JPH1151856A - Method for measuring photoreflectance mapping - Google Patents

Abstract

PROBLEM TO BE SOLVED: To measure a PR(photoreflectance) spectrum and the intensity distribution of PR signal with high accuracy by extracting the intensity of PR signal from all signal intensities taken into a photodetector. SOLUTION: A total signal intensity distribution data 1 is represented by I(x, y, λ), a PL(photoluminescence) + excitation scattering light intensity distribution data 2 is represented by IPL+ S(x, y), a probe light reflection intensity distribution data 3 is represented by R(x, y, λ) when the exciting light is turned off, and a PR signal intensity distribution data 4 is represented by IPR(x, y, λ). The PR signal intensity IRP is a function of the wavelength λ and the measuring position (x, y) on a sample and represented by a formula; IPR (x, y, λ)= I(x, y, λ)-IPL+ S(x, y)}/IPR(x, y, λ). A true PR signal intensity distribution can be obtained by performing the calculation of signal intensity at each point using a computer.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

The present invention relates to a photoreflectance mapping measurement method for evaluating the quality of a crystal for a semiconductor device.

[0002]

2. Description of the Related Art Non-destructive evaluation techniques for the electrical and optical properties of crystals for semiconductor devices have become increasingly important. This technology is extremely important in sorting semiconductor wafers and managing quality and reliability in semiconductor device fabrication. Heretofore, a photoluminescence (PL) method, which is a noncontact optical method, has been mainly applied to the evaluation of an epitaxially grown crystal of a compound semiconductor. PL
According to the method, the band gap of a semiconductor crystal can be measured, and impurities can be identified and quantified. This technique is to observe light emission due to recombination of electrons and holes generated by photoexcitation, and is a technique widely used mainly for evaluation of compound semiconductor crystals. This method is also applied to a mapping method for measuring the distribution of the PL intensity on the sample and the peak wavelength of the PL spectrum. The mapping method is capable of distributing the film thickness and mixed crystal composition of a sample, and is a very effective method for evaluating the uniformity of an epitaxial wafer.

[0003] However, although the PL method can mainly obtain information on a flat band region in a crystal, it has a problem that it is difficult to obtain information on a region where an internal electric field exists. In a semiconductor epitaxial crystal having an actual electronic or optical device structure, a region where an electric field exists, such as a pn junction or a heterointerface, exists, and plays an important role in device operation. In such a region, an electric field is present, so that electrons and holes generated by photoexcitation are separated, and recombination of these electrons is unlikely to occur. Therefore, in order to obtain information in this area by the PL method, the sample must be irradiated with strong excitation light. In such a case, the influence of the generation of crystal defects due to the irradiation of the high-intensity light may be considered, and there is a concern that the original nondestructive measurement may be lost.

In order to overcome the problems of the PL method, it is expected to establish a nondestructive measurement method capable of measuring a sample in which an internal electric field is present, even with a very weak excitation which is hardly affected by the measurement. I was Under the circumstances described above, research on nondestructive measurement methods other than the PL method has become active. Photoreflectance method (Photoreflectance: PR) or PR method and spectral ellipsometry (S)
E) Photo-ellipsometry (P)
The E) method is a typical example. In the PR method and the PE method, a sample is irradiated with two beams of excitation light and probe light. The method modulates the internal electric field of the crystal by the accumulation of excess electrons and holes generated by the irradiation of the excitation light, and measures the change in the reflectivity of the probe light due to the electric field modulation. This method has a great feature that information on a region where an internal electric field exists can be extracted, and it is possible to measure a band gap of a semiconductor crystal, an internal electric field strength, and a crystal defect density serving as a recombination center. For this reason, it is preferable for evaluating a device epitaxial crystal having an internal electric field. In addition, since the measurement can be performed with a weak excitation intensity as compared with the PL method, it can be said that the measurement method is essentially non-destructive.

If this PR method is applied to mapping measurement, it becomes possible to evaluate the composition of a mixed crystal semiconductor, the distribution of crystal defects, and the like in a sample surface such as a semiconductor wafer. Further, by obtaining the PR spectrum at each point, it becomes possible to measure the electric field intensity distribution which cannot be evaluated by the PL method. We have realized for the first time the mapping of semiconductor wafers by the PR method, which has not been reported before, and have confirmed the effectiveness of this method as described above. However, in the process of realizing the PR mapping method, we remove the PL signal generated by the excitation light irradiation and the scattered light of the excitation light from the sample surface and extract only the PR signal in the usual PR spectrum measurement and PR mapping measurement. I found that the procedure of doing was important.

In an actual PR measurement, a PR signal is expressed by equation (2). ΔR (λ) = R _on (λ) −R _Off (λ) (1) ΔR (λ) / R (λ) (2) R _on is the reflectance of the probe light when the excitation light is irradiated, and R _Off is the excitation light Is the difference between the probe light reflectances R _on and R _Off , and λ is the wavelength.

However, not only the reflection intensity due to the reflectance R _on of the probe light, which should be originally captured, but also the PL signal from the sample at the time of excitation light irradiation is captured by the photodetector in actual measurement. Is a big problem. That is, the measured ΔR includes the PL signal intensity I _PL . In addition, the excitation scattered light intensity I _{S of} the excitation light on the sample surface may be taken into the photodetector. The reflection intensity due to the reflectance R _on of the probe light when the excitation light is irradiated is represented by Ip.
R _On , the reflectance R of the probe light when the excitation light is not irradiated
_The reflection intensity due to _Off is IpR _Off and the reflectance R of the probe light is
Assuming that the reflection intensity due to is IpR, the total signal intensity I _total detected by the photodetector at the time of excitation light irradiation is I _total = I pR _On + I _PL + I _S (3) Therefore, the true PR signal ΔR _real / R is ΔR _real / R = (R _on −R _Off ) / R = {(I _total −I _PL −I _S ) −I pR _Off } / IpR (4)

In a normal PR measurement system, the wavelength of the probe light is made monochromatic by a spectroscope and irradiated, and the PR spectrum is obtained by scanning the wavelength. As described above, the photodetector simultaneously detects the scattered light and the PL signal generated by the reflection of the probe light from the sample surface and the irradiation of the excitation light. In order to remove the scattered light of the excitation light and the PL signal, a filter having an appropriate wavelength characteristic is provided immediately before the photodetector. However, when measuring a PR signal appearing at a wavelength near the band gap of each semiconductor, it is inevitable to take in the PL signal of the semiconductor to be measured, and it is difficult to remove it with a filter. This problem can be avoided by using a PR signal in a wavelength range other than the vicinity of the band gap, but in that case, the measurement becomes difficult because the signal intensity becomes small.

As described above, in order to obtain a true PR signal, it is essential to remove the intensity of the PL signal and the scattered excitation light. Since the intensity of the excitation light originally used in the PR method is weaker than that of the ordinary PL method, it has been considered that the PL signal intensity I _PL taken into the photodetector can be sufficiently ignored. However, through experiments, we have found that, depending on the sample, the intensity of the PL signal taken into the photodetector may be greater than the PR signal, which is a great obstacle to PR measurement.

[0010] This problem is particularly apparent in mapping measurement. In the PR spectrum measurement, the above PL
Increment in strength due to the signal intensity and the excitation scattered light intensity I _PL + I _S takes a constant intensity that is independent of wavelength. Therefore, in the spectral measurement, because of the observed spectral shape of the difference ΔR of the reflectance being measured at any time, that the PL signal intensity and the excitation scattered light intensity I _PL + I _S is included in the total signal strength It's easy to know.

[0011]

However, in the conventional photoreflectance mapping measurement method, during the measurement of the intensity distribution of the PR signal, the PL signal intensity and the excitation scattered light included in the distribution of the _total signal intensity I _total are measured. It is not easy to separate the distribution of the intensity I _PL + I _S. In particular, as shown in the examples described later, in a sample in which the PL signal intensity is very large compared to the PR signal intensity, the distribution of the _total signal intensity I _total , the PL signal intensity and the excitation scattered light intensity I _PL + I _S Is sometimes hardly changed, which is a problem for knowing the true PR signal intensity distribution. That is, there is a problem that it is difficult to extract only the PR signal intensity with high accuracy and remove other signals.

[0012] The present invention has been made in view of such a point, and the PR from the total signal intensity taken in by the photodetector.
It is an object of the present invention to provide a photoreflectance mapping measurement method capable of extracting a signal intensity and measuring a PR spectrum and a PR signal intensity distribution with high accuracy.

[0013]

The above object is achieved by the following methods (1) to (3). (1) In the PR signal intensity distribution measurement, the PL signal intensity distribution not including the PR signal and the scattered light intensity distribution are separately measured. Then, by computer processing, the PL
Subtract signal intensity distribution and scattered light intensity distribution to obtain true PR
Obtain the signal strength distribution. (2) An ordinary PR measurement system is provided with an optical system for detecting only the PL signal and the excitation scattered light separately from the PR signal detection system,
A true PR signal intensity is extracted by subtracting the PL signal intensity and the excitation scattered light intensity from the signal intensity of the PR measurement system. (3) In the PR measurement system, a PL signal + excitation scattered light is separated by chopping the excitation light and the probe light to detect a true PR signal. As a result of experiments and considerations, the present inventors have confirmed that the PR signal and other signals can be separated by the above-described method.

[0014]

DESCRIPTION OF THE PREFERRED EMBODIMENTS In order to solve the above-mentioned problems, a photoreflectance mapping measurement method according to the present invention irradiates two beams, excitation light and probe light, and turns on / off the excitation light.
In a photoreflectance mapping measurement method for measuring the distribution of electrical and optical characteristics on a semiconductor crystal wafer from a change in the reflectivity of probe light due to ff, photoluminescence is obtained from all signal intensity distribution data I (x, y, λ). By subtracting the signal intensity distribution data I _PL (x, y) and the excitation scattered light intensity distribution data I _S (x, y), only the photoreflectance signal is extracted and the photoreflectance signal intensity distribution data I _PR (x , Y, λ).

Further, the photoreflectance mapping measurement method of the present invention irradiates two beams of excitation light and probe light, and changes the reflectivity of the probe light due to the on / off of the excitation light to determine the electric power on the semiconductor crystal wafer. In the photoreflectance mapping measurement method for measuring the distribution of optical and optical characteristics, the reflection intensity IpR _On of the probe light when the excitation light is irradiated by the photodetector 15 of the optical measurement system and the probe when the excitation light is not irradiated detecting the reflection intensity of light iPr _Off and photoluminescence signal intensity I _PL and excitation scattered light intensity I _S, detects a photoluminescence signal intensity I _PL and excitation scattered light intensity I _S by the optical detector 22 of the other optical measuring system And
It is characterized in that only the photoreflectance signal is extracted by subtracting the signal intensity detected by the photodetector 22 from the signal intensity detected by the photodetector 15.

Further, the photoreflectance mapping measurement method of the present invention irradiates two beams of excitation light and probe light, and changes the reflectivity of the probe light due to on / off of the excitation light to determine the electric power on the semiconductor crystal wafer. In the photoreflectance mapping measurement method for measuring the distribution of optical and optical characteristics, the irradiation cycle of the excitation light is set by the chopper 18, the irradiation cycle of the probe light is set by the chopper 26, and the excitation light is The reflection intensity IpR _On of the probe light when illuminated, the reflection intensity IpR _Off of the probe light when the excitation light is not illuminated, the photoluminescence signal intensity I _PL and the excitation scattered light intensity I _S are detected. performing synchronous detection by chopping cycle of light to remove the photoluminescence signal intensity I _PL and excitation scattered light intensity I _S By removing the reflective intensity iPr _Off of the probe light when not irradiated with performing synchronous detection excitation light chopping cycle of the excitation light in the lock-in amplifier 28, characterized in that to extract only the photo-reflectance signal ing.

[0017]

An embodiment of the present invention will be described below with reference to the drawings. FIG. 1 is a diagram showing a procedure of a photoreflectance mapping measurement method in the first embodiment of the present invention. In the figure, 1 is the total signal intensity distribution data, 2 is the PL + excitation scattered light intensity distribution data, 3 is the probe light reflection intensity distribution data when the excitation light is off, and 4 is the PR signal intensity distribution data. In normal PR mapping measurement, the PR signal intensity distribution for each wavelength is measured by fixing the wavelength of the probe light and scanning the irradiation position on the sample. The total signal intensity I and the reflectance R of the probe light at the time of the excitation light off by this measurement are represented by the wavelength λ and the measurement position (x,
y). Also, stop by the shutter irradiation of the probe light to the sample, the excitation scattered light intensity I _S from PL signal intensity I _PL and the surface due to excitation light irradiation is a function of the measurement position on the sample (x, y) .

The total signal intensity distribution data 1 is I (x, y,
λ), PL + excitation scattered light intensity distribution data 2 is I
_{PL + S} (x, y), probe light reflection intensity distribution data 3 at the time of excitation light off is represented by R (x, y, λ), and PR signal intensity distribution data 4 is represented by I _PR (x, y, λ). . The PR signal intensity I _PR is calculated based on the wavelength λ and the measurement position (x,
y). That is, I _PR (x, y, λ) = {I (x, y, λ) −I _{PL + S} (x, y)} / R (x, y, λ) (5) By calculating the above equation on a computer for the signal strength of, a true PR signal strength distribution can be obtained.

FIG. 2 is a graph showing a P value for implementing the photoreflectance mapping measurement method according to the second embodiment of the present invention.
It is a schematic diagram of an R measurement system. In the figure, 11 is an excitation light source, 12 is a probe light source, 13 is a spectroscope, 14 is a filter, 15 is a photodetector, 16 is a preamplifier, 17 is a lock-in amplifier, 18 is a chopper, and 19 is XY for mapping. A movable stage, 20 is a sample, 21 is a filter,
22, a photodetector; 23, a preamplifier; 24, a lock-in amplifier; and 25, a computer.

Next, the operation will be described. The photodetector 15 for detecting the ordinary PR signal intensity has a reflection intensity IpR _On of the probe light when the excitation light is irradiated, a reflection intensity IpR _Off of the probe light when the excitation light is not irradiated, a photoluminescence signal intensity I _PL, The excitation scattered light intensity _IS is detected. The lock-in amplifier 17 performs synchronous detection using the signal of the pump light from the chopper 18 as a reference signal to remove IpR _Off . On the other hand, a photodetector 22 for detecting the PL signal intensity I _PL and the excitation scattered light intensity I _S is provided for the PR measurement system. The photodetector 22 is disposed at a position optically equivalent to the photodetector 15 with respect to the irradiation position of the probe light and the excitation light on the sample 20. If amplification is performed by the lock-in amplifier 24 using the signal of the excitation light from the chopper 18 as a reference signal, only the PL signal intensity I _PL and the excitation scattered light intensity I _S can be extracted. Further, the signal processing by the computer 25 extracts the difference between the signal I ₂₂ from the signal I ₁₅ and the lock-in amplifier 24 from the lock-in amplifier 17. _{That, I 15 = IpR On -IpR Off} + I PL + I S (6) I 22 = I PL + I S (7) I PR = (I 15 -I 22) / R = (IpR On -IpR Off) / R ( 8) where I _PR represents the PR signal strength. By performing mapping according to the equation (8) obtained by this procedure, a true PR signal intensity distribution can be obtained. True PR signal ΔR / R
Is represented by the following equation. ΔR / R = I _PR (9)

FIG. 3 is a diagram showing the relationship between the chopping cycle of the excitation light and the probe light and the signal intensity according to the photoreflectance mapping measurement method in the third embodiment of the present invention. In the third embodiment, not only the excitation light but also the probe light is chopped. In the figure, the chopping frequency f _E of the pump light and the chopping frequency f _{P of the} probe light are shown.
The relationship between, and _{f E = 2f P (10)} , timing the respective light irradiation. At this time, the signal intensity I is incorporated into the _{photodetector, (a) I = IpR On} + (I PL + I S), (b) I = IpR Off, (c) I = I PL + I S, (d) I = 0. Therefore, (a) the difference _{(c) (IpR On + I} PL + I S) - by taking (I _PL + I _S), and extracted iPr _On, further and this IpR _On (b)
By obtaining the difference between iPr _Off, it is possible to extract the _{ΔR / R = (R on -R} Off) / R = (IpR On -IpR Off) / IpR (11) next to the true PR signal.

FIG. 4 is a diagram showing a P for implementing the photoreflectance mapping measurement method according to the third embodiment of the present invention.
It is a schematic diagram of an R measurement system. In the figure, 26 is a chopper, 27 is a lock-in amplifier, and 28 is a lock-in amplifier. Elements denoted by the same reference numerals as those in FIG. 2 indicate the same elements, and the description thereof will be omitted. First, signals IpR _On + I _PL + I _S , IpR _Off , I
_PL + I _S is synchronously detected by the lock-in amplifier 27 at the chopping frequency f _P of the probe light, and I _PL + I S
Remove _S. The analog signal IpR obtained by this
_On is further input to the lock-in amplifier 28. In the lock-in amplifier 28, the chopping frequency f _{E of the} pump light
To perform synchronous detection to remove IpR _Off . By performing the above-described signal processing, the PL signal intensity I _PL and the surface scattered light intensity I _S of the excitation light are removed, and the PR signal intensity I _PR can be measured with high accuracy. The use of this signal enables the mapping of the true PR signal strength.

FIG. 5 shows an example of PR signal extraction by the photoreflectance mapping measurement method according to the embodiment of the present invention, and shows the result of PR mapping measurement of an epitaxial wafer having an InAlAs / InGaAs-HEMT structure on an InP substrate. The measurement was performed at 1400 nm, which is a wavelength at which a signal from the InGaAs layer was detected. (A) is P
ΔR ′ including L intensity and scattering intensity (= I pR _On + I _PL + I
_S- IpR _Off ), (b) shows the distribution of PL and excitation scattered light intensity (I _PL + I _S ), and (c) shows the intensity distribution of the extracted PR signal ΔR = (R _on −R _Off ). FIG.
(A) and (b) show almost the same distribution. This indicates that at this wavelength, the PL intensity from the InGaAs layer is largely dominant in all the signals, so that the intensity distribution of the PR signal is buried in the (PL intensity + scattered light) distribution.
Therefore, the distribution of (c) is obtained by subtracting the distribution of (b) from the distribution of (a). This shows the intensity distribution of the PR signal. By performing this processing, the true intensity distribution due to the PR signal, which was not seen in the unprocessed data, came to be clearly observed. Although this is based on the first embodiment of the present invention, similar effects can be obtained in the second embodiment and the third embodiment.

[0024]

As described above, the photoreflectance mapping measurement method of the present invention can extract a photoluminescence signal and a true PR signal from which excitation scattered light is removed from the total signal intensity taken into the photodetector. So P
R measurement becomes more accurate. Further, the technique of separating the PL signal and the PR signal can be applied to the combination of the PL measurement device and the PR measurement device. INDUSTRIAL APPLICATION This invention can be applied to evaluation of various heterogeneous semiconductor laminated structure samples, and has the effect that the practical application range is wide.

[Brief description of the drawings]

FIG. 1 is a diagram showing a procedure of a photoreflectance mapping measurement method according to a first embodiment of the present invention.

FIG. 2 is a schematic diagram of a PR measurement system for performing a photoreflectance mapping measurement method according to a second embodiment of the present invention.

FIG. 3 is a diagram illustrating a relationship between a chopping cycle of excitation light and probe light and a signal intensity according to a photoreflectance mapping measurement method according to a third embodiment of the present invention.

FIG. 4 is a schematic diagram of a PR measurement system for implementing a photoreflectance mapping measurement method according to a third embodiment of the present invention.

FIG. 5 is an example of PR signal extraction by a photoreflectance mapping measurement method according to an embodiment of the present invention;
(A) is an intensity distribution of ΔR ′ (= R _on + I _PL + I _S −R _Off ), (b) is a distribution of PL and excitation scattered light intensity (I _PL + I _S ), and (c) is an extracted PR signal ΔR. = (R _on -R
FIG. 4 is a diagram showing an intensity distribution of _Off ).

[Explanation of symbols]

Reference Signs List 1 Total signal intensity distribution data 2 PL + excitation scattered light intensity distribution data 3 Probe light reflection intensity distribution data when excitation light is off 4 PR signal intensity distribution data 11 Excitation light source 12 Probe light source 13 Spectroscope 14 Filter 15 Photodetector 16 Preamplifier 17 Lock-in amplifier 18 Chopper 19 XY movable stage for mapping 20 Sample 21 Filter 22 Photodetector 23 Preamplifier 24 Lock-in amplifier 25 Calculator 26 Chopper 27 Lock-in amplifier 28 Lock-in amplifier R _on. Reflectance of probe light when irradiated R _Off: Reflectance of probe light when not irradiated with excitation light ΔR: Difference in reflectance of probe light between when and not irradiated with excitation light R _on -R _Off R ····· Probe light reflectance ΔR / R ··· PR signal I _{1 5} ... Output signal (I) from lock-in amplifier 17
pR _On -IpR _Off + I _PL + I _S ) I ₂₂ ... Output signal from lock-in amplifier 24 (I
_PL + I _S ) I _PL ··· PL intensity captured by the photodetector I _S ··· Excitation scattered light intensity captured by the photodetector I _PR ··· PR signal intensity IpR ··· Probe light of reflection intensity IpR _On reflection intensity of the probe light when not irradiated with reflection intensity IpR _Off ·· excitation light of ... the probe light when irradiated with excitation light f _E ···· chopping frequency f _P ···· Chopping frequency

──────────────────────────────────────────────────の Continued on the front page (51) Int.Cl. ⁶ Identification code FI H01L 21/66 H01L 21/66 L

Claims (3)

[Claims] 1. A photometer for irradiating two beams of an excitation light and a probe light and measuring a distribution of electrical and optical characteristics on a semiconductor crystal wafer from a change in a reflectance of the probe light due to on / off of the excitation light. In the reflectivity mapping measurement method, the photoluminescence signal intensity distribution data (I _PL (x, y)) is converted from the total signal intensity distribution data (I (x, y, λ)).
By subtracting the excitation scattered light intensity distribution data (I _S (x, y)), only the photoreflectance signal is extracted, and the photoreflectance signal intensity distribution data (I _PR (x, y)
y, λ)) is measured. 2. A photometer which irradiates two beams, excitation light and probe light, and measures the distribution of electrical and optical characteristics on a semiconductor crystal wafer from a change in reflectivity of the probe light due to on / off of the excitation light. In the reflectivity mapping measurement method, the reflection intensity (IpR _On ) of the probe light when the excitation light is irradiated by the photodetector (15) of the optical measurement system and the reflection intensity (IpR _Off ) of the probe light when the excitation light is not irradiated ) and photoluminescence signal intensity (I _PL) and detects the excitation scattered light intensity (I _S), photoluminescence signal strength on other optical measuring system of the optical detector (22) (I _PL) and excitation scattered light intensity ( I _s ), and from the signal intensity detected by the photodetector (15), the photodetector (2
A photoreflectance mapping measurement method characterized in that only the photoreflectance signal is extracted by subtracting the signal intensity detected in 2). 3. A photo-irradiator that irradiates two beams, excitation light and probe light, and measures the distribution of electrical and optical characteristics on the semiconductor crystal wafer from a change in the reflectance of the probe light due to the on / off of the excitation light. In the reflectivity mapping measurement method, the irradiation cycle of the excitation light is set by the chopper (18), the irradiation cycle of the probe light is set by the chopper (26), and the probe when the excitation light is irradiated by the photodetector (15) The lock-in amplifier detects the light reflection intensity (IpR _On ), the probe light reflection intensity (IpR _Off ), the photoluminescence signal intensity (I _PL ), and the excitation scattered light intensity (I _S ) when the excitation light is not irradiated. (27) performing synchronous detection by chopping cycle of the probe light to remove the photoluminescence signal intensity (I _PL) and excitation scattered light intensity (I _S) in, locked By removing the reflective intensity of the probe light when not irradiated with excitation light by performing synchronous detection by chopping cycle of the excitation light in-amplifier (28) (IpR _Off), characterized by extracting only the photo-reflectance signal Photoreflectance mapping measurement method.