I have a suite of rocks that have nearly identical major element chemistry, but a single sample has quite different trace element concentrations. All the rocks are trachydacites. I know the analysis was done corrected. What process could cause this?enter image description here

  • $\begingroup$ What mass were the individual aliquots and samples? How were they prepared? $\endgroup$ Commented May 20, 2014 at 9:57
  • $\begingroup$ Unknown, the data was presented with no knowledge of preparation other than the values are correct $\endgroup$ Commented May 20, 2014 at 10:13
  • $\begingroup$ This is preliminary work making up part of potential future reseach $\endgroup$ Commented May 20, 2014 at 11:07
  • $\begingroup$ How do you know the sampling, sample management & analysis was done correctly if you have "no knowledge of preparation other than the values are correct"? To me, that does not add up. $\endgroup$
    – blunders
    Commented May 20, 2014 at 13:15

1 Answer 1


Under the assumption that we don't know anything about sampling, preparation and analysis we have to proceed with caution. Any of these steps can introduce effects that can produce stronger signals than the natural geochemistry of a rock sample.

Just to give you an idea of the errors:

  • Sampling a heterogeneous rock can produce very different whole rock chemistry. If you take your sample from a mica-rich layer you will get different results than from a feldspar-rich layer.
  • During sample preparation you can inadvertently through grinding, sieving and pouring produce aliquots which don't have the same mineral composition. Micas pour different from one vessel into another than quartz-grains. A certain mineral you want to analyze might have a smaller grain size and sit at the bottom of your vessel or in a different sieve your not analyzing. Each step in the lab can result in chemical fractioning!
  • Analytics is a whole science on its own. Without knowing how the mineral sample was dissolved (did everything dissolve?) and values for detection limit and standard deviation you will have uncertainties in your data that might be make these values meaningless.

Now assuming that the data is somewhat trustworthy:

  • The left plot shows you the chondrite normalized rare earth element pattern. The light rare earths are 100 to 1000 times enriched relative to the chondrite (should actually be cited which one). Heavy earth elements are also enriched but less. The sample CCS01 has lower enrichment and a positive Eu-anomaly. This could indicate that the sample has a different mineral composition than the other samples. Maybe a rare-earth-element-enriched mineral has a lower concentration in the CCS01-sample.

  • The right plot the CCS01-sample sticks out again. It has lower P-concentrations. Think again of how your sample might have fewer phosphates. The relative depletion in Ce and La might correlate with P. The sample might just have less monazite (But other minerals could cause this too). Lower Nb and Ta might indicate that the sample has fewer columbite and tantalite (but the Elements occur in other minerals too).


  • Instead of guessing you should find out everything you can about the samples history (sampling, preparation, analysis). If possible suggest more analytics. A day at the microprobe could show you which minerals are contained in your sample (rock slide or strewn grain concentrate). If not available a polarizing microscope could also help you figure out the differences in the sample (excluding very small grains).
  • $\begingroup$ Thank you very much for the response. I'll try do more analytics and head to the microscope room. $\endgroup$ Commented May 20, 2014 at 16:43

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.