RU2431216C1 - Method of determining ionisation energy of deep levels in semiconductor barrier structures and device for realising said method - Google Patents

Abstract

FIELD: physics.

SUBSTANCE: sample of a semiconductor barrier structure is put into a measurement cell. Temperature dependency of relaxation time upon external action of reverse-bias voltage pulses with amplitude V₁ is determined. The signal obtained from the sample is multiplied by a reference signal F(t). Selection based on relaxation time is carried out and the maximum output voltage ΔU and temperature of the maximum of the spectrum peak DLTS (T_max) is determined. An additional reverse-bias voltage pulse with amplitude V₂>V₁ is used, where |V₂-V₁| >>k T_max/e. Two amplitude values of spectrum peaks DLTS ΔU₁ and ΔU₂ which correspond to the same constant relaxation time and different amplitude of the reverse-bias voltage V₁ and V₂ are determined. The ionisation energy of the deep level (AE) is determined using the disclosed formula. The device which realises this method has a square-pulse generator of an irregular shape connected to a deep level relaxation spectroscopy device with possibility of varying temperature of the analysed sample.

EFFECT: improved method and device.

2 cl, 2 dwg

Description

The invention relates to the field of electronic technology, micro- and nanoelectronics and can be used to study the energy spectrum of electronic states, the study of defects with deep levels in semiconductor materials and barrier structures.

There is a method of studying the energy spectrum of electronic states and defects with deep levels in semiconductor barrier structures by the method of relaxation spectroscopy of deep levels (RSGU), based on the study of the temperature dependence of relaxation of electric current, charge or capacity of the barrier structure [1-4].

The disadvantages of this method are:

1. The ionization energy of deep levels is determined by the slope of the Arrhenius graph, which is plotted at different temperatures under the assumption that the capture cross section (σ) is independent of temperature. In fact, σ can depend on temperature, which introduces an additional error in determining the ionization energy.

2. In some cases, to construct an Arrhenius graph, it is necessary to conduct a temperature scan of the sample several times, which increases the measurement time.

Closest to the proposed method (prototype) is the Lang method [1], which consists in the fact that a sample of the semiconductor barrier structure is placed in the measuring cell, the temperature dependence of the relaxation time is determined under external influences by reverse bias voltage pulses with amplitude V ₁ , the received signal from the sample multiply by the reference signal F (t) using the relaxation time selection device (DLTS spectrometer) and determine the maximum output voltage ΔU and the peak temperature maximum of the DLT spectrum S (T _max ). The disadvantage of this method is the obligatory in some cases multiple temperature scanning, since the ionization energy of deep levels in the DLTS method is traditionally determined according to the Arrhenius chart [1]. Devices for implementing the Lang method, known in the English abbreviation as DLTS spectrometers, are manufactured abroad. The domestic industry does not produce such devices.

The proposed method allows to determine the ionization energy of deep levels as a result of a single temperature scan and without building a direct Arrhenius line and eliminate the disadvantage of the prototype.

The essence of the method for determining the ionization energy of deep levels is as follows. In the well-known Lang method [1], namely, that a sample of a semiconductor barrier structure is placed in a measuring cell, the temperature dependence of the relaxation time is determined under external influences by reverse bias voltage pulses with amplitude V ₁ , the obtained signal from the sample is multiplied by the reference signal F (t ), using the relaxation time selection device (DLTS spectrometer) and determine the maximum output voltage ΔU and the peak temperature maximum of the DLTS spectrum, add an additional voltage pulse back th offset with an amplitude of V ₂ > V ₁ , and | V ₂ -V ₁ | >> k T _max / e. The reverse bias voltage pulses V ₁ and V ₂ are alternately applied to the investigated semiconductor structure (Fig. 1a). After each pulse, a current relaxation process is obtained (Fig. 1b). A signal proportional to these processes is multiplied by a reference signal F ₁ (t) (FIG. ₁ c) and F ₂ (t) (FIG. 1 d). As a result, selection is made according to the relaxation time and two values of the amplitude peaks of the DLTS spectrum ΔU ₁ and ΔU ₂ are determined, corresponding to the same relaxation time constant and different amplitude of the pulses of the reverse bias voltage V ₁ and V ₂ .

The concentration of defects with deep levels (deep centers) in the case of the current version of the DLTS method is determined by the amplitude of the peak ΔU of the spectrum of deep levels (GU) using the relation [7]:

where b is a factor depending on the magnitude of the reverse bias voltage on the sample [4];

m is the transmission coefficient of the through path of the DLTS spectrometer;

S is the Schottky barrier contact area;

d is the thickness of the SCR;

θ is the factor taking into account the effect of the SCR boundary shift in the process of current relaxation [7];

λ - the thickness of the layer of incomplete ionization of a deep level, is determined by the formula

We write formula (1) for two cases of the amplitude of the voltage of pulses of reverse biases V ₁ and V ₂ at b = 1 and θ = 1:

Under the condition of a uniform distribution of the space charge of deep and shallow centers in the SCR:

whence after transformations we have:

where d ₁ and d ₂ are the thickness of the SCR for the amplitude of the voltage of the pulses of the reverse bias V ₁ and V ₂ .

Replace in the formula (2) E _F -E _t = (E _C -E _t ) - (E _C -E _F ) = ΔE- (E _C -E _F ) (3)

We use the relation known from the literature for calculating the Fermi level [2]:

From formulas (2), (3), (4), taking into account the fact that n ₀ = N _∂м , we obtain:

.When calculating ΔЕ by formula (5), instead of T we use the value of T _max , which is determined by the location of the peak in the DLTS spectrum. If there are several deep levels in the sample, all information about the ionization energy is obtained as a result of one temperature scan. To calculate the density of states in the conduction band (valence band) N _{C (V), we} use the standard technique. So for silicon

.

The values of d ₁ , d ₂ and N _dm can be easily found from the corresponding CV - dependencies.

Thus, the claimed method meets the criteria of the invention of "novelty", because in known sources, the proposed method for determining the ionization energy of deep levels was not found. Therefore, the sequence of operations in the study of the energy spectrum of electronic states differs from existing ones, and the proposed technical solution has significant differences.

This method is proposed for implementation by scientific laboratories, enterprises and organizations involved in research in the field of micro- and nanoelectronics.

To implement the method, a device is proposed that contains a rectangular pulse generator of complex shape, a measuring cell with the ability to change the temperature of the test sample, and a DLTS spectrometer.

The prototype of a DLTS spectrometer can be, for example, a spectrometer manufactured by Sula Technologies, USA [6].

The invention and possible implementations of the proposed method are illustrated by the following drawings:

figure 1 - timing diagrams explaining the operation of the device;

figure 2 is a structural diagram of a device that implements the proposed method.

The pulse voltage of complex shape (figa) comes from block 1 to block 2 of the measuring cell. The prototype of the measuring cell can be a cryostat from JANIS [5]. Next, the relaxation signal of an electric current, capacitance or charge from block 2 enters block 3, in which the analysis and determination of the amplitudes of the peaks of the DLTS spectrum are carried out at various values of depletion voltage pulses. The rectangular pulse generator (1) contains an additional generator of a sequence of rectangular pulses V ₂ with the possibility of separate adjustment of their amplitude and duration, and in the device of deep-level relaxation spectroscopy (3), the output of the time constant selector is connected to two low-pass filters (LPF) through a switch controlled by a rectangular pulse generator (1).

The technical and economic result is to reduce the measurement time and increase the reliability of information about the parameters of deep centers in semiconductor barrier structures and materials.

Literature

[1] Lang D.V. Deep level transient spectroscopy: a new method to characterize traps in semiconductors // J Appl. Phys. 1974.V.45. P.3023-3032.

[2] Berman L.S., Lebedev A.A. Capacitive spectroscopy of deep centers in semiconductors. L .: Nauka, 1981, 176 p.

[3] Denisov A.A., Laktyushkin V.N., Sadofiev Yu.G. Relaxation spectroscopy of deep levels // Reviews on electronic technology. 1985. Ser. 7. Issue 15 (1141), 52 p.

[4] ASTM standard F 978-02 Standard test method for characterizing semiconductor deep levels by transient capacitance techniques.

[5] www.janis.com.

[6] www.sulatech.com.

[7] Zubkov M.V. Determination of the concentration of deep centers taking into account the field dependence of the current relaxation time // Electronic Engineering. Ser. 10. Microelectronic devices. Issue 6 (78), 1989 .-- S. 42-45.

Claims (2)

1. The method of determining the ionization energy of deep levels, which consists in the fact that a sample of the semiconductor barrier structure is placed in the measuring cell, the temperature dependence of the relaxation time under external influences by voltage pulses of reverse bias with amplitude V ₁ is determined, the obtained signal from the sample is multiplied by the reference signal F ( t), produced by selection of the relaxation time and determine the maximum output voltage ΔU of the spectrum and the maximum peak temperature DLTS (t _max), characterized in that the supplement is administered flax reverse bias voltage pulse with amplitude V _2> V _1, and | V ₁ -V ₂ | >> _max kT / e, is defined by two values DLTS spectrum peak amplitude ΔU ΔU ₁ and ₂ corresponding to the same relaxation time constant and different amplitude pulses of reverse bias voltage V ₁ and V ₂ , determine the ionization energy of the deep level (AE) by the formula:

,
where d ₁ and d ₂ are the thickness of the SCR for the amplitudes of the voltage of the pulses of the reverse bias V ₁ and V ₂ ,
e is the electron charge,
ε is the relative dielectric constant of the semiconductor,
ε ₀ is the dielectric constant,
k is the Boltzmann constant,
T _max - peak temperature of the DLTS spectrum,
N _m - the concentration of small centers in the SCR. 2. A device for implementing a method for determining the ionization energy of deep levels, comprising a series-connected generator of rectangular pulses of reverse bias voltage, a measuring cell with the ability to change the temperature of the sample and a relaxation spectroscopy device of deep levels, characterized in that the generator of rectangular pulses contains an additional generator of a sequence of rectangular pulses V ₂ with the possibility of separate adjustment of their amplitude and duration In the deep-level relaxation spectroscopy device, the output of the relaxation time selector is connected to two low-pass filters (LPFs) through a switch controlled by a square-wave pulse generator.

Images