UDK 533.5:621.793:669.71 ISSN 1318-0010 Izvirni znanstveni članek KZLTET 33(6)501(1999) M. ČEKADA ET AL.: CHARACTERIZATION OF PVD CRC, CR(C,N) AND CRN HARD COATINGS CHAR AC TER IZA TION OF PVD Cr-C, Cr(C,N) AND Cr-N HARD COAT INGS KARAKTERIZACIJA Cr-C, Cr(C,N) IN Cr-N TRDIH PREVLEK, PRIPRAVLJENIH S PVD POSTOPKOM Miha Čekada 1, Marijan Maček 1 ,2 , Pe ter Panjan 1, Boris Navinšek 1 1 Jožef Stefan Institute, Jamova 39, 1000 Ljubljana, Slovenia 2 University of Ljubljana, Faculty of Electrotechnical Engineering, Tržaška 25, 1000 Ljubljana, Slove nia Prejem rokopisa – received: 1999-11-17; sprejem za objavo – accepted for publication: 1999-11-22 Cr-N hard coat ings have been widely in ves ti gated for over 10 years and suc cess fully im ple me nted in in dus try. Re cently, work has been done on Cr-C hard coat ings and some in dus trial ap pli ca tions were re ported, in par t ic u lar, the pro tec tion of tools for alu minium die-casting. How ever, in for ma tion about the ter nary Cr(C,N) hard coat ing is scarce . We pre pared sev eral Cr(C,N) hard coat ings with dif fer ent C:N ra tios by vary ing the N 2 and C 2H2 flow. Stoichiometric, as well as nonstoichiometric Cr-N and Cr-C coat ings were pre pared for com par i son. The coat ings were d e pos ited by evap o ra tion in a ther mi onic arc ion plat ing BAI730M (Balzers) ap pa ra tus. Pol ished tool steel disks, sil i con wa fers and pol ished alu m i n a ce ram ics were used as sub strates. Ox i da tion of the coat ings was car ried out by heat ing the sam ples at 750-900°C in an ox y gen at mo sphere while mea sur ing the weight gain and the thick ness in crease. The depth pro files of the sam ples, ox i dized at var i ous tem per a tures, were mea sured by Au ger elec tron spec trom e ter (AES). Micro hard ness, ad he sion and sur face rough ness were al so mea sured. We in ves ti gated the in flu ence of the N 2 and C 2H2 par tial pres sures on the above men tioned phys i cal and chem i cal prop er ties. The re sults show that Cr-C coat ings do have the po ten tial for in dus trial ap pli ca tion. Key words: hard coat ings, phys i cal vapour de po si tion, micro hard ness, ad he sion, ox i da ti on re sis tance Raziskave Cr-N trdih prevlek potekajo že več kot 10 let in so se uspešno uveljavile v industriji. Z adnjih nekaj let se raziskuje tudi Cr-C trde prevleke in prišlo je že do prvih industrijskih aplikacij, konkretno za prekrivanje orodij za tlačno litje aluminija. Toda informacij o ternarni Cr(C,N) trdi prevleki je malo. Pripravili smo več Cr(C,N) trdih prevlek z različnim atomskim razmerjem C:N tako, da smo spreminjal i pretok N 2 in C 2H2. Za primerjavo smo pripravili tudi stehiometrične in nestehiometrične Cr-N in Cr-C prevleke. Prevleke s mo pripravili z naparevanjem v termionskem sistemu BAI730M (Balzers) za ionsko prekrivanje. Za podlage smo uporabil i polirane ploščice iz orodnega jekla, silicijeve rezine in polirano aluminij-oksidno keramiko. Vzorce smo oksidirali tako, da smo jih segrevali pri temperaturah 750-900°C v kisikovi atmosferi in merili prirastek mase in debeline. Globinski profil vzorcev, ki smo jih oksidirali pri različnih temperaturah, smo izmerili z Augerjevim elektronskim spektrometrom (AES). Izmerili smo tudi mikrotrdoto, adhezijo in hrapavost površine. Študirali smo vpliv delnega tlaka N 2 in C 2H2 na zgoraj omenjene fizikalne in kemijske lastnosti. Rezultati kažejo, da so Cr-C prevleke potencialno uporabne v industrijske namene. Ključne besede: trde prevleke, vakuumski postopki nanašanja, mikrotrdota, adhezija, oksidacijska ob stojnost 1 IN TRO DUC TION PVD Cr-N coatings have been widely investigated for over 10 years and successfully implemented in industry for the last decade. There are numerous references to their properties and applications 1-6 . PVD Cr-C coatings, on the other hand, are relatively new in this field. Recently, some industrial applications were reported 7 ,8 . However, information about the ternary PVD Cr(C,N) coating is scarce. These papers on Cr(C,N) concentrate on the deposition parameters and phase composition 9,10 , as well as on more application-oriented measurements such as microhardness, adhesion, oxidation and wear resistance 11,12 . We deposited several Cr(C,N) coatings by varying the partial pressures of the reactive gases nitrogen and acetylene. Keeping the working gas (argon) pressure constant the total pressure was varied as well. We investigated the influence of these process parameters on the physical and chemical properties of the coatings. These properties were: atomic composition, KOVINE, ZLITINE, TEHNOLOGIJE 33 (1999) 6 microhardness, adhesion and resistance to oxidation. Rather than being linear, these relationships could exhibit a peak or valley, which would have a great influence on possible applications. Besides the ternary carbonitrides Cr(C,N), the binary coatings Cr-N and Cr-C were also deposited, stoichiometric, as well as substoichiometric. The next step in our work would be the determination of the optimum deposition parameters for possible application of these coatings in industry. 2 EX PER I MEN TAL PRO CE DURE The coatings were deposited by evaporation in a Balzers BAI730M thermionic arc ion plating apparatus at a relatively high temperature (450 ±50°C). The deposition rate was about 4 µm/h and the bias on the substrates was -125V. Details of the deposition procedure are described elsewhere 13 . Chromium was used as the target material, argon served as the working gas and the source (reactive) gases 501 M. ČEKADA ET AL.: CHARACTERIZATION OF PVD CRC, CR(C,N) AND CRN HARD COATINGS for nitrogen (in nitrides) and carbon (in carbides) were N2 and C 2H2, respectively. The working gas (argon) pressure was kept constant at 1.5 ×10 -3 mbar. Four series of coatings were deposited with total pressures of 2, 2.5, 3 and 3.5 ×10 -3 mbar. Subtracting the constant argon pressure, the combined pressures of both reactive gases were 0.5, 1, 1.5 and 2 ×10 -3 mbar. For convenience, from now on the term "total pressure" is used for this quantity. In each series, several coatings were deposited by varying the nitrogen-to-acetylene ratio. Putting the nitrogen partial pressure to zero and introducing only acetylene as the reactive gas, a chromium carbide (Cr-C) coating formed. Introducing only nitrogen gas into the chamber, on the other hand, caused the deposition of a chromium nitride (Cr-N) coating. If both reactive gases were present, a chromium carbonitride (Cr(C,N)) coating formed. The coatings were deposited on polished high-speed steel disks (DIN PMS6-5-3-8) for microhardness and adhesion measurements, and on polished alumina ceramics (R a=25nm) for oxidation tests. Oxidation was studied by annealing the samples in a tube furnace with an oxygen flow. Periodically, the specimens were taken out of the furnace and weighed using a precise microbalance (Mettler Toledo UMT2). For the microhardness measurements a Mitutoyo MVK-H2 microhardness tester was used. Adhesion was evaluated by the scratch test method using a Revetest Scratch Tester (loading rate 100N/min, scratching rate 10mm/min, loading range 0-100N). Thickness and roughness were measured with a Taylor-Hobson Talysurf 2 profilometer. The atomic ratios of the coatings were determined by Auger electron spectrometer (AES). 3 RE SULTS AND DIS CUS SION Atomic composition Fig. 1 shows the partial pressures of both reactive gases (N 2 and C 2H 2) during the deposition of the coatings. The coatings situated on the x-axis are chromium carbides since there was no nitrogen in the chamber. The coatings situated on the y-axis were deposited without acetylene and are chromium nitrides. The coatings in between are chromium carbonitrides. The coatings on the same inclined lines were deposited at the same total pressures. Most of our analysis is based on the dependence of a particular property on the nitrogen-to-acetylene ratio at a certain total pressure. The dependence of the atomic ratios of the coatings on the pressures is presented in more detail in Fig. 2 . The upper plot (Fig. 2a) shows the atomic percentage of chromium in the coatings as a function of the reactive gases ratio. At higher total pressures (1.5 and 2 ×10 -3 mbar) the percentage of chromium is constant at 50%, regardless of the reactive gases composition. This means 502 p(Ar)= 1,5.10Č mbar pfCjHČIIOČmbar] Fig ure 1: Par tial pres sure of acet y lene (x-axis) and par tial pres sure of ni tro gen (y-axis) for dif fer ent de po si tion runs. Atomic composition was determined by AES. Each dot rep re sents one de po si tion run Slika 1: Delni tlak acetilena (os x) in parcialni tlak dušika (os y) pri različnih depozicijah. Atomska sestava je bila določena z AES. Vsaka točka predstavlja eno depozicijo. that we always get a stoichiometric coating. At lower total pressures however, the flux of reactive gases is too low to build a stoichiometric nitride or carbide coating. Instead, a substoichiometric coating is formed. This effect is more pronounced in the carbides. At the lowest acetylene pressure (0.5 ×10 -3 mbar), no carbon is incorporated into the coating at all – in this case a pure chromium coating is formed. In Fig. 2b the atomic percentage of nitrogen in the coating is presented as a function of the reactive gases 100% <> 40% 20% 0% p(N2+C3H3)Š10Čmbar] 0 0,5 Q 1 ? 1,5 A 2 Q% Cr-G 25% 50% 75% p(Ns]/p(NČCsH;.) Cr-N ino% 100% _ 80% z ď 60% U S" 40% ° 20% 0% (b)0% Č ? 1,5 Cr-C 25% 507ů 75% 0(N2]/p(NČC2H2) Cr-N 100% Fig ure 2: a) Atomic per cent age of chro mium as a func tion of re ac tive gases ra tio; b) Atomic per cent age of ni tro gen as a func tion of re ac tive gases ra tio Slika 2: a) Atomski delež kroma kot funkcija razmerja reaktivnih plinov; b) Atomski delež dušika kot funkcija razmerja reaktivnih plinov KOVINE, ZLITINE, TEHNOLOGIJE 33 (1999) 6 M. ČEKADA ET AL.: CHARACTERIZATION OF PVD CRC, CR(C,N) AND CRN HARD COATINGS ratio. A linear relationship would be expected – the higher the percentage of acetylene in the chamber, the higher the percentage of carbon in the coating. From the figure it is clear (especially for 1.5 and 2 ×10 -3 mbar total pressures) that this relationship is not linear; it has a rather S-like dependence. Small quantities of acetylene in a predominantly nitrogen atmosphere do not alter the nitride significantly. Conversely, small quantities of nitrogen in a predominantly acetylene atmosphere have little influence on the carbide. Mechanical properties The microhardness was measured for all samples. In order to avoid substrate influence the indentation load should be as low as possible, however such indentations are very small and difficult to evaluate. The lowest load applied was 25g force, which still involves some substrate influence. We determined that the Cr-C coatings have a smaller microhardness than the Cr-N coatings. The results for a 4 µm thick coating deposited on HSS steel (900 HV 0.025 ) are as follows: Cr-N 2500-2700 HV 0.025 , Cr-C 1600-2000 HV 0.025 . The intermediate Cr(C,N) phases have similar values to the Cr-N, but there is a significant drop at the carbon-rich side. Differences between the coatings prepared at different total pressures (but with the same reactive gases ratio) are not observable. The adhesion of the coatings was also evaluated on an HSS substrate. There are many criteria for the evaluation of the adhesion but the most reproducible quantity is the critical force for total delamination of the coating, which correlates with the scratching force jump. Due to high scattering of the results, only basic assumptions can be made. The critical force for the Cr-N and the Cr-C coatings is around 50N. The adhesion of intermediate Cr(C,N) coatings is 10-20 N lower, which could present an obstacle to possible applications. Oxidation The oxidation was studied in the range 750-900°C. In all cases, the oxidation proceeded parabolically as described by the diffusion theory for at least one hour of the oxidation, so the coefficient of parabolic oxidation could be determined. The higher the coefficient, the faster the oxidation. Fig. 3 presents this coefficient as a function of the reactive gases ratio for the total pressure of 1.5 ×10 -3 mbar. The behavior at other total pressures does not differ significantly. Two basic assumptions can be made. First, the oxidation resistance of the ternary phases is much weaker than that of the binary coatings Cr-N and Cr-C. At all temperatures there is a peak, i.e. the weakest oxidation resistance, at an approximately equal concentration of carbon and nitrogen. The second conclusion is that the oxidation resistance of the binary 25% 50% 75%|Cr-N| 100% p