SELF CONTACTING BIT LINE TO MRAM CELL

Abstract

Embodiments of the invention disclose magnetic memory cell configurations in which a magnetic storage structure is coupled to an upper metal layer with minimal overlay margin. This greatly reduces a size of the memory cell.

Description

FIELD

Embodiments of the invention relate to MRAM (Magnetic Random Access Memory) semiconductor devices.

BACKGROUND

MRAM (Magnetic Random Access Memory) cells may be fabricated during BEOL (Back End Of Line) after a MOS FET device process. The minimum feature size of an MRAM cell is often 1.5× larger than that of FEOL (Front End of Line). It is therefore difficult to shrink memory size compared with other FEOL based memories.

SUMMARY

Embodiments of the invention disclose a plurality of self-aligned structures that save the overlay margin.

The first embodiment discloses a MTJ cell wherein the MTJ stack is directly coupled to the upper metal without the requirement of a via. Sidewalls of individual MTJ elements are protected with dielectric film spacer to prevent from PIN-Switch layer shorting 10 through the tunnel oxide layer. The top layer of MTJ is exposed to upper metal. Overlay margin in this embodiment is required only for upper metal coverage over MTJ. The upper metal width comes to f+2∂, saving 2∂ compared to previous art. Putting MTJ feature size equal to that of FEOL, the memory size becomes competitive to FEOL based memory.

The second embodiment comprises an electrically conductive material such as Titanium Nitride, which is used as a hard mask. The hard mask is for MTJ stack etch and remains on top of MTJ pillar after the etching. Inter layer oxide is deposited over the MTJ pillar. The hard mask remained on MTJ is exposed with CMP. Metal like as Al/Cu is deposited and patterned with conventional lithography and Reactive Ion Etching. The same reduction in memory cell size as the first embodiment is provided by the second embodiment.

The third embodiment discloses a self-aligned via which replaces the hard mask. Silicon nitride is used as hard mask as an example. The hard mask is for MTJ stack etch and remains on top of MTJ pillar after the etching. Inter layer oxide is deposited over the MTJ pillar. The hard mask remained on MTJ is exposed with CMP or Dual Damascene oxide trench etch. The exposed hard mask is removed by hot phosphoric acid followed by upper metal deposition. The same squeezing memory cell size as the first embodiment is expected on the structure.

The fourth embodiment is of self-aligned etching. MTJ is to be etched twice along word line direction first and bit line direction 2^nd. Putting dielectric film, nitride preferred, spacer on MTJ pillar to prevent PIN layer—Fix layer short. Oxide is deposited and planerized by CMP. The oxide is recessed until MTJ appeared. Upper metal layer is deposited patterned. MTJ and bottom read lead is etched with the same mask as upper metal. The upper metal is wrapping around MTJ pillar. It works to help induce magnetic field. The upper metal width can be same size as MTJ pillar. It saves 4∂ compared with prior arts.

The fifth embodiment is also of self-aligned patterning. It is different in read electrode connecting to top of MTJ instead of bottom of the pillar. MTJ is connected to lower metal (write word line). Top metal is electrically isolated from MTJ with a thin dielectric film. The upper metal also wraps around MTJ. It enhances magnetic field induction for switching. It saves cell footprint also by 4∂.

BRIEF DESCRIPTION OF THE DRAWINGS

While the appended claims set forth the features of the present invention with particularity, the invention, together with its objects and advantages, will be more readily appreciated from the following detailed description, taken in conjunction with the accompanying drawings, wherein:

FIG. A illustrates a cross-sectional view of prior arts.

FIG. B illustrates a top view of prior arts.

FIG. 1A illustrates a cross-sectional view of 1^st preferred embodiment

FIG. 1B illustrates a top view of 1^st preferred embodiment.

FIG. 1.1 to FIG. 1.8 illustrate cross sectional views along bit line direction at individual process steps to the 1^st embodiment

FIG. 1.3 s to FIG. 1.8s illustrate cross sectional views along other direction of FIG. 1.3s to FIG. 1.8s.

FIG. 2A illustrates a cross-sectional view of 2^nd preferred embodiment.

FIG. 2B illustrates a top view 2″^d preferred embodiment.

FIG. 2.1 to FIG. 2.6 illustrate cross sectional views along bit line direction at individual process steps to the 2^nd embodiment.

FIG. 2.3 s to FIG. 2.6s illustrate cross sectional views along other direction of FIG. 2.3 to FIG. 2.6.

FIG. 3A illustrates a cross-sectional view of 3^rd embodiment.

FIG. 3B illustrates a top view 3^rd embodiment.

FIG. 3.1 to FIG. 3.8 illustrate cross sectional views along bit line direction at individual process steps to the 3^rd embodiment.

FIG. 3.5 s to FIG. 3.8s illustrate cross sectional views along other direction of FIG. 3.5 to FIG. 3.8.

FIG. 4A illustrates a cross-sectional view of 4^th embodiment.

FIG. 4B illustrates a top view 4^th embodiment.

FIG. 4.1 to FIG. 4.8 illustrate cross sectional views along bit line direction at individual process steps to the 4^th embodiment.

FIG. 4.3 s to FIG. 4.8s illustrate cross sectional views along other direction of FIG. 4.3 to FIG. 4.8.

FIG. 5A illustrates a cross-sectional view of 5^th preferred embodiment.

FIG. 5B illustrates a top view 5^th embodiment.

FIG. 5.1 to FIG. 5.5 illustrate cross sectional views along bit line direction at individual process steps to the 5^th embodiment.

FIG. 5.2 s to FIG. 5.5s illustrate cross sectional views along other direction of FIG. 5.2 to FIG. 5.5.

DETAILED DESCRIPTION

In the following description, for purposes of explanation, numerous specific details are set forth in order to provide a thorough understanding of the invention. It will be apparent, however, to one skilled in the art that the invention can be practiced without these specific details. In other instances, structures and devices are shown in block diagram form only in order to avoid obscuring the invention.

Reference in this specification to “one embodiment” or “an embodiment” means that a particular feature, structure, or characteristic described in connection with the embodiment is included in at least one embodiment of the invention. The appearance of the phrase “in one embodiment” in various places in the specification are not necessarily all referring to the same embodiment, nor are separate or alternative embodiments mutually exclusive of other embodiments. Moreover, various features are described which may be exhibited by some embodiments and not by others. Similarly, various requirements are described which may be requirements for some embodiments but not other embodiments.

Moreover, although the following description contains many specifics for the purposes of illustration, anyone skilled in the art will appreciate that many variations and/or alterations to said details are within the scope of the present invention. Similarly, although many of the features of the present invention are described in terms of each other, or in conjunction with each other, one skilled in the art will appreciate that many of these features can be provided independently of other features. Accordingly, this description of the invention is set forth without any loss of generality to, and without imposing limitations upon, the invention.

Prior Art FIG. A shows a cross-sectional view through a prior art MRAM cell, whereas Prior Art FIG. B shows a plan view of the MRAM cell. As can be seen the MRAM cell includes a MTJ (Magnetic Tunnel Junction) as a memory element. The MTJ is connected to upper and lower metals through via holes where overlay margin ∂ is required on the both edges of via hole landing area. The MTJ cell is designed to be bigger than the upper through hole to upper metal by 2∂. Since the upper metal should cover the MTJ, the upper metal becomes bigger than the MTJ by 2∂. The upper metal width consequently becomes 4∂ bigger than a feature size f of the via hole. Overlay margin is estimated to be 20% to 30% of the minimum 1−b. As will be seen, the upper metal 113 is directly connected feature size. The metal width would be twice bigger than minimum feature size.

FIG. 1A shows a cross sectional view of a first embodiment of an MRAM cell. A top view of the first embodiment is shown in FIG. 1B. As will be seen, the upper metal 113 is directly connected to the top of MTJ. Overlay margin of MTJ to via is not necessary so that upper metal width becomes f+2∂ considering overlay margin of upper metal to MTJ. Thus, the first embodiments saves 2∂ compared to conventional structure showed in Prior Art FIG. A and Prior Art FIG. B.

As shown in FIG. 1.1, a lower metal as write word line 101 and landing pad 102 to read device are patterned after the FEOL process is completed. The surface over write word line is planerized with CMP. Bottom read lead 104, MTJ Pin layer 105, tunnel oxide 106, MTJ fixed layer 107 and hard mask layer are subsequently deposited as shown in FIG. 1.2. Patterning photo resist 108 with MTJ pillar mask in FIGS. 1.3 and 1.3s, MTJ stack (107, 106, 105) is etched with ion milling or reactive ion etch with end point at read lead metal 104 surface. Read lead metal is patterned with photo resist mask 109 and etched also with ion milling or reactive ion etch as shown in FIGS. 1.4 and 1.4s. A dielectric layer having enough etch selectivity to oxide such as nitride is deposited and vertically etched as shown in FIGS. 1.5 and 1.5s to put dielectric spacer 110 on MTJ sidewall to protect the junction 106. Oxide 111 as an inter dielectric layer is deposited and planerized as shown in FIGS. 1.6 and 1.6s. Trench line 112 is formed in oxide 111 using conventional damascene process. The trench etch goes until top of MTJ surface completely appears as shown in FIGS. 1.7 and 1.7s. Seed layer is deposited and copper 112 is plugged in trench with electro plating. Conventional copper CMP is used to remove excess copper out side of the trench as shown in FIGS. 1.8 and 1.8s.

A cross sectional view of the 2^nd embodiment is shown in FIG. 2A. Top view is in FIG. 2B. The MTJ pillar is coupled to the upper metal 213 without the need of a via. Overlay margin of MTJ to via is thus not necessary so that upper metal width becomes f+2∂ considering overlay margin of upper metal to MTJ as discussed in the first embodiment. This embodiment saves 2∂ compared to conventional structure shown in Prior Art FIG. A and FIG. B.

As shown in FIG. 2.1, lower metal as write word line 201 and landing pad 202 to read device are patterned after FEOL process is completed. The surface over write word line is planerized with CMP. Bottom read lead 204, MTJ Pin layer 205, tunnel oxide 206, MTJ fixed layer 207 and hard mask layer 208 consisting of oxide and Titanium nitride are subsequently deposited as shown in FIG. 2.2. Titanium Nitride layer and oxide layer 208 are patterned using conventional lithography and mask etch as shown in FIGS. 2.3 and 2.3s. Vertical ion etching with Ion milling or reactive ion allows to transfer the hard mask patter into MTJ stack as in FIGS. 2.4 and 2.4s, with end point at read lead metal 204 surface, followed by read lead metal patterning similar to the first embodiment. Oxide 210 as an inter dielectric layer is deposited as shown in FIGS. 2.5 and 2.5.s. CMP is allowed until Titanium nitride appears on surface as shown in FIGS. 2.6 and 2.6s, followed by conventional metal dry etch process.

A cross-sectional view of the 3^rd embodiment is shown in FIG. 3A. A top view of the 3^rd embodiment is shown in FIG. 3B. A self-aligned via connects the MTJ pillar/stack to the upper metal. Overlay margin of MTJ to via is not necessary so that upper metal width becomes f+2∂ considering overlay margin of upper metal to MTJ as discussed in the first embodiment. It save 2∂ compared to conventional structure showed in Prior Art FIG. A. and FIG. B.

As shown in FIG. 3.1, lower metal as write word line 301 and landing pad 302 to read device are patterned after FEOL process is completed. The surface over write word line is planerized with CMP. Bottom read lead 304, MTJ Pin layer 305, tunnel oxide 306, MTJ fixed layer 307 and hard mask layer consisting of bottom oxide 308 and nitride 309 are subsequently deposited as shown in FIG. 3.2. Nitride layer and oxide layer are patterned using conventional lithography and mask etch as shown in FIG. 3.3. Vertical ion etching with Ion milling or reactive ion etch allows to transfer the hard mask patter into MTJ stack as in FIG. 3.4, with end point at read lead metal 304 surface, followed by read lead metal patterning similar to the first embodiment. Oxide 310 as an inter dielectric layer is deposited and planerized as shown in FIGS. 3.5 and 3.5s. Trench line 311 is formed in oxide 310 using conventional damascene process. The trench etch goes until top of hard mask nitride surface completely appears as shown in FIGS. 3.6 and 3.6s. Exposed nitride 309 is removed with hot phosphoric acid as shown in FIGS. 3.7 and 3.7s. The self aligned via structure 312 delivered. Adding oxide etch, the oxide 308 over MTJ is etched and MTJ surface appears. Seed layer is deposited and copper 313 is plugged in trench with electro plating. Conventional copper CMP remove excess copper out side of the trench as shown in FIGS. 3.8 and 3.8s.

A cross-sectional view of the 4^th embodiment is shown in FIG. 4A. A top view of the 4^th embodiment is shown in FIG. 4B. MTJ is patterned twice. Firstly along the word line direction and secondly along the bit line direction. At 2^nd patterning, upper metal layer, MTJ and bottom read lead are patterned with one mask. No overlay margin is required so that upper metal width becomes same feature size as MTJ. This embodiment saves 4∂ compared to conventional structure showed in Prior Art FIG. A and FIG. B. The structure has other benefit than cell size. The upper metal wraps around the MTJ. The current flowing the metal induces stronger magnetic field than straight metal line. It works better to switch the pin layer direction.

As shown in FIG. 4.1, lower metal as write word line 401 and landing pad 402 to read device are patterned after FEOL process is completed. The surface over write word line is planerized with CMP. Bottom read lead 404, MTJ Pin layer 405, tunnel oxide 406, MTJ fixed layer 407 and hard mask layer are subsequently deposited as shown in FIG. 4.2. With the same process step as previous embodiments, MTJ stack 408 is patterned as a line along word line direction as shown in FIGS. 4.3 and FIG. 4.3s. Nitride spacer 409 is placed on side wall of MTJ line as shown in FIGS. 4.4 and 4.4s. Oxide 410 is deposited and planerized as shown in FIGS. 4.5 and 4.5s. The planerized oxide is recessed with vertical ion etching until top of MTJ line appears enough as shown in FIGS. 4.6 and 4.6s. Remained oxide 411 in FIG. 4.6 is to insulate upper metal from bottom read lead metal. Upper metal 412 like as aluminum/Cu alloy is deposited as shown in FIGS. 4.7 and 4.7s. Patterning photoresist, the upper metal is etched with conventional metal etching process by reaching to insulation oxide 411. Subsequent Ion milling etches oxide, MTJ and bottom read lead metal to get self-aligned structure 413 as shown in FIGS. 4.8 and 4.8s.

A cross-sectional view of the 5^th embodiment is shown in FIG. 5A. A top view of the 5^th embodiment is shown in FIG. 5B. MTJ is connected lower metal line (write word line) instead of connecting upper metal as adapted in previous embodiments. Read lead is connected to top of MTJ different from previous 4 embodiments. Thin oxide separates upper metal and read lead/MTJ electrically. MTJ is also patterned twice along word line direction first and bit line direction 2^nd as was the case with the 4^th embodiment. At 2^nd patterning, upper metal layer, MTJ and bottom read lead are patterned with one mask. No overlay margin is required so that upper metal width becomes same feature size as MTJ. It save 4∂ compared to conventional structure showed in Prior Art FIG. A and FIG. B. The structure has other benefit than cell size. The upper metal wraps around the MTJ. The current flowing the metal induces stronger magnetic field than straight metal line. It works better to switch the pin layer direction.

As shown in FIG. 5.1, lower metal as write word line 501 and landing pad 502 to read device are patterned after FEOL process is completed. The vias 503 and 504 to be connected to MTJ and read lead metal are opened over 501 and 502.

Tungsten is deposited and allows CMP to make the surface smooth. MTJ Pin layer 505, tunnel oxide 506, MTJ fixed layer 507 and hard mask layer are subsequently deposited as previous embodiments. The stack is patterned as a line along the word line direction and followed by spacer oxide protect the MTJ sidewall as shown in FIGS. 5.2 and 5.2s. Read metal 509 is deposited and patterned as shown in FIGS. 5.3 and 5.3s. With the same process step as previous embodiments, Thin oxide 510 is deposited to insulate MTJ/Read Metal and upper metal (Bit line). Upper metal 511 like as aluminum/Cu alloy is deposited as shown in FIGS. 5.4 and 5.4s. Patterning photoresist, the upper metal is etched with conventional metal etching process by reaching to insulation oxide 510. Oxide 510 can be removed by wet etch. Subsequent Ion milling etches read lead metal, MTJ as shown in FIGS. 5.5 and 5.5s.

Although the present invention has been described with reference to specific exemplary embodiments, it will be evident that the various modification and changes can be made to these embodiments without departing from the broader spirit of the invention. Accordingly, the specification and drawings are to be regarded in an illustrative sense rather than in a restrictive sense.

Claims

1. A magnetic memory cell, comprising: a memory element; andan upper metal layer; wherein the memory element is directly coupled to the upper metal layer without the use of a via.

2. The magnetic memory cell of claim 1, wherein the memory element comprises a Magnetic Tunnel Junction (MTJ) stack.

3. The magnetic memory cell of claim 2, wherein a size of the MTJ stack is f and a width of the upper metal layer is f+2∂, where ∂ is an overlay margin for the upper metal layer.

4. The magnetic memory cell of claim 2, further comprising dielectric sidewalls formed on opposed sides of the MTJ stack.

5. The magnetic memory cell of claim 1, further comprising: a hard mask layer etched and patterned to sit atop the memory element.

6. The magnetic memory cell of claim 5, wherein the memory element comprises a Magnetic Tunnel Junction (MTJ) stack.

7. The magnetic memory cell of claim 5, wherein a size of the MTJ stack is f and a width of the upper metal layer is f+2∂, where ∂ is an overlay margin for the upper metal layer.

8. The magnetic memory cell of claim 5, wherein the hard mask layer comprises a material selected from the group consisting of titanium nitride, titanium, aluminum, and tantalum.

9-11. (canceled)

12. A magnetic memory cell, comprising: a magnetic memory element;a read lead directly coupled to a lower end of the magnetic memory element;a bit line directly coupled to an upper end of the magnetic memory element; anda word line positioned under the magnetic memory element such that the read lead passes between the word line and the magnetic memory element.

13. The magnetic memory cell of claim 12, wherein a size of the magnetic storage element is f and a width of the bit line is f+2∂, where ∂ is an overlay margin for the bit line.

14. The magnetic memory cell of claim 12, wherein the magnetic memory element comprises a Magnetic Tunnel Junction (MTJ) stack.

15. The magnetic memory cell of claim 13, further comprising dielectric sidewalls formed on sides of the MTJ stack.

16. The magnetic memory cell of claim 12, further comprising a hard mask layer etched and patterned to sit atop the memory element.

17. A magnetic memory cell, comprising: a Magnetic Tunnel Junction (MTJ) stack including a plurality of layers;a read lead directly coupled to a lower layer of the MTJ stack;a bit line directly coupled to an upper layer of the MTJ stack; anda word line positioned under the MTJ stack such that the read lead passes between the word line and the MTJ stack.

18. The magnetic memory cell of claim 17, wherein a size of the MTJ stack is f and a width of the bit line is f+2∂, where ∂ is an overlay margin for the bit line.

19. The magnetic memory cell of claim 17, wherein the plurality of layers of the MTJ stack comprises a pin layer, a tunnel oxide layer, and a fixed layer.

20. The magnetic memory cell of claim 17, further comprising dielectric sidewalls formed on sides of the MTJ stack.

21. The magnetic memory cell of claim 17, further comprising a hard mask layer etched and patterned to sit atop the MTJ stack.

22. The magnetic memory cell of claim 21, wherein a size of the MTJ stack is f and a width of the upper metal layer is f+2∂, where ∂ is an overlay margin for the upper metal layer.

23. The magnetic memory cell of claim 21, wherein the hard mask layer comprises a material selected from the group consisting of titanium nitride, titanium, aluminum, and tantalum.