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

Подождите немного. Документ загружается.

the tape leader is threaded into the drive over a read/

write head to the take-up reel by a pentagon thread-

ing mechanism. A decoupler column placed near the

cartridge entrance decouples any tape vibrations that

may occur inside the cartridge. The tension is sensed

and controlled by a tension transducer. In a rigid disk

drive as shown in Fig. 3(a), a stack of disks is mounted

on a sealed, grease lubricated ball-bearing spindle. The

disk stack is rotated by a d.c. motor at speeds ranging

from a few thousand rpm to a maximum speed of

about 10 000 rpm. A head slider is supplied for each

disk surface. The slider suspension assembly is actu-

ated by a stepping motor or a voice-coil motor for

read/write operation (Bhushan 1996a, 2000).

Magnetic heads consist either of conventional in-

ductive or inductive/magnetoresistive (MR) devices

(Bhushan 1993, 1996a, 2000). If MR head design is

used, it is only used for read purposes. Air-bearing

surfaces of tape heads are generally cylindrical in

shape (Fig. 2(b)). The tape is slightly underwrapped

over the head surface to generate hydrodynamic lift

during read/write operations. The head sliders used in

rigid disk drives are either two- /three-rail, taper-ﬂat

design or two-rail, step, negative pressure design

(Fig. 3(b)). The sliders are supported by a nonmag-

netic steel leaf spring (ﬂexing) suspension to allow

motion along the vertical, pitch, and roll axes. The

front taper or step pressurizes the air lubricant, while

some air leaking over the side boundaries of the rail

results in a pitch angle.

Magnetic media fall into two categories: (i) partic-

ulate ﬂexible media, where magnetic particles are dis-

persed in a polymeric matrix and coated onto the

polymeric substrate (Fig. 2(c)); and (ii) thin-ﬁlm media,

where continuous ﬁlms of magnetic materials are

deposited by vacuum techniques onto the polymeric

substrate or ﬂexible media (Fig. 2(c)) or onto a rigid

substrate such as aluminum, glass, or glass ceramics

(Fig. 3(c)) (Bhushan 1993, Miller and Bhushan 1996).

2. Macrotribology of Magnetic Storage Devices

Friction and wear of two surfaces in relative motion

is governed by the real area of contact and interfacial

adhesion. In addition, for wet contacts with a tiny

amount of liquid present at the interface, ‘‘stiction’’

(or high static friction) can occur owing to meniscus/

viscous effects (Fig. 4). Generally, any liquid that

wets or has a small contact angle on surfaces will

condense from vapor in the form of an annular-

shaped capillary condensate in the contact zone. The

negative (attractive) pressure of the liquid ﬁlms of the

capillary condensates or pre-existing ﬁlm of lubricant

can significantly increase the adhesion between solid

bodies. Liquid-mediated adhesive forces can be

divided into two components: meniscus force (F

)

owing to surface tension and a rate-dependent vis-

cous force (F

). The total tangential force (F) re-

quired to separate the surfaces during sliding is equal

to intrinsic friction force (F

) and stiction force (F

combination of friction force owing to the meniscus

Figure 3

Schematics of (a) a rigid disk drive, (b) an inductive/MR

thin-ﬁlm picoslider, and (c) sectional view of a thin-ﬁlm

rigid disk.

Figure 4

Schematic of a wet interface with meniscus formation,

showing various forces being applied during sliding. F

is the meniscus force, W is the normal force, and F

and

are intrinsic friction and stiction forces, respectively.

550

Magnetic Reco rding Devices: Head/Medium Interface

effect and the peak viscous force) (Bhushan 1999b):

F ¼ F

þ F

¼ m

ðW þF

ÞþF

ð1Þ

where m

is the ‘‘true’’ static coefﬁcient of friction and

W is the normal the load. It turns out that since the

surfaces in magnetic storage devices are smooth and

the normal load being applied is relatively low, fric-

tional effects because of meniscus and viscous forces

are significant. Optimization of surface topography

of head and medium and overcoats and lubricant

ﬁlms needs to be achieved for long-term reliability of

magnetic storage devices.

2.1 Contact Modeling

Numerical techniques are used to solve contact prob-

lems of rough surfaces under dry and wet conditions

(Bhushan 1996b, 1996c, 1998, 1999b, Tian and

Bhushan 1996a, 1996b, Peng and Bhushan 2001).

The numerical model does not require an asperity

model and makes no probabilistic assumptions such

as distribution of asperity heights, slopes, and curva-

tures, as required in analytical models. The numerical

techniques allow analysis of a large number of con-

tact spots that go through elastic and elastic/plastic

deformations. The numerical model provides useful

information on the number of contact spots, their

sizes and distributions, and the spacing between con-

tacts. Furthermore, wet contact analysis for contact

of rough surfaces with a liquid ﬁlm can be carried

out. In the wet contact analysis, elastic/plastic dry

contact of rough surfaces is ﬁrst analyzed. In the next

step, a liquid ﬁlm of known mean thickness is intro-

duced over the deformed rough surfaces. Wetted

areas are determined by selecting the area where as-

perities of both contacting surface touch the liquid.

The total projected meniscus area is then used for the

calculations of meniscus forces. The model can be

used to develop optimum surface roughness distri-

butions for the head and medium surfaces. Figure 5

shows representative contact area and meniscus area

maps for a computer-generated random rough sur-

face in contact with a smooth surface in the presence

of water ﬁlm. As expected, the meniscus area is larger

than the contact area and the meniscus force is three

times that of the normal force.

Tian and Bhushan (1996b) and Poon and Bhushan

(1996) used the model to predict the effects of relative

humidity and liquid ﬁlm thickness on the meniscus

force (see also Bhushan 1998). The effect of relative

humidity on the meniscus force at an interface is

shown in Fig. 6(a). The effect of liquid ﬁlm thickness

and interface roughness on computer-generated

rough surfaces in contact with a smooth surface is

shown in Fig. 6(b). An increase in either relative hu-

midity or liquid ﬁlm thickness increases the liquid

present at the interface. The thicker a liquid ﬁlm, the

more asperities touch the liquid surface and the more

menisci form on a larger number of asperities. In

addition, with a thicker ﬁlm, a larger volume of liquid

is present around the asperities resulting in a greater

amount of meniscus volume accumulated at the con-

tact interface and a greater meniscus height. These

effects lead to larger meniscus forces. There is a crit-

ical ﬁlm thickness for a surface with given roughness

above which the meniscus force increases rapidly.

The critical ﬁlm thickness is of the order of three-

quarters of the composite r.m.s. roughness of the

mating surfaces (s).

Bhushan and Chilamakuri (1996) and Chilamakuri

and Bhushan (1998a) used the numerical model to

study contact between ﬂat rigid surfaces and non-

Gaussian rough surfaces with known skewness (Sk)

and kurtosis (K). Rough surfaces with different skew-

ness and kurtosis were generated using a two-dimen-

sional digital ﬁlter technique. Figure 7 shows the

probability density functions of surfaces with various

skewness and kurtosis values. Figure 8(a) shows the

effect of skewness and kurtosis on the fractional real

area of contact (A

, where A

is the contact area

and A

is the apparent area) and relative meniscus

force (F

/W) at different nominal pressures (p). A

positive skewness between 0 and 0.2 at low pressure

and about 0.2 at higher pressures results in the lowest

real area of contact and meniscus force. Contact area

and meniscus force decrease with an increase in the

kurtosis. Fewer peaks present on a surface with a

range of positive skewness values or high kurtosis can

explain the trends. Figure 8(b) shows the variation of

relative meniscus force with h/s ratio for different

skewness and kurtosis values. Note that the sensitiv-

ity of h/s to the meniscus force decreases at a range of

Figure 5

Contact area and meniscus area for the case of

computer-generated rough surface (s ¼1 nm,

b* ¼0.5 mm) in contact with a smooth, ﬂat surface with

a composite elastic modulus of 100 GPa and a water ﬁlm

(g ¼73 dyne cm

1

(7.3  10

4

Ncm

1

), y ¼601)

thickness of 1 nm and meniscus height of 1 nm (after

Poon and Bhushan 1996).

551

Magnetic Reco rding Devices: Head/Medium Interface

positive skewness of 0–0.2, and kurtosis of about ﬁve

or larger is optimum.

Laser-textured disks are becoming increasingly

popular in the design and manufacture of rigid disks.

In the laser-texturing technique, discrete topographic

features with either rounded, dome-like protrusions

(sombrero shaped) or craters (V or W types or donut

shaped) usually known as bumps are generated over

the disk surface. In laser-textured disk surfaces, these

microscale features are placed in the landing zone to

reduce friction/stiction during the start and stop ope-

rations. Typical sombrero-, V-, and W-type bumps

have about 10–15 mm rim diameter and height of

about 15–25 nm with a 50 mm 50 mm pitch. This

gives of the order of 4 10

bumps in the landing

zone of a 95 mm diameter disk. V- or W-type bumps

are more commonly used because they can be rep-

roducibly manufactured. The stiction phenomena at

the head/disk interface can be controlled by control-

ling the size and shape of the laser bumps. Studies to

identify an optimum bump shape and the relationship

between the optimum number of bumps as a function

of the bump geometry for a given bump shape have

been conducted (Chilamakuri and Bhushan 1997,

1998b, 1999a, 1999b, Chilamakuri et al. 2000).

The following design methodology has been used to

determine an optimum number of bumps. For a given

size and shape, the meniscus force increases with an

increase in the number of bumps, so it is important to

minimize the number of bumps to control the stiction

phenomena. However, too few bumps can result in

Figure 6

(a) Effect of relative humidity on the relative meniscus

force for a glass/ceramic disk substrate in contact with a

smooth surface (after Tian and Bhushan 1996b). (b)

Effect of water ﬁlm thickness and surface roughness on

the relative meniscus force for computer-generated

Gaussian surfaces (correlation distance b* ¼0.5 mm) in

contact with a smooth surface. The dotted line deﬁnes

the critical ﬁlm thickness for different s (after Poon and

Bhushan 1996).

Figure 7

Probability density functions for surfaces with (a)

different skewness and (b) different kurtosis values (after

Chilamakuri and Bhushan 1998a).

552

Magnetic Reco rding Devices: Head/Medium Interface

the failure of the disk surface owing to plastic defor-

mation and wear. To prevent the bumps from yield-

ing, the number of bumps on the disk surface should

be greater than a certain minimum, which would

also give a lower bound for the stiction. Numerical

analyses of computer-generated surfaces are used to

predict an optimum number of bumps. Figure 9

shows typical computer-generated surfaces with var-

ious bump types and variable heights. The three-

dimensional numerical contact model mentioned

earlier is used to perform contact analysis of a com-

puter-generated laser-textured surface against a ﬂat

surface. Figure 10 shows an optimum number of

bumps, contact area, and meniscus force as a func-

tion of yield stress at different values of coefﬁcient

of friction. As expected, the optimum number of

bumps increases with an increase in the coefﬁcient of

Figure 8

(a) Fractional real area of contact and relative meniscus

force as a function of skewness and kurtosis at various

nominal pressures, and (b) relative meniscus force as a

function of h/s for different skewness and kurtosis

values, for an interface in the presence of

perﬂuoropolyether ﬁlm (g ¼25 dyne cm

1

(2.5 10

4

Ncm

1

), y ¼101) (after Chilamakuri and

Bhushan 1998a).

Figure 9

Three-dimensional surface plots and two-dimensional

line plots of sombrero-, V-, and W-type laser surfaces

rim

¼7.5 mm, H

mean

¼25 nm, s ¼5 nm) with a

Gaussian height distribution.

553

Magnetic Reco rding Devices: Head/Medium Interface

friction. Contact area and meniscus force also

follow similar trends. Measurements made on exper-

imental disks verify numerical predictions (Bhushan

2000).

2.2 Overcoats and Lubricants

The carbon overcoat is almost exclusively used as an

overcoat for thin-ﬁlm disks and metal tapes (Bhushan

1999c). It has to be chemically active in order to attach

to lubricant molecules, whereas the lubricant mole-

cules should be hydrophobic so that they repel water

from the environment. Lubricants can attach to the

carbon surface by one or more of the following

mechanisms: physisorption, conformational rearrange-

ment, and chemisorption. The type and degree of

attachment depends on the molecular structure and

end groups of the lubricants and surface groups on

the carbon surface. A hydrogenated carbon surface is

highly functional and consists of many oxygen-

containing surface groups which include –C–O–C–

(ether), 4CQO (carbonyl), –COOR (ester), –COOH

(carboxyl), and –COH (hydroxyl) groups, as well as

unpaired electrons (dangling bonds). Figure 11 shows

schematically the adsorption of a polar perﬂuoro-

polyether (PFPE) (Z-Dol) molecule on a carbon sur-

face at 20 1C. The CQO bond and the hydroxyl

groups in the case of Z-Dol interact with functional

groups and dangling bonds on the carbon surface. The

Figure 10

Effect of coefﬁcient of friction, m, on number of bumps, contact area, and meniscus force for sombrero-, V-, and

W-type bumps at s ¼3 nm.

554

Magnetic Reco rding Devices: Head/Medium Interface

functional groups behave as an anchor, i.e., an im-

mobile fraction of the lubricant ﬁlm. The addition of

an alcohol (–OH) end group to a lubricant molecule

adds an interaction energy of the order of

20 kJmol

1

, even if no chemical reaction occurs with

the surface. The attachment of lubricant molecules

with or without functional groups is enhanced after

sliding as long as the lubricant chains are exposed to

the disk surface (Zhao et al. 1999).

The PFPE backbone mainly interacts with surfaces

through van der Waals forces. The van der Waals

interaction energy of PFPE molecules with a DLC

surface is of the order of 2 kJmol

1

or less. This en-

ergy is sufﬁciently high so that molecules lie ﬂat on

the surface when the lubricant ﬁlm thickness is com-

parable to the polymer chain diameter of 0.7 nm.

There are two approaches for chemisorption that

have been shown to be successful in chemical bonding

of the liquid monolayer to the overcoat. The ﬁrst re-

lies on exposure of the disk lubricated with neutral

PFPE to various forms of radiation, such as low-

energy x rays, electron or ion beam, nitrogen plasma,

or far UV. The improved afﬁnity of the lubricant for

the surface by obtained by UV radiation is believed

to result from enhanced bonding by photochemical

alteration of the lubricant, the substrate, or the air-

deposited organics. The PFPEs undergo dissociative

electron attachment caused by photoelectrons that

are generated by the interaction of the UV light with

the substrate. This dissociative electron attachment

results in the formation of a negative ion and a rad-

ical. It is suggested that subsequent radical propaga-

tion and termination steps cross-link the PFPE and

bond it to the substrate. Another approach uses

chemically active PFPE molecules through either the

above-mentioned treatments or a post-thermal treat-

ment, in which the various functional end groups

offer the opportunity of strong attachments to a

specific interface if a high enough temperature is ap-

plied to overcome the activity barrier (Zhao and

Bhushan 1996, Zhao et al. 2000). The activation en-

ergy of the thermal attachment of Z-Dol to a carbon

surface is of the order of 30 kJmol

1

. Surface clean-

liness and chemical activity of the overcoat affect

the degree of bonding. The length of the thermal

treatment and the temperature for PFPE lubricants

with polar end groups are important factors in de-

termining the adsorption level of lubricants to the

disk surface (Zhao and Bhushan 1996).

The lubricant ﬁlm thickness (h) and composite

roughness (s) are generally known to have an oppo-

site effect on static and kinetic friction and durability

in terms of desirability, i.e., an increase in h or a

decrease in s results in an increase in the values of

/W, F

/W, and durability. F

/W, F

/W, and dura-

bility data obtained on various disks with a lubricant

C (Z-Dol) as a function of h/s is shown in Fig. 12

(Bhushan and Zhao 1999). In Fig. 12, h represents a

mobile fraction of the total lubricant ﬁlm thickness in

static and kinetic friction because only the mobile

fraction forms menisci; h is a total ﬁlm thickness in

durability because both mobile and immobile frac-

tions affect wear life. The Z-Dol lubricants used are

untreated, partially bonded, and fully bonded. The

normalized static friction force (F

/W) and normal-

ized kinetic friction force (F

/W) increase rapidly

above a critical value of h/s (B0.7) and durability

decreases above the critical ﬁlm thickness. Trends in

experimental results for static friction force are com-

parable with the results of the numerical contact

model presented earlier (see Fig. 6(b)). A rapid de-

crease in durability above a critical ﬁlm thickness

occurs because of a large meniscus being formed

around slider edges and the presence of the stick/slip

phenomena (Bhushan and Zhao 1999).

Zhao et al. (2000) and Kajdas and Bhushan (1999)

have presented detailed discussion of various degra-

dation mechanisms of Z-Dol lubricant which include

thermal decomposition, catalytic degradation, elec-

tron-mediated degradation, and mechanical degra-

dation processes. Comparing various degradation

mechanisms it is believed that the catalytic degra-

dation mechanism proposed in the literature is not

relevant because kinetics is slow at asperity temper-

atures. Based on a wide variety of experimental data

it is reasonable to emphasize that anionic inter-

mediates (negative ions and/or negative-ion/radical

species) produced by low-energy electrons play an

important part in both (i) the electron-mediated deg-

radation process of PFPE lubricants, and (ii) chem-

ical bonding of PFPE lubricant ﬁlms with DLC

surfaces under sliding conditions.

3. Microtribology of Magnetic Storage Devices

At most interfaces of technological relevance, contact

occurs at multiple asperity contacts. Consequently,

the importance of investigating single-asperity con-

tacts in studies of the fundamental micromechanical

and tribological properties of surfaces and interfaces

has long been recognized. The emergence and prolif-

eration of proximal probes, in particular tip-based

microscopy (the atomic force microscope/friction

Figure 11

Schematic of Z-Dol molecules adsorbed on a carbon

surface at 20 1C.

555

Magnetic Reco rding Devices: Head/Medium Interface

microscope (AFM/FFM)), and of computer simula-

tions of tip/surface interactions and interfacial prop-

erties, provide nanoscale realizations of single-

asperity contacts, and allow systematic investigations

of interfacial problems of materials with high reso-

lution as well as ways and means for modiﬁcations

and manipulations of nanoscale structures. These

advances provide the impetus for intensive research

endeavors aimed at fundamental understanding of

the nature and consequences of the interactions

between materials on the atomic scale, and guide

the ‘‘rational’’ design of material for technological

applications. The advent of these techniques and

methodologies has led to the development of the ﬁeld

of micro/nanotribology and micro/nanomechanics

(Bhushan 1999a, Bhushan et al. 1995). This ﬁeld per-

tains to experimental and theoretical investigations of

processes ranging from the atomic and molecular

scales to the microscale, that arise as a consequence

of interfacial interactions.

3.1 Friction

Friction has been measured on a microscale (Bhush-

an 1999a), and it has been reported that the coefﬁ-

cient of friction on a microscale is lower than the

macrofriction. Table 1 gives typical values of coefﬁ-

cients of friction of metal-particle (MP) and metal-

evaporated (ME) tapes, poly(ethylene terephthalate)

(PET) tape substrate, and thin-ﬁlm rigid disks. Fric-

tion values on the microscale are much lower than

those on the macroscale, which is believed to be be-

cause of a lower plowing contribution (almost no

damage on the disk surface as measured). When

measured for the small contact areas and very low

loads used in microscale studies, mechanical proper-

ties are higher than at the macroscale, this reduces the

degree of wear. In addition, the apparent area of

contact reduces the number of particles trapped at

the interface, and thus minimizes the plowing contri-

bution to the friction.

Gray-scale plots of coefﬁcients of local friction of a

thin-ﬁlm disk as the FFM tip is scanned in either

direction are shown in Fig. 13. Corresponding gray-

scale plots of slope of roughness proﬁles are also

shown in Fig. 13. Since the sign of the friction and the

slope reverses with the change in the scanning direc-

tion, the sign of one set of images needs to be reversed

before data can be compared. The left-hand sections

correspond to the sample sliding from the left to-

wards the right, the center sections correspond to the

sample sliding from the right toward the left, and the

right-hand sections correspond to the ﬁrst set with

the sign of slope and friction values reversed. We note

a general correlation between the slope and friction

proﬁles. We note that generally the points that have

high friction in the left to right scan also have high

friction as the sliding direction is reversed (after sign

change). This relationship is not true at some loca-

tions. We observe some differences in the friction

proﬁles that may result from the asymmetrical as-

perities and/or asymmetrical transfer of wipe material

during manufacturing of the disk. The directionality

effect in friction on a macroscale has been observed in

some magnetic tapes (Bhushan 1999a).

3.2 Scratching and Wear

Microscratches made on MP tape and on an unlu-

bricated thin-ﬁlm rigid disk at various loads are

Figure 12

Normalized static and kinetic friction forces and

durability in revolutions as a function of the ratio of the

lubricant ﬁlm thickness to composite roughness (h/s).

Data on untreated, partially bonded, and fully bonded

Z-Dol lubricants are presented. For static and kinetic

friction data, h used is a mobile fraction; for durability, h

used is a total ﬁlm thickness (after Bhushan and Zhao

1999).

556

Magnetic Reco rding Devices: Head/Medium Interface

shown in Fig. 14. All scratches are made with 10

cycles. We notice that the scratch depth increases

with an increase in the normal load. Figure 15 shows

the microwear proﬁles at 20 mN load and at various

cycles of an unlubricated thin-ﬁlm disk. We note that

wear is not uniform and is largely initiated at the

texture marks present on the disk surface. This sug-

gests that surface defects act as initiated sites for

wear. Figure 16(a) shows the wear depth as a function

of the number of cycles for unlubricated and lubri-

cated thin-ﬁlm disks at 10 mN and 20 mN loads. Wear

Figure 13

Gray-scale plots of the slope of the surface roughness

and the friction force proﬁles for a lubricated disk with

FFM tip sliding in different directions. Higher points are

shown by lighter color (after Bhushan 1999a).

Figure 14

Surface proﬁles for scratched, unlubricated thin-ﬁlm

rigid disk and MP tape (after Bhushan 1999a).

Table 1

Surface roughness (r.m.s.), microscale and macroscale friction, and nanohardness data of thin-ﬁlm magnetic rigid

disk, magnetic tape, and magnetic tape substrate (PET) samples.

Roughness (r.m.s.) (nm)

Coefﬁcient of

microscale

friction

Coefﬁcient of

macroscale

friction

NOP

AFM

Sample

250 mm 

250 mm

1 mm 

1 mm

10 mm 

10 mm

1 mm 

1 mm

10 mm 

10 mm

Mn–Zn

ferrite

–

TiC

Nanohardness/normal

load (GPa)/(mN)

Unlubricated disk 2.2 3.3 4.5 0.05 0.06 0.26 21/100

Lubricated disk 2.3 2.3 4.1 0.04 0.05 0.19

Metal-particle tape 6.0 5.1 12.5 0.08 0.06 0.19 0.30/50

Metal-evaporated

tape

9.3 4.7 5.1 0.05 0.03 0.18 0.7 to 4.3/75

PET tape substrate 33 5.8 7.0 0.05 0.04 0.55 0.3/20 and 1.4/20

a Noncontact optical proﬁler. b Atomic force microscope. c Scan area. d Polymer and particulate regions, respectively.

557

Magnetic Reco rding Devices: Head/Medium Interface

initially takes place slowly with a sudden increase

after about 40 cycles at 10 mN and after about

10 cycles at 20 mN. The rapid increase is associated

with the breakdown of the carbon coating. Similar

behavior has been reported for ME tapes. Wear rates

for particulate tapes (Fig. 16(b)) and PET tape sub-

strates are approximately constant for various loads

and number of cycles (Bhushan 1999a).

We note that scratches and wear proﬁles can be

produced with very shallow depths, thus the AFM

technique can be used to measure scratch resistance

and wear resistance of ultrathin ﬁlms.

3.3 Lubrication

To study lubricant depletion during microscale meas-

urements, the microscale friction as a function of

Figure 15

Surface proﬁles of unlubricated thin-ﬁlm rigid disk

showing the worn region (center 2 mm 2 mm). The

normal load and the number of test cycles are indicated

(after Bhushan 1999a).

Figure 16

Wear depth as a function of number of cycles for

(a) unlubricated and lubricated thin-ﬁlm rigid disks at

10 mN and 20 mN loads, and (b) MP tape at a normal

load of 2 mN (after Bhushan 1999a).

558

Magnetic Reco rding Devices: Head/Medium Interface

number of cycles of a virgin Si(100) surface and a

silicon surface lubricated with about 2 nm thick Z-15

and Z-Dol PFPE lubricants are shown in Fig. 17.

These experiments were conducted in an AFM/FFM

by sliding a Si

tip against the samples. Z-Dol is a

PFPE lubricant with hydroxyl end groups. Its lubri-

cant ﬁlm was thermally bonded at 150 1C for 30

minutes and washed off with a solvent to provide a

chemically bonded layer of the lubricant ﬁlm. We

note from Fig. 17 that the friction force in a virgin

silicon surface decreases after a few cycles after the

natural oxide and other contaminant ﬁlm present

on the silicon surface gets removed. In the case of the

Z-15-coated silicon sample, the friction force is

initially low and then approaches that of an unlubri-

cated silicon surface after a few cycles. The increase

in friction of the lubricated sample suggests that the

lubricant ﬁlm gets worn and the silicon underneath is

exposed.

In the case of the Z-Dol-coated silicon sample, the

friction force starts out low and remains low during

the 100 cycles of the test. It suggests that Z-Dol does

not get displaced/depleted as readily as Z-15. Fur-

thermore, the initial friction value for Z-Dol is low as

compared to that of Z-15. Z-15 gets displaced locally

by condensed water molecules permeated through the

lubricant ﬁlm or condensed over exposed areas,

whereas Z-Dol does not get displaced as readily by

water because it is chemically bonded to the subst-

rate. The absence of mobile lubricant and lack of

water molecules minimize the possibility of formation

of meniscus bridges, which is responsible for low

friction in the case of Z-Dol. From this example, it is

clear that AFM methods can be used to study the

performance of lubricants in simulated single-asper-

ity contacts.

4. Closure

The magnetic recording process involves relative

motion between a magnetic medium and a station-

ary or rotating read/write magnetic head. Under

steady operating conditions, a load-carrying air ﬁlm

is formed. There is a physical contact between the

medium and the head during starting and stopping.

In modern high-end computer tape and disk drives,

the surfaces have roughness of the order of 2–5 nm

r.m.s. and the head-to-medium separation is of the

order of 25–50 nm. The need for higher recording

densities requires the surfaces be as smooth as pos-

sible and the ﬂying height be as small as possible. The

very smooth surfaces and ultralow ﬂying heights re-

quired present challenges and opportunities unparal-

leled in any other industry.

Analytical models and experimental data available

of the tribology of the head/medium interface are

plentiful. Understanding of tribology on nanoscales

is beginning to emerge. Understanding of micro/

nanotribology of ultrasmooth surfaces and small

components in future head/medium interfaces is lack-

ing. AFM/FFM studies are expected to play a major

role in computer tribology.

See also: Magnetic Recording: Rigid Media, Tribo-

logy; Magnetic Recording Technologies: Overview

Bibliography

Bhushan B 1993 Magnetic slider/rigid disk substrate materials

and disk texturing techniques—status and future outlook.

Adv. Info. Storage Syst. 5, 175–209

Bhushan B 1996a Tribology and Mechanics of Magnetic Stor-

age Devices, 2nd edn. Springer, New York

Bhushan B 1996b Contact mechanics of rough surfaces in

tribology: single asperity contact. Appl. Mech. Rev. 49,275–98

Bhushan B 1996c Methodology for roughness measurement

and contact analysis for optimization of interface roughness.

IEEE Trans. Magn. 32, 1819–25

Bhushan B 1998 Contact mechanics of rough surfaces in

tribology: multiple asperity contact. Tribol. Lett. 4, 1–35

Bhushan B (ed.) 1999a Handbook of Micro/Nanotribology 2nd

edn. CRC Press, Boca Raton, FL

Bhushan B 1999b Principles and Applications of Tribology.

Wiley, New York

Bhushan B 1999c Chemical, mechanical and tribological char-

acterization of ultra-thin and hard amorphous carbon coat-

ings as thin as 3.5 nm: recent developments. Diamond Relat.

Mater. 8, 1985–2015

Bhushan B 2000 Mechanics and Reliability of Flexible Magnetic

Media, 2nd edn. Springer, New York

Bhushan B, Chilamakuri S K 1996 Non-Gaussian surface

roughness distribution of magnetic media for minimum fric-

tion/stiction. J. Appl. Phys. 79, 5794–6

Bhushan B, Israelachvili J N, Landman U 1995 Nanotri-

bology: friction, wear and lubrication at the atomic scale.

Nature 374, 607–16

Bhushan B, Zhao Z M 1999 Macroscale and microscale tribo-

logical studies of molecularly thick boundary layers of

Figure 17

Friction force as a function of number of cycles for

unlubricated and lubricated silicon samples (after

Bhushan 1999a).

559

Magnetic Reco rding Devices: Head/Medium Interface