Contents · transmission electron microscopy, x-ray diffractometry and atom probe tomography to obtain information of the microstructure and the composition. Furthermore, nanoindentation
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Paper 1 .................................................................................................................................................77
Paper 2 .................................................................................................................................................83
Paper 3 .................................................................................................................................................93
Paper 4 ...............................................................................................................................................101
Paper 5 ...............................................................................................................................................123
1
1. Introduction
Contrary to popular belief, thin films are all around us in our daily life. It can be in the
appearance of a decorative and wear resistance coating on a phone or a wristwatch, or as an
electrical contact inside an electrical gadget. New applications using thin film technology are
constantly evolving and the world market is growing. The important market for this thesis is
the protective coatings used in the cutting tool industry. These coatings are expected to have
high hardness and stiffness, in combination with good chemical inertness. One of the
coatings fulfilling these requirements is Ti1-xAlxN, the material explored in this thesis.
1.1. Research goals
The main objective of this thesis is to understand the behavior of Ti1-xAlxN/TiN multilayer
coatings, or more explicitly, the influence from the lamellar structure on the mechanical
properties, thermal stability and cutting performance. The unstable c-Ti1-xAlxN transforms to
nano-sized domains rich in AlN and TiN by spinodal decomposition, which results in
improved mechanical properties. I herein study the details of this isostructural
decomposition which are not fully understood. I also examine the possibilities to control the
decomposition behavior with a multilayer architecture, and through this improve the
mechanical properties. Hence, with this work some light might be shed on how internal
interfaces influence the high temperature behavior of Ti1-xAlxN. The specimens were
deposited with the physical vapor deposition (PVD) technique reactive cathodic arc
evaporation, using a full-scale industrial system. Characterization was performed by analytical
transmission electron microscopy, x-ray diffractometry and atom probe tomography to
obtain information of the microstructure and the composition. Furthermore,
nanoindentation and cutting tests were performed to investigate the mechanical properties
and differential scanning calorimetry was used to examine the thermal stability.
Introduction
2
1.2. Outline of the thesis
The second chapter gives an introduction to the materials used for hard coatings, especially
the ones of interest for this work. The structure of multilayers and the resulting hardening
mechanisms are described in chapter 3. Chapters 4 and 5 deal with phase transformations,
decomposition behavior and how the coatings in this work were deposited. This is followed
by a description of the characterization techniques used in this thesis. Chapter 7 gives a short
introduction to phase-field simulations and how the method has supported the experimental
results in this work. Chapter 8 explains the wear mechanisms studied in this work and
chapter 9 presents data, not found in the appended papers, showing epitaxial stabilization
effects and mechanical properties of an arc evaporated c-Ti0.25Al0.75N/TiN multilayer.
Chapter 10 gives a summary of the appended papers and their contribution to the field. The
final chapter contains some suggestions of future work, based on the results in this thesis.
3
2. Hard coatings
To give a perspective of how hard the coatings in this work are, an overview with their
hardnesses in comparison to steel, c-BN and diamond is given in Figure 1. It is seen that
steel, a common engineering material, and cemented carbide (WC), a typical cutting tool
material, is softer compared to the hard coatings. At the right end of the graph the hardest
material known is found, diamond. The second hardest material in the graph is c-BN, a
modern cutting tool material.
Figure 1. Approximate hardness of stainless steel [1],
The most common method used today is by utilizing an IR-CCD camera which provides a
temperature map offering relatively good resolution [149]. One should, however, be aware
that even the modern thermograph methods suffer from uncertainties. The generation of
Metal cutting
48
heat has been attributed to the shearing of the work material and the sliding of the chip, and
is thus closely related to the cutting speed, i.e. a higher cutting speed increases the
temperature [150]. Temperatures in the range of 700-1000 °C is usually measured at the hot
spot during modern metal machining operations [95]. Furthermore, it has been revealed that
the temperature is reached rapidly after which a steady temperature is observed [149, 151].
These annealing conditions, particularly the heating rates, are intricate to imitate in a lab-
furnace but the setup applied in paper 4 is relatively close.
The high temperatures of the cutting insert during operation results in a higher diffusivity
and increased chemical interaction between the coating and the substrate and the coating
and the work piece. An effect of the substrate-coating interaction is frequently seen as a peak
corresponding to Co in x-ray diffractograms of annealed coatings deposited on WC [7, 13].
A proof of this is illustrated in Figure 25, showing how Co has diffused in the grain
boundary into the coating after heat treatments. A higher diffusion rate of Co in the grain
boundaries is expected, considering the local higher defect density. The figure also gives an
example how a multilayer structure can alter the chemical interaction.
Another factor to consider during cutting is the stress distribution. High stresses have been
revealed both by calculations [152, 153] and experiments [95] to influence the
thermodynamics of the cubic solid solution TiAlN. The determination of the stresses
prevailing at the cutting edge is however more complex compared to the temperature
distribution. A model where chip thickness, contact lengths, shear strength of the work piece,
and cutting forces are considered has been used successfully [146]. The stresses during metal
machining has by this model been estimated to 2-6 GPa [95, 154].
8.2. Wear mechanisms
It is not an easy task to make a complete description of the wear during metal machining.
Often a combination of theoretical modeling, materials science, chemistry, heat transfer,
mechanics, and tribology has to be applied to give a complete story. The wear types
presented here are in the dimensions that they are easily to study with the most common
methods for characterization available in the industry, i.e. optical microscopy (OM) and
SEM. There are however more wear mechanisms at a microscopic level, such as micro
cracking and dislocation motions, requiring advanced characterization methods such as FIB
and analytical TEM.
The three most common wear mechanisms discussed in the literature are crater wear, flank
wear, and notch wear. The wear types are illustrated schematically in paper 3. The material
removed from the work piece during metal cutting is called the chip. When the chip slides
along the rake face of the cutting insert, it will cause a significant increase of the temperature.
Due to the increased chemical interactions and the abrasive wear a crater will be formed. An
Metal cutting
49
example of crater wear, from paper 3, is seen in Figure 26. Kramer et al. [155] early suggested
that such wear behavior depends mainly on the solubility of the tool material into the work
piece material i.e. a higher solubility result in increased crater wear. The crater wear is,
logically, also dependent on the mechanical properties of the work piece material and the
combination of adhesion and abrasion. At time of writing, it is not clear exactly which
mechanisms dominate the crater wear in relation to cutting parameters. The break down of
the cutting edge, attributed to crater wear, occurs first when the crater reaches the edge. The
crater wear has consequently no detrimental effect on cutting performance of the insert until
this occurs.
The flank wear is, in contrast to the crater wear, a continuous wear of the cutting edge.
This wear mechanism is generally related to the constant abrasive wear from hard second
phases such as inclusions of carbides and oxides present in the work material.
When machining a work piece material that has been strain-hardened from previous
cutting, e.g. stainless steels, especially in combination with high temperature, notch wear is
likely to occur at the depth of cut. Notch wear primarily depends on the insert geometry and
the oxidation properties of the coating and is not considered in this work.
Figure 25. STEM micrograph and EDS elemental map of Co, Al and
Ti of a heat treated TiAlN coating, showing grain boundary diffusion
of Co. The Co containing substrate is located below the image.
Metal cutting
50
8.3. Cutting performance of TiAlN coatings
When the TiAlN coating was introduced to the hard coating community the improved
cutting performance, compared to TiN, was mainly attributed to the fundamental advantage
that it forms a dense, well adhered and protective Al2O3 film on its surface, and an inner
TiO2 when exposed to high temperatures [15, 156, 157]. The oxides were reported to
prevent diffusion of oxygen into the coating material and thus improving the performance
[15, 156, 157]. It was, however, not clear if the coating actually was subjected to an oxidizing
atmosphere at the insert/work piece contact area. Another early reported advantage for
TiAlN was attributed to its relatively low thermal conductivity, allowing for more heat to
dissipate through the chip removal, resulting in a lower thermal load on the insert [158]. A
feasible explanation of the improved cutting performance were however lacking in the
literature for a long time. In 2003 Mayrhofer et al. made a breakthrough by showing that the
hardness of the TiAlN coating increases upon annealing [6]. The hardness increase was
attributed to the decomposition of the TiAlN into coherent c-AlN and c-TiN domains.
Later comprehensive investigations of the cutting performance and decomposition behavior
were performed where the improvements were assigned to the age hardening [7, 8, 30].
Evidence of an active decomposition of TiAlN during metal cutting was first shown in paper
3 in this thesis, however in a multilayer structure. Recent work by Norrby et al. [95] confirms
that the decomposition is present also in a monolithic Ti0.60Al0.40N after metal cutting.
Furthermore, it has been shown that the high stresses prevailing during cutting can affect
cutting performance of TiAlN by an alteration of the thermal stability. Alling et al. [153]
showed theoretical that the favorable spinodal decomposition will occur earlier. This was
also recently confirmed experimentally by Norrby et al. [95].
Figure 26. Crater wear of (a) a monolithic Ti0.34Al0.66N and (b) a Ti0.34Al0.66N /TiN
multilayer coating (Λ=6+12 nm) from paper 3.
Metal cutting
51
8.4. Cutting performance of multilayer coatings
The improved wear mechanism of multilayers has been explained by, for example, increased
hardness which decreases the abrasive wear resistance [159, 160], altered friction coefficient
[161, 162], and decreased tool/coating interactions [163]. TiAlN based multilayer coating
found in the literature showing improved wear properties compared to monoliths are, for
example, TiAlN/TiAlCN [164], TiAlN/TiNbN [165], TiAlN/TiN [13, 31, 32, 161],
AlN/TiN/TiAlN [166], TiAlN/CrN [49, 167, 168] and TiAlN/Mo [169, 170]. However, in
the majority of those publications the conclusions are based on results from tribological test
methods, such as the pin on disk, which are not directly comparable to metal machining. In a
metal machining operation the coatings are continuously subjected to virgin material, in
contrast to most tribological test methods. However, there exist publications showing an
alteration of the crack mechanisms, during metal machining, due to the interfaces. For
example, in a TiN/TiCN multilayer coating, used in interrupted-cut machining, both the
crack formation and propagation was reported to be suppressed by the layered structure
compared to the monoliths of its constituents [171]. Furthermore, Prengel et al. [16] tested a
multilayer, consisting of layers of TiAlN with different Al content, for a milling operation of
ductile and gray cast iron, with and with out cooling. For the high speed dry milling the
coating was reported to perform significantly better compared to the monolithic TiAlN, due
to its ability to resist micro chipping. The TiAlN-based multilayers in this thesis were tested
with in continuous turning operation on AISI 316L stainless steel. Stainless steel is generally
considered to be a complex material to machine, due to that the chip has a strong tendency
to weld to the flank face of the cutting insert [159]. We showed that the cutting performance
is closely related to the multilayer period, i.e. when the period is decreased both the flank and
crater wear are decreased.
To summarize, the wear mechanisms of a coating during metal machining are very
complex, especially for a multilayer structure. Hence, much research remains to be done on a
microstructural level for both multilayer and monolithic coatings.
53
9. Stabilization of c-Ti0.25Al0.75N
The results presented in this chapter are unpublished and not part of the appended papers
and report on the structure of as-deposited monolithic Ti0.25Al0.75N and Ti0.25Al0.75N/TiN
multilayer coatings. The solid solution c-Ti1-xAlxN can be deposited for x<67 at.% using arc
evaporation [7, 23]. A higher Al content result in deposition of a hexagonal phase or a
mixture of amorphous, hexagonal and cubic phase depending on the deposition technique [7,
23, 24]. Here we investigate if it is possible to deposit cubic layers of Ti0.25Al0.75N in a
multilayer coating with TiN as the second layer type, utilizing the epitaxial stabilization effect.
It has been shown that structures not allowed by the phase diagram can be grown by this
technique [53, 54]. To our knowledge there is no work in the literature doing this by
cathodic arc evaporation for the presented multilayer system. Such study is of interest
because epitaxial stabilization in coatings deposited with industrial arc evaporation systems is
relatively unexplored, but can result in attractive properties [172]. The drawback using
cathodic arc evaporation is the presence of macro particles (see paragraph 5.1.2 for a more
detailed description) which breaks the periodic layer growth and acts as nucleation points
[112] for growth of a not cubic Ti0.25Al0.75N. The study is also attractive since Tantardini et al.
[173] showed that a non-isostructural c-Ti0.7Al0.3N/h-Ti0.3Al0.7N multilayer coatings deposited
by unbalanced dc magnetron sputter, exhibited a hardness of almost 50 GPa.
9.1. Deposition conditions
Coatings were deposited using the Sulzer/Metaplas MZR-323 reactive cathodic arc
evaporation system operating in a N2 atmosphere of 2 Pa, a base pressure of 0.5 mPa and a
substrate bias of -40 V. For the growth of the monolithic coating three 63 mm compound
cathodes of Ti0.25Al0.75 were used. For the multilayer growth the Ti0.25Al0.75 cathodes were
placed opposite to three cathodes of Ti. Cleaned cemented carbide pieces, polished to a
mirror like surface, were used as substrates. To achieve the desired stabilization effect a
multilayer consisting of ~6 nm thick Ti0.25Al0.75N layers were deposited. This was made with
Stabilization of c-Ti0.25Al0.75N
54
a drum rotation of 4 revolutions per minute based on the growth rate of the multilayer
periods seen in paper 2. To control the initial growth to a cubic structure, a ~50 nm thick
layer of TiN was deposited before starting the multilayer deposition.
Figure 27. X-ray diffractograms of monolithic and multilayer
Ti0.25Al0.75N in as-deposited state.
9.2. Microstructure
Figure 27 shows the x-ray diffractograms of the monolithic and multilayer Ti0.25Al0.75N
coatings. The monolithic Ti0.25Al0.75N only shows peaks corresponding to the substrate. This
is what can be expected for an arc evaporated Ti0.25Al0.75N and similar to what Hörling et al.
[7] observed. The multilayer coating show a higher intensity peak between the TiN and c-
Ti0.25Al0.75N both at the 111 and 200 planes. This is similar to what was observed for the
Ti0.34Al0.66N/TiN multilayer with shortest period in paper 2 and is assigned to super lattice
reflections.
Stabilization of c-Ti0.25Al0.75N
55
Figure 28. TEM cross sectional images of (a) monolithic Ti0.25Al0.75N and (b)
multilayer Ti0.25Al0.75N/TiN. Both images are acquired in the same magnifications.
Figure 28 shows cross sectional TEM images (a) of monolithic and (b) multilayer coatings.
The monolithic Ti0.25Al0.75N exhibits a dense fine grained microstructure. The multilayer, on
the other hand, shows a columnar structure. Figure 29 shows a HR-TEM image of a
Ti0.25Al0.75N layer, the neighboring TiN layers, and the corresponding fast Fourier transform
(FFT). The thinner lines to the left indicate the positions of the interfaces between the TiN
and Ti0.25Al0.75N layers. The image reveals coherency across the layers and confirms the
epitaxial growth of a cubic structure expected from the x-ray diffractograms, Figure 27.
9.3. Mechanical properties
Figure 30 shows hardness of the monolithic and multilayer Ti0.25Al0.75N coatings measured
with nanoindentation with a 25 mN load. The data was analyzed by the method of Oliver
and Pharr [121]. The hardness values of cubic Ti0.50Al0.50N and Ti0.50Al0.50N/TiN are inserted
to be used as references. A low hardness of the monolithic Ti0.25Al0.75N is expected since
Tantardini et al. [173] reported a hardness of 16.4 GPa of Ti0.30Al0.70N produced by
unbalanced DC magnetron sputtering. Poor mechanical properties of this coating
composition is also reported from cutting test by Hörling et al. [8] showing less than half of
the tool life time (7 min) compared to the c- Ti0.34Al0.66N (20 min). When the Ti0.25Al0.75N is
layered with TiN an increase of ~5 GPa in hardness is obtained, as seen in Figure 30. This
hardness increase can be attributed to several effects. The first effect arises from the fact that
we have a multilayer coating with several hundred of interfaces acting as crack deflectors and
dislocation barriers. One can expect a large difference in E-modulus between the two layers
[145] i.e. fulfilling a requirement for Koehler hardening [56]. The multilayer effects are
described in more details in paragraph 3.2. There is also a effect from coherence between the
Stabilization of c-Ti0.25Al0.75N
56
layers, which will increase the hardness [12, 26]. The last effect is the contribution from the
overall polycrystalline cubic structure present in the multilayer but not in the monolith. The
cubic phase of Ti1-xAlxN is well known to be harder than the one with a present hexagonal
phase. [6, 13, 27, 30].
Figure 29. HR-TEM image of a Ti0.25Al0.75N and the neighboring TiN
layers showing coherence across the layers. The thinner lines to the
left indicate the positions of the interfaces between the TiN and
Ti0.25Al0.75N layers. Inset shows corresponding FFT pattern with
zone axis [110].
To summarize, in this chapter we show that isostructual c-Ti0.25Al0.75N/TiN multilayers can
be grown by cathodic arc evaporation using the epitaxial stabilization effect. This is
confirmed by both x-ray diffraction, showing peaks corresponding to the cubic phase, and
HR-TEM showing a coherent growth with the c-TiN layer. The FFT of the layers further
confirm this showing a (110) cubic pattern. We also reveal that the hardness of the c-
Ti0.25Al0.75N/TiN multilayer is comparable to monolithic c- Ti0.50Al0.50N i.e. ~5 GPa higher
than monolithic Ti0.25Al0.75N.
Stabilization of c-Ti0.25Al0.75N
57
Figure 30. Hardness of monolithic and multilayer Ti0.25Al0.75N
measured with nanoindentation. Hardness values of isostructural
Ti0.50Al0.50N, multilayer and monolithic, are added as reference.
59
10. Summary of papers and contribution to the field
This chapter gives a summary of the included papers and my opinion how the results may
contribute to the community.
10.1. Paper 1
Direct observations of the decomposed microstructure of Ti0.34Al0.66N, in terms of elemental
contrast, were lacking in the literature before this publication. Cubic metastable
Ti0.34Al0.66N/TiN multilayers with layer thicknesses of 25 and 50 nm, respectively, were
grown by reactive arc evaporation using Ti0.33Al0.67 and Ti cathodes in a N2-atmosphere.
XRD and TEM revealed that the metastable c-Ti0.34Al0.66N layers decompose into c-TiN rich
and c-AlN rich domains with retained lattice coherency after annealing at 900 °C for 2 h.
Elemental mapping by EDS showed a homogenous distribution of Ti and Al in the as-
deposited 25 nm Ti0.34Al0.66N layers. In the annealed specimen the Ti0.34Al0.67N had
decomposed into domains of high Al content surrounded by areas of low Al and high Ti
content. The resolution of the STEM/EDS image is sufficient to expose chemical diffuse
boundaries from an expected spinodal decomposition process. However, in these
experiments possible projection of overlapping particles contributing to the diffuse
boundaries could not be ruled out. Thus, in the investigations of the interfaces by EDS and
HR-TEM there was nothing that contradicted the presence of spinodal decomposition.
The results in this paper showed that the TiAlN-layer decompose to well defined AlN and
TiN domains. This gave an estimation of the size and shape of the domains after 2 hour of
annealing. The observation motivated time resolved studies of the microstructure evolution
of Ti0.34Al0.66N, but also investigations on how the decomposition is affected by the
multilayer interfaces.
Summary of the papers and contribution to the field
60
10.2. Paper 2
There exist numerous publications showing how the mechanical properties of a coating are
altered with a multilayer structure, see paragraph 3.2 and 8.4. However, there are few
investigating how the multilayer structure and the period length influence the thermal
stability and age hardening of the coatings. To investigate this, cubic monoliths of
Ti0.34Al0.66N and multilayers of Ti0.34Al0.66N /TiN with three different periods were grown by
reactive arc evaporation. The multilayers were synthesized by mounting the substrates on a
single axis rotating drum set to rotate 1, 2 and 4 times per minute. This resulted in multilayer
periods of 25/50, 12/25 and 6/12 nm with the thinner layer being the Ti0.34Al0.66N.
X-ray diffraction revealed that the Ti0.34Al0.66N in the multilayer decomposes in the same
two steps seen in the monolith i.e. first to c-AlN and c-TiN followed by a transformation to
h-AlN [6, 7, 26]. DSC showed that the first step of decomposition in the multilayers is
shifted towards lower temperatures. The multilayer coatings further showed, in contrary to
the monolith, increasing h-AlN diffraction peak intensity between the diffractograms of
films heat treated at 1000 and 1100 °C. This suggested that the transformation occurred later,
or slower, in the multilayers. The DSC measurements confirmed the XRD data, showing
that the phase change was shifted to higher temperatures compared to the monolithic
Ti0.34Al0.66N. It was also shown that the hardness drop occurred at higher temperature in the
multilayer coatings, which was in line with the measured heat responses. STEM showed that
h-AlN domains in the multilayers are confined by the TiN layers, i.e. the growth was stopped
in the direction perpendicular to the multilayer interfaces. With this study we showed that
the age hardening and decomposition behavior of Ti0.34Al0.66N can be significantly affected
by a multilayer structure.
10.3. Paper 3
The aim of this study was to investigate how the change in thermal stability and age
hardening, seen in paper 2, affects the cutting performance of the Ti0.34Al0.66N/TiN coating
compared to monoliths of Ti0.34Al0.66N and TiN. The multilayer structures of the coatings as
investigated in paper 2, were deposited on pressed and sintered WC-Co milling inserts
(geometry CNMG120408-MR3). The cutting performance of the inserts was evaluated with
continuous turning of AISI 316L stainless steel with a cutting speed of 250 m/min, feed of
0.15 mm/rev and with a 2 mm depth of cut. TEM specimens from the coating on the worn
cutting insert were prepared by a FIB.
A decrease of multilayer period resulted in both improved resistance to flank and crater
wear. The multilayer with period Λ=6+12 nm showed similar flank wear resistance as a
monolithic Ti0.34Al0.66N coating deposited under identical deposition conditions. All the
Summary of the papers and contribution to the field
61
multilayers, regardless of multilayers period, showed improved crater wear resistance to the
Ti0.34Al0.66N monolith. TEM studies revealed a retained multilayer structure with a varied
defected density in the coating exposed to 15 min of continues wear. HR-TEM showed local
coherency over the multilayer interfaces both in as-deposited state and after the continuous
turning. Further, both coherency and incoherency inside the Ti0.34Al0.66N after the cutting test
was observed. STEM imaging and EDS mapping revealed that the layer has decomposed to
Al-rich and Ti-rich areas.
With our study we showed that there is a connection between the multilayer period and
the cutting performance. Furthermore it revealed that there is a stress relaxation and
decomposition of Ti0.34Al0.66N active during metal cutting. The results are important for the
increased understanding of the cutting behaviour of the widely used Ti0.34Al0.66N coatings.
10.4. Paper 4
The aim of this study was to increase the understanding of the microstructural evolution
during the isostructural decomposition of TiAlN and how it is influenced by composition
and isothermal annealing. Two compositions, Ti0.33Al0.67N and Ti0.50Al0.50N, were studied by
in-situ small angle x-ray scattering (SAXS) using a synchrotron source. Phase-field simulations
were used to understand the experimental results. We showed that the isostructural
decomposition occurs in two stages; spinodal decomposition (initial stage) and coarsening
(latter stage). During the initial stage, spinodal decomposition, of the Ti0.50Al0.50N alloy, the
phase separation proceeded with a constant compositional wavelength of ~2.8 nm of the
AlN- and TiN-rich domains. The time of the initial stage depended on the temperature as
well as the composition, and was shorter for the Ti0.33Al0.67N coating. Following the initial
stage, the AlN- and TiN-rich domains coarsened. The coarsening process is kinetically
limited by the diffusion, which allowed us to estimate of the diffusivity constant and the
activation energy for the metals in the coatings.
From an application point of view, these findings are important because they imply that
already after a short time of metal cutting, considering that the temperatures may reach
above 900 °C [147], the microstructure of the coating is in a coarsening stage.
10.5. Paper 5
In this study, the presence of surface directed spinodal decomposition in arc evaporated Ti1-
xAlxN/TiN multilayers, with two compositions, x=0.67 and x=0.50, was investigated using a
combination of experiments and phase-field simulations. Such study is of interest since
simulations shows that the kinetics of the spinodal decomposition and the resulting evolving
microstructure can be significantly affected by the presence of an interface or a surface. The
characteristics of interface controlled decomposition are the formation of a layered
Summary of the papers and contribution to the field
62
microstructure parallel to the interface i.e. a dominant wave vector directed normal to the
surface.
DSC revealed that the isostructural spinodal decomposition to c-AlN and c-TiN in the
multilayers occur at the same temperature regardless of composition. The onset was located
at a lower temperature compared to the monolithic coatings. Z-contrast STEM imaging
confirmed this by showing a decomposed structure of the multilayers at a temperature where
it was not present in the monoliths. Furthermore, the thermograms show that the
decomposition occurs over a larger temperature range in the multilayers, in comparison to
the monoliths. This is in accord with the phase-field simulations showing longer
decomposition time of the multilayers. 3D atom probe measurements revealed an AlN rich
layer followed by an enriched TiN-layer at the interface in the decomposed Ti0.34Al0.66N/TiN
multilayer, which is in close agreement with the simulated microstructure using large
elemental fluctuations in the initial stage.
The results in this work propose an underlying mechanism for the altered thermal stability
of the multilayer coatings. Since it has been shown that microstructural features such as grain
boundaries might initiate SDSD [104], the understanding of the decomposition type is
important also when considering monolithic TiAlN.
63
11. Future work
This chapter gives an outlook of the possibilities for future work, based on the results
presented within this thesis.
11.1. In-situ decomposition studies
The results in paper 4 show that the coarsening rate of the domains in TiAlN resulting from
spinodal decomposition is significantly increased with temperature. It is further observed,
that the Ti0.50Al0.50N has a period of time of the spinodal decomposition with a constant
compositional wavelength. What is lacking in the literature at the moment is in-situ imaging
of the decomposition, i.e. a motion picture of the evolving microstructure. A modern STEM
equipped with a high temperature sample holder, can provide this. Paper 4 and Figure 11 in
chapter 4, is of great importance for such study, since they allow for selection of appropriate
temperatures and magnifications. Furthermore, an in-situ STEM study of the decomposition
in the multilayers could possibly resolve the evolving SDSD nanostructure, discussed in
paper 5.
11.2. Wear behavior
The wear behavior of TiAlN/TiN multilayers with different periods was investigated in
paper 3. A more detailed study of the microstructure after cutting should be performed, to
increase the understanding of the cutting behavior of multilayers. Such study should contain
investigations of multilayer coated cutting tools, exposed to a series of much shorter
machining times compared to the ones used in paper 3. This is based on the results in paper
4 and 5, showing that the decomposition of the Ti0.34Al0.66N, especially in multilayers, occur
at very short annealing times. This is in line with the results of Norrby et al. [95] showing a
coarsened decomposed microstructure of Ti0.40Al0.60N after only 10 minutes of continuous
cutting. In such study, also the chemical interaction between the cutting insert and the work
piece with the coatings should be considered and investigated. The motivation for this is that
Future work
64
multilayer structures can work as diffusion barriers as discussed in chapter 8. Such
investigation will give a more detailed explanation of the improved crater wear resistance of
the multilayer coatings.
11.3. Mechanical properties
In paper 2 we showed that the age hardening of Ti0.34Al0.66N/TiN multilayers was more
pronounced than monolithic Ti0.34Al0.66N, i.e. it occurred over a wider temperature range and
the relative hardness increase was larger. A study should be performed using FIB sample
preparation and TEM on an indent. Cross sections of indents allow for investigation of the
contact induced deformation mechanisms of coatings. Recently Verma et al. [174] showed
that columnar TiAlN/TiN multilayers, similar to the ones investigated in this work, provides
a more distributed columnar sliding, which reduced the shear cracking. Furthermore, they
showed that interfacial dislocations provide a stress relief mechanism by enabling lateral
movement of material. It has also been shown that at higher loads the main fracture
mechanism consists of crack propagation along the columns while lower loads results in
plastic yielding of the top layers [175]. A comparative study, of as-deposited and
decomposed multilayers, investigating the crack propagation and micro mechanisms during
contact deformation, can give a more detailed explanation of the improved mechanical
properties upon annealing. A similar study on an age hardened Ti0.34Al0.66N monolith, i.e. an
investigation of the crack behavior after annealing, is also interesting and lacking in the
literature.
11.4. Surface directed spinodal decomposition
Paper 5 investigates the presence of SDSD in TiAlN / TiN multilayers. A layer rich in AlN
was observed at the multilayer interfaces. The throughout periodicity which has been
observed in simulations and some experimental results of other material system undergoing
SDSD was, however, not present. This is due to the high initial elemental fluctuations and
high defect density introduced during growth. A similar study should be performed on
TiAlN/TiN multilayers with lower as-deposited elemental fluctuations and dislocations
density. A more homogenous coating can be grown with changed deposition parameters,
such as bias and substrate temperature. An alternative is to use reactive sputtering, allowing
for growth of coatings with much lower defect densities and better interface quality
compared to the ones investigated in paper 5. Such study will explore to what extend the
decomposing structure can be influenced by an interface.
Future work
65
11.5. Improved thermal stability by alloying
It was shown in paper 2 that the unfavorable transformation from c-AlN to h-AlN is
suppressed in the multilayer coatings compared to the monolithic coating. A similar
alteration has been observed in TiAlN alloyed with Cr [176]. Furthermore, other studies
have shown that the spinodal decomposition can be significantly influenced by alloying [34-
38]. Based on these publications a study of a TiAlXN/TiN multilayer should be performed
to investigate if there is a possibility for cumulative attractive properties from the multilayer
structure and the alloying elements. Another approach is to replace the TiN layer which have
poor mechanical properties and low oxidation resistance. The multilayer should, based on
the results in paper 2 and Ref. [33, 176], have a period of ~15 nm and a relatively low
percent of the X element. The characterization should be performed using nanoindentation,
DSC and STEM investigation.
67
12. Bibliography
[1] V. Muthukumarana, V. Selladuraib, S. Nandhakumarb, M. Senthilkumarc, Materials & Design (2010) 2813.
[2] X.H. Ji, S.P. Lau1, G.Q. Yu, W.H. Zhong, B.K. Tay, J. Phys. D: Appl. Phys. 37 (2004) 2543.
[3] X. Jiang, J. Philip, W.J. Zhang, P. Hess, S. Matsumoto, J. Appl. Phys. 93 (2003) 1515.
[4] C.M. Sung, M. Sung, Mat. Chem. and Phys. 43/1 (1996) 1.
[5] Konstantinos-Dionysios, B. Michailidis, G. Skordaris, E. Bouzakis, D. Biermann, R. M’Saoubi, CIRP Annals - Manufacturing Technology Article in press (2012).
[6] P.H. Mayrhofer, A. Hörling, L. Karlsson, J. Sjölén, C. Mitterer, L. Hultman, Appl. Phys. Lett. 83/10 (2003) 2049.
[7] A. Hörling, L. Hultman, M. Odén, J. Sjölén, L. Karlsson, J. Vac. Technol. A 20 (2002) 1815.
[8] A. Hörling, L. Hultman, M. Odén, J. Sjölén, L. Karlsson, Surf. Coat. Technol. 191 (2005) 384.
[9] c-TiN, PDF No.38-1420 JCPDS, - International center for diffraction data, 1998.
[10] J.-E. Sundgren, B.O. Johansson, A. Rocket, S.A. Barnett, J.E. Green, Physics of and Chemisty of protecive coatings, American insitute of physics, Universal City, 1985.
[11] H. Ljungcrantz, M. Odén, L. Hultman, J.E. Greene, J.-E. Sundgren, J. Appl. Phys. 80 (1996) 6725.
[12] A. Flink, T. Larsson, J. Sjolen, L. Karlsson, L. Hultman, Surf. Coat. Technol. 200 (2005) 1535.
[13] A. Knutsson, M.P. Johansson, L. Karlsson, M. Oden, J. Appl. Phys. 108 (2010) 044312.
Bibliography
68
[14] L. Karlsson, A. Hörling, M.P. Johansson, L. Hultman, G. Ramanath, Acta Mater. 20 (2002) 5103.
[15] H.A. Jehn, S. Hofmann, V.-E. Rückborn, W.-D. Münz, J. Vac. Technol. A 4 (1986) 2701.
[21] H. Holleck, Surf. Coat. Technol. 36 (1988) 151.
[22] B. Alling, A.V. Ruban, A. Karimi, O.E. Peil, S.I. Simak, L. Hultman, I.A. Abrikosov, Phys. Rev. B 75 (2007) 045123.
[23] M. Zhou, Y. Makino, M. Nose, K. Nogi, Thin Solid Films 339 (1999) 203.
[24] U. Wahlström, L. Hultman, J.E. Sundgren, F. Adib, I. Petrov, Thin Solid Films 235/62 (1993) 62.
[25] P.H. Mayrhofer, D. Music, J.M. Schneider, Appl. Phys. Lett. 88 (2006) 0171922.
[26] L. Rogström, J. Ullbrand, J. Almer, L. Hultman, B. Jansson, M. Odén, Thin Solid Films 520/17 (2012) 5542.
[27] R. Rachbauer, S. Massi, E. Stergar, D. Holec, D. Kiener, J. Keckes, J. Patscheider, M. Stiefel, H. Leitner, P.H. Mayrhofer, J. Appl. Phys. 110 (2011) 023515.
[28] A.E. Santana, A. Karimi, V.H. Derflinger, A. Schütze, Tribology Letters 17/4 (2004) 689.
[29] R.F. Zhang, S. Veprek, Materials Science and Engineering: A 448/1-2 (2007) 111.
[80] M. Shinn, S.A. Barnett, Appl. Phys. Lett. 64 (1994) 61.
[81] J.M. Molina-Aldareguia, S.J. Lloyd, M. Oden, T. Joelsson, L. Hultman, W.J. Clegg, Philos. Mag. A 82 (2002) 1983.
[82] L. Karlsson, L. Hultman, J.-E. Sundgren, Thin Solid Films 371/1-2 (2000) 167.
[83] A.H. Cottrell, Philos. Mag. 46 (1951) 1169.
[84] K.-D. Bouzakis, N. Michailidis, G. Skordaris, E. Bouzakis, D. Biermann, R. M’Saoubi, CIRP Annals - Manufacturing Technology, Article in press (2012).
[85] A. Hara, T. Asai, H. Sakanou, K. Hirose, Y. Doi, S. Atr. Machine Tool Review 5 (1983) 54.
[86] Metal working products - Sandvik Coromant, Sweden, 1989.
[87] B. Kellock, Mach. Prod. Eng. 148 (1990) 61.
[88] D.B. Wagner, Iron and Steel in Ancient China: Second Impression, 1993.
[89] P.S. Lysaght, J.C. Woicik, M.A. Sahiner, B.-H. Lee, R. Jammy, J. Non-Cryst. Solids 354 (2008) 399.
[90] G. Pant, A. Gnade, M.J. Kim, R.M. Wallace, B.E. Gnade, M.A. Quevedo-Lopez, P.D. Kirsch, Appl. Phys. Lett. 88/3 (2006) 032901.
[91] A.J. Bradely, Proc. Phys. Soc. 52/80 (1940).
[92] J.W. Cahn, Acta Metallurgica 9/7 (1961) 625.
Bibliography
72
[93] J.W. Cahn, Acta Matellurgica 14 (1966) 1685.
[94] D.A. Porter, K.E. Easterling, Phase Transformation in Metals and Alloys, 2nd edition (1992).
[95] N. Norrby, M.P. Johansson, R. M'Saoubi, M. Odén, Surf. Coat. Technol. 209 (2012) 203.
[96] L.J.S. Johnson, M. Thuvander, K. Stiller, M. Odén, L. Hultman, Thin Solid Films 520/13 (2012) 4362.
[97] R. Rachbauer, S. Massl, E. Stergar, P. Felfer, P.H. Mayrhofer, Surf. Coat. Technol. 204/11 (2010) 1811.
[98] T.J. Rappl, N.P. Balsara, The Journal of chemical physics 122 (2005) 214903.
[141] S. Allen, J.W. Cahn, Acta. Metall. 27 (1979) 1084.
[142] K.A. Grönhagen, J. Ågren, Acta Mater. 54 (2006) 1241.
[143] J. Kundin, R. Kumar, A. Schlieter, M.A. Choudhary, T. Gemming, U. Kühn, J. Eckert, H. Emmerich, Computational Materials Science 63 (2012) 319.
[144] B. Alling, A.V. Ruban, A. Karimi, L. Hultman, I.A. Abrikosov, Phys. Rev. B 83/3 (2011) 10420.
[145] F. Tasnádi, I.A. Abrikosov, L. Rogström, J. Almer, M.P. Johansson, M. Odén, Appl. Phys. Lett. 97 (2010) 231902.
[146] H. Chandrasekaran, A. Thuvander, Mach. Sci. Technol. 2 (1998) 355.
[147] M.A. Davies, T. Ueda, R. M'Saoubi, B. Mullany, A.L. Cooke, CIRP Annals - Manufacturing Technology 52/2 (2007) 581.
[148] G. Sutter, L. Faure, A. Molinari, N. Ranc, V. Pina, Int. J. Mach. Tools Man. 43/7 (2003) 679.
[149] R. M´Saoubi, H. Chandrasekaran, Machine Tools and Manufacture/44 (2004) 213.
[150] A. Liljerehn, V. Kalhori, M. Lundblad, Mach. Sci. Technol. 13/4 (2009) 488.
[151] R. M’Saoubi, S. Ruppi, Manufacturing Technology 58 (2009) 57.
[152] D. Holec, F. Rovere, P.H. Mayrhofer, P.B. Barna, Scr. Mater. 62/6 (2010) 349.
[153] B. Alling, M. Odén, L. Hultman, I.A. Abrikosov, Appl. Phys. Lett. 95/18 (2009) 181906.
Bibliography
75
[154] K.-D. Bouzakis, G. Skordaris, S. Gerardis, G. Katirtzoglou, S. Makrimallakis, M. Pappa, E. LilI, R. M'Saoubi, Surf. Coat. Technol. 204 (2009) 1061.
[155] B.M. Kramer, N.P. Suh, J. Engineering for Industry 102 (1980) 303.
[156] W.D. Mûnz, Vac. Sci. Technol. A 4 (1986) 2717.
[157] G. Beensh-Marchwicka, L. Krol-Stepniewska, W. Posadowski, Thin Solid Films 85/3-4 (1981) 543.
[158] T. Leyendecker, O. Lemmer, E. Esser, Ebberink, Surf. Coat. Technol. 48 (1991) 175.
[159] T.I. Selinder, M.E. Sjöstrand, M. Nordin, M. Larsson, Å. Östlund, S. Hogmark, Surf. Coat. Technol. 105 (1998) 52.
[160] M. Nordin, M. Larsson, S. Hogmark, Surf. Coat. Technol. 106 (1998) 234.
[161] K.N. Andersen, E.J. Bienk, K.O. Schweitz, H. Reitz, J. Chevallier, P. Kringshoj, J. Bottiger, Surf. Coat. Technol. 122/2-3 (2000) 219.
[162] C. Ducros, C. Benevent, F. Sanchette, Surf. Coat. Technol. 163-164 (2003) 681.
[163] M. Nordin, R. Sundström, T.I. Selinder, S. Hogmark, Surf. Coat. Technol. 133-144 (2000) 240.