Disposable spacer for symmetric and asymmetric Schottky contact to SOI MOSFET

Abstract

A silicon on insulator transistor is disclosed which has a Schottky contact to the body. The Schottky contact may be formed on the source and/or drain side of the gate conductor. A spacer, with at least a part thereof being disposable, is formed on the sidewalls of the gate conductor. Extension regions are provided in the substrate which extend under the spacer and the gate conductor. Source and drain diffusion regions are implanted into the substrate adjacent to the extension regions. The disposable part of the spacer is then removed to expose a portion of the extension region. A metal layer is formed at least in the extension regions, resulting in the Schottky contact.

Description

FIELD OF THE INVENTION

The present invention generally relates to a structure for minimizing floating body effects in silicon on insulator (SOI) technology utilizing a Schottky diode and a method of forming the same.

BACKGROUND OF THE INVENTION

SOI technology is becoming an increasingly important field in the manufacture of integrated circuits. SOI technology deals with forming transistors in a layer of semiconductor material which overlies an insulating layer. The insulating layer is formed on an underlying substrate. The transistor comprises source and drain regions implanted into the semiconductor material. A gate structure is formed on the semiconductor material between the source and drain regions. The source and drain regions are typically formed deep into the semiconductor material and reach the insulating layer. A part of the semiconductor material isolated between the source and drain regions and underlying the gate structure is referred to as the body of the transistor. Due to the difficulty of forming a body contact, the body of the transistor in SOI is typically left electrically floating. A floating body can sometimes adversely affect the characteristics of the transistor. For example, when a high voltage is applied on the drain of the transistor, there is a charge accumulation in the body. Normally, in bulk transistors, the charge accumulation is carried away by having the wafer grounded. However, in SOI, the body is isolated by the insulating layer so a charge accumulates in the body area. As a result, a lateral bipolar transistor is formed with the source acting as the emitter and the body acting as the base. When a positive voltage is applied to the drain, a positive charge accumulates in the body. A diode formed by the junction of the source and body will turn on at a certain voltage. At this voltage, the current conducted through the transistor begins to increase exponentially. Thus, as the voltage on the drain is gradually increased, at a certain voltage, for example about 2.5 volts, the current through the transistor will begin to increase exponentially.

Although there is some conduction of current through the diode formed by the body and the source, the conduction is not enough to remove all the charge accumulating in the body. In essence, there is a race between the amount of current drawn off from the body by the source and the amount of charge accumulated in the body as a result of the high voltage applied to the drain.

This characteristic of SOI technology may cause difficulty during burn-in of a chip. Burn-in of a chip refers to the practice of operating the chip at an elevated voltage in order to test it. For example, if a chip is designed to operate at 1.8 volts, a voltage approximately 1.5 times the operating voltage, that is approximately 2.7 volts, is applied to the chip during burn-in. This is done so if any incipient defects exist in the transistor, the transistor will fail during bum-in and not during operation in the field. However, due to the above described tendency of the body to accumulate charge, the transistor may not be able to accommodate the higher burn-in voltage. Thus, there is a need to accelerate the amount of charge removed from the body to the source in order to keep the charge of the body low. This will then allow the diode to operate normally during burn-in.

SUMMARY OF THE INVENTION

A field effect transistor (FET) on an SOI substrate and method of forming the same is provided. An SOI substrate having a silicon base substrate, an insulating layer on the base substrate, and a semiconductor material on the insulating layer is provided. A conductive gate is formed on the semiconductor material. First spacers are formed on sidewalls of the gate and on the semiconductor material. Diffusion extensions are formed in the semiconductor material adjacent to and extending under a part of the gate. Second spacers are then formed on the first spacers and on the semiconductor material. Deep diffusions are implanted into the semiconductor material. The deep diffusions are implanted adjacent to the second spacers, close to the insulating layer and abutting the diffusion extensions. The semiconductor material between the deep diffusions defines a body region of the FET. At least a part of one of the second spacers is removed from the first spacers on the sidewall of the gate to expose a portion of the diffusion extension in the semiconductor material. A metal layer is formed in the semiconductor material at least in the exposed portion of the diffusion extension.

BRIEF DESCRIPTION OF THE DRAWINGS

The present invention will now be further described in the following pages of specification when taken in conjunction with the attached drawings, in which:

FIG. 1 is a graph of the current-voltage characteristics of an ideal diode and a Schottky diode;

FIGS. 2-7 are cross sectional views illustrating the fabrication of a transistor according to one embodiment of the invention; and

FIG. 8 is a cross sectional view illustrating a transistor fabricated according to another embodiment of the invention.

DETAILED DESCRIPTION OF THE INVENTION

The I-V characteristics of a Schottky diode and an ideal diode are shown in FIG. 1 . When a forward bias is placed on either a Schottky or an ideal diode, the current rises at a particular rate. For an ideal diode, for example, for every increase of about 60 mv the current increases by approximately a factor of 10. Also, as shown on the right side of FIG. 1 , when an ideal diode is reverse biased, the current initially increases a small amount and then maintains a relatively constant value. In comparison, a Schottky diode has a higher starting current. This is illustrated by Point 1 in FIG. 1 which shows at about zero volts the Schottky diode has a higher current than an ideal diode. Additionally, the Schottky diode has a steeper slope under forward bias. For example, for every increase in voltage of approximately 40mv the current increases by a factor of 10 . Therefore, for a given voltage V 1 , the Schotiky diode will conduct a greater amount of current I, than an ideal diode I 2 .

The present invention utilizes a Schottky diode in order to control the floating body effects of SOI. A Schottky diode is preferably formed only on the source side of an SOI transistor. The Schottky diode conducts current from the body to the source of the transistor, keeping a charge accumulating in the body low. The Schottky diode may be formed using a spacer which is at least partially disposable.

Referring now to FIGS. 2-7 , a method of forming a Schottky contact for an SOI FET will be described. An SOI substrate comprises an insulating layer 8 formed on a base substrate 6 . The insulating layer is typically silicon dioxide and the base substrate is usually silicon. A semiconductor material 10 in which the transistors are to be formed is arranged on the insulating layer 8 . The semiconductor material 10 is also usually silicon. A gate conductor 12 , which will define the gate of a MOS transistor, is formed on a top surface of the semiconductor material 10 . First sidewall spacers 14 are formed on sides of the gate conductor 12 and on the top surface of the semiconductor material 10 . The resulting structure is shown in FIG. 2 . Alternatively, a gate insulating layer (not shown) may be formed between the semiconductor material 10 and the gate conductor 12 .

Next, as shown in FIG. 3 , an extension region 16 is formed in the semiconductor material 10 . The extension region 16 is usually formed by implanting an impurity at a doping concentration of about 10 18 to 10 19 dopants/cm 3 . The implantation of the impurities may be conducted in a known manner to form the extension region 16 . Ideally, the extension region 16 has a dopant concentration lower than that of the later formed diffusion regions and is formed over the surface of the semiconductor material 10 on both sides of the gate conductor 12 . Also, the extension region 16 should extend under both the first spacers 14 and the gate conductor 12 as shown in FIG. 3 . Preferably, the extension region 16 extends about 100-200 Å under the gate conductor 12 . The depth the extension region 16 is formed into the semiconductor material 10 may vary depending upon the particular design of the transistor.

Referring now to FIG. 4 , second spacers 18 are formed on the first spacers 14 . The second spacers 18 should also be formed on the top surface of the semiconductor material 10 to cover a portion of the extension region 16 . The first 14 and second 18 spacers are preferably formed so that a portion of the second spacers are disposable. This may be accomplished by forming the first and second spacers from different materials to facilitate the removal of the second spacers 18 , if desired, at a later point in the process. For example, the first spacers 14 may be made of nitride and the second spacers 18 may be made of oxide. Additionally, the first and second spacers may be combined into a single structure, a portion of which is disposable.

Next, diffusion regions 20 , corresponding to the source and drain of a transistor, are formed in the semiconductor region 10 as shown in FIG. 5 . The diffusion regions 20 may be formed by implanting an impurity at a doping concentration of approximately 10 20 -10 21 dopants/cm 3 . The implantation of the impurities is preferably done in a known manner directly over the extension regions 16 using the second spacers 18 as a mask. The diffusion regions 20 should not extend very far under the second spacers 18 , if at all. The portions of the extension region 16 formed under the second spacer 18 are protected from additional dopant implantation during the formation of the diffusion regions 20 . Additionally, the diffusion regions 20 are formed deep into the semiconductor material 10 close to the insulating layer 8 . Ideally, the diffusion regions 20 should reach the insulating layer 8 , however in practice, the distance between the diffusion regions and the insulating layer 8 may be approximately 100 Å.

The process up to this point can be carried out using known techniques in the field of semiconductor manufacture. Various steps not directly related to the present invention have been omitted for clarity. However, it should be noted that the materials for the first and second spacers should be selected with the subsequent portion of the process in mind.

At this point, as shown in FIG. 6 , a body 21 of the transistor has been defined in the semiconductor material 10 between the diffusion regions 20 and under the gate 12 . Due to the presence of the insulating layer 8 , the body 21 is isolated from the base substrate 6 . Thus, the body 21 is typically electrically floating.

The extension regions 16 are also located between the diffusion regions 20 at the top surface of the semiconductor material 10 . The first 14 and second 18 spacers are arranged on the semiconductor material 10 , above the extension region 16 . A portion of the extension region 16 at the top surface of the semiconductor material 10 can be exposed by removing a part of one of the spacers. The second spacer 18 was formed to be disposable as described above to facilitate the formation of the Schottky diode. The Schottky diode is formed by contacting the extension region with a metal layer. Preferably, the second spacer on source side is removed entirely or in part. The amount of the extension region 16 exposed can be varied by changing the size of the second spacer. A larger area of the extension region 16 covered by the second spacers 18 will result in more of the extension region 16 being exposed when the second spacer 18 is removed.

FIG. 6 shows the device with the second spacer 18 on one side of the gate conductor 12 removed. The second spacer may be removed by forming a mask exposing only the source side of the transistor. A removal procedure selective to material of the second spacer, such as a DBF dip, is performed to remove all or part of the second spacer, exposing a portion 17 of the extension region 16 . The second spacer 18 on the drain side of the transistor preferably remains. The mask is then removed and a metal layer 22 is formed at least in the exposed portion 17 of the extension region 16 to form a Schottky diode. As shown in FIG. 7 , the metal layer should be self-aligned to the first spacer 14 . Any contact between the metal layer 22 and the extension region 16 should form a Schottky diode. The presence of the Schottky diode on the source side of the transistor will improve the leakage current between the body 21 and the source which, in turn, reduces the charge accumulated in the body 21 of the transistor. Ideally, the metal, preferably a silicide, is formed on the entire exposed surface of the substrate, that is, in both the exposed portion of the extension region 16 and in the diffusion region 20 . By also forming the metal layer 22 in the diffusion region 20 , the normal function of the metal layer of lowering the resistance of the source is achieved.

In the embodiment shown in FIG. 7 , the metal layer 22 is formed deeper into the semiconductor material 10 than the extension region 16 . The metal layer 22 thereby contacts the body 21 of the transistor. A metal layer 22 thus formed contacts the portion 17 of the extension region 16 exposed by removal of the second spacer 18 , the body 21 of the transistor, and the diffusion region 20 that was abutting the exposed portion 17 of the extension region 16 . As mentioned above, as long as the metal layer contacts a portion of the extension region 16 , a Schottky diode should be formed. Thus, the metal layer may contact the body 21 of the transistor as shown in FIGS. 7 , but this is not required. The Schottky diode may have more conduction if the metal layer contacts the body, but it will have a significant effect even if contact to the body is not established.

Thus, an asymmetric Schottky contact to an SOI transistor has been formed. The Schottky contact is formed by removing at least a portion of a disposable spacer from the source side of the transistor, while the spacer on the drain side remains. By removing the disposable spacer, a portion of a low doped extension region is exposed. A metal layer, such as a suicide, is then formed at least in the exposed extension region to form the Schottky contact. The metal layer may also be formed in the diffusion region adjacent to the exposed portion of the extension region, as well as in the diffusion region on the drain side of the transistor to lower the resistance of the source and/or drain. Any significant contact between the metal layer and the low doped extension region should result in the formation of a Schottky diode.

For extremely small gates where the lithography alignment to the gate is difficult, it is possible to remove the disposable spacers from both the source and drain sides of the transistor. The metal layer is then formed on both sides of the transistor resulting in symmetric Schottky contact to both the source and the drain of the transistor. FIG. 8 shows a transistor with symmetric Schottky contacts. The method of forming this device is substantially the same as the process for forming the asymmetric device described above except the step of forming the mask to expose only the source side of the transistor is omitted. Thus, in this case no extra lithography for exposing only the source is required. The second spacers are removed from both sides of the transistor. Therefore, when the metal layer 22 is formed, it is formed in the extension region 16 on both the source and drain side of the transistor, as shown in FIG. 8 . Although forming symmetric Schottky contacts to both the source and the drain of a transistor requires fewer lithographic steps, there are some associated problems. The presence of the Schottky diode slightly increases the resistance of the transistor. Therefore, it is preferable that only one Schottky diode be formed to keep the resistance of the transistor at a minimum. Additionally, the quality of the Schottky contact to the drain should be controlled carefully in order to minimize the reverse bias leakage current from the drain. This can be achieved by numerous methods previously known, such as implanting the source and drain with germanium before forming the metal layer.

Accordingly, a method and structure for minimizing the floating body effects of an SOI device has been provided. A Schottky contact to the source and/or drain is formed to reduce the charge accumulated in the body of the transistor. The Schottky contact is formed using disposable spacers such that the Schottky contact is self aligned. The Schottky diode enhances the forward bias leakage of the transistor and will hold the body of the transistor at a lower potential than it would otherwise sit. Thus, the range of body voltages of the transistor is narrowly bounded.

While a preferred embodiment of the invention has been described above, since variations in the invention will be apparent to those skilled in the art, the invention should not be construed as limited to the specific embodiments described above.

Claims

1. A method of forming a transistor comprising:forming a gate conductor on a top surface of a substrate; forming spacers on sides of the gate conductor; providing extension regions in the substrate extending at least under the spacers and the gate conductor; forming diffusion regions in the substrate adjacent to the extension regions; removing a part of at least one of the spacers to expose a portion of the extension region at the top surface of the substrate; and contacting the exposed portion of the extension region with a metal layer.

2. The method of claim 1 wherein the spacers comprise a permanent portion adjacent to the gate conductor and a disposable portion provided on the permanent portion.

3. The method of claim 2 wherein the permanent portion is nitride and the disposable portion is oxide.

4. The method of claim 2 wherein the removing step removes the disposable portion of the spacer.

5. A method of forming an FET on an SOI substrate, comprising:providing an SOI substrate having a silicon base substrate, an insulative layer on the base substrate, and a silicon layer on the insulative layer; forming a conductive gate on the silicon layer; forming first spacers on sidewalls of the gate and on the silicon layer; implanting diffusion extensions into the silicon layer adjacent to and extending under the gate; forming second spacers on the first spacers and on the silicon layer; implanting deep diffusions into the substrate layer adjacent to the second spacers and abutting the diffusion extensions, wherein the silicon layer between the deep diffusions defines a body region of the FET; removing the second spacer from at least one of the first spacers on the sidewalls of the gate including exposing a portion of the diffusion extension in the silicon layer under the second spacer; and forming a metal layer in the silicon layer at least in the exposed portion of the diffusion extension, wherein the metal layer is formed deep enough to contact the diffusion extension.

6. The method of claim 5 wherein the deep diffusions are implanted down to the insulative layer.

7. The method of claim 5 wherein the metal layer is formed deeper into the silicon layer than the diffusion extension.

8. The method of claim 7 wherein the metal layer abuts each of: the exposed portion of the diffusion extension, the body region of the FET, and the deep diffusion that was abutting the diffusion extension.

9. The method of claim 5 wherein the metal layer contacts a dopant concentration of less than or equal to 1×1019 dopants/cm3 in the extension region.

10. A method of forming a transistor comprising:providing an SOI substrate; forming a gate conductor on the substrate; forming first spacers on sidewalls of the gate conductor; forming extension regions in the substrate, the extension regions extending under the first spacers and the gate conductor; forming second spacer on the first spacers; implanting diffusion regions into the substrate adjacent to the second spacers; removing at least one of the second spacers to expose a portion of the extension region; and forming a metal layer in the substrate contacting at least the extension region.

11. The method of clam 10 wherein the second spacers are formed on a portion of the extension region.

12. The method of claim 10 wherein the diffusion regions are formed down to an insulating layer.

13. The method of claim 10 wherein the diffusion regions define a body therebetween.

14. The method of claim 10 wherein the metal layer is also formed in the diffusion regions.

15. The method of claim 13 wherein the metal layer extends through the extension region to contact the body.

16. The method of claim 10 wherein both second spacers are removed.

17. The method of claim 10 wherein the metal layer is formed on both sides of the gate.

18. The method of claim 10 wherein the metal layer is self-aligned to the first spacer.

19. The method of claim 10 wherein the first and second spacers are formed from different materials.

20. The method of claim 19 wherein the first spacers is a nitride and the second spacer is an oxide.

21. The method of claim 19 wherein the removing step comprises:forming a mask exposing one side of the transistor; and performing a dip to selectively remove the material of the second spacer.