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

Подождите немного. Документ загружается.

recording media can be approximated by such a

model. If a recording is stressed in time by heat or

applied ﬁelds, this manifests itself as decay in the

signal sensed. If the track averaged output is plotted

against the logarithm of time, nearly linear decay

may result and can be used to predict the signal that

will remain after a given time, for a given tempera-

ture, at a given recording density (whose demagnet-

izing ﬁelds act to reduce the stability of the grain’s

magnetization).

4.14 Imaging Internal to the Drive

The scale of the magnetic features in a magnetic re-

cording system can range from smaller than a single

bit to many thousands of bits down-track or thou-

sands of tracks wide. A millimeter long scratch, a

thousandth of a millimeter wide, on a surface con-

taining a million bits per square millimeter recorded

with a bit aspect ratio of ten, may affect as many as

ten thousand bits. Such a defect may be difﬁcult to

observe without taking the recording system apart. A

sophisticated measurement that uses the read head as

a probe of the surface and then plots the output from

several adjacent tracks to build up a spatial picture of

the surface can assist in the determination of systemic

problems without having to disassemble the record-

ing system. This type of functional imaging of the

medium’s surface can have resolution comparable to

that of MFM and can be nondestructive.

See also: Longitudinal and Perpendicular Recording:

Thermal Stability; Magnetic Recording Devices:

Inductive Heads, Properties; Magnetic Recording

Materials: Tape Particles, Magnetic Properties; Mag-

netic Recording Media: Particulate Dispersions

Bibliography

Aharoni A 1996 Introduction to the Theory of Ferromagnetism.

Oxford University Press, New York

Ashar K G 1997 Magnetic Disk Drive Technology: Heads, Me-

dia, Channel, Interfaces, and Integration. IEEE Press, Piscat-

away, NJ

Bertram H N 1994 Theory of Magnetic Recording. Cambridge

University Press, New York

Bozorth R M 1993 Ferromagnetism. IEEE Press, Piscataway,

Brown W F 1963 Micromagnetics. Wiley, New York

Camras M 1988 Magnetic Recording Handbook. Van Nostrand

Reinhold, New York

Chikazumi S 1964 Physics of Magnetism. Wiley, New York

Craik D J 1995 Magnetism: Principles and Applications. Wiley,

New York

Hoagland A S, Monson K E 1991 Digital Magnetic Recording.

Wiley, New York

Jorgensen F 1988 The Complete Handbook of Magnetic Re-

cording, 3rd edn. Tab Books, Ridge Summit, PA

Jorgensen F 1996 The Complete Handbook of Magnetic

Recording, 4th edn. McGraw-Hill, New York

Mallinson J C 1993 The Foundations of Magnetic Recording,

2nd edn. Academic Press, San Diego, CA

Mee C D, Daniel E D 1990 Magnetic Recording Handbook:

Technology and Applications. McGraw-Hill, New York

Mee C D, Daniel E D 1996 Magnetic Recording Technology,

2nd edn. McGraw-Hill, New York

Morrish A H 2001 The Physical Principles of Magnetism. IEEE

Press, Piscataway, NJ

O’Grady K, Chantrell R W, Popplewell J, Charles S W 1981

Time dependent magnetization of a system of ﬁne cobalt

particles. IEEE Trans. Magn. 17 (6), 2943–5

Sharrock M P, Flanders P J 1990 Time-dependent magnetic

phenomena and particle-size effects in recording media.

IEEE Trans. Magn. 26 (1), 193–7

Wang S X, Taratorin A M 1999 Magnetic Information Storage

Technology. Academic Press, San Diego, CA

Wallace R L 1951 The reproduction of magnetically recorded

signals. Bell Syst. Tech. J. 30, 1145–73

Watkinson J 1990 Coding for Digital Recording. Focal Press,

Stoneham, MA

White R M 1985 Introduction to Magnetic Recording. IEEE

Press, Piscataway, NJ

R. S. Indeck and M. W. Muller

Washington University, St. Louis, Missouri, USA

Magnetic Recording Media: Advanced

The development of advanced metal particle (AMP)

media was initially driven by the need to match the

performance of metal evaporated (ME) media in

terms of areal data density and recording perform-

ance (output, frequency response, etc.). In this article,

the term ‘advanced MP media’ is taken to mean a

magnetic recording system containing a thin magne-

tic layer, an underlayer, a substrate, and a back coat.

A schematic representation of such a system is shown

in Fig. 1. The primary role of the back coat is to act

as a conductive layer to disperse any static charge

that may build up on the tape. The back coat will

not be addressed in this article other than to say it

consists of carbon black (graphitic carbon) particles

dispersed in a binder matrix with a typical layer

thickness of 0.5 mm.

The technological advantage of AMP media, ini-

tially introduced by Fuji Photo Film in 1992 (Shibata

et al. 1992) with their

Advanced Thin Layer High

Output Metal Media (ATOMM) and subsequently

adopted by many other magnetic media companies

producing advanced data recording products, is a

tribute to the technical excellence of the media man-

ufacturers. Details of the exact processes used by the

media manufacturers to obtain such uniform thin

magnetic coating layers remains very much proprie-

tary information.

600

Magnetic Reco rding Media: Advanced

ATOMM technology is deﬁned by the use of a dual

layer front coating comprising a thin (o0.25 mm)

magnetic layer on top of a nonmagnetic underlayer

(o2.0 mm). The dual coat strategy confers the consi-

derable advantages of reducing the surface roughness

of the magnetic coating after post-coat calendering

and improving the general durability of the media

due to the design properties of the underlayer. Tech-

nological developments in the material science of

metal particles, underlayer properties, and the sub-

strate have contributed to the continued development

of AMP media since the ﬁrst introduction of a dual

layer media product. The remainder of this article

will address the development of each of these key

areas, and address the impact of these developments

on the recording performance of the system.

1. Metal Particle Development

The manufacture of metal particles is highly propri-

etary within the very few companies that have large-

scale production capability. The general technology

trend for the particle manufacturers is to continually

evolve and improve the properties of the metal par-

ticles to meet the requirements of a two-fold data

capacity increase in AMP media every 2–3 years.

The ﬁrst metal particles for digital data storage,

developed in the mid-1980s, had a coercivity of ap-

proximately 1500 Oe (119 kAm

1

), a particle length

of 230 nm, aspect ratios of 410:1, and were based on

a magnetic iron core protected by a passivating ox-

idation shell. Such a particle is shown schematically

in Fig. 2. The corrosion resistance (Ds

) of these early

metal particles, which measured the decrease in spe-

cific magnetization (s

), when subjected to 7 days

storage at the environmental extremes of 60 1C and

90% relative humidity (RH), was typically around

20–25%. The loss of magnetic signal deﬁned by Ds

was initially a concern, as over time this corresponds

to a reduction in the signal output from the tape. The

early specialized surface treatments did manage to

reduce this to o10%, but with a corresponding loss

in s

which compromised recording output of the

media.

The incorporation of rare-earth elements in the

protective oxidation shell improved the particle’s

corrosion resistance. The increased effectiveness of

the outer shell in protecting the magnetic core from

the degrading effects of oxidation allows a reduction

in the average outer shell thickness. This reduction in

the outer shell thickness increases the s

value by al-

lowing an increase in the ratio of the volume of the

magnetic core to the volume of the oxidation shell.

The use of magnetic alloys, using 10–30% weight ra-

tio cobalt alloyed with iron as the magnetic core, has

also allowed an increase in coercivity from 1500 Oe to

2400 Oe (119 to 191 kAm

1

) for large-scale produc-

tion particles. Current AMP formats such as DDS-3

and DDS-4 media use metal particles with coercivi-

ties as high as 2300 Oe (183 kAm

1

), particle lengths

of o100 nm and aspect ratios from 4:1 to 5:1.

Figure 1

Schematic diagram of advanced MP media.

Figure 2

Schematic diagram of advanced metal particle.

601

Magnetic Reco rding Media: Advanced

Metal particle manufacturing technology continues

to advance at a pace that addresses the requirements

of high output, low noise, and acceptable Ds

for

AMP media, Okamoto et al. (1996). Metal particles

can be customized to meet the requirements of spe-

cific applications. The current high output types have

a s

of B150 emu g

1

which is approximately 25%

higher than the ﬁrst metal particles introduced for

digital data storage cartridges in the 1980s, such as

DLTIII. This is achieved while maintaining or

improving other important properties such as Ds

The very latest metal particle types have also been

reported to have particle lengths in the range of

40–60 nm to meet the requirements for low noise

applications. Table 1 illustrates the evolution of the

metal particle since the introduction in the 1980s to

the AMP media applications.

The advantages of higher coercivity, smaller par-

ticle size, and increased s

gave considerable advan-

tages in recording density and output. For example,

a ‘state of the art’ single-coated MP data cartridge

(DLTIII) of 1989 had a storage capacity of 10 GB

and a transfer rate of 1.5 MBs

1

. This compares to

a year 2000 data cartridge with 100 GB and transfer

rate of B15 MBs

1

. As the ﬂexible media recording

industry moves to recording systems using magneto-

resistive (MR) head technologies the demand for

high-output metal particles is being replaced by the

need for lower noise systems, i.e., smaller particles

with better dispersion properties.

As the particle size continues to reduce it becomes

more difﬁcult to maintain the coercivity and narrow

switching ﬁeld distribution of acicular particles, which

depend upon shape anisotropy for their coercivity.

The need for maintaining the higher coercivity and

narrow switching ﬁeld distribution of particles in the

sub-30 nm size range have led researchers to inves-

tigate nano-sized spherical particles, which depend

upon crystalline anisotropy for their coercivity.

Examples of these spherical nano-particles, with

particle sizes of less than 20 nm, are FePt based

particles (Hattori 2004) and iron-nitride based parti-

cles (Sasaki 2005).

2. Underlayer Technologies

The role of the underlayer is a crucial one for both

the recording and tribological properties of AMP

media. It is the underlayer in the dual coating pro-

cess, which allows for the uniform thin, smooth mag-

netic coating required for the improved recording

performance.

Saitoh et al. (1995) suggested that the role of the

underlayer is to realize advantages in durability by

replenishing the surface with lubricant from the un-

derlayer. The head/media contact can also be adjust-

ed by controlling the physical properties of the

underlayer, and conductive agents in the underlayer

can be used to control the electrical properties of

the ﬁlm. Initially the underlayer comprised of small,

spherical TiO

particles to meet the required me-

chanical and durability properties. Saitoh et al.

(1995) reported significant improvements in record-

ing performance for the dual layer system compared

to conventional single layer media.

The role of the underlayer was conﬁrmed (Sasaki

et al. 1997) and it was further reported that the fric-

tion coefﬁcient of the dual layer tape could be lower

than that of a single layer tape despite having a lower

surface roughness. More recently, advanced non-

spherical particles (nonmagnetic) have been used to

control the interface between the magnetic layer and

the underlayer (Inaba et al. 1998). The reliability/

durability of recording systems using AMP media has

been reported (Maness and Bradshaw 2000).

The underlayer may also be used to play a direct

role in the recording properties of the media. Mag-

netic particles have been used in the underlayer to

modify the long wavelength recording properties and

Table 1

Advanced media metal particle properties.

Classiﬁcation Year

Particle length

(nm)

Coercivity

(Oe)

(emu g

1

) Ds

Particle volume

(nm

)

1st MP data type particle

1985 230 1500 122 20 49 000

MP þ

1989 170 1600 121 20 38 600

MP þþ AMP media

1993 100 1850 130 12 24 000

MP þþþ AMP media

High output 1995–1998 B100 2300 150 10–15 21 000

AMP media

Low output 1999 B75 2400 110–140 B20 B10 000

602

Magnetic Reco rding Media: Advanced

to improve the ability of the system to overwrite

lower frequency signals with higher frequency signals

(Veitch et al. 1999).

3. Substrates

The material science of substrates, or base ﬁlms, has

evolved with the quest for increased storage capacity

per unit volume. Increasing the media length by re-

ducing the substrate thickness has immediate advan-

tages for increased storage capacity, provided that

the transportation mechanics of the drive can handle

the thinner media and that the desired conformance

to the head contour (for stable head/media separa-

tion) can be satisﬁed. The development of linear re-

cording formats using multiple data channels for

higher transfer rate applications has led to the need

for advanced substrates with very stable dimensional

properties (Richards and Sharrock 1998).

3.1 Dimensional Stability

Dimensional stability is deﬁned as the percentage

change in either the transverse or the longitudinal

direction when subjected to changes in temperature

(thermal) or relative humidity (hygroscopic).

Polyethylene teraphthalate (PET) was the substrate

material of choice for many generations of data tape,

ranging in thickness from 39 mm (1960s, 9 track) to

6.2 mm (mid-1990s DLT). To achieve a 6 mm thickness

in PET suitable for AMP media requires extensive

processing in terms of increasing the ﬁlm stiffness

in either the longitudinal or transverse direction.

The ﬁlm process known as tensilization induces an

orientation dependent crystallinity and increases the

mechanical stiffness (Young’s modulus, creep, etc.).

PET has a lower thickness limit of 6 mm and the de-

mand for even thinner media and increased stiffness

has been met by the introduction of substrate mate-

rials such as polyethylene naphthalate (PEN) and

polyaramid (PA). Table 2 illustrates typical proper-

ties of AMP media substrates.

PEN and PA substrates have lower thermal and

hygroscopic expansion coefﬁcients than PET. This

improves the dimensional stability when subjected to

environmental extremes and is a key requirement for

AMP multitrack, linear recording systems where the

effects of thermal and hydroscopic expansion can

cause a dimensional change both in the transverse

and longitudinal direction of the substrate. For ex-

ample, the transverse dimensional change has to be

controlled within a tolerance that allows sufﬁciently

accurate alignment of the recording heads with the

data tracks across the operating environment of the

system. This is particularly important for AMP me-

dia in a system such as LTO Ultrium, which has a

total of 384 data tracks across the 12.65 mm tape

width (Simmons 2000).

It can be calculated for this particular record-

ing system that the transverse dimensional change

must be limited to a maximum of 0.12% across the

operating environment of the system (10–45 1C and

10–80% RH). The primary properties of the sub-

strate are dictated by the chemistry of the substrate,

but these can be inﬂuenced by the relative degree of

tensilization in the longitudinal and transverse direc-

tion. This balance of tensilization, determined in the

manufacturing process, is critically important in

determining the ﬁnal dimensional stability properties

of the substrate. Table 3 illustrates the effect of tensi-

lization on dimensional stability.

Table 2

Advanced media substrate properties.

Property

PET

(tensilized in M/D) PEN PA

Minimum thickness

AMP media 6.2 mm 5.0 mm 3.2 mm is commercially available

Creep:

Load 28 MPa at 20 1C 0.7% 1.3% 0.8%

Tear strength

M/D 350 gmm

1

150 gmm

1

200 gmm

1

Heat shrinkage

M/D 100 1C, 30 min 1.8% 0.1% 0%

Transverse dimensional stability 0.13% 0.09% 0.1%

Young’s modulus

M/D 7200 MPa 7800 MPa 11300 MPa

T/D 5800 MPa 8100 MPa 14700 MPa

Typical application DLT IV (1996) LTO Ultrium (2000) DDS-3 and DDS-4 (1997 and 1999)

603

Magnetic Reco rding Media: Advanced

3.2 Dual Surface Substrate

The requirement for higher and higher recording

densities on thinner substrates have led the substrate

manufacturers to focus on making smoother subst-

rate surfaces to allow for a reduction in underlayer

thickness. This has resulted in substrate surface

roughness (R

) decreasing from B40 nm (1970s au-

dio/video tape) to o10 nm for the current generation

of AMP media. Surface roughness is controlled by

the addition of organic and inorganic ﬁller particles

with a tightly controlled particle size distribution.

These ﬁller particles create protrusions at the sur-

face of the substrate and a deﬁned surface roughness,

which is primarily dictated by the ﬁller particle size.

Smoother substrates ultimately give improved re-

cording properties but smoother substrates are more

difﬁcult to transport and guide through the media

manufacturing process. This presents the dilemma of

satisfying the competing requirements of reducing the

surface roughness for optimum recording perform-

ance while maintaining the substrates handling prop-

erties, required for the media manufacturer to

precisely control the tension and winding of the

substrate during a complex multistage coating oper-

ation. The major substrate manufacturers have met

this challenge with the development of dual surface

substrates, using their proprietary technologies that

allow both of these requirements to be met.

The dual layer substrate results in an extremely

smooth substrate ‘front coat’ surface (R

¼4 nm) and

a rougher ‘back coat’ surface (R

¼9 nm), which aids

processability of the substrate at the media manufac-

turer. The use of dual layer substrates, in place of

single layer substrates (R

¼6 nm), may yield an out-

put increase of up to 1.5 dB. AMP media products

are beginning to migrate to the use of dual layer

substrates for production of media with improved

recording properties.

3.3 Other Substrate Properties

Other important substrate properties include creep

and physical properties such as Young’s Modulus

and tear strength when subjected to load and thermal

stresses typical of those the media is subjected to

during normal drive winding and search operations.

These properties are dominated by the polymer type

but can be inﬂuenced by the processing in the same

way as the dimensional stability characteristics can be

customized for a particular application.

In a competitive data storage environment the

substrate cost is important and undoubtedly is in-

creasing due to the complex and evolving technology

of substrate design and manufacturing.

4. Concluding Remarks

The combination of the key developments in all of

these component (particles, substrate, and underlay-

ers) technologies has enabled AMP media to match

or exceed the recording performance of thin ﬁlm

(ME) tape media, and development continues at a

rapid pace. The emergence of MR read and thin ﬁlm

write head technology in multitrack linear tape re-

cording systems promises a long future of continued

growth in AMP media and will require further de-

velopment of smaller particles and thinner magnetic

layers (o100 nm).

See also: Magnetic Recording: Particulate Media,

Micromagnetic Simulations; Magnetic Recording

Technologies: Overview; Metal Evaporated Tape;

Metal Particle versus Metal Evaporated Tape; Mi-

cromagnetics: Basic Principles

Bibliography

Hattori Y, Waki K, Ogawa K, Furusawa G 2004 Synthesis of

FePt and FePtCu nanoparticles by a reverse micelle method

and studies of magnetic recording media using them. Trans.

Magn. Soc. Japan. 4, 85–8

Inaba A, Ejiri K, Masaki K, Kitahara T 1998 Development of

an advanced metal particulate tape. IEEE Trans. Magn. 34,

1666–8

Maness B, Bradshaw R 2000 Extensive testing/design features

support LTO reliability/interoperability. Data Storage 30–3

Okamoto O, Okazaki Y, Nagai N, Uedaira S 1996 Advanced

metal particles technologies for magnetic tapes. J. Magn.

Magn. Mater. 155, 60–6

Table 3

Effect of substrate tensilization on dimensional stability.

PEN Substrate

Classiﬁcation MD tensilized MD semi-tensilized Balanced

Young’s modulus (kg mm

2

)MD

900 800 700

600 650 730

Dimensional stability to thermal and

hydroscopic expansion

Very good Good Fair

Fair Good Very good

Note: MD ¼machine direction (longitudinal), TD ¼transverse direction.

604

Magnetic Reco rding Media: Advanced

Richards D B, Sharrock M P 1998 Key issues in the design of

magnetic tapes for linear systems of high track density. IEEE

Trans. Magn. 34, 1878–82

Saitoh S, Inaba H, Kashiwagi A 1995 Developments and ad-

vances in thin layer particulate recording media. IEEE Trans.

Magn. 31, 2859–64

Sasaki Y, Nishida Y, Okamoto K, Kondo H 1997 Role of the

under layer on surface smoothness and friction for double-

layered particulate tape. IEEE Trans. Magn. 33, 3064–6

Sasaki Y, Usuki N, Matsuo K, Kishimoto M 2005 Develop-

ment of NanoCAP (Nano Compost Advanced Particles)

technology for high density recording. Presented at Intermag.

2005 (paper GR-02) to be published in IEEE Trans. Magn.

Shibata N, Chino T, Chikamasa H 1992 Simultaneous video

double layer coating technology. Fujiﬁlm Res. Develop. 37,

107–11

Simmons R 2000 HDD technologies key to Ultrium tape. Data

Storage 25–8

Vietch R J, Ilmer A, Lenz W 1999 Improved overwrite per-

formance of dual magnetic tape layer coatings. IEEE Trans.

Magn. 35, 2787–9

G. Spratt

Hewlett Packard, Boise, USA

P. Hobby

Philmore Group Limited, Monmouthshire, UK

Magnetic Recording Media: Particulate

Dispersions

Magnetic particle dispersions or inks consist of mag-

netic particles suspended in a mixture of organic

solvents by a wetting polymer (O’Grady et al. 1991).

The particles are dispersed in a mill that applies

shear forces to break apart aggregates. The wetting

polymer adsorbs to the particle surface to provide

a steric barrier that inhibits particle ﬂocculation by

magnetic attraction. The dispersion also contains

lubricants and alumina particles to control the

tribology (see Magnetic Recording: Rigid Media,

Tribology) of the head–tape interface. Carbon black

may be added to raise the conductivity of the coat-

ing, allowing the dissipation of static electricity. A

polyisocyanate cross-linker is added immediately

before coating. In this article we will consider the

science of simple particulate dispersions that contain

only magnetic particles, wetting binders, and organic

solvents.

The ideal dispersion would have each particle sep-

arated from every other particle, the particles would

have an adsorbed polymer layer, and the adsorbed

polymer layer would be solvated by the organic sol-

vents. The ideal dispersion of individually dispersed

acicular particles is rare (Veitch et al. 1999) and more

likely the particles are found in bundles.

1. Particle–Binder Interactions

1.1 Particle Size Distributions

The preparation of commercial iron particles is de-

scribed in Magnetic Recording: Rigid Media, Prepa-

ration. The particles have a lognormal distribution of

lengths, diameters, and aspect ratios. For particles

used in high-end video tapes, our measurements

found the average length was 190 nm, the average

diameter was 22 nm, and the average aspect ratio was

9.8. The polydispersity of the distribution was 30.5%.

The effect of this distribution on the physical prop-

erties of magnetic dispersions is not known. No one

has ever prepared acicular particles with a monodis-

perse distribution of lengths and diameters. However,

polydispersity in sizes does have an effect on the

rheological properties of other colloidal systems

(Roscoe 1952).

A suggestion of the effect of polydispersity in par-

ticle lengths comes from computer simulations of

particle dynamics during orientation by an external

magnetic ﬁeld. A monodisperse distribution of sizes

assembles into a smectic ordered phase consisting of

layers with the particles oriented perpendicular to the

layer (Visscher and Gunal 1997). When the particles

have a lognormal distribution with a polydispersity

of greater than 10%, the ability of the particles to

form the ordered phases is suppressed (Visscher and

Gunal 1998). Such ordering has never been observed

in dispersions containing acicular magnetic particles.

Commercial particles have a polydispersity ranging

from 29% up to 53% (Inaba et al. 1998). It is inter-

esting to imagine the potential beneﬁts of highly or-

dered particle dispersions and the potential to achieve

high degrees of orientation in the resulting magnetic

coatings. One would ﬁrst need to discover methods to

produce monodisperse acicular magnetic particles.

1.2 Particle Surface Chemistry

Commercial iron particles have a thin oxide layer that

serves three roles: (i) to prevent sintering during par-

ticle preparation, (ii) to modify the magnetic prop-

erties, and (iii) to passivate the iron surface. The

thickness of the oxide layer is about 4 nm (Rosler

et al. 1996). The oxide contains mostly aluminum,

but can contain bismuth, boron, cerium, cobalt, neo-

dymium, silicon, titanium (Gokturk and Maki 1997),

or yttrium (Okamoto et al. 1996). Except for com-

positional data obtained from x-ray photoelectron

spectra, little is known about the chemistry of the

oxide layer.

The layer is amorphous, as one would expect, since

the particles are processed at temperatures less than

500 1C. This is much lower than the crystallization

temperature for alumina. If one considers the oxide

layer to be a doped amorphous alumina phase, then

the literature for alumina chemistry can provide

605

Magnetic Reco rding Media: Particulate Dispersions

insight into the types of binding sites available for

adsorbates (Henrich 1994). Alumina has both acidic

and basic sites. The dopants play a role in determin-

ing the nature of the sites. Commercial iron particles

are basic (Gokturk and Maki 1997, Veitch et al.

1999), which means there are more basic sites than

acidic sites. The acid sites are coordinately unsatu-

rated aluminum cations and the base sites are coor-

dinately unsaturated oxygen anions (Hsaio et al.

1998).

1.3 Wetting Binders

The wetting binders are moderate molecular weight

polymers containing anionic functional groups, e.g.,

carboxylates, sulfonates, or phosphonates that have a

high afﬁnity for the particle surface. The binders

chemisorb to the particle surface, anchored primarily

by the anionic functional groups. Other functional

groups in the binder, such as alcohols, also adsorb to

the particle surface. The binder also has loops and

tails that extend from the anchor points away from

the particle. This provides the steric stabilization of

the dispersion. Early wetting binders had the wetting

groups randomly distributed along the polymer

chains. Subsequently the use of block copolymers

has been reported (Weingart et al. 1996). One block is

responsible for adsorption and the other block for

steric stabilization.

The two most important wetting binders for MP

tape have been poly(vinyl chloride) (PVC) and ther-

moplastic polyurethanes (TPUs). The PVC is a co-

polymer of vinyl chloride and vinyl acetate, where a

portion of the acetate groups has been hydrolyzed to

give vinyl alcohol groups. The most important PVC

is MR-110 from Nippon Xenon (Jeon et al. 1997),

which contains 0.5 wt.% vinyl alcohol. The alcohol

groups have an afﬁnity for the particle surface;

however, MR-110 also contains 0.7 wt.% sulfonate

groups that have a high afﬁnity for the surface.

The weight average molecular weight (M

)is

about 26 000. On average, each polymer chain con-

tains about one alcohol and one sulfonate wetting

group. TPUs, such as Estanes from BFGoodrich or

Morthanes from Morton International, have either

sulfonate or phosphonate wetting groups. These are

block copolymers containing polyester blocks cou-

pled together with urethane linkages. The sulfonate

groups are introduced into the polyester blocks via

copolymerization of 5-sulfoisophthalic acid (Yatsuka

et al. 1997).

1.4 Solvents

The solvents used in magnetic dispersions include

ketones (methyl ethyl ketone (MEK), methyl isobutyl

ketone (MIBK), or cyclohexanone), tetrahydrofuran

(THF), and toluene. These are typically mixtures

containing a highly volatile solvent (MEK or THF)

and a less volatile solvent (MIBK, cyclohexanone, or

toluene). The choice of solvent mixture is determined

by dynamics of the drying process. After coating and

orienting in a magnetic ﬁeld, the more volatile solvent

leaves the coating. This causes a large increase in

viscosity that locks the particles in place. The less

volatile solvent plasticizes the binder, allowing any

stresses developed during drying to be relieved.

A problem with the organic solvents used in mag-

netic tape manufacture is their potential to cause air

pollution. The US Environmental Protection Agency

(EPA) has listed MEK, MIBK, and toluene as haz-

ardous air pollutants. The regulations for the US

magnetic tape industry have been published in the

1994 National Emissions Standards for Hazardous

Air Pollutants (Federal Register 1994). In modern

tape manufacturing plants the solvent vapors from

the drier are recovered by use of a carbon bed ad-

sorber. The carbon bed is stripped by steam, and the

steam is condensed or put through a distillation col-

umn. This allows recovery and reuse of the individual

components. Typically 93–95% of the solvents are

captured and recycled. The coating operations are

safe and the air emissions meet EPA standards. Con-

siderable capital equipment is required to contain and

recycle the organic solvents. With increasing world-

wide concern for the environment, there will be in-

creasing pressure to minimize or eliminate air

pollution in tape coating operations.

1.5 Adsorption

Adsorption of molecular or macromolecular disper-

sants onto commercial iron particles follows a Lang-

muir isotherm behavior. The amount depends on the

choice of anionic anchoring group and the solvent

mixture. Adsorption of stearic acid from THF solu-

tion gives a Langmuir capacity of 52 mg g

1

, while

adsorption from toluene solution gives a capacity of

86 mg g

1

(Papirer et al. 1997). Clearly THF is com-

peting with stearic acid for the surface sites.

Adsorption of bis(2-ethylhexyl) phosphate from

THF solution gives a compact monolayer on the par-

ticle surface with a Langmuir capacity of 36 mg g

1

THF does not compete with bis(2-ethylhexyl) phos-

phate for binding sites. Incorporating sulfonate

groups into the polyester blocks of TPUs increases

the amount of polymer adsorbed to the particle sur-

face (Table 1). For a polymer having no sulfonate

groups, adsorption from 64/36 wt.% toluene/2-buta-

none gives a capacity of 80 mg g

1

. When 2.5 mol%

5-sulfoisophthalic acid is incorporated into the poly-

mer, the adsorption amount increases to 185 mg g

1

For both polymers the amount adsorbed is signiﬁ-

cantly lower when cyclohexanone is included in the

solvent mixture. Cyclohexanone is competing with the

polymer for binding sites on the particles.

606

Magnetic Reco rding Media: Particulate Dispersions

2. Characterization of Magnetic Inks

2.1 Microstructure of Magnetic Inks

The high concentration of magnetic particles in the ink

results in strong magnetic and steric interparticle in-

teractions. Figure 1(a) is a cryo-transmission electron

microscopy (TEM) image of an MP ink (Chen and

Nikles 2001). A microstructural simulation of an MP

ink is shown in Fig. 1(b) (Visscher and Gunal 1997). A

schematic that can be used as a model of the micro-

structure is shown in Fig. 1(c). Individual particles are

usually held together in small bundles of particles in a

side-by-side orientation of their magnetic moments.

Weaker dipole–dipole magnetic attraction between the

ends of these bundles link them together in a larger

network or ﬂoc. The attractive magnetic interactions

are countered by the steric repulsion force from the

polymer adsorbed on the particle surfaces.

This microstructure determines the physical be-

havior of the ink and ultimately the properties of the

ﬁnal tape product. The interactions between aggre-

gates determine the colloidal stability of the ink. The

viscoelastic properties determine how the ink ﬂows in

the coating head and onto the web or onto an un-

derlayer in the case of double coating. The net mag-

netic moment and size of the particles and aggregates

and their interaction with other aggregates deter-

mines their ability to rotate in an orienting ﬁeld.

These properties also affect how quickly the particles

and aggregates will disorient due to elastic recovery

and Brownian motion once the web moves beyond

the orienting ﬁeld.

2.2 Physical Properties of the Magnetic Inks

Many researchers have characterized the properties

of magnetic ink using various rheological and mag-

netic techniques. The most common and meaningful

test is to simply make a hand-drawn laboratory tape

sample by spreading the ink on a plastic ﬁlm or glass

plate with a calibrated knife coater. When dry this

sample can be observed optically with a gloss meter

or microscope and tested in a VSM or BH loop tracer

for magnetic properties such as coercivity, square-

ness, and switching ﬁeld distribution. These tests are,

of course, only an indirect measure of ink properties.

(a) Optical methods

It is difﬁcult to use normal optical methods to view

the ink because it is too thick to transmit light. Figure

1(a) was obtained by spreading a thin ﬁlm of wet ink

at its normal solids concentration onto a TEM grid

with immediate quenching in liquid nitrogen to pre-

vent the solvent from evaporating. It was kept frozen

in a specialized TEM sample holder and observed at

10 000% magniﬁcation. Both the particle agglomer-

ates and their network structure can be observed.

However, one must search the TEM grid for regions

thin enough to see through, leaving some doubt as to

how representative any one region is. The ink can

also be diluted with more solvent or other particles

and dried on a TEM grid (Scholten and Felius 1990).

The aggregates can be observed but their network

structure is destroyed. These methods give qualitative

insight into the structure of the ink.

(b) Rheology

It is common to test the rheology of the ink, especially

the viscosity at high and low shear rates. In rheology

Table 1

Effect of the presence of sulfonate groups and the choice of solvent mixture on the adsorption capacity of

thermoplastic polyurethanes.

Polymer

Toluene (wt.%) 2-Butanone (wt.%) Cyclohexanone (wt.%) Capacity (mg g

1

)

1 36.5 36.5 27 19

150 50 0 64

164 36 0 80

2 36.5 36.5 27 140

2 50 50 0 170

2 64 36 0 180

Source: Yatsuka et al. (1977).

a Polymer 1 has no 5-sulfoisophthalate groups, polymer 2 has 2.5 mol% 5-sulfoisophthalate.

Figure 1

MP dispersion: (a) cryo-TEM photograph;

(b) simulation; (c) schematic of dispersion

microstructure with small bundles of particles connected

in a network structure.

607

Magnetic Reco rding Media: Particulate Dispersions

measurements a force is applied to the surface of a

ﬂuid and the rate at which it moves is measured. A

sample of ink is subjected to shear at a controlled

strain rate and the resulting stress is measured in a

controlled rate instrument or vice versa in a controlled

stress instrument. There are many variations of this

experiment with shear applied in a steady, transient,

or oscillatory fashion, at various amplitudes and fre-

quencies, in various geometries, and in the presence of

other shear and magnetic forces.

The viscosity as a function of shear rate for an MP

ink is shown in Fig. 2 (Chae, unpublished data, 2001).

The viscosity is the ratio of shear stress to shear rate

and should be constant for a Newtonian ﬂuid. These

inks are decidedly non-Newtonian because of the in-

teractions between particles. At low shear rates the

network structure breaks and reforms on a time scale

faster than the shear rate resulting in a constant vis-

cosity. But high shear rates tear apart the structure

faster than it can reform, making the ﬂuid easier to

move. This is called shear-thinning behavior.

The shear can be applied in a sinusoidal fashion to

obtain the storage (G

) and loss modulus (G

) that

reﬂect the elastic and viscous nature of the ink. If the

amplitude is kept small, the particle microstructure

remains intact and can be probed. Figure 3 shows G

and G

of a Co–g-Fe

ink at various resin concen-

trations (Potanin et al. 1998). Magnetic inks are G

dominant, unlike many polymer solutions. The elas-

ticity of the particle network is very strong. Further-

more, the moduli are much higher than for the resin

solution without particles, indicating that the visco-

elastic properties of the ink are largely determined by

the particle interactions. Note that the moduli under-

go a minimum at around 5 wt.% polymer concentra-

tion. At low polymer concentration, the magnetic

attraction between particles is very strong in the ab-

sence of good steric protection. At high polymer con-

centration, excess polymer builds up on the particles,

giving them a different type of elastic response.

In a magnetic measurement an applied magnetic ﬁeld

acts on the individual particles inside the ink. A VSM

measurement of a g-Fe

ink showed a ‘‘coercivity’’

of 100 Oe that is due to particle rotation in the wet

ink and not magnetic switching (which can be seen as

a much smaller transition at 800 Oe) (Greaves et al.

1996a). The complication in interpreting magnetic

measurements of magnetic ink is that the particles

readily move, a phenomenon that does not happen

when measuring tape.

Methods have been demonstrated that effectively

prevent the particles from moving in response to the

magnetic ﬁeld so that magnetic switching information

can be obtained. In one technique, the ink is frozen at

liquid nitrogen temperatures (with or without ﬁrst

applying an orienting ﬁeld) and kept frozen during a

VSM measurement. Figure 4 shows the interaction

parameter, dM, for a dispersion of barium ferrite

with different amounts of milling (Greaves et al.

1996b). The positive interactions (due to stacking of

the hexagonal platelets) decrease as the time of mill-

ing increases. In another technique, a special pulse

magnetometer is constructed to apply a ﬁeld over a

period of 50 ns, too fast for the particle to physically

move, but slow enough for the magnetic moment to

switch (Greaves et al. 1996a).

Much information about the ink can be obtained

by observing the motion of the particles in a magne-

tic ﬁeld. As with rheology there are many varia-

tions on this experiment with the ﬁeld applied in

a steady (d.c.), transient, or oscillatory (a.c.) fashion,

at various amplitudes and frequencies, in various

geometries, and in the presence of other shear and

magnetic forces.

In a.c. susceptibility measurements, a low-ampli-

tude a.c. ﬁeld wiggles the particles while their motion

Figure 2

Viscosity as a function of shear rate for an MP

dispersion.

Figure 3

Storage (G

, m) and loss (G

, ’) modulus as a function

of polymer concentration. G

(*) is also shown for the

polymer solution.

608

Magnetic Reco rding Media: Particulate Dispersions

is sensed by a pickup coil. If a transverse d.c. ﬁeld is

applied, the particles rotate perpendicular to the a.c.

ﬁeld, increasing the signal. This motion is presumably

larger for better dispersed particles. However, in a

dispersion of Co–g-Fe

particles with 3.75% or

15% wetting binder, the susceptibility (at an a.c. fre-

quency of 100 Hz) improves with milling for the low-

resin ink but decreases for the high-resin ink. Figure 5

shows the in-phase (w

) and out-of-phase (w

) com-

ponents of the a.c. susceptibility (with no transverse

d.c. ﬁeld) over a large frequency range (Peikov et al.

1999). The data are ﬁt and extrapolated using a

Cole–Cole model. The low-frequency a.c. suscepti-

bility shows that both dispersions are actually im-

proved with milling. The relaxation time, as indicated

by the peak of the w

curve, shows a shift to longer

times (lower frequencies) consistent with a buildup of

polymer on the particles.

3. Conclusions

Much progress has been made in the understanding

of the physical properties of magnetic inks and the

underlying particle microstructure. It is still difﬁcult,

though, to determine exactly what makes a good ink

become a good tape. Most techniques are demon-

strated with very good and very bad dispersions

(milling time, polymer concentration, solids content)

and the differences are large and interpretable. But of

more practical interest to industry is the difference

between a very good and an excellent dispersion. This

requires more fundamental research and develop-

ment of a large body of experience.

See also: Magnetic Recording Technologies: Over-

view; Magnetic Recording Media: Advanced; Mag-

netic Recording Materials: Tape Particles, Magnetic

Properties; Magnetic Recording: Rigid Media, Prep-

aration; Metal Particle versus Metal Evaporated

Tape

Bibliography

Chen M, Nikles D E 2001 Surface characterization of iron

particles treated with a polyvinylchloride wetting binder.

IEEE Trans. Magn. 37, 1637

Federal Register 1994, 59 (240), 64180–612

Gokturk H, Maki K 1997 Surface properties and particle

interactions of metal magnetic particles. J. Magn. Magn.

Mater. 167, 166–78

Greaves S J, Hancock C P, O’Grady K, Lane A M 1996a Pulse

and low temperature magnetometry of g-Fe

dispersions.

IEEE Trans. Magn. 32 (5), 4040–2

Greaves S J, O’Grady K, Lane A M 1996b Interaction effects in

barium ferrite dispersions using low temperature magneto-

metry. IEEE Trans. Magn. 32 (5), 4046–8

Henrich V E 1994 The Surface Science of Metal Oxides. Cam-

bridge University Press, London

Hsaio G S, Erley W, Ibach H 1998 Probing the surface acidity

of single crystal alumina surfaces: CO adsorption as meas-

ured by infrared reﬂection absorption spectroscopy. Surf.

Sci. 405, L465–71

Inaba H, Ejiri K, Masaki K, Kitahara T 1998 Development of

an advanced metal particle tape. IEEE Trans. Magn. 34 (4),

1666–9

Jeon K S, Shrauti S M, Arnold D W, Lane A M 1997 Trans-

verse susceptibility and polymer adsorption of magnetic ink.

IEEE Trans. Magn. 33, 3067–9

O’Grady K, Gilson R G, Hobby P C 1991 Magnetic pigment

dispersions (a tutorial review). J. Magn. Magn. Mater. 95,

341–55

Okamoto K, Okazaki Y, Nagai N, Uedaira S 1996 Advanced

metal particles technologies for magnetic tapes. J. Magn.

Magn. Mater. 155, 60–6

Papirer E, Walter E, Vidal A, Sifert B 1997 Adsorption of steric

acid and diethylhexyl phosphate on magnetic metallic iron

pigments: electrical surface charging and adsorption compe-

tition. J. Coll. Interfac. Sci. 187 , 529–38

Peikov V T, Jeon K S, Lane A M 1999 Characterization of

magnetic inks by measurements of frequency dependence of

AC susceptibility. J. Magn. Magn. Mater. 193, 307

Figure 4

Interaction parameter, dM, as a function of applied d.c.

ﬁeld for barium ferrite inks milled for various times. The

cryo-VSM technique was used to keep the ink frozen in

order to eliminate particle motion.

Figure 5

In-phase, w

, and out-of-phase, w

, a.c. susceptibility of a

Co–g-Fe

ink prepared with different resin

concentrations and milling time.

609

Magnetic Reco rding Media: Particulate Dispersions