Classifications

• • H—ELECTRICITY
  • H01—ELECTRIC ELEMENTS
  • H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
  • H01L27/00—Devices consisting of a plurality of semiconductor or other solid-state components formed in or on a common substrate
  • H01L27/02—Devices consisting of a plurality of semiconductor or other solid-state components formed in or on a common substrate including semiconductor components specially adapted for rectifying, oscillating, amplifying or switching and having potential barriers; including integrated passive circuit elements having potential barriers
  • H01L27/04—Devices consisting of a plurality of semiconductor or other solid-state components formed in or on a common substrate including semiconductor components specially adapted for rectifying, oscillating, amplifying or switching and having potential barriers; including integrated passive circuit elements having potential barriers the substrate being a semiconductor body
  • H01L27/08—Devices consisting of a plurality of semiconductor or other solid-state components formed in or on a common substrate including semiconductor components specially adapted for rectifying, oscillating, amplifying or switching and having potential barriers; including integrated passive circuit elements having potential barriers the substrate being a semiconductor body including only semiconductor components of a single kind
  • H01L27/085—Devices consisting of a plurality of semiconductor or other solid-state components formed in or on a common substrate including semiconductor components specially adapted for rectifying, oscillating, amplifying or switching and having potential barriers; including integrated passive circuit elements having potential barriers the substrate being a semiconductor body including only semiconductor components of a single kind including field-effect components only
  • H01L27/088—Devices consisting of a plurality of semiconductor or other solid-state components formed in or on a common substrate including semiconductor components specially adapted for rectifying, oscillating, amplifying or switching and having potential barriers; including integrated passive circuit elements having potential barriers the substrate being a semiconductor body including only semiconductor components of a single kind including field-effect components only the components being field-effect transistors with insulated gate
  • H01L27/092—Devices consisting of a plurality of semiconductor or other solid-state components formed in or on a common substrate including semiconductor components specially adapted for rectifying, oscillating, amplifying or switching and having potential barriers; including integrated passive circuit elements having potential barriers the substrate being a semiconductor body including only semiconductor components of a single kind including field-effect components only the components being field-effect transistors with insulated gate complementary MIS field-effect transistors
  • H01L27/0928—Devices consisting of a plurality of semiconductor or other solid-state components formed in or on a common substrate including semiconductor components specially adapted for rectifying, oscillating, amplifying or switching and having potential barriers; including integrated passive circuit elements having potential barriers the substrate being a semiconductor body including only semiconductor components of a single kind including field-effect components only the components being field-effect transistors with insulated gate complementary MIS field-effect transistors comprising both N- and P- wells in the substrate, e.g. twin-tub
• • H—ELECTRICITY
  • H01—ELECTRIC ELEMENTS
  • H01L—SEMICONDUCTOR DEVICES NOT COVERED BY CLASS H10
  • H01L21/00—Processes or apparatus adapted for the manufacture or treatment of semiconductor or solid state devices or of parts thereof
  • H01L21/70—Manufacture or treatment of devices consisting of a plurality of solid state components formed in or on a common substrate or of parts thereof; Manufacture of integrated circuit devices or of parts thereof
  • H01L21/77—Manufacture or treatment of devices consisting of a plurality of solid state components or integrated circuits formed in, or on, a common substrate
  • H01L21/78—Manufacture or treatment of devices consisting of a plurality of solid state components or integrated circuits formed in, or on, a common substrate with subsequent division of the substrate into plural individual devices
  • H01L21/82—Manufacture or treatment of devices consisting of a plurality of solid state components or integrated circuits formed in, or on, a common substrate with subsequent division of the substrate into plural individual devices to produce devices, e.g. integrated circuits, each consisting of a plurality of components
  • H01L21/822—Manufacture or treatment of devices consisting of a plurality of solid state components or integrated circuits formed in, or on, a common substrate with subsequent division of the substrate into plural individual devices to produce devices, e.g. integrated circuits, each consisting of a plurality of components the substrate being a semiconductor, using silicon technology
  • H01L21/8232—Field-effect technology
  • H01L21/8234—MIS technology, i.e. integration processes of field effect transistors of the conductor-insulator-semiconductor type
  • H01L21/8238—Complementary field-effect transistors, e.g. CMOS
  • H01L21/823892—Complementary field-effect transistors, e.g. CMOS with a particular manufacturing method of the wells or tubs, e.g. twin tubs, high energy well implants, buried implanted layers for lateral isolation [BILLI]

Definitions

• the invention is related to integrated circuit structures and methods of fabricating them, and is particularly related to integrated circuits fabricated upon a "twin well” or “twin tub” structure.
• CMOS complementary metal-oxide-semiconductor
• a p well may be implanted or diffused into an n substrate at a concentration high enough to overco pensate the n substrate and give good control over the resultant p-type doping.
• the doping level in the p well is five to ten times higher than that in the n-type substrate to attain this control.
• this high p-tub doping causes undesirable effects in the n-channel transistor such as increased back-gate bias effects and increased source/drain to p-well capacitance.
• n-channel transistors An alternate approach is to employ an n well to form the p-channel transistors.
• the n-channel device is formed in the p-type substrate and the n-well is compatible with standard NMOS processing.
• the n well tends to overcompensate the p substrate and the p-channel devices suffer from excessive doping effects.
• CMOS complementary metal-oxide-semiconductor
• This "twin well" CMOS approach permits the doping profiles in each well region to be independently designed, so that neither type of device suffers from excessive doping effects.
• Such an approach has been used on lightly doped n-type and p-type substrates.
• FIG. 1 Shown in FIG. 1 is a CMOS structure having a lightly doped n substrate 10 within which are formed a p well 12 and an n well 14 having an n source 16 and a p source 18 respectively.
• the devices are isolated from each other via field oxide regions 20 and interconnection is provided by polycrystalline silicon (polysilicon) layer 22.
• polycrystalline silicon (polysilicon) layer 22 It has been discovered with CMOS devices that when the wells are driven in, appreciable lateral diffusion occurs and interdiffusion of the wells forms a depletion region 24 under the isolation regions 20 and that if the depletion region 24 is wide enough or the substrate is too lightly doped, the depletion regions of the sources 16 and 18 and drains of adjacent transistors may punchthrough to each other and form an undesirable parasitic transistor. This effect prevents the devices from being placed close to each other and requires the use of additional chip area.
• channel stop regions or "chan stops” 26 and 27 which are more highly doped regions between the active device sources 16 and 18, as seen in FIG. 2.
• the p-channel stop 26 and n-channel stop 27 effectively narrow the area of the depletion region 24 and permit the devices to be placed closer together.
• One process which employs a lift-off technique to help form combined channel stop/well regions is described by J. Y. Chen in "Quadruple-Well CMOS — A VLSI Technology," Tech. Digest, IEEE IEDM, 1982, pp. 791-792. See also U.S. Pat. No. 4,558,508 issued to Kinney, et al. on December 17, 1985.
• channel stops are often misaligned as seen in the case of misaligned channel stops 28 and 29 of FIG. 3. Because the channel stops 28 and 29 in a twin well or single well structure must be individually placed at the edge of each well, the potential for misalignment is great and the depletion region 24 is not narrowed as much as desired, causing a greater possibility of punchthrough. Using a two mask channel stop process has even greater potential for misalignment.
• FIG. 4 illustrates a more typical process of forming twin well regions and an isolation oxide and channel stop between the regions.
• the future n well region of the substrate 30 is receiving a phosphorus implant, represented with the Xs, through pad oxide 32 while the future p well region is shielded by nitride pattern 34.
• the structure in FIG. 4B is produced by driving in the phosphorus to produce n well 36, growing thick oxide layer 38 and stripping the nitride pattern 34.
• a boron implant is then conducted as illustrated by the dots in FIG. 4B, with n well 36 being protected by the thick oxide layer 38.
• the thick oxide layer 38 is stripped and a new oxide layer 40 is grown during which the p well 42 is driven in as seen in FIG. 4C.
• Oxide layer 40 is stripped, a new pad oxide 41 is grown and second nitride pattern 44 is formed, after which a second boron implant is conducted in the region to be the channel stop as defined by the nitride pattern 44.
• Photoresist pattern 46 is formed to permit the second phosphorus implant into the n well channel stop region as seen in FIG. 4D.
• the photoresist pattern 46 is removed and the impurities are driven in slightly to form p chan stop 48 and n chan stop 50 during or after which field oxide isolation region 52 is grown.
• the second nitride pattern 44 is then removed to give the structure seen in FIG. 4E.
• the process described relative to FIG. 4 has a number of disadvantages, not the least of which is the fact that two mask steps are used, with their attendant risks of mis ⁇ alignment and the disadvantages caused by channel stop mis ⁇ alignment as discussed earlier.
• the depletion region is not as narrow as would be desired and the devices may not be placed as closely together as is desired.
• FIG. 5 An alternative process has been proposed by Hillenius and Parrillo in U.S. Pat. No. 4,554,726 (incorporated by reference herein) , as illustrated in FIG. 5 which uses a single mask and permits the use of a high/low doping profile for the wells.
• the process begins by the future n well region of the substrate 54 receiving both an arsenic, as represented by the triangles in FIG. 5A, and a phosphorus implant, as represented by the Xs, through a blanket pad oxide 56.
• the future p well region is shielded by nitride pattern 58, which is the only pattern needed for the graded well formation process.
• a "graded well” is meant a well with a high-low doping profile; the relatively higher doping concentration being nearer the surface.
• a thick oxide layer 60 is formed in the n-well region and the nitride pattern 58 is stripped.
• the slow diffusing arsenic does not diffuse as far into the silicon with respect to the relatively fast diffusing phosphorus to give n-well 62 and highly doped surface region 64.
• these regions 62 and 64 are driven-in simultaneously, it is difficult to achieve a shallow depth of highly doped arsenic surface region 64 if the n-well 62 is driven-in to its desired depth.
• the highly doped arsenic region 64 is not confined to surface proximity as would be desired. P-channel devices subsequently formed in n-well 62 perform extremely poorly.
• the next step is the first boron p-well implant, as symbolized by the dots.
• This implant is driven-in to form p-well 66, after which a second boron implant for the highly concentrated surface region is conducted as shown in FIG. 5C. Since the drive-in of the second boron implant is performed as a separate step, the depth of highly doped boron surface region 68 may be independently determined and may be actually restricted to just below the surface unlike highly doped arsenic region 64, which must be jointly driven-in with the phosphorus. Stripping the thick oxide layer 60, forming a uniform dielectric oxide layer 70 and depositing a nitride pattern 72 produces the structure seen in FIG. 5D.
• Forming field oxide region 74 by high pressure oxidation or other means gives the finished isolation boundary seen in FIG. 5E. While the n-channel devices subsequently made in the p-well 66 of this structure perform well, the p-channel devices in n-well 62 suffer from excessive doping effects such as source/drain to well capacitance and body effect due to the fact that heavily doped arsenic layer 64 is too deep.
• the channel stops produced by these implants would permit close spacing between adjacent NMOS and PMOS devices due to a low parasitic leakage, since the depletion region beneath the interface field oxide would be narrow.
• Another object of the present invention is to provide a twin well structure and process for making the same in which relatively higher doped (e.g. 5 x 10 16 atoms/cm 3 ) surface layers are present in each well to give a graded well structure. It is another object of the present invention to provide a twin well structure and process therefor in which the depth of the relatively higher doped surface layers in each well may be shallow and precisely controlled.
• relatively higher doped e.g. 5 x 10 16 atoms/cm 3
• Still another object of the present invention is to provide a twin well structure and process therefor in which relatively higher doped surface regions may be introduced into each well by using only one mask pattern.
• an integrated circuit built on a semiconductor substrate having a planar surface, which has formed therein a plurality of deep, low-doped wells of a first conductivity type in selected areas of the semiconductor substrate surface and a plurality of deep, low-doped wells of a second conductivity type in selected areas of the substrate different from those of the first conductivity type.
• shallow, relatively higher-doped layers of a first conductivity type are present in selected ones of the plurality of deep, low-doped wells of the first conductivity type; and shallow, relatively higher-doped layers of a second conductivity type are present in selected ones of the plurality of deep, low-doped wells of the second conductivity type.
• FIG. 1 is a cross-sectional illustration of a prior art boundary between two wells of different conductivity types
• FIG. 2 is a cross-sectional illustration of a prior art boundary between two wells of different conductivity types which have precisely aligned channel stop regions present at the boundary;
• FIG. 5 is a cross-sectional illustration of a prior art process for forming a high-low well doping profile in two adjacent wells of different conductivity types using one mask
• FIG. 6 is a cross-sectional illustration of the method of this invention demonstrating how high-low well doping profiles in two adjacent wells of different conductivity types may be formed with independent and precise depth control of each high-low well portion, and one mask.
• a semiconductor substrate 76 is first provided, upon which a blanket pad oxide 78 is formed by any of the well-known techniques.
• a dopant of one type is next blanket implanted into the substrate 76 through pad oxide 78.
• this first implant is a shallow boron implant to form the eventual p-well, although it is conceivable that the n-well implant could be conducted first and the dopant conductivity types throughout this description would be reversed. However, in some cases this may not be desired.
• a layer of silicon nitride 80 is formed and patterned using photoresist layer 82 to give the structure shown in FIG. 6B, wherein the exposed region overlies the future n-well and the silicon nitride pattern 80 overlies the future p-well.
• the next part of the procedure involves removing the boron impurity from the future n-well region and introducing phosphorus impurity therein to enable the production of the n-well.
• the first option is illustrated in FIGs. 6C-1 and 6D-1 and involves first stripping the photoresist pattern 82, and then via high-pressure oxidation, or other means, forming thick silicon oxide region 84.
• the region of silicon substrate 76 in the future n-well having the boron doping is absorbed into this thick silicon oxide region 84 which is next physically removed.
• a screen oxide layer 86 is next formed through which phosphorus is ion implanted as represented by the Xs in FIG. 6D-1.
• the p-well region is shielded from the phosphorus by nitride pattern 80.
• the p-well 88 and n-well 90 are driven in simultaneously to give the structure of FIG. 6E.
• the second option involves implanting the phosphorus impurity at a high energy through pad oxide layer 78 to a depth much deeper than that of the boron impurity, if a high pressure oxidation is employed subsequently, shown in FIG. 6C-2. If a different oxidation technique is used, the implant can be conducted at a lower energy and the implant depth need not be deep since the phosphorus piles up in the silicon and the boron tends to migrate to the oxide. Now the resist layer 82 is stripped and the thick oxide region 84 is grown to absorb that portion of the silicon 76 which is boron-doped, as seen in FIG. 6D-2. Finally, the oxide layer 84 is stripped and the p-well 88 and n-well 90 are driven in simultaneously. Screen oxide layer 86 is formed last, as shown in FIG. 6E.
• the portion of the silicon substrate 76 having the boron impurity is etched away to give the structure of FIG. 6C-3.
• Screen oxide 86 is formed and phosphorus is implanted as shown in FIG. 6D-3.
• the simultaneous drive-in step to form the p-well 88 and the n-well 90 is conducted as before to give the structure of FIG. 6E.
• This third option has the advantages of being sure that nearly all of the boron is removed from the future n-well area and not requiring a thick oxide formation step. On the other hand, this option exposes the silicon to a reactive ion etch (RIE) , which may not be desirable, and requires precise control of the silicon etch.
• RIE reactive ion etch
• the p-well 88 and n-well 90 have been formed, there remains to be created the relatively high/low well or graded well doping profile with the p-well 88 and the n-well 90 as the well regions having a low doping concentrations.
• arsenic represented by the triangles, is ion implanted as a channel stop/punchthrough layer in a relatively high concentration but at low enough energy so that it does not penetrate the nitride layer 80 or penetrate deeply into the n-well 90. This implant could also be phosphorus. Note that p-well region is once again shielded by nitride pattern 80.
• a drive-in step forms shallow, relatively highly-doped n-layer 92, after which the n-well region 90 is oxidized to form thick oxide layer 94, or these two steps could be performed simultaneously.
• the nitride pattern 80 is stripped and a second boron implant is conducted at relatively high concentration and low energy to a shallow depth as seen in FIG. 6F. This second boron implant is driven-in to give a shallow, highly-doped p-layer 96 which serves as a channel stop or punchthrough prevention layer.
• the thick oxide layer 94 and thin oxide 78 are stripped and isolation regions are formed in one embodiment by using high pressure oxidation according to methods well-known in the art.
• a uniform oxidation process in which there is a direct pattern etch of the active regions will produce the structure seen in FIG. 6G which has island isolation regions.
• LOCOS local oxidation of silicon
• FIG. 6H which has thin gate oxide regions 100 and thicker isolation regions 102.
• the shallow, highly- doped layers 92 and 96 will be divided into areas of somewhat different doping concentrations.
• the surrounding p-layer 96 and active n-area 106 are expected to have a some ⁇ what lower dopant concentration than the active p-area 104 and n-layer 92, respectively.
• this new CMOS process produces self-aligned twin wells 88 and 90 as well as more relatively heavily doped surface layers 96 and 92 in each well using only one lithographic mask, not counting the mask used in the isolation procedure.
• the two types of wells and surface layers are essentially independent of each other, so little compensation is required of an n-layer 92 over a p-layer 96, for example.
• the more heavily doped surface layers 92 and 96 serve simultaneously as channel-stop layers, punchthrough prevention implants and/or highly doped surface layers to produce the desired transistor threshold voltages.
• high-pressure oxidation using a short thermal cycle is recommended for use; however, the major benefits may also be obtained using atmospheric oxidation.
• the method of this invention achieves the purposes of providing high/low impurity profiles in each well to optimize the NMOS and PMOS active transistors; providing close NMOS to PMOS transistor spacing; avoiding a channel-stop mask level and avoiding a threshold adjustment/punchthrough mask level.
• the depth of the shallow, heavily-doped surface layers may be independently achieved, which is an important advantage over the process of U.S. Pat. No. 4,554,726 illustrated in FIG. 5.
• the pad oxide grown before FIG. 6A may be approximately 300 Angstroms of silicon dioxide and the first boron implant conducted there may be approximately 10 keV, for example.
• the silicon nitride formed and patterned in FIG. 6B could be about 1200 Angstroms.
• the thick oxide region 84 is preferably grown using a high pressure, low temperature step to cause the boron in the future n-well area to be removed. Since boron will diffuse little during this oxidation, an oxide region 84 of about 2000 Angstroms should be sufficient in most cases.
• the phosphorus implant seen in FIG. 6D-1 could be done at an energy of 30 keV, for example, whereas the one conducted under the second option in FIG. 6C-2 should be done at a higher energy, say around 200 keV.
• the twin wells 88 and 90 may be driven-in to give the structure of FIG. 6E using a mildly oxidizing ambient, to prevent too thick of an oxide growth over the n-well 90, for example to drive the wells to approximately 2.5 um deep.
• the thickness of thick oxide layer 94 may range around 2000 to 3000 Angstroms or more.
• p-layer 96 may be driven in a bit before field oxidation.
• the shallow, relatively higher doped surface layer can be made shallower than is achievable by the simultaneous graded well drive-in of U.S. Pat. No. 4,554,726, due to the independent control of the present process.
• a possible high-low doping profile may be 3 x 10 16 to 5 x 10 16 atoms/cm 2 in the shallow, highly-doped layers and about 5 x 10 15 atoms/cm 2 in the deep, low-doped wells.

Landscapes

• Engineering & Computer Science (AREA)
• Power Engineering (AREA)
• Physics & Mathematics (AREA)
• Condensed Matter Physics & Semiconductors (AREA)
• General Physics & Mathematics (AREA)
• Computer Hardware Design (AREA)
• Microelectronics & Electronic Packaging (AREA)
• Manufacturing & Machinery (AREA)
• Metal-Oxide And Bipolar Metal-Oxide Semiconductor Integrated Circuits (AREA)

Applications Claiming Priority (2)

Publications (1)

Family

ID=25284733

Family Applications (1)

Country Status (2)

Families Citing this family (4)

Family Cites Families (8)

• 1986
  • 1986-11-21 EP EP19860907209 patent/EP0260271A1/de not_active Withdrawn
  • 1986-11-21 WO PCT/US1986/002503 patent/WO1987005443A1/en unknown

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events