JPS6152569B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an inspection method in a semiconductor annealing process.

In recent years, various new methods and techniques have been introduced in the process of creating semiconductor materials and processing them into devices. Ion implantation into semiconductors is one such technique and is well known to those skilled in the art. It is known that when ions are implanted into a semiconductor, the structure of the semiconductor in the ion implanted region and the surface region is significantly disturbed. If the semiconductor is single crystal before ion implantation, the surface region of the semiconductor is amorphous or polycrystalline after ion implantation. Furthermore, impurity ions implanted into a semiconductor do not function as electrically active carriers such as donors or acceptors if they are simply introduced into the semiconductor. Therefore, in order to recover from the damage caused by ion implantation and to make the implanted impurities electrically active, so-called annealing is performed. Semiconductor annealing has traditionally been performed by applying heat in an electric furnace, but recently a so-called laser annealing method has been developed in which the semiconductor is irradiated with a laser beam. Annealing methods using electron beams in addition to laser beams have also been developed.

In these semiconductor annealing processes, it is necessary to apply a certain amount of energy to the semiconductor material in order to recover from damage and activate the implanted impurities, but if too much energy is applied, the crystal structure may deteriorate. This not only disturbs the environment, but is also disadvantageous in terms of material or device processing. Therefore, if there were a method to evaluate the degree of disorder of the crystal structure of the semiconductor or the activation of impurities at the same time as annealing or during a short period of time when annealing is temporarily interrupted, or after annealing, it would be possible to carry out annealing at an appropriate stage. It can be stopped or a suitably annealed material can be selected.

On the other hand, in connection with the development of inexpensive solar cells, silicon (Si) is being used on various substrates such as metal and glass.
Research is being actively conducted to form various amorphous semiconductors such as . In this case as well, annealing is performed in the same manner as in ion implantation for the purpose of improving the characteristics and crystallizing the amorphous semiconductor. Moreover, it is extremely desirable to obtain information regarding the crystallinity of the semiconductor during annealing, just as in the case of ion implantation.

The present invention relates to a method of monitoring various characteristics of a semiconductor during an annealing process by utilizing the Raman effect of a semiconductor, and determining and inspecting the degree or quality of annealing. The present invention relates to an inspection method suitable for use in semiconductor annealing. In the present invention, the Raman scattered light generated when a semiconductor is irradiated with laser light contains information about various characteristics such as the structural disorder of the semiconductor and the carrier concentration in the semiconductor. This is based on the discovery that the wavelength (wave number), intensity, half-width, etc. of

The present invention will be described in detail with reference to the drawings.

Figure 1 shows a Raman spectrum obtained when a GaAs single crystal (100) plane is irradiated with laser light, and one peak P ₁ is observed at a wave number of approximately 292 cm ^-1 . This peak is known to correspond to the Raman band caused by the LO phonon, which is unique to GaAs, and only the LO phonon is observed as Raman light from the (100) plane of a relatively perfect GaAs single crystal. , the wave number is unique to GaAs single crystal.

FIG. 2 shows the configuration of a laser annealing device that also serves as a Raman effect monitor used in an embodiment of the present invention. The Raman effect is measured using a commercially available laser Raman spectrophotometer 1 (JRS-400T manufactured by JEOL Ltd.), and the measurement results are displayed on a recorder 2 or a cathode ray tube display 3. One argon laser oscillator 4 was used for both annealing the semiconductor sample 5 and measuring the laser Raman effect.
The wavelength of the argon laser can be changed to 5145 Å or 4880 Å, and the maximum power is 2W. The laser beam diameter can be varied from less than 50 μm to several tens of mm.

Figure 3 shows the Raman analysis of a sample in which a relatively high dose (2×10 ¹³ /cm ² ) of Si ions was implanted into the surface of a semi-insulating GaAs single crystal during annealing using the laser annealing device described above. Shows changes in spectrum. Specifically, the laser output
Annealing was performed at 2 W, and in the middle of the annealing, the laser output was reduced to 500 mW to substantially stop the annealing, and a wavenumber sweep was performed to measure the Raman spectrum. The measurement plane is the (100) plane.

From Figure 3, the intensity of the Raman band peak P 1 at a wave number of 292 cm ^-1 due to LO phon increases as the annealing is repeated from pre-annealing (i) to annealing periods of 1, 2 minutes (ii), and ₂ minutes (iii). together with the LO
It can be seen that the difference in wave numbers indicating the band width of the Raman band corresponding to the phon, eg, the half-width of the peak _P1 or the intensity of the peak intensity 1/e, becomes gradually narrower. However, when annealing is continued further, as shown in (iv), the intensity of peak P ₁ decreases and the half-width widens again, and a new peak P ² appears at a wave number of around 272 cm ^-1 .

FIG. 4, FIG. 5, and FIG. 6 are diagrams showing the relationship between the annealing period and the peak intensity of _P1 , the half-width of _P1 , and the peak intensity of _P2, respectively. Electrical carrier concentration measurement using the Hall effect shows that the change in period T ₁ in FIGS. 4 and ₅ is simply a crystal structure repair stage with no change in carrier concentration; It has been confirmed that the changes in Figures 5, 6, and 6 are due to the influence of plasmons caused by the impurity in the ion-implanted semiconductor becoming electrically active as a donor or acceptor and the carriers in the semiconductor increasing. Ta. That is, the peak P ₂ is due to a Raman band corresponding to plasmons in the semiconductor, and the intensity of the peak P ₁ in the period T ₂ decreases due to the increase in the plasmons.
Half price width increases.

Therefore, the intensity of the peak P ₂ of the plasmon mode in the Raman spectrum corresponds to the carrier concentration, so by observing the intensity of the peak P ₂ , it can be determined that the impurity implanted into the semiconductor exhibits the desired electrical behavior. It is possible to judge whether the child has become old.

Figure 7 shows the relationship between the peak intensity of the plasmon mode and the carrier concentration measured using the Hall effect for a sample in which 2×10 ¹⁴ /cm ² of Si was ion-implanted into semi-insulating GaAs. be. Then, the peak intensity of the plasmon mode was monitored while performing laser annealing on another new sample, and the annealing was stopped when the peak intensity reached 1.0 in Fig. 7, that is, 1 × 10 ¹⁹ /cm ³ in terms of carrier concentration. When the carrier concentration was measured using the Hall effect, the carrier concentration was 9.5×10 ¹⁸ /cm ³ , indicating that the carrier concentration can be determined from the peak intensity of the plasmon mode using the relationship shown in Figure 7. It was done. Moreover, since the method of monitoring the peak intensity uses a fixed spectrometer, the measurement can be carried out almost instantaneously, and therefore the carrier concentration can be accurately monitored while annealing is being performed.

It should be noted here that the effect of plasmons in the semiconductor appears as a change in the Raman spectrum only when the carrier concentration in the semiconductor is approximately 1×10 ¹⁷ /cm ³ or higher. Therefore, in the case of an ion-implanted sample with a low toes amount (specifically, for example, 2×10 ¹² /cm ² or less) where the carrier concentration is lower than this, changes in the peak and half-width due to plasmon will not occur. Therefore, it is not possible to monitor the carrier concentration according to the present invention as described above.

The next thing to note is that when the crystal structure begins to be destroyed due to excessive annealing, a peak similar to that in period T ₂ will appear in period T ₃ , as shown in FIGS. 4 and 5. This is the point at which a decrease and an increase in half-width appear. However, this period
The change at T ₃ can be distinguished from the change at period T ₂ due to the influence of plasmons as follows. In other words, in the ion-implanted sample with a low ion dose, the decrease in the peak and the increase in the half-width due to plasmons do not appear as described above, but the decrease in the peak and the increase in the half-width due to the destruction of the crystal structure do not occur in the ion-implanted sample with a high dose. Appears similar to the injected sample.
FIG. 8 shows only the decrease in peaks. Therefore, if the time at which the crystal structure begins to break down is measured in advance using a low-dose ionized sample, it can be determined that the decrease in the peak and the increase in the half-width before that time are due to the influence of plasmons.

As described in detail above, according to the present invention, the degree of annealing in a semiconductor annealing process or the quality of the annealing result can be determined and inspected by utilizing the Raman effect.

In the apparatus shown in FIG. 2 described above, one laser oscillator is used for annealing and Raman measurement, but the invention is not limited to this. Separate laser oscillators may be provided for annealing and inspection, or even for annealing and Raman measurement. It goes without saying that this may be performed using other types of annealing equipment that do not use lasers.

[Brief explanation of the drawing]

Figure 1 shows the Raman spectrum of the GaAa single crystal (100) plane, Figure 2 shows the configuration of a laser annealing device that also serves as a Raman effect monitor, and Figure 3
The figure shows the changes in the Raman spectrum during the annealing process. Figures 4, 5, and 6 show the intensity of P ₁ , half-width of P ₁ , and P ₂ during the annealing process.
Figure 7 is a diagram showing the relationship between the intensity of P ₂ and carrier concentration, and Figure 8 is a diagram showing the change in intensity of P ₁ during the annealing process for an ion implanted sample with a low ion dose. It is.

Claims (1)

[Claims] 1. The surface of the semiconductor that has undergone annealing treatment is irradiated with laser light, the resulting Raman light is introduced into a spectrophotometer, and the impurities added to the semiconductor in the spectrum of the Raman light are detected by the annealing treatment. An annealing process characterized in that the degree of the annealing process is inspected by detecting the peak intensity of a Raman band due to plasmons generated in the semiconductor as carriers increase due to electrical activation. related semiconductor inspection methods.