The following is a discussion of fluid inclusion dating that was held during early April 1997


 

****************************************************************
Hi,

Several of us are looking at techniques for discriminating between remobilized and syn-tectonically emplaced ore deposits. I seem to recall seeing a paper on direct dating of fluid inclusions. Can anyone recall this and provide the reference?

Many thanks,

Brian.

Professor Brian Marshall, Tel: +61-2-9514 1775
Department of Applied Geology, Fax: +61-2-9514 1755
University of Technology - Sydney, Email: Brian.Marshall@uts.edu.au
P O Box 123 Broadway,
NSW 2007,
AUSTRALIA.
****************************************************************
Perhaps:

Pettke, T.; Diamond, L.W. (1995): Rb-Sr isotopic analysis of fluid inclusions in quartz: Evaluation of bulk extraction procedures and geochronometer systematics using synthetic fluid inclusions. Geoch. Cosm. Acta 59, 4009-4027.

Wayne, D.M.; Miller, M.F.; Scrivener, R.C.; Banks, D.A. (1996): U-Pb and Rb-Sr isotopic systematics of fluids associated with mineralization of the Dartmoor granite, southwest England. Geoch. Cosm. Acta 60, 653-666.

Especially in the latter you'll find some very useful refs. concerning isotope analysis on FI's.

Best regards, Martijn Moree

mailto: morm@geo.vu.nl
Martijn Moree Martijn Moree
Vakgroep P.I. Department P.I.
Fac. Aardwetenschappen Fac. of Earth Sciences
Vrije Universiteit Free University
De Boelelaan 1085 De Boelelaan 1085
1081 HV Amsterdam 1081 HV Amsterdam
tel: ()31-204447282 tel: ()31-204447282
fax: ()31-206462457 fax: ()31-206462457
****************************************************************
>Hi, >

>Several of us are looking at techniques for discriminating between
>remobilized and syn-tectonically emplaced ore deposits. I seem to
>recall seeing a paper on direct dating of fluid inclusions. Can
>anyone recall this and provide the reference?

Here are a couple on fluid inclusion dating using the 40Ar-39Ar technique:

Turner G. and Bannon M. (1992) Argon isotope geochemistry of inclusion fluids from granite-associated mineral veins in southwest and northeast England. Geochim. Cosmochim. Acta 56, 227-243.

Qui H-N (1996) 40Ar-39Ar dating of the quartz samples from two mineral deposits in western Yunnan (SW China) by crushing in vacuum. Chem. Geol. 127, 211-222.

___________________
Ray Burgess
Dept of Earth Sciences
University of Manchester
Manchester, M13 9PL, UK
Tel +44 (0)161 275 3958
Fax +44 (0)161 275 3941
mailto:rburges@mh1.mcc.ac.uk
****************************************************************
Brian,

Getting good geochronological data from quartz is no mean task -- It is a black art. The technical details published in the literature make it seem straightforward.

I've outlined some of the potential pitfall below:

The first problem is good sample selection in the field. In our experience at BGS, "magmatic" quartz with high salinity NaCl-H2O inclusions has the best chance for success, while quartz associated with basinal brines the lowest (NaCl-CaCl2-H2O inclusions - High Sr low Rb, not enough spread in Rb/Sr). "Metamorphic" quartz with low salinity NaCl-H2O (with volatiles) is intermediate. Large quartz structures (>1m*50m) also seem to give better results than "vein swarms" intermingled with wallrock probably as a result of less mixing between fluid and wallrock Rb and Sr.

Before, you embark on collecting samples you should be aware of the complex nature of the general analytical procedure. First, you need to "hand pick" about 1g of acid cleaned quartz FREE of all mineral impurities for about 10 samples (5 to 10 days work). The separates are then electrolytically cleaned for about three weeks in millpore water with regular changes to the water -- To be sure that you've removed all adsorbed Rb and Sr aliquots of the "electrolyte" should be analysed by ICP-MS or other means. Only after this is the sample ready for analysis by isotope dilution, and here arises another problem, how much spike do you add when you don't know the approximate concentrations in the fluid inclusions/quartz? -- You could get this from crush-leach ICP-MS, but this requires more sample for analysis, or you could try trial and error on your precious samples. You then have to decide -- do you want a fluid inclusion isochron or a "quartz whole rock isochron". That is a crush-leach extraction or a total dissolution. With a crush-leach extraction you need to dope the leachant with Lanthanum Chloride -- What does this do to your blanks? If you decide to go for a total dissolution are your samples totally free of mineral inclusions (e.g. 15*m micas)?

Also, there is the mass-spectrometry of extremely small samples to deal with. 8 to 10 weeks work and the mass spec "vapourizes" 5 of your precious samples before you get any data!! Finally, do you want to work in a lab where people are regularly analysing samples with 100's of ppm Sr & Rb compared to your 1-10ppm (by weight in quartz).

Forewarned is forarmed

Best of luck

Jon

-----------------------------------------------------------------
Jon Naden
Fluid Inclusion Researcher, Geochemist and Ore Petrologist
British Geological Survey, Keyworth
Notts NG12 5GG
Tel: +44 (0)115 936 3163
Fax: +44 (0)115 936 3302
E-mail j.naden@bgs.ac.uk
****************************************************************
At 05:19 8-4-97 -0500, Jon Naden wrote:

> With a crush-leach extraction you need to dope the
>leachant with Lanthanum Chloride -- What does this do to your blanks?

You can solve that by cleaning the La-solution before you use it.

> If you decide to go for a total dissolution are your samples totally
>free of mineral inclusions (e.g. 15*m micas)?

This depends on what you want to know. Do you want to know the composition of the vein, or do you want to know the composition of the quartz in the vein? (Better: the fluids inside the quartz, assuming the quartz doesn't contain Rb-Sr-Pb-U-Th, etc.). Assuming you want to get an isochron, you better leach out the carbonate (if present), crush the sample to get the fluids, and dissolve the quartz. A three point isochron is better than a two-point isochron. Whatever mica is present in the quartz is irrelevant, then.

>Also, there is the mass-spectrometry of extremely small samples to
>deal with. 8 to 10 weeks work and the mass spec "vapourizes" 5 of
>your precious samples before you get any data!!

That's a risk you got to take. :-) But a state-of-the-art TIMS is capable of analyzing Rb and Sr isotopes to very low contents. For Pb isotopes you really need very saline samples.

>Finally, do you want
>to work in a lab where people are regularly analysing samples with
>100's of ppm Sr & Rb compared to your 1-10ppm (by weight in quartz).

That is a big problem. But you can avoid that by allocating a place in the lab that is dedicated to low blank stuff. Keeping it that way is just a matter of discipline.

I must say that I am not disagreeing with Jon's concerns. He's just a bit too glum, IMHO. It *is* possible to analyze isotopes from fluid inclusions. It is just a lot of work. Most of this work is to keep the samples free from contamination.

Martijn Moree

****************************************************************
Dear Brian

It is good to see you getting such an enthusiastic response from the fluid inclusion daters, but I think you need to bear in mind the nature of the hypotheses that you are trying to test. The successes of fluid inclusion dating are mostly from magmatic hydrothermal systems (and even so, may sometimes be solid inclusion dates). The problem with looking at deposits which formed or were remobilized during metamorphism is that fluids with long crustal residence times will have undergone extensive exhange with the immediate wall rocks. For example, Banks et al. (GCA v. 55, p.1021-1030, 1991) found a close correlation between cation and radiogenic isotope composition of fluid inclusions and wall rock chemistry (fluids a from thrust zone), even though halogen contents indicated a common fluid source irrespective of wall rock. It is impossible to use Rb-Sr in such settings, because differences in Rb/Sr ratio will reflect different wall rocks, and hence the samples will have had different initial Sr ratios. Likewise, Ar methods will be extremely susceptible to excess Ar. So, if the ores simply remobilized during metamorphism, it is difficult to see why any age would necessarily be that of their original formation. On the other hand, if the ores were formed only during the metamorphic event, it seems highly unlikely that isotopic analysis of the fluid inclusions would actually yield a date of formation, although a mixing line to confuse you is a distinct possibility.

I'm afraid that I would earnestly advise you not to attempt to date the fluid inclusions in your samples, because the work involved will be so great that you will feel obliged to make something of the results, even though they will be meaningless!

Please feel free to regard these comments as somewhat provocative - I'd like to know if the daters can rebut them for this type of problem!

Bruce Yardley
mailto: B.Yardley@earth.leeds.ac.uk
****************************************************************
Hi,

I didn't want to go into the pitfalls of interpreting the data. Bruce is absolutely right about mixing lines versus isochrons and the potential for "wallrock contamination".

Tomas Pettke and Larryn diamond have produced an excellent paper concerning mixing versus isochrons (1/Sr plots etc.), recently in Economic Geology (v 91 p951-956). Although this paper is for sphalerite dating it is also applicable to quartz. The same criteria that Rb-Sr geochronlogists use for assessing mixing in their data are applicable to "fluid inclusion" Rb-Sr

With regard to wallrock contamination. This can be overcome by careful sampling. If you choose large quartz structures (1x50m) fluid inclusions trapped in this quartz towards the centre of these structures will have had the least interaction with the wallrocks. A sample suite should contain material from along the quartz structure, from the hanging and footwalls, and wallrock. A comprehensive sample suite will allow you asses the different roles of mixing and wallrock contamination. Anyone involved with Rb-Sr isotope systematics in rocks will tell you that correct sampling procedure is the key.

There are three types of "fluid inclusion" geochronlogy.

1. Data from fluids alone (crush-leach)
2. Data from fluids and co-existing mineral (crush-leach and total dissolution)
3. Data from total dissolution

The last of these is not strictly "fluid inclusion" geochronolgy and should be though more along the lines of "quartz-whole-rock" geochronology. This is the technique that we've use at BGS.

1 and 3 are the only methods not susceptible to the argument that you only have a two point isochron. 2 is analogous to a single mineral-rock isochron and "carries only apparent age significance a priori" (Pettke and Diamond Econ Geol) in the first instance -- steps have to be taken to prove your data make up a multiphase isochron.

There is no doublt in my mind that is possible to obtain valid "fluid inclusion" isochrons, and in a number of circumstances this is the only method for obtaining absolute ages. However, Bruce's point about the amount of effort involved and the potential rewards should always be borne in mind before embarking on a programme of research.

I know of several "fluid inclusion" isochrons in the literature that provide "golden spikes" for the metallogenetic evolution of an ore field. Carrock Fell Tunsten deposit, The great sulphur vein in the North Pennines, and a variety of data from Cornwall. I would remind Bruce, that if we ignored all difficult science there would be little progress.

Outwith " fluid inclusion" geochronlogy U-Pb zircon dating is making movements into the field of ore and structural geology -- this could provide another option to the originator of this thread.

Cheers

Jon
****************************************************************
Hi,

It's good to know that many people is working on fluid inclusion dating and, also, I found that the references and advices you're giving are a "heaven's gift" to me because I beared in mind to try to date some fluid inclusions too. I think my problem is a little bit different: I need to date some pre-cambrian quartz veins enclosed in granitoids and metasediments from the north African Shield in Morocco. In this case, all primary fluid inclusions contain euhedral trapped muscovite crystals that seem to me to precipitate from the same brine. Brines are hypersaline NaCl-dominated (no crushing analyses yet!!). I was thinking to use either Rb-Sr after total quartz dissolution or Ar-Ar after irradiating the samples. The Rb/Sr problem that Bruce stated I don't think will be applicable here because I wont (try to) date the fluid but the hydrothermal muscovite and the contribution of Rb and Sr from the fluid, I think, would be negligeable. Any more advices?

And the other problem is that we cannot run this kind of samples in Barcelona yet so, any lab suggestions will be welcome!!!

And... thanks for the information you already sent and for any forthcoming comment.

Cheers,

Jordi Tritlla

Jordi Tritlla i Cambra
Dep. C.M.D.M.
Facultat de Geologia
Universitat de Barcelona
c/Mart=ED i Franqu=E8s s/n
08028 Barcelona (Spain)
Tl: 34 3 4021357
Fax: 34 3 4021340
mailto: jordit@natura.geo.ub.es
****************************************************************
At 14:47 8-4-97 -0500, Jordi Tritlla wrote:

>In this case, all
>primary fluid inclusions contain euhedral trapped muscovite crystals that
>seem to me to precipitate from the same brine. Brines are hypersaline
>NaCl-dominated (no crushing analyses yet!!). I was thinking to use either
>Rb-Sr after total quartz dissolution or Ar-Ar after irradiating the
>samples. The Rb/Sr problem that Bruce stated I don't think will be
>applicable here because I wont (try to) date the fluid but the hydrothermal
>muscovite and the contribution of Rb and Sr from the fluid, I think, would
>be negligeable. Any more advices?

Hmm. I don't think you can neglect the fluid Rb and Sr contents. I don't know by head what the muscovite-hydrous fluid partition coefficients are, but it won't surprise me if they are very funny.

Another thing: If you are dating daughter muscovite with Rb/Sr, you didn't obtain the trapping time of the fluid inclusions, not even the forming time of the muscovite. The age you get is the time the muscovite system got closed for Rb and Sr. So what does that really tell you? Moreover, you'll have a non-isochron age or a two-point isochron (if you date the fluid too). I don't think it's worth the effort.

In agreement with what Jon Naden said: when micas are involved, things become very tricky.

Martijn Moree

****************************************************************
Subject: Ar-Ar fluid inclusion dating

Jordi and others

If you want to date the muscovite inclusions to obtain a crystallisation age you will need to use a laser technique in order to discriminate between the different Ar components. If re-heating above muscovite closure temperature has occurred since crystallisation then some 40Ar will have diffused from the muscovite and be trapped in the inclusion. Therefore Ar should be released in a two-stage process:

(1) drill to the muscovite inclusion and release Ar from the fluid.
(2) melt the muscovite to obtain a cooling age.

Combining Ar released during steps (1) and (2) should give the crystallisation age. So you get two ages (crystallisation and cooling) from one analysis which has to be good value! Of course if some 40Ar was trapped with the NaCl fluid at the time of formation then this will complicate matters but can probably be resolved through 40Ar-K-Cl correlations (Cl is measured as a by-product of the Ar-Ar technique).

I agree with Bruce Yardley that excess Ar is a problem for direct dating of fluid inclusions but it can (and has) been done. Excess 40Ar shows a good correlation with Cl and therefore the 40Ar/Cl ratio can be used to correct 40Ar/K ratios. This was done by Turner and Bannon (1992) referred to in my previous e-mail.

_____________
Ray Burgess
Dept of Earth Sciences
University of Manchester
Manchester, M13 9PL, UK
Tel +44 (0)161 275 3958
Fax +44 (0)161 275 3941

****************************************************************
Dear all

I would like to contribute some comments and ideas to the recent (?, I know I am somewhat late) discussion on discriminating between remobilized and syn-tectonically emplaced ore deposits, and the apparently preferred approach of dating the hydrothermal activity (i.e. finding temporal evidence for discriminating between the above possibilities) using fluid inclusions (flincs).

First, getting precise data of a given bulk fluid inclusion sample - to me - IS straightforward, but analytically demanding and time-consuming.

However, the use of such data for dating purposes is a completely different story. Here are some reasons, why:

a) there are hardly any hydrothermal systems which do NOT evolve with time, i.e. particularly in initial isotopic composition of e.g. Sr or Pb (recall that for isochron dating, the initial isotopic composition MUST be constant). This is what has to be expected in rock-dominated environments (such as syntectonically emplaced ore bodies). Various studies in literature document this fact (also the scarcity of multi-mineral isochrons for hydrothermal environments!), and it has been nicely outlined by Bruce Yardley.

b) using bulk flincs alone implies that one tries to go for a single-phase isochron approach, but there is a priori no reason why a single phase should have variable parent/daughter ratios at all. Therefore, if anything, using flincs as an additional phase (like conventional multi-mineral approaches) appears to me more promising (as implicit in Martijn Morees suggestions). Consider that if there is a process which affects the parent/daughter ratio in a hydrothermal fluid (fluid mixing, variable fluid:rock interaction, ...) this will almost certainly also affect the initial isotopic composition of the daughter element in the fluid, and the fluid will be trapped at any time during quartz growth (primary vs. pseudosecondary flincs...).

I personally would seek for minerals suitable for conventional single-mineral approaches such as Ar-Ar on mica, K-feldspar (but NOT on quartz-hosted fluid inclusions because of the likelihood of trapped Ar with variable isotopic signatures (one might have a look at: Pettke, T., R. Frei, J. D. Kramers, and I. M. Villa, 1997, Isotope systematics in vein gold from Brusson, Val d'Ayas (NW Italy), 2. (U+Th)/He and K/Ar in native Au and its fluid inclusions: Chem. Geol., v. 135, p. 173-187)) or U-Pb on rutile, sphene, ...

Specifically for the problem of remobilization, one could also try to attack it with the approach of isotopic tracing. This, however, requires knowledge of what the original isotopic signature of a given deposit was (or should be) in order to decide whether or not the obtained data may be deduced from the original signature (i.e. remobilization vs. new input). Feasibility of such an approach strongly depends on how well a given system is characterized, and this knowledge also allows to choose the most promising tracer isotopic system.

I think that facing analytical problems like low blanks, uncertainty of successful analyses and so on is common for non-standard approaches in isotope geochemistry; the limitations of such approaches are given by the lab-infrastructure and should not be dictated by the degree of expected success - where would this lead to? And fluid inclusion - residue approaches e.g. for sphalerite document success beside magmatic systems (see also most recently: Christensen, J. N., A. N. Halliday, K. E. Leigh, R. N. Randell, and S. E. Kesler, 1995, Direct dating of sulfides by Rb-Sr: A critical test using the Polaris Mississippi Valley-type Zn-Pb deposit: Geochim. Cosmochim. Acta, v. 59, p. 5191-5197).

These are a few general thoughts I have. Please feel free to throw in more specific questions.

Good Luck!

Thomas

######################################
Thomas Pettke
Dept. Geol. Sci.
The University of Michigan
2354 C.C. Little Building
Ann Arbor, MI 48109-1063
USA fax: (+1) 313 763 4690
phone: (+1) 313 763 9368
######################################