**3. Methods of preparation and investigation of II-VI films**

508 Advanced Aspects of Spectroscopy

or at *g=*1 - <sup>1</sup> . <sup>4</sup>

Than

 

*Fm*

*d N e dE kT*

*F E*

Thus, building the function 1/*ed*

found from the function 1/*ed*

it is easy to obtain

The value of the last function at the maximum (*EF=Et*) is

.

*t*

<sup>2</sup>

(24)

*t*

<sup>1</sup> , 1 *Fm*

*d gN e dE kT g*

*F E*

*/dEF* - *EF* and finding the maximums by using (24) gives the

(25)

 *- EF*. In case of the mono-level where the Fermi quasi-

*N h E dE* . Under reconstruction of

*/dEF* - *EF* and at low

*t E*

) the relationship

*n*

2

  */dEF* - *EF* may

*t*

*N*

*<sup>g</sup>* ,

 

<sup>1</sup> 4 .

*e dE* 

concentration of discrete monoenergetical levels. The energy position of the maximum

If the LS monotonically distributed by energy *h*(*E*)*=ANt=* const are in the gap of the material

<sup>1</sup> () . *<sup>t</sup>*

In other words the trap concentration in the sample under such distributions is immediately

In general case when LS distribution in the gap of the material is described by the arbitrary

such distributions from the SCLC CVC these distributions are energetically broadening depending on the temperature of experiment [43-45]. LS energy positions are again

*<sup>d</sup> N hE*

*t*

immediately produces energy positions of these levels.

temperatures can be found from the relationship

determined by the maximums of the curve.

If the LS distribution is a Gaussian function (

be checked out by using the function

then (1 )[ ] *Fm*

*/dEF* - *EF*.

function their concentration is defined by the area under the curve 1/*ed*

The correct determination of the trap concentration from the dependence 1/*ed*

level coincides with the LS energy position, it is easy to obtain from (22) - [ ] <sup>1</sup> *Fm*

Earlier [43-45] we have described the effect of experimental factors on accuracy of determining parameters of the deep centers by IC method. In Ref. [44, 45, 48] it was shown

*N gn t t <sup>E</sup>* . If *g*=1 then [] 2 *Fm tE t n N* , [2 ] *Fm*

for determination of *Nt* is analogous to that described above.

*<sup>d</sup> N kT*

*Fm*

*F*

2

1 ( ) *E t E*

*N n t tE* .

*N EE h E*

2 2 *t t t t*

1 2 ( ) exp

 

*e dE* 

*F E*

Thin films CdTe, ZnS, ZnTe were prepared on glass substrates in vacuum by close-spaced vacuum sublimation (CSVS) [49-50]. For further electrical investigations we have deposited hard-melted metal conductive layers on the main substrate by electron beam evaporation (Mo – for CdTe, ZnS; Cr, Ti – for ZnTe). The up-source contact (In(Ag) or Cr in dependence on the conductivity type of the semiconductor) was deposited by the vacuum thermal evaporation. Under condensation of the films of binary compounds the chalcogenide stoichiometric powders were used.

The common temperature of the evaporator was *Т<sup>e</sup>* = 973 К for zinc telluride, *Тe* = (1200÷1450) K for zinc sulfide and *Тe* = (933÷1023) K for cadmium telluride. The substrate temperature was changed in a wide range *Т<sup>s</sup>* = (323÷973) К. Time of deposition was varied: *t* = (10÷30) min.

Morphology of the samples' surfaces was investigated by optical and electron microscopy. Jeffries' method was used to determine the arbitrary grain size (*D*) in the condensates. The films' thickness (*d*) was measured by fractography and interferential methods. The element composition of the layers was studied by X-ray spectroscopy (XRS) analysis using the energy-dispersed X-ray analysis (EDAX) unit or by Rutherford back scattering (RBS) technique (if it was possible). Structural examinations of the films were carried out by the XRD-unit in Ni-filtered Kα radiation of Cu-anode. The XRD patterns were registered in the range of Bragg angles from 200 to 800. Phase analysis was provided by comparison of interplane distances and arbitrary intensities from the samples and the etalon according to the ASTM data [51]. Structural properties of II-VI films are investigated in [20, 49-50, 54-56].

Dark CVC at different temperatures and  *– Т* dependencies of the sandwich-structures (MSM were examined in vacuum by standard techniques (Fig. 1) [21-22].

The power of electronic scheme was estimated by source of stable voltage AIP 120/0.75 that provided a possibility of precise voltage regulation in electric circle in the ranfe of *U* =0.1 ÷ 120 V.

A current that passed throught samples in the range of *I* = (10-9÷10-5) A measured by digital nanoampermeter. Voltage drop on sample was fixed by digital multimeters APPA-108N and UT70B. Sample temperature at inveatigation of electro-physical properties controlled by authomatic feedback temperature controller "OVEN TRM10", input signal from it fed from chromel-alumel thermocouple.

**Figure 1.** Typical electrical-type scheme for CVC and *σ-T* characteristic investigations of II-VI semiconductors films: 1 – heater holder; 2 – heater; 3 – glass substrate; 4 – lower conductive layer (Мо, Cr, Ti); 5 – collectors; 6 – thermocouple; 7 – II-VI film

The current mechanisms were identified by the differential method developed in [52-53]. This technique completely analyses *j–U, γ–U* and d(log)/d(log*U*)-*U* functions, where =d(log*j*)/d(log*U*) and differentiates satellite and concurrent current mechanisms in the structures and defines the high-field mechanisms among all of them. When the CVCs of multilayered structures were determined by unipolar injection from the source contact the experimental curves were numerically studied by using low-temperature and hightemperature approximations of the IS method [43-45, 48].

PL spectra of CdTe, CdSe and ZnTe films were studied using the spectrometer SDL-1 undеr excitation of the samples by Ar-laser (*λ*=514 nm for CdTe and *λ*=488,8 nm for ZnTe). PL spectra from ZnS films are registered by MPF-4 Hitachi and xenon bulb (*λ*=325 nm). The temperature in all experiments was stable in the range 4.7÷77 К by using the system "UTREX" [49]. The films CdTe, ZnTe were investigated in the range of edge luminescence, the films ZnS were studied in the impurity energy range.

At interpretation of the PL data it was suggested that the radiation had appeared as a result of electrons' transfer from the conduction (valence) band or shallow donor (acceptor) levels to the deep LS in the gap of the material. Then the activation energy of the processes are defined from the expression:

$$
\Delta E = \hbar \nu = E\_x - E\_i = E\_y - (E\_a + E\_d) \tag{26}
$$

where *Ea, Ed* are energy levels of the donors and acceptors in the gap of the material.

The set of methods for defining parameters of LS in the gap allowed to enhance the accuracy of data obtained and to examine traps and recombination centers with wide energy range.

#### **4. Determination of LS parameters of polycrystalline chalcogenide films by injection spectroscopy method and analysis of**  *– T* **functions**

#### **4.1. General description of CVC and**  *– T* **functions**

510 Advanced Aspects of Spectroscopy

chromel-alumel thermocouple.

120 V.

The power of electronic scheme was estimated by source of stable voltage AIP 120/0.75 that provided a possibility of precise voltage regulation in electric circle in the ranfe of *U* =0.1 ÷

A current that passed throught samples in the range of *I* = (10-9÷10-5) A measured by digital nanoampermeter. Voltage drop on sample was fixed by digital multimeters APPA-108N and UT70B. Sample temperature at inveatigation of electro-physical properties controlled by authomatic feedback temperature controller "OVEN TRM10", input signal from it fed from

**Figure 1.** Typical electrical-type scheme for CVC and *σ-T* characteristic investigations of II-VI semiconductors films: 1 – heater holder; 2 – heater; 3 – glass substrate; 4 – lower conductive layer

The current mechanisms were identified by the differential method developed in [52-53].

=d(log*j*)/d(log*U*) and differentiates satellite and concurrent current mechanisms in the structures and defines the high-field mechanisms among all of them. When the CVCs of multilayered structures were determined by unipolar injection from the source contact the experimental curves were numerically studied by using low-temperature and high-

PL spectra of CdTe, CdSe and ZnTe films were studied using the spectrometer SDL-1 undеr excitation of the samples by Ar-laser (*λ*=514 nm for CdTe and *λ*=488,8 nm for ZnTe). PL spectra from ZnS films are registered by MPF-4 Hitachi and xenon bulb (*λ*=325 nm). The temperature in all experiments was stable in the range 4.7÷77 К by using the system "UTREX" [49]. The films CdTe, ZnTe were investigated in the range of edge luminescence,

)/d(log*U*)-*U* functions, where

(Мо, Cr, Ti); 5 – collectors; 6 – thermocouple; 7 – II-VI film

This technique completely analyses *j–U, γ–U* and d(log

temperature approximations of the IS method [43-45, 48].

the films ZnS were studied in the impurity energy range.

Dark CVC of sandwich structures current-conductive substrate-film-upper drain contact were measured at different temperatures for examining electrical properties of Zn and Cd chalcogenide films and determination of parameters for LS in the gap of material. Besides that, the function conductivity-temperature was studied in ohmic sections of the CVC and in some cases in the square section of the CVC. Energy positions of donor (acceptor) centers in the films were found from dependencies log *=*ƒ*(*103*/T)* taking into account their Arrheniuslike character [21-22].

As was shown by the study, the CVC of multilayered structures MSM is defined by the condensation conditions of chalcogenide films, their crystal structure, and material of bottom and upper metallic contacts. CVC of multilayered structures based on lowtemperature condensates of II-VI compounds were linear or sublinear. For ZnTe-based MSM structures the CVC were defined by the Pool-Frenkel mechanism, and the data were linearized in the coordinates 1 2 log *IU U* [52].

Fig. 2 plots typical double-log CVC measured at different temperatures. This figure also shows the function  *– T* measured at the ohmic section of the CVC.

It is found out that the  *– T* function of low-temperature condensates are linear with the slope to the *T* axis decreasing at lowering the measurement temperature. These features are typical for the material with various types of donor (acceptor) impurities with different activation energy. The CVC of high-temperature condensates were somewhat others (Fig. 2). The linear sections are reveled, their slope to the *T* axis increases as the measurement temperature decreases. It is typical for compensated materials [21-22]. The compensation effect appears more visible under sufficiently low experimental temperatures when the electron concentration becomes close to that of acceptor centers. The slope of the straight lines to the *T*-axis increases from the value *Ea*/2*k* up to the value *Ea*/*k*, making it possible to define activation energy for donor and acceptor centers [21-22].

**Figure 2.** CVC of the structure Cr/ZnTe/Ag at various temperatures: ● – *Т* = 298 К; ▲ – *Т* = 303 К; ▼ – *Т* = 308 К; ► – *Т* = 313 К; – *Т* = 318 К; \* – *Т* = 323 К, and the dependence log – 1/*T* obtained from the ohmic section of the CVC. The film is prepared at *Т<sup>e</sup>* = 973 К and *Т<sup>s</sup>* = 823 К

CVC of multilayered structures where chalcogenide films are prepared at *Ts* > (500÷600) K were superlinear. As is analytically shown, they are determined by the unipolar injection from the drain contact. Typical SCLC CVCs of the examined films are plotted in Figs. 2-3. CVCs of high-temperature condensates in the range of high field strength a set of linear sections with various slopes to the *U*-axis was observed. As a rule, the sections with functions: *I – U*, *I – U 2*, *I – U 3-5*, *I – U 8-10* were the most pronounced. In some cases after superlinear sections we have observed a square dependence *I* on *U*, which had further changed again to the supelinear one with a very large slope ( 13–25). The current jump was revealed and the samples were turn on the low-ohmic state as an irreversible process.

**Figure 3.** Double-log SCLC CVC of multilayered structures Mo/CdTe/Ag and results of their differentiation. CdTe films are prepared at *Te* = 893 К and various *Ts*: 723 К (а); 823 К (b)

The features of the CVCs are clearly shown in functions – log*U* giving a possibility to reveal a fine structure of the CVCs (Fig.3). Each point of this graph defines the slope of the CVC in double-log scale to the voltage axis. Dependencies – log*U* were obtained by differentiating the CVC in every experimental point. As it was mentioned above, the problem mathematically reduces to the building smoothing cubic spline which approximates experimental data and its differentiation at the sites.

512 Advanced Aspects of Spectroscopy

process.

**Figure 2.** CVC of the structure Cr/ZnTe/Ag at various temperatures: ● – *Т* = 298 К; ▲ – *Т* = 303 К; ▼ – *Т*

CVC of multilayered structures where chalcogenide films are prepared at *Ts* > (500÷600) K were superlinear. As is analytically shown, they are determined by the unipolar injection from the drain contact. Typical SCLC CVCs of the examined films are plotted in Figs. 2-3. CVCs of high-temperature condensates in the range of high field strength a set of linear sections with various slopes to the *U*-axis was observed. As a rule, the sections with functions: *I – U*, *I – U 2*, *I – U 3-5*, *I – U 8-10* were the most pronounced. In some cases after superlinear sections we have observed a square dependence *I* on *U*, which had further

jump was revealed and the samples were turn on the low-ohmic state as an irreversible

**Figure 3.** Double-log SCLC CVC of multilayered structures Mo/CdTe/Ag and results of their differentiation. CdTe films are prepared at *Te* = 893 К and various *Ts*: 723 К (а); 823 К (b)

 (

– 1/*T* obtained from the

13–25). The current

= 308 К; ► – *Т* = 313 К; – *Т* = 318 К; \* – *Т* = 323 К, and the dependence log

ohmic section of the CVC. The film is prepared at *Т<sup>e</sup>* = 973 К and *Т<sup>s</sup>* = 823 К

changed again to the supelinear one with a very large slope

The curves  *–* log*U* resulting from the working-out of the SCLC CVC showed 1-4 maximums in correspondence to the sections of sharp current increase in the *I – U* dependencies. The most often values of ext were 8–10. Sometimes the functions  *–* lоg*U* were practically revealed.

Horizontal sections with the almost constant slope  *>* 2 were also observed. It may be explained by the presence in the samples of sets of monoenergetical or quasimonoenergetical levels traps of various energy position and concentration or by availability of the exponential (or other form) LS energy distribution. The specific points of CVCs were used for calculating trap parameters in the material, the ohmic sections helped to find specific conductivity of the layers = (104÷105) m. As a result we obtained the concentrational distributions of the traps in the gap of material *h*(*E*)-*E*, their energy position (*Еt*) and concentration (*Nt*).

At the high voltage the CVCs are typical for the unipolar injection, but, according to [52-53] there some other current mechanisms leading to qualitatively similar current-voltage functions. Thus, we had to identify them additionally according to the procedure described in [52] by analyzing functions log*I –* log*U*,  *–* log*U* and log *–* log*U*. It allowed identifying high-voltage current mechanisms in the samples and defining (in some cases) their type.

For further definition of the dominant current mechanism in the base chalcogenide layer we calculated the discrimination coefficient *Qext* in the extremum points of the function  *–* log*U* and compared it with coefficients typical for other mecahisms [52]. We have found *Qext* > 106÷107 almost in all cases, what is significantly larger than the values of *Qext* typical for the field trap ionization and the barrier –involved current mechanism in the material. This, in turn, points out [52-53] that the extremums in functions  *–* log*U* are caused by filling-in of the traps in the material with charge carriers injected from the metallic contact. Using various analytical methods allows to conclude with a good reliability that the CVC's features for multilayered structures with high-temperature chalcogenide layers (*Ts* > 500 К), were caused namely by the SCLC mechanism. Further we have worked out the CVCs due to injection currents only.

Fig. 4. illustrates a typical example of the CVC working-out. It is easy to see that the LS distributions are obtained under analysis of two different CVCs and they are in a good correlation.

To make the distribution more precise we have plotted in the same picture the Gaussian curve. It is seen that for examined polycrystalline CdTe films there are trap distributions in the gap with a form closed to that of the Gaussian one with a small half width *t*. Broadening energy levels in CdTe layers prepared by the vacuum condensation may be due to statististical dispersion of polar charge carriers' energy caused by fluctuative irregularities of the film crystalline lattice. This effect is enhanced near the substrate where the most defective layer of the film is grown. This region was an object for determining LS parameters by the method of SCLC CVC.

**Figure 4.** SCLC CVC and its derivative (*U*) for CdTe-based sandwich structures (а), and trap distribution in the gap of cadmium telluride (b): – *j(U)*; ▲ – (*U*) (а); the energy trap distribution is resulted from the high-temperature IS method (b) ( – first measurement; - repeating measurement at somewhat other temperature); the Gaussian distributions (solid line) are presented for comparison.

#### **4.2. LS parameters from CVC and**  *– T* **functions**

SCLC CVC was used for determination of trap parameters in the films. The low level of scanning LS spectrum was defined by the position of the equilibrium Fermi level *EF0*, i.e. its position without charge carrier injection in the sample (ohmic section of the CVC), the upper limit was defined by the position of the Fermi quasi-level at the turn-on of the multilayered structures into low-ohmic state. The start position of the Fermi level was pre-defined by the equilibrium carrier concentration in the material, respectively, by the conductivity of the films. The calculations showed the position of the equilibrium Fermi level *EF0* was coincide or was close to the energy of the deepest LS in the corresponding samples. The Fermi level is fixed by the traps because the concentration of free carriers in the films is close to the full concentration of LS located at grain boundaries and in bulk crystallites of condensates. As a result, the deepest trap levels located lower than the energy of the equilibrium Fermi level were not revealed in chalcogenide films by the SCLC CVC method.

The possibility of revealing shallow traps in the samples (*Et*  0.21 eV, for ZnTe films) is restricted by their turn-on into the low-ohmic state stimulated namely by the LS. Thus, the SCLC CVC method had revealed the traps with energy higher positions. However, the traps with different energies also may exist in the samples as shown by the data from the slope of conductivity-temperature functions in ohmic and square sections of the CVCs and luminescence spectra.

#### *4.2.1. CdTe films*

514 Advanced Aspects of Spectroscopy

parameters by the method of SCLC CVC.

**Figure 4.** SCLC CVC and its derivative

**4.2. LS parameters from CVC and** 

were not revealed in chalcogenide films by the SCLC CVC method.

distribution in the gap of cadmium telluride (b): – *j(U)*; ▲ –

(*U*) for CdTe-based sandwich structures (а), and trap

(*U*) (а); the energy trap distribution is

resulted from the high-temperature IS method (b) ( – first measurement; - repeating measurement at somewhat other temperature); the Gaussian distributions (solid line) are presented for comparison.

 *– T* **functions** 

SCLC CVC was used for determination of trap parameters in the films. The low level of scanning LS spectrum was defined by the position of the equilibrium Fermi level *EF0*, i.e. its position without charge carrier injection in the sample (ohmic section of the CVC), the upper limit was defined by the position of the Fermi quasi-level at the turn-on of the multilayered structures into low-ohmic state. The start position of the Fermi level was pre-defined by the equilibrium carrier concentration in the material, respectively, by the conductivity of the films. The calculations showed the position of the equilibrium Fermi level *EF0* was coincide or was close to the energy of the deepest LS in the corresponding samples. The Fermi level is fixed by the traps because the concentration of free carriers in the films is close to the full concentration of LS located at grain boundaries and in bulk crystallites of condensates. As a result, the deepest trap levels located lower than the energy of the equilibrium Fermi level

The possibility of revealing shallow traps in the samples (*Et*  0.21 eV, for ZnTe films) is restricted by their turn-on into the low-ohmic state stimulated namely by the LS. Thus, the SCLC CVC method had revealed the traps with energy higher positions. However, the traps with different energies also may exist in the samples as shown by the data from the slope of

Broadening energy levels in CdTe layers prepared by the vacuum condensation may be due to statististical dispersion of polar charge carriers' energy caused by fluctuative irregularities of the film crystalline lattice. This effect is enhanced near the substrate where the most defective layer of the film is grown. This region was an object for determining LS

> Table 1 presents some results of IS calculations for deep centers in polycrystalline and monocrystalline CdTe films. In the gap of the polycrystalline material are LS with *E1* = (0.68÷0.70) eV*; E2* = (0.60÷0.63) eV; *E3* = (0.56÷0.57) eV; *E4* = (0.51÷0.53) eV; *E5* = (0.45÷0.46) eV; *E6* = (0.39÷0.41) eV and concentration *N* = (1018÷1020) m-3. The concentration of these LS is in the range *Nt* = (1018 ÷1021) m-3 and mostly increases with closing their energy positions to the bottom of the conduction band. The traps by the profile 2 2 ( ) 2 exp 2 *t t hE N E* are similar to the mono-energetical ones with a half width *<sup>t</sup>* (0.011÷0.015) eV. The dominant LS affecting SCLC CVC are the LS with energies *Еt* = (0.60÷0.63) eV; *Еt* = (0.56÷0.57) eV; *Еt* = (0.45÷0.46) eV. Only the traps (if revealed) with *Еt* = 0.40 eV had the larger concentration.

> The LS were registered not only in polycrystalline films but also in monocrystalline layers. We have resolved the traps with *Еt* = (0.56÷0.57) eV; *Еt* = (0.52÷0.53) eV; *Еt* = (0.45÷0.46) eV and *Еt* = (0.40÷0.41) eV in the gap of the material. The monocrystalline condensates had lower resistance that the polycrystalline layers (10÷100 times), the equilibrium Fermi level in these films was placed more closely to the conduction (valence) band than that in polycrystalline films. Thus, the deepest traps were not revealed by SCLC CVC method in monocrystalline layers. So, the traps *Et*  0.70 eV and *Et*  0.62 eV found in polycrystalline films may be presented in lower-resistive monocrystalline films.

> Ionization energies of the defects in the gap of CdTe were determined from the slope of functions conductivity-temperature in coordinate's log-1/*T* [21-22]. Table 2 lists the results for polycrystalline and monocrystalline CdTe films. In high-temperature polycrystalline condensates the following activation energies were observed for conductivity: *Et*=0.15; 0.33; 0.40÷0.41; 0.46; 0.60÷0.61, 0.80 eV. In the monocrystalline films the LS had smaller activation energy: *Et*=0.06÷0.07; 0.13÷0.14; 0.22÷0.23; 0.29; 0.40; 0.46 eV. Activation energy *Et* = (1.50÷1.52) eV is typical for high temperatures of the experiment and corresponds to the gap of the material. The comparison of the LS energy levels from the SCLC CVC and  *– Т* functions is carried out in Table 2. The values *Et* from the  *–Т* functions correlate with those observed in CdTe films by SCLC CVC method.

> The wide range of the traps revealed in CdTe condensates is obviously caused by investigation of disordered transition layer of the films formed under the film condensation near the substrate. In this layer may be presented foreign impurities adsorbed from the substrate and residual atmosphere under film condensation. Besides that for CdTe the concentration of uncontrolled residual impurities in the charge mixture can be *Nt* =(1020-1021) m-3 which is behind the sensitivity of the IS method. These impurities can form a chain of complexes impurity-native defect producing deep levels in the gap of the semiconductor.



**Table 2.** Energetical positions of LS levels for defects in the gap of CdTe

As the chalcogenide films were not doped in-advance all LS found here are corresponding to native defects and their complexes with uncontrolled impurities. The interpretation is a challenge while the energy spectrum of PD in the gap of tellurium is studied not enough and identification in the most cases is not satisfactory (Table 3). For example, in [62] the levels *Е<sup>t</sup>* of LS are studied by photoinduced currents (PICTS) and authors give more than 150 values of deep levels, where the sufficient part of them is caused by the native defects. More reliables are some theoretical works where energies *Еt* are calculated («*ab anitio*») [57- 61]. We have used namely the data Wei [57-58] obtained from the first principles. Table 3 summarizes our results.

According to calculations the deep centers with energy position 0.71 eV are belonging to 2 *VTe* . We have experimentally observed the level *Et* = (0.68÷0.70) eV which may be caused by this defect. Analogically, the LS with energies (0.60÷0.63) eV may be ascribed to the antistructural defect 2 *TeCd* (0.59 eV), and (0.56÷0.57) eV and (0.45÷0.46) eV to the interstitial cadmium in different charge states: *<sup>c</sup>* <sup>2</sup> *Cdi* (0.56 eV), *<sup>c</sup> Cdi* (0.46 eV). The level 0.29 eV is also formed by the native defect bound with cadmium *<sup>a</sup>*<sup>2</sup> *Cdi* (0.33 eV). Different ionization energies of interstitial cadmium are due to its place in octo- or tetrahedral position in the crystal lattice of the material.

#### *4.2.2. ZnTe films*

516 Advanced Aspects of Spectroscopy

2 (1st measurement)

2 (2nd measurement)

6 (monocrystalline)

Polycrystalline films

Sample number *d*, m *Тs*, К *Тe*, К *Et*, eV *Nt*, m-3

3 12 748 968 0.68

4 9 723 893 0.62

5 12 823 893 0.62

7 15 753 953 0.60

8 26 758 978 0.61

0.68-0.70 0.80 - <sup>2</sup> *VTe*

0.60-0.63 0.60 - <sup>2</sup>

0.56-0.57 0.57 - *<sup>c</sup>*<sup>2</sup> *Cdi*

0.51-0.53 - - <sup>2</sup> *VTe*

0.45-0.46 0.46 0.46 *<sup>c</sup> Cdi*


Interpretation From SCLC CVC From

*-T* dependencies

Monocrystalline films

**Table 1.** Parameters of LS revealed in CdTe films by high-temperature IS

*Et*, eV

Polycrystalline films

0.39-0.40 0.40÷0.41 0.40



1 8 743 1023 0.63 4.41019 0.030

0.45

0.45

0.62 0.53

0.56

0.57

0.52 0.41

0.52 0.46 0.41

0.56 0.52

19 748 948 0.61

19 748 948 0.62

11 753 933 0.62

*<sup>t</sup>*, eV

0.031 0.028

0.035 0.032

0.023 0.023 0.027

0.021 0.016

0.023 0.015

0.019 0.009 0.016

0.019 0.020 0.020 0.015

0.023 0.015 0.015

(0.71 eV [57-61]

(0.56 eV) [57-61]

(0.50 eV) [57-61]

(0.46 eV) [57-61]

(0.20 eV) [57-61]

[57-61]

(0.40 eV)

<sup>2</sup> *TeCd*

. *VTe*

*Cd Te* (0.59 eV) [57-61]

1.71019 7.31019

1.51019 8.11019

7.81018 1.51019 6.11019

6.61018 4.41019

2.01018 1.71019

4.61018 1.31019 1.11020

2.31018 3.61018 8.61018 1.41019

3.61018 3.01019 7.41019

Table 4.3 summarizes the results of calculations for ZnTe condensates in dependence on physical technical growth conditions. SCLC CVC method reveals set of trap groups with the most probable energy position *Еt*1 = 0.21 eV; *Еt*2 = (0.32÷0.34) eV, *Еt*3 = 0.57 eV; *Еt*4 = (0.41÷0.42) eV; *Еt*5 = 0.89 eV. The concentration of the revealed LS is in interval *Nt* = (1020 ÷1021) m-3. The LS with energy *Е<sup>t</sup>* = (0.32÷0.33) eV are dominant in the most samples and they determine the CVCs of the films.

The trap spectrum in ZnTe films can be partially checked by investigation of temperature – conductivity functions for the condensates. As shown by analysis of  *– Т* functions in the Ohmic section of the CVC for high-temperature ZnTe condensates the following conductivity activation energies are typical: 0.05 eV; (0.14÷0.15) eV; (0.20÷0.21) eV; (0.33÷0.34) eV; (0.42÷0.43) eV; (0.51÷0.52) eV; (0.57÷0.58) eV; (0.69÷0.70) eV and 0.89 eV (Table 4). Set of *Et* values from the  *– Т* functions is in a good correlation with those in ZnTe films defined by SCLC CVC method (Table. 3) and low-temperature luminescence (Table. 2).

As the films ZnTe as CdTe layers were not doped in-advance all the calculated LS are due to native PD, their complexes, uncontrolled impurities and their complexes with native defects.

The LS in monocrystals and films ZnTe were studied by SCLC CVC in [24, 63, 64]. Authors [24] have found the trap parameters in monocrystalline samples by the voltage of complete trap filling-in: *Et* = 0.17 eV and *Nt* = 1022 m-3. On the other hand, measurements of  *– Т* dependencies in the square section of the CVC gave *Et* = 0.14 eV and *Nt* = 1023 m-3 taking in mind the presence of traps in the material authors [64] calculated the LS density in ZnTe

films prepared by laser evaporation technique: *Nt* = (4.2÷8.4)1022 m-3. The trap energy is not defined in this work. In ZnTe films obtained by the electro-deposition the trap concentration is *Nt* = 3.61021 m-3 [63].


**Table 3.** LS parameters in ZnTe films from SCLC CVC and *-Т* functions

As is seen from the Table 3, the trap concentration in ZnTe films is significantly lesser than that in condensates prepared by laser evaporation, electro-deposition methods and even in the monocrystalline material [23, 63- 64]. It shows a high structural perfectness and stoichiometry of the layers.

Nevertheless, the most levels found in ZnTe films may be identified with some probability. The level *Е<sup>1</sup>* = 0.05 eV is commonly bound with single-charged dislocation *VZn* , and the level *Е<sup>2</sup>* = 0.15 eV is bound with a double-charged 2 *VZn* Zn vacancy [65, 66]. In later works the second level is ascribed to Cu as to a traditional residual impurity in ZnTe, and the doublecharged Zn vacancy is supposed to have a more deeper energy level 0.21 eV [66]. It is thought that the energy activation (0.36÷0.40) eV [67, 68] is for the common substitution impurity in ZnTe, namely *OTe*. The most deepest level 0.58 eV authors [67] ascribe to the Te vacancy 2 *VTe* (interstitial zinc <sup>2</sup> *Zni* ). The possible interpretation of LS in ZnTe films is listed in table 5.7. Other energy levels on our opinion are belonging to the uncontrolled impurities and complexes native defect-impurity.

We have revealed trap levels with energy position *Еt* = (0.22 – 0.25) eV and concentration *Nt* = (5.01020 ÷ 1.51021) m-3 by the analysis of SCLC CVC in ZnS films. These LS may be localized in the gap due to presence of the interstitial Zn atom <sup>2</sup> *Zni* . LS with energy position *Еt* = (0.24÷0.25) eV were observed in [70] the thermo stimulated current technique.


**Table 4.** Energy positions of LS for defects in ZnTe gap determined from the slope of *–1/Т* functions

According to the Arrhenius equation  *- Т* functions allowed calculating conductivity activation energies in linear sections: *Е*1=0.03 eV; *Е*2 = (0.07÷0.08) eV, *Е*3 = 0.15 eV; *Е*4 = (0.23÷0.24) eV; *Е*5 = 0.33 eV; *Е*6 = 0.46 eV; *Е*7 = 0.87 eV.

#### *4.2.3. ZnS films*

518 Advanced Aspects of Spectroscopy

is *Nt* = 3.61021 m-3 [63].

Parameters of film condensation

> *Т<sup>s</sup>* = 623 К, *Te* = 973 К

> *Т<sup>s</sup>* = 673 К,

*Т<sup>s</sup>* = 723 К, *Te* = 973 К

*Т<sup>s</sup>* = 773 К,

*Т<sup>s</sup>* = 823 К,

*Т<sup>s</sup>* = 873 К

*Т<sup>s</sup>* = 623 К *Te* = 973 К

stoichiometry of the layers.

vacancy 2 *VTe*

films prepared by laser evaporation technique: *Nt* = (4.2÷8.4)1022 m-3. The trap energy is not defined in this work. In ZnTe films obtained by the electro-deposition the trap concentration

*Te* = 973 <sup>К</sup> 0.33 5.31020 - -

*Te* = 973 <sup>К</sup> 0.32 5.31020 - -

*Te* = 973 <sup>К</sup> 0.42 2.11020 - -

*Te* = 973 <sup>К</sup> 0.32 1.51021 - -

As is seen from the Table 3, the trap concentration in ZnTe films is significantly lesser than that in condensates prepared by laser evaporation, electro-deposition methods and even in the monocrystalline material [23, 63- 64]. It shows a high structural perfectness and

Nevertheless, the most levels found in ZnTe films may be identified with some probability.

second level is ascribed to Cu as to a traditional residual impurity in ZnTe, and the doublecharged Zn vacancy is supposed to have a more deeper energy level 0.21 eV [66]. It is thought that the energy activation (0.36÷0.40) eV [67, 68] is for the common substitution impurity in ZnTe, namely *OTe*. The most deepest level 0.58 eV authors [67] ascribe to the Te

in table 5.7. Other energy levels on our opinion are belonging to the uncontrolled impurities

We have revealed trap levels with energy position *Еt* = (0.22 – 0.25) eV and concentration *Nt* = (5.01020 ÷ 1.51021) m-3 by the analysis of SCLC CVC in ZnS films. These LS may be

*Еt* = (0.24÷0.25) eV were observed in [70] the thermo stimulated current technique.

8.81020

*-Т* functions

0.35 0.37

The level *Е<sup>1</sup>* = 0.05 eV is commonly bound with single-charged dislocation *VZn*

localized in the gap due to presence of the interstitial Zn atom <sup>2</sup> *Zni*

**Table 3.** LS parameters in ZnTe films from SCLC CVC and

*Е<sup>2</sup>* = 0.15 eV is bound with a double-charged 2 *VZn*

(interstitial zinc <sup>2</sup> *Zni*

and complexes native defect-impurity.

From SCLC CVC From

*Et*, eV *Nt*, m-3 *Et*, eV *Nt*, m-3


0.34 2.91020 0.33 4.11020 - - 0.57 5.51020 - - 0.89 8.41020

*-Т* dependencies


Zn vacancy [65, 66]. In later works the

). The possible interpretation of LS in ZnTe films is listed

, and the level

. LS with energy position

Table 5 summarizes LS parameters calculated by SCLC CVC method and from the  *- Т* functions in ZnS condensates prepared under various physical technical conditions. Reference data are presented for comparison. The table shows a correlation between our results and data obtained by other authors [71-74]. Besides that, there is a coincidence of defect energy positions defined from the SCLC CVC and  *- Т* functions.


**Table 5.** LS parameters defined by analysis of SCLC CVC and  *- Т* functions at ohmic section of the CVC for ZnS films prepared under various physical technical condensation modes

All the LS found here were not identified because of absence of corresponding reference data. Only the levels with activation energy *Е1* = 0.15 eV and *Е2* = (0.22÷0.25) eV may be bound with single- *Zni* and double charged 2 *Zni* interstitial Zn atom.
