1. This might be due to the fact that the samples with poorer crystalline quality have higher density of unwanted parasitic shunt paths for charge carriers  leading to lower resistivities.
Average resistivities of the CdTe samples with different Cl doping concentrations and different positions within the ingots. Mean values for each row/column were also calculated and presented in the Table.
|Top [Ω·cm]||Middle [Ω·cm]||Bottom [Ω·cm]||Mean [Ω·cm]|
|Undoped||5.20 × 104||1.33 × 104||1.60 × 104||2.71 × 104|
|Cl doped, 4.97 × 1019cm−3||3.56 × 106||5.70 × 109||2.07 × 109||2.59 × 109|
|Cl doped, 9.94 × 1019cm−3||1.06 × 109||8.03 × 109||3.35 × 109||4.15 × 109|
|Cl doped, 1.99 × 1020cm−3||2.19 × 109||1.44 × 1010||2.11 × 109||6.22 × 109|
|Mean [Ω·cm]||8.15 × 108||7.02 × 109||1.88 × 109|
Although the exact mechanisms of the role of Cl dopant during the CdTe growth are not clearly understood at this moment, the following can be stated. Based on the structure-property relationships, Cl dopant in CdTe may take an important role to improve the crystalline quality of CdTe. In other words, XRD measurement results (Fig. 2) show that the doped samples exhibit reduced, thus almost negligible (5 1 1) peaks. Our preliminary XRF (X-ray fluorescence) measurements results for the undoped and doped (1.99 × 1020 cm−3) samples (not shown) also showed that the Cl dopants were clearly observed and properly incorporated within the host CdTe material. In this work, it has been shown that Cl dopant helped reduce Te inclusions as shown in Fig. 3 and Fig. 4. In addition, effects of Cl dopant in enhancing the CdTe material quality from the optical point of view are observed in Fig. 5. In other words, the optical bandgap and gamma value are larger and smaller, respectively, for doped samples as can be observed in Fig. 5. Finally, from resistivity calculations, Cl dopant was shown to successfully compensate native point defects, resulting in the semi-insulating CdTe, as shown in Fig. 6. Thus, although the optimum doping concentrations were reported by different research groups to be slightly different [9, 13, 19], we can assume that Cl doping is an effective way to enhance the material properties, which can be beneficial for making high-performance radiation detectors.
CdTe crystals doped with different amounts of Cl were characterized for use in radiation detection applications. It was shown that the CdTe crystals from the middle parts of the ingots doped at concentrations of 9.94 × 1019 cm−3 and 1.99 × 1020 cm−3 exhibited better crystalline qualities as observed from XRD, IR mapping, and SE measurement and analysis results. It was further observed that the CdTe samples doped with a doping concentration of 1.99 × 1020 cm−3 from the middle part of the ingot exhibited an average resistivity of 1.44 × 1010 Ω.cm. These findings clearly demonstrate that our Cl-doped CdTe samples can successfully be used for fabricating high-performance radiation detectors.
We are grateful to Prof. Tae Jung Kim and Prof. Young Dong Kim, Mr. Han Gyeol Park (all with the Kyung Hee University) for fruitful discussions regarding SE analysis. The authors would like to gratefully acknowledge the financial support by the Nuclear R&D Programs of the Ministry of Science, the ICT & Future Planning (MSIP) of South Korea (NRF-2010-0026096, NRF-2013M2A2A4023359), a grant from the Korea Atomic Energy Research Institute (Grant No.: 523280-15), and by the WCU (World Class University) Program through the National Research Foundation (NRF) of Korea funded by the Ministry of Education, Science and Technology (R31-2008-10029).
Meyer B.K. Hofmann D.M. Stadler W. Salk M. Eiche C. Benz K.W.Proceedings of the MRS Spring Meeting San Francisco 1993.
Weigel E. Miiller-Vogt G. Steinbach B. Wendl W. Stadler W. Hofmann D.M. Meyer B.K.Mater. Sci. Eng. B-Adv. 16 (1993) 17.
Korbutyak D.V. Krylyuk S.G. Tkachuk P.M. Tkachuk V.I. Korbutjak N.D. Raransky M.D.J. Cryst. Growth 197 (1999) 662.
Popovych V.D. Virt I.S. Sizov F.F. Tetyorkin V.V. Tsybrii(Ivasiv) Z.F. Darchuk L.O. Parfenjuk O.A. Ilashchuk M.I.J. Cryst. Growth 308 (2007) 63.
Bolotnikov A.E. Camarda G.S. Carini G.A. Cui Y. Kohman K.T. Li L. Salomon M.B. James R.B.IEEE T. Nucl. Sci. 54 (2007) 821.