1. Introduction

In general, infrared optical materials are insulators or semiconductors as judged by their band-gaps and resistivity. Photons of light corresponding to energy greater than the band-gap of the solid are strongly absorbed at the surface. [1-2] As the wavelength is increased and the photon energy is decreased below the band-gap, light is transmitted through the solid. The beginning of light transmission of a solid occurs at the wavelength that corresponds to the band-gap energy. The absorption of the photon is a very strong, quantized electronic transition.

One may think of this energy as representing the average ionization energy for the primary chemical bonds formed between the atoms that make up the solid. If the required ionization energy is large enough, transmission begins in the ultraviolet region of the spectrum, as in the case for alkali halides or alkaline earth halides. Then the solid appears water-clear or colorless. If it occurs in the visible region, the solid appears colored. If the absorption edge occurs in the infrared region, the solid appears metallic because all visible light is strongly absorbed and reflected.

2. Experimental

Considerable change in optical, electrical, and mechanical properties of CdTe thin films is produced by doping and annealing conditions [3]. Moreover p-type and n-type doping of CdTe is easy to achieve. All samples we studied are CdTe doped with Indium and CdZnTe. All samples were grown by the Vertical gradient freeze method directly from the melt [3], where In doping was added to the samples. The doping concentration was 5x10¹⁵ cm^-3. Electrical conductivity – mobility was measured using the Van der Paw method. The data tabulated in Table 1 shows the electrical properties of the samples and the annealing conditions of each sample.

Table 1. Samples specifications No Specifications of the samples Sample Annealing Resistivity BNG1 CdZnTe 15%Zn, undoped after Cd-rich annealing 24 hours, 700/610℃ n-typeinhomogeneous BNG5 CdZnTe 15%Zn, undoped as-grown high resistivity BNG6 CdZnTe 15%Zn, undoped after Cd-rich annealing 48 hours, 600/430℃ n-type,n=1.4e15cm-3, low resistivity ro=3.1 Ohm.cm E33C3C CdTe doped with In as-grown n-type, n=4.2e15cm-3, low resistivity ro=1.0 Ohm.cm E33K2E CdTe doped with In after Te-rich annealing p-type, p=1.1e16cm-3, ro=5.7 Ohm.cm

Raman spectroscopy can give information on the composition, the stress-strain state, crystal symmetry and orientation, and crystalline defects in a material. Raman spectroscopy is commonly used to determine the molecular structure of organic and inorganic compounds for contamination analysis, material classification, and stress measurements. Also, to characterize carbon polymorph layers (graphitic vs diamond) or for sense covalent bonding (complexes, metal bonding).

Additionally, to classify material orientation (random vs organized structure) and to identify organic functional groups and often specific organic compounds [4-6]. It requires very little sample preparation and is generally non-destructive. We used the Renishaw Ramascope RM 1000 with an He-Ne laser of 632.8nm excitation wavelength to reduce the fluorescence signature in samples that show strong fluorescence at shorter wavelengths [7, 8]. The laser spot on the surface had a diameter of approximately 1 μm and a power of ~4mW. The spectral resolution of about 0.5cm^-1 was determined by measuring the Rayleigh line of the He-Ne laser.

The laser is focused on the sample and viewed simultaneously through the lens with our USB camera. The advantage is the ideal solution when sample molecules or trace impurities emit fluorescence a condition which can overshadow the effectiveness of Raman spectroscopy. Eliminating or reducing fluorescence emission requires a laser excitation wavelength that does not have the energy to excite molecular fluorescence.

It uses a He-Ne laser of 632.8nm excitation laser to reduce the fluorescence signature in samples that show strong fluorescence at shorter wavelengths.

Spectra can be saved with an automated baseline correction algorithm or in raw form. All spectra are referenced with the laser off to subtract ambient light. The spectral viewer allows users to interpret the most recently acquired spectrum. Spectra may be zoomed or expanded and peak heights and frequencies are reported on the screen.

In Fig. 1-5, the micro-Raman spectra of each sample acquired in the spectral range from 180 to 2000cm^-1, 10sec X 10 accumulations are presented. In addition we did micro-Raman measurements for all the samples in the range of 180-2000 cm^-1, 20sec X 2 accumulations and in the range of 180-500 cm^-1, 20sec X 10 accumulations and of 180-500 cm^-1, 50sec X 10 accumulations.

Figure 1. A micro-Raman spectrum of CdZnTe 15%Zn, undoped, as-grown

Figure 2. A micro-Raman spectrum of CdZnTe 15%Zn, undoped, after Cd-rich annealing (24 hours)

Figure 3. A micro-Raman spectrum of CdZnTe 15%Zn, undoped, after Cd-rich annealing (48 hours)

Figure 4. A micro-Raman spectrum of CdTe doped with In, after Te-rich annealing

Figure 5. A micro-Raman spectrum of CdTe doped with In, as-grown

3. Results and Discussion

First, we studied in details the samples of CdZnTe 15% Zn, in the range of 150-550cm^-1. Their typical Raman peaks were confirmed, according to the literature [9,10]. A possible origin for a Raman peak identified could be a secondary phonon with A1 symmetry and E symmetry of Te inclusions. A minor difference of the vibrational frequencies related to the annealing temperature towards lower frequencies, as shown in the Figs 6-8 was observed.

Figure 6. A micro-Raman spectrum of CdZnTe 15%Zn, undoped, as-grown, in the range 150-550

Figure 7. A micro-Raman spectrum of CdZnTe 15%Zn, undoped, after Cd-rich annealing (24 hours), in the range 150-550

Figure 8. A micro-Raman spectrum of CdZnTe 15%Zn, undoped, after Cd-rich annealing (48 hours), in the range 150-550

Additionally, we can observe from the Figs. 1, 2and 3 that the peak intensity and the area under the (LO) phonon peaks increases and the FWHM values decreases with annealing temperature, which is due to the improvement of crystallinity of the films.

In the Figures 9-10 the Raman spectra for the In doped CdTe in the range of 180-2000cm^-1 are presented. After the Te-rich annealing, changes in the peaks of the CdTeIn samples were observed. Under these conditions defecting scattering by one longitudinal optical (LO) phonon as well as Fröhlich-induced two-LO phonon scattering is observed. We present the results of (LO) phonon frequency as function of the temperature. In both cases scattering is found to be strongly affected by In doped and related to the annealing temperature.

Figure 9. A micro-Raman spectrum of CdTe doped with In, after Te-rich annealing, in the range 180-2000

Figure 10. A micro-Raman spectrum of CdTe doped with In, as-grown, in the range 180-2000

As shown in the Figures 9 and 10, the peaks of triangle Te inclusions in Te-rich annealing wafers are shifting to high energy after Te-rich annealing, and this is in good agreement with the literature. [11] This shows the compressive stress exists around Te inclusions.

The Raman measurements in our wafers revealed that in those with Te-rich annealing, the Te aggregates, at least part of them, formed particles with a crystalline structure. The crystals are most likely to deposit in the form of inclusions around the grain boundaries and dislocations. Grain boundary movement and dislocations slipping could produce a stress field resulting in the formation of a compressive stress around the Te inclusions.

4. Concluding Remarks

It is well known that the resistivity of CdTeIn strongly depends on the In doping concentration and growth and post-annealing conditions. In this research work we studied In-doped CdTe and CdZnTe after annealing at various temperatures using Raman spectroscopy. We present the results of (LO) phonon frequency as function of the temperature.

Additional changes are observed in the spectral range above 200 cm^-1, where the clean CdTe surface shows the 2LO multiphonon structure at 314 cm^-1 in accordance with bibliography [9-10]. The increase in the peak intensity and the decrease in FWHM of the Raman peak for the annealed films indicate the improvement in the crystallinity of the films during thermal annealing.

ACKNOWLEDGEMENTS

This research work is supported by the Greek General Secretariat of Research and Technology (project code 11CZ_32_ΕΤ33, 7497/19-6-12) Joint Research and Technology Programme Greece-Czech Republic 2011 – 2013. The authors would like to express their special thanks to Dr. M. Perraki for the Micro-Raman measurements.