METHOD FOR EVALUATING DEFECT IN MONOCLINIC GALLIUM OXIDE

Abstract

Disclosed is a qualitative evaluation method of a volumetric defect density due to other grains having different crystal orientations from a single crystal matrix in a (001) monoclinic gallium oxide sample or a (010) monoclinic gallium oxide sample.

The method includes the steps of: forming an etch pit by etching an observation plane of a single crystal; and selecting a quadrilateral etch pit formed by volumetric defects except for void defects.

Claims

1. A qualitative evaluation method of a volumetric defect density due to other grains having different crystal orientations from a single crystal matrix in a (001) monoclinic gallium oxide sample or a (010) monoclinic gallium oxide sample, the method comprising the steps of: forming an etch pit by etching an observation plane of a single crystal; and selecting a quadrilateral etch pit formed by volumetric defects except for void defects.

2. The method according to claim 1, wherein additional defects are contained in the grains.

3. The method according to claim 1, wherein a monoclinic gallium oxide sample is a (001) sample, and the quadrilateral etch pit is formed in a rectangular rod shape.

4. The method according to claim 3, wherein the rectangular major axis direction is a (010) direction.

5. The method according to claim 1, wherein the monoclinic gallium oxide sample is a (010) sample; and the quadrilateral etch pit is formed in a parallelogram shape.

6. A qualitative evaluation method of a volumetric defect density due to other grains having different crystal orientations to a single crystal matrix in a (001) monoclinic gallium oxide sample or a (010) gallium oxide sample, the method comprising the steps of: forming an etch pit by etching an observation plane of a single crystal; selecting a quadrilateral etch pit formed by volumetric defects except for void defects; and evaluating a volumetric defect density from the shape of the etc pit due to the volumetric defects by the following equation, volumetric defect density=Σkaibi/area of sample, wherein 0<k≤1, ai is the width of the etch pit by the i-th volumetric defect; bi is the height of the etching pit by i-th volumetric defect.

7. The method according to claim 6, wherein additional defects are contained in the crystal grains.

8. The method according to claim 6, wherein a monoclinic gallium oxide sample is a (001) sample, then k is 0<k≤½.

9. The method according to claim 8, wherein a thickness of the (001) monoclinic gallium oxide sample is equal to or greater than 1.5 μm, then k=½.

10. The method according to claim 6, wherein a monoclinic gallium oxide sample is a (010) sample, then k=1.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0022] For more complete understanding of the present invention and its advantages, reference is now made to the following description taken in conjunction with the accompanying drawings, in which like reference numerals represent like parts:

[0023] FIG. 1 is an image showing a Reflection High Energy Electron Diffraction (RHEED) pattern of a gallium oxide epi-layer according to Preparation Example 1.

[0024] FIG. 2 is an image showing an RHEED pattern of a gallium oxide epi-layer according to Preparation Example 2.

[0025] FIG. 3 is a SEM image showing an etch pit of a (001) gallium oxide epi-layer.

[0026] FIG. 4 is a TEM image showing a cross-section of an etch pit formed on a (001) gallium oxide epi-layer.

[0027] FIG. 5 is a high magnification cross-sectional TEM image for analyzing origin of an etch pit formed on a (001) gallium oxide epi-layer.

[0028] FIG. 6 is a high resolution TEM image and TEM diffraction patterns of a matrix and defect regions of a (001) gallium oxide epi-layer.

[0029] FIG. 7 is a schematic diagram showing an etch pit derived from a volumetric defect in a (001) gallium oxide epi-layer.

[0030] FIG. 8 is a SEM image showing an etch pit formed on a (010) gallium oxide epi-layer.

[0031] FIG. 9 is a cross-sectional TEM image of an etch pit formed on a (010) gallium oxide epi-layer.

[0032] FIG. 10 is a high magnification cross-sectional TEM image for analyzing origin of an etch pit formed on a (010) gallium oxide epi-layer.

[0033] FIG. 11 is a high resolution TEM image and TEM diffraction patterns of a matrix and defect regions of a (010) gallium oxide epi-layer.

[0034] FIG. 12 is a schematic diagram showing an etch pit derived from a volumetric defect in a (010) gallium oxide epi-layer.

DETAILED DESCRIPTION

[0035] To fully understand the present invention, the operational advantages of the present invention, and the objects achieved by the embodiments of the present invention, reference should be made to the accompanying drawings and the contents described in the accompanying drawings, which illustrate preferred embodiments of the present invention.

EXAMPLES

Example 1: Homoepitaxial Growth of Gallium Oxide

[0036] For the epi-layer growth of gallium oxide, single crystal substrates include (010), (100), (001), and (−201) substrates. Among these substrates, since the (100) substrate has a very slow growth rate of epi-layer, the practicality as a substrate for the single crystal growth is low, and currently, a (−201) substrate is commercially available as a 2-inch substrate, and the growth rate of epi-layer is also pertinent, but there is a problem that the frequency of twin defects may be high. The gallium oxide epi-layer was then grown using a (001) substrate which commercialized to up to 4 inches which expected to be most widely used in the future for gallium oxide epitaxial growth for device applications, and the (010) substrate which is currently the most commonly used for device fabrication demonstrations, although the size of commercial substrates is limited to 10 mm×15 mm segmented substrates, and defects in each substrate were analyzed.

Preparation Example 1: (001) Homoepitaxial Growth on Gallium Oxide Substrate

[0037] The commercial gallium oxide (001) substrate was cleaned by ultrasonic stirring for 10 minutes sequentially using acetone, methanol, and deionized water (DI water). The cleaned substrate was chemically cleaned with a 1:1:4 mixtures in a volume ratio of Di water, 30% hydrogen peroxide, and 96% sulfuric acid for 5 minutes and then further cleaned with DI water and dried using nitrogen gas. The dried substrate was introduced into a Plasma Assisted Molecular Beam Epitaxy (PAMBE) growth chamber and thermally cleaned at 850° C. for 30 minutes using oxygen radicals generated with plasma excited by an RF power of 300 W and an oxygen gas flow rate of 2 sccm to prepare a substrate.

[0038] A plasma-assisted molecular beam epitaxy growth chamber equipped with a substrate, elemental gallium (Ga) having a purity of 7 N was supplied through a Sumo Effusion cell, (Veeco, USA), and oxygen was flowed through a plasma cell (SVAT, USA) to generate oxygen radicals by plasma, whereby a gallium oxide thin film was grown by the method of PAMBE. The growth chamber pressure before gas introduction was about 2×10.sup.−9 Torr, the pressure during growth was 2×10.sup.−5 Torr to 4×10.sup.−5 Torr, and the conditions for thin film growth are as shown in Table 1.

TABLE-US-00001 TABLE 1 Epitaxial growth temperature (°C.) 850 Oxygen plasma (W-sccm) 300-2 Ga cell temperature (°C.) 600 Epitaxial growth time (hr) 4

Preparation Example 2: (010) Homoepitaxial Growth on Gallium Oxide Substrate

[0039] A monoclinic gallium oxide was homoepitaxially grown on a (010) gallium oxide substrate using the same method of Preparation Example 1, except that the (010) substrate was used and the epi-growth temperature was set to 800° C.

Example 2: Confirming Growth of Gallium Oxide Epi-Layer

[0040] Before evaluating crystal defects, the crystal structure of the thin film grown in Example 1 was analyzed by Reflection High Energy Electron Diffraction (RHEED) and confirmed that the gallium oxide epi-layer was grown first. RHEED was analyzed under conditions of an electron energy voltage of 18 kV and a beam current of 39 μA.

[0041] FIG. 1 and FIG. 2 are images showing RHEED patterns of gallium oxide layer grown on a (001) substrate and a (010) substrate by Preparation Example 1 and 2, respectively. The gallium oxide layer grown in FIG. 1 and FIG. 2 exhibits the same RHEED pattern as of the substrate, respectively, and confirmed that a monoclinic gallium oxide epi-layer has grown on the substrate.

Example 3: Etch Pit Analysis

[0042] The gallium oxide epi-layer prepared in Example 1 was etched at 140° C. with 85% phosphoric acid for 2 minutes, and the etched surface was observed with an optical microscope and SEM. Optical microscope observation was carried out applying 1000× using the Nikon Eclipse LV100ND, and SEM observation used the a focused ion beam (FIB) from the FEI Helios NanoLab system. Hereinafter, an etch pit of gallium oxide epi-layer produced in Preparation Examples 1 and 2 is analyzed.

Preparation Example 1

[0043] FIG. 3 is a planar SEM image showing an etch pit of a (001) gallium oxide epi-layer grown on the (001) substrate of Preparation Example 1. The most primary etch pit in FIG. 3 is a rod-shaped etch pit having a major axis in the [010] direction, and the rod-shaped minor axis length (a) was approximately 200 nm and the major axis direction length (b) was approximately several hundred nanometers to several micrometers. Hereinafter, the length in the minor axis direction referred to as width, and the length in the major axis direction is referred to as height.

[0044] A TEM image of a cross-section including the etch pit was observed to identify the correlation between an etch pit and a defect. FIG. 4 is a low magnification bright-field TEM image of a TEM cross-sectional specimen fabricated by a FIB along the red dotted line of FIG. 3, and the numerals on the image correspond to the numerals indicated for the etch pit of FIG. 3. The presence of defect under each etch pit was confirmed.

[0045] FIG. 5 is a high magnification bright-field TEM image of the third etch pits, which is identified that a defect is present under the etch pit in FIG. 5(a), and that the etch pit width reflects the width of the defect. In addition, the defect starts on tens of nanometers above the interface between the substrate and the epi-layer and propagate to the surface, and the width of the defects gradually increases as the substrate grows. FIG. 5(b) is a further magnified image of FIG. 5(a) showing that an incoherent interface is formed at the interface between the epi-layer and the defect and that there is an additional incoherent interface within the defect. Further, it can be identified from the TEM image FIG. 5(c) obtained by further magnifying the image that a plurality of stacking faults are produced inside the crystal defect due to the incoherent interface.

[0046] In order to clarify the cause of the defect, the defect region is observed with a high-resolution TEM, and the result is shown in FIG. 6. FIG. 6(b) and FIG. 6(c) are images showing a TEM diffraction pattern of a defect region and a matrix region represented by yellow and green quadrangles in the TEM image of (a), respectively, and the difference in the crystal orientations of the defect region and the matrix region can be seen clearly. The crystal orientations of the defect region and the matrix region were illustrated in FIG. 6(a), from which it may be seen that the defect regions are rotated with respect to the normal matrix region.

[0047] Overall, a rod-shaped etch pit, which is the primary etch pit is mainly caused by growth in which the crystal orientation of the matrix region is rotated. Rotation of the crystal orientation results in an incoherent interface in the matrix region and the defect region, which causes additional defects, such as rotational orientation or stacking faults in the defect region. This is in contrast to a twin defect, in which the crystal orientation is symmetrical, and the twin defect merely includes additional defect, and only the twin boundary is regarded as a defect and classified as a plane defect called a twin boundary defect (usually referred to as twin defect). The defects identified in the present embodiment are not the twin defect that classified into the planar defect, does not have a symmetric plane such as twinning, have a specific asymmetric rotational direction relationship, and include a plurality of additional defect, so that when compared with twin defects, and it is predicted that the effect of degrades in a device is expected to be significant. Therefore, it is reasonable to consider defect density in terms of volumetric defect rather than considering such defect as a planar defect. There is no such complex volumetric defect in a gallium oxide epi-layer has been reported.

[0048] Further, the defect density of the observed volumetric defect was calculated from the etch pit. The calculation of defect density using a conventional etch pit was limited to dislocations, and the inventors have registered a method for calculating defect density from the etch pit for a twin defect which is a planar defect, as in Korean Patent Registration No. 10-2012809. However, the etch pit has not been reported in the volumetric defect, and naturally, calculation of defect density has not been attempted.

[0049] FIG. 7 is a volumetric defect and a schematic diagram for calculating the defect density of identified defects in the present embodiment. After identifying a TEM cross-section of various samples of a rod-shaped etch pit, a defect started in the middle of the epi-layer showed a triangular prism shape whose width gradually increases as the epi-layer grows. Accordingly, the volume for one of the volumetric defects may be determined from the following equation: ½×abc, wherein a is the width of the etch pit, b is the height of the etching pit and c is the depth of the volumetric defect. Since the depth of volumetric defect cannot be greater than the thickness of the epi-layer and expressed as 0<c≤z, where z is the thickness of the epi-layer. Therefore, the volume of the volumetric defects may be determined from the following equation: kabz, where 0<k≤½.

[0050] Typically, the defect of the epitaxial layer grown by homoepitaxy is caused by a defect of a substrate, so that the defect is started near the interface between the substrate and the epitaxial layer. Therefore, when the thickness of the epitaxial layer is large, the thickness is converged from c to z, and thus k converges to ½.

[0051] Thus, the total volume V, where V=xyz, the density of volumetric defect may be determined from the following equation:

Volumetric defect density=Δka.sub.ib.sub.ic/xyz=Σka.sub.ib.sub.iz/xyz=Σka.sub.ib.sub.i/xy=Σka.sub.ib.sub.i/area of sample

[0052] wherein a.sub.i is the width of the etch pit by the i-th volumetric defect; b.sub.i is the height of the etching pit by i-th volumetric defect, where 0<k≤½.

[0053] As the thickness of the epitaxially grown single crystal sample increases, k converges to ½, so that the volumetric defect density also converges to (½×Σa.sub.ib.sub.i/area of sample).

Preparation Example 2

[0054] FIG. 8 is a planar SEM image showing an etch pit of a gallium oxide epi-layer grown on the (010) substrate of Preparation Example 2. The primary etch pit of the epi-layer grown on the substrate (010) was in a parallelogram shape as can be identified in FIG. 8, different from the result that the most etch pits observed in the epi-layer grown on a substrate (001) was in a rod shape. The width (a) in the parallelogram-shaped etch pits and the length (b) in the direction perpendicular to the width (i.e., the height of the parallelogram) were in the range of 100 to 200 nm, respectively.

[0055] A TEM image of a cross-section including the etch pit was observed to identify the correlation between an etch pit and a defect. FIG. 9(a) is an SEM image of a surface including an etch pit on which a FIB TEM specimen was prepared, and (b) is a low magnification bright-field SEM image of the FIB cross-sectional TEM specimen along a yellow dotted line. In FIG. 9(b), crystal defects are present under the etch pit, and the crystal defects start at the interface of the substrate and the epi-layer and propagate to the surface of the epi-layer to a size almost identical to that at the interface.

[0056] FIG. 10 is a high magnification bright-field TEM image of a cross-section specimen for the etch pit, showing that FIG. 10(a) defects are present under the etch pit, and the etch pit width reflects the width of the defect as it is. It is also shown that the defects begin at the interface of the substrate and the epi-layer and propagate to the epi-layer, and there is no significant difference in the width of the defects as the epi-layer is grown, and the width of the defects remains almost constant as that of at the interface. FIG. 10(b) is a further magnified image of FIG. 10(a), and may identify the crystal orientation of the epi-layer and the defect is different from each other and a plurality of stacking faults are included in the crystal.

[0057] In order to clarify the cause of the defect, the defect region is observed with a high-resolution TEM and the results are shown in FIG. 11. FIG. 11(b) and FIG. 11(c) are images showing a TEM diffraction pattern of a defect region and a matrix region represented by yellow and green quadrangles in the TEM image of (a), respectively, and the difference in the crystal orientations of the defect region and the matrix region may be seen clearly. FIG. 11(a) illustrated the crystal plane orientations of the defect region and the matrix region, from which it may be seen that the defect regions have completely different crystal orientations with respect to the normal matrix region and are rotated in a specific direction. As like in Preparation Example 1, the rotation of the crystal orientations results in an incoherent interface in the matrix region and the defect region, which causes additional defects, such as rotational orientation or stacking faults inside the defect region. Thus, in FIG. 11(a) not only the incoherent interface due to the rotation in the crystal direction but also the stacking faults in the crystal defect are observed.

[0058] Overall, a parallelogram-shaped etch pit, which is the primary etch pit is mainly caused by growth in which the crystal orientations of the matrix region are rotated. In addition, a significant number of stacking faults were observed in the crystal defects, which was higher in frequency than the defects observed in the volumetric defect of the epi-layer grown in Preparation Example 1.

[0059] Accordingly, the primary defect observed in the gallium oxide (010) epi-layer is also predicted that the entire defect region may attribute degradation of a device severely as observed in a (001) epi-layer. Thus, the defect may also be treated with a volumetric defect, and the defect density is to be considered in terms of the volumetric defects.

[0060] Since the volumetric defects observed from the (010) epi-layer are propagated from the interface between a substrate and the epi-layer, the defect density may be calculated more simply than the volumetric defects observed from the (001) epi-layer. FIG. 12 is a schematic diagram to calculate the defect density of the volumetric defect, which is a primary defect observed in the (010) epi-layer. The volumetric defects in the (010) epi-layer can be typified in a form of cuboids with a parallelogram in cross-section. Accordingly, the volume for one of the volumetric defects can be determined from the following equation: abc, wherein a is the width of etch pit, b is the length of etch pit, and c is the depth of the volumetric defect. The length of etch pit is the length in the directions perpendicular to the width of the etch pit, which corresponds to the height of a parallelogram. The volumetric defect propagates from the interface between the substrate and the epi-layer, therefore c is the same as the thickness of the epi-layer.

[0061] Therefore, the total volume V, where V=xyz, the density of volumetric defect may be determined from the following equation:

Volumetric defect density=Σa.sub.ib.sub.iz/xyz=Σa.sub.ib.sub.i/xy=Σa.sub.ib.sub.i/area of sample,

[0062] wherein a.sub.i is the width of the etch pit by the i-th volumetric defect; b.sub.i is the height of the etching pit by i-th volumetric defect.

[0063] As described above, according to a method for evaluating a monoclinic gallium oxide sample of the present invention, a volumetric defect density which is a primary defect that is observed from the optical microscope after simply etching the sample may be evaluated qualitatively/quantitatively, and used for the analysis of a monoclinic or monoclinic thin film sample.

[0064] Accordingly, the method for evaluating the volumetric defects, for example, a crystal for manufacturing the semiconductor device or for setting an optimized condition that may reduce the volumetric defects, may be used efficiently.