HIGHLY PHYSICAL ION RESISTIVE SPACER TO DEFINE CHEMICAL DAMAGE FREE SUB 60NM MRAM DEVICES

Abstract

A magnetic tunneling junction (MTJ) structure comprises a pinned layer on a bottom electrode. a barrier layer on the pinned layer, wherein a second metal re-deposition layer is on sidewalls of the barrier layer and the pinned layer, a free layer on the barrier layer wherein the free layer has a first width smaller than a second width of the pinned layer, a top electrode on the free layer having a same first width as the free layer wherein a first metal re-deposition layer is on sidewalls of the free layer and top electrode, and dielectric spacers on sidewalls of the free layer and top electrode covering the first metal re-deposition layer wherein the free layer and the top electrode together with the dielectric spacers have a same the second width as the pinned layer wherein the dielectric spacers prevent shorting between the first and second metal re-deposition layers.

Description

TECHNICAL FIELD

This application relates to the general field of magnetic tunneling junctions (MTJ) and, more particularly, to methods for preventing shorts and sidewall damage in the fabrication of sub 60 nm MTJ structures.

BACKGROUND

Fabrication of magnetoresistive random-access memory (MRAM) devices normally involves a sequence of processing steps during which many layers of metals and dielectrics are deposited and then patterned to form a magnetoresistive stack as well as electrodes for electrical connections. To define the magnetic tunnel junctions (MTJ) in each MRAM device, precise patterning steps including photolithography and reactive ion etching (RIE), ion beam etching (IBE) or their combination are usually involved. During RIE, high energy ions remove materials vertically in those areas not masked by photoresist, separating one MTJ cell from another. However, the high energy ions can also react with the non-removed materials, oxygen, moisture and other chemicals laterally, causing sidewall damage and lowering device performance. To solve this issue, pure physical etching techniques such as pure Ar RIE or ion beam etching (IBE) have been applied to etch the MTJ stack. However, due to the non-volatile nature, pure physically etched conductive materials in the MTJ and bottom electrode can form a continuous path across the tunnel barrier, resulting in shorted devices. A new approach to remove these two kinds of sidewall damage is thus needed for the future sub 60 nm MRAM products.

Several patents teach two-step methods of etching MTJ stacks, including U.S. Pat. No. 9,087,981 (Hsu et al), U.S. Pat. No. 9,406,876 (Pinarasi), and U.S. Pat. No. 9,728,718 (Machkaoutsan et al), but these methods are different from the present disclosure.

SUMMARY

It is an object of the present disclosure to provide a method of forming MTJ structures without chemical damage on the MTJ sidewalls or shorting of MTJ devices.

Another object of the present disclosure is to provide a method of forming MTJ structures having chemical damage free MTJ sidewalls and eliminating conductive metal re-deposition induced shorted devices.

Another object of the present disclosure is to provide a method of forming MTJ structures having chemical damage free MTJ sidewalls and eliminating conductive metal re-deposition induced shorted devices using a spacer assisted pure physical etch.

In accordance with the objectives of the present disclosure, a method for fabricating a magnetic tunneling junction (MTJ) structure is achieved. A MTJ stack is deposited on a bottom electrode wherein the MTJ stack comprises at least a pinned layer, a barrier layer on the pinned layer, and a free layer on the barrier layer. A top electrode layer is deposited on the MTJ stack. A hard mask is deposited on the top electrode layer. The top electrode layer and the free layer not covered by the hard mask are etched, stopping at or within the barrier layer. Thereafter, the hard mask, top electrode layer, and free layer are encapsulated with an encapsulation layer. A spacer layer is deposited over the encapsulation layer and the spacer layer is etched away on horizontal surfaces leaving spacers on sidewalls of the encapsulation layer wherein sidewalls of the free layer are covered by a combination of the encapsulation layer and spacers. Thereafter, the barrier layer and pinned layer are etched to complete formation of the MTJ structure.

Also in accordance with the objectives of the present disclosure, a magnetic tunneling junction (MTJ) structure comprises a pinned layer on a bottom electrode. a barrier layer on the pinned layer, wherein a second metal re-deposition layer is on sidewalls of the barrier layer and the pinned layer, a free layer on the barrier layer wherein the free layer has a first width smaller than a second width of the pinned layer, a top electrode on the free layer having a same first width as the free layer wherein a first metal re-deposition layer is on sidewalls of the free layer and top electrode, and dielectric spacers on sidewalls of the free layer and top electrode covering the first metal re-deposition layer wherein the free layer and the top electrode together with the dielectric spacers have a same second width as the pinned layer wherein the dielectric spacers prevent shorting between the first and second metal re-deposition layers.

BRIEF DESCRIPTION OF THE DRAWINGS

In the accompanying drawings forming a material part of this description, there is shown:

FIGS. 1 through 8 illustrate in cross-sectional representation steps in a preferred embodiment of the present disclosure.

DETAILED DESCRIPTION

In the present disclosure, a spacer assisted pure physical etch can create chemical damage free MTJ sidewalls and also eliminate conductive metal re-deposition induced shorted devices. More specifically, the free layer is physically etched by pure Ar RIE or IBE, then covered by a spacer. Next, the pinned layer is physically etched using the spacer as a hard mask. The spacer material can be made of carbon or TaC, which is highly resistant to this type of etch, thus ensuring that enough of the spacer remains to protect the free and barrier layers. This method is particularly useful for high density sub 60 nm MRAM devices, where chemical damage and re-deposition on the MTJ sidewall become very severe for these smaller MRAM chips.

In a typical MRAM fabrication process, the whole MTJ stack consisting of free, barrier, and pinned layers is patterned by one single step etch, either by chemical RIE or physical IBE. It therefore creates either chemical damage or physical shorts on the MTJ sidewall. However, in the process of the present disclosure, we firstly etch the free layer by pure Ar RIE or IBE, cover it with a highly physical etch resistant spacer, and then etch the pinned layer by pure Ar RIE or IBE using the spacer as a hard mask. By this method, both issues are solved simultaneously, greatly enhancing the device performance.

The preferred embodiment of the present disclosure will be described in more detail with reference to FIGS. 1-8. FIG. 1 illustrates a bottom electrode layer 12 formed on a semiconductor substrate, not shown. Now, the MTJ stack, comprising at least a pinned layer 14, a tunnel barrier layer 16, and a free layer 18, are deposited on the bottom electrode. A top electrode 20 comprising Ta, TaN, Ti, TiN, W, Cu, Mg, Ru, Cr, Co, Fe, Ni or their alloys is deposited over the MTJ stack to a thickness h1 of 10-100 nm, and preferably≥50 nm. A dielectric hard mask 22 of SiO₂, SiN, SiON, SiC or SiCN is deposited on the top electrode to a thickness of ≥20 nm. Finally a photoresist mask 24 is formed over the hard mask 20 forming pillar patterns with size d1 of approximately 70-80 nm and height h2 of ≥200 nm.

Now, as shown in FIG. 2, the hard mask 22 is etched by a fluorine carbon based plasma such as CF₄ or CHF₃ alone, or mixed with Ar and N₂. O₂ can be added to reduce the pillar size further to d2 of about 50-60 nm.

Next the top electrode is etched by RIE or IBE, followed by a pure Ar RIE or IBE etch of the free layer. If RIE is used to etch the top electrode, the top electrode and free layer etching must be in separate steps since RIE causes chemical damages and cannot be applied to the free layer. If IBE is used, the top electrode and free layer can be etched by one single etch step using the same recipe. The free layer etch step can stop at the interface between the free layer 18 and the tunneling barrier 16 or within the tunneling barrier. Because of the nature of a physical etch, there is no chemical damage after this etching step, but only a thin layer of conductive metal re-deposition 26 on the free layer's sidewall, as shown in FIG. 3.

The photoresist 24 is stripped away by oxygen alone or mixed with N₂ or H₂O. Then, as illustrated in FIG. 4, an encapsulation layer 28, made of Al₂O₃, SiON, or SiN, is deposited over the partially etched MTJ stack and the bottom electrode by chemical vapor deposition (CVD), physical vapor deposition (PVD), or atomic layer deposition (ALD) to a thickness of 5-30 nm. The deposition may be either in-situ or ex-situ. This non-conductive encapsulation layer will protect the free layer from shorting by encapsulating the conductive metal re-deposition 26.

Now, as shown in FIG. 5, a spacer material layer 30 is deposited over the encapsulation layer 28. The spacer material 30 is carbon, TaC, or Al which has a very low etch rate under physical etching. The spacer material layer 30 is deposited in-situ or ex-situ by CVD, PVD, or ALD to a thickness of 10-30 nm.

Next the portion of the spacer layer 30 that is on horizontal surfaces is etched away by RIE, leaving spacers 32 having a thickness of 5-20 nm only on the sidewalls of the pattern, as shown in FIG. 6. Dependent on the material that is used for the spacer, different plasmas can be used for this step. For example, O₂ can be applied for carbon, fluorine carbon such as CF₄ or a Halogen such as Cl₂ can be used for TaC, and a Halogen such as Cl₂ can be used for Al.

Referring to FIG. 7, using the spacer 32 left on the encapsulated free layer sidewall as a self-aligned hard mask, the barrier layer 16 and pinned layer 14 are etched by the same type of physical etch as was used to etch the free layer. By doing this, one can again avoid forming any chemical damage layer but only generate a thin layer of conductive metal re-deposition 34 on the pinned layer's sidewall. Since the pinned layer is thin, the re-deposition from it would not cover the whole spacer but at most the bottom portion of the spacer 32. Here it should be noted that, as shown in Table 1, carbon's IBE etch rate is only ˜60 A/s, much lower than those commonly used metals in the MTJ hard mask and stack, which are usually larger than 200 A/s. That is, the spacer's etch rate would be ≤⅓ the etch rate of the pinned layer. Therefore after this etch step, enough of the spacer 32, ≥1 nm, would remain to protect the encapsulated free layer.

TABLE 1
Summary of various materials' IBE etch rate in Angstroms/minute
(from http://www.microfabnh.com/ion_beam_etch_rates.php)
          Etch Rate
Material  (A/min)
Ag        1050
Al        48
Au        630
AZ 1350   117
C         64
CdS       1283
Co        262
Cr        309
Cu        513
Fe        204
Si        216
SiC       204
SiO2      192
Hf        385
InSb      887
Ir        344
Ge        537
Mg        131
Mn        507
Mo2C      163
Nb        274
Ni        309
NiCr      309
Pb        1517
PbTe      2199
Pd        642
Rb        2333
Re        303
Rh        420
Riston 14 146
Ru        356
Sb        1889
Ni80Fe2O  292
Ni        309
Zr        332
Ta        245
Ta2O5     350
TaC       87
TaN       233
Ti        192
Ti or TiW 195
W         198
Y         554
Zr        332

The re-deposition from the free and pinned layer etches, 26 and 34, respectively, are separated by the encapsulation 28 and spacer 32 materials, without forming a continuous path to short the devices. This approach is of particular use for sub 60 nm MRAM devices where the spacer has to be thin enough to maintain the pattern geometry for the self-aligned etch, but still be capable of protecting the previously defined free layer. Another benefit of this spacer etch is that the pinned layer has a larger volume than the free layer, about 50-60 nm for the pinned layer and about 40-50 nm for the free layer, so that the pinned layer has strong enough pinning strength to stabilize the magnetic state in the free layer.

After the pinned layer etch, the whole device can be filled with dielectric material 36 and flattened by chemical mechanical polishing (CMP) to expose the top electrode 20, as shown in FIG. 8. The remaining spacer 32 on the sidewall can stay within the structure without affecting the device integrity and device performance. A top metal contact 38 contacts the top electrode 20.

The process of the present disclosure employs a physical etch to eliminate chemical damage on the MTJ sidewall and prevents the conductive re-deposition from shorting the devices. It has been considered to be difficult to achieve these two results simultaneously, but the process of the present disclosure provides these results.

The process of the present disclosure will be used for MRAM chips of size smaller than 60 nm as problems associated with chemically damaged sidewalls and re-deposition from the bottom electrode become very severe for these smaller sized MRAM chips.

Although the preferred embodiment of the present disclosure has been illustrated, and that form has been described in detail, it will be readily understood by those skilled in the art that various modifications may be made therein without departing from the spirit of the disclosure or from the scope of the appended claims.

Claims

1. A device comprising: a stack of magnetic tunneling junction (MTJ) layers, the stack of MTJ layers including a first portion and a second portion;a first metal re-deposition layer disposed on the first portion of the stack of MTJ layers,a second metal re-deposition layer disposed on the second portion of the stack of MTJ layers; anda first dielectric layer disposed between the first and second metal re-deposition layers thereby electrically isolating the first metal re-deposition layer from the second metal re-deposition layer, wherein the first metal re-deposition layer has a first surface and the second metal re-deposition layer has a second surface that faces the first surface, and wherein the first dielectric layer physically contacts the first surface of the first metal re-deposition layer and the second surface of the second metal re-deposition layer.

2. The device of claim 1, wherein the first portion of the stack of MTJ layers includes a pinned layer over a tunnel barrier layer, wherein the second portion of the stack of MTJ layers includes a top electrode layer over a free layer.

3. The device of claim 1, wherein the first portion extends a first width along a first direction, the second portion extends a second width along the first direction, and the first width is greater than the second width.

4. The device of claim 1, further comprising: a third metal re-deposition layer disposed on the first portion of the stack of MTJ layers, wherein the first portion of the stack of MTJ layers is disposed laterally between the first and the third metal re-deposition layers; anda fourth metal re-deposition layer disposed on the second portion of the stack of MTJ layers, wherein the second portion of the stack of MTJ layers is disposed laterally between the second and the fourth metal re-deposition layers.

5. The device of claim 4, wherein the first and the third metal re-deposition layers are spaced apart at a first distance, the second and the fourth metal re-deposition layers are spaced apart at a second distance, and the first distance is greater than the second distance.

6. The device of claim 1, wherein the first dielectric layer includes a nitride material.

7. The device of claim 1, further comprising: a second dielectric layer extending along sidewalls of the first dielectric layer and landing on a horizontal surface of the first dielectric layer.

8. The device of claim 7, wherein the first and the second dielectric layer includes different materials, and the second dielectric layer includes carbon or aluminum.

9. The device of claim 1, further comprising: a bottom electrode layer; anda top metal contact layer,wherein the first portion of the stack of MTJ layers lands on the bottom electrode layer, and the top metal contact layer lands on the second portion of the stack of MTJ layers.

10. The device of claim 9, wherein each of the bottom electrode layer and the top metal contact layer extends a greater width along a first direction than that of the first portion of the stack of MTJ layers.

11. A device comprising: a bottom electrode;a stack of magnetic tunneling junction (MTJ) layers over the bottom electrode;first metal re-deposition layers along sidewalls of first portions of the stack of MTJ layers;second metal re-deposition layers along sidewalls of second portions of the stack of MTJ layers;a dielectric layer separating the first metal re-deposition layers from the second metal re-deposition layers; anda metal contact over the stack of the MTJ layers.

12. The device of claim 11, wherein the first portions of the stack of MTJ layers includes a pinned layer having a first width, the second portions of the stack of MTJ layers includes a free layer having a second width, and the first width is different from the second width.

13. The device of claim 12, wherein the first portions further includes a tunnel barrier layer having the first width, and the tunnel barrier layer is vertically disposed between the pinned layer and the free layer.

14. The device of claim 12, wherein the second portions further includes a top electrode having the second width, and the top electrode is vertically disposed between the free layer and the metal contact.

15. The device of claim 11, wherein the dielectric layer includes a spacer layer disposed on an encapsulation layer, wherein the spacer layer includes carbon or aluminum.

16. The device of claim 15, wherein the first metal re-deposition layers directly contact the encapsulation layer and a bottom portion of the spacer layer, and the second metal re-deposition layers directly contact the encapsulation layer without contacting the spacer layer.

17. The device of claim 11, wherein the metal contact lands on the stack of the MTJ layers, the second metal re-deposition layers, and the dielectric layer.

18. A device comprising: a first electrode;a pinned layer on the first electrode, the pinned layer having a first width;a barrier layer on the pinned layer, the barrier layer having the first width;a free layer on the barrier layer, the free layer having a second width different from the first width;a second electrode on the free layer, the second electrode having the second width;a first metal re-deposition layer on sidewalls of the barrier layer and the pinned layer; anda second metal re-deposition layer on sidewalls of the free layer and the second electrode, wherein the first metal re-deposition layer is physically separated from the second metal re-deposition layer.

19. The device of claim 18, further comprising: a dielectric layer disposed between and in direct contact with the first and the second metal re-deposition layers, and the first metal re-deposition layer is physically separated from the second metal re-deposition layer by the dielectric layer.

20. The device of claim 19, further comprising: a spacer disposed along sidewalls of the dielectric layer, wherein the first width plus a total width of the dielectric layer and the spacer equals the second width.