Embedded DRAM cells having capacitors within trench silicide trenches of a semiconductor structure
09831248 ยท 2017-11-28
Assignee
Inventors
Cpc classification
H01L27/0886
ELECTRICITY
H01L29/41791
ELECTRICITY
H01L28/91
ELECTRICITY
H01L27/1211
ELECTRICITY
International classification
H01L21/00
ELECTRICITY
H01L23/522
ELECTRICITY
Abstract
A semiconductor structure includes an array of fins extending horizontally across a substrate. A plurality of transistors are embedded in the fins. The transistors include a 1.sup.st S/D region and a 2.sup.nd S/D region defining a channel region therebetween. The transistors have a gate structure disposed over the channel region and extending perpendicular to the fins. An ILD layer is disposed over the structure. The ILD layer includes a plurality of TS trenches disposed over the 1.sup.st and 2.sup.nd S/D regions. The TS tranches extend parallel to the gate structures. A plurality of storage capacitors are disposed within the TS trenches. The storage capacitors include a 1.sup.st metal terminal electrically connected to one of the 1.sup.st and 2.sup.nd S/D regions, a 2.sup.nd metal terminal and a capacitor dielectric disposed therebetween. Each transistor is electrically connected to a single storage capacitor to form an eDRAM cell.
Claims
1. A semiconductor structure comprising: an array of fins extending horizontally across a substrate; a plurality of transistors embedded in the fins, the transistors including a 1.sup.st S/D region and a 2.sup.nd S/D region defining a channel region therebetween, the transistors having a gate structure disposed over the channel region and extending perpendicular to the fins; an ILD layer disposed over the structure, the ILD layer including a plurality of TS trenches disposed over the 1.sup.st and 2.sup.nd S/D regions and extending parallel to the gate structures, and a plurality of storage capacitors disposed within the TS trenches, the storage capacitors including a 1.sup.st terminal electrically connected to one of the 1.sup.st and 2.sup.nd S/D regions, a 2.sup.nd terminal and a capacitor dielectric disposed therebetween; and wherein each transistor is electrically connected to a single storage capacitor to form an eDRAIVI cell.
2. The semiconductor structure of claim 1 comprising: a plurality of TS structures alternatingly disposed in parallel with the storage capacitors, each TS structure connected to the other of the 1.sup.st and 2.sup.nd S/D regions of the transistors.
3. The semiconductor structure of claim 2 comprising: each TS structure being electrically connected to a bit line through an interconnect system of the semiconductor structure; each TS structure being electrically connected to the other of the 1.sup.st and 2.sup.nd S/D regions of a pair of transistors to form a pair of eDRAM cells; and each pair of eDRAIVI cells being bordered by an isolation region.
4. The semiconductor structure of claim 1 wherein: each gate structure is electrically connected to a word line through an interconnect system of the semiconductor structure; each storage capacitor is connected to a system ground though the interconnect system; and each of the other 1.sup.st and 2.sup.nd S/D regions of the transistor are connected to a bit line through the interconnect system.
5. The semiconductor structure of claim 1 wherein the storage capacitor includes: the 1.sup.st terminal being formed from a metal liner layer, the metal liner layer being disposed upon and electrically connected to the one of the 1.sup.st and 2.sup.nd S/D regions of the transistors; and the 2.sup.nd terminal being formed from TS conduction metal, the TS conduction metal being electrically connected to the system ground.
6. The semiconductor structure of claim 5 wherein the metal liner layer conformally coats an entire perimeter of the TS trenches.
7. The semiconductor structure of claim 1 wherein the storage capacitor includes: the 1.sup.st terminal being formed from TS conduction metal, the TS conduction metal being disposed upon and electrically connected to a silicide layer at the bottom of the TS trenches, the silicide layer electrically connected to the one of the 1.sup.st and 2.sup.nd S/D regions of the transistors; and the 2.sup.nd terminal being formed from a CA capacitor contact disposed within a CA capacitor opening, the CA capacitor opening extending into the TS conduction metal within the TS trenches, the CA capacitor contact being electrically connected to the system ground.
8. The semiconductor structure of claim 7 wherein the CA capacitor contact is a plurality of CA capacitor contacts within the TS trenches.
9. The semiconductor structure of claim 7 comprising the 1.sup.st terminal also being formed from a silicide metal layer disposed against sidewalls of the TS trenches.
10. The semiconductor structure of claim 7 comprising: portions of the capacitor dielectric being disposed over sidewalls of the CA capacitor openings and over top surfaces of the TS trenches; and the CA capacitor contact filling the CA openings and covering the portions of the capacitor dielectric that are disposed on the top surfaces of the TS trenches.
11. A method comprising: providing a semiconductor structure, the structure including: an array of fins extending horizontally across a substrate, a plurality of transistors embedded in the fins, the transistors including a 1.sup.st S/D region and a 2.sup.nd S/D region defining a channel region therebetween, the transistors having a gate structure disposed over the channel region and extending perpendicular to the fins, and an ILD layer disposed over the structure, the ILD layer including a plurality of TS trenches disposed over the 1.sup.st and 2.sup.nd S/D regions and extending parallel with the gate structures, and forming a plurality of storage capacitors within the TS trenches, the storage capacitors including a 1.sup.st terminal, a 2.sup.nd terminal and a capacitor dielectric disposed therebetween; and electrically connecting one of the 1.sup.st and 2.sup.nd S/D regions of each transistor to the 1st terminal of a single storage capacitor to form a plurality of eDRAM cells.
12. The method of claim 11 comprising: forming a plurality of TS structures alternatingly disposed in parallel with the storage capacitors; electrically connecting each TS structure to the other of the 1.sup.st and 2.sup.nd S/D regions of each transistor.
13. The method of claim 12 comprising electrically connecting a TS structure to the other of the 1.sup.st and 2.sup.nd S/D regions of a pair of transistors to form a pair of eDRAM cells.
14. The method of claim 12 comprising: electrically connecting the 2.sup.nd terminals of the storage capacitors to a system ground through an interconnect system of the semiconductor structure; electrically connecting the TS structures to a bit line through the interconnect system; and electrically connecting the gate structures to a word line through the interconnect system.
15. The method of claim 11 comprising: disposing a metal liner layer over the semiconductor structure, including the TS trenches; disposing an ODL layer over the structure to fill the TS trenches; recessing back the ODL layer to expose the metal liner layer on the upper portion of the TS trenches; removing the exposed portions of the metal liner layer to form the 1.sup.st terminal; removing the ODL layer to expose the metal liner layer within the TS trenches; disposing a capacitor dielectric layer over the structure, including the exposed metal liner layer, to form the capacitor dielectric; disposing TS conduction metal over the semiconductor structure to fill the TS trenches; and planarizing the TS conduction metal to form the 2.sup.nd terminal.
16. The method of claim 15 comprising: disposing a block mask layer over the semiconductor structure to expose only select TS trenches utilized to form TS structures; removing the dielectric layer and metal liner layer from the select TS trenches to expose the S/D regions at the bottom of the select TS trenches; forming a silicide layer over the exposed S/D regions; and removing the block mask layer; wherein disposing and planarizing the TS conduction metal forms the 2.sup.nd terminals and the TS structures.
17. The method of claim 11 comprising: disposing a silicide metal layer over the perimeter of the TS trenches; heating the silicide metal layer to form a silicide layer at the bottom of the TS trenches; disposing TS conduction metal over the semiconductor structure to fill the TS trenches; planarizing the TS conduction metal down to the top level of the ILD layer; disposing a first litho stack over the semiconductor structure; patterning a plurality of CA capacitor openings disposed over a first set of TS trenches utilized to form the storage capacitors; etching the CA capacitor openings into the TS conduction metal disposed in the first set of TS trenches; conformally coating the CA capacitor openings in the first set of TS trenches with a capacitor dielectric layer; and filing the CA capacitor openings with CA capacitor contacts.
18. The method of claim 17 comprising: disposing a second litho stack over the semiconductor structure; patterning a plurality of CA contact openings disposed over a second set of TS trenches utilized to form TS structures; etching the CA contact openings down into the second litho stack to land on top surfaces of the second set of TS trenches; and filling the CA contact openings with CA contacts.
19. The method of claim 17 wherein etching the CA capacitor openings into the TS conduction metal comprises not etching the silicide metal layer disposed on sidewalls of the TS trenches.
20. The method of claim 17 comprising: disposing portions of the capacitor dielectric layer over top surfaces of the TS trenches; and covering the portions of the capacitor dielectric layer disposed on the top surfaces of the TS trenches with CA contacts.
Description
DRAWINGS
(1) The invention will be more fully understood from the following detailed description taken in conjunction with the accompanying drawings, in which:
(2)
(3)
(4)
(5)
(6)
(7)
(8)
(9)
(10)
(11)
(12)
(13)
(14)
(15)
(16)
(17)
(18)
(19)
(20)
(21)
(22)
(23)
(24)
(25)
(26)
DETAILED DESCRIPTION
(27) Certain exemplary embodiments will now be described to provide an overall understanding of the principles of the structure, function, manufacture, and use of the methods, systems, and devices disclosed herein. One or more examples of these embodiments are illustrated in the accompanying drawings. Those skilled in the art will understand that the methods, systems, and devices specifically described herein and illustrated in the accompanying drawings are non-limiting exemplary embodiments and that the scope of the present invention is defined solely by the claims. The features illustrated or described in connection with one exemplary embodiment may be combined with the features of other embodiments. Such modifications and variations are intended to be included within the scope of the present invention.
(28)
(29) Referring to
(30) In operation, a voltage is applied to the word line 34 to activate the gate 22 and turn on the channel 20, therefore electrically connecting the storage capacitor 14 to the bit line 32. This allows high and low logic voltage levels (bits) to be transferred between the storage capacitor 14 and the bit line 32.
(31) DRAM cells 10 are typically arranged in logic arrays (not shown) of rows and columns that can range from a plurality of a few cells to many thousands of cells in an array. The rows of the arrays generally function as the word lines 34, which may be electrically connected in common with many gates 22 of many transistors 22. The columns of the arrays generally function as the bit lines 32, which may be electrically connected in common with many sources 16 of many transistors 22. However, each DRAM cell 10 has only one storage (or cell) capacitor 14 associated with it and not connected in common with either word line 34 or bit line 32. It is these logic arrays that make up the main memory of many electronic devices such as computers or similar.
(32) Referring to
(33) Referring more specifically to
(34) Though this specific exemplary embodiment illustrates the fins 102 formed on a bulk substrate structure 104, one skilled in the art would recognize that the fins could also be formed on a silicon-on-insulator (SOI) structure as well. In that case, the fins would be formed from a silicon layer on top of a buried oxide (BOX) layer. The BOX layer would be disposed on the substrate 104.
(35) Extending perpendicular to the fins 102 are a plurality of trench-silicide (TS) structures 108 and storage capacitors structures 110, both of which are disposed in a plurality of trench-silicide (TS) trenches 111. Also extending perpendicular to the fins 102 are a plurality of gate structures 112, which are alternatingly disposed in parallel with the TS trenches 111.
(36) Also extending perpendicular to the fins 102 are a plurality of dummy gate structures 113. As will be explained in greater detail herein, the dummy gates 113 are disposed above isolations regions 115 (best seen in
(37) Disposed over the semiconductor structure 100 is an interlayer dielectric (ILD) layer 114, which may be an oxide layer. From this top cross-sectional view, the ILD layer 114 covers the fins 102, gates 112 and dummy gates 113, so they are illustrated in dotted line format. Also from this top view, the TS structures 108 and storage capacitor structures 110 extend vertically through the ILD layer 114 so their top surfaces are illustrated in solid line format.
(38) In this embodiment, there is one TS structure 108 repetitively alternating with a pair of adjacent capacitors 110 across the entire semiconductor structure 100. Bordering the adjacent capacitors 110 are a pair of dummy gates 113A, 113B (collectively 113) that are utilized to provide electrical isolation. As will be discussed in greater detail herein, the pairing of two adjacent capacitors 110 with one TS structure 108 is done to save valuable space on the structure 100. However, this pairing is not necessary to form an eDRAIVI cell 116 (best seen in
(39) Referring more specifically to
(40) Just as schematically illustrated in the DRAM cell diagram 10 of
(41) In this embodiment, for the particular transistor 118 illustrated in
(42) The transistor 118, also includes the gate structure 112. Gate 112 is disposed over the channel 124 and is operable to control electrical continuity through the channel 124 and between the S/D regions 120, 122. The gate 112 includes gate metal (or gate metal stack) 126 disposed between a pair of gate spacers 128.
(43) The gate spacers 128 are composed of a dielectric material such as SiN, SiBCN or similar. The gate metal 126 is typically a stack of gate metals, metal nitride, and high-k dielectric, which generally includes three main groups of structures (not shown). Those three main structures are: the gate dielectric layers (typically a high-k dielectric material), the work-function metal structures (typically TiN, TaN, TiCAl, other metal-nitrides or similar materials) and the gate electrode metal (typically Al, W, Cu or similar metal). The gate dielectric layers are used to electrically insulate the work-function metal structures and the gate electrodes from the substrate. The work-function metal structures are generally metal-nitrides that provide the work-function needed for proper FinFET operation, but typically have 10 to 100 times larger resistivity than the gate electrodes. The gate electrodes are metals with a very low resistivity.
(44) The storage capacitor 110 of the eDRAM cell 116 is disposed in one of the TS trenches 111, which were formed earlier in the process flow of structure 100.
(45) The storage capacitor 110 includes a first metal (1st) terminal 132 a second metal (2nd) terminal 134 and a capacitor dielectric 136, which combined form a metal-insulator-metal (MIM) type capacitor. The 1st terminal 132 is disposed as a metal liner and may be composed of such materials as titanium nitride (TiN), titanium (Ti), nickel (Ni), tungsten (W) or similar. The 2nd terminal 134 is formed from the same TS metallization process utilized to form the TS structures 108 disposed in their respective TS trenches 111 and therefore is composed of the same conduction metal 140. Such conduction metal 140 may be titanium nitride TiN, tantalum nitride (TaN) and/or such bulk conducting materials as tungsten (W), cobalt (Co), ruthenium (Ru) or similar. The capacitor dielectric 136 may be a high-k dielectric liner, which may be composed of such material as hafnium dioxide (HfO2), nitride hafnium silicates (HfSiON) or the like.
(46) The TS structure 108 is also disposed in a TS trench 111 and formed with the same material as the 2nd terminal 134 of capacitor 110. TS structure 108 is disposed upon and electrically connected with the 2nd S/D region 122 of transistor 118. The TS structures (or TS layers) 108 are disposed over the S/D regions 120, 122 by a process of TS metallization. The TS metallization process may include formation of a bottom silicide layer 138 over the S/D regions 120, 122 followed by deposition of a top TS conduction metal layer 140. The bottom silicide layer 138 may be composed of NiSi, TiSi2, CoSi2, WSi2 or other like silicides. The TS conduction metal layer 140 may be composed of TiN, TaN and bulk conducting materials such as W, Co, Ru or similar. The TS structures 108 extend longitudinally across the active region of the semiconductor structure 100 in order to ensure proper electrical contact with the S/D regions 120, 122 in the array of fins 102 even under worst case misalignment conditions.
(47) The 1st terminal 132 of capacitor 110 is disposed upon and electrically connected to the 1st S/D region 120 of transistor 118. The 2nd terminal 134 of capacitor 110 is connected to a source/drain (CA) contact 142. CA contact (or CA contact metal) 142 is a part of an interconnect system of structure 100, which connects the 2nd terminal of capacitor 110 to a system ground 30 per schematic diagram 10.
(48) The gate metal 126 of the gate structure 112 of transistor 124 is electrically connected to a gate (CB) contact (or CB contact metal) 144, which is also part of the interconnect system of structure 100 and which connects the gate structure 112 to a word line 34 in similar fashion as that of schematic diagram 10. Additionally, the TS structure 108 is electrically connected to another CA contact 146, which is also a part of the interconnect system of structure 100 and which connects to a bit line 32. Therefore the 2nd S/D region 122 is also electrically connected through TS structure 108 to the system bit line 32 in accordance with schematic 10.
(49) In operation, a voltage is applied to the word line 34 to activate the gate 112 and turn on the channel 124, therefore electrically connecting the storage capacitor 110 to the bit line 32 through transistor 118. This allows high and low logic voltage levels (bits) to be transferred between the storage capacitor 110 and the bit line 32.
(50) Unlike prior art eDRAM cells 116, the capacitor 110 is a MIM capacitor formed within a TS trench 111 of structure 100. Advantageously, with the storage capacitor 110 disposed within the TS trench 111, the eDRAM cell 116 is much more compact than prior art eDRAM cells. Further, as will be discussed in more detail herein, the flow process for manufacturing such cells is simplified relative to prior art eDRAM cells.
(51) Additionally, the process temperature ranges utilized to form the TS structures 108 and the capacitors 110 are significantly lower than the process temperature ranges utilized in the formation of prior art capacitors for prior art eDRAM cells. Typically the process temperature for formation of the TS structures 108 and capacitors 110 range less than 450 degrees centigrade (C).
(52) Referring more specifically to
(53) Referring more specifically to
(54) However, the transistors 118A and 118B each must be connected to their own unique capacitor 110A and 110B respectively in order to function as two separate eDRAM cells 116A and 116B. In this manner (i.e. 2 adjacent eDRAM cells sharing one TS structure 108), further space can be saved over the structure 100.
(55) It is important to also note that the eDRAM pair 116A and 116B are bounded by a pair of dummy gates 113A and 113B (collectively the dummy gates 113), which are disposed over a pair of isolation regions 115A and 115B (collectively the isolation regions 115) The combination of the dummy gates disposed over isolation regions 115 function as single diffusion breaks (SDBs) that isolate the storage capacitors 110A, 110B from other storage capacitors 110 in adjacent eDRAM cell pairs (only partially shown).
(56) Though this embodiment illustrates an SDB, one skilled in the art would recognize that the eDRAM cells 116A, 116B could be separated with other types of isolation regions. For example, the eDRAM cells could be isolated with a well-known double diffusion break (DDB) where the isolation regions 115 would typically be much larger in volume than in an SDB.
(57)
(58) Referring to
(59) Referring to
(60) Referring to
(61) The ODL 150 is then recessed (or etched) back to expose the metal liner layer 148 disposed over the top surface of the IDL layer 114 as well as the metal liner layer 148 disposed over the upper portion (e.g., top half or top eighth) of the trenches 111. This may be done using a reactive ion etching (RIE) process or similar.
(62) Referring to
(63) Referring to
(64) Referring to
(65) Once the block mask layer 152 is applied, the high k dielectric layer 136 and the metal liner 1st terminal layer 132 are removed from the exposed TS trenches 111. This may be done by a series of wet or dry etching processes or similar. By removing the dielectric layer 136 and metal liner 1st terminal layer 132 from the selected TS trenches 111, the S/D areas 122, 120 are exposed at the bottoms of the selected TS trenches.
(66) Referring to
(67) The silicide formation may be done by first disposing a silicide metal layer (not shown) over the structure 100 and annealing the silicide metal layer to form the silicide layer 138 over the S/D regions 120 that are in direct contact with the silicide metal layer 138. The silicide metal layer may be composed of such metals as titanium (Ti), nickel (Ni), cobalt (Co), tungsten (W) or the like. The silicide layer 138, which has been formed by the heat induced reaction of the silicide metal layer (not shown) with the silicon (Si) in the S/D regions 120, may be composed of TiSi2, NiSi, CoSi2, WSi2 or similar. In this embodiment, the un-reacted silicide metal layer is then removed, leaving only the silicide layer 138 disposed at the bottom of the TS trenches 111.
(68) Please note that in this specific embodiment, the silicide layer 138 is shown exclusively over the 2nd S/D regions 122 that are located under the TS structures 108 (best seen in
(69) Next the TS conduction metal is disposed over the entire structure to fill all of the TS trenches 111 and planarized down to complete the formation of the eDRAIVI cells 116. The conduction metal 140 disposed in the TS trenches 111 utilized to form the capacitors 110 becomes the 2nd terminal 134 of those capacitors. The conduction metal 140 disposed in the TS trenches 111 utilized to form the TS structure 108 forms an ohmic contact with the S/D regions 122 through the silicide layer 138 to complete the formation of the TS structures 108. (Note: in this flow, only the bottom of TS structures has silicides 138).
(70) Note also that in this embodiment, there are two fully functioning transistors 118A and 118B that share the same TS structure 108. That is single TS structure 108 is electrically connected to the S/D regions 122 of a pair of transistors 118A, 118B to form a pair of eDRAM cells 116A, 116B.
(71) However, the transistors 118A and 118B each must be connected to their own unique capacitor 110A and 110B respectively in order to function as two separate eDRAM cells 116A and 116B. In this manner (i.e. 2 adjacent eDRAM cells sharing one TS structure 108), further space can be saved over the structure 100.
(72) Referring to
(73) Referring to
(74) More specifically, the capacitors 110 of eDRAM cells 116 included a 1st terminal 132 formed from a metal liner layer 148 and a 2nd terminal 134 formed from the TS conduction metal 140 (best seen in
(75) Herein, normal CA contacts 226 are those contacts that perform the conventional function of electrically activating the S/D regions 122 through the TS structures 202. By contrast herein, CA capacitor contacts 222 are similar to CA contacts 226 in material composition and structure, but connect to capacitor 204 and function herein as the 2nd metal terminal 222 of the MIM capacitor 204. As such, the CA contacts 226 land on TS structures 202 only and the CA capacitor contacts 222 land on storage capacitors 204 only.
(76) At the stage of the process flow illustrated in
(77) However, the partially formed TS structures 202 and storage capacitors 204 within the TS trenches 111 are already of different construction than that of the previously discussed TS structures 108 and storage capacitors 110. More specifically, the TS trenches 111 have been lined with a silicide metal liner layer 206, which has been annealed to form silicide layers 208 in the bottoms of each of the trenches 111.
(78) The silicide metal liner layer 206 may be composed of such metals as Ti, Ni, Co, W or the like. The silicide layer 208, which has been formed from the heat induced reaction of the metal liner 206 with the silicone (Si) in the S/D regions 120, 122, may be composed of TiSi2, NiSi, CoSi2, WSi2 or other similar silicides. Contrary to the previous embodiment of structure 100, the silicide metal liner 206 of structure 200 has not been removed.
(79) TS conduction metal 210 has also been disposed in all of the TS trenches 111 and planarized down to the top level of the ILD layer 114. The TS conduction metal 210 is substantially the same composition as the TS conduction metal 140 in semiconductor structure 100.
(80) Referring to
(81) Referring more specifically to
(82) Once the 1st litho stack 214 is disposed over structure 200, a plurality of CA capacitor openings 218 are patterned into the 1st resist layer 212 through well-known lithographic techniques. The CA capacitor openings 218 are disposed in predetermined positions where CA capacitor contacts 222 (best seen in
(83) Referring more specifically to
(84) Referring more specifically to
(85) Referring to
(86) The CA photo resist layer 212 can be removed by a wet etch or similar process. Additionally, one or more of the layers in stack 216 may also be removed through well-known processes. For purposes of this embodiment, it will be assumed that only a bottom oxide layer now composes the remainder of stack 216.
(87) Referring to
(88) Next a capacitor dielectric layer 220 is disposed over the CA capacitor openings 218 to conformally coat the CA capacitor openings 218 in the TS trench 111. The capacitor dielectric 220 is a high-k dielectric liner layer, which may be composed of such material as hafnium dioxide (HfO2), nitride hafnium silicates (HfSiON) or the like.
(89) Referring to
(90) At this point in the process flow, the storage capacitor 204 is now fully formed. The capacitor 204 is a MIM capacitor 204 disposed within the TS trench 111. The capacitor 204 includes the TS conduction metal 210 as a 1st terminal, the CA capacitor contact metal 222 as a 2nd terminal and the capacitor dielectric 220 therebetween. Moreover, the plurality of CA capacitor openings 218 that are conformally coated with the capacitor dielectric 220 is advantageous because they can potentially increase the overall area of the storage capacitor 204 relative to the storage capacitor 110 in the previous embodiment.
(91) Referring to
(92) Herein, normal CA contacts 226 are those contacts that perform the conventional function of electrically activating the S/D regions 122 through the TS structures 202. To form such CA contacts 226 a second litho stack (not shown) is disposed over the structure 200. CA contact openings 224 are then patterned into the stack 216 and disposed over the TS structure 202.
(93) Referring to
(94) Referring to
(95) Referring more specifically to
(96) This capacitor area can be significantly affected during the etching process of the CA openings 218 (best seen in
(97) Comparing
(98) By contrast, the capacitor area in
(99) Referring to
(100) Referring to
(101) Although the invention has been described by reference to specific embodiments, it should be understood that numerous changes may be made within the spirit and scope of the inventive concepts described. Accordingly, it is intended that the invention not be limited to the described embodiments, but that it have the full scope defined by the language of the following claims.