RU2080611C1 - Method for determining electrophysical properties of semiconductors - Google Patents

Abstract

FIELD: semiconductor engineering; local check-up of deep center parameters. SUBSTANCE: semiconductor strip under check is placed between two conducting plates one of which is transparent plate, nonequilibrium potential difference is set up across barrier junction by irradiating semiconductor plate through transparent plate with electromagnetic rays and generating photoelectromotive force. Data on relaxation processes are taken by using capacitive coupling between plates and surfaces of semiconductor strip. Tp ensure desired locality, electromagnetic radiation is recorded. In case of semiconductor strips of uniform thickness (such as half-insulating substrates), surface-to-volume semiconductor junction is used as barrier junction. In implementing this method, it will be most natural to irradiate semiconductor strip with square-wave pulses of electromagnetic radiation of fixed intensity and then to determine amplitude and waveform of pulses across conducting plates; parameters of relaxation processes should be between calculated according to parameters of pulses across plates. EFFECT: facilitated procedure. 4 cl, 3 dwg

Description

 The invention relates to techniques for controlling the parameters of semiconductors. It is most advisable to use the present invention for local control of parameters of deep centers.

A number of known methods for determining the parameters of semiconductors. The spectrophotometric method is based on recording the absorption of light quanta during stimulation of impurity-zone transitions [1] This method is non-contact and allows one to determine the main parameters of the impurity. The disadvantage of this method is the relatively low sensitivity (10 ¹⁶ -10 ¹⁷ cm ^-3 ), low resolution (0.2 eV) and low locality (1 mm).

 Closest to the proposed invention is a method for determining the parameters of semiconductor wafers having a barrier transition based on relaxation spectroscopy of deep levels (RSGU) [2] When using this method, a voltage pulse of one or another polarity is applied to the semiconductor barrier transition. As a result of exposure to a nonequilibrium potential difference, a change in the distribution of potential occurs in the barrier transition, capture or release of charge carriers from deep levels, and a change in the width and capacitance of the barrier transition. The process of relaxation to a stationary state, i.e. the process of capture or ejection of charge carriers is recorded by a change in the capacitance of the barrier transition. Measurements are taken over a temperature range. In this case, using a device for selecting a signal by the relaxation time, a signal with a certain time constant is extracted and the dependence of this signal on temperature is determined. Then the selection device is adjusted to a different value of the relaxation time and the procedure is repeated. The position of the maxima of these dependences, as well as their width, determine all the basic parameters of the deep level.

The advantage of the RSGU method is that it allows you to determine all the basic parameters of deep levels. The method has high sensitivity (up to 10 ⁸ -10 ⁹ cm ^-3 ) and high resolution (better than 10 ^-2 eV). The disadvantage of this method is that it is destructive, since it requires the creation of barrier and ohmic contacts on the semiconductor surface, as well as low locality associated with the need to create a contact with a diameter of about 1 mm.

 The aim of the invention is primarily to provide non-destructive testing by eliminating irreversible effects on the semiconductor.

 The goal is achieved by the fact that in the known method for determining the electrophysical parameters of semiconductors, which includes cooling and (or) heating a semiconductor wafer containing a barrier transition in a temperature range at which the relaxation time constant of deep levels is within a range that allows it to be determined with sufficient accuracy (usually in the range of 90 450 K, creation and change of the nonequilibrium potential difference at the semiconductor barrier junction, registration of relaxation processes occurring in the semiconductor On the other hand, determining the dependence of the parameters of relaxation processes on the temperature of the semiconductor and calculating from these dependences the parameters of the deep impurity centers of the semiconductor, the semiconductor wafer is placed between two conducting parallel planes (plates), one of which is transparent, parallel to these planes. irradiation of a semiconductor wafer through a transparent lining with electromagnetic radiation, the quantum energy of which is above the threshold g neratsii free carriers.

 The change in the nonequilibrium potential difference is carried out by periodic or pulsed changes in the radiation intensity over time. The registration of relaxation processes is carried out by determining the amplitude and shape of the stress arising on the plates, and then calculating the values of the parameters of the relaxation processes by the amplitude and shape of the voltage on the plates, taking into account the time dependence of the intensity of electromagnetic radiation.

 To ensure the required measurement locality, electromagnetic radiation is focused so that the focal spot is on the surface of the semiconductor and its diameter does not exceed the required measurement locality.

 In a number of cases (for example, when controlling semi-insulating semiconductor substrates) as a semiconductor barrier transition, it is advisable to use the surface-internal volume of the semiconductor; to create this barrier transition does not require any additional technological operations.

 When implementing the proposed method for determining the parameters of semiconductors, it is most natural to irradiate a semiconductor wafer with rectangular pulses of electromagnetic radiation of fixed intensity, then determine the amplitude and shape of the pulses on the conductive plates, and calculate the values of the parameters of the relaxation processes from the values of the parameters of the voltage pulses on the plates. In this case, it is advisable to use a signal selection device for the relaxation time to process the signal, which allows you to select a signal with a certain relaxation time constant.

 The proposed method for determining the electrophysical parameters of semiconductors is one of the variants of the RSGU method, which is based on the study of the processes of emission of charge carriers from deep levels located in the space charge layer (POP) (or carrier capture at these levels). The space charge layer is associated with some kind of barrier transition (pn junction, Schottky barrier, MIS structure). In the equilibrium state at this transition there is a potential difference, the width of the POPs is determined by this potential difference. When an external voltage is applied, the potential difference at the barrier transition changes, the width of the POP changes and the emission of free charge carriers from deep levels located in the POP (or capture to these levels) occurs. In the proposed method, the change in the equilibrium barrier potential difference occurs due to the barrier photoEMF arising at the barrier transition. The processes occurring in POPs during the emergence and disappearance of photoDES are similar to the processes occurring in POPs when an external voltage is applied to it. There is also a change in the width of POPs and recharging of deep levels. The amplitude and shape of the stresses at the barrier junction depend on the amplitude and time dependence of the intensity of electromagnetic radiation with which the semiconductor is irradiated, as well as on the value of the barrier potential difference (barrier height) and the parameters of relaxation processes in POPs. The voltage from the barrier junction is applied to the input circuits of the measuring amplifier via capacitive coupling. In this case, capacitances formed by the surfaces of the semiconductor wafer and the conductive plates, which are galvanically connected to the measuring amplifier, are used. Thus, the controlled semiconductor wafer does not have any galvanic contact with the input of the measuring amplifier; mechanical contact can also be eliminated. Any irreversible effects on the semiconductor are excluded.

 According to a second embodiment of the invention, focused electromagnetic radiation is used, for example, in the infrared range, which is focused into a spot with a diameter of several microns, which ensures high locality of control.

 According to a third embodiment of the invention, the surface-to-internal volume of the semiconductor and, accordingly, surface POPs are used as a semiconductor barrier transition. This significantly expands the scope of the method, since it becomes possible to control homogeneous semiconductors, in particular semi-insulating gallium arsenide substrates.

 According to a fourth embodiment of the invention, the semiconductor is irradiated with rectangular pulses of electromagnetic radiation of a fixed intensity, and the parameter values of the relaxation processes are calculated from the parameter values of the voltage pulses on the conductive plates. Excitation of the semiconductor by such radiation pulses allows the use of methodological developments and hardware solutions of traditional RSGU. In particular, in this case, it is advisable to use a signal selection device for the relaxation time to process the signal, which allows you to select the signal with a certain relaxation time constant.

 The combination of four technical solutions in one application is due to the fact that all four solutions solve one problem: the determination of the electrophysical parameters of semiconductors, more specifically the parameters of deep centers, while the second technical solution provides high locality of control, the third solution specifies the type of controlled semiconductors makes it possible to control homogeneous semiconductors and the fourth - specifies the parameters of the used electromagnetic radiation.

 In FIG. 1 shows a functional block diagram of a device that implements the proposed method: in FIG. 2 equivalent measurement scheme: in FIG. 3 is a graph of the voltage at the input of a measuring amplifier.

 A device that implements the proposed method for determining the electrophysical parameters of the semiconductor wafer 1, consists of two conductive planes (plates) 2 and 3, one of which (2) is transparent to the used electromagnetic radiation. A semiconductor wafer 1 is located between these plates and is irradiated through the optical fiber 4 by LED 5. The LED 5 is fed from the rectangular pulse generator 6. The signal from the plates is fed to the input of the amplifier 7 and then to the signal selection device according to the relaxation time 8. The plate and plates are placed in the thermostat 9 with a thermocouple 10. A signal from a thermocouple 10 is fed to input x of the recorder 11. An output signal from a selection device 8 is supplied to input y of the recorder 11. A lens system 12 focuses electromagnetic radiation.

In FIG. 2 ε photo-emf generated by a light pulse, R _pp is the semiconductor resistance, C is the capacitance between the upper and lower surfaces of the semiconductor wafer 1 and the upper 2 and lower 3 plates, R _in is the input resistance of the amplifier 7.

In FIG. 3a, a graph of a light pulse of intensity I _o and duration T _{o is} shown. In FIG. 3, b is a graph of the voltage U _I on the resistance R _I versus time; t _{1 the} duration of the transition process of the occurrence of photoEMF without taking into account the deep centers; τ _{o the} duration of the relaxation processes of recharging of the deep centers in the event of photo-emf; τ _{2 the} duration of the transition process of the decline in photoEMF without taking into account the deep centers; τ _{3 the} duration of relaxation processes with the disappearance of photo-emf; ε _{1 is the} value of phoEMF excluding deep centers; ε _{2 is the} value of the photoEMF taking into account the deep centers; ε ₂ -u ₃ voltage drop at the input of the amplifier due to the discharge of capacitances C with a time constant R _in • C / 2.

The method is implemented as follows. For concreteness, we consider the case when a homogeneous semiconductor wafer of type n is irradiated with pulses of electromagnetic radiation of duration T _o and intensity I _o : the surface potential is higher than the volume potential of the semiconductor. (In the case of the pn junction or Schottky barrier, the physical processes are similar). Let v _b denote the potential difference between the surface and the volume of the semiconductor. The thickness of the surface POPs is [4]
h = [2ε _d ε _o v _b / e (n + N _i )] ^1/2 , (1)
where ε _{o is the} dielectric constant
ε _{d the} dielectric constant of the semiconductor grating,
e is the charge of an electron,
n concentration of small donors,
N ₁ concentration of ionized deep centers.

где k постоянная Больцмана,
T температура полупроводинка,
j₁=eg,
g темп генерации пар электрон дырка,
j_s плотность тока насыщения при обратном включении барьерного перехода поверхность объем, определяемая соотношением [5]
j_s=AT²exp(-e v_б/kT), (3)
где A постоянная Ричардсона. Из (2), (3)

Рассмотрим переходные процессы при включении освещения (см. фиг. 21). Будем считать, что ε ≪ v_б Постоянная времени τ₁ определяется длительностью дрейфовых процессов при возникновении фотоЭДС. Будем считать, что τ₁≪ τ₂ Зависимость v_б(t) при t ≥ τ₁ может быть представлена в виде
v_б= v_бо+Δv(t). (5)
Здесь v_бо барьерная разность потенциалов, обусловленная наличием мелких доноров, а Δv барьерная разность потенциалов, обусловленная перезарядкой глубоких центров. Подставив (5) в (4) получим при t ≥ τ₁

где ε₁= (kg/AT)exp(ev_бо/KT). В свою очередь
Δv(t) = Δv_к[1-exp(-t/τ₂)],)](7)
где

постоянная времени релаксационного процесса перезарядки глубокого уровня E_ti, a Δv_к, согласно [4] определяется соотношением

где N_дг концентрация глубоких уровней, U разность между потенциалом глубокого уровня и потенциалом Ферми в объеме полупроводника. Таким образом

При t< R_вх•C ε равно напряжению на входе усилителя 7; при t ≈R_вх•C необходимо учитывать влияние разряда емкостей C. При n > N_дг соотношение (9) упрощается

Отметим что при включении освещения происходит заполнение глубокого уровня.When one of the semiconductor surfaces is irradiated, light is absorbed at a distance l = 1 / α, where α is the absorption coefficient. At 1 ≅ h, electron – hole pairs are generated in the POPs, with the electrons being carried away into the semiconductor bulk, and the holes to the surface. Between the illuminated and unlit surfaces of the semiconductor, a barrier photoEMF equal to [5]

where k is the Boltzmann constant,
T temperature semiconductor,
j ₁ = eg,
g rate of generation of electron-hole pairs,
j _s the saturation current density when the barrier junction is switched back on, the surface is the volume defined by the relation [5]
j _s = AT ² exp (-ev _b / kT), (3)
where A is Richardson's constant. From (2), (3)

Consider the transients when the lighting is turned on (see Fig. 21). We assume that ε ≪ v _b The time constant τ ₁ is determined by the duration of the drift processes when photo-emf occurs. We assume that τ ₁ ≪ τ ₂ The dependence v _b (t) for t ≥ τ ₁ can be represented as
v _b = v _bo + Δv (t). (5)
Here v _bo barrier potential difference due to the presence of shallow donors and Δv barrier potential difference caused by charge of deep centers. Substituting (5) into (4) we obtain for t ≥ τ ₁

where ε ₁ = (kg / AT) exp (ev _bo / KT). In turn
Δv (t) = Δv _to [1-exp (-t / τ ₂ )],)] (7)
Where

the time constant of the relaxation process of recharging a deep level E _ti , a Δv _k , according to [4] is determined by the relation

where N _{dg is the} concentration of deep levels, U is the difference between the potential of the deep level and the Fermi potential in the semiconductor bulk. Thus

When t <R _I • C ε is equal to the voltage at the input of amplifier 7; at t ≈ R _in • C, it is necessary to take into account the influence of the discharge of capacitances C. For n> N _dg, relation (9) is simplified

Note that when the lighting is turned on, a deep level is filled.

At t> τ _{2 / the} drop in the peak of the pulse is determined by the discharge of capacitances C through the resistances R _I and R _pp . The time constant of this decline is 0.5 (R _I + R _pp ) • C.

When the radiation source is turned off (at t> T _o ), the voltage drop is determined by the same processes as when the forward bias is turned off at the asymmetric pn junction. In accordance with this, τ ₂ is equal to half the lifetime of minority carriers, and τ ₃ = τ _o is determined by the relaxation processes of recharging deep centers (when the lighting is turned off, deep centers are emptied).

где σ_i сечение захвата носителей заряда,
<v_n >- средняя тепловая скорость электронов,
N_o эффективная плотность состояний в зоне проводимости полупроводника,
g_i коэффициент вырождения глубокого уровня.The time constant τ _o is associated with the parameters of the deep level E _{ti by the} relation [2, 3, 4]

where σ _{i is} the carrier capture cross section,
<v _n > is the average thermal velocity of electrons,
N _o effective density of states in the conduction band of the semiconductor,
g _{i the} coefficient of degeneracy of the deep level.

E _ti deep level ionization energy.

To determine E _ti, it is necessary to carry out measurements in the temperature range when tuning the signal selection device for the relaxation time to two values of the signal time constant τ ₀₁ and τ ₀₂ The value of E _ti , according to (10), is determined by the relation
E _ti = [ln (τ ₀₁ ) -ln (τ ₀₂ )] / [1 / kT _m1 -1 / kT _m2 ], (11)
where T _m1 and T _{m2 are} the temperature values at which the output signal from the selection device is maximum when it is tuned to the times τ ₀₁ and τ _02, respectively. The capture cross section σ _i can be calculated from (10) using the known values of N _o , E _ti , <v _n >.

где (Δv)_min чувствительность определения изменения напряжения.The amplitude of the signal from the selection device is proportional to the concentration of charge carriers captured at deep levels (or emitted from these levels), i.e., the concentration of deep levels. In addition, it is influenced by a number of other factors such as a deep level, the method of recording the signal, the value of the time constant, temperature, etc. In this regard, only relative measurements of the concentration of deep levels are possible. Sensitivity by concentration can be estimated by the ratio

where (Δv) _{min is the} sensitivity of determining the change in voltage.

Примем величину входного сопротивления усилителя R_вх=100 МОм, а зазор обкладка полупроводник 0,1 мм. Тогда постоянная спада вершины импульса равна 10^-2 с; при этом имеется возможность фиксировать релаксационные процессы длительностью меньшей ≃ 10^-2 с, т.е. контролировать содержание в арсениде галлия большинства глубоких уровней. Электромагнитное излучение с длиной волны 0,7 мкм легко сфокусировать в пятно диаметром ≈2 мкм. Таким образом предлагаемый способ позволяет контролировать глубокие уровни и связанные с ними дефекты в арсенале галлия с локальностью несколько микрон без каких-либо контактов или необратимых воздействий. Предварительные экспериментальные исследования это подтвердили.As an example, we consider the possibility of controlling gallium arsenide structures. Gallium arsenide of n-type has at room temperature v _b ≈0.75 B, α ≈ 5 μm ^{-1 The} condition 1 / α <h is satisfied at a carrier concentration of n ≅ 3 • 10 ¹⁶ cm ^-3 , that is, there is a possibility control gallium arsenide structures with a lower concentration, including semi-insulating substrates. As a radiation source, we choose a semiconductor LED with a pulsed power P _o = 100 μW and a wavelength of 0.7 μm. The value of the quantum yield is taken equal to q = 0.1. The area of the semiconductor S _o = 10 cm ² . The rate of generation of electron-hole pairs is

We take the value of the input impedance of the amplifier R _I = 100 MΩ, and the clearance gap of the semiconductor is 0.1 mm. Then the constant decline of the peak of the pulse is 10 ^-2 s; in this case, it is possible to fix relaxation processes with a duration of less than ≃ 10 ^-2 s, i.e. control the content of gallium arsenide at most deep levels. Electromagnetic radiation with a wavelength of 0.7 μm is easy to focus into a spot with a diameter of ≈2 μm. Thus, the proposed method allows you to control deep levels and related defects in the gallium arsenal with a locality of several microns without any contacts or irreversible effects. Preliminary experimental studies have confirmed this.

Claims (4)

 1. A method for determining the electrophysical parameters of semiconductors, including cooling and (or) heating a semiconductor wafer containing a barrier junction, creating and modifying a nonequilibrium potential difference at a semiconductor barrier junction, recording relaxation processes occurring in a semiconductor, determining the dependence of relaxation process parameters on the semiconductor temperature, and calculating according to these dependences of the parameters of deep levels, characterized in that the semiconductor wafer is placed m between two parallel conductive plates, one of which is transparent, parallel to these plates, the nonequilibrium potential difference at the barrier junction is created by irradiation through the transparent plate of the semiconductor wafer with electromagnetic radiation, the quantum energy of which is higher than the energy threshold of generation of free charge carriers in the semiconductor, a change in the nonequilibrium potential difference carried out by periodic or pulsed changes in radiation intensity over time, and registration relaxation processes carried out by determining the amplitude and shape of voltage produced in conductive plates, and the subsequent calculation of the parameter values of the relaxation processes in the amplitude and shape of voltage given time depends on the intensity of the electromagnetic radiation.  2. The method according to claim 1, characterized in that the used electromagnetic radiation is focused so that the focal spot is on the surface of the semiconductor wafer and its diameter does not exceed the required locality of the measurements.  3. The method according to claim 1 or 2, characterized in that as a semiconductor barrier transition, a surface-internal volume transition of the semiconductor is used.  4. The method according to claim 1, or 2, or 3, characterized in that the semiconductor wafer is irradiated with rectangular pulses of electromagnetic radiation of a fixed intensity, the amplitude and shape of the voltage pulses arising on the conductive plates are determined, and the values of the parameters of the relaxation processes are calculated from the values of the parameters of the pulses voltage on these plates.

Images