«STRUCTURE AND PROPERTIES OF ELECTRODEPOSITED NANOCRYSTALLINE NI AND NI-FE ALLOY CONTINUOUS FOILS by Jason Derek Giallonardo A thesis submitted in ...»
In this chapter, the results from the synthesis method used to produce the nanocrystalline Ni and Ni-Fe alloy foils along with a series of microstructural characterizations is presented. The microstructural characterization is grouped into four main categories: (1) scanning electron microscopy, (2) transmission electron microscopy, (3) Xray diffraction, and (4) thermal analysis. The main objective is to divulge as much information as possible in order to explain the macroscopic behaviour of these materials to be discussed in later chapters, including, the hardness, Young’s modulus and internal stresses.
4.2. Synthesis of Nanocrystalline Ni and Ni-Fe Alloys A summary of the materials produced for this study, using the method described in Chapter 3 (Section 3.2), is given in Table 4.1. It includes the compositional analysis determined using EDS for Fe and the LECO analyzer for quantification of S and C.
Synthesis of the Ni samples was carried out by varying the concentration of Nanovate™ A24 in the electroplating solution. As a result, there is a clear trend in the relationship between the Nanovate™ A24 concentration in the solution and the bulk S concentration in the deposit. Note that even when there is no Nanovate™ A24 in the electrolyte, i.e. for sample no. 1, there is still a measureable quantity of S. The source of the S in this case is likely the SO42- ions which are present in the electrolyte and/or the proprietary wetting agent (Nanovate™ B16) which is a mixture of organic species that may contain S. Fig. 4.3 presents a plot of the S concentration in the deposit vs. the Nanovate™ A24 concentration in the electrolyte. Initially, there is a sharp increasing S concentration
trend with increasing Nanovate™ A24 concentration that eventually plateaus. The C concentration in the deposit for samples 1, 2, 3, 4 and 5 do not display any notable trend.
S concentration in the nanocrystalline Ni deposit vs. the Nanovate™ A24 concentration in the electrolyte used to produce nanocrystalline Ni samples no. 1-5.
The nanocrystalline Ni-Fe samples were produced by holding the Nanovate™ A24 concentration constant at 2 g/L and varying the concentration of FeSO4·5H2O in the electrolyte. Fig. 4.4 presents a plot of the Fe concentration in the deposit vs. the FeSO4·5H2O concentration in the electrolyte. Similar to the findings of Cheung et al. (1995) and Li and Ebrahimi (2003), effectively increasing the Fe metal content in the electrolyte increases the Fe concentration in the deposit. In this particular case, the relationship between the Fe concentration in the deposit and the FeSO4·5H2O in the electrolyte is approximately linear. The S in the deposit for the nanocrystalline Ni-Fe samples remains relatively constant
Plot of Fe concentration in the deposit vs. FeSO4·5H2O in the electrolyte used to produce nanocrystalline Ni-Fe samples no. 2, 6, 7, 8 and 9.
Nanovate™ A24 concentration was held at a constant value of 2 g/L. On the other hand, the C concentration is somewhat variable and does not seem to have any trend. In the case of all samples produced, it should be noted that there are multiple sources of C, including the proprietary grain refiner/stress reducer (Nanovate™ A24), the proprietary wetting agent (Nanovate™ B16), and specific to the Ni-Fe electrolyte, the proprietary complexing agent (Nanovate™ C77). It is likely that the variability of the C concentration is a result of the varying “free” complexing agent concentration. The “free” complexing agent concentration is considered to be the concentration of complexing agent which has not formed a ligand with Fe3+ (ferric) ions. The concentration of Fe3+ ions can depend on a number of factors including the O2 concentration in the electrolyte. The consequence of high O2 concentrations
in the electrolyte is an increased concentration of Fe3+ ions due to the oxidation of the Fe2+ ion. Thus, it is likely that the variability of the C in the deposit for the nanocrystalline Ni-Fe alloys is dependent on the effective concentration of organic species in the electrolyte.
4.3. Deposit Quality and Morphology SEM images of various samples were taken to evaluate their surface quality and morphology. Images of the “free” anode facing surface were taken. Shown in Fig. 4.5-4.7 are examples of the surfaces observed at 60,000 X magnification. Additional images of these and the other samples may be found in Appendix A. Sample no. 1 (Ni), shown in Fig. 4.5, displays evidence of faceting which is indicative of relatively large grain sizes. The rest of the samples (Fig. 4.6-4.7) displayed generally smooth surfaces with no evidence of faceting indicating that the grain sizes are relatively small. All samples were without any signs of pores or microcracks. Some surface contamination (dark patchy areas) is visible in the images, which becomes more prominent with increasing Fe concentration.
In previous studies, electrodeposited nanocrystalline Ni [El-Sherik and Erb (1995)] were found to have a cauliflower-like morphology that was affected by the saccharin concentration in the electrolyte. As cauliflower-like features were not observed in the current images, differences in processing parameters, e.g., the current density at which the electrodeposition process was operated is likely to be one of several influencing factors and therefore, should be taken into consideration. High current densities tend to induce the formation of nodules while low current densities tend to promote leveling or smoothing of the surface. In addition to this, past studies [e.g., El-Sherik and Erb (1995)] used a stationary Chapter 4 – Materials Synthesis and Characterization
flat Ti cathode in a confined space to produce samples whereas in the present case, a rotating Ti curved (drum) surface was employed. Sample size may also be a factor, where in this case, foils were substantially larger compared to the coupon samples produced by El-Sherik and Erb (1995).
In the case of the Ni-Fe samples, there is some consistency with previous studies, e.g., Cheung et al. (1995), Czerwinski (1998) and Ebrahimi and Li (2003), whereby the surface was in fact generally smooth. Similar to the current study, Cheung et al. (1995) observed no signs of microcracking. However, in the case of Czerwinski (1998) and Ebrahimi and Li (2003) significant microcracking was observed. The presence of microcracking was owed to the internal stresses in the electrodeposits which are high enough to exceed the fracture strength of the material [Czerwinski (1996), Czerwinski (1998), Czerwinski and Kedzierksi (1997)]. The microcracking observed by Czerwinski (1998) and Ebrahimi and Li (2003) was owed to the absence of organic additives in the electrolyte which behave as stress reducers. In the present study and that of Cheung et al. (1995) a stress reducer was employed in part to prevent the onset of microcracking.
4.4. Transmission Electron Microscopy (TEM) 4.4.1. Conventional TEM bright-field (DF), dark-field (DF) images, and selected area diffraction (SAD) patterns are presented in Fig. 4.8 for sample no. 5 (Ni, 23 nm) and Fig. 4.9 for sample no. 7 (Ni-16wt.%Fe, 12 nm). The images for the remaining samples can be found in
0.10 0.05 0.00
0.2 0.1 0.0
TEM images of sample no. 7 (Ni-16wt.%Fe, 12 nm): (a) BF image, (b) DF image, (c) SAD pattern, and (d) grain size distribution.
Appendix B. All diffraction patterns display characteristics of an fcc structured material.
For samples no. 2-9, the dark field images were used to determine the average grain sizes by determining at least 200 equivalent-circle diameters for each material. The grain sizes that are presented in Table 4.2 have error values corresponding to one standard deviation and are rounded to the nearest whole number.
The grain size for sample no. 1 (*) was extrapolated from the Hall-Petch relationship that will be discussed in Chapter 5. Included in Table 4.2 are the respective total intercrystal volume fraction, f ic, grain boundary volume fraction, f gb, and triple junction volume fraction, f tj, for the purpose of illustrating the significance in the presence of these defects with progressively decreasing grain size. Volume fractions were calculated by using a grain boundary thickness, Δ, of 1 nm using Eq. 2-1, 2-2, and 2-3 [Palumbo et al. (1990)]. The intercrystal volume fraction increases rapidly at grain sizes less than 20 nm. As an example,
at a grain size of 10 nm, approximately 27% of the material consists of intercrystal defects which represent a significant fraction of the total volume.
distribution is typically evident. In order to further illustrate the distribution of grain sizes within a given sample a cumulative volume fraction vs. grain size plot is presented in Fig.
4.10. The use of this analysis for illustrating grain size distribution is important since it provides a more realistic interpretation of volume fractions of grains with a given grain size present in the material in comparison the grain size distributions presented in Fig. 4.8 and
4.9. That is, when converted to volume fraction, the distribution of size of grains which make-up of the material becomes more evident [Soong (2009)]. In this case, it can be seen
that samples no. 7-9 have a relatively narrow grain volume fraction distribution, however, as the average grain size increases greater than 20 nm a progressively broader grain volume fraction distribution is observed for samples no. 2-5.
During the synthesis of the samples, changes were made to the chemical constituents in the electrolyte that can have a considerable effect on the microstructure of the resulting materials (see Table 4.2). The nanocrystalline Ni samples (no. 1-5) were produced by gradually increasing the Nanovate™ A24 concentration in the electrolyte. In addition to the increase of S concentration in the deposit, gradually increasing the Nanovate™ A24
Plot of grain size vs. Nanovate™ A24 concentration in the electrolyte used to produce the nanocrystalline Ni samples no. 1-5.
concentration in the electrolyte was also found to cause a substantial decrease in grain size which is similar to the findings of El-Sherik and Erb (1995). Fig. 4.11 presents a plot of resulting grain size vs. Nanovate™ A24 concentration in the electrolyte. In the case of the Ni-Fe samples (samples no. 2, 6-9), the Nanovate™ A24 concentration was held constant at 2 g/L, however, as the FeSO4·5H2O concentration was increased there was a corresponding grain refinement effect. This result is consistent with those of Cheung et al. (1995) and Li and Ebrahimi (2003) who also found that with an effectively increasing Fe/Ni ion ratio in the electrolyte, there is a decrease in grain size along with an increase in the Fe concentration in the deposit. Fig. 4.12 describes this effect by plotting the grain size vs. the Fe concentration in the deposit. With small additions of Fe (15wt.%) in Ni, the grain size is dramatically decreased. Additions greater than 15wt.% have a lesser effect on decreasing the grain size.
4.4.2. High-Resolution* A defects analysis was carried out on the series of HR-TEM images. Samples no. 5, 7, and 9 were used for this purpose and examples of HR-TEM images are shown in Fig. 4.13, 4.14, and 4.15, respectively. In general, the HR-TEM images for all three materials revealed small grains (100nm) with lattice resolution. The grain sizes obtained using conventional TEM image analysis corresponded well with observations using HR-TEM. There are no observable signs of coarse porosity which is consistent with previous studies on materials synthesized using the electrodeposition technique [e.g., Van Petegem et al. (2003), Haasz et al. (1995), Zhou et al. (2009)]. Further analysis would be required to confirm the presence of low porosity using a more appropriate technique such as positron annihilation
The key findings presented in this section were previously published in the following refereed journal article:
J.D. Giallonardo, G. Avramovic-Cingara, G. Palumbo and U. Erb, “Microstrain and growth fault structures in electrodeposited nanocrystalline Ni and Ni-Fe alloys”, Journal of Materials Science, 48 (2013) 6689.
Chapter 4 – Materials Synthesis and Characterization
spectroscopy [e.g., Van Petegem et al. (2003), Zhou et al. (2009)]. However, it should be noted that electrodeposited nanocrystalline metals produced in a similar manner as the samples in the current study are typically about 99-100% dense [Van Petegem et al. (2003)], and may possess excess free volumes at the grain boundaries with sizes smaller than a vacancy in a perfect lattice [Zhou et al. (2009)].
There was also no evidence of co-deposited second phases in these materials. This is consistent with the SAD patterns produced using conventional TEM (e.g., Fig. 4.8c and 4.9
c) which show no signs of the presence of second phases. The absence of second phases is also demonstrated in the XRD patterns that are presented later (e.g., Fig. 4.26 and Fig. 4.27).