The controversy about the use of selective or weak extractions (SWE) to detect buried or blind mineralization continues unabated. Proponents of the methods often appear to be associated with the companies that offer the analytical services (Birrell, 1996; Clark, 1993; Mann et.al., 1998) whilst studies that reveal the difficulties in using such methods are primarily from arms-length institutions (Bajc, 1998; Gray et.al., 1999; Seneshen et.al., 1999; Smee, 1997). I expect the debate will continue into the foreseeable future.
The principal uncertainty with SWE is the lack of a solid understanding of the geochemical transport processes that might give rise to interpretable element patterns. This understanding is critical in order to produce predictable results from a particular mineralized target. Additionally, without this knowledge it is not possible to select the most appropriate SWE method for each climatic and geological environment, or choose the most revealing method of interpretation.
Figure 6: Copper contour overlain on calcium as 3D relief map, Oyu Tolgoi, Mongolia. The major copper anomalies caused by sulphide mineralization are surrounded by high calcium. Calcium is absent from the immediate area of the mineralization.
Whatever the transport process, at least one ion of interest must move through cover in order to be used for mineral exploration. Ions might move in solution, by seismic pumping along fractures (a form of diffusion), by gaseous transport (another form of diffusion), or by true diffusion along overburden grain boundaries.
If ions move by hydromorphic processes, (i.e. ions dissolved in water), then a careful application of “normal” geochemical methods should detect responses to mineralization where the water meets the surface boundary. More likely, any ore-forming elements in interstitial soil or overburden waters would be in the form of colloids or adsorbed on soil grain surfaces, and would be transported by diffusion as well.
The theoretical transport velocities of cations due to diffusion through clay and calcareous soils was investigated by Smee, (1983a), and validated by radioisotopic measurements (Smee, 1983b). It was shown that concentration profiles can be calculated for the soil and boundary conditions important to applied geochemists by the relationship:
C(x,t) is the concentration of diffusing substance at distance x (cm) from the interface and time t (s) from the start of diffusion, in M cm-3. Co is the initial concentration of diffusing ion, M cm-3. l is the distance from the initial diffusion boundary (base of overburden) to the upper diffusion boundary (usually the soil surface or site at which ions are fixed by either chelation (organics or sesquioxides) or precipitation (high pH) and therefore become immobile.
De is the effective diffusion coefficient, which must be determined by taking into account the media tortuosity, moisture content, ion stability in the Eh-pH regime, and partition coefficient between particles and interstitial water for the ion in question and the soil substrate.1
Equation 1 can be used to calculate concentration profiles for any initial condition of Co and for any time, t for chelatable cations such as Zn2+, Cu2+, Mn2+, and Fe3+. Chelatable ions become fixed at the upper boundary by some process (adsorption, chelation, precipitation) and change the concentration gradient as diffusion progresses. Diffusing ions that are not chelatable or otherwise rendered immobile in the near surface soils such as H+, K+, and Na+, require a different boundary definition:
However, more important to applied geochemists is the total mass per unit area (Mt/A) of diffusing ion that will pass into an organic or other soil horizon over a length of time, t, at a distance l, from the initial boundary. This expansion is given by Crank (1975) as:
Equation 3 was used by Smee (1983a) for a number of ions in the glaciolacustrine clay belt of Canada using the following conditions for mass transport:
1. The De in soils can be 5 orders of magnitude less than the published Do in aqueous solutions. The De values for the number of cations can be found in carious soil science publications.
2. Assuming total fixation of cations in a 10 cm thick organic horizon of denisty 1 g/cc.
The time t is 8000 years, an estimate of the period for a sulphide body to undergo oxidation beneath a till and clay cover since the last Canadian glaciation. Co concentrations were actually measured in clays and waters from the base of the overburden. De values (with the exception of Zn, which was actually measured using a radioisotope, Zn65) were obtained from various publications, usually in the soil physics, plant nutrition, or radioactive waste disposal disciplines. The De for H+ is nearly 300 times higher than for the ore-forming cations. The boundary length l varies from 100 to 500 cm and is the distance to the upper boundary (surface) from the lower boundary.
The calculations show that only H+ can move through anything more than 100 cm of clay cover in a mass per unit area that could be measured by any form of SWE or instrumentation. All other ore-forming cations occur in concentrations less than 1 x 10-10 M cm-2, much lower than what can be separated from normal background noise. This situation has not changed in the 20 years since this work was published, regardless of leach or analytical sensitivity (Smee, 1983b). Unless a form of transport other than diffusion is active (such as hydromorphic migration through porous soils, or mechanical transport) a direct anomaly formed from the target cations in the glaciolacustrine-covered areas will not be detected through more than 5 m of impervious cover. The target ions never make it to the surface at all. Other parts of the world have their own geochemical challenges. Semi-arid to arid areas are usually highly alkaline and oxidized. Under these Eh-pH conditions, ore-forming cations are normally stable in the form of oxides and carbonates (hence oxide caps). The ore-forming cations move very little because De is extremely small, even though time is large. Again, no direct geochemical response to mineralization should be expected in arid areas with appreciable exotic overburden cover.
The geochemical method that should be most consistently successful, regardless of the environmental conditions, is one that seeks an indirect response to sulphide mineralization. An indirect geochemical response is a change in the distribution of major or trace elements at the overburden-surface boundary that is in response to another stimulus associated with the underlying mineralization. Examples of an indirect pattern have been published for years by many workers in both the mineral and petroleum exploration industries (Donovan (1974); Smee (1983a,b; 1997, 1998, 1999, 2001); Hamilton (1998), Hamilton et.al., (2001b)).
A further form of response to buried or blind mineralization could be called a secondary indirect indicator. This is a pattern formed by base or precious metal cations that appear to be caused by their movement through overburden above or on the margins of mineralization. These patterns, which are often highlighted by SWE methods, are actually formed by the scavenging of available cations in the surface soil by reprecipitating pH sensitive cations. Positive responses in SWE Cu in soils, for instance, may correlate perfectly with a SWE/Hot Acid Fe, Mn, or Ca ratio. A regression of these two variables completely eliminates the Cu response. Thus many of the published successes of SWE methods are actually related to changes in pH in the surface soils.
Figure 1: A ratio of EDTA Cu (a SWE method) to hot acid extraction (HA) Cu, crrected for the total amount of C in the soil, which controls the amount of chelating sites and thus the SWE response. It appears as though the Cu has formed a valid response above the Magusi sulphide deposit that is overlain by >5m of varved glaciolacustrine clay (Smee, 1983a).
This phenomenon is illustrated in Figures 1 and 2. The example in Figure 1 is from the Magusi River Cu-Zn massive sulphide deposit in Western Quebec, Canada. The sulphide horizon subcrops beneath 5-10 m of glaciolacustrine sediment and till. The SWE Cu, done by cold EDTA/AAS is ratioed against a hot aqua regia extraction to show that the Cu anomaly is not caused by a simple increase in total Cu. That ratio is then corrected for the total number of ion binding sites in the organic-rich soil by ratioing against organic C. The resulting pattern makes it appear that the SWE methods detected a clear double-peak Cu response from the margins of the sulphide body.
Figure 2: The SWE Cu response in Figure 1, ratio to SWE Fe response in the same samples. The Cu anomaly is fully accounted for by the soil Fe.
Figure 2 shows the result of taking the above responseand regressing against a similar treatment for SWE Fe. Ifthe Cu actually traveled through the clay to surface, therewould be a residual Cu pattern above the sulphide. In fact,the Fe accounts for all the Cu, leaving no trace of Cuattributable to the upward migration of that element.
The model for the formation of at-surface indirect element anomalies such as Ca, Fe, Mg and Mn in responseto a change in H+ in Canadian Shield conditions was developed by Smee (1983a,b) and confirmed by Hamilton et al. (2001b). Whether the geochemical gradient in overburden is caused principally by a change in Eh or pH is irrelevant to the method of transport of the resulting ionic products. All products must move by some form of diffusion and are thus governed by diffusion principles.
Smee (1997, 1998) published a similar model of ion transport and indirect anomaly formation for arid conditions which showed the added possibility of gaseous transport in dry overburden. The De for gases traveling through a dry porous medium is usually much larger than for ionic species in water, thus producing the interesting possibility of indirect geochemical anomalies that have formed through a significant thickness of cover.
All geochemical contributors attribute these indirect geochemical patterns to changes in Eh-pH conditions induced by the oxidation of the mineralization, or reduction of the overlying rock by hydrocarbons. Workers recognize that a reproducible measurement of Eh (Oxidation-Reduction Potential ORP) is a difficult and time consuming undertaking (Smee, 1983a; Hamilton et al., 2001a). As well, workers recognize that an in-field pH measurement is simple, fast, reproducible and cost effective, and produces similar or clearer responses to mineralization than does a measurement of Eh or ORP.
Smee (1983a) suggested that a simple field pH measurement, that reveals patterns in the pH sensitive cations such as Ca+2, would be an effective indirect geochemical exploration tool in the Canadian Shield. Smee (1998, 1999) further illustrated that field pH measurements could be a viable direct and indirect geochemical tool for arid terrain using an example from Nevada. A pH meter and a source of distilled water are all that is necessary to locate sulphide mineralization.
A few companies began testing the concept either in the field or by measuring pH routinely in the laboratory. Some of these results have been released with permission.
Soil H+ profiles over gold-bearing structures, Chile
The Atacama desert of northern Chile is one of most arid places in the world. Much of the terrain is overlain with salt deposits consisting of calcite, gypsum, nitrates and halogen salts of various compositions. These salt deposits may form over shallow colluvial cover or valley-filled alluvium that may reach several hundreds of metres thickness. The surficial aridosols are almost universally alkaline and may exceed a pH of 10 in some instances. As previously mentioned, most base metal cations are stable in these conditions, so the De for all of these cations is extremely small. It is unlikely that base metal cations are transported through this alkaline cover in any concentration that could be detected by soil sampling.
In this environment, like the Nevada example, a portable pH meter and some water may be all that is required to detect the products (both direct and indirect) of buried or blind oxidizing sulphides. The direct indicator would be a change in soil pH, whilst the indirect indicator would be a change in concentration of the pH sensitive cations such as Ca, Fe and Mn.
Plate 1: Field Portable pH meter measuring surface soil pH, Atacama desert.
Experience has shown that the latter style of anomaly is not easily reproduced in the Atacama desert because the presence of gypsum masks the possible movement of calcite, and the formation of gypsum is not sensitive to pH conditions. The amount of Ca in the form of calcite is not easily separated from that of Ca that is in the form of gypsum, when only performing multi-element ICP-ES or MS analyses. Attempts at screening ICP analyses by using S and ratios with Ca have not been successful.
A series of field tests have been carried out using a portable pH meter and distilled water with soils collected on traverses over known mineralized structures. Approximately a teaspoon of soil was mixed with 50 ml of distilled water for 2 minutes, before reading the pH, as shown in Plate 1.
Results were converted to moles H+ to remove the log scale before plotting. A single traverse over a buried mineralized structure is shown in Figure 3. Soil samples were spaced at 50 m over the surface projection of the mineralization, and 100 m away from the mineralization. The H+ clearly shows the position of the sulphide mineralization as a double-peak response. There is a 10-fold change in H+ concentration compared to background, however there is no discernable visible change in the surface soil textures or mineral components.
Figure 3: Profile of water slurry H’ in soils above a sulphide-rich vein, Atacama desert, Chile.
The relationship of soil Ca to soil Cu Oyu Tolgoi, Mongolia
The Oyu Tolgoi (Turquoise Hill) project is located within the South Gobi desert in Mongolia. The climate is classified as semi-arid with low shrubs and small trees supported by meager precipitation. Soils are typical aridosols featuring a ubiquitous calcite-enriched horizon near surface. This caliche is not indurated into a hard layer or crust, but can be easily outlined with a few drops of dilute HCl. Mineralization is still being discovered, and is taking the form of multiple disseminated copper-gold-molybdenum bodies within or around intrusive rocks. Sulphide mineralization is highly weathered and forms both oxide and secondary sulphide layers above or in proximity to the primary sulphide mineralization.
The presence of copper in this area has been known since the bronze age. The initial modern exploration was performed by BHP Minerals in the mid to late 1990s. As part of that exploration, BHP conducted a soil sampling orientation program over then known mineralization. Samples were analyzed for pH, strong acid soluble elements by ICP-ES, and most of the available SWE methods. All of the extraction methods reveal the sub-cropping mineralization as shown by copper in Figure 4.
Figure 4: Aqua regia soluable copper in soils ppm, Oyu Tolgoi, Mongolia. Survey clearly reveals the main mineralization. Area to the east is covered by sand and overburden.
A commercial laboratory did the soil pH measurements over a period of three years. The pH analytical method either varied from year to year or the internal controls were not stable over the sampling period. This produced pH data that is obviously influenced by sampling episode, and resulted in “along-line” patterns that do not represent geological features. Nonetheless, soil pH (when converted to Moles H+) is useful in showing northeast and northwest striking structures in the overburden-covered eastern areas. This area was sampled and analyzed in one campaign. Curvilinear responses on the edges of known mineralization occur in the central portion of the grids as shown in Figure 5.
Figure 5: Contours of H’ in soils as done by commercial laboratory over a three-year period, Oyu Tolgoi, Mongolia. Note the linear alongline patterns that show changes in methods or quality control factors. This analytical instability makes the use of this data problematic. Linear anomalous features on the east side of the grid have been related to bedrock structure.
Copper contours superimposed on a 3D relief image of aqua regia soluble calcium in soils are shown in Figure 6 (page 1). Although the calcium shows a great deal of noise in other portions of the sampled area, it is absent from soils over the main areas of mineralization as shown by the copper. A weak two-line copper response in the northern portion of the grid is also devoid of calcium. This is the area of the newly discovered North Oyu mineralization.
There is therefore both a direct and indirect indication of mineralization in the aqua regia multi-element information. A single aqua regia and ICP-ES analyses would give both indications of mineralization for a nominal cost of analyses. A field measurement of soil pH, or a well-controlled laboratory measurement should produce information helpful in locating oxidizing sulphides and water-bearing structures.
The use of a simple pH measurement on soils in the field can produce both direct and indirect indications of sulphide mineralization. The method is not specific to any deposit type; only the presence of sulphides is required. More importantly, the distribution patterns shown by H+ and the pH sensitive elements such as Ca, Fe, Mn, and Mg indicate that a specific mode of element transport is operational through many types of soil and in many climatic conditions. Over the past 25 years I have observed similar patterns of element distribution from the high arctic through the boreal forests and the Atacama desert. This is a clear indication that a specific transport mechanism is operational and should be examined in detail to fully understand the geochemical methods that are useful when searching for blind deposits.
The data required to understand these mechanisms are already available, but not in the applied geochemical field. Soil and forestry physicists, nuclear waste management scientists, and other experts in ion movement have an existing bank of data and experience available for the having. Several years ago I put forward a proposal to a Canadian mineral research organization to hire an ion migration specialist from outside the field of applied geochemistry who would compile the required data from the various sources. This proposal was rejected out-of-hand as not being relevant to solving the problems of exploration for blind mineral deposits. Now that many hundreds of thousands of dollars have been spent, and we appear to be where we were twenty-five years ago, I suggest this idea be revisited by an applied geochemical research organization. If we as a group of applied geochemists are to survive, we must broaden our knowledge of sister fields. This cross-fertilization of ideas will, I believe, give the impetus we require to make significant advances in searching for mineral deposits.
I would like to express my appreciation to Dave Heberlein of Barrick Gold and Charles Forster of Ivanhoe Mines for technical and moral support and for permission to present the data from the Atacama desert, Chile and Oyu Tolgoi, Mongolia respectively.
by: Barry W. Smee
Smee and Associates Consulting Ltd.
4658 Capilano Road, North Vancouver B.C. V7R 4K3
Phone 1 (604) 929-0667 Fax 1(604) 929-0662
Bajc, A.F.: 1998: A comparative analyses of enzyme leach and mobile metal ion selective extractions; case studies from glaciated terrain, northern Ontario. J. Geochem. Explor., 61: 113-148.
Birrell, R., 1996: MMI geochemistry: mapping the depths. Min. Mag. May: 306-307.
Crank, J., 1975: The mathematics of diffusion. 2nd Edition, Clarendon Press, Oxford, 414p.
Clarke, J.R., 1993: Enzyme-induced leaching of B-horizon soils for mineral exploration in areas of glacial overburden. Trans. IMM, 425: B19-29.
Donovan, T.J., 1974: Petroleum microseepage at Cement Oklahoma: evidence and mechanism. Am. Assoc. Petrol. Geol. Bull. 58: 3: 429-446.
Govett, G.J.S., 1973: Differential secondary dispersion in transported soils and post-mineralizatiion rocks: an electrochemical interpretation. In: Geochemical Exploration 1972: Papers presented at the Fourth International Geochemical Exploration Symposium, London, April 17-20, 1972, M.J. Jones, (Editor), Inst. Min. Metall. (London), p. 81-91.
Govett, G.J.S., 1976b. Detection of Deeply Buried and Blind Sulphine Deposits by Measurement of H+ and Conductivity of Closely Spaced Soil Samples. JGE 6 (3), p. 359-382.
Govett, G.J.S. and Atherden, P.R., 1987. Electrogeochemical Patterns in Surface Soils – Detection of Blind Mineralization Beneath Exotic Cover, Thalanga, Queensland, Australia. JGE 28 (1/3), p. 210-218.
Govett, G.J.S., Dunlop, A.C. and Atherden, P.R., 1984. Electrochemical techniques in deeply weathered terrain in Australia. JGE 21 (1/3), 311-331.
Gray, D.J., Wildman, J.E., and Longman, G.D., 1999: Selective and partial extraction analyses of transported overburden for gold exploration in the Yilgarn Craton, Western Australia. J. Geochem. Explor. 67: 1-3: 51-66.
Hamilton, S.M., 1998: Electrochemical mass-transport in overburden: a new model to account for the formation of selective leach geochemical anomalies in glacial terrain. J. Geochem. Explor. 63: 155-172.
Hamilton, S.M., Cameron, E., McClenaghan, B., and Hall, G., 2001a: A comparative study of selective leaches and evaluation of element migration processes through thick overburden, Abitibi Clay Belt, Ontario. CAMIRO Report 18, Deep Penetrating Geochemistry.
Hamilton, S.M., McClenaghan, B., Hall, G., and Cameron, E., 2001b: Preliminary Report, Cross Lake. CAMIRO Report 8, Deep Penetrating Geochemistry.
Mann, A.W., Birrell, R.D., Mann, A.T., Humphreys, D.B., and Perdrix, J.L., 1998: Application of the mobile metal ion technique to routine geochemical exploration. J. Geochem. Explor. 61: 87-102.
Seneshen, D.M., Alcock, R.A., and Nicol, I., 1999: The application of partial extractions to mineral exploration in areas of transported overburden. Symp. Prog. and Abst., AEG 19th IGES, p. 49-50.
Smee, B.W., 1983a: Laboratory and field evidence in support of the electrogeochemically enhanced migration of ions through glaciolacustrine sediment. Ph.D. Thesis, Univ. New Brunswick. 477 p.
Smee, B.W., 1983b: Laboratory and field evidence in support of the electrogeochemically enhanced migration of ions through glaciolacustrine sediment. In: G.R. Parslow (Editor), Geochemical Exploration 1982. J. Geochem. Explor., 19: 277-304.
Smee, B.W., 1997. The formation of surficial geochemical pattern over buried epithermal gold deposits in desert environments. Results of a test of partial extraction techniques In Exploration 97 Symposium Volume, Toronto: 301-314.
Smee, B.W., 1998: A new theory to explain the formation of soil geochemical responses over deeply covered gold mineralization in arid environments. J. Geochem. Explor., 61: 149-172
Smee, B.W., 1999: The effect of soil composition on weak leach solution pH: a potential exploration tool in arid environments. Explore: 102, 4-7.
Smee, B.W., 2001: Travels of a transient metal ion, the shocking truth revealed: things to be aware of when using the “new” geochemistry. Vancouver MEG Workshop notes, Dec. 2001.