A widely accepeted way to estimate the dislocation density is X-Ray diffraction but which plane do I consider for estimating the disloaction density in the material. Similarly for TEM and EBSD method, which plan is to consider for estimate the dislocation density.
It depends how many dislocation you have. For semiconductors: Light microscope, for steel possibly XRD. Because of the small size in TEM the dislocation characterization is comparatively easy but the statistic is horrible. Therefore, precision good, accuracy bad. For XRD it should be vice versa. Accuracy acceptable, precision not very good. For peak analysis you need to make some assumptions...but if they are calibrated, the results are from my point of view the most trustful. Using SEM you need to "calibrate" your investigation technique since you cannot see them (except in semiconductors or at very rare cases). And you have to make sure that the dislocation you observe are not the result of your preparation (this is the most critical problem, also in TEM where everything can be relaxed and the visible dislocation density is only a part of the real one). Therefore, accuracy bad, precision bad...since you do not observe them directly.
However, everything is relative... You have to specify your problem more precisely.
Dear Akram, thanks for your opinion, but may I know why TEM is best among all these mention technique.
It depends how many dislocation you have. For semiconductors: Light microscope, for steel possibly XRD. Because of the small size in TEM the dislocation characterization is comparatively easy but the statistic is horrible. Therefore, precision good, accuracy bad. For XRD it should be vice versa. Accuracy acceptable, precision not very good. For peak analysis you need to make some assumptions...but if they are calibrated, the results are from my point of view the most trustful. Using SEM you need to "calibrate" your investigation technique since you cannot see them (except in semiconductors or at very rare cases). And you have to make sure that the dislocation you observe are not the result of your preparation (this is the most critical problem, also in TEM where everything can be relaxed and the visible dislocation density is only a part of the real one). Therefore, accuracy bad, precision bad...since you do not observe them directly.
However, everything is relative... You have to specify your problem more precisely.
By "Light microscope", if you mean EPD prior or some other method of accentuating the strain field to create the optical "contrast", then there are many related issues to contend with:
TEM & XRD are the only two method that may be reliably used to estimate dislocation density. However, there are distinct differences between the statistics and spatial resolution among these two methods. XRD has the distinct advantage for in situ measurements while TEM requires destructive specimen prep and may introduce additional defects. XRD is more "global" while TEM is more "local". In my opinion, a combination of XRD followed by TEM of preselected sampling areas to calibrate the XRD measurements would yield the best as well as expedient results.
By XRD, I do not mean the archaic 0D (point counter) methodology that has existed since the Braggs for over a century. Instead, I mean the more modern version using a real time 2D high resolution XRD imaging system.
The challenge with EBSD as I have read from Gert on prior occasions, is the relatively poor spatial resolution as well as angular resolution. Is that correct?
https://www.researchgate.net/post/What_is_the_likely_Nanostructural_origin_for_observations_of_the_LiF_113-XRD_reflection_rocking_curve_integrated_breadth_squared_b2_map?_tpcectx=profile_questions
Ravi, when I mentioned light microscopy then I had in mind semiconductor compounds like GaAs. There KOH melt is used to generate etch pitches which you can count very nicely using light microscope. For GaP you can even use infrared microscopy.
Regarding EBSD I have to confirm your assumptions. The precision is very bad, although we could improve it up to a factor of 10, cf. the link. On the other hand, the misorientation can only image the geometrically necessary dislocation which are estimated to contribute only by about 10% to the entire amount of dislocations. The major part is "hidden" in the band profiles which are not at all considered until know. Presently, it also looks like a too complicated problem because of the still rudimentary understanding of the patterns. On the other hand, I assume that not all dislocations equally contribute to observations so that the question, how many dislocations are in a material, might be to unspecific.
Presentation Significant Improvement of the orientation resolution by EBS...
The universal challenge with any etchant is its ability to reveal those defects terminating at the exposed surface preferentially. Defects below the surface may not manifest at all. XRD would sense the entire strain field in the sampling voxel. The trick is interpreting the Bragg profile and its deviation from the IDEAL Bragg profile. Tamas Ungar's lecture series (I-IV) at IIT Kanpur were very instructive for me in this regard.
https://www.researchgate.net/post/Is_it_possible_to_determine_the_type_morphology_of_defects_in_single_crystals_by_the_shape_of_the_Bragg_XRD_Rocking_Curve_Profile
https://www.youtube.com/watch?v=kGQb8EiSQAw
Also I think with xrd we measure total dislocation density but with ebsd we measure only gnd density, kindly clarify if I am wrong.
Yes Ravi, about this I was talking about. He is already working for decades on this subject. It is amazing what you can extract from interference profiles. In Germany Klimanek was and now Daniel Rafaja is also very active in this field.
https://www.researchgate.net/profile/D_Rafaja/publications?sorting=newest&page=15
The answer given by Prof. Gert is most accurate during experiments. Although I truly believe it is almost impossible to accurately quantity to good statistics using any method like XRD, TEM or EBSD given the issues associated with measurements. The possible way forwards for good repeatable levels of accuracy would be use a multiple combination of methods and check whether they give the right 'order' for the amount of dislocations present in the material.
Nice answer by PRof Gert. For calculating dislocation density in epitaxial thin films, one must do rocking curve scans for the planes that one is interested in. From these rocking, curves, we can get the FWHM. Then if we know what's the Burger's vector for the set of planes that are being rocked, we can estimate the dislocation density. Then one can compare the results with TEM. We just recently published an article about using XRD and TEM to study dislocation and stacking faults in non-polar nitride semiconductors.
"Integrated Breadth Squared", (Beta2, β2), will yield a far more reliable estimate of defect density than the FWHM. Defects tend to affect the Bragg profile disproportionately below the FWHM. The FWHM would be insensitive to perturbations at the lower shoulder and tail area of the Bragg profile. See this example for LiF mono crystal. A comparison of the FWHM and β2 is also shown. The image contrast in the β2 map is far superior! Computing β2 directly from the digital data would avoid the need to assume any particular shape for the Bragg profile. Otherwise, one needs to convert the FWHM to integrated breadth appropriately for computations.
BTW as mentioned earlier, β(hkl)2, would be proportional to the projection of the strain field due to defects on to the (hkl) plane.
https://www.researchgate.net/post/What_is_the_likely_Nanostructural_origin_for_observations_of_the_LiF_113-XRD_reflection_rocking_curve_integrated_breadth_squared_b2_map?_tpcectx=profile_questions
https://www.flickr.com/photos/85210325@N04/34003566875/
https://www.flickr.com/photos/85210325@N04/36273871102/
I can recommend the XRD and the TEM ... on the same spot. Both has its weekness but they are not the same so you can get good informations after u compare the results.
However, all XRD, TEM and EBSD have its significance. But among them from my point of view XRD is the more accurate method to estimate the dislocation density. In XRD you can choose highly intense plane to calculate the dislocation density or more precisely you can take average of all the planes.
In my opinion, precision should be no challenge with XRD. Meaning, one can get the same answer (relative values) reproducibly. However, accuracy depends on how well one can calibrate XRD using known standards and some sort of TEM verification.
The in situ nature of XRD examinations is a key advantage!
- Badges
- Science method