Dunn Colin E. Biogeochemistry in Mineral Exploration

Подождите немного. Документ загружается.

quantities of some elements but less of others. For each leach, the detection limit

varies. Figure 6-15 shows similar amounts of Cu extracted by the three leaches, but

with the AAc7 leach extracting a little more from the ﬁr needles than from the pine

bark. Comparative data for these leaches are presented for approximately 50 ele-

ments on the accompanying CD in the ‘Halogen Study’.

This observation demonstrates that for optimal metal extraction by weak leaches,

there is the need to select an appropriate leach for each individual plant species and

plant tissue. However, also of importance for biogeochemical exploration is the fact

that the relative amounts of each element extracted remain much the same – i.e., the

patterns of relative concentrations are very similar. This is evident from the plot of

zinc in the two tissue types by the three leaches (Fig. 6-16), with the AAc7 yielding

more Zn than the other two leachates, but with appreciably more extracted from the

bark than from the needles. Note, however, that the element proﬁles across the

1VE (AR) v H

O v AAc7

500

1000

1500

2000

12345678910

Sequence number

Concentration

Ni_AR ppb

Ni_H

O ppb

Ni_AAc7 ppb

Fig. 6-12. Nickel in dry ﬁr needles determined by ICP-MS after three types of leach – one

strong acid digestion (nitric acid/aqua regia) and two weak digestions (water and pH 7-

controlled ammonium acetate).

1VE (AR) v H

O v AAc7

2000

4000

6000

8000

10000

12000

12345678910

Sequence number

Concentration

Rb_AR ppb

Rb_H

O ppb

Rb_AAc7 ppb

Fig. 6-13. Rubidium in dry ﬁr needles determined by ICP-MS after three types of leach – one

strong acid digestion (nitric acid/aqua regia) and two weak digestions (water and pH 7-

controlled ammonium acetate).

182 Sample Preparation and Decomposition

sampling proﬁle remain almost identical, so any of the leaches would provide the

same patterns of element distributions, but different absolute concentrations.

For other elements, such as potassium, the AAc7 leach extracts proportionally far

less K from the needles than from the bark (Fig. 6-17).

Each element has its own characteristics, and there is a wealth of information that

can be extracted from the digital data comprising part of the ‘Halogen Study’ on the

CD that accompanies this book.

RUBIDIUM - H

O v 2%HNO

v AAc7

2000

4000

6000

8000

10000

12000

12345678910111213141516

Concentration

Rb_H

O ppb

Rb_2%HNO

ppb

Rb_AAc7 ppb

Dry Dou

las-fir needles Dry LP Pine bark

High Au intreetops

High Au

in bark

Fig. 6-14 . Rubidium extracted from ﬁr needles and pine bark by three weak leachates –

water, dilute nitric acid and ammonium acetate (pH 7). Determinations by ICP-MS. Note the

virtually identical amounts extracted by each dilute leach.

COPPER - H

O v 2%HNO

v AAc7

0.2

0.4

0.6

0.8

1.2

1.4

1.6

1.8

12345678910111213141516

Concentration

Cu_H

O ppm

Cu_2%HNO

ppm

Cu_AAc7 ppm

Dou

las-fir needles Dr

LP Pine bark

HighAu in treetops

High Au

in bark

Fig. 6-15. Copper in ﬁr needles (left) and pine bark (right) extracted by three weak leachates.

Determinations by ICP-MS.

183Biogeochemistry in Mineral Exploration

The results from these test s have an important bearing on the decision of whether

or not samples should be washed prior to drying and analysis. Clearly, intense

washing will remove a portion of elements contained within the plant structure. If

any washing seems necessary, only gentle rinsing should be performed.

ZINC- H

O v 2%HNO

v AAc7

12345678910111213141516

Concentration

Zn_H

O ppm

Zn_2%HNO

ppm

Zn_AAc7 ppm

Dry Douglas-fir needles Dry LP Pine bark

High Au in treetops

High Au

in bark

Fig. 6-16. Zinc in ﬁr needles and pine bark by three weak leachates. Determinations by ICP-

MS.

ZINC- H

O v 2%HNO

v AAc7

1000

2000

3000

4000

5000

6000

7000

8000

1 2 3 4 5 6 7 8 9 101112131415 16

Concentration

K_H

O ppm

K_2%HNO

ppm

K_AAc7 ppm

Dou

las-fir needles Dr

LPPine bark

High Au in treetops

High Au

in bark

Fig. 6-17. Potassium in ﬁr needles and pine bark extracted by three weak leachates. Deter-

minations by ICP-MS.

184 Sample Preparation and Decomposition

FUSIONS

In the exploration for kimberlites, some of the HFSE can be useful ‘pathﬁnder ’

elements (e.g., Nb, Ta). Some are only partially extracted from an aqua regia leach,

and a fusion with a ﬂux such as lithium metaborate is required to facilitate bringing

them into solution.

Hall (1995) states that the great efﬁciency of fusion co mpared to acid attack is due

to the effect of the high temperature (500–1100 1C). Heterogeneous reactions

taking place in the melt are of two types: acid–base and oxidation–reduction.

Alkaline ﬂux reagents include Na and K carbonate and bicarbonate, Na and K

hydroxide, and sodium tetraborate; acid ﬂuxes include Na and K hydrosulphate, Na

and K pyrosulphate, boron trioxide and hydroﬂuoride. Oxidative reagents comprise

largely Na

, KNO

and KClO

while carbonaceous substan ces such as ﬂour and

starch are added to ﬂux mixtures for a reducing action. The drawbacks of a fusion

are: (1) the potential addition of contaminan ts due to the high ﬂux:sample ratio

(3:1–10:1); (2) the high salt concentration introduced and subsequent need for a

higher dilution factor; and (3) the difﬁculty in streamlining the operation for high

throughput.

Hall et al. (2001) conducted a study using ashed samples to determine the dif-

ferences found in elemental concentrations when the common HF–HClO

–

HNO

–HCl digestion and lithium metaborate fusion (LiBO

) were applied. The

118 samples for which both datase ts are available comprised mostly twigs of paper

birch (Betula papyrifera), green alder (Alnus rugosa), speckled alder (Alnus crispa),

balsam poplar (Populus balsamea), balsam ﬁr (Abies balsamea) and trembling aspen

(Populus tremuloides). The agreement between the HF and fusion methods was ex-

tremely good for Co, Cs, Ga, Rb, Sr, Th, U, V and the REEs represented by La and

Ce. Given that sample weights used were 0.5 g and extremely different decompo-

sitions were employed in different laboratories with different ICP-MS instruments

and procedures, this agreement is quite remarkable. However, Hf and Zr by the HF-

acid digestion were significantly low (and noisy), at 0.6270.68 ppm ( cf

4.5670.43 ppm by fusion) and 25.8725.0 ppm ( cf 173724 ppm), respectively.

The two sets of results for the ashed vegetation corroborated the similarities and

differences found for the control V6. For Cs and Sr in the mixed tissue vegetation the

agreement between the two methods was excellent, with slopes of the regression lines

close to unity and R

values of 0.80 for Cs and 0.97 for Sr. The data for Nb were

rather noisy as many values were close to detection limits, although there was no

obvious bias. Results for the ashed vegetation for Hf, Zr and probably Ta were low,

and randomly so, by the HF-acid digestion.

The low results by acid digestion for Hf, Zr and Ta were also evident in another

project that focused on the use of vegetation for diamond explora tion (Seneshen

et al., 2005). This low recover y may be due to (1) refractory mineral matter intimately

associated with the ash itself, and/or (2) formation of ‘insoluble’ forms of these

elements during the ashing procedure. It is unlikely that Hf and Zr would exist in

185

Biogeochemistry in Mineral Exploration

‘insoluble’ forms within the dry vegetation unless they were contained within an

inorganic dust component.

STANDARD REFERENCE MATERIALS AND ANALYTICAL CONTROLS

Prior to considering in Chapter 7 the instrumentation availab le for analysis of

plant materials, controls upon analytical quality need to be discussed, because of

their vital role in any analytical protocol.

There is an ever-increasing number of standard or certiﬁed reference materials

(SRMs or CRMs), supplied by a wide array of institutions, which are suitable for

assessing the accuracy of biogeochemical data. Ihnat (1998a) lists 16 major producers

and suppliers of reference materials for elemental composition quality control in

plant analysis, and notes that there are other (unspeciﬁed) distributors. In a separate

34 page table, Ihnat (1998b) lists the many elements in reference plant materials for

which data are available, with concentrations sorted by increasing order of abun-

dance. Included in Ihnat’s lists are the two main suppliers of vegetation controls: (1)

NIST (National Institute of Standards and Technology, SRM Program, Gait-

hersburg, MD 20899, USA), and (2) BCR (Community Bureau of Reference, IRMM

[Institute for Reference Materials and Measurements], Brussels, Belgium). Other

agencies that offer vegetation controls include the Japanese National Institute for

Environmental Study (NIES No. 1 – Pepperbush leaves (Clethra barbinervis) NIES

No. 7 – tea leaves; NIES No. 8 – seaweed Sargassum), and in China the National

Research Centre for Certiﬁed Reference Materials produ ces a number of controls

(e.g., GBW 07602 and 07603 – bush branches and leaves; GBW 07604 – poplar leaves

[Populus spp.]).

A comprehensive list of reference control materials was compiled by the IAEA

(International Atomic Energy Agency) in 2003, and can be viewed at web page

www.naweb.iaea.org/nahu/nmrm/nmrm2003/index.htm. Another web page is

www.comar.bam.de/pdf/comarﬂyer.pdf. Many of the controls that are listed are

not of particular use for exploration biogeochemical purposes (e.g., pig kidney or

bovine muscle) and many are speciﬁc to the food industry (e.g., cabbage, wheat, rice).

Controls of particular relevance to biogeochemical exploration include:



BCR controls such as beech leaves (BCR-CRM-100), spruce needles (BCR-CRM-

101), white clover (BCR-CRM-402),



NIST controls such as pine needles (NIST-SRM-1575a),



The NIES controls from Japan, listed above,



The GBW controls from China, listed above, and



CANMET (Canadian Certiﬁed Reference Materials, Ottawa, ON, Canada) con-

trols CLV-1 (black spruce twigs) and CLV-2 (blac k spruce needles). These two

controls are from the uraniferous Cluff Lake area of northern Saskatchewan,

186

Sample Preparation and Decomposition

Canada. Although they are somewhat more enriched in a number of trace elements

(notably U) than the average black spruce, they are appropriate for many studies

in that black spruce is the most common tree species of the boreal forest.

Whenever possible, it is advisable to use an SRM of similar matrix to the sample

medium that is the target of a biogeochemical exploration programme. So if leaves

are the sample medium, a leaf SRM (e.g., NIST-SRM-1515, apple leaves, or BCR-

CRM-100, beech leaves) would be a preferred control; if twigs are colle cted, a twig

SRM (e.g., CANMET CLV-1) would be preferred. However, this is in an ideal world

and there simply is not a wide enough array of control materials that have been

characterized for their multi-element content. There is not, for example, a control for

outer bark from conifers, and so in this case the twig SRM would be preferred to that

of a leaf, since the composition of bark is closer to that of twigs than foliage.

There are limitations with regard to using SRMs.



They are expensive. The cost of including a large number of SRM samples (e.g.,

one for each 20 survey samples) in a large exploration programme would be pro-

hibitive.



Each SRM is only ‘Certiﬁed’ for a limited number of elements. Suppliers of SRMs,

such as NIST, generally provide ‘Reference Values’ for some elements, indicating

that only a best estimate of the true value is available, but usually a level of

uncertainty is assessed. In addition, there are ‘Information Values’ which are un-

certiﬁed and have no uncertainty assessed.



For many SRMs there are relatively few Certiﬁed and Reference Values, and some

elements, both commodity metals and pathﬁnder elements, of interest to a mineral

exploration programme may have only Inform ation Values published or no values

at all. Elements for which there are few or no data include Au, Be, Bi, Hg, PGEs,

Re, Sb and Tl. Taking the NIST SRM 1575a pine needles as an example, Certiﬁed

Values are provided for only nine trace elements and minor constituents; Reference

Values are provided for an additional 10 trace elements and Mg; Information

Values are given for two trace elements. Values for many additional elements are

listed, but the degree of validity of the data is lower than in the above three

categories. As an exampl e, the ofﬁcial NIST listings show that ﬁve laboratories

provided data for Au, all by INAA, yet returned average values (n from 2 to 9) of

0.56–2.6 ppb Au and so this is not a very useful ‘control’ on analytical data. For

some trace elements, e.g., Re and platinum-group elements, no data are provided.

Of the commodity and pathﬁnder short list shown abo ve (Au, Be, Bi, Hg, PGEs,

Re, Sb and Tl), only Hg has a Certiﬁed Value and only As has a Reference Value,

and none fall into the category of Information Value.

Given the various limitations of SRMs with respect to biogeochemical explora-

tion, it becomes necessary to use them sparingly. Published data provide an excellent

guide as to the accuracy of the analytical data for a limited number of elements that

are acquired from a set of biogeochemical survey samples. For reasons of economy,

187

Biogeochemistry in Mineral Exploration

SRMs need to be supplemented by a set of secondary reference samples for mon-

itoring quality control (QC). Analytical laboratories typically develop their own bulk

sample in-house control materials for all aspects of their analytical protocols. A

suitable rock material will be used for monitoring lithogeochemical data; a suitable

sediment for a soil or sediment survey; and a suitable vegetation medium for a

biogeochemical survey. Some laboratories use, as vegetation analytical controls, such

media as peat moss, ﬂour or bulk samples of a common tree species. The laboratories

characterize these secondary controls by reference to a few samples of an appropriate

SRM. The result is high quality data that are closely controlled.

For an independent assessment of QC by the exploration biogeochemist, a bulk

sample of a plant tissue can be collected from an appropriate ﬁeld location, and, after

drying, milling to a ﬁne powder and homogenizing, portions can be inserted as ‘blind’

controls. The data shown on some of the ensuing pages are compiled from a bulk

sample (designated V6) prepared using expedient measures that did not fully con-

form to the international SRM protocols. The procedure to prepare this material was

to ﬁrst collect a few tens of kilograms (sufﬁcient for several years) of pine twigs from

a convenient location near Ottawa, Canada, and pass it through a commercial

‘chipper-shredder’ of the type that can be used for shredding branches up to 2 cm in

diameter. This shredded material was passed through a Retsch

mill to reduce the

particle size to less than 1 mm. It was then homogen ized in a proprietary rotating

drum at the GSC, quartered several times, mixed at the CANMET laboratories and

bottled. It has been used continuously for 15 years and analysed several thousand

times – mostly by INAA, ICP-ES and ICP-MS. V6 is not commercially available, but

the data obtained by several methods over the years provide an indication of the level

of precision that can be expected from an in-house bulk control sample. The level of

precision obtained is very close to that for analyses by ICP-MS for NIST 1575a (dry

pine needles) inserted in the same batches as the V6 con trols, using the same an-

alytical procedures.

Compilations of the data from samples of V6 inserted as blind controls are in-

cluded on the CD as tables



Table 6-IID – INAA data on 10 g pellets of dry tissue.



Table 6-IIID – INAA data on 0.5 g portions of ash.



Table 6-IVD – ICP-ES (AR) data on 0.5 g portions of ash. The high variability of

As exempliﬁes the low precision typically obtained at concentrations below 10 ppm

As; over the years the laboratory changed the detection level for Ag; sensitivity to

Pb varied.



Table 6-VD – ICP-MS (HNO

+AR) on 1 g portions of dry tissue.



Table 6-VID – ICP-MS (HNO

+AR) on 0.25 g portions of ash.

In the following chapters consideration of analytical quality is based primarily on

analyses of dry V6 by ICP-MS, but with supplementary observations from other

controls and analytical duplicates.

188

Sample Preparation and Decomposition

SUMMARY

This chapter has discussed the pros and cons of many options involved in the

preparation of vegetation samples prior to their introduction into the analytical

instrumentation. The literature on these topics is voluminous and the ﬁne points

pertain more to the chemists than the exploration biogeochemists. However, it is as

well for the explorationist to be aware of these options and the poten tial pitfalls in

order to optimize conditions and provide meaningful interpretation of the results

that emerge from an exploration biogeochemical survey.

The wealth of information available can be summarized in just a few recom-

mendations.



Washing: Washing of samples from a non-dusty area creates another possible

source of error, and a relevant subtle biogeochemical signature could be removed

by vigorous washing. If samples are obviously dusty, they should be rinsed in

running water. However, in most situations, sample washing is not a requirement.



Milling: Samples should be milled to a ﬁne particle size (at least o1 mm) that is as

consistent as is practically attainable.



Ashing: For most elements this is no longer a requirement, because modern an-

alytical instrumentation (e.g., INAA and ICP-MS) is sufﬁciently sensitive to de-

termine a wide range of elements on small samples of dry tissue. A temperature of

475 1C is recommended. Elements that are typically at or close to detection 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 can be added because

data are commonly imprecise at low concentrations in dry vegetation. Some lab-

oratories provide at nominal cost an optional ashing ‘add-on’ to a multi-elem ent

package to analyse dry tissue. This is well worth considering.



Digestion: Aqua regia, preferably with a nitric acid pre-wetting, provides data for

the greatest range of elements at the lowest cost.

This protocol generates robust data and spatial patterns with an excellent cost-to-

beneﬁt ratio. A recurring theme in this book is to remember to ask ‘are the data ﬁt

for the purpose’? (Bettenay and Stanley, 2001). Given the low concentrations and

mostly high precision obtained by modern analytical methods, minor imprecision

can be tolerated without compromising the integrity of a survey.

189

Biogeochemistry in Mineral Exploration

This page intentionally left blank

Chapter 7

PLANT ANALYSIS

This chapter provides a summary acco unt of the principal methods used in the

analysis of plant materials, so that the reader has a general unde rstanding of the

various options that are offered by commercial laboratories. Emphasis is on what the

geologist/geochemist can expect to receive from a low-cost multi-element analysis

now available at one of the many co mmercial laboratories that produce data of

outstanding quality and value.

For those seeking information on sophisticated methods using state-of-the-art

analytical equipment there is a wealth of literature and some excellent reviews of

instrumentation available at ‘User Facilities’, i.e., state-of-the-art research facilities

that have been specifically constructed and operated for use by the general scientiﬁc

community (Sutton, 2006). This instrumentation is not generally available to the

explorationist looking for good to excellent quality data on many elements at low

cost. Concise reviews of ‘User Facility’ locations and instrumentation are given by

Brown et al. (2006) and Sutton et al. (2006).

A recent overvi ew by Ayrault (2005) gives descriptive accounts of the more widely

used methods.



Atomic absorption spectrometry (AAS),



Inductively coupled plasma – atomic emission spectrometry (ICP-AES), sometimes

referred to as ICP-ES (or, to the dismay of chemists simply ‘ICP’, because the ICP

component of the description is just the power source, which is the same as that

used for the more advanced generation of instrumentation, ICP-MS),



Inductively coupled plasma-mass spectrometry (ICP-MS),



Instrumental neutron activation analysis (INAA), and



X-ray ﬂuorescence per se is not described, since the principles are included in the

description of Synchrotron X-ray Fluorescence (SXRF).

Ayrault (2005) also includes su mmaries of some advanced and very expensive

instruments, such as SXRF and proton induced X-ray emission (PIXE). SXRF has

huge potential for resolving many biogeochemical questions (e.g., results of an in-

triguing Se speciation study by Pickering et al., 2000), but since access to the syn-

chrotron is not readily available to the explorationist it is not given further

consideration in this chapter.

During the 1990s, the Association of Exploration Geochemists (now renamed the

Association of Applied Geochemists) sponsored a number of short courses on