Semiconductor on Insulator Transistor

The present invention relates to the field of integrated circuit fabrication. More specifically, the present invention relates to the field of transistor fabrication in semiconductor-on-insulator structures.

Semiconductor-on-insulator structures are often used in applications where resistance to radiation effects on integrated circuits are required. Semiconductor-on-insulator structures include a layer of single crystal semiconductor, usually silicon, formed on an insulating structure such as sapphire or silicon dioxide. The single crystal layer is usually divided into separate regions or mesas to provide complete electrical isolation between the components formed in a mesas and components formed in other mesas. Semiconductor-on-insulator structures owe their robustness to radiation effects to the isolation of the electrical devices from the substrate. When a radiation particle, such as an alpha particle strikes a normal integrated circuit, the interaction between the electrical device and the substrate can cause errors in the operation of the integrated circuit. With semiconductor-on-insulator structures, this interaction is prevented by the insulating layer.

One problem presented by the use of semiconductor-on-insulator structures is the body effect on the channel of field effect transistors. Field effect transistors operate by coupling a voltage to a channel region which causes conduction between source and drain regions. The point at which the conduction between the source and drain occurs is called the threshold voltage. Because the channel is formed in an insulated structure, trapped charge can cause variations in the threshold voltage and thus inconsistencies in the operation of the field effect transistor. It is common practice to provide a ground contact to the channel to prevent these effects. However, the effects of a floating channel region are not entirely negative. For example, a floating channel region can provide higher drive current through the channel region. This provides for faster operation of integrated circuits using transistors with higher drive currents. The present invention is directed to providing a structure and method to allow for a floating channel region during the "on" condition of the transistor when the higher drive current is useful and to provide a control voltage on the channel when the transistor is to be off to provide consistent threshold control.

The described embodiments of the present invention provide a method and structure for actively controlling the voltage applied to the channel of field effect transistors. In the described embodiments, a transistor connected to the channel region is fabricated. The channel transistor has opposite conductivity type to the transistor using the main channel region. The source of the channel transistor is connected to the channel and the drain of the channel transistor is connected to a reference voltage. The same gate is used to control the channel transistor and the main transistor. When a voltage which causes the main transistor to be on conductive is applied, the channel transistor is off nonconductive, thus allowing the channel to float and allowing higher drive current. On the other hand, when a voltage to turn off the main transistor is applied, the channel transistor is turned on, thus clamping the channel region to the reference voltage. This allows for consistent threshold voltage control of the main transistor.

In a preferred embodiment, the channel of the main transistor is used as the source of the channel transistor and the gate of the main transistor extends onto the channel region of the channel transistor. The reference voltage is then connected to the drain region which is formed on the opposite side of the channel transistor channel region from the main transistor's channel. In preferred embodiments, the drain region is positioned adjacent to the source region of the main transistor. In most non-pass transistor circuits, the source of the main transistor is established. In this embodiment, the source of the main transistor and the drain of the channel transistor can be connected together locally, thus providing a simpler interconnection structure.

FIG. 1 is a plan view diagram of one embodiment of the present invention;

FIG. 2 is an electrical schematic diagram showing the operation of the structure shown in FIG. 1;

FIGS. 3A and 3B are section diagrams of sections AA and sections BB, respectively, of FIG. 1;

FIGS. 4A through 4K are side view schematic diagrams showing the processing steps necessary for fabricating the embodiment shown in FIG. 1;

FIG. 5 is a plan view diagram of a preferred embodiment of the present invention;

FIG. 6 is an electrical schematic diagram showing the electrical operation of the embodiment shown in FIG. 5;

FIGS. 7A and 7B are section diagrams along sections AA and BB, respectively, of FIG. 5;

FIGS. 8A through 8D are side view schematic diagrams showing the processing steps necessary to fabricate the embodiment shown in FIG. 5; and

FIGS. 9 and 10 are plan view diagrams of preferred embodiments of the invention.

FIG. 1 is a plan view diagram of one embodiment of the present invention. FIG. 2 is an electrical schematic diagram showing the operation of the embodiment shown in FIG. 1. FIGS. 3A and 3B are section diagrams of the embodiment shown in FIG. 1. FIGS. 4A through 4K are side view schematic diagrams showing the fabrication of the embodiment shown in FIG. 1.

The embodiment shown in FIG. 1 comprises two merged transistors. The first transistor is the main transistor 56 and the second transistor is the body or channel transistor 58. This structure is formed in mesa 14 on insulating layer 12. Insulating layer 12 may be silicon dioxide, sapphire or any number of insulating structures known in the art. Gate 42 controls the current between source 54 and drain 52 of main transistor 56. Source 54 and drain 52 are connected to other devices in the integrated circuit containing the present embodiment of the invention to provide the function of a field effect transistor. Gate 42 controls the current through channel region 34 thus providing the control of the current from drain 52 to source 54. Channel region 34 is a p-type region which is self-aligned to gate 42. P+ source region 48 is electrically connected to channel region 34 by virtue of their similar conductivity types. N-channel 48 is controlled by gate 42, which in turn controls the current between drain region 48 and source region 46. These components comprise body transistor 58. Source region 46 is generally connected to a reference potential to provide a fixed voltage to channel region 34 when body transistor 58 is on. The structure of FIG. 1 is designed so that when transistor 56 is on, channel region 34 is allowed to float and thus increased drive current is available for the circuit including transistor 56. When transistor 56 is off, channel region 34 is connected to a reference potential so that a consistent threshold voltage is available for predictable, accurate operation of transistor 56 in the circuitry containing transistor 56.

FIG. 2 is an electrical schematic diagram showing the electrical operation of the structure of FIG. 1. Source region 54 and drain region 52 are connected to other devices in the integrated circuit. Gate 42 controls both transistor 56 and transistor 58. Drain region 48 is connected to channel 34 of transistor 56. Source region 46 is connected a reference potential. Transistor 56, in this embodiment, is an n-channel transistor and transistor 58 is a p-channel transistor. When a high voltage is applied to gate 42, transistor 58 is off and transistor 56 is on. Thus, the channel region is allowed to float on transistor 56. When a low voltage is applied to gate 42, transistor 58 is on and transistor 56 is off. Because transistor 58 is on, the voltage level on channel region 34 is controlled by the reference potential connected to source region 46.

FIGS. 3A and 3B are cross-sectional diagrams along section lines AA and BB, respectively, of FIG. 1. As can be seen from FIG. 3A, insulator layer 12 is formed on the surface of substrate 10. In this particular embodiment, substrate 10 is a crystalline silicon substrate and insulating layer 12 is a layer of silicon dioxide formed by the implanted oxygen method. Descriptions of the implanted oxygen method (SIMOX) may be found in U.S. Pat. Nos. 3,855,009 and 4,241,359. Sidewall oxide regions 28 are provided on the sidewalls of mesa 14 to stabilize the conduction at the surface of the edge of the mesa. Gate 42 is insulated from channel regions 34 and 38 by silicon dioxide layer 30. Gate 42 controls the conductivity between drain region 48 and source region 46 as shown in FIG. 3A. Gate region 42 also controls the conductivity between source region 54 and drain region 52 as shown in FIG. 3B.

FIGS. 4A through 4K are side view schematic diagrams corresponding to the view provided by FIG. 3A showing the processing steps necessary for the fabrication of the embodiment shown therein. Crystalline silicon substrate 10 is implanted with oxygen ions and annealed to form insulating layer 12. Epitaxial layer 14 is then formed on the surface of the remaining structure using the single crystal silicon structure remaining above silicon dioxide layer 12 for nucleation sites. Substrate 10 and thus, subsequently, epitaxial layer 14 may be n or p-type layers using a 100 crystalographic orientation. The doping concentration is generally on the order of 3 to 6 ohm centimeters for n-type and 10 to 12 ohm centimeters for p-type. The process described herein assumes epitaxial layer 14 is n type.

A layer of silicon dioxide 16 is then grown to a thickness of approximately 350 Angstroms using thermal oxidation techniques. A layer of silicon nitride 18 is then deposited by low pressure chemical vapor deposition to a thickness of approximately 1,700 Angstroms. A layer of silicon dioxide 20 is then deposited on the surface of silicon nitride layer 18 to a thickness of approximately 3,200 Angstroms using low pressure chemical vapor deposition. A layer of photoresist 22 is then deposited on the surface of silicon dioxide layer 20. Using known lithographic techniques, photoresist layer 22 is optically exposed and patterned. Using photoresist layer 22 as an etch mask, silicon dioxide layer 20, silicon nitride layer 18 and silicon dioxide layer 16 are etched. Layers 16, 18 and 20 are etched using reactive ion etching techniques known in the art to provide anisotropic etching properties.

Photoresist layer 22 is then removed using common wet removal techniques. The structure of FIG. 4B is then subjected to an ion implantation using two steps. The first step is implantation of boron ions having an energy of approximately 30 kiloelectron volts and a density of approximately 3.times.10.sup.12 ions/centimeter.sup.2. The second step uses boron ions at an energy of approximately 80 kiloelectron volts and a density of approximately 5.times.10.sup.12 ions/centimeter.sup.2. This forms channel stop regions 14' as shown in FIG. 4B. Silicon dioxide layer 20 is then removed using any of a number of etching techniques such as a wet deglaze using hydrofluoric acid. A layer of silicon dioxide 24 is then deposited using low pressure chemical vapor deposition to a thickness of approximately 1,000 Angstroms. Silicon dioxide layer 24 is then etched using an anisotropic etching process, such as reactive ion etching using a fluorine based chemistry (such as carbon tetrafluoride), to provide sidewall oxide regions 26 as shown in FIG. 4D. Silicon nitride layer 18, silicon dioxide layer 16 and silicon sidewall oxide layers 26 are then used as an etch mask for the etching of epitaxial silicon layer 14. Epitaxial silicon layer 14 is etched using reactive ion etching using hydrochloric acid to provide the structure of FIG. 4E. FIG. 4E includes sidewall passivating layers 14' which prevent unwanted conduction along the edges of mesa 14. The formation of such sidewall passivating layers is shown in Matloubian, U.S. Pat. No. 4,753,896, issued June 28, 1988 and assigned to the assignee of this application. The structure of FIG. 4E is then subjected to oxidation to provide approximately 250 Angstroms of silicon dioxide on the sidewalls of mesa 14. A further layer of low pressure chemical vapor deposited silicon dioxide having a thickness of approximately 2,500 Angstroms is deposited and reactive ion etched to provide sidewall silicon dioxide layers 28 as shown in FIG. 4F. In subsequent diagrams, passivation regions 14' are omitted for clarity of the drawings.

Silicon nitride layer 18, silicon dioxide layer 16 and sidewall regions 26 are removed using a two-step process of etching using hot phosphoric acid to remove silicon nitride layer 18 and anisotropic etching in a fluorine based chemistry to remove silicon dioxide layer 16 and sidewall silicon dioxide layers 26. Because of the anisotropy of reactive ion etching, silicon dioxide layers 28 will remain. A layer of silicon dioxide 30 is then grown by thermal oxidation of the surface of mesa region 14 to provide silicon dioxide layer 30 as shown in FIG. 4G. A photoresist layer 32 is then deposited and patterned using common photolithographic techniques to provide the structure of photoresist layer 32, as shown in FIG. 4G. The structure of FIG. 4G is then subjected to a first ion implantation of boron ions at an energy of approximately 80 kiloelectron volts and a density of approximately 3.5.times.10.sup.12 ions/centimeter.sup.2. This sets the back threshold voltage of p region 34 to a voltage in excess of 25 volts. The front threshold voltage of p region 34 may then be adjusted using an ion implantation of boron ions at an energy of approximately 25 kiloelectron volts and a density selected to provide the selected threshold voltage. The edge of photoresist layer 32 is selected to fit in the region .DELTA. which is an overlap region where p+ drain region 48 will be formed. Because p+ region 48 will be formed in this area, the alignment of photoresist layer 32 within the region .DELTA. is not critical.

Photoresist layer 32 is then removed and a second photoresist layer 36 is deposited and patterned as shown in FIG. 4H. The structure of FIG. 4H is then subjected to an ion implantation of boron having an energy of approximately 25 kiloelectron volts and a density of approximately 1.0.times.10.sup.12 ion/centimeter.sup.2. This sets the front threshold voltage at approximately -1 volt. A second implantation of phosphorous ions having an energy of approximately 180 kiloelectron volts and a density of approximately 1.2.times.10.sup.12 ions/centimeter.sup.2 is then performed to set the back threshold voltage at approximately -13 volts. As is well known in the art, the energies and densities of these ion implantations may be selected to provide particular characteristics desirable in particular circumstances. In the structure shown in FIG. 4H, this ion implantation leaves a gap 40 between n-channel region 38 and p-channel region 34. This is a region within the region .DELTA. as shown in FIG. 4G and n-channel region 38 and p-channel region 34 may provide a gap as shown in FIG. 4H or may overlap without adverse effects on the operation of the integrated circuit.

Photoresist layer 36 is then removed and silicon dioxide layer 30 is stripped and a second silicon dioxide layer 31 is formed to a thickness of approximately 250 Angstroms using thermal oxidation. A layer of polycrystalline silicon 42 is then deposited to a thickness of approximately 4,500 Angstroms and patterned to provide the structure shown in FIG. 4I. A layer of photoresist 44 is then deposited and patterned as shown in FIG. 4J. The structure of FIG. 4J is then subjected to an ion implantation of boron ions having an energy of approximately 20 kiloelectron volts and a density of approximately 2.times.10.sup.15 ions/centimeter.sup.2. This ion implantation forms source region 48 and drain region 56 as shown in FIG. 4J. Photoresist layer 44 is then removed and photoresist layer 50 is deposited on the surface of the structure as shown in FIG. 4K. The structure of FIG. 4K is then subjected to an ion implantation of phosphorous ions having an energy of approximately 140 kiloelectron volts and a density of approximately 5.times.10.sup.14 ions/centimeter.sup.2 and a second ion implantation of arsenic ions having an energy of approximately 150 kiloelectron volts and a density of approximately 3.5.times.10.sup.15 ions/centimeter.sup.2. These ion implantations form source region 54 and drain region 52 as shown in FIG. 1. Thus, the embodiment shown in FIG. 1 is fabricated.

A second preferred embodiment is shown in plan view in FIG. 5. An electrical schematic of the operation of the embodiment shown in FIG. 5 is shown in FIG. 6. Cross-sectional diagrams along sections AA and BB are shown in FIGS. 7A and 7B, respectively. The fabrication steps necessary to fabricate the embodiment shown in FIG. 5 are shown in FIGS. 8A through 8B.

The embodiment of FIG. 5 includes two transistors, a main transistor 156 and a body transistor 158. In the main transistor 156, gate 142 controls current between source 154 and drain 152. This is controlled by controlling the conductivity of channel region 134. Channel region 134 abuts n-channel region 148. P-channel region 134 serves as the source and p+ region 146 serves as the drain of body transistor 158. Conductivity between p-channel region 134 and p+ drain 146 is controlled by gate 142 by controlling the conductivity of n-channel region 148. This comprises body transistor 158.

An electrical schematic diagram of the structure of FIG. 5 is shown in FIG. 6. Gate 142 controls the conductivity of n-channel transistor 156 and p-channel transistor 158. When a high voltage signal is applied to gate 142, p-channel transistor 158 is off and n-channel transistor 156 is on. Because p-channel transistor 158 is off, channel region 134 is allowed to float and maximum drive current is provided by transistor 156. When a low voltage signal is applied to gate 142, p-channel transistor 158 is on and n-channel transistor 156 is off. Because p-channel transistor 158 is on, channel region 134 is tied to a reference potential and a consistent threshold voltage and controllable on/off characteristics are provided by transistor 156.

FIGS. 7A and 7B are cross-sectional diagrams of cross sections AA and BB, respectively, of FIG. 5. As can be seen in FIG. 7A, gate 142 controls the conductivity of channel region 148 and thus the conductivity between p region 134 and p+ region 146. As can be seen from FIG. 7B, gate 142 also controls the conductivity of p region 134 which controls the conductivity between source region 152 and drain region 154.

FIG. 8A is a side view schematic diagram showing an intermediate processing step for fabricating an embodiment shown in FIG. 5. FIG. 8A corresponds to the step in the procedure shown in FIG. 4G and the steps used to fabricate the structures shown in FIG. 4G are also used to fabricate the structures shown in FIG. 8A where the reference numeral plus 100 from FIG. 4G is used as a reference numeral to corresponding components in FIG. 8A (for example, sidewall oxide regions 128 correspond to sidewall oxide regions 28 in FIG. 4G).

The structure of FIG. 8A is subjected to a first ion implantation of phosphorous ions having an energy of approximately 180 kiloelectron volts and a density of approximately 1.2.times.10.sup.12 ions/centimeter.sup.2. This adjusts the back threshold voltage and in part adjusts the front threshold voltage. The front threshold voltage is then further adjusted by an ion implantation of boron ions having an energy of approximately 25 kiloelectron volts and a density of approximately 1.45.times.10.sup.12 ions/centimeter.sup.2. A layer of photoresist 132 is then deposited and patterned, as shown in FIG. 8B. The structure of FIG. 8B is then subjected to an ion implantation of boron ions having an energy of approximately 85 kiloelectron volts and a density of approximately 3.7.times.10.sup.12 ions/centimeter.sup.2. This counterdopes the back threshold implantation used to form n region 138. An additional ion implantation of boron having an energy of 25 kiloelectron volts may be performed to adjust the front threshold voltage of p region 134 using a density selected to produce the desired threshold voltage. Photoresist layer 132 is then removed as is silicon dioxide layer 130. Silicon dioxide layer 130 is removed using any one of numerous silicon dioxide etching techniques such as reactive ion etching in a fluorine base chemistry.

A gate oxide is then thermally grown using thermal oxidation to a thickness of approximately 250 Angstroms, as shown in FIG. 8C. A layer of polycrystalline silicon 142 is then deposited and patterned, as shown in FIG. 8C. A layer of photoresist 136 is then deposited on the surface of patterned layer 142 and patterned to provide the structure shown in FIG. 8C. The structure of FIG. 8C is then subjected to an ion implantation of boron ions having an energy of approximately 20 kiloelectron volts and a density of approximately 2.times.10.sup.15 ions/centimeter.sup.2 to form p+ region 146, as shown in FIG. 8C. Photoresist layer 136 is then removed and photoresist layer 150 is deposited and patterned as shown in FIG. 8D. The structure of FIG. 8D is then subjected to an ion implantation of phosphorous ions having an energy of approximately 140 kiloelectron volts and a density of approximately 5.times.10.sup.14 ions/centimeter.sup.2 and a second ion implantation of arsenic ions having an energy of approximately 150 kiloelectron volts and a density of approximately 3.5.times.10.sup.15 ions/centimeter.sup.2 to form n+ source and drain regions 154 and 152, as shown in FIG. 5.

FIGS. 9 and 10 are plan view diagrams of preferred embodiments of the invention. In FIG. 9, main transistor 256 comprises drain 252, source 254, channel region 234 and gate 242. The channel transistor comprises main channel 234, channel transistor channel 248, channel transistor drain 246. In addition to a more compact layout, the structure of FIG. 9 has the channel transistor drain 246 and the main transistor source 254 formed adjacent to each other. In many circuits, the main transistor source 254 and the channel transistor drain 246 will be connected to the same reference potential. The structure of FIG. 9 allow those to regions to be easily strapped together by a silicide layer of titanium, molybdenum or other refractory metal, or by titanium nitride interconnection or other local interconnection. This provides a very compact structure by limiting the necessary connections to the transistors. In FIG. 10, main transistor 356 comprises drain 352, source 354, channel region 334 and gate 342. The channel transistor comprises main channel 334, channel transistor channel 348, channel transistor drain 346. The structure of FIG. 10 provides a slightly more compact layout than that of FIG. 9 with the same advantages.

Although specific embodiments of the present invention are herein described, they are not to be construed as limiting the scope of the invention. Numerous modifications to the present invention will become obvious to those skilled in the art in light of this specification. For example, to provide transistors having opposite conductivity characteristics, regions of opposite conductivity to those shown in the specification may be used. In addition, the structures herein used may find applicability in structures other than silicon-on-insulator structures. For example, when field effect transistors are formed in isolated wells in BiCMOS integrated circuits, where the wells are completely isolated by buried regions beneath the wells, this smart contact scheme may be advantageously used. The scope of the present invention is only limited by the claims appended hereto.