Method of Making Vertical Bipolar Transistor Having Base Above Buried Nitride Dielectric Formed by Deep Implantation

The present invention relates to the field of integrated circuit fabrication. Specifically, the present invention relates to a structure and method for fabricating bipolar transistors.

One of the properties which inhibits the high frequency and high speed digital operation of bipolar transistors is the capacitive coupling between the base and the emitter and the base and the collector of the bipolar transistor. The capacitive coupling occurs across the depletion regions of the respective junctions. This phenomenon is well known and is explained in Sze, PHYSICS OF SEMICONDUCTOR DEVICES, pp. 79-81(1981). Because the base to emitter junction controls the current flow of the transistor, most transistors are fabricated so that the base to emitter junction has a minimum of interface area. The interface area is that region of the base and emitter which are in contact with each other. Using a simple parallel plate capacitor model, the equation for the capacitance of a junction is

where,

C is the capacitance of the junction,

k is the dielectric constant of the material between the "plates" of the capacitor,

60 is the permitivity of free space,

A is the area of the "plates", and

d is the width of the junction depletion region.

Thus the capacitance of a junction is directly proportional to the area of the junction and is inversely proportional to the width of the junction depletion region. Thus there are three ways of decreasing the capacitance of a junction: decreasing the dielectric constant of the junction material or a portion of the the junction, decreasing the area of the junction and increasing the thickness of the junction. Because the active junction must consist of the semiconductor material which forms the bipolar transistor, it is usually impractical to change the dielectric constant of the overall junction (active and parasitic) in a bipolar transistor. Therefore, in order to reduce the capacitance of a junction the active junction area must be decreased, the effective junction thickness must be increased and/or the dielectric constant of the parasitic junction regions must be decreased.

FIG. 1 is a side view of a prior art vertical bipolar transistor. Buried collector 3 is formed in substrate 1. An N-type epitaxial layer is formed on substrate 1 and isolation oxide regions 2 are formed in this epitaxial layer. Base region 5 is formed on epitaxial layer 4 and emitter region 6 is formed in base region 5. Contact diffusion 9 allows base contact 8 to make ohmic contact. Emitter contact 7 contacts emitter region 6 directly. In this structure, and in most vertical bipolar transistors, the base to emitter interface area is much less than the base-collector interface area. Therefore by reducing the base to collector capacitance, the overall parasitic capacitance of a vertical bipolar transistor may be minimized. Therefore, it is an object of the present invention to minimize the base to collector capacitance of a vertical bipolar transistor.

In one embodiment of a vertical bipolar transistor constructed in accordance with the teachings of this invention, oxygen is implanted into the vertical bipolar transistor to provide a silicon dioxide layer between the parasitic extrinsic base region and collector of the vertical bipolar transistor. This silicon dioxide layer reduces the actual interface area of the base to collector junction, thereby decreasing the capacitance of the base collector junction. In addition, the thickness and dielectric constant of the silicon dioxide layer is such that the capacitance across the silicon dioxide layer, and thus between the base and collector, is minimal relative to the base to collector capacitance provided by the base to collector junction itself, because the dielectric constant of silicon dioxide is approximately 3.9 (which is much less than the dielectric constant of crystalline silicon which is approximately 11.7). In another embodiment, nitrogen ions are implanted to form silicon nitride regions rather than silicon dioxide regions.

FIG. 1 is a schematic side view showing the construction of a prior art vertical bipolar transistor;

FIGS. 2A-2H are schematic side view drawings depicting the processing steps necessary to construct one embodiment of the present invention;

FIGS. 3A-3D are schematic side view drawings depicting the processing steps necessary to fabricate another embodiment of the present invention.

FIGS. 2A-2H are schematic side view drawings showing the processing steps necessary to produce a vertical bipolar transistor with reduced base to collector capacitance in accordance with one embodiment of the present invention. Diffusion region 3 of FIG. 2A is an N+ diffusion formed in P-type substrate 1 using techniques well known in the art. Silicon epitaxial layer 10 of FIG. 2B is formed on the surface of substrate 1 to a thickness of approximately 1 micron using techniques well known in the art. Silicon dioxide layer 11 and silicon nitride layer 12 are formed and patterned on the surface of epitaxial layer 10 using techniques well known in the art. Epitaxial layer 10 is then anisotropically etched to provide the structure shown in FIG. 2C. Silicon dioxide layers 2 of FIG. 2D are then thermally grown in an steam ambient at a temperature of approximately 975 degrees C. for approximately 8 hours. Silicon nitride layer 12 and silicon dioxide layer 11 are then removed using techniques well known in the art. Silicon dioxide layer 14 is then thermally grown in an oxygen ambient at a temperature of 950 degrees C. for approximately 30 minutes to a thickness of approximately 300 angstroms. Polycrystalline silicon layer 14A is then formed by chemical vapor deposition to a thickness of approximately 2,000 angstroms. Silicon nitride layer 13 is then formed to a thickness of approximately 2,000 angstroms using chemical vapor deposition. Silicon dioxide layer 13A is then deposited by chemical vapor deposition or plasma deposition to a thickness of 6,000 to 8,000 angstroms. Silicon dioxide layer 14, silicon dioxide layer 13A, polycrystalline silicon layer 14a and silicon nitride layer 13 are then patterned and anisotropically etched using techniques well known in the art. Epitaxial layer 10 is then subjected to an ion implantation of oxygen ions(O2) at an energy of approximately 300 kiloelectron volts and a density of 6E17 ions per centimeter squared. After this oxygen implant, silicon dioxide layer 13A is removed by wet etching in dilute 10% hydrofluoric acid which selectively etches undensified silicon dioxide layer 13A without significantly undercutting silicon dioxide layer 14. The oxygen implant will have no effect on isolation regions 2 and, after annealing at approximately 1150 degrees C. for approximately two hours, will form silicon dioxide regions 15 of FIG. 2F which are approximately 4000 angstroms below the surface of epitaxial layer 10.

In a preferred embodiment, epitaxial layer 10 is implanted with nitrogen ions(N2) rather than oxygen ions. These nitrogen ions are implanted at an energy of approximately 150 to 200 kiloelectron-volts and a density of approximately 5E17 to 1.5E18 ions per centimeter squared to form silicon nitride regions approximately 4000 angstroms below the surface of silicon epitaxial layer 10. When annealed, these implanted ions will form silicon nitride regions instead of silicon dioxide regions 15 of FIG. 3B. Experimental evidence has shown that implanted nitrogen used to form buried insulator layers produces fewer defects in the crystal structure of epitaxial layer 10. The subsequent processing steps for this embodiment are the same as for the embodiment using implanted oxygen ions.

Returning to the embodiment including implanted oxygen ions, protective silicon dioxide layer 15A is then thermally grown on the exposed surfaces of epitaxial layer 10 to a thickness of approximately 1500 angstroms.

The structure of FIG. 2F is then subjected to a boron ion implantation having an energy of approximately 50 kiloelectron volts and a density of approximately 1E15 ions per centimeter squared. Silicon nitride layer 13 is then removed using a solution of phosphoric acid. Polycrystalline silicon layer 14A is then removed by selective wet etching or by selective plasma etching in, for example, a carbon tetrafluoride-oxygen plasma. A second boron implantation is then performed at an energy of approximately 50 to 70 kiloelectron volts and at a density of approximately 1E13 to 1E14 ions per centimeter squared. These boron ion implantations are then driven in to form P+ regions 16 and P region 17 of FIG. 2G thus forming a base region on silicon dioxide regions 15 providing a base-on-insulator structure. Silicon dioxide layer 14 is removed by wet etching in a 10% hydrafluoric acid solution such that a substantial portion of silicon dioxide layers 15A remain. N+ region 18 is then formed in the surface of epitaxial layer 10 by implanting an appropriate dopant ion such as phosphorous or arsenic ions to yield the structure of FIG. 2H. Contacts (not shown) are then made to collector diffusion 3 and extrinsic base regions 6 by techniques well known in the art. This structure limits the parasitic capacitance between the base and the collector by decreasing the dielectric constant at the extrinsic base-collector junction.

FIGS. 3A through 3D are side view schematic diagrams depicting the processing steps necessary to form another embodiment of the present invention. This embodiment provides a similar self-aligned process but provides a polycrystalline silicon emitter structure. Polycrystalline silicon layer 20 is formed by low pressure chemical vapor deposition to a thickness of approximately 2000 angstroms. Silicondioxide region 21 is thermally grown to a thickness of approximately 300 angstroms. Silicon nitride layer 22 is formed by low pressure chemical vapor deposition to a thickness of 2000 to 3000 angstroms. Silicon dioxide layer 23 is formed by plasma or chemical vapor deposition to a thickness of approximately 6000 to 8000 angstroms. These layers are then patterned and anisotropically etched using techniques well known in the art to provide the structure shown in FIG. 3A. A layer of silicon dioxide (not shown) approximately 3000 angstroms thick is then deposited by low pressure chemical vapor deposition on the surface of epitaxial layer 10. The silicon dioxide layer (not shown) is then anisotropically etched to form sidewall oxide layers 24. The structure of FIG. 3A is then subjected to an implant of oxygen ions (O2) at an energy of approximately 300 kiloelectron volts and a density of approximately 6E17 atoms per centimeter squared. This ion implantation is annealed to provide slicon dioxide regions 15 of FIG. 3B. Sidewall oxide layers 24 and plasma oxide layer 22 are then selectively etched using 10% dilute hydroflouric.

In a preferred embodiment, the structure of FIG. 3A is implanted with nitrogen ions (N2) rather than oxygen ions. These nitrogen ions are implanted at an energy of approximately 150 to 200 kiloelectron-volts and a density of approximately 5E17 to 1.5E18 ions per centimeter squared to form silicon nitride regions approximately 4000 angstroms below the surface of epitaxial layer 10. When annealed, these implanted ions will form silicon nitride regions instead of silicon dioxide regions 15 of FIG. 3B. Experimental evidence has shown that implanted nitrogen used to form buried insulator layers produces fewer defects in the crystal structure of epitaxial layer 10. The subsequent processing steps for this embodiment are the same for the embodiment using implanted oxygen ions.

Returning to the embodiment using implanted oxygen, silicon dioxide layer 25 is then formed by thermal oxidation to a thickness of approximately 1500 angstroms. Of importance, silicon dioxide layer 21 has a thickness of 300 angstroms which is much less than the thickness of silicon dioxide layer 25. The structure of FIG. 3B is then subjected to a boron ion implantation having an energy of approximately 70 kiloelectron volts and at a density of approximately 1E15 ions per centimeter squared. Silicon nitride layer 22 (FIG. 3B) is removed using an etching process which selectively removes silicon nitride while not etching silicon dioxide, such as wet etching in phosphoric acid. Silicon dioxide layer 21 is then removed by wet etching in a solution of 10% dilute hydrofluoric acid. The etching period is selected so as to be sufficient to remove oxide layer 21 but leave a substantial portion of silicon dioxide layer 25. Thus polycrystalline silicon region 20 is exposed.

The structure of FIG. 3C is subjected to an implant of N type dopant ions such as arsenic ions having an energy of approximately 50 kiloelectron volts and a density of approximately 1E15 to 1E16 ions per centimeter squared. This ion implantation is annealed to provide a very shallow N+ emitter region 26 of FIG. 3D and also drives in P+ regions 16. In addition, the ion implantation heavily dopes polycrystalline silicon region 20 to lower the resistivity of polycrystalline silicon region 20 to a negligible value. An ion implantation of boron ions is then performed at an energy of approximately 140 kiloelectron-volts and a density of approximately 1E13 to 1E14 ions per centimeter squared and annealled to provide base region 17. Base contact 27 opening is etched in silicon dioxide region 25 using techniques well known in the art. A contact (now shown) is also made to collector diffusion 3 by techniques well known in the art.

Thus transistor 50 is formed with silicon dioxide regions 15 using a self-aligned process. Silicon dioxide regions 15 reduce the capacitance of the base to collector junction for the reasons stated here and before.

The present invention provides the technical advantage of reducing the base to collector capacitance of vertical bipolar transistors. This reduction in capacitance allows vertical bipolar transistors constructed according to the teachings of this invention operate at increased frequency and to operate at increased speed when used in digital electronic circuits.