Chapter 1: Trace Element Analysis: Methodology, Instrumentation, Analytical Performance and Application Free

The term trace was earlier used to designate a minute quantity of a substance known to be present in the sample but not quantitatively determined. As per Hillebrand,¹  the amount of a constituent indicated as trace was supposed to be below the limit of quantitative determination in the amount of samples taken for analysis and, in general, for the sake of completeness and accuracy, was supposed to be indicated as less than 0.02 or even 0.01%. Determination below this limit was rarely done to some extent because of little demand and, to some, because these levels were not readily determined by then existing classical methods such as gravimetric and volumetric. As 0.01% more or less marks the lower limit of gravimetric or volumetric methods as commonly applied, the figure 0.01% was considered as an approximate upper limit of a trace constituent, though the boundary was not very rigid²  and a few hundredths of a percent was at times considered to be a trace constituent for the sake of convenience; for example, Cu in silicate rocks, though usually ranging from 10 to 500 parts per million (ppm), was permitted to be referred to as a trace constituent in the silicate rocks.

The revolutionary advances in analytical techniques brought down the detection and even determination limits of several present-day techniques for many elements easily to the ppm level. This has resulted in regarding the figure 0.01% as the highest limit of trace amount and ppm and sub-ppm are the present-day concept of trace amount and, in general, an element present at this level is considered a trace element.

Out of the elements listed in the Periodic Table, eight elements, namely, oxygen, silicon, aluminum, iron, calcium, sodium, potassium and magnesium, make up nearly 99% and constitute major rock-forming minerals and decide the composition of the rock. The rest, normally occurring at less than 1% by weight, make up the group of minor and trace elements.

Trace elements follow a definite pattern of distribution in a particular environment. Based on geochemical behaviour, many minor and trace elements in igneous rocks originating from crystal–liquid processes could be categorized into various groups. Almost all of the transition elements belonging to the 3d group, with an atomic number from 21 to 30, for instance, Sc to Cu, get enriched in the mafic or basic fraction and exhibit medium to high interelement correlations. Y and several other heavier elements tend to be preferentially partitioned into the melt phase during the fractional crystallization of a basic magma. Platinum group elements are excluded at this stage. Cu, Fe, Ni, Co, Ag, Au, Se, Te, In, Ti and Re prefer the sulfide fraction. The immiscible aqueous phase tends to get enriched in Na, K, Ca, Mg, Cl and S (as sulfide); Li, B, P and C (as bicarbonate or carbonate). Trace elements can be conveniently grouped based on their charge and size. The elements with compatible charge and size are readily incorporated into minerals crystallizing from magma; for example, elements like V, Ni and Co enter readily into the crystal structure of magnetite. On the other hand, elements with unusual ionic size and charge tend to get enriched in the residual melt; e.g., elements like boron due to small size (20 pm) or tungsten due to high charge (+6) get rejected by major rock-forming minerals. The elements Be⁺² (35 pm), Nb⁺⁵ (69 pm), Ta⁺⁵ (68 pm), Sn⁺⁴(71 pm), Th⁺⁴ (102 pm), U⁺⁴ (97 pm), Pb⁺² (120 pm), Cs⁺ (167 pm), Li⁺ (68 pm), Rb⁺ (147 pm), Sr⁺² (112 pm) and rare earth elements (+3, 114–85 pm) get enriched in residual magmatic fluids and crystallize in later phases and concentrate in pegmatite,³  the only economic source for many of them. These elements are often referred to as incompatible elements. Elements like Ti, Zr and Nb possess high ionic potential, i.e., charge radii ratio, and are referred to as high-field-strength elements. These elements are not transported in fluids in general. Their transport is facilitated by the presence of strongly electronegative elements such as fluorine, as their presence increases the compatibility of high-field-strength elements by reducing their field strength with the negative charge on them. Such elements (high-field-strength elements) are not affected by metasomatic alteration. The trace elements with large ionic radii and lithophile tendency are referred to as large ion lithophiles.

Based on their behaviour in liquid–liquid equilibria between melts, the elements may be classified as lithophile, chalcophile or siderophile⁴  depending upon their affinity for oxygen, sulfur or iron, respectively. The gaseous elements, due to their affinity for the atmosphere, are classified as atmophile. The lithophiles are the elements with high positive electrode potential (1–3 volts), such as alkalis and alkaline earth metals; the siderophiles, such as noble metals with high negative electrode potential and the chalcophiles are generally with intermediate potentials.⁵  Some elements show affinity for more than one group⁶  as the distribution of an element depends to some extent upon the pressure, temperature and chemical behaviour of the system as a whole. For instance, chromium is a strong lithophile element in the earth's crust, but if oxygen is deficient, as in some meteorites, chromium is decidedly chalcophile and enters almost exclusively into the sulfo-spinel daubréelite, FeCr₂S₄. Similarly, carbon and phosphorous, though lithophile in the earth's crust, are siderophile under strongly reducing conditions. The mineralogy of an element, although a general guide, may not be altogether indicative of its geochemical character; e.g., although all thallium minerals are sulfides, the greater parts of the thallium in the earth's crust are contained in potassium minerals, in which Tl⁺ ion proxies for the K⁺ ion.

The signature of trace elements is revealed by their abundance, enrichment and depletion. The abundances of trace elements in principal types of sedimentary and igneous rocks are presented in:⁴⁸ 

The enrichment/depletion of trace elements depend upon their behaviour during the evolution of magma, characterized by their:

• a) Partition coefficient (K_d): K_d = C_mineral/C_liquid/melt

  • K_d < 1: in case of incompatible elements; preferentially such elements are concentrated in the melt phase,

  • K_d > 1: in case of compatible elements.

    These are preferentially concentrated in the solid phase structure (crystals).

• b) Mobility: higher mobility results in depletion of trace elements in parent rock and enrichment in secondary favourable geological environment.

• c) Geochemical affinity.

• d) Tectonic activities such as mountain building, volcanic activity, faulting and folding. These processes cause enhancement of mobility of certain trace elements due to changes in pressure, temperature conditions and action of chemically active fluids. This may cause relative enrichment or depletion of certain trace elements in a specific geological environment (mid-oceanic ridge, subduction zone, continental rift zone, etc.)

Primary magma in equilibrium with typical upper mantle mineralogy has high Ni (> 400 to 500 ppm) and Cr (>1000 ppm) and silica not more than 50%. However, if they are not derived from normal mantle but from the metasomatized source region, this criterion may not be applicable.

During the course of fractional crystallization of basaltic magma, Sr is concentrated in plagioclase, whereas Rb remains in the residual magma. Consequently, the Rb/Sr ratio of magma increases gradually during the course of progressive crystallization. Thus, a suite of cogenetic igneous rocks related by the process of fractional crystallization to a parent magma will tend to have an increasing Rb/Sr ratio with increasing degree of differentiation. However, in a closed system, all members of the suite would have an identical initial isotope ratio depending upon their age and Rb/Sr ratio.⁴⁹ 

U and Th are both preferentially concentrated in silicate melt compared to Pb and consequently, the U/Pb and Th/Pb ratios of crustal rocks are higher than those of the mantle. Additionally, U and Th are concentrated in upper crustal rocks and consequently, the upper and lower crust have distinctly different Pb isotope signature.⁴⁷ 

Mid-oceanic ridge basalts (MORB) are depleted in Cs, Rb, K, Ba, Pb and Sr relative to oceanic island and continental tholeiites. The Zr/Nb ratio serves particularly as a useful discriminant for normal type MORB from plume type. The former have higher ratios (>30) and the latter contain a low ratio of about 10.⁵⁰  Islands that are basalt, in general, have low Ni content, which suggests that they are not primary magma and have undergone olivine fractionation en route to the surface. In these, selective enrichment of Sr, K, Rb, Ba ± Th and low abundance of Ta, Nb, Ce, P, Zr, Hf, Sm, Ti, Y, Sc and Cr compared to normal type MORB is observed. MORB and uncontaminated intraplate basalt concentrate Ta and Th equally and plot within a well-defined band with a slope of unity in a Th/Yb–Ta/Yb variation diagram.⁶¹  In contrast, island arch basalt and active continental margin basalt display a higher Th/Yb ratio reflecting the influence of subduction zone fluid enriched in Th in their petrogenesis (Figure 1.1).

Trace element analysis plays a prominent role in various fields viz. mineralogical, geological, medical to biological sciences, in living organisms and in various materials of commerce from semiconductors to food, environmental samples, etc.⁵¹  The significance in geochemical exploration is obviated by the fact that the number of samples analysed is very large, in excess of thousands of millions, annually on a worldwide basis⁴⁹  in applied geochemistry alone, because of the multifaceted role played by the trace elements.

Trace elements in geochemical exploration act as pathfinders in the discovery of certain ore bodies because of their association with the mineralization due to their relative mobility and geochemical affinity. As they are widely dispersed than the ore element sought, they form targets that are larger and hence less easily missed by the sampling strategy. A few examples are as follows (Figure 1.2):

• i) Mo, V, Se and U itself for uranium deposits in sedimentary terrain (Se lags, Mo advances while V accompanies U due to their relative mobility)⁶² 

• ii) V, Cr, Co, Ni, Cu, Pb, Mo, Fe³⁺ in sandstone type

• iii) Co, Ni, As, Au, Pt group elements in unconformity type

• iv) Co, Ni, Cu, Fe^3+/Fe²⁺ in vein type

• v) V, Mo, As, Se, C, Fe^3+/Fe²⁺ in pyrites

• vi) Cu, Ni, Co, Mo and As in hydrothermal deposits

• vii) Zn, Mn, Au, Pb, Re, Ti, Te for porphyry copper deposits

• viii) As for U and Au

Trace elements have been successfully utilized in deciphering the depositional environment of sediment, for instance.

1. B⁵²  > 100 ppm indicates marine environment

   B < 60 ppm indicates freshwater environment

2. Low Fe/Mn ratio of 10 indicates marine shales and a high ratio of 42 to 62⁵³  indicates continental deposits, while the average ratio of Fe/Mn is 12.5

Different trace element discrimination diagrams have been used to decipher the tectonic settings for basic⁵⁴  and acidic igneous rocks.⁵⁵ 

• a. Basic rocks:

Basic rocks can be grouped into ocean floor basalt, low K-tholeiite calc alkaline basalt and within plate basalt based upon the ternary diagrams (Figure 1.3 and Figure 1.4):⁶³ 

• i) Ti–Zr–Y

• ii) Ti–Zr–Sr

• b. Granitic rocks:

Granites can be classified into syncollision granite, volcanic arc granite, within plate granite and ocean ridge granite based upon the binary diagrams⁶⁴  between (Figure 1.5 and Figure 1.6):

• iii) Nb–Y

• iv) Rb–Y + Nb

Rb, Ba and Sr are widely distributed in acidic igneous rocks and all three are known to replace K, particularly in feldspars. The ternary relation between these elements proved to be useful in tracing differentiation trends in acidic suites and it has the potentiality for being used as a criterion in genetic problems of granite.⁵⁶ 

According to Turekian and Wedephol⁴⁸  1961, values of 420 and 840 ppm Ba for high and low calcium granite involve an increase of Ba with differentiation. In the differentiation sequence from diorite–quartz diorite–granodiorite–normal granite, the Rb/Sr ratio increases on account of the decrease in Sr relative to Ba, while the Rb content tends to increase from normal granite to strongly differentiated granite, the Rb/Sr ratio also increases (Figure 1.7).⁶⁵ 

• i) The anomalous Zr content >250 ppm shows crystal origin for granite.

• ii) High Rb/Sr ratio is also indicative of crustal source for granites.

• iii) Low Rb/Sr indicates mantle source of magma.

Sc content in biotite and Ti content in magnetite indicate its temperature of formation.

Estimation of Rare Earth elements, a coherent group of fourteen elements, is essential to understand the petrogenesis of rocks, in identifying complex inorganic anions like Cl⁻, CO$32$ − and F⁻ in hydrothermal solutions/fluids responsible for the weathering and enrichment of various elements like U, W, Sn, etc.

Trace element analysis in geological samples constitutes a prime component in geology, as detailed in Section 1.4.2. However, analysis in applied geochemistry has very distinctive characteristics that tend to set it apart from other analytical applications in geology.⁵⁷  This difference stems from the recognition that few users of trace element data in this field need analytical results which are better than ±5% relative accuracy, i.e., a coefficient of variation (COV) of 5%. This is a result of the fact that the sampling strategy, the analysis and the interpretation of data all of them are components of an integrated activity, each of which has been optimized as a whole. Sampling error, sampling density, analytical error and cost factors are all taken into account in a system that maximizes cost-effectiveness. As accuracy and precision cost money, the analyst should not attempt to improve them beyond their useful levels. This is the philosophy that is consciously adopted in the applied geochemical analysis. Thus, it is a highly developed field carefully balanced for cost effectiveness, which meets a specific need in a competitive market.

The starting point of applied geochemistry and still its most important application is in mineral exploration. In this enterprise, samples of soil or other material are systematically collected from large tracts of land, with a view to identifying small areas (known as “anomalies”) where particular elements have unusually high concentrations. These high concentrations are often associated with mineral deposits that have no visible surface expression. The elements most often sought are the principal ore elements (Cu, Pb, Zn, Sn, W, Au, etc.), but occasionally other elements called pathfinders, which become more widely dispersed than the ore element, are also sought, and therefore form targets which are larger and hence less easily missed by the sampling strategy. The analysis for the great majority of samples has only one outcome, namely that the area represented by the sample is eliminated from further consideration. This happens when the required elements are at background levels, i.e., they do not indicate mineralization (Rose et al. 1979⁵⁸ ).

New mineral deposits become progressively harder to detect by this simple method as the easy targets are discovered. Thus, anomalies are sought which are more subtle in their expression. This subtlety may reside purely in the lower contrast of anomaly over the background, which calls for a greater degree of precision in the data, both sampling and analytical precision. Alternatively, anomalies in specific forms of an element, such as the “cold-extractable” fraction, may be determined. Such forms are reckoned to be more closely related to secondary dispersion from a mineral body than the total content of the metal in a sample. A third possibility is the more detailed study of the composition of the sample as a whole, in which case the ore metals are seen in their total context. This calls for a total analysis with reasonable precision.

In addition to exploration orientation, applied geochemistry has of late begun to diversify into environmental studies such as industrial pollution and the study of trace element disorders in crops, livestock and even man. The development resulted from a recognition that data collected for exploration purposes could be utilized directly in environmental studies and that the sampling and analytical philosophy in the two fields were very similar. Multielement data is especially valuable in environmental geochemistry because of the diversity of elements of interest and because of the probability of interactions between these elements in determining their availability and other effects on the biosphere.

Geochemical data contain variance from three sources, viz: (i) sampling error, (ii) analytical error and (iii) geochemical variation. In order to be recognizable, geochemical variation must be significantly greater than (i) and (ii) combined. Thus, in principle, the two sources of error should be reduced to a minimum, to expose the maximum amount of geochemical variation. Sampling error, however, is notoriously difficult to reduce. For exploration purposes, simple (and therefore cheap) sampling strategies are adopted. Their faults are accepted, and partly compensated for, by adjusting the sampling density. Thus, in the sampling of media such as stream sediment or soil, sampling precisions are rarely better than 10% for elements that are present. Precisions are rarely better than 10% for elements that are present as traces in a major component of the sample (e.g., zinc in clay minerals). For elements that occur in trace minerals (e.g., gold or tin in cassiterite), sampling precision may be much worse than combined by summing the variances, and there is no point in achieving analytical precision, which is substantially better than the sampling precision. Thus, it can be taken as a rule of thumb that a COV of 5% is suitable for most aspects of applied geochemistry. In some cases, even higher COVs are quite acceptable.

Some explorationists have claimed that absolute accuracy is not required in applied geochemistry, so long as consistency is maintained and significant patterns of variation can be seen in the data. While this may be true within limited areas, it is a dangerous policy to pursue, because of the possibilities for variation in accuracy in long-term work. This variation can stem from several separate sources: (i) It can be the product of variable bias between batches of analysis due to systematic error accidentally introduced. (ii) It can also result from variations in sample type, e.g., when one type of sample matrix (say limestone) causes an interference in the analyte determination which is not present in samples from another area (say shale). (iii) It can be present when one method of analysis is compared with or replaced by another. (iv) Finally, it can be present when samples are analysed in a number of different laboratories. The only safe policy is to strive towards an acceptably low degree of bias at all times. Probably a suitable guideline is to ensure that any bias fails within the range defined by the precision.

For successful determination, the analysis should have background concentrations at least three times their detection limit. This level is sometimes known as the limit of determination or the lowest quantifiable concentration. For a method with an ultimate precision of 10% (k = 0.05), the precision at the limit of determination is 37% (COV = 19%) according to Thompson and Howarths (1978) model of precision.⁵⁹ 

In many cases, the analysis is either at the detection limit or below the detection limit. This and the highest possible throughput required in applied geochemistry necessitate the development of more sensitive methods and methods for direct estimation of the analyte at the abundance level.

In the early days, the traces of elements were determined by gravimetric method by taking a very large amount of samples. To quote a few examples, 50 g of sample was taken to determine <10 ppm of Ga in Al.⁷  Fire assay procedures for noble metals still require 5–100 g of samples.⁸  Potentiometric titrations for the determination of iodine, silver, etc. and extractive titration of heavy metal dithizonates into organic solvent for determining their amount were used sometimes in the upper ranges of trace analysis. But the procedures were time-consuming, laborious and required a highly experienced analytical hand. The trace analysis today prefers methods other than these except fire assay, which even today enjoys a special position in the case of noble metals.

The present-day trace analysis methods may broadly be categorized as follows:

• i) Destructive technique (DT) and

• ii) Non-destructive technique (NDT).

The DT mainly involves solution preparation of the samples and thus provides better homogeneity, relative freedom from matrices in many cases, and an easy way of dilution without effecting the sample's homogeneity and without increasing salt concentration rather by decreasing it and providing the only approach to water analysis as compared to NDTs. But the analytical figures for DT are often quoted on solution basis and to reach at actual working analytical figures, dilution factors are to be taken into account as against the NDT, where the figures are quoted on solid basis.

The DT relates mainly to optical spectrometry, but polarography, differential pulse anodic stripping voltammetry (DPASV), ion chromatography, etc., also fall under this head.

Optical spectrometry comprises of absorption spectrometry, fluorescence spectrometry and emission spectrometry (ES)—the three branches of analytical spectrometry. The techniques derive analytical information from the molecular or atomic spectra in the optical region of the electromagnetic spectrum. The spectra in the region originate from energy transition in the outer electronic shells of the molecules or atoms.

Absorption spectrometry may further be divided into molecular absorption spectrometry and atomic absorption spectrometry depending upon whether the absorption of the electromagnetic spectrum in the optical region is affected by molecules or ground state or low-level lying atoms.

The technique comprises UV–vis spectrophotometry and allied techniques. The allied techniques involve the conversion of the component being assayed into a sparingly soluble compound in the form of a suspension, with subsequent measurement of the (i) intensity of the scattered light in nephelometry and (ii) the drop in the intensity of the incident light in turbidimetry.

The intensity of the scattered light is measured at right angles to the incident beam. The techniques are of rare use. However, they are useful in the estimation of anions like Cl⁻ and SO$4−2$.

The technique may be categorized as flame AAS or non-flame AAS depending upon whether the energy required for atomization is provided by flame or an electrical source.

• Flame AAS

The commonly used flames are the combination of

• (i) Air–acetylene: 2450 K

• (ii) N₂O–acetylene: 3200 K

• Non-flame AAS

The technique utilizes variation in atom reservoirs for efficient atomization. Graphite tube furnace/electrothermal atomization/carbon rod atomizer, graphite platform, cup in tube,⁹  wireloop,¹⁰  double furnace¹¹  allowing separate optimization of the analyte volatilization and atomization, etc. have been used. Matrix modifiers, pyrolytic coating of graphite tube furnaces or using tubes entirely made of pyrolytic graphite,¹²  etc., have been utilized to achieve a certain isoformation of samples and a like partial removal of matrix interference.

Hydride generation¹³  coupled with flame and cold vapor techniques represents a special combination of chemical separation and pre-enrichment, independent of nebulization efficiency with a determination by AAS. The power of detection can be increased and the interference can be reduced by isolating the hydride, although the signals are transient, as against in flame AAS, causing poor precision.

The techniques comprising fluorescence spectrometry are described in the following subsections.

The spectrometry utilizes either conventional instruments or laser or EDL-based instruments for higher sensitivity and low detection limits. The instruments utilize a narrow spectrum peaked at a fixed wavelength for excitation. The fluorescent emission is isolated by filters and measured at right angles to the excitation beam usually in solution attachments while for reflectance attachments, the angle between primary and secondary radiation is nearly 0°. However, gratings are used in spectrofluorometers to isolate the exciting and fluorescent radiations. The instruments employed are conventional fluorimeter, laser-induced fluorimeter (puled N₂ laser), LED fluorimeter and spectrofluorometer.

The spectrometry comprises the benefits of atomic absorption spectrometry and inductively coupled plasma (ICP). Further, the signal is measured against a nil background which results in low detection limits.

The benefits may be listed as follows:

• (1) High sensitivity and detection power

• (2) Large linear dynamic range

• (3) Spectral selectivity

• (4) Minimal chemical interference

• (5) Cost and simplicity of use of systems

Some of the systems used in atomic fluorescence spectrometry (AFS) are hollow cathode lamp-excited ICP-AFS¹⁴  atomizer, source ICPs in AFS,^15,60  laser-ICP-AFS¹⁶  and laser-excited atomic or ionic non-resonance ICP-AFS.¹⁷ 

Emission spectra are observed by dispersing the radiation, emitted by the atoms or ions on excitation, according to wavelength so that photons of different frequencies appear in the focal plane of a spectroscopy apparatus as an array of monochromatic images of the entrance slit. The images are characterized by their wavelengths and are called spectral lines. Atomic spectra are composed of discrete spectral lines. The wavelength of a line is related to the frequency and energies (E_q, E_p) of the atomic levels (q, p) between which the transition takes place by

where c is the speed of light and h is the Planck constant.

The excitation sources principally used in ES (under destructive test) are as follows.

• (i) Flames

  Air–LPG: in flame photometers

  Air–acetylene and N₂O–acetylene using an emission facility available with an atomic absorption spectrometer.

• (ii) Plasma

Plasma is a luminous volume of gas produced by an electrical discharge, laser impact or wire or foil explosion having a fraction of its atoms or molecules ionized with the following characteristics.

1. Dimensions ≫ Debye length (λ_d)

   λ_d = 6.9 (T_e/n_e)^½ (2 × 10⁻⁴ mm in ICP).

   where T_e and n_e stand for excitation temperature and the number of electrons, respectively.

2. The particle density should be so large that there are many particles within a sphere of radius λ_d.

3. (Quasi)-electro neutrality.

4. Interaction should be taking place between charged particles mutually or between charged and neutral particles.

The various types of plasma utilized in ES are directly coupled plasma (DCP),^18,19  microwave induced plasma (MIP) alone or with a graphite furnace²⁰  or wire loop,²¹  capacitively coupled plasma (CMP), at present hardly used for practical analysis because of poor detection limit as the aerosol does not penetrate the plasma efficiently, and ICP (and hyphenated techniques like hydride generation coupled with ICP).

• (a) Polarography

• (b) DPASV

• (a) Inductively coupled plasma-mass spectrometry (ICP-MS)²² 

• (b) Thermal ionization mass spectrometry (TIMS)

• (c) Isotope dilution mass spectrometry (IDMS)

• (d) Resonance ionization mass spectrometry (RIMS)

Ion chromatography²³  is a technique for multielement determination of inorganic anions.

Radiochemical NAA.

May be broadly categorized into two categories:

• a) Chemical techniques: non-nuclear analytical techniques (non-NAT)

• b) NAT

These techniques involve vaporization/excitation principally by DC-arc,²⁴  spark,²⁵  and glow discharge,²⁶  lasers²⁷  and exploding conductors.²⁸ 

XRS involves producing X-rays of suitable energy (1 to 2 keV, i.e., 12–0.6 Å) by the interaction of high-energy X-ray photons with atoms of the sample, causing ionization of the discrete orbital electron and subsequent measurement of the intensity of characteristics fluorescence X-ray produced by the atoms on electronic rearrangement in order to regain stability.

The chief techniques under XRS are listed below.

• (i) X-ray fluorescence spectrometry

  • a) Energy dispersive spectrometry (EDS)

  • b) Wavelength dispersive spectroscopy (WDS)

• (ii) Synchrotron X-ray fluorescence

• (iii) Electron probe microanalysis (EMPA)

  EMPA involves the interaction of an electron beam, accelerated in the range of 15–30 keV, with the sample to produce X-rays characteristic of the atoms of the excited samples and subsequent measurement of the intensity of the characteristic X-rays.

• (iv) Proton-induced X-ray emission spectroscopy (PIXE)

MS involves the formation of ions, their separation according to their charge to mass ratio and subsequent measurement of the characteristic signals produced by the respective ions.

IDMS involves the addition of an enriched isotope, i.e., labelling of analyte isotopically. The isotope composition in the mixture measured by MS provides the concentration of the analyte in the sample.

Some of the recent developments in MS are

• (a) Glow discharge mass spectrometry (GD-MS)²⁹ 

• (b) Mass spectrometry of sputtered neutral atoms (SNMS)³⁰ 

• (c) Laser micro-mass analyser (LAMMA) permits the determination of absolute amounts down to 10⁻¹⁸ g^31,32 

• (d) Spark–MS

• (e) Secondary ionization mass spectrometry (SIMS)

The technique involves activation of samples irradiation with neutrons, usually in a nuclear reactor and subsequent determination by measurement of the gamma-ray activity of the specific isotope of the analyte produced on irradiation.

The techniques involve two main procedures:

• (i) Instrumental neutron activation analysis (INAA) and

• (ii) Radiochemical neutron activation analysis (RNAA).

The neutrons mainly used are

1. Thermal neutrons (with energy 0–0.5 eV, most probable energy 0.025 eV)

2. Epithermal neutrons (with energies in the range of 0.5 eV to 10 keV)

3. Delayed neutron fission (DNF)

NAA is potentially applicable to the determination of almost seventy elements with high sensitivity, minimal sample preparation.

Auger electron spectroscopy (AES) is a technique used for the study of a surface or, more commonly, area for probing the chemical and compositional surface environment, specially in metallurgy, etc.

Reduced analysis time in NDT encourages developments to use solid samples at temperature/energy sufficient to vaporize, atomize and excite solid samples in recent destructive tests also. At the same time, some of the NDT are employing solution preparation of the samples to combine the benefits of the destructive test. To quote a few examples, in the first category mention may be made of the direct solid insertion device (DSID)^33–36  and graphite rod ICP-AES,³⁷  electrothermal vaporization (ETV)-MIP-AES,^20,21  furnace AAS, laser/furnace AFS, etc. while in the second category mention may be made of RNAA in which the detection limits are one to two order lower³⁸  than INAA because of isolation of specific trace elements from the more active matrix being possible by chemical separation. These would lead to the reduction of differences between DT and NDT and the evolution of techniques combining the benefits of both.

A universal technique³⁹  may be defined as the one which is useful for every element in the presence of all other elements with a similar power of detection, concentration range, precision and accuracy and without interference or interelement effect whatever be the form of sample (solid, liquid, gas) and in chemical compound form.

To date, no technique has evolved which may be categorized as a universal technique or near it. This makes an assessment of the analytical performance of the existing techniques inevitable. A rigorous, unambiguous and unbiased comparison is virtually impossible. However, it may be of use to list the order of magnitude of the power of detection, to give an indication about the matrix effects, assess multielement capability and make some statements about the possibilities for local, micro and in-depth analyses, high precision analyses and about the state of samples (liquid, solid, gas). The listing is provided in Table 1.1.⁴⁰ 

The order of magnitude listed in the table for different techniques is expressed by C_L, i.e., the limit of detection expressed on 2σ_B (two times the standard deviation of the background, usually determined by the readings obtained on analysing ten times the distilled water or, at the maximum, a process blank solution). The measurements at this level have an RSD of 50% and the determination at this level for routine concentrations in real samples should never be attempted.⁴¹  This limit or limit based on 3σ_B measurements as defined by IUPAC is useful only for comparison between different instruments or between differing operating conditions on a single instrument and are unrealistically low for application to most practical analysis and should be qualified as instrumental detection limits⁴²  (IDLs 3σ_B). The limit of quantitative determination C_D is the concentration determined with 10% RSD and usually exceeds 5C_L, true,⁴³  elsewhere referred to as limit of determination.⁴⁴  Although, Potts⁴⁵  has defined the limit determination on 6σ_B and the limit of quantitation on a 10σ_B basis. The measurements are made above the mean background.

Furthermore, the power of detection even for the same technique varies from element to element and analytical figures of merit are quoted on a solution basis in DT and the dilution factor has to be taken into account to arrive at actual figures for samples in the solid state. This necessitates either enrichment of an analyte to bring its concentration in the geological samples above the quantitative limit of determination or the development of more sensitive methods. The need is further accentuated because of the accessibility to the selected technique due to the high-cost factor involved. Among the commonly accessible techniques for inorganic analysis, the prominent techniques used are UV–vis spectrophotometry, atomic absorption spectrometry (more commonly); inductively coupled plasma-optical emission spectroscopy (ICP-OES), XRF (less commonly) and NAA (rarely). Most of the burden of geological sample analyses are shared by these techniques. A comparison between UV–vis spectrophotometry and AAS and between ICP-OES and XRF is shown in Tables 1.2 and 1.3, respectively, while the advantages and limitations of ICP-OES are discussed separately; Table 1.4 and Chart 1.1 show some of the ion sources for MS and flowsheet and capability of NAA, respectively. In brief, these were techniques very much used for specific trace elements at ppm to sub-ppm level prior to realization of ICP-OES (during 1970) and continue to do so even now in developing countries.

1. High temperature in ICP-OES (10 000 °C near the coil region) results in virtually complete atomization, and therefore, chemical interference because of the formation of stable compounds such as oxides or carbides common in flame or furnace AAS is absent.

2. Electrodeless excitation and therefore no contamination from the electrode.

3. Free atoms may be generated in the hottest zone of plasma while observation can be made in a lower temperature zone where background emission is lower.

4. Continuous temperature gradient from 9000 K to very low temperature provides greater latitude in selecting an optimal temperature.

5. Effective injection of the samples into the hot portion of plasma is possible.

6. Relatively long residence time for the sample in plasma is available.

7. Skin effect of plasma is responsible for cooler temperature in the central tube where the samples are injected. This causes the absence of self-absorption and self-reversal effect (fundamentally different from other flame techniques where the temperature falls dramatically away from the centre), resulting in a linear dynamic range of 4 to 6 orders.

8. The technique has multielement capability.

1. Nebulizing efficiency is only 1–2% as against 10% in AAS.

2. The technique cannot tolerate a high total dissolve solid (TDS) content. A TDS up to the level of 10 mg mL⁻¹ could be tolerated comfortably; higher TDS can be tolerated with humidified argon.

3. The low argon flow rate (1 L per min) necessitates the use of a peristaltic pump for sample aspiration as opposed to in AAS.

4. Halide and rare gas analysis is not possible. Although free halogens are possible, the energy provided by plasma is not enough to dissociate the halides and then to excite.

The techniques employed mainly are as follows:

• i) INAA

• ii) RNAA (DL – 1–2 order better than INAA)

La, Ce, Nd, Sm, Eu, Tb, Yb, Lu; Sc, Co, Cs, Hf, Ta, Th, U – routine determination down to DL in ppm to sub-ppm range by NAA.

U, Ta, Sb, Mo, Au, Ir – by epithermal neutron with energy >0.5 ev (0.5 ev to 10 keV).

Delayed neutron fission activation analysis (DNFAA) – U, Th (less sensitive).

NAT utilizes nuclear properties usually associated with the phenomena of ionizing radiation and isotopes. In NAT, elemental concentrations are determined based on the measurement of radiations from isotopes. Alpha spectrometry, gamma-ray spectrometry, NAA, charged particle activation analysis (CPAA), ion beam analysis (IBA), radioimmunoassay and other radiotracer–based techniques are some of the popular NATs. Besides, techniques like PIXE and XRF are considered as NATs due to the use of similar equipment in these techniques.

Further, NAT uses nuclear properties of isotopes, such as their half-life, energy and intensity of emitted radiations. Nuclear radiations like gamma rays being highly penetrating in the matter can be effectively utilized for diagnostic purposes in both industries and medical sciences. However, in some cases, IBA and CPAA techniques provide information mainly about the surface of the material (due to very low penetrating power of the ions).

NATs are highly useful in view of their multielement analytical capability in a variety of samples with good detection limits for a number of elements besides being free from contamination in most of the cases. Sensitivities of NATs such as NAA are superior for some elements and are comparable with many non-nuclear techniques like AAS, ICP-AES and ICP-MS. Besides, most of the NATs are capable of multielement analysis. Table 1.5 gives the comparison of capabilities of some NATs with other non-nuclear techniques.⁴⁶  The techniques being non-destructive could be used for art objects, coins, forensic samples, archaeological specimens, semiconductors, etc.

The main source of error in NAA is due to self-shielding, unequal neutron flux during exposure of sample and standard, counting uncertainties and error in counting due to scattering, absorption and variation in geometry of sample and standard.⁴⁷  The errors from these sources can easily be reduced to 10%. Frequently, uncertainties range within 1 to 3% (at nanogram to picogram level), giving comparable or better detection limits of several elements usually in demand as compared to techniques: flame AAS, hydride generation-AAS, GF-AAS, ICP-OES and ICP-MS.