I am polishing and etching pure cast copper and are trying to measure dendrite arm spacing. I cant see features that clearly indicate dendrite arms, although I am seeing a finer grain structure, see attached.
Are dendrites able to create a dense structure? And when casting, the structure is dense, except for the shrinkage shell. If you need dendrites, then it is necessary to analyze the area of the shrinkage shell. I think so.
Thanks for the reply. My understanding of the theory is that dendrites are the fundamental mechanism behind solidification and that in my case, the dendrites are there but that they are densely packed.
I have observed Copper dendrites readily in alloys containing high % of copper. I am thinking that the differences in melting point between the two elements in an alloy enable phase separation and for the dendrites to be seen after etching. So, for high purity metals with one melting point its dendrites can't be resolved?
It will be my plan next to characterise the observable macro grain structure in different regions of the cast rod.
Dendrites may be easily seen in cast alloys - this is obvious. They appear because of constitutional undercooling, which stops the grain growth in local principal direction, and provokes crystallisation in lateral direction. That's a simple explanation of how dendrites are formed. However, you're talking about pure metals. Pure metals are not alloys, and they do not exhibit constitutional undercooling due to homogeneous chemical composition (by definition of a pure metal). I believe you will not see dendrites in pure copper, provided "pure" is really pure. The microstructure you presented doesn't feature dendrite arms, because they can not be formed due to a good chemical homogeneity of your casting.
And just a remark: in the middle of the casting one should expect equiaxed grains caused by isotropic heat transfer. Closer to the mould, one can see columnar grains even in pure metals, however, they should not be confused with dendrites. Columnar grains are formed by unidirectional heat flow towards the cool mould wall, and this is a universal phenomenon, regardless of local segregation of alloy's constituents, even if we deal with an alloy, not a pure metal. Even in a cast alloy, one will hardly see dendrites in the columnar grains zone, because the temperature gradient will override the constitutional undercooling and lateral arms will not form.
Thanks for your answer Jarosław Ferenc it is very informative.
What due you think are the 'spot' features observed within the large grain in the bottom right of my attached image. A sub-grain structure? Pitting due to acid etch? A copper oxide phase?
Thanks
Given the lack of information which etchant you used, of information on the image magnification and my little experience with copper alloys, I'm far from giving immediate answer. The options you've proposed are possibe, I would bet on the artefacts from etching and inadequate rinsing, or contamination (such as oxides). I'm pretty sure that the sample was a bit too etched, see greenish freckles on the bright grains aligned horizontally, particularly on the small, 2nd from left grain there is a "twin" spot. Also, the grain boundaries have greenish colour, which is the effect of slightly excessive etching (GBs are prone to faster etching, this is perfectly normal).
If you wish to know what are these features, re-polish the sample (don't etch), put into a SEM and map the elements by EDS, then etch the sample and repeat the SEM observations with EDS mapping. Such tedious procedure will allow recognition of elements' distribution in copper, if this copper is not perfectly pure; and you will see what are the features observed in light microscope after etching. If you fear that the amount of impurites in your sample is high, consider XRD of the polished sample and repeat it on the etched sample. However, I think the latter would be the last thing to do: you will look for small traces just popping out from the background, so long time per step is necessary to distinguish weak peaks from bkgd noise, and XRD will hardly detect phases which are less than 1%.
Once again thanks for your response and some useful advise.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Materials Science