The following is a short discussion of apparent low eutectic temperature observations that was held during June 1997

See also the 2nd discussion.


I have found some nice inclusions in fluorite, from a fault zone. They are 2 phase inclusions, 5 vol.% bubble. Eutectic melting appears at about -80 deg. C.
Does anyone know which chemical system that might be ?


Ronald J. Bakker
Geologisch-Palaontologisches Institut
Universitat Heidelberg
Im Neuenheimerfeld 234
D-69120 Heidelberg
tel (49) 6221-544834
Couldn't it be metastable Te in the system H2O-NaCl-CaCl2 (cf. Oakes et al., PACROFI mtg 1992)?



In these "high salinity basinal brine (NaCl-CaCl2-H2O) type fluids". Low eutectics seem to be a common phenomenon. Some people think this may be due to the presence of lithium. However, these fluids show some strange metastable phenomena. In the same inclusion you can get stable melting:

L+V+hydrate+ice -> L+V+ice -> L+V (Tice ~-25; Thyd ~-10)

which obeys all of the "topological" constraints of the system. But when you try to repeat the experiment the inclusion follows a completely different melting path, where hydrate is completely absent:

L+V+ice -> L+V (ice melting at ~-40).

This type of metastable behavior may indicate that that low eutectics may also be metastable.

Anyone else observed this? A topic for discussion at ECROFI



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
Hello to all,

These low temperature "eutectics" have always bothered me because they seem to require exotic compositions and even those compositions rarely have experimentally documented eutectics at temperatures as low as inclusionists' observe.

Jon Naden raised the problem of metastability. This is probably the key issue, one that Ed Roedder addressed many years ago. Now that you all have feasted on Ed's reprints, be sure to read the one on metastability.

There is one metastable phenomenon that has not been discussed much in the fluid inclusion literature: the glass transition temperature. Actually, this is not a metastable transition, but let's leave semantics aside. Most high salinity aqueous fluids "freeze" to a glass on rapid cooling and only nucleate crystals on warming. The temperature of devitrification of aqueous electrolyte glasses is typically -100 deg. C to -80 deg. C or so ( I don't remember the exact numbers, but check Austen Angell's publications). I suspect that many "low-T" eutectics are observations of the crystal growth from the glassy solid.

Naturally, I could be wrong . . . however, I do know that the kinetics of crystallization of highly saline aqueous fluids is a complex issue. If you want a fun time, go to the Gordon Conference on water (no, not the hydrothermal geochemistry one that you all go to, but just plain water). It is an amazing conference.

Have fun with this topic. I've got to go to the field and collect some inclusion samples where the mean annual temperature is -10 deg. C and humans are not the top of the food chain.



Robert C. Burruss
U. S. Geological Survey
Energy Resources Program
Box 25046 MS 973
Denver, CO 80225 USA
telephone: 303-236-5795
FAX: 303-236-3200

As Jon points out, Ron's observation is consistent with the topology of the NaCl+CaCl2+H2O system. The observation is also consistent with NaCl+MgCl2+H2O and probably other ternary or higher order systems containing geologically common salts. The metastable relationships Jon cites are indeed common in both 'naturally' formed inclusions and in lab synthesized flincs. As Pierfranco pointed out, this has been discussed before (by myself and others, cf Davis et al, GCA 1990 p591-602). Not that I mean to be inflammatory but Ron should've done his homework.

The business of invoking lithium- or strontium-chlorides (two systems that some folks insist on invoking to produce 'eutectecs' at < 50 deg. C) is tenuous to say the least. Aqueous-electroyte systems containing either metal in sufficient quantity as to produce visible melt at T < -60 in fluid inclusions are not common in nature (off the top of my head I'd say pegmatites would be the only viable example). Given that Ron's flincs are in "fluorite, from a fault zone" choosing between simple metastability and high concentrations of exotic salts seems a bit of a no-brainer.


Charlie Oakes


Hello all,

I agree that metastable first melting is probably almost always observed in high salinity fluids, particularly of the NaCl-CaCl2-H2O type. Recorded "first" melting temperatures can be well below the stable eutectic for saline inclusions - however a rapid burst of melting at, or close to, the stable eutectic can be observed.

I haven't seen melting behaviour of the type described by Jon but it's an interesting phenomenon. An observation that's always struck me as interesting is the fact that ice melting (in systems where hydrate is the final phase to melt) commonly seems to be "fixed" at c. -24C (or displays a narrow range around that temperature). Perhaps this is a reaction point in the quaternary Na-Ca-K system? In addition, hydrate commonly melts above +0.1 deg. C - metastability or modified topology of the Na-Ca-Cl system due to the presence of K? I've certainly had hydrates apparently stable for several hours above zero which melted rapidly once their apparent stable melting temperature (c. +7 deg. C) was reached.

On a point of order I'm not sure that Charlie's somewhat inflammatory comments are what we want to see on the list. To those of us aware of a possible explanation, even a likely explanation, for something, some questions may seem a bit naive. However, I think it is the purpose of the list to answer such questions for the benefit of all. Besides, these simple questions can often generate an interesting, and all too rare, discussion.

[Editor: I agree wholeheartedly with Jamie here on his point of order. PEB]

Jamie Wilkinson


As an afterthought on this one, I'd like to comment on Charlie Oakes's point about the unlikelihood of Li, Sr for example as important cations in inclusion fluids.

Crustal fluid chemistry can show very big departures from rock chemistry, and the major elements in one may be traces in the other. In the case in point, Charlie is of course right that Ca is the obvious culprit. But it is very misleading to cite MgCl2 as a likely cause of low eutectics, because Mg is usually present only at very low concentrations in rock-equilibrated fluids: generally much less than Fe and Mn and not uncommonly below Sr. Except under strongly reducing conditions, Mn is often comparable to or greater than Fe. Li is actually quite an important cation in many magmatic fluids, and in dilute, well fractionated pegmatite fluids can be much more important than any of the divalent species. Remember that the proportion of divalent to monovalent cations goes up rapidly with salinity for any given rock buffer. Other remarkably abundant species in crustal fluids can include As..... So, in interpreting eutectics, etc., think fluid chemistry, not rock or mineral chemistry!

Bruce Yardley



Thanks for making your very valid point. Fluid inclusion compositions represent the EXTREME end of differentiation; as a good example, note the occurrence of obscure daughter minerals, such as avogadrite (for those without a Glossary, this is potassium cesium borofluoride) in pegmatite inclusions. There is no reason to exclude ANY element from "possible" occurrence in inclusion fluids. I am sure that there are many new minerals, yet to be discovered, that occur ONLY as dms in inclusions, and these only represent a part of the elements that must have been present in the original trapped fluid. LONG LIVE FLUID INCLUSIONS!!

Ed Roedder



Regarding Bruce's and Ed's comments on the low eutectic theme I'll make some semantic points and some additions to the scienctific discussion. I'll ignore earlier misrepresentations of my statements that were non-science related.

So, let's review what I said:

"Aqueous-electroyte systems containing either metal [refering to Li and Sr] in sufficient quantity as to produce visible melt at T < -60 in fluid inclusions are not common in nature (off the top of my head I'd say pegmatites would be the only viable example)."


"The observation is also consistent with NaCl+MgCl2+H2O and probably other

ternary or higher order systems containing geologically common salts."

I did not discuss probabilities of solute occurrences or 'likely' causes of low eutectics. Nor do my statements advocate exclusion of chemical systems from '"possible" occurrence in inclusion fluids'. My first quoted statement above refers to lever-rule requirements and geologic processes which could generate fluids which would satisfy those requirements in fluid inclusions. For Bob Bakker's case and many other published cases the simplest explanation of apparent Te at < -60 C is metastability. While the relative abundances of Li, Sr, Mn... may be much greater in a rock-equilibrated fluid than in the rock the concentrations of these elements in fluid inclusions must be large if they are to be invoked as the solutes responsible for discernible melt at T < -60 C. It is implicit that these elements be present in solution - not merely present (e.g. as a daughter mineral with low to neglible solubility at low temperature).

So the bottom line is - everybody's comments regarding this topic are valid (if my latest are not I'm sure somebody will point it out) and variable; so please read the words as written.

Charlie Oakes