Hi Brenton,

I work with Sedimentary rocks for provenance analysis.  I usually collect about 5 kilos (a big block) of sample in the field and break down about half of that for analysis.  In sedimentary rocks it depends on the sample, some have more Zircons than others and also diagenesis plays a part in degrading the sample so you need plenty to account for that, Granite should have a higher ratio of Zircons.

I use heavy mineral separation techniques using heavy liquids and magnetic separations to obtain the zircon fraction.  From a kilo of sample you usually end up with about 0.1- 0.5g of zircon fraction sample which usually contains about 300-500 zircons.

For analysis you generally need a minimum of 120 analysable zircons (See Vermeesch 2004), having more allows for any errors -I've lost sample when polishing the slides for U-Pb etc so having more helps.

I would suggest that you get a similar sample size for granites in the field in order to have sufficient sample for analysis and future reference (depending on how accessible your area is)

Have you also considered modern river samples for source characterisation?  The granite will only tell you the zircons formed in that specific intrusion, for general source characterisation a modern river sample would give a broader spectrum for the catchment area and would include the source bedrock.

I use the same sample for U-Pb and Hf which works very well.  Also Pieter Vermeesch has worked extensively with U-Pb and Hf and has published a good technique for analysis See:

Vermeesch, P., Sherlock, S.C., Roberts, N.M.W., Carter, A. (2012). A simple method for in-situ U-Th-He dating. Geochimica et Cosmochimica Acta, 79 140-147. doi:10.1016/j.gca.2011.11.042

He also published a study on the number of grains needed for a provenance study:

Vermeesch, P. (2004). How many grains are needed for a provenance study?. EARTH PLANET SC LETT, 224 (3-4), 441-451. doi:10.1016/j.cpsl.2004.05.037

I hope that helps

Tam

Thanks for your answer Tamsin.

I have a suite of U-Pb ages for a number of sedimentary samples. And I am aware of the separation and analysis process in terms of sedimentary rocks. The number of zircons available would tend to be higher in sediments because the sedimentary process allows for the concentration of heavy minerals. This is not the case for granites. So I would imagine that I would require a larger sample for granitic material (my question is: how large?). Granites inherit zircons. So if you were to perform core-rim analyses, you might find that you have zircon cores in the granite that are older than the cooling/crystallization age of the pluton.  This would provide clues about the terrane in which the pluton developed. The Vermeesch et al. (2012) paper is about U-Th-Helium dating, not hafnium.  Thanks again for your contribution. Much appreciated.

Brenton.

Apologies, I didn't look up your history and answered in a hurry,

I know Eric Cowgill has done analysis on Granites -  Cowgill et al (2003) Reconstruction of the Altyn Tagh fault based on U-Pb geochronology: Role of back thrusts, mantle sutures, and heterogeneous crustal strength in forming the Tibetan Plateau (2003), he is on Researchgate - not sure how to link him - He is at the University of California and is very approachable.

Also Alex Robinson - University of Houston has worked on this too (mostly gneisses but I believe he dated Granites too) - again very approachable.

They should be able to answer your question in a quick email.

Good luck!

Tam

No problem. Thanks for your input!

Brenton

Dear colleagues,

the technical part I can fully agree with but as far as the stream sediments are concerned I would like to make a warning message. It is correct that you get an insight  into the heavy mineral suite of a certain area. Placer deposits are "baskets" collecting all the "chunk" of denudation in time and space. It is necessary to know very well what the evolution of the drainage system looked like in time and space. The modern-day stream sediments cannot be used as a provenance indicator for the existing riverine system only and have to be tracked down into deeper parts of the geological history. I did not sample the area for age dating but for the placers in their own sake and for geomorphological and sedimentological studies of the Cenozoic landscape. I append two papers that may shed some light on the complexity of this issue.

It was only an annotation. Often highly-sophisticated methods such as age dating fail due to simple mistakes made during sampling and often we even do not become aware of this.

Best regards

H.G.Dill

Dear Brenton,

this is a very difficult question, because the abundance of zircon is highly variable in granites, which means there is no specific  "strating mass" of sample. Even if you have determined the Zr-content of your granite before heavy mineral separation, you can not say anything about the amount of zircon per gramm granite, etc.

Sometimes, you can have samples with very few zircon, I had some with only about 10 grains from 5kg granite and the other way round with 200 grains from a matchbox sized sample.

Furthermore the number of zircon populations depends on the type of granite. An I-type granite usually has less zircon populations than a S-type granite.

Normally, I treat granites like sediments, which means I pick about 120 to 200 zircon grains. During the first measurements of age determination (let's say 30 to 60 grains) you will see, if there is much difference in the signals. If yes, it would be good to continue, if not it could be enough data for the specific sample.

Anyway, during the picking process you can have a look on the shape of the zircon crystal faces (see for example Pupin, 1980) or the roundness of the grains. If they vary a lot, you can expect various populations. If they are all more or less the same, there is a quite high chance that you have only one or few populations.

Hope this helps a bit with this difficult question,

Andreas

Granite is a type of felsic magmatic rocks that are rich in zircon than mafic igneous rocks. Usually you may acquire 100-200 grains from just a hand specimen size of granite samples. Of course, the abundance of zircon is highly variable in granites, so that it is safe to sample 5-10 kg samples for zircon separation.

Thank you all for your contributions. Sounds like it will be a trial and error kind of job. Andreas - excellent answer - thanks very much - very useful indeed.

Apologies for the late answer.

Short answer: A litre-size sample (3 kg) should be enough in most cases.

As others pointed out, it depends on the granite. If it is likely to have sedimentary source (IE S-Type granite with white mica) then you may have quite a difficult time getting "all" populations.

Typically elongated zircons (if you have any) are less likely to have cores and you should easily be able to get the magmatic age with just 18 analyses. I picked around 30 grains from s-type granites and got relatively homogenous populations from all. If you want to date all the cores I would pick in excess of around 100-140 grains with rounded morphology, or better yet, make a mount where you just sprinkle the zircons indiscriminately on tape before mounting them in epoxy, this way you avoid the sampling bias of picking. Then do proper CL work to determine where to hit them with the laser.

How much samples i need to study chronology of a dyke rock?

First of all you need a fat dike: considering dike volume vs contact area, the thinner dike the more it is contaminated. If dike mafic (say, c.30 m thick) - seek for pegmatitic schlieren in the later phase may contain own zircon, otherwise, baddeleite can present.

For sample weight - chek, please, Article Effects of sampling and mineral separation on accuracy of de...