Geochemical exploration using palaeo-hydrothermal fluids

Kingsley Burlinson

A presentation at the SGA conference, Uppsala, Sweden, August 2013

Many mineral deposits are formed by hydrothermal processes. To explore for these we make extensive use of geophysics and geochemistry but rarely do we use the fluids themselves in exploration, despite the fact these fluids are preserved as fluid inclusions. With carefully chosen analytical methods we can easily derive very useful information from the fluids themselves to use in exploration for hydrothermal mineral deposits.

Some typical inclusions trapping the palaeo-hydrothermal fluids are shown here.

Aqueous inclusions are very common, but not usually useful for mineral exploration.
CO₂ rich fluid inclusions are frequently associated with mesothermal gold deposits and are often a good indication of deep-sourced fluids which may have transported and deposited gold and other economic minerals.
Highly saline inclusions with a daughter crystal of halite are common in the core zone of porphyry copper systems, or other intrusion related deposits.

fluid inclusion types

This model for the formation of mesothermal gold deposits shows that CO₂ rich fluids are often derived from metamorphic de-volatilisation. These fluids may have dissolved gold from the source region. As the fluids ascend to the surface their temperature drops and CO₂ may ex-solve as the pressure decreases and these changes can lead to deposition of the gold in solution to form a deposit. The CO₂ also buffers the fluid in a pH range which favours the solution and transport of gold. (Phillips & Evans) Hence the presence of CO₂ rich fluids is a good exploration guide. Using CO₂ as an exploration guide provides a larger and more consistent target than trying to use geochemical analyses or mineralogical zoning. It is also advantageous in the detection of blind deposits which are otherwise difficult to locate.

Traditional microthermometric methods to determine CO₂ contents are slow and tedious and usually require petrographic sections. But for exploration we can use the baro-acoustic decrepitation method to easily and quickly determine approximate CO₂ contents. This method uses a computerised instrument and is completely objective as it avoids the need for visual observation with its potential for bias. Analyses are done on crushed grain samples and there is no need to prepare petrographic sections. The analysis is rapid and takes just 30 minutes per sample and so large numbers of samples from a spatial array can be analysed in the same manner than geochemical surveys are undertaken. The presence of CO₂ in the sample is shown by a distinctive peak in the baro-acoustic decrepigram result.

This is the model 105 decrepitation instrument in current use.
model 105 decrepitometer

The analysis result is a histogram of counts versus temperature.
typical decrepigrams

This shows the decrepitation curve for 2 different quartz samples, one (blue) without CO₂ and the other (red) with CO₂ rich fluid inclusions. The CO₂ causes a peak at unusually low temperature which is characteristic of the presence of CO₂ and the peak amplitude is an approximate estimate of the CO₂ amount. (Other non-condensible gases such as CH₄ also contribute to this low temperature decrepitation peak.) The green result is an analysis of quartz that has previously been analysed. It shows no response at all and confirms that the measurements are of fluid inclusions and not crystallographic effects. Fluid inclusion decrepitation is destructive and hence irreversible, but many crystallographic transitions are reversible and would also be detected on a re-analysis of the previously analysed quartz.

PT graph for h20 and co2

Using this P-T graph, we can easily explain why CO₂ rich inclusions cause a distinctive low temperature peak on the decrepitation results.
Consider 2 inclusions formed at the "formation point" of 380 C and 1000 bars, one with only water and the other with only CO₂.
At room temperature the aqueous inclusion will have condensed to a liquid with a vapour bubble. As it is heated the internal pressure (blue line) does not rise much until after the liquid expands and eliminates the vapour bubble at the "homogenisation point". The pressure will then rise quickly with further heating, following the green isochore line until it decrepitates near 350 C.
In contrast, the CO₂ inclusion does not condense and remains as a gas phase. When heated the internal pressure is determined from the gas law equation (PV=nRT) and rises linearly as shown by the magenta line. Decrepitation occurs at the much lower temperature of 240 C, giving the characteristic low temperature decrepitation peak due to gas rich inclusions.

An additional more detailed explanation of the cause of low temperature decrepitation is here.

The baro-acoustic decrepitation method exploits this behaviour to provide an easy way to determine the CO₂ content of inclusion fluids.

Using CO₂ in gold exploration at Woods Point, Vic., Australia

The Morning Star mine at Woods Point is about 120 Km north-east of Melbourne. It has produced over 900,000 oz of gold since discovery in 1861.

The gold mineralisation is associated with the intrusion of a late Devonian aged dyke swarm within the Silurian and early Devonian sedimentary host rocks.

(Map modified from "A geochronological framework for orogenic gold mineralisation in central Victoria, Australia" by Bierlein, Arne, Foster & Reynolds, Mineralium Deposita (2001) V36:741-767). KEY: Major intrusives are: WRG White Rabbit Granite; SG Stawell Granite; MAG Mt Ararat Granite; MB Mt Bute; CBG Cobaw/Pyalong Granite; TP Tarnagulla Pluton; HG Harcourt Granite; SBG Strathbogie Granite

Samples from the Morning Star mine and adjacent areas were collected by Caitlyn Hoggart as part of her thesis work. 34 samples were analysed by baro-acoustic decrepitation.
Most samples had a prominent low temperature decrepitation peak indicating the presence of CO₂ rich fluid inclusions as seen here in samples from the Morning Star mine adit.

woodspoint decrepigrams

Each sample result curve was de-convoluted into component skewed-gaussian curves as described here.
This mathematical procedure provides consistent and reliable values for the temperature and height of the decrepitation peaks in each sample to facilitate inter-sample comparisons.

This is an example of the results of de-convolution of sample 512 into 4 component curves. The black line (frequently hidden beneath the red line) is the raw data while the red line is the mathematically calculated best fit to the raw data.

de-convolution of a woodreef sample

Comparison of all the CO₂ peak data from all the samples shows that the temperature does not vary much across the field. But there are significant and potentially informative variations in the amplitude of the CO₂ peak, reflecting variations in the abundance of CO₂ rich inclusion populations in each sample.

woodsreef T and CO2 summary

This plot compares the gold analyses with the low temperature CO₂ caused decrepitation peak height. All except one sample (sequential sample #3 in this plot) containing more than 10 ppm gold had a high CO₂ peak. (The magenta lines connect all the above background Au results and their CO₂ analysis.)
Because CO₂ rich fluid inclusions are widely dispersed around mineralisation they form a large anomaly target. Exploration for these fluids is better than relying on gold results which are less widely dispersed and often erratic due to nugget effect irregularities. But this study is incomplete due to the lack of distal unmineralised comparison samples.

woodsreef Au analyses and CO2

Saline fluids in porphyry copper and intrusion related systems

The relationship between highly saline fluids and the core zone of porphyry copper systems has been widely documented, including this old summary from 1981. As the parent intrusion crystallizes, incompatible minerals concentrate in the last stage residual aqueous fluids. Salt also concentrates in these last stage fluids, which then form the economically interesting mineral deposits as they migrate away from the intrusion. These saline fluids can be used to identify potentially mineralised zones as the saline fluids are dispersed more widely than the mineralization itself. They can be used to vector in towards the core zone of the intrusion and its associated mineralisation.

Although measuring precise salinities of fluid inclusions can be complicated and slow, such detailed measurements are not necessary. In an exploration programme it is sufficient to merely observe the presence of daughter halite crystals in the fluid inclusions as these form when the fluid salinity exceeds NaCl saturation of about 26 wt. %. Quick and easy observations are adequate to recognize these important saline fluids which directly indicate the proximity to the potentially mineralised core zone of the hydrothermal fluid system.

This depositional model diagram shows the relationship between an intrusive magma and the saline fluids which concentrate in its late stage core fluids. Saline fluid inclusions occur above and peripheral to the economically mineralised zones and assist in locating the mineralised core zone and also blind deposits.

porphyry deposit model

It is very easy to make these observations and it is not even necessary to prepare petrographic sections or use a polarizing microscope. Crushed and sized grains (approx <420 microns [40 mesh] and >200 microns [80 mesh]) immersed in an oil with the same refractive index as quartz (clove oil) are quite suitable for observation on a transmitted light microscope with magnification of about 600 (40* objective, 15* eyepiece).

Despite the ease of measurement and great benefit of using these saline inclusions to assist in exploration, fluid inclusions are not used often if at all and the method was completely ignored at Cadia, NSW, Australia.

This image shows fluid inclusions in crushed grains in oil. The right hand image is at low magnification of about 60 times and the numerous dark spots are abundant fluid inclusions, each about 5 to 20 microns across. At high magnification you can easily see the contents of the inclusions as in the left hand image. (Unfortunately there are no halite daughter crystals in this image as I do not have a suitable photograph.)

A literature survey (1981) about using fluid inclusions in porphyry deposit exploration is here.

microscopy on grains in oil

Using fluid temperature measurements in mineral exploration

Academic fluid inclusion studies invariably measure numerous fluid inclusion homogenisation temperatures to determine the temperature of formation of the system. Such studies invariably record great complexity with varying types of fluid inclusions emplaced at different stages (primary, pseudo-secondary, secondary) in mineral host grains of differing paragenesis. Temperatures are usually painstakingly recorded with 0.1 C resolutions. The resulting studies are extremely detailed, but curiously they almost always summarise the temperatures as very broadly averaged histograms with very poor temperature resolution. The end result of these slow and tedious studies is the realisation that mineralised quartz veins form in similar or identical temperature ranges as barren veins and that temperature measurements are consequently of little or no use in an exploration context.

In this astonishingly comprehensive study, Tomilenko et.al measured the temperatures of 5025 quartz samples from both mineralised and barren quartz veins in the Sovetskoye gold deposit, Siberia, Russia. Their summary histograms show that there is no significant temperature difference between mineralised and barren quartz veins.

fi temperatures at sovetskoye

Measurements of fluid inclusion temperatures do not provide useful information to guide regional exploration. Such data are primarily of use in forensic studies of the genesis of deposits that have already been discovered.

However, fluid inclusion temperatures may be useful in carefully controlled studies of an individual deposit to outline zonation.

Temperature zonation within the Malanjkhand copper mine, India

Malanjkhand is a large open pit copper mine in central India

malanjkhand mine location map

The copper occurs within an extensive quartz reef within granitoid host rocks. A description of the deposit is given in: "The Malanjkhand copper (+molybdenum) deposit, India: mineralization from a low-temperature ore-fluid of granitoid affiliation" by M.K. Panigrahi and A. Mookherjee in Mineralium Deposita (1997) V32:p133-148.

Although sometimes classified as a "porphyry copper" type deposit, quartz is the dominant accessory mineral (almost the exclusive accessory mineral) in the ore zone and this is quite unlike typical porphyry copper deposits elsewhere. But the abundance of quartz allows detailed fluid studies throughout the pit.

malanjkhand mine geology map

Samples were collected from the pit itself and from adjacent areas where possible. 8 locations were sampled and are geo-located on this satellite image. At each location multiple samples were collected to examine fluid variations on both local and regional scales. The prefix MJ together with these site numbers is used in the following diagrams to refer to the sample collection locations.

satellite image malanjkhand

These 8 samples collected at sample location MJ4 on the eastern wall of the main pit in the ore zone are typical of all the results from Malanjkhand. There is no low temperature decrepitation indicating that the fluids are aqueous without significant gas content. Some differences in the temperature of the decrepitation peak near 450 C are apparent between samples.

h2240 decrepitation result

To accurately determine a temperature for each sample to enable comparison they were all de-convoluted to their component skewed-gaussian distributions, as described here. The mode temperature of each peak was used for inter-sample comparison. This temperature is only approximately related to the homogenisation and formation temperatures of the sample fluids but it is a convenient and consistent temperature for comparison of a suite of similar samples.

In this fit plot, the black line is the raw (smoothed) analytical data. This has been fitted by 2 gaussian curves in cyan and green. The yellow curve is the mathematical sum of the 2 fitted gaussian component curves. The red curve, which is almost completely concealed beneath the yellow curve, is the regression fit curve using the Levenberg -Marquardt algorithm. Further discussion of the mathematical fitting methods is here.

fitted h2240 result

The mode temperatures of the fitted gaussian curves for each sample are plotted in the diagram below. Samples are in groups according to their geographic location MJ number. The green samples are repeat gaussian fits of the same raw data. These replicate results confirm that the mathematical fit procedure is stable and robust.

There are significant temperature differences across the pit with temperatures ranging from 454 to 508 C. Temperatures from the northern section of the pit tend to be higher than in the south of the pit. And the active ore zone at location MJ4 also has lower temperatures.

These temperature variations could be showing zonation within the ore. Samples from the Molybdenum rich north end of the pit at location MJ3 show high temperatures as might be expected in a zone associated with molybdenite deposition.

Temperature zonation such as this might be useful in mine scale mapping. However, there are numerous late stage dykes crosscutting the pit (refer to the geology map above) which may also be influencing the temperatures so it is not possible from this data to be completely certain about the cause of the observed temperature zonation.

Note that distal background unmineralised samples from the town area at locations MJ5 and MJ6 have the same temperature as mineralised samples. Temperatures alone are not diagnostic of mineralisation and are only meaningful on a carefully collected spatial array of samples.

malanjkhand summary results

The temperature variations from the above plot indicate a temperature difference across the current mine pit with higher temperatures at the north end of the pit, where Molybdenite also occurs, and lower temperatures at the south and central areas of the pit, although the zonation is indistinct and possibly affected by overprinting from the later crosscutting dykes. These differing temperature zones are superimposed on the geology map below.

malanjkhand pit with temperature anomalies

malanjkhand pit with temperature anomalies

Microthermometric temperature measurements are too slow, tedious and subjective to be useful in exploration. However, the baro-acoustic decrepitation method can be used to determine relative temperatures for inter-sample comparison on mine scale projects or for detailed zonation studies.

Conclusions

The gas content (CO₂) of fluid inclusions is a guide to many hydrothermal systems, particularly mesothermal gold deposits

Highly saline fluids can be used to locate the core zones of intrusions for porphyry copper and intrusion related gold exploration

Temperatures of hydrothermal fluids are less useful in regional exploration but may outline local zonation. Temperatures alone are not diagnostic as barren and mineralised quartz frequently form at the same temperatures

Hydrothermal fluids provide a larger and more easily recognized halo around mineralization than by using mineralogical or trace element data

Fluid inclusion data easily identifies important features of hydrothermal fluids which are highly relevant to mineral exploration

But not all the academic methods are appropriate or useful in mineral exploration

There are simple methods to measure gas contents, salinity and temperature which are quick and easy and are appropriate for exploration application

Microscope observations on crushed grains are easy and ideal for locating zones of saline fluid inclusions

Baro-acoustic decrepitation provides data on gas contents and temperature zonation

Excessive un-focused data collection and paranoia about precision is counter-productive and will conceal the useful information in an ocean of irrelevant data