FURNACE FOR SEEDED SUBLIMATION OF WIDE BAND GAP CRYSTALS

Abstract

An apparatus for physical vapor transport growth of semiconductor crystals having a cylindrical vacuum enclosure defining an axis of symmetry; a reaction-cell support for supporting a reaction cell inside the vacuum enclosure; a cylindrical reaction cell made of material that is transparent to RF energy and having a height Hcell defined along the axis of symmetry; an RF coil provided around exterior of the vacuum enclosure and axially centered about the axis of symmetry, wherein the RF coil is configured to generate a uniform RF field along at least the height Hcell; and, an insulation configured for generating thermal gradient inside the reaction cell along the axis of symmetry. The ratio of height of the RF induction coil, measured along the axis of symmetry, to the height Hcell may range from 2.5 to 4.0 or from 2.8 to 4.0.

Claims

1. An induction furnace apparatus for growing semiconductor crystals by seeded sublimation growth, comprising: a quartz vacuum chamber; a cylindrical RF induction coil positioned coaxially with the quartz vacuum chamber; an RF power supply coupled to the RF induction coil; a reaction cell configured for containing a seed crystal and source material, the reaction cell defining an axial length measured as the reaction cell height along its axis of rotational symmetry; an arrangement of insulation layers around the cell configured for generating a thermal gradient inside the reaction cell; a support for placing the reaction cell inside the quartz vacuum chamber; wherein the RF induction coil is configured for generating a uniform electromagnetic field around the reaction cell when the reaction cell is positioned co-axially with the induction coil, coaxially to the quartz vacuum chamber, and near or at the center of the coil with respect to its axial length; and, wherein a ratio of height of the RE induction coil, measured along the axis of rotational symmetry, to the axial length of the reaction cell is from 2.5 to 4.0.

2. (canceled)

3. The apparatus of claim 1, wherein a ratio of height of the RF induction coil, measured along the axis of rotational symmetry, to the axial length of the reaction cell is from 2.8 to 4.0

4. The apparatus of claim 1, wherein the reaction cell and the insulation are made of graphite.

5. The apparatus of claim 1, further comprising a mechanism for forced air flow around the exterior wall of the quartz vacuum chamber.

6. The apparatus of claim 1, wherein a height of the RF induction coil, measured along the axis of rotational symmetry, is smaller than height of the quartz vacuum chamber, measured along the axis of rotational symmetry, but longer than the axial length of the reaction cell.

7. The apparatus of claim 6, wherein the ratio of the height of the quartz vacuum chamber to the reaction cell height ranges from 5.9 to 6.5.

8. The apparatus of claim 6, wherein the ratio of the height of the quartz vacuum chamber to the reaction cell height ranges from 7.1 to 7.8.

9. The apparatus of claim 1, wherein the RF induction coil has an internal diameter of from 330 to 725 mm.

10. The apparatus of claim 1, further comprising a water jacket positioned on exterior wall of the quartz vacuum chamber.

11. The apparatus of claim 1, wherein the support is configured so that it does not absorb energy from the RF fields and for supporting a reaction cell inside the vacuum enclosure, such that the reaction cell is positioned axially centrally to the axis of symmetry.

12. The apparatus of claim 11, wherein a ratio of height of the RF induction coil, measured along the axis of symmetry, to the height Hcell ranges from 2.8 to 4.0.

13. The apparatus of claim 12, wherein the RF coil has internal diameter of from 330 to 550 mm.

14. A method for physical vapor growth of semiconductor crystal, comprising: placing a seed and source material inside a reaction cell having a height Hcell, defined as the height along the reaction cell's axis of symmetry; providing an arrangement of insulation layers around the cell configured to generate thermal gradient inside the reaction cell; placing the reaction cell inside a quartz vacuum chamber; generating vacuum inside the quartz vacuum chamber; and, generating an RF field inside the vacuum chamber, wherein the RF electromagnetic field is consistent for a region at least as long as the height Hcell when the reaction cell is placed near the center of the coil, by making the RF coil having a height, defined along the axis of symmetry, that is from 2.6 to 4.0 times longer than the height Hcell.

15. (canceled)

Description

BRIEF DESCRIPTION OF THE DRAWINGS

[0026] The accompanying drawings, which are incorporated in and constitute a part of this specification, exemplify the embodiments of the present invention and, together with the description, serve to explain and illustrate principles of the invention. The drawings are intended to illustrate major features of the exemplary embodiments in a diagrammatic manner. The drawings are not intended to depict every feature of actual embodiments nor relative dimensions of the depicted elements, and are not drawn to scale.

[0027] FIG. 1 is a schematic illustrating a furnace construction according to one embodiment, while FIG. 1A illustrates an embodiment utilizing convection cooling.

[0028] FIG. 2 is a plot illustrating distribution of the standard deviation of temperature stability during the experiment of example 1.

[0029] FIG. 3 is a plot of comparison of sliced wafer bow for example 1.

[0030] FIG. 4 is a plot of comparison of sliced wafer warp for example 1.

[0031] FIG. 5 is a plot that compares the distribution of micropipe density associated with each coil type used in example 2.

[0032] FIG. 6 is a plot of distribution of etch pit density for 76-100 mm n+ 4H-SiC polished wafers of example 3.

DETAILED DESCRIPTION

[0033] Various disclosed embodiments relate to the design of an inductive PVT reaction furnace. For proper growth of SiC, thermal gradient needs to be established, wherein the area occupied by the source material is hotter than the area occupied by the SiC seed. According to embodiments disclosed below, the thermal gradient is achieved by designing an induction coil that generates a rather uniform heating field within the area occupied by the reaction cell. However, the reaction cell and the insulation are designed to cause a non-uniform heat loss from the reaction cell.

[0034] According to some specific examples, the range of minimum ratio value of coil height to reaction cell height defined as

Ratio=(axial length of coil)/(axial length of reaction cell)

is 1.8 to 4.0, 2.0 to 5.0, 2.0-4.0, and preferred value range of nominally 2.8-4.0. Larger ratios will not impact crystal growth but will make the construction of large furnaces more expensive and difficult to assemble and maintain. When the application is required for a furnace design to accommodate growth of larger crystals, it is discovered that the critical length ratio (defined above) is invariant to increased diameter.

[0035] FIG. 1 illustrates one embodiment of a growth chamber utilizing features of the invention. In the example of FIG. 1, an enclosure 110 is supported by cabinet 105, which forms the base of the apparatus. A support, in the form of tubular pedestal 115 supports the reaction cell 120, and is movable in the vertical direction. In FIG. 1 the tubular pedestal 115 is illustrated in its retracted position. However, in its extended position the tubular pedestal 115 places the reaction cell 120 within the processing chamber's inner tube 125, wherein the bottom base plate 135 contacts and seals the chamber base 130, so that the interior of the inner chamber tube 125 can be maintained in vacuum. The chamber base 130 and the base plate 135 may be water cooled. Also, an outer sleeve 127 may be provided over inner tube 125 to provide water cooling, and a chamber cap 113 seals the top of the inner tube 125 and may also be water cooled. The base plate 135 may be attached to a seal adapter 137, which may comprise a Ferrofluidic seal. Seal adapter 137 may also incorporate a rotation motor to rotate the reaction cell during the growth process, and fiberoptic pyrometer head for temperature measurement. Neither the rotation motor nor the fiberoptic pyrometer head are specifically illustrated in FIG. 1. A top pyrometer 117 measures the temperature at the top of the insulated reaction cell. In this example, the pyrometer 117 is attached to a pyrometer X-Y translation stage to enable measurements at different locations.

[0036] The cylindrical RF coil 140 is around and coaxially aligned to the inner tube 125. The coil 140 is designed so as to impart uniform electromagnetic field within the inner tube 125. In particular, the coil will create an electromagnetic field such that perturbations of the field by the presence of the reaction cell are very small as along as the reaction cell is positioned near the center of the coil. In the example of FIG. 1, the ratio of the axial length of the coil (marked as ALcoil) to the height, i.e., the axial length of the reaction cell (marked as ALcell) set to 2.0-4.0. In this manner, the effect of the induction coil design is such that the design of the reaction cell (geometry, wall thickness, insulation 123, etc.) will primarily determine the degree and uniformity of the temperature in the sublimation zone. In order to minimize effects of the vacuum chamber on the RF fields in the cell, the flanges of the chamber should be made of aluminum or austenitic grades of steel. The ratio of the height of the chamber, including the endcaps, to the height of the coil should be 1.7 to 2.0 times the ALcoil or larger.

[0037] The coil design, according to disclosed embodiments, generates a uniform field along the axial length of the inner tube for a length that is at least equal the height of the reaction crucible. The reaction crucible is then placed within this uniform field for the duration of the crystal growth. In some embodiments the uniform field is generated over a length that is longer than the height of the reaction cell, so as to provide safety margin.

[0038] In some embodiments, a mechanism for forced air flow around the exterior wall of the quartz vacuum chamber is provided. In the embodiment of FIG. 1, the forced air mechanism comprises air pump 111 coupled to air conduits 112, and delivering air from bottom to top. The forced air flow exits the enclosure 110 at the top through exhaust connection 144. Conversely, in FIG. 1A the air cooling is achieved by fabricating air intakes 142 at the bottom of the chamber cover, to enable air to enter from the bottom, as exemplified by the arrows. In this example, the air intakes are covered with louvers to assist in laminar flow. An exhaust connection 144 can be led to a blower assisted exhausted ventilation system, to assist in disposing of the hot air.

[0039] The internal diameter of the reaction cell is designed to accommodate crystal growth with diameters of from 76 to 200 mm. Similarly, the internal diameter of the RF induction coil is designed to accommodate this growth and may be set to, for example, from 330 to 550 mm or from 330 to 725 mm. In order to maintain the RF field uniformity within the reaction cell, the support of the reaction cell is made of material to which the RF fields do not effectively couple.

EXAMPLES

Example 1

[0040] 76 mm diameter 4H-SiC Crystal Growth using a traditional coil and new induction coil.

[0041] Two identical vacuum furnaces were each fit with a different induction coil. Multiple reaction cells with insulation were prepared identically for SiC sublimation crystal growth using 76 mm 4H-SiC seeds. In the “control” coil furnace, the ratio of the coil axial length to reaction cell axial length was 2.0, while in the “new” coil furnace the ratio of the coil axial length to reaction cell axil length was 3.6. The same process to grow N-doped 4H-SiC crystals was executed in each furnace. The reaction cells were centered within the axial length of the coil and supported by material to which the RF fields do not effectively couple.

[0042] Temperature control stability was tracked during each growth for the steady state temperature value held during the growth stage. The distribution of the standard deviation of this temperature value is plotted in FIG. 2 for all the growth tests in each furnace. FIG. 2 shows that the run-to-run temperature stability was improved with the “new” coil (Ratio=3.6). Temperature is in units of Celsius.

[0043] Wafers were sliced using a multiwire slicing system from all the crystals produced using an identical process. FIG. 3 shows the distribution of slice wafer bow and FIG. 4 shows the distribution of slice wafer warp as compared for each furnace. The units of bow and warp are microns. FIGS. 3 and 4 indicate that the crystals from the new furnace exhibit a much tighter distribution of bow and lower value of warp. Since the slicing process imparted the same surface damage to each wafer, the changes in the bow/warp distributions can be attributed to a lower inherent stress in the crystals produced using the new furnace (ratio=3.6).

Example 2

[0044] 76 mm diameter 4H-SiC Crystal Growth using a control coil and “new” induction coil.

[0045] Two identical vacuum furnaces were each fit with a different induction coil. Multiple reaction cells with insulation were prepared identically for SiC sublimation crystal growth using 76 mm 4H-SiC seeds. In the “control” coil furnace, the ratio of the coil axil length to reaction cell axial length was 2.0, while in the “new” coil furnace the ratio of the coil axial length to reaction cell axial length was 3.6. The same process to grow N-doped 4H-SiC crystals was executed in each furnace. The reaction cells were centered within the axil length of the coil and supported by material to which the RF fields do not effectively couple.

[0046] The produced crystals were cut into slices and the slices were fully processed into polished wafers. Each wafer was examined using a KLA-Tencor CS2 laser light scanning spectrometry system which is capable to detect micropipes in the polished wafers (J. Wan, S.-H. Park, G. Chung, and M. J. Loboda, “A Comparative Study of Micropipe Decoration and Counting in Conductive and Semi-Insulating Silicon Carbide Wafers,” J. Electronic Materials, Vol. 34 (10), p. 1342 (2005)). The measurement determines the total count of micropipes on the wafer and divides that value by the total measurement area. In these measurements the entire wafer was measured with the exception of 2 mm edge exclusion. FIG. 5 compares the distribution of micropipe density associated with each coil type. In FIG. 5 Coil type “0” is the “control” coil and coil type “1” is the “new” coil. Units of the micropipe density are defects/cm2. Inspection of FIG. 5 reveals that the crystals produced with the “new” coil (ratio=3.6) consistently show lower micropipes and have tighter distribution.

Example 3

[0047] 76-100 mm diameter 4H-SiC Crystal Growth using a traditional coil and “new” induction coil.

[0048] Sets of identical vacuum furnaces were each fit with one of 2 different induction coils. Multiple reaction cells with insulation were prepared identically for SiC sublimation crystal growth using 76 mm 4H-SiC seeds. Another set of multiple reaction cells with insulation were prepared identically for SiC sublimation crystal growth using 100 mm 4H-SiC seeds. In the “control” coil furnace, for 76 mm growth the ratio of the coil length to reaction cell length was 2.0, while in the “new” coil furnace the coil length to reaction cell length was 3.6. In the “control” coil furnace, for 100 mm growth the ratio of the coil length to reaction cell length was 1.6, while in the “new” coil furnace the coil length to reaction cell length was 2.9. The reaction cells were centered within the length of the coil and supported by material to which the RF fields do not effectively couple. All the crystals grown were doped with nitrogen corresponding to resistivity of 0.016-0.028 ohm-cm range.

[0049] Slices were cut from each crystal and etched in molten KOH to reveal dislocation etch pits corresponding to basal, edge, and screw dislocations. The total number of dislocation etch pits were counted at 9 locations on each wafer and then the total count was divided by the measurement area. FIG. 6 shows the trend in the value of etches pit density comparing the traditional and new RF coils.

[0050] FIG. 6 shows that as the coil to reaction cell length ratio was increased, there was a correlated drop in the etch pit density in the 4H-SiC crystals, both at 76 mm and 100 mm. Dislocation reduction is typically associated with reduced stress in the crystal during growth. Reduction of parasitic temperature gradients resulting from non-uniform RF fields will lead to reduction of stress during crystal growth.

Example 4

[0051] 150 mm diameter 4H-SiC Crystal Growth

[0052] An induction furnace was constructed to support growth of crystals up to 200 mm diameter. An insulated reaction cell was constructed with a 4H-SiC seed wafer and the design corresponds to a coil to reaction cell length ratio of approximately 3.5. A 155 mm diameter 4H-SiC crystal (6B13470010) doped with nitrogen was grown. The crystal was sliced into wafers and examined by x-ray topography. The screw dislocation and basal plane dislocation count was evaluated at 9 sites on a sliced wafer. The dislocation density was determined at each site by dividing the dislocation count by the measurement area. The basal plane dislocation density ranged 3.1-6.2×10.sup.3/cm.sup.2 and the screw dislocation density ranged from 0.25-3.75×10.sup.2/cm.sup.2.

[0053] It should be understood that processes and techniques described herein are not inherently related to any particular apparatus and may be implemented by any suitable combination of components. Further, various types of general purpose devices may be used in accordance with the teachings described herein. It may also prove advantageous to construct specialized apparatus to perform the method steps described herein.

[0054] The present invention has been described in relation to particular examples, which are intended in all respects to be illustrative rather than restrictive. Those skilled in the art will appreciate that many different combinations of hardware, software, and firmware will be suitable for practicing the present invention. Moreover, other implementations of the invention will be apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the invention being indicated by the following claims.