91
Tribology Letters, Vol. 20, No. 1, September 2005 (Ó 2005) DOI: 10.1007/s11249-005-7796-y
Subsurface zone in pure magnesium studied by positron lifetime spectroscopy J. Dryzeka,*, E. Dryzeka, T. Suzukib and R. Yub b
a Institute of Nuclear Physics PAN, ul. Radzikowskiego 152, PL-31-342 Krako´w, Poland Radiation Science Center, High Energy Accelerator Research Organization KEK, 1-1 Oho, 305-0801 Tsukuba, Japan
Received 10 February 2005; accepted 31 July 2005
This paper presents the results of investigation of a subsurface zone created during dry sliding in pure magnesium. Positron annihilation and microhardness techniques were used to detect the defect distribution below the worn surface. Near to the surface there was found a strong gradient of dislocations and defects associated with them, which were extended more than a hundred of micrometers. The decrease of microhardness at a depth less than 30 lm from a worn surface suggested that dynamic recovery took place in pure magnesium. The defect profiles were also studied for samples exposed to compression and blasting. KEY WORDS: subsurface zone, positron annihilation, defect depth profile, magnesium
1. Introduction Positron annihilation spectroscopy has many applications in studies of various condensed matter problems [1]. However, for several years it has been used in tribological investigations. The positron technique is a suitable tool for detecting open volume defects that are also created during the sliding of two bodies. Using this technique, a well defined defect depth distribution was found in the zone below a surface exposed to a friction treatment [2,3]. There are several advantages of using positrons in studies of defects in the subsurface zone. First, this technique is extremely sensitive to vacancies, small vacancy clusters, dislocations and grain boundaries and thus allows us to trace even quantitative changes in the crystal lattice. Second, the positrons prior to annihilation scan a certain region of the sample. Hence, the information about defects is averaged over this region, and this is desired in the case of samples that are characterized by microscopic irregularities on the surface and inhomogeneities below a worn surface. This can be also treated as a disadvantage, but by using a low energy positron beam [1] or a new experimental technique called DSIP (Depth Scanning of the Positron Implantation Profile) [4] one can reduce this problem. Third, the samples after a friction treatment can be directly measured without any further preparation. We should also mention that the physical outline of the positron annihilation technique is well-understood. In former studies we applied successfully the positron technique to studies of the subsurface zone created during the sliding of copper [3] or aluminium [5] against steel. *To whom correspondence should be addressed. E-mail:
[email protected]
First of all, we demonstrated that there is an exponential relationship between the defect concentration and depth below a worn surface. Such a dependency was not found for the case that a sample was exposed to normal loading. The onset of the subsurface zone, which was usually located at a depth of more than one hundred micrometers, is well correlated with the von Misses criterion for the yield [6], and depends on the sliding conditions, i.e., the sliding speed, applied load and distance or time. The type of defect that has been found in the subsurface zone depends mainly on the worn material. The present investigations focused on a study of the subsurface zone in pure magnesium created during a friction treatment. Magnesium has a close-packed hexagonal crystalline structure, low strength and high damping due to the easy motion of dislocations at room temperature. Nevertheless similarly like other hexagonal metals with c/a>1.633 magnesium has only three independent slip systems which may occur on either non closepacked planes or in non close-packed direction. This is not sufficient to accommodate an arbitrary change in the shape as it is in the case of copper or aluminium which posses at least 12 independent slip systems. Thus, we expect different than in these metals mechanism of formation and properties of the subsurface zone. Magnesium and its alloys offer lightweight alternative to conventional metallic alloys, and consequently find structural applications in the automotive industry [7,8]. Despite a growing interest in magnesium alloys, very little data exist on their friction and wear behaviors. However, before we use the positron technique to study magnesium alloys, we first wish to collect data about pure magnesium that in the near future will be treated as a reference. 1023-8883/05/0900–0091/0 Ó 2005 Springer Science+Business Media, Inc.
92
J. Dryzek et al./Subsurface zone in magnesium
In our studies we applied the conventional positron lifetime spectroscopy, which seems to be more convenient for magnesium. The studies were completed using another experimental technique, which is commonly used in studies of the subsurface zone i.e., a microhardness test.
2. Outline of used experimental techniques 2.1. Positron annihilation technique The positron annihilation technique is based on the fact that when an energetic positron enters a solid it rapidly loses energy and becomes thermalized in a time of about 10 ps. Having thermal energy, the positron continues its random walk through the lattice, and finally annihilates with an electron of the sample, emitting mainly two photons of energy near 511 keV in opposite directions. The average lifetime ranges from 100 to 500 ps, depending on the material. If the sample contains defects, like vacancies, microvoids or dislocations that have a negative effective charge, the positron may be trapped and form a bound state. Due to the lower electron density inside these defects than in the host, the positron lifetime is longer than in the bulk. Therefore, measuring the positron lifetime constitutes an appropriate technique for the detection of such defects. However, defects repelling positrons, like interstitial atoms, do not appreciably modify its lifetime. The method is sensitive to defect concentrations as low as 10)7 atomic sites. For recording the positron lifetime spectra we used the conventional fast-fast spectrometer. The spectrometer was constructed of BaF2 based detectors and standard ORTEC electronic units; the time resolution of the system was 240 ps (FWHM). As a positron source 22 Na isotope enveloped into a 7 lm thick kapton foil was used. Its activity was 32 lCi. One should note that the positrons emitted from the source have sufficient energy (Emax=544 keV) to penetrate a certain depth of the sample. For magnesium, the linear absorption coefficient for positrons is equal to ca. 1/144 lm)1 [9]. A distance of 144 lm can be taken as the thickness of the layer penetrated by positrons during measurements, because 84.7% of emitted from this source are stopped in this layer. For that reason this technique is not sensitive to surface defects and we will ignore near surface inhomogeneities, mechanically mixed layer and oxide effects in future consideration. All obtained spectra were deconvoluted using the LT code, subtracting the background and the source component [10].
2.2. Microhardness test For measurements of the microhardness profile of the subsurface zone a universal microhardness test in the
range of microloads was performed. This test is based on a measurement of the indentation depth under a dynamic load and for a controlled load-unload cycle [11]. It provides understanding of not only the total hardness of the tested sample, but also of its plastic component, Young’s modulus and percent elastic recovery. This test is suitable to study of the thin films, and due to the low load is more sensitive to change of the material properties at the atomic scale, and hence is also suitable for studies of the subsurface zone. In our experiment the commercially available Mikro– Combi–Tester produced by CSEM was used. A Vickers indenter with the maximum load of 20 mN was applied. The velocity of the load increase was equal to 40 mN/min.
2.3. The experimental procedure Samples of pure magnesium (99.9% purity) had a cuboid shape of size 10 mm 10 mm 5 mm, and before treatments they were annealed in the flow of N2 gas at a temperature of 400 °C for 3 h, and then slowly cooled to room temperature. After annealing, the samples were etched in a 5% solution of acetic acid in distilled water to reduce their thickness by 100 lm and clean their surface. This is important in order to remove defects that occurred during manufacturing. In fact, in measuring of the positron lifetime spectrum for such virgin samples, only one lifetime component equal to 226±1 ps was resolved. This value corresponds well with the data reported in the literature as the bulk value for magnesium [12]. After that, the sample was located in a tribotester, and a rotated disc made from the martensitic steel (steel SW18 hardness about 670 HV0.1) of diameter 50 mm was pushed to its surface with certain load. The treatment was performed in air, no oxidation was observed. The velocity of the disc relative to the surface of the sample was equal to 5 cm/s. During this test we obtained the value of the friction coefficient equal to 0.24 and the specific wear rate, defined as worn volume per unit sliding distance per unit load equal to (8.0±0.8)10)13 m3N)1 m)1. After sliding, the positron source was located between two identical magnesium samples, and this sandwich was positioned in front of the scintillator detectors of the positron lifetime spectrometer. The positron lifetime spectrum was measured during 24 h to obtain more than 2 106 counts in the spectrum. The samples were then etched to reduce their thickness by about 30 lm, and the next measurement of the positron lifetime spectrum was preformed. Such sequenced measurements allowed us, after deconvolution of each spectrum, to obtain the positron lifetime depth profile. For measurements of the microhardness, the samples were cut down with a low speed diamond saw along the axis normal to the worn surface and polished. The microhardness profile along this axis was measured.
J. Dryzek et al./Subsurface zone in magnesium
3. Experimental results 3.1. Positron lifetime profile In all of the measured spectra, only one lifetime component was resolved. This was a surprise, because usually in deformed metals or alloys, due to the existence of several kinds of defects, two or three components are observed. Magnesium is an exception. The values of the positron lifetime ranged from the bulk value (226 ps) to about 248 ps. One should notice that the lifetime of the trapped positron at the single vacancy in magnesium host is equal to 253 ps [13]. Thus the detected values did not originate from this defect. We can argue that this is due to dislocations decorated by vacancies, or jogs. Such defects located near a dislocation are deformed due to a stress field, and the positron lifetime becomes lower than that in the host vacancy. We can conclude this by analogy to other metals, i.e., aluminium [14], nickel [15] and iron [16] because no theoretical data for magnesium can be found in the literature. The net dislocation line is a weak trap for positrons, but it is accompanied by jogs, vacancies or interstitial atoms produced during its motion caused by plastic deformation. The positrons annihilating in all of those defects and in the bulk contribute to the lifetime spectrum, but due to the finite time resolution of the spectrometer one can resolve only the average value. Thus, the positron lifetime obtained from the deconvolution procedure may be treated as the mean lifetime. The main deformation mode in magnesium is a basal slip, i.e., a slip on the (0001) plane with a (11 20) Burgers vector. Because this vector lies in the basal plane, no plastic strain parallel to the c-axis is present. However,
93
such a strain can be produced by twinning; the easiest and most common is {10 12} [17]. Also, a prismatic slip {10 10} (11 20) and a pyramidal slip {10 11}(11 20) was observed, but their critical resolved shear stress at room temperature was one hundred-times greater than that for basal plane [18]. In figure 1 we depict the dependency of the measured positron lifetime depth profile for four values of the applied load during the friction treatment. In all cases, the sliding distance was equal to 9 m. A common feature is that with an increase of a depth, the positron lifetime gradually decreases and at a certain depth reaches the bulk value. It tags the total range of the subsurface zone, which depends on the applied load. For the lower value of the load, 25 N, it is about 150 lm, and for the highest, 150 N, it is 440 lm (see table 1). The experimental dependencies can be well described by a simple exponential function, sðZÞ ¼ s0 þ a expðz=d0 Þ
ð1Þ
where z is the depth or the thickness of the material etched away and s0, a and d0 are the fitted parameters. The solid line in figure 1 presents the best fit of relation (1) to the experimental points obtained for the highest applied load of 150 N. In table 1 we have gathered the fitted parameters obtained for the other samples. The parameter d0 characterizes the subsurface zone. It is worth noticing that its value increases when the load increases (table 1).The measured positron lifetime can be used to estimate the dislocation density. According to the trapping model, the mean positron lifetime is related to the dislocation density, qd, as follows:
Figure 1. Depth profile of the positron lifetime measured for pure magnesium samples exposed to friction treatment against a rotated martensitic steel disc during 3 min (sliding distance 9 m) with different applied loads. The solid line presents the best fit of relation (1) to the experimental points obtained for a load of 150 N. The dashed line corresponds to the estimated dislocation density depth profile (load 150 N) obtained from relation (2), taking into account relation (1). The shaded region represents the bulk positron lifetime in pure magnesium.
94
J. Dryzek et al./Subsurface zone in magnesium
Table 1. Values of the parameters from relation (1) fitted to the experimental points from figures 1, 2 and 5. The surfaces of well-annealed magnesium samples were damaged in the tribotest by dry sliding against martensitic steel and blasted; the sliding conditions are stated in the first column. Surface Treatment Load 25 N, distance 9 m Load 50 N, distance 9 m Load 100 N, distance 9 m Load 150 N, distance 9 m Load 25 N, distance 9 m Load 25 N, distance 45 m Blasting
s0 (ps)
a (ps)
d0 (lm)
Total range of the subsurface zone (lm)
225.9±0.6 226.2±1.5 225.3±1.7 227.0±1.8 228.6±1.1 230.1±1.0 227.6±0.4
18.5±1.3 16.6±1.5 19.4±1.7 19.0±1.4 21.3±2.2 18.4±1.3 17.8±0.5
40±6 82±20 98±23 137±28 31±8 50.8±9.0 130±10
150±30 200±30 240±30 440±30 150±30 >270 ±30 >500±30
1 þ ssat nrd =b s ¼ sbulk 1 þ sbulk nrd =b where sbulk=226 ps and ssat is the lifetime for a dislocation saturated sample; in our case we assume it is equal to 248 ps. Further, b is the Burger’s vector and m is the trapping efficiency; in the case of a dislocation in magnesium it is equal to 3.23 10)6 s)1cm3 [19]. If we substitute (1) in (2) as the mean positron lifetime, we can estimate the dislocation’s density profile. In figure 1 the dashed line follows such a profile for a sample exposed to sliding with a load of 150 N. It shows that a strong gradient of the dislocation concentration exists in the subsurface zone. One can also notice that the range of the subsurface zone increases with increasing sliding distance. Figure 2 presents the positron lifetime profile obtained for two samples exposed to the sliding treat-
ment with the same load of 100 N, but with different sliding distance. Note that, in both cases the value of the positron lifetime measured directly on the worn surface is c.a. 249±1 ps; thus we can conclude that close to the worn surface the density of the dislocations quickly becomes so high that the positron trapping saturates. With an increase of the sliding distance the gradient slightly decreases.
3.2. Microhardness test The measurement of the microhardness also showed the existence of the subsurface zone. In figure 3(a) we present profiles measured for two selected samples. It is interesting to notice that the microhardness initially increases with a depth from the worn surface, and at a certain depth reaches the maximum, and then decays to
Figure 2. Depth profile of the positron lifetime measured for the pure magnesium samples exposed to the friction treatment against a martensitic steel disc with a load of 25 N at the sliding distance 9 m and 45 m, respectively. The shaded region represents the bulk positron lifetime in pure magnesium.
J. Dryzek et al./Subsurface zone in magnesium
95
Figure 3. Depth profiles of the microhardness (a) and Young’s modulus (b) measured for pure magnesium samples exposed to the friction treatment against a rotated martensitic steel disc during 3 min (sliding distance 9 m) with different applied loads.
the bulk value. The total depth of the microhardness profile is about 120 lm, which is lower than that detected by the positron technique (see table 1). We have also observed a similar fact in the aluminium alloys [20]. We argue that the microhardness test is not a suitable tool for detecting the total range of the subsurface zone induced during surface treatments. This test, in comparison to positron measurements is less sensitive to point defects, a great amount of which is created during, for instance plastic deformation.
Nevertheless, the existence of the maximum of the microhardness profile at a certain depth from a worn surface seems to be interesting. Note, this depth slightly increases with increases of the applied load. This may indicate that a dynamic recovery process takes place in magnesium. Certainly, the recovery process is accompanied mainly by a rearrangement of the dislocation structure, which effect significantly on the microhardness. The rearrangement of the dislocations cannot reduce the point defects concentration and hence this is not observed
96
J. Dryzek et al./Subsurface zone in magnesium
in the positron lifetime profile (figures 1 and 2), which exhibits a monotonic dependency in this depth range. The depth profile of the Young’s modulus presented in figure 3b exhibits a rather weak depth dependency. For a load of 100 N, the maximum in the profile was observed at a depth of 30 lm, but the total changes were within the measurement accuracy. It should be noted that the average value of the Young’s modulus measured using this technique was equal to 33.2±1 GPa, which is lower than the value 45 GPa reported in textbooks. This discrepancy is surprising, because for other tested materials, aluminium alloys and steels both quantities coincidences. Nevertheless, the conventional measurement of the elastic modulus is performed in the macro scale but using the Mikro–Combi–Tester it is in the micro scale. One cannot exclude a discrepancy for materials like magnesium for which the crystalline lattice exhibits some anisotropy. 4. Other surface treatments 4.1. Uniform pressing Well annealed magnesium samples were compressed in a flat geometry between two martensitic steel plates using a press in which pressures equal to 6 MPa, 10 MPa and 12.5 MPa and strains equal to 4, 11 and 17%, respectively, were achieved. After 15 s the pressure was released and the positron experiment was performed as mentioned above. Only one lifetime component was resolved in the spectra, its value versus the depth is depicted in figure 4. The common feature is that the measured profiles depend very weakly on a depth, which is slightly different from the previous results obtained in
aluminium [21] . Note, that an increase of the strain induces an increase of the values of the positron lifetime measured just below the surface. We think that this dependency can be useful if we wish to estimate the strain in the subsurface zone. 4.2. Blasting In our former studies of copper [6] we found similarities between the subsurface zone created during friction and blasting. We decided to find out if in magnesium a similar phenomenon takes place. The surfaces of well annealed magnesium samples were blasted by silicon carbide particles having a diameter less than 0.5 mm with a pressure of 6.5 bar. The positron experiment described above was then performed. Again, in the measured spectra we have found only one lifetime component, which value versus the depth is depicted in figure 5. Note, its value decays with a depth increase, and finally reaches the bulk value. The dependency is well described by an exponential relation (1), and the fitted parameters are presented in table 1. It is visible that the profile of the positron lifetime is almost identical with the profile induced by the friction treatment (figure 1). From this, we can conclude that the subsurface zone induced by sliding results from impacts of the sliding body due to its surface unevenness, or debris particles, and finally induces a depth defect distribution similar to that in the blasting process. Such impacts can distribute the plastic deformation much deeper than the shear stress at the surface. Certainly, they also act directly on the surface, and probably affect the wear process as well. Nevertheless, it is interesting to show that in magnesium this phenomenon is also clearly visible.
Figure 4. Depth profiles of the positron lifetime measured for pure magnesium samples after uniform normal pressing for the values of the thickness reduction equal to 4, 11 and 17%.
J. Dryzek et al./Subsurface zone in magnesium
97
Figure 5. Depth profile of the positron lifetime measured for pure magnesium sample whose surface was blasted with small SiC particles. The solid line presents the best fit of relation (1) to the experimental points. The shaded region represents the bulk positron lifetime in pure magnesium.
5. Conclusions
References
In conclusion, it may be stated that in pure magnesium during friction treatments only dislocations with accompanied defects, like vacancies and jogs, are created. Their concentration decays exponentially with an increases of a depth. The total range of the subsurface zone created during friction treatments is more than one hundred micrometers, and depends on the applied load and sliding distance of the treatment. Measurements of the microhardness indicate that dynamic recovery at a depth less than 40 lm from a worn surface takes place. The subsurface zone created during the blasting process exhibits similarities to the subsurface zone created during friction treatments. This points out that the impacts of the body during the friction treatment can be responsible for creating the subsurface zone below a worn surface.
[1] R. Krause-Rehberg and H.S. Leipner, Positron Annihilation in Semiconductors (Springer-Verlag, Berlin, 1999). [2] J. Dryzek, E. Dryzek, T. Stegemann and B. Cleff, Tribol. Lett. 11 (2001) 29. [4] J. Dryzek, Applied Physics A 81 (2005) 1099. [5] J. Dryzek and E. Dryzek, Tribol. Lett. 17 (2004) 147. [6] J. Dryzek and E. Dryzek, Tribol. Lett. 15 (2003) 309. [7] D.S. Mehta, S.H. Masood and W.Q. Song, J. Mater. Process. Technol. 155–156 (2004) 1526. [8] P.J. Blau and W. Walukas, Tribol. Int. 33 (2000) 573. [9] M. Mourino, H. Lo¨bl and R. Paulin, Phys. Lett. A 71 (1979) 106. [10] J. Kansy, J. Nucl. Instr. and Meth. in Phys. Res. A 374 (1996) 235. [11] W.C. Oliver and G.M. Pharr, J. Mater. Res. 7 (1992) 1564. [12] A. Seeger and F. Banhart, Phys. Stat. Sol. (a) 102 (1987) 171. [13] H.-E. Schaefer, Phys. Stat. Sol. (a) 102 (1987) 47. [14] H. Ha¨kkinen, S. Ma¨kinen and M. Manninen, Europhys. Lett. 9 (1989) 809. [15] T. Riewcharoon, Z. Tang, Y. Nagai and M. Hasegawa, Mater. Sci. Forum 363–365 (2001) 167. [16] T. Onitsuka, M. Ohmura, M. Takenaka and E. Kuramoto, Mater. Sci. Forum 363–365 (2001) 163. [17] E. Tenckhoff, Deformation mechanisms, texture and anisotropy in zirconium and zircaloy (ASTM, Philadelphia, 1988). [18] E.W. Kelley and W.F. Hosford, Trans. AIME 242 (1968) 5. [19] M. Abdelrahman and E. Badawi, Jpn. J. Appl. Phys. 35 (1996) 4728. [20] E. Dryzek and J. Dryzek (in press in Tribology International). [21] J. Dryzek and E. Dryzek, Mater. Sci. Forum 445–446 (2004) 72.
Acknowledgments Measurements of the microhardness were performed in Laboratory for Tribology and Surface Engineering, Faculty of Mechanical Engineering and Robotics, AGH University of Science and Technology Krako´w. This research was supported in the frame of the program COST Action 532 by Committee of Scientific Research (Poland) project No. 620/E-77/SPB/COST/T-08/DWM 49/2004-2005.