Depends what you are looking for! PAAS and NASC can give you some idea how different your coal samples are from an "average" shales (however, there is no average shale!) - typical sedimentary stuff. Do you need to compare coals from different localities, regions, environments or maturities, it all depends...

Dear Arash V. Moradi

Resulted from chondrite normalized could help you to identify the source or mineral origin of the coal that possibly from part of carbonatite  and  granite. The bulky of granite which deposited in sample are showed some negative in Ce and mostly negative  Eu anomaly. However for further mineral investigation need other tests by using i.e XRF for finding the species of rock, mineral, particle and elements.

Best regard's

Zolhizir bin Daud

Considering that you are looking at sediments, PAAS is likely to be best. However, your remark about the Ce-anomaly is a bit mysterious - since PAAS itself does not have either a positive or a negative Ce-anomaly when normalised over chondrite, I cannot really see how your samples' Ce-anomaly changes from normalising over chondrite to normalising over PAAS.

But in general, Alvar is right in saying that you first have to decide what you want to do, and then choose the appropriate normalisation.

I'm not sure where the samples came from. Just assuming if the sample in tectonic environment as refer to Henderson (1984) better you using the chondrite, Anyway that true as Dr Marlina and Dr Alvar Soesoo saying depend on type of samples  and normalized requirement.

I think, you can find the answers of your questions from this paper; 

Rare earth element sources and modification in the Lower Kittanning coal bed, Pennsylvania: implications for the origin of coal mineral matter and rare earth element exposure in underground mines

Also, this paper can help you to interpret the behaviour of Ce and Eu elements under the different redox conditions;

RARE EARTH ELEMENT (REE) GEOCHEMISTRY AND GENESIS OF OIL SHALES AROUND DAĞHACILAR VILLAGE, GÖYNÜK-BOLU, TURKEY

Best wishes

I second others in that your choice of the reference values depends on what process you're trying to unravel. PAAS would be a better choice as it's more relevant to the types of processes that can affect the REE composition of your samples. I suggest using the most updated values for CI-chondrites and PAAS from the following:

Article A novel extraction chromatography and MC-ICP-MS technique fo...

Thanks, Ali, for the reference! I have to admit that I missed that paper.

Sure, I hope you find it useful :-)

I would agree that chondrite normalisation is not likely to be applicable. One of the shale normalised patterns would be more suitable and my preference would be for the NASC normalisation. However, as noted by other commentators it all depends upon what your aim is.

When looking into the normalization value from the view of points of PAAS, NASC and UCC, these value are quasi similar especially on LREE which from La to Eu i.e please refer to Hugh Rollinson (1993) in using geochemical data: evaluation, presentation and interpretation. But anyway based on your cases, if the samples was from your place it' s quite enough to finalized the normalizing value (using PAAS, NASC, UCC)?, but  the origin of the sample is far away from Australian plate or American shale. that why from my idea the fist try using the Chondrite is rationale compared to other. But it does not mean you ignoring the other normalized. The important is you need to deep interpretation by using Chondrite because this could settle down or much much analysis especially rock species where not only concentrate on shale, clay, aeolian or diamictite etc. but generously more to early of rock crystallization of carbonatite for origin and provenance of coal sample you are.

Regards

Zolhizir bin Daud

Thank you all for your comments. They are good and helpful.

the main objection of the study is to interpret paleoredox conditions (also using Ce and Eu anomalies) and also to determine the source rock(s) and that whether the coal seams have similar source rock or not. 

special thanks to Ali for his useful reference.

Dear Arash V. Moradi 

I suggest you to use the data by Prof. Yudovich. He is a recognized leader of coal geochemistry researching in Russia and this data on background contents are based on the studies which were carrying out for many years. 

Please find attached a link leading to the full-text. See the Table 3 in this paper.

With the best, warm regards from Russia,

Alexey

Article Estimations of Clarkes for Carbonaceous biolithes: World ave...

Thank you so much dearAlexey

Dear Mr. Moradi,

how did you determine your Ce and Eu anomalies by visual inspection only or by calculation? I use PAAS and chondrite and regardless of the approach taken the values are comparable as being calculated. On the other hand did you use the whole rock samples or did you make use of incineration under plasma? As the ash is the most crucial aspect in the concentration of these elements, I would suggest the latter step prior to analysis of the REE ( DILL, H.G. and PÖLLMANN, H. (2002) Chemical composition and mineral matter of paralic and limnic coal types of lignite through anthracite rank. Upper Carboniferous coal in comparison with Mesozoic and Cenozoic coals (Germany).- Canadian Society of Petroleum Geologists Memoir, 19: 851-867).

You might also increase the total amount of your REE. But be aware of the situation that the in-situ parameters essential for an environment analyses might be modified.

Ce- and Eu anomalies are strongly impacted by the rock-forming non-coalified matter and might give you answer on a  different question hitherto not raised by you.

Best regards

Harald G. Dill

It doesn't matter what you normalise to, as long as you are aware of what the differences between the different normalisation data sets do to your REY patterns (in particular with respect to the Eu anomaly - it does neither affect Ce nor Y anomaly). Normalisation is just a way to make the interpretation of REY data easier and help you illustrate your message. Make sure, however, you use a REY data set of good analytical quality that itself produces a smooth normalised REY pattern. If the normalised REY pattern produced from your data is not smooth (except for the anomalies that make "chemical" sense), the analytical quality of your data is highly questionable and you must not use such data!

(Note that Pourmand et al's "new" PAAS is based on different samples than the "old" PAAS, and hence their REY concentrations - though not the shape of the REY pattern, of course - are rather different.)

I recommend Michael's answer

Best Regards