JPWO2008123242A1 - Epitaxial growth substrate and epitaxial growth method - Google Patents

Abstract

Provided are an epitaxial growth substrate and an epitaxial growth method capable of making the PL characteristics of an epitaxial layer constant when an epitaxial layer is grown on a III-V group compound semiconductor substrate such as InP. In the process of vapor-phase growth of an epitaxial layer made of a compound semiconductor such as an InGaAsP layer on a III-V group semiconductor substrate such as InP, the substrate temperature (long diameter) is caused by the size (major axis) of the elliptical etch pits on the back surface of the substrate. The set temperature was appropriately controlled in consideration of the change in the growth temperature.

Description

  The present invention relates to an epitaxial growth substrate and an epitaxial growth method for growing an epitaxial layer on the substrate, and more particularly to a technique for improving photoluminescence (hereinafter abbreviated as PL (Photoluminescence)) characteristics of the grown epitaxial layer.

Conventionally, for semiconductor device applications for optical devices, an epitaxial layer made of a compound semiconductor such as an InGaAs layer, an AlGaAs layer, an AlInAs layer, an AlInGaAs layer, or an InGaAsP layer is formed on an InP substrate by metal organic chemical vapor deposition (MOCVD method). Semiconductor devices grown by the above are widely used. Among these, In _1-x Ga _x As _y P _1-y can be lattice-matched with the InP substrate relatively easily, and thus can be said to be suitable for obtaining a semiconductor element having excellent element characteristics.

  In order to use the semiconductor element made of the above-described III-V group compound semiconductor for a laser or the like, it is desirable that the PL peak wavelength and the PL emission intensity are constant. However, it is known that even if epitaxial growth is performed under exactly the same conditions, PL characteristics after epitaxial growth vary depending on the characteristics of the substrate used for epitaxial growth.

  Further, it has been found by experiments so far that it is important to make the substrate thickness and carrier concentration used for epitaxial growth constant in order to make the PL characteristics of the epitaxial growth layer constant.

For example, in Patent Document 1, a carrier concentration at a room temperature of a semiconductor substrate is measured in advance, and the actual substrate surface temperature becomes a desired temperature regardless of the carrier concentration of the semiconductor substrate. Thus, the set temperature of the substrate is controlled to grow the epitaxial layer.
JP-A-2005-231909

  However, there is a problem that even if the substrate thickness and the carrier concentration of the substrate are unified as in the prior application technique, the PL characteristics of the obtained semiconductor elements vary.

  An object of the present invention is to provide an epitaxial growth substrate and an epitaxial growth method capable of making the PL characteristics of an epitaxial layer constant when an epitaxial layer is grown on a III-V group compound semiconductor substrate such as InP. And

The present invention has been made to solve the above problems, and is a substrate for epitaxial growth made of a III-V group compound semiconductor, wherein the maximum value of the reflection density measured in the [011] direction is 0.5 to 2. 0, the maximum value of the reflection density measured in the [0-11] direction is in the range of 0.5 to 1.5, and the standard deviation of the measured value in each measurement direction is within 10% of the average value. It is characterized by being.
Preferably, the maximum value of the reflection density measured in the [011] direction is from 0.5 to 1.6, and the maximum value of the reflection density measured in the [0-11] direction is from 0.5 to 1.3. To be.

  By epitaxial growth using such a substrate, variations in the PL emission intensity of the epitaxial layer can be made within a predetermined range. In particular, by defining the standard deviation as an index of the variation in the reflection density (measured value) on the back surface of the substrate, it is possible to effectively reduce the variation in the PL emission intensity of the epitaxial growth layer when epitaxially growing using this substrate. it can.

  Here, the “reflection density” is defined by a common logarithm of the reciprocal of the reflectance, and is defined by JIS B9622.

  The reflection density data shown in the present application is measured in a monochrome mode of DM-400 “Color / monochrome reflection densitometer” (measurement range: 0.5 to 2.5, measurement area: 4 mmφ) manufactured by Dainippon Screen. It is a numerical value. For example, the reflection density is measured at several locations (for example, four locations in 90 ° increments) at the central portion of the substrate and the peripheral portion, and the maximum value, average value, and standard deviation are calculated from these measured values.

  In addition, when the value is negative, the direction is generally expressed by adding “-” on the number, but in this specification and the claims, “-” is added before the number. It shall be expressed as

  In addition, the major axis of the elliptical etch pits existing on the back surface of the substrate is in the range of 5 to 40 μm, and the standard deviation of the major axis of each etch pit is within 20% of the average value. There is a correlation between the degree of the major axis of the elliptical etch pit on the back side of the substrate and the difference between the set temperature during epitaxial growth and the actual substrate temperature, and the growth conditions for achieving the desired growth temperature by using this relationship (Setting temperature) can be easily determined. In addition, by defining the standard deviation of the major axis as an index of the major axis variation of the elliptical etch pits existing on the back surface of the substrate, the PL wavelength variation of the epitaxial growth layer when this substrate is epitaxially grown can be reduced. Can do.

  In the epitaxial growth substrate having the above-described properties, the back surface of the substrate is etched at a temperature of 15 to 100 ° C. with an etching solution comprising 1 to 10 parts of phosphoric acid, 1 to 10 parts of hydrogen peroxide solution, and 0 to 10 parts of water. Can be obtained.

  Further, in the epitaxial growth method, the epitaxial growth conditions are controlled according to the reflection density on the back surface of the substrate and the long diameter of the elliptical etch pits to grow the epitaxial layer. Thereby, the variation in the PL emission intensity of the epitaxial layer can be made within a predetermined range.

  Further, when two or more epitaxial growth substrates are simultaneously epitaxially grown in the same furnace, the epitaxial growth substrate is a substrate having the above-described properties, and the average value of the reflection density on the back surface of each substrate is all. Epitaxial growth is performed using a substrate that is within ± 10% of the average of the input substrates and whose average length of elliptical etch pits existing on the back surface of the substrate is within ± 20% of the average of all input substrates. I did it. Thereby, even when epitaxial growth is simultaneously performed on a plurality of substrates, the PL characteristics of the epitaxial layer can be kept constant.

The following is a brief description of how the present invention was completed.
First, Patent Document 1 discloses that when the carrier concentration of the substrate is high, the surface temperature of the substrate becomes low even at the same set temperature, and as a result, the PL peak wavelength becomes long. It is also disclosed that there is a good correlation between the carrier concentration of the substrate and the PL peak wavelength. That is, it is known that the carrier concentration of the substrate is important as a substrate characteristic for determining the epitaxial growth conditions (specifically, the set temperature).

  However, it has been found that even when epitaxial growth is performed by controlling the growth conditions based on the carrier concentration of the substrate, the PL characteristics of the obtained semiconductor elements vary. Therefore, the present inventors have speculated that there are substrate characteristics that affect the epitaxial growth conditions in addition to the carrier concentration of the substrate, and studied the substrate back surface characteristics.

  Specifically, the substrate back surface properties are quantitatively grasped by measuring the reflection density on the back surface of the substrate, the shape and size of the back surface etch pits, and the PL characteristics (PL emission intensity, PL of the semiconductor element obtained by epitaxial growth). Wavelength).

The results are shown in FIGS.
FIG. 1 is an explanatory diagram showing the relationship between the reflection density (maximum value) on the back surface of a substrate and the PL emission intensity of a semiconductor element obtained by epitaxial growth. FIG. 1 shows the maximum value of the reflection density measured from the [011] direction.

  As shown in FIG. 1, there is a good correlation between the maximum value of the reflection density on the back surface of the substrate and the PL emission intensity. Thus, by using a substrate having a maximum reflection density on the back surface of the substrate in a predetermined range (for example, 0.5 to) and growing an epitaxial layer on the substrate, the variation in the PL emission intensity is predetermined. It turned out to be in range.

  Furthermore, when the experiment was repeated, the maximum value of the reflection density measured in the [011] direction was 0.5 to 2.0, and the maximum value of the reflection density measured in the [0-11] direction was 0.5 to 1. It was found that when the substrate in the range of 5 is used, the variation in PL emission intensity of the epitaxial layer can be extremely reduced.

  FIG. 2 is an explanatory diagram showing the relationship between the variation in the reflection density on the back surface of the substrate and the PL emission intensity of the semiconductor element obtained by epitaxial growth. In FIG. 2, the minimum / maximum value of the reflection density measured from the [011] direction is used as an index of the variation in reflection density. In FIG. 85 (standard deviation / average value is 7%, small variation), “♦” uses a substrate having a reflection density minimum / maximum value of 0.65 (standard deviation / average value is 17%, large variation) The normalized PL light emission intensity when it is present is shown.

  The number of samples is the same when the minimum value / maximum value of the reflection density is about 0.85 and when the minimum value / maximum value is about 0.65.

  As shown in FIG. 2, when the minimum value / maximum value of the reflection density is about 0.85, the PL emission intensity variation (standard deviation / average value) is 7%, and the minimum / maximum value of the reflection density is In the case of about 0.65, the PL emission intensity variation (standard deviation / average value) is 17%. That is, it can be seen that the variation in the PL emission intensity is remarkably improved when the variation in the reflection density is small.

  Further, as a result of repeated experiments, when the standard deviation / average value of the reflection density is within 10%, the variation in the PL emission intensity of the semiconductor element can be effectively reduced. Specifically, the PL emission intensity is reduced. It was found that the standard deviation / average value could be 10% or less.

  FIG. 3 is an explanatory diagram showing the relationship between the major axis of the back surface etch pits of the substrate and the epitaxial growth temperature (substrate temperature). FIG. 3 shows the actual substrate temperature when the growth temperature is set to 640 ° C.

  As shown in FIG. 3, there is a predetermined range (for example, etch pits) between the size of the back surface etch pit of the substrate (specifically, the major axis of the elliptical etch pit) and the actual substrate temperature during growth. There is a good correlation between the major axis and the major axis. Thus, by controlling the shape of the back surface etch pit of the substrate (the major diameter of the elliptical etch pit), it becomes possible to accurately set the temperature for obtaining the desired growth temperature.

  FIG. 4 is an explanatory diagram showing the relationship between the epitaxial growth temperature and the PL peak wavelength of the semiconductor element obtained by epitaxial growth. As shown in FIG. 4, there is a good correlation between the epitaxial growth temperature (actual substrate temperature) and the PL peak wavelength.

  Therefore, a semiconductor element having a desired PL peak wavelength can be obtained by controlling the shape of the back surface etch pits of the substrate and performing epitaxial growth at a desired growth temperature. Since the surface temperature of the substrate changes depending on the substrate thickness and heating method (vapor phase growth apparatus used), the relationship between the set temperature and the actual substrate temperature for the backside etch pit shape of the semiconductor substrate is grasped for each. If this is done, it is easy to set the temperature to bring the substrate temperature to a desired temperature.

  FIG. 5 is an explanatory diagram showing the relationship between the variation in the pit major axis on the back surface of the substrate and the PL peak wavelength of the semiconductor element obtained by epitaxial growth. In FIG. 5, the minimum / maximum value of the pit long diameter is used as an index of variation in the pit long diameter. In FIG. 5, “■” indicates that the minimum / maximum value of the pit long diameter is 0.65 (standard deviation / average value). Is the peak wavelength when using a substrate with a minimum / maximum pit length of 0.33 (standard deviation / average is 34%, large variation). Show.

  Note that the number of samples is the same when the minimum / maximum value of the pit major axis is about 0.65 and when the minimum / maximum value is about 0.33.

  As shown in FIG. 5, when the minimum value / maximum value of the pit major axis is about 0.65, the PL peak wavelength variation (standard deviation / average value) is 0.07%, and the minimum / maximum pit major axis value is 0.07%. When the value is about 0.33, the PL peak wavelength variation (standard deviation / average value) is 0.15%. That is, it can be seen that the variation in the PL peak wavelength is remarkably improved when the variation in the pit major axis is smaller.

  Specifically, when the minimum / maximum pit length is about 0.33, the PL peak wavelength is 1547 to 1555 nm, whereas the minimum / maximum pit length is about 0.65. In the case, the PL peak wavelength was 1548 to 1551 nm.

  Further, as a result of repeated experiments, when the standard deviation / average value of the pit major axis is within 20%, it is possible to effectively reduce the variation in the PL peak wavelength of the epitaxial layer. It was found that the standard deviation / average value could be 0.1% or less.

  Next, the back surface characteristics (reflection density, etch pit major axis) of the substrate and etching conditions were examined.

FIGS. 6 and 7 are explanatory views showing the relationship between the etching amount when using a phosphoric acid-based etching solution and the reflection density on the back surface of the substrate, and the composition (mixing ratio) of the used etching solution is different. That is, FIG. 6 shows the results when the etching solution 1 mixed with phosphoric acid: hydrogen peroxide water: water = 20: 3: 20 is used, and FIG. 7 shows phosphoric acid: hydrogen peroxide water: water. The result when using the etching solution 2 mixed at = 1: 1: 1 is shown.
6 and 7 show the reflection densities measured from the [011] direction (measurement direction 1) and the [0-11] direction (measurement direction 2). Further, this is data when the minimum value / maximum value of the pit major axis is 0.65.

  As shown in FIGS. 6 and 7, there is a good correlation between the etching amount and the reflection density on the back surface of the substrate, and the reflection density increases as the etching amount increases. This shows that the reflection density can be adjusted by the etching amount.

  However, it should be noted that if the measurement direction is different, the reflection density slightly changes even if the etching amount is the same. Moreover, when the composition ratio of the etching solution is different, the reflection density is changed even when the etching amount is the same, but the etching solution is mixed in a ratio of phosphoric acid: hydrogen peroxide: water = 1 to 10: 1 to 10: 0 to 10. By using this, it is possible to realize a desired reflection density and range of etch pits, and also to control the length of the etch pit major axis.

  FIG. 8 is an explanatory diagram showing the relationship between the etching amount when using the phosphoric acid-based etching solutions 1 and 2 and the major diameter of the etch pits on the back surface of the substrate. As shown in FIG. 8, there is a good correlation between the etching amount and the major axis of the etch pit on the back surface of the substrate. This shows that the major axis of the etch pit on the back surface of the substrate can be adjusted by the etching amount.

  As described above, the PL emission intensity and the PL peak wavelength, which are important characteristics of a semiconductor element obtained by epitaxial growth, can be optimized by adjusting the reflection density on the back surface of the substrate and the major axis of the elliptical etch pit on the back surface of the substrate. As a result, the present invention has been completed.

  According to the present invention, in the process of vapor-phase growth of an epitaxial layer made of a compound semiconductor such as an InGaAsP layer on a III-V group semiconductor substrate such as InP, it is caused by the shape and size of the back surface etch pit of the substrate. Considering that the substrate temperature (growth temperature) changes, the substrate temperature can be kept constant at a desired temperature by appropriately setting the heating conditions, so that a semiconductor device having a constant PL peak wavelength can be stably There exists an effect that it can manufacture.

  Further, since a substrate having a reflection density on the back surface of the substrate in a predetermined range is used, a semiconductor element having a PL emission intensity in a predetermined range can be realized.

  Furthermore, when the epitaxial layer is grown on two or more substrates at the same time, the reflection density on the back surface of the substrate and the size (major axis) of the elliptical etch pits are set within a certain range. The PL characteristic of the semiconductor element obtained by epitaxial growth on the substrate is constant.

It is explanatory drawing which shows the relationship between the reflective density | concentration of a substrate back surface, and PL light emission intensity of the semiconductor element obtained by epitaxial growth. It is explanatory drawing which shows the relationship between the variation in the reflection density of a board | substrate back surface, and normalized PL light emission intensity. It is explanatory drawing which shows the relationship between the major axis of the etch pit of a substrate back surface, and epitaxial growth temperature (substrate temperature). It is explanatory drawing which shows the relationship between epitaxial growth temperature and PL peak wavelength. It is explanatory drawing which shows the relationship between the variation of the pit major axis of a substrate back surface, and PL peak wavelength. It is explanatory drawing which shows the relationship between the etching amount at the time of using the phosphoric acid type etching liquid 1, and the reflection density of a board | substrate back surface. It is explanatory drawing which shows the relationship between the etching amount at the time of using the phosphoric acid type etching liquid 2, and the reflection density of a board | substrate back surface. It is explanatory drawing which shows the relationship between the etching amount when using the phosphoric acid type etching liquids 1 and 2, and the major axis of the etch pit of a substrate back surface.

  Hereinafter, as a preferred embodiment of the present invention, a case where a semiconductor device is manufactured by epitaxially growing a group III-V compound semiconductor will be described. Note that the target PL peak wavelength of the semiconductor device to be manufactured is set to 1285 nm.

First, a (100) InP substrate having a predetermined carrier concentration (for example, 1 to 2 × 10 ¹⁸ cm ⁻³ ) was manufactured by a liquid encapsulated Czochralski (LEC) method.

  Next, the InP substrate was etched with a phosphoric acid-based etchant. Specifically, an etching solution mixed at a composition ratio of phosphoric acid: hydrogen peroxide water: water = 1: 1: 1 was used, and etching treatment was performed at 80 ° C. for 12 minutes. At this time, the etching amount was 10 μm. Although the etching amount can be adjusted by the etching temperature and the etching time, the etching temperature is preferably 15 to 100 ° C.

  With respect to the InP substrate after the etching treatment, when the reflection density at a plurality of positions on the back surface was measured, the maximum value of the reflection density from the [011] direction was 0.9, and the maximum reflection density from the [0-11] direction was measured. The value was 0.7. In addition, the standard deviation of the measured reflection density was within 10% of the average reflection density.

  Further, elliptical etch pits are formed on the back surface of the substrate, the major axis is 22 μm on average, and the standard deviation is within 20% of the average value.

Then, an undoped InP layer (film thickness: 0.3 μm), an undoped In _1-x Ga _x As _y P _1-y layer (film thickness: 0.3 μm), and an undoped InP layer (film thickness: 0.3 μm) are formed on the substrate by metal organic chemical vapor deposition. A semiconductor element was manufactured by sequentially growing a film thickness of 0.3 μm. The growth pressure at this time was 40 torr, and the target surface temperature of the substrate was 650 ° C.

  In this embodiment, for the InP substrate, the major axis of the elliptical etch pit on the back surface of the substrate is measured in advance, and the set temperature is set so that the actual substrate temperature is constant at 650 ° C. based on the average major axis of the etch pit. An InGaAsP layer was grown by adjusting.

  Specifically, since an InP substrate having an average major axis of elliptical etch pits on the back surface of the substrate of about 22 μm is used, the substrate heating condition was set to 660 ° C.

  Note that this temperature setting is effective in the present embodiment, and it goes without saying that the temperature setting varies depending on, for example, the vapor phase growth apparatus to be used. In other words, with regard to the vapor phase growth apparatus to be used, if the relationship between the set temperature for the major axis of the elliptical etch pit on the back surface of the substrate and the actual substrate temperature is known, the temperature setting for setting the substrate temperature to the desired temperature is Can be easily determined. The same applies to the substrate thickness.

  When PL characteristics of the semiconductor element obtained by the above-described method were measured, the PL emission intensity was 4300 CU and the PL peak wavelength was 1.3 μm, and the target PL characteristics could be achieved.

  When epitaxial layers are simultaneously grown on two or more substrates, the reflection density on the back surface of the substrate and the size (major axis) of the elliptical etch pits are adjusted to a certain range, specifically 22 μm. The PL characteristics of the semiconductor elements obtained by epitaxial growth on these substrates were constant.

  As described above, in this embodiment, in the process of vapor-phase growth of an epitaxial layer made of a compound semiconductor such as an InGaAsP layer on a III-V group semiconductor substrate such as InP, the shape and size of the back surface etch pits of the substrate are set. The growth temperature is set in consideration of the resulting change in the substrate temperature (growth temperature). As a result, the actual substrate temperature can be kept constant at a desired temperature, and as a result, a semiconductor element having a desired PL characteristic can be stably manufactured.

  Further, since a substrate having a reflection density on the back surface of the substrate in a predetermined range is used, a semiconductor element having a PL emission intensity in a predetermined range can be realized.

  As mentioned above, although the invention made by this inventor was concretely demonstrated based on embodiment, this invention is not limited to the said embodiment, It can change in the range which does not deviate from the summary.

  For example, in this embodiment, an example in which an InGaAsP layer is grown on an InP substrate by metal organic vapor phase epitaxy has been described, but the heating conditions are set according to the size of the elliptical etch pits on the back surface of the substrate, A method of epitaxial growth with the substrate temperature being constant at the desired temperature can be expected to have the same effect regardless of the type of epitaxial layer to be grown. Also, the type of substrate used and the growth method are not limited, and it is apparent from the above description that the present invention can be applied to other substrates and growth methods.

Claims (5)

An epitaxial growth substrate made of a III-V compound semiconductor,
The maximum value of reflection density measured in the [011] direction is 0.5 to 2.0, and the maximum value of reflection density measured in the [0-11] direction is in the range of 0.5 to 1.5.
And,
A substrate for epitaxial growth, wherein a standard deviation of measured values in each measuring direction is within 10% of an average value. The major axis of the elliptical etch pit present on the back surface of the substrate is in the range of 5 μm to 40 μm,
And,
2. The epitaxial growth substrate according to claim 1, wherein the standard deviation of the major axis of each etch pit is within 20% of the average value.   The substrate back surface obtained by etching at a temperature of 15 to 100 ° C with an etching solution comprising 1 to 10 parts of phosphoric acid, 1 to 10 parts of hydrogen peroxide, and 0 to 10 parts of water. Epitaxial growth substrate.   An epitaxial growth method comprising growing an epitaxial layer by controlling epitaxial growth conditions in accordance with a reflection density on the back surface of a substrate and a major axis of an elliptical etch pit. When simultaneously performing epitaxial growth using two or more epitaxial growth substrates in the same furnace,
As the epitaxial growth substrate, the substrate according to any one of claims 1 to 3,
The average value of the reflection density on the back side of each substrate is within ± 10% of the average reflection density of all input substrates,
And,
5. The epitaxial growth method according to claim 4, wherein epitaxial growth is performed using a substrate in which an average value of the major axis of the elliptical etch pits existing on the back surface of the substrate is within ± 20% of an average major axis of all input substrates. .

Images