Dunn Colin E. Biogeochemistry in Mineral Exploration

Подождите немного. Документ загружается.

the samples to ash prior to analysis, elements are concentrated to levels above the

detection limit and the data structure can be observed. Commonly, this exercise

reveals data distribution patterns that can be helpful for focusing on exploration

targets. Elements that are typically at or close to the detection limit by ICP-MS,

and for which ashing may provide valuable insight, include Be, Bi, Ga, Ge, In, Nb,

Pd, Pt, Re, Se, Te, Tl, U, V and W. Gold and As data can be quite imprecise at low

levels in dry vegetation; therefore these two elements can be added to this list.

0.005

0.01

0.015

0.02

0.025

Conc.

In ppm

In_Norm ppm

0.005

0.01

0.015

0.02

0.025

0.03

0.035

0.04

0.045

Conc.

Te ppm

Te_Norm ppm

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

Conc.

Nb_Norm ppm

Nb ppm

135791113151719212325272931333537394143454749515355

Sequence #

Averages

135791113151719212325272931333537394143454749515355

Sequence #

Averages

1 3 5 7 9 1113151719212325272931333537394143454749515355

Sequence #

Averages

Dry (diamond) v Ash (square)

Fig. 6-9. Examples of elements best determined on ash. Comparison of element concentra-

tions determined by ICP-MS (aqua regia digestion) in dry Acacia tissues compared to ashed

portions of the same samples. Concentrations in dry tissue (diamond symbol) are either all or

mostly below the detection limit, plotted here at half the detection limit. Ash data are nor-

malized to dry weight basis (detailed spreadsheet on CD – Table 6-ID).

172 Sample Preparation and Decomposition

Losses of elements during ashing may vary according to the plant species. Girling

and Petersen (1978) showed that in plants containing cyanogenic glycosides (e.g.,

those of the prunus and rose families) the Au volatilizes as Au cyanide well before the

normal temperature of ashing is attained. Most plants used in biogeochemical pros-

pecting do not contain cyanogenic compounds.

Ashing to 4751C results in the loss of some Cr in species such as Acacia (Table 6-

ID on CD), whereas many plants of the boreal and temperate forests lose much of

their Cr content. Figure 6-10 shows a comparison of Cr in pine bark samples (Pinus

contorta) from British Columbia. More than 70% of the Cr volatilized during ashing,

yet the sites of enrichmen t are the same, and in fact the correlation coefﬁcient (r)

between the two datase ts is 0.965 (n ¼ 55). Fortunately for prospecting, this strong

relationship between patterns derived from the analysis of ash and those from the

analysis of dried tissue are typically very similar for most elements, attesting to the

robustness of the biogeochemical method.

Much of the advice given by Kovalevsky (1987) on ashing and analysis was based

upon his many years experience and tens of thousands of samples, and remains sound

and well worth heeding. However, it is 20 years since his last book in English was

published, during which time considerable advances in our understanding have been

made as multi-element analytical technology has surged forward. In fact Kovalevsky

(1987) states that

it should be appreciated that analytical methods are improving rapidly y analyses of

unashed plant material are at present used to a small extent due to insufﬁcient sensitivity

for determining background and minimum anomalous contents of most indicator elements.

The exploration biogeochemist now has available the technology to determine

most elements in dry tissue, and only a few ultra-trace elements still need to be

determined on ash.

Pine bark sample

Concentration

Cr (dry) ppm Cr (Dry equiv. from ash) ppm

Fig. 6-10. Chromium in dry lodgepole pine bark (Pinus contorta) compared to bark ash

normalized to dry weight.

173Biogeochemistry in Mineral Exploration

A SPECIAL CASE: VEGETATION FROM SITES NEAR SMELTERS OR OTHER SITES

OF POINT-SOURCE METAL EMISSIONS

As already stressed, when using ashed plant tissues for biogeochemical explora-

tion it is important that, for meaningful interpretation of survey data, there are fairly

consistent losses of each element. Near mine sites, smelters and other point-source

sites of metal emissions, when compared to areas remote from such locations, these

losses on ignition are greater. This may be because during roasting of ores the

relatively volatile fractions dominate the emissions from the stack and settle on the

ground and vegetation. During reduction to ash of plant tissues from such an en-

vironment, the surface particulates that may be complexed as relatively volatile

compounds are released more readily than elements absorbed through root systems

(Dunn et al., 2002).

Two sites were studied near the Giant Mine, locat ed in an area of discontinuous

permafrost, in Yellowknife, Northwest Territories, Canada. From 1948 until 2004

when it was decommissioned, the Giant Mine produced 200,000 kg of gold. The

refractory ore, containing arsenopyrite, was mined from underground and roasted to

facilitate the recovery of gold. Over 50 years of operation resulted in the accumu-

lation of more than 265,000 tons of arsenic trioxide-bearing dust from the roasting

process which was stored in underground ch ambers (Thompson and Schultz, 2001).

These authors note that the As

dust included the following.



Arsenic: 36–67%;



Antimony: 0.30–2.13%;



Iron: 0.78–2.62%;



Gold: 2–80 (ppm), averaging 15 ppm Au.

Dr. D. Kerr at the Geological Survey of Canada (GSC) initiated a project to

assess the effects of the mine operations on the local environment and bulk samples

were collected from two sites in accord with established protocols (Dunn et al., 2002).

Previously, Dave Nickerson (personal communication) had established the extent of

As and Au enrichments down-wind from the mine (Fig. 6-11).

Two sites for the GSC study, both underlain by metavolcanic rocks, were selected

for the collection of bulk samples of vegetation. Site A was located 1.5 km nor th of

the Giant Mine mill north of Yellowknife and Site B was approximately 250 m north

of the mill. At each site collections comprised two of the most common plant species

that occur in the boreal forests – black spruce (Picea mariana) and Labrador tea

(Ledum groenlandicum). Each collection consisted of 1 kg of outer bark scales from

several black spruce trees and 1 kg of Labrador tea stems from several shrubs. These

were sent to Activation Laboratories Ltd. (Ancaster, ON, Canada) for sampl e prep-

aration, processing and analysis for 35 elements by INAA of dried and ashed tissue,

washed and unwashed. The analytical scheme is summarized in Table 6-VII.

174

Sample Preparation and Decomposition

Gold in Black Spruce Outer Bark

y = 1113.2x

-0.684

500

1000

1500

2000

2500

3000

0642 8 10 12 14 16 18

0642 8 10 12 14 16 18 0 642 8 10 12 14 16 18

Distance (km) from Giant Mine

Au ppb in ash

500

1000

1500

2000

2500

3000

0642 8 10 12 14 16 18

Distance (km) from Giant Mine

Au ppb in ash

Gold in Labrador Tea Stems

y = 1056.4x

-0.8163

Arsenic in Black Spruce Outer Bark

y =1050.5x

-0.4384

200

400

600

800

1000

1200

1400

1600

1800

2000

Distance (km) from Giant Mine

As ppm in ash

200

400

600

800

1000

1200

1400

1600

1800

2000

As ppm in ash

Arsenic in Labrador Tea Stems

y =762.2x

-0.6373

Distance (km) from Giant Mine

Fig. 6-11. Gold and As in spruce bark and Labrador tea stems. Changes in content with increasing distance from Giant Gold Mine and

mill.

175Biogeochemistry in Mineral Exploration

The data obtained from these samples provided insight into elemental losses

during sampl e washing and ashing, from sites affected by more than half a century of

mining. Among the observations, some trends were apparent with respect to the

principal elements associated with the As

. These are summarized in Table 6-VIII .

From the data in Table 6-VIII there are several features of note.



There is considerable enrichment of Au, As, Fe and Sb in all samples.



As would be expected, samples from close to the mine yielded higher concentra-

tions than those from a distance of 1500 m.



Unwashed samples yielded greater losses of Au, As and Fe from ashed tissue near

the mine than from a distance of 1500 m, implying that surface particulates derived

from the mine are relatively volatile.



Washed samples showed a similar trend of more As loss from ashed tissue near the

mine than from the more distant site; however, Au losses were similar at both sites.

In summarizing the Giant Mine study it was noted that during ashing Hg vol-

atilizes completely and some 80–90% of Br is lost. For many elements the amount

of loss is within the general range of 20–40%, with a few elements (Ba, Ca, Rb, Sr

TABLE 6-VII

Protocols for preparation and analysis of samples by INAA (Giant Mine)

Step 1

(subdivided)

Step 2 Step 3 Step 4 INAA (n ¼ 5

sub-samples of each

prepared sample)

Site A (1.5 km north of mill)

Black spruce bark For washing Milled Direct analysis 15 g pellets (dry)

Milled Reduced to ash 1 g ash

No washing Milled Direct analysis 15 g pellets (dry)

Milled Reduced to ash 1 g ash

Labrador tea stems For washing Milled Direct analysis 15 g pellets (dry)

Milled Reduced to ash 1 g ash

No washing Milled Direct analysis 15 g pellets (dry)

Milled Reduced to ash 1 g ash

Site B (~250 m north of mill)

Black spruce bark For washing Milled Direct analysis 15 g pellets (dry)

Milled Reduced to ash 1 g ash

No washing Milled Direct analysis 15 g pellets (dry)

Milled Reduced to ash 1 g ash

Labrador tea stems For washing Milled Direct analysis 15 g pellets (dry)

Milled Reduced to ash 1 g ash

No washing Milled Direct analysis 15 g pellets (dry)

Milled Reduced to ash 1 g ash

176 Sample Preparation and Decomposition

and Zn) indicating apparent slight gains that are may be attributable to some im-

precision related to instrumental calibration of the two methods of INAA – dry

pellets and ash.

The implications for biogeochemical exploration are that close to sites of mining

activities, especially where ore-roasting or smelting has taken place, the natural

biogeochemical signature from a mineral deposit may not be discernible from the

superimposed imprint of an anthropogenic signature. Consequently, the signature

from any undiscovered mineral deposit may be masked. In the case of the Giant

Mine, no biogeochemical study was undertaken prior to mining operations, and so

the extent of any naturally occurring anomalous concentrations of metals is un-

known. A survey area should be widely reviewed for past and present activities

before a decision is made to undertake a biogeochemical exploration survey. It might

be possible to devise a procedure for separating out the natural signature, but this is

rarely a simple and reliable exercise. Procedures such as normalizing biogeochemical

data to ‘conservative’ elements, as can be done wi th lithogeochemical data, are in-

variably too impr ecise, because many of the lithogeochemically conservative ele-

ments can be absorbed to varying degrees by certain plant tissues. It is not sufﬁcient

to invoke, for example, any of the high ﬁeld strength elements (HFSE – e.g., Hf, Nb,

REE, Ta, Ti, Zr) against which to normalize data for dust contamination, because

many studies have indicated that these elements can on occasion be absorbed and

even utilized by some plants (notably REE in ferns). They are not, therefore,

TABLE 6-VIII

Metal losses from black spruce outer bark (Picea mariana) at two sites near the Giant Au mine,

NWT, Canada. Each value is the mean of ﬁve determinations by INAA

250 m N of mine 1500 m N of mine

Mean

(dry)

Ash recalc.

to dry

% loss Mean

(dry)

Ash recalc.

to dry

% loss

Unwashed

Ash yield (%) 6.40% 5.5

Au (ppb) 210 122 42 97 68 30

As (ppm) 644 294 54 122 77 37

Fe (%) 0.366 0.21 42 0.252 0.17 34

Sb (ppm) 81 53 35 16.8 10.5 38

Washed

Ash yield (%) 4.80% 5%

Au (ppb) 111 79 28.7 80 56 29.8

As (ppm) 412 235 43 126 83 33.9

Fe (%) 0.193 0.15 23 0.217 0.15 32

Sb (ppm) 71 46 36 17.2 9.7 43.5

177Biogeochemistry in Mineral Exploration

‘immobile’ in the lithogeochemical sense of the word. That said, an examination of

the HFSE content can provide an indication of whether a biogeochemical dataset

might be contaminated from dust particulates, but this is us ually difﬁcult to quantify.

The study emphasizes the importance of washing samples from close to poin t-

source areas of metal emissions, always bearing in mind the caveats expressed in the

section on ‘washing’ in that some of the natural signature may be removed. It shows,

too, that the relatively high loss on ignition of elements associated with such point-

source emissions suggests that they are present in the environment as the volatile

phases. During ore-roasting the volatile phases dominate the particulates that even-

tually precipitate on the ground and plant surfaces. Since they are mobile, they are

the ﬁrst components to be absorbed and/or adsorbed by plant tissues and subse-

quently they are the ﬁrst to be released during ashing.

WET DECOMPOSITION

Hall (1995) provided a detailed and comprehensive account of methods of wet

chemical decomposition and the remainder of this chapter leans heavily upon the

information that she supplied, supplemented with data from personal observations

and data compilations. Additional information has been extracted from Gorsuch

(1970), Bock (1979), Fletche r (1981) and Batley (1989). Hall (1995) notes that, fol-

lowing ashing, a sample is normally digested in a dilute acid such as HCl or HNO

or, more vigorously, in aqua regia. The dissolution of silicate material (which is a

structural component of some plants such as horsetails (Equisetum spp.), bamboo

and grasses) requires evaporation with HF whereupon SiF

volatilizes. Hot aqua

regia is an effective solvent for numerous sulphides, arsenides, selenides, sulphosalts,

simple oxides and their hydrates, Ca phosphates and most sulphates (except barite).

Evaporation to dryness with HCl converts salts to chlorides, ready for ﬁnal sol-

ubilization in dilute (0.5–1 M) mineral acid which is compatible with the analytical

technique. In the previous sectio n it was noted that the process of ashing changes the

forms of the elements and, rather than facilitate subsequent solubilization, it can

hinder it (e.g., Pd). This explains why some element determinations by a ‘total’

method (XRF or INAA) may yield higher results than by an acid digestion from

which some elements may be only partiall y dissolved.

If an exploration pro gramme requires total dissolution of all elements, a stronger

oxidant can assist in this process. Those most commonly used are high purity hy-

drogen peroxide (H

) or perchloric acid (HClO

). Hydroﬂuoric acid (HF) is re-

quired for bringing the silica component of plant tissues into solution, but care must

also be taken to avoid losses by volatili zation when evaporating off acids and chem-

ical species such as SiF

. The valency in which the element is present plays a role; for

example, the halides of As (III) are much more volatile than those of As (V) and thus

pre-oxidation in the acid mixture prior to evaporation is necessary. Low recoveries

178

Sample Preparation and Decomposition

may sometimes be explained by adsorption or coprecipitation of the analyte on to the

solid material remaining after digestion.

For non-exploration oriented decomposition of organic matter, nitric acid is the

most widely used as it reacts readily with both aromatic and aliphatic groups. The

normal acid boils at about 120 1C; a factor which assists in its removal after ox-

idation but which correspondingly limits its effectiveness. Thus, nitric acid is used in

the presence of sulphuric acid which partially degrades the more resistant material, or

with perchloric acid (boiling point 2031C) which continues the oxidation after nitric

acid has been removed. Although sulphuric has similar properties to perchloric acid,

it has not found such widespread application probably due to the interfer ence

effects created by SO

in AAS and to the low solubilities of alkaline earth and Pb

sulphates. Care must be exercised in using perchloric acid as it is such a powerful

oxidising agent and explosions have occurred. Arafat and Glooschenko (1981)

reported complete recovery for As, Al, Fe, Zn, Cr and Cu in various plant

tissues using 0.5 g samples and a mixture of 10 ml of HNO

, 5 ml of H

and 2 ml

of 70% H

. Haas and Krivan (1984) successfully determined As, Bi, Cd, Cr, Hg,

Pb, Sb, Se and Tl in lichens, pine needles and grasses using a mixture of HNO

, HCl

and H

Adeloju (1989) compared the efﬁcacy of wet digestion and dry ashing methods for

voltammetric analysis of leaves. He found the direct dry ashing method without an

ashing aid proved to be the most suitable for the determination of Bi, Cd, Co, Cu,

Pb, Ni, V and Zn, but As and Se required the incorporation of an ashing aid or wet

digestion with HNO

and H

MICROWAVE DIGESTION

It needs only a brief review of the literature on methods to decompose plant

tissues to realise that the chemistry involving wet decomposition can be complex and

fraught with potential problems. Pressure digestion in closed or semi-closed vessels

placed in industrial quality microwave ovens lowers the risks of contamination

from the atmosphere and of volatilization losses in addition to decreasing the

amount of reagents necessary (hence lower blank levels and better detection limits).

Decomposition times are reduced to minutes as reactions carried out at elevated

pressures and temperatures require considerably less time to reach completion.

Moreover, substances that ordinarily would not be decomposed by these acids at

their normal boili ng points react at elevated temperature and pressure. Some of the

pioneering work in this ﬁeld was reported by Kingston and Jassie (1986), Mat-

usiewicz and Sturgeon (1989), Matusiewicz et al. (1989), Bettinelli et al. (1989)

and Sah and Miller (1992). A brief review and details of procedures are given in

Miller (1998).

Despite the encouraging results of these early workers and subsequent research-

ers, especially using PTF E (Teﬂon) ‘bombs’, microwave oven s are still not in

179

Biogeochemistry in Mineral Exploration

common use for digesting exploration-oriented biogeochemical samples. The pri-

mary reasons seem to be



Cost: the pressure resistant vessel s are expensive.



The limited capacity of most ovens compared to the many beakers which can be

placed on a hot plate. This becomes a critical consideration for time-sensitive

exploration activities, since a quick turn-around from sample collection (com-

monly many hundreds of samples) to results for follow-up work can be of par-

amount importance.

Microwave decomposition has distinct advantages for the dissolution of dry

vegetation where maintaining both a minimum sample-to-reagent volume ratio and

contamination level to optimize detection capability is particularly advantageous

when elements of interest are present at low levels.

Many tests have been conducted on the decomposition of dry plant tissues, re-

sulting in voluminous literature in chemical journals and textbooks on the preferred

techniques to be employed in research laboratories. Now that modern analytical

instrumentation is capable of detecting very small amounts of many elements in dry

plant tissue from a single digestion, tests on ashed tissue are largely, but not entirely,

of academic interest. There remains the suite of elements discussed above, present at

sufﬁciently low levels that reduction to ash is required to determine ‘relief’ in the

geochemical data. These elem ents include Be, Bi, Ga, Ge, In, Re, Te, Tl, U, V, W,

many of the REE and all of the platinum-group elements.

SELECTIVE LEACHING

In order to pursue the use of biogeochemical sampling in mineral exploration to the

fullest extent, we must attempt to understand the mechanisms of accumulation

and translocation taking place in the species under investigation y We should be de-

signing experiments to answer such questions as: (1) What form(s) of element X is

preferentially taken up by this species? (2) Where does the element accumulate in the

plant? (3) How does the element exist in the plant, in which forms in which organs?

(Hall, 1995)

Examples of some of the early attempts to address these questions for issue s affecting

biogeochemical explora tion include the following.



Girling and Peterson (1978) examined the effect of the form of Au on its uptake

in a range of species (e.g., Phacelia sericea, Hordeum vulgare) and found the

cyanide complex to be absorbed to a much greater degree than the chloride

or thios ulphate form. Autoradiography was employed to identify the location of

Au accumulation within the plant and the study established the acropetal tendency

(movement toward growing tips) of Au with enrichments occurrin g toward

leaf tips.

180

Sample Preparation and Decomposition



A study using electrophoresis examined the inﬂuence of humates on the Au uptake

by perennial ryegrass (Jones and Peterson, 1989). Both complexed and ionic forms

of Au were found to exist in the humic acid solution and the uptake of this Au

depended upon (1) humic acid concentration; (2) pH; and (3) size fraction (e.g.,

more uptake from unﬁltered solutions). Far less Au was absorbed from the humic

solutions than from solutions of AuCl





Nickel in hyperaccumulator plants is bound to citric, malic and malonic acids and

their derivatives. This association was found by Brooks (1983) in a study of Ni in

Alyssum serpyllifolium from the Iberian Peninsula; more than half the Ni was

soluble in water and dilute acid, indicating its presence also as polar complexes.

There remains a gargantuan task that is beyond the scope of this book to discuss

the many answers provided, yet many questions still remaining, since Hall’s sugges-

tions were made. Progress is steadily being made at many institutions, and key results

are published in many professional journals (e.g., Biogeochemistry; Chemosphere;

Science of the Total Environment, to mention but a few). Adriano (2001) provides a

comprehensive review of trace elements in terrestrial environments, and the publi-

cation ‘Biogeochemistry’ (Schlesinger, 2005) delves deeply into the biogeochemistry

of organic compounds, and every two years since 1990, a wealth of research is

summarized as extended abstracts in the Proceedings volumes of the International

Conference on the Biogeochemistry of Trace Elements (ICOBTE). The web page

www.isteb.org/default.asp?id=14&lid=2 provides details of these conferences and

their respective multitude of abstracts, some of which are directly relevant to

biogeochemical exploration, but many deal with the ﬁne points of biogeochemical

processes in many environments – including plants, animals, soils, waters, atmos-

pheric combustion and bio-solids in general.

A compari son of vegetation samples leached by aqua regia (AR), pH 7-controlled

ammonium acetate (AAc7) and distilled water provides further insight to the degree

by which elements are bound within plant tissues (Dunn et al., 2006a). Whereas the

amount of each element extracted by the aqua regia leach is for most elements, as

would be expected, considerably greater than from the weak leaches, the relative

amounts of each element extracted remain much the same, such that the patterns are

consistent. Again, it is the relative concentrations that are of interest. In the case of

Ni (Fig. 6-12) the patterns are almost identical.

Highly soluble elements, such as Rb are mostly extracted by just a simple water

leach. Figure 6-13 shows that more than 80% Rb is removed by the water leach.

Another selective leach test compared three weak leaches – distilled water, dilute

[2%] nitric acid and pH 7-controlled ammonium acetate. Two types of tissue from

the same location were tested – Douglas-ﬁr needles, and lodgepole pine outer bark.

For some elements (e.g., Rb – Fig. 6-14) the amount extracted by each leachat e was

virtually identical.

In general there is little difference in the amount of each element extra cted by a

water leach and a dilute (2%) nitric acid leach. The AAc7 leach extracts higher

181

Biogeochemistry in Mineral Exploration