Dunn Colin E. Biogeochemistry in Mineral Exploration

Подождите немного. Документ загружается.

As a second example, Table 6-V shows the effect of different grinding times, hence

ﬁneness of particle size, on the analysis of Acacia stems from samples collected in

central Africa.

In Table 6-V gold shows lack of hom ogeneity at the sub-ppb levels. This is typical

for one-gram samples of milled tissue, although gold is more evenly distributed in

tissues of some species than in others. Gold has an acropetal tendency in leaves – i.e.,

it tends to migrate toward leaf tips. Consequently, the ﬁne fraction of mil led leaves is

TABLE 6-V

Stems of Acacia milled for different durations in a grinder with rotary cutting blades. One-

gram samples digested in HNO

then aqua regia with ICP-MS ﬁnish

20 s grind 40 s grind 60 s grind

Coarse Medium Sieved ﬁnes Coarse residue Fine+coarse

Ag (ppb) 2 4 4 3 4

Au (ppb) o 0.2 0.6 0.3 0.2 0.7

B (ppm) 9 11 19 7 14

Ba (ppm) 28 29 62 23 44

Ca (%) 0.99 1.05 2.2 0.9 1.47

Ce (ppm) 0.03 0.02 0.07 0.03 0.05

Co (ppm) 0.01 0.01 0.02 0.01 0.03

Cr (ppm) 1.37 1.55 1.92 1.52 1.68

Cu (ppm) 3.98 4.78 6.57 3.88 5.1

Fe (%) 0.002 0.002 0.004 0.002 0.003

K (%) 0.99 1.12 1.57 1.02 1.33

La (ppm) 0.06 0.04 0.11 0.06 0.09

Li (ppm) 0.02 0.01 0.04 0.03 0.06

Mg (%) 0.089 0.106 0.223 0.087 0.139

Mn (ppm) 11 12 20 13 13

Mo (ppm) 0.47 0.33 0.66 0.43 0.6

Na (%) 0.007 0.008 0.016 0.01 0.01

Nb (ppm) o0.001 0.001 0.009 0.002 0.006

Ni (ppm) 1 1.3 1.8 0.9 1.6

P (%) 0.094 0.106 0.185 0.084 0.132

Pb (ppm) 0.02 0.07 0.02 0.08 0.07

Rb (ppm) 12.4 14.3 19.3 12.1 15.8

S (%) 0.1 0.1 0.17 0.08 0.15

Sb (ppm) 0.05 0.08 0.12 o0.02 0.12

Se (ppm) 0.5 0.4 0.6 0.3 0.6

Sr (ppm) 209 219 443 197 271

Y (ppm) 0.01 0.006 0.017 0.007 0.015

Zn (ppm) 10.5 11.6 22.3 12.4 16

162 Sample Preparation and Decomposition

likely to contain the relatively soft and fragile leaf tips, whereas the coarser residue is

made up of the harder vein tissues which typically contain less gold. The result is that

higher gold values tend to be obtained from ﬁne fractions. This relationship holds,

too, for other tissue types and other elements. Pickering et al. (2000, 2003) used a

Synchrotron to study the distribution and chemical speciation of Se in leaves of the

Se hyperaccumulator plant Astragalus bisulcatus (two-grooved poison or milk vetch).

This fundamental investigation showed, among other observations, that selenate is

located almost exclusively in the soft tissue of mature leaves, rather than the veins of

the leaves (Fig. 6-6).

SAMPLE DECOMPOSITION

Once samples have been prepared for analysis, the next consideration becomes

whether to analyse dry tissue directly or ﬁrst reduce tissues to ash. If low and ac-

ceptable detection levels can be obtained from the direct analysis of dry tissue, and

the selected method (e.g., ICP-MS, INAA) can generate high-quality data (good

precision and accuracy) for all elem ents of interest, then reducing samples to ash is an

unnecessary step. However, plant tissues contain only minute traces of many ele-

ments, and so from the direct analysis of dry tissue by such methods as AAS and

ICP-ES many values, particularly some of those of relev ance to mineral exploration,

are recorded as ‘below detection’ and ashing may be warranted. The various an-

alytical methods and their merits are discussed in the next chapter, in which dis-

cussions centre on the sophisticated analytical instruments that are becoming

increasingly available for determining ‘ultra-trace’ levels of almost all elements.

Consequently, the multi-element analysis of dry tissues is steadily gaining a foothold

as the method of choice.

Historically, no single analytical instrument was sufﬁciently sensitive to provide

data for a comprehensive range of elemen ts of relev ance to mineral exploration. The

non-destructive methods of INAA and XRF require no sample decomposition and

provide excellent data for many elements. However, the main limitations for INAA

are (a) the requirement of access to a nuclear reactor; (b) the inability to provide data

for certain elements; and (c) the need for special irradiations to obtain data for

certain elements. Included in the list of elements that are problematic or not pos sible

to determine by INAA are Be, Bi, Cd, Cl , Cu, Ga, Ge, I, In, Li, Mg, Mn, P, Pb, Pd,

Pt, Re, S, Te, Ti, Tl and V. Other elements have detection levels that are too high for

most vegetation samples – e.g., Ag, Ni and Sr. Yet others may require substantial

corrections for ﬁssion products, so if a sample has a high U content, then it may be

impossible to report values for Ba, Mo, Te and some of the rare-earth elements

(REE) (Hoffman, 1992). Furthermore, high level s of Na or Br in halophyte species

necessitate additional corrections and therefore higher detection levels. Similarly,

XRF has its limitations. These non-destructive methods are discussed in the next

163

Biogeochemistry in Mineral Exploration

Fig. 6-6. Selenium concentration images of different parts of Astragalus bisulcatus. From left to right: optical micrograph; effective tissue

thickness; concentration of organoselenium; concentration of selenate. Top to bottom: mature, intermediate and young leaves and roots. The

mature leaves, young leaves and roots were taken from the same plant, the mature leaves from the lowest shoot and the young leaves from the

highest shoot of the same plant branch. Reproduced with permission from National Academy of Sciences, USA, Ref. www.pnas.org,

Pickering et al. (2000).

164 Sample Preparation and Decomposition

chapter, and the remainder of this section deals with procedures required to decom-

pose plant tissues prior to their introduction into analytical instruments.

A number of the elements listed above can be determined by other techniques

(e.g., ICP-ES or AAS), but without careful pre-concentration procedures multi-el-

ement scans generate data that are mostly below the level of detection. Many an-

alytical methods have been developed that can address these problems, but they

involve costs that are too high for the implementation of a biogeochemical explo-

ration programme for minerals.

A solution to the problem of plant tissues having very low levels of many elements

of exploration interest is to pre-concentrate them. This can be done by ﬁrst reducing

samples to ash by controlled ignition to drive off organic compounds, so that only

the inorganic constituents remain for analysis. This ‘ashing’, as it will be referred to

here, substantially reduces the volume of the plant material so that a sample that has,

for example 1 ppm Ni in twig tissue, will have 50 ppm Ni in ash, because the ash yield

of the twig is only 2% (i.e., a 50-fold concentration). From an exploration point of

view, it is more comforting to see a value of 50 ppb Au (in ash) than 1 ppb Au (in dry

material). In practical exploration terms, it is more appeali ng to an exploration

manager to see an ‘anomaly’ of 50 ppb Au than the same ‘anomaly’ at 1 ppb Au in

dry-weight equivalent.

Prior to discussing the merits of analysis of ‘dry’ versus ‘ash’, it is worth noting

that modern analytical equipment (ICP-MS) can now provide remarkably good

multi-element data from the analysis of just 1 g of dry tissue, and so the need to ash

samples is dimi nishing. However, for the reasons discussed above, results of past

surveys were quite commonly reported as concentrations in ash and, since the ex-

pensive state-of-the-art instrumentation will not be universally available for some

years to come, the ashing process warrants discussion.

As is noted repeatedly in this book, in biogeochemical exploration, the absolute

concentrations of elements are of lesser importance than their distribution patterns.

Provided there is a direct relationship between the element content of dry tissue and

the element co ntent of ashed tissue, it makes no difference to the spatial distribution

of anomalous zones of concentrations, and the data therefore provide the same focus

for deﬁning exploration targets.

DRY ASHING

Element volatilization

The most important consideration in dry ashing is the potential loss of some

elements due to volatilization. For Hg, Br, Cl, F and I only a small percentage, if any,

remains in ash after controlled ignition at 475 1C. Although there is partial loss of

many elements during ashing the critical factor for exploration is that, for a given

plant species, the amount of an element lost remains constant. The question is, if, for

165

Biogeochemistry in Mineral Exploration

example, twigs from species ‘A’ are found to lose 20% of their As at one sample

station, do they also lose 20% of their As at another? Provided this loss remains

constant, then the relative concentrations remain the same and it becomes possible to

generate realistic plots of As distribution patterns. These plots of relative concen-

trations are critical to interpreting biogeochemical exploration data, because they

may deﬁne stratigraphic, lithological and structural trends, in addition to zones of

mineralization.

Volatilization depends on the chemical form of the element as well as the com-

position of the matrix (i.e., the type of plant tissue). Lead remains in ash if it is present

as its sulphate, nitrate or oxide, but significant loss occurs if it is present as the

chloride (Hall, 1995). Furthermore, during heating to decompose vegetation tissues an

element may react with an organic or inorganic constituent to produce a volatile

chemical species. Various ashing aids can be added to control chemical reactions. The

complexity of potential chemical reactions during the heating phase is enormous and

chemists can point to many combinations of physico-chemical parameters that could

take place to render the ashing process an untenable option. Some of these are

summarized in Hall et al. (1991) and Hall (1995). Fortunately, as will be shown in the

following pages, the actual losses of elements are surprisingly consistent and quan-

tiﬁable (for most species), conﬁrming that ashing is a viable option. Consistency in the

approach to ashing is, like so many situations in science, the key to obtaining mean-

ingful results that the exploration biogeochemist can plot to reveal structure in the

data that can be attributed to underlying geological conditions.

In order to minimize and standardize elemental losses, the ashing parameters need

to be closely controlled. Ashing should take place by slowly ramping up the tem-

perature of a mufﬂe furnace (or kiln), dedicated to vegetation, to abou t 475 1C. It

rarely makes any difference if a temperature of 500 1C is selected, but it is as well to

be consistent, and 475 1C is about the lowest temperature at which vegetation will be

reduced to ash without a long period of charring. If the temperature is maintained at

450 1C, most types of plant tissue go through a long phase as charcoal before all of

the organic components are released. Increasing the temperature by just 25 1Cto

475 1C greatly accelerates the ashing process.

Borosilicate glass beakers are suitable for ashing vegetation samples; although

after repeated usage the glass be comes etched from reaction with the hot ash and a

few ppm B is transferred from the borosilicate to the ashed samples. However, since

B is commonly present in plant ash in hundreds of ppm, the slight addition of boron

from the beakers does not usually significantly affect natural distribution patterns of

boron.

Quartz vessels are reported to be unsatisfactory for dry-ashing of biological ma-

terials, because Zn (and perhaps other metals) has been found to penetrate the

surface and cannot be readily extracted with acid (Spitzy and Dosudil, 1962). An

alternative is to use large aluminium trays, since they are suitable for ashing mod-

erately large samples (50–100 g) of dry tissue. They can be rinsed between batches of

samples and they can be re-used for many years. Dry tissues of plants commonly

166

Sample Preparation and Decomposition

used in biogeochemical exploration contain tens to hundreds ppm Al. On reduction

to ash, these co ncentrations are magniﬁed 30- to 100-fold, depending on the type of

tissue, so that the content of ash is frequently in the range of 0.1–1% Al. Thus, as in

the case of B, the potential addition to samples of a few or tens ppm Al from contact

with the Al trays is not significant for most purposes. Of course, other vessels should

be used if low-level Al determinations are required.

Temperature control can be particu larly important for a number of reasons, not

least of which is that aluminium trays melt at around 550 1C. Also, at around this

temperature, the carbonate component of the ash may start to dissociate releasing

(and possibly some higher temperature phases of metals contained within the

ash), especially if the ash contains a magnesite component. Kovalevsky (1987) notes

that the chemical composition of plant ash approximates that of dolomi tized car-

bonate rocks and so plant ash can be treated as if it is a carbonate.

Tests on dry black spruce samples show that between 100 and 475 1C there is a

weight loss of approximately 98% for twigs and outer bark scales and 97% for

needles; between 475 and 700 1C there is a further loss from the 475 1C ash weight of

15–20%. As the temperature increase continues, additional weight losses occur as

some elemental oxide bonds break down (e.g., PdO), and by the time 900 1Cis

reached all of the carbonate has dissociated releasing CO

, and for some tissues a

fused pellet has formed. Elements that show significantly higher concentrations in

ash heated to 900 1C, mostly due to breaking of oxide bonds, include Al, Au, Ba, Eu,

Ga, Ge, Li, Na, Pb, Pd, Yb and W. If ceramic crucibles are used for high-temper-

ature ashing, some tissues are sufﬁciently reactive at 900 1C for the remaining ash to

further complic ate the situation by fusing with the crucible itself, introducing con-

taminants from the crucible.

The door to the mufﬂe furnace or kiln should remain closed throughout the

ashing process, because if it is opened a rush of fresh oxygen can cause partially

decomposed samples to ignite. This sudden and localized increa se in heat can cause

differential losses of elements among the samples (because of differing temperatures

of element dissociation) with the result that analytical results cannot be effectively

compared. Furthermore, ﬂash ﬁres in the furnace can cause partial or even complete

melting of some aluminium trays. In the former Soviet Union large scale biogeo-

chemical exploration programmes sometimes used open ﬁres to rapidly reduce sam-

ples to ash in the ﬁeld, claiming that they were able to treat 200–1000 samples per

shift and that the temperature of ashing ranged from 400 to 700 1C with ashing time

ranging from 20 min to 4 h (Kovalevsky, 1987). Kovalevsky noted that

the percentage losses of volatile elements for background and mineralized samples under

standard ashing conditions are similar and have no effect on the main results of ex-

ploration: i.e., the shape, degree of contrast, and intensity of biogeochemical anomalies

and haloes.

Given what is now known about element volatilization at increasing temperatures,

these conditions, although expeditious in the ﬁeld, are rather too semi-quantitative for

167

Biogeochemistry in Mineral Exploration

some elements and probably too much valuable multi-element information is lost by

not closely controlling the ashing conditions.

Nonaka et al. (1981) determined losses of elements during dry ashing of standard

reference material (SRM ) orchard leaves and bamboo leaves, at stepwise tempe rature

increases from 200 to 800 1C. Their principal ﬁndings were [with added comments in

square parentheses] as follows.



Loss of Hg began at 110 1C and increased steadily thereafter [studies using pine

twigs yielded similar results, see Table 6-VI].



[Partial] losses of As and Sb occurred at 200 1C.



There was a sha rp loss of Br, Cl, Cr and Se at 200 1C and again above 500 1C [this

provides further strength to the argument that 475 1C should be a preferred tem-

perature for ashing].



There were no losses of alkaline earths, REE, or Al, Co, Fe, Mn, V and Zn.



Losses of alkali elements were dependent upon the crucible material (e.g., severe in

a silica dish) and sample species.

Once samples have been reduced to ash, care should be taken not to expose them

to highly volatile elements such as Br, Cl, F, Hg, I and S, because ash is a good

absorbent of gases (Kovalevsky, 1987). This phenomenon explains why some multi-

element ICP-MS determinat ions of ash occasionally yield a few ppb Hg.

On occasion a perceived ashing loss can actually be an inadequate digestion of the

ash, or the formation of poorly soluble oxide bonds that require high temperatures to

dissociate. Also, apparent ‘ele ment losses’ can be the result of element incorporation

into the insoluble residue, which is dominated by silica in some plant species (Hoenig

and de Borger, 1983), and the HCl and HNO

normally used to digest the ash do not

dissolve silica nor the elements associated with it.

TABLE 6-VI

Loss of Hg from pine twigs (Pinus banksiana)

on heating for 24 h at progressively higher tem-

peratures (material comprising control V6).

Analysis by ICP-MS after aqua regia digestion

Temperature

(1C)

Hg (ppb)

Air-dried 40

70 40

80 40

110 30

150 30

200 o3

168 Sample Preparation and Decomposition

Although the ramping rate does not appear to be critical, an increase in tem-

perature of about 100 1 C per hour has proved to be appropriate and, depending on

the amount and nature of the material to be ashed, the furnace is held at 475 1C for

about 24 h. For small samples of leaf tissue 12 h is sufﬁcient. For cones or large

chunks of wood the ashing may take up to 48 h.

These examples serve to demonstrate the complexity of problems that can exist

during reduction of plant tissues to ash at various temperatures. However, to re-

iterate a point made earlier ‘in order to minimize and standardize elemental losses,

the ashing parameters need to be closely controlled’.

Dry ashing – the realities

The last section outlined many of the potential problems that can arise from

ashing samples. They appear sufﬁcient in number and complexity that the explo-

rationist might dismiss the idea of ashing. In hindsight, it is fortuitous that ashing

was once the only feasible approach for determining concentrations of many ele-

ments, until the commercialization in the 1980s of multi-element INAA of pelletized

dry tissues and, more importantly, the introduction by commercial laboratories over

the past 10 years or so of multi-element analysis of small samples of dry tissue by

ICP-MS. The historic necessity ‘to ash’ provided a large amount of information, and

showed that, pr ovided ash ing takes place under controlled conditions, the biogeo-

chemical signatures are remarkably robust. Modern comparisons of ‘ash’ versus ‘dry’

analyses of the same samples permit evaluation of concentrations (and attendant

losses) of a wide array of elements.

Figures 6-7, 6-8 and 6-9 provide some examples of various Acacia tissues from

Western Australia that were analysed both as 50 g samples reduced to ash (from which

a 0.25 g portion was taken for analysis) and a 1 g sample of milled dry tissue. Both were

digested in aqua regia with an ICP-MS ﬁnish. The data from analysis of the ash have

been normalized to a dry weight basis by taking the concentration in ash, dividing by

100 and multiplying by the ash yield. These data are shown on the accompanying CD

as Table 6-ID permitting the user to examine the relationship of ash versus dry for all

51 elements that were determined. The digital table also permits sorting into relative

concentrations of different plant tissues. In the examples illustrated here, samples are

sorted from lowest to highest concentrations, with the average for the dataset (n ¼ 53)

plotted as single symbols on the right of each chart. They illustrate three features.



Figure 6-7 shows examples of elements that exhibit little or no loss from con trolled

ignition to 475 1C. The proﬁles are almost identical. Gold gives somewhat erratic

results, partly because it is not always all dissolved from dry tissue. This is dis-

cussed in the section on data quality.



Figure 6-8 shows examples of elements that partially volatilize during ashing.



Figure 6-9 shows elements for which data derived from the analysis of dry tissue

are too close to the detection limit for data structure to be determined. By reducing

169

Biogeochemistry in Mineral Exploration

Dry (diamond) v Ash (square)

135791113151719212325272931333537394143454749515355

Sequence #

Conc.

0.1

0.05

0.15

0.2

0.25

0.3

0.35

Conc.

Cu ppm

Cu_Norm ppm

As ppm

As_Norm ppm

Au ppb

Au_Norm ppb

Cd ppm

Cd_Norm ppm

Averages

135791113151719212325272931333537394143454749515355

Sequence #

Averages

135791113151719212325272931333537394143454749515355

Sequence #

Averages

135791113151719212325272931333537394143454749515355

Sequence #

Avera

Fig. 6-7. Examples of elements showing little or no loss on ignition. Comparison of element

concentrations determined by ICP-MS (aqua regia digestion) in dry Acacia tissues compared

to ashed portions of the same samples. Data normalized to dry weight basis (details on CD –

Table 6-ID).

170 Sample Preparation and Decomposition

S %

S_Norm %

0.02

0.04

0.06

0.08

0.1

0.12

0.14

0.16

Sb ppm

Sb_Norm ppm

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

Se ppm

Se_Norm ppm

0.05

0.1

0.15

0.2

0.25

0.3

Sn ppm

Sn_Norm ppm

Dry (diamond) v Ash (square)

0.05

0.1

0.15

0.2

0.26

0.3

135791113151719212325272931333537394143454749515355

Sequence #

Conc.

Conc.Conc.Conc.

Averages

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55

Sequence #

Averages

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55

Sequence #

Averages

1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55

Sequence #

Avera

Fig. 6-8. Examples of elements showing moderate to significant loss on ignition. Comparison

of element concentrations determined by ICP-MS (aqua reiga digestion) in dry Acacia tissues

compared to ashed portions of the same samples. Data normalized to dry weight basis (details

on CD – Table 6-ID).

171Biogeochemistry in Mineral Exploration