Circuit and system of a high density anti-fuse

Abstract

A high density anti-fuse cell can be built at the cross points of two perpendicular interconnect lines, such as active region lines, active and polysilicon lines, active and metal lines, or polysilicon and metal lines. The cell size can be very small. At least one of the anti-fuse cells have a thin oxide fabricated before, after, or between a diode in at least one contact holes at the cross points of the interconnect lines. The thin oxide of the anti-fuse cells at the cross points can be selected for rupture by applying supply voltages in the two perpendicular lines. In some embodiments, a diode can be created after thin oxide is ruptured so that explicitly fabricating a diode or opening a contact hole at the cross-point may not be necessary.

Description

BACKGROUND OF THE INVENTION

Anti-fuse is one of the One-Time Programmable (OTP) devices that can only be programmed once. Particularly, an anti-fuse has a high impedance state after fabrication and a low impedance state after being programmed. On the contrary, a fuse has a low impedance state after fabrication and a high impedance state after being programmed. The most commonly used anti-fuses are based on MOS gate oxide breakdown, metal-dielectric-metal breakdown, metal-dielectric-silicon breakdown, or silicon-dielectric-silicon breakdown, etc. Silicon dioxide (SiO₂) is the most commonly used dielectric for breakdown in anti-fuses. However, Silicon-Oxide-Nitride (SON), Silicon Nitride (SiN_x), Oxide-Nitride-Oxide (ONO), or other type of metal oxides, such as Aluminum Oxide (Al₂O₃), MgO, HfO₂, or Cr₂O₃, can also be used.

MOS gate oxide breakdown is based on applying a high voltage to break down the gate oxide to create a programmed state. However, there is a mechanism called soft-breakdown, other than the desirable hard-breakdown, which makes the dielectric film appear to be broken down, but the film may heal by itself after cycling or burn-in. The reliability may be a concern for practical applications.

Dielectric breakdown anti-fuses have been proven in manufacture. One of conventional dielectric breakdown anti-fuse is shown in FIGS. 1(a), 1(b), and 1(c). This anti-fuse is based on metal-dielectric-silicon with a diode constructed by P+ active region over N+ bar as program selector. FIG. 1(a) shows a portion of process steps by using a first Local Oxidation (LOCOS) to define an N+ bar area. FIG. 1(b) shows a second LOCOS step to further define active regions within each N+ bar in a perpendicular direction. The cell is patterned by two LOCOS steps so that the cell size is determined by the pitches of active regions in the X- and Y-directions. The cell size is generally referred to 4F², where F stands for figure size. After the active region of the cells is determined, a P type dopant is implanted, a thin silicon dioxide is grown, and then a metal is built on top of each cell as shown in FIG. 1(c). The equivalent circuit of the anti-fuse cell is a capacitor in series with a diode at an X and Y cross-point as shown in FIG. 1(d). For additional information see, e.g., Noriaki, et. al, “A New Cell for High Capacity Mask ROM by the Double LOCOS Techniques,” International Electronics Device Meeting, December, 1983, pp. 581-584.

The anti-fuse cell in FIGS. 1(a), 1(b), and 1(c) is very complicated to fabricate, as it requires three more masks and two LOCOS steps over standard CMOS processes. Fabricating LOCOS requires a mask for field implant, nitride deposition, and a long thermal cycle to grow field oxide. Accordingly, there is a need for an anti-fuse cell that is more compatible with standard CMOS process to save costs.

SUMMARY OF THE INVENTION

The invention pertains to an anti-fuse device and memory based on dielectric breakdown formed at the cross points of two perpendicular conductors that has minimum process steps or masks over standard CMOS logic processes.

The general structures of the devices in this invention has a dielectric film for rupture and a diode as program selector in a cell defined at a cross-point of two conductor (conductive) lines in perpendicular directions. There are various embodiments that are within the scope and spirit of this invention. The dielectric film can be fabricated from silicon dioxide (SiO₂), silicon nitride (SiN_x, or Si₃N₄ particularly), silicon oxide-nitride (SON), or silicon oxide-nitride-oxide (ONO). Alternatively other types of metal oxides, such as Al₂O₃, HfO₂, MgO, ZrO₂, or Cr₂O₃, can be used, but they may be more expensive, difficult to manufacture, and have a higher breakdown voltage. The diode can be a junction diode constructed from bulk silicon, a diode constructed from polysilicon, a diode constructed from bulk silicon and polysilicon, or a p-i-n diode with an intrinsic layer between P and N type silicon or polysilicon. An intrinsic layer means it is not intentionally P or N doped but can be slightly N or P type due to out-diffusion or contamination. The dielectric film may be fabricated before, after, or between the N or P terminals of the diode. The conductors in perpendicular directions can be both active regions, active and polysilicon, polysilicon and metal, or active and metal, in various embodiments. The cross point may be formed at the junction of two perpendicular conductors, or inside a contact hole at the junction of two perpendicular conductors.

Another key concept of this invention is to use core logic or I/O devices in the peripheral of the anti-fuse memory that are the same devices as those built in the rest of integrated circuits. In the past, the programming voltage of an anti-fuse is very high, about 12V or 18V, that needs special high voltage devices in the peripheral circuit to design an anti-fuse memory. As a result, more mask layers and more process steps are needed, and hence the fabrication costs are very high. One aspect of this invention can eliminate the need to use high voltage devices to lower the breakdown voltage of the dielectric film so that core or I/O devices in an integrated circuit can be used. Another embodiment is to use a combination of different voltages in two perpendicular conductors to achieve high voltage for a selected cell to program, while the unselected cells are operated within a low voltage range.

Though there are many different and useful embodiments of the invention, the size of the anti-fuse can be 4F², where F stands for figure size that is the width or space of the conductors to form an anti-fuse cell. It will be apparent to those skilled in the art that various modifications and variations can be made in the present invention without departing from the spirit or scope of the invention. Thus, it is intended that the present invention covers the modifications and variations of this invention provided they come within the scope of the appended claims and their equivalents.

The invention can be implemented in numerous ways, including as a method, system, device, or apparatus (including graphical user interface and computer readable medium). Several embodiments of the invention are discussed below.

As an anti-fuse memory, one embodiment can, for example, include a plurality of anti-fuse cells. At least one of the anti-fuse cells can include a dielectric film coupled to a first supply voltage line, and a diode including at least a first type of silicon and a second type of silicon. The first type of silicon can have a first type of dopant and the second type of silicon can have a second type of dopant. An intrinsic layer may be inserted between the first and the second types of silicon. The first type of silicon can provide a first terminal of the diode and the second type of silicon can provide a second terminal of the diode. The first type of silicon can also be coupled to the dielectric film, and the second type of silicon can be coupled to a second supply voltage line. The first and second type of silicon can be fabricated at the cross points of two perpendicular conductor lines. The conductor lines can be implemented by any combinations of metal, active region, buried layer, or polysilicon. The diode can be constructed explicitly or by itself after oxide breakdown. If one of the conductor lines is metal while the other is an active region, buried layer, or polysilicon, a diode can built explicitly in the active region, buried layer or polysilicon with a first and a second type of silicon. If the two perpendicular conductor lines are buried layer and polysilicon with a first and second type of silicon, respectively, a diode can be constructed by itself once the oxide film is broken down. The dielectric film can be configured to be programmable by applying voltages to the first and second supply voltage lines to thereby change the resistance of the dielectric film into a different logic state. Alternatively, the dielectric film can be coupled to the second type of silicon, or in between the first and the second type of silicon in other embodiments.

As an electronic system, one embodiment can, for example, include at least a processor, and an anti-fuse memory operatively connected to the processor. The anti-fuse memory can include at least a plurality of anti-fuse cells for providing data storage. Each of the anti-fuse cells can include at least a dielectric film coupled to a first supply voltage line, and a diode including at least a first type of silicon and a second type of silicon. The first type of silicon can have a first type of dopant and the second type of silicon can have a second type of dopant. An intrinsic layer may be inserted between the first and the second types of silicon. The first type of silicon can provide a first terminal of the diode and the second type of silicon can provide a second terminal of the diode. The first type of silicon can be coupled to the dielectric film and the second type of silicon can be coupled to a second supply voltage line. The first and second type of silicon can be fabricated at the cross points of two perpendicular conductor lines. The conductor lines can be implemented by any combinations of metal, active region, buried layer, or polysilicon. The diode can be constructed explicitly or by itself after oxide breakdown. If one of the conductor lines is metal while the other is an active region, buried layer, or polysilicon, a diode can be built explicitly in the active region, buried layer or polysilicon with a first and a second type of silicon. If the two perpendicular conductor lines are buried layer and polysilicon with a first and second type of silicon, respectively, a diode can be constructed by itself once the oxide film is broken down. The dielectric film can be configured to be programmable by applying voltages to the first and the second supply voltage lines to thereby change the resistance of the dielectric film into a different logic state. Alternatively, the dielectric film can be coupled to the second type of silicon, or in between the first and the second type of silicon in other embodiments.

As a method for providing an anti-fuse memory, one embodiment can, for example, include at least providing a plurality of anti-fuse cells, and programming a logic state into at least one of the anti-fuse cells by applying voltages to the first and the second voltage lines. The at least one of the anti-fuse cells can include at least (i) a dielectric film coupled to a first supply voltage line, and (ii) a diode including at least a first type of silicon and a second type of silicon. The first type of silicon can have a first type of dopant and the second type of silicon can have a second type of dopant. An intrinsic layer may be inserted between the first and the second types of silicon. The first type of silicon can provide a first terminal of the diode and the second type of silicon can provide a second terminal of the diode. The first type of silicon can be coupled to the dielectric film and the second type of silicon can be coupled to a second supply voltage line. The first and second type of silicon can be fabricated at the cross points of two perpendicular conductor lines. The conductor lines can be any combinations of metal, active region, buried layer, or polysilicon. The diode can be constructed explicitly or by itself after oxide breakdown. If one of the conductor lines is metal while the other is an active region, buried layer, or polysilicon, a diode can be built explicitly in the active region, buried layer or polysilicon with a first and a second type of silicon. If the two perpendicular conductor lines are buried layer and polysilicon with a first and second type of silicon, respectively, a diode can be constructed by itself once the oxide film is broken down. The dielectric film can be configured to be programmable by applying voltages to the first and the second supply voltage lines to thereby change the resistance of the dielectric film into a different logic state. Alternatively, the dielectric film can be coupled to the second type of silicon, or in between the first and the second type of silicon in other embodiments.

As an electronic system, one embodiment can, for example, include at least a battery and an integrated circuit operatively connected to the battery. The at least one battery can provide a voltage between 1.0 and 2.0V in the nominal conditions. The integrated circuit can include at least an anti-fuse memory including at least a plurality of anti-fuse cells. Each of at least the plurality of the anti-fuse cells can be constructed at one of the cross points and each of at least the plurality of anti-fuse cells can include at least: (i) a plurality of conductive lines with a first type of dopant; (ii) a plurality of metal lines being substantially perpendicular to the polysilicon lines; (iii) a layer of isolation oxide fabricated between the metal and conductive lines; (iv) a plurality of contacts being open at the cross points of metal and conductive line; and (v) a silicon diode and a layer of thin oxide fabricated in each contact hole before placing metal lines. The anti-fuse memory can be configured to be selectively programmable by applying a first supply voltage to the metal lines and a second supply voltage to the conductive lines to rupture the thin oxide.

BRIEF DESCRIPTION OF THE DRAWINGS

The present invention will be readily understood by the following detailed descriptions in conjunction with the accompanying drawings, wherein like reference numerals designate like structural elements, and in which:

FIG. 1(a) shows a first LOCOS step in a conventional dielectric breakdown anti-fuse.

FIG. 1(b) shows a second LOCOS step in a conventional dielectric breakdown anti-fuse.

FIG. 1(c) shows a cross section of a conventional metal-oxide-silicon anti-fuse using a diode as program selector.

FIG. 1(d) shows a conventional equivalent circuit of the anti-fuse cell.

FIG. 2(a) shows a top view of a dielectric-breakdown anti-fuse defined by active regions in X- and Y-directions according to one embodiment.

FIG. 2(b) shows a cross section of a dielectric-breakdown anti-fuse defined by active regions in the Y direction.

FIG. 2(c) shows a cross section of a dielectric-breakdown anti-fuse defined by active regions in the X direction.

FIG. 3(a) shows step (a) of fabricating a dielectric-breakdown anti-fuse defined by active regions (ref. FIG. 2(a)-2(c)) after buried N+ layer.

FIG. 3(b) shows step (b) of fabricating a dielectric-breakdown anti-fuse defined by active regions (ref. FIG. 2(a)-2(c)) after field implant.

FIG. 3(c) shows step (c) of fabricating a dielectric-breakdown anti-fuse defined by active regions (ref. FIG. 2(a)-2(c)) after building isolation for active regions.

FIG. 3(d) shows step (d) of fabricating a dielectric-breakdown anti-fuse defined by active regions (ref. FIG. 2(a)-2(c)) after a P+ implant as the P terminal of a diode.

FIG. 3(e) shows step (e) of fabricating a dielectric-breakdown anti-fuse defined by active regions (ref. FIG. 2(a)-2(c)) after growing an interlayer dielectric and etching contact holes.

FIG. 3(f) shows step (f) of fabricating a dielectric-breakdown anti-fuse defined by active regions (ref. FIG. 2(a)-2(c)) after depositing a thin layer of dielectric for breakdown.

FIG. 3(g) shows step (g) of fabricating a dielectric-breakdown anti-fuse defined by active regions (ref. FIG. 2(a)-2(c)) after depositing an adhesive layer.

FIG. 3(h) shows step (h) of fabricating a dielectric-breakdown anti-fuse defined by active regions (ref. FIG. 2(a)-2(c)) after depositing, patterning, and etching a metal layer.

FIG. 4 shows a top view of another embodiment of dielectric breakdown anti-fuse defined by metal and polysilicon according to one embodiment.

FIG. 5(a) shows step (a) of fabricating a dielectric breakdown anti-fuse defined by metal and polysilicon (ref. FIG. 4) after depositing and siliciding polysilicon.

FIG. 5(b) shows step (b) of fabricating a dielectric-breakdown anti-fuse defined by metal and polysilicon (ref. FIG. 4) after growing an interlayer dielectric and etching contact holes.

FIG. 5(c) shows step (c) of fabricating a dielectric-breakdown anti-fuse defined by metal and polysilicon (ref. FIG. 4) after depositing an adhesive layer.

FIG. 5(d) shows step (d) of fabricating a dielectric-breakdown anti-fuse defined by metal and polysilicon (ref. FIG. 4) after growing a p-i-n silicon diode inside a contact hole.

FIG. 5(e): shows step (e) of fabricating a dielectric-breakdown anti-fuse defined by metal and polysilicon (ref. FIG. 4) after growing an oxide film on top of a silicon diode.

FIG. 5(f): shows step (f) of fabricating a dielectric-breakdown anti-fuse defined by metal and polysilicon (ref. FIG. 4) after depositing another adhesive layer.

FIG. 5(g): shows step (g) of fabricating a dielectric-breakdown anti-fuse defined by metal and polysilicon (ref. FIG. 4) after depositing, patterning, and etching a metal layer.

FIG. 6 shows using dummy polysilicon to raise contact-hole height in an anti-fuse defined by metal and active lines.

FIG. 7 shows a top view of dielectric breakdown anti-fuse defined by active and polysilicon lines according to one embodiment.

FIG. 8(a) shows step (a) of fabricating a dielectric breakdown anti-fuse defined by active and polysilicon (ref. to FIG. 7) after building active isolation (e.g. STI/LOCOS).

FIG. 8(b) shows step (a) of fabricating a dielectric breakdown anti-fuse defined by active and polysilicon (ref. to FIG. 7) after buried N+ implant.

FIG. 8(c) shows step (a) of fabricating a dielectric breakdown anti-fuse defined by active and polysilicon (ref. to FIG. 7) after growing intrinsic and P type silicon.

FIG. 8(d) shows step (a) of fabricating a dielectric breakdown anti-fuse defined by active and polysilicon (ref. to FIG. 7) after growing oxide.

FIG. 8(e) shows step (a) of fabricating a dielectric breakdown anti-fuse defined by active and polysilicon (ref. to FIG. 7) after depositing and etching polysilicon.

FIG. 9 shows a top view of fabricating a dielectric breakdown anti-fuse defined by active region and metal with a piece of polysilicon at each cross-point.

FIG. 10(a) shows step (a) of fabricating a dielectric breakdown anti-fuse defined by active and metal (ref. FIG. 9) after building buried N+ lines and depositing gate oxide.

FIG. 10(b) shows step (b) of fabricating a dielectric-breakdown anti-fuse defined by active and metal (ref. FIG. 9) after stripping the gate oxide and depositing P type polysilicon.

FIG. 10(c) shows step (c) of fabricating a dielectric-breakdown anti-fuse defined by active and metal (ref. FIG. 9) after depositing interlayer dielectric, opening contact holes, and deposit an adhesive layer.

FIG. 10(d) shows step (d) of fabricating a dielectric-breakdown anti-fuse defined by active and metal (ref. FIG. 9) after growing or depositing an oxide film.

FIG. 10(e): shows step (e) of fabricating a dielectric-breakdown anti-fuse defined by active and metal (ref. FIG. 9) after depositing another adhesive layer.

FIG. 10(f): shows step (f) of fabricating a dielectric-breakdown anti-fuse defined by active and metal (ref. FIG. 9) after depositing, patterning, and etching a metal layer.

FIG. 11(a) shows one high voltage and one core logic or I/O voltage voltages applied to the perpendicular conductors to the selected and unselected cells according to one embodiment.

FIG. 11(b) shows positive and negative supply voltages applied to the perpendicular conductors to the selected and unselected cells according to one embodiment.

FIG. 11(c) shows a high and a half-value program supply voltages applied to the perpendicular conductors to the selected and unselected cells according to one embodiment.

DETAILED DESCRIPTION OF THE INVENTION

This invention concerns a dielectric breakdown anti-fuse cell using diode as program selector defined at a cross-point of two conductor lines in perpendicular directions. Various embodiments about dielectric materials, diode structures, conductor types, process steps, devices employed, and cell select schemes will be disclosed and that are within the scope of this invention.

FIG. 2(a) shows a top view of a dielectric breakdown anti-fuse 10 according to one embodiment. Each anti-fuse cell 15 is defined by active regions in X- and Y-directions. An array of buried N+ lines 11 serve as wordlines are built running in the X-direction before active regions 12 is defined. The active regions can be isolated in X- and Y-directions by Local Oxidation (LOCOS) or Shallow Trench Isolation (STI). A contact hole 13 in each active region is etched after depositing an interlayer dielectric. Then, N type and P type dopants are subsequently implanted into the contact hole in each active region to act as N and P terminals of a diode. A thin film of dielectric, such as SiO₂, Si₃N₄, SiN_x, silicon-oxide-nitride (SON), or silicon oxide-nitride-oxide (ONO), is deposited or grown into each contact hole for rupture by programming. A metal is built on top of the contact hole to serve as a bitline.

FIG. 2(b) shows a cross section 30 of the dielectric breakdown anti-fuse in FIG. 2(a) along the Y-direction. Buried N+ lines 34 between LOCOS or STI 32 connect active regions along the X-direction and also serve as the N terminal of each diode. The N+ lines should be deeper than the isolation between active regions. Then, an interlayer dielectric 31 (normally silicon dioxide) is deposited to isolate the active regions from the upper conductor layers and then contact holes are etched. Each contact hole is implanted with P+ 33, the same implant for source and drain of PMOS devices, to serve as the P terminal of a diode. FIG. 2(c) shows a cross section 20 of the same dielectric breakdown anti-fuse in FIGS. 2(a) and 2(b) along the X-direction. Buried N+ lines 24 between LOCOS or STI 22 connect active regions along the X-direction and also serve as the N terminal of each diode. Then, the interlayer dielectric 21 (normally silicon diode) is deposited to isolate the active regions from the upper conductor layers and then contact holes are etched. Each cell is implanted with P+ 23, the same implant for sources or drains of PMOS devices, to serve as the P terminal of a diode. The N+ lines can be fabricated in one or two steps. The first step is to implant a heavy N+ line deep into the silicon to connect the cells in the X-direction, and then a second shallower N type dopant is implanted into each cell to reach the deep N+ line and to serve as the N terminal of a diode.

FIGS. 3(a)-3(h) show cross sections 40 in a portion of process steps related to the dielectric breakdown anti-fuse shown in FIG. 2(a)-2(c). The cross sections are shown along the Y-direction. FIG. 3(a) shows a cross section after buried N+ lines 41 are implanted and annealed. The N+ lines 41 act as the N terminals of a diode in each cell, interconnect the cells in the X-direction, and serve as wordlines. The N+ lines shall be deeper than the isolation for active regions and can be heavily doped near the bottom. The desirable embodiment for N+ lines is a shallow retrograde N well, similar to the N well in common CMOS processes. FIG. 3(b) shows a field implant 42 before growing thermal oxide in LOCOS or etching shallow trenches in STI process. FIG. 3(c) shows after active region isolations 43, either LOCOS or STI, are built. Alternatively, the buried N+ 41 in FIG. 3(a) can be implanted into silicon after LOCOS or STI 43 is formed in FIG. 3(c). FIG. 3(d) shows after a P+ implant 44, the same as source or drain implant of a PMOS, to act as the P terminal of the diode in each cell. FIG. 3(e) shows after an interlayer dielectric 45 is deposited and contact holes 45A are etched. FIG. 3(f) shows after a thin oxide film 46 is deposited for rupture upon programming. The thickness of the SiO₂ is generally around 30-80 Å to have breakdown voltage about 6-15V. FIG. 3(g) shows after an adhesive layer 47 is deposited to provide a better attachment for metal on top. The adhesive layer can be a 200 Å TiN film or other material. FIG. 3(h) shows after a metal layer 48 is deposited, patterned, and then etched. Two additional masks may be needed: one for defining and building N+ lines and the other for growing a dielectric film for breakdown.

FIG. 4 shows a top view of dielectric breakdown anti-fuse cells 50 defined by metal lines 52 and polysilicon lines 51 according to embodiment. Inside each contact hole 53 at the junctions of the polysilicon lines 51 and metal lines 52 builds an anti-fuse cell 55. The cell size is determined by metal pitch in the X-direction and polysilicon pitch in the Y-direction so that the cell size is 4F².

FIGS. 5(a)-5(g) show cross sections 60 in a portion of process steps related to the dielectric breakdown anti-fuse defined by metal and polysilicon in FIG. 4. The cross sections are along the Y-direction. FIG. 5(a) shows a cross section after polysilicon lines 61 are built and silicide 62 is grown on top. The polysilicon lines interconnect the cells in the X-direction, and serve as wordlines. The silicide on top reduces resistance of the polysilicon lines. FIG. 5(b) shows after an interlayer dielectric 64 is deposited and contact holes 64A are etched therein. FIG. 5(c) shows after an adhesive layer 65 is deposited in the contact holes. FIG. 5(d) shows after a silicon diode 66 is built with N type, intrinsic, and P type dopants. The diode can be fabricated by Chemical Vapor Deposition (CVD) with dopants changed in situ, (i.e. dopants changed from N type, intrinsic, and to P type along the course of CVD processing). The composition of the diode in one embodiment can be an intrinsic layer between the P and N type silicon, the so-called p-i-n diode. The intrinsic layer means there is no intention to add any dopants but it could be slightly N or P type due to out diffusion or contamination. The thickness of the intrinsic layer determines the breakdown voltage of the diode and shall be high enough to prevent premature programming. Another embodiment to control the breakdown voltage of the diode is to use lesser doped N or P type, rather than intrinsic, in between the heavily doped P and N terminals of the diode. The thickness of the diode is about 3,000-5,000 Å to fit into the contact height. Alternatively, the p-i-n diode can be fabricated from depositing silicon and then followed by implanting N or P type dopants in different steps. FIG. 5(e) shows after a thin layer of oxide 67 is deposited as an anti-fuse film after the formation of the P and N portions of the diode 66. The dielectric film can be fabricated from silicon dioxide (SiO₂), silicon nitride (Si₃N₄), silicon oxide-nitride (SON), or silicon oxide-nitride-oxide (ONO). Alternatively other types of metal oxides, such as Al₂O₃, HfO₂, ZrO₂, or Cr₂O₃, can be used. If SiO₂ is used, the thickness is generally around 30-80 Å to have breakdown voltage about 6-15V. The oxide for breakdown 67 can also be fabricated before or inserted between the P, intrinsic, or N portions of the diode 66. FIG. 5(f) shows after another adhesive layer 68 is deposited. FIG. 5(g) shows after a metal layer 69 is deposited, patterned, and etched. The metal layer 69 serves as the bitline running in the Y-direction. In this embodiment, one additional mask is needed to open contact holes, build p-i-n or P/N silicon diode, grow oxide inside the contact holes. The adhesive layers serve to provide better attachment between metal and different materials in the contact holes that can be one, two, or none.

Embodiments as illustrated in FIGS. 4, 5(a)-5(g) are about anti-fuses defined by metal and polysilicon lines. There are some embodiments about building an oxide film external to a p-i-n diode or between P and N terminals of a diode. The required processing steps are about the same. But the breakdown mechanism in p-i-n diode is an oxide breakdown between silicon and metal, while the breakdown mechanism in the p-oxide-n sandwich structure is an oxide breakdown between P and N type silicon. Breaking down between silicon and metal may be more reliable than the mechanism between P and N types of silicon. Therefore, a p-i-n diode is a more desirable embodiment over a P-oxide-N sandwich structure.

Polysilicon lines can be readily replaced by active region lines while the other perpendicular conductor lines are metal in other embodiment, which can be anti-fuse cells having a p-i-n diode with external oxide film or having a p-oxide-n sandwich structure. FIG. 4 and FIG. 5(a)-5(g) can be readily applied with the polysilicon lines replaced by active region lines. Using polysilicon lines, rather than active region lines, as wordlines allows them to be biased at a negative voltage, since the wordlines are isolated by oxidation rather than P/N junctions.

However, anti-fuse cells defined by metal and active region lines in other embodiments allow variations in adjusting contact height by placing dummy polysilicon between active regions on the field. FIG. 6 shows a cross section 80 of anti-fuse cells defined by metal and active lines. Active region lines 81 are built and grown with silicide 82 on top. Dummy polysilicon lines 84 with silicide 83 on top are placed between active region lines, but on the field oxide. Because of the dummy polysilicon, the height of the interlayer dielectric 85 to the silicon surface is raised by the thickness of the polysilicon, or about 2,000-5,000 Å, so is the height of contact holes. This allows more depth to build p-i-n diode or p-oxide-n sandwich structure inside a contact hole 86, and hence provides one more parameter for optimization.

FIG. 7 shows a top view of dielectric breakdown anti-fuse cells 90 defined by active lines 92 and polysilicon lines 91 according to one embodiment. At the junctions of polysilicon 91 and active region 92 build anti-fuse cells 95. The polysilicon 91 lines are blocked spacers on each side so that the memory cell size can be smaller. The spacer is a technique to allow forming a lightly-doped source/drain (LDD) region near the gate to alleviate the short channel effects. The cell size is determined by active region pitch in the X-direction and polysilicon pitch in the Y-direction so that the cell size is 4F². The active lines 92 can be doped with N type dopant and the polysilicon lines 91 can be doped with P type dopant so that a diode can be created after the oxide layer is broken down. The active lines 92 can also be doped with P type dopant after being doped with the N type dopant to form an explicit diode before fabricating an oxide layer. Similarly, the active lines 92 can be doped with an intrinsic layer after doped with the N type dopant and before fabricating the oxide layer to create a p-i-n diode after the oxide is broken down.

FIGS. 8(a)-8(e) show cross sections 100 in a portion of process steps related to dielectric breakdown anti-fuse cells defined by active region and polysilicon lines in FIG. 7. In this embodiment, the gate oxide between active and polysilicon acts as an anti-fuse for rupture. The cross sections are along the X-direction. FIG. 8(a) shows a cross section after active isolation, such as LOCOS or STI 104, is built on a P type substrate 101 according to standard CMOS processes. Then, the surface is implanted by N+ 106 to create buried N+ lines to serve as bitlines in FIG. 8(b). The buried N+ lines tend to be heavily N+ doped in the bottom and slightly lighter doped near the surface and may have silicide on top to reduce bitline resistance. They tend to be strapped by metal in fixed intervals to further reduce the resistance. FIG. 8(c) shows a cross section after a P type implant 107 is introduced. The P type dopant and the buried N+ constitute a P/N junction diode. FIG. 8(d) shows a cross section after a thin layer of oxide 108 as dielectric is deposited or grown. Then, a polysilicon 109 is deposited, patterned, implanted by P+, and etched to serve as wordlines of the anti-fuse cells, running in the X-direction in FIG. 8(e). The polysilicon 109 may be silicided to reduce resistance but is blocked spacers on each side so that memory cell can be smaller. The rest of process steps follow the standard CMOS processes. With a proper voltage applied between buried N+ and polysilicon lines, the gate oxide can be rupture so that at the cross-point of the polysilicon and buried N+ lines acts as the P and N terminals of a diode. One additional mask may be needed to build buried N+ lines with heavy N+ implant.

One of the embodiments is to eliminate the P type dopant. Without explicit P type implant to build a P/N junction diode in FIG. 8(c), an implicit diode, constructed from P type polysilicon and N type buried lines, may still be created after the oxide is ruptured. Another embodiment is to fabricate an intrinsic layer before the P type dopant on the silicon surface to create a p-i-n diode. Intrinsic means no intentionally N or P doped but can be slightly N or P type due to out-diffusion or contamination. Yet another embodiment is to fabricate an intrinsic layer without any P type dopant on the silicon surface before an oxide is fabricated and a polysilicon is deposited. The step to grow or deposit oxide 108 in FIG. 8(d) may be omitted by sharing the gate oxide growing process in standard CMOS processes. These are many variations to build an implicit or explicit P/N junction diodes at the cross points in other embodiments.

FIG. 9 shows a top view of dielectric breakdown anti-fuse cells 110 defined by active lines 111 and metal lines 114 with a piece of polysilicon 112 at the cross-point in according to 5one embodiment. At the junction of metal 114 and active region 111 builds an anti-fuse cell 115, where the polysilicon and active region serves as the P and N terminals of a diode. The cell size is determined by active region pitch in the X-direction and metal or polysilicon pitch in the Y-direction so that the cell size is 4F².

FIGS. 10(a)-10(f) show cross sections 120 in a portion of process steps related to the dielectric breakdown anti-fuse cells defined by active and metal with a piece of polysilicon in each contact hole as shown in FIG. 9. In this embodiment, a diode constructed from a polysilicon patch and an active line is formed after the gate oxide is stripped. An oxide film is then grown or deposited inside the contact holes. The cross sections are along the Y-direction. FIG. 10(a) shows a cross section after N type active lines 121 are built and gate oxide 122 is grown on top of a silicon substrate. The active lines interconnect the cells in the X-direction, and serve as bitlines. FIG. 10(b) shows after the gate oxide is striped and pieces of polysilicon 123 are built over the active regions. The polysilicon is P type and the active is N type so that a diode is built as program selector in an anti-fuse cell. FIG. 10(c) shows after an interlayer dielectric 124 is deposited, contact holes 124A are etched, and a thin adhesive layer 125 is deposited. FIG. 10(d) shows after an oxide film 126 is fabricated. The oxide film, such as SiO₂, Si₃N₄, SiN_X, silicon oxide-nitride (SON), silicon oxide-nitride-oxide (ONO), or other types of metal oxide, can be deposited or grown by oxidation or deposit. The thickness of the oxide layer determines the breakdown voltage of the anti-fuse for programming. If SiO₂ is employed, the thickness is generally around 30-80 Å to have breakdown voltage about 6-15V. FIG. 10(e) shows after another adhesive layer 128 is deposited. FIG. 10(f) shows after a metal layer 129 is deposited, patterned, and etched. The metal serves as a bitline running in the Y-direction. In this embodiment, two additional masks are needed to strip the gate oxide and open contact holes for building diodes and an oxide film. The gate oxide strip mask can be shared in a CMOS process that offers more than one kind of gate-oxide devices. The adhesive layers serve to have better attachment between different materials to metal in the contact holes such that the adhesive layer number can be one, two, or none.

Though the anti-fuse can be fabricated in a few more masks over the standard CMOS process, more masks may be needed to fabricate high voltage devices in the peripheral circuit considering the programming voltage tend to be very high, about 10-15V. As a rule of thumb, the rupture voltage for a SiO₂ film is 2V for every 10 Å. For example, the breakdown voltage for a 30 Å SiO₂ film is about 6V. Reducing the thickness of the dielectric film can lower the program voltage so that high voltage devices are not needed in the peripheral circuits. Novel half-select schemes also help to alleviate the requirements of using high voltage devices so that the core logic or I/O devices in the other parts of the integrated circuit can be used for embedded applications.

FIG. 11(a) shows one embodiment of a 4×5 anti-fuse array 200 to alleviate the requirements for high voltage supplies in the peripheral. Suppose the supply voltage of the peripheral and the rest of integrated circuit is 5V, while the program voltage for anti-fuse is 8V, the voltage swings for two perpendicular conductors are 0-5V, and 5-8V, respectively. The selected cell 202 has 0V in the horizontal line and 8V in the vertical line so that the resultant voltage at the cross point is 8V to program the selected cell. However, for those unselect cells, the voltage across the cell is either 5V, or 3V so that no programming could happen. The 8V program voltage is low enough for the peripheral devices to sustain over the maximum programming time. By doing this way, not only high voltage devices are not necessary but also the internal voltage generators. This cell select scheme does not require any negative voltages that may need additional masks to isolate NMOS from a P substrate.

FIG. 11(b) shows another embodiment of a 4×5 anti-fuse array 300 to alleviate the requirements for high voltage supplies in the peripheral. Suppose the supply voltage of the peripheral and the rest of integrated circuit is 4V, while the program voltage for anti-fuse is 8V, the voltage swings for two perpendicular conductors are 0-4V, and −4-0V, respectively. The selected cell 302 has −4V in the horizontal line and 4V in the vertical line so that the resultant voltage at the cross point is 8V to program the cell. However, for those unselect cells, the voltage across the cell is either −4V, or 4V so that no programming could happen. If the 8V program voltage is too high for the peripheral devices to sustain, this combination of positive and negative supply is a desirable embodiment. By doing this way, not only high voltage devices are not necessary but also the internal voltage generators. This cell select scheme requires a negative voltage supply that may need additional masks to isolate NMOS from a P substrate.

FIG. 11(c) shows another half-select scheme 400 by swinging the row and column voltages between 0 to 4V. For a selective cell 402, the voltage applied to the column is 8V while the voltage to the row is 0V, so that the resulting voltage drop on the cell 402 is 8V for programming. On the other hand, the unselected cell 401 has either both 4V's on the rows and columns, or one 4V and the other 0V on row or column, so that the resulting voltage drop is either 4V or 0V to prevent from programming. In this scheme, a voltage generator to generate half of the program voltage may be needed.

There are many variations in the embodiments of this invention. For example, the substrate can be N type rather than P type through the above discussions. The N type or P type dopant can be reversed so that a p-i-n diode and n-i-p diode can be considered equivalent. So are the p-oxide-n and the n-oxide-p sandwich structures. Some process steps may be omitted, such as adhesive layers. And the order of fabricating oxide film and p-i-n or P/N diode may be reversed. The dielectric film for rupture can be fabricated before, after, or in between P type and N type of the diode. The polysilicon and active may not be silicided in an older process. For those skilled in the art understand that various embodiments are possible and they are still within the scope of this invention.

The many features and advantages of the present invention are apparent from the written description and, thus, it is intended by the appended claims to cover all such features and advantages of the invention. Further, since numerous modifications and changes will readily occur to those skilled in the art, it is not desired to limit the invention to the exact construction and operation as illustrated and described. Hence, all suitable modifications and equivalents may be resorted to as falling within the scope of the invention.

Claims

1. An anti-fuse memory in an integrated circuit, comprises: a plurality of anti-fuse cells, at least one of the cells being constructed at the cross points of: a plurality of conductive lines, the conductive lines being polysilicon lines;a plurality of metal lines being substantially perpendicular to the conductive lines;a layer of isolation oxide fabricated between the metal lines and the conductive lines;a plurality of contacts holes at the cross points of the metal lines and the conductive lines; anda silicon diode and a layer of thin oxide fabricated in at least one of the contact holes before placing the metal lines, the layer of thin oxide being provided over the silicon diode between the silicon diode and at least one of the metal lines;wherein the anti-fuse memory is configured to be selectively programmable by applying a first supply voltage to the metal lines and a second supply voltage to the conductive lines to rupture the thin oxide.

2. An anti-fuse memory as recited in claim 1, wherein the conductive lines are active region lines with an implant before CMOS source/drain implants.

3. An anti-fuse memory as recited in claim 1, wherein the layer of thin oxide is fabricated after the silicon diode is fabricated in the at least one of the contact hole.

4. An anti-fuse memory as recited in claim 1, wherein the layer of thin oxide is fabricated between the first and second portions of the silicon diode in at least one of the contact holes, where the first and second portions of the silicon diode have different types of dopants that serve as the P and N terminals of the silicon diode.

5. An anti-fuse memory as recited in claim 1, wherein the layer of thin oxide comprises a silicon dioxide (SiO2), silicon nitride (SiNx), silicon oxide-nitride (SON), silicon oxide-nitride (SON), silicon oxide-nitride-oxide (ONO), or any metal oxide.

6. An anti-fuse memory as recited in claim 1, wherein the thickness of the layer of thin oxide in the at least one of the contact holes is less than 50 Å.

7. An anti-fuse memory as recited in claim 1, wherein the layer of thin oxide is fabricated between first and second portions of the silicon diode in at least one of the contact holes, where the first and second portions of the silicon diode have different types of dopants that serve as the P and N terminals of the silicon diode, andwherein the thickness of the layer of thin oxide in each of the contact holes is less than 50 Å.

8. An anti-fuse memory as recited in claim 1, wherein the conductive lines are active region lines with an implant before CMOS source/drain implants.

9. An anti-fuse memory as recited in claim 1, wherein the layer of thin oxide is fabricated after the silicon diode is fabricated in the at least one of the contact holes.

10. An anti-fuse memory as recited in claim 9, wherein the layer of thin oxide is fabricated between first and second portions of the silicon diode in at least one of the contact holes, where the first and second portions of the silicon diode have different types of dopants that serve as the P and N terminals of the silicon diode.

11. An anti-fuse memory as recited in claim 10, wherein the layer of thin oxide comprises a silicon dioxide (SiO2), silicon nitride (SiNx), silicon oxide-nitride (SON), silicon oxide-nitride-oxide (ONO), or any metal oxide.

12. An anti-fuse memory as recited in claim 9, wherein the thickness of the layer of thin oxide in the at least one of the contact holes is less than 50 Å.

13. An anti-fuse memory as recited in claim 1, wherein devices in a peripheral portion of the anti-fuse memory are built from the same as core logic or 1/O devices in another part of the integrated circuit.

14. An electronic system, comprises: at least one battery; andan integrated circuit operatively connected to the at least one battery, the integrated circuit including at least an anti-fuse memory comprising a plurality of anti-fuse cells, each of at least the plurality of the anti-fuse cells being constructed at one of the cross points and each of at least the plurality of anti-fuse cells includes at least:a plurality of conductive polysilicon lines with a first type of dopant;a plurality of metal lines being substantially perpendicular to the conductive polysilicon lines;a layer of isolation oxide fabricated between the metal and conductive polysilicon lines;a plurality of contact holes at the cross points of the metal and conductive polysilicon lines; anda silicon diode and a layer of thin oxide fabricated in at least one of the contact holes before placing the metal lines, the layer of thin oxide being provided over the silicon diode between the silicon diode and at least one of the metal lines,wherein the anti-fuse memory is configured to be selectively programmable by applying a first supply voltage to the metal lines and a second supply voltage to the polysilicon conductive lines to rupture a portion of the layer of thin oxide.

15. An electronic system as recited in claim 14, wherein devices formed in a peripheral portion of the anti-fuse memory are built from the same as core logic or I/O devices in another part of the integrated circuit.

16. An electronic system as recited in claim 14, wherein the polysilicon conductive lines are active region lines with an implant before CMOS source/drain implants.

17. An electronic system as recited in claim 14, wherein the layer of thin oxide is fabricated after the silicon diode is fabricated in each of the contact holes.

18. An electronic system as recited in claim 14, wherein the layer of thin oxide is fabricated between first and second portions of the silicon diode in at least one of the contact holes, where the first and second portions of the silicon diode have different types of dopants that serve as the P and N terminals of the silicon diode.

19. An electronic system as recited in claim 14, wherein the layer of thin oxide comprises a silicon dioxide (SiO2), silicon nitride (SiNx), silicon oxide-nitride (SON), silicon oxide-nitride-oxide (ONO), or any metal oxide.

20. An electronic system as recited in claim 14, wherein the thickness of the layer of thin oxide in each of the contact holes is less than 50 Å.

21. An electronic system as recited in claim 14, wherein the at least one battery provides a voltage between 1.0 and 2.0V in the nominal conditions.