hard yet tough nanocomposite coatings – present status and future trends

Full Paper

Hard yet Tough Nanocomposite Coatings –Present Status and Future Trends

Sam Zhang,* Hui Li Wang, Soon-Eng Ong, Deen Sun, Xuan Lam Bui

Toughness and hardness are important aspects for coating applications in manufacturingindustry. Extensive theoretical and experimental efforts have been made to synthesize andstudy nanocomposite coatings with super hardness and high toughness. The materials can behardened through various or combined hard-ening mechanisms. However, for engineeringapplications, coating toughness is as importantas, if not more than, super hardness. At present,there is neither a standard test procedure, nor astandalone methodology for the assessment ofthin film toughness. The determination of thetoughness is still a difficult task, and verymucha fully open problem. In this article, we reviewthe hardening and toughening mechanismsof nanocomposite films, and the toughnesscharacterization techniques. Based on thesereviews, an outlook will be presented in theconcluding remarks.

Introduction

Nanostructured or nanocomposite coatings are a new

branch of materials that possess unique physical and

mechanical properties. A nanocomposite coating com-

prises of at least two phases: a nanocrystalline phase and

an amorphous phase, or two different nanocrystalline

phases. These nanocomposite coatings represent a new

class of materials, whose mechanical and tribological

properties are not subjected to volume mixture rules, but

depend on grain boundary effects, and on synergetic

interactions of the composite constituents owing to the

size effect.[1] Now, it is accepted that materials can be

S. Zhang, H. L. Wang, S. E. Ong, D. Sun, X. L. BuiSchool of Mechanical and Aerospace Engineering, NanyangTechnological University, 50 Nanyang Avenue, Singapore 639798,Republic of SingaporeFax þ65 6791 1859; E-mail: [email protected]

Plasma Process. Polym. 2007, 4, 219–228

� 2007 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

classified as hard, superhard or ultrahard for hardness over

20, 40 or 80 GPa, respectively.[2] Hard, superhard or even

ultrahard nanocomposites have been a hot research topic

recently.[2–10] Hard coatings are used inmany applications,

for example, cutting and polishing tools, molds, dies, hard

disk and other wear-resistant applications. However, for

engineering applications, hardnessmust be complimented

with high toughness, which is a property of equal impor-

tance as hardness.[11–13] Therefore, it is vital to master the

formation of hard filmswith high toughness (see Figure 1).

Toughness is an important mechanical property related

to the materials resistance against shock loads, and it

describes the resistance against the formation of cracks

resulting from stress accumulation in the vicinity of

structure imperfections. In an energetic context, toughness

is the ability of a material to absorb energy during

deformation up to fracture.[11] According to this definition,

toughness encompasses the energy required to create a

crack and to enable the crack to propagate until fracture.

DOI: 10.1002/ppap.200600179 219

S. Zhang, H. L. Wang, S. E. Ong, D. Sun, X. L. Bui

Figure 1. Classification of nanocomposites according to theirhardness and toughness.[14]

220

Therefore, a high toughness coating has high resistance to

formation of cracks under stress, and in the mean time,

high energy absorbance to deter crack propagation,

whereby preventing chipping, flaking or catastrophic

failure. Creating tough, fracture-resistant ceramics has

been a focal point of ceramics research for decades.

Ceramics have inherently low fracture toughness, and are

thus subject to brittle fracture (in contrast with ductile

metals, which are much tougher because of the plasti-

city induced by dislocation motion). It is known that

mechanical properties of a solid depend strongly on the

density of dislocations, interface-to-volume ratio and

grain size.[14]

An enhancement in damping capacity of a nanocompo-

site solidmay be associatedwith grain-boundary sliding,[15]

or with energy dissipation mechanism localized at inter-

faces.[16] A decrease in the grain size significantly affects the

yield strength and hardness. The grain boundary structure,

boundary angle, boundary sliding and movement of

dislocations are important factors that determine the

mechanical properties of the nanocomposites. In most

cases, coatings are applied to protect surfaces from the

consequences of mechanical loads. A greatly simplified

examination of the abrasive wear resistance in a variety of

materials as a function of the fracture toughness (as applied

to bulk materials) reveals a dependence, namely the

maximumwear resistance results from a specific, favorable

combination of hardness and toughness.[17]

Design Methods for Hard yet ToughNanocomposites

Hardening Mechanisms in Nanocomposite Coatings

Hardness is defined as the resistance of a material to

plastic deformation. Plastic deformation of crystalline



materials occurs predominantly by dislocation movement

under applied load. It means that a material with en-

hanced hardness has a higher resistance to dislocation

movement. Very often, several of the classical hardening or

strengthening mechanisms are active in hard coatings

deposited by plasma-assisted vapor deposition techniques.

Such hardening mechanisms work by providing obstacles

for the dislocation movements, which can also be applied

to some extent to hard films. Hindering of dislocation

movements can be achieved by: (i) grain boundary harden-

ing, (ii) solid solution hardening, (iii) age hardening and (iv)

compressive stress hardening.

Grain Boundary Hardening

In a recent review,[18] Veprek has analyzed the design

criteria to produce superhard or even ultrahard nanocom-

posite coatings. From material-selection point of view,

combinations of nanocrystalline transition-metal nitrides,

for example TiN, W2N and VN, with amorphous Si3N4 or

BN as the grain boundary phase, have a potential to

achieve coating hardness exceeding 40–50 GPa. This

hardening behavior, especially in the regime where the

crystallite size d is less than 10 nm, is because the tiny

crystallites will restrict the grain boundary sliding in case

of a thin amorphous grain boundary.[2] One of the best

known theories based on dislocation pile-up is described

by the Hall-Petch equation.[19,20] The relationship indicates

that the hardness of the material is inversely proportional

to the square root of the grain size. However, as grain sizes

are reduced to the nanometer scale, and the percentage of

grain boundary atoms increases correspondingly, this

traditional view of dislocation-driven plasticity in poly-

crystalline metals needs to be reconsidered. In a sample

with grain diameters of 20 nm, 10% of atoms are located at

grain boundaries. Dislocation sources and pileups are

hardly expected to exist in such a material, and deforma-

tion is believed to be carried mostly by the grain

boundaries via a certain accommodation mechanism.[21]

For very small grains, the numbers of atoms at

inter-granular boundaries and inside grains become

comparable. In this case, softening may happen via

inter-granular slips, as the inter-granular boundaries cease

to be barriers for the motion of dislocations.[22] The search

for very hard materials is coupled with the study of

low-compressibility solids, which have high values of

the bulk (Kb) or shear (G) modulus. Transition-metals-

containing compounds have large cohesive energies and

high Kb, associated with the distribution of their valence

electrons between bonding and anti-bonding states

within the partly filled electronic bands. Transition metal

carbides, nitrides, borides, or oxides have very low

compressibility and also often possess high hardness.

Based on thermodynamically driven spinodal segregation

in a binary or ternary system, Veprek et al. proposed a

DOI: 10.1002/ppap.200600179

Hard yet Tough Nanocomposite Coatings – Present Status and Future Trends

design to achieve superhard or even ultrahard coatings;

a highly thermal stable nanocomposite structure with

�10 nm size crystallites of the transition metal nitride

separated by about one monolayer of a-Si3N4 was

synthesized (see ref.[2] and references therein). The

approximately one monolayer thin tissue of Si3N4 acts

as a ’glue’ for the MeN nanocrystals to achieve ultra-

hardness, thus avoiding the reverse Hall-Petch effect.

Though the original value given was 103 GPa, more

recently studies show that the Fischerscope used normally

gives too high hardness values for superhard coatings.[23]

Nevertheless, the design concept should be on the right

track. Subsequently, many superhard nanocomposites

with similar design have been synthesized, but none

approaches the hardness of diamond. Figure 2 shows the

hardness of nc-TiN/a-Si3N4 coatings deposited by close-

field unbalancedmagnetron sputtering as a function of the

silicon nitride fraction. Silicon-free TiN forms elongated

crystallites that are several hundreds of nanometers long

and tens of nanometers wide (inset A in Figure 2).[24,25] TiN

crystallites become very small at increased Si3N4 contents.

At 15–20 at.-% Si3N4, the mean grain size of TiN does not

exceed 7 nm; this size is too small for dislocation activities.

Therefore, under mechanical loading, such a material can

react only by grain-boundary sliding (i.e., by moving

single, undeformed TiN nanocrystallites against one

another). This process requires more energy than deforma-

tion by dislocation movement; hence a higher hardness

can be achieved for such coating structures. Estimations of

the mean grain separation at this composition show that

only a few monolayers of silicon nitride separate the

nanocrystallites (inset B in Figure 2). At high silicon nitride

fractions, themean grain separation becomes sowide, that

ordinary crack propagation in Si3N4 takes place, and the

hardness approaches that of Si3N4 (inset C in Figure 2).

Figure 2. Hardness of nc-TiN/a-Si3N4 as a function of siliconnitride fraction.[24,25] Insets illustrate the schematic nanostruc-ture for different compositions.



Solid Solution Hardening

Solid solution hardening comes from the lattice distortion

as a result of insertion of atoms of an alloying element in

the interstitial location or substation for some of the host

atoms. This is, perhaps, one of the oldest hardening

methods used in bulk materials. In thin films or coatings,

the same principle works too. TiCN,[26] TiAlN,[27] CrAl-

N,[28]and CrZrN[29] are a few popular examples. Formation

of a non-equilibrium supersaturated solid solution of B in

TiN may contribute to further strengthening of the

material,[30] since the gliding of the dislocations eventually

formed inside the crystallites is hindered by the strain

exerted by the insertion of B. To increase the solubility of B

in TiN, non-equilibrium growth processes, such as physical

vapor deposition (PVD), or plasma-assisted chemical vapor

deposition (PACVD) have been used, where a maximum B

solubility of 17.4 at.-% has been reported for PACVD[30]

with hardness up to 43 GPa. For reactive arc evaporation

from Ti-B compound targets, the highly ionized flux of

film-forming species was utilized to synthesize a promis-

ing nanocrystalline metastable supersaturated solid solu-

tion of B in TiN at lower nitrogen fractions.[31] The

maximum in hardness (34.5 GPa) is obtained when the

crystallites are 6–8 nm in size, and are strained by

the formation of a substitutional solid solution of B in TiN.

The formation of a metastable solid solution was

confirmed by XRD-observed widening of the lattice upon

B substitution of N as compared to TiN. Therefore, the

mobility of deformation-induced dislocations is hindered

in these small crystallites that are initially dislocation-free.

At high N2 fractions, the substitutional B sites in the TiN

crystallites are replaced with additional N to form

energetically favorable unstrained TiN crystals. This

unstrained TiN crystals combined with high fraction

Figure 3. Indentation hardness and reduced modulus values forTi-B-N coatings for different N2 fractions.[31]

www.plasma-polymers.org 221


222

of less strongly bonded BN amorphous phase lead to a

hardness drop (Figure 3). Another example is the super-

hard Ti-Al-Si-N film (maximal hardness� 55 GPa) synthe-

sized by a hybrid coating system of arc ion plating and

sputtering method.[32] Based on XPS and XRD results, it

was suggested that the Ti-Al-Si-N films were a composite

consisting of solid-solution (Ti,Al,Si)N crystallites and

amorphous Si3N4.

Age Hardening

Age hardening is yet another ‘‘aged’’ hardening mechan-

ism in strengthening bulk materials (for example Al alloys

and steels). Age hardening achieves higher hardness

through precipitation over time of a small and uniformly

distributed strengthening phase in the form of particles

from a supersaturated solid solution. This works equally

well in hard coatings, as demonstrated in ref.[33–35]

Supersaturated phases can be easily obtained by

non-equilibrium deposition techniques such as PVD or

PACVD due to limited atomic kinetics during deposition.

As-deposited (TiAl)N[34] and Ti(BN)[35] films show a dense

columnar microstructure of a supersaturated NaCl struc-

tured phased based on fcc TiN whose Ti location is partly

substituted by Al to form (TiAl)N, and N is partly

substituted by B to form Ti(BN). During annealing, (TiAl)N

and Ti(BN) coatings undergo spinodal decomposition and

transform into coherent cubic-phase nanometer-size

domains, resulting in hardness increase.

Compressive Stress Hardening

Ion bombardment during deposition at low temperatures

has been utilized to increase the density and modify the

morphology of the films.[36] During growth, stresses are

generally induced by energetic particle bombardment. The

lattice defect arrangements induced by impinging ener-

getic ions are responsible for the stress. As the incoming

ion or knock-on atoms possess enough kinetic energy, they

knock atoms out of their lattice positions and create

secondary collisions. In this way, collision cascades.

The atom first knocked absorbs most of the energy. The

resulting strong atomic motion along the trajectory of

the ions leads to a rearrangement of the lattice atoms.[37]

The hardness enhancement results from a complex

synergistic effect of the decrease of crystallite size,

densification of grain boundaries, and compressive stress

upon ion bombardment. Musil et al. reported enhance-

ment of hardness of TiN (up to 80 GPa) and (TiAlV)N (up to

100 GPa) during deposition by means of unbalanced

magnetron sputtering at negative substrate bias,[38] where

a correlation between the hardness enhancement and

biaxial compressive stress exists.[39] However, upon

annealing at a certain temperature, the hardness may

decrease back to ‘‘normal’’ when the induced defects are

annealed and the compressive stress relaxes.



Toughening Mechanisms in Nanocomposite Coatings

In order to obtain high hardness in nanocomposite

coatings, usually plastic deformation is designed to be

prohibited, and dislocationmovement and grain boundary

sliding are prevented, thus causing a loss in ductility.

Ductility is related to toughness, which is very important

for hard coatings to avoid catastrophic failure. Veprek[40]

indented a 6–10 mm-thin coating on a soft steel substrate

with a high load of 1 000 mN and found no cracks. This,

however, does not necessarily mean it has high ‘‘fracture

toughness’’: the classical definition of ‘‘fracture toughness’’

measures the resistance to crack propagation – a crack

exists first, then how difficult it is to make the crack

propagate. The coating in the above example is ‘damage

tolerant’ owing to the difficulty in generating a crack, but

once a crack is generated, it might propagate in a catastro-

phic manner and thus result in a brittle failure if it is not

truly ‘‘tough’’. In bulk ceramics, various toughening

methods are available: ductile phase toughening, fiber

and whisker toughening, transformation toughening,

microcrack toughening, etc.[41] In hard coatings, however,

investigation on toughening mechanisms lags far behind.

Ductile Phase Toughening

In order to toughen hard ceramic films, a straightforward

method is to incorporate a ductile phase. Ductile phase

toughening arises from two major mechanisms: the

relaxation of the strain field around the crack tip through

the ductile phase, and the yielding and bridging of cracks

by ligaments of the ductile phase. These two mechanisms

can increase the work for plastic deformation. Zhang

et al.[42] doped Al into a-C films by co-sputtering of

graphite and Al. The composite films exhibit a reduction in

the residual stress and also an improvement in toughness.

However, this came at huge expense of hardness, where

the hardness determined by nanoindentation dropped

from 31.5 to 8.8 GPa. In order to restore the hardness,

nanocrystalline TiC (or nc-TiC) was embedded into the a-C

matrix doped with Al to form nc-TiC/a-C(Al) nanocompo-

site film. The hardness is restored to �20 GPa, while

toughness remained high (indentation plasticity of 55%)

and residual stress low (only 0.5 GPa).[43] Although

toughness is not measured due to lack of proper

measurement facilities for such a high toughness coating,

the nc-TiC/a-C(Al) nanocomposite film is evidently very

tough: the optical comparison of scanning scratch tracks

between nc-TiC/a-C (Figure 4a) film and nc-TiC/a-C(Al)

(Figure 4b) shows that without doping of Al, the coating

basically fails in a brittle way (Figure 4a); with doping of

Al, however, the scratching tip ploughs into the coating

causing the scanning amplitude to reduce along the

scratch direction (Figure 4b). Co-sputtering of Ti, TiNi

DOI: 10.1002/ppap.200600179


Figure 5. Comparison of Vicker’s indenter impressions intoZr-Cu-O coatings with different residual stress: a) 0.3 GPa;b) –0.1 GPa; and c) –0.3 GPa.[50]

Figure 4. Optical image of scratch tracks on: a) nc-TiC/a-C andb) nc-TiC/a-C(Al) coatings.[43]

and Si3N4 targets produced nanocomposite coatings where

the amorphous metallic Ni increased the toughness of

nc-TiN/a-SiNx coating.[44] Doping from 0 to �40 at.-% Ni

brings about an increase in toughness from 1.15 to

2.60 MPa �m1/2, but at the expense of hardness, which

diminishes from 30 to 14 GPa.

Phase Transformation Toughening

Some phase transformation occurs under stress and is

accompanied by absorption of a large quantity of energy,

which could be used to toughen a material. Transforma-

tion from tetragonal to the monoclinic phase takes place

under stress. The process occurs in the stress field around

the tip of the crack and is accompanied by a volume

increase of about 4%. The resultant strain during the

transformation relieves the stress field and absorbs the

fracture energy, whereby toughens the material. One

example is ZrO2-toughened ZrB2 composite.[45] In order to

facilitate the phase transformation, retention of the

high-temperature tetragonal phase is the key, which can

be easily realized during deposition. Sprio et al.[46]

deposited thin zirconium oxide films by radiofrequency

(RF) magnetron sputtering. The films have a mixture of

tetragonal and monoclinic zirconia phases. It was shown

that increasing the substrate bias power disrupted the

columnar grain growth. TEM confirmed a reduction

in the intergranular porosity, but also an increase in

lateral defects.[47] These defects are hypothesized to be

stress-induced microcracks caused by a tetragonal to



monoclinic phase transformation. Tetrahedral zirconia has

a high toughness and is proposed to increase properties of

brittle substrates.[48] Trinh et al.[49] prepared g-Al2O3-ZrO2

thin films by RFmagnetron sputteringwith hardness up to

30 GPa. However, owing to the lack of proper toughness

measurement instrumentation, toughness was not mea-

sured.

Compressive Stress Toughening

Generally cracking is initiated by tensile stress, thus

compressive residual stress has to be overcome first before

a crack is initiated in tension. When a coating has a high

compressive residual stress to start with, the coating will

be able to take more tensile strain before fracture, as is

demonstrated in Zr-Cu-O coatings (see Figure 5).[50]

Although a certain level of compressive stress can increase

toughness, excessively large compressive stress can cause

film delamination or cracking.[51]

Nanotube Toughening

The discovery of carbon nanotube (CNTs) has opened

up new avenues in producing unique carbon-based

materials. Novel mechanical tests on individual

CNTs[52–54] and atomistic calculations[55–58] suggest that

CNTs have ultra-high elastic modulus approaching 1 TPa,

and exceptional tensile strengths, in the range of

20–100 GPa. Due to their outstanding properties, carbon

nanotubes have attracted a growing interest and are

considered to be the most promising materials for

applications in nanoengineering.[11,59–65] Toughness

enhancement can be obtained through crack deflection,

crack bridging and CNT pull-out. CNT toughened bulk

ceramics are attractive, and outstanding results have been

successfully achieved, for instance, in Al2O3 matrix,[65] and



224

BaTiO3 matrix.[66] However, nanotube toughened thin

films is yet to be realized, due to difficulties in controlled

alignment, interface reaction and reasonable volume

fractions. Xia et al.[63,67] anodized aluminum in order to

Figure 6. Micrographs of deformations in CNT-alumina compo-sites fabricated using ordered porous alumina templates:a) transverse crack deflection at the CNT/matrix interface;b) CNTs bridging longitudinal matrix cracks; c) nanotube pullout,and subsurface nanotube bridging.[63,67]



obtain an amorphous alumina matrix having pores with

diameters in the 30–40 nm range. Multiwall carbon

nanotubeswere grown by CVD in this matrix to realize CT/

alumina composite layers of 20 to 90 mm in thickness.

Crack deflection at the CNT/matrix interface (Figure 6a),

crack bridging by CNT (Figure 6b), CNT pull-out (Figure 6c)

are all observed.

Composition or Structure Grading Toughening

Graded inter-layering is a known effective method to

reduce cracks concentration and enhance adhesion be-

tween coating and substrate. In the gradient design, the

substrate is first coated with a high adhesion layer, and

then the coating constituents are allowed to vary homo-

geneously or heterogeneously while the coating thickness

builds up. With biased graded (varying substrate bias

voltage while coating thickness builds up) in magnetron

sputtering, Zhang et al.[68] prepared a 1.5 mm a-C gradient

coating on tool steels with moderately high hardness

(25 GPa), but very high toughness (plasticity 57.6%). During

deposition, substrate bias gradually increased from –20 to

–150 V, and a graded sp2/sp3 was achieved through

coating thickness. The bottom layer has the highest sp2

fraction for high adhesion, whereas the top layer posses

the highest sp3 fraction to render high hardness.

Bias-graded coating design enhances the toughness and

the adhesion of the coating on tool steel (Figure 7). The

adhesion strength increased more than two times as

compared to the same coating deposited at constant bias

(–150 V). Pei et al.[69,70] changed the substrate bias in order

to change the microstructure of the nc-TiC/a-C:H coatings

prepared by close field unbalanced magnetron sputtering.

Increasing the bias resulted in a clear transition from

columnar to glassymicrostructure in the coating (Figure 8).

Substrate bias also simultaneously and greatly enhan-

ced the coating hardness, toughness and adhesive

strength. The coating with glassy microstructure exhibits

Figure 7. Scratch adhesion strength of the non-hydrogenatedDLCs deposited under various constant bias or bias-graded con-ditions.[68]

DOI: 10.1002/ppap.200600179


Figure 8. SEM micrographs showing the fractured cross-sectionsof nanocomposite coatings deposited at different substrate bias:a) floating, b) �60, c) �100, and d) �150 V.[69]



a substantial toughening. However, a word of caution on

the actual toughness value given in the paper – the value

of few tens of MPa �m1/2 is unbelievably high, as also

pointed out by the authors, probably due to experimental

limitations. Ion bombardment can densify the structure

and prevent the formation of columns which are detri-

mental in terms of microcrack initiation and propagation

under load. The induced residual stresses grow as bias

increasing.

Toughness Characterization

In a classical definition, toughness measures the ability of

amaterial to resist the crack propagation until fracture. For

bulk materials and very thick films, toughness can be

easily measured according to ASTM standards.[71,72] For

thin films, there is no proper definition for toughness, since

in thin films or coatings, the cracks do not necessarily

pre-exist, and thus energy has to be used to initiate the

crack. Thin film toughness measurements are difficult or

tedious in practice due to the thickness limitations.

Although many proposed methods can be used in some

cases, a standard procedure or a universally accepted

methodology does not exist yet for ceramic thin film or

coating toughness measurements. Qualitatively, tough-

ness can be simply estimated by the plasticity[73,74], the

MDP (microhardness dissipation parameter)[75,76], and the

scratch crack propagation resistance.[77,78] The plasticity or

MDP are measured as the ratio of the plastic over the total

displacement or work. The scratch crack propagation

resistance, or CPR, is determined by measuring the lower

critical load, i.e. cohesive failure load Lc1 and the upper

critical load or adhesion failure load Lc2 and then calculated

by CPR¼ Lc1 (Lc2� Lc1 ). The CPR measures the scratch

resistance to crack initiation and propagation. More

detailed quantitative measurement of toughness via

bending, buckling, indentation and tensile methods can

be found in ref.[78] The most widely method used in the

assessment of thin film toughness is indentation. When

the stress exceeds a critical value during indentation, a

crack or spallation is generated. Failure of the coating is

manifested by the formation of a kink or plateau in the

load-displacement curve, or by the crack formation in the

indent impression. The length, c, of radial cracks is related

to the fracture toughness KIC through the following:[79]

KIC ¼ dE

H

� �1=2 P

c3=2

� �(1)

where P is the applied indentation load; E and H are the

elastic modulus and the hardness of the coating,

respectively. d is an empirical constant which depends

on the geometry of the indenter. Major uncertainties in



226

this method come from the crack formation threshold, the

substrate effect and the crack length measurement.

Another approach based on indentation is through

determining the fracture energy. The energy difference

before and after through-thickness cracking is believed to

be responsible for the coating fracture. The fracture

toughness of the film is calculated by:[80]

Tab

Ma

TiC

TiN

Ni-

YSZ

Na

Ni/

ta-

Plasma

� 2007

KIC ¼ E

ð1� y2f Þ2pCR

DU

t

� �1=2

(2)

where E is the elastic modulus of the coating, nf is Poisson’s

ratio of the coating, 2p CR the planar crack perimeter, t the

effective coating thickness, and DU is the strain energy

difference responsible for the cracking. This method also

inherited the uncertainties for the substrate effect on the

deformation fields, the crack size measurement and

negligent of the interface debonding.

Tensile testing of free-standing thin films with a

pre-crack induced by indentation was recently used to

measure the toughness of 0.52 and 1 mm tetrahedral

amorphous diamond-like carbon (ta-C).[81] During pre-

paration of the stand-alone film, wet etching of SiO2

interlayer between the coating and silicon substrate has

been employed. This method has difficulties in obtaining

freestanding film without curling, precrack sharpness and

accurate measurement of the precrack dimension. A

simple two-step tensile method, based on the strain

energy difference between two tensile processes is

proposed by Zhang et al.[82] The ‘‘two-step’’ tensile method

extends a very elastic substrate coated with the ceramic

film in question until the coating fracture. A kink is

observed in load-extension curve. The load is then released

to zero but re-loaded to the previous extension. The energy

difference is considered to be the energy used in creating

the fracture. In this method, nc-TiN/a-SiNx nanocomposite

layer with thickness of 3.0 mm gives a toughness value of

2.6 MPa �m1/2. The method requires very a good elasticity

for the substrate (so that when the coating is fractured the

le 1. Toughness values for some hard coatings.

terials Toughness

MPa �m1/2

xNy/SiCN 1.0

/SiNx 1.15

doped nc-TiN/a-SiNx 2.60

/Au 1.6

nolayered Cr/a-C 1.81–3.49

Al2O3 2.4

C 4.25–4.4

Process. Polym. 2007, 4, 219–228

WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim

substrate is still in the elastic deformation regime), and

that the thickness of the substrate to be small enough (for

the kink to be visible); plus, there should not be a coating

delamination during tension. Currently, there are rela-

tively few toughness data available for thin nanocompo-

site films due to the difficulties in forming a sensible crack

while reducing the substrate effect in the thin films.

Toughness values of some nanocomposite are listed in

Table 1.

Future Trends and Directions

For engineering applications, the zest for sole super-

hardness will subside, while a combination of hardness

and toughness will be pursued more by the academics

driven by the industrial needs. A sensible route is to

toughen superhard coatings in order to achieve what we

call ‘‘hard yet tough’’ coatings. In principle, all toughening

mechanisms can be applied to the toughening of ceramic

thin films or coatings, the real challenge lies in the

implementation. In-situ synthesis of carbon nanotu-

be-imbedded nanocomposite (such as nc-TiC/a-C(CNT) or

nc-TiN/a-Si3N4(CNT)) or nanocrystals imbedded in a

complete matrix of carbon nanotubes (for instance,

nc-TiC/CNT, or even nc-Diamond/CNT) may hold the key

(at least, one of the keys) to this ideal ‘‘hard yet tough’’

coating, where the CNT-imbedded matrix or the CNT

matrix provides the ultimate ‘‘bulk’’ elasticity and tough-

ness to stop or block crack propagations, and the

nanocrystalline phase provides the superior hardness.

In respect to the toughness measurements, the lack of a

universally accepted toughness measurement methodo-

logy and instrumentation for thin films or coating has

effectively hindered the progress of the research. It is

hoped that in a not too far future, a universally accepted

methodology will be born, and that this will accompany

the birth of a toughness measurement apparatus for

ceramic thin films or coatings; a stand-alone machine, or

Evaluation Method Reference

indentation [83]

indentation [44]

two-step tensile [82]

indentation [84]

indentation [85]

indentation [86]

tensile with precrack [81]

DOI: 10.1002/ppap.200600179


perhaps, as an integral part of a hardness testing system.

To catalyze the birth of the universal toughness measure-

ment for ceramic thin films or coatings, the community

needs first to have a consensus on the definition of the

toughness or ‘‘fracture toughness’’ for thin films or

coatings (which we believe it should take into account

the energy needed to generate the cracks), and regular

international seminars should be organized especially on

this topic.

Received: September 27, 2006; Revised: December 19, 2006;Accepted: December 21, 2006; DOI: 10.1002/ppap.200600179

Keywords: coatings; hardening; nanocomposites; superhard;toughening

[1] A. I. Gusev, Usp. Fiz. Nauk 1998, 168, 55.[2] S. Veprek, J. Vac. Sci. Technol. A 1999, 17, 2401.[3] S. Veprek, M. Jilek, Vacuum 2002, 67, 443.[4] J. Musil, Surf. Coat. Technol. 2000, 125, 322.[5] S. Zhang, D. Sun, Y. Fu, H. Du, Surf. Coat. Technol. 2003, 167,

113.[6] S. Veprek, S. Reiprich, S. H. Li, Appl. Phys. Lett. 1995, 66, 2640.[7] S. Veprek, M. Haussmann, S. Reiprich, J. Vac. Sci. Technol. A

1996, 14, 46.[8] P. Karvankova, H.-D. Mannling, C. Eggs, S. Veprek, Surf. Coat.

Technol. 2001, 146–147, 280.[9] P. Karvankova, M. G. J. Veprek-Heijman, O. Zindulka,

A. Bergmaier, S. Veprek, Surf. Coat. Technol. 2003, 163–164,149.

[10] I. W. Park, K. H. Kim, A. O. Kunrath, D. Zhong, J. J. Moore,A. A. Voevodin, E. A. Levashov, J. Vac. Sci. Technol. B 2005, 23,588.

[11] S. Zhang, D. Sun, Y. Fu, H. Du, Surf. Coat. Technol. 2005, 198, 2.[12] J. Musil, P. Baroch, J. Vlcek, K. H. Nam, J. G. Han, Thin Solid

Films 2005, 475, 208.[13] A. A. Voevodin, J. S. Zabinski, C. Muratore, Tsinghua Sci.

Technol. 2005, 10, 665.[14] J. Musil, Acta Metall. Sin. (Engl. Ed.) 2005, 18, 433.[15] B. S. Berry, W. C. Pritchet, Thin Solid Films 1976, 33, 19.[16] C. M. Su, R. R. Oberle, M. Wuttig, R. C. Cammarata, Mater. Res.

Soc. Symp. Proc. 1993, 280, 527C.[17] H. Holleck, M. Lahres, P. Woll, Surf. Coat. Technol. 1990, 41,

179.[18] S. Veprek, M. G. J. Veprek-Heijman, P. Karvankova,

J. Prochazka, Thin Solid Films 2005, 476, 1.[19] E. O. Hall, Proc. Phys. Soc. B 1951, p. 747.[20] N. Petch, J. Iron Steel Inst. 1953, 174, 25.[21] H. Van Swygenhoven, Science 2002, 296, 66.[22] J. Schiotz, F. D. Di Tolla, K.W. Jacobsen,Nature 1998, 391, 561.[23] A. C. Fischer-Cripps, P. Karvankova, S. Veprek, Surf. Coat.

Technol. 2006, 200, 5645.[24] J. Patscheider, Mater. Res. Soc. Bull. 2003, 28, 180.[25] M. Diserens, J. Patscheider, F. Levy, Surf. Coat. Technol. 1999,

120–121, 158.[26] J. E. Sundgren, H. T. G. Hentzell, J. Vac. Sci. Technol. A 1986, 4,

2259.[27] F. H. W. Loffler, Surf. Coat. Technol. 1994, 68–69, 729.[28] J. C. Sanchez-Lopez, D. Martinez-Martinez, C. Lopez-Cartes,

A. Fernandez, M. Brizuela, A. Garcia-Luis, J. I. Onate, J. Vac.Sci. Technol. A 2005, 23, 681.



[29] G. Kim, B. Kim, S. Lee, J. Hahn, Surf. Coat. Technol. 2005, 200,1669.

[30] P. H. Mayrhofer, M. Stoiber, C. Mitterer, Scr. Mater. 2005, 53,241.

[31] J. Neidhardt, Z. Czigany, B. Sartory, R. Tessadri, M. O’Sullivan,C. Mitterer, Acta Mater. 2006, 54, 4193.

[32] I. W. Park, S. R. Choi, J. H. Suh, C. G. Park, K. H. Kim, Thin SolidFilms 2004, 447, 443.

[33] P. H. Mayrhofer, C. Mitterer, L. Hultman, H. Clemens, Prog.Mater. Sci. 2006, 51, 1032.

[34] H. M. Paul, H. Anders, K. Lennart, S. Jacob, L. Tommy,M. Christian, H. Lars, Appl. Phys. Lett. 2003, 83, 2049.

[35] J. G. Wen, P. H. Mayrhofer, C. Mitterer, J. E. Greene, I. Petrov,Microsc. Microanal. 2006, 12, 720.

[36] D. M. Mattox, J. Vac. Sci. Technol. A 1989, 7, 1105.[37] R. W. Cahn, P. Haasen, ‘‘Physical Metallurgy’’, North-Holland,

Amsterdam, New York 1996.[38] J. Musil, S. Kadlec, J. Vyskocil, V. Valvoda, Thin Solid Films

1988, 167, 107.[39] P. H. Mayrhofer, F. Kunc, J. Musil, C. Mitterer, Thin Solid Films

2002, 415, 151.[40] S. Veprek, S. Mukherjee, P. Karvankova, H. D. Mannling,

J. L. He, K. Moto, J. Prochazka, A. S. Argon, J. Vac. Sci.Technol. A 2003, 21, 532.

[41] J. D. Kuntz, G. D. Zhan, A. K. Mukherjee, Mater. Res. Soc. Bull.2004, 29, 22.

[42] S. Zhang, X. L. Bui, Y. Fu, Thin Solid Films 2004, 467, 261.[43] S. Zhang, X. L. Bui, X. T. Zeng, X. Li, Thin Solid Films 2005, 482,

138.[44] S. Zhang, D. Sun, Y. Fu, Y. T. Pei, J. T. M. De Hosson, Surf. Coat.

Technol. 2005, 200, 1530.[45] B. Basu, T. Venkateswaran, D. Y. Kim, J. Am. Ceram. Soc. 2006,

89, 2405.[46] S. Sprio, S. Guicciardi, A. Bellosi, G. Pezzotti, Surf. Coat. Technol.

2006, 200, 4579.[47] J. R. Piascik, J. Y. Thompson, C. A. Bower, B. R. Stoner, J. Vac. Sci.

Technol. A 2006, 24, 1091.[48] D. E. Ruddell, B. R. Stoner, J. Y. Thompson, Thin Solid Films

2003, 445, 14.[49] D. H. Trinh, H. Hogberg, J. M. Andersson, M. Collin, I. Reineck,

U. Helmersson, L. Hultman, J. Vac. Sci. Technol. A 2006, 24,309.

[50] M. Jirout, J. Musil, Surf. Coat. Technol. in press[51] S. H. N. Lim, D. G. McCulloch, M. M. M. Bilek, D. R. McKenzie,

Surf. Coat. Technol. 2003, 174–175, 76.[52] M.-F. Yu, O. Lourie, M. J. Dyer, K. Moloni, T. F. Kelly, R. S. Ruoff,

Science 2000, 287, 637.[53] M.-F. Yu, B. S. Files, S. Arepalli, R. S. Ruoff, Phys. Rev. Lett. 2000,

84, 5552.[54] E. W. Wong, P. E. Sheehan, C. M. Lieber, Science 1997, 277,

1971.[55] B. I. Yakobson, P. Avouris, Carbon Nanotubes 2001, 80,

287.[56] M. B. Nardelli, B. I. Yakobson, J. Bernholc, Phys. Rev. Lett. 1998,

81, 4656.[57] T. Belytschko, S. P. Xiao, G. C. Schatz, R. S. Ruoff, Phys. Rev. B

2002, 65, 2354301.[58] K. I. Tserpes, P. Papanikos, Composites B 2005, 36, 468.[59] H. Yanhong, A. S. Olga, H. Zushou, W. P. Clifford, W. B. Donald,

Rep. Prog. Phys. 2006, 69, 1847.[60] C. B. Mo, S. I. Cha, K. T. Kim, K. H. Lee, S. H. Hong, Mater. Sci.

Eng. A 2005, 395, 124.[61] S. I. Cha, K. T. Kim, K. H. Lee, C. B. Mo, S. H. Hong, Scr. Mater.

2005, 53, 793.



228

[62] S. I. Cha, K. T. Kim, S. N. Arshad, C. B. Mo, S. H. Hong, Adv.Mater. 2005, 17, 1377.

[63] Z. Xia,W. A. Curtin, B.W. Sheldon, J. Eng. Mater. Technol. 2004,126, 238.

[64] A. B. Dalton, S. Collins, J. Razal, E. Munoz, V. H. Ebron,B. G. Kim, J. N. Coleman, J. P. Ferraris, R. H. Baughman,J. Mater. Chem. 2004, 14, 1.

[65] G. D. Zhan, J. D. Kuntz, J. L. Wan, A. K. Mukherjee, Nat. Mater.2003, 2, 38.

[66] Q. Huang, L. Gao, J. Sun, J. Am. Ceram. Soc. 2005, 88, 3515.[67] Z. Xia, L. Riester, W. A. Curtin, H. Li, B. W. Sheldon, J. Liang,

B. Chang, J. M. Xu, Acta Mater. 2004, 52, 931.[68] S. Zhang, X. L. Bui, Y. Fu, D. L. Butler, H. Du, Diamond Relat.

Mater. 2004, 13, 867.[69] Y. T. Pei, D. Galvan, J. T. M. De Hosson, C. Strondl, J. Eur. Ceram.

Soc. 2006, 26, 565.[70] Y. T. Pei, D. Galvan, J. T. M. De Hosson, Acta Mater. 2005, 53,

4505.[71] G. D. P. Cherepanov, ‘‘Mechanics of brittle fracture’’,McGraw-

Hill, New York 1979.[72] American Society for Testing and Materials. Philadelphia, PA

1987.[73] A. A. Voevodin, S. V. Prasad, J. S. Zabinski, J. Appl. Phys. 1997,

82, 855.



[74] S. Zhang, X. L. Bui, Y. Fu, Surf. Coat. Technol. 2003, 167, 137.[75] G. S. Fox-Rabinovich, B. D. Beake, J. L. Endrino, S. C. Veldhuis,

R. Parkinson, L. S. Shuster, M. S. Migranov, Surf. Coat. Technol.2006, 200, 5738.

[76] G. S. Fox-Rabinovich, S. C. Veldhuis, V. N. Scvortsov,L. S. Shuster, G. K. Dosbaeva, M. S. Migranov, Thin Solid Films2004, 469–470, 505.

[77] S. Zhang, D. Sun, Y. Fu, H. Du, Q. Zhang, J. Metastable Nano-cryst. Mater. 2005, 23, 175.

[78] S. Zhang, D. Sun, Y. Fu, H. Du, Surf. Coat. Technol. 2005, 198, 74.[79] B. R. Lawn, A. G. Evans, D. B. Marshall, J. Am. Ceram. Soc. 1980,

63, 574.[80] X. Li, D. Diao, B. Bhushan, Acta Mater. 1997, 45, 4453.[81] K. N. Jonnalagadda, S. Cho, I. Chasiotis, T. A. Friedmann,

J. P. Sullivan, Mater. Res. Soc. Symp. 2005, p. U9.7.1.[82] S. Zhang, D. Sun, Y. Fu, H. Du, Thin Solid Films 2004, 469–470,

233.[83] P. Jedrzejowski, J. E. Klemberg-Sapieha, L. Martinu, Thin Solid

Films 2004, 466, 189.[84] A. A. Voevodin, J. J. Hu, J. G. Jones, T. A. Fitz, J. S. Zabinski, Thin

Solid Films 2001, 401, 187.[85] Z. Q. Qi, E. I. Meletis, Thin Solid Films 2005, 479, 174.[86] S. Neralla, D. Kumar, S. Yarmolenko, J. Sankar, Composites B

2004, 35, 157.

DOI: 10.1002/ppap.200600179

hard yet tough nanocomposite coatings – present status and future trends

Documents