In this study, diamond coating adhesion on cobalt-cemented tungsten-carbide (WC–Co) substrates was investigated using scratch testing. In particular, a methodology was applied to evaluate the effects of the coating thickness and the interlayer on diamond coating delaminations. In the coating thickness effect research, substrate surface preparations, prior to diamond depositions, were common chemical etching using Murakami solutions. On the other hand, to study the interlayer effect, Cr/CrN/Cr and Ti/TiN/Ti, were deposited to WC–Co substrate surfaces (no chemical etching) by using a commercial physical vapor deposition (PVD) system in a thickness architecture of 200 nm/1.5 μm/1.5 μm, respectively. Diamond films were synthesized by using a hot-filament chemical vapor deposition (HFCVD) reactor at a fixed gas mixture with varied deposition times.Scratch testing was conducted on the fabricated specimens using a commercial instrument. It is noted that the onset of coating delamination can be clearly identified by high-intensity acoustic emission (AE) signals and high tangential force fluctuations when such events occur, which can be used to determine the criticalload for coating delaminations. Scratched track morphology was also characterized by scanning electron microscopy and white light interferometry.The results show that the adhesion of diamond coatings on WC–Co substrates increases with the increased coating thickness, with a nearly linear relation. The trend is consistent with the findings from a previous study that experimentally evaluated the coating thickness effect on the diamond-coated tool performance. Scratched tracks were characterized by diamond coating cracking and coating delaminations once the adhesion critical load is reached. For the two types of interlayer materials tested, neither of them seems to be effective and the diamond coating with the Ti-interlayer shows poor adhesion compared to the Cr-interlayer coatings.
Contents lists available at SciVerse ScienceDirect
Surface & Coatings Technology
j ourna l homepage: www.e lsev ie r .com/ locate /sur fcoat
Coating thickness and interlayer effects on CVD-diamond film adhesion tocobalt-cemented tungsten carbides
Ping Lu a, Humberto Gomez b,d, Xingcheng Xiao c, Michael Lukitsch c, Delcie Durham b, Anil Sachdeve c,Ashok Kumar b, Kevin Chou a,⁎a Mechanical Engineering Department, The University of Alabama, Tuscaloosa, AL 35487, USAb Department of Mechanical Engineering, University of South Florida, Tampa, FL 33620, USAc Chemical Sciences & Materials Systems Laboratory, General Motors R&D Center, MI 48090, USAd Departamento de Ingeniería Mecánica, Universidad del Norte, Barranquilla, Colombia
In this study, diamond coating adhesion on cobalt-cemented tungsten-carbide (WC–Co) substrates wasinvestigated using scratch testing. In particular, a methodology was applied to evaluate the effects of thecoating thickness and the interlayer on diamond coating delaminations. In the coating thickness effect re-search, substrate surface preparations, prior to diamond depositions, were common chemical etching usingMurakami solutions. On the other hand, to study the interlayer effect, Cr/CrN/Cr and Ti/TiN/Ti, were deposit-ed to WC–Co substrate surfaces (no chemical etching) by using a commercial physical vapor deposition(PVD) system in a thickness architecture of 200 nm/1.5 μm/1.5 μm, respectively. Diamond films weresynthesized by using a hot-filament chemical vapor deposition (HFCVD) reactor at a fixed gas mixturewith varied deposition times.Scratch testing was conducted on the fabricated specimens using a commercial instrument. It is noted thatthe onset of coating delamination can be clearly identified by high-intensity acoustic emission (AE) signalsand high tangential force fluctuations when such events occur, which can be used to determine the criticalload for coating delaminations. Scratched track morphology was also characterized by scanning electronmicroscopy and white light interferometry.The results show that the adhesion of diamond coatings on WC–Co substrates increases with the increasedcoating thickness, with a nearly linear relation. The trend is consistent with the findings from a previous studythat experimentally evaluated the coating thickness effect on the diamond-coated tool performance. Scratchedtracks were characterized by diamond coating cracking and coating delaminations once the adhesion criticalload is reached. For the two types of interlayer materials tested, neither of them seems to be effective and thediamond coating with the Ti-interlayer shows poor adhesion compared to the Cr-interlayer coatings.
Chemical vapor deposition (CVD) of diamond on cemented carbideshas been an ideal approach in enhancing cutting tool life and improvingtheir machining performance due to exceptional diamond propertiessuch as superior hardness, low coefficient of friction, and chemicalstability. CVD diamond-coated tools provide significant advantages interms of cost and fabrication flexibility when compared to syntheticpolycrystalline diamond (PCD) tools [1], which are also commonlyused in the manufacturing industry. The ability to form a conformalcoating on a carbide tool surface, the relative synthesis simplicity as aresult of the newadvances in industrial CVD reactors, and the possibility
rights reserved.
to produce different film structural characteristics (micro or nano-crystalline), represent a significant benefit of CVD diamond coatings[2]. However, under cutting operations represented by harshmachiningconditions or high-strength workpiece materials, diamond coating de-lamination remains to be the primary wear mechanism that results incatastrophic tool failures [3]. For cemented carbide substrates likeWC–Co, diamond delamination is due to insufficient adhesion betweenthe coating and the substrate, partially as the result of the formation ofnon-diamond compounds at the substrate–diamond film interface dueto the cobalt–carbon interdifussion at CVD deposition temperatures.
1.2. Interface engineering
Several interface engineering approaches have been reported in thelast 15 years with the aim to reduce the undesired catalytic effect ofcobalt on diamond adhesion [4–6]. In order to maximize the practicaladhesion of diamond coatings on cemented carbides, any approach
Fig. 1. Surface characteristics of the WC–Co (6%) surface as received from supplier,a) SEM image of surface morphology representing the finishing marks on the tooland b) optical interferometry image of the surface representing the roughness valueand pattern.
Table 2Sample details for diamond coated inserts in scratch tests.
273P. Lu et al. / Surface & Coatings Technology 215 (2013) 272–279
must halt the interdifussion effect of cobalt. The most widely successfultechniques discussed in the literature are related to the cobalt removalin depths ranging in about 3 to 10 μm from the substrate surface byusing chemical etching methods [7], or by halting the cobalt effect onthe surface by depositing interdifussion barrier layers [8] that alsodiminish the thermal stresses caused during the diamond growth.
Interface engineering techniques are specifically targeted to im-prove diamond coating adhesion. Since an increased surface roughnesshas been correlated to enhancing the diamond nucleation density andpromoting a film interlocking behavior, surface pretreatment effortscan also be tailored accordingly besides suppressing the cobalt catalyticeffects. In addition to the improvements in the diamond growth condi-tions, the substrate surface plays an important role in the final adhesionbehavior of diamond coatings. Surface textures and surface/subsurfacedamage characteristics on the substrate have a direct impact to the sub-sequent diamond adhesive quality and wear failure modes; hence, thefinal diamond coating adhesion behavior depends on surface pretreat-ments used and their resulting effects on the substrate surface, whichare ultimately the interface characteristics in the substrate-coatingcomposite system. This interface requires the formation of strong
Table 1Conditions of the pretreatments of the substrate.
Denotation Pretreatment Conditions
T Chemical etching 1. Acetone cleaning2. K3(Fe(CN)6)+KOH+H2OHNO3+H2O2
interfacial chemical bonds between the diamond crystallites nucleatedat the surface and the atoms at the substrate surface. Moreover, amechanical interlocking effect is also desired in order to enhance thecoating addition.
The effects of chemical etchings on the surface characteristics ofWC–Co substrates have been studied by far [9,10] and represent thepretreatment method used in most of the commercial diamond coatedcemented carbides in the industry. This method has the purpose to pro-duce a selective etching of the cobalt binder by using a two step processcomposed by an initial wet treatment in a Murakami solution with theaim of reconstructing and roughing the surface by attacking the WCgrains and exposing the Co binder [9]. Then, a second wet etching inan acid solution (H2SO4 or HNO3 with H2O2) is used to reduce theexposed cobalt in a depth determined by the etching time [10].
Another approach to avoid the catalytic effect of cobalt is the depo-sition of carbide and nitride intermediate layers (CrN, TiN, TiC, SiC, AlN,etc.) on the substrate before final diamond depositions. These inter-layers normally deposited by physical vapor deposition (PVD)methodsmust remain stable during the diamond deposition, have a low thermalexpansion coefficient to minimize internal stresses, and provide acarbide formation layer to improve diamond nucleation [11]. Theseconditions may also be improved by using nanometer sized metal thinlayers like Cr and Ti at the top or bottom of the interlayer architecture.Additional diamond particles may be peened in the top interlayersurface to provide additional diamond nucleation sites and serve asanchors to final diamond coatings [12].
1.3. Coating adhesion and scratch testing
There are several methods commonly used to examine the adhe-sion of coatings in general [13]. Scratch testing is one of the mostpractical approaches in evaluating the adhesion of a hard-thin coatingon a substrate [14,15], since it is reliable, simple to perform, and withno special specimen geometry or preparation requirements. Coatingadhesion is measured as a correlation between the occurrences ofcritical load at the coating failure instant. In the event of an adhesivefailure, this critical normal load is taken as a measure of the coating–substrate interface adhesion or used to calculate the work of adhesion[16,17]. During a scratch test, a spherical indenter tip slides over the
Fig. 2. Raman spectrum corresponding to the structure of the CVD films (Sample T-2.5)depicting the microcrystalline diamond structure represented by the 1332 cm−1 broadpeak.
Fig. 3. SEM image corresponding to the diamond film deposited in the sample (SampleT-2.5) depicting the faceted diamond polycrystals.
0
1000
2000
3000
4000
5000
0
2
4
6
8
0 2000 4000 6000 8000 10000
Tan
gent
ial F
orce
(mN
)
Aco
ustic
Em
issi
on(%
)
Normal Force(mN)
AE FtSpot 1
Fig. 5. Acoustic emission (AE) and tangential force (Ft) vs. normal load (Fn) for a max-imum load of 10 N.
274 P. Lu et al. / Surface & Coatings Technology 215 (2013) 272–279
surface of the coating to generate a groove at an incremental or con-stant normal load mode. In addition, the tangential force can be mea-sured during the test and the morphology of the scratch track can beobserved simultaneously or afterwards. Moreover, an acoustic emis-sion sensor can be used to capture the coating delaminations duringscratch tests. When the resolved compressive mean stress exceeds acritical value, the coating detaches from the substrate decreasingthe elastic energy stored in the coating [18]. The work of adhesionat the interface between the coating and the substrate is equal tothe energy release rate from the coating at the instant of detachmentas a function of the compressive mean stress of the coating over thedelaminating area. Thus, the critical compressive coating mean stressresponsible for the detachment could be a measure of the coating–substrate adhesion. On the other hand, diamond coatings are verybrittle. While a coating can withstand compressive stresses inducedby the indenter to a certain extent, it may fracture if a high tensileor shear stress field is induced simultaneously, e.g., at the interfacecausing delaminations. It is known that diamond coatings have ahigher critical compressive stress than tensile and shear stresses,but stresses less than the critical compressive stress may result incoating delaminations during scratch testing.
1.4. Adhesion characterization
As discussed, the adhesion of coatings is measured by the criticalload under coating failures, and there are different ways to determinethe critical load [13]. Microscopic observations are the most reliablemethod to detect the coating delaminations. This technique can dis-tinguish the cohesive failure within the coating and the adhesivefailure at the interface of the coating–substrate system. The use ofacoustic emission (AE) sensors, which is insensitive to mechanicalvibration frequencies of the instrument, represents another optionto detect the elastic waves generated as a result of the formationand propagation of micro-cracks in the diamond coating during ascratch test. The adhesion of CVD diamond coatings on molybdenum
Fig. 4. Digital microscopic images of scratch grooves on sample T-1.5, 20 N load.
substrates has been investigated by scratch testing [11], and theresults displayed critical normal load values in the range of 16 N to40 N for CVD diamond films grown for 4 h. at a CH4/H2 ratio of0.5%. However, diamond films grown for 24 h at a methane concen-tration of 0.5% do not exhibit any failure when the force increasedto75 N. Moreover, scratch tests have been applied and able to providea direct qualitative comparison of the adhesion of diamond coatingson steel and copper substrates [12], and attempted to investigatethe effect of metal substrates (copper and steel) and film thicknesson the adhesion The results showed that the diamond coatings onsteel exhibits a higher critical load than on the copper, and thickerfilms displays a higher critical load than thinner films for the samekind of substrates.
1.5. Objectives and approaches
This study aims at better understanding the adhesion of diamond-coated carbide cutting tools by micro-scratch testing, and the criticalload for coating delaminations was used to evaluate the adhesion ofdiamond-coated carbide tools, where corresponding process signalswould help identify the coating delamination. It is intended to investi-gate the effects of coating thickness and interlayers on the adhesion ofdiamond coatings for cutting tool applications.
2. Experimental details
2.1. Substrate preparation
Experimental samples correspond to WC–Co (6%) square-shapedcemented carbide substrates. The surfaces of the tools display surfacecharacteristics represented by feed marks resulting from their manu-facturing process. These preferential marks are depicted in Fig. 1and constitute the as-receive state of samples before any surfacepretreatment.
The use of chemical etching pretreatment and the pre-depositionof interdiffussion barrier layers were applied to the samples to modifythe as-ground surface before the diamond deposition, improving thecoating adhesion by halting the effect of the cobalt binder in thecemented carbide substrate. The conditions of the pretreatments aresummarized in Table 1 and were selected from previous work bythe authors [10].
2.2. Interlayer preparation
Two different interlayers, Cr/CrN/Cr and Ti/TiN/Ti were depositedto WC–Co substrate surfaces by using a commercial PVD coating sys-tem in the thickness architecture of 200 nm/1.5um/1.5um, respec-tively. This physical barrier prevents the diffusion of carbon into theunderlying cobalt phase and the subsequent graphite formation thatis so deleterious to diamond film adhesion. The barrier also providesa stress relaxation barrier layer [19]. Additional treatments after theinterlayer deposition were applied to the top of the surface in order
Fig. 7. Acoustic emission (AE) and tangential force (Ft) vs. normal load (Fn) for a max-imum load of 20 N.
a) Optical: at around 8N
b) Optical, end of scratch
c) SEM image
100 µm
100 µm
Cracks
Fig. 6. Scratch track images (T-1.5 sample at a 10 N maximum load) at different loca-tions: (a) optical, around 8 N load, and (b) optical, at end of scratch. (c) SEM, an exam-ple showing coating cracks.
275P. Lu et al. / Surface & Coatings Technology 215 (2013) 272–279
to improve the surface roughness for diamond nucleation. Thissurface treatment corresponds to an additional shot peening to thefinal Cr and Ti layers using diamond powder particles (1 μm).
2.3. Coating deposition
Pretreated samples were subjected to a seeding process prior todiamond depositions. The seeding method was performed using aslurry solution, consisting of 1.2 g titanium nanopowder, 1.2 g nano-crystalline diamond powder, and 100 ml of methanol. Diamond films
were synthesized by using an HFCVD reactor at a pressure of 20 Torr,two filaments located at the top of the sample operating at 90 V, anda gas mixture of 6 sccm of CH4 and 60 sccm of H2, with a depositionrate of about 0.3 μm/h. In order to investigate the effect of coating thick-ness on the adhesion of diamond coatings, three different coating sam-ples (coded T-1.5, T-2.5 and T-4.5) were prepared at the same workingparameters except the deposition time, whichwill result in thin coatingthickness ranged from 1.5 μm to 4.5 μm. Table 2 shows the sampledetails for the diamond-coated cutting tools used in the scratch testsdetailed below. For the specimens with either the Ti or Cr interlayer,the deposition thickness estimated was about 2 to 4 μm.
2.4. Scratch test setup, procedure and data acquisition
A Micro-scratch tester from CSM Instruments, model Micro-Combi,wasused for the experiments at room temperature, by using an indenterwith a tip radius of 50 μm, and a scratch speed of 2 mm/min with aprogressive loading method to determine the critical load of diamond-coated tools. The scratch distance for each test was set at 5 mm. Duringthe scratch test, tangential force values, acoustic emission (AE) signals,and the resulting depth of the scratch were acquired. A KEYENCE digitalmicroscope (VHX-600×) was used to observe the scratch tracks andcoating delaminations after the test. In addition, a white-light interfer-ometer (WLI) was used to characterize the morphology of the scratchgrooves, and a scanning electron micrograph instrument (PhilipsXL30) was used to present the coating delamination phenomena andpropagations.
3. Results and discussion
3.1. Characterization information of diamond coatings
Raman spectroscopy was performed to all CVD diamond-coatedsamples, corresponding to a microcrystalline diamond structure repre-sented by the 1332 cm−1 broad peak observed in the Raman spectrumand shown in Fig. 2. The large background in the spectrum was due toother carbon bond structures, e.g., sp2 (graphite), amorphous carbon,and disordered sp3 and sp2, which exist in typical CVD diamond films.The crystal structure of the diamond film corresponding to faceted(100){111} polycrystals is shown in Fig. 3.
3.2. Scratch test results for diamond coated WC with different thicknesses
Scratch tests conducted on the T-1.5 included four repeated scratchtests at a maximum load of 10 N, and one scratch test of a maximumload of 20 N. Fig. 4 shows an example of the scratch grooves at 20 Nload.
Fig. 5 shows the AE signal and tangential force (Ft) vs. the appliednormal load (Fn) during the 4th scratch test at the load of maximum10 N. It is observed that the tangential force increases smoothly,though varies slightly until reaching 10 N. It is also found that onlya few isolated high peaks (i.e. Spot 1) of AE signals exist before the
a) 200X
b) 800X
Fig. 9. The SEM image of the coating delamination on sample T-1.5 (20 N load):(a) 200× and (b) 800×.
a) Spot 1
b) Spot 2
c) End of scratch
100 µm
100 µm
100 µm
Fig. 8. The digital microscopic images: (a) spot 1 around 9 N, (b) spot 2 around 13 N,and (c) the end of the scratch for scratch test under maximum load of 20 N.
276 P. Lu et al. / Surface & Coatings Technology 215 (2013) 272–279
normal force reaches 10 N. This implies that coating delaminationsmay not have initiated under such a load.
To understand the isolated high peaks, Spot 1, the scratch groovewas observed in the digital microscope at 1000× after testing. Fromthe correlation between the load and the distance, the location corre-sponding to Spot 1 can be physically located to verify whether thecoating delamination has initiated at that point. Fig. 6a shows the dig-ital microscopic image of Spot 1 and the corresponding normal load isaround 8.0 N, and Fig. 6b shows the digital microscopic image at theend of the scratch test. It can be observed that the cracks on the
coating have been formed at the spot without coating delaminations.Fig. 6c is an SEM image at a location showing crack formations.
For Sample T-1.5 at a 20 Nmaximum load, Fig. 7 shows the AE signaland tangential force (Ft) vs. the applied normal load (Fn) during thetest. Similar results are found that the tangential force increasessmoothlywith the normal force, if less than 9 N, but followed by consid-erable fluctuations. An abrupt amplitude increase of AE signals (Spot 1)occurs at the load around 9 N, followed by a series of continuoushigh-amplitude AE peaks. In addition, the highest amplitude of AEpeak occurred at Spot 2, also drastic fluctuations of tangential forceswere observed at Spot 2, corresponding to a normal force of 13 N.
Fig. 8a and b shows the digitalmicroscopic images at Spot 1 and Spot2, showing the cracks initiation but without coating delamination atSpot 1 (9 N load), and coating delaminations at Spot 2 (13 N) haveinitiated at such a force, with clearly exposing the substrate layer ofWC, near Spot 2. Fig. 8c shows the digital microscopic image at theend of scratch test. It is shown that coating delaminations continued,once initiated, to the end of the final load, with a comparable delamina-tion width. Fig. 9 displays the SEM images of Spot 1 at 200× and 800×,clearly showing multiple micro-cracks on the coating surface nearSpot 1, followed by coating material removals from the substrate,where the coating delamination formed.
The coating delamination sizes have also been evaluated by WhiteLight Interferometry (WLI), NT1100 from Veeco Metrology. As anexample, a WLI image and the scratch track 2D profile plot areshown in Fig. 10. The analysis result shows that the delaminationwidth at the end of the scratch (20 N maximum load) is around0.8 mm with a depth of around 15 μm.
10864202468
10
0 100 200 300 400 500 600 700 800 900
Y Profileµm
µm
798.92µm
Fig. 10. WLI analysis of scratch track at 20 N load location, showing coating delamination with 2D profile.
20 µm
a) Spot 1 around 14.6N
b) End of scratch
277P. Lu et al. / Surface & Coatings Technology 215 (2013) 272–279
The results of testing on the sample T-4.5 at a maximum load of20 N, repeated three times, are discussed below. Fig. 11 shows theAE signal and tangential force (Ft) vs. the applied normal load (Fn)during the 1st scratch test. Similar to earlier observations, the transi-tion at a location (Spot 1) is found at the normal load of 14.6 N. andan abrupt amplitude increase of AE signals (Spot 1) exists at that lo-cation, followed by a series of continuous high-amplitude AE peaks,implying the coating delamination.
Fig. 12a shows the digital microscopic image at Spot 1 with corre-sponding normal load around 14.6 N. Coating delaminations have ini-tiated at such a force with a clearly exposed substrate layer of WC,near Spot 1. Fig. 12b shows the digital microscopic image at the endof scratch test, showing continued coating delaminations.
Fig. 13 shows the SEM images of scratch tracks at the final loadlocations (20 N) of the samples with three different coating thick-nesses. All the results confirm that coating spallation formed alongthe scratching process, and severe coating detachments were notedat the end of the scratch. It was also found that T-4.5 had the slightestcoating detachment compared to the other two samples.
Fig. 14 shows the critical load of coating delaminations vs. thecoating thickness. It can be noted that the critical load increaseswith the coating thickness, increasing from 11.2 N to 14.5 N for thecoating thickness of 1.5 μm vs. 4.5 μm. It demonstrates that thediamond coating adhesion increases with the coating thickness, forthe range in this study. In scratch testing, the contact between theindenter and the coating produces a high tensile stress right atthe contact circle, resulting in Hertzian ring cracks. Meanwhile, thebending of the coating on a compliant substrate incurs compressivestresses at the top surface of the coating, but tensile stresses at thebottom surface. The combination of the two mechanisms may affectfailures in an opposite way at different coating thicknesses. For thincoatings used in this, the radial cracking is more dominant, and thecritical load will increase with the coating thickness. Further, theexperimental results of the coating thickness effect are consistentwith a previous work that evaluated diamond-coated tool perfor-mance [20]. In the previous work [20], diamond-coated carbidecutting tools with 3 levels of coating thickness (4 to 29 μm) weretested in machining A359/SiC-20p composite bars at specific cutting
0
2000
4000
6000
8000
10000
0
2
4
6
8
10
12
14
0 5000 10000 15000 20000
Tan
gent
ial F
orce
(mN
)
Aco
ustic
Em
issi
on(%
)
Normal Force(mN)
AE Ft Spot 1
Fig. 11. Acoustic emission (AE) and tangential force (Ft) vs. normal load (Fn) for amaximum load of 20 N, sample T-4.5.
conditions. The results clearly indicated that the delamination wear re-sistance of diamond-coated tools increases with the coating thickness.
3.3. Scratch test results on diamond coating with different interlayers
In the interlayer study, Fig. 15a and b shows the AE signal and tan-gential force (Ft) vs. the applied normal load (Fn), with a maximum5 N load, during the test on the I-Ti and I-Cr samples, respectively.It is concluded, based on the force data, AE signal and optical images,that the coating delamination initiated around 1.0 N for I-Ti and 3.5 Nfor I-Cr.
Fig. 16 displays the SEM images at the scratch end location (5 Nload) of the two samples: I-Ti and I-Cr. The same results of coatingspallation along the scratch test after coating delaminations, and
20 µm
Fig. 12. The digital microscopic images: (a) spot 1 around 14.6 N, and (b) the end ofthe scratch for sample T-4.5 at a maximum load of 20 N.
a) T-1.5
b) T-2.5
c) T-4.5
Fig. 13. SEM images at the scratch end location (20 N): (a) T-1.5, (b) T-2.5, and (c) T-4.5.
a) I-Ti
b) I-Cr
030060090012001500
020406080
100
Tan
gent
ial F
orce
(mN
)
Aco
ustic
Em
issi
on(%
)
Normal Force(mN)
AE Ft
05001000150020002500
0
10
20
30
0 1000 2000 3000 4000 5000
Normal Force(mN)0 1000 2000 3000 4000 5000 T
ange
ntia
l For
ce(m
N)
Aco
ustic
Em
issi
on(%
)
AE Ft
Spot 1
Spot 1
Fig. 15. Acoustic emission (AE) and tangential force (Ft) vs. normal load (Fn) at a maxi-mum load of 5 N: (a) I-Ti and (b) I-Cr.
a) I-Cr
b) I-Ti
278 P. Lu et al. / Surface & Coatings Technology 215 (2013) 272–279
more severe coating detachments were observed on the sample ofI-Ti than that on I-Cr.
The results indicated that the critical load is only 1 N for the I-Ti,and 3.5 N for I-Cr. Thus, both are not effective with the current
8
10
12
14
16
0 1 2 3 4 5
Cri
tical
load
(N)
Coating thickness( m)
Fig. 14. Critical load for coating delamination vs. coating thickness.
approach. The sample of I-Ti has the poorest adhesion compared tothe samples discussed in this research. The Cr-interlayer provides aslightly better adhesion than the Ti-interlayer. The possible reasonis that the carbon diffusion in Cr is relatively low compared to Ti,which should improve the adhesion of diamond to the seeded sub-strate. However, the adhesion is still rather poor compared to the dia-mond coating samples without the specific interface approach. This ispossibly due to a defective chromium carbide layer formed during thedeposition. It is known that the multi-phase system could provide aneasy path for the micro-cracks due to transformation among these
Fig. 16. SEM images at the scratch end location (5 N): (a) I-Cr, and (b) I-Ti.
279P. Lu et al. / Surface & Coatings Technology 215 (2013) 272–279
phases [19,21]. In addition, the mismatch of the coefficients of thermalexpansion in the multilayer structure aggregates the propagation ofmicro-cracks in the scratch test that eventually leads to delaminations.It is also suggested that TiC directly as the interlayer would be a betterchoice than using Ti because it could avoid the potential phase transfor-mation to form graphite.
4. Conclusions
Scratch testing of thin diamond films deposited on WC–Co sub-strates has been carried out using a micro-scratch tester. The objectiveis to evaluate the coating thickness and the interlayer effects on the de-lamination critical load of diamond-coated carbide substrates. Duringthe scratch tests, the tangential forces, the acoustic emission signalsand the penetration depth were acquired to identify the delaminationinitiation event. After scratch tests, the scratched tracks were alsoobserved by digital microscopy, SEM as well as WLI. The results aresummarized as the following.
(1) In the scratching process, coating delaminations can be clearlydetected by AE signals. It was observed that the abrupt AE peakjumps followed by several continuous AE high-amplitude peaksare associated with the coating delamination. In addition, thetangential force increases smoothly with the normal force beforethe initiation of coating delamination, but fluctuates consider-ably once coating delaminations initiated.
(2) Scratched tracks were characterized by coating cracking andcoating delaminations once the adhesion critical load is reached.Further, the size of coating delaminations increases with theincreased normal loads after the delamination is initiated, con-firmed by the scratched track images.
(3) The adhesion of the diamond coating increases with the in-creased coating thickness in the range tested. The trend is consis-tent with the findings from a previous study that experimentallyevaluated the coating thickness effect on the diamond-coatedtool performance.
(4) The attempted Ti-interlayer and Cr-interlayer do not seem to beeffective in the interface adhesion enhancement compared toother chemically etched samples.
Acknowledgments
This research is supported by NSF, CMMI 0928627 — GOALI/Collaborative Research: Interface Engineered Diamond Coatings forDry Machining, between The University of Alabama, General Motorsand University of South Florida. P. Lu would also like to thank theGraduate School of The University of Alabama for offering a GraduateCouncil Fellowship to partially support his study.
References
[1] J. Hu, Y.K. Chou, R.G. Thompson, Trans. NAMRI/SME 35 (2007) 177.[2] P. Lu, Y.K. Chou, R.G. Thompson, in: Proc. 2009 Int. Manuf. Sci. and Eng. Conf,
2009, (MSEC2009-84372).[3] F. Qin, J. Hu, Y.K. Chou, R.G. Thompson, Wear 267 (2009) 991.[4] R. Polini, M. Barletta, Diamond Relat. Mater. 17 (2008) 325.[5] R. Polini, F.P. Mantini, M. Barletta, R. Valle, F. Casadei, Diamond Relat. Mater. 15
(2006) 1284.[6] Z. Xu, L. Lev, M. Lukitsch, A. Kumar, Diamond Relat. Mater. 16 (2007) 461.[7] J. Oakes, X.X. Pan, R. Haubner, B. Lux, Surf. Coat. Technol. 47 (1991) 600.[8] E. Cappelli, F. Pinzari, P. Ascarelli, G. Righini, Diamond Relat. Mater. 5 (1996) 292.[9] F. Deuerler, H. van den Berg, R. Tabersky, A. Freundlieb, M. Pies, V. Buck, Diamond
Relat. Mater. 5 (1996) 1478.[10] H. Gomez, D. Durham, X. Xiao, M. Lukitsch, P. Lu, K. Chou, A. Sachdev, A. Kumar,
J. Mater. Process. Technol. 212 (2012) 523.[11] J.G. Buijnsters, P. Shankar, W. Fleischer, W.J.P. van Enckevort, J.J. Schermer, J.J. ter
Meulen, Diamond Relat. Mater. 11 (2002) 536.[12] F.J.G. Silva, A.P.M. Baptista, E. Pereira, V. Teixeira, Q.H. Fan, A.J.S. Fernandes, F.M.
Cost, Diamond Relat. Mater. 11 (2002) 1617.[13] A.J. Perry, Thin Solid Films 107 (1983) 167.[14] S.J. Bull, in: W. Gissler, H.A. Jehn (Eds.), ECSC, EEC, EAEC, 1992, p. 31.[15] H. Ollendorf, D. Schneider, Surf. Coat. Technol. 113 (1999) 86.[16] R. Jaworski, L. Pawlowski, F. Roudet, S. Kozerski, F. Petit, Surf. Coat. Technol. 202
(2008) 2644.[17] S. Nakao, J. Kim, J. Choi, S. Miyagawa, Y. Miyagawa, M. Ikeyama, Surf. Coat. Technol.
201 (2007) 8334.[18] P. Lu, X. Xiao, M. Lukitsch, A. Sachdev, Y.K. Chou, Surf. Coat. Technol. 206 (2011)
1860.[19] X. Xiao, B.W. Sheldon, E. Konca, L.C. Lev, M.J. Lukitsch, Diamond Relat. Mater. 18
(2009) 1114.[20] F. Qin, Y.K. Chou, D. Nolen, R.G. Thompson, Surf. Coat. Technol. 204 (2009) 1056.[21] B. Huang, M. Zhao, T. Zhang, Phil. Mag. 84 (2004) 1233.
