Method for Forming a Polysilicon to Polysilicon Capacitor

This invention relates in general to integrated electronic devices and methods of fabrication thereof. More particularly, the present invention relates to a method for forming a polysilicon to polysilicon capacitor and an apparatus formed therefrom.

CMOS and BiCMOS are rapidly evolving as the premiere technology for integrating highly complex analog-digital subsystems on a single chip. Such single chip subsystems require precision capacitors. Polysilicon to polysilicon capacitors have been increasingly used to provide this necessary precision.

In prior art devices, several double level polysilicon (DLP) processes have often been developed to form the polysilicon to polysilicon capacitors. In particular, the LinEPIC DLP process uses a two-mask approach to define a capacitor bottom plate. The first polysilicon layer is masked and etched separate from the second polysilicon layer. Due to the separate masking and etching steps, this process has been considerably complicated and time consuming. Additionally, the topography that is associated with this DLP process requires an additional step of planarization prior to depositing metal on the appropriate contact points.

A second process known as the 4/3 Linear process uses a single mask approach to define a capacitor bottom plate. The first level of polysilicon serves as both the bottom plate and the CMOS gate. After the interlevel dielectric is formed, the second polysilicon layer is deposited to form the capacitor top plate. In order to eliminate filaments from the bottom plate edges and the CMOS gate edges, a large overetch is required. If there is a negative slope on the bottom plate edge, filaments will be trapped under the bottom plate edge. Since this is a DLP process, the added topography also requires additional planarization prior to metalization.

Therefore, a need has arisen for a simplified method of fabricating polysilicon to polysilicon capacitors used on CMOS and BiCMOS. Additionally, it would be an advantageous to reduce the complications and expenses associated with DLP processes. There is also a need to eliminate filaments from the bottom plate edges and CMOS gate edges without requirement of a large overetch which may cause system failure. Finally, there is a need to eliminate the step of planarization prior to applying metal to the appropriate contacts.

The present invention provides a method for forming a polysilicon to polysilicon capacitor used in CMOS and BiCMOS devices and apparatus formed therefrom. One aspect of the present invention is a method for forming the CMOS. Initially, a layer of field oxide is formed over a substrate. This field oxide layer has a gate formed therein for the future CMOS transistor of the device. A first layer of polysilicon is deposited onto the field oxide at a thickness of approximately 2,000 angstroms. After the first polysilicon layer has been deposited, the temperature of the system is increased to approximately 900 degrees Centigrade (.degree. C.) which forms a thin layer of oxide over the entire polysilicon layer. A layer of silicon nitride is then deposited over the layer of oxide, and a first photoresist layer is selectively deposited onto the silicon nitride layer. After the photoresist layer has been deposited, the exposed surfaces are etched to form an interlevel dielectric from the oxide and nitride layers formed thereon. After the first photoresist layer has been removed, a second polysilicon layer is deposited over the dielectric and the first polysilicon layer. A second photoresist layer is selectively deposited over the polysilicon layer such that when anisotropically etched the exposed surfaces are removed to form a polysilicon to polysilicon capacitor, a CMOS transistor, and a bottom plate contact for the capacitor.

In accordance with the present invention, the interlevel dielectric functions as an insulator when used with the capacitor, and as a barrier to protect the underlying polysilicon layer during anisotropic etching. By using the interlevel dielectric as an insulator and barrier, the present invention eliminates one series of steps of masking and etching. The elimination of this series of processing steps also reduces the topography discrepancies and requirement to planarize prior to deposition of metals within the contacts.

The present invention discloses an apparatus which has a polysilicon to polysilicon capacitor connected thereto. Initially, the polysilicon layer is deposited on a substrate. The interlevel dielectric is formed on a portion of the first polysilicon layer. The second polysilicon layer is formed on the first polysilicon layer and the dielectric such that a capacitor and a contact for the capacitor are created.

The present invention presents technical advantages over the prior method of fabricating a polysilicon to polysilicon capacitor used in both a CMOS and a BICMOS. This novel invention has a dielectric layer used as a barrier to protect the underlying polysilicon, and as an insulator for the polysilicon to polysilicon capacitor. It also has the technical advantage of reducing the processing steps associated with fabricating a polysilicon to polysilicon capacitor. Additionally, the present invention is capable of eliminating the planarization step associated with prior art techniques.

Other aspects of the invention and their advantages may be appreciated with reference to the following Detailed Description taken in conjunction with the appended Drawings in which:

FIG. 1 is an enlarged cross-sectional view of a workpiece illustrating the configuration of a first layer of polysilicon having an oxide and a silicon nitride layer connected thereon;

FIG. 2 is an enlarged cross-sectional view of the workpiece as shown in FIG. 1, illustrating the method of forming a dielectric by selectively etching the workpiece;

FIG. 3 is an enlarged cross-sectional view of the workpiece as shown in FIG. 2, illustrating a second layer of polysilicon deposited thereon;

FIG. 4 is an enlarged cross-sectional view of the workpiece as shown in FIG. 3, illustrating the selective deposition of the photoresist prior to etching;

FIG. 5 is an enlarged cross-sectional view of the workpiece as shown in FIG. 4, showing a capacitor, a contact for the capacitor, and a CMOS transistor as formed after etching in accordance with the invention;

FIG. 6 is an enlarged cross-sectional view of the workpiece, illustrating the configuration of a first layer of polysilicon having an oxide and a silicon nitride layer connected thereon;

FIG. 7 is an enlarged cross-sectional view of the workpiece as shown in FIG. 6, illustrating the method of forming a dielectric by selectively etching the workpiece;

FIG. 8 is an enlarged cross-sectional view of the workpiece as shown in FIG. 7, showing the selective deposition of a photoresist which forms a via therein;

FIG. 9 is an enlarged cross-sectional view of the workpiece as shown in FIG. 8, illustrating a second polysilicon layer formed thereon;

FIG. 10 is an enlarged cross-sectional view of the workpiece as shown in FIG. 9, illustrating the selective deposition of the photoresist prior to etching; and

FIG. 11 is an enlarged cross-sectional view of the workpiece as shown in FIG. 10, showing the formation of a capacitor, a contact of the capacitor, and an npn junction as formed in accordance with the present invention.

FIG. 1 is an enlarged cross-sectional view of a workpiece generally designated 10. In its preferred embodiment, workpiece 10 is designed to make a complimentary metal-oxide semiconductor (CMOS). Workpiece 10 generally comprises a substrate 12 which is preferably made of silicon. Substrate 12 is selectively shaped by depositing a photoresist and etching to form a gate region 14.

After the etching process has been performed, an oxide layer 15 is formed over gate region 14, along with a field oxide layer 16 on substrate 12. This oxide isolation formation process is well-known in the art. After field oxide layer 16 has been formed, a first polysilicon layer 18 is formed over field oxide layer 16 and gate region 14. It is preferable that polysilicon layer 18 be deposited to a thickness of approximately 2,000 angstroms. After polysilicon layer 18 has been deposited, an arsenic dopant is implanted into the polysilicon at approximately 1 .times. 10.sup.16 atoms per centimeter squared (atoms/cm.sup.2). This concentration level requires that there be 85 KeV applied to workpiece 10. After polysilicon layer 18 has been properly doped with an arsenic implant dopant, a thin oxide layer 19 is formed over polysilicon layer 18. This polysilicon oxidation process is performed by increasing the temperature of workpiece 10 to approximately 900.degree. C. for twelve (12) minutes, while workpiece 10 is exposed to a mixture of oxygen and nitrogen. By utilizing these parameters, oxide layer 19 is formed having a thickness of approximately 200 angstroms.

Once polysilicon layer 18 has had oxide layer 19 formed thereon, a silicon nitride layer 20 is evenly deposited over the entire surface of oxide layer 19. This silicon nitride deposition is performed in accordance with methods well-known in the art. It is preferable to form layer 20 approximately 250 angstroms thick.

Referring now to FIG. 2, an enlarged cross-sectional view of workpiece 10, illustrating subsequent steps can be seen. A masking layer or photoresist layer 22 is selectively deposited over layer 20, such that when an etchant 24 is applied to workpiece 10, a dielectric layer 26 formed from layers 19 and 20 is formed under the photoresist 22. In its preferred embodiment, photoresist layer 22 is a Novolak.TM. composition. Polysilicon layer 18 is not severely damaged by etchant 24 since etchant 24 has a high selectivity to polysilicon. Etchant 24 is a plasma etchant which etches oxide and nitride with high selectivity to polysilicon. After etching, photoresist layer 22 is then removed.

FIG. 3 illustrates a second polysilicon layer 28 which has been evenly deposited onto first polysilicon layer 18 and over dielectric layer 26. In its preferred embodiment, the second polysilicon layer 28 is deposited to a thickness of approximately 3,000 angstroms. After second polysilicon layer 28 has been deposited, a dopant is diffused into second polysilicon layer 28. It is preferable that a phosphorous (POCl.sub.3) dopant be diffused into second polysilicon layer 28.

FIG. 4 illustrates the selective masking and depositing of a photoresist layer 30 onto second polysilicon layer 28 such that three separate regions are formed thereon.

Referring now to FIG. 5, a final CMOS product of the present invention can be seen. In accordance with the present invention, an anisotropic etchant, with a selectivity to oxide and nitride of approximately 100 to one, is applied to the structure of FIG. 4 to form the embodiment of the present invention. Once the etchant is applied to workpiece 10, a gap 32 is formed in second polysilicon layer 28 to form a polysilicon to polysilicon capacitor 34 and a contact 36 for the bottom plate of capacitor 34. As can be appreciated, dielectric layer 26, which is exposed at gap 32, functions as an insulator for the polysilicon to polysilicon capacitor 34 and as a barrier to protect the underlying polysilicon layer 18 during anisotropic etching. Also, workpiece 10 has a CMOS transistor 38 formed from layers 15, 18, 28 and gate region 14 in accordance with the present invention.

Referring now to FIG. 6, a workpiece 40 is illustrated which will create a BiCMOS structure. Initially, the method of fabricating is quite similar to the steps shown in FIGS. 1-2. A substrate 42 has a field oxide layer 46 formed thereon. Formed in the space between field oxide layer 46 is a gate region 44 over which a thin layer of oxide is formed. A first polysilicon layer 48 is formed over field oxide layer 46 and gate region 44. After polysilicon layer 48 is deposited, an arsenic dopant is implanted into polysilicon 48 at a concentration of approximately 1 .times. 10.sup.16 atoms/cm.sup.2. After the implantation of the arsenic into the polysilicon, an oxidation layer 49 is formed over the polysilicon by increasing the temperature of workpiece 40 to approximately 900.degree. C. for twelve (12) minutes. This oxidation step is performed by exposing workpiece 40 to a mixture of oxygen and nitrogen and the resultant thickness is approximately 200 angstroms. After polysilicon layer 48 has been thermally treated, a silicon nitride layer 50 is deposited to approximately 250 angstroms.

FIG. 7 illustrates the patterning and etching of workpiece 40 which creates a dielectric 56 from layers 49 and 50. Dielectric 56 is formed by having a photoresist 52 selectively deposited onto the exposed surface. After photoresist 52 has been deposited, an appropriate etchant 54 is applied to workpiece 40 which etches oxide and nitride with high selectivity to polysilicon. Photoresist 52 is then removed.

FIG. 8 shows a photoresist layer 58 deposited over polysilicon layer 48 and dielectric 56. Once photoresist 58 has been deposited and patterned, an etchant 62 is applied to workpiece 40 to form a continuous via 60 through polysilicon layer 48 and which exposes gate region 44. Gate region 44 acts as the future emitter contact for the present invention.

Referring now to FIG. 9, workpiece 40 is shown with a second polysilicon layer 64 deposited over workpiece 40 and into contact with gate region 44 through via 60. It is preferable to have polysilicon layer 64 deposited to a thickness of approximately 3,000 angstroms. Once polysilicon layer 64 has been deposited, a phosphorous implant dopant is applied.

Referring to FIG. 10, the selective deposition of three areas a photoresist layer 66 can be seen. These regions will be subsequently used to define portions of the BiCMOS device.

Referring to FIG. 11, the BiCMOS device formed in accordance with the present invention can be seen. This device is shown after an appropriate etchant is applied to workpiece 40 and the photoresist layers 66 are removed. This etchant is preferably an anisotropic etchant which is highly selective to silicon oxide and silicon nitride, and is preferably approximately 100 to 1 in its selectivity. The present invention is shown with a polysilicon to polysilicon capacitor 70 which is separated from a contact 72 by a gap 74. This gap 74 is the exposed region of dielectric 56. In accordance with the present invention, dielectric 56 functions as a barrier during anisotropic etching to protect polysilicon 48 layer thereunder, and as an insulator for capacitor 70. Also, in accordance with the present invention, a npn transistor 68 is formed over gate region 44.

The present invention presents technical advantages over the prior art. Before the present invention, double level polysilicon (DLP) processes were used to make polysilicon to polysilicon capacitors. In particular, the LinEPIC DLP process used a two-mask approach to define a capacitor bottom plate. Initially, the first mask was used to ,etch a frame around the bottom plate without removing the polysilicon diffusion area. A sidewall oxide deposition and etch followed to form a slope surface at the edge of the bottom plate. The purpose of the sidewall oxide was to help prevent polysilicon filament formation when the top plate was defined. After the interlevel dielectric was formed, a second mask was used to protect the bottom plate, while allowing the interlevel and first polysilicon to be removed from all other areas. The second polysilicon deposition, patterning, and etching formed the capacitor top plate and CMOS gates. While this approach helped eliminate polysilicon filaments, it is considerably more complicated and expensive than the present invention. Additionally, a second deposition patterning and etching step is necessary when using this invention. Additionally, since this is a DLP process, the topography problems associated with it require planarization of the entire surface prior to metalizing the contacts.

A second DLP process, the 4/3 linear process, used a single mask approach to define a capacitor bottom plate. The first level of the polysilicon served as both the bottom plate and CMOS gates. After the interlevel dielectric formation, a second polysilicon was deposited, patterned, and etched to form the capacitor top plate. To eliminate filaments from the bottom plate edges and CMOS gate edges, a large overetch was used. If there was a negative slope on the bottom plate edge, filaments may become trapped. Also, since this was a DLP process, the added topography requires additional planarization prior to metalization.

In accordance with the present invention, there is disclosed an improved method for fabricating polysilicon to polysilicon capacitors used for CMOS and BiCMOS devices. This method of manufacturing reduces the steps for manufacturing such devices and minimizes required planarization prior to metalizing the surface of the workpiece.

While the preferred embodiment of the present invention and its advantages have been disclosed in the above Detailed Description, the present invention is not limited thereto, but only by the spirit and scope of its appended claims.