Classifications

• • H—ELECTRICITY
  • H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
  • H10B—ELECTRONIC MEMORY DEVICES
  • H10B43/00—EEPROM devices comprising charge-trapping gate insulators
  • H10B43/30—EEPROM devices comprising charge-trapping gate insulators characterised by the memory core region
• • H—ELECTRICITY
  • H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
  • H10B—ELECTRONIC MEMORY DEVICES
  • H10B43/00—EEPROM devices comprising charge-trapping gate insulators
  • H10B43/40—EEPROM devices comprising charge-trapping gate insulators characterised by the peripheral circuit region
• • H—ELECTRICITY
  • H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
  • H10B—ELECTRONIC MEMORY DEVICES
  • H10B69/00—Erasable-and-programmable ROM [EPROM] devices not provided for in groups H10B41/00 - H10B63/00, e.g. ultraviolet erasable-and-programmable ROM [UVEPROM] devices

Definitions

• the present invention relates to a manufacturing method for NROMs with trench transistors and separate bit lines.
• the smallest non-volatile memory cells are required for the highest integration density in multimedia applications.
• DE 100 39 441 A1 describes a memory cell with a trench transistor which is arranged in a trench formed on an upper side of a semiconductor body.
• An oxide-nitride-oxide layer sequence (O ⁇ O layer) is present as a storage layer between the gate electrode introduced into the trench and the source region adjoining it on the side and the drain region adjoining it on the other side. This sequence of layers is intended for trapping charge carriers (hot electrons) at the source and drain.
• DE 101 29 958 describes a memory cell arrangement in which a further reduction in the dimensions of the memory cells while at the same time keeping the access time for writing and reading sufficiently short is achieved by designing the bit lines to be sufficiently low-resistance.
• separate layers or layer sequences structured as strips are arranged as bit lines on the doped source / drain regions of the individual memory transistors corresponding to the bit lines.
• These layer sequences can comprise doped polysilicon or a metallic layer.
• the metallic layer can be a siliconized metal layer which is produced by the process known under the name "Salicide” as an abbreviation of Seif-aligned-Silicide.
• NROM memory cells are described in the publication by B. Eitan et al .: "NROM: A Novel Localized Trapping, 2-Bit Nonvolatile Memory Cell” in IEEE Electron Device Letters 21, 543 (2000). Because of the special material properties, source / drain voltages of 4 to 5 volts are typically required for such memory cells during programming and erasing. The channel lengths of the memory transistor can therefore not be produced significantly below 200 nm. However, it would be desirable if, despite this channel length of 200 nm, the width of the bit lines could be reduced so that a cell area of less than 5 F 2 is possible.
• bit lines with a sufficiently low electrical resistance are desirable so that multiple connection of the bit lines at intervals within the memory cell array (bitline strapping) could be dispensed with, so that no contact holes for the electrical connection of the bit lines would have to be made between the word lines and as a result the area required between the word lines could be reduced.
• the memory transistor is formed in a trench on an upper side of a semiconductor body or a semiconductor layer.
• the gate electrode is in this Trench introduced and is separated from the laterally adjacent source / drain regions by a storage layer, in particular an ONO layer.
• electrically conductive layers are arranged above the source / drain regions, which preferably comprise several layer layers. It is important to be able to determine the trench depth relative to the depth of the source / drain regions, so that the position at which the lower boundary surface of the source / drain regions adjoins the trench, the so-called junction, can be set precisely. As a result, the channel length between the two-sided junctions is set very precisely according to the specified value.
• this is achieved by introducing an implantation for determining the position of the junctions after the structuring of the bit line layer and before the etching of the trench, or the structuring of the bit line layer after implantation of the source / drain regions using a etching stop layer arranged on the semiconductor material. In this way it is achieved that after the structuring of the low-resistance bit lines, the distance between the top of the semiconductor material into which the trench is etched and the depth of the position of the junctions measured therefrom has exactly the predetermined value.
• the relevant position of the upper side of the semiconductor material results when the bit lines are etched.
• the depth of the position of the junctions is then set by means of a separately introduced implantation of dopant, with which the source / drain regions are finally formed. If the implants for the source / drain regions were introduced before the bit lines were produced, an etching stop layer ensures that the original top side of the semiconductor material remains intact during the structuring of the bit lines, so that In this case, too, the distance of this upper side from the junctions maintains the original value.
• etch stop layer which is initially applied over the entire surface
• a good electrical transition between the bit lines and the regions of source and drain can be produced by partially removing the etch stop layer under the bit lines on both sides and the resulting gaps with an electrical one conductive contact layer, e.g. B. made of conductive doped polysilicon.
• Figures 1.1 to 5.1 show cross sections through intermediate products after various steps of a preferred first embodiment of the method.
• Figures 2.2 to 4.2 show cross sections corresponding to Figures 2.1 to 4.1 for a further embodiment of the method.
• FIGS. 2.3 to 4.3 show cross sections corresponding to FIGS. 2.1 to 4.1 for a further exemplary embodiment of the method.
• a preferred exemplary embodiment of the method begins in accordance with the cross section shown in FIG. 1.1, starting from a semiconductor body or a semiconductor layer applied to a substrate, to which or first a pad oxide / nitride is applied in a manner known per se.
• the semiconductor body 1 preferably has a basic doping which is weakly p-conductive.
• An n + -doped well is formed on the upper side provided with the pad oxide by introducing dopant.
• the pad Oxide layer later serves as etch stop layer 2.
• An oxide is preferred here, although in principle any material is suitable as etch stop layer 2, with respect to which the material of the bit line layers to be applied can be selectively etched.
• STI shallow trench isolation
• Memory cells are provided which run at regular intervals in front of and behind the drawing plane and parallel to the drawing plane with respect to the cross section shown in FIG. 1.1. Trough implants for the formation of CMOS transistors in the control periphery can also be introduced at this stage of the method. These method steps are carried out in a manner known per se, as in the production of conventional memory cell arrays.
• a resist mask 21 is then applied, which has openings in the areas of the bit lines to be produced. Using this resist mask 21, the etching stop layer 2, here the pad oxide, is removed in some areas.
• At least one electrically conductive bit line layer is then applied to the upper side.
• a layer stack is preferably applied here, which first comprises a first bit line layer 3 made of polysilicon, a second bit line layer 4 made of a metal or metal silicide and a hard mask layer 5.
• a thin anti-reflective layer is preferably applied in a manner known per se, which is not shown in the figure.
• the hard mask layer 5 is first structured by means of a photolithography, so that the second bit line layer 4 and the first bit line layer 3 can then be etched back using the hard mask produced in this way.
• the etching of the first bit line layer 3, which here is preferably polysilicon, can be continued a little further to ensure that all remaining portions of the polysilicon have been removed.
• the structure shown in FIG. 2.1 is obtained, in which the first n + implant for forming the n + well 19 in the p-type semiconductor body 1, indicated by the dashed line, is also shown.
• bit line webs laterally it is expedient to cover the bit line webs laterally with a thin oxide layer 6.
• a thin oxide layer 6 This is shown in cross section in FIG. 3.1, in which it is assumed that the first bit line layer 3 is polysilicon and the second bit line layer 4 is a metallic layer, in particular a metal silicide. These layers are therefore oxidized on the surface, so that the thin oxide layer 6 covers the semiconductor material and the flanks of the bit line webs.
• the hard mask layer 5, the z. B. is a nitride is not oxidized or only slightly.
• spacers 7 are then produced on the flanks of the bit line webs, preferably by first depositing a uniform thickness of a nitride layer over the entire surface and then etching this layer back in an anisotropic etching step to such an extent that the spacers shown in FIG. 4.1 7 left.
• the thin oxide layer 6 again serves as an etching stop layer, so that the top of the semiconductor body 1 is not attacked.
• the trenches provided for the memory transistors are then etched out between the spacers 7 produced. This is done using A so-called break-through step, in which, as usual, in several successive etching steps, the thin oxide layer 6 is first removed and then the semiconductor material is etched out in trench form.
• the trench 8 shown in FIG. 4.1 is thus formed. Because of the previously existing etch stop layer 2 or the oxide layer 6, the upper side of the semiconductor body 1 was located at a precisely defined distance from the lower boundary surface of the source / drain regions before the trench etching, as shown in broken lines in FIG. 4.1. Where this interface meets the trench walls are the so-called junctions, which define the beginning and the end of the channel area arranged between them. The channel area is located on the top of the semiconductor material between the junctions in the area of the trench bottom. After the trench 8 has been etched, the walls and the bottom of the trench can be improved by applying a sacrificial layer made of a thin oxide, which is subsequently removed. The storage layer provided can then be applied to the surface of the semiconductor material which has been improved in this way.
• FIG. 5.1 shows in cross section that the storage layer 9 is applied over the entire area to the structure shown in FIG. 4.1.
• the storage layer 9 is preferably an oxide-nitride-oxide layer sequence in which the nitride layer is provided as the storage medium and the two oxide layers are provided as boundary layers for trapping charge carriers.
• the memory layer 9 can be removed photolithographically and replaced by suitable dielectric layers as the gate oxide of the drive transistors.
• a first word line layer 10 made of doped polysilicon is then preferably applied to produce the gate electrodes of the memory transistors. That part of this The first word line layer 10, which fills the respective trench, forms a relevant gate electrode 18.
• STI isolation trenches can have been introduced into the semiconductor material in parallel with the word lines. The trenches are therefore interrupted in the longitudinal direction by the insulating material, in particular silicon dioxide, so that in this last method step the material of the first word line layer 10 is only introduced into the trenches 8 of the memory transistors between the STI insulation trenches.
• the gate electrodes 18 produced in this way are separated from the source / drain regions 15 by the storage layer 9.
• the channel region 17 is located between the junctions 16 directly below the storage layer 9 in the semiconductor material.
• a second word line layer 11, which is applied to the top of the first word line layer 10, is preferably a metal silicide, in particular tungsten silicide (WSi).
• a further hard mask layer 12 applied thereon serves to structure the word lines as strips which run from left to right in FIG. 5.1 within the plane of the drawing. to
• FIG. 2.2 A cross section corresponding to the method step in FIG. 2.1 is shown in FIG. 2.2. It is here the n + tub 19 for training the
• the structuring of the bit line strips which here likewise comprise a first bit line layer 3 (preferably conductively doped polysilicon), a second bit line view 4 (preferably tungsten silicide) and a hard mask layer 5, is etched into the semiconductor material. Is between the bit line strips Therefore, the top of the semiconductor body 1 is lowered accordingly, so that here the distance between the lower interface of the n + well and the top of the semiconductor body 1 is reduced. In order to safely separate the bit line strips from one another, the etching process is continued here until all material of the first bit line layer 3 has been removed.
• the n + well 19 is initially formed here only with a small depth, which but is sufficient to achieve a good electrical transition between the bit line strips and the semiconductor material underneath. Only after the bit line webs have been etched does the actual n + doping take place, with which the regions of source and drain are produced and the positions of the junctions are determined.
• the further source / drain implantation is shown in cross section in FIG. 3.2 with the doped n + region 20.
• the somewhat flatter profile of the dopant concentration under the bit line webs is indicated by the curved lower dashed line. It can be seen from this that the further tub implant 20 was only inserted after the bit line webs had been produced.
• the implantation dose is set here in such a way that the lower boundary surface of the further tub implant 20 has an intended distance from the upper side of the semiconductor material between the bit line webs. This is then followed in the manner described above to produce a thin oxide layer 6 which covers the flanks of the bit line webs.
• the cross section shown in FIG. 4.2 corresponds to the cross section of FIG. 4.1 after the production of the spacers 7 and the etching of the trench 8.
• the position of the junctions is determined by the position of the lower boundary surface of the further tub implant on the walls of the trench 8, is located in the intended distance from the top of the semiconductor body in the region between the bit line webs, so that the etching depth is also here when the trench 8 is etched can be set so precisely that the intended channel length is produced.
• Another exemplary embodiment of the method is based on an etching stop layer 2 covering the entire surface.
• the bit line layers are on the etch stop layer 2, z. B. the pad oxide layer applied.
• the cross section which is shown in Figure 2.3, shows the arrangement after the etching of the bit line webs. It is shown here that even when an etching stop layer 2 is used, the implantation of the dopant, which is provided for the regions of source and drain, can be carried out in two steps before the bit line layers are applied and after the bit line layers are applied. An n + tub 19 and a further tub implant 20 are therefore also shown here.
• the etching stop layer 2 is present over the entire area, there is initially only insufficient electrical contact between the n + well 19 and the first bit line layer 3 (preferably conductively doped polysilicon). The etch stop layer 2 is therefore removed, so that only a small proportion of the etch stop layer 2 remains below the bit line webs.
• a contact layer 13 is applied over the entire surface, which is preferably a thin, electrically conductive polysilicon layer. With this contact layer 13, the spaces on both sides between the bit line strip and the semiconductor body 1 are filled. In this way, there is a good electrical transition between the bit line webs and the semiconductor material of the n + well 19. Die remaining portions of the contact layer 13 on and between the bit line webs are removed.
• FIG. 4.3 the structure achieved with this exemplary embodiment of the method is shown in cross section in accordance with the method step of FIG. 4.1. Below the bit line webs there is a remaining portion of the etch stop layer 2 and the remaining portions 14 of the contact layer 13. The structure shown otherwise corresponds to the structure according to FIG. 4.1, the same reference numerals designating the same parts.

Landscapes

• Semiconductor Memories (AREA)
• Electrodes Of Semiconductors (AREA)
• Non-Volatile Memory (AREA)

Applications Claiming Priority (3)

Publications (1)

Family

ID=29718894

Family Applications (1)

Country Status (8)

Families Citing this family (8)

Family Cites Families (9)

• 2002
  • 2002-06-07 DE DE10225410A patent/DE10225410A1/de not_active Ceased
• 2003
  • 2003-05-13 TW TW092113006A patent/TWI227937B/zh not_active IP Right Cessation
  • 2003-05-15 KR KR1020047019839A patent/KR100675195B1/ko not_active IP Right Cessation
  • 2003-05-15 CN CNB038130033A patent/CN100337324C/zh not_active Expired - Fee Related
  • 2003-05-15 JP JP2004512201A patent/JP2005531920A/ja active Pending
  • 2003-05-15 WO PCT/DE2003/001573 patent/WO2003105231A1/de active Application Filing
  • 2003-05-15 EP EP03735305A patent/EP1512179A1/de not_active Withdrawn
• 2004
  • 2004-12-07 US US11/006,049 patent/US7205195B2/en not_active Expired - Fee Related

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events