Recording physics of perpendicular media: soft underlayers
-
Upload
dmitri-litvinov -
Category
Documents
-
view
229 -
download
2
Transcript of Recording physics of perpendicular media: soft underlayers
Journal of Magnetism and Magnetic Materials 232 (2001) 84–90
Recording physics of perpendicular media: soft underlayers
Dmitri Litvinov*, Mark H. Kryder, Sakhrat Khizroev
Seagate Research, 2403 Sidney Street, Pittsburgh, PA 15203, USA
Received 24 January 2001; received in revised form 13 March 2001
Abstract
The results of both theoretical and experimental studies of some of the key aspects of the recording physics specific to
perpendicular media with a soft underlayer are presented. Some of the discussed issues are the material requirements forthe preferred soft underlayer design such as the role of magnetic anisotropy, the proper choice of the magnetic moment,and the control of the soft underlayer noise. # 2001 Elsevier Science B.V. All rights reserved.
Keywords: Perpendicular magnetic recording; Perpendicular recording media; Soft underlayer
1. Introduction
As the storage industry ramps the areal bitdensities at increasingly higher rates, thermalinstabilities in recording media begin to manifestthemselves [1]. Perpendicular recording technology[2] while being technically close to conventionallongitudinal recording and the least difficulttechnology to make the transition to, if necessary[3–5], addresses the issue of thermal stability forareal bit densities exceeding 100 Gbit/in2 [6].
One of the key aspects of perpendicular record-ing that makes it superior to the longitudinalrecording with respect to superparamagneticeffects is the utilization of media with a softunderlayer (SUL). A single-pole head/media witha soft underlayer perpendicular recording systemenables write fields in excess of 80% of 4pMS ofthe pole head/soft underlayer material [7]. This
doubles the fields available in longitudinal record-ing, thus opening the possibility to write onsubstantially higher anisotropy media and leadingto better thermal stability. Acting as a magneticmirror, soft underlayer effectively doubles therecording layer thickness, facilitating substantiallystronger readout signals [8]. Also, the effectivethickness increases due to the mirroring effects bya soft underlayer which leads to the reduction ofthe demagnetizing fields with a potential to furtherimprove thermal stability [11].
While the use of perpendicular media with a softunderlayer should make it possible to postpone thesuperparamagnetic limit, the soft underlayer in-troduces a number of technical challenges. Theissues related to the presence of the soft underlayerare the subject of this work.
2. Saturation moment
The importance of the relation between themoment of the soft underlayer material and the
*Corresponding author.
E-mail address: dmitri [email protected]
(D. Litvinov).
0304-8853/01/$ - see front matter # 2001 Elsevier Science B.V. All rights reserved.
PII: S 0 3 0 4 - 8 8 5 3 ( 0 1 ) 0 0 2 1 6 - 5
Information Storage: Basic and Applied
moment of the recording pole tip material wasdiscussed previously [7]. It was shown thatsaturation of the soft underlayer can lead to adramatic deterioration of the trailing field gradi-ents. It follows from Maxwell’s laws, div B ¼ 0and r�H ¼ 0, that to avoid the soft underlayersaturation under the pole tip, the followinginequality has to hold
4pMS soft underlayer � Asoft underlayer effective
54pMS pole tip � AABS pole tip; ð1Þ
where Asoft underlayer effective is the effective area ofthe soft underlayer into which the magnetic fluxemanating from the pole tip enters andAABS pole tip is the area of the air bearing surface(ABS) of the pole tip.
Fig. 1 shows a schematic of the recording poletip/soft underlayer combination. When the separa-tion between the ABS of the pole tip and thesoft underlayer (d �10–20 nm) is substantiallysmaller than the lateral dimensions of the poletip (lpole tip �300–1000 nm, wpole tip �50–150 nm),the effective area of the soft underlayer,Asoft underlayer effective, is approximately equal tothe ABS area of the pole tip, AABS pole tip. Itfollows that to avoid saturation effects in the softunderlayer the moment of the soft underlayermaterial has to be higher than or equal to themoment of the recording pole tip material.
Although it is possible to generate strongrecording fields (with the magnitude approaching4pMS of the pole tip) even if the soft underlayer
has lower magnetic moment than the pole tip,saturation of the soft underlayer will lead to asubstantial deterioration of the trailing fieldgradients. The results of boundary element model-ing for two different head/soft underlayer combi-nations are presented in Fig. 2 [7]. The trailinggradients in the case of the Permalloy based softunderlayer are substantially worse than the trailinggradients in the case when an FeAlN based softunderlayer is used.
3. Thickness
The following approximate equation gives aminimum thickness of the soft underlayer thatallows to achieve maximum magnitude of therecording fields and the trailing gradients of therecording fields (it is assumed that the length,lpole tip, of the pole tip is substantially greater thanthe width, wpole tip, of the pole tip (see Fig. 1)
tsoft underlayer MIN �1
2
MS pole tip
MS soft underlayerwpole tip: ð2Þ
The validity of the above equation can be shownusing arguments similar to the argument used inthe previous section and is a consequence of theconservation of the magnetic flux.
Fig. 3 shows the results of the boundary elementmodeling of the magnitude and the trailinggradient of the field generated a pole tip
Fig. 1. A schematic of the recording head pole tip/soft under-
layer combination. lpole tip is the dimension along the track and
wpole tip is the dimension across the track, i.e. wpole tip defines the
track width.
Fig. 2. Trailing fields from a single pole perpendicular write
head made out of FeAlN (4pMS=20 kG) for FeAlN and
Permalloy (4pMS=10 kG) soft underlayers.
D. Litvinov et al. / Journal of Magnetism and Magnetic Materials 232 (2001) 84–90 85
Information Storage: Basic and Applied
(wpole tip ¼ 0:4 mm; lpole tip ¼ 2mm) of a recordinghead and soft underlayers of different thicknesses.Both the soft underlayer and the pole tip areassumed to be made of FeAlN with 4pMS valuesof 20 kG. It can be seen that when the softunderlayer thickness becomes smaller than 0.2 mm,the performance of the recording system deterio-rates.
Media with almost identical recording layersand different soft underlayers were fabricated totest the validity of the theoretical results outlinedabove. The pole tip was made of Ni45Fe55 alloywith a 4pMS of 16 kG and the soft underlayer wasmade of FeAlN with a 4pMS of 20 kG. The trackwidth (� wpole tip) was 0.8 mm. According toEq. (2), tsoft underlayer MIN � 0:3 mm.
Fig. 4 shows the dependence of the saturationcurrent on the thickness of the soft underlayer. Forvalues of the soft underlayer thickness below0.3 mm, saturation current starts to increase, whichis an expected effect when the soft underlayer getspartially saturated.
Fig. 5 presents roll-off curves for three differentvalues of the soft underlayer thickness. In agree-ment with the theoretical consideration andconferring with the result in Fig. 4, the resolutionof the recording system starts to deteriorate whenthe soft underlayer thickness is less than 0.3 mm.
One of the consequences of the above discussionis the fact that to achieve an efficient write processit is not necessary to make a soft underlayerexcessively thick. Soft underlayers with thickness
values above 1 mm have been routinely reported inliterature [9]. For example, for a 100 Gbit/in2
recording density, utilizing a 4 : 1 bit aspect ratio(track width of 160 nm), if an FeAlN pole tip iswith a saturation magnetization of 20 kG andCoFe soft underlayer is with saturation magneti-zation of 24 kG, then a soft underlayer can be asthin as �70 nm. This makes the deposition of thesoft underlayer in conventional media depositiontools a fairly straightforward process.
4. Soft underlayer to the air-bearing surface of the
recording head distance
Apart from the clear advantages of having thesoft underlayer as close as possible to the
Fig. 4. Saturation current, ISAT, versus different soft underlayer
thicknesses. FeAlN soft underlayer and 0.8mm wide Ni45Fe55
pole tip were used.
Fig. 5. Roll-off curves for FeAlN soft underlayers of different
thicknesses. 0.8 mm wide Ni45Fe55 pole tip was used.
Fig. 3. The amplitude and the gradient of the trailing field
generated by a single pole perpendicular write head made of
FeAlN (4pMS=20 kG) and an FeAlN soft underlayer versus
the soft underlayer thickness.
D. Litvinov et al. / Journal of Magnetism and Magnetic Materials 232 (2001) 84–9086
Information Storage: Basic and Applied
recording head for better write field gradients [7],certain consideration related to the playbackperformance of the perpendicular recording sys-tem should be taken into account while designingperpendicular media with a soft underlayer. Fig. 6shows the dependence of the PW50 and theamplitude of the read sensitivity function of a30 nm thick reader as a function of the softunderlayer to the air-bearing surface of therecording head (SUL-to-ABS) distance.
Two effects can be seen in Fig. 6. As expected,the amplitude of the read sensitivity functiondecreases with the increase of the separationbetween the soft underlayer and the ABS, i.e. thecloser the soft underlayer to the ABS, the higherplayback signal can be expected. This trendsuggests to minimize the SUL-to-ABS spacing.
The other observation, which is critical tooptimizing the perpendicular system’s resolution,is that there exists a maximum in the value of thePW50 as the SUL-to-ABS distance is varied. FromFig. 6, it is obvious that a special care should beundertaken to correctly choose SUL-to-ABS dis-tance. It will be seen below from the experimentaldata, that this effect can substantially deterioratethe playback performance of a perpendicularrecording system, in which media with a softunderlayer is used, thus making it, potentially,even worse than the playback performance of aperpendicular recording system in which mediawithout a soft underlayer is used.
Fig. 7 compares roll-off curves for the mediawith and without a soft underlayer (close toidentical CoCr alloy based recording layers wereused). The dramatic effect of using perpendicularmedia with a soft underlayer in which the SUL-to-ABS distance is not optimized is clearly observed.
5. Anisotropy: micromagnetics of a soft underlayer
This section addresses the micromagnetics of asoft underlayer in the presence of stray fieldsgenerated by a recording layer. It is confirmed thatrelatively high anisotropy soft underlayer materi-als need to be utilized to optimize the performanceof the recording system [10]. As was elaborated inRef. [10] when the characteristic bit size in therecording layer becomes comparable to the char-acteristic in-plane length, d, in the soft underlayerfilm (often referred to as the domain wallthickness), the imaging ability of the soft under-layer deteriorates affecting the playback perfor-mance. This characteristic length d is defined bythe following relation:
d �ffiffiffiffiffiffiffiffiffiffiffiffiffiA=KU
p¼
ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiA=2pHKMS
p; ð3Þ
and its value along with other selected materialproperties are summarized in Table 1 for severalsoft underlayer candidates.
The inability of the soft underlayer to imagehigh frequency spacial variations of the magneti-zation in the recording layer causes an effectiveformation of a magnetically dead layer at the topof the soft underlayer [10] leading to an effective
Fig. 7. Roll-off curves for media with and without soft under-
layers.
Fig. 6. The dependence of the amplitude of the read sensitivity
function and of the PW50 on the SUL-to-ABS distance.
D. Litvinov et al. / Journal of Magnetism and Magnetic Materials 232 (2001) 84–90 87
Information Storage: Basic and Applied
increase of the SUL-to-ABS (air-bearing surface)distance. As shown in Fig. 6, the increase of theSUL-to-ABS distance will lead to accelerated roll-offs.
The experimental data presented below demon-strates the micromagnetic behavior of the softunderlayer leading to accelerated roll-offs, indeed,takes place and strongly depends on anisotropy ofthe soft underlayer material.
Media with close to identical recording layers(Co/Pd multilayer based) and different soft under-layers were spin-stand tested. The similarity ofrecording layer was confirmed via measuringM–H loops and microstructure of the films. Allthe measured parameters for the recording layerssuch as coercivity, saturation moment, nucleationfield, grain size, grain size distribution, etc. werewithin 1–2% error bar for all recording layers.
Fig. 8 shows the roll-off curves for three mediawith different soft underlayers. It is clear that theroll-off of the playback signal is significantly moredramatic if soft underlayer materials with lowerHK ’s are used.
It should be emphasized that the decrease in softunderlayer relative permeability from 2000 forPermalloy to 320 for high anisotropy Ni45Fe55
cannot explain the observed behavior. Boundaryelement based macromagnetic calculations showthat lowering relative permeability of the softunderlayer from infinity to �50 has virtually noeffect on the playback characteristic of the record-ing system [8].
In summary, as outlined above and in Ref. [10],the inability of a soft underlayer with lowmagnitude of HK to perfectly image the magneticbits in a recording layer at high linear density,may lead to distortions in the playback signal.The deteriorated imaging results in the loss of
resolution (increased PW50 and decreased ampli-tude of the read sensitivity function) and, subse-quently, in the deterioration of recording systemperformance.
6. Magnetic biasing
Among the technical challenges introduced bythe presence of a soft underlayer is the fact that itcontributes an additional source of noise [9]. A notproperly optimized soft underlayer material canintroduce a significant amount of noise into theplayback signal. The noise results from effectivecharges resulting from domain walls in the softunderlayer. A solution to this problem is magneticbiasing of the soft underlayer [10].
Fig. 9 shows a schematic of the experimentalsetup to study the effect of magnetic biasing of thesoft underlayer on the noise. The magnetic biasingwas achieved using two NdFeB permanent mag-nets placed in the vicinity of the media. The
Fig. 8. Roll-off curves for media with identical recording layers
and different soft underlayers.
Table 1
Selected material properties for Permalloy, FeAlN, Ni45Fe55 with high stress induced anisotropy, and FeCo where HK}anisotropy
field, 4pMS}saturation moment, A}exchange constant, LD}linear density corresponding to dimension d
Material HK (Oe) 4pMs (kG) A � 10�6 (erg/cm) d (nm) LD (kfci)
Ni81Fe19 5 10 1.0 112 226
FeAlN 15 20 1.7 60 420
Ni45Fe55 50 16 1.5 34 737
Fe65Co35 100 24 2.0 23 1100
D. Litvinov et al. / Journal of Magnetism and Magnetic Materials 232 (2001) 84–9088
Information Storage: Basic and Applied
placement of the magnets was such that it allowedto achieve complete saturation of the soft under-layer underneath the reader. Special care wasnecessary to arrange the magnets sufficiently farfrom the recording head (� 2cm away) in ordernot to affect the properties of the read element.
Fig. 10 shows the playback signals from the twomedia with the as-deposited non-biased (a) andmagnetically biased (b) soft underlayers. A sub-stantial level of noise attributed to the presence ofa large number of domain walls (confirmed bymagnetic force microscopy) in the soft underlayercan be seen in Fig. 10a. A drastic reduction of thenoise (by at least 10 dB) is clearly observed inFig. 10b where the soft underlayer is magneticallybiased.
The magnetic biasing saturates soft underlayerfilm forcing it into a pseudo-single domain state.Thus magnetic biasing effectively sweeps the
domain walls out of the soft underlayer material.This results in the elimination of the soft under-layer noise.
7. Summary
Various aspects of perpendicular media record-ing physics related to the use of a soft underlayerhave been discussed. It is shown that to fullyoptimize the performance of a recording system,the saturation moment of the soft underlayermaterial has to be equal to or higher than thesaturation moment of the recording pole tip.
The importance of a proper choice of SUL-to-ABS distance for the optimized performance of aperpendicular recording system, in which mediawith a soft underlayer is used is stressed. It isshown, that if that distance is not chosen properly,the resolution of the recording system becomessubstantially worse than the resolution of anequivalent perpendicular recording system opti-mized for the same areal densities, in which mediawithout a soft underlayer is used.
Furthermore, it is demonstrated that to improvethe playback efficiency, it is desirable to use highanisotropy soft underlayer materials. Also, thereduction of the soft underlayer noise via softunderlayer magnetic biasing has been demon-strated. Finally, it is shown, both experimentallyand theoretically, that for high density recordingsystems it is not necessary to make the softunderlayer excessively thick as it was proposedby various authors in earlier publications.
Fig. 10. Playback signal from two media with different soft underlayers. (a) Soft underlayer with a large number of stripe domains.
(The presence of stripe domains was confirmed using magnetic force microscopy.) (b) Biased soft underlayer with domain walls swept
out from the soft underlayer material.
Fig. 9. A schematic of experimental setup to magnetically bias
soft underlayer film.
D. Litvinov et al. / Journal of Magnetism and Magnetic Materials 232 (2001) 84–90 89
Information Storage: Basic and Applied
A proper choice of the hard and the easy axes ofthe soft underlayer material should be emphasized.This is applicable to both write and read processes.To avoid hysteretic behavior and associated non-linearities, such as the Barkhausen noise, it isstrongly desirable to have the easy axis of a softunderlayer material to be radially aligned. Whenthe easy axis is radially aligned, the magnetizationswitching during the write process will be accom-plished via magnetization rotation, while duringthe read process, the application of the radialbiasing field will efficiently wipe the domain wallsout of the soft underlayer.
In summary, although there are many simila-rities between perpendicular and conventionallongitudinal recording, the specifics native toperpendicular media should be clearly understoodand taken into account to design an efficientperpendicular recording system.
Acknowledgements
The authors would like to acknowledge the helpof their co-workers at Seagate Research for usefuldiscussions and for their help in some of theexperimental arrangements.
References
[1] S.H. Charap, IEEE Trans. Magn. 33 (1996) 978.
[2] S. Iwasaki, Y. Nakamura, IEEE Trans. Magn. 13 (1977)
1272.
[3] D.A. Thompson, J. Magn. Soc. Japan 21 (1997) 91.
[4] S. Khizroev, M. Kryder, Y. Ikeda, K. Rubin, P. Arnett, M.
Best, D. Thompson, IEEE Trans. Magn. 35 (1999) 2544.
[5] S. Khizroev, D. Litvinov, Perpendicular Systems above
100 Gbit/in2 density, Presented at Joint MMM-Intermag
2002 in San Antonio, TX, to be published IEEE Trans.
Magn.
[6] N.H. Bertram, M. Williams, IEEE Trans. Magn. 36 (1999)
4.
[7] S. Khizroev, R. Chomko, Y. Liu, K. Mountfield, M.
Kryder, D. Litvionv, Presented at INTERMAG 2000, CB-
02, April 11th, 2000, to be published.
[8] S. Khizroev, J. Bain, M. Kryder, IEEE Trans. Magn. 33
(1997) 2893.
[9] W. Cain, A. Payne, M. Baldwinson, R. Hempstead, IEEE
Trans. Magn. 32 (1996) 97.
[10] D. Litvinov, R. Chomko, L. Abelman, K. Ramstock, S.
Khizroev, IEEE Trans. Magn. 36 (2000) 2483.
[11] S. Khizroev, D. Litvinov, Perpendicular Recording,
Presented at Data Storage Systems Center of Carnegie
Melon University Seminar Series, 1999.
D. Litvinov et al. / Journal of Magnetism and Magnetic Materials 232 (2001) 84–9090
Information Storage: Basic and Applied