CHAPTER 2: LITERATURE REVIEW

2.1 INTRODUCTION

In this section, a review of platinum group metals and gold chemistry, base metals

chemistry, solid sample introduction techniques, slurry technique, slurry sampling

applications and slurry rheology is given. Areas such as sample handling,

analytical frame work, developed methods limitations and analytical figures of

merits are examined on published literature. Briefly, the chapter provides the

backbone of the study by highlighting other studies done in the past relevant to

this field.

2.2 PLATINUM GROUP METALS

South Africa is the world’s leading producer of precious metals. The members of

the platinum group metals (PGM); ruthenium, rhodium, palladium, osmium,

iridium, and platinum and gold are called ‘precious’ or ‘noble’ metals. These

names have roots in the unique physical and chemical properties of the metals,

table 2.1, owing to their low abundance and high economic value.

Nobility and catalytic activity are unique properties of precious metals that result

in their wide applications, for example, as catalysts in various chemical processes,

in electrical and electronic industries, and in jewelry. Growing demand for

rhodium, palladium and platinum has resulted from autocatalyts production. Since

1978, platinum complexes, cis-platinum and second-generation compounds, have

12
been applied in chemotherapy as effective anticancer drugs (Keppler, 1993). The

investigation of ruthenium complexes as an alternative to platinum cancer

inhibitors has recently become a subject of extensive studies.

Table 2.1 Physical properties of platinum group metals

Platinum Palladium Rhodium Iridium Ruthenium Osmium

Chemical symbol Pt Pd Rh Ir Ru Os
Density (g cm-3) 21.45 12.02 12.41 22.65 12.45 22.61
Melting point (ºC) 1,769 1,552 1,960 2,443 2,310 3,050
Vickers hardness 41 41 101 220 240 350
no.
Electrical resistivity 9.85 9.93 4.33 4.71 6.80 8.12
(microohm.cm at
0ºC)
Thermal 73 76 150 148 105 87
conductivity
(watts/metre/ºC)
Tensile strength (kg 14 17 71 112 165 -
mm-1)
Best solvents Alkaline Alkaline Hot Conc. Conc. Aqua
oxidizing oxidizing conc.H2SO4; HCl + HNO3, HCl regia
fusion fusion conc. HCl + NaClO3 + Cl2
NaClO3 at at 125 -
125 - 150ºC 150ºC
Source: Cotton and Wilkinson, 2001.

The six platinum metals are chemically very similar. Precious metals are colored

and lustrous, exceptionally stable, hard, malleable, electrically resistant and inert

to chemical attacks. Platinum, iridium and rhodium are the densest known metals,

platinum being 11 per cent denser than gold and twice the weight of the same

13
volume of silver or lead. Palladium, rhodium and ruthenium are lighter, palladium

having about the same density as silver.

2.2.1 Common features of platinum group metals chemistry

The chemistry of PGMs has some common features, but there are nevertheless

wide variations depending on differing stabilities of oxidation states and

stereochemistry among other things. The principle areas of general similarity

include:

o Binary compounds – there are a large number of oxides, sulfides and

phosphides besides others, but the most important are the halides.

o Aqueous chemistry – this chemistry is almost exclusively that of complex

compounds. Aqua ions of Ruп, Ruш, Rhш, Irш, Pdп and Ptп exist, but

complex ions are formed in the presence of anions other than ClO4-, BF4- or

p-toluenesulfonate. The precise nature of many supposedly simple solutions

like rhodium sulfate is complicated and often unknown. Vast arrays of

complex ions, predominantly with halides or nitrogen donor ligands, are

water soluble.

o Compounds with п-Acid Ligands

14
 Binary carbonyls are formed by all PGMs but Pd and Pt, the majority of

them polynuclear. Substituted polynuclear carbonyls are known for Pd and

Pt, and all six elements give carbonyl halides and a wide variety of

carbonyl complexes containing other ligands and carbonyl anions.

o Nitric oxide complexes are a feature of the chemistry of ruthenium.

o The formation of complexes with trialkyl- and triarylphosphines and related

phosphates, and to a lesser extent with R3As and R2S is a widely studied

area. The most important are those with triphenylphosphine and methyl-

substituted phosphines. The later are more soluble in organic solvents than

PPh3 complexes, and have proved particularly useful for the determination

of configuration by NMR.

o Mixed complexes of PR3 with CO, alkenes, halides, and hydride ligands in

at least one oxidation state are common for all of the elements.

o All these elements have a strong tendency to form bonds to carbon,

especially with alkenes and alkynes; Ptп and PtIV, and to a lesser extent Pdп

have a strong tendency to form sigma bonds, and Pdп very readily forms п-

allyl species.

15
 o A highly characteristic feature is the formation of hydrido complexes, and

M-H bonds may be formed when the metal halides in higher oxidation

states are reduced, especially in the presence of tertiary phosphines or other

ligands. Hydrogen abstraction from reaction media such as alcohols and

DMF is common.

o For the d8 ions RhI, IrI, PdII and PtII, the normal coordination is square

(though five-coordinate species are fairly common) and oxidative addition

reactions are of great importance (Cotton and Wilkinson, 2001).

Almost all complex and binary compounds of the elements give the metal when

heated above 200ºC in air or oxygen. Apart from Pd(NO3)2(OH)2 formed on

dissolution of Pd in concentrated nitric acid, the complexes are mainly the

octahedral halide anions.

Platinum group metal chemistry is an exceedingly active area of research, and

even omitting patents, which are very numerous, the research papers number in the

many hundreds per year.

Nobility and catalytic activity are unique properties of precious metals that result

in their wide applications, for example, as catalysts in various chemical processes,

in electrical and electronic industries, and in jewelry (Hartley, 1991). As observed

16
before, growing demand for rhodium, palladium and platinum has resulted from

auto-catalyts production. Since 1978, platinum complexes, cisplatin and second-

generation compounds, have been applied in chemotherapy as effective anticancer

drugs (Keppler, 1993). The investigation of ruthenium complexes as an alternative

to platinum cancer inhibitors has recently become a subject of extensive studies.

The large variety and complexity of examined materials, wide concentration

ranges to be determined (from ng g-1 and sub-ng g-1 in geological, environmental

and clinical samples to percentage levels in some industrial products), high

inertness of noble metals towards many chemical reagents and high chemical

similarities of numerous chemicals formed make the choice of the analytical

methodology for their determination in the sample of interest a challenge.

Extensive studies on the methods for the determination of the metals in various

matrices with satisfactory sensitivity, selectivity and reliability have been carried

out. Spectrometric techniques, ultra-violet visible (UV-VIS) spectrometry, atomic

absorption spectrometry (both flame (FAAS) and graphite furnace (GFAAS)

techniques), inductively coupled plasma combined with atomic emission

spectrometry (ICP-AES) or mass spectrometry (ICP-MS), are widely applied in

the analysis of a variety of samples containing noble metals over a large range of

concentrations (Balcerzak, 2002).

17
2.2.2 Sample preparation and analytical methods for PGMs and gold

The PGMs can occur as discrete mineral phases or in solid solution in mineral

phases (Juvenen et al., 1994, Talkington and Watkinson, 1986, Scoats and

Eckstrand, 1986, Hulbert, 2001). The distribution of PGMs can be erratic within

samples, particularly if they are present as discrete platinum group minerals. This

‘nugget’ effect for Platinum Group Metals is similar to that known to occur

commonly for gold, but perhaps not quite as extreme as the Platinum Group

metals minerals are usually not completely PGMs where as native gold particles

would be almost entirely gold. The required sub-sample size that should be chosen

for analysis will depend on PGMs particle size and the mineralogy of the sample

(Hoffman and Dunn, 1999).

There are many methods for determination of PGMs. The most common method is

the standard lead fire assay using silver as a collector for the platinum and

palladium followed by inductively coupled plasma-optical emission spectrometry

(ICP-OES), or direct current plasma spectroscopy (DCP) for final analytical finish.

A variation on the lead fire assay using gold as a collector is required to determine

rhodium quantitatively, as rhodium forms insoluble alloys in the silver collector

matrix that cannot be dissolved quantitatively (Hoffman and Dunn, 1999). These

techniques are rapid and accurate. For the ultimate in sensitivity, ICP-MS is used

for the analytical finish. The ICP-MS method has great application for large scale

lake bottom sediment surveys and litho-geochemical surveys looking for depletion

18
and enrichment in magmatic stratigraphy (Hulbert, 2001). The least used analytical

method for PGMs in modern laboratories are atomic absorption spectroscopy

(AAS) and graphite furnace atomic absorption spectrometry (GFAAS). ICP

methods are faster, cheaper and have less elemental interference (Hoffman and

Dunn, 1999).

The nickel sulfide fire assay technique is effective for the determination of all

PGMs and gold. After collection in the nickel sulfide matrix, the button produced

is dissolved in concentrated hydrochloric acid and the residue analyzed by

instrumental neutron activation analysis (INAA), or it can be further digested in

aqua regia and analyzed by ICP-MS. Tellurium co-precipitation may be required

to ensure that the PGMs are not partially dissolved. These procedures are usually

slower and more costly than the standard lead fire procedure, but they are the only

practical methods for quantitative determination of all PGMs and gold (Hoffman

and Dunn, 1999).

Therefore it is evident that for many years, a number of analytical techniques have

existed for elemental analysis of PGMs. Conventionally; this is done by

nebulization of dissolved samples. For many samples, however, the sample

preparation is protracted, and may be hazardous to the samples, depending on the

preparation procedure. More importantly, however, is the extended analysis time.

The direct introduction of solids would eliminate these problems, as well as

19
reducing the preparation time, and thus allowing quick monitoring. In Addition the

sample preparation cost is much cheaper than that of solid preparation.

2.3 GOLD

Gold is a soft, shiny, yellow, dense, malleable, ductile (trivalent and univalent)

transition metal. It does not react with most chemicals but is attacked by chlorine,

fluorine and aqua regia. Heat, moisture, oxygen, and most corrosive agents have

very little chemical effect on gold, making it well-suited for use in coins and

jewelry; conversely, halogens will chemically alter gold, and aqua regia dissolves

it. Common oxidation states of gold include +1 (gold (I) or aurous compounds)

and +3 (gold (III) or auric compounds). Gold ions in solution are readily reduced

and precipitated out as gold metal by the addition of virtually any other metal as

the reducing agent. The added metal is oxidized and dissolves allowing the gold to

be displaced from solution and be recovered as a solid precipitate. Gold is a good

conductor of heat and electricity, and is not affected by air and most reagents.

Pure gold is too soft for ordinary use and is hardened by alloying with silver,

copper, and other metals. Gold and its many alloys are most often used in jewelry,

coinage and as a standard for monetary exchange. Because of its high electrical

conductivity and resistance to corrosion and other desirable combinations of

physical and chemical properties, gold also emerged as an essential industrial

metal, table 2.2.

20
Table 2.2 Properties of gold

Properties Gold Properties Gold

Chemical symbol Au Number 79
Metal series Transition Density (g cm-3) 19.3
Group, Period, Block 11, 6, d Melting point (ºC) 1064.18
Appearance-metallic Yellow Boiling point (ºC) 2856
Atomic mass g mol-1 196.966569(2) Heat of fusion (Kj mol-1) 12.55
Electron configuration [Xe] 4f14 5d10 6s1 Heat of vapourization (kJ mol-1) 324
-1
Electrons per shell 2, 8, 18, 32, 18, 1 Heat capacity (25ºC) (kJ mol ) 25.418

Although gold is a noble metal, it can form many compounds, auric chloride

(AuCl3) and chlorauric acid (HAuCl4) being the most common. Gold compounds

can be aurous (univalent, +1) or auric (trivalent, +3). Gold also can under extreme

conditions form a +5 state with fluorine (gold pentafluoride, AuF5), as well as

(unusually for a metal), a -1 state. Such compounds containing the Au- anion are

called aurides and include caesium auride, CsAu, rubidium auride, RbAu, and

tetramethylammonium auride, (CH3)4N+ Au-.

Due to its relative chemical inertness gold is usually found as the native metal or

alloy. Occasionally large accumulations of native gold (also known as nuggets)

occur but usually gold occurs as minute grains. These grains occur between

mineral grain boundaries or as inclusions within minerals. Common gold

associations are quartz often as veins and sulfide minerals. The most common

sulfide associations are pyrite, chalcopyrite, galena, sphalerite, arsenopyrite,

21
stibnite and pyrrhotite. Rarer mineral associations are petzite, calaverite, sylvanite,

muthmannite, nagyagite and krennerite.

The primary source of gold is usually igneous rocks or surface concentrations. A

deposit usually needs some form of secondary enrichment to form an

economically viable ore deposit: either chemical or physical processes like erosion

or solution or more generally metamorphism, which concentrates the gold in

sulfide minerals or quartz. There are several primary deposit types, common ones

are termed reef or vein. Primary deposits can be weathered and eroded, with most

of the gold being transported into stream beds where it congregates with other

heavy minerals to form placer deposits. In all these deposits the gold is in its

native form. Another important ore type is in sedimentary black shale and

limestone deposits containing finely disseminated gold and other platinum group

metals.

2.4 BASE METALS

Base metals are common or at least inexpensive metals. Frequently, the term is

used to refer to those metals that oxidize or corrode relatively easily, and react

variably with dilute hydrochloric acid to form hydrogen. Examples include iron,

nickel, lead, copper and zinc. Copper is considered a base metal as it oxidizes

relatively easily, although it does not react with hydrochloric acid. These metals

form part of the ore body that include the Bushveld Igneous Complex (BIG), the

22
Merensky reef, Upper Group 2 (UG2) and Plat Reef. They occur in close

association with the PGMs and gold minerals, for example, Pyrrhotite (Fe1-xS),

Pentlandite (Fe,Ni)9S8 and Chalcopyrite (CuFeS2).

Table 2.3 Properties of base metals

Nickel Copper Iron Cobalt

Number 28 29 26 27
metal series Transition Transition Transition Transition
Group, Period, Block 10, 4, d 11, 4, d 8, 4, d 9, 4, d
Appearance-metallic Lustrous Brown Lustrous Gray tinge
Atomic mass g mol-1 58.6934(2) 63.545(3) 55.845(2) 58.933195(5)
8 2 10 1 6 2
Electron configuration [Ar] 3d 4S [Ar] 3d 4S [Ar] 3d 4S [Ar] 3d7 4S2
Electrons per shell 2, 8, 16,2 2, 8, 18, 1 2, 8, 14, 2 2, 8, 15, 2

Table 2.4 Physical properties of base metals

Nickel Copper Iron Cobalt

Chemical symbol Ni Cu Fe Co
Density (g cm-3) 8.908 8.96 7.86 8.90
Melting point (ºC) 1455 1084.62 1538 1495
Boiling point (ºC) 2913 2562 2861 2927
Heat of fusion (kJ mol-1) 17.48 13.26 13.81 16.06
-1
Heat of vapourization (kJ mol ) 377.5 300.4 340 377
-1
Heat capacity (25ºC) (kJ mol ) 26.07 24.44 25.10 24.81

Nickel is a silvery white metal that takes on a high polish. It belongs to the iron

group, and is hard, malleable, and ductile. The most common oxidation state of

nickel is +2, though 0, +1 and +3 Ni complexes are observed. It occurs combined

23
with sulfur in millerite, with arsenic in the mineral niccolite, and with arsenic and

sulfur in nickel glance. It is inert to oxidation and is permanently in air. The bulk

of nickel mined comes from two types of ore deposits. The first are laterites where

the principal ore minerals are nickeliferous limonite: (Fe,Ni)O(OH) and garnierite

(a hydrous nickel silicate): (Ni,Mg)3Si2O5(OH). The second are magmatic sulfide

deposits where the principal ore mineral is pentlandite: (Ni,Fe)9S8.

Copper is a reddish-colored metal, with a high electrical and thermal conductivity

(among pure metals at room temperature, only silver has a higher electrical

conductivity). Copper has its characteristic color because it reflects red and orange

light and absorbs other frequencies in the visible spectrum, due to its band

structure. Common oxidation states of copper include the less stable copper (I)

state, Cu+1; and the more stable copper (II) state, Cu+2, which forms lovely blue or

blue-green salts. Under unusual conditions, a +3 state can be obtained. Copper can

be found as native copper in mineral form. Minerals such as the carbonates azurite

(Cu3(CO3)2(OH)2) and malachite (Cu2CO3(OH)2) are sources of copper, as are

sulfides such as chalcopyrite (CuFeS2), bornite (Cu5FeS4), covellite (CuS),

chalcocite (Cu2S) and oxides like cuprite (Cu2O).

Iron is notable for being the final element produced by stellar nucleosynthesis, and

thus the heaviest element which does not require a supernova or similarly

cataclysmic event for its formation. It is therefore the most abundant heavy metal

24
in the universe. It is believed to be the tenth most abundant element in the

universe. It is also the most abundant (by mass, 34.6%) element making up the

earth; the concentration of iron in the various layers of the earth ranges from high

at the inner core to about 5% in the outer crust, contributing to the earths magnetic

field. Iron is a metal extracted from iron ore, and is hardly ever found in the free

(elemental) state. In order to obtain elemental iron, the impurities must be

removed by chemical reduction.

Most of iron is found in various iron oxides, such as the minerals hematite,

magnetite and taconite. The earth’s core is believed to consist largely of metallic

iron –nickel alloy. About 5% of the meteorites similarly consist of iron-nickel

alloy. Although rare, these are the major form of natural metallic iron on the

earth’s surface. Iron is also one of the least reactive metals, and therefore, is

sometimes found pure in nature.

Cobalt is a hard ferromagnetic silver-white element. Common oxidation states of

cobalt include +2, and +3, though +1 is also seen. Due to the various oxidation

states, there are an abundant number of compounds. Oxides are anti-ferromagnetic

at low temperature CoO (Neel temperature: 291 K) and Co3O4 (Neel temperature:

40 K). It is frequently associated with nickel, and both are characteristic

ingredients of meteoric iron. Cobalt is not found as a free metal and is generally

found in the form of ores. Cobalt is usually not mined alone, and tends to be

25
produced as a by-product of nickel and copper mining activities. The main ores of

cobalt are cobaltite, erythrite, glaucodot, and skutterudite.

2.5 SOLID SAMPLES INTRODUCTION TECHNIQUES

A variety of techniques have been developed for the introduction of solid samples

into ICP-OES. Techniques in combination with ICP-OES/MS and AAS such as

direct sample insertion, electrothermal vapourization, arc and spark ablation, laser

ablation and slurry nebulization bring new opportunities for trace/ultratrace

analysis relative to the nebulization of aqueous samples.

Direct sample insertion (DSI) and electro thermal vaporization (ETV) share many

common attributes. A few milligrams of samples is pre-weighed into a graphite or

metal support and then heated electrically or by the plasma, such that the entire

sample is used up. The solid sample may be in any form. The signal behavior is

time dependent and ETV provides limits of detection in the µg g-1 to pg g-1 range.

Precision is typically 5 to 10% RSD, with linear dynamic ranges of three to four

orders of magnitude (Montaser, 1992).

The ablation techniques, arc, spark and laser, differ from DSI and ETV in that

material is removed from the sample surface without pre-weighing. While the arcs

and sparks require the sample to be conductive, lasers do not. Laser ablation offers

an additional advantage over the other techniques in that very small surface area

26
may be ablated, providing spatial information about sample heterogeneity. For the

ablation techniques, the mass sampled is approximately 1 µg, with less than 1 µg

necessary for laser ablation. Detection limits are in the µg g-1 to ng range, and

precisions are typically 1 to 5% RSD if the sampled surface area is homogeneous.

Linear dynamic ranges are three to four orders of magnitude, similar to the DSI

and ETV methods.

A common attribute for all the techniques is the relative independence of

calibration on the matrix, provided by the Argon ICP-OES/MS. Consequently,

calibration curves may be constructed with good linearity using a variety of solid

materials having different elemental composition. In spark and laser ablation,

solution standards may be substituted in some circumstances, greatly simplifying

the calibration procedure. Solution standards may be used in slurry nebulization,

but optimum particle size is different for various samples. Ultrasonic deposition of

slurry on the ETV devices has certainly reduced the particle size effect based on

the studies reported thus far as reviewed by Montaser (1992).

Aside from solid sampling approaches already discussed, a number of other

techniques have potential for solid and slurry sampling, or have been explored

with varying degrees of success in ICP spectrometry. These approaches include

mechanical agitation/vibration followed by entrainment of particles in a carrier gas

stream (Meyer and Barnes, 1985), collection of atmospheric particles by

27
electrostatic precipitation in a graphite tube followed by analysis with ETV-

ICP/MS (Bitterli et al.,1997), fractionation of car exhaust particles by a cascade

impactor and subsequent analysis by ETV-ICP/MS (Ludke et al., 1996), use of

powder injector device (Zerezghi and Caruso, 1984), use of an acoustically driven

device (Starn and Hieftje, 1992) and fluidized-bed sampling (De Silva and

Guevremont, 1990).

The greatest challenge to be met is that of finding new methods to introduce

liquids and solids into inductively coupled plasmas. There is still a considerable

scope for devising better and more general methods for introducing materials into

plasma without the need for tedious chemical procedures (Montaser, 1992).

Slurry nebulization ICP-OES provides an exciting tool especially for the

exploration geochemists. The small minimum sample size required allows for

rapid quantification of metals like PGMs and gold in heavy mineral concentrates

or separates and in a variety of industrial materials. Such data would usefully

augment existing bulk-sample assays.

2.6 SLURRY TECHNIQUE

A number of studies have been undertaken on slurry nebulization of certain

geological materials (coal and Kaolin) into ICP-AES (Ebdon and Collier, 1988a,

28
1988b, Ebdon and Wilkinson, 1987b). Many of the factors affecting the efficiency

of slurry nebulization (for example, particle size, slurry viscosity, sample uptake)

are similar for ICP-AES and ICP-MS because of their common sample

introduction systems. Ebdon and Wilkinson (1987a) discussed the optimization of

whole coal analysis by ICP-OES which was then used to analyze five coal samples

for Cu, Fe, Mn, Ni and V (Ebdon and Wilkinson, 1987b). The quality of the

results was assessed by comparison with certified values and results obtained by

ashing and acid digestion followed by AAS analysis. The accuracy of the slurry

nebulization results varied considerably between samples and could only be

considered to be semi-quantitative at best.

Ebdon and Collier (1988a) investigated the effect of a range of dispersing agent

concentrations on the emission signal of Mg by slurry nebulization ICP-OES, and

concluded that the dispersing agent concentration must be matched in the samples,

standards and blanks to minimize differences in transport efficiencies. Later, based

on a study of five elements in a kaolin sample, Ebdon and Collier (1988b)

demonstrated that the analysis of a slurry containing particles < 8 µm could be

calibrated successfully against aqueous standards.

A simple method for preparing slurries for ICP-MS was first described by

Williams et al. (1987) for the analysis of reference soil. This study showed that

aqueous slurries were valid for instrumental calibration. The application of this

29
method to the analysis of geological samples was explored by Jarvis and Williams

(1989), Jarvis I (1992) and Jarvis I and Jarvis K.E. (1992), who demonstrated that,

provided a maximum particle size of < 5 – 10 µm and a total suspended and

dissolved solids content of < 2000 µg mL-1 were achieved, slurry nebulization

ICP-MS is a viable method for the determination of trace elements in geological

samples. Jarvis, I and Jarvis, K. (1992) analyzed 13 geochemical standard

reference materials (SRMs) for 18 elements with acceptable accuracy and

precision.

Ebdon et al. (1988) applied ICP-MS for semi-quantitative analysis of coal SRMs

as slurries using a rhodium internal standard. Quantitative determinations were

based on internal standardization with calibration curves prepared using aqueous

multi-element standards. It was reported that slurries of 1% m/v could be

nebulized for up to 4 hours without cone blockage. Mochizuki et al., (1989)

determined rare earth metals in silicate samples ground to fine particle sizes of < 3

µm. Method development studies examined the effect of grinding time, slurry

concentration, and argon injector gas flow rate. Differences in sensitivity between

the slurry and aqueous solutions were noted, but the use of aluminium as an

internal standard permitted valid calibration based on aqueous standards. Results

for the determination of lanthanide elements in several SRMs were in good

agreement with certified values (Montaser, 1997). The method was concluded to

be promising particularly for the determination of volatile elements and/or the

30
analysis of refractory minerals that commonly yield unsatisfactory results by

conventional solution-based techniques. However, one of the most serious

limitations of direct solid sampling is the lack of suitable solid standards. This is a

serious limitation for all of the plasma-based techniques.

The total solid contents of sample solutions must also be controlled.

Concentrations of solids above 0.2% have resulted in excessive deposition of salts

on sample and skimmer orifices of the instrument and blockage of sample delivery

systems in analytical instruments. Dilution of sample solutions and the use of

internal standards have been recommended as means of alleviating this problem.

Also, flow injection analysis (FIA) procedures have been used effectively (Jarvis,

1992).

2.7 SLURRY SAMPLING APPLICATIONS

The analysis of samples in the slurry form has been applied in a number of studies

with varying degrees of success. A number of examples employing the use of

different spectroscopic techniques have been documented and include applications

in atomic absorption spectroscopy (AAS), electrothermal atomic absorption

spectroscopy (ETAAS), atomic fluorescence spectroscopy (AFS), inductively

coupled plasma optical emission spectroscopy (ICP-OES) and inductively coupled

plasma mass spectroscopy (ICP-MS). Most of the methods developed involved the

31
use of hyphenated techniques. Slurry rheology has also been investigated

extensively by a number of studies (Ploetgaerts et al., 1997, Ebdon et al., 1995,

Saba et al., 1981).

2.7.1 Applications in Atomic Fluorescence Spectroscopy (AFS)

Atomic fluorescence spectroscopy is reported as the least commonly used method

of all slurry analysis methods. This is confirmed by the few publications available

on slurry slurry Atomic Fluorescence Spectroscopy.

Cava-Montesinos et al. (2004) developed an analytical procedure for the

determination of As, Sb, Se, Te and Bi in milk samples by hydride generation

atomic fluorescence spectrometry (HG-AFS). Samples were treated with aqua

regia for 10 minutes in an ultrasonic water bath and pre-reduced with KBr for total

Se and Te determination or with KI and ascorbic acid for total As and Sb. The

determination of Bi was possible in all with or without pre-reduction. Slurries of

samples, in the presence of antifoam A, were treated with NaBH4 in HCl medium

to obtain the corresponding hydrides, and AFS measurements were processed

using external calibrations prepared and measured in the same way as samples.

Results obtained by the developed procedure compared well with those found after

microwave-assisted complete digestion of samples. The proposed method was

found to be simple, fast and only 1 mL of milk was needed. The values obtained

for detection limit were 2.5, 1.6, 3, 6 and 7 ng L-1 for As, Sb, Se, Te and Bi

32
respectively in the diluted samples, with average relative standard deviation values

of 3.8, 3.1, 1.9, 6.4 and 1.2% for three independent analysis of a series of

commercially available samples of different origin.

2.7.2 Applications in Atomic Absorption Spectrometry (AAS)

AAS has been widely employed in slurry sampling studies. Several procedures

involving its use have been published, majority of which are interfaced with

various sample introduction techniques.

Ultrasonic slurry sampling graphite furnace Atomic Absorption Spectroscopy

(GFAAS) analytical consideration for analysis of high density glass was studied

by Miller-Ihli (1997). On average results were found to be 20% low as compared

with mean reference concentrations. In consequence the glass sample was

evaluated and its physical properties characterised, including particle size

distribution and density. A systematic evaluation of analysis parameters was

conducted, including slurry preparation, sample volume, pipetting accuracy and

sample depth. The data suggested that the low analytical values are likely due to

sedimentation resulting especially from high density of 2.6 mg cm-3 of the

materials. The study suggested that the optimum depth for sampling for these high

density materials should be approximately 10 mm below the surface. The

feasibility of reducing particle size to < 38 µm to allow sampling at a range of

depths was also demonstrated.

33
Lopez-Garcia et al. (1997) developed a slurry procedure for determination of

mercury in soils and sediments by GFAAS. The detection limit for Hg in 125 µL

mL-1 suspension was obtained as 0.1 µg g-1. Calibration was performed using

aqueous standards. The reliability of the procedure was proved by analysing

certified reference materials.

Dobrowolski and Mierzwa (1997) studied the slurry sampling hydride generation

atomic absorption spectrometry (HGAAS) for determination of extractable or

soluble arsenic in sediment samples. The accuracy and precision of slurry

sampling HGAAS method was studied using certified reference materials. The

relative standard deviation of the full (overall) analytical procedure was 8.5% and

an absolute limit of detection of 2.75 ng was achieved. Factors which influence the

reliability of this method were identified as the choice of slurry liquid phase

(extraction medium), sample homogeneity and, in particular, very effective mixing

of the slurry.

Lopez-Garcia (1997) further applied slurry sampling in the determination of

arsenic and antimony in soils and sediments by graphite furnace Atomic

absorption spectroscopy (GFAAS). The detection limit for Sb for 100 mg mL-1

suspension was found to be 0.03 µg g-1. The results obtained for five certified

34
reference materials using both direct calibration against aqueous standards and

standard addition method demonstrate the reliability of the procedures.

A flow injection procedure was developed by Güryelük et al. (2002) for

determination of acid-extractable arsenic in soils by HGAAS. Several parameters,

including acid and borohydride concentrations, exposure time to microwave

energy, and the microwave power applied, were optimized. The on-line

microwave extraction increased the recovery of the adsorbed arsenic significantly;

whereas, preparation of the slurry in 10% hydrochloric acid instead of water

increased the recovery only when the microwave oven was off. Low recoveries,

which were investigated by removing the hydride generation manifold and

connecting the system via an on-line filter to an inductively coupled plasma mass

spectrometer, were due to the irreversible adsorption of arsenic on the soil. This

irreversible adsorption does not allow standard additions calibration. The solution

detection limit was 0.2 µg L-1 for a 212-µL injection loop, corresponding to 7 ng g-
1
solid, for 2.5% m/v slurry prepared in 25 mL.

Marco et al. (2003) developed a simple procedure for the determination of

manganese in different sections of human brain samples by graphite furnace

atomic absorption spectrometry (GFAAS). Brain sections included cerebellum,

hypothalamus, frontal cortex, vermix and encephalic trunk. Two sample

preparation procedures were evaluated, namely, slurry sampling and microwave-

assisted acid digestion. Brain slurries (2% w/v) were prepared in de-ionized water,

35
with good stability for up to 30 minutes. Brain samples were also digested in a

domestic microwave oven using 5 mL of concentrated HNO3. A mixed palladium

and magnesium nitrate chemical modifier was used for thermal stabilization of the

analyte in the electro-thermal atomizer up to pyrolysis temperatures of 1300oC,

irrespective of the matrix. Quantitation of manganese was conducted in both cases

by means of aqueous standards calibration. The detection limits were 0.3 and 0.4

ng mL-1, for the slurry and the digested samples, respectively. The accuracy of the

procedure was checked by comparing the results obtained in the analysis of

slurries and digested brain samples, and by analysis of the NIST Bovine Liver

standard reference material (SRM 1577a). It was concluded that the ease of slurry

preparation, together with the conventional set of analytical and instrumental

conditions selected for the determination of manganese ensured that the method

used was suitable for routine clinical applications.

Dobrowolski (1996) demonstrated the determination of two base metals, nickel

and chromium, in soil by slurry GFAAS. Later, Kamogawa et al. (2001)

investigated trace metal content in hair as an index for exposure to potentially

toxic elements. Direct determination of three base metals, cadmium, cupper and

lead, in slurries of hair samples was investigated using AAS with Zeeman-effect

background correction. The analyte addition technique was used for quantification

of Cd, Cu and Pb in hair slurry samples. A reference material (GBW076901) was

analyzed and a paired t-test showed that the results for all elements obtained with

36
the slurry sampling procedure were in agreement at 95% confidence level with the

certified values. The cryogenic grinding was established as an effective method for

grinding the samples.

2.7.3 Applications in Electro-Thermal Atomic Absorption Spectrometry

(ETAAS)

Slurry sampling technique is today a well established methodology for direct

determination of trace metals in solid samples using electrothermal atomic

absorption spectrometry (SS-ETAAS). ETAAS is reported as the most commonly

used method of all the slurry analysis methods. This is confirmed by the extensive

literature published on metal content in varied matrices some of which are outlined

below. The direct introduction of solid into ETAAS atomiser offers many

advantages over the conventional techniques in terms of shortening sample

preparation time, simplicity, small reagent consumption, reduced risk of sample

contamination and improved detection limits (Baralkiewicz et al., 2005). A very

important advantage of SS ETAAS method is the fact that it may employ aqueous

solution for quantitative calibration. These advantages have popularized the slurry

atomization technique, and have been highlighted in a number of publications.

Direct cold vapour generation from aqueous slurries of environmental (marine

sediment, soil, coal) and biological (human hair, seafood) samples have been

developed using a batch mode generation system coupled with electro-thermal

37
atomic absorption spectroscopy by Moreda-Pineiro et al. (2002). The effects of

several variables affecting the cold vapour generation efficiency from solid

particles (hydrochloric acid and sodium tetrahydroborate concentrations, argon

flow rate, acid solution volume and mean particle size) was evaluated using a

Plackett-Burman experimental design. In addition, variables affecting cold vapour

trapping and atomization efficiency on Ir-treated graphite tubes (trapping and

atomization temperatures and trapping time) were also investigated. Atomization

and trapping temperatures, trapping time and hydrochloric acid concentration were

the significant variables. The 22+star and 23+star central composite designs have

been used to obtain optimum values of the variables selected. The accuracy of

methods was verified by using several certified reference materials (PACS-1,

GBW-07410, NIST-1632c, CRM-397 and DORM-2). A characteristic mass of 390

pg was achieved. The detection limits of methods were in the range of 40 - 600 ng

g-1. A particle size of less than 50 mm was adequate to obtain total cold vapour

generation of Hg content in the aqueous slurry particles.

Bermejo-Barrera et al. (1999) applied slurry sampling technique to the

determination of total mercury in human hair by electro-thermal atomic absorption

spectrometry (ETAAS). In addition, an acid leaching procedure with hydrochloric

acid was developed in order to determine Me-Hg+ in human hair with ETAAS.

Palladium reduced by ascorbic acid was found to be adequate as a chemical

modifier for both total mercury determination in hair slurries and methyl-mercury

38
determination in hydrochloric acid solutions, at concentrations of 20 and 50 mg L-
1
for palladium and ascorbic acid, respectively. The acid leaching procedure was

carried out by mechanical stirring with hydrochloric acid at 4%, and a leaching

time of 2 minutes was selected. The limits of detection were 0.04 and 0.10 mg g-1

for methyl-mercury and total mercury determination, respectively. The within-

batch precision was between 1.8 and 8.8% for total mercury determination, and

between 1.5 and 6.5% for methyl-mercury. The method was applied to 20 human

hair samples from healthy people, and it was also applied to the CRM 397 human

hair reference material, and the results agreed well with the certified ones.

More recently, Baralkiewicz et al. (2005) determined mercury in sewage sludge

by direct sampling graphite furnace atomic absorption spectrometry. KMnO4 + Pd

were used as a modifier. The minimum sample amount required for slurry

preparation with respect to sample homogeneity was evaluated by weighing

masses between 3 and 30 mg directly into the auto-sampler cups. The method was

validated using CRM 007-040 and CRM 144R certified reference materials of

sewage sludge. Two sewage sludge samples from Poznaň (Poland) were analysed

using acid digestion and direct sampling GFAAS. Similar results were obtained

for both methods within limits of statistical error. The results obtained in this work

showed that slurry sampling is a rapid and economical alternative to the

conventional acid digestion procedures for sewage sludge analysis.

39
Vieira et al. (2002), developed a method for the determination of arsenic in

sediment and in coal without sample digestion, based on hydride generation from

slurry samples. After grinding the samples to a particle size of < 50 µm, the

sample powder was mixed with aqua regia and hydrofluoric acid in an ultrasonic

bath for 30 minutes. After diluting the mixture with hydrochloric acid, the slurry

was allowed to stand for 48 hours, and an aliquot was used for hydride generation

with sodium borohydride. More than 80% of the arsenic was leached to the

aqueous phase under these conditions, except for one sediment sample with very

high silica content. The generated arsine was collected in a graphite tube, treated

with 0.5 mg of iridium as a permanent modifier, and the arsenic determination was

carried out by ETAAS. The same tube could be used for at least 160 cycles

without any re-treatment. The greatest advantage of the method was that minimum

of reagents was used and minimal sample handling was required, reducing the

risks of contamination and/or analyte loss. However, the addition calibration

technique had to be used in order to obtain results within the 95% confidence level

for 11 certified reference materials, 5 sediments, 5 coals and one coal fly ash. One

certified sediment slurry was spiked with the analyte and the resulting addition

calibration curve was used for the analysis of the certified sediments. Similarly,

one certified coal was used to obtain the addition calibration curve for the coal and

coal fly ash samples. The recoveries of the certified values were between 91 and

115%. The limits of detection in the samples were 0.54 and 0.7 mg g-1 for the coal

40
and sediment samples, respectively, obtained for 1 mL of slurry containing 1 mg

of sample.

The determination of selenium in serum by ETAAS using longitudinal Zeeman-

effect background correction combined with automated slurry sampling was

studied by Wen-Kang et al., (1997). The standard reference material bovine serum

(NIST, SRM, 1598) and second generation biological-freeze dried human serum

were analysed to verify the accuracy and precision of this technique. The average

accuracy values for certified reference serum samples and the recovery values of

spiked samples indicated this method to be an efficient and rapid technique for

determining selenium in biological samples.

ETAAS determination of selenium, cadmium and lead in different types of baby

food using slurry sampling was also studied by Vinas et al. (2000). Calibration

with aqueous standard solution was used for selenium and lead determination

while standard addition method was used for cadmium determination. The

detection limits were 5.2, 3.4 and 0.4 ng g-1 for selenium, lead and cadmium,

respectively. The reliability of the procedure was established by comparing the

results obtained with those obtained for five fish-based baby foods using a

previous microwave oven mineralization stage and by analysing six biological

certified reference materials. The lead concentration was below the detection limit

of all the baby food tested.

41
A method for determination of tin in slurry samples of sediments and coal by

hydride generation graphite furnace electro-thermal atomic absorption

spectrometry (HGGF-ETAAS) was studied by Vieira et al. (2004). The accuracy

of the method was verified by analysing five certified sediments and three coal

samples. By using external calibration against aqueous solutions, the results

obtained were in agreement with certified values only for the sediment samples.

For the coal samples, an additional calibration curve, obtained for one certified

coal, was necessary to achieve accurate results. The obtained limits of detection

were 0.03 µg g-1 for sediments and 0.09 µg g-1 for coal. Thorium was used as a

permanent modifier. The relative standard deviations were less than 15%,

demonstrating an adequate precision for slurry analysis.

A method for the determination of lead in human hair slurries by ETVAAS was

optimized by Bermejo-Barrera et al. (1999). Particle size reduction was achieved

with a vibration mill ball equipped with zirconia cups, 20 minutes being sufficient

grinding time to achieve an adequate particle diameter (< 1 µm). The use of

different thickening agents, namely glycerol, Triton X-100 and Viscalex HV30,

was studied and glycerol was found to be the best. The use of Pd and Mg(NO3)2 at

optimum concentrations of 20 and 25 mg L−1 respectively was found to be

satisfactory for stabilizing lead at 1100°C. A limit of detection of 0.21 mg kg−1

was obtained. The limit of detection can be reduced to 0.05 mg kg−1 without loss

of analytical performance by increasing four-fold the amount of hair sample.

42
Accuracy was studied by analysis of a CRM 397 human hair reference material

with a certified lead content of 33.0 ± 1.2 mg kg−1. Standard addition method was

used for the determination of lead in hair samples from healthy people, the levels

being between 2.3 and 35.5 mg kg−1.

Later, Baralkiewicz and Stepak (2001) determined lead concentrations in lake

sediment samples by means of ultrasonic slurry sampling electrothermal atomic

absorption spectrometry (USSS-ETAAS). Soil samples were suspended (0.025 –

0.15% w/v slurry) in three different liquid media containing 0.5% v/v nitric acid,

0.5% v/v nitric acid + Triton X-100 and 5% v/v nitric acid + Triton X-100. Effects

of instrumental operating conditions and slurry preparation on the signal were

examined. Palladium and magnesium were used as modifiers to improve the signal

quality. The procedure was validated by analysis of the certified reference lake

sediments materials. All analytical recoveries for lead were concluded to be

satisfactory and varied from 95% to 104%. Relative standard deviation values

were reported as 4.8, 4.7, 4.5 and 5.5. The detection limits (LOD) of lead were

obtained as 0.52, 0.45, 0.35 and 0.22 µg g-1 for sample masses of 0.025, 0.050,

0.10 and 0.15 g respectively.

Felipe-Sotelo et al. (2004) developed a new method for cobalt determination in

coal fly ash by slurry extraction – electrothermal atomic absorption spectrometry.

43
Quantitative extraction was obtained from their study with good accuracy (coal fly

ash SRM 1633a and SRM 1633b) and precision (RSD < 11.5%). The limit of

detection of the method was 0.02 µg g-1 (100 mg ml-1 slurry), and the characteristic

mass was 15 ± 1.0 pg). The slurry extraction procedure was validated for soil and

sediments analysis.

Baralkiewicz and Gramowska (2004) studied the parameters associated with the

determination of cadmium in lake sediments by SS-ETAAS. In their study,

cadmium concentration in the lake sediments was determined by suspending the

solid samples in a solution containing 5% v/v concentrated nitric acid and 0.1%

v/v Triton X-100. Three modifiers were tested for direct determination. The

furnace temperature programmes and appropriate amount for each modifier were

optimized and the highest signal and best separation between the atomic and

background signals obtained. Drying was performed by programming 400º C

temperature using a ramp time of 25 seconds and hold time of 10 seconds. Ashing

was avoided and platform atomization was performed at 2200ºC. The best

chemical modifier for cadmium determination was W-Rh permanent modifier

combined with conventional modifier using 10 µL of 0.05% w/v NH4H2PO4

solution. The modifier also acted as liquid medium for slurry and therefore

simplified the procedure further. Calibration was performed using aqueous

standards ranging from 1 – 5 µg L-1. The optimized method gave a detection limit

of 0.56 ng mL-1, characteristic mass of 10.1 ± 0.8 pg for aqueous standards, 9.6 ±

44
0.7 pg for slurry samples containing different Cd concentrations and good

precision (7.6 – 5.2%). The method was validated using four certified reference

lake sediment materials: LKSD-1, LKSD-2, LKSD-3 and LKSD-4. Satisfactory

recoveries were obtained of between 90.0 – 96.3%. No statistical differences were

reported between the experimental and the certified cadmium concentration. The

developed method was used to determine cadmium in three sediment samples

from lakes in Wielkopolski National Park area in Poland.

Detcheva et al. (2002) successfully applied slurry sampling ETAAS to the

determination of sodium impurities in single crystals of potassium titanyl

phosphate (KTP) and potassium gadolinium tungstate (KGW). Platform atomizers

coated with titanium carbide or tungsten carbide, respectively, were used in order

to avoid sensitivity drift due to the changes in the composition and the structure of

the platform surface. They indicated that calibration curves with aqueous

standards could be used for the KGW slurry (no matrix effects); analysis of KTP

slurry required the standard additions method. The precision of the proposed

method was better than 3% RSD. The results obtained by the method showed a

good agreement with those obtained by an independent method-flame atomic

absorption spectrometry (AAS) after sample digestion.

Simple and rapid slurry sampling ETAAS method was also developed by

Gentscheva et al., (2003) for determination of sodium and iron in single crystals of

45
rubidium titanyl phosphate (RTP). The finely ground crystal material was

dispersed in 5 mL of 0.2% (v/v) HNO3 containing 0.005% (v/v) Triton X-100. No

matrix interference was registered for either of the analytes in slurries containing 4

– 10 mg mL-1 of RTP, which permitted simple calibration against aqueous

standards. The precision was about 2% RSD. The results for Na and Fe in RTP

were in good agreement with those obtained by flame atomic absorption

spectroscopy (AAS) after sample digestion.

SS-ETAAS determination of aluminium and chromium in vegetables using

hydrogen peroxide as a matrix modifier was developed by Vinas et al. (1995). For

suspensions containing 0.1% m/v of ground vegetables, the relative standard

deviation (RSD) was about 0.5% for aluminium. For chromium determination, the

use of 0.5 – 1.0% suspensions lead to relative standard deviations close to 0.4%.

Calibration was done using aqueous standards. The aluminium and chromium

contents of a number of vegetable samples obtained by using the slurry approach

agreed with those obtained by means of conventional procedure based on the total

dissolution of the samples. The reliability of the procedure was also confirmed by

analyzing two certified reference materials.

Gonzales et al. (1998) applied slurry technique to the determination of Cr, Ni and

Co in wheat flour by ETAAS. They concluded that direct slurry sampling is

suitable for determination of Cr and Ni in wheat flour samples at levels of 30 – 70

46
and 150 – 450 ng g-1, respectively, as it provides results similar to those obtained

by ashing the sample. However, the typically low levels of Co in these samples

precluded its determination by this method (the study was made in a SRM spiked

whole meal flour), at least in those samples that were contaminated with elevated

levels of metals (> 90 ng of Co g-1 of flour). The method provided a relative

standard deviation of 6, 8 and 4% for Ni, Cr and Co respectively.

Slurry sampling with rapid atomization and an open focussed microwave digestion

using conventional atomization were compared for determination of Cu, Mn and

Ni in an algae matrix by Noorbasha and Sunil (1999). Transverse heated ETAAS

with a longitudinal magnetic field was used for the analysis. No significant

difference between results obtained by both methods was observed, when the

students t-test was applied at 95% confidence level. The limits of detection by

rapid slurry atomization were obtained as 1.4, 1.5, 1.3 ng mL-1 while those of

microwave digestion were 8.0, 3.6 and 6.5 ng mL-1 for Cu, Mn and Ni,

respectively. The process blank in slurry rapid atomization was found to be twice

lower for Mn and six times lower for Cu and Ni as compared to open microwave

digestion.

Kamogawa et al. (2001) applied slurry sampling to direct determination of Cd, Cu

and Pb in hair samples using an atomic absorption spectrometer with Zeeman-

effect background correction. The samples were pulverized in a freezer/mill for 13

47
minutes, and hair slurries with 1.0 g L-1 for the determination of Cu and Pb, and

5.0 g L-1 for the determination of Cd, respectively, were prepared in three different

media: 0.1% v/v Triton X-100, 0.14 mol L-1 HNO3, and 0.1% v/v of CFA-C, a

mixture of tertiary amines. The easiest way to manipulate the hair samples was in

CFA-C medium. The optimum pyrolysis and atomization temperatures were

established with hair sample slurries spiked with 10 mg L-1 Cd2+, 30 mg L-1 Pb2+,

and 10 mg L-1 Cu2+. For Cd and Pb, Pd was used as a chemical modifier, and for

Cu no modifier was needed. The analyte addition technique was used for

quantification of Cd, Cu, and Pb in hair sample slurries. A reference material

(GBW076901) was analyzed, and a paired t-test showed that the results for all

elements obtained with the proposed slurry sampling procedure were in agreement

at a 95% confidence level with the certified values. The cryogenic grinding was

established as an effective strategy to efficiently pulverize hair samples.

Baralkiewicz and Siepak, (2001) applied slurry sampling for determination of Cr,

Ni, Pb and Cd in sewage sludge by ETAAS. They optimized graphite furnace

temperature program, prepared suspensions in a medium containing 0.5 and 5.0%

(v/v) HNO3 and 0.5 and 5.0 HNO3 with 0.5% (m/v) glycerol. The prepared

suspensions were introduced directly into the furnace. The reliability of the

procedure was assessed by analyzing certified reference material, CRM 144R,

sewage sludge. The contents of Cr, Ni, Pb and Cd in two sludge samples from

Poland (Poznan area) ranged between 1.2 and 7.2 mg kg-1 for Cr; 4.5 and 44.9 mg

48
kg-1 for Ni; 0.93 and 103.1 mg kg-1 for Pb and 0.28 and 1.65 mg kg-1 for Cd. The

detection limits were calculated and obtained as 0.087, 0.054, 0.068 and 0.005 mg

kg-1 for Cr, Ni, Pb and Cd, respectively. The relative standard deviation (RSD)

measurements for the slurry sampling method was 2.7 - 4.5% for CRM 144R and

7.1 - 9.4% for sewage sludge from Poland. They concluded that slurry sampling

ETAAS provides adequate accuracy and precision for the determination of Cr, Ni,

Pb and Cd in sewage sludge, provides a simpler and faster method for trace

elements determination than wet digestion and is therefore attractive for routine

determination of elements in sewage sludge.

Bermejo-Barrera et al. (1997) developed a slurry method for determination of the

five metals, As, Cd, Pb, Hg and Sn, by ETAAS in marine sediment slurry. A

mixture of Palladium and Magnesium nitrate was used as chemical modifier.

Precision and accuracy of the method were investigated using the marine sediment

reference material PACS-1 (National Research Council Canada). The detection

limits were 44, 0.08, 60, 70 and 57 µg kg-1 for As, Cd, Pb, Hg and Sn respectively.

Slurry sampling technique was also used effectively for determination of Ni in

human scalp hair by the same author (Barmejo-Barrera, et al., 1997). Nickel was

determined by ETAAS using D2 arc background correction combined with

chemical modification (magnesium nitrate at 10 mg L-1). Matrix effects were not

observed. Aqueous calibration that covered the range 0.8 - 30µg L-1 Ni was found

49
to be suitable to develop the measurements. A characteristic mass of 22.5 pg was

obtained.

A comparative determination of Ba, Cu, Fe, Pd and Zn in tea leaf samples by two

atomic spectrometric techniques was studied by Mierzwa et al. (1998). At first,

ETAAS was applied. The results of Ba and Pb determination were calculated

using the method of standard additions, and results of Cu, Fe and Zn from the

calibration graphs based on aqueous standards. These results were compared with

the results obtained after microwave-assisted wet acid (nitric-hydrochloric-

hydrofluoric acids) digestion in closed vessels followed by ICP-OES

determinations calibrated using aqueous standards. The accuracy of the studied

methods was checked by the use of the certified reference material tea leaves,

GBW-07605. The recoveries of the analytes varied in range from 91 to 99% for

slurry sampling ETAAS, and from 92.5 to 102% for liquid sampling ICP-OES.

The advantages of the SS-ETAAS method used included the simplicity of sample

preparation procedure and the very good sensitivity displayed. Slurry sampling

ETAAS method was found to be relatively fast. However, for determination of

several elements in one sample, microwave-assisted digestion procedure with ICP-

OES determination was found to be a shorter method. Amin et al. (2003), applied

slurry sampling ETAAS for the determination of copper in herbal medicine

samples with molybdenum tube atomizer successfully.

50
SS-ETAAS method for the determination of base metals, Al, Cr, Cu, Fe, K, Li,

Mg, Mn and Na, at trace and ultra-trace level in high purity quartz samples was

developed by Hauptkorn and Krivian (1996). For their technique, the achievable

limits of detection was in the range of 2 (Mg) to 500 (Fe) ng g-1. Thus the method

was concluded to be suited for ultra-trace analysis of high purity quartz for

microelectronic applications.

Schäffer and Krivan (1996) applied SS-ETAAS to the analysis of graphite powder.

A direct method for determination of a number of base metals, Al, Ca, Cr, Cu, Fe,

K, Mg, Na, Ni, Pb, Si, Sn and Zn, in graphite powder was developed. The

influence of refractory matrix on the atomization behaviour of analyte elements

was examined. The method was optimised in respect of slurry preparation,

pyrolysis and atomization parameters. Except for Mg and Si, no chemical

modification was required. The limits of detection were in the range of 8 – 30 ng

g-1 for Ca, Mg, Cr, Cu, Na and Zn, between 50 – 200 ng g-1 for Al, Fe, Ni, Pb and

Sn. For Si, a detection limit of 800 ng g-1 was achieved.

Fast heating programs for determination of titanium and tin in soil, sediments and

sludge using ETAAS with slurry sampling was developed by Lopez-Garcia et al.

(2004). The tin and titanium content of a number of samples obtained by using

slurry sampling agreed with those obtained by digesting the samples. The

reliability of the procedure was also confirmed by the analysis of several certified

51
reference materials. The same authors also used slurry sampling for determination

of silver and gold in soil samples using ETAAS.

A procedure for determining silver and gold in soil and sediment samples using

ETAAS was developed by Lopez-Garcia et al. (2003). Slurries were prepared

from the samples and fast-program methodology was used for the heating cycle.

To determine silver, suspensions were prepared by weighing 5 - 200 mg of sample

and adding 25 ml of a solution containing 3% v/v concentrated nitric acid, 50%

v/v concentrated hydrogen peroxide and 25% v/v concentrated hydrofluoric acid.

For gold determination, slurries are prepared by weighing up to 200 mg of sample

and then adding 1 mL of concentrated hydrofluoric acid. For both cases, the

slurries were submitted to a 10 minutes mild heating stage. After cooling to

ambient temperature, the suspensions were introduced directly into the atomizer.

No modifier other than hydrofluoric acid and hydrogen peroxide (for silver

determination) were required and calibration was carried out using aqueous

standards. The detection limits were 0.02 and 0.01 mg kg-1 for silver and gold,

respectively.

Cal-Prieto et al. (2000) reviewed literature from 1990 to 2000 on slurry sampling

for direct analysis of solid materials by ETAAS. They concluded that the

determination of trace metals in solid samples has traditionally been performed by

acid digestion and subsequent measurement by a suitable instrumental technique.

52
This dissolution step is time-consuming and it shows important drawbacks. For

these reasons, in the past years many efforts have been focused on the direct

analysis of solid samples. Among the developed methodologies, slurry sampling-

electrothermal atomic absorption spectrometry combines the significant

advantages of the solid and liquid sampling methods, and it can be already

considered as a mature technique, that is widely utilized for metal determination in

both organic and inorganic matrices, even for routine analysis.

2.7.4 Applications in Inductively Coupled Plasma - Optical Emission

Spectroscopy (ICP-OES)

Analytical plasma emission spectroscopy has dramatically grown in recent years

as a multi-element analysis technique. The analytical advantage of ICP-OES is

now widely acknowledged. The excellent performance of the ICP-OES as a multi-

element source allows the rapid acquisition of data on samples introduction into

the plasma and therefore significantly saves on analysis time. Several procedures

have been published for determination of metals by slurry technique using ICP-

OES.

Ebdon and Collier, 1988b, developed a simple and rapid method for analysis of

whole kaolin samples based on nebulization of slurries into ICP-OES. They

discussed the importance of nebulizer design and particle size on transport

efficiency. Silicon was used as an internal standard in their study and effectively

53
corrected for variable atomization efficiencies associated with different particle

sizes. The value of dispersants in improving atomization efficiency was

demonstrated. Lanthanum was used to correct for the effects of viscosity arising

from different slurry concentrations of above 10% m/v.

Varga et al. (1996) investigated the colloidal stability of silicon nitride, silicon

carbide and boron carbide aqueous slurries used for slurry nebulization ICP-OES

in the pH range 2 – 10 by electrophoretic mobility and particle size measurements

and sedimentation tests. They concluded that the mean particle size of silicon

carbide and silicon nitride suspensions changed with increase of pH showing a

maximum at isoelectric points (pH 5.5 and 7.5 respectively).

Ploetgaerts et al. (1997) studied the introduction of solid environmental samples as

slurry into ICP-OES as an alternative to other techniques. They investigated a

number of approaches to improve efficiency of slurry analysis: various preparation

modes, study of optimal slurry concentration, optimization of instrumental

parameters, modification of spray chamber geometry and introduction of a

molecular gas into the plasma. They concluded that easy and efficient routine

slurry–ICP-OES analysis could not be reasonably considered for most elements

studied.

54
Merten et al. (1997) discussed the purpose, the design principle and the usage of

computer software for modelling the evaporation behaviour of particles in ICP-

OES. Computer software was developed in FORTRUN 77 language to simulate

the evaporation behaviour of particles of refractory materials such as encountered

in the analysis of advanced ceramic powders by slurry nebulization ICP-OES.

Applications for the evaporation of Al2O3 and SiC document the usefulness of the

model (for the case of 1.5 kW argon ICP of which the temperature above the load

coil has been determined to be 1600 K). The model predicts the maximum particle

size of Al2O3 and SiC that can be transported to be between 10 – 15 µm and

evaporated for a given efficiency under given experimental conditions. For both

Al2O3 and SiC, two ceramic powders of different grain sizes were investigated.

The median particle covered the range typical of ceramic powders.

The use of a low-power ICP (0.65 kW) and a mini-torch (outer tube: 12.6 mm i.d.)

for slurry sampling of Al2O3 powders was studied by Hyun et al., (1997). In order

to find the optimum conditions for mini-torch ICP-OES, the influence of particle

size and aerosol gas flow on the signal intensities was studied. The emission

intensities obtained with slurries increased as the particle size decreased and

maximized in aqueous solution. The intensities obtained with slurries also

increased with aerosol flow rate, but were still smaller than those from solution.

The intensity difference between particles of size 0.2 - 30 µm and solutions

increased with increased aerosol flow rate up to a maximum at 0.8 L min−1.

55
Dependence of the emission intensity and its relative standard deviation on the pH

of the slurries in the mini-torch ICP-OES was also studied. The best signal-to-

noise ratio was obtained at pH 3, when the pH of the suspension was varied in the

range 1 - 5. The analysis of Japan Ceramic Reference Materials (JCRM 032)

Al2O3 powders for trace impurities using slurry nebulization under the optimized

conditions provided a recovery of > 85% for MgO and Fe2O3 and 67% for Na2O

and SiO2, with precision of < 4.3% RSD. The analytical results were also

compared with those estimated for the same material, using decomposition method

in sealed PTFE vessels at high pressure.

Renfrow et al. (1997) discussed the direct injection of solids-water slurries into the

torch of ICP-OES as a method that offers a quick and easy method for the analysis

of major and minor elements in various solids. In their project, aqueous slurries of

solids were prepared by wet milling with a stirred-ball slurry attractor’s mill and

analyzed. Slurries of standard reference materials, prepared in the same way, were

used as calibration standards. The best results were obtained with solids having a

small particle size (2 - 5 µm) and a close distribution of particle sizes. The average

percent relative error for slurry-ICP analysis of most major and minor elements

was found to be generally less than 5% for the solids studied.

Marjanovic et al. (2004) investigated a simplified generalized standard additions

method (GSAM) as an alternative method for the ICP-OES analysis of solid

materials, introduced into the plasma in the form of slurries. The method was an

56
expansion of the conventional standard additions method. It was based on the

principle of varying both the sample mass and the amount of standard solution

added. The relationship between the sample mass, standard solution added and

signal intensity was assumed to be linear. Concentration of the analyte was found

either geometrically from the slope of the two-dimensional response plane in a

three-dimensional space or mathematically from the ratio of the parameters

estimated by multiple linear regressions. The analysis of a series of certified

reference materials (cement CRM-BCS No 353, gypsum CRM-GypA and basic

slag CRM No 382/I) introduced into the plasma in the form of slurry was

described. The slurries contained glycerol and hydrochloric acid and were placed

in an ultrasonic bath to ensure good dispersion. “Table curve 3D”software was

used to fit the data. Results obtained showed that the method could be successfully

applied to the analysis of cement, gypsum and slag samples, without the need to

dissolve them. In this way, the use of hazardous chemicals (concentrated acids)

could be avoided, incomplete dissolution and loss of some volatiles. The

application of the simplified GSAM for the analysis did not require a certified

reference material with similar chemical and mineralogical properties for the

calibration of the instrument.

Galbacs et al. (2000) developed a quick, novel method for tin determination in

organotin chemicals by slurry nebulization inductively coupled plasma optical

emission spectrometry (ICP-OES). The method was tested by the measurement of

57
five organotin carboxylate complexes of known composition for obtaining simple

stoichiometric data. The slurries were prepared by first dissolving the organotins

in an adequate solvent (methanol, pyridine or acetone) well miscible with water

and then adding this solution drop-by-drop to a 0.005% TX-100 tenzide solution

while maintaining intensive mixing. Dynamic laser light scattering experiments

showed that the average equivalent particle size in the resulting slurry was ~ 0.3

mm for all samples and solvents. Under suitable ICP-OES measuring conditions,

the signal recoveries were found to be between 101.8 and 106.6%, which allowed

direct nebulization and calibration against aqueous solutions. Typically, 70 mg L-1

detection limit and 1 - 5% relative standard deviation (RSD) on five replicates was

achieved by the method.

Ebdon and Wilkinson, 1987b, applied slurry atomisation to the analysis of

aqueous suspensions (10% m/v) of whole powdered coal dispersed in Triton X-

100 (1% v/v) using ICP-OES. Coals were powdered to a particle size of less than

30 µm and sprayed into a conventional torch using a high-solids type nebulizer.

Slurry atomization results for Cu, Fe, Mn, Ni and V in a range of reference

materials and other coal samples were obtained and compared with results

obtained by ashing and digestion. Acceptable precision and percentage recoveries

were obtained.

58
A method for the analysis of Ca, Cu, Fe, K, Mg, Mn, Na, P and Zn in infant

formulas, milk powders and liquid milk was developed using ICP-OES by

McKinstry et al., (1999). Samples were prepared as slurries and determined

against aqueous standards following La internal standard compensation for

potential matrix effects. The instrument used was radially configured and

incorporated a charge-transfer device detector allowing simultaneous acquisition

of multi-element and multi-line measurements. Instrument and method

performance parameters were estimated for precision, linearity, background

equivalent concentrations, detection limits and accuracy. The method was found to

be suitable for routine quality control monitoring of infant formulas and milk

powders, thereby improving sample throughput and analytical confidence.

Cintia et al. (2001) evaluated the feasibility of Portland cement analysis by

introduction of slurries in ICP-OES with axial viewing. After a fast manual

grinding of the cement samples, owing to the pulverized state of this material,

0.1% m/v slurries were prepared in 1% v/v HCl. Calibration was performed in two

ways: based on slurries prepared from different masses (50, 75, 100 and 125 mg)

of a Portland certified reference material (NIST SRM 1881) and the other based

on aqueous standard solutions. A complete analysis of cement for major (Al, Ca,

Mg, Fe and Si), minor and trace elements (Mn, S, P, Sr and Ti) was accomplished.

Both strategies led to accurate results except for Si and Ti, for which the

59
calibration using aqueous solutions produced low recovery values. Paired t-test

showed that most results were within a 95% confidence level.

Tianyou et al. (2001) developed Fluorination-assisted electro-thermal

vaporization-inductively coupled plasma-atomic emission spectrometry (FETV-

ICP-OES) for the direct determination of trace amounts of refractory impurity

elements in silicon carbide ceramic powders using slurry sampling. Investigations

indicated that a poly-tetra-fluoro-ethylene (PTFE) emulsion is a useful fluorinating

reagent for the destruction of silicon carbide and simultaneous vaporization of the

refractory impurities like B, Mo, Ti, and Zr. The vaporization behaviors of the

analytes in slurry and solution were comparatively investigated in the presence of

PTFE. The fluorinating vaporization processes and the influence factors for this

method were also studied in detail. The experimental results indicated that 80 mg

silicon carbide (10 mL of 0.8% (m/v) slurry) could be destroyed and vaporized

completely with 600 mg of PTFE under the selected conditions. Calibration was

performed using the standard addition method with aqueous standard solutions.

The accuracy was checked by comparison of the results with those obtained by

solution fluorination-assisted ETV-ICP-AES and pneumatic nebulization (PN)-

ICP-AES involving a wet-chemical decomposition of the sample. Detection limits

between 0.5 mg g-1 (B) and 0.2 mg g-1 (Mo) were achieved. In most cases, the

precision expressed as relative standard deviation (RSD) was better than 8%.

60
Matusiewicz and Golich (2004) used slurry sampling technique for simultaneous

multi-element analysis by micro-wave induced ICP-OES. Slurry samples from a

spray chamber were fed directly into the microwave cavity-torch assembly (power

300W) with no desolvation apparatus. The performance of SST-MIP-OES was

demonstrated by determination of macro (Na, K, Ca, Mg, P) and trace (Cd, Cu,

Mn, Sr, Zn) elements in three biological certified reference materials using V-

groove, clog-free Barbington-type nebulizer. The method offered relatively good

precision (RSD ranged from 7% to 11%) with measured concentrations being in

satisfactory agreement with certified values. The concentrations of Na, K, Ca, Mg,

P and Cd, Cu, Mn, Sr, Zn were determined in the range 90 – 22000 µg g-1 and 1 –

420 µg g-1, respectively.

Mokgalaka et al. (2001) evaluated the efficiency of internal standardization by

varying the operating conditions of ICP-OES. A method for the determination of

Au, Pt, Pd, Ru and Rh in a converter matte sample using ICP-OES with Y or Sc as

internal standard was developed. They found that changes in the emission signals

of the PGMs due to power variations could be effectively compensated for, using

Sc, Y or Ar lines as internal standards. For variations in aerosol gas flow rate, both

Sc and Y could only compensate for Pd and Au between flow rates of 0.60 and

0.80 L min-1 and for Pt, Ru and Rh at 0.80 L min-1. They also studied the effect of

sodium on the determination of Au and PGMs, and the use of Sc and Y as internal

standards to compensate for Na matrix effect on accuracy. They concluded that

61
emission intensities of the analytes decreased with increasing Na concentration

and that Sc could compensate for Pt, Ru and Rh up to 5 g L-1 Na as an internal

standard whereas Y could only compensate for the same elements up to 1 g L-1.

Recently, a new method that involves slurry nebulization ICP-OES analysis and

permits accurate estimation of platinum group metals (PGMs) and gold in

converter matte was developed by Mokgalaka et al. (2004). The method involves

reduction of the PGMs and gold with SnCl2 in concentrated HCl medium by

heating on a laboratory hotplate and subsequent filtration of the hot mixture

through a 0.45 µm membrane filter to separate the accompanying base metals. The

resulting metallic precipitate of Au and PGMs was quantitatively transferred into a

micronizing mill, ground for 30 minutes, diluted to 1000 mL with distilled-de-

ionized water, and placed in an ultrasonic bath for 15 minutes to ensure

homogeneity. An aliquot of this concentrated slurry was diluted to give a slurry of

0.005% m/v concentration prepared in 5.0% v/v hydrochloric acid and 1.0% m/v

Triton X-100 as a dispersant and was directly nebulized into the ICP-OES for the

determination of Au, Ir, Pd, Pt, Rh and Ru. The concentrations of Au and PGMs

determined using ICP-OES were validated employing the standard additions

method and by comparing with literature values that were obtained utilizing

reductive separation and conventional dissolution procedure. The method, which

allows separation of the PGMs and gold from a complex matrix, offers the

advantage of reduced analysis time by approximately 60% and minimized

62
contamination from additional sample preparation steps involved in the traditional

procedure.

2.7.5 Applications in Inductively Coupled Plasma - Mass Spectrometry

(ICP-MS)

Inductively coupled plasma-mass spectrometry (ICP-MS) is one of the most

sensitive, multi element technique, which is particularly well suited to the

determination of metals in trace quantities. An increasingly large number of

publications have demonstrated the application of slurry ICP-MS to the

determination of metals in a wide range of materials.

Ultrasonic slurry sampling electrothermal vaporization inductively coupled plasma

mass spectrometry (USS-ETV-ICP-MS) was applied to the determination of

mercury in several fish samples by Ming-Jyh et al., (1997). The effects of

instrument operating conditions and slurry preparation on the ion signals were

reported. Palladium was used as modifier to delay the vaporization of mercury in

this study. As the vaporization behavior of mercury in fish slurry and aqueous

solution is quite different, the standard addition method was used for the

determination of mercury in reference materials. The detection limit of mercury

estimated from the standard addition curve was in the range 0.002 - 0.004 µg g−1

for different samples. This method has been applied to the determination of

63
mercury in dogfish muscle reference material (DORM-1 and DORM-2) and

dogfish liver reference material (DOLT-1). Accuracy was better than 4% and

precision was better than 7% with the USS-ETV-ICP-MS.

Liao and Jiangu (1999) applied USS-ETV-ID-ICP-MS to the determination of Cd,

Hg and Pb in coal fly ash samples. Thioacetamide (TAC) was used as the

modifier. Since the sensitivities of the elements studied in coal fly ash slurry and

aqueous solution were quite different, isotope dilution method was used for the

determination of Cd, Hg and Pb in these coal fly ash samples. The isotope ratios of

each element were calculated from the peak areas of each injection peak. This

method was applied to the determination of Cd, Hg and Pb in NIST SRM 1633a

coal fly ash reference material and a coal fly ash sample collected from Kaohsiung

area. Analysis results of reference sample NIST SRM 1633a coal fly ash agreed

satisfactorily with the certified values. Precision was better than 6% for most of

the determinations and accuracy was better than 4% with the USS-ETV-ID-ICP-

MS method. Detection limits estimated from standard addition curves were in the

range of 24 - 58, 6 - 28 and 108 - 110 ng g-1 for Cd, Hg and Pb, respectively.

USS-ETV-ICP-MS was also applied by Yi-Ching et al. (1998) to the

determination of trace metals, Ge, As, Se, Cd and Pb, in plant samples. The

influences of instrument operating conditions and slurry preparation on the signals

were studied. Since the sensitivities of the elements in various plant slurries were

64
quite different, standard addition was used. Detection limits of Ge, As, Se, Cd and

Pb estimated in different samples from the standard addition curve were in the

range 2.1 ± 14, 3.8 ± 6.5, 14 ± 18, 1.8 ± 12, and 7.8 ± 18 ng g-1, respectively. This

method was applied to the determination of Ge, As, Se, Cd and Pb in tomato

leaves reference material (NIST SRM 1573) and a ganoderma sample obtained

from the local market. The analysis results of the tomato leaves reference sample

agreed with the certified values. Precision was better than 24% for all

determinations with the USS-ETV-ICP-MS method.

Coedo et al. (1999) used a commercial atomic absorption graphite furnace

(AAGF), with a self-made adapter and valve system, as a slurry sampling cell for

electrothermal vaporization inductively coupled plasma mass spectrometry (ETV-

ICP-MS). The system was applied to the determination of As, Sn, Sb, Se, Te, Bi,

Cd, V, Ti and Mo in steelmaking flue dusts. Experimental conditions with respect

to ETV and ICP-MS operating parameters were optimized. Compared to aqueous

solutions, slurry samples were found to present better analyte transport.

Microgram amounts of Rh were used to reduce the difference in analyte response

in sensitivity for aqueous solutions of the tested analytes. No such increasing

effect was observed for slurry samples and aqueous standards. An added quantity

of Rh acting as modifier/carrier resulted in an increase for the same analytes in

matrix-slurry solutions, even the addition of an extra Rh quantity resulted in a

decrease in the signals. The effect of Triton X-100 used as a dispersant agent on

65
analyte intensity and precision was also studied. External calibration from aqueous

standards spiked with 100 µg ml-1 Rh was performed to quantified 0.010 g/100 ml

slurry samples. Results were presented for a certified reference electrical arc

furnace flue dust (EAF): CRM-876-1 (Bureau of Analysis Samples Ltd.,

Cleveland, UK), a reference sample of coke ashes X-3705 (from AG der Dillinger

Huttenwerke, Germany), and a representative sample of EAF flue dust from a

Spanish steelmaking company (CENIM-1). For the two reference materials, an

acceptable agreement with certificate values was achieved, and the results for the

CENIM sample matched those obtained from conventional nebulization solution.

A procedure for the determination of As, Pb, Se and Sn in sediment slurries by

electrothermal vaporization inductively coupled plasma mass spectrometry was

developed by Dias et al., (2002). The slurry, 1 mg mL-1, was prepared by mixing

the sample ground to a particle size < 50 µm with 5% v/v nitric and 1% v/v

hydrofluoric acids in an ultrasonic bath. The slurry was homogenized with a

constant flow of argon in the auto sampler cup, just before transferring an aliquot

to the graphite furnace. The tube was treated with Ru as a permanent modifier, and

an optimized mass of 1 mg of NaCl was added as a physical carrier. The pyrolysis

temperature was optimized through pyrolysis curves, and a compromised

temperature of 800 oC was used; the vaporization temperature was 2300 oC. The

effect of different acid concentrations in the slurry on the analyte signal intensities

was also evaluated. The accuracy of the method was assured by the analysis of

66
certified reference sediments MESS-2, PACS-2 and HISS-1 from the National

Research Council Canada, SRM 2704 and SRM 1646a from the National Institute

of Standards and Technology and RS-4 from a round robin test, using external

calibration with aqueous standards prepared in the same medium as the slurries.

The obtained concentrations were in agreement with the certified values according

to the Student's t-test for a confidence level of 95%. The detection limits in the

samples were: 0.17 mg g-1 for As; 0.3 mg g-1 for Pb; 0.05 mg g-1 for Se and 0.28

mg g-1 for Sn. The precision found for the different sediment samples, expressed

as RSD was 1 - 8% for As, 2 - 9% for Pb, 6 - 12% for Se and 3 - 8% for Sn (n =

5).

Later, Ribeiro et al. (2004) developed a method for determination of trace metals,

As, Hg, Sb, Se and Sn, in environmental and in geological reference materials, as

acidified slurries, by flow injection (FI) coupled to a hydride generation system

(HG) and detection by inductively coupled plasma mass spectrometry (ICP-MS).

The HG unit had a gas liquid separator and a drying unit for the generated vapor.

The slurries were prepared by two procedures. Approximately 50 mg of the

reference material, ground to a particle size < 50 µm, was mixed with acid

solutions in an ultrasonic bath. In Procedure A, the medium was a hydrochloric

acid solution while in Procedure B; the medium was aqua regia plus a

hydrochloric acid solution. The conditions for the slurry formation and the

instrumental parameters were optimized. Harsh conditions were used in the slurry

67
preparation in order to reduce the hydride forming analytes to their lower

oxidation states, As (III), Se (IV), Sb (III) and Sn (II), before reacting with sodium

tetrahydroborate. To test the accuracy, 10 certified reference materials were

analyzed (four sediments, three coals, one coal fly ash and two sewage sludges),

with the analyte concentrations mostly in the µg g-1 level. Good agreements with

the certified values were obtained for Hg, Sb and Sn in the sediments using

Procedure A and calibration against aqueous standard solutions. Using Procedure

B, good results were obtained for Hg, Se and Sn in the sediment samples, for Se in

the coal and coal fly ash samples and for Hg in the sewage sludge samples, also

using external calibration with aqueous standard solutions. For As in sediments,

coals and coal fly ash, Procedure B and the analyte addition calibration was

required, indicating matrix effects. The relative standard deviations were found to

be lower than 5%, demonstrating a good precision for slurry analysis. The limits

of quantification (10 x the standard deviation; n = 10), in the samples, in ng g-1,

were: 20 for As, 60 for Hg, 80 for Sb, 200 for Se and 90 for Sn. The method

required small amounts of reagents and reduced contamination and losses.

Lu and Jiang (2001) applied USS-ETV-ICP-MS to the determination of Zn, Cd, Tl

and Pb in several soil samples. A mixture of 1% m/v EDTA and 1.5% m/v of

ascorbic acid was used as the modifier to enhance the ion signals. The influences

of instrument operating conditions and slurry preparation on the ion signals were

reported. In this study, a relatively low vaporization temperature was used which

68
separated the analyte from the major matrix components and improved the ion

signals significantly. With ETV sample introduction device, spectroscopic and

non-spectroscopic interferences were alleviated significantly. Since the

sensitivities of Zn, Cd, Tl and Pb in various soil slurries and aqueous solution were

different, standard addition method and isotope dilution method were used for the

determination of Zn, Cd, Tl and Pb in soil samples. The method was applied to the

determination of Zn, Cd, Tl and Pb in NIST SRM 2711 Montana soil reference

material and several soil samples collected from Kaohsiung area. The analysis

results were found to be in good agreement with the certified values. The precision

between sample replicates was found to be better than 9% with USS-ETV-ICP-

MS method. Detection limits estimated from standard addition curves were about

260 - 620, 3 - 5, 0.3 - 1 and 9 - 20 ng g-1 for Zn, Cd, Tl and Pb, respectively, in

different samples.

Wagner et al. (1999) studied the analytical strategy for the evaluation of base

metals, iron and copper; distribution in ancient manuscript by slurry sampling

graphite furnace atomic absorption spectrometry GFAAS and laser ablation

inductively coupled plasma mass spectrometry (LA ICP-MS). The aim of the

study was the investigation of ink corrosion in the written area of the ancient

manuscript. GFAAS was used to determine the amount of both metals in micro-

samples of paper. Determination was based on aqueous standards and the

precision using slurry sampling varied in range from 0.8 to 10.0% RSD for iron

69
and from 0.3 to 7.6% RSD for copper. The content of iron and copper varied

significantly between written (32.8 +/- 0.9 mg g-1 of iron, 15.2 +/- 0.5 mg g-1 of

copper) and non-written (3.3 +/- 1.3 mg g-1of iron, 1.6 +/- 0.8 mg g-1of copper)

areas of the investigated manuscript. LA ICP-MS was used to define the

distribution patterns of iron and copper on the surface of the manuscript.

Ultrasonic slurry sampling electrothermal vaporization isotope dilution inductively

coupled plasma mass spectrometry (USS-ETV-ID-ICP-MS) was applied to the

determination of base metals, Cu, Zn, Cd and Pb, in several fish reference samples

by Yi-Ching and Shiuh-Jen (1998). NH4NO3 was added to the sample solution to

work as a modifier. Since the sensitivity of the elements studied in fish slurry and

aqueous solution was quite different, isotope dilution was used for the

determination of Cu, Zn, Cd and Pb in reference samples. The influences of

instrument operating conditions, slurry preparation, non-spectroscopic and

spectroscopic interferences on the ion signals and accuracy and precision of

isotope ratio determination were reported. The isotope ratios of each element were

calculated from the peak areas of each injection peak. Precision of isotope ratio

determination was better than 6%. This method was applied to the determination

of Cu, Zn, Cd and Pb in dogfish muscle reference material (DORM-2), dogfish

liver reference material (DOLT-1) and oyster tissue (NIST SRM 1566a). Accuracy

was better than 19% and precision was better than 14% with ultrasonic slurry

sampling and electrothermal vaporization isotope dilution inductively coupled

70
plasma mass spectrometry. Detection limits estimated from standard addition

curves were in the range of 5 – 50, 200 – 500, 8 – 20 and 20 – 50 ng g−1 for Cu,

Zn, Cd and Pb, respectively, in different samples.

Totland et al. (1992), determined slurry PGMs and gold in solid samples by slurry

nebulization ICP-MS. Samples were ground to < 5 µg using Zr beads in

polypropylene bottles. Tetrasodium pyrophosphate was used as the dispersant.

Modified instrumental conditions included increased forward power and nebulizer

flow rate and the use of a 3 mm torch injector tube. Calibration was done using a

blank and tetrasodium pyrophosphate spiked with appropriate amounts of

standards. The limits of detection for the method were obtained as 0.04–0.2 µg g-1

in samples. Reproducibility was obtained as 2 - 30% relative standard deviation

for most materials. Analytical accuracy was assessed by comparing the slurry

results with reference concentration. The method was concluded to offer a means

for rapid determination of the precious metals in mineral concentrates, PGMs ores

and allied industrial materials.

2.8 SLURRY RHEOLOGY

A number of investigations have been done on the dependency of analytical

accuracy upon the particle size distribution (PDS) of slurry and attempts have been

made to deconvolute atomization interferences from transport effects. McCurdy

and Fry (1986) examined the determination of sulphur in coal using slurry

71
nebulization direct current plasma optical emission spectroscopy (DCP-OES) and

ICP-OES (1985). Their treatment of comparative slurry/aerosol transport assumed

that the aerosol size distribution (ASD) exiting a spray chamber was indicative of

the aerodynamic properties of that spray chamber. For slurry to be transported as

effectively as a solution, its particle size distribution must lie within the ASD of

the aerosol exiting the spray chamber. This PSD was modified according to

Stokes’ law to account for density differences between slurry and the calibration

solution. Transport efficiencies through a horn-type spray chamber were estimated

with the impact bead in place (89%) and removed (94%) for coal slurry with a

particle size distribution of d50 = 6.9 µm and d90 = 15.8 µm. Analytical recoveries

were obtained as 56% and 65%, respectively, yielding atomization efficiencies of

66% and 69%, respectively. This was with the assumption that the slurry was

homogeneous. The model’s applicability to heterogeneous materials like coal may

not be viable because analytes may be concentrated in phases with PSD and

densities very different to those assumed for the bulk slurry and therefore

transportable fractions cannot be estimated for these specific analytes.

Saba et al. (1981) studied transport efficiencies of iron particles through a number

of spray chambers commonly used for either DCP or ICP spectrometry. This was

the first in-depth study to distinguish quantitatively between the ability of a sample

introduction system to transfer large particles and the ability to transfer as

efficiently as a calibrant solution (Goodall et al., 1993). Iron particles were

72
classified into broad size bands by centrifugation and sedimentation before

suspension in base ester oil. The maximum size of an iron particle that could pass

through a standard DCP spray chamber was obtained as 17 µm. 14 µm particle

was obtained for a single pass, cylindrical spray chamber with baffle (ICP

spectroscopy). Transport efficiency decreased as particle size increased, a <10 µm

fraction was transported only 59% as efficiently as a < 1 µm fraction.

Fractionation of analyte-rich phases was demonstrated convincingly by Ebdon et

al. (1988). Determination of iron in sulphide ores was investigated and low

recoveries were observed. From the analysis done on the aerosol leaving the

sample introduction system, the low recovery of iron from the sulphide ore could

be attributed to preferential rejection of an iron-rich phase. Transport properties of

alumina and silica slurries were also examined. A sample introduction system

consisting of a high solids V-groove nebulizer, Scott-type double pass spray

chamber and a 3 mm in diameter injector rejected all alumina particles of > 5 µm.

Recoveries versus aqueous calibration showed a clear particle size distribution

(PSD) dependency.

An integration of these studies suggests that accurate analyses by slurry

nebulization ICP-OES requires a slurry whose PSD is 100% (by volume) < 5 µm

and possibly the bulk < 3 µm.

73
Atomization and transportation effects were further investigated. Silica-based

SCX ion exchange materials of known particle size (10 µm and 5 µm), which had

been completely converted to the magnesium form were used. It was assumed that

the recovery of silica would be indicative of atomization whilst that of magnesium

would be mainly dependent upon transport effects. Atomization and transport

efficiencies of 20 - 25% and 3 - 4%, respectively, were estimated for a 10 µm

particle, whilst values of 60 - 70% and 14 - 15%, respectively, estimated for a 5

µm particle.

Raeymaekers et al. (1988) studied the transport properties of slurries of refractory

oxide powders. A GMK Babington Nebulizer with impact bead/single pass spray

chamber was used. The droplet size distributions of slurries of alumina (1% m/v

Al2O3) and solution (0.8% m/v Zr) exiting from this sample introduction system

were measured using a cascade impactor. A 10 mL bore injector was employed to

direct the aerosol droplets to the impactor. Droplet volume fraction distribution

showed peak values in the 5 µm diameter size range for argon carrier gas flow

rates of 0.6 – 1.0 L min-1. Nebulizer efficiencies were found to be dependent on

the carrier gas flow rate and the presence or absence of an impact bead. Values for

the transport efficiency between 0.13 and 0.66% were obtained for solutions and

slurries.

74
Van Borm et al. (1991) studied droplet size distributions, analyte transport

efficiencies and analyte mass transfer rates of slurries and solutions with high

dissolved solids content at different sample aspiration rates using a GMK

nebulizer with impact bead, spray chamber and cascade impactor assembly as

previously described. Differences between high dissolved solids solutions and

slurry aerosols were observed under certain measurement conditions, such as

variations of liquid flow rate, slurry standard addition and the presence of an

impact bead. It was concluded that the use of simple aqueous standard solutions to

obtain calibration in slurry nebulization ICP spectrometry was limited by

fundamental differences in aerosol production processes for pure solution droplets

and droplets containing particulate matter.

Goodall et al. (1999) examined how the particle size distribution of the slurry

affects analytical accuracy and precision in slurry nebulization ICP-OES. An

empirical upper diameter for the particle size distribution of slurry of 2.0 - 2.5 µm

was derived from the analysis of a range of slurries possessing different size

distributions. A model for slurry particle transport was suggested that gives good

agreement with experimental data. The model assumes that for arbitrarily defined

but realistic standard slurry, the maximum allowable particle size is that which

allows the occupation of every aerosol droplet by one solid particle – the single

occupancy diameter. Theoretical considerations and empirical data suggest that for

accurate analyses by slurry nebulization, the particle size distribution of the slurry

75
should not exceed a value determined by density, of 2.9 µm for a material of

density 1g cm-1 falling to 1.5 µm for a material of density 7 g cm-3. Certain

carbonaceous materials were shown to be susceptible to micro-flocculation,

resulting in 5 - 10 member assemblies formed from primary particles.

This constraint on particle size may be seen as the fundamental limitation upon

slurry nebulization with direct calibration with simple aqueous solutions. The

advent of new approaches to sample grinding may allow more rapid preparation of

samples. The use of desolvation devices to provide additional aerosol processing is

still under investigation.

An often ignored property of slurry, the degree of dispersion of the solid within

the liquid, was shown to be an important limiting factor in slurry nebulization ICP

spectrometry, with micro-flocculation of very fine primary solid particles leading

to reduced analytical accuracy. This can be overcome by increasing the

concentration of surfactant in the aqueous dispersant allowing the successful

determination of a number of analytes in a variety of environmental certified

reference materials.

It is evident that sample preparation steps are of paramount importance to ensure a

good quality of analytical results. The introduction of slurry samples by pneumatic

nebulization devices has revealed some critical aspects. Problems associated with

76
sample introduction into ICP are completely different from those observed using

GF-AAS techniques where the whole sample introduced is volatilized during the

atomization step. In the ICP-OES analysis, only a little part of the sample reaches

the plasma due to low nebulization efficiency. Slurry transport may be considered

to be most efficient where every aerosol droplet is occupied by a solid particle

representative of the bulk, determined by the solid’s density and aerosol size

distribution (Ebdon et al., 1995). If this condition is achieved, then aqueous

standards may be used for calibration.

Aerosols formed from solutions and slurries do not have the same behaviour in

plasma where the sample is only present for a limited residence time of several

milliseconds. Consequently, the time necessary for the energy transfer between the

plasma and the sample is reduced. Direct analysis of slurries using ICP-OES may

then be influenced by errors if the efficiency of volatilization, emission and

ionization processes depends on the particle size. It is also expected that the

transport efficiency of small particles in a gas or liquid stream will be better than

for larger particles. With such size discrimination, the fraction transported to the

plasma may be insufficiently representative of the actual sample composition.

Measurement errors may also be expected if the emission intensity resulting from

a constant mass of the analyte depends on the size, form and density of transported

particles. These errors may result from incomplete evaporation and excitation of

large particles (Ploetgaerts et al., 1997).

77
The poor characteristics of pneumatic nebulization devices in slurry analysis and

short residence time of species in the plasma are responsible for size segregation

of particles and their incomplete dissociation. This results in a suppression of

emission signals and consequently in poor recoveries when a simple calibration

against aqueous standards is applied.

78