TECHNICAL FIELD
The disclosed structures and processes relate to ways to improve the performance of a TAMR feature by making the separation between the edge plasmon generator and the plasmon shield less than the separation between the edge plasmon generator and the optical wave-guide.
BACKGROUND
Thermally assisted magnetic recording (TAMR) is expected to be one of the future generation magnetic recording technologies that will enable recording at 1˜10 Tb/inch2 data density. TAMR converts optical power into localized heating of the magnetic recording medium to temporarily reduce the field needed to switch the magnetizations of the medium grains. With a sharp temperature gradient alone, or together with the magnetic field gradient when both gradients are aligned correctly, data storage density can be further improved over the current state-of-the-art magnetic recording technology.
A TAMR head usually comprises, in addition to its conventional magnetic recording structure, an optical wave-guide (WG) and a Plasmon generator (PG). The WG serves to guide external laser light to the PG, where the optical mode is coupled to the Plasmon mode of the PG. After being converted to plasmon mode the optical energy then concentrated at the location where heating of the medium is required. When the heating spot is correctly aligned relative to the write field of the magnetic recording structure, TAMR is achieved.
We refer now to the prior art air bearing surface (ABS) view shown in FIG. 1a. This illustrates a TAMR head located at the end of main pole 10, integrated with Edge Plasmon generator (EPG) 15 and having, in cross-section, triangular shape 16. This shaped edge is placed in the vicinity of optical waveguide 11 where it supports the very confined Edge Plasmon (EP) mode. The optical energy in WG 11 is efficiently transformed to edge plasmon mode through evanescent coupling and its energy is directed towards the ABS. The local confinement of the edge plasmon mode is determined by the angle and radius of 16's triangular corner, by the noble metal from which the EPG is formed, as well as by the dielectric material that surrounds tip 16.
Referring next to FIG. 1b, for a 25 nm tip radius, the size of optical spot 18 in recording medium 9 is about 100 nm across its half-maximum intensity area. By placing plasmon shield 12 a small dielectric gap distance from EPG 15, optical spot 18 can be further reduced in size since, in the presence of plasmon shield 12, the spot size is mainly determined by PSG which is gap distance 13 between edge plasmon generator 15 and plasmon shield 12. For example, a 50 nm optical spot size can be achieved if this gap distance is less than 40 nm.
Plasmon shield 12 is placed at the front of wave-guide 11. Using the current (prior art) process, the top surface of plasmon shield 12 is at the same level as the top surface of WG 11 or even slightly lower than the top surface of WG 11 due to different CMP rates for the Au of layer 15 and the Ta2O5 used for WG 11.
When PSG 13 is scaled down, WEG 31 (the gap between WG 11 and EPG 15) will also be reduced. One consequence of a reduced WEG is poorer optical coupling efficiency between WG 11 and EPG 15, so some optical power will be wasted as a result. This coupling efficiency cannot be improved by fine turning the length of EPG 15 when WEG 31 is less than 25 nm. Thus, to simultaneously achieve both good optical efficiency and a small optical spot, it is important to have both a large fixed WEG as well as a small PSG.
SUMMARY
It has been an object of at least one embodiment of the present disclosure to simultaneously achieve both good optical efficiency and a small optical spot in a TAMR magnetic write head.
Another object of at least one embodiment of the present disclosure has been to devise a structure in which the gap between the plasmon shield and the plasmon generator is smaller than the gap between the waveguide and the plasmon generator.
Still another object of at least one embodiment of the present disclosure has been to provide a process for manufacturing said structure
These objects have been achieved in three different ways:
In the first structure that is disclosed, a triangular indentation is formed in the top surface of the wave-guide so WEG becomes the distance between the floor of this indentation and the edge of the plasmon generator. Since the top surface of the plasmon shield is coplanar with the top surface of the wave-guide, it follows that PSG is smaller than WEG.
In the second structure that is disclosed, the sharp lower edge of the plasmon generator comprises two seamlessly connected parts. The first part is directly above the plasmon shield and is also closer to the plane of the plasmon shield's top surface than the second part is.
In the third structure that is disclosed, there is a ‘dummy’ a layer of dielectric material on the optical wave-guide's top surface with the plasmon shield located between the dummy layer and the ABS whereby any plasmon radiation propagating towards the ABS through the dummy layer will be blocked;
BRIEF DESCRIPTION OF THE DRAWINGS
FIGS. 1
a and 1b show two views of a TAMR feature of the prior art.
FIGS. 2
a-2c illustrate a first embodiment of the disclosed structure
FIGS. 3
a-8c describe a process for manufacturing this first embodiment of the disclosed structure
FIG. 9 shows a second embodiment of the disclosed structure
FIGS. 10
a-13b describe a process for manufacturing this second embodiment of the disclosed structure
FIGS. 14
a-14c show a third embodiment of the disclosed structure
FIGS. 15-17 describe a process for manufacturing this third embodiment of the disclosed structure
DETAILED DESCRIPTION
A first embodiment of the disclosed methodology is shown in FIG. 2a. A first novel feature is how to increase WEG 21. This is achieved by forming indentation 22 which is a trench of triangular cross-section that extends downwards from the top surface of optical wave-guide 11, starting at plasma shield 12, and, as seen in cross-section in FIG. 2b, extending away therefrom for a sufficient distance to no longer be directly below edge plasmon generator 15. The thickness of PS 12 is 50-200 nm, and its top surface is aligned with the top surface of WG 11. The indentation can also be in other shapes, for example, rectangular or half circular.
The second novel feature is how to simultaneously decrease PSG 13. This is achieved by allowing PS 12 to extend above the floor of indentation 22, as shown in FIG. 2b. Also apparent from the view given in FIG. 2b is that WEG 21, the physical gap between WG 11 and EPG 15, is larger than PSG 13 (the gap between PS 12 and EPG 15). WEG 21 is kept in the 20-30 nm range to ensure efficient WG to EPG coupling.
It can be seen in FIG. 2c that plasma shield 12 is slightly wider than wave-guide 11. If so desired, PS 12 could even be made wide enough to connect with a suitably located heat sink.
The process to manufacture the first embodiment is as follows:
The process starts, as shown in FIGS. 3a and 3b, with the provision of optical wave-guide 11 onto whose top surface is deposited layer 31 of a low dielectric constant material such as alumina. Photoresist mask 32 is then formed on layer 31, following which, as shown in FIGS. 4a and 4b, triangular indentation 22 is etched all the way through layer 31 as well as an additional short distance (between about 5 and 40 nm) into wave-guide 11.
Control of the latter depth is achieved through over-etching layer 31 using an angled Ion Beam Etching (IBE) for which tantala has a faster etch rate than alumina. Note that FIG. 4a represents a cross-section made through the floor (i.e. apex) of indentation 22 so dielectric layer 31 and photoresist mask 32 do not appear in FIG. 4a even though they are still present in the structure at this point.
Following the removal of mask 32 and dielectric layer 31, the structure has the appearance shown in FIGS. 5a and 5b.
Referring next to FIGS. 6a and 6b, photoresist mask 62 is laid down to define the future location of trench 61 as extending inwards away from the ABS, following which trench 61 is formed by etching wave-guide 11 to a depth of 20 to 500 nm and then leaving photoresist layer in place.
Then, as shown in FIG. 7a, trench 61 is just filled with gold (or other suitable metal (such as Ru, Zr, Cr, Ta Ni, Co and their alloys), thereby forming plasmon shield 12, and photoresist 62 is lifted off to give the structure the appearance shown in FIGS. 7a and 7b. Finally, as illustrated in FIGS. 8a-8c. layer 81 of alumina is deposited to a thickness that exceeds the height of PS 12 above wave-guide 11's top surface, thereby determining values for both WEG and PSG (designated as gaps 31 and 13, respectively, in earlier figures). FIG. 8c is an ABS view of the completed structure before processing the EPG.
In a second embodiment, in the region directly over the plasma shield, the EPG lower edge is selectively brought closer to the plasma shield, as shown in FIG. 9 while the remainder of the EPG's bottom surface is left at its normal level of higher than the tip portion, whereby WEG 21 becomes larger than PSG 13. The top surface of PS 12 is at the same level as the top surface of the WG in this case.
Referring now to FIGS. 10a and 10b the process for the second embodiment begins with the provision of optical wave-guide 11 Then, at wave-guide 11's ABS end, plasma shield 12 is formed, as described earlier for the first embodiment (see FIGS. 6a to 8b) but with the CMP step continued until the top surface of PS 12 is coplanar with the top surface of WG 11.
Following the deposition of alumina layer 118 on the top surface of wave-guide 11, photoresist mask 116 is formed on layer 118 and cavity 112 is formed by under-etching layer 118 with mask 116 so that the floor of cavity 112 is located a distance above the top surface of wave-guide 11, as shown in FIG. 11b.
Referring next to FIGS. 12a and 12b, after protecting the area immediately above PS 13 with photoresist mask 126, first gold layer 121 is laid down to a thickness in the range of from 20 to 80 nm Following the removal of photoresist 126, second gold layer 121 is laid down to a thickness in the range of from 5 to 50 nm, giving the structure the appearance shown in FIGS. 9, 13a, and 13b.
In a third embodiment, a blocked layer is used to reduce PSG while leaving WEG unchanged. This is illustrated in FIGS. 14a-c. FIG. 14a shows dielectric layer 141 inside which plasmon radiation will have been induced by edge plasmon generator 15 (see FIG. 1) but which is unable to reach the recording medium since it is blocked by PS 12. FIG. 14b is a cross-section through WG 11 made some distance away from PS 12 while FIG. 14c is a cross-section through WG 11 made at the intersection of PS 12 with blocked layer 141. Note that blocked layer 141 is made of a low dielectric constant material such as silica or alumina, and its refractive index should be lower than that of WG 11.
As illustrated in FIG. 15, the process for manufacturing the blocked layer begins with the deposition of layer 141 on the top surface of WG 11. Layer 141 should have a thickness in a range of 5 to 25 nm with a thickness in a range of 10 to 20 nm being preferred. Next, as shown in FIG. 16, trench 161 is formed at the ABS end of WG 11. Finally, following the deposition of sufficient gold to fill trench 161, the structure is planarized to remove all gold outside trench 161, giving the completed structure illustrated in FIG. 17 which shows the top surface of PS 12 to be at the same level as the top surface of blocked layer 141.
The advantages of the reduced Plasmon shield gap structures and processes include:
- 1. The ability to reduce and shape the optical spot, thereby reducing the size of the thermal spot in the recording medium, resulting in a higher thermal gradient which achieves narrower track, higher BPI and greater areal density;
- 2. A larger wave-guide to Plasmon Generator gap together with a smaller Plasmon Shield to Plasmon Generator gap whereby there is minimal loss in optical efficiency;
- 3. The processes that have been disclosed for the manufacture these structures are simple to achieve as well as being suitable for mass production.