Microscopy

The following is an excerpt from Dewar - Characterization and Evaluation of Aged 20Cr32Ni1Nb Stainless Steels, and references a stainless steel used in this thesis. The content of this article can also be applied to a general case, and should not be limited to to the presented example.

1. Scanning Electron Microscopy (SEM/EDS)

SEM microscopy offers higher spatial resolution over optical microscopy (~ 1μm spatial resolution), resolving topographical features of the sample, and providing atomic number differences over the sample. An image of the sample is produced by rastering a concentrated electron beam over the surface of the sample, where certain types of electrons will be emitted from the microstructure and collected by the appropriate detector. The accelerating potential of these emitted electrons ranges from 15-30 kV, but will generally be 20kV for the images presented in this study unless otherwise specified. Topological features of a sample can be viewed with the photoemission mode of the SEM, where Secondary Electrons (SEs) will be emitted from inelastic collisions with electrons in the k-orbital of the specimen’s atoms. Backscattered Electrons (BSEs) are elastically scattered where the scattering effect is dependent on the atomic number, Z, of the atom. The higher the atomic number the more backscattered electrons that will be elastically scattered and collected by the detector, showing a lighter contrast compared to atoms with a lower atomic number. BSE mode is useful for viewing different phases and precipitates in the microstructure, and does not require the sample to be etched.



Fig. 1: Various modes of electron emission from incident x-rays or electrons. Secondary electrons are inelastically scattered from the emission of valence electrons, backscattered electrons are elastically scattered where larger atoms deflect more electrons increasing collection rates. Auger and x-ray effects are caused by the ejection of inner shell electron compensated for by an electron in a higher orbital dropping down and filling the vacant energy level where then excess energy will either be given off in the form of a characteristic x-ray or the ejection of an outer orbital electron.

The interaction volume for the various scattering types, and imaging modes is illustrated in Figure 2 where BSE electrons are seen to have a much larger interaction volume than SE electrons. The larger excitation volume results in BSE mode having a lower resolution than SE mode (BSE resolution ~ 1μm, SE resolution ~ 10nm).



Fig. 2: Interaction volumes is the depth range probed by the various types of scattered electrons and x-rays, where imaging resolution is dependent on the volume generated by the emission type[5].


Quantitative analysis of individual points, and constituents on the sample can be achieved with EDS, which collects characteristic x-rays emitted from excited atoms(Figure 1). Each element produces an x-ray with a unique amount of energy that is collected by the EDS detector which causes a current to flow proportional to the energy of the x-ray. These current pulses are converted to voltage pulses, where once a large enough sample of x-rays are collected the intensities of voltage pulses are reported for further analysis in determining elemental concentrations at the point of interest. EDS is efficient at detecting elements with an atomic number, Z, greater than 9, however a windowless detector can detect elements greater than Z = 5. Due to instrument limitation for detecting carbon and nitrogen (due to the X-ray energy peak overlap, absorption, and attenuation by the detector window material), as well as surface contamination from carbon, carbon and nitrogen are generally omitted from EDS analysis. Depth resolution for EDS is 2-5μm, depending on the atomic weight of the sample and its microconstituent. Spatial resolution for EDS in SEM is ~ 2μm, where the analyzed particles should be greater than 2μm in order to obtain proper quantification without overlap from another constituent in its interaction volume.

Element maps can also be generated using EDS for SEM, where x-ray intensities are measured while the beam is rastered over a specified area of the sample. The resulting image is a brightness intensity map showing points of high concentration, and low concentration of a specific element. Element maps can be used to qualitatively determine which phases contain which elements, or quantitatively produce phase fractions if the individual phases have been properly characterized. EDS has a drastically lower collection time than WDS, but the results are less accurate and the collection efficiency of EDS is much lower.

For the in-service 2032Nb samples SEM was conducted on a Zeiss EVO MA 15 Scanning Electron Microscope (SEM) with a LaB6 crystal source. It is equipped with a Bruker Silicon Drift Detector for EDS with a peak resolution of 125 eV. SE detection has a 3nm resolution at 30kV, and BSE have a 4.5nm resolution at 30kV The samples were left uncoated, and connected to the conductive stage with copper tape. SEM for the commerical 2032Nb and Super-2032Nb samples was done with a Tescan Vega-3 Scanning Electron Microscope fitted with a Oxford Instruments Inca energy system EDXS detector. The Vega-3 is a semi-automated tabletop SEM with an accelerating voltage of 200 eV to 30 keV, capable of both BSE and SE detection. SE resolution on the Vega is 3nm at 30KV, while for BSE the resolution is 3.5nm at 30kV. Micrographs produced for this study were obtained at a 20keV operating voltage at a 11mm working distance for BSE, and 13mm working distance for EDX.

2.Auger Electron spectroscopy (AES)

AES has many advantages over SEM, and EDS as it can achieve higher depth, and spatial resolutions (< 100Å and 300nm respectively), where micron and sub-micron sized features can be inspected. AES is also sensitive to lighter elements than SEM, and EDS where atomic numbers of Z < 15 up to helium can be inspected. AES does have some disadvantages to SEM as characterization of the bulk of the sample cannot be achieved since AES is a surface sensitive technique (Figure 2). Mapping procedures should typically be done with EDS or WDS for phase fraction calculations since a large enough area cannot be covered to give an accurate representation of the sample. Collection times for Auger electrons are also much longer than for EDS, where a dozen point could take over an hour to process.

The Auger effect shown in Figure 1 occurs when the sample is excited with an x-ray or an electron, where the energy from the incident electron knocks out an electron from the atoms k-shell. Higher energy electrons will drop down to maintain a minimum energy level, where the excess energy is then either released as a characteristic x-ray or an outer shell electron is ejected. An Auger electron is considered to be the ejected outer shell electron, whereas the characteristic x-ray is associated with x-ray fluorescence, and is used in XRD and EDS. From Figure 2 Auger electrons are only emitted from the first few nanometers of the sample making it a surface sensitive technique. Similar to EDS, AES can display the electron energy spectra for quantitative analysis. Since the interaction volume is much smaller, a more accurate composition of a specific feature can be resolved, as other phases or precipitates will not overlap within the excitation volume.

AES was performed on a JAMP-9500F Scanning Auger Microscope with 3nm SE resolution, and an 8nm probe diameter for Auger analysis. For the samples in this study the accelerating potential was set to 15keV, and the surface of the sample was ion sputtered with argon for 20 seconds to remove any surface contaminants that might have formed on the sample from the open atmosphere. However, since carbon is a significant constituent in the primary and secondary carbide phases it is necessary to calibrate its atomic sensitivity factor (ASF) to reduce the error from hydrocarbon surface contamination. The ASF for carbon was calibrated through the quantification of M23C6 carbides, since it has been shown that carbon is the primary element to constitute the second sublattice of M23C6, and any nitrogen soluble in this phase is dilute [6]. The ASF value was adjusted until the carbon concentration in the quantification of M23C6 was 20.7 at.pct through Eq. 1, where Ccarbon is the concentration of carbon, and Imeasured is the measured intensity of the carbon peak. The ASF value calibrated for carbon was 0.326.

Ccarbon = Imeasured
          ASF
(1)

3.Electron Probe Microanalyzer (EPMA/WDS)

EPMA is a similar spectroscopy technique to SEM, however it works on the principle of WDS by using several crystal spectrometers to collect characteristic x-rays from individual elements. Various diffracting crystals with a known inter-atomic plane spacing, d, are calibrated to diffract characteristic x-rays emitted from the sample to the x-ray detectors in the spectrometers. The characteristic x-rays diffracted by the crystals are representative of predetermined elements, or components of the sample. For the 2032Nb system the EPMA instrument must be calibrated for chromium, nickel, niobium, silicon, and carbon elements. For chromium, and nickel a LLIF crystal was used, a PET crystal was used to collect characteristic niobium x-rays, a LTAP crystal for silicon x-rays, and a PC1 crystal for carbon x-rays.

WDS works on the principle of x-ray florescence where characteristic x-rays from the sample are emitted via the process shown in Figure 1, and scattered in all directions. A small percentage of these characteristic x-rays reach the crystals in each of the spectrometers situated around the sample. These x-rays are incident upon the crystal, and any x-rays that obey Bragg’s law will be diffracted out of the crystal towards the spectrometers detector where an argon gas will be ionized and a voltage is recorded. Bragg’s law is defined as,

nλ = 2dsinθ
(2)

where λ is the wavelength of the incident x-ray, θ is the angle between the x-ray and the atomic planes, and n is an integer. Due to the x-ray separation process WDS can resolve a better signal to noise ratio, and produce higher quality x-ray maps. WDS resolution is better than EDS, however collection times are longer due to lower collection efficiency. WDS can also separate x-ray peaks that are unresolved with EDS methods. WDS can detect elements as low as Z = 5, however precise quantitative chemical analysis is difficult for atomic numbers less than 8 (Oxygen). This implies that although WDS is more accurate than EDS for detecting carbon and nitrogen concentrations, there will still be a significant error associated with it. WDS analysis requires that the user knows which components are present in the sample, whereas EDS does not require prior knowledge of which elements are of interest.

For the purposes of this study EPMA was chosen primarily for mapping individual elements for qualitative segregation analysis, and quantitative phase fraction analysis. Mapping with WDS will give much better results than mapping with EDS due its higher peak-to-background ratio. While EPMA can map a large enough area to give an accurate representation of the samples microstructure, its spatial resolution is limited to ~ 1μm.

EPMA analysis was conducted on a CAMECA SX100 electron probe microanalyzer (EPMA) operating at 20kV. The resolution of the elemental maps was 1μm∕pixel, at a dwell time of 20ms. Multiple samples were analyzed through various cross-sections for a more accurate representation of the bulk material. The area fractions were calculated from a total cross-sectional area of hundreds of mm2 to get a number representative of the entire sample.

4. X-Ray Diffraction (XRD)

EDS and WDS methods are important for characterizing a material, and provide a large insight into the potential microstructural constituents (i.e. phases) of the sample. However, compositional analysis of a phase provides no insight into its crystal structure, which is necessary to define a phase with significant degree of certainty. In XRD a monochromatic beam of x-rays with a fixed wavelength are incident upon a samples surface, or powder. If Bragg’s law is satisfied a certain portion of those incident x-rays will be diffracted to a detector which will record their intensities as a function of diffraction angle, 2θ. The sample is rotated 180 where a plot of intensity vs 2θ is output as shown in Figure 3. The resulting I vs 2θ output is compared to a database of powder diffraction files which contain the experimentally determined relative intensity values, and diffraction angles for various phases and all of their diffracting planes. The XRD profiles for the powder diffraction files can determine the types of atoms, and the position of the atoms by the relative intensities of diffracted x-rays, and the spacing between inter-atomic diffracting planes by the diffraction angle. Unit cell spacing, and the space group for the compound can be determined from this information. If the powder diffraction file fits the XRD profile then their is a significant probability that that compound is present in the sample. The amount of each compound in the sample will be represented by the absolute intensities shown in the XRD profile. A search/match routine can be utilized in XRD data processing software (i.e. JADE), but it is useful to know the potential compounds present in the system beforehand to filter out any false positives.



Fig. 3: X-ray diffraction spectra of precipitates extracted from the austenite matrix of the static cast 2032Nb Stainless Steel alloy.

Since in the 2032Nb alloy the microconstituents only consist of ~ 5% of the alloy, with some precipitates having only a 1wt% concentration, the constituents will be vastly overshadowed by the matrix phase if the sample is used as polished. The interdendritic, and intradendritic precipitates will have to be extracted from the matrix by dissolving the matrix electrolytically in acid leaving behind the alloys precipitates. For a 2032Nb alloy a solution of 5% HCl, and 5 × 10-3g∕cm3 of citric acid in methanol was used to electrolytically etch the sample at 1.5V for 12hrs, as outlined by Piekarski [7]. The cathode used for the phase extraction was a Cr-Ni wire mesh made from Chromel alloy found in K-type thermocouples. The residue left behind in the electrolyte was filtered out and dried at 70℃ . Powder diffraction was then carried out with a Rigaku Geigerflex Powder Diffractometer using Co-Kα radiation.


5. References

[1]    Jasiński, W.. Solutionizing of IN 519 superalloy after long-term exposure. Adv Manuf Sci Technol 2008;32(1):69–80.

[2]    Balk, T., Nychka, J.. Microscopy lessons. Tech. Rep.; University of Kentucky; 2006.

[3]    Shinozaki, K., Kuroki, H., Nakao, Y., Nishimoto, K., Inui, M., Takahashi, M.. Deterioration of weldability of long term aged HP heat-resistant cast steel containing Nb, Mo and W. Quarterly Journal of the Japan Welding Society 1998;16(2):223–232.

[4]    Berghof-Hasselbcher, E., Gawenda, P., Schorr, M., Schtze, M., Hoffman, J.. Atlas of Microstructures. Materials Technology Institute; 2008.

[5]    Werner, W.. Interaction of electron beams with matter. 2012.

[6]    Landon, P., Caligiuri, R., Duletsky, P.. The influence of the M23(C,N)6 compound on the mechanical properties of type 422 stainless steel. Metall Mater Trans A 1983;14(7):1395–1408.

[7]    Piekarski, B.. Effect of Nb and Ti additions on microstructure, and identification of precipitates in stabilized Ni-Cr cast austenitic steels. Mater Charact 2001;47(3-4):181–186.

[8]    Zhang, T., Wang, Q., Song, X., Fu, H.. Effect of electromagnetic centrifugal casting on solidification microstructure of cast high speed steel roll. Materialwissenschaft und Werkstofftechnik 2011;42(6):557–561.

[9]    Wolczynski, W., Krajewski, W., Ebner, R., Kloch, J.. The use of equilibrium phase diagram for the calculation of non-equilibrium precipitates in dendritic solidification. theory. Calphad 2001;25(3):401–408.

[10]    Shi, S., Lippold, J.. Microstructure evolution during service exposure of two cast, heat-resisting stainless steels – HP-Nb modified and 20-32Nb. Mater Charact 2008;59(8):1029–1040.

[11]    Gonzalez, R.C., Woods, R.E.. Digital Image Processing. Addison-Wesley Longman Publishing Co., Inc.; 1992.

[12]    Suzuki, S., be, K.. Topological structural analysis of digitized binary images by border following. Computer Vision, Graphics, and Image Processing 1985;30(1):32–46.


6. Glossary *

treatment
A specific level of a factor in a factorial design.. 3

7. Acronyms *

AES
Auger Electron Microscope. 4, 9, 10
BSE
Backscattered Electron. 7
EDS
Energy Dispersive X-ray Spectroscopy. 6, 8–10
EPMA
Electron Probe Microanalysis. 4, 10
SE
Secondary Electron. 6, 7
SEM
Scanning Electron Microscope. 4, 6, 9
WDS
Wavelength Dispersive Spectroscopy. 9, 10
XRD
X-ray Diffraction. 4, 10

 

© 2012 CCWJ. All Rights Reserved.