CORRELATION OF TEXTURE WITH MAGNETIC PROPERTIES OF THIN FILM STORAGE MEDIA

Thin coherent metal films of Co–Cr alloys on Si or glass substrates were used for high density 
magnetic storage media. They were made by sputtering or evaporation. Variation of the deposition 
parameters will change the micro structure especially the texture of the thin magnetic layers. We 
could relate the crystallographic data (rocking curves and pole figures) with the magnetic data 
(hysteresis loops). Correlations between deposition parameters and crystallographic and magnetic 
investigations will be discussed in detail.


INTRODUCTION
The requirement for large amounts of data and the trend towards ever smaller equipment for storing them requires storage media with ever higher storage densities.In the case of digital magnetic recording, this calls for smaller storage regions for the individual bits.
In the magnetic recording media normally used at present, a film of organic lacquer in which magnetic pigments (usually Fe or Cr oxides) are dispersed, is applied to a substrate.In these media the storage of information takes place in the so-called longitudinal mode.In this case the required increase in storage density calls for a considerable reduction in the film thickness of the magnetic recording film.This is achieved by going over from the particulate media to coherent thin-film media.These metallic films are produced by chemical or electrochemical deposition (plating), vapor deposition or sputtering.Mostly Co-TM alloys (TM Ni, Cr-ia.) are used which attain the necessary high magnetic flux and coercivity for stable storage.The metallic thin-film rigid disks being launched on the market attain storage densities which are four times higher than that of conventional ones.
Still higher storage densities can be achieved with vertical magnetic recording which is being developed throughout the world.Here, the objective is a ten-fold increase in the linear recording density, e.g. in the case of Co-Cr alloys, compared with the present conventional recording media.
The basic parameters for magnetic storage media, such as saturation magnetization, remanent magnetic flux or coercivity, are obtained from the magnetic hysteresis curve.This curve, measured with a vibrating sample magnetometer, represents the magnetic moment as a function of the field strength.
In the case of Co-TM alloys the magnetic easy direction (direction in which magnetization is reversed at storage,) coincides with the direction of the crystallographic c axis.Therefore it is essential to determine the structure of the storage films.This is achieved primarily with X-ray texture studies which determine the orientation of the crystallites and their distribution.
Since the orientation is already fixed during production in the case of coherent metallic thin-film media, the relationships between the production parameters, the magnetic properties and the structure are of appreciable significance for the efficacious adjustment of certain desired magnetic parameters.The saturation magnetization depends primarily on the chemical composition of the storage film.All the other static magnetic parameters, such as remanence and coercivity can be altered over a wide range by altering the morphology of the film while keeping the composition the same.We should like to use the example of sputtered 0.5 #m thick CoaoCr20 films for vertical recording to explain these relationships.

PREPARATION OF SAMPLES
The Co-Cr films were sputtered in the RF diode mode from a target having the composition CoaoCr20.The sputtering parameters were an Ar pressure of 5 x 10 -3 mbar, a substrate-target distance of 55 mm and a sputtering rate of 18 nm/min (target diameter 150mm, RF power 450 W).The back-pressure was better than 10-'mbar (turbo pump and liquid nitrogen trap).All the films which are described here have a thickness of 500nm and were deposited on glass substrates and monocrystalline Si wafers with (100) and (111) orientations.The substrates were cleaned by sputter etching.

MAGNETIC MEASUREMENTS
Figure 1 gives an impression how the hysteresis loops in the magnetic hard and magnetic easy direction of a sample with uniaxial magnetic anisotropy looks like, mag.easy direction ' ag.hard dlrectlon Figure 1 Hysteresis loops of magnetic uniaxial sample (ideal case).ligare Z Hysteresis loops of a magnetic uniaxial layer with perpendicular orientation.(a) Hysteresis loop perpendicular to layer (ideal case).(b) Hysteresis loop parallel to the layer (ideal case).(c) Hysteresis loops perpendicular () and parallel (---) to the plane of a real Co-Cr-layer on glass. in the ideal case.In the easy direction one gets a rectangular shaped curve.The area of this curve is determined by the coercivity Hc and the saturation magnetization Ms.The hysteresis loop in the magnetic hard direction has an area of zero.So also the relative remanence in the film plane (mr(para)= Mr must be zero.
Figure 2 shows the deviation of the actual curves from the ideal curves.An ideal oriented perpendicular film of uniaxial Co-Cr crystallites has a sheared hysteresis loop for the magnetic easy direction.This is a consequence of the film geometry which is the source of a large demagnetizing field perpendicular to the film plane.Typical measured hysteresis loops perpendicular and parallel to the Co-Cr film plane are contrasted with these ideal curves.
The hysteresis loop for the magnetic easy direction is quite similar to the ideal case.Since the back-shearing is not without problems, we use only Hc(perpen) from that curve to characterize the sample.On the other hand the hysteresis loop for the magnetic hard direction shows a large deviation from the ideal case.There is some remanent magnetization (mr(para)>0), indicating magnetic moments which are not perpendicular to the film plane.So mr(para) seems a good parameter for characterizing the magnetic perpendicular orientation of the Co-Cr film. 1,2RAY DIFFRACTION Co-Cr films with Cr contents below 30 at.% 3 crystallize, like l'-Co, 4 with a hexagonal closest packing (hcp).They form substitution mixed crystals whose crystallographic data do not differ substantially from those of the a-Co.Figure 3 shows a diffraction diagram of 0.5/m thick Co-Cr film on Si.In the 20 range of 35-55 the Bragg reflexions (100), ( 002) and (101) of Co-Cr are located.The planes at which diffraction takes place are shown in Figure 4.
Usually preliminary X-ray diagrams in back-reflexion are taken of thin films.
The primary and secondary beam emerge on the same side of the sample and are gragg-angte 28 Figure 3 X-ray diffraction diagram of a Co-Cr film on a Si substrat with a high perpendicular orientation of the c axis.
in the same plane as the normal to the sample surface (Figure 5).The sample is rotated during diffraction by 0 with respect to the primary beam, the detector by 20.Since only a strong (002) reflexion occurs in the case of Figure 3, this means that, in almost all the crystallites, the c axis is perpendicular oriented to the sample surface.The precise orientation of the crystallites and the distribution of their orientation is found from X-ray texture measurements using a texture diffractometer.
( In this apparatus (Figure 6), the normal to the sample can be tilted through the angle ;t with respect to the diffraction plane which is formed by the primary and the secondary beam.In addition, a rotation through ff around the normal to the sample is possible.Sample rotations during diffraction at a reflexion were represented as a pole figure in a stereographic projection (Figure 7).The reflexion intensities during the diffraction are shown as contour lines in this pole figure.Figure 8 shows the (002) pole figure of a typical Co-Cr sample.In this case, the maximum in the diffraction intensity is situated at ;t 90 at the center of the pole figure, i.e. the c axis orientation is mainly perpendicular to the film surface.This distribution of the crystallite orientations is sharp, i.e. the orienta- tions of the most of the crystallites differ only slightly from the perpendicular direction.Besides that the distribution is symmetrical, i.e. there is no preferred position in the rotation around the normal to the sample.We find this case of a sharp fiber texture in the most of our Co-Cr samples.The sharpness of the distribution of the orientation characterizes the quality of perpendicular crystallographic orientation in Co-Cr films.For the purpose of a precise measurement, the sample is tilted through 0 (with small apertures) in the diffraction plane (with fixed g and 9 values) and the diffraction intensity is recorded.The result obtained is a so-called rocking curve (Figure 9).In this example of a 0.5/m thick Co-Cr sample on a glass substrate, a full width at half maximum (FWHM) is obtained of A0 3.2 , indicating a very narrow distribu- tion of the c axis orientation.

EFFECTS DUE TO SUBSTRATE CLEANING BY SPUTFER ETCHING AND TO SUBLAYERS
To achieve a good perpendicular orientation of the Co-Cr films, the substrates have to be cleaned.This can be done very effectively by RF sputter etching.With a power density of 0.2 W/cm 2 3 and an Ar pressure of 5 x 10-mbar, the etching rate of glass and Si with our equipment is about 1.5 nm/min.The hysteresis loops (in plane and perpendicular) of the 0.5/zm thick Co-Cr films on glass and Si substrates indicate a strong increase of perpendicular orientation as a conse- quence of cleaning (Figure 10).This can be seen particularly well from the mr(para) values (Figure 11).These values decrease in the case of Si and in the case of glass substrates (but by a different amount) with increasing etching time.
Especially for etched Si substrates the hysteresis loop is very near the ideal case shown in Figure 1.
Optimum surface conditions for growing a perpendicular film are obtained only after an etching time of 15 min.This means almost 20 nm of substrate had to be removed.ESCA measurements revealed only a 2 nm thick SiO film on the uncleaned Si substrates, indicating a further effect in optimizing the surface through etching, besides the removal of surface contamination and oxide layers.Perhaps the change in the surface roughness during sputter etching plays a part.
A delay of 15 min was allowed after sputter etching before sputtering on Co-Cr in order to bring substrates back to room temperature.Figare 10 Hysteresis loops (perpendicular (--) and parallel (---)) of Co-Cr films on glass and Si substrates, with and without etching.The fact that the substrates affect the properties of the Co-Cr films is shown by the different mr(para) values of the films on glass and Si (Figure 11).However, there is no hint of epitaxial growth in this case since the values of the films deposited on Si(100)-and Si(111)-textured wafers are equal.
The substrates may be altered by depositing sublayers. 5Thus, it is known that Ge sublayers on plastic films increase the perpendicular orientation of the Co-Cr films, while Bi sublayers reduce it. 5We investigated the effect of Cr and Ti sublayers of various thickness on glass and Si substrates.They were sputtered on with the same sputtering parameters as the Co-Cr films.
It can be seen from Figure 12 that Ti sublayers have only a weak influence on mr(para) of the Co-Cr films.For dTi > 50 nm one observes a slight increase of m,(para).
In the case of Cr sublayers (Figure 12), m,(para) of the Co-Cr films increases very markedly to 0.4 even for very thin (5 nm) films and then drops slightly to 0.32 at 200nm.The body centered cubic (bcc) structure of Cr 4 is shown in Figure 13.Epitaxial considerations reveal that the atoms of the Co(ll0) plane fit particularly well on to the atoms of the Cr(200) plane (Figure 14).We know from our studies of the Cr films that the Cr films do not have any preferred texture.Thus, the Co-Cr films on a Cr sublayer should have a random orientation of the c axis which yields high m(para) values.This effect of Cr sublayers is made use of in preparing longitudinal magnetic media with Cr sublayers and Co alloys as the magnetic film.Figare 12 mr(para) of 0.5 pm thick Co-Cr films as a function of the thickness of sublayers of dcr or dTt.Used substrates are glass and Si.
Fixture 13 Unit cell of bee Cr.
The Hc values of the Co-Cr magnetic films are also affected by the sublayers.
Without a sublayer, the coercivity in the plane Hc(para) is low (25 kA/m), while the coercivity perpendicular to the plane Hc(perpen) is high (90 kA/m) (Figure 15).The latter is nearly independent from dcr whereas H(para) increases markedly to 50 kA/m for dcr 50 nm and remains constant up to dcr 200 nm.lle 5 Coecivity H of 0.5/m thick Co-C films as a function of the thickness dc of the C sublayer.
For Ti sublayers the situation is quite different (Figure 16).Hc(perpen) drops from 90 kA/m to 50 kA/m for dTi 5Onto.With increasing thickness of the sublayer, Hc(perpen) rises to 145kA/m at dTi=200nm.An increase of H(perpen) with increasing di was also published in Refs 6 and 7.The reason shall be a morphological change towards a columnar Co-Cr layer, using a Ti sublayer.Fibre 16 Coercivity He of 0.5 ttm thick Co-Cr films as a function of the thickness dTi of the Ti sublayer.particularly in view of the mr(para) values.Figure 21 shows both the hysteresis loops and also the (002) pole figure of Co-Cr on Si(111) substrates at room temperature and for a substrate temperature of 250C during sputtering.An extremely sharp perpendicular orientation of the c axis is found at room temperature.At Tp 250C, the orientation distribution of the c axis is very wide and two preferred orientations are found at an X angle of 60.This deviation from the perpendicular orientation of the c axis begins at Tp 230C.Although the c axis of the crystallites have a preferred perpendicular orientation up to 220C, the distribution of this orientation becomes increasingly wider with higher temperatures, as the full width at half maximum (FWHM) of the rocking curves of the Co-Cr film show (Figure 22).On Si substrates, this widening from 1 to 2 is very small, whereas, in the case of glass substrates, it increases from a value of 3 at  Figare 24 (a) m,(para) of Co-Cr films on glass and Si as a function the full width at half maximum (FWHM) of the rocking curves.(Samples were made at different Tp; (---) represents theoretical values). (b) Connection between c axis distribution (magnetic moment parallel to c axis) and magnetic moment parallel to the film plane (M(para)).(c) The two possibilities of the orientation of Co-Cr crystallites at high Tp.
FWHM values.On Si, m,(para) increases for higher substrate temperatures although FWHM remains almost constant.Two possible orientations of the Co-Cr crystallites (Figure 24c) are used for the purpose of interpretation: a) crystallites whose c axis are preferentially perpendicular to the plane of the film, b) crystallites whose c axis are in the plane of the film.Crystallites of type a) contribute to the parallel remanent magnetization M,(para) as a result of their projection, i.e. the rocking curve (Eq. 1) has to be weighted E. HJDICKE, A. WERNER AND H. HIBST with sin(0) (Figure 24b).The integration over 0(A0 in radian measure) yields: m,(para) 0.35A0 (4) This relationship is entered in Figure 24a as dotted curve and shows that the increase in m,(para) which is observed in the case of Co-Cr on glass is the result of the broader orientation distribution of the c axis at higher substrate temperatures.
Crystallites of type b) contribute fully to M,(para).This probably explains the increase in m,(para) at higher substrate temperatures in the case Co-Cr on Si.The number of crystallites with a c axis orientation parallel to the sample surface probably increases with increasing substrate temperature.From the literature it is known that the boundary layer of the substrate, in particular, is built up of such misoriented Co-Cr crystallites.The thickness of this layer probably increases at elevated temperature.

CONCLUSION
The structure of the magnetic film should be elucidated for the purpose of systematically developing coherent magnetic storage media with high storage density and certain magnetic properties.In particular, the parameters which are important for storage, such as the coercivity or the remanent magnetic flux, may be varied within a wide range by making use of the structure of the magnetic film.
By means of examples it has been shown how the morphology and the magnetic properties are altered by sputter etching, by sublayers and by substrate temperature in the case of perpendicularly oriented sputtered Co-Cr films on glass and Si substrates.A main part plays the orientations of the crystallites which is particularly well characterized by means of X-ray texture investigations.

Figure 8 (Figure 9
Figure 8 (002) pole figure of a Co-Cr film on glass.
time[rain]   Hgare 11 In plane relative magnetization m,(para) of Co-Cr films on glass and Si as a function of etching time.

Figure 15 Figare 19 Figure 20
Figure15Co-Cr films on Si substrates (a) m,(para) as a function of Tap (b) H as a function of Tap.
Figare 22 Full width at half maximum (FWHM) of Co-Cr films on Si(ll1) and glass substrates as a function of