The present invention relates to memory cells and, particularly, to embodiments of an eight-transistor (8T) static random access memory (SRAM) cell with high device density and symmetry.
A conventional six-transistor (6T) static random access memory (SRAM) cell includes: two n-type pass-gate transistors (PG1 and PG2), two n-type pull-down transistors (PD1 and PD2), and two p-type pull-up transistors (PU1 and PU2). Area scaling of the 6T SRAM cell is limited due to the need to pattern four semiconductor bodies (i.e., one for PG1 and PD1, one for PU1, one for PU2, and one for PD2 and PG2) and further due to the need for L-shaped local interconnects to provide electrical connections required for storage node creation. Recently, a high performance eight-transistor (8T) SRAM cell has been developed. In the 8T SRAM cell, PG1 and PG2 are replaced by two transmission gates (TG1 and TG2). Each transmission gate includes two transistors and, particularly, parallel-connected n-type and p-type pass-gate transistors with gates connected to two different wordlines, respectively, where the signals on the different wordlines are inverted with respect to each other. Replacing PG1 and PG2 with TG1 and TG2 reduces leakage current and delay and improves data stability but increases chip area consumption.
Generally, disclosed herein are embodiments of a memory structure including a substrate and, on the substrate, an array of memory cells arranged in columns and rows. Each memory cell can include a first semiconductor body and a second semiconductor body, which is adjacent and parallel to the first semiconductor body. An isolation region can be positioned laterally between the first semiconductor body and the second semiconductor body. The memory cell can further include four gate structures, which traverse the first semiconductor body, the isolation region, and the second semiconductor body. The memory cell can further include eight transistors and, particularly, four p-type transistors on the first semiconductor body and four n-type transistors on the second semiconductor body. Each of the four gate structures can be common to one of the p-type transistors and to adjacent an adjacent one of the n-type transistors. That is, each gate structure can have a first section and, particularly, a gate for a p-type transistor on the first semiconductor body and a second section and, particularly, a gate for an n-type transistor on the second semiconductor body.
More specifically, disclosed herein are embodiments of a memory structure including a substrate and, on the substrate, an array of eight-transistor (8T) static random access memory (SRAM) cells. Each 8T SRAM cell can include a first semiconductor body and a second semiconductor body, which is adjacent and parallel to the first semiconductor body. An isolation region can be positioned laterally between the first semiconductor body and the second semiconductor body. The 8T SRAM cell can further include four gate structures, which are perpendicular to and traverse the first semiconductor body, the isolation region, and the second semiconductor body. The 8T SRAM cell can further include eight transistors and, particularly, four p-type transistors on the first semiconductor body and four n-type transistors on the second semiconductor body. The four p-type transistors can include, in order across the first semiconductor body, a first p-type pass-gate transistor, a first p-type pull-up transistor adjacent to the first p-type pass-gate transistor, a second p-type pull-up transistor adjacent to the first p-type pull-up transistor, and a second p-type pass-gate transistor adjacent to the second p-type pull-up transistor. The four n-type transistors can include, in order across the second semiconductor body, a first n-type pass-gate transistor, a first n-type pull-down transistor adjacent to the first n-type pass-gate transistor, a second n-type pull-down transistor adjacent to the first n-type pull-down transistor, and a second n-type pass-gate transistor adjacent to the second n-type pull-down transistor. Each of the four gate structures can be common to one of the p-type transistors and to adjacent an adjacent one of the n-type transistors. That is, each gate structure can have a first section and, particularly, a gate for a p-type transistor on the first semiconductor body and a second section and, particularly, a gate for an n-type transistor on the second semiconductor body.
Within the 8T SRAM cell, a common p-type source/drain region of the first p-type pass-gate transistor and the first p-type pull-up transistor can be electrically connected (e.g., by a silicide bridge or a contact strap) to common n-type source/drain region of the first n-type pass-gate transistor and the first n-type pull-down transistor to form a first storage node. The first storage node can further be electrically connected to a common gate structure of the second p-type pull-up transistor and the second n-type pull-down transistor. Additionally, a common p-type source/drain region of the second p-type pass-gate transistor and the second p-type pull-up transistor can be electrically connected (e.g., by a silicide bride or a contact strap) to a common n-type source/drain region of the second n-type pass-gate transistor and the second n-type pull-down transistor to form a second storage node. The second storage node can further be electrically connected to a common gate structure of the first p-type pull-up transistor and the first n-type pull-down transistor.
The present invention will be better understood from the following detailed description with reference to the drawings, which are not necessarily drawn to scale and in which:
In view of the foregoing, disclosed herein are embodiments of a memory structure including a memory cell and, particularly, an eight-transistor (8T) static random access memory (SRAM) cell with high device density and symmetry. The layout of the 8T SRAM cell in particular avoids a significant increase in area consumption over a conventional 6T SRAM cell despite the added transistors for transmission gates. Specifically, in the 8T SRAM cell, two elongated and parallel semiconductor bodies can be separated by an isolation region. Four gate structures can traverse the semiconductor bodies and the isolation region between them. Four p-type transistors (pFETs), including two p-type pass-gate transistors and two pull-up transistors between the two p-type pass-gate transistors, can be on one semiconductor body. Four n-type transistors (nFETs), including two n-type pass-gate transistors and two pull-down transistors between the two n-type pass-gate transistors, can be on the other semiconductor body. Each pair of adjacent p-type and n-type transistors can share a corresponding one of the gate structures (i.e., can have a common gate structure). As described in greater detail below, various interconnects (including, but not limited to, silicide bridges and/or contact straps) provide the required internal and external electrical connections for making the 8T SRAM cell operational and for incorporating the 8T SRAM cell into an array of such cells.
The 8T SRAM cell 300 can operate in three different stages: standby, write and read. In the standby state, the cell 300 is idle. In the write stage, a data value of “1” or “0” is written onto SN1. Specifically, if a data value of “1” (i.e., a high data value) is to be written onto SN1315, a high voltage level (e.g., VDD) is applied to BL1 and BL2 is discharged to ground. Then, WL is activated (e.g., driven to VDD) and WLB is discharged to ground, thereby concurrently turning on nPG1311 and pPG1312 of TG1 and nPG2321 and pPG2322 of TG2. As a result, current will only flow from BL1 (which is high) through TG1 to SN1315, thereby storing a “1” on SN1315. The high voltage level on SN1315 will turn off PU2323 and turn on PD2324, thereby ensuring SN2325 is maintained at a low voltage level. The low voltage level on SN2325 will turn on PU1313 and turn off PD1314, thereby ensuring that SN1315 is maintained at the high voltage level. Contrarily, if a “0” (i.e., a low data value) is to be written to the SN1315, BL1 is discharged to ground and a high voltage level (e.g., VDD) is applied to BL2. Then, WL is activated (e.g., driven to VDD) and WLB is discharged to ground, thereby concurrently turning on nPG1311 and pPG1312 of TG1 and nPG2321 and pPG2322 of TG2. As a result, current will only flow from BL2 (which is high) through TG2 to SN2325, thereby causing SN2325 to go high. The high voltage level on SN2325 will turn off PU1313 and turn on PD1314, thereby ensuring SN1315 is maintained at a low voltage level (i.e., stores a “0”). The low voltage level on SN1315 will turn on PU2323 and turn off PD2324, thereby ensuring that SN2325 is maintained at the high voltage level. In the reading stage, the data value stored at SN1315 is read out. Specifically, BL1 and BL2 are both pre-charged to high voltage levels. Next, WL is activated (e.g., driven to VDD), and WLB is discharged to ground, thereby concurrently turning on TG1 and TG2. When a data value of “1” is stored in the 8T SRAM cell 300, SN1315 is at a high voltage level and SN2325 is at a low voltage level. As a result, BL1 will remain charged at its pre-charged high voltage level, but the voltage level on BL2 will drop through TG2320 and PD2324. When a data value of “0” is stored in the 8T SRAM cell 300, SN1315 is at a low voltage level and SN2325 is at a high voltage level. As a result, BL2 will remain charged at its pre-charged high voltage level, but the voltage level on BL1 will drop through TG1310 and PD1314. A sense amplifier can sense whether BL1 or BL2 is higher after the read and thereby will sense the stored data value.
As mentioned above, replacing the pass gate transistors of a 6T SRAM cell (e.g., as shown in
More specifically, disclosed herein are embodiments of a memory structure 400 (see
It should be noted that, for purposes of illustration,
Each 8T SRAM cell 300 can include a first semiconductor body 303.1 and a second semiconductor body 303.2, which is parallel to the first semiconductor body 303.1. The first and second semiconductor bodies 303.1 and 303.2 can be elongated essentially rectangular-shaped semiconductor bodies. In some embodiments, an isolation region 308 between the two semiconductor bodies 303.1 and 303.2 can include a combination of shallow trench isolation (STI) regions 306 and semiconductor extensions 303.3 extending laterally between and perpendicular to adjacent sections of the first and second semiconductor bodies 303.1 and 303.2 (see detailed discussion below and the cross-sections DD of
More specifically, the structure can include shallow trench isolation (STI) regions 306. The STI regions 306 can include trenches, which extends vertically through the semiconductor layer 303 to the insulator layer 302 in the case of a semiconductor-on-insulator structure, as illustrated (or, alternatively, which extends vertically into the upper portion of the semiconductor substrate to a buried well region in the case of a bulk semiconductor structure). In some embodiments, the trenches can be patterned so that the first and second semiconductor bodies 303.1 and 303.2 are elongated essentially rectangular-shaped semiconductor bodies with some adjacent sections of the first and second semiconductor bodies being physically separated by trenches and with other adjacent sections of the semiconductor bodies having semiconductor extensions 303.3 extending laterally therebetween (see
It should be noted that, within each row R0-Rn, first semiconductor bodies 303.1 for adjacent 8T SRAM cells can be in end-to-end alignment and physically separated by an STI region 306 and second semiconductor bodies 303.2 for the adjacent 8T SRAM cells can similarly be in end-to-end alignment and physically separated by an STI region 306. End-to-end alignment of semiconductor bodies refers to the idea that elongated semiconductor bodies arranged (e.g., patterned) in a line with the end of one semiconductor body being adjacent to the end of the next semiconductor body in the line, and so on. As mentioned above, within each 8T SRAM cell, the first and second semiconductor bodies 303.1-303.2 are positioned laterally adjacent to each other and physically separated by STI 306. Additionally, it should be noted that, from row to row, placement of the first and second semiconductor bodies 303.1-303.2 alternates. For example, in the first row (R0), the first semiconductor bodies 303.1 can be above the second semiconductor bodies 303.2 (as oriented on the drawing sheet); in the second row (R1), the second semiconductor bodies 303.2 can be above the first semiconductor bodies 303.1 (as oriented on the drawing sheet) such that they are adjacent to the second semiconductor bodies 303.2 of R0; in the third row (R2), the first semiconductor bodies 303.1 can be above the second semiconductor bodies 303.2 (as oriented on the drawing sheet) such that they are adjacent to the first semiconductor bodies 303.1 of R1; and so on. Thus, in each column, the order of semiconductor bodies would be first semiconductor body, two second semiconductor bodies, two first semiconductor bodies, and so on. For purposes of illustration, the 8T SRAM cell 300 shown in detail
Each 8T SRAM cell 300 can further include four gate structures (i.e., a first gate structure (G1), a second gate structure (G2), a third gate structure (G3) and a fourth gate structure (G4)). The four gate structures G1-G4 can be parallel, equally spaced, and essentially perpendicular to and overlaying the first and second semiconductor bodies 303.1-303.2. That is, the four gate structures G1-G4 can traverse the first semiconductor body 303.1, the isolation region 308 (and, particularly, STI region(s) 306 in the isolation region 308), and the second semiconductor body 303.2 and can be immediately adjacent to at least the top surface of the first and second semiconductor bodies 303.1-303.2. Gate sidewall spacers can be positioned laterally adjacent to the sidewalls of each of the four gate structures G1-G4.
Each 8T SRAM cell 300 can include eight field effect transistors (FETs) and, specifically, four p-type field effect transistors (pFETs) on the first semiconductor body 303.1 and four n-type field effect transistors (nFETs) on the second semiconductor body 303.2. Each one of the four gate structures can be common to one of the four pFETs and an adjacent one of the four nFETs. That is, the four pFETs can include first sections of the four gate structures, respectively (i.e., those sections immediately adjacent to the first semiconductor body) and the four nFETs can include second sections of the four gate structures, respectively (i.e., those sections immediately adjacent to the second semiconductor body).
For purposes of illustration, these eight FETs are described below and illustrated in the drawings as being planar FETs, where the top surfaces of the STI regions 306 and the semiconductor bodies 303.1 and 303.2 at the FET channel regions are essentially co-planar and where the four gate structures are only adjacent to the top surfaces of the first and second semiconductor bodies 303.1-303.2. However, it should be understood that the salient aspects of the disclosed memory structure embodiments do not rely on the type of FETs used (e.g., planar or non-planar) but rather on the novel layout of the 8T SRAM cell as well as the various interconnects that provide the internal and external electrical connections required to make the 8T SRAM cell operational. Thus, it should be understood that the embodiments described above and illustrated in the drawings are not intended to be limiting. For example, the same 8T SRAM cell layout with two semiconductor bodies and four gate structures for eight FETs could be employed in alternative embodiments where the eight FETs are non-planar FETs (e.g., fin-type FETs (finFETs), gate-all-around FETs (GAAFETs), etc.).
For example, instead of being essentially planar semiconductor bodies with top surfaces co-planar with the top surface of the isolation region (as illustrated), the first and second semiconductor bodies could be first and second semiconductor fins with top surfaces that are above the level of the top surface of the isolation region. In this case, each of the four gate structures can traverse the semiconductor fins and, particularly, be positioned adjacent to the top surface and the opposing sides of each semiconductor fin such that the eight FETs are finFETs. Alternatively, instead of being essentially planar semiconductor bodies with top surfaces co-planar with the top surface of the isolation region (as illustrated), the first and second semiconductor bodies could be first and second nanosheets or nanowires (or first and second stacks of nanosheets or nanowires) that are above the level of the top surface of the isolation region. In this case, each of the four gate structures can traverse the first and second nanosheets or nanowires and, particularly, can wrap around the top surface, bottom surface and opposing sides of the first and second nanosheets or nanowires such that the eight FETs are GAAFETs.
In any case, the four pFETs can include, in order across the first semiconductor body 303.1, a first p-type pass-gate transistor (pPG1) 312, a first p-type pull-up transistor (PU1) 313, a second p-type pull-up transistor (PU2) 323, and a second p-type pass-gate transistor (pPG2) 322. pPG1312 and PU1313 can be connected in series between a first bitline (BL1) and a first voltage rail (e.g., a positive supply voltage rail at VDD or, alternatively, at VCS). PU2323 and pPG2322 can be connected in series between the first voltage rail 391 and a second bitline (BL2). The four nFETs can include, in order across the second semiconductor body 303.2, a first n-type pass-gate transistor (nPG1) 311, a first n-type pull-down transistor (PD1) 314, a second n-type pull-down transistor (PD2) 324, and a second n-type pass-gate transistor (nPG2) 321. nPG1311 and PD1314 can be connected in series between BL1 and a second voltage rail 392 (e.g., a ground or negative voltage rail at VSS). PD2324 and nPG2321 can be connected in series between the second voltage rail 392 and BL2. Interconnects that provide the electrical connections to the first voltage rail 391, the second voltage rail 392, BL1 and BL2 are described in greater detail below.
As mentioned above, each one of the four gate structures G1-G4 can be common to one of the four pFETs and to an adjacent one of the four nFETs. Specifically, the first and second sections of the first gate structure G1 can be the gates 312g and 311g of pPG1312 and nPG1312, respectively. The first and second sections of the second gate structure G2 can be the gates 313g and 314g of PU1313 and PD1314, respectively. The first and second sections of the third gate structure G3 can be the gates 323g and 324g of PU2323 and PD2324, respectively. Finally, the first and second sections of the fourth gate structure G4 can be the gates 322g and 321g of pPG2323 and nPG2321, respectively.
The first and second sections of each gate structure G1-G4 can include at least a gate dielectric layer and a gate conductor layer on the gate dielectric layer. However, since the first sections and second sections are employed as the gates for pFETs and nFETs, respectively, the gate structures can, optionally, be dual work function gate structure wherein the first and second sections are different (e.g., comprise different materials) for optimal pFET and nFET performance.
More specifically, in some embodiments, each gate structure G1-G4 can include the same materials in the first and second sections including a gate dielectric layer and a gate conductor layer, which is suitable for either pFET or nFET performance.
In some embodiment, each gate structure G1-G4 can include: a gate dielectric layer, such as a silicon dioxide gate dielectric layer, in both the first and second sections; a p-doped polysilicon gate conductor layer in the first section for optimal pFET performance; and an n-doped polysilicon gate conductor layer in the second section for optimal nFET performance.
In still other embodiments, each gate structure G1-G4 can include: a high-K gate dielectric layer in both the first and second sections; one or more work function metal layers in the first section and preselected to achieve an optimal pFET work function; one or more different work function metal layers in the second section and preselected to achieve an optimal nFET work function; and, optionally, conductive fill material on the work function metal layers in both the first section and the second section. Those skilled in the art will recognize that a high-K gate dielectric layer refers to a layer of dielectric material with a dielectric constant that is greater than the dielectric constant of silicon dioxide (i.e., greater than 3.9). Exemplary high-K dielectric materials include, but are not limited to, hafnium (Hf)-based dielectrics (e.g., hafnium oxide, hafnium silicon oxide, hafnium silicon oxynitride, hafnium aluminum oxide, etc.) or other suitable high-k dielectrics (e.g., aluminum oxide, tantalum oxide, zirconium oxide, etc.). Those skilled in the art will further recognize that the optimal work function for a gate conductor of pFET will be, for example, between about 4.9 eV and about 5.2 eV. Exemplary metals (and metal alloys), which have a work function within this range include, but are not limited to, ruthenium, palladium, platinum, cobalt, and nickel, as well as metal oxides (aluminum carbon oxide, aluminum titanium carbon oxide, etc.) and metal nitrides (e.g., titanium nitride, titanium silicon nitride, tantalum silicon nitride, titanium aluminum nitride, tantalum aluminum nitride, etc.). Those skilled in the art will further recognize that the optimal work function for a gate conductor of an nFET will be, for example, between 3.9 eV and about 4.2 eV. Exemplary metals (and metal alloys), which have a work function within this range include, but are not limited to, hafnium, zirconium, titanium, tantalum, aluminum, and alloys thereof, such as, hafnium carbide, zirconium carbide, titanium carbide, tantalum carbide, and aluminum carbide. The conductive fill material layer can be, for example, doped poly silicon or any suitable metal or metal alloy fill material including, but not limited to, tungsten, a tungsten alloy (e.g., tungsten silicide or titanium tungsten), cobalt, or aluminum. Various configurations for dual work function gate structures are known in the art and, thus, the details thereof have been omitted from this specification in order to allow the reader to focus on the salient aspects of the disclosed embodiments.
In any case, within the first gate structure G1, which is the common gate structure of pPG1312 and nPG1311, a gate cut isolation region 380 can electrically isolate the first section from the second section (i.e., the gate 312g of pPG1312 from the gate 311g of nPG1312) (see
Those skilled in the art will recognize that a gate cut isolation region refers to an isolation structure that physically separates and electrically isolates the gate conductor material in adjacent sections of a gate structure in order to electrically isolate those sections. Thus, the gates of pPG1 and nPG1 can be electrically connected to and controlled by two different wordlines (WLB and WL, respectively) and the gates of pPG2 and nPG2 can also be electrically connected to and controlled by the two different wordlines (WLB and WL, respectively), where the signals on the different wordlines are inverted with respect to each other. Depending upon the processing techniques used to form the gate structure, the structure of the cut isolation region may vary. For example, each gate cut isolation region 380 could include ILD material that completely fills the spaces between patterned sections of the gate conductor material that are in end-to-end alignment (as illustrated). Alternatively, each gate cut isolation region 380 could include sidewall spacers positioned laterally adjacent to end walls of the patterned sections of the gate conductor material that are in end-to-end alignment and ILD material between the sidewall spacers. Alternatively, each gate cut isolation region 380 can have any other structure suitable for providing isolation between sections of gate conductor material that are in end-to-end alignment. End-to-end alignment of sections of a gate structure refers to the idea that elongated sections of a gate structure are arranged (e.g., patterned) in a line with the end of one section being adjacent to the end of the next section in the line, and so on.
Within the second gate structure G2, which is the common gate structure of PU1313 and PD1314, the first and second sections are not electrically isolated from each other by a gate cut isolation region and, thus, the gates 313g and 314g of PU1313 and PD1314 are electrically connected (see
It should be noted that the same four gate structures (G1-G4) can extend across the first and second semiconductor bodies of each 8T SRAM cell in the same column. Additional gate cut isolation regions 380 can provide gate-to-gate isolation, as needed. For example, within G2 of column (c), the gates 313g and 314g of PU1313 and PD1314 of 8T SRAM cell 300 can be electrically isolated by gate cut isolation regions 380 from the gates of a pull-up transistor and a pull-down transistor, respectively, of the 8T SRAM cells 300(c:r−1) and 300(c:r+1) in the adjacent rows. Similarly, within G3 of column (c), the gates 323g and 324g of PU2323 and PD2324 of 8T SRAM cell 300 can be electrically isolated by gate cut isolation regions 380 from the gates of a pull-up transistor and a pull-down transistor, respectively, of the 8T SRAM cells 300(c:r−1) and 300(c:r+1) in the adjacent rows. However, within G1, the gate 312g of pPG1 is continuous with the gate of a p-type pass-gate transistor of the 8T SRAM cell 300(c:r−1) and the gate 311g of nPG1 is continuous with the gate of an n-type pass-gate transistor of the 8T SRAM cell 300(c:r+1). Similarly, within G4, the gate 322g of pPG2 is continuous with the gate of a p-type pass-gate transistor of the 8T SRAM cell 300(c:r−1) and the gate 321g of nPG2 is continuous with the gate of an n-type pass-gate transistor of the 8T SRAM cell 300(c:r+1).
The four pFETs can further include: channel regions in the first semiconductor body 303.1 immediately adjacent to the first sections of the gate structures; and source/drain regions in and/or on the first semiconductor body on opposing sides of the first sections of the gate structures (see
The channel regions 312c, 313c, 323c and 322c of the pFETs can be intrinsic (i.e., undoped) or, alternatively, doped so as to have an n-type conductivity at a relatively low conductively level (e.g., n− conductivity). The source/drain regions 312s/d, 313s/d, 323s/d, and 322s/d of the pFETs can be doped so as to have p-type conductivity at a relatively high conductivity level (e.g., p+ conductivity). Additionally, adjacent pFETs and can have a common p-type source/drain region (also referred to herein as a shared source/drain region) between their respective gates. That is, one of the p-type source/drain regions 312s/d of pPG1312 can abut one of the p-type source/drain regions 313s/d of PU1313, thereby forming a common p-type source/drain region of pPG1312 and PU1313; one of the p-type source/drain regions 313s/d of PU1313 can abut one of the p-type source/drain regions 323s/d of PU2323, thereby forming a common p-type source/drain region of PU1313 and PU2323; and one of the p-type source/drain regions 323s/d of PU2323 can abut one of the p-type source/drain regions 322s/d of pPG2322, thereby forming a common p-type source/drain region of PU2323 and pPG2322.
In some embodiments, the source/drain regions 312s/d, 313s/d, 323s/d, and 322s/d of the pFETs can include epitaxial semiconductor material on the first semiconductor body 303.1. The epitaxial semiconductor material can be grown on the top surface of the first semiconductor body 303.1 on either side of the first sections of the gate structures G1-G4 and physically separated therefrom by the gate sidewall spacers. Alternatively, the epitaxial semiconductor material can be grown within source/drain trenches that have been etched into the first semiconductor body 303.1 on either side of the first sections of the gate structures G1-G4 and physically separated therefrom by the gate sidewall spacers. In either case, the epitaxial semiconductor material can be the same semiconductor material used in the first semiconductor body (e.g., silicon) or some other suitable semiconductor material or material(s) (e.g., silicon germanium or layers of silicon germanium and silicon). In other embodiments, the source/drain regions 312s/d, 313s/d, 323s/d, and 322s/d of the pFETs could be dopant implant regions within the first semiconductor body 303.1. In any case, as mentioned above, the p-type source-drain regions can be doped so as to have a p-type conductivity at a relatively high conductivity level (e.g., so as to have p+ conductivity).
The four n-FETs can further include: channel regions in the second semiconductor body 303.2 adjacent to the second sections of the gate structures G1-G4; and source/drain regions in and/or on the second semiconductor body 303.1 on opposing sides of the second sections of the gate structures G1-G4 (see
The channel regions 311c, 314c, 324c and 322c of the nFETs can be intrinsic (i.e., undoped) or, alternatively, doped so as to have a p-type conductivity at a relatively low conductively level (e.g., p− conductivity). The source/drain regions 311s/d, 314s/d, 324s/d, and 321s/d of the nFETs can be doped so as to have n-type conductivity at a relatively high conductivity level (e.g., n+ conductivity). Additionally, adjacent nFETs can have a common n-type source/drain region (also referred to herein as a shared source/drain region) between their respective gates. That is, one of the n-type source/drain regions 311s/d of nPG1311 can abut one of the n-type source/drain regions 314s/d of PD1314, thereby forming a common n-type source/drain region of nPG1311 and PD1314; one of the n-type source/drain regions 314s/d of PD1314 can abut one of the n-type source/drain regions 324s/d of PD2324, thereby forming a common n-type source/drain region of PU1314 and PU2324; and one n-type source/drain regions 324s/d of PD2324 can abut one of the n-type source/drain regions 321s/d of nPG2321, thereby forming a common p-type source/drain region of PD2324 and nPG2321.
In some embodiments, the source/drain regions 311s/d, 314s/d, 324s/d, and 321s/d of the nFETs can include epitaxial semiconductor material on the second semiconductor body 303.2. The epitaxial semiconductor material can be grown on the top surface of the second semiconductor body 303.2 on either side of the second sections of the gate structures G1-G4 and physically separated therefrom by the gate sidewall spacers. Alternatively, the epitaxial semiconductor material can be grown within source/drain trenches that have been etched into the second semiconductor body 303.2 on either side of the second sections of the gate structures G1-G4 and physically separated therefrom by the gate sidewall spacers. In either case, the epitaxial semiconductor material can be the same semiconductor material used in the second semiconductor body (e.g., silicon) or some other suitable semiconductor material or material(s) In other embodiments, the source/drain regions 311s/d, 314s/d, 324s/d, and 321s/d of the nFETs could be dopant implant regions within the second semiconductor body 303.2. In any case, as mentioned above, the n-type source-drain regions can be doped so as to have an n-type conductivity at a relatively high conductivity level (e.g., so as to have n+ conductivity).
Various interconnects, discussed below, can provide external and internal electrical connections to and between the FETs within the 8T SRAM cell in order to make each 8T SRAM cell operational and further between FETs of adjacent 8T SRAM cells within the columns in order to efficiently incorporate the cells into an array.
Specifically, each 8T SRAM cell 300 can include the following internal electrical connections. The common p-type source/drain region of PU1313 and PU2323 (see abutting p-type source/drain regions 313s/d and 323s/d) can be electrically connected to the first voltage rail 391 (e.g., at VDD or VCS). Additionally, the common n-type source/drain region of PD1314 and PD2324 (see abutting n-type source/drain regions 314s/d and 324s/d between the gates 314g and 324g) can be electrically connected to the second voltage rail 392 (e.g., at VSS).
The common p-type source/drain region of pPG1312 and PU1313 (see abutting p-type source/drain regions 312s/d and 313s/d between the gates 312g and 313g) and the common n-type source/drain region of nPG1311 and the PD1314 (see abutting n-type source/drain regions 311s/d and 314s/d between the gates 311g and 314g) can be electrically connected to each other to form a first storage node (SN1) 315 for the 8T SRAM cell 300 and SN1315 can be electrically connected to the third gate structure G3, which is the common gate structure of the PU2323 and PD2324 (i.e., to gates 323g and 324g). Similarly, the common p-type source/drain region of pPG2322 and PU2323 (see abutting p-type source/drain regions 322s/d and 323s/d between the gates 322g and 323g) and the common n-type source/drain region of nPG2321 and the PD2324 (see abutting n-type source/drain regions 321s/d and 324s/d between the gates 321g and 324g) can be electrically connected to each other to form a second storage node (SN2) 325 for the 8T SRAM cell 300 and SN2325 can be electrically connected to the second gate structure G2, which is the common gate structure of PU1313 and the PD1314 (i.e., to gates 313g and 314g).
Additionally, the p-type source/drain region 312s/d of pPG1312 at one end of the first semiconductor body 303.1 can be electrically connected to the adjacent n-type source/drain region 311s/d of nPG1311 at the same end and further electrically connected to BL1. Similarly, the p-type source/drain region 322s/d of pPG2322 at the opposite end of the first semiconductor body 303.1 can be electrically connected to the adjacent n-type source/drain region 321s/d of nPG2321 at the same end and further electrically connected to BL2. Thus, as mentioned above, pPG1312 and PU1313 are electrically connected in series between BL1 and the first voltage rail 391, PU2323 and pPG2322 are electrically connected in series between the first voltage rail 391 and BL2, nPG1311 and PD1314 are electrically connected in series between BL1 and the second voltage rail 392, and PD2324 and nPG2321 are electrically connected in series between the second voltage rail 392 and BL2. As a result, pPG1312 and nPG1311 are also electrically connected in parallel between BL1 and SN1315 and since the gates of pPG1 and nPG1 are controlled by two different wordline signals on WLB and WL, respectively, that are inverted with respect each other, pPG1311 and nPG1312 form a first transmission gate (TG1) that provides access to SN1315. Similarly, pPG2322 and nPG2321 are electrically connected in parallel between BL2 and SN2325 and since the gates of pPG2322 and nPG2321 are also controlled by the same wordline signals on WLB and WL, respectively, pPG2322 and nPG2321 form a second transmission gate (TG2) that provides access to SN2325. Such transmission gates TG1 and TG2 reduce leakage current and delay and improve data stability.
Each 8T SRAM cell 300 can further include the following external electrical connections to adjacent 8T SRAM cell(s) (e.g., to 8T SRAM cells 300(c:r−1) and/or 300(c:r+1) in the adjacent row(s) of the same column. The common p-type source/drain region of PU1313 and PU2323 (see abutting p-type source/drain regions 313s/d and 323s/d) of the 8T SRAM cell 300 can be electrically connected to the common p-type source/drain region of the pull-up transistors of an adjacent 8T SRAM cell (e.g., see adjacent 8T SRAM cell 300(c:r−1)), which is also connected to the first voltage rail 391. Additionally, the common n-type source/drain region of PD1314 and PD2324 (see abutting n-type source/drain regions 314s/d and 324s/d) of the 8T SRAM cell 300 can be electrically connected to the common n-type source/drain region of the pull-down transistors of an adjacent 8T SRAM cell (e.g., see adjacent 8T SRAM cell 300(c:r+1)), which is also connected to the second voltage rail 392. It should be understood that 8T SRAM cells that are in rows between the first row (R0) and the last row (Rn) will be connected to two adjacent 8T SRAM cells, as described. However, 8T SRAM cells in the first row (R0) and the last row (Rn) will only be connected to one adjacent 8T SRAM cell.
The embodiments of the memory structure (e.g., see memory structure 400 of
Referring specifically to
Additional silicide bridge(s) 371.5-371.6 can provide the electrical connections to adjacent 8T SRAM cells within the same column, as necessary. For example, as mentioned above, the common p-type source/drain region of PU1313 and PU2323 (see abutting p-type source/drain regions 313s/d and 323s/d) of the 8T SRAM cell 300 can be electrically connected to the common p-type source/drain region of the pull-up transistors of an adjacent 8T SRAM cell (e.g., see adjacent 8T SRAM cell 300(c:r−1)) by a fifth silicide bridge 371.5. It should be noted that a single contact 397 landing on the common p-type source/drain region of PU1313 and PU2323 of the 8T SRAM cell 300 (as illustrated), on the fifth silicide bridge 371.5, or on the common p-type source/drain region of the pull-up transistors of the adjacent 8T SRAM cell could be employed to provide the connection to the first voltage rail. Furthermore, the common n-type source/drain region of PD1314 and PD2324 (see abutting n-type source/drain regions 314s/d and 324s/d) of the 8T SRAM cell 300 can be electrically connected to the common n-type source/drain region of the pull-down transistors of an adjacent 8T SRAM cell (e.g., see adjacent 8T SRAM cell 300(c:r+1)) by a sixth silicide bridge 371.6. It should be noted that a single contact 397 landing on the common n-type source/drain region of PD1314 and PD2324 of the 8T SRAM cell 300 (as illustrated), on the sixth silicide bridge 371.6, or on the common n-type source/drain region of the pull-down transistors of the adjacent 8T SRAM cell could be employed to provide the connection to the second voltage rail.
For purposes of this disclosure, a silicide bridge refers to a portion of a silicide layer 305, which is immediately adjacent to a pn junction and, thereby electrically connects the semiconductor material on both sides of the junction. In the 8T SRAM cell, each pair of adjacent p-type and n-type source/drain regions can include p-type epitaxial semiconductor material 304.1 on a section of the first semiconductor body 303.1 and n-type epitaxial semiconductor material 304.2 on an adjacent section of the second semiconductor body 303.2.
In some embodiments, for any pair of adjacent p-type and n-type source/drain regions that are required to be electrically connected, semiconductor extensions 303.3 can be perpendicular to and extend between the adjacent sections of the first semiconductor bodies 303.1 and 303.2 below (see the cross-sections DD of
In other embodiments, however, the isolation region 308 includes an STI region 306 that completely separates the first and second semiconductor bodies 303.1 and 303.2. That is, the structure may be devoid of semiconductor extensions between the semiconductor bodies. In these embodiments, for any pair of adjacent p-type and n-type source/drain regions that are required to be electrically connected, the p-type and n-type epitaxial semiconductor materials for the source/drain regions can be grown during processing on the first and second semiconductor bodies 303.1 and 303.2, respectively, and further overgrown so that they extend onto the adjacent STI region 306 and meet, thereby creating the required p-n junctions (see the alternative cross-section DD shown in
In either case, when silicide layers 305 are subsequently formed on the top surfaces of all p-type and n-type source/drain regions including over any pn junctions the silicide bridges 371.1-371.6 are formed. It should be noted that, as mentioned above, the common p-type source/drain region of PU1 and PU2 is not to be electrically connected to the adjacent common n-type source/drain region of PD1 and PD2. Thus, during processing, masking layers prevent overgrowth of the epitaxial semiconductor material onto STI region(s) 306 between these source/drain regions and thereby prevent the formation of a pn junction. As a result, when the silicide layers 305 are formed, the silicide layer on the common p-type source/drain region of PU1 and PU2 is physically separated from the silicide layer on the adjacent common n-type source/drain region of PD1 and PD2 and no silicide bridge is formed (see space 372 filled with dielectric material 390 shown in
Referring specifically to
Additional contact strap(s) 396.5-396.6 can provide the electrical connections to adjacent 8T SRAM cells within the same column, as necessary. For example, as mentioned above, the common p-type source/drain region of PU1313 and PU2323 (see abutting p-type source/drain regions 313s/d and 323s/d) of the 8T SRAM cell 300 can be electrically connected to the common p-type source/drain region of the pull-up transistors of an adjacent 8T SRAM cell (e.g., see adjacent 8T SRAM cell 300(c:r−1)) by a fifth contact strap 396.5. Furthermore, the common n-type source/drain region of PD1314 and PD2324 (see abutting n-type source/drain regions 314s/d and 324s/d) of the 8T SRAM cell 300 can be electrically connected to the common n-type source/drain region of the pull-down transistors of an adjacent 8T SRAM cell (e.g., see adjacent 8T SRAM cell 300(c:r+1)) by a sixth contact stray 396.6.
For purposes of this disclosure, a contact strap (also referred to herein as a contact bar) is a middle of the line (MOL) source/drain contact. However, instead of being a conventional MOL source/drain contact with a circular or square shape having a width that is essentially equal to its length (or depth), a contact strap 396 is elongated with a length that is greater than its width so that it can land across multiple source/drain regions (e.g., on a p-type source/drain region, on the isolation region, and on an adjacent n-type source/drain region, as shown in
Referring specifically to
As in the memory structure 400, in the memory structure 600 additional silicide bridge(s) 371.5-371.6 can provide the electrical connections to adjacent 8T SRAM cells within the same column, as necessary. For example, as mentioned above, the common p-type source/drain region of PU1313 and PU2323 (see abutting p-type source/drain regions 313s/d and 323s/d) of the 8T SRAM cell 300 can be electrically connected to the common p-type source/drain region of the pull-up transistors of an adjacent 8T SRAM cell (e.g., see adjacent 8T SRAM cell 300(c:r−1)) by a fifth silicide bridge 371.5. It should be noted that a single contact 397 landing on the common p-type source/drain region of PU1313 and PU2323 of the 8T SRAM cell 300 (as illustrated), on the fifth silicide bridge 371.5, or on the common p-type source/drain region of the pull-up transistors of the adjacent 8T SRAM cell could be employed to provide the connection to the first voltage rail. Furthermore, the common n-type source/drain region of PD1314 and PD2324 (see abutting n-type source/drain regions 314s/d and 324s/d) of the 8T SRAM cell 300 can be electrically connected to the common n-type source/drain region of the pull-down transistors of an adjacent 8T SRAM cell (e.g., see adjacent 8T SRAM cell 300(c:r+1)) by a sixth silicide bridge 371.6. It should be noted that a single contact 397 landing on the common n-type source/drain region of PD1314 and PD2324 of the 8T SRAM cell 300 (as illustrated), on the sixth silicide bridge 371.6, or on the common n-type source/drain region of the pull-down transistors of the adjacent 8T SRAM cell could be employed to provide the connection to the second voltage rail.
It should be understood that in the embodiments described above a semiconductor material refers to a material whose conducting properties can be altered by doping with an impurity. Exemplary semiconductor materials include, for example, silicon-based semiconductor materials (e.g., silicon, silicon germanium, silicon germanium carbide, silicon carbide, etc.) and III-V compound semiconductors (i.e., compounds obtained by combining group III elements, such as aluminum (Al), gallium (Ga), or indium (In), with group V elements, such as nitrogen (N), phosphorous (P), arsenic (As) or antimony (Sb)) (e.g., GaN, InP, GaAs, or GaP). A pure semiconductor material and, more particularly, a semiconductor material that is not doped with an impurity for the purposes of increasing conductivity (i.e., an undoped semiconductor material) is referred to in the art as an intrinsic semiconductor. A semiconductor material that is doped with an impurity for the purposes of increasing conductivity (i.e., a doped semiconductor material) is referred to in the art as an extrinsic semiconductor and will be more conductive than an intrinsic semiconductor made of the same base material. That is, extrinsic silicon will be more conductive than intrinsic silicon; extrinsic silicon germanium will be more conductive than intrinsic silicon germanium; and so on. Furthermore, it should be understood that different impurities (i.e., different dopants) can be used to achieve different conductivity types (e.g., P-type conductivity and N-type conductivity) and that the dopants may vary depending upon the different semiconductor materials used. For example, a silicon-based semiconductor material (e.g., silicon, silicon germanium, etc.) is typically doped with a Group III dopant, such as boron (B) or indium (In), to achieve P-type conductivity, whereas a silicon-based semiconductor material is typically doped a Group V dopant, such as arsenic (As), phosphorous (P) or antimony (Sb), to achieve N-type conductivity. A gallium nitride (GaN)-based semiconductor material is typically doped with magnesium (Mg) to achieve P-type conductivity and with silicon (Si) or oxygen to achieve N-type conductivity. Those skilled in the art will also recognize that different conductivity levels will depend upon the relative concentration levels of the dopant(s) in a given semiconductor region.
It should be understood that the terminology used herein is for the purpose of describing the disclosed structures and methods and is not intended to be limiting. For example, as used herein, the singular forms “a”, “an” and “the” are intended to include the plural forms as well, unless the context clearly indicates otherwise. Additionally, as used herein, the terms “comprises” “comprising”, “includes” and/or “including” specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof. Furthermore, as used herein, terms such as “right”, “left”, “vertical”, “horizontal”, “top”, “bottom”, “upper”, “lower”, “under”, “below”, “underlying”, “over”, “overlying”, “parallel”, “perpendicular”, etc., are intended to describe relative locations as they are oriented and illustrated in the drawings (unless otherwise indicated) and terms such as “touching”, “in direct contact”, “abutting”, “directly adjacent to”, “immediately adjacent to”, etc., are intended to indicate that at least one element physically contacts another element (without other elements separating the described elements). The term “laterally” is used herein to describe the relative locations of elements and, more particularly, to indicate that an element is positioned to the side of another element as opposed to above or below the other element, as those elements are oriented and illustrated in the drawings. For example, an element that is positioned laterally adjacent to another element will be beside the other element, an element that is positioned laterally immediately adjacent to another element will be directly beside the other element, and an element that laterally surrounds another element will be adjacent to and border the outer sidewalls of the other element. The corresponding structures, materials, acts, and equivalents of all means or step plus function elements in the claims below are intended to include any structure, material, or act for performing the function in combination with other claimed elements as specifically claimed.
The descriptions of the various embodiments of the present invention have been presented for purposes of illustration but are not intended to be exhaustive or limited to the embodiments disclosed. Many modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The terminology used herein was chosen to best explain the principles of the embodiments, the practical application or technical improvement over technologies found in the marketplace, or to enable others of ordinary skill in the art to understand the embodiments disclosed herein.
Therefore, disclosed above are embodiments of a memory structure including a memory cell and, particularly, an eight-transistor (8T) static random access memory (SRAM) cell with high device density and symmetry. The layout of the 8T SRAM cell in particular avoids a significant increase in area consumption over a conventional 6T SRAM cell despite the added transistors for transmission gates. Specifically, in the 8T SRAM cell, two elongated and parallel semiconductor bodies can be separated by an isolation region. Four gate structures can traverse the semiconductor bodies and the isolation region between them. Four p-type transistors (pFETs), including two p-type pass-gate transistors and two pull-up transistors between the two p-type pass-gate transistors, can be on one semiconductor body. Four n-type transistors (nFETs), including two n-type pass-gate transistors and two pull-down transistors between the two n-type pass-gate transistors, can be on the other semiconductor body. Each pair of adjacent p-type and n-type transistors can share a corresponding one of the gate structures (i.e., can have a common gate structure). Various interconnects (including, but not limited to, silicide bridges and/or contact straps) provide the required internal and external electrical connections for making the 8T SRAM cell operational. While a conventional 6T SRAM cell requires two gate structures and the disclosed 8T SRAM cell requires four gate structures, the conventional 6T SRAM cell requires four semiconductor bodies and the disclosed 8T SRAM cell only requires two. Thus, the device density within the memory cell is increased and the difference, if any, in area consumption is minimal. Furthermore, the high degree of symmetry within the cell layout simplifies future size scaling.
Number | Name | Date | Kind |
---|---|---|---|
5831897 | Hodges | Nov 1998 | A |
6341083 | Wong | Jan 2002 | B1 |
7126837 | Banachowicz | Oct 2006 | B1 |
7336533 | Hunter et al. | Feb 2008 | B2 |
8294212 | Wang | Oct 2012 | B2 |
9842634 | Jeong et al. | Dec 2017 | B2 |
9922702 | Shanmugam | Mar 2018 | B1 |
20050111251 | Liaw | May 2005 | A1 |
20050265070 | Liaw | Dec 2005 | A1 |
20060038234 | Liaw | Feb 2006 | A1 |
20070280026 | Hunter et al. | Dec 2007 | A1 |
20080137436 | Salter | Jun 2008 | A1 |
20100315859 | Lee | Dec 2010 | A1 |
20110075504 | Chan | Mar 2011 | A1 |
20200027869 | Yeh | Jan 2020 | A1 |
Entry |
---|
Paul, et al., “Implementation of High Performance of SRAM Cell Using Transmission Gate”, International Journal & Magazine of Engineering, Technology, Management and Research, vol. 4, Issue 2, 2017, pp. 859-863. |
Ramesh et al., “Design of Power Efficient High Performance SRAM Cell using Transmission Gates”, International Journal of VLSI System Design and Communication Systems, vol. 4, Issue 8, 2016, pp. 569-663. |
Sharma et al., “Implementation of high performance SRAM Cell Using Transmission Gate”, IEEE Fifth International Conference on Advanced Computing & Communication Technologies, 2015, pp. 257-260. |
Number | Date | Country | |
---|---|---|---|
20220310629 A1 | Sep 2022 | US |