Interface reactions in the Mg-Cu thin films may be referred from the Mg-Cu equilibrium phase diagram, as shown in Figure 3-1. The Mg-Cu amorphous and intermetallic phases might form in the binary thin film.
The binary metallic thin films were fabricated using sputtering deposition. The microstructure was examined by X-ray diffraction (XRD), scanning electron microscopy (SEM) with energy dispersive X-ray spectrometry (EDS), and transmission electron microscopy (TEM). The thermal stability of the Mg-Cu metallic thin films was examined using a differential scanning calorimeter (DSC). The mechanical properties of the Mg-Cu metallic thin films were evaluated by nanoindentater. The flow chart of the experimental procedure is shown in Figure 3-2.
3-1 Materials
Mg and Cu targets used in this study were purchased from Well Being Enterprise Co., Ltd, Taipei, Taiwan. The purity level of the magnesium and copper targets are 99.99%, 99.99%, respectively.
3-2 Sample preparation
3-2-1 Substrate preparation
36
In this experiment, films were deposited on is P-type (100) silicon wafer. In order to avoid some other impurities and greasy dirt adhering to the substrates, the following surface cleaning procedure was adopted.
(1) With deionized (DI) water, an ultrasonic cleaner was used to clean the substrates for 10 minutes to remove the dust and impurities on the surface of substrate.
(2) An ultrasonic cleaner was used to clean the substrates in alcohol for 10 minutes to remove the greasy on the surface of substrate.
(3) After rinsing in alcohol, an ultrasonic cleaner was used to clean it in acetone for 10 minutes.
(4) After rinsing in acetone, we used an ultrasonic cleaner is used to clean the substrates in DI water for 10 minutes.
(5) Air was used to dry the substrates.
3-2-2 Film preparation
In this study, monolayer and multilayer thin films were prepared by DC and RF magnetron sputtering with a target of 50.8 mm in diameter at a working pressure of 3×10-3 torr. There were two kinds of procedures which are listed below:
(1) Mg-Cu monolayer thin films using cosputtering,
(2) Multilayer thin films composing of alternative Mg and Cu layers with the atomic ratios of 3(Mg): 2(Cu) and 1(Mg):4(Cu).
The first type can be fabricated using the co-sputtering deposition process with magnesium and copper targets. A rotary pump was used to achieve medium vacuum, and a cryo-pump was used to achieve a base pressure less than 1×10-6 torr. There was a load-lock chamber, for quick and convenient exchanging of substrates without venting the main
37
chamber.
After achieving the base pressure, Ar was introduced into the chamber, and targets were pre-sputtered by inserting a movable shutter for one minute. During sputtering, the magnesium target was set up on the RF gun with 100 and 50 W, and the copper target was set up on the DC gun with 150, 100, 50, 25, and 15 W, respectively. The substrate is set on the sample holder, 80 mm higher than the target. Working gas was pure argon. Argon flow rate was fixed at 30 standard cubic centimeters per minute (sccm). During deposition, the substrate was rotated with an average speed of 15 rpm for the uniform distribution of the film thickness.
The second type can by fabricated using layer-by-layer deposition with alternative magnesium and copper. The details of the monolayer and multilayer sputtering conditions are summarized in Tables 3-1 and 3-2, respectively. During deposition, the substrate was rotated with an average speed of 15 rpm for the uniform distribution of the film thickness.
The multilayered specimens with the different atomic ratios are named as 20T32, 40N32, 20T14, 40T14, respectively, as shown in Table 3-3. The front numbers of the names represent that the specimens consist of 20 or 40 layers, respectively. Then, the latter numbers of the names mean the calculated compositions of the multilayer. Then, T mean the multilayered films consisting of thick Mg and Cu individual layers, and N mean the multilayered films consisting of thin Mg and Cu individual layers. The co-sputtered specimens of the 100 series under different conditions are named as 100-150, 100-100, 100-50, 100-25, and 100-15. The front part was the power of Mg gun set at 100 W. However, the latter represents the set power of DC Cu gun, 150, 100, 50, 25, and 15 W, respectively. Then, the co-sputtered films of the 50 series under the different conditions are named as 50-150, 50-100, 50-50, and 50-25. The
38
front part represents the power of Mg set at 50 W. The latter numbers mean the power of Cu, as shown in Table 3-4.
3-2-3 Post-deposition treatments
According to the DSC curves of 100-150, the temperature of the post heat-treatment was set at 423 K for the 100-150, 100-100, 100-50 and 100-25 specimens. By using the isothermal heat-treatment, the structural transformation can be understood in different compositions of Cu. Moreover, thermal stabilities at this temperature of these specimens can also be evaluated qualitatively. The process of the heat treatment under a base pressure below 10-3 torr was divided into five steps listed below, and is also shown in Figure 3-3:
(1) 298 K to 343 K at 20 K/min (2) Isothermal at 343 K for 5 minutes (3) 343 K to 423 K at 60 K/min (4) Isothermal at 423 K for 60 minutes
(5) The XRD data were collected per hour after the specimens were cooled down.
In the multilayered case, owning to the discrepancy between the melting points of Mg and Cu, it can be expected that Mg atoms can diffuse into individual Cu layers at about 0.2Tm
(403 K). According to the research of Arcot et al. [62], the diffusion between individual 50-nm-thick Mg and Cu layers would occur around 363 K, much lower than 403 K, due to the large different free energy of Mg2Cu than the Mg-Cu amorphous phase. Hence, the multilayered films were annealed under the low temperature of 363 K, in an attempt to avoid the formation of intermediate phase between the metal and compound phases.
The temperature of the heat treatment in this study was set to be at 363 or 413 K, judged
39
from the DSC curve of the Mg-Cu multilayered thin films [63]. First, the process of the post heat-treatments of 20T32 and 40N32 at 413 K under a base pressure below 10-3 torr was divided to five steps, as shown in Figure 3-4:
(1) 298 K to 343 K at 20 K/min (2) Isothermal at 343 K for 5 minutes (3) 343 K to 413 K at 60 K/min (4) Isothermal at 413 K for 30 minutes
(5) The XRD data were collected per hour after the specimens were cooled down.
In order to remove H2O in the chamber and the specimens, the temperature was first hold at 343 K for 5 minutes. Then, the temperature was raised to the set point, and held for a period of time.
For the specimens of 20T14 and 40N14, another process of the heat treatment at 363 K under a base pressure below 10-3 torr, omitting the isothermal step for remove H2O, were applied, as shown in Figure 3-5:
(1) 298 K to 363 K at 60 K/min (2) Isothermal at 363 K for 30 minutes
(3) The XRD data were collected per hour after the specimens were cooled down.
3-3 Property measurements and analyses
3-3-1 X-ray diffraction
The nature of the Mg-Cu binary thin film fabricated by the magnetic sputtering deposition was examined by X-ray diffraction (XRD). The SIEMENS D5000 X-ray
40
diffractometer with Cu Kα radiation (λ = 1.5406 A。), operated at 40 kV and 30 mA, and equipped with 0.02 mm graphite monochrometer, was utilized. The ranges of the diffraction angle 2θ of co-sputtered and multilayered specimens were within 20o to 60o and 17o to 55o, respectively, at the scanning rate of 0.1o per four seconds.
3-3-2 Preparation of TEM specimens of co-sputtered and multilayered thin films
In order to understand the microstructure of Mg-Cu metallic thin film, plane-view TEM specimens of co-sputtered 100-150 and 100-100 samples were prepared via ion-thinning technique. Then, to prepare the plane-view TEM specimen, free-standing 100-150 and 100-100 samples were bound on a 3-mm slot grid and thinned until the thickness is lower than 100 nm using an ion miller (GATAN PIPS-691 Ion-Miller). Finally, the plane-view specimens were examined using the JOEL 3010 analytical scanning transmission electron microscope (AEM) at 300 kV.
To prepare the cross-section TEM (XTEM) specimens of 20T32 and annealed 20T32, at 413 K under a base pressure below 10-3 torr for 120 min, advanced focus-ion-beam (FIB) technique was employed. The focus ion beam instrument (SMI 3050) belongs to the dual-beam type of FIB, where one beam can provide a second-electron image to observe the appearance of the sample, and another can provide the etching function by Ga+ ion for the XTEM specimen, cross-section image, and patterns, and so on. First, these specimens, fixed on an aluminum holder and cleaned by nitrogen, were moved into the FIB chamber. As shown in Figure 3-6, the approximate 1 μm × 3 μm area was deposited with a carbon film to protect the damage from the Ga+ ion during the process of preparation. Then, the slope-etching was employed at the upper and under areas. At this moment, the archetype of
41
the XTEM specimen in the trapezoid shape was fabricated. The XTEM archetype was retouched two parallel planes, and the thickness of the XTEM specimen was thinned out less than 100 nm. Finally, bottom cutting was employed, and both sides of the XTEM specimen were cutted. The finished XTEM specimens were moved on a carbon-coated Cu grid, as shown Figure 3-7 (a) and (b). The plane-view specimens were examined using the JOEL 3010 AEM at 200 kV.
3-3-2 Qualitative and quantitative constituent analysis
In order to identify the constituent component and confirm the composition percentage of the binary metallic thin films, the samples were characterized by a scanning electron microscopy (SEM) with energy dispersive X-ray spectrometry (EDS). The film surface was selected to examine the quantity of the designed compositions by EDS.
3-3-3 Thermal analysis using differential scanning calorimetry (DSC)
The Perkin Elmer, Pyris Diamond DSC was used for thermal analysis to determine the temperatures of the phase transformation. The as-deposited films were first peeled off, avoiding the effect of the substrate. In DSC analysis, for the non-isothermal heating course, the binary thin films were heated to 523 K at the heating rate of 5 K/min.
3-3-4 Nano-mechanical analysis using nanoindenter
The MTS XP nanoindenter was used to evaluate the nanomechanical properties of the 4-μm-thick Mg-Cu metallic thin films coated on the Si (001) wafer. During the process of indentation, the deeper penetration depth, the higher effect of the substrate. Generally
42
speaking, the examined nano-mechanical properties without the substrate effect were tested within about 10% thickness of the film. Hence, 100-150, 100-100, 100-50 with the Si substrate were tested by the nanoindenter operated in the continuous stiffness measurement (CSM) mode, which involves the application of a very small oscillation force, equal to 45 Hz, to the loading force at a high frequency. From the oscillation of the resulting depth signals one can continuously measure the contact stiffness and calculate the hardness and modulus values [64]. In this study, the 100-150, 100-100, and 100-50 specimens were indented into 10% thickness (400 nm) under a constant strain rate of 5×10-3 s-1.
43