US20020001219A1 - Discrete devices including eaprom transistor and nvram memory cell with edge defined ferroelectric capacitance, methods for operating same, and apparatus including same - Google Patents
Discrete devices including eaprom transistor and nvram memory cell with edge defined ferroelectric capacitance, methods for operating same, and apparatus including same Download PDFInfo
- Publication number
- US20020001219A1 US20020001219A1 US09/461,009 US46100999A US2002001219A1 US 20020001219 A1 US20020001219 A1 US 20020001219A1 US 46100999 A US46100999 A US 46100999A US 2002001219 A1 US2002001219 A1 US 2002001219A1
- Authority
- US
- United States
- Prior art keywords
- ferroelectric capacitor
- memory cell
- coupled
- recited
- gate
- 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.)
- Granted
Links
Images
Classifications
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C11/00—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor
- G11C11/005—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor comprising combined but independently operative RAM-ROM, RAM-PROM, RAM-EPROM cells
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C11/00—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor
- G11C11/21—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements
- G11C11/22—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements using ferroelectric elements
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C11/00—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor
- G11C11/21—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements
- G11C11/34—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements using semiconductor devices
- G11C11/40—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements using semiconductor devices using transistors
- G11C11/401—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements using semiconductor devices using transistors forming cells needing refreshing or charge regeneration, i.e. dynamic cells
- G11C11/403—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements using semiconductor devices using transistors forming cells needing refreshing or charge regeneration, i.e. dynamic cells with charge regeneration common to a multiplicity of memory cells, i.e. external refresh
- G11C11/404—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements using semiconductor devices using transistors forming cells needing refreshing or charge regeneration, i.e. dynamic cells with charge regeneration common to a multiplicity of memory cells, i.e. external refresh with one charge-transfer gate, e.g. MOS transistor, per cell
-
- H—ELECTRICITY
- H10—SEMICONDUCTOR DEVICES; ELECTRIC SOLID-STATE DEVICES NOT OTHERWISE PROVIDED FOR
- H10B—ELECTRONIC MEMORY DEVICES
- H10B53/00—Ferroelectric RAM [FeRAM] devices comprising ferroelectric memory capacitors
-
- G—PHYSICS
- G11—INFORMATION STORAGE
- G11C—STATIC STORES
- G11C11/00—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor
- G11C11/21—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements
- G11C11/22—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements using ferroelectric elements
- G11C11/223—Digital stores characterised by the use of particular electric or magnetic storage elements; Storage elements therefor using electric elements using ferroelectric elements using MOS with ferroelectric gate insulating film
Definitions
- the present invention relates to discrete elements such as electrically alterable programmable read-only memory (EAPROM) and nonvolatile random access memory (NVRAM) devices. More specifically, the present invention relates to EAPROM and NVRAM cells where data is stored using a ferroelectric capacitor incorporated in the cells. Methods for operating, writing to and reading from, such DRAM, logic devices, memory arrays, and systems containing these cells are also disclosed.
- EAPROM electrically alterable programmable read-only memory
- NVRAM nonvolatile random access memory
- DRAMs Dynamic random access memories
- DRAMS are the most cost-effective, high-speed memory used with present day computers. DRAMS last nearly indefinitely. Moreover, DRAMs are available in very high density configurations, e.g., 64 megabytes (MB). However, DRAMS can only store information for a limited time without constant refreshing and lose all knowledge of their state once power is removed.
- MB 64 megabytes
- Ferroelectric capacitors are currently being used as nonvolatile memory devices, as disclosed in U.S. Pat. No. 4,888,733, which is entitled “Nonvolatile Memory Cell and Sensing Method” and discloses a two transistor, one capacitor (2T/1C) memory cell.
- ferroelectric capacitors are often employed in nonvolatile random access memories (NVRAMs).
- Memory cells having structures approximating that of DRAMs, i.e., arranged in the conventional one transistor, one capacitor (1T/1C) memory cell pattern, are disclosed in U.S. Pat. No. 5,600,587 (Ferroelectric Random Access Memory), U.S. Pat. No.
- FIG. 1( a ) the capacitor in a 1T/1C DRAM structure is replaced by a ferroelectric capacitor, as illustrated in FIG. 1( a ).
- the memory cell 1 of FIG. 1( a ) includes a wordline 10 , a digitline 12 , a transistor 14 and a ferroelectric capacitor 16 .
- the remnant charges are detected on a bit line in a manner similar to the operation of a conventional DRAM.
- these structures are not without attendant problems.
- the RAM data is volatile and needs to be refreshed on a regular basis, or this aspect of the cell acts like a DRAM.
- the electrically alterable read-only memory (EAROM), e.g., an electrically erasable programmable ROM (EEPROM) data is nonvolatile.
- EAROM electrically alterable read-only memory
- EEPROM electrically erasable programmable ROM
- the FRAM acts like a DRAM with a “repressed” nonvolatile read only “memory,” or “repressed memory.”
- ferroelectric capacitors as part of a stacked gate structure similar not to DRAMs but rather to EEPROM or flash memory devices. See, for example, U.S. Pat. No. 5,541,871 (“Nonvolatile Ferroelectric Semiconductor Memory) and U.S. Pat. No. 5,856,688 (“Integrated Circuit Memory Devices Having Nonvolatile Single Transistor Unit Cells Therein), which patents are also incorporated herein by reference for all purposes. It will be appreciated that in these devices, however, the remnant charge or polarization charge of the ferroelectric element is used to store information rather than electrons injected onto or removed from a floating gate by hot electron effects and/or tunneling. Instead, the charge differences are not differences in the number of electrons trapped on the gate but rather the polarization charge of the upper ferroelectric capacitor.
- a memory cell 1 ′ includes a wordline 10 , a digitline 12 , a transistor 14 , and a ferroelectric capacitor 20 , one plate of the ferroelectric capacitor forming the gate of the transistor 14 .
- FIG. 1( c ) illustrates a memory cell 1 ′′, which includes the wordline 10 , digitline 12 , transistor 14 , a high dielectric capacitor 30 and a ferroelectric capacitor 32 . It will be noted that the lower plate of capacitor 30 forms the control gate of the transistor 14 .
- the ferroelectric elements have very high electric permittivities as, for instance, 80 and 150 k. It will be appreciated that these permittivity values are 20 to 40 times higher than that of silicon dioxide. It will also be appreciated that, if comparable thicknesses of materials as are used in constructing the ferroelectric and conventional capacitors, only a small fraction, e.g., 2% to 5%, of the voltage applied across the series capacitors will appear across the ferroelectric capacitor. Thus, if the ferroelectric capacitor has a coercive voltage, Vc, i.e., the voltage required for programming, of 1V or 3V, then the total word line voltage required for programming the memory cell will be on the order 20V to 150V. It will be noted that these are far in excess of voltages used on current CMOS-integrated circuits.
- U.S. Pat. No. 5,856,688 seeks to solve this problem by using a “C” shaped floating gate and two control gates.
- One control gate is a plate of the ferroelectric capacitor and the other control gate is a plate of another capacitor fabricated using a high dielectric constant material.
- the two capacitors in series, one with a high dielectric constant ferroelectric and the other with a high dielectric constant insulator, are used to program the ferroelectric capacitor at lower voltages. Since the capacitances are more or less comparable, the programming voltage will divide more equally, resulting in a significant fraction across the ferroelectric capacitor.
- U.S. application Ser. No. 09/385,380 describes a memory cell 1 ′′′ having first and second operating modes.
- the memory cell 1 ′′′ includes a charge transfer transistor 14 having a gate adjacent to a channel region coupling source and drain regions, a digitline 12 coupled to one of the source and drain regions, a storage capacitor 16 coupled to the other of the source and drain regions, a ferroelectric capacitor 32 , and a wordline 10 coupled to the gate by the ferroelectric capacitor 32 .
- the polysilicon vertical interconnect connecting the ferroelectric capacitor 32 with oxide layer covering the gate of the transistor 14 forms an intrinsic capacitor 30 . See FIG. 1( d ).
- Data is written to and read out of the storage capacitor during the first operating mode, i.e., a normal DRAM access speeds, and written to and read out of the ferroelectric capacitor during the second mode of operation, i.e., at speeds several times slower than normal DRAM access speeds.
- a memory cell or device could be developed with all of the positive features of DRAMS, i.e., cost, size, speed, availability, etc., which also is nonvolatile, i.e., maintains its memory state with power removed. It would be highly desirable to have a memory element and corresponding system that functions at speeds comparable to that of today's DRAMs. It would also be desirable if at least one logic device could be implement using a modified form of the memory cell.
- the present invention provides a one transistor/one capacitor (1T/1C) memory cell including a ferroelectric capacitor operatively coupled to the gate electrode of a charge amplifier.
- the ferroelectric capacitor has a small surface area relative to the surface area of the gate electrode and, most preferably, the ferroelectric capacitor is an edge defined ferroelectric capacitor.
- the charge amplifier is a transistor.
- the present invention provides a memory cell including a charge amplifier having a gate adjacent to a channel region coupling source and drain regions, a digitline coupled to one of the source and drain regions, a ground lead coupled to the other of the source and drain regions, a ferroelectric capacitor coupled to the gate, and a wordline coupled to the ferroelectric capacitor.
- the charge amplifier can be a transistor.
- the gate is coupled to the ferroelectric capacitor by polysilicon, the junction formed at the gate has an intrinsic capacitance, and the capacitance of the ferroelectric capacitor is based on the magnitude of the intrinsic capacitance.
- the gate is coupled to the ferroelectric capacitor by polysilicon, the junction formed at the gate has an intrinsic capacitance, and the physical size of the ferroelectric capacitor is based on the magnitude of the intrinsic capacitance.
- the thickness of a ferroelectric material layer in the ferroelectric capacitor can be based on the magnitude of the intrinsic capacitance.
- the present invention provides a one transistor/one capacitor (1T/1C) memory cell array, wherein each of the 1T/1C memory cells is a ferroelectric capacitor operatively coupled to the gate electrode of a charge amplifier.
- the charge amplifier is a transistor and, most preferably, the charge amplifier is a CMOS transistor.
- the memory cell array is N rows by M columns of memory cells, and wherein the each of the M columns is defined by an associated digitline and an associated source line.
- the present invention provides a memory cell array, including a plurality of memory cells organized as an array of rows and columns, at least one of the memory cells being a non-volatile memory cell including a charge transistor having a gate opposing a channel region coupling source and drain regions, a digitline coupled to the drain region, a source line coupled to the source region, a ferroelectric capacitor which stores data coupled to the gate, a wordline coupled to the ferroelectric capacitor, and a sense amplifier coupled to the digitline of the at least one memory cell.
- the present invention provides a processor based system, including a processor, and a memory cell array coupled to the processor, the memory cell array including a plurality of memory cells organized as an array of rows and columns, at least one of the memory cells being a non-volatile memory cell including a charge amplifier including a gate opposing a channel region coupling source and drain regions, a digitline coupled to the drain region, a source line coupled to the source region, a ferroelectric capacitor coupled to the gate, a wordline coupled to the ferroelectric capacitor, and a sense amplifier coupled to the digitline of the at least one memory cell.
- a charge amplifier including a gate opposing a channel region coupling source and drain regions, a digitline coupled to the drain region, a source line coupled to the source region, a ferroelectric capacitor coupled to the gate, a wordline coupled to the ferroelectric capacitor, and a sense amplifier coupled to the digitline of the at least one memory cell.
- the present invention provides a memory module, including a substrate including a circuit board, a plurality of memory chips mounted on the die substrate, wherein one or more of the memory chips comprise a memory cell array fabricated on the semiconductor chip communicating with a processor, the memory cell array including at least one non-volatile memory cell including a ferroelectric capacitor operatively coupled to the gate of a charge amplifier.
- the present invention provides a method for reading information from a one transistor/one capacitor (1T/1C) memory cell having a charge amplifier including a gate disposed adjacent to a channel region coupling source and drain regions, a digitline coupled to drain region, a source line coupled to the source region, a ferroelectric capacitor coupled to the gate, and a wordline coupled to the ferroelectric capacitor, the method including grounding the source line, asserting the wordline and digitline, amplifying the charge on the ferroelectric capacitor, and determining the resultant charge on the digitline.
- a charge amplifier including a gate disposed adjacent to a channel region coupling source and drain regions, a digitline coupled to drain region, a source line coupled to the source region, a ferroelectric capacitor coupled to the gate, and a wordline coupled to the ferroelectric capacitor, the method including grounding the source line, asserting the wordline and digitline, amplifying the charge on the ferroelectric capacitor, and determining the resultant charge on the digitline.
- the present invention provides a method for writing information to a one transistor/one capacitor (1T/1C) memory cell having a charge amplifier including a gate disposed adjacent to a channel region coupling source and drain regions, a digitline coupled to drain region, a source line coupled to the source region, a ferroelectric capacitor coupled to the gate, and a wordline coupled to the ferroelectric capacitor, the method including asserting the source line, asserting the wordline to thereby apply a potential greater than the coercive voltage of the ferroelectric capacitor across the ferroelectric capacitor and the charge amplifier.
- a charge amplifier including a gate disposed adjacent to a channel region coupling source and drain regions, a digitline coupled to drain region, a source line coupled to the source region, a ferroelectric capacitor coupled to the gate, and a wordline coupled to the ferroelectric capacitor, the method including asserting the source line, asserting the wordline to thereby apply a potential greater than the coercive voltage of the ferroelectric capacitor across the ferroelectric capacitor and the charge amplifier.
- the present invention provides a logic element, including N+1 source/drain regions, N channels, each channel coupling adjacent ones of the N+1 source/drain regions, N gates disposed adjacent to the N channels, N ferroelectric capacitors, each of the N ferroelectric capacitors coupled to a respective one of the N gates, N wordlines, each of the N wordlines coupled to a respective one of the N ferroelectric capacitors, a digitline couple to an end one of the N+1 source/drain regions, and a source line coupled to the other end one of the N+1 source/drain regions.
- the logic element is operated by assertion of all of the N wordlines.
- the ferroelectric capacitors are edge defined ferroelectric capacitors.
- FIG. 1( a ) illustrates a first use of a ferroelectric capacitor in a conventional memory cell
- FIG. 1( a ) illustrates a second use of a ferroelectric capacitor in a conventional memory cell
- FIG. 1( c ) illustrates a third use of a ferroelectric capacitor in a conventional memory cell
- FIG. 1( d ) illustrates a fourth use of a ferroelectric capacitor in a memory cell
- FIG. 2( a ) illustrates a NVRAM memory cell employing a ferroelectric capacitor according to the present invention
- FIG. 2( b ) illustrates a NVRAM memory cell array employing a ferroelectric capacitor according to the present invention
- FIG. 3( a ) illustrates selected portions of the components used in the memory cell depicted in FIGS. 2 ( a ) and 2 ( b );
- FIG. 3( b ) illustrates the hysteresis characteristic of the ferroelectric capacitor employed in the memory cell illustrated in FIGS. 2 ( a ) and 2 ( b );
- FIG. 4 is a high-level block diagram of a memory array device which can be fabricated using the memory cell depicted in FIG. 2( b );
- FIG. 5 is a flow chart illustrating a method for accessing the memory cell depicted in FIG. 2( a );
- FIGS. 6 ( a ), 6 ( c ) 6 ( e ) illustrate the application of various voltages to selected portions of the circuitry illustrated in FIG. 2( a );
- FIGS. 6 ( b ), 6 ( d ) 6 ( f ) are hysteresis diagrams illustrating the effects on the ferroelectric capacitor of the voltages specified in FIGS. 6 ( a ), 6 ( c ) 6 ( e ), respectively;
- FIG. 7 illustrates a NAND logic device which is formed using the basic NVRAM cell according to the present invention
- FIG. 8 is a flow chart illustrating the process flow for fabricating the memory cell illustrated in FIG. 2( a );
- FIGS. 9 ( a ) through 9 ( o ) illustrate one of plan and section view of a workpiece at various steps of the process flow illustrated in FIG. 8;
- FIG. 10 is a plan view of a memory module having memory chips constructed according to the present invention.
- FIG. 11 is a high-level block diagram of a processor-based system employing NVRAM memory chips constructed according to the present invention.
- FIG. 2( a ) illustrates the basic transistor and memory cell 100 electrical configuration according to the present invention, which preferably includes a transistor 114 connected between a source line 116 and a digitline 112 .
- a stacked capacitor gate structure is provided, which includes a polysilicon gate capacitance 120 and a small area ferroelectric capacitor 118 .
- the memory cell 100 according to the present invention includes a wordline 110 , a digitline 112 , a transfer device, e.g., a transistor, 114 , a source line 116 and a ferroelectric capacitor 118 .
- the transition between the active region of the transfer device 114 and the polysilicon interconnect 122 to the wordline 110 is an oxide layer 120 a, which forms a gate capacitor 120 .
- the basic memory cell 100 advantageously can be employed in the memory cell array 142 illustrated in FIG. 2( b ), which is discussed in greater detail below.
- FIG. 3( a ) A more detailed illustration of the transfer device 114 , intrinsic gate capacitor 120 , polysilicon interconnect 122 and ferroelectric capacitor 118 is provided in FIG. 3( a ). It should be mentioned that the ferroelectric capacitor 118 preferably includes Platinum (Pt) electrodes 118 a, 118 c sandwiching a ferroelectric layer 118 b, which will be described in greater detail below. The characteristics of the ferroelectric layer 118 b and the gate oxide layer 120 a are listed in the table below.
- FIG. 3( b ) illustrates a hysteresis loop characteristic of ferroelectric material 118 b in comparison to the linear response of an oxide material 120 a.
- V C indicates coercive voltage
- Q r is remnant charge
- Q s is saturation charge
- the Y axis represents charge/centimeter squared in micro-coulombs. It will be appreciated that as long as the applied voltage is less than V c , the charge will follow the linear profiles indicated for layers 118 b and 120 a. Once V c has been exceeded, which corresponds to an applied voltage of approximately 1 V, the layer 118 b follows the hysteresis profile.
- the layer 118 b will exhibit a remnant charge Q r , which advantageously can be positive or negative. Since Q r will effect the threshold voltage V th of transistor 114 , the polarity of Q r can be determined by ascertaining the effect on V th , as discussed in greater detail below.
- ferroelectric characteristics depicted in FIG. 3( b ) and Table I are that of a common ferroelectric, e.g., SrBiTaO, although other ferroelectric materials advantageously can be used, as discussed in greater detail below.
- the capacitor 120 has the normal gate capacitance of a minimum-sized transistor 114 in CMOS technology.
- the values presented in Table I are characteristic of 1 micron CMOS technology employing 5V power supplies. It will be appreciated that these values advantageously can be scaled, as appropriate, to smaller CMOS technologies with critical dimensions smaller than 1 micron.
- the gate capacitance, i.e., the intrinsic capacitance of capacitor 120 of a transistor 114 in this technology with a gate oxide thickness of approximately 100 ⁇ is about 3.3 fF.
- the charge storage element of a ferroelectric memory cell consists of a ferroelectric film sandwiched between two electrodes.
- PZT film processing can be done in combination with standard integrated circuit processing, as discussed in the article by R. Cuppens, et al., entitled “Ferroelectrics for Non-volatile Memories” (Microelectronic Engineering, 19 (1992), pp.
- Nonvolatile memory devices include BaMgF 4 and Bi 4 Ti 3 O 12 for application with the ferroelectric material gate insulator in a field effect transistor (FET).
- FET field effect transistor
- BaTiO 3 , SrTiO 3 and (Sr,Ba)TiO 3 advantageously can be employed for application as dielectrics in DRAMs.
- Bi-layer ferroelectric oxides have large polarization along the a or b axis, but no or little polarization along the c axis.
- Bi 4 Ti 3 O 12 is a typical bi-layer ferroelectric material.
- the obtained remnant polarization is very small, i.e., less than 10 ⁇ C/cm 2 and fatigue endurance is less than optimal.
- SrBi 2 Ta 2 O 9 or SrBi 2 Nb 2 O 9 hve demonstrated acceptable endurance.
- the ferroelectric capacitor 118 is illustrated as a capacitor on top of the gate of the transistor 114 , which is connected to the gate of transistor 114 by polysilicon interconnect 122 , which is, in an exemplary case, approximately 2000 ⁇ thick.
- the ferroelectric capacitor 118 is fabricated by depositing a ferroelectric material between two Pt plates or electrodes, one of the electrodes, e.g., electrode 118 a, being a vertical Pt wordline.
- the ferroelectric capacitor 118 advantageously has an effective plate area of only about one-fifth that of the gate oxide capacitor 120 .
- the ferroelectric dielectric 118 b will typically be thicker, approximately 800 ⁇ , than the gate oxide 120 a, approximately 100 ⁇ . It will be appreciated that by properly sizing the ferroelectric capacitor 118 and the intrinsic gate capacitor 120 , the gate oxide capacitance and the ferroelectric capacitances can be made comparable. Thus, when the coercive or programming voltage V c of the ferroelectric capacitor 118 is on the order of 1V, then the memory cell 100 can be programmed with a voltage as low as ⁇ 5V. It will be appreciated that both of these values are consistent with and, in fact, less than used in 1 micron flash memory CMOS technology.
- one preferred embodiment of the present invention is a one transistor/one capacitor (1T/1C) memory cell including a ferroelectric capacitor operatively coupled to the gate electrode of a charge amplifier.
- the ferroelectric capacitor has a small surface area relative to the surface area of the gate electrode and, most preferably, the ferroelectric capacitor is an edge defined ferroelectric capacitor, i.e., a capacitor formed by running the ferroelectric film over a step in the workpiece and then anisotropically etching the film in the vertical direction.
- the charge amplifier is a transistor.
- An exemplary embodiment according to the present invention provides a memory cell including a charge amplifier having a gate adjacent to a channel region coupling source and drain regions, a digitline coupled to one of the source and drain regions, a ground lead coupled to the other of the source and drain regions, a ferroelectric capacitor coupled to the gate, and a wordline coupled to the ferroelectric capacitor.
- the charge amplifier can be a transistor.
- the gate is coupled to the ferroelectric capacitor by polysilicon, the junction formed at the gate has an intrinsic capacitance, and the capacitance of the ferroelectric capacitor is based on the magnitude of the intrinsic capacitance.
- the gate is coupled to the ferroelectric capacitor by polysilicon, the junction formed at the gate has an intrinsic capacitance, and the physical size of the ferroelectric capacitor is based on the magnitude of the intrinsic capacitance.
- the thickness of a ferroelectric material layer in the ferroelectric capacitor can be based on the magnitude of the intrinsic capacitance.
- the ferroelectric capacitor is a dielectric layer of lead zirconate titanate (PZT).
- PZT lead zirconate titanate
- the PZT forming the ferroelectric capacitor is of the form PbZr x Ti 1 ⁇ x O 3 with x ranging between 0.4 and 0.53
- the memory array 142 advantageously can be formed from the NVRAM memory cell illustrated in FIG. 2( a ).
- the memory cell array 142 includes a plurality of NVRAM memory cells, e.g., cells 100 a, 100 b, 100 c, connected to one another by a common word line W1.
- Each of the NVRAM memory cells, e.g., NVRAM memory cell 100 a advantageously can be addressed by a corresponding digit line, e.g., digit line D1, which preferably addresses other NVRAM memory cells, e.g., cell 100 d, and which can be recharged for read out via a pullup transistor, e.g., transistor T1.
- transistor T1 is switched for readout in the normal manner.
- the source terminals for a column of memory cells 100 are connected to a ground via a switch S1. It will be appreciated that all of the switches advantageously can be closed during a read operation of the memory array 142 but can be selectively operated during a write operation, i.e., in order to write data to a selected one of the memory cells in the memory array 142 , the corresponding source line is asserted.
- each of the 1T/1C memory cells is a ferroelectric capacitor operatively coupled to the gate electrode of a charge amplifier.
- the charge amplifier is a transistor and, most preferably, the charge amplifier is a CMOS transistor.
- the memory cell array is N rows by M columns of memory cells, and wherein the each of the M columns is defined by an associated digitline and an associated source line.
- the DRAM 140 contains a memory cell array 142 , row and column decoders 144 , 148 and a sense amplifier circuit 146 .
- the memory cell array 142 consists of a plurality of memory cells 100 (constructed as illustrated in FIG. 2( a )) whose word lines 110 and digitlines 112 are commonly arranged into rows and columns, respectively.
- the digitlines 112 of the memory array 142 are connected to the sense amplifier circuit 146 , while its word lines 110 are connected to the row decoder 144 .
- the source lines for all memory cells 100 is grounded, preferably by source line 116 . Address and control signals are input into the DRAM 140 and connected to the column decoder 148 , sense amplifier circuit 146 and row decoder 144 and are used to gain read and write access, among other things, to the memory array 142 .
- the column decoder 148 is connected to the sense amplifier circuit 146 via control and column select signals.
- the sense amplifier circuit 146 receives input data destined for the memory array 142 and outputs data read from the memory array 142 over input/output (I/O) data lines.
- Data is read from the cells of the memory array 142 by activating a word line (via the row decoder 144 ), which couples all of the memory cells 100 corresponding to that word line 110 to respective digitlines 112 , which define the columns of the array.
- One or more of the digitlines 112 are also activated.
- the sense amplifier circuit 146 When a particular wordline-digitline pair is activated, the sense amplifier circuit 146 connected to a digitline column detects and amplifies the data bit impressed on the capacitor 118 of the memory cell 100 to its respective digit line 112 by measuring the potential difference between the activated digitline and a reference line which may be an inactive bit line.
- a reference line which may be an inactive bit line.
- a memory cell array includes a plurality of memory cells organized as an array of rows and columns, where at least one of the memory is a non-volatile memory cell having a charge transistor having a gate opposing a channel region coupling source and drain regions, a digitline coupled to the drain region, a source line coupled to the source region, a ferroelectric capacitor which stores data coupled to the gate, a wordline coupled to the ferroelectric capacitor, and a sense amplifier coupled to the digitline of the at least one memory cell.
- the ferroelectric capacitor has a surface area which is small relative to the surface area of the gate.
- the ferroelectric capacitor is an edge defined ferroelectric capacitor.
- the data is written into the ferroelectric capacitor when the wordline and source line are asserted and the data is read from the non-volatile memory cell when the wordline, the digitline and the source line are asserted.
- an asserted source line is a grounded source line.
- the charge amplifier amplifies the charge stored in the ferroelectric capacitor by at least one order of magnitude. In any event, the charge amplifier amplifies the charge stored in the ferroelectric capacitor by a predetermined gain factor.
- the charge amplifier advantageously can be a transistor.
- a check is first performed to determine whether the memory cells 100 are being accessed for a read operation during step 502 . If the answer is affirmative, all of the source lines are coupled to ground potential during step 504 by operation of switches S 1 , S 2 , etc., and then a read operation is performed in the normal manner for a DRAM, i.e., the wordline and digitlines for a particular memory cell are asserted, during step 506 , which read operation is well known to one of ordinary skill in the art and will not be described further in the interest of brevity. However, the principal which permits the memory cell 100 to be read out normally will be discussed below.
- step 502 is then repeated. It will be appreciated that as long as the voltage applied to the wordline 110 remains in the normal operating range of the transistor 114 , e.g., 3 V, the coercive voltage of the ferroelectric capacitor 118 is not exceeded and the state of the ferroelectric capacitor is unchanged. See FIGS. 6 ( a ) and 6 ( b ).
- the source line for all of the memory cells are deasserted, i.e., the switches S 1 , S 2 , etc., are opened during step 508 .
- a voltage in excess of the coercive voltage Vc for the ferroelectric capacitor 118 is applied to the wordline 110 of memory cell 100 while the source 116 for the selected cell(s) is(are) grounded and all other source lines 112 are open circuited.
- a voltage in excess of the coercive voltage Vc for the ferroelectric capacitor 118 is thus applied by assertion of the wordline 110 and source line 116 for the selected memory cell. It will be appreciated that the value stored in the ferroelectric capacitor 118 is determined by the polarity of the voltage applied.
- the read operation is achieved by driving the word line to some voltage less than the coercive or programming voltage, and reading the conductivity of the transistor.
- the capacitance of the wordline 110 of each memory cell 100 advantageously be comparable to that in DRAMs since the gate capacitances are in series and, in the exemplary case under discussion, about 1fF.
- the transistor 114 is not a passive transfer gate, but rather provides gain and charge multiplication. If the charge differences on the ferroelectric capacitor 118 are 10 ⁇ C/cm 2 , and the area is 2 ⁇ 10 ⁇ 9 cm 2 , then the charge difference in the memory cell 100 itself is 2 fC.
- this charge is multiplied during the read operation by the transistor 114 .
- the difference in charge states of the ferroelectric capacitor 118 results in a drain current difference of 100 uA in the transistor 114 during a read cycle of 1 nanosecond (ns)
- the differences in charge on the data or bit line 112 during read operation is approximately 100 fC. It will be appreciated that this value is comparable to that obtained using current DRAMs equipped with 30 ferroelectric capacitor storage capacitors and 3 V potential differences.
- one exemplary embodiment of the present invention is a method for reading information from a one transistor/one capacitor (1T/1C) memory cell having a charge amplifier including a gate disposed adjacent to a channel region coupling source and drain regions, a digitline coupled to drain region, a source line coupled to the source region, a ferroelectric capacitor coupled to the gate, and a wordline coupled to the ferroelectric capacitor, the method including grounding the source line, asserting the wordline and digitline, amplifying the charge on the ferroelectric capacitor, and determining the resultant charge on the digitline.
- a voltage less than the coercive voltage of the ferroelectric capacitor is asserted on the wordline while a predetermined voltage, e.g., V DD , is asserted on the digitline.
- a complementary embodiment of the present invention is a method for writing information to a one transistor/one capacitor (1T/1C) memory cell having a charge amplifier including a gate disposed adjacent to a channel region coupling source and drain regions, a digitline coupled to drain region, a source line coupled to the source region, a ferroelectric capacitor coupled to the gate, and a wordline coupled to the ferroelectric capacitor, the method including asserting the source line, and asserting the wordline to thereby apply a potential greater than the coercive voltage of the ferroelectric capacitor across the ferroelectric capacitor and the charge amplifier.
- a charge amplifier including a gate disposed adjacent to a channel region coupling source and drain regions, a digitline coupled to drain region, a source line coupled to the source region, a ferroelectric capacitor coupled to the gate, and a wordline coupled to the ferroelectric capacitor, the method including asserting the source line, and asserting the wordline to thereby apply a potential greater than the coercive voltage of the ferroelectric capacitor across the ferroelectric
- ferroelectric capacitor/transistor pairs advantageously can be fabricated into a NAND gate 175 , as illustrated in FIG. 7.
- a plurality of ferroelectric capacitor FCA, FCB, FCC, and FCD are coupled to the gates of respective transistors TA, TB, TC, and TD.
- the output of NAND gate 175 can be sensed on digitline 179 when all of the gates are asserted.
- all of the gates of transistor TA-TD are asserted when the wordlines 177 , e.g., wordlines W1, W2, W3, and W4, are all asserted.
- the source of transistor TA advantageously can be coupled to ground.
- a logic device includes N+1 source/drain regions, N channels, each channel coupling adjacent ones of the N+1 source/drain regions, N gates disposed adjacent to the N channels, N ferroelectric capacitors, each of the N ferroelectric capacitors coupled to a respective one of the N gates, N wordlines, each of the N wordlines coupled to a respective one of the N ferroelectric capacitors, a digitline couple to an end one of the N+1 source/drain regions, and a source line coupled to the other end one of the N+1 source/drain regions.
- the logic element is operated by assertion of all of the N wordlines.
- the ferroelectric capacitors are edge defined ferroelectric capacitors.
- the ferroelectric capacitor 118 advantageously can be an edge-defined ferroelectric capacitor. Fabrication of the ferroelectric capacitor will be discussed with respect to FIG. 8. However, before describing an exemplary process for the manufacture of memory cell 100 , a brief review of the currently available technology, from which preferred methodology for the fabrication of one type of edge-defined capacitors, will be presented. It will be appreciated that the method discussed immediately below advantageously may be applied to other types of edge-defined capacitors.
- the charge storage element 118 of a ferroelectric memory cell 100 consists of a ferroelectric film 118 b sandwiched between two electrodes 118 a, 118 c.
- Other thin-film materials which advantageously can be employed in nonvolatile memory devices include BaMgF 4 and Bi 4 Ti 3 O 12 for application with the ferroelectric material gate insulator in FET.
- BaTiO 3 , SrTiO 3 and (Sr,Ba)TiO 3 have demonstrated usefulness in application as dielectrics in DRAMs.
- Alternative techniques for the deposition of PZT films include Metal Organic Deposition (MOD) and a sol-gel process.
- MOD Metal Organic Deposition
- sol-gel process the ferroelectric film is formed by the thermal decomposition of a solution of organo metallic compounds which can be spun onto the substrate.
- the subsequent drying, firing and annealing processes of the film strongly influence the morphology and ferroelectric properties of the PZT. It should be mentioned that thicker films can be obtained by repeating the spinning, drying and firing sequence.
- the main advantage of these processes is the ease by which complex oxide films can be prepared.
- OMCVD Organo Metallic Chemical Vapor Deposition
- OMCVD Organo Metallic Chemical Vapor Deposition
- an ozone jet reactive evaporation technique has been developed to deposit thin lead zirconate titanate films. It will be appreciated that because lead has a low affinity with oxygen and high volatility, lead may re-evaporate from the film and/or diff-use into the electrode and leave lead vacancies in PZT films. In a sol-gel or sputtering process, excess lead is often added to compensate for this lead loss.
- MBE molecular beam epitaxy
- the electrodes are the electrodes. Because of the chemical reactivity of PZT and related materials and relatively high processing temperatures (>600° C.), a very inert bottom electrode is required. In general, Pt-based electrodes are used which are deposited onto silicon oxide or nitride by sputtering using an adhesive layer such as TiN, Ti, or the two in combination. Although relatively inert, the Pt bottom electrode does affect the morphology and the ferroelectric properties of the PZT films and related materials.
- Bi-layer ferroelectric oxides have large polarization along the a or b axis, but no or little polarization along the c axis.
- Bi 4 Ti 3 O 12 is a typical bi-layer ferroelectric, and the synthesis of this material has been reported. However, the obtained remnant polarization is very small, less than 10 ⁇ C/cm 2 and fatigue endurance is poor.
- other types of bi-layer oxides such as SrBi 2 Ta 2 O 9 or SrBi 2 Nb 2 O 9 were found to have good endurance.
- FIG. 8 a process flow for integration of ferroelectric capacitors into the memory devices 140 will now be described. It should be mentioned that the process flow is geared to produce a memory cell pair; the process advantageously can be employed in fabricating the memory array 142 or the NAND gate 175 by proper selection of various process masks. It should also be mentioned at this point that such integration should not degrade the ferroelectric properties of the film.
- One example of process integration with a stacked capacitor structure was shown by Onish, et al., in the article entitled “A Half-Micron Ferroelectric Memory Cell Technology with Stacked Capacitor Structure,” (1994 IEDM Digest, pp. 843-846, 1994).
- the process flow starts with a workpiece 500 , e.g., a wafer fragment, following well formation and growth of a thin (e.g., ⁇ 10 nm) pad of thermal oxide 502 followed by deposition of a thick pad of nitride 504 ( ⁇ 0.2 ⁇ m) for formation of STI (shallow trench isolation) during step 600 .
- a thin (e.g., ⁇ 10 nm) pad of thermal oxide 502 followed by deposition of a thick pad of nitride 504 ( ⁇ 0.2 ⁇ m) for formation of STI (shallow trench isolation) during step 600 .
- STI shallow trench isolation
- step 604 the pad nitride 504 and oxide 502 is removed to expose bare silicon in the active areas, preferably recessed relative to the STI surface.
- step 606 gate oxide 508 is grown, prior to depositing polysilicon gate material 510 and subsequent CMP finishing to leave structure as shown in FIGS. 9 ( c ) and 9 ( d ).
- step 608 a resist 512 is applied, the gate pattern is exposed, and the wafer is etched to produce the intermediate product depicted in FIGS. 9 ) e ) and 9 ( f ).
- step 610 the resist 512 is removed and the source/drain pockets 514 are ion implanted and annealed.
- the wafer can be finished to restore surface planarity by chemical vapor deposition (CVD) of intrinsic polysilicon 516 followed by another round of CMP.
- step 614 the ferroelectric capacitor 118 is formed by depositing a Pt layer 118 a, a ferroelectric material layer 118 b, and a final Pt cap layer 118 c on the surface of the intrinsic polysilicon 516 and gate material 510 . It will be appreciated that any of the deposition processes discussed in detail above advantageously can be employed in depositing the ferroelectric material layer.
- ferroelectric capacitor material 118 b advantageously can be chosen from the non-limiting list of materials presented above.
- intrinsic polysilicon 516 is deposited to a thickness equal to the desired height of the final spacer defined word lines (e.g., 0.2 ⁇ m). Then a resist 512 ′ is applied and masked to define word line edge pattern illustrated in FIGS. 9 ( g ) and 9 ( h ) during step 620 .
- the intrinsic polysilicon mandril material is directionally etched, preferably stopping on underlying Pt cap layer 118 c and then the resist 512 ′ is removed.
- a suitable word line conductor 518 e.g., doped polysilicon, refractory metal or silicide
- CVD chemical vapor deposition
- conductor 518 is deposited by CVD and then directionally etched to leave conductor 518 as a spacer on the vertical edges of the mandril, i.e., to form edge-defined ferroelectric capacitors as shown in FIGS. 9 ( i ) and 9 ( j ).
- a suitable selective etchant is employed to remove intrinsic polysilicon 516 mandril material.
- the exposed Pt-ferroelectric material-Pt sandwich 118 a, 118 b, 118 c is etched away.
- the workpiece 500 is selectively etched to remove any remaining intrinsic polysilicon 516 over the source/drain pockets 514 , leaving the structure shown in FIGS. 9 ( k ) and 9 ( l ).
- step 630 a thick oxide interlayer dielectric is deposited, planarized to form contact holes and wiring levels as in conventional processing. It will be appreciated that when the wordlines 110 are formed, these wordlines will be considerably wider than the underlying ferroelectric capacitor 118 , as illustrated in FIG. 9( m ). It will be appreciated that FIG. 9( n ), which is a schematic diagram of the memory cells illustrated in FIG. 9( m ) is included to orient the reader.
- the memory array 142 advantageously could be fabricated as a number of subarrays, each having an associated block select switch, e.g., BS 1 , BS 2 , etc., serially coupling the local wordlines LW 1 , LW 2 , etc., to a global wordline GWL, as illustrated in FIG. 9( o ).
- FIG. 10 is plan view of a memory module 600 having memory chips 60 - 68 including semiconductor memory devices constructed in accordance with the present invention. That is, chips 60 - 68 have a NVRAM cell such as described in connection with FIG. 2.
- Memory module 600 is a SIMM (single in line memory module) having nine memory chips (IC's) 60 - 68 aligned on one side of a printed circuit board substrate. The number of such memory chips in the SIMM typically will vary between 3 to 9.
- the circuit board 601 has an edge connector 602 along one longitudinal edge to permit it to plug into a memory socket on a computer motherboard of conventional design (not shown).
- a wiring pattern (not shown), which can be a conventionally known design for this purpose, is formed on the board 601 and connects the terminals or leads shown comprising the edge connector 602 to the memory chips 60 - 68 .
- Small ceramic decoupling capacitors 69 are also mounted on substrate 601 to suppress transient voltage spikes.
- the general layout of the SIMM 600 can be a conventional construction.
- another exemplary embodiment of the present invention is a memory module, including a substrate in, for example, a circuit board, a plurality of memory chips mounted on the die substrate, wherein one or more of the memory chips comprise a memory cell array fabricated on a semiconductor chip for communicating with a processor, the memory cell array including at least one non-volatile memory cell including a ferroelectric capacitor operatively coupled to the gate of a charge amplifier.
- the gate is coupled to the ferroelectric capacitor by polysilicon, the junction formed at the gate has an intrinsic capacitance, and the capacitance of the ferroelectric capacitor is based on the magnitude of the intrinsic capacitance.
- the gate is coupled to the ferroelectric capacitor by polysilicon, the junction formed at the gate has an intrinsic capacitance, and the physical size of the ferroelectric capacitor is based on the magnitude of the intrinsic capacitance.
- the ferroelectric capacitor includes a dielectric layer of lead zirconate titanate (PZT), which advantageously can be of the form PbZr x Ti 1 ⁇ x O 3 with x ranging between 0.4 and 0.53.
- the non-volatile memory cell includes the charge amplifier having a gate disposed adjacent to a channel region coupling source and drain regions, a digitline coupled to drain region, a source line coupled to the source region, the ferroelectric capacitor coupled to the gate, and a wordline coupled to the ferroelectric capacitor.
- FIG. 11 is a block diagram of a processor-based system 400 utilizing a memory 412 constructed in accordance with the present invention. That is, the memory 412 utilizes the memory cells 100 illustrated in FIG. 2( a ), which advantageously can be provided in the form of the memory device 140 .
- the processor-based system 400 may be a computer system, a process control system or any other system employing a processor and associated memory.
- the system 400 includes a central processing unit (CPU) 402 , e.g., a microprocessor, that communicates with the RAM 412 and an I/O device 408 over a bus 420 .
- CPU central processing unit
- bus 420 may be a series of buses and bridges commonly used in a processor-based system, but for convenience purposes only, the bus 420 has been illustrated as a single bus.
- a second I/O device 410 is illustrated, but is not necessary to practice the invention.
- the processor-based system 400 also includes an additional memory 414 , which could be either a read-only memory (ROM) or another of the memory devices 140 according to the present invention, since, as discussed above, the memory cells are non-volatile devices.
- ROM read-only memory
- the processor based system may include peripheral devices such as a floppy disk drive 404 , a compact disk (CD) ROM drive 406 , a display (not shown), a key board (not shown), and a mouse (not shown), that communicate with the CPU 402 over the bus 420 as is well known in the art.
- peripheral devices such as a floppy disk drive 404 , a compact disk (CD) ROM drive 406 , a display (not shown), a key board (not shown), and a mouse (not shown), that communicate with the CPU 402 over the bus 420 as is well known in the art.
- the memories 412 and 414 advantageously can be the memory device 140 according to the present invention, the number of discrete memory devices needed for the processor based system 400 can be reduced. Stated another way, since the memory array 142 according to the present invention is populated by non-volatile memory cells 100 , a predetermined portion of the memory device 142 advantageously can be employed as the ROM in the processor based system 400 .
- another preferred embodiment of the present invention includes a processor based system, including a processor, and a memory cell array coupled to the processor, the memory cell array including a plurality of memory cells organized as an array of rows and columns, at least one of the memory cells being a non-volatile memory cell including a charge amplifier including a gate opposing a channel region coupling source and drain regions, a digitline coupled to the drain region, a source line coupled to the source region, a ferroelectric capacitor coupled to the gate, a wordline coupled to the ferroelectric capacitor, and a sense amplifier coupled to the digitline of the at least one memory cell.
- a processor based system including a processor, and a memory cell array coupled to the processor, the memory cell array including a plurality of memory cells organized as an array of rows and columns, at least one of the memory cells being a non-volatile memory cell including a charge amplifier including a gate opposing a channel region coupling source and drain regions, a digitline coupled to the drain region, a source line coupled
- the ferroelectric capacitor is an edge defined ferroelectric capacitor.
- the ferroelectric capacitor is a dielectric layer of lead zirconate titanate (PZT) and, most preferably, the PZT forming the ferroelectric capacitor is of the form PbZr x Ti 1 ⁇ x O 3 with x ranging between 0.4 and 0.53.
Abstract
Description
- This is a Continuation-in-Part of U.S. application Ser. No. (09/385,380), which is entitled “DYNAMIC RANDOM ACCESS MEMORY (DRAM) CELLS WITH REPRESSED FERROELECTRIC MEMORY, METHODS OF READING SAME, AND APPARATUSES INCLUDING SAME, and which was filed on Aug. 30, 1999.
- 1. Field of the Invention
- The present invention relates to discrete elements such as electrically alterable programmable read-only memory (EAPROM) and nonvolatile random access memory (NVRAM) devices. More specifically, the present invention relates to EAPROM and NVRAM cells where data is stored using a ferroelectric capacitor incorporated in the cells. Methods for operating, writing to and reading from, such DRAM, logic devices, memory arrays, and systems containing these cells are also disclosed.
- 2. Description of Related Art
- Dynamic random access memories (DRAMs) are the most cost-effective, high-speed memory used with present day computers. DRAMS last nearly indefinitely. Moreover, DRAMs are available in very high density configurations, e.g., 64 megabytes (MB). However, DRAMS can only store information for a limited time without constant refreshing and lose all knowledge of their state once power is removed.
- Ferroelectric capacitors are currently being used as nonvolatile memory devices, as disclosed in U.S. Pat. No. 4,888,733, which is entitled “Nonvolatile Memory Cell and Sensing Method” and discloses a two transistor, one capacitor (2T/1C) memory cell. In addition, ferroelectric capacitors are often employed in nonvolatile random access memories (NVRAMs). Memory cells having structures approximating that of DRAMs, i.e., arranged in the conventional one transistor, one capacitor (1T/1C) memory cell pattern, are disclosed in U.S. Pat. No. 5,600,587 (Ferroelectric Random Access Memory), U.S. Pat. No. 5,572,459 (Voltage Reference for a Ferroelectric 1T/1C Based Memory), U.S. Pat. No. 5,550,770 (Semiconductor Memory Device Having FE Capacitor Memory Cells with Reading, Writing, and Forced Refreshing Functions and a Method of Operating the Same), U.S. Pat. No. 5,530,668 (Ferroelectric Memory Sensing Scheme Using Bit Lines Precharged to a Logic One Voltage), and U.S. Pat. No. 5,541,872 (Folded Bit Line Ferroelectric Memory Device). It should be mentioned that all of the patents cited above are incorporated by reference herein for all purposes.
- In the above-mentioned patents, the capacitor in a 1T/1C DRAM structure is replaced by a ferroelectric capacitor, as illustrated in FIG. 1(a). The
memory cell 1 of FIG. 1(a) includes awordline 10, adigitline 12, atransistor 14 and aferroelectric capacitor 16. In operation, the remnant charges are detected on a bit line in a manner similar to the operation of a conventional DRAM. However, these structures are not without attendant problems. - There is an extensive body of literature on both the electrical characteristics of ferroelectric capacitors and applications in cells similar to conventional DRAMs, except that these cells are classified as nonvolatile RAMs (NVRAMs) instead of DRAMs. One recent publication by K. Asaril et al., entitled “Multi-Level Technologies for FRAM Embedded Reconfigurable Hardware” (IEEE Int. Solid-State Circuits Conf., San Francisco 1999, pp. 108-109), describes the use of a ferroelectric capacitor in a ferroelectric RAM (FRAM)which is employed with low voltage to store and access RAM data superimposed on less-frequently accessed read-only memory (ROM) data in the same cell. See U.S. Pat. No. 5,539,279, entitled “Ferroelectric Memory.” The RAM data is volatile and needs to be refreshed on a regular basis, or this aspect of the cell acts like a DRAM. The electrically alterable read-only memory (EAROM), e.g., an electrically erasable programmable ROM (EEPROM) data is nonvolatile. In other words, the FRAM acts like a DRAM with a “repressed” nonvolatile read only “memory,” or “repressed memory.”
- Other applications use the ferroelectric capacitors as part of a stacked gate structure similar not to DRAMs but rather to EEPROM or flash memory devices. See, for example, U.S. Pat. No. 5,541,871 (“Nonvolatile Ferroelectric Semiconductor Memory) and U.S. Pat. No. 5,856,688 (“Integrated Circuit Memory Devices Having Nonvolatile Single Transistor Unit Cells Therein), which patents are also incorporated herein by reference for all purposes. It will be appreciated that in these devices, however, the remnant charge or polarization charge of the ferroelectric element is used to store information rather than electrons injected onto or removed from a floating gate by hot electron effects and/or tunneling. Instead, the charge differences are not differences in the number of electrons trapped on the gate but rather the polarization charge of the upper ferroelectric capacitor.
- It should also be mentioned that the prior art referenced in U.S. Pat. No. 5,541,871 describes a basic structure consisting of a ferroelectric capacitor in series with a gate capacitance, which are both planar structures. As shown in FIG. 1(b), a
memory cell 1′ includes awordline 10, adigitline 12, atransistor 14, and aferroelectric capacitor 20, one plate of the ferroelectric capacitor forming the gate of thetransistor 14. - FIG. 1(c) illustrates a
memory cell 1″, which includes thewordline 10,digitline 12,transistor 14, a highdielectric capacitor 30 and aferroelectric capacitor 32. It will be noted that the lower plate ofcapacitor 30 forms the control gate of thetransistor 14. - The problem with this disclosed structure is that the ferroelectric elements have very high electric permittivities as, for instance, 80 and 150 k. It will be appreciated that these permittivity values are 20 to 40 times higher than that of silicon dioxide. It will also be appreciated that, if comparable thicknesses of materials as are used in constructing the ferroelectric and conventional capacitors, only a small fraction, e.g., 2% to 5%, of the voltage applied across the series capacitors will appear across the ferroelectric capacitor. Thus, if the ferroelectric capacitor has a coercive voltage, Vc, i.e., the voltage required for programming, of 1V or 3V, then the total word line voltage required for programming the memory cell will be on the order 20V to 150V. It will be noted that these are far in excess of voltages used on current CMOS-integrated circuits.
- U.S. Pat. No. 5,856,688 seeks to solve this problem by using a “C” shaped floating gate and two control gates. One control gate is a plate of the ferroelectric capacitor and the other control gate is a plate of another capacitor fabricated using a high dielectric constant material. The two capacitors in series, one with a high dielectric constant ferroelectric and the other with a high dielectric constant insulator, are used to program the ferroelectric capacitor at lower voltages. Since the capacitances are more or less comparable, the programming voltage will divide more equally, resulting in a significant fraction across the ferroelectric capacitor.
- It should be mentioned that other repressed memory devices have been proposed. For example, a repressed memory where the NVRAM function is provided by a flash memory type structure for the gate of the transfer device is described in the commonly assigned, copending application entitled “DRAM AND SRAM MEMORY CELLS WITH REPRESSED MEMORY” (Ser. No. (09/362,909, filed Jul. 29, 1999), which application is incorporated herein by reference for all purposes. It will be appreciated that these memories do not function like the shadow RAM disclosed in U.S. Pat. No. 5,399,516 (“Method of Making a Shadow Ram Cell Having a Shallow Trench EEPROM”), storing the same information on both memory planes.
- In contrast, U.S. application Ser. No. 09/385,380, describes a
memory cell 1′″ having first and second operating modes. Thememory cell 1′″ includes acharge transfer transistor 14 having a gate adjacent to a channel region coupling source and drain regions, adigitline 12 coupled to one of the source and drain regions, astorage capacitor 16 coupled to the other of the source and drain regions, aferroelectric capacitor 32, and awordline 10 coupled to the gate by theferroelectric capacitor 32. The polysilicon vertical interconnect connecting theferroelectric capacitor 32 with oxide layer covering the gate of thetransistor 14 forms anintrinsic capacitor 30. See FIG. 1(d). Data is written to and read out of the storage capacitor during the first operating mode, i.e., a normal DRAM access speeds, and written to and read out of the ferroelectric capacitor during the second mode of operation, i.e., at speeds several times slower than normal DRAM access speeds. - It would be very desirable if a memory cell or device could be developed with all of the positive features of DRAMS, i.e., cost, size, speed, availability, etc., which also is nonvolatile, i.e., maintains its memory state with power removed. It would be highly desirable to have a memory element and corresponding system that functions at speeds comparable to that of today's DRAMs. It would also be desirable if at least one logic device could be implement using a modified form of the memory cell.
- Based on the above and foregoing, it can be appreciated that there presently exists a need in the pertinent art which mitigates the above-described deficiencies.
- In one aspect, the present invention provides a one transistor/one capacitor (1T/1C) memory cell including a ferroelectric capacitor operatively coupled to the gate electrode of a charge amplifier. Preferably, the ferroelectric capacitor has a small surface area relative to the surface area of the gate electrode and, most preferably, the ferroelectric capacitor is an edge defined ferroelectric capacitor. In an exemplary embodiment, the charge amplifier is a transistor.
- In another aspect, the present invention provides a memory cell including a charge amplifier having a gate adjacent to a channel region coupling source and drain regions, a digitline coupled to one of the source and drain regions, a ground lead coupled to the other of the source and drain regions, a ferroelectric capacitor coupled to the gate, and a wordline coupled to the ferroelectric capacitor. Advantageously, the charge amplifier can be a transistor. Preferably, the gate is coupled to the ferroelectric capacitor by polysilicon, the junction formed at the gate has an intrinsic capacitance, and the capacitance of the ferroelectric capacitor is based on the magnitude of the intrinsic capacitance. Alternatively, the gate is coupled to the ferroelectric capacitor by polysilicon, the junction formed at the gate has an intrinsic capacitance, and the physical size of the ferroelectric capacitor is based on the magnitude of the intrinsic capacitance. In this case, the thickness of a ferroelectric material layer in the ferroelectric capacitor can be based on the magnitude of the intrinsic capacitance.
- In yet another aspect, the present invention provides a one transistor/one capacitor (1T/1C) memory cell array, wherein each of the 1T/1C memory cells is a ferroelectric capacitor operatively coupled to the gate electrode of a charge amplifier. Preferably, the charge amplifier is a transistor and, most preferably, the charge amplifier is a CMOS transistor. In an exemplary embodiment, the memory cell array is N rows by M columns of memory cells, and wherein the each of the M columns is defined by an associated digitline and an associated source line.
- In still another aspect, the present invention provides a memory cell array, including a plurality of memory cells organized as an array of rows and columns, at least one of the memory cells being a non-volatile memory cell including a charge transistor having a gate opposing a channel region coupling source and drain regions, a digitline coupled to the drain region, a source line coupled to the source region, a ferroelectric capacitor which stores data coupled to the gate, a wordline coupled to the ferroelectric capacitor, and a sense amplifier coupled to the digitline of the at least one memory cell.
- In another aspect, the present invention provides a processor based system, including a processor, and a memory cell array coupled to the processor, the memory cell array including a plurality of memory cells organized as an array of rows and columns, at least one of the memory cells being a non-volatile memory cell including a charge amplifier including a gate opposing a channel region coupling source and drain regions, a digitline coupled to the drain region, a source line coupled to the source region, a ferroelectric capacitor coupled to the gate, a wordline coupled to the ferroelectric capacitor, and a sense amplifier coupled to the digitline of the at least one memory cell.
- In a still further aspect, the present invention provides a memory module, including a substrate including a circuit board, a plurality of memory chips mounted on the die substrate, wherein one or more of the memory chips comprise a memory cell array fabricated on the semiconductor chip communicating with a processor, the memory cell array including at least one non-volatile memory cell including a ferroelectric capacitor operatively coupled to the gate of a charge amplifier.
- In still another aspect, the present invention provides a method for reading information from a one transistor/one capacitor (1T/1C) memory cell having a charge amplifier including a gate disposed adjacent to a channel region coupling source and drain regions, a digitline coupled to drain region, a source line coupled to the source region, a ferroelectric capacitor coupled to the gate, and a wordline coupled to the ferroelectric capacitor, the method including grounding the source line, asserting the wordline and digitline, amplifying the charge on the ferroelectric capacitor, and determining the resultant charge on the digitline.
- In yet another aspect, the present invention provides a method for writing information to a one transistor/one capacitor (1T/1C) memory cell having a charge amplifier including a gate disposed adjacent to a channel region coupling source and drain regions, a digitline coupled to drain region, a source line coupled to the source region, a ferroelectric capacitor coupled to the gate, and a wordline coupled to the ferroelectric capacitor, the method including asserting the source line, asserting the wordline to thereby apply a potential greater than the coercive voltage of the ferroelectric capacitor across the ferroelectric capacitor and the charge amplifier.
- In another aspect, the present invention provides a logic element, including N+1 source/drain regions, N channels, each channel coupling adjacent ones of the N+1 source/drain regions, N gates disposed adjacent to the N channels, N ferroelectric capacitors, each of the N ferroelectric capacitors coupled to a respective one of the N gates, N wordlines, each of the N wordlines coupled to a respective one of the N ferroelectric capacitors, a digitline couple to an end one of the N+1 source/drain regions, and a source line coupled to the other end one of the N+1 source/drain regions. Preferably, the logic element is operated by assertion of all of the N wordlines. Advantageously, the ferroelectric capacitors are edge defined ferroelectric capacitors.
- These and various other features and aspects of the present invention will be readily understood with reference to the following detailed description taken in conjunction with the accompanying drawings, in which:
- FIG. 1(a) illustrates a first use of a ferroelectric capacitor in a conventional memory cell;
- FIG. 1(a) illustrates a second use of a ferroelectric capacitor in a conventional memory cell;
- FIG. 1(c) illustrates a third use of a ferroelectric capacitor in a conventional memory cell;
- FIG. 1(d) illustrates a fourth use of a ferroelectric capacitor in a memory cell;
- FIG. 2(a) illustrates a NVRAM memory cell employing a ferroelectric capacitor according to the present invention;
- FIG. 2(b) illustrates a NVRAM memory cell array employing a ferroelectric capacitor according to the present invention;
- FIG. 3(a) illustrates selected portions of the components used in the memory cell depicted in FIGS. 2(a) and 2(b);
- FIG. 3(b) illustrates the hysteresis characteristic of the ferroelectric capacitor employed in the memory cell illustrated in FIGS. 2(a) and 2(b);
- FIG. 4 is a high-level block diagram of a memory array device which can be fabricated using the memory cell depicted in FIG. 2(b);
- FIG. 5 is a flow chart illustrating a method for accessing the memory cell depicted in FIG. 2(a);
- FIGS.6(a), 6(c) 6(e) illustrate the application of various voltages to selected portions of the circuitry illustrated in FIG. 2(a);
- FIGS.6(b), 6(d) 6(f) are hysteresis diagrams illustrating the effects on the ferroelectric capacitor of the voltages specified in FIGS. 6(a), 6(c) 6(e), respectively;
- FIG. 7 illustrates a NAND logic device which is formed using the basic NVRAM cell according to the present invention;
- FIG. 8 is a flow chart illustrating the process flow for fabricating the memory cell illustrated in FIG. 2(a);
- FIGS.9(a) through 9(o) illustrate one of plan and section view of a workpiece at various steps of the process flow illustrated in FIG. 8;
- FIG. 10 is a plan view of a memory module having memory chips constructed according to the present invention; and
- FIG. 11 is a high-level block diagram of a processor-based system employing NVRAM memory chips constructed according to the present invention.
- Illustrative embodiments and exemplary applications will now be described with reference to the accompanying drawings which disclose the advantageous teachings of the present invention.
- While the present invention is described herein with reference to illustrative embodiments for particular applications, it should be understood that the invention is not limited thereto. Those having ordinary skill in the art and access to the teachings provided herein will recognize additional modifications, applications, and embodiments and equivalent substitutions within the scope of the invention and additional fields in which the present invention would be of significant utility.
- FIG. 2(a) illustrates the basic transistor and
memory cell 100 electrical configuration according to the present invention, which preferably includes atransistor 114 connected between asource line 116 and adigitline 112. In addition, a stacked capacitor gate structure is provided, which includes apolysilicon gate capacitance 120 and a small areaferroelectric capacitor 118. More specifically, thememory cell 100 according to the present invention includes awordline 110, adigitline 112, a transfer device, e.g., a transistor, 114, asource line 116 and aferroelectric capacitor 118. It will be appreciated that the transition between the active region of thetransfer device 114 and thepolysilicon interconnect 122 to thewordline 110 is anoxide layer 120 a, which forms agate capacitor 120. Thebasic memory cell 100 advantageously can be employed in thememory cell array 142 illustrated in FIG. 2(b), which is discussed in greater detail below. - A more detailed illustration of the
transfer device 114,intrinsic gate capacitor 120,polysilicon interconnect 122 andferroelectric capacitor 118 is provided in FIG. 3(a). It should be mentioned that theferroelectric capacitor 118 preferably includes Platinum (Pt)electrodes ferroelectric layer 118 b, which will be described in greater detail below. The characteristics of theferroelectric layer 118 b and thegate oxide layer 120 a are listed in the table below. - FIG. 3(b) illustrates a hysteresis loop characteristic of
ferroelectric material 118 b in comparison to the linear response of anoxide material 120 a. In FIG. 3(b), VC indicates coercive voltage, Qr is remnant charge, Qs is saturation charge; the Y axis represents charge/centimeter squared in micro-coulombs. It will be appreciated that as long as the applied voltage is less than Vc, the charge will follow the linear profiles indicated forlayers layer 118 b follows the hysteresis profile. Thus, by applying a relatively large voltage, e.g., ±5.0 V, thelayer 118 b will exhibit a remnant charge Qr, which advantageously can be positive or negative. Since Qr will effect the threshold voltage Vth oftransistor 114, the polarity of Qr can be determined by ascertaining the effect on Vth, as discussed in greater detail below.TABLE I Ferroelectric Gate Oxide Layer Layer 118bLayer 120a Thickness 800Å 100Å Relative Permittivity εr 80 4 Area 0.2 μm2 1.0 μm2 Capacitance Per Unit Area 8.85 fF/μm2 3.54 fF/μm2 Capacitance of Device 1.77 fF 3.54 fF - It should also be mentioned that the ferroelectric characteristics depicted in FIG. 3(b) and Table I are that of a common ferroelectric, e.g., SrBiTaO, although other ferroelectric materials advantageously can be used, as discussed in greater detail below.
- It will be appreciated that since the effective dielectric constant of the ferroelectric material is so large, only a small area is required for the
ferroelectric capacitor 118. In contrast, thecapacitor 120 has the normal gate capacitance of a minimum-sized transistor 114 in CMOS technology. For purposes of illustration, the values presented in Table I are characteristic of 1 micron CMOS technology employing 5V power supplies. It will be appreciated that these values advantageously can be scaled, as appropriate, to smaller CMOS technologies with critical dimensions smaller than 1 micron. As mentioned above, the gate capacitance, i.e., the intrinsic capacitance ofcapacitor 120, of atransistor 114 in this technology with a gate oxide thickness of approximately 100 Å is about 3.3 fF. - As mentioned above, the charge storage element of a ferroelectric memory cell consists of a ferroelectric film sandwiched between two electrodes. Lead zirconate titanates, PbZrxTi1−xO3 (PZT) with x=0.4-0.53 are most studied materials for the ferroelectric film. It will be appreciated that this is because PZT films with thickness from 500 to 4000 Å can fill the ferroelectric material requirements to a substantial degree. Furthermore, it has been shown that PZT film processing can be done in combination with standard integrated circuit processing, as discussed in the article by R. Cuppens, et al., entitled “Ferroelectrics for Non-volatile Memories” (Microelectronic Engineering, 19 (1992), pp. 245-252, 1992). Other thin-film materials potentially usable for nonvolatile memory devices include BaMgF4 and Bi4Ti3O12 for application with the ferroelectric material gate insulator in a field effect transistor (FET). Moreover, BaTiO3, SrTiO3 and (Sr,Ba)TiO3 advantageously can be employed for application as dielectrics in DRAMs.
- Another important class of ferroelectric materials is the bismuth (Bi)-layer oxides. Bi-layer ferroelectric oxides have large polarization along the a or b axis, but no or little polarization along the c axis. Bi4Ti3O12 is a typical bi-layer ferroelectric material. However, the obtained remnant polarization is very small, i.e., less than 10 μC/cm2 and fatigue endurance is less than optimal. Recently, other types of bi-layer oxides, such as SrBi2Ta2O9 or SrBi2Nb2O9hve demonstrated acceptable endurance. Amanuma, et al., recently reported a simple solution deposition process of SrBi2Ta2O9 (SBT) thin films in the article entitled “Preparation and Ferroelectric Properties of SrBi2Ta2O9 Thin Films” (Appl. Phys. Lett., 66 (2), pp. 221-223, 1995).
- Referring back to FIG. 3(a), the
ferroelectric capacitor 118 is illustrated as a capacitor on top of the gate of thetransistor 114, which is connected to the gate oftransistor 114 bypolysilicon interconnect 122, which is, in an exemplary case, approximately 2000 Å thick. Preferably, theferroelectric capacitor 118 is fabricated by depositing a ferroelectric material between two Pt plates or electrodes, one of the electrodes, e.g.,electrode 118 a, being a vertical Pt wordline. Theferroelectric capacitor 118 advantageously has an effective plate area of only about one-fifth that of thegate oxide capacitor 120. Moreover, theferroelectric dielectric 118 b will typically be thicker, approximately 800 Å, than thegate oxide 120a, approximately 100 Å. It will be appreciated that by properly sizing theferroelectric capacitor 118 and theintrinsic gate capacitor 120, the gate oxide capacitance and the ferroelectric capacitances can be made comparable. Thus, when the coercive or programming voltage Vc of theferroelectric capacitor 118 is on the order of 1V, then thememory cell 100 can be programmed with a voltage as low as ±5V. It will be appreciated that both of these values are consistent with and, in fact, less than used in 1 micron flash memory CMOS technology. - Thus, one preferred embodiment of the present invention is a one transistor/one capacitor (1T/1C) memory cell including a ferroelectric capacitor operatively coupled to the gate electrode of a charge amplifier. Preferably, the ferroelectric capacitor has a small surface area relative to the surface area of the gate electrode and, most preferably, the ferroelectric capacitor is an edge defined ferroelectric capacitor, i.e., a capacitor formed by running the ferroelectric film over a step in the workpiece and then anisotropically etching the film in the vertical direction. In an exemplary embodiment, the charge amplifier is a transistor.
- An exemplary embodiment according to the present invention provides a memory cell including a charge amplifier having a gate adjacent to a channel region coupling source and drain regions, a digitline coupled to one of the source and drain regions, a ground lead coupled to the other of the source and drain regions, a ferroelectric capacitor coupled to the gate, and a wordline coupled to the ferroelectric capacitor. Advantageously, the charge amplifier can be a transistor. Preferably, the gate is coupled to the ferroelectric capacitor by polysilicon, the junction formed at the gate has an intrinsic capacitance, and the capacitance of the ferroelectric capacitor is based on the magnitude of the intrinsic capacitance. Alternatively, the gate is coupled to the ferroelectric capacitor by polysilicon, the junction formed at the gate has an intrinsic capacitance, and the physical size of the ferroelectric capacitor is based on the magnitude of the intrinsic capacitance. In this case, the thickness of a ferroelectric material layer in the ferroelectric capacitor can be based on the magnitude of the intrinsic capacitance. In an exemplary case, the ferroelectric capacitor is a dielectric layer of lead zirconate titanate (PZT). Most preferably, the PZT forming the ferroelectric capacitor is of the form PbZrxTi1−xO3 with x ranging between 0.4 and 0.53
- Referring now to FIG. 2(b), the
memory array 142 advantageously can be formed from the NVRAM memory cell illustrated in FIG. 2(a). Preferably, thememory cell array 142 includes a plurality of NVRAM memory cells, e.g.,cells NVRAM memory cell 100 a, advantageously can be addressed by a corresponding digit line, e.g., digit line D1, which preferably addresses other NVRAM memory cells, e.g.,cell 100 d, and which can be recharged for read out via a pullup transistor, e.g., transistor T1. It will be appreciated that transistor T1 is switched for readout in the normal manner. In addition, the source terminals for a column ofmemory cells 100 are connected to a ground via a switch S1. It will be appreciated that all of the switches advantageously can be closed during a read operation of thememory array 142 but can be selectively operated during a write operation, i.e., in order to write data to a selected one of the memory cells in thememory array 142, the corresponding source line is asserted. - Thus, an exemplary one transistor/one capacitor (1T/1C) memory cell array is provided by the present invention. In the exemplary case under discussion, each of the 1T/1C memory cells is a ferroelectric capacitor operatively coupled to the gate electrode of a charge amplifier. Preferably, the charge amplifier is a transistor and, most preferably, the charge amplifier is a CMOS transistor. In an exemplary embodiment, the memory cell array is N rows by M columns of memory cells, and wherein the each of the M columns is defined by an associated digitline and an associated source line.
- Referring to FIG. 4, an
exemplary DRAM circuit 140 is illustrated. TheDRAM 140 contains amemory cell array 142, row andcolumn decoders sense amplifier circuit 146. Thememory cell array 142 consists of a plurality of memory cells 100 (constructed as illustrated in FIG. 2(a)) whose word lines 110 anddigitlines 112 are commonly arranged into rows and columns, respectively. Thedigitlines 112 of thememory array 142 are connected to thesense amplifier circuit 146, while itsword lines 110 are connected to therow decoder 144. The source lines for allmemory cells 100 is grounded, preferably bysource line 116. Address and control signals are input into theDRAM 140 and connected to thecolumn decoder 148,sense amplifier circuit 146 androw decoder 144 and are used to gain read and write access, among other things, to thememory array 142. - The
column decoder 148 is connected to thesense amplifier circuit 146 via control and column select signals. Thesense amplifier circuit 146 receives input data destined for thememory array 142 and outputs data read from thememory array 142 over input/output (I/O) data lines. Data is read from the cells of thememory array 142 by activating a word line (via the row decoder 144), which couples all of thememory cells 100 corresponding to thatword line 110 torespective digitlines 112, which define the columns of the array. One or more of thedigitlines 112 are also activated. When a particular wordline-digitline pair is activated, thesense amplifier circuit 146 connected to a digitline column detects and amplifies the data bit impressed on thecapacitor 118 of thememory cell 100 to itsrespective digit line 112 by measuring the potential difference between the activated digitline and a reference line which may be an inactive bit line. The operation of DRAM sense amplifiers is described, for example, in U.S. Pat. Nos. 5,627,785, 5,280,205, and 5,042,011, all of which are assigned to Micron Technology Inc., and all of which are incorporated herein by reference for all purposes. - Thus, a memory cell array according to the present invention includes a plurality of memory cells organized as an array of rows and columns, where at least one of the memory is a non-volatile memory cell having a charge transistor having a gate opposing a channel region coupling source and drain regions, a digitline coupled to the drain region, a source line coupled to the source region, a ferroelectric capacitor which stores data coupled to the gate, a wordline coupled to the ferroelectric capacitor, and a sense amplifier coupled to the digitline of the at least one memory cell. Preferably, the ferroelectric capacitor has a surface area which is small relative to the surface area of the gate. Most preferably, the ferroelectric capacitor is an edge defined ferroelectric capacitor. In an exemplary case, the data is written into the ferroelectric capacitor when the wordline and source line are asserted and the data is read from the non-volatile memory cell when the wordline, the digitline and the source line are asserted. In an exemplary case, an asserted source line is a grounded source line. Preferably, the charge amplifier amplifies the charge stored in the ferroelectric capacitor by at least one order of magnitude. In any event, the charge amplifier amplifies the charge stored in the ferroelectric capacitor by a predetermined gain factor. The charge amplifier advantageously can be a transistor.
- The operation of the
memory cell 100 according to the present invention will now be described while referring to FIGS. 2(b), 5, and 6(a)-6(f). - Referring now to FIG. 5, after the system in which the
memory device 142 is located is initialized, a check is first performed to determine whether thememory cells 100 are being accessed for a read operation duringstep 502. If the answer is affirmative, all of the source lines are coupled to ground potential duringstep 504 by operation of switches S1, S2, etc., and then a read operation is performed in the normal manner for a DRAM, i.e., the wordline and digitlines for a particular memory cell are asserted, duringstep 506, which read operation is well known to one of ordinary skill in the art and will not be described further in the interest of brevity. However, the principal which permits thememory cell 100 to be read out normally will be discussed below. Irrespective of whetherstep 506 or step 510 (discussed below) is performed,step 502 is then repeated. It will be appreciated that as long as the voltage applied to thewordline 110 remains in the normal operating range of thetransistor 114, e.g., 3 V, the coercive voltage of theferroelectric capacitor 118 is not exceeded and the state of the ferroelectric capacitor is unchanged. See FIGS. 6(a) and 6(b). - When the answer at
step 502 is negative, the source line for all of the memory cells are deasserted, i.e., the switches S1, S2, etc., are opened duringstep 508. Then, a voltage in excess of the coercive voltage Vc for theferroelectric capacitor 118 is applied to thewordline 110 ofmemory cell 100 while thesource 116 for the selected cell(s) is(are) grounded and allother source lines 112 are open circuited. A voltage in excess of the coercive voltage Vc for theferroelectric capacitor 118 is thus applied by assertion of thewordline 110 andsource line 116 for the selected memory cell. It will be appreciated that the value stored in theferroelectric capacitor 118 is determined by the polarity of the voltage applied. For example, as illustrated in FIGS. 6(c) and 6(d), when a +5V potential is applied towordline 110, a value corresponding to one digital level, e.g., “one,” is stored in theferroelectric capacitor 118. In contrast, application of a −5V potential to the wordline results in the opposite digital level, i.e., “zero,” being stored in theferroelectric capacitor 118, as illustrated in FIGS. 6(e) and 6(f). It will also be appreciated that the correspondence between the stored charge polarity and the digital value is arbitrary; a positive potential advantageously can be used to write a digital “zero” intoferroelectric capacitor 118. It should be mentioned that the time needed to write to thememory cell 100 inmemory array 142 is substantially to the read time for the memory cell. - Thus, the read operation is achieved by driving the word line to some voltage less than the coercive or programming voltage, and reading the conductivity of the transistor. The capacitance of the
wordline 110 of eachmemory cell 100 advantageously be comparable to that in DRAMs since the gate capacitances are in series and, in the exemplary case under discussion, about 1fF. It should be mentioned at this point that unlike the transistor in a conventional DRAM, thetransistor 114 is not a passive transfer gate, but rather provides gain and charge multiplication. If the charge differences on theferroelectric capacitor 118 are 10 μC/cm2, and the area is 2×10−9 cm2, then the charge difference in thememory cell 100 itself is 2 fC. However, it will be appreciated that this charge is multiplied during the read operation by thetransistor 114. Thus, if the difference in charge states of theferroelectric capacitor 118 results in a drain current difference of 100 uA in thetransistor 114 during a read cycle of 1 nanosecond (ns), then the differences in charge on the data or bitline 112 during read operation is approximately 100 fC. It will be appreciated that this value is comparable to that obtained using current DRAMs equipped with 30 ferroelectric capacitor storage capacitors and 3 V potential differences. - Thus, one exemplary embodiment of the present invention is a method for reading information from a one transistor/one capacitor (1T/1C) memory cell having a charge amplifier including a gate disposed adjacent to a channel region coupling source and drain regions, a digitline coupled to drain region, a source line coupled to the source region, a ferroelectric capacitor coupled to the gate, and a wordline coupled to the ferroelectric capacitor, the method including grounding the source line, asserting the wordline and digitline, amplifying the charge on the ferroelectric capacitor, and determining the resultant charge on the digitline. Preferably, a voltage less than the coercive voltage of the ferroelectric capacitor is asserted on the wordline while a predetermined voltage, e.g., VDD, is asserted on the digitline.
- A complementary embodiment of the present invention is a method for writing information to a one transistor/one capacitor (1T/1C) memory cell having a charge amplifier including a gate disposed adjacent to a channel region coupling source and drain regions, a digitline coupled to drain region, a source line coupled to the source region, a ferroelectric capacitor coupled to the gate, and a wordline coupled to the ferroelectric capacitor, the method including asserting the source line, and asserting the wordline to thereby apply a potential greater than the coercive voltage of the ferroelectric capacitor across the ferroelectric capacitor and the charge amplifier.
- It will be appreciated that several ferroelectric capacitor/transistor pairs advantageously can be fabricated into a
NAND gate 175, as illustrated in FIG. 7. In theNAND gate 175, a plurality of ferroelectric capacitor FCA, FCB, FCC, and FCD are coupled to the gates of respective transistors TA, TB, TC, and TD. From FIG. 7, it will be noted that the output ofNAND gate 175 can be sensed ondigitline 179 when all of the gates are asserted. In an exemplary case, all of the gates of transistor TA-TD are asserted when thewordlines 177, e.g., wordlines W1, W2, W3, and W4, are all asserted. The source of transistor TA advantageously can be coupled to ground. - Stated another way, a logic device according to the present invention includes N+1 source/drain regions, N channels, each channel coupling adjacent ones of the N+1 source/drain regions, N gates disposed adjacent to the N channels, N ferroelectric capacitors, each of the N ferroelectric capacitors coupled to a respective one of the N gates, N wordlines, each of the N wordlines coupled to a respective one of the N ferroelectric capacitors, a digitline couple to an end one of the N+1 source/drain regions, and a source line coupled to the other end one of the N+1 source/drain regions. Preferably, the logic element is operated by assertion of all of the N wordlines. Advantageously, the ferroelectric capacitors are edge defined ferroelectric capacitors.
- As previously mentioned, the
ferroelectric capacitor 118 according to the present invention advantageously can be an edge-defined ferroelectric capacitor. Fabrication of the ferroelectric capacitor will be discussed with respect to FIG. 8. However, before describing an exemplary process for the manufacture ofmemory cell 100, a brief review of the currently available technology, from which preferred methodology for the fabrication of one type of edge-defined capacitors, will be presented. It will be appreciated that the method discussed immediately below advantageously may be applied to other types of edge-defined capacitors. - As discussed above, the
charge storage element 118 of aferroelectric memory cell 100 consists of aferroelectric film 118 b sandwiched between twoelectrodes - It will be appreciated that the use of sputtering for the deposition of PZT, Bi4Ti3O12 and oxides was developed in the 1970s. RF diode, magnetron and ion-beam sputtering techniques have all been used with metallic and oxide targets. However, it is difficult to prepare large, dense ceramic targets, which are necessary to obtain uniform layers on large substrates. Therefore, reactive sputtering using metallic targets is more commonly used. Because of the volatility of Pb and PbO, the control of stoichiometry is a major problem, which problem can be minimized by the use of multi-target or multi-target ion beam sputtering. This technique is suitable for edge-defined capacitors.
- Alternative techniques for the deposition of PZT films include Metal Organic Deposition (MOD) and a sol-gel process. With these processes, the ferroelectric film is formed by the thermal decomposition of a solution of organo metallic compounds which can be spun onto the substrate. The subsequent drying, firing and annealing processes of the film strongly influence the morphology and ferroelectric properties of the PZT. It should be mentioned that thicker films can be obtained by repeating the spinning, drying and firing sequence. The main advantage of these processes is the ease by which complex oxide films can be prepared. These techniques are applicable to edge-defined capacitors.
- Yet another technique which is useful in the fabrication of ferroelectric capacitors is Organo Metallic Chemical Vapor Deposition (OMCVD), which offers advantages such as good step coverage, high deposition rate, and low-temperature deposition, and is particularly suitable to a type of edge-defined capacitors which might have ferroelectric material on a sidewall. Additionally, an ozone jet reactive evaporation technique has been developed to deposit thin lead zirconate titanate films. It will be appreciated that because lead has a low affinity with oxygen and high volatility, lead may re-evaporate from the film and/or diff-use into the electrode and leave lead vacancies in PZT films. In a sol-gel or sputtering process, excess lead is often added to compensate for this lead loss. However, problems resulting from lead diffusion into surrounding electrode material was not elevated by this practice, i.e., the use of excess lead during sputtering. For a chemical vapor deposition (CVD) process, since the wafer is being heated during film deposition, unreacted lead may easily diff-use into the electrode material. To mitigate the problems associated with metallic lead or less oxidized lead in the film, it was common to employ high-concentration ozone to oxidize lead vapor at low temperature, and obtain a fully oxidized amorphous films by the co-evaporation method. Rapid thermal annealing (RTA), 650° C. for 120 seconds in oxygen ambient, is then used to crystallize the amorphous film into perovskite structure without lead re-evaporation or lead diffusion. It will be appreciated that this technique is particularly well suited to the fabrication of edge-defined capacitors. Other methods for the preparation of ferroelectric thin films include molecular beam epitaxy (MBE), although application of this fabrication method makes it difficult to comply with competing requirements of IC production processes such as large wafers, high throughput, etc.
- It will also be appreciated that an important parameter, one which will influence the properties of the ferroelectric capacitors, are the electrodes. Because of the chemical reactivity of PZT and related materials and relatively high processing temperatures (>600° C.), a very inert bottom electrode is required. In general, Pt-based electrodes are used which are deposited onto silicon oxide or nitride by sputtering using an adhesive layer such as TiN, Ti, or the two in combination. Although relatively inert, the Pt bottom electrode does affect the morphology and the ferroelectric properties of the PZT films and related materials.
- It has been reported in the literature that better hysteresis properties of PZT films (by sol-gel process) are achievable when the PZT is deposited onto a multilayer stack of Pt/RuO2/PZT/RuO2/Pt. It is speculated that the microstructure, in particular the grain size of PZT, of the PZT films were influenced by RuO2. Beneficially, films with smaller grain sizes had larger coercive fields. In addition, the effect of microstructure on the electrical characteristics of sol-gel derived PZT thin films deposited on a Pt electrode have been demonstrated using different solutions, i.e., films with different degree of preferred orientation of the perovskite crystal as well as different grain size were obtained. It is believed that while smaller grains are generally desired to obtain good device to device uniformity, the electrical properties of these smaller grains require further evaluation before they can be deemed suitable for either NVRAM MEMORY CELL, in particular, or DRAM applications, in general.
- As previously mentioned, another class of ferroelectric materials is the bismuth (Bi)-layer oxides. Bi-layer ferroelectric oxides have large polarization along the a or b axis, but no or little polarization along the c axis. Bi4Ti3O12 is a typical bi-layer ferroelectric, and the synthesis of this material has been reported. However, the obtained remnant polarization is very small, less than 10 μC/cm2 and fatigue endurance is poor. Recently, other types of bi-layer oxides, such as SrBi2Ta2O9 or SrBi2Nb2O9 were found to have good endurance.
- Referring now to FIG. 8, a process flow for integration of ferroelectric capacitors into the
memory devices 140 will now be described. It should be mentioned that the process flow is geared to produce a memory cell pair; the process advantageously can be employed in fabricating thememory array 142 or theNAND gate 175 by proper selection of various process masks. It should also be mentioned at this point that such integration should not degrade the ferroelectric properties of the film. One example of process integration with a stacked capacitor structure was shown by Onish, et al., in the article entitled “A Half-Micron Ferroelectric Memory Cell Technology with Stacked Capacitor Structure,” (1994 IEDM Digest, pp. 843-846, 1994). However, an alternative exemplary process flow is described below which advantageously can be employed in fabricating the ferroelectric capacitors within spacer-definedwordlines 518. It should also be mentioned that the process flow described below advantageously can be employed in fabricating memory cell pairs sharing a common digitline. - The process flow starts with a
workpiece 500, e.g., a wafer fragment, following well formation and growth of a thin (e.g., ˜10 nm) pad ofthermal oxide 502 followed by deposition of a thick pad of nitride 504 (˜0.2 μm) for formation of STI (shallow trench isolation) duringstep 600. Duringstep 602, resist is applied and masked, isolation trenches are etched and filled withoxide 506 and the structure is subjected to chemical mechanical processing (CMP), as in other, conventional process flows. This leaves the structure as shown in FIGS. 9(a) and 9(b). - During
step 604, thepad nitride 504 andoxide 502 is removed to expose bare silicon in the active areas, preferably recessed relative to the STI surface. Next, duringstep 606,gate oxide 508 is grown, prior to depositingpolysilicon gate material 510 and subsequent CMP finishing to leave structure as shown in FIGS. 9(c) and 9(d). - During
step 608, a resist 512 is applied, the gate pattern is exposed, and the wafer is etched to produce the intermediate product depicted in FIGS. 9)e) and 9(f). - During
step 610, the resist 512 is removed and the source/drain pockets 514 are ion implanted and annealed. Duringoptional step 612, the wafer can be finished to restore surface planarity by chemical vapor deposition (CVD) ofintrinsic polysilicon 516 followed by another round of CMP. Duringstep 614, theferroelectric capacitor 118 is formed by depositing aPt layer 118 a, aferroelectric material layer 118 b, and a finalPt cap layer 118 c on the surface of theintrinsic polysilicon 516 andgate material 510. It will be appreciated that any of the deposition processes discussed in detail above advantageously can be employed in depositing the ferroelectric material layer. Moreover, theferroelectric capacitor material 118 b advantageously can be chosen from the non-limiting list of materials presented above. Duringstep 616,intrinsic polysilicon 516 is deposited to a thickness equal to the desired height of the final spacer defined word lines (e.g., 0.2 μm). Then a resist 512′ is applied and masked to define word line edge pattern illustrated in FIGS. 9(g) and 9(h) duringstep 620. - During
step 620, the intrinsic polysilicon mandril material is directionally etched, preferably stopping on underlyingPt cap layer 118 c and then the resist 512′ is removed. Duringstep 622, a suitable word line conductor 518 (e.g., doped polysilicon, refractory metal or silicide) is deposited by CVD and then directionally etched to leaveconductor 518 as a spacer on the vertical edges of the mandril, i.e., to form edge-defined ferroelectric capacitors as shown in FIGS. 9(i) and 9(j). - During
step 624, a suitable selective etchant is employed to removeintrinsic polysilicon 516 mandril material. Instep 626, the exposed Pt-ferroelectric material-Pt sandwich workpiece 500 is selectively etched to remove any remainingintrinsic polysilicon 516 over the source/drain pockets 514, leaving the structure shown in FIGS. 9(k) and 9(l). - Finally, during
step 630, a thick oxide interlayer dielectric is deposited, planarized to form contact holes and wiring levels as in conventional processing. It will be appreciated that when thewordlines 110 are formed, these wordlines will be considerably wider than the underlyingferroelectric capacitor 118, as illustrated in FIG. 9(m). It will be appreciated that FIG. 9(n), which is a schematic diagram of the memory cells illustrated in FIG. 9(m) is included to orient the reader. - It should also be mentioned that the process discussed above advantageously could be employed in fabricating other structures. For example, the
memory array 142 advantageously could be fabricated as a number of subarrays, each having an associated block select switch, e.g., BS1, BS2, etc., serially coupling the local wordlines LW1, LW2, etc., to a global wordline GWL, as illustrated in FIG. 9(o). - FIG. 10 is plan view of a
memory module 600 having memory chips 60-68 including semiconductor memory devices constructed in accordance with the present invention. That is, chips 60-68 have a NVRAM cell such as described in connection with FIG. 2.Memory module 600 is a SIMM (single in line memory module) having nine memory chips (IC's) 60-68 aligned on one side of a printed circuit board substrate. The number of such memory chips in the SIMM typically will vary between 3 to 9. The circuit board 601 has anedge connector 602 along one longitudinal edge to permit it to plug into a memory socket on a computer motherboard of conventional design (not shown). A wiring pattern (not shown), which can be a conventionally known design for this purpose, is formed on the board 601 and connects the terminals or leads shown comprising theedge connector 602 to the memory chips 60-68. Smallceramic decoupling capacitors 69 are also mounted on substrate 601 to suppress transient voltage spikes. Other than the memory device structures formed in accordance with the inventive method and used in memory chips 60-68, the general layout of theSIMM 600 can be a conventional construction. - Thus, another exemplary embodiment of the present invention is a memory module, including a substrate in, for example, a circuit board, a plurality of memory chips mounted on the die substrate, wherein one or more of the memory chips comprise a memory cell array fabricated on a semiconductor chip for communicating with a processor, the memory cell array including at least one non-volatile memory cell including a ferroelectric capacitor operatively coupled to the gate of a charge amplifier. Preferably, the gate is coupled to the ferroelectric capacitor by polysilicon, the junction formed at the gate has an intrinsic capacitance, and the capacitance of the ferroelectric capacitor is based on the magnitude of the intrinsic capacitance. Alternatively, the gate is coupled to the ferroelectric capacitor by polysilicon, the junction formed at the gate has an intrinsic capacitance, and the physical size of the ferroelectric capacitor is based on the magnitude of the intrinsic capacitance. Preferably, the ferroelectric capacitor includes a dielectric layer of lead zirconate titanate (PZT), which advantageously can be of the form PbZrxTi1−xO3 with x ranging between 0.4 and 0.53. In an exemplary case, the non-volatile memory cell includes the charge amplifier having a gate disposed adjacent to a channel region coupling source and drain regions, a digitline coupled to drain region, a source line coupled to the source region, the ferroelectric capacitor coupled to the gate, and a wordline coupled to the ferroelectric capacitor.
- FIG. 11 is a block diagram of a processor-based
system 400 utilizing amemory 412 constructed in accordance with the present invention. That is, thememory 412 utilizes thememory cells 100 illustrated in FIG. 2(a), which advantageously can be provided in the form of thememory device 140. The processor-basedsystem 400 may be a computer system, a process control system or any other system employing a processor and associated memory. Thesystem 400 includes a central processing unit (CPU) 402, e.g., a microprocessor, that communicates with theRAM 412 and an I/O device 408 over abus 420. It must be noted that thebus 420 may be a series of buses and bridges commonly used in a processor-based system, but for convenience purposes only, thebus 420 has been illustrated as a single bus. A second I/O device 410 is illustrated, but is not necessary to practice the invention. The processor-basedsystem 400 also includes anadditional memory 414, which could be either a read-only memory (ROM) or another of thememory devices 140 according to the present invention, since, as discussed above, the memory cells are non-volatile devices. The processor based system may include peripheral devices such as afloppy disk drive 404, a compact disk (CD)ROM drive 406, a display (not shown), a key board (not shown), and a mouse (not shown), that communicate with theCPU 402 over thebus 420 as is well known in the art. - It should be noted that since the
memories memory device 140 according to the present invention, the number of discrete memory devices needed for the processor basedsystem 400 can be reduced. Stated another way, since thememory array 142 according to the present invention is populated bynon-volatile memory cells 100, a predetermined portion of thememory device 142 advantageously can be employed as the ROM in the processor basedsystem 400. - Thus, another preferred embodiment of the present invention includes a processor based system, including a processor, and a memory cell array coupled to the processor, the memory cell array including a plurality of memory cells organized as an array of rows and columns, at least one of the memory cells being a non-volatile memory cell including a charge amplifier including a gate opposing a channel region coupling source and drain regions, a digitline coupled to the drain region, a source line coupled to the source region, a ferroelectric capacitor coupled to the gate, a wordline coupled to the ferroelectric capacitor, and a sense amplifier coupled to the digitline of the at least one memory cell. Preferably, data is written into the ferroelectric capacitor when the wordline and the source line are both asserted, while data is read out of the memory cell when the wordline, the digitline and the source line are asserted, under the control of the processor. Advantageously, the ferroelectric capacitor is an edge defined ferroelectric capacitor. Preferably, the ferroelectric capacitor is a dielectric layer of lead zirconate titanate (PZT) and, most preferably, the PZT forming the ferroelectric capacitor is of the form PbZrxTi1−xO3 with x ranging between 0.4 and 0.53.
- It should again be mentioned that all of the patents, patent applications, and articles cited to or discussed above are incorporated herein by reference for all purposes.
- Although presently preferred embodiments of the present invention have been described and illustrated, it should be clearly understood that many variations and/or modifications can be made without departing from spirit and scope of the invention. Accordingly, the invention is not limited by the foregoing description but is only limited by the scope of the attached claims.
Claims (55)
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US09/461,009 US6337805B1 (en) | 1999-08-30 | 1999-12-15 | Discrete devices including EAPROM transistor and NVRAM memory cell with edge defined ferroelectric capacitance, methods for operating same, and apparatuses including same |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US09/385,380 US6141238A (en) | 1999-08-30 | 1999-08-30 | Dynamic random access memory (DRAM) cells with repressed ferroelectric memory methods of reading same, and apparatuses including same |
US09/461,009 US6337805B1 (en) | 1999-08-30 | 1999-12-15 | Discrete devices including EAPROM transistor and NVRAM memory cell with edge defined ferroelectric capacitance, methods for operating same, and apparatuses including same |
Related Parent Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US09/385,380 Continuation-In-Part US6141238A (en) | 1999-08-30 | 1999-08-30 | Dynamic random access memory (DRAM) cells with repressed ferroelectric memory methods of reading same, and apparatuses including same |
Publications (2)
Publication Number | Publication Date |
---|---|
US20020001219A1 true US20020001219A1 (en) | 2002-01-03 |
US6337805B1 US6337805B1 (en) | 2002-01-08 |
Family
ID=46276583
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US09/461,009 Expired - Fee Related US6337805B1 (en) | 1999-08-30 | 1999-12-15 | Discrete devices including EAPROM transistor and NVRAM memory cell with edge defined ferroelectric capacitance, methods for operating same, and apparatuses including same |
Country Status (1)
Country | Link |
---|---|
US (1) | US6337805B1 (en) |
Cited By (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20040125629A1 (en) * | 2002-12-31 | 2004-07-01 | Scheuerlein Roy E. | Programmable memory array structure incorporating series-connected transistor strings and methods for fabrication and operation of same |
US20040152342A1 (en) * | 2003-02-04 | 2004-08-05 | Micron Technology, Inc. | Method of eliminating residual carbon from flowable oxide fill |
US20050023624A1 (en) * | 2002-06-05 | 2005-02-03 | Micron Technology, Inc. | Atomic layer-deposited HfAlO3 films for gate dielectrics |
US20050132308A1 (en) * | 2003-12-15 | 2005-06-16 | Bruce Whitefield | Method for calculating high-resolution wafer parameter profiles |
US6917380B1 (en) * | 2000-10-05 | 2005-07-12 | Ess Technology, Inc. | One time programmable solid-state device |
US20060091437A1 (en) * | 2004-11-02 | 2006-05-04 | Samsung Electronics Co., Ltd. | Resistive memory device having array of probes and method of manufacturing the resistive memory device |
US20060148180A1 (en) * | 2005-01-05 | 2006-07-06 | Micron Technology, Inc. | Atomic layer deposited hafnium tantalum oxide dielectrics |
US20060223337A1 (en) * | 2005-03-29 | 2006-10-05 | Micron Technology, Inc. | Atomic layer deposited titanium silicon oxide films |
US20060281330A1 (en) * | 2005-06-14 | 2006-12-14 | Micron Technology, Inc. | Iridium / zirconium oxide structure |
US20070059881A1 (en) * | 2003-03-31 | 2007-03-15 | Micron Technology, Inc. | Atomic layer deposited zirconium aluminum oxide |
US20070101929A1 (en) * | 2002-05-02 | 2007-05-10 | Micron Technology, Inc. | Methods for atomic-layer deposition |
US8084370B2 (en) | 2006-08-31 | 2011-12-27 | Micron Technology, Inc. | Hafnium tantalum oxynitride dielectric |
US8501563B2 (en) | 2005-07-20 | 2013-08-06 | Micron Technology, Inc. | Devices with nanocrystals and methods of formation |
Families Citing this family (25)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20050094457A1 (en) * | 1999-06-10 | 2005-05-05 | Symetrix Corporation | Ferroelectric memory and method of operating same |
EP1126525A3 (en) * | 2000-02-15 | 2005-10-19 | Matsushita Electric Industrial Co., Ltd. | Semiconductor memory device, method for driving the same and method for fabricating the same |
US6587365B1 (en) | 2000-08-31 | 2003-07-01 | Micron Technology, Inc. | Array architecture for depletion mode ferroelectric memory devices |
US6515889B1 (en) * | 2000-08-31 | 2003-02-04 | Micron Technology, Inc. | Junction-isolated depletion mode ferroelectric memory |
US7088002B2 (en) * | 2000-12-18 | 2006-08-08 | Intel Corporation | Interconnect |
US6953730B2 (en) * | 2001-12-20 | 2005-10-11 | Micron Technology, Inc. | Low-temperature grown high quality ultra-thin CoTiO3 gate dielectrics |
US6768144B2 (en) * | 2001-12-31 | 2004-07-27 | Texas Instruments Incorporated | Method and apparatus for reducing leakage current in an SRAM array |
ITTO20020118A1 (en) * | 2002-02-08 | 2003-08-08 | St Microelectronics Srl | INTEGRATING DEVICE A NON VOLATILE MEMORY MATRIX AND A VOLATILE MEMORY MATRIX. |
US7589029B2 (en) * | 2002-05-02 | 2009-09-15 | Micron Technology, Inc. | Atomic layer deposition and conversion |
US6970370B2 (en) * | 2002-06-21 | 2005-11-29 | Micron Technology, Inc. | Ferroelectric write once read only memory for archival storage |
US7193893B2 (en) * | 2002-06-21 | 2007-03-20 | Micron Technology, Inc. | Write once read only memory employing floating gates |
US6888739B2 (en) | 2002-06-21 | 2005-05-03 | Micron Technology Inc. | Nanocrystal write once read only memory for archival storage |
US7154140B2 (en) * | 2002-06-21 | 2006-12-26 | Micron Technology, Inc. | Write once read only memory with large work function floating gates |
US6996009B2 (en) | 2002-06-21 | 2006-02-07 | Micron Technology, Inc. | NOR flash memory cell with high storage density |
US6804136B2 (en) * | 2002-06-21 | 2004-10-12 | Micron Technology, Inc. | Write once read only memory employing charge trapping in insulators |
US7221586B2 (en) * | 2002-07-08 | 2007-05-22 | Micron Technology, Inc. | Memory utilizing oxide nanolaminates |
US7221017B2 (en) * | 2002-07-08 | 2007-05-22 | Micron Technology, Inc. | Memory utilizing oxide-conductor nanolaminates |
US7233201B2 (en) * | 2004-08-31 | 2007-06-19 | Micron Technology, Inc. | Single-ended pseudo-differential output driver |
JP4748978B2 (en) * | 2004-12-02 | 2011-08-17 | 日本碍子株式会社 | Piezoelectric / electrostrictive element and manufacturing method thereof |
US7560335B2 (en) * | 2005-08-30 | 2009-07-14 | Micron Technology, Inc. | Memory device transistors |
US7709402B2 (en) * | 2006-02-16 | 2010-05-04 | Micron Technology, Inc. | Conductive layers for hafnium silicon oxynitride films |
JP2008153479A (en) * | 2006-12-19 | 2008-07-03 | Rohm Co Ltd | Semiconductor device with ferroelectric field-effect transistor, and semiconductor integrated circuit device using the same |
US7916513B2 (en) * | 2008-11-05 | 2011-03-29 | Seagate Technology Llc | Non-destructive read back for ferroelectric data storage device |
US10636471B2 (en) * | 2016-04-20 | 2020-04-28 | Micron Technology, Inc. | Memory arrays, ferroelectric transistors, and methods of reading and writing relative to memory cells of memory arrays |
US10957373B2 (en) * | 2018-07-05 | 2021-03-23 | Samsung Electronics Co., Ltd. | Semiconductor memory device |
Family Cites Families (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US5303182A (en) * | 1991-11-08 | 1994-04-12 | Rohm Co., Ltd. | Nonvolatile semiconductor memory utilizing a ferroelectric film |
JPH0745794A (en) * | 1993-07-26 | 1995-02-14 | Olympus Optical Co Ltd | Drive method for ferroelectric memory |
JP3505758B2 (en) * | 1993-12-28 | 2004-03-15 | ローム株式会社 | Non-volatile semiconductor memory |
JP3570692B2 (en) * | 1994-01-18 | 2004-09-29 | ローム株式会社 | Non-volatile memory |
US5640030A (en) * | 1995-05-05 | 1997-06-17 | International Business Machines Corporation | Double dense ferroelectric capacitor cell memory |
JP2800745B2 (en) * | 1995-11-10 | 1998-09-21 | 日本電気株式会社 | Ferroelectric memory |
JPH09205181A (en) * | 1996-01-26 | 1997-08-05 | Nec Corp | Semiconductor device |
JP3766181B2 (en) * | 1996-06-10 | 2006-04-12 | 株式会社東芝 | Semiconductor memory device and system equipped with the same |
-
1999
- 1999-12-15 US US09/461,009 patent/US6337805B1/en not_active Expired - Fee Related
Cited By (29)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6917380B1 (en) * | 2000-10-05 | 2005-07-12 | Ess Technology, Inc. | One time programmable solid-state device |
US20070101929A1 (en) * | 2002-05-02 | 2007-05-10 | Micron Technology, Inc. | Methods for atomic-layer deposition |
US7670646B2 (en) | 2002-05-02 | 2010-03-02 | Micron Technology, Inc. | Methods for atomic-layer deposition |
US20050023624A1 (en) * | 2002-06-05 | 2005-02-03 | Micron Technology, Inc. | Atomic layer-deposited HfAlO3 films for gate dielectrics |
US7505321B2 (en) * | 2002-12-31 | 2009-03-17 | Sandisk 3D Llc | Programmable memory array structure incorporating series-connected transistor strings and methods for fabrication and operation of same |
US20040125629A1 (en) * | 2002-12-31 | 2004-07-01 | Scheuerlein Roy E. | Programmable memory array structure incorporating series-connected transistor strings and methods for fabrication and operation of same |
US20040212036A1 (en) * | 2003-02-04 | 2004-10-28 | Micron Technology, Inc. | Method of eliminating residual carbon from flowable oxide fill |
US7205248B2 (en) | 2003-02-04 | 2007-04-17 | Micron Technology, Inc. | Method of eliminating residual carbon from flowable oxide fill |
US20040152342A1 (en) * | 2003-02-04 | 2004-08-05 | Micron Technology, Inc. | Method of eliminating residual carbon from flowable oxide fill |
US20070059881A1 (en) * | 2003-03-31 | 2007-03-15 | Micron Technology, Inc. | Atomic layer deposited zirconium aluminum oxide |
US20050132308A1 (en) * | 2003-12-15 | 2005-06-16 | Bruce Whitefield | Method for calculating high-resolution wafer parameter profiles |
US20060091437A1 (en) * | 2004-11-02 | 2006-05-04 | Samsung Electronics Co., Ltd. | Resistive memory device having array of probes and method of manufacturing the resistive memory device |
US7687838B2 (en) * | 2004-11-02 | 2010-03-30 | Samsung Electronics Co., Ltd. | Resistive memory device having array of probes and method of manufacturing the resistive memory device |
US20060148180A1 (en) * | 2005-01-05 | 2006-07-06 | Micron Technology, Inc. | Atomic layer deposited hafnium tantalum oxide dielectrics |
US8278225B2 (en) | 2005-01-05 | 2012-10-02 | Micron Technology, Inc. | Hafnium tantalum oxide dielectrics |
US8524618B2 (en) | 2005-01-05 | 2013-09-03 | Micron Technology, Inc. | Hafnium tantalum oxide dielectrics |
US20100029054A1 (en) * | 2005-01-05 | 2010-02-04 | Ahn Kie Y | Hafnium tantalum oxide dielectrics |
US20070181931A1 (en) * | 2005-01-05 | 2007-08-09 | Micron Technology, Inc. | Hafnium tantalum oxide dielectrics |
US8399365B2 (en) | 2005-03-29 | 2013-03-19 | Micron Technology, Inc. | Methods of forming titanium silicon oxide |
US7687409B2 (en) | 2005-03-29 | 2010-03-30 | Micron Technology, Inc. | Atomic layer deposited titanium silicon oxide films |
US8076249B2 (en) | 2005-03-29 | 2011-12-13 | Micron Technology, Inc. | Structures containing titanium silicon oxide |
US20060223337A1 (en) * | 2005-03-29 | 2006-10-05 | Micron Technology, Inc. | Atomic layer deposited titanium silicon oxide films |
US20060281330A1 (en) * | 2005-06-14 | 2006-12-14 | Micron Technology, Inc. | Iridium / zirconium oxide structure |
US7510983B2 (en) | 2005-06-14 | 2009-03-31 | Micron Technology, Inc. | Iridium/zirconium oxide structure |
US8501563B2 (en) | 2005-07-20 | 2013-08-06 | Micron Technology, Inc. | Devices with nanocrystals and methods of formation |
US8921914B2 (en) | 2005-07-20 | 2014-12-30 | Micron Technology, Inc. | Devices with nanocrystals and methods of formation |
US8084370B2 (en) | 2006-08-31 | 2011-12-27 | Micron Technology, Inc. | Hafnium tantalum oxynitride dielectric |
US8466016B2 (en) | 2006-08-31 | 2013-06-18 | Micron Technolgy, Inc. | Hafnium tantalum oxynitride dielectric |
US8759170B2 (en) | 2006-08-31 | 2014-06-24 | Micron Technology, Inc. | Hafnium tantalum oxynitride dielectric |
Also Published As
Publication number | Publication date |
---|---|
US6337805B1 (en) | 2002-01-08 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US6337805B1 (en) | Discrete devices including EAPROM transistor and NVRAM memory cell with edge defined ferroelectric capacitance, methods for operating same, and apparatuses including same | |
US6141238A (en) | Dynamic random access memory (DRAM) cells with repressed ferroelectric memory methods of reading same, and apparatuses including same | |
US6700146B2 (en) | Semiconductor memory device and method for producing the same | |
US6190957B1 (en) | Method of forming a ferroelectric device | |
US5371699A (en) | Non-volatile ferroelectric memory with folded bit lines and method of making the same | |
US5751037A (en) | Non-volatile memory cell having dual gate electrodes | |
US5416735A (en) | Non-volatile random access memory with ferroelectric capacitor | |
US6303958B1 (en) | Semiconductor integrated circuit and method for manufacturing the same | |
US6967365B2 (en) | Ferroelectric memory cell with angled cell transistor active region and methods for fabricating the same | |
US7425738B2 (en) | Metal thin film and method of manufacturing the same, dielectric capacitor and method of manufacturing the same, and semiconductor device | |
US5262343A (en) | DRAM stacked capacitor fabrication process | |
US6952362B2 (en) | Ferroelectric write once read only memory for archival storage | |
US6908802B2 (en) | Ferroelectric circuit element that can be fabricated at low temperatures and method for making the same | |
JPH08111510A (en) | Semiconductor device | |
US5963466A (en) | Ferroelectric memory having a common plate electrode | |
JP3249470B2 (en) | Nonvolatile semiconductor memory device and method of manufacturing the same | |
US6511877B2 (en) | Semiconductor integrated circuit and method for manufacturing the same | |
US6807082B2 (en) | Semiconductor device and method of manufacturing the same | |
US7015531B2 (en) | FeRAM having bottom electrode connected to storage node and method for forming the same | |
US7050322B2 (en) | Device integrating a nonvolatile memory array and a volatile memory array | |
WO2003103050A2 (en) | Variable capacitances for memory cells within a cell group | |
US20010011739A1 (en) | Ferroelectric random access memory device | |
JP3715551B2 (en) | Manufacturing method of semiconductor device | |
US6114214A (en) | Method for forming a high-density dram cell with a double-crown rugged polysilicon capacitor | |
JPH09213899A (en) | Non-volatile memory device provided with ferroelectric film |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
AS | Assignment |
Owner name: MICRON TECHNOLOGY, INC., IDAHO Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:FORBES, LEONARD;AHN, KIE Y.;NOBLE, WENDELL P.;REEL/FRAME:010466/0009;SIGNING DATES FROM 19991203 TO 19991208 |
|
CC | Certificate of correction | ||
FPAY | Fee payment |
Year of fee payment: 4 |
|
FEPP | Fee payment procedure |
Free format text: PAYER NUMBER DE-ASSIGNED (ORIGINAL EVENT CODE: RMPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY Free format text: PAYOR NUMBER ASSIGNED (ORIGINAL EVENT CODE: ASPN); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY |
|
FPAY | Fee payment |
Year of fee payment: 8 |
|
REMI | Maintenance fee reminder mailed | ||
LAPS | Lapse for failure to pay maintenance fees | ||
STCH | Information on status: patent discontinuation |
Free format text: PATENT EXPIRED DUE TO NONPAYMENT OF MAINTENANCE FEES UNDER 37 CFR 1.362 |
|
FP | Lapsed due to failure to pay maintenance fee |
Effective date: 20140108 |