JPS628022B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an apparatus for measuring various characteristics of a semiconductor using a photovoltaic method.

The photovoltaic method itself has been used for a long time in the semiconductor measurement field, and has the advantage of being a non-contact measurement method compared to, for example, the four-probe method for measuring resistivity. FIG. 1 is a diagram for explaining the basic principle of a conventionally well-known method for measuring the resistivity distribution of a semiconductor sample using a light beam.

When the light beam 1 is irradiated onto the surface 2' of the semiconductor sample 2, which has a two-dimensional spread, the sample 2 generally
A carrier pair of holes 3 and electrons 4 is generated on the surface 2' of the sample 2, and these carriers diffuse toward the back surface 2'' of the sample 2 as shown by arrows 3' and 4'. However, for example, As is well known in the case of Si, the mobility of electron 4 is greater than the mobility of hole 3, so more electrons 4 move to the back surface 2'' than hole 3, so the semiconductor Many positively charged holes 3 remain on the surface 2' of the sample 2, and eventually the surface 2' of the sample 2 becomes positively charged. This phenomenon was reported by H.Dember of Germany in 1931 and has since been known as the "Dember effect". However, the voltage generated by the Denba effect, ie, the Denba voltage, is much smaller than the voltage generated when a pn junction is irradiated with light, and thus has not been utilized in the past.

According to the considerations of the present inventors, the following results can be obtained for n-type Si wafers.

ΔV _D =(b-1)/Sρ(0)eIαL _P ² /αL _P +1v _P /Sf+v _P (1) Here, ΔV _D means Denba voltage,
The meaning of each signal is as follows.

b: Electron mobility/hole mobility S: Wafer area ρ(0): Wafer surface resistivity e: Electron charge I: Light beam intensity (photon number/second) α: Light Beam absorption coefficient L _P : Diffusion length of minority carriers V _P : Diffusion speed of minority carriers Sf : Recombination speed of carriers on the wafer surface As can be seen from equation (1), the density voltage depends on many factors. However, if everything other than the resistivity ρ(0) can be considered constant, the result is ΔV _D =K·ρ(0) (2) where K is a constant. That is, if the light beam 1 is focused narrowly and scanned with it over a semiconductor sample (wafer) 2 without bonding, and the distribution of photovoltaic force at that time is measured,
Since this is equal to the Denba voltage, the distribution of resistivity on the surface of the sample 2 is being measured after all.

Conventionally, a Schottky junction has been used to determine the resistivity distribution. Figure 2 shows its basic principle. An ohmic electrode 6 is attached to the back surface 2'' of the semiconductor sample 2, a metal probe 5 is set up on the surface, and the vicinity of the probe 5 is irradiated with the light beam 1.As is well known, in the shot junction 5', A photoelectromotive force is generated, which can be measured with the voltmeter 7. Since the magnitude of the photovoltaic force usually depends on the resistivity of the sample 2 in the portion facing the metal probe 5, the voltmeter 7 The indication is proportional to the resistivity.The metal probe 5 can be moved over a wide wafer surface 2', but this is difficult in practice, so a mesh-shaped probe 5 as shown in FIG. The electrode 8 is crimped onto the sample 2 to form a shot joint 8'' on the entire surface. By scanning the light beam 1 over the sample 2 in this manner, the distribution of resistivity on the surface 2' can be determined.

However, the conventional method shown in FIG. 3 has several drawbacks. That is, firstly, the shot bonding 8'' depends on the pressure bonding force of the metal, the condition of the metal surface (protrusions, oxide film, etc.), and the semiconductor surface condition (oxide film, oxide film, etc.).
(moisture, dust, etc.), making it difficult to create a uniform bond over a wide surface.Secondly, there is a portion covered by the net-like electrode 8, making it impossible to irradiate the entire surface of the sample 2 with the light beam 1. , Thirdly, Ohmic electric board 6
Attaching a mark will cause a kind of damage to sample 2, and it cannot be called a complete non-destructive test.
There are drawbacks such as.

In addition, using the same principle as the method shown in Fig. 3, a shot junction was formed on one electrode using an electrolyte 13 such as Na ₂ SO ₄ as shown in Fig. 4, and the characteristics of sample 2 were measured. Examples have been reported. However, the electrolyte 13 is generally difficult to use as a transparent electrode, and there is also an ohmic electrode 6 on the back surface 2''.
Similar to the prior art described above, there is the inconvenience of having to attach a . In FIG. 4, reference numeral 12 indicates an electrode, and reference numeral 14 indicates a side wall of the electrolyte tank.

In this way, without damaging the sample to be measured,
For example, a method for measuring the resistivity distribution within the plane of a Si wafer using photovoltaic voltage was completely unknown.

It is therefore an object of the present invention to provide an apparatus capable of measuring the distribution of various properties in the plane of a semiconductor sample using photovoltaic force without causing any damage to the semiconductor sample.

To achieve the above object, the device according to the invention comprises electrodes, at least one of which is transparent, placed at intervals on both sides of a semiconductor sample whose properties are to be measured.
A thin pulsed light beam is scanned over the surface of the semiconductor sample through the transparent electrode, and the photoelectromotive force generated between the two electrodes due to capacitance coupling is measured, and the in-plane distribution of the characteristics of the semiconductor sample is measured. The main point is that the system is configured to observe the following.

In a more advantageous device configuration of the invention, the electrodes are transparent on both sides, so that the light transmitted through the semiconductor specimen is also detected, and the transmitted light and the photovoltaic force are measured simultaneously.

Hereinafter, the present invention will be described in detail based on the drawings.

FIG. 5 shows the principle of the measuring method according to the invention.
According to the present invention, when the light beam 1 is irradiated onto an arbitrary location on the sample 2, the Dember voltage can be accurately measured. As mentioned above, since the Denba voltage is generated between the front surface 2' and the back surface 2'' of the sample 2, the electrodes 8 and 9 are placed facing each other with a gap between them.The light beam 1 is pulsed. Since the generated voltage is also pulsed, the electrodes 8 and 9
Even if it is separated from the sample 2, the density voltage can be detected by capacitance coupling by the floating capacitances 10 and 11 that are formed.

FIG. 6 shows an equivalent circuit for the state shown in FIG. Voltage source (sample) generated by irradiation with light beam 1
There are capacitances 10 and 11 above and below 2, through which the voltmeter 7 is connected.

In FIG. 5, the electrode 8 can be made transparent by, for example, baking indium onto a glass surface, and the light beam 1 can be guided onto the sample 2 without significant absorption. Note that the electrode 9 may be a transparent electrode like the electrode 8, or may be an opaque electrode.

After all, in principle, the semiconductor sample 2 has electrodes 8,
9 is placed without direct contact with the electrodes 8, 9, it is clear that this method can be a completely non-destructive method.

FIG. 7 shows an embodiment of the apparatus of the present invention. In this example, a cathode ray tube 17 is used as the light source of the light beam 20. The light beam 20 is aligned to a suitable wavelength by an optical filter 18, focused by an optical lens 19, and irradiated onto the semiconductor sample 2. The light beam 20 is scanned by deflecting and scanning an electron beam (not shown) in the cathode ray tube 17. The scanning speed and scanning range are determined by appropriately adjusting the voltage from the scanning power source 31 using the regulator 32, converting the voltage into a current, and supplying the converted current to the deflection coil 16. The same scanning signal is supplied to the deflection coils 27 and 29 of the cathode ray tubes 26 and 28 for displaying scanned images. In particular, in the case of the cathode ray tube 28, the signal from the semiconductor sample 2 is converted into a deflection current by an adder 30. It is designed so that it can be superimposed on In other words, the scanning of the cathode ray tube 28 is performed in synchronization with the electron beam of the cathode ray tube 17 serving as a light source.
At locations where no photovoltage is generated on the sample surface, the cathode ray tube 17 performs scanning similar to the electron beam of the cathode ray tube 28.
is carried out on the screen. At a location where a photovoltage is generated on the sample surface, the photovoltage signal is added to the scanning signal by an adder, so the added signal is applied to the deflection coil 29, which changes the electron beam to correspond to the photovoltage signal. and deflect it. Thus, an amplitude modulated scanned image is obtained on the screen which represents the change in the photovoltage on the sample surface.

In FIG. 7, the sample 2 has two transparent electrodes 8,
8' shows the state where it is sandwiched. The purpose is to know the intensity and wavelength distribution of the light beam 20' transmitted through the sample 2. That is, if the transmitted light 20' is detected and analyzed by the detector 21 consisting of a photodiode and amplified by the amplifier 22, information on the impurity concentration can be obtained based on a well-known principle. As a result, it becomes more reliable to determine which of the factors related to the power voltage shown in equation (1) has the greatest effect, and the synergistic effect is large. For example, if there is no change in the intensity of the transmitted light 20', but the Denver voltage changes greatly, this is due to a sudden change in the surface recombination speed Sf, rather than due to the resistivity ρ(0). It is thought that he did.

As already mentioned, in the present invention, the density voltage is measured by capacitance coupling, and for this purpose,
Light beam 20 is pulsed. Pulsing is performed by pulsing the electron beam of the cathode ray tube 17, and this is done by modulating the brightness of the cathode ray tube 17 with the pulse power source 15.
This pulse voltage is used for synchronous detection of signals. That is, this is used as a reference voltage for the synchronous detector 25. By doing so, the signal-to-noise ratio of the signal is significantly improved.

One of the amplified and synchronously detected signals is
It is used to modulate the electron beam intensity of the display cathode ray tube 26. Since the scanning of the electron beam is synchronized with the scanning of the electron beam of the cathode ray tube 17,
The brightness-modulated scanned image on the screen of No. 6 represents the photovoltage distribution on the sample surface. On the other hand, the synchronously detected signal is transmitted to the display cathode ray tube 28.
is also applied to the deflection coil of , and as described above, an amplitude-modulated scanning image corresponding to the photovoltage signal is obtained. A brightness modulation image is convenient for grasping the outline of the photovoltage distribution, but it is difficult to grasp it quantitatively. On the other hand, the amplitude-modulated scanning image is convenient for quantitatively understanding the photovoltage, and in this embodiment, since both scanning images are used together, the cathode ray tube
6 and 28 are arranged.

In FIG. 7, spacers 39, 39' are inserted between the electrodes 8, 8' and the sample 2;
This is to keep it close to mica, myra,
A light-transmissive insulating film such as polyethylene is used.
The appropriate thickness is several tens of micrometers or less.

The embodiment of the present invention has been described above based on FIG. It goes without saying that a method such as moving a mirror may also be used as a scanning method.

In addition, as an application example, we have only mentioned its effectiveness in measuring resistivity distribution, but as is clear from equation (1), it is also possible to measure other characteristics, and furthermore, it is possible to measure other characteristics such as ion implantation. wafers with p-n junctions made of oxides or junctions made of regions of the same conductivity type but with different impurity concentrations; wafers with oxide films, especially oxide films containing fixed charges therein; It is clear that the present apparatus can also be applied to wafers having interface (interface) levels. For example, p-n
In the case of bonding, the uniformity of bonding is displayed on the display tube in a short time, making it easy to determine whether the wafer is to be used in the process of solid-state circuit elements.
The industrial benefits are great.

[Brief explanation of the drawing]

Figure 1 shows the basic principle of measuring the resistivity distribution of a semiconductor sample using a light beam, and Figures 2 to 4 show the principle of a conventional device for measuring the characteristics of a semiconductor sample using a light beam. 5 is a diagram showing the principle of the present invention, FIG. 6 is an equivalent circuit diagram corresponding to FIG. 5, and FIG. 7 is a block diagram showing an embodiment of the semiconductor characteristic measuring device according to the present invention. be. 1, 20, 20'...Light beam, 2...Semiconductor sample, 2'...Surface, 2''...Back surface, 3...Hole,
4...electron, 5,12...tip, 5,8''...shot junction, 6...ohmic electrode, 7...voltmeter, 8,8'...transparent electrode, 9...opaque electrode, 10 , 11... floating capacitance, 13...
... Electrolyte, 14 ... Side wall, 15 ... Pulse power supply,
16, 27, 29... Deflection coil, 17, 26,
28...Cathode ray tube, 18...Optical filter, 19
...Optical lens, 20, 21...Photodetector, 2
2, 24...Amplifier, 23...Switch, 25...
... Synchronous detector, 30 ... Adder, 31 ... Scanning power supply, 32 ... Adjuster, 39, 39' ... Spacer.

Claims (1)

[Scope of Claims] 1. A light beam irradiation means for scanning and irradiating the surface of a semiconductor sample with a pulsed and focused light beam; Electrodes are arranged at intervals on the front and back sides of the sample in order to take out the photovoltaic force generated between the back surfaces by capacitance coupling, and among these electrodes, at least the electrode on the side irradiated with the light beam is irradiated with the light beam. the electrode consisting of a transparent electrode; and a synchronous detection circuit that extracts only the signal component of the photovoltaic force by comparing and referencing the output signal extracted by the electrode with the pulsed signal of the light beam. and observation means for inputting the signal component as a modulation signal in order to observe changes in the output from the synchronous detection circuit. 2. The light beam irradiation means includes a cathode ray tube, a pulse power supply connected to the cathode ray tube for brightness modulation, a scanning power supply connected to a deflection coil of the cathode ray tube for deflection, and the above for convergence. A semiconductor characteristic measuring device according to claim 1, comprising a cathode ray tube and an optical lens disposed between the sample. 3. The semiconductor characteristic measuring device according to claim 2, wherein an optical filter is added between the cathode ray tube and the optical lens for aligning the wavelengths of the light beams. 4. The semiconductor characteristic measuring device according to claim 1, wherein the electrode is placed on the sample via a light-transmitting electrically insulating plate. 5. The semiconductor characteristic measuring device according to claim 1, wherein the observation means is a cathode ray tube whose brightness is modulated by the signal component. 6. The semiconductor characteristic measuring device according to claim 1, wherein the observation means is a cathode ray tube whose amplitude is modulated by the signal component. 7 A light beam irradiation means that scans and irradiates the surface of a semiconductor sample with a pulsed and focused light beam, and a photovoltaic device that generates between the front and back surfaces of the sample by the irradiation of the light beam. light-transmissive electrodes arranged at intervals on the front and back sides of the sample for extracting electric power through capacitance coupling; and for detecting the intensity of light transmitted through the sample by irradiation with the light beam. A light detection means disposed on the back side of the sample and an output signal taken out by the electrode or an output signal from the light detection means are compared and referenced with a pulsed signal of the light beam. It comprises a synchronous detection circuit that extracts only the signal component of the photovoltaic force or the signal component of the transmitted light, and an observation means that inputs the signal component as a modulation signal in order to observe changes in the output from the synchronous detection circuit. A semiconductor characteristic measuring device characterized by the following.