JPS63257199A - Method of measuring anions in plasma - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION (Field of Industrial Application) The present invention relates to a method for measuring information regarding negative ions such as temperature, density, and energy distribution of negative ions in plasma.

(Prior Art and Problems) Plasma processing has recently become popular and is useful for modifying various solid surfaces. In order to efficiently control these plasmas, it is important to control the energy distribution of charged particles, especially electrons. However, since processing plasma uses gases such as hydrogen, oxygen, hydrocarbon, and halogen, the density of negative ions in the plasma cannot be ignored. Therefore, if such a plasma is analyzed by the conventional probe method, there is a risk that the charged particle density will be highly erroneously evaluated. Furthermore, since negative ions in plasma are as useful for surface modification as positive ions, it is often essential to understand their density, temperature, energy distribution, etc. However, the conventional Drivestein method (M, J, Druyvestein: Z, Physik
64 (1930) 787), it was possible to measure the energy distribution of negative ions only under very limited conditions. That is, when the noise is low and the negative ion density is much larger than the electron density (J, B. Thompson:
Proc, Roy, SOC, A262 (1961)
503), when the mass of the negative ion is not very large (H,
Amemiya: Jpn, J, App1, Phys,
25 (1986) 595), it was difficult to detect negative ions when the ratio of negative ion density/electron density was small or when the mass of negative ions was large.

(Means for Solving the Problems) In general, in weakly ionized plasma, the electron temperature is several hundred degrees K, but the specific baion temperature is from several hundred@K to several thousand degrees at most.

Therefore, in the current-voltage characteristics of negative ions, the saturation current value is much smaller than that of electrons, but the rise in the voltage repulsion region is much steeper.

The present invention focuses on this and takes the higher-order derivative of current with respect to voltage to make the signal of negative ions higher than the signal of electrons, thereby allowing the Langminiar probe to separate and measure electrons and negative ions. made possible.

(Function) When a flat probe is inserted into plasma, a probe current characteristic tp (vp) that changes non-linearly with respect to the applied probe voltage 2 as shown in FIG. 2 is obtained. In Fig. 2, ■5 is the space potential near where the probe is located in the plasma. Each region r, n, m of this nonlinear curve
The characteristics of , are expressed by the following equation when the velocity distribution of charged particles in the plasma follows the mask well distribution.

1: i, = i + = n, eS exp (-1/2)
(kTe/M+) “2;Vp<<Vs (1) II: i,=n,eS(kTa/2 yrm)”
2exp (-e(Vs-Vp)/kTa)+ne-e
S(kT-/2 πM-) ”'exp (-e(V,
−V, )/kTi; V, <V,
(2) III: i, = n*5 (kTe/2 πm)
"'+n-eS(kT-/2yrlJ-)"";
Vp〉Vs (3) Here, n, , n, , n- are positive ions, electrons,
n in negative ion density. =n, + n-, T+ + Ts
+T- is the temperature of positive ions, electrons, and negative ions, respectively;
M, m, and M- are the masses of positive ions, electrons, and negative ions, and S
is the probe surface area, e is the electron charge, and k is the Boltzmann constant. Next, for tp(vp), the higher order differential coefficient I
F") N'p) becomes 0 in regions (2) and (2), and a signal is obtained only in region H, resulting in the following equation.

U: ip” (Vp)=ts (″)(Vp)”t
−” (Vp) (4a)rs ”' (vp
) = n, eS (k'r, /2 xmν" (e/k
T,) ”xexp (-e(V,-V,)/kT,
) (4b)i-(h) (V,)=n-eS (k
T-/2 yr M-) '/2(e/kT-) ”
xexp (-e(Vs-V,)/kT-3(4C)(
As can be seen from equations 4a) and 4c), even if you take the nth derivative, both electrons and negative ions are ■, and each temperature T a +
Although the exponential change determined by T- does not change, the value of its peak changes with each differentiation such that it becomes larger at lower temperatures. That is, 1@”' (Vp), i-(re'
The peak value of (V,) is io "' max, l-"'
max, the ratio becomes the following formula.

i-(+″'max/i,”)max=(n-/n
, )(m7M−ν″X(Ta/TJ −”2(5)) From this equation, the ratio of higher order differentials of electrons and negative ions is T8
It can be seen that in the case of −, the negative ion peak appears as a peak larger than the electron peak by a factor of (T, /T) for each differentiation.

The minute alternating current method, which is a specific means for obtaining this differential current value, will be explained below. Probe voltage■.

When a minute AC voltage ■ is superimposed on the Taylor expansion, 1
．． changes as follows.

Here, let v=a−s inωt and convert this into equation (6)%
The frequency components proportional to the formula % are each 5in3ωt.

5in4ωt, 5in5ωt, -. Furthermore, v=
If aI sin ω, t+b-sina+2 t, then
1, (3) , 1, (U , 1, (s)
,... are each frequency components proportional to s i n
(2ω1-C2)t (or 5in(2ω2-ω,)t)
, sir+(C13ω2)t(or S l n(2ω,
2 Q) 2) t, sin(3ω, −(I
J2) t), s 1n(3ω, 2ω,)
t (or 5in(3ω2-2co,)t, s
1n(4ω, -ω,)t, sin(4ω1-C
2) t), .... Therefore, each high-order differential coefficient can be obtained by receiving these high-frequency components. Strictly speaking, each of the frequency components includes higher-order differentials higher than higher-order differential coefficients. For example 5in3
The components of ω mainly depend on I, (3), but also i, (S), i, ('l).

It also depends on... However, in the case of the Maxwellian distribution, as is clear from equation (4), each higher-order differential has a similar exponential function term, so the shape of the peak is kept the same as in the case of only 1p(3), and these higher-order derivatives Distortion is not a problem.

(Effects of the Invention) According to the present invention, detection of characteristic values such as density, temperature, and energy distribution of negative ions, which has been difficult in the past, can be more easily achieved.

(Example) An example of the present invention is shown in FIG. Fv is a variable frequency generator that generates an alternating current signal 5ln2πfvt of frequency fv. Fc is an oscillator whose frequency fc is fixed (preferably crystal) and generates an alternating current signal 5in2πfct. However, fv
>fc. The signals from Fv and Fc are introduced into a carrier removal mixer M, and two alternating current signals with frequencies fv+fc and fv-fc are generated. The phases of these two alternating current signals have a fixed relationship. B is a buffer amplifier for setting the amplitude of the signal composed of the two AC signals to a certain constant value. The output of the buffer amplifier B is applied to the primary side of the coil T2, and the above AC signal appears with a constant amplitude on the secondary side, and is superimposed on the variable DC voltage (probe voltage).

The voltage formed by this superposition is applied to the probe P inserted into the plasma. A probe characteristic as shown in FIG. 2 is obtained at both ends of the resistor r, which changes non-linearly with respect to the applied voltage. When an alternating current voltage is superimposed on the probe voltage, the probe current also has an alternating current component, and this alternating current component occurs on the secondary side of the coil T. In this embodiment, the frequency components corresponding to the 3rd, 4th, and 5th derivatives of the AC components are transmitted to each selection amplifier A3. Δ4. A5
Selective amplification is carried out by. In this case, the superimposed frequency component ω1 /2π=
fv〒f c, C2/2π=f v−f c is sufficiently prevented from influencing the measurement system. Since fv is variable, the resulting two frequencies fv+fc are variable. Therefore, a programmable filter is used as the band elimination filter BEF, and the removal frequency is set so as to remove the fv+fc component according to the value of fv. Selection amplifier A3. A4. The signal passed through A5 is introduced into the first input terminal 1 of the phase sensitive detector LOCK. On the other hand, the beat generating circuits C3 and C4°C generate the frequencies corresponding to the third, fourth, and fifth derivatives of the signals from the variable frequency oscillator Fv and fixed frequency oscillator Fc.
5 to generate frequency components corresponding to the higher-order differentials. These signals are introduced into the other input terminal 2 of the phase detector. The signal from input 1.2 is multiplied by a phase tU detector (phase synchronized detection) for each higher order differential i, (3),
Signals corresponding to i, (4), i, (S), . . . are output. In this case, the beat wave generation circuits C3, C4, and C5 convert the frequency component corresponding to the above-mentioned higher-order differential into a square wave or a narrow pulsed square wave from a sine waveform, and convert the frequency component corresponding to the above-mentioned higher-order differential into a square wave or a narrow pulsed square wave in a short time. The signal-to-noise ratio can be improved by phase-locked sampling of signal 1. Figure 3 shows the higher order differential i, (2)
, i, (4) , i, (5ゝ example is shown.a
is a negative ion peak, and b is an electron peak.

The output of the phase detector gives information about electrons and negative ions in separate form in the region (■) in Figure 2, which shows the probe characteristics, but the higher the order of differentiation, the more the electrons are suppressed and the negative ions are emphasized. The obtained curve, , (nlyp) is obtained. From a measurement perspective, the lower the order of differentiation, the larger the signal output, but the electron suppression ratio is (T-/T, )"
Only low. In practice, it is sufficient to measure the 3rd to 5th order differential and select the order to obtain appropriate data. Although the above method is limited to the 3rd to 5th order differentiation, it goes without saying that higher order differentiation may be taken as long as the electronic circuit permits, and the idea of the present invention is not limited to the 3rd to 5th order. do not have. Further, although a flat probe was used in the above embodiment, the present invention is also applicable to a cylindrical or spherical probe.

[Brief explanation of drawings]

FIG. 1 is a circuit diagram for negative ion detection according to an embodiment of the present invention, and shows the probe characteristics and their higher-order differential i, (3)
, i, (4) , i, (5) Figure 2 is a diagram showing the probe characteristics, Figure 3 is the higher order differential of the probe current obtained by the circuit in Figure 1, i, (3 ), i, (4), 1p(S)
Diagram showing. P...Flat probe, A...Counter electrode, Fv...
Variable frequency generator, Fc...fixed oscillator, B...buffer amplifier, T2...coil, A3. A4. A5・
...Selection amplifier, L...Phase detector, C3, C4, C5...P)1 generation circuit.

Claims (7)

[Claims] (1) A minute AC voltage is superimposed on the DC probe voltage applied to the probe inserted into the plasma, and a signal corresponding to the third or higher differential coefficient of the probe current with respect to the probe voltage is detected, and negative ions in the plasma are detected. A method for measuring negative ions in plasma to obtain information on (2) The method for measuring negative ions in plasma according to claim (1), wherein the alternating current voltage is a sine wave voltage. (3) The detection of the signal corresponding to the differential coefficient of order N (an integer of 3 or more) or higher is performed by detecting a harmonic of N times or more the sine wave voltage frequency. A method for measuring negative ions in plasma according to scope (2). (4) The method for measuring negative ions in plasma according to claim 1, wherein the AC voltage is a sum signal of a sine wave voltage of frequency f_1 and a sine wave voltage of frequency f_2. (5) Detection of the signal corresponding to the third-order differential coefficient is 2f_1−
A method for measuring negative ions in plasma according to claim (4), which is carried out by detecting a frequency component of f_2 or 2f_2-f_1. (6) Detection of the signal corresponding to the fourth-order differential coefficient is f_1-3
A method for measuring negative ions in plasma according to claim (4), which is carried out by detecting frequency components f_2, 3f_1-f_2, or 2f_1-2f_2. (7) Detection of the signal corresponding to the fifth-order differential coefficient is 3f_1-
The method for measuring negative ions in plasma according to claim 4, which is carried out by detecting frequency components 2f_2, 3f_2-2f_1, 4f_1-f_2, or 4f_2-f_1.