Wear 196 (1996) 197-201
Tribological study of nitrogenimplantedniobium T.M. Wing, X.J. Wang, W.J. Wang *, J. Shi Departmentof MaterlalsScience.Lanz~u Universily,Lanzhoa.730000.PR China Received7 April1995;accepted7 December1995
Abstract Niobiumsinglecrystalswith ( I 1! ) erientationwere implantedby 1I0 key nitrogenionsup to a dose of 5 x 10~ ions cm- ' at a temperature less than 100 °C. Before and after the implantation,~bological propertiesinvolvingm i ~ , f~'fion coefficientand wear rate against hardened GCrl5 steel ball were determined.The wear tracks and the scars were also examined by scanning electron microscopyand an electronprobe micro-ana!yzer.The structureof the implantedlayer was characterizedby ghneing-angleX-raydiffraelion.The resultsshowed that the I l0 keV nitrogen implantationled to an increase of 18.5% in microhantnessand big reduc~onsof both friction coefficient(from unimplanted0,95 to implanted0.20 ~ 0.21 ) and wearrate ( from?.8 × 10- 7to !. I x i0 -7 mm3 nun- t), For the unimphmed niobium,features indicatingadhesionand/or severewear were observedon wear tracks,whilefor the implantedniobium,only parallelgmovesand micmeracks with higher Fe content were formedon wear tracks.The tribologicalimprovementwas explainedin terms ofimplantntion-inducedfemmtion of niobium nitrides which was revealedby glancing-angleX-mydiffractionanalysis.
Keywords:ton implantation; Tiibology;Niobiumcr/stal
It is known that nitrogen ion (N +) implantation can improve the tri~logicai properties of many me~s. Because of this, many nitrogen-metal systems have been investigated in order to clarify the formation mechanisms of various metal nitrides and their influences on surface tribological properties [ 1,2]. For the refractory metal niobium, however, although some nitrogen ion implantation experiments have been made in the past ten years, the main purposes of most investigators were not related to implantation-induced tribological property changes but to structural changes and physical property changes. In particular, many investigators stressed the superconducting properties because the NbN phase is a superconducting phase. For example, to the authors' knowledge, G. Linker [ 3 ] studied the N + implantation-induceddefectstructure which was thought to be the mason for changes of the superconducting transition temperature. Based on Linker's experimental data, S.L Ran et ai. [4,5] explained the defect structure of N + implanted Nb films. M.I. Guseva and G.V. Gordeeva  determined the phase transitions in N + implanted layers and found that NbzN and NbN phases were formed.They alsobelievedthattheformationofthesehi,des are responsiblefor hardnessimprovement and wear resis* Corresponding author.Tell: +86-931-8843000-3806; fax: +86-931-
8881996; e-mall:[email protected]
0043-1~819615i5.00© 1996FJsev~ ScienceS.A./allrigl~smen'ed SSD10043.1648( 95 ) 06905-4
lance of the implanted layers, but no detailed results concerning wear or tribologicai behavior have been published. Recently, by means of Rutherfonl b a c k ~ g spectroscopy, Auger electron spectroscopy and X-ray diffraction analysis, T. Fujihanaet. al.  repomdthe N + implantationinduced phases in niobimn sheets and found that different niobium nitrides were formed in the implanted layers. Niobium nitrides me hard materials, and like other metal nitrides, such as TiN, they have potential in the applicmi~ area of surface medifxeation. Because the niobium nitrides can be formed in the near surface layer of niobium by using N + implantation, so it is inter-~ing m investigate the tribelogicalproperty changes of niobium beforeand afterN + implantations.Recently.we have intphnted the niobium sheetsby ll0keV N + and chmctetized the tribological behaviorsof the sheetsbefot'eand afterthe implantation.In the present paper, we will report some preliminary results and discuss the wear mechanisms briefly.
2. SaperUum~detaes The subsumes used were singlecrystalNb sheetsof 18× 1.8mm ~ with a putty of 99.99~. which were mechanicallypolished with alumina followed by electro-polishing in a solution of HI: and HzSO,. The roughness of the mbslrate
EK Wanget al./ Wear196(1996)197-201
surfaces was better than 0.1 p.m, The orientation of the surface was (111). After polishing and ultrasonic cleaning, the substrates were implanted by 110 keV N + up to a dose of 5× 10s7 ions cm -2. The implantation was performed by using an ion accelerator in the Institute of Modern Physics, Chinese Academy of Sciences. During the implantation, the chamber pressure was fixed as 1.3 × 10 -3 Pa. The beam current density was 30p, A cm -2. The substrate holder was cooled by flowing water and the temperature rise of the substratus was lower than ~ 100 K. In order to avoid the channeling effect, the angle between the incident ion beam and the ( 111 ) direction of the substrates was 7°. By using TRIM 91, we estimated that the projected range of the ions in Nb was about 100 nm. After N + implantation, glancing-angle X-ray diffraction (GXRD) was used to characterize the structure of the implanted layer. The glancing angle between the incident Xray and the specimen surface was 5 °. Before and after N + implantation, a Vickers microhardness tester with a load of 10 g was used to determine the microhardness of the specimen surfaces and, for each specimen, the microhardness given below was the arithmetic mean of 20 measured values. A ball-on-disc tfibometer was used to measure the friction coefficient between the hardened GCrl5 steel ball and specimen and the wear of specimen. The balls had the same roughness as the substrate surfaces and their Vickers microhardness was 710 kg mm- 2 ( for detailed compositions of the steel, see ref, ). Fig. t shows the schematic diagram of the bali-on-disc apparatus. During this test, the GCrI5 steel ball of 3 mm in diameter was slid in a single direction under a load of 300 g. The sliding speed and the track length were 1.5 mm s-t and 5 mm respectively. The friction coefficient vs. sliding number was given automatically by the tester. After a certain amount of sliding, the wear area across the wear track was determined by using a surface profiIometer. The morphology and the elemental distributions of the wear track were then examined by scanning electron microscopy (SEM) and an electron probe micro-analyzer (EPMA). In order to get the relation of wear area vs. sliding number, several tracks with the final sliding number of 50, 100, 150, 250, 400, 600 and 800 were tested and the wear rate was then calculated after plotting wear area vs. sliding number. For each track with final sliding number, a new GCrl5 steel ball was used. For the implanted and the unimplanted specimens, the above tribological experiments were run three times and the results obtained showed good agreement with one another.
S Cl ,
Fig. I. Schematicdiagramofthebali-on-discapparatus.
3. Results and discussion 3.1. X-ray diffraction analysis During the GXRD measurement, the specimen holder did not rotate, while specimen-position-independent peaks were observed except for the specimen-dependent diffraction peaks of BCC-Nb. A typical diffraction pattern for the ion implanted surfaces is shown in Fig. 2, The pattern was obtained under a certain specimen position at which the diffraction peaks ofNb substrate (and also of the solid solution BCC-Nb(N)) were the lowest in intensities. The identification results of the peaks are listed in Table 1. It can be seen that, except for the diffraction peaks of BCC-Nb (and of BCC-Nb(N)), the GXRD pattern suggests the formation of one or a mixture of such niobium nitrides as cubic-Nb4N3.~, tetra-Nb4N3, cubic-HbN and HCP-Nb2N. In order to separate the diffraction peaks, we tried doing experiments several times but failed because the GXRD peaks of these niobium nitrides lie very close to each other according to the JCPDS cards, So, from the GXRD analysis, we can only conclude that one or more niobium nitrides as cubic-Nb4N3.92, tetraNb4N3, cubic-NbN and HCP-Nb2Nwas/were formed after the implantation. But we cannot decide the exact niobium
~ 2 TEETA|lhlg,I
Fig, 2, X-my diffraction pat~ typically observed after N"~ ion implamtiou, Table 1 Phaseidentificationresultsof GXRDpatternshownin Fig.2 Peaknuml~r
P~eaad (h~) identified
2.5364 2.5180 2.5360 2.4740 2.3380 2.3330 2.1965 2.1910 2.1960 1.5533 !.5510 1.5530 1.5280 1.3249 1.3210 1,3240 1,3230
Cubic-Nb, N3.~(! I 1) Tetm-Nb4N3(I 11) Cubic-IVoN(l 11) HCP-Nb2N(002 ) BCC-Nb(I I0) HCP-Nb2N(101) Cubic-Nb4Ns.n(200) Tetra.Nb4Ns(2(}0 ) Cubic-NbN(200) Cubic-Nb4N3.~(220) Tetm-Nb4N3(220) Cebio.NbN(220) HCP-Nb.zN (I 10) Cub|c-Nb4N3.s~(31 I) Tetra-Nb4N3(311) Cebic-NbN( 31I) HCP-Nb2N(200)
' The sumdazdvaluescomefromthe powderdiffractionfilesof the Joint Committeeon PowderDiffractionStandardsUCPDS).
TM, Wanget al,/ Wear196(1996)197-201
nitride phases formed. This is the same as the report by Fujihanaet al. .
199 N~ 7O
3.2. Tribologicai characterization
Measurementindicatedthat, due to the implantation,the microhauiness of Nb surfaces changed from 144 to 171 kg mm-z, i.e. an increase of 18.5% was achieved. (Because the implantedlayer is very thin, so the measured microhardness is only a mean value of the layer plus the substrate.Theintrinsichardnessof the layershouldbe higher than the measuredvalue.) The friction coefficientmeasurement showed that, in the range of sliding numberwhich we used (maximum: 800 times), the friction coefficientof the implantedsurfacehas a very stablevalueof 0.20-O.21; while for the unimplantedone, the frictioncoefficienthas an unstable stage at the very beginningof the sliding and then, after about 400 slidings, it reaches a steady value of about 0.95, see Fig. 3. Thus, compared with the original subsUate,the implantation improved the friction coefficient greatly. Becauseof this improvement,we expectthat the wear should alsobe reduced.Measurementconfirmedthisdeductionwell, as indicatedin Fig. 4. In this figure,we findthat the wearrate of the implanted specimen is much lower than that of the unimplantedone (the wear rates for implantedand unimplanted specimens,which were calculatedfrom the figure, are 1.1XlO-~mm~mm-~ and 3.8xlO-Tmm3mm-I respectively).We think that the improvementsin microhardheSS, friction coefficientand wear rate were caused by the implantation-induced formation of niobium nitfides as describedabove. Of com'se,the carboncontaminationof the substrate surface during the implantation(see next section) may also have made a certaincontribution. 3.3, Examination of wear tracks
Fig. 5 shows the SEM and EPMA results for the unimplanted specimens.For the track of 50 slidings,we find the features indicating adhesion, plastic deformation and the existence of wear debris along the wear track and both its sides (Fig. 5, parts (a) and (b)). Elementalmaps of C, O,
ooOOOOOO • •
oo O0 w 0,6 II4PI,,~IkN~EO & A & & & ,t
~2 JdlkA && I
m ~ m 10o ~ m S~mNG I ~ R
Fig. 3. F~tion coefficient~;s. slidingnumberbeforeand after N+
11 INKtI~ & A
i i n i i i 600 gg9 11100
~ i O l l tUq~q Fig. 4. Wear anmvs. slidingnumb~beforeand ~ e r N÷ i m p .
Fe and Nb showthat,exceptfor uniformdistributionsof trace C and 0 whichwere absorbedby the surface, the elemental distributions are not tmifonn for Fe and Nb, as shown in Fig. 5, parts (c) and (d). This implies that no oxidation occurred during the wear and proves the idea of adhesion. For the track of 600 slidings,however, there exists a large amount of debris along the track and both its sides, Fig. 5, parts (e) and (f). The Fe mapin Fig. 5(g) "~owsthat alarge material transferoccurredbetweenthe C-Crl5 steel ball and the specimen, for a non-uniform distribution of a large amount of Fo can be observed on the wear track. Similarly, except for the uniformtrace existenceof C and O, the distribution of bib on the wear track is also not uniform (see Fig. 5(h)). Thesesuggestthat the wear mechanismbetween the hardenedGCrI5 steelball and the unimplantedNb sheets seems to be an adhesive type at the very beginningof the sliding. With the increase of sliding nember, a lot of wear debris was produced,and then severe abrasivewear would occur.But from the trackfeatureswe tend to believethat both the adhesiveand the abrasivetypes existed at higher sliding numbers. The SEM and EPMA results for the implanted Nb are shownin Fig. 6. It is clear that the characteristicsof the tracks are totallydifferentfromthoseof the unimplantedones.After 50 cycles, the track looks smooth. Song shallow paratl¢l grooveswereformedalongthe slidingdirection.Meanwhile, an outstandingfeature is the presenceof some microcracks on the tracks, Fig. 6, parts (a) and (b). Exuminafionfor C, O, Fe and Nb elementsshowedthat, exceptfor uniformtrace O, there existsa higher,uniformC componenton both worn and unwornparts, Fig. 6(c). This indicatesthat carbon contaminationoccurredduringthe implamtionand no oxidation happenedduring the wear. Besides,there exists a higher Fe element only in the microcracks,Fig. 6, lxa'ts (d) and (e). After 600 cycles, the featureof the wear track is almost the same except for deeper grooves and denser microeracks, Fig. 6, parts (f) and (g). These strongly suggest that the adhesive tendency between the GCrI5 steel ball and the Nb sheetswas subslaa(iallyreducedbecauseof the implantation.AccouUng to Dearnaley ,we considerthat,after
T~M.Wang etal,i Wear 196 (19o~)197-201
(a) 2oox, (b) 6oox; (c) and (d) ckmentalngl)Sofl:c and I¢0c ~ g to (b); (e) ud (f) microgrg)lnof arrack after600 slidinp, originalmagnifcalm (e) 200X, (f) 600x; (g) and (h) ~meatal raps ol'Fe s~l Nb coaespomlingto (f).
Fig. 6. SEM~ FJ3MA~UI~ fro"implmstcd~ : (a) and ( b ) ~ of it track ~ 50 slidings,originalmagnification(a) 200x, (b) 600X; (c)-(e) dcazmal mapsof C, Fe a,'KINI)comnpondingto (b); (f) micrognlphof a trick after600 slidings,odgiw,l magnification600x; (g) elementalmap ofl~ coacspolldingto (f).
T.M. Wang etal./Wear196 f1996)197-201
implantation, the wear mechanism is a mild abrasive type, no oxidation wear or other mechanisms occur. As to contributions to the reduction, we believe that, in ~tdition to the formation of implantation-induced niobium nitrides as described above, carbon contamination might also have had some beneficial effects. Because the implanted layer was harder than the Nb substrate, it could not fit the shear deformation of the substrate during the sliding, so the formation of local cracks on the worn surfaces is understandable. The existence of a higher Fe content in these cracks can be explained in terms of the scraping effect of the cracks on the GCrl5 steel ball.
4. Conclusions 110 key N + implantation on single crystal niobium sheets led to an increase of microhardness and big reductions both in friction coefficient and in wear rat.e against hardened GCrl5 steel ball. Because of the implantation,the wear mechanism between Nb sheets and GCrI5 steel ball changed from adhesive and/or adhesive plus abrasive type (sliding timedependent) to mild abrasive type. It is believed that the implantation-induced niobium nitrides and carbon contamination were responsible for the above tribological
changes. Acknowledgements This research was supported by both the National Natural Science Foundation of China and the Research Foundation of the Laboratory of Solid Lubrication, Lan~ou Institute of Chemical physics, Chinese Academy of Sciences. Prof. X.S. Zhang and Dr. J. Zhang at the Laboratory of Solid Lubrication, Lanzhou Institute of Chemical Physics, Chinese Academy of Sciences are acknowledged for their help in measuring tribological properties. Prof. Z.L. Xiao and Mr. J.Y. Wang at the Institute of Modem Physics, Chinese Academy of Sciences are also thanked for their help in performing nitrogen ion implantation.
References [ 1] H. Dimigen, K. Kobs, R. Leutcneckea,H. Ryssel and P. Eichinger, Mater.$ci.Eng,,69 (1985) 181.
 DJ.,. Willianuon,L. Waag, R. Wci and PJ. Wilbur, Mater. l,at, 9 (1990) 302.  G. Linker,lCu:l. lnstrur~ tdeth., 182-183 (1981) 501.  S.I.Rao,CR. HousImand K.S.G t a b o ~ N~d. Ins/rum.M¢~, BI8 (1986) 47. [51 S.l.Rao, C.R. Houska realK.S.Gtabowski,Nucl.Instru, m. Metb.,B27 (.1987) 396.  M.I.Gusevaand G.V.Gordeeva, P~s.Stat, Sol,.~95 (1986) 385.  T. Fujibana, Y. Okabe, K. Takahmlfiand M. Iwaki, Nat/. Instna~ MetL B45 (1990) 669.  K. Yu.H.D.Li, X.Z.ZlmngandJ.H.Tt~,Nuci.lnstru~ Melg,209210 (1983) 1063.  O. DeamaIey,Mater.Sci. Eng.,69 (1985) t39.
Biographies Professor T.M. Wang was born in 1940. Fie graduated from the Department of Physics, Lanzhou University in 1964. Since then, he has worked successively in the Depa'anent of Physics and the Department of Mmrials Science of Lanzhou University. In 1987, he was awarded a doctorate in engineering at the University of Tokyo, Japan. He is now working in the Department of Materials Science, L a n ~ University and has been head of the depm'Uncntfor seven years. His main research fields are uibology of the ion implantalioninduced layers of ma~als and defects in metals. Associate Professor WJ. Wang was born in 1964. He graduated from the Depamncm of Physics, Lanzhou Univmity in 1987 and obtained his Master's dcgn= in science at the same time. Since then, he has worked in the Ikpamnont of Materials Science of Lanzhou University. In 1992, he was appointed the younge~ associate professor of the university. He will receive a doctorate in science in 1997. His present research fields include: ion implantation modification of materials' surfaces and thin film physics and ~hniques. X.J. Wang is a postgraduate student, born in 1968. When he undertook the present study, he was a graduate student in the Department of Materials Science, Lanzhou University. Now he is studying for a doctorate as a postgradum¢ student at East China Normal University. J. Shi is a postgraduate student, born in 1964.While he undertook the present study, he was a lecturer in the Depamnent of Materials ~ienoe, Lanzhou University. He is now studying for a doctorate in engineering at the DeparUnentof Metallurgical E n g i ~ n g , Tokyo Institute of Technology, Japan.