EP1145320A1 - Dispositif de cellule dram et son procede de fabrication - Google Patents

Dispositif de cellule dram et son procede de fabrication

Info

Publication number
EP1145320A1
EP1145320A1 EP99967861A EP99967861A EP1145320A1 EP 1145320 A1 EP1145320 A1 EP 1145320A1 EP 99967861 A EP99967861 A EP 99967861A EP 99967861 A EP99967861 A EP 99967861A EP 1145320 A1 EP1145320 A1 EP 1145320A1
Authority
EP
European Patent Office
Prior art keywords
depression
recess
produced
memory cell
substrate
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Withdrawn
Application number
EP99967861A
Other languages
German (de)
English (en)
Inventor
Bernd Goebel
Emmerich Bertagnolli
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Infineon Technologies AG
Original Assignee
Infineon Technologies AG
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Infineon Technologies AG filed Critical Infineon Technologies AG
Publication of EP1145320A1 publication Critical patent/EP1145320A1/fr
Withdrawn legal-status Critical Current

Links

Classifications

    • H—ELECTRICITY
    • H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10B—ELECTRONIC MEMORY DEVICES
    • H10B12/00—Dynamic random access memory [DRAM] devices
    • H—ELECTRICITY
    • H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10B—ELECTRONIC MEMORY DEVICES
    • H10B12/00—Dynamic random access memory [DRAM] devices
    • H10B12/01—Manufacture or treatment
    • H10B12/02—Manufacture or treatment for one transistor one-capacitor [1T-1C] memory cells
    • H10B12/03—Making the capacitor or connections thereto
    • H10B12/038—Making the capacitor or connections thereto the capacitor being in a trench in the substrate
    • H10B12/0383—Making the capacitor or connections thereto the capacitor being in a trench in the substrate wherein the transistor is vertical
    • H—ELECTRICITY
    • H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
    • H10B—ELECTRONIC MEMORY DEVICES
    • H10B12/00—Dynamic random access memory [DRAM] devices
    • H10B12/30—DRAM devices comprising one-transistor - one-capacitor [1T-1C] memory cells
    • H10B12/39—DRAM devices comprising one-transistor - one-capacitor [1T-1C] memory cells the capacitor and the transistor being in a same trench
    • H10B12/395—DRAM devices comprising one-transistor - one-capacitor [1T-1C] memory cells the capacitor and the transistor being in a same trench the transistor being vertical

Definitions

  • the invention relates to a DRAM cell arrangement, i. H. a memory cell arrangement with dynamic random access, and a method for their production.
  • Memory cell used which comprises a transistor and a capacitor.
  • the information of the memory cell is stored in the form of a charge on the capacitor.
  • the capacitor is connected to the transistor, so that when the transistor is driven via a word line, the charge on the capacitor can be read out via a bit line.
  • the general aim is to produce a DRAM cell arrangement that has a high packing density.
  • US Pat. No. 5,208,657 describes a DRAM cell arrangement in which a memory cell comprises a transistor and a capacitor. To increase the packing density, the transistor is arranged on four flanks of a depression in which a storage node of the capacitor is arranged.
  • the recess is arranged under an area in which a word line and a bit line of the memory cell cross.
  • the transistor is designed as a vertical transistor and its gate electrode is arranged in the recess above the storage node. The space requirement of the
  • Memory cell is at least 6.25 F, where F is the minimum structure size that can be produced in the technology used.
  • the invention is based on the problem of specifying a DRAM cell arrangement whose memory cells have transistors and capacitors in comparison with the prior art can have improved electrical properties without having to reduce the packing density of the DRAM cell arrangement.
  • a method for producing such a DRAM cell arrangement is also to be specified.
  • a DRAM cell arrangement which has memory cells which each comprise at least one vertical transistor and one capacitor.
  • An upper source / drain region, a channel region and a lower source / drain region of the transistor are arranged one above the other and each adjoin both a first flank of a first depression and a second depression. At least part of the first flank of the first depression is provided with a capacitor dielectric of the capacitor, which has a cutout in the region of the lower source / drain region.
  • a storage node of the capacitor is arranged in the first depression and adjoins the lower source / drain region in the recess.
  • a gate electrode of the transistor is arranged in the second depression.
  • the memory cells are connected to word lines and to bit lines which run transversely to the word lines.
  • the problem is also solved by a method for producing a DRAM cell arrangement, in which memory cells are produced, each having a vertical transistor and a capacitor.
  • a lower source / drain region, a channel region and an upper source / drain region are produced in a substrate, so that they are arranged one above the other.
  • a first depression is produced in the substrate and adjoins the lower source / drain region, the channel region and the upper source / drain region with a first flank.
  • the first recess is provided with a capacitor dielectric of the capacitor.
  • the capacitor dielectric is on the first flank of the first Provide a recess in the area of the lower source / drain region.
  • a storage node of the capacitor is created in the first recess, which is adjacent to the lower source / drain region in the recess.
  • a second depression is created which is adjacent to the upper source / drain region, the channel region and the lower source / drain region.
  • a gate electrode of the transistor is produced in the second depression. Word lines and bit lines that run across the word lines are generated and connected to the memory cells.
  • the second depression of the memory cell lies outside the first depression of the memory cell.
  • the DRAM cell arrangement can have a high packing density since the transistor is designed as a vertical transistor, the storage node is arranged in a recess and a connection between the storage node and the lower source / drain region is made possible by a recess in the capacitor dielectric, which does not requires additional space.
  • the quality of a boundary layer of the channel region, at which a gate dielectric of the transistor is generated, generally has a great influence on electrical ones
  • the transistor can be manufactured with improved electrical properties compared to US Pat. No. 5,208,657, since different recesses are provided for the capacitor and for the transistor, so that the boundary layer of the channel region can be spared process steps for producing the first recess.
  • the provision of two different depressions also offers the advantage that the geometry of the boundary layer of the channel region is independent of one geometry an area on which the capacitor dielectric is produced.
  • the boundary layer of the channel region is preferably flat so that it has a defined orientation with respect to the crystal lattice of the substrate so that the gate dielectric can grow homogeneously.
  • the surface on which the capacitor dielectric is produced is preferably curved, so that the capacitor dielectric has no edges on which field distortions can lead to leakage currents. Both the transistor and the capacitor can have particularly good electrical properties.
  • a horizontal cross section of the first depression is, for example, circular or elliptical.
  • first depressions and the second depressions of the memory cells are arranged in such a way that the second depression of a first of the memory cells adjoins the storage node which is arranged in the first depression of a second of the memory cells.
  • the memory cells directly adjoin one another or overlap.
  • the storage node is at least initially generated in such a way that it also adjoins the substrate at least on a second flank of the first depression opposite the first flank of the first depression in the region of a further recess.
  • the capacitor dielectric is deposited substantially conformally, for example after the first depression has been produced, so that surfaces of the first depression are covered without the first depression being filled in the process.
  • the first recess is then filled with conductive material.
  • the conductive material is etched back to a lower level.
  • exposed parts of the Capacitor dielectric removed.
  • the first depression is filled again by separating further conductive material.
  • the conductive material is then etched back to an upper height that is above the lower height. In this way, the recess of the capacitor dielectric between the lower height and the upper height is created not only on the first flank of the first depression, but in particular also the further recess of the capacitor dielectric is formed on the second flank of the first depression. If the first depression adjoins the substrate with further flanks between the upper height and the lower height, then recesses are also produced on these flanks.
  • the storage node is produced from the conductive material.
  • the conductive material is, for example, doped silicon.
  • a further possibility of producing the storage node is to fill the first depression with conductive material after the capacitor dielectric has been separated off and to scratch the conductive material back up to the upper level.
  • Exposed parts of the capacitor dielectric and parts of the capacitor dielectric which are arranged between the upper height and the lower height are subsequently removed by isotropic etching.
  • a further gap between the conductive material and the substrate is filled by depositing and scratching off further conductive material.
  • Amorphous doped silicon is particularly suitable as a further conductive material, since it can be deposited conformally and does not cause any defects in the adjacent substrate.
  • a tempering step is carried out in which dopant diffuses from the storage node m at least in the region of the recess.
  • the lower source / drain region is produced by structuring a doped layer of the substrate.
  • a middle layer doped by a first conductivity type can be arranged between two layers doped by a second conductivity type opposite to the first conductivity type.
  • the layers can be generated by in situ doped epitaxy or by implantation.
  • the channel region is produced from the middle layer and the upper source / drain region and the lower source / drain region are produced from the other layers. This has the advantage that a channel length of the transistor can be set precisely, since the process inaccuracy with regard to the thickness of the layers is small, particularly in the case of epitaxy. If an implantation is carried out, this can also be done after the recesses have been created.
  • a low-doped part of the lower source / drain region is produced by structuring the doped layer, and a high-doped part of the lower source / drain region is produced by out-diffusion of dopant from the storage node.
  • the low doped part surrounds the highly doped
  • Such a DRAM cell arrangement has soft p-n junctions and consequently fewer leakage currents, since only the low-doped part of the lower source / drain region adjoins the channel region and the substrate.
  • the highly doped part is adjacent to the channel region but not to the substrate or to the substrate but not adjacent to the channel region.
  • the storage node does not adjoin the substrate on the second flank of the first depression. This enables a distance between each other to be reduced adjacent first depressions without leakage currents occurring between the associated storage nodes.
  • the storage node is first produced in such a way that it also adjoins the substrate on the second flank of the first depression.
  • the second recess is created so that its bottom is lower than a lower edge of the recess.
  • the second depression separates the memory node of the second memory cell from the substrate. It is consequently possible to dispense with the creation of a mask to prevent the further cutout, at the same time achieving a high packing density.
  • the lower source / drain region is produced in the form of a layer which first extends from the first depression of the first memory cell to the first depression of the second Storage cell extends.
  • the lower source / drain region is structured such that it no longer adjoins the first depression of the second memory cell and thus no longer abuts the storage node of the second memory cell.
  • an upper surface of the storage node lies in the region of the recess, for example at its upper edge, and an insulating structure is arranged in the first depression on the storage node. Since the storage node does not reach higher, for example not to a surface of the substrate, a capacitance between the storage node and the upper source / drain region or the channel region or the lower source / drain region of the transistor is avoided.
  • the insulating structure enables an interface between the storage node of the second memory cell and the second recess of the first memory cell to be reduced, so that a capacitance between the gate electrode of the first memory cell or a word line, which is partially in the second
  • the second depression is provided with a gate dielectric before the gate electrode is produced, and otherwise no further, thicker, capacitance-reducing insulating structure is provided in larger parts of the interface in the second depression.
  • the second depression of the first memory cell is laterally offset with respect to the first depression of the second memory cell, so that the second depression of the first memory cell is arranged partially in the first depression of the second memory cell and partially in the substrate.
  • the width of the insulating structure is at least large enough to prevent the transistor of the second memory cell from being activated by the gate electrode of the first memory cell and / or by the word line, which is partially arranged in the second recess of the first memory cell.
  • Structure size that can be produced by technology.
  • a distance between the first depressions can be F.
  • a distance between the second well and the first well of the same memory cell may be less than F.
  • At least the channel area and the lower one can be used first
  • Source / drain region of the transistor of the first memory cell are generated so that they are adjacent to the first recess of the first memory cell and to the first recess of the second memory cell before generating the second recess of the first memory cell.
  • a mask is produced which is arranged over the first flanks of the first depressions and does not cover regions over the second flanks of the first depressions.
  • the second depression is produced with the aid of the mask, with at least both the substrate and the conductive material being etched.
  • the conductive material is patterned through the second recess so that the storage node is created. It is within the scope of the invention if the upper source / drain region also adjoins the first depression of the first memory cell and the first depression of the second memory cell before the second depression is produced.
  • the second depression of the first memory cell is arranged in the substrate and outside the first depression of the second memory cell and adjoins the second flank of the first depression of the second memory cell.
  • the substrate is etched with the aid of a mask which is arranged over the first flanks of the first depressions when the second depression is produced.
  • the capacitor dielectric can also have the further recess in the finished DRAM cell arrangement, so that the second recess of the first Storage cell adjoins the storage node of the second storage cell in the region of the further recess.
  • a mask can first be created which covers the first depressions to be produced. Trenches are produced between the first depressions to be produced by creating spacers on the flanks of the mask and etching the substrate selectively with respect to the mask and the spacers. The trenches are filled with insulating material. Material is then deposited and etched back so that the material is located between parts of the mask. The mask is removed and the first recesses are created by etching the substrate selectively to the material. The substrate is covered with insulating material. A portion of the substrate that is adjacent to the first flank of the first depression of the first memory cell is exposed.
  • the substrate is etched isotropically, the trench filled with the insulating material acting as a lateral etching stop, so that a cutout is produced in the substrate, which recess adjoins the first flank of the first depression of the first memory cell.
  • the recess is filled with insulating material.
  • the insulating material and structure is partially replaced by the mask for the second wells by etching back the insulating material and structure and depositing and planarizing material until a portion of the substrate is exposed that contacts the second flank of the first Deepening of the second storage cell and adjacent to the trench.
  • the second wells are created using this mask by etching the substrate selectively to the material.
  • an expansion of the substrate perpendicular to the channel plane can also be set precisely when a distance between adjacent first depressions is only F.
  • This expansion determines the threshold voltage of the transistor. In this case it is through given the width of the spacers since the trenches are formed between mutually adjacent spacers, and the trenches, by acting as an etching stop, determine how large an area is that covers the mask for the second depressions.
  • the second depression of the first memory cell is arranged in the first depression of the second memory cell and shares part of the second flank of the first depression of the second memory cell with the first depression of the second memory cell.
  • the extent of the substrate perpendicular to the channel plane is determined by a first mask, which is used to produce the first depressions.
  • a first layer is applied to the substrate and structured in accordance with the first depressions.
  • a second layer is applied and structured so that it is arranged over the first flanks of the first depressions and does not cover the regions over second flanks of the first depressions opposite the first flanks.
  • the first layer and the second layer act as a mask in the production of the second depressions, in which the insulating structure and the conductive material are etched.
  • the first layer and the second layer consist of a material which is selectively etchable for the insulating structure. If the insulating structure consists of SiC> 2, the first layer and the second layer can consist of silicon nitride, for example.
  • a layer is produced on the first layer that can be etched selectively to the second layer.
  • the second layer is created on the layer.
  • the layer can act as an etch stop so that the first layer is not attacked.
  • a layer is applied to the substrate and structured in accordance with the first depressions. Spacers are produced on the first flanks of the first depressions.
  • the second depressions are created by etching the conductive material selectively to the layer and the spacers.
  • the spacers can be the insulating structures. Alternatively, the spacers are removed and replaced by the insulating structures. In both cases, a width of the second depressions is determined by the thickness of the spacers.
  • the DRAM cell arrangement has so-called folded bit lines.
  • the signal of the bit line via which the information is read out is compared with the signal of a bit line adjacent to the bit line, the signal of which consists of background noise. In this way the background noise can be filtered out. So that the signal of the adjacent bit line consists only of background noise, no memory cell which is connected to the adjacent bit line may be connected to the word line to which the memory cell to be read out is connected.
  • the second depression is part of a word line trench in which two different word lines are arranged.
  • the gate electrode of the transistor is part of one of the word lines.
  • conductive material can be deposited and etched back, so that the word lines in the form of
  • Spacers are formed on the flanks of the word line trench.
  • the second depression is partially arranged in the substrate and partially in the first depression, since the second depression can have at least a width of F with a high packing density at the same time, so that the two word lines have space in the same word line trench.
  • Word line trenches only a single word line is arranged. In this case one speaks of so-called open bit lines.
  • a dopant source can be generated in the first depression before the storage node is generated, from which dopant diffuses into the substrate in a tempering step.
  • Capacitor electrode is a doped region in the substrate and surrounds at least part of the first recess.
  • a dopant source z.
  • B. arsenic glass is suitable, which is deposited so that surfaces of the first wells are covered, but the first wells are not filled. Then the first wells with a polymer, e.g. B. photoresist, filled that is etched back to a height which is below the height of the lower source / drain regions to be generated. Exposed arsenic glass is then removed. Through a tempering step, arsenic diffuses from the arsenic glass into the substrate. If a distance between adjacent first depressions is sufficiently small, adjacent capacitor electrodes grow together and form a common capacitor electrode.
  • the capacitor electrode can also be produced by plasma immersion. Ions of a plasma diffuse into the substrate.
  • the common capacitor electrode can also be formed as a doped layer of the substrate before the first depression is produced be generated. This layer is created, for example, by epitaxy or by implantation.
  • the lower source / drain region is diffused out of dopant from e.g. the storage node or through
  • Insulations are arranged. This prevents a word line in the word line trenches of the first memory cell from driving the transistor of the second memory cell in the region of edges of the first recess which adjoin the first edge of the first recess.
  • the isolations prevent the upper source / drain region from adjoining the word line trenches.
  • the upper source / drain region, the channel region and the lower source / drain region are arranged along the direction of the bit lines between the first depression and the second depression.
  • isolation trenches are created after the creation of the first depression, which run essentially parallel to one another and to the bit lines.
  • the first depression is cut by two of the isolation trenches that are adjacent to one another.
  • Transistor of the second memory cell is only on the first edge of the first depression of the second Arranged memory cell and can not be controlled by a word line of the second recess of the first memory cell.
  • the isolation trenches are filled with the isolations by separating isolating material. Word line trenches which run essentially parallel to one another are then produced by etching at least both the substrate and the insulation.
  • the isolating structures can be created before, after or together with the isolations.
  • the dimension of the first recess parallel to the distance between the two insulations is more than F.
  • the first depression is preferably produced in a self-aligned manner adjacent to the insulation.
  • the substrate is selectively etched to the insulation using a strip-shaped mask, the strips of which run transversely to the isolation trenches.
  • the dimension of the first depression F can be parallel to the distance between the insulations.
  • the capacitor electrode is preferably not produced by out-diffusion.
  • the bottoms of the isolation trenches are lower than the lower source / drain regions.
  • the lower source / drain region is delimited on two sides by the isolation trenches and delimited on the remaining two sides by the first depression or by the second depression.
  • the channel area which is consequently a floating body. Since the gate electrode is preferably arranged on a flank of the second depression only in the region of the channel region, it is expedient if the word line trench is flatter than the insulation trenches.
  • the word line can protrude from the word line trenches. This is advantageous since parts of the word line arranged outside the word line trench can be produced from metal, so that the word line has an increased electrical conductivity. In addition, such
  • Word line are structured together with gate electrodes of transistors of the periphery of the DRAM cell arrangement, which means a process simplification.
  • at least one conductive material e.g. B. doped polysilicon, deposited, and structured with the aid of a strip-shaped mask, the strips of which run essentially parallel to the word line trenches and which do not cover the at least parts of the word line trenches.
  • a protective layer for. B. is generated in the generation of the gate dielectric, is arranged, which serves as an etch stop.
  • a material with a high electrical conductivity for. B. a metal or a metal silicide, are deposited on the conductive material and then structured together with the conductive material.
  • bit lines are generated in such a way that they run above the word lines, it is advantageous to encapsulate the word lines in order to avoid short circuits between the bit lines and the word lines.
  • an insulating material for example silicon nitride, is deposited and etched back, so that protective spacers are produced on the flanks of the protruding parts of the word line.
  • the word line can also be covered with insulating material by separating the insulating material before structuring the conductive material of the word line and structuring it together with the conductive material. Spaces between the word lines can be used to create a planar
  • the insulating layer is selectively etched to the insulating material with the aid of a mask, which does not cover regions above the upper source / drain regions. Since the
  • the DRAM cell arrangement can be produced with a high packing density. A slight misalignment of the mask does not lead to a short circuit of the word lines with the bit lines. Contacts are created in the contact holes.
  • the contacts and the bit lines are produced by depositing conductive material and structuring with the aid of a strip-shaped mask, the strips of which run transversely to the word lines and which at least partially do not cover the contacts. To topology problems due to a too high
  • the word lines are generated, for example, by conducting material, eg. B. doped polysilicon, is deposited so that the word line trenches are filled, and then etched back until the conductive material is removed outside the word line trenches.
  • conducting material eg. B. doped polysilicon
  • the first depressions are arranged such that storage nodes which adjoin the word line trenches alternately adjoin a first flank and a second flank of the word line trench from adjacent memory cells. If a first word line borders the first edge of the word line trench and a second word line borders the second edge of the word line trench, the DRAM cell arrangement has folded bit lines. The first word line is only connected to every second of these memory cells. The second word line is connected to the rest of these memory cells, so that memory cells which are connected to adjacent bit lines are not connected to the same word line.
  • a memory cell of the DRAM cell arrangement can have a space requirement of 5-6 F 2 .
  • first depressions are arranged such that storage nodes which adjoin the word line trenches adjoin adjacent memory cells to the same flank of the word line trench.
  • a distance between adjacent word lines and a distance between each other adjacent bit lines can be F, so that an effective space requirement per memory cell 4 can be F 2 .
  • Word lines have protuberances which are arranged in the second depressions.
  • the word line is arranged in the second depression of the first memory cell and in the second depression of the second memory cell.
  • the substrate is preferably a semiconductor substrate comprising monocrystalline silicon and / or germanium.
  • the substrate can contain GaAs.
  • the substrate can comprise epitaxially grown layers of semiconductor material.
  • FIG. 1 shows a cross section through a first substrate after depressions, a capacitor electrode, a first part of a capacitor dielectric, a second part of a capacitor dielectric and storage nodes have been produced.
  • FIG. 2a shows a cross section from FIG. 1 after the first insulating structures, isolation trenches, isolations, upper source / drain regions and lower source / drain regions of transistors have been produced.
  • Figure 2b shows a cross section perpendicular to the cross section of Figure 2a through the first substrate.
  • FIG. 2c shows a top view of the first substrate, in which the depressions and the isolation trenches are shown.
  • FIG. 3a shows the cross section from FIG. 2a after a first layer, a second layer and a mask have been produced and structured from photoresist.
  • Figure 3b shows the top view of Figure 2c, in which the
  • Figure 4 shows the cross section of Figure 3a after
  • Word line trenches, second insulating structures, a gate dielectric, word lines and third insulating structures were produced.
  • FIG. 5a shows the cross section from FIG. 4 after fourth insulating structures, fifth insulating structures, an insulating layer, contacts and bit lines have been produced.
  • FIG. 5b shows the cross section from FIG. 2b after the process steps from FIG. 5a.
  • FIG. 5c shows the top view from FIG. 3b, in which the depressions, the isolation trenches, the
  • Word line trenches, the contacts and the bit lines are shown.
  • Figure 6 shows a cross section through a second substrate after recesses, a capacitor electrode
  • Capacitor dielectric storage node first insulating structures, isolation trenches with Insulations (not shown), word line trenches, second insulating structures, a gate dielectric, upper source / drain regions, channel regions, lower source / drain regions, word lines, third insulating structures, spacers, an insulating one
  • FIG. 7 shows a cross section through a third substrate, after depressions, a capacitor electrode, a capacitor dielectric, storage nodes, first insulating structures, isolation trenches with insulation (not shown), word line trenches, second insulating structures, a gate dielectric, upper source / drain regions, channel regions, lower source / drain regions, word lines, fourth insulating structures, fifth insulating structures, an insulating layer, contacts and bit lines were generated.
  • FIG. 8a shows a cross section through a fourth substrate after isolation trenches with insulations, depressions, a capacitor electrode and a capacitor dielectric have been produced and conductive material has been deposited and etched back.
  • FIG. 8b shows a top view of the fourth substrate after the process steps from FIG. 8a.
  • FIG. 9a shows a cross section through a fifth substrate after isolation trenches with isolations (in FIG. 9
  • depressions a capacitor electrode, a capacitor dielectric, storage nodes first insulating structures, word line trenches, second insulating structures, a gate dielectric, upper source / drain regions,
  • FIG. 9b shows a cross section through the fifth substrate perpendicular to the cross section from FIG. 9a.
  • FIG. 9c shows a top view of the fifth substrate, in which the isolations, the first depressions, the word line trenches, the bit lines and regions which are not covered by a mask are shown.
  • Figure 10a shows a cross section through a sixth
  • Substrate after a first layer, a second layer and a third layer have been produced.
  • FIG. 10b shows the cross section from FIG. 10a after depressions, a capacitor electrode, a capacitor dielectric, storage nodes, first insulating structures, isolation trenches (not shown), upper source / drain regions, channel regions, lower source / drain regions, word line trenches, second insulating trenches Structures, a gate dielectric, word lines, third insulating structures, fourth insulating
  • FIG. 11 shows a cross section through a seventh substrate after isolation trenches with isolations (not shown), a first layer made of silicon nitride, depressions, a capacitor dielectric, a capacitor electrode, storage nodes, first insulating structures, upper source / drain regions,
  • Channel areas, lower source / drain areas, a layer of SiO 2, a second layer of Silicon nitride, word line trenches and second insulating structures were produced.
  • FIG. 12 shows the cross section from FIG. 11 after a gate dielectric, word lines, third insulating
  • FIG. 13 shows the cross section through an eighth substrate after a layer of silicon nitride
  • Figure 14 shows the cross section of Figure 13 after
  • Word line trenches, spacers, second insulating structures, a gate dielectric and word lines were generated.
  • FIG. 15 shows the cross section from FIG. 14 after third insulating structures, an insulating layer, contacts and bit lines have been produced.
  • FIG. 16a shows the cross section through a ninth substrate which comprises a layer after a mask, first spacers and trenches have been produced.
  • FIG. 16b shows a cross section perpendicular to the cross section from FIG. 16a through the ninth substrate after the process steps from FIG. 16a.
  • FIG. 16c shows a top view of the ninth substrate, in which the mask, the first spacers and the trenches are shown.
  • FIG. 17a shows the cross section from FIG. 16a after first insulating structures and second insulating structures have been produced.
  • FIG. 17b shows the cross section from FIG. 16b after the process steps from FIG. 17a.
  • FIG. 18a shows the cross section from FIG. 17a after first depressions, a capacitor electrode
  • Capacitor dielectric, storage nodes, third insulating structures and second spacers were generated.
  • FIG. 18b shows the cross section from FIG. 17b after the process steps from FIG. 18a.
  • FIG. 19a shows the cross section from FIG. 18a after the third insulating structures have been enlarged, the first spacers, upper parts of the first insulating structures, the second insulating structure and parts of the second spacers have been removed, isolation trenches with isolations (shown in FIG. 19b), fourth insulating structures, upper source / drain regions, channel regions, lower ones
  • Source / drain areas and cutouts were created.
  • FIG. 19b shows the cross section from FIG. 18b after the process steps from FIG. 19a.
  • FIG. 20a shows the cross section from FIG. 19a after the fourth insulating structures and the second spacers have been removed and fifth insulating structures and a further mask have been produced.
  • FIG. 20b shows the cross section from FIG. 19b after the process steps from FIG. 20a.
  • FIG. 21a shows the cross section from FIG. 20a, after second depressions, sixth insulating structures, a gate dielectric, word lines, seventh insulating structures, third spacers, an insulating layer,
  • FIG. 21b shows the cross section from FIG. 20b after the process steps from FIG. 21a.
  • F 150 nm, where F is the minimum structure size that can be produced in the technology used.
  • a largely n-doped first substrate S made of silicon which has an approx. 1 ⁇ m thick p-doped layer P comprises a
  • depressions V approximately 10 ⁇ m deep are produced in the first substrate S.
  • an etchant such. B. HBr + HF suitable.
  • the depressions V each have two opposing flat flanks. Furthermore, the depressions V have two further opposing flanks which are curved, so that the depressions V have no edges or corners.
  • a y-axis y runs parallel to a surface f of the first substrate S and parallel to the flat flanks of the depressions V.
  • An x-axis x runs perpendicular to the y-axis y and parallel to the surface f of the first substrate S.
  • a distance between one of the two flat flanks Wells V is approximately 190 nm.
  • Rows are each formed by wells V, which are adjacent to one another along the x-axis x.
  • a distance between two depressions V of a row, which are adjacent to one another, is approximately 185 nm. The projections of every other row on the x-axis x correspond to one another.
  • Projections of mutually adjacent lines on the x-axis x are shifted in a translation-symmetrical manner in the direction of the x-axis x so that a depression V of a first of the rows is arranged between two mutually adjacent depressions V of a second row adjacent to the first row.
  • a distance between the curved flanks of the depression V with respect to the y-axis y is approximately 300 nm.
  • a distance parallel to the y-axis y between a depression V in a row and a depression V in the row after next is approximately 450 nm (see FIG. 2 c).
  • the first mask made of photoresist is removed. Then arsenic glass is deposited to a thickness of approximately 50 nm, so that surfaces of the depressions V are covered with arsenic glass without the depressions V being filled (not shown). Subsequently, photoresist is deposited to a thickness of approximately 500 nm and etched back to a first height h (see FIG. 1), which is approximately 1.5 ⁇ m below the surface f of the first substrate S. As an etchant such. B. 02 ⁇ plasma suitable. Then exposed parts of the arsenic glass with z. B. HF removed. The photoresist is z. B. 02 ⁇ plasma removed.
  • dopant diffuses from the arsenic glass into the first substrate S, so that an n-doped capacitor electrode E is produced in the first substrate S, which surrounds parts of the depressions V, into the p-doped layer P, and one
  • n-doped polysilicon is deposited in situ to a thickness of approximately 500 nm, so that the depressions V are filled.
  • the polysilicon is planarized by chemical-mechanical polishing until the surface f of the first substrate S is exposed.
  • the polysilicon is then etched back to a second height H, which is below the first height h and approximately 2 ⁇ m below the surface f of the first substrate S (see FIG. 1).
  • an etchant such. B. C2 5 + O2 suitable.
  • Capacitor dielectric is deposited by a TEOS process SiO 2 in a thickness of about 25 nm and with z.
  • B. CHF3 + O2 etched back so that V spacer-shaped structures are produced on the flanks of the depressions.
  • n-doped polysilicon is deposited in situ to a thickness of approx. 500 nm, planarized by chemical mechanical polishing until the surface f of the first substrate S is exposed, and up to an upper height o which is above the first height h and approx 400 nm below the surface f of the first substrate S, etched back. Then the spacer-shaped structures made of SiO 2 with z. B. HF up to a lower height u, which is about 80 nm below the upper height o removed.
  • the part of the spacer-shaped structure which is removed between the lower height u and the upper height o is replaced by amorphous n-doped silicon by depositing the amorphous silicon in a thickness of approximately 20 nm and then etching 30 nm isotropically (see FIG. 1). .
  • Remaining parts of the Spacer-like structures form the second part Kb of the capacitor dielectric.
  • the capacitor dielectric Ka, Kb has cutouts in a region between the lower height u and the upper height o.
  • the polysilicon and the amorphous silicon form storage nodes Sp, which are each arranged in one of the depressions V and adjoin the substrate S at the cutouts.
  • first insulating structure Ia SiO 2 is deposited in a thickness of approx. 200 nm in a TEOS process and z. B. CHF3 + O2 etched back until the surface f of the first substrate S is exposed.
  • the first insulating structures Ia are arranged in the depressions V and on the storage node Sp (see FIGS. 2a and 2b).
  • isolation trenches GI are produced in such a way that the depressions V are each of two of the isolation trenches GI, which are adjacent to one another, are cut.
  • Capacitor dielectric etched.
  • the isolation trenches GI are approximately 800 nm deep.
  • As an etchant such. B. NF3 + Ar suitable (see Figures 2b and 2c).
  • the second mask is removed.
  • isolations IS are produced in the isolation trenches GI by depositing SiO 2 in a thickness of approximately 200 nm and planarizing by chemical mechanical polishing until the surface f of the first substrate S is exposed.
  • implanted dopant is activated by a tempering step.
  • the upper source / drain regions SDo are approximately 100 nm deep and have a dopant concentration of approximately 5 x 10 20 cm "3. Due to the tempering step, dopant diffuses from the storage nodes Sp in the region of the cutouts into the first substrate S, see above that lower source / drain regions SDu of the transistors are produced, which are each arranged between two of the depressions V and between two of the isolation trenches GI.
  • SiO 2 is deposited on the surface f of the first substrate S in a thickness of approximately 30 nm.
  • 2 polysilicon are deposited in a thickness of approximately 30 nm to produce a second layer.
  • a stripe-shaped third mask Mc is produced from photoresist over the second layer 2, the stripes of which are approximately 225 nm wide, spaced approximately 150 nm apart and run parallel to the y-axis y (see FIG. 3b).
  • the strips of the third mask Mc overlap the upper source / drain regions SDo and the first insulating structures Ia (see FIG. 3a). Parts of the upper source / drain regions SDo and the first insulating structures Ia, which are in the region of first flanks Fa
  • the second layer 2, the first layer 1 and then the first substrate S, the first insulating structures Ia, and the second parts Kb of the capacitor dielectric are etched, and the storage nodes Sp and the isolations IS are structured, so that between the strips of the third mask Mc word line trenches GW are generated, the bottoms of which are approximately 800 nm below the surface f of the first substrate S (see FIG. 4).
  • NF3 + Ar is suitable as an etchant.
  • the Word line trenches GW adjoin the storage nodes Sp in the region of second flanks Fb of the depressions V opposite the first flanks Fa of the depressions V.
  • the bottoms of the word line trenches GW are lower than the cutouts of the capacitor dielectric Ka, Kb and higher than bottoms of the isolation trenches GI.
  • the third mask Mc is removed.
  • SiO 2 is deposited to a thickness of approximately 200 nm and etched back approximately 500 nm deep with CHF3 + O2 (see FIG. 4).
  • An approximately 4 nm thick gate dielectric Gd is produced by thermal oxidation and also covers the second layer 2.
  • n-doped polysilicon is deposited in situ to a thickness of approximately 50 nm and z.
  • the gate dielectric Gd protects the second layer 2.
  • the first of the word lines W adjoin the first flanks of the word line trenches GW and the second of the word lines W adjoin the second flanks of the word line trenches GW.
  • Word lines W are separated from the storage nodes Sp by the gate dielectric Gd and the second insulating structures Ib.
  • the storage node Sp of the capacitor of a first memory cell adjoins the lower source / drain region SDu of the transistor of the first memory cell in the region of the recess in the capacitor dielectric Ka, Kb, which lies on the first flank Fa of the associated recess V.
  • the lower source / drain region SDu is delimited by two of the isolation trenches GI, by the depression V and by one of the word line trenches GW.
  • the Word line trenches GW separate the lower source / drain region SDu from the depression V of a second memory cell which is adjacent to the first memory cell.
  • a part of the first substrate S which is arranged between the lower source / drain region SDu and the upper source / drain region SDo of the transistor, serves as the channel region KA of the transistor (see FIG. 4).
  • a part of one of the word lines W which is arranged in the word line trenches GW and is separated from the channel region KA of the transistor by the gate dielectric Gd, acts as the gate electrode of the transistor.
  • Storage nodes Sp which adjoin the word line trenches GW, of adjacent memory cells alternately adjoin a first edge and a second edge of the word line trench GW.
  • the capacitor dielectric Ka, Kb only has the cutout on the first
  • the third insulating structures Ic are arranged in the word line trenches GW between the word lines W.
  • silicon nitride is deposited in a thickness of approximately 100 nm and approximately 120 nm deep with z.
  • the fourth insulating structures Id are arranged in the word line trenches GW and cover the word lines W (see FIG. 5a).
  • S1O2 is deposited to a thickness of approximately 200 nm and planarized by chemical mechanical polishing until the Surface f of the first substrate S is exposed. This creates a planar surface. The second layer 2 and the first layer 1 are removed.
  • the planar surface can also be formed by the fourth insulating structures Id by chemical-mechanical polishing instead of etching back.
  • the fifth insulating structure le can then be dispensed with.
  • SiO 2 is deposited in a thickness of approximately 250 nm.
  • Contact holes are etched into the insulating layer I using a fourth mask made of photoresist, which does not cover the square regions Q, which have a side length of approximately 150 nm and overlap the upper source / drain regions SDo (see FIG. 5c).
  • Contacts K are produced in the contact holes by depositing tungsten with a thickness of approximately 100 nm and chemical-mechanical polishing until the insulating layer I is exposed.
  • bit lines B are thereby generated which contact the upper source / drain regions SDo (see FIGS. 5a, 5b and 5c).
  • No two memory cells connected to adjacent bit lines B are connected to the same word line W.
  • a DRAM cell arrangement comprising the memory cells consequently has folded bit lines.
  • F 150nm is the minimum structure size that can be produced in the technology used.
  • the associated word line is driven and the signal, which is determined by the charge of the associated capacitor, is read out via the associated bit line. In order to filter out background noise, this signal is compared with a signal on a bit line B adjacent to bit line B.
  • the associated word line W is driven and a voltage is applied to the bit line B, which depending on the information to be stored is e.g. Is 0 V or 1.8 V.
  • a second substrate IS is provided which corresponds to the substrate S of the first exemplary embodiment.
  • the second layer is removed.
  • a gate dielectric LGD is generated by thermal oxidation.
  • n-doped polysilicon is in situ in one
  • Thickness of approx. 200 nm Tungsten nitride is deposited in a thickness of approx. 200 nm. Tungsten is deposited in a thickness of approx. 100 nm. Silicon nitride is deposited in a thickness of approx. 100 nm. With the aid of a strip-shaped photoresist mask (not shown), the strips of which are arranged above the word line trenches IGW, silicon nitride, Tungsten, tungsten nitride and polysilicon are etched until the first insulating structures Ila are exposed on a surface lf of the second substrate IS (see FIG. 6).
  • a word line IW is generated in each word line trench IGW which protrudes from the word line trench IGW and consists of polysilicon, tungsten nitride and tungsten.
  • Third insulating structures 11c are formed from silicon nitride and cover the word lines IW.
  • silicon nitride is deposited to a thickness of approximately 50 nm and etched back, so that spacers IC are produced on the flanks of parts of the word lines IW which protrude from the second substrate IS (see FIG. 6).
  • An insulating layer II, contacts 1K and bit lines IB are then produced.
  • the third insulating structures 11c and the spacers IC protect the word lines IW.
  • a DRAM cell arrangement generated in this way has open bit lines.
  • a third substrate 2S is provided, which corresponds to the first substrate S of the first exemplary embodiment.
  • n-doped polysilicon is deposited in situ to a thickness of approximately 100 nm and z. B. C2F5 + O2, until a word line 2W is generated in each word line trench 2GW, which is approximately 70 nm below a surface 2f of the third substrate 2S (see FIG. 7).
  • fourth insulating structures 2Id made of silicon nitride, fifth insulating structures 2Ie made of Si02c, an insulating layer 21, contacts 2K and bit lines 2B are produced (see FIG. 7).
  • a DRAM cell arrangement generated in this way has open bit lines.
  • a fourth substrate 3S is provided which corresponds to the first substrate S of the first exemplary embodiment.
  • the first mask is then removed.
  • Isolation trenches 3GI are filled with isolations 3IS by depositing S1O2 m with a thickness of approximately 200 nm and planarizing by chemical mechanical polishing until a surface 3f of the fourth substrate 3S is exposed (see FIG. 8a).
  • a second mask 3Ma is then produced from photoresist.
  • the second mask 3Ma is composed of jagged strips (see FIG. 8b), so that when the third substrate 3S is etched, selective depressions 3V are generated in the isolation trenches 3GI in the isolation trenches 3GI, the dimensions of which are parallel to the x-axis x and approximately 180 nm and their size dimensions parallel to the y-axis y are approximately 150 nm (see FIGS. 8a and 8b).
  • depressions 3V adjacent to one another form along the x-axis x, projections of which on the x-axis of every second line coincide with one another.
  • the projection of a depression 3V in one row is adjacent to the projections of two depressions in an adjacent row.
  • the depressions 3V are approx. 10 ⁇ m deep.
  • As an etchant such. B. HBr + HF suitable.
  • silicon nitride is first deposited to a thickness of approximately 20 nm. Photoresist is applied in a thickness of approx. 500 nm and z. B. 02- plasma etched back approx. 2 ⁇ m deep. Exposed parts of the silicon nitride are z. B. H3PO4 removed. The photoresist is then removed so that flanks of the depressions 3V are exposed between a second height H, which is approximately 2 ⁇ m below the surface 3f, and the surface 3f, while they are covered by silicon nitride below the second height H. Thermal oxidation produces between the second height H and the surface 3f the approximately 25 nm thick second part of the capacitor dielectric 3 Kb.
  • the oxidized silicon nitride is then removed.
  • a capacitor electrode 3E that surrounds the recesses 3V is produced by plasma immersion.
  • first insulating structures, upper source / drain regions, channel regions, lower source / drain regions, word line trenches, a gate dielectric, word lines, further insulating structures, an insulating layer, contacts and bit lines are produced (not shown).
  • a fifth substrate 4S is provided which corresponds to the first substrate S of the first exemplary embodiment.
  • isolation trenches 4GI are produced, with the difference that the isolation trenches 4GI are approximately 150 nm wide (see FIG. 9c).
  • the isolation trenches 4GI are filled with isolations 4IS.
  • a strip-shaped mask 4Ma is then produced from photoresist, the strips of which are approximately 150 nm wide, spaced approximately 150 nm apart and run parallel to the y-axis y (see FIG. 9c).
  • the mask 4Ma is used to selectively etch the fifth substrate 4S to the insulations 4IS, so that depressions 4V are produced between the isolation trenches 4GI, which have a square horizontal cross section with a side length of approximately 150 nm.
  • Recesses 4V adjacent to each other along the x-axis form a row. Rows adjacent to one another are arranged in such a way that adjacent memory cells form columns along the y axis (see FIG. 9c).
  • a capacitor electrode 4E As in the fourth embodiment, a capacitor electrode 4E, a capacitor dielectric 4Ka, 4Kb and storage nodes 4Sp are produced. Then, as in the second exemplary embodiment, insulating structures 41a, 41b, 41c, word line trenches 4GW, a gate dielectric 4Gd, word lines 4W, spacers 4C, an insulating layer 41, contacts 4K and bit lines 4B are produced (see FIGS. 9a and 9b).
  • a DRAM cell arrangement produced in this way has memory cells with a space requirement of only 4 F 2 .
  • Substrate 5S made of monocrystalline n-doped silicon with a dopant concentration of approximately 10 15 cm -3 is provided. Implantation with p-doping ions produces an approximately 500 nm thick first layer a, which has a dopant concentration of approximately 10 18 cm -3 .
  • an approximately 200 nm thick n-doped second layer b is produced on the first layer a, which has a dopant concentration of approximately 5 * 10 18 cm "3.
  • an approximately 300 nm thick layer is produced by epitaxy p-doped third layer c produces one
  • Capacitor electrode 5E a capacitor dielectric 5Ka, 5Kb, storage node 5Sp, first insulating structures 51a, isolation trenches with insulation (not shown), upper source / drain regions 5SDo of transistors and channel regions 5KA of the transistors.
  • the upper source / drain regions 5S / Do are produced in upper parts of the third layer c. Other parts of the third layer c form the channel regions 5KA.
  • the dopant diffuses out of the
  • the tempering step is shorter in duration than the corresponding one Tempering step carried out in the first embodiment, so that the dopant of different storage nodes 5Sp do not meet (see FIG. 10b).
  • At first flanks of the recesses 5FA 5V are thus highly doped parts of 5S in the second layer b / Du generated by the lower source / drain regions of the transistors that have a dopant concentration of approximately l ⁇ l9cm-3.
  • Other parts of the second layer b form lightly doped parts of the lower source / drain regions.
  • a DRAM cell arrangement produced in this way has lower leakage currents compared to the DRAM cell arrangement of the first exemplary embodiment, since the highly doped parts 5SDu of the lower source / drain regions do not directly adjoin the channel regions 5KA or the first layer a.
  • the DRAM cell arrangement has softer p-n junctions due to the low-doped parts of the lower source / drain regions.
  • a channel length of the transistors can be set more precisely in comparison to the first exemplary embodiment, since it is caused by epitaxy and by the implantation depth of the upper one
  • Source / drain areas 5S / Do is determined.
  • the channel length in the first exemplary embodiment is determined by an etching depth combined with out-diffusion and the implantation depth of the upper source / drain regions S / Do.
  • a seventh substrate 6S is provided which corresponds to the fifth substrate 4S of the fifth exemplary embodiment.
  • isolation trenches are produced and filled with isolations (not shown).
  • silicon nitride is deposited in a thickness of approximately 50 nm.
  • depressions 6V are produced, the first layer N1 being additionally structured from silicon nitride (see FIG. 11).
  • Silicon nitride is then deposited to a thickness of approximately 4 nm. Above this, photoresist is applied in a thickness of approx. 500 nm and etched back approx. 2 ⁇ m deep with 02-Plas a.
  • silicon nitride is removed in a thickness of approximately 4 nm, so that the first layer N1 of silicon nitride is retained, but silicon nitride is exposed from flanks of the depressions 6V, which lie above the photoresist. The photoresist is then removed.
  • Thermal oxidation produces a second part 6Kb of a capacitor dielectric between a height H, which corresponds to the second height H of the first exemplary embodiment, and a surface 6f of the seventh substrate 6S.
  • the silicon nitride below the height H is oxidized and forms a first part 6Ka of the capacitor dielectric (see FIG. 11).
  • a capacitor electrode 6E, storage node 6Sp, first insulating structures 61a, upper source / drain regions 6SD0, Channel regions 6KA and lower source / drain regions 6SDu are generated (see FIG. 11).
  • an approximately 10 nm thick layer 0 ' is produced from S1O2 (see FIG. 11).
  • An approximately 3 nm thick second layer N2 of silicon nitride is produced above this.
  • the second layer N2 made of silicon nitride is structured with the aid of a strip-like mask (not shown) which corresponds to the third mask Mc of the first exemplary embodiment.
  • the second layer N2 made of silicon nitride covers areas over first flanks 6Fa of the depressions ⁇ V.
  • the layer 0 'made of S1O2 acts as an etch stop. The mask is then removed.
  • second insulating structures 61b, a gate dielectric 6Gd, word lines 6W and third insulating structures 61c are produced (see FIG. 12).
  • silicon nitride is deposited to a thickness of approximately 50 nm and coated with e.g. B. C2Fg + O2 approximately 50 nm. Parts of the first layer N1 are thereby made Silicon nitride and the second layer N2 of silicon nitride removed (see Figure 12).
  • an insulating layer 61, contacts 6K and bit lines 6B are produced (see FIG. 12).
  • an eighth substrate 7S is provided which corresponds to the seventh substrate 6S of the seventh exemplary embodiment.
  • Capacitor electrode 7E a capacitor dielectric 7Ka, 7Kb, storage node 7Sp, first insulating structures 71a, upper source / drain regions 7SDo, channel regions 7KA, lower source / drain regions 7SDu and isolation trenches with insulation (not shown) (see FIG. 13).
  • the first insulating structures 71a in the depressions 7V and parts of the insulation (not shown) are removed, so that word line trenches 7GW which cross the depressions 7V run transversely to the isolation trenches (not shown).
  • the word line trenches 7GW are narrowed by spacers 7C in that silicon nitride is deposited to a thickness of approximately 50 nm and etched back (see FIG. 14).
  • second insulating structures 71b, a gate dielectric 7Gd and word lines 7W are produced (see FIG. 14).
  • the spacers 7C and the layer N1 'made of silicon nitride are removed.
  • third insulating structures 71c are produced which adjoin and cover the word lines 7W (see FIG. 14).
  • an insulating layer 71, contacts 7K and bit lines 7B are produced (see FIG. 15).
  • a largely n-doped ninth substrate 8S which comprises an approximately 1.2 ⁇ m thick p-doped layer 8P, which has a dopant concentration of approximately 10 18 cm 3 .
  • SiO 2 is deposited in a thickness of approximately 300 nm on a surface 8f of the ninth substrate 8S.
  • a mask 8M from SiO 2 using a first photoresist mask (not shown)
  • a plan view of the mask 8M corresponds to a plan view of the depressions V of the first exemplary embodiment, with the difference that a distance between parts of the mask 8M which are adjacent to one another along the x-axis x is approximately 150 nm.
  • a distance between one parallel to the y-axis y Part of the mask 8M of a line and part of the mask 8M of a line after the next is approximately 450 nm.
  • the first photoresist mask is then removed.
  • silicon nitride is deposited to a thickness of approximately 70 nm and etched back to a depth of approximately 100 nm.
  • the first spacers 8C1 adjoin flanks of the mask 8M. Upper parts of the flanks of the mask 8M are exposed (see FIGS. 16a to 16c).
  • first insulating structures 81a SiO 2 is deposited in a thickness of approx. 10 nm and approx. B. HF etched, so that the first insulating structures 81a are produced in the trenches G between parts of the mask 8M which are adjacent to one another along the x axis (see FIG. 17a).
  • silicon nitride is deposited to a thickness of approximately 200 nm and coated with e.g. B. C2F5 + O2 etched back approx. 200 nm deep.
  • the second insulating structure 81b covers the spacers 8C1 and the first insulating structures 81a and fills up parts of the trenches G (see FIGS. 17a and 17b).
  • the mask 8M is etched with z. B. CHF3 + O2 selectively removed to silicon nitride.
  • the second insulating structure 81b serves as a mask (see FIGS. 18a and 18b).
  • Capacitor electrode 8E a capacitor dielectric 8Ka, 8Kb and storage node 8Sp generated (see Figures 18a and 18b).
  • SiO 2 is deposited to a thickness of approximately 200 nm and etched back to a depth of approximately 550 nm.
  • the third insulating structures 81c are arranged in the first depressions 8V on the storage node 8Sp and extend to a height of approximately 50 nm below a surface 8f of the ninth substrate 8S (see FIGS. 18a and 18b).
  • second spacers 8C2 silicon nitride is deposited to a thickness of approximately 20 nm and etched back.
  • the second spacers 8C2 are arranged in the first depressions 8V (see FIGS. 18a and 18b).
  • the third insulating structures 81c are then enlarged by depositing SiO 2 in a thickness of approximately 200 nm and planarizing together with the silicon nitride by chemical mechanical polishing until the surface 8f is exposed (see FIGS. 18a and 18b).
  • the first spacers 8C1, upper parts of the first insulating structures 81a, upper parts of the second spacers 8C2 and the second insulating structure 81b are removed (see FIGS. 19a and 19b).
  • isolations 8IS are generated in the isolation trenches 8GI, and upper source / drain regions 8SD0 and lower source / drain regions 8SDu of transistors are produced by implantation or outdiffusion (see FIGS. 19a and 19b).
  • Fourth insulating structures 8Id are generated on the upper source / drain regions 8SD0 by thermal oxidation (see FIGS. 19a and 19b).
  • silicon is etched isotropically to silicon nitride and SiO 2 approximately 100 nm deep, so that recesses A are produced in the ninth substrate 8S, which adjoin the first flanks 8Fa of the first depressions 8V (see FIG. 19a).
  • the recesses A in the ninth substrate 8S are approximately 70 nm wide along the x-axis x, since the third insulating structures 81c and the first insulating structures 81a act as an etching stop.
  • the formation of corresponding recesses in the ninth substrate 8S on second flanks 8Fb of the first depressions 8V opposite the first flanks 8Fa of the first depressions 8V is prevented by the remaining second spacers 8C2 and the fourth insulating structures.
  • the third photoresist mask is then removed.
  • the fourth insulating structures 8Id are removed by approximately 20 nm deep etching back of SiO 2.
  • the remaining second spacers 8C2 are z. B. H3PO4 removed.
  • fifth insulating structures 8Ie SiO 2 is deposited to a thickness of approximately 200 nm and planarized by chemical mechanical polishing until the surface 8f is exposed (see FIG. 20a).
  • the fifth insulating structures 8Ie fill the recesses A in the ninth substrate 8S.
  • the insulations 8IS, the first insulating structures 81a, the third insulating structures 81c and the fifth insulating structures 8Ie are removed in a thickness of approximately 80 nm and replaced by a further mask 8M 'by silicon nitride in 200 nm thick and planarized by chemical mechanical polishing until the surface 8f is exposed (see FIGS. 20a and 20b).
  • the capacitor dielectric 8Ka, 8Kb has further cutouts on the second flanks 8Fb of the first depressions 8V which are not overlaid by the second depressions 8V.
  • Sixth insulating structures 8If are produced on the bottoms of the second depressions 8V by depositing SiO 2 in a thickness of approx. 200 nm and etching back approx. 600 nm deep (see FIGS. 21a and 21b).
  • n-doped polysilicon is deposited in situ to a thickness of approximately 100 nm, so that the second depressions 8V are filled.
  • Tungsten nitride is deposited in a thickness of approx. 20 nm.
  • Tungsten is deposited in a thickness of approx. 100 nm.
  • Silicon nitride is deposited in a thickness of approx. 100 nm.
  • word lines 8W which are covered by seventh insulating structures 81g, silicon nitride, tungsten, tungsten nitride and polysilicon is patterned with the aid of a strip-shaped photoresist mask (not shown) which corresponds to the corresponding photoresist mask from the second exemplary embodiment (see FIGS. 21a and 21b).
  • third spacers 8C3 are produced by depositing silicon nitride in a thickness of approximately 50 nm and etching it back.
  • an insulating layer 81, contacts 8K and bit lines 8B are produced (see FIGS. 21a and 21b).
  • the word line can be generated by first depositing in-situ doped polysilicon to a thickness which does not fill up the word line trenches and then a material with a higher electrical conductivity, e.g. B. tungsten, is deposited so that the word line trenches are filled.
  • the tungsten and the polysilicon can be structured together and form the word lines.
  • the lower source / drain regions can be produced by structuring a doped layer of the substrate.
  • the doped layer and further layers from which the channel regions and the upper source / drain regions are produced can be produced by epitaxy. The same applies to the capacitor electrode.
  • the first layer a, the second layer b and the third layer c of the sixth exemplary embodiment can alternatively be produced by carrying out an implantation with n-doping ions with an energy starting from the sixth substrate 5S such that the second layer b than in the sixth Substrate 5S buried layer is generated at a depth between 200 nm and 400 nm.
  • the implantations can also be carried out after the wells have been created.
  • the fourth and the sixth exemplary embodiment can be modified in such a way that instead of folded bit lines, instead open bit lines are produced which either protrude from the substrate, as in the second exemplary embodiment, or, as in the third
  • Exemplary embodiment are buried in the substrate.
  • the fifth exemplary embodiment can be modified so that the DRAM cell arrangement has folded bit lines or open bit lines with buried word lines.

Landscapes

  • Engineering & Computer Science (AREA)
  • Manufacturing & Machinery (AREA)
  • Semiconductor Memories (AREA)

Abstract

Des cellules de mémoire comprennent chacune un transistor et un condensateur. Un noeud mémoire (Sp) du condensateur est disposé dans un premier évidement (V), tandis qu'une électrode de grille du transistor est disposée dans un deuxième évidement. Une région supérieure source-drain (Sdo), une région canal (KA) et une région inférieure source-drain (Sdu) du transistor sont superposées et sont chacune adjacentes à un premier flanc (Fa). Le premier flanc (Fa) est muni d'un diélectrique de condensateur (Ka, Kb) présentant, dans la zone de la région inférieure source-drain (Sdu), une cavité pour laquelle le noeud mémoire (Sp) est adjacent à la région inférieure source-drain (Sdu). Le deuxième évidement d'une première des cellules de mémoire peut être adjacent au noeud mémoire (Sp) disposé dans le premier évidement (V) d'une deuxième des cellules de mémoire. Les deuxièmes évidements peuvent être des parties de tranchées canal mot (GW) s'étendant transversalement par rapport aux tranchées d'isolation. Au-dessus de la cavité, il est prévu de préférence une structure isolante (Ia) dans le premier évidement (V), adjacente à deux, mutuellement voisines, des tranchées d'isolation. Des lignes de bits (B) viennent en contact avec la région supérieure source-drain (Sdo) par l'intermédiaire de contacts (K).
EP99967861A 1998-12-02 1999-12-01 Dispositif de cellule dram et son procede de fabrication Withdrawn EP1145320A1 (fr)

Applications Claiming Priority (3)

Application Number Priority Date Filing Date Title
DE19855688 1998-12-02
DE19855688 1998-12-02
PCT/DE1999/003840 WO2000033383A1 (fr) 1998-12-02 1999-12-01 Dispositif de cellule dram et son procede de fabrication

Publications (1)

Publication Number Publication Date
EP1145320A1 true EP1145320A1 (fr) 2001-10-17

Family

ID=7889783

Family Applications (1)

Application Number Title Priority Date Filing Date
EP99967861A Withdrawn EP1145320A1 (fr) 1998-12-02 1999-12-01 Dispositif de cellule dram et son procede de fabrication

Country Status (6)

Country Link
US (1) US6586795B2 (fr)
EP (1) EP1145320A1 (fr)
JP (1) JP3702180B2 (fr)
KR (1) KR100444791B1 (fr)
TW (1) TW469599B (fr)
WO (1) WO2000033383A1 (fr)

Families Citing this family (17)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
DE19941401C1 (de) 1999-08-31 2001-03-08 Infineon Technologies Ag Verfahren zur Herstellung einer DRAM-Zellenanordnung
US6335247B1 (en) * 2000-06-19 2002-01-01 Infineon Technologies Ag Integrated circuit vertical trench device and method of forming thereof
US6339241B1 (en) * 2000-06-23 2002-01-15 International Business Machines Corporation Structure and process for 6F2 trench capacitor DRAM cell with vertical MOSFET and 3F bitline pitch
DE10038728A1 (de) * 2000-07-31 2002-02-21 Infineon Technologies Ag Halbleiterspeicher-Zellenanordnung und Verfahren zu deren Herstellung
DE10143650A1 (de) * 2001-09-05 2003-03-13 Infineon Technologies Ag Halbleiterspeicher mit einen vertikalen Auswahltransistor umfassenden Speicherzellen sowie Verfahren zu seiner Herstellung
DE10318625B4 (de) * 2003-04-24 2006-08-03 Infineon Technologies Ag Vertikale Speicherzelle und Verfahren zu deren Herstellung
US7271433B1 (en) * 2004-09-02 2007-09-18 Micron Technology, Inc. High-density single transistor vertical memory gain cell
US7531464B2 (en) * 2005-12-20 2009-05-12 Texas Instruments Incorporated Semiconductive device fabricated using a substantially disassociated chlorohydrocarbon
TWI300975B (en) * 2006-06-08 2008-09-11 Nanya Technology Corp Method for fabricating recessed-gate mos transistor device
KR101334174B1 (ko) * 2007-01-12 2013-11-28 삼성전자주식회사 배선 구조체 및 상기 배선 구조체를 포함한 반도체 소자
TWI455314B (zh) 2011-01-03 2014-10-01 Inotera Memories Inc 具有浮置體的記憶體結構及其製法
US9401363B2 (en) * 2011-08-23 2016-07-26 Micron Technology, Inc. Vertical transistor devices, memory arrays, and methods of forming vertical transistor devices
JP6077251B2 (ja) * 2012-09-28 2017-02-08 エスアイアイ・セミコンダクタ株式会社 半導体装置
US10269804B2 (en) * 2016-05-11 2019-04-23 Micron Technology, Inc. Array of cross point memory cells and methods of forming an array of cross point memory cells
US11177266B2 (en) 2019-08-26 2021-11-16 Micron Technology, Inc. Array of capacitors, an array of memory cells, a method of forming an array of capacitors, and a method of forming an array of memory cells
US11557593B2 (en) 2020-11-30 2023-01-17 Micron Technology, Inc. Array of memory cells, methods used in forming an array of memory cells, methods used in forming an array of vertical transistors, and methods used in forming an array of capacitors
US11355531B1 (en) 2020-11-30 2022-06-07 Micron Technology, Inc. Array of capacitors, an array of memory cells, method used in forming an array of memory cells, methods used in forming an array of capacitors, and methods used in forming a plurality of horizontally-spaced conductive lines

Family Cites Families (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5208657A (en) 1984-08-31 1993-05-04 Texas Instruments Incorporated DRAM Cell with trench capacitor and vertical channel in substrate
US4763180A (en) * 1986-12-22 1988-08-09 International Business Machines Corporation Method and structure for a high density VMOS dynamic ram array
JPS63240061A (ja) * 1987-03-27 1988-10-05 Matsushita Electric Ind Co Ltd 半導体装置およびその製造方法
US5008214A (en) * 1988-06-03 1991-04-16 Texas Instruments Incorporated Method of making crosspoint dynamic RAM cell array with overlapping wordlines and folded bitlines
US5316962A (en) * 1989-08-15 1994-05-31 Matsushita Electric Industrial Co., Ltd. Method of producing a semiconductor device having trench capacitors and vertical switching transistors
JPH0414868A (ja) * 1990-05-09 1992-01-20 Hitachi Ltd 半導体記憶装置とその製造方法
US5034787A (en) * 1990-06-28 1991-07-23 International Business Machines Corporation Structure and fabrication method for a double trench memory cell device
JPH05291528A (ja) * 1992-04-09 1993-11-05 Toshiba Corp 半導体記憶装置およびその製造方法
JPH07130871A (ja) * 1993-06-28 1995-05-19 Toshiba Corp 半導体記憶装置
US5529944A (en) * 1995-02-02 1996-06-25 International Business Machines Corporation Method of making cross point four square folded bitline trench DRAM cell
US5831301A (en) * 1998-01-28 1998-11-03 International Business Machines Corp. Trench storage dram cell including a step transfer device
US6225158B1 (en) * 1998-05-28 2001-05-01 International Business Machines Corporation Trench storage dynamic random access memory cell with vertical transfer device

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See references of WO0033383A1 *

Also Published As

Publication number Publication date
US20020017671A1 (en) 2002-02-14
KR100444791B1 (ko) 2004-08-21
WO2000033383A1 (fr) 2000-06-08
TW469599B (en) 2001-12-21
JP2002531951A (ja) 2002-09-24
JP3702180B2 (ja) 2005-10-05
KR20010081063A (ko) 2001-08-25
US6586795B2 (en) 2003-07-01

Similar Documents

Publication Publication Date Title
DE68922819T2 (de) Ultradichte DRAM-Zelle-Matrix und ihr Herstellungsverfahren.
DE102007003583B4 (de) Verfahren zum Herstellen eines Transistors
EP0987764A1 (fr) Arrangement de cellules de mémoire programmables électriquement non-volatiles
DE10306281A1 (de) Anordnung und Verfahren zur Herstellung von vertikalen Transistorzellen und transistorgesteuerten Speicherzellen
WO2000033383A1 (fr) Dispositif de cellule dram et son procede de fabrication
DE19943760C1 (de) DRAM-Zellenanordnung und Verfahren zu deren Herstellung
EP0875937A2 (fr) Réseau de cellules DRAM et son procédé de fabrication
DE19811882A1 (de) DRAM-Zellenanordnung und Verfahren zu deren Herstellung
EP1116270A1 (fr) Ensemble circuit integre comportant des transistors verticaux, et son procede de production
DE19845004C2 (de) DRAM-Zellenanordnung und Verfahren zu deren Herstellung
EP0945901A1 (fr) Réseau de cellules DRAM à transistor vertical et son procédé de fabrication
EP1125328B1 (fr) Procede de production d'un ensemble de cellules de memoire ram dynamique
EP1129482B1 (fr) Procédé de production d'une configuration de cellules DRAM
WO2000055904A1 (fr) Agencement de cellules de memoire ram dynamique, et son procede de realisation
EP0917203A2 (fr) Structure de type gain-cell pour DRAM et son procédé de manufacture
DE19929211B4 (de) Verfahren zur Herstellung eines MOS-Transistors sowie einer DRAM-Zellenanordung
DE10109564A1 (de) Grabenkondensator und Verfahren zu seiner Herstellung
DE19954867C1 (de) DRAM-Zellenanordnung und Verfahren zu deren Herstellung
EP0920060B1 (fr) Réseau de cellules SRAM et méthode de fabrication
WO1998011604A1 (fr) Procede de production d'une matrice cellulaire a dram
WO2000060666A1 (fr) Ensemble cellules memoires et son procede de production
DE19923262C1 (de) Verfahren zur Erzeugung einer Speicherzellenanordnung
DD280851A1 (de) Verfahren zur herstellung von graben-speicherzellen
EP1118116B1 (fr) Substrat dote d'un creux adapte a une configuration de circuit integre et son procede de production
EP1097471B1 (fr) Circuit integre comportant au moins un transistor et un condensateur, et son procede de production

Legal Events

Date Code Title Description
PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

17P Request for examination filed

Effective date: 20010406

AK Designated contracting states

Kind code of ref document: A1

Designated state(s): AT BE CH CY DE DK ES FI FR GB GR IE IT LI LU MC NL PT SE

RBV Designated contracting states (corrected)

Designated state(s): DE FR GB IE IT

17Q First examination report despatched

Effective date: 20080723

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: THE APPLICATION IS DEEMED TO BE WITHDRAWN

18D Application deemed to be withdrawn

Effective date: 20081202