Patent application title: FLASH MEMORY HAVING A HIGH-PERMITTIVITY TUNNEL DIELECTRIC
Leonard Forbes (Corvallis, OR, US)
IPC8 Class: AH01L21336FI
Class name: Having insulated gate (e.g., igfet, misfet, mosfet, etc.) having additional gate electrode surrounded by dielectric (i.e., floating gate) multiple interelectrode dielectrics or nonsilicon compound gate insulator
Publication date: 2009-07-30
Patent application number: 20090191676
Patent application title: FLASH MEMORY HAVING A HIGH-PERMITTIVITY TUNNEL DIELECTRIC
LEFFERT JAY & POLGLAZE, P.A.;Attn: Thomas W. Leffert
Origin: MINNEAPOLIS, MN US
IPC8 Class: AH01L21336FI
A high permittivity tunneling dielectric is used in a flash memory cell to
provide greater tunneling current into the floating gate with smaller
gate voltages. The flash memory cell has a substrate with source/drain
regions. The high-k tunneling dielectric is formed above the substrate.
The high-k tunneling dielectric can be deposited using evaporation
techniques or atomic layer deposition techniques. The floating gate is
formed on top of the high-k dielectric layer with an oxide gate insulator
on top of that. A polysilicon control gate is formed on the top gate
1. A method for fabricating a flash memory cell, the method
comprising:creating a plurality of source/drain regions by doping
portions of a substrate;forming an atomic layer deposition (ALD) tunnel
dielectric layer with an atomic layer deposition technique on the
substrate substantially between the plurality of source/drain regions,
the ALD tunnel dielectric layer having a dielectric constant that is
higher than silicon dioxide;depositing a floating gate on the evaporated
tunnel dielectric layer;depositing an oxide insulator material on the
floating gate; andforming a control gate on the oxide insulator material.
2. The method of claim 1 wherein forming an ALD tunnel gate insulator comprises depositing one of AlOX, LaAlO3, HfAlO3, Pr2O3, Lanthanide-doped TiOx, HfSiON, Zr--Sn--Ti--O films using TiCl4 or TiI4, ZrON, HfO2/Hf, ZrAlXOy, CrTiO3, or ZrTiO.sub.4.
3. A method for fabricating a memory device on a substrate, the memory device having a plurality of memory cells, the method comprising:doping a source region and a drain region into the substrate for each memory cell;forming a tunnel dielectric layer over the substrate substantially between each pair of source and drain regions, each tunnel dielectric layer having a barrier height that is less than silicon dioxide;forming a floating gate over each tunnel dielectric layer;forming a gate insulator over each floating gate; andforming a gate over each gate insulator.
4. The method of claim 3 wherein the tunnel dielectric layer comprises a high permittivity gate dielectric.
5. The method of claim 4 wherein the tunnel dielectric layer has a band gap of 6.6 eV and a conduction band offset of 2.1 eV.
6. The method of claim 3 wherein the tunnel dielectric layer is formed using atomic layer deposition using sequential deposition of individual monolayers or fractions of a monolayer where gaseous precursors are introduced one at a time to the substrate.
7. The method of claim 3 wherein the floating gate comprises a polysilicon layer.
8. The method of claim 3 wherein the gate insulator comprises an interpoly oxide insulator layer.
9. The method of claim 3 wherein the tunnel dielectric layer is less than 15 Å thick.
10. A method for fabricating a memory device, the method comprising:forming a pair of source/drain regions in the substrate for each memory cell of the memory device;forming a tunnel dielectric with atomic layer deposition, using an oxygen source of one of hydrogen peroxide, ozone, or nitrous oxide, over the substrate and substantially between each pair of source/drain regions, each tunnel dielectric layer having a barrier height that is less than silicon dioxide and is comprised of a metal compound;forming a floating gate over each tunnel dielectric;forming a gate insulator over each floating gate; andforming a gate over each gate insulator.
11. The method of claim 10 wherein the tunnel dielectric is formed by a plurality of reaction steps, including a first precursor that is saturatively chemisorbed at the substrate surface, in a flow-type reactor.
12. The method of claim 11 and further including purging the first precursor and introducing a second precursor that is introduced on the substrate for films growth reaction.
13. The method of claim 12 and further including purging byproducts and the second precursor.
14. The method of claim 10 wherein the tunnel dielectric is formed from one of source materials: zirconium tetrachloride (ZrCl4) for a Zr film, titanium tetraisopropoxide (Ti(OCH(CH3)2)4) for a Ti film, trimethyl aluminum (Al(CH3)3) for an Al film, chromyl chromide (CrO2Cl2) for a Cr film, praseodymium chloride (PrCl3) for a Pr film, and hafnium chloride (HfCl4) for an Hf film.
15. The method of claim 14 wherein thin oxide films are deposited on the substrate at a temperature that is high enough such that it is absorbed to the substrate surface.
16. The method of claim 15 wherein vaporized source material reacts with a molecular layer of a second source material.
17. The method of claim 16 wherein the vaporized source material becomes absorbed and reacts with the second source material directed to the substrate surface in a subsequent fabrication step.
18. The method of claim 10 wherein each memory cell is scaled below 50 nm.
19. The method of claim 10 wherein the oxygen sources further comprise water and alcohols.
20. The method of claim 10 wherein tunnel dielectric materials include one of Si3N4, Al2O3, Y2O3, La2O3, Ta2O3, TiO2, HfO2, ZrO.sub.2.
This Application is a Divisional of U.S. application Ser. No. 11/209,128, titled "FLASH MEMORY HAVING A HIGH-PERMITTIVITY TUNNEL DIELECTRIC," filed Aug. 22, 2005 (Allowed), which is a Divisional of U.S. application Ser. No. 10/739,253, filed Dec. 18, 2003, now U.S. Pat. No. 7,157,769 issued on Jan. 2, 2007, which are commonly assigned and incorporated herein by reference.
The present invention relates generally to memory devices and in particular the present invention relates to flash memory device architecture.
BACKGROUND OF THE INVENTION
Memory devices are typically provided as internal, semiconductor, integrated circuits in computers or other electronic devices. There are many different types of memory including random-access memory (RAM), read only memory (ROM), dynamic random access memory (DRAM), synchronous dynamic random access memory (SDRAM), and flash memory. One type of flash memory is a nitride read only memory (NROM). NROM has some of the characteristics of flash memory but does not require the special fabrication processes of flash memory. NROM integrated circuits can be implemented using a standard CMOS process.
Flash memory devices have developed into a popular source of non-volatile memory for a wide range of electronic applications. Flash memory devices typically use a one-transistor memory cell that allows for high memory densities, high reliability, and low power consumption. Common uses for flash memory include personal computers, personal digital assistants (PDAs), digital cameras, and cellular telephones. Program code and system data such as a basic input/output system (BIOS) are typically stored in flash memory devices for use in personal computer systems.
The performance of flash memory transistors needs to increase as the performance of computer systems increases. To accomplish a performance increase, the transistors can be reduced in size. This has the effect of increased speed with decreased power requirements.
However, a problem with decreased flash memory size is that flash memory cell technologies have some scaling limitations. For example, stress induced leakage typically requires a tunnel oxide above 60 Å. This thickness results in a scaling limit on the gate length. Additionally, this gate oxide thickness limits the read current and may require large gate widths.
For the reasons stated above, and for other reasons stated below which will become apparent to those skilled in the art upon reading and understanding the present specification, there is a need in the art for a more scalable, higher performance flash memory transistor.
BRIEF DESCRIPTION OF THE DRAWINGS
FIG. 1 shows a cross-sectional view of a flash memory cell transistor of the present invention.
FIG. 2 shows an energy-band diagram in accordance with a write operation to the transistor structure of FIG. 1.
FIG. 3 shows an energy-band diagram in accordance with an erase operation from the transistor structure of FIG. 1.
FIG. 4 shows a plot of tunneling current dependence on barrier height for various electric fields in accordance with the transistor structure of FIG. 1.
FIG. 5 shows a block diagram of an electronic system of the present invention.
In the following detailed description of the invention, reference is made to the accompanying drawings that form a part hereof and in which is shown, by way of illustration, specific embodiments in which the invention may be practiced. In the drawings, like numerals describe substantially similar components throughout the several views. These embodiments are described in sufficient detail to enable those skilled in the art to practice the invention. Other embodiments may be utilized and structural, logical, and electrical changes may be made without departing from the scope of the present invention. The following detailed description is, therefore, not to be taken in a limiting sense, and the scope of the present invention is defined only by the appended claims and equivalents thereof.
FIG. 1 illustrates a cross-sectional view of a flash memory cell transistor of the present invention. The transistor is comprised of two source/drain regions 101 and 102 in a silicon substrate 111. Which region 101 or 102 functions as source and which functions as drain is determined by the direction of operation of the transistor.
In one embodiment, the source/drain regions 101 and 102 are n+ doped regions in a p-type substrate 111. An alternate embodiment may use p+ doped source/drain regions in an n-type substrate. The present invention is not limited to any one conductivity type for the source/drain regions or the substrate.
A high-permittivity (high-k) tunnel gate dielectric 103 is formed on top of the substrate 111 between the source/drain regions 101 and 102. A polysilicon floating gate layer 105 is formed on top of the tunnel gate dielectric layer 103. An interpoly oxide insulator layer 107 is formed on top of the floating gate 105. A polysilicon control gate 109 is formed on top of the oxide insulator 107.
In one embodiment, a high dielectric constant is considered to be a dielectric constant that is greater than that of SiO2. In one embodiment, LaAlO is used as the high-k tunneling gate dielectric 103 instead of the prior art SiO2 tunneling gate dielectric. Alternate embodiments use other dielectrics having other dielectric constants. These dielectric materials include Si3N4, Al2O3, Y2O3, La2O3, Ta2O3, TiO2, HfO2, ZrO2, or other high dielectric constant materials. The characteristics of these materials are well known to those skilled in the art and are not discussed further.
The LaAlO dielectric material is a high-k tunneling dielectric that has a band gap of 6.6 eV and a conduction band offset of 2.1 eV. Both of these are smaller than the prior art SiO2. Even though the barrier height of the LaAlO dielectric is less than that of SiO2 (i.e., 3.2 eV), the leakage current resulting from tunneling is still low enough to meet the requirements of intrinsic storage of non-volatile memory.
The lower tunneling barrier height of high-k dielectric gate insulators provides larger tunneling current into the floating gate 105 with a smaller gate voltage. Additionally, larger tunneling current out of the floating gate is accomplished with smaller control gate 109 voltages.
Another advantage of high-k gate insulators is that smaller write and erase voltages are necessary due to the reduced thickness of the SiO2 layer 107 between the control gate 109 and the floating gate 105. This layer can be made less than 15 Å thick. Additional advantages include increased cell current with respect to prior art flash memory cells, transistors can be scaled below 50 nm, drain turn-on, short-channel effects, and punchthrough are substantially eliminated.
In one embodiment, the high-k gate dielectric layer 103 of FIG. 1 is fabricated using atomic layer deposition (ALD). As is well known in the art, ALD is based on the sequential deposition of individual monolayers or fractions of a monolayer in a well controlled manner. Gaseous precursors are introduced one at a time to the substrate surface and between the pulses the reactor is purged with an inert gas or evacuated.
In the first reaction step, the precursor is saturatively chemisorbed at the substrate surface and during subsequent purging the precursor is removed from the reactor. In the second step, another precursor is introduced on the substrate and the desired films growth reaction takes place. After that reaction, byproducts and the precursor excess are purged from the reactor. When the precursor chemistry is favorable, one ALD cycle can be performed in less than one second in a properly designed flow-type reactor.
ALD is well suited for deposition of high-k dielectrics such as AlOx, LaAlO3, HfAlO3, Pr2O3, Lanthanide-doped TiOX, HfSiON, Zr--Sn--Ti--O films using TiCl4 or TiI4, ZrON, HfO2/Hf, ZrAlxOy, CrTiO3, and ZrTiO4.
The most commonly used oxygen source materials for ALD are water, hydrogen peroxide, and ozone. Alcohols, oxygen and nitrous oxide have also been used. Of these, oxygen reacts poorly at temperatures below 600° C. but the other oxygen sources are highly reactive with most of the metal compounds listed above.
Source materials for the above-listed metals include: zirconium tetrachloride (ZrCl4) for the Zr film, titanium tetraisopropoxide (Ti(OCH(CH3)2)4) for the Ti film, trimethyl aluminum (Al(CH3)3) for the Al film, chromyl chromide (CrO2Cl2) for the Cr film, praseodymium chloride (PrCl3) for the Pr film, and hafnium chloride (HfCl4) for the Hf film. Alternate embodiments use other source materials.
Thin oxide films are deposited at a temperature that is high enough such that, when it is adsorbed to the substrate surface, the vaporized source material reacts with a molecular layer of a second source material or that the vaporized source material becomes absorbed and reacts with the second source material directed to the substrate surface in the subsequent step. On the other hand, the temperature should be low enough such that thermal breakdown of the source material does not occur or that its significance in terms of the total growth rate of the film is very small. Regarding the above-listed metals, the ALD process may be carried out at a temperature range of approximately 200-600° C. Alternate embodiments use other temperature ranges.
In another embodiment of the flash memory transistor of the present invention illustrated in FIG. 1, the high-k dielectric layer 103 can be fabricated using evaporation techniques. Various evaporation techniques are subsequently described for the high dielectric constant materials listed above.
Very thin films of TiO2 can be fabricated with electron-gun evaporation from a high purity TiO2 slug (e.g., 99.9999%) in a vacuum evaporator in the presence of an ion beam. In one embodiment, an electron gun is centrally located toward the bottom of the chamber. A heat reflector and a heater surround the substrate holder. Under the substrate holder is an ozonizer ring with many small holes directed to the wafer for uniform distribution of ozone that is needed to compensate for the loss of oxygen in the evaporated TiO2 film. An ion gun with a fairly large diameter (3-4 in. in diameter) is located above the electron gun and argon gas is used to generate Ar ions to bombard the substrate surface uniformly during the film deposition to compact the growing TiO2 film.
A two-step process is used in fabricating a high purity HfO2 film. This method avoids the damage to the silicon surface by Ar ion bombardment, such as that encountered during Hf metal deposition using dc sputtering. A thin Hf film is deposited by simple thermal evaporation. In one embodiment, this is by electron-beam evaporation using a high purity Hf metal slug (e.g., 99.9999%) at a low substrate temperature (e.g., 150°-200° C.). Since there is no plasma and ion bombardment of the substrate (as in the case of sputtering), the original atomically smooth surface of the silicon substrate is maintained. The second step is oxidation to form the desired HfO2.
The first step in the deposition of CoTi alloy film is by thermal evaporation. The second step is the low temperature oxidation of the CoTi film at 400° C. Electron beam deposition of the CoTi layer minimizes the effect of contamination during deposition. The CoTi films prepared from an electron gun possess the highest purity because of the high-purity starting material. The purity of zone-refined starting metals can be as high as 99.999%. Higher purity can be obtained in deposited films because of further purification during evaporation.
A two step process in fabricating a high-purity ZrO2 film avoids the damage to the silicon surface by Ar ion bombardment. A thin Zr film is deposited by simple thermal evaporation. In one embodiment, this is accomplished by electron beam evaporation using an ultra-high purity Zr metal slug (e.g., 99.9999%) at a low substrate temperature (e.g., 150°-200° C.). Since there is no plasma and ion bombardment of the substrate, the original atomically smooth surface of the silicon substrate is maintained. The second step is the oxidation to form the desired ZrO2.
The fabrication of Y2O3 and Gd2O3 films may be accomplished with a two step process. In one embodiment, an electron gun provides evaporation of high purity (e.g., 99.9999%) Y or Gd metal followed by low-temperature oxidation technology by microwave excitation in a Kr/O2 mixed high-density plasma at 400° C. The method of the present invention avoids damage to the silicon surface by Ar ion bombardment such as that encountered during Y or Gd metal deposition sputtering. A thin film of Y or Gd is deposited by thermal evaporation. In one embodiment, an electron-beam evaporation technique is used with an ultra-high purity Y or Gd metal slug at a low substrate temperature (e.g., 150°-200° C.). Since there is no plasma or ion bombardment of the substrate, the original atomically smooth surface of the silicon substrate is maintained. The second step is the oxidation to form the desired Y2O3 or Gd2O3.
The desired high purity of a PrO2 film can be accomplished by depositing a thin film by simple thermal evaporation. In one embodiment, this is accomplished by an electron-beam evaporation technique using an ultra-high purity Pr metal slug at a low substrate temperature (e.g., 150°-200° C.). Since there is no plasma and ion bombardment of the substrate, the original atomically smooth surface of the silicon substrate is maintained. The second step includes the oxidation to form the desired PrO2.
The nitridation of the ZrO2 samples comes after the low-temperature oxygen radical generated in high-density Krypton plasma. The next step is the nitridation of the samples at temperatures >700° C. in a rapid thermal annealing setup. Typical heating time of several minutes may be necessary, depending on the sample geometry.
The formation of a Y--Si--O film may be accomplished in one step by co-evaporation of the metal (Y) and silicon dioxide (SiO2) without consuming the substrate Si. Under a suitable substrate and two-source arrangement, yttrium is evaporated from one source, and SiO2 is from another source. A small oxygen leak may help reduce the oxygen deficiency in the film. The evaporation pressure ratio rates can be adjusted easily to adjust the Y--Si--O ratio.
The prior art fabrication of lanthanum aluminate (LaAlO3) films has been achieved by evaporating single crystal pellets on Si substrates in a vacuum using an electron-beam gun. The evaporation technique of the present invention uses a less expensive form of dry pellets of Al2O3 and La2O3 using two electron guns with two rate monitors. Each of the two rate monitors is set to control the composition. The composition of the film, however, can be shifted toward the Al2O3 or La2O3 side depending upon the choice of dielectric constant. After deposition, the wafer is annealed ex situ in an electric furnace at 700° C. for ten minutes in N2 ambience. In an alternate embodiment, the wafer is annealed at 800°-900° C. in RTA for ten to fifteen seconds in N2 ambience.
FIG. 2 illustrates an energy-band diagram in accordance with a write operation in the transistor structure of FIG. 1 while FIG. 3 is the energy-band diagram for an erase operation. The diagrams show the conduction band edge, EC, and the valence band edge, EV. Between EC and EV is the band gap where there are no states for electrons. The energy barrier, Φ, is the discontinuity in the conduction bands.
The high-k tunnel gate dielectric of the present invention reduces the barriers between the substrate and gate insulator and/or between the floating gate and the gate insulator. FIG. 4 illustrates a plot of tunneling current dependence on barrier height for various electric fields in accordance with the transistor structure of FIG. 1. This plot shows that the tunneling current at a fixed electric field can be increased by orders of magnitude as a result of reducing the barriers.
In the specific case of Fowler-Nordheim tunneling, the expression that describes the conduction in the insulator is J=AE2exp(-B/E) where J is the current density in amps/cm2, E is the electric field in the insulator in volts/cm and A and B are constants for a particular insulator. The constants depend on the effective mass and the electron barrier energy of the insulator and are scaled with the barrier energy, Φ, as A∝(1/Φ) and B∝(Φ)3/2.
For the case of the commonly used gate insulator, SiO2, the equation above renders A(SiO2--Si)=5.5×10-16 amps/volt2 and B(SiO2--Si)=7.07×107 V/cm. If a new barrier of <)=1.08 eV is utilized, likely values for A and B can be extrapolated from the above equations. In this case, A(Φ=1.08 eV)=1.76×10-15 amps/volt2 and B(Φ=1.08 eV)=1.24×107 V/cm.
Curves of J versus the barrier energy, F, are shown in FIG. 4 for several values of E. For a given tunneling current, lower barriers require lower electric fields. As an example, an SiO2 barrier of 3.2 eV has an electric field of 6×106 V/cm while for the same tunneling current, a high-k dielectric with a 1.08 eV barrier requires only an electric field of 7×105 V/cm. If the thicknesses of the two dielectrics are the same then the voltage required will be about 8.6 times less for the same current. If the high-k dielectric has a dielectric constant of 28, then the equivalent oxide thickness (EOT) will be 7 times less than the actual thickness of the high-k dielectric.
The flash memory transistors of the present invention can thus be designed with very small equivalent oxide thicknesses and scaled into the 50 nm dimensions without drain turn-on problems, short-channel effects, and punchthrough. Additionally, retention times will decrease due to more thermal excitation and emission of electrons over the smaller barriers.
FIG. 5 illustrates a functional block diagram of a memory device 500 that can incorporate the flash memory cells of the present invention. The memory device 500 is coupled to a processor 510. The processor 510 may be a microprocessor or some other type of controlling circuitry. The memory device 500 and the processor 510 form part of an electronic system 520. The memory device 500 has been simplified to focus on features of the memory that are helpful in understanding the present invention.
The memory device includes an array of flash memory cells 530 that can be floating gate flash memory cells. The memory array 530 is arranged in banks of rows and columns. The control gates of each row of memory cells is coupled with a wordline while the drain and source connections of the memory cells are coupled to bitlines. As is well known in the art, the connection of the cells to the bitlines depends on whether the array is a NAND architecture or a NOR architecture.
An address buffer circuit 540 is provided to latch address signals provided on address input connections A0-Ax 542. Address signals are received and decoded by a row decoder 544 and a column decoder 546 to access the memory array 530. It will be appreciated by those skilled in the art, with the benefit of the present description, that the number of address input connections depends on the density and architecture of the memory array 530. That is, the number of addresses increases with both increased memory cell counts and increased bank and block counts.
The memory device 500 reads data in the memory array 530 by sensing voltage or current changes in the memory array columns using sense/buffer circuitry 550. The sense/buffer circuitry, in one embodiment, is coupled to read and latch a row of data from the memory array 530. Data input and output buffer circuitry 560 is included for bi-directional data communication over a plurality of data connections 562 with the controller 510). Write circuitry 555 is provided to write data to the memory array.
Control circuitry 570 decodes signals provided on control connections 572 from the processor 510. These signals are used to control the operations on the memory array 530, including data read, data write, and erase operations. The control circuitry 570 may be a state machine, a sequencer, or some other type of controller.
The flash memory device illustrated in FIG. 5 has been simplified to facilitate a basic understanding of the features of the memory. A more detailed understanding of internal circuitry and functions of flash memories are known to those skilled in the art.
In summary, the flash memory transistors of the present invention use a high-k tunnel gate dielectric to enable the transistor to be reduced in size without performance problems. The high-k dielectric enables smaller write and erase voltages to be used and eliminates drain turn-on problems, short-channel effects, and punchthrough. The high-k dielectric can be deposited either with an evaporation process or an atomic layer deposition process.
The flash memory cells of the present invention may be NAND-type cells, NOR-type cells, or any other type of flash memory array architecture.
Although specific embodiments have been illustrated and described herein, it will be appreciated by those of ordinary skill in the art that any arrangement that is calculated to achieve the same purpose may be substituted for the specific embodiments shown. Many adaptations of the invention will be apparent to those of ordinary skill in the art. Accordingly, this application is intended to cover any adaptations or variations of the invention. It is manifestly intended that this invention be limited only by the following claims and equivalents thereof.
Patent applications by Leonard Forbes, Corvallis, OR US
Patent applications in class Multiple interelectrode dielectrics or nonsilicon compound gate insulator
Patent applications in all subclasses Multiple interelectrode dielectrics or nonsilicon compound gate insulator