Buschow K.H.J. (Ed.) Concise Encyclopedia of Magnetic and Superconducting Materials

Подождите немного. Документ загружается.

increases the processing costs. An alternative route is

the production of net-shape parts by die-pressing. The

alloy powder is ﬁlled into the cavity of a die, aligned

by an axial magnetic ﬁeld and compacted (see Fig. 1,

right-hand side). Since the easy-axes of the powder-

particles are aligned parallel to the pressing direction,

the alignment is disturbed during the compaction.

Hence, the remanent polarization of axial ﬁeld die-

pressed magnets is in general about 8% lower in

comparison to isostatically pressed magnet blocks.

By axial ﬁeld die pressing a large variety of dimen-

sions and shapes can be manufactured economically.

For instance, net-shape Nd–Fe–B magnets for motor

applications achieve a remanent polarization of 1.08 T

and a coercivity H

(20 1C) of 28.7 kA cm

1

(36 kOe)

and a maximum energy density of 225 kJ m

3

(28 M GOe). At 150 1C, the H

of such magnets

amount to 12 kA cm

1

(15 kOe). The TCs of the re-

manent polarization and of the coercivity amount to

0.08% K

1

and 0.51% K

1

in the temperature

range of 20–100 1C. At room temperature such

Nd–Fe–B magnet grades exceed sintered Sm

(Co,

Cu, Fe, Zr)

magnets, but at 200 1C the temperature

stability of Sm

(Co, Cu, Fe, Zr)

magnets is superior.

2.3 Transverse Field Die-pressed (TP) Near

Net-shape Parts

Another compaction technology is pressing of parts

in a transverse magnetic ﬁeld (Fig. 1, right-hand

side). In this case the alignment of the powder par-

ticles is not disturbed so strongly during compaction

resulting in a remanent polarization similar to iso-

statically pressed magnet blocks. The alignment co-

efﬁcients range from 94% to 98% for rectangular

blocks and from 93% to 97% for near net-shape

parts.

By transverse ﬁeld die pressing mainly rectangular

blocks or near net-shape parts can be manufactured.

After sintering, the blocks are cut into thin magnet

plates according to customer speciﬁcations. However,

near net-shape parts will need some contour grinding

before cutting. Hence the machining costs increase in

comparison to axial ﬁeld die-pressed magnets.

2.4 Rubber Isostatic Pressing (RIP)

Near net-shape magnet blocks with an excellent tex-

ture, i.e., an alignment coefﬁcient similar to that of

isostatically pressed blocks, are now manufactured by

rubber isostatic pressing (Sagawa et al. 2000). The

alignment of the precompacted alloy-powder parti-

cles in an elastic rubber mold is performed by a set of

d.c. and a.c. ﬁeld pulses with a maximum ﬁeld

strength of 24 kA cm

1

(30 kOe) for instance. After-

wards the aligned alloy powder is compacted by axial

pressing. But the axial pressing forces are diverted

homogeneously by the elastic rubber mold, so that

the compaction occurs almost isostatically.

Figure 2

Demagnetization curves J(H) and B(H) at different temperatures of sintered Nd–Fe–B magnets, grade Vacodym

722 HR, with a maximum energy density of 420 kJ m

3

(53 M GOe), manufactured by cold isostatic or rubber isostatic

pressing, respectively.

880

Magnets: Sintered

Optimization of the alloy composition, alloy pow-

der milling, aligning and compaction technology, and

sintering and annealing treatments now enables the

commercial production of near net shape Nd–Fe–B

parts with a maximum energy density of 420 kJ m

3

(53 M GOe). Hence, sintered Nd–Fe–B magnets ex-

ploit the magnetic properties of the Nd

B com-

pound, J

¼1.6 T, theoretical (BH)

¼509 kJ m

3

(63 M GOe), to more than 80% and have gained a

similar maturity as Sm–Co magnets.

3. Samarium–Cobalt Magnets

Although the temperature stability and the corrosion

behavior of sintered Nd–Fe–B magnets can be

improved by increasing the coercivity and by small

additions of aluminum, copper, cobalt, etc., the tem-

perature and corrosion stability of sintered Sm

(Co,

Cu, Fe, Zr)

magnets remains superior at tempera-

tures above 200 1C. The demagnetization curves J(H)

at different temperatures of isostatically pressed mag-

nets demonstrate, that at 20 1C the remanent polar-

ization and the coercivity amount to 1.11 T and

25.3 kA cm

1

(31.6 kOe) resulting in a maximum en-

ergy density of 225 kJ m

3

(28 M GOe, Fig. 3). At

200 1C, Sm

(Co, Cu, Fe, Zr)

magnets still achieve a

coercivity H

(200 1C) of 14 kA cm

1

(17.5 kOe).

Maximum continuous operating temperatures up to

350 1C are possible for such magnets. Recently de-

veloped Sm

(Co, Cu, Fe, Zr)

magnets even achieve

a straight demagnetization curve B(H) at 550 1C

(Walmer et al. 2000, see Magnets: High-temperature).

4. Concluding Remarks

Powder metallurgy enables the production of sintered

net-shape RE–TM magnets for applications in mo-

tors, electronic devices, speakers, microphones, a

large variety of sensors, bearings, couplings, NMR

analysis equipment, MRI tomography, beam guid-

ance devices, magnetic separation, switches, and re-

lays (see Magnetic Materials: Domestic Applications).

In order to decrease the processing costs and im-

prove the magnetic performance, the powder process-

ing, aligning and compaction technologies, sintering

and annealing treatments, as well as machining and

coating processes need continuous optimization. Sin-

tered Nd–Fe–B magnets with a maximum energy

density of 420 kJ m

3

(53 M GOe) have now come

into commercial production, while by laboratory

facilities a maximum energy density of 444 kJ m

3

(55 M GOe) has been achieved (Kaneko 2000). In the

year 2000, in total about 16000 t of sintered Nd–Fe–B

magnets and 1000 t of sintered Sm–Co magnets were

produced (Luo 2000).

Besides the production of anisotropic RE–TM

magnets by sintering, some alternative processing

routes have been developed and evaluated. These in-

clude: hot-pressing and hot-deformation of rapidly

solidiﬁed ﬂakes (Croat et al. 1984, Panchanathan

1995), hot-backward extrusion of radially aligned

ring-magnets (Yoshikawa et al. 1994), hot-rolling

(Shimoda et al. 1989, Yuri and Ohki 1994, see Tex-

tured Magnets: Deformation-induced), and vacuum

plasma spraying (Overfelt 1986, Rieger et al. 2000).

But the sintering technologies still meet the technical

requirements very efﬁciently and economically.

Figure 3

Demagnetization curves J(H) at different temperatures of sintered Sm

(Co, Cu, Fe, Zr)

magnets, grade Vacomax

225 HR, with a maximum energy density of 225 kJ m

3

(28 M GOe), manufactured by cold isostatic pressing.

881

Magnets: Sintered

See also: Hard Magnetic Materials: Basic Principles

of; Magnetic Materials: Hard; Magnets: Bonded

Permanent Magnets

Bibliography

Croat J J, Herbst J F, Lee R W, Pinkerton F E 1984 Pr–Fe and

Nd–Fe-based materials: a new class of high-performance

permanent magnets. J. Appl. Phys. 55, 2078–82

Harris R, McGuiness P, Short C, Book D, Gutﬂeisch O, Fujita

A, Verdier M, Nagel H 1994 The use of hydrogen in the

processing and in the characterization of Nd–Fe–B magnets

and alloys: an update. Trans. Mater. Res. Soc. Jpn. 14B,

969–74

Herget C 1985 Metallurgical ways to NdFeB alloy—permanent

magnets from Co-reduced NdFeB. In: Strnad K J (ed.) Proc.

8th Int. Workshop on RE Magnets and their Applications.

University of Dayton, OH, pp. 407–22

Hirose Y, Hasegawa H, Sasaki S 1998 Microstructure of strip

cast alloys for high performance NdFeB magnets. In: Schultz

L, Mu

ller K-H (eds.) Proc. 15th Int. Workshop on RE Mag-

nets and their Applications. Dresden, Germany. Werkstoff-

Informationsgesellschaft, Frankfurt, Germany, pp. 77–86

Kaneko Y 2000 Rare-earth magnets with high energy products.

In: Kaneko H, Homma M, Okada M (eds.) Proc. 16th Int.

Workshop on RE Magnets and their Applications. Sendai,

Japan. Japan Institute of Metals, Sendai, Japan, pp. 83–98

Luo Y 2000 Further development of rare earth magnet industry

in China. In: Kaneko H, Homma M, Okada M (eds.) Proc.

16th Int. Workshop on RE Magnets and their Applications.

Sendai, Japan. Japan Institute of Metals, Sendai, Japan, pp.

31–40

Overfelt R A, Anderson C D, Flanagan W F 1986 Plasma

sprayed Fe

permanent magnets. Appl. Phys. Lett.

49, 1799–801

Panchanathan V 1995 Magnequench magnets status overview.

J. Mater. Eng. Perf. 4, 423–9

Rieger G, Wecker J, Rodewald W, Sattler W, Bach F W, Duda

T, Unterberg W 2000 Nd–Fe–B permanent magnets (thick

ﬁlms) produced by a vacuum-plasma-spraying process. J.

Appl. Phys. 87, 5329–31

Rodewald W, Blank R, Wall B, Reppel G W, Zilg H D 2000

Production of sintered Nd–Fe–B magnets with a maximum

energy density of 53 MGOe. In: Kaneko H, Homma M, Ok-

ada M (eds.) Proc. 16th Int. Workshop on RE Magnets and

their Applications. Sendai, Japan. Japan Institute of Metals,

Sendai, Japan, pp. 119–26

Sagawa M, Nagata H, Watanabe T, Itani O 2000 Rubber

isostatic pressing (RIP) of powders for magnets and other

materials. Mater. Des. 21, 243–9

Shimoda T, Akioka K, Kobayashi O, Yamagami T, Ohki T,

Miyagawa M 1989 Hot-working behavior of cast Pr–Fe–B

magnets. IEEE Trans. Magn. 25, 4099–104

Yoshikawa N, Yamada H, Iwasaki Y, Nagata K, Kasai Y 1994

Magnetic properties and multi-pole magnetizability of a huge

radially oriented Nd–Fe–B ring magnet for EV driving mo-

tors. In: Manwaring C A, Jones D G E, Williams A J, Harris

J R (eds.) Proc. 13th Int. Workshop on RE Magnets and their

Applications. University of Birmingham, UK, pp. 635–44

Yuri T, Ohki T 1994 Crystal alignment in Pr–Fe–B hot-rolled

magnets. In: Manwaring C A, Jones D G E, Williams A J,

Harris J R (eds.) Proc. 13th Int. Workshop on RE Magnets

and their Applications. University of Birmingham, UK,

pp. 645–54

Walmer M S, Chen C H, Walmer M H, Liu S, Kuhl E 2000

Magnetic hardening, thermal stability and microstructures

for a new class of Sm-TM high temperature magnets. In:

Kaneko H, Homma M, Okada M (eds.) Proc. 16th Int.

Workshop on RE Magnets and their Applications. Sendai,

Japan. Japan Institute of Metals, Sendai, Japan, pp. 119–26

W. Rodewald

Vacuumschmelze GmbH, Hanau, Germany

Metal Evaporated Tape

The ﬁrst generation of commercial Metal-Evaporated

(ME) tapes was introduced in 1989 for the Hi-8 video

system (Chiba et al. 1989, Hokkyo et al. 1993) and

consisted of an oblique-deposited Co–Ni–O thin ﬁlm

tape. This was followed by the scoopman digital re-

corder with a stamp-sized cassette (Hinemo and

Katahu 1994). In 1993, a new digital video system

was proposed (Digital Video Cassette (DVC)) (Yam-

amitsu et al. 1995). This system uses an improved ME

tape with a 4 dB higher C/N (Futamoto and Honda

1994). Besides Co–Ni–O-deposited tapes, Co–O ﬁlms

with oblique anisotropy appeared on the market as

the tape for DVC (Kawana et al. 1995). It is also

planned that this type is used for large-capacity data

storage tape for computer systems. More than

1 Gbin

2

is being developed (Ishida et al. 2000) in a

helical scanning tape storage system based on Co–O

thin ﬁlm material.

Important in system design is the trade-off between

the amount of data reduction, playing time, and cas-

sette size. For both the Hi8 and DVC systems, equiv-

alent metal particle tapes (MP tapes) have been

developed which meet these requirements. There is an

ongoing competition between MP and ME tapes. In

general, ME tape gives a promising signal-to-noise

ratio and is a potential candidate for future applica-

tions if much smaller tracks will be written (Draaisma

et al. 1999).

Another advantage of ME tape is that the record-

ing layer thickness can be reduced below 100 nm

without changing the magnetic properties too much.

Moreover, the properties can be tailored with the

deposition parameters of the coating machine. ME

tape also has a low noise because of the decreased

ﬁlm thickness (Kagawa et al. 1997). The need for

extremely small sizes combined with long playing

times requires a tape with a very high signal-to-noise

ratio and very thin magnetic and total tape thickness.

In 1998, a very important paper was published dis-

cussing the key issues in the design of magnetic tapes

(Richards and Charrock 1998).

Although the paper is not specifically focused on

ME tape, most of the issues discussed are directly

applicable to it. In the next paragraphs, only the basic

882

Metal Evaporated Tape

properties of the oblique-deposited magnetic layer,

the related microstructure, and the magnetic behavior

of ME tape will be discussed. Apart from the devel-

opment of the magnetic layer, numerous other as-

pects have also to be considered such as tape

mechanics, tape transport, friction and wear, head

design, long term stability, etc.

In Fig. 1, a cross-section of an advanced ME tape

is shown schematically.

In Fig. 1, two magnetic layers are presented. In

recent years ME tape manufacturers have experi-

mented with several magnetic coating layers. One of

the reasons is an increased coercivity (by magnetic

coupling) of so-called double layers. In Ashton and

Thompson (1998), several possibilities are shown

which can enhance the magnetic properties. Up to

four magnetic sublayers were even reported in

Kawana et al. (1995).

1. Oblique Evaporation and Columnar Separation

When depositing a thin magnetic layer, by either

sputtering or evaporation, a columnar morphology is

usually observed (Smith 1995, Abelmann and Lodder

1997). Since these columns are in close contact with

each other, a strong exchange coupling between ad-

jacent columns is present resulting in an almost in-

stant switching behavior of the ﬁlm. This is highly

unfavorable for a magnetic recording medium since it

prevents writing a stable bit structure with sharp

transitions in such a medium, so inducing a columnar

separation is necessary to enable the recording of in-

formation in a thin magnetic ﬁlm. One possibility for

obtaining this columnar separation is to make use of

the so-called shadowing effect as schematically pre-

sented in Fig. 2.

From Fig. 2(a) it is observed that due to the height

of the initial nuclei, a ‘‘shadow’’ is thrown behind

these nuclei and therefore the rest of the vapor ﬂux

can not reach the area directly behind them. Since

this ﬂux is deposited under an angle with respect to

the normal of the substrate, subsequent stages of ﬁlm

growth will result in a tilted columnar morphology,

which is schematically shown in Fig. 2(b). Here the

columns make an angle b with respect to the normal

of the substrate. The relation between both angles is

very much determined by the process conditions, i.e.,

substrate temperature, base pressure, energy of the

arriving atoms, etc. (see, e.g., Abelmann and Lodder

1997). In principle, both sputtering (Bijker et al.

1997) as well as evaporation can be used in exploiting

the effect of shadowing.

In general, atoms tend to migrate across the

substrate surface once they have become adatoms.

Therefore, the oblique incidence effect might be over-

come by the transfer of thermal energy of the subst-

rate to these adatoms, once the substrate temperature

becomes high. Hence, low substrate temperatures are

beneﬁcial for this type of process. The thin-ﬁlm mor-

phology of an oblique-deposited magnetic layer with

magnetically uncoupled grains can lead to relatively

high coercivities. It is known from literature (Feuer-

stein et al. 1985) that the deposited materials iron,

cobalt, and nickel show a remarkable increase in co-

ercivity as a function of the angle of incidence (a in

Fig. 2). Such an example is shown in Fig. 3. Similar

results were reported in Kita et al. (1981) for

62.5

37.5

and Co

as well.

The sharp increase in longitudinal coercivity at a

high angle of incidence is believed to arise from

a strongly reduced exchange coupling between the

columns.

2. Manufacturing Process

To produce magnetic tapes, a mass-production equiv-

alent of the mentioned static-angle experiments is

schematically shown in Fig. 4.

In this ﬁgure, an unwinding and winding reel are

shown, which guide the base ﬁlm over the evaporation

source. Therefore, the angle a (from Fig. 2) is changed

from 901 to 601 (as indicated in Fig. 4). Thus, instead

of making a ﬁxed angle with the atom ﬂux with re-

spect to the substrate, a continuously varying angle of

incidence (CVI) is introduced. The use of evaporation

has led to the general term metal-evaporated tape or

ME tape. The CVI method results in the typical mor-

phology for ME tape, which is shown in Fig. 5.

Figure 1

Schematic cross section of an ME tape

Figure 2

Physical principle of the shadowing effect. (a) shows the

actual shadowing of the deposited atoms impinging

under an angle a with the substrate. In (b) the columnar

inclination angle b is shown.

883

Metal Evaporated Tape

Comparing the coating thickness and coercivity

of ME and metal particle (MP) tapes, it can be

concluded that thinner coatings can be achieved with

usually better coercivities. Hence, according to

the Williams and Compstock (1971) model, smaller

values of the transition width (between two written

bits) can be achieved. However, recently the partic-

ulate coatings have also been improved by developing

better particles, which can be better aligned at smaller

coating thickness, which have eventually resulted in

comparable recording properties (see, e.g., Lalbaha-

doersing et al. 1997, Richter and Veitch 1995). Dif-

ferences in easy-axis (MP (in-plane) and ME (tilted),

and reversal mechanism of MP and ME tape might

correspond with a smaller transition length, which

results in a better overwrite behavior for ME tape

(Lalbahadoersing et al. 1997). From the above, the

principal process of making ME tapes was made

clear.

Many publications can be found about this subject

and more detailed information of the subject can be

acquired from Luitjens et al. (1996), Onodera et al.

(1996), Kunieda et al. (1985), Richter (1993a). Apart

from the development of the magnetic thin ﬁlm or

coating, numerous other aspects have also to be con-

sidered. These involve issues such as tape mechanics,

tape transport, friction and wear, head design, long

term stability, etc. When comparing ME tape with

the more conventional particulate tape (e.g., Cr

and Fe

), the manufacturing process is the most

striking point of difference. As we explained above,

ME tape is produced via an evaporation process. In

Feuerstein and Mayer (1984), Feuerstein et al. (1985),

and Wright (1987), several aspects of the vacuum

deposition process are shown. In Fig. 6, a cross-sec-

tion of a typical deposition machine is shown (taken

from Feuerstein and Mayer 1984).

In Fig. 6, the base ﬁlm is unwound from reel (a) via

an intermediate coating station (b), towards the ob-

lique deposition zone (c). Here the electron beam (e)

Figure 4

Principle of ME tape mass production.

Figure 3

Effect of the angle of incidence of the vapor ﬂux of the

three ferromagnetic elements (after Feuerstein et al.

1985).

Figure 5

Scanning electron microscope image of an experimental

ME tape showing the banana-shaped columnar

structure (after Bijker 1998).

884

Metal Evaporated Tape

evaporates the magnetic material in the crucible (f).

Finally, the tape is wound onto a reel (d). Further-

more, a separation between the unwinding zone and

the deposition zone is necessary to sustain a reason-

able pressure in the deposition zone. The separation

is indicated in the ﬁgure by the grey area. Some sim-

ple calculations can be undertaken to calculate the

necessary web speed and the deposition rate. The

base material is typically supplied with a width of

1.20 m and a length of 10 km. For these base ﬁlm

lengths, a total coating time of around a few hours is

applied. In total, about one hour is needed for pump-

ing, accelerating the web to its full speed, heating up

the evaporation source, etc. So, in approximately

three hours, the total web length should be coated,

which results in a web speed of around 1 ms

1

60 m min

1

. Together with the drum diameter of one

meter, and an opening angle of 301, the necessary

deposition rate can be calculated to be B0.5 mms

1

In 1999, Maezawa et al. (1999) published data of

Co–O tape prepared at deposition rates of 1 mm/s.

Such deposition rates require e-beam evaporation

sources with astronomically large power consumption.

Usually two e-beam guns are installed in a coater

with a web width of 1.20 m (see Leybold 1992). These

would be of around 0.5 MW each. The crucible also

has a width of around 1.20 m, usually made of a

ceramic. These large deposition rates require

the coating material to be constantly replenished. In

the early designs, a rod feed system was designed, but

in later designs slugs of coating material were fed via

a vibration system into the melt. This allows the use

of a crucible, which can be parked in a separate vac-

uum chamber during charging of the vacuum ma-

chine. This eliminates a complete cooling down of the

crucible and saves time. To obtain a homogeneous

melt, the e-beams are swept over the melt with a fre-

quency of about 400 Hz. Winding a thin polymer

substrate in vacuum conditions is not trivial, let alone

in a bi-directional manner. Base ﬁlms do not neces-

sarily have similar winding characteristics in air com-

pared to vacuum conditions. As modern substrates

tend to become thinner and thinner, more sophisti-

cated winding systems are needed. Usually the drum

and the winding and unwinding reels are controlled

by measuring the torque exerted on the base ﬁlm. To

prevent wrinkling of the base ﬁlm a slightly V-shaped

roll, or capstan, is present in the winding system.

Furthermore, when such large rolls of base ﬁlm are

unwound in a vacuum system, they tend to release

considerable amounts of water vapor.

3. Microstructure and Magnetic Structure

From Fig. 3 it can be concluded that the necessary

high coercivity can only be obtained by oblique dep-

osition if the angle of incidence is very high. Conse-

quently, that is also the area in which the deposition

process becomes very inefﬁcient. To overcome this

problem depositions are made at smaller angles

but with the addition of oxygen during deposition.

The result is a very complicated microstructure (see

Fig. 7) but with an increase of the coercivity.

The incorporation of the oxygen is hard to under-

stand because companies do have their own specific

processes, which are not published. However, in, e.g.,

Kunieda et al. (1985) and Hokkyo et al. (1993), the

general trends are indicated: an increase in oxygen

Figure 6

Schematic cross section of a production roll coater. See

text for explanation of the items.

Figure 7

Banana-shaped columnar morphology of ME tape of

thickness d. Two possible microstructures are also

drawn.

885

Metal Evaporated Tape

ﬂow, whilst keeping the deposition rate constant, will

lead to an increase in longitudinal coercivity (H

) and

a decrease in apparent saturation magnetization

). From Gau et al. (1986), a local optimum in

as a function of the oxygen ﬂow can be observed.

A similar behavior has been observed by Bijker et al.

(1999).

As can be seen in Fig. 7, apart from an oblique

banana-shaped columnar morphology, which is due

to the specific deposition geometry, the columns can

either exist of ferromagnetic grains in close contact

with the nonmetallic grains, or apart from the pres-

ence of nonmetallic grains, the ferromagnetic grains

have a small oxide skin around them. The oxidation

process is most likely diffusion limited (Mujumdar

et al. 1986), and therefore could explain the correla-

tion between the used deposition rate and the oxygen

content in the upper part of the ME tapes which is

accompanied by the larger grains sizes in this part of

the ﬁlm as well. This is also reﬂected by the H

of the

ﬁlms, which become larger with increasing oxide-shell

thickness, thereby reducing the magnetic coupling

between the ferromagnetic grains (Bijker 1998). In

general, it is not straightforward to relate the micro-

structure of ME tapes to the switching behavior of

the layer. Another distinct property of ME tape is the

oblique easy axis of the tape. The easy axis of mag-

netization is known to exhibit a strong correlation

with both the oxygen ﬂow and the deposition geo-

metry. More oxygen leads to increased columnar in-

clination angles (Nagao et al. 1991, Bijker 1998).

The strong relation on the deposition geometry

was shown in, e.g., Ogawa et al. (1994) and Yano

et al. (1994). The magnetic anisotropy (see Magnetic

Films: Anisotropy) is found to be largely due to the

shape anisotropy of the columns (see, e.g., Richter

and Hibst 1991), although some contribution from

the crystal anisotropy may be expected in view of the

presence of small h.c.p. cobalt crystallites. The ques-

tion remains, however, of whether the crystal anisot-

ropy can contribute significantly since all the

observed crystals seem to be very small (B5 nm)

and randomly oriented. No clear texture was ob-

served from the microstructural analysis. From the

Auger depth-proﬁling experiments, it becomes clear

that an effective magnetization distribution over the

ﬁlm thickness must exist. When the packing density

distribution is known as a function of the ﬁlm thick-

ness, the effective magnetization proﬁle could be cal-

culated. However, judging from the oxygen proﬁle

alone, a maximum magnetization is reached in the

center section of the tape. A similar conclusion was

made in Richter (1993b), although assessed with a

combination of recording and VSM measurements.

Apart from a distribution of the saturation magnet-

ization, the longitudinal coercivity should also have a

distribution over the thickness of the tape, since the

columns are better separated in the lower parts of the

tape by the geometry of the deposition process. The

aid of oxygen leads to a better lateral decoupling of

the columns in the upper parts of the tape.

4. Anisotropy and Recording

The anisotropy is strongly related to the morphology

and microstructure of the deposited ﬁlm. The extreme

uniaxial anisotropy associated with the mesoscopic

columnar structure in ME tapes gives rise to an ob-

lique recording process which is close to one-dimen-

sional in nature (Nouchi et al. 1986).

An oblique easy-axis angle gives rise to asymmetric

pulse shapes. Since the columnar structure of ME

tape is asymmetrical with respect to the recording

head ﬁeld, it might be expected that the recording

behavior is somewhat different in both recording di-

rections. The output spectrum of the Hi8 ME tape is

shown in Fig. 8, where indeed a difference in output

between the two recording directions is observed.

Similarly, the good and bad recording directions

can be understood by considering the projection of

the head ﬁelds to the oblique medium easy axis in the

write process. If writing is assumed to occur at the

trailing edge of the head where the ﬁeld is equal to

, in the good direction of head-to-tape motion, the

magnitude of the projected total ﬁeld gradient is

larger, leading to sharper transitions (and thus larger

output) than the bad direction. The effect of record-

ing in the good and bad direction on the recorded bits

can be studied with magnetic force microscopy

(MFM) observations.

In Fig. 9, written bits are shown for recording in

both directions (see also Luitjens et al. 1996, Porthun

et al. 1998). The same current is used for writing the

bits. The distance between transitions ( ¼bit length)

is about 0.33 um. The good direction shows sharp bits

and track edges (Fig. 9, left). In the less favorable

Figure 8

Difference in recording output depending on the

recording direction of the Hi8 ME tape.

886

Metal Evaporated Tape

direction, less sharp transitions and interrupted bits

are visible (Fig. 9, right). The intrinsic magnetic

structures beside the tracks have more or less the

same dimensions as the written bits.

The direction, which leads to a higher output, is

generally referred to as the good direction. Subse-

quently the other direction is called the bad direction.

The difference is about 4.2 dBV (see Fig. 8). Apart

from the effect on output, the read-back pulses are

also known to be asymmetric (Krijnen et al. 1988).

From Fig. 9 it becomes clear that the transition

regions of the bits written in the good direction are

far less irregular compared to the image on the right

hand side where the bits are written in the bad re-

cording direction. On the right hand side, some sub-

domains are visible within the written bit structure

and the bits are connected to the outside domain

pattern. These irregularities contribute to the lower

read-back voltage in the bad recording direction.

Various microscopic models have been developed,

e.g., Richter et al. (1993). Most macroscopic models

of the magnetic reversal process give rise to inverse

cosine dependence. Incoherent reversal can be incor-

porated in a Stoner-Wohlfarth model by appropriate

truncation of the switching asteroid (Cramer 1993,

Lalbahadoersing 1999). In the limit of very strong

interactions, mean-ﬁeld models are not appropriate.

This does not seem to be the case for ME. A key

feature which needs to be incorporated, however, are

the cooperative interactions (which give rise to very

sharp transitions). The results of the different nu-

merical simulations are in good agreement with each

other and conﬁrm the earlier-mentioned one-dimen-

sional nature of the recording process in these ma-

terials. Both the one-dimensional and the more

sophisticated models indicate that there is little or

no improvement in the recorded output signal in ME

tape associated with the transition from longitudinal

to oblique recording (Stupp et al. 1994, Richter et al.

1993, Tagawa et al. 1991, Cramer 1993).

In Hokkyo et al. (1993), a clear relation was dem-

onstrated between the noise level of an ME tape and

the oxygen ﬂow. When more oxygen is incorporated

into the ﬁlm, the cobalt grains become smaller and a

simultaneous increase of the coercivity suggests a bet-

ter decoupled structure. Thus, the more oxygen is in-

corporated, the more cobalt grains per unit of volume

exist. From Mallinson (1993, p. 116), the signal-

to-noise ratio (SNR) for thin ﬁlm media is shown

to be:

SNR ¼

nWdl ðdBÞð1Þ

with n being the number of particles per unit of vol-

ume, W the track width, d the medium thickness, and

l the recording wavelength. From Eqn. (1) it becomes

clear that increasing the number of particles per unit

volume will lead to enhanced SNRs. Thus if the art of

making ﬁne decoupled cobalt particles is mastered, it

will lead to an increase in SNRs.

Some authors have addressed the question of

whether there exists exchange coupling over the

length of the column, so over the thickness of the

tape. Some clue is given by measurements of the ac-

tivation volume (Richter 1993a, Suzuki and Tajiri

1997). Although care has to be taken with the use of a

quantity like the activation volume, it might contrib-

ute in the discussion about the role of the exchange

coupling. In both articles it was found that switching

volumes are generally larger than the crystallites in

the columns but smaller than the column size as a

whole.

5. Co–Ni(–O) and Co(–O) ME Tape

In the early ME tape, Co–Ni–O is always used as the

thin ﬁlm material. The material is deposited using a

continuously varying angle of vapor incidence in an

oxygen environment, which leads to the mesoscopic

banana-shaped column morphology as shown in

Fig. 5. The dependence of the saturation magnetiza-

tion M

, coercivity H

, signal, and noise on the pro-

cess parameters is discussed in Hokkyo et al. (1993).

There is a direct relationship between H

and the O

content (due to the increased decoupling of grains)

and an inverse relationship between M

and the O

content (due to a reduction in the cobalt content and

an increase in the CoO content). To further study the

relation between deposition angle, oxygen concentra-

tion, magnetic properties, and recording perform-

ance, a mini roll coater has been developed (ten Berge

et al. 1994) and basic experiments have been per-

formed. Auger depth proﬁling showed that the Co/Ni

ratio is constant over the total thickness of the ﬁlm.

Figure 9

MFM measurements of the recorded magnetization

pattern of the Hi8 ME sample in both recording

directions. On the left are the bit transitions in the good

recording direction; the right hand side shows the

transitions in the bad recording direction.

887

Metal Evaporated Tape

The general picture that emerges from the TEM

and XRD analysis is that the ME tapes consist of a

banana-shaped columnar structure where small crys-

tallites are the building blocks of the entire column.

This has also been established in literature (see, e.g.,

Hokkyo et al. 1993). A schematic representation is

given in Fig. 7 as a result of combined TEM and

XRD analysis. A similar sketch can be found in

Kaneda (1997). In Fig. 10 a TEM cross section is

given from a Hi8 Co–Ni(–O) ME tape. From the

darkﬁeld observation, the more-or-less homogenous

distribution of the small crystals can be observed.

The O concentration is enhanced at the top (30 nm)

and the bottom (10 nm). The relation between the

deposition rate, the magnetic properties, and structur-

al feature has been studied by ten Berge et al.(1994),

Bijker (1998), Bijker et al. (1997), and Tohma et al.

(1997). Between 1993 and 2001, not only Co–Ni–O

tapes but also Co–O tapes have been prepared. The

so-called advanced ME tape has been produced by

oblique deposition of Co–O layers (Futamoto and

Honda 1994, Yoshida et al. 1994, Sugita et al. 1999,

Ishida et al. 2000). Experimental observations have

shown that these ﬁlms have a microstructure similar

to the Co–Ni–O ﬁlms: an hcp cobalt phase and Co–O

crystals of about 5 nm are found. The easy axis of

magnetization is about 50 degrees from the tape

normal.

These tapes have shown sharp transitions at 100

kfrpi (kilo ﬂux reversal per inch). The high recording

densities are a result of the microstructure as well

as the higher M

values (due to the higher cobalt

Figure 10

Cross sectional TEM views of a Hi8 ME tape. Both the brightﬁeld (left) and the darkﬁeld (right) are shown. In the

brightﬁeld view, the initial columnar inclination angle is also indicated.

Figure 11

MFM images of the recorded magnetization at 200 kfrpi (a) and 400 kfrpi (b) (after Ishida et al. 2000).

888

Metal Evaporated Tape

content) in these media. In Fig. 11 MFM images are

shown (Ishida et al. 2000) of the magnetization pat-

terns taken at 200 and 400 kfrpi. The white rectangular

regions indicate the bit size corresponding to an areal

density of 10 G b in

2

. Experimental work has been

started (2000) by studying the effect of a CoO under-

layer for the Co–O obliquely deposited ﬁlms. The

granular texture of the underlayer promotes separa-

tion between the obliquely grown columns and also

decreases the size of the columns (Tohma et al. 1997).

The difference between Co–Ni–O and Co–O tape is

that the latter has higher remanent magnetization

and crystalline anisotropy (and thus coercivity). This

results in better read/write characteristics and excel-

lent recording performance. This is also due to the

very thin DLC protective coatings, which can now be

prepared (Yoshida et al. 1994, Yoshida et al. 1995,

Shinohara 1994).

See also: Magnetic Recording: Flexible Media,

Tribology; Magnetic Recording: Patterened Media;

Magnetic Recording Technologies: Overview; Mag-

netic Recording: VHS Tapes; Metal Particle versus

Metal Evaporated Tape; Micromagnetics: Basic Prin-

ciples

Bibliography

Abelmann L, Lodder J C 1997 Oblique evaporation and surface

diffusion. Thin Solid Films 305, 1–21

Ashton G R, Thompson S J 1998 NML Thin Film Media Final

Report, http://www.nta.org/Publications/TechnicalReports/

MediaEvaluationAndRD/ThinFilmMediaFinal/

Bijker M D 1998 Oblique sputtered and evaporated magnetic

thin ﬁlms for tape recording. PhD thesis, University of

Twente, Enschede, The Netherlands

Bijker M D, Abelmann L, Lodder J C, Popma T J A 1999 The

role of oxygen in obliquely deposited tapes. J. Magn. Magn.

Mater. 193, 357–61

Bijker M D, Visser E M, Lodder J C, Popma T J A 1997

Oblique sputtered cobalt and Co80Ni20 thin ﬁlms. J. Magn.

Soc. Jpn. 21, 169–72

Chiba K, Sato K, Ebine Y, Sasaki T 1989 Metal-evaporated

tape for high band 8 mm video system. IEEE Trans. Con-

sumer Electron 35, 421–8

Cramer H A J 1993 On the hystheresis and the recording pro-

cess in magnetic media. Ph.D. Thesis, University of Twente,

Enschede, The Netherlands

Draaisma E A, Cumpson S R, Jansen T P H G, Hidding P,

Luitjens S B 1999 Performance of advanced tapes with

DigaMax thin ﬁlm heads. J. Magn. Magn. Mater. 193 (1–3),

384–7

Feuerstein A, La

mmermann H, Mayer M, Ranke H 1985

Videotape manufacture by high-vacuum evaporation aspects

of machine technology. Vacuum 7, 277–81

Feuerstein A, Mayer M 1984 High-vacuum evaporation of

ferromagnetic materials—a new production technology for

magnetic tapes. IEEE Trans. Magn. 20 (1), 51–6

Futamoto M, Honda Y 1994 Microstructures and recorded

magnetization microstructures of cobalt-based thin ﬁlm re-

cording media. J. Magn. Soc. Jpn. 18 (S1), 485

Gau J S, Spahn R G, Majumdar D 1986 The Role of oxygen in

oblique-deposited CoNi thin ﬁlms. IEEE Trans. Magn.

MAG-22 (5), 582–4

Hinemo T, Katahu T 1994 High-density recording magnetic

tape recording by nontracking method. J. Magn. Magn. Ma-

ter. 134, 255–61

Hokkyo J, Suzuki T, Chiba K, Sato K, Arisaka Y, Sasaki T,

Ebine Y 1993 Co-Ni obliquely evaporated tape and its mag-

netic and recording characteristics. J. Magn. Magn. Mater.

120, 281–5

Ishida T, Tohma K, Yoshida H, Shinohara K 2000 More than 1

Gb/in

recording on obliquely oriented thin ﬁlm tape. IEEE

Trans. Magn. 36 (1), 183–8

Kagawa K, Okabe A, Yoshida A, Iii T, Sugawara N, Takiguchi

M, Okazaki Y, Hayashi K 1997 Recording performance of

the combination of magnetoresistive head and thin metal-

evaporated tape. J. Appl. Phys. 81 (8), 4527–9

Kaneda Y 1997 Tribology of metal-evaporated tape for high-

density magnetic recording. IEEE Trans. Magn. 33, 1058–68

Kawana T, Onodera S, Samoto T 1995 Advanced metal-evap-

orated tape. IEEE Trans. Magn. 31 (6), 2865–70

Kita E, Tagawa K, Kamikubota M, Tasaki A 1981 Magnetic

recording media prepared by oblique incidence. IEEE Trans.

Magn. MAG-17 (6), 3193–5

Krijnen G, Luitjens S B, de Bie R W, Lodder J C 1988 Cor-

relation between anisotropy direction and pulse shape for

metal-evaporated tape. IEEE Trans. Magn. 24 (2), 1817–9

Kunieda T, Shinohara K, Tomago A 1985 Metal-evaporated

video tape. J. Inst. Electron. Radio Engrs. 55 (6), 217–22

Lalbahadoersing S 1999 Overwrite and track edges in high-

density tape recording: simulation and analysis. PhD Thesis,

University of Twente, Enschede, The Netherlands

Lalbahadoersing S, Groenland J P J, Luitjens S, Lodder J C

1997 Comparison of recording properties of ME tape and

thin MP tape with respect to overwrite behaviour. IEEE

Trans. Magn. 35 (5), 3061–3

Leybold Information Brochure 1992 Vacuum web coating,

13-230.02

Luitjens S B, Stupp S E, Lodder J C 1996 Metal-evaporated

tape: state of the art and prospects. J. Magn. Magn. Mater.

155, 261–5

Maezawa Y, Yoshida H, Shinohara K 1999 Characteristics of

ME tape made at 1 um/s deposition rate. IEEE Trans. Magn.

35 (5), 2688–90

Mallinson J C 1993 The Foundations of Magnetic Recording.

Academic Press, New York

Mujumdar D, Gau J S, Spahn R G 1986 Oxidation behavior of

obliquely deposited Co-Ni magnetic thin ﬁlms. Thin Solid

Films 145, 241–7

Nagao M, Yasunaga T, Dohshita H, Kanazawa M, Miura T

1991 Effect of oxygen gas on columnar structures and mag-

netic properties of Co-O evaporated thin ﬁlm. J. Magn. Soc.

Jpn. 15, (S2), 915–20

Nouchi N, Yoshida H, Shinohara K, Tomago A 1986 Analysis

due to vector magnetic ﬁeld for recording characteristics of

ME tape. IEEE Trans. Magn. 22 (5), 385

Ogawa Y, Yano A, Kitagaki N, Mizumura T, Okuwaki T,

Wakai K 1994 Recording characteristics of obliquely

evaporated thin Co-O ﬁlms. J. Magnics Soc. Jpn. 18 (S1),

413–34

Onodera S, Kondo H, Kawana T 1996 Materials for magnetic

tape media. MRS Bull. 21 (9), 35–41

Porthun S, Abelmann L, Lodder J C 1998 Magnetic force mi-

croscopy of thin-ﬁlm media for high-density magnetic re-

cording. J. Magn. Magn. Mater. 182 (1–2), 238–73

889

Metal Evaporated Tape