Electrothermal vaporization—inductively coupled plasma emission spectrometry

Electrothermal vaporization—inductively coupled plasma emission spectrometry

110 trends in analytical chemistry, vol. 5, ~0. 5, I986 trends Electrothermal vaporization-inductively coupled plasma emission spectrometry Joseph ...

590KB Sizes 2 Downloads 230 Views

110

trends in analytical chemistry, vol. 5, ~0. 5, I986

trends

Electrothermal vaporization-inductively coupled plasma emission spectrometry Joseph Sneddon and Fredrick Bet-Pera Pomona, CA, U.S.A. The use of electrothermal vaporization for sample introduction to the inductively coupled plasma for emission spectrometrk determination of elements in complex samples is discussed. The advantages and disadvantages, analytical performance characteristics, and ease of operation as compared to conventional pneumatic nebulization and other sample introduction techniques, are described.

Introduction

The inductively coupled plasma (ICP) has become widely used and accepted as an excitation source for ultratrace and trace element determination in a variety of materials by emission spectrometry. This is due, in part, to widespread availability, ease of operation, long linear dynamic range, low detection limits, high sample throughput, inherent multi-element capability, excellent freedom from interferences, good accuracy and acceptable precision. A conventional ICP is shown in Fig. 1 and was originally developed as an excitation source for emission spectrometry independently in the late 1960s and early 1970s by Greenfield and Fassel before becoming commercially available in 1975. The torch consists of three concentric tubes with the outer two usually made of quartz. An outer gas of argon or nitrogen is delivered through the outer tubes and cools the outside of the plasma and protects the torch. The intermediate gas flow may be omitted, can propagate the plasma or be the plasma gas. The central tube is used for the injector gas and consists of argon plus the sample as an aerosol. The diameter of the torch is usually either 27 or 18 mm (the ‘Greenfield Torch’ and the ‘Fassel Torch’, respectively). A radiofrequency generator produces between 2 and 30 kW forward power at between 5 and 50 MHz, typically a few kW at 27.12 MHz. When the power is on to the 2- or 3-turn induction coil, an a.c. magnetic field is generated axially through the coil. When the argon gas is flowing, the torch is seeded using a spark and these electrons accelerate in the field. Rapidly the electrons 0165-9936/86/$02.00.

reach ionizing energies and collisions with the gas in the field produce further breakdown and an avalanche effect. This occurs almost instantaneously and the magnetic field causes the ions and electrons to flow in closed circular horizontal paths. These eddy currents heat the neutral argon by collisional energy exchange and a white hot fireball is produced. The temperature of the plasma is in the region of SOOO”C,which gives an excellent excitation source for emission spectrometry. However a limitation of ICP methods is that the sample is almost exclusively introduced as a solution via a pneumatic nebulizer. Pneumatic nebulization of solution has a low efficiency and relatively high sample require\ I : + Tall Flame I :, *Viewing zone I

i

A, !:

Annular

plasma

fire- ball

Induction coil (2or,3 tut=s copper tubing)

flow (organ or nitrogen)

Aeroiol of orgon t somple (injector flow)

Fig. I. Conventional inductively coupled plasma. OElsevier

Science Publishers

B.V.

trends inqzalyticalchemistry,

TABLE

I. Advantages and disadvantages

Technique Pneumatic nebulization

111

vol. 5, no. 5,1986

of sample introduction

techniques for inductively coupled plasma emission spectrometry

Advantage ’

Good precision (l-3%) Widespread availability easy to operate.

Disadvantage for aqueous type solutions. and use. Inexpensive and

Laser ablation

Good for direct solid microgram and microliter analysis. Sample does not require to be electrically conducting.

Spark and arc

Simpler line spectra, narrower lines and improved signal-to-noise ratio compared to conventional spark emission spectrometry. High temperature (cu. 30 000 K) gives complete vaporization.

Electrothermal vaporization

Microliter and milligram samples analyzed. High analyte density and preconcentration is possible. Direct sample pretreatment, e.g., ashing is possible. Relatively easy to interface the ETV with the ICP. Modest cost although sophisticated and more expensive systems are available. Good sensitivity and interference free analysis. Modest cost although more expensive and sophisticated systems available.

Direct vaporization via hydride generation

Direct injection nebulizer

Good detection limits and relative freedom from interelement effects. Micro volume sampling and clogfree operation with 100% sample introduction efficiency. The signal decays to baseline more rapidly than with normal continuous sample.

ment (1.0 ml/min). Other sample introduction techniques to the ICP have been proposed with varying degrees of success and often for specific samples. The technique of direct injector nebulization interfaced to an ICP has been recently shown to have great promise particularly for high pressure liquid chromatographic effluents and flow injection analysis applications’. The advantages and disadvantages of these techniques are summarized in Table I. General sample introduction techniques for atomic spectroscopic s stems have been reviewed by Browner and Boorn Y. Many determinations are required to be performed on solid samples which can require fusion (silicates) or extensive and complex digestion procedures (ores) to obtain the sample in a suitable form. This can greatly increase the analysis time, dilute the sample which may take it beyond the analytical capabilities of the ICP and introduce error through contamination. A further disadvantage of pneumatic nebulization is that the solvent in a solution (partic-

Minimum volume of 1.0 ml/min required. About 90% of sample is removed to a drain. Dissolved solids and high salt concentration will block nebulizer. Organic solvents may perturb the plasma and reduce accuracy. Memory effects are possible in some cases. Ruby laser means that integration is over one shot which can give poor precision (ca. 50%) particularly if inhomogeneities in small solid samples exist. Low power-high repetition rate. Nd:YAG lasers can give improved precision (
Limited to elements which form gaseous covalent hydrides such as arsenic, selenium, lead, tin, tellurium, antimony, germanium and bismuth. Different species will require sample pretreatment prior to vaporization. Recent technique which has not received wide acceptance at present. Principally used in conjunction with high performance liquid chromatographic effluents and flow injection analysis applications where it functions best at these carrier flow rates.

ularly organic solvents) can perturb the plasma and reduce accuracy. In order to overcome some of these limitations, the use of an electrothermal vaporizer (ETV) as an alternative and complimentary technique to pneumatic nebulization for sample introduction to the ICP has been proposed by a number of workers. The ETV as an atomization source in atomic absorption spectrometry (AAS) is widely used due to microvolume sampling and much lower detection limits, in some cases lower by 2 orders of magnitude over conventional flame AAS. In this report we describe the latest developments in the use of the ETV for sample introduction to the ICP. Experimental aspects The main objectives in coupling an ETV to an ICP has been to optimize the efficiency of transfer of volatilized sample and to have a system which can be easily switched to conventional electrothermal

112

trends in analytical chemistry, vol. 5, eo. 5,1986

TO PLASMA 4

NITROGEN Fig. 2. A connecting system between an ETV and an ICP.

atomization-atomic absorption spectrometry and pneumatic nebulization-ICP. The most straight forward approach is shown in Fig. 2 and consists of positioning a glass funnel directly over an ETV (a carbon rod atomizer type, CRA 90). The sample is placed in the cup of the rod, volatilized and a flow of nitrogen or argon carries the sample to the ICP. However, the transfer efficiency can be poor due to the vaporized sample condensing on the funnel or transfer tube, and the system being open to the atmosphere which can introduce air to the system, particularly if the flow-rate of nitrogen or argon was not critically controlled. The air can dilute the sample and reduce sensitivity and perturb the plasma which will decrease accuracy. This problem can be overcome by constructing a quartz cell to enclose the ETV or use ETVs which are not open to the atmosphere3. In general, the ETV should be small to allow the sample to be reproducibly deposited, minimize vapor dilution due to diffusion and allow quick heating to vaporize the sample in one pulse for maximum sensitivity. A short glass or plastic tube between the ETV and ICP would prevent dilution of the sample but optimum flow rates are required to prevent sample condensation on this connection. Glass chamber volumes of up to one liter have been used in the interface to prevent sample condensation but at the expense of sensitivity due to dilution. Sample transport efficiency of less than 40% for cadmium has been obtained with the losses attributed to condensation. This was improved to over 60% with addition of selenium which formed a volatile cadmium-selenium compound and improved transport efficiency by reducing condensation. It is desirable to allow the sample vapor to cool immediately after leaving the hot

ETV surface to prevent condensation and maximize transport efficiency. The use of carbon or pyrolytic carbon as a surface in the ETV is the most widely used due to its relative chemical inertness, cost, machinability and strength, particularly at the high temperatures of around 3000 K obtained in an ETV. For certain determinations, it may be more desirable to use non-graphite metallic surfaces, e.g., carbide and forming elements such as molybdenum tungsten, although these metallic surfaces may react with the matrix of a sample or oxygen and have considerably less lifetime than carbon surfaces. As the maximum attainable temperature of an ETV is approximately 3000 K, then samples or elements which require higher temperatures for vaporization cannot be introduced to the ICP using the ETV, e.g., a steel sample or high boiling point elements such as refractory elements are not vaporized efficiently from the ETV. They can leave a residue behind or lead to meTABLE Element

II. Comparison d (nm)

ETV-ICP (carbon rod) IlOp1 volume]

Ag Al AS Au Be Bi Ca Cd co cu Fe Ga Ge Hg K In Li Mg Mn Ni P Pb Re Rb Sn Sb Tl Zn

328.1 396.2 228.8 242.8 267.6 234.9 289.8 422.7 228.8 238.9 324.8 259.9 372.0 417.2 303.9 253.7 404.7 325.6 670.8 279.6 285.2 257.6 279.5 352.4 213.6 405.8 346.0 420.2 317.5 259.8 377.6 535.0 213.8

0.1 2 200 1 0.1 _ 0.002 3 0.2 2 _ 1 6 2 0.4 0.01 0.1 20 10 10 30 6 0.2

of detection

limits (ng/ml)5

ETV-ICP (carbon

NebulizedICP [continuous introduction]

cup) [lop1 volume] 0.3 2 20 1 10 1 3 1 3 _ 1 1 300 0.6 0.01 0.1 4 10 280 5 5 0.3

4 40 40 -

ETVAAS [5 ~1 volume]

0.04 6 20 2

0.5 -

0.4

0.07 2 -

0.02 1.2

11.4 5 -

0.6

14 200 30 0.7 7 40 8 200 200 200 2

20 0.18 1 0.01 0.1 2

1.2 12 0.6 0.02

113

trends in analytical chemistry, vol. 5, no. 5,1986

Electrothermal Vaporization 10 ppm Pb (54)

Solution Nebulization :lOOppmPb

TIME-4

Fig. 3. Comparison of ETV-ICP ICP analyte emission intensity.

and solution nebulization-

mory effects which require blank firings of the ETV to reduce the signal back to background levels. No blank firings are required for relatively volatile elements or matrices. Volatilization of some refractory elements by chemical conversion to the more volatile halide is possible, e.g., use of a halocarbon argon gas converts boron, chromium, molybdenum, tungsten and zirconium into more volatile halides. If the element is very volatile, e.g., mercury, arsenic and selenium, it may be vaporized during a drying stage but can be overcome by addition of sodium sulfide, nickel salts or iodine which prevents vaporization at low temperatures4. Performance of ETV for sample introduction to ICP An advantage of an ETV system for sample introduction to the ICP is the formation of a high density

TABLE

of transient analyte atoms. The detection limits obtained using a carbon rod and carbon cup type ETV are shown in Table II and compared to pneumatic nebulization-ICP and conventional ETV-AAS’. While a direct comparison of detection limits between different systems can be misleading due to operating parameters, it can be clearly shown that improvement of between two and three orders of magnitude is obtained with the ETV-ICP over conventional pneumatic nebulization-ICP and comparable with ETV-AAS. Further improvement of apparent detection limits for the ETV-ICP may be obtained by preconcentration. The linearity of calibration obtainable with the ETV-ICP usually extends over four to six orders of magnitude. The precision of an analysis would be typically 3.0% or better if the stability of the injector gas flow, stability of the coupled radiofrequency power, reproducibility of introducing the sample to the ETV and the reproducibility of the vaporization of the sample is controlled and optimized. A typical analyte emission signal obtained with ETV-ICP and solution nebulization-ICP is shown in Fig. 3. The ETV-ICP signal gives a higher emission intensity and subsequently would improve the detection limit. The ETV-ICP signal is time dependent and transient in nature compared to the steadystate signal for the solution nebulization. In evaluating the ETV-ICP system, interferences, both physical and chemical, must be investigated. Incomplete vaporization, memory effects and physical effects due to the design criteria of the interface have been discussed earlier. Matrix interferences have

III. Selected applications of ETV-ICP

Elements determined

Matrix

Comments

Ca, Cd, Cu, Fe, K Mg, Na, P, Pb and Zn

Bovine liver, wastewater and plasma

Hg

Drinking water

Ni and Mn

Whole blood and animal muscle digests

Pb and Zn

Motor oil and gasoline

Al, Be, Ca, Cr, Cu, Fe, Mg, Mn, Na, Pb, Sn, Sr

Aqueous solutions

Microliter volumes or milligram masses simultaneously determined with no ashing or sample preparation required. Good accuracy obtained when compared with standard additions and certified values in standards. Addition of ammonium sulfide allowed use of ashing temperatures of over 400°C to be used without loss of mercury. Detection limit of 0.8 ng/ml and relative standard deviation of 3.1% at 100 ng/ml level obtained with accuracy confirmed by standard additions. The method was rapid and direct with sensitivity comparable to conventional techniques. Sample preconcentration not required and lack of interferences precluded the need for protein precipitation. Twenty-six solvents were introduced without extinguishing the plasma. Addition of iodine to the gasoline samples allowed tetramethyllead and tetraethyllead to be determined without difficulty. Principally concerned with simultaneous signal-vs.-time profiles.

Reference 6

7

8

9

10

114

trends in analytical chemistry, vol. 5, no. 5,1986

been noted in several determinations particularly in the presence of easily ionizable elements, which necessitates the use of matrix matching standards or use of standard additions to ensure accuracy. For certain determinations it may be possible to remove the matrix by ashing although care would need to be exercised to prevent removing the analyte at the same time. A sample trans ort model for the ETV-ICP has been developed s based on the ratio of the amount of analyte entering the ICP to the amount of analyte consumed by the ETV, exercised as a percentage. A more useful parameter may be the mass of analyte reaching the ICP per unit time and may be calculated for a specific element in a particular matrix. Real applications ETV-ICP has been applied to the determination of many elements in complex matrices and selected applications are shown in Table III. The major-advantage of ETV systems are the use of microliter volumes or milligram masses and the ability to pretreat the sample in situ which can allow a higher pyrolysis temperature to allow removal of a possible interfering matrix or preconcentration. A further potential use of the ETV could be for speciation studies, e.g., it may be possible to differentiate between different species of organolead or mercury compounds by optimizing the ashing/atomization stages. Conclusion In selecting a sample introduction system to an ICP the following criteria should be considered: (a) the state of the sample, (b) levels of the analyte, (c) accuracy and precision, (d) amount of sample available. If sufficient sample is available in a low matrix level solution, then conventional nebulization-ICP would be sufficiently accurate and precise in most cases. However for certain determinations, e.g., solid, high matrix level, or high dissolved solids then

ETV-ICP may be more acceptable for accurate and ’ precise analysis. The ETV would be a complimentary technique for sample introduction in many laboratories which require analyses to be performed on a wide variety of samples. References 1 K. E. LaFreniere, G. W. Rice and V. A. Fassel, Spectrochim. Actu, 40B (1985) 1495. 2 R. F. Browner and A. W. Boorn, Anal. Chem., 56 (1984) 786A and 875A.

3 G. Crabi, P. Cavalli, M. Achilli, G. Rossi and N. Omenetto, Atom. Spectros., 3(4) (1982) 81. 4 K. C. Ng and J. A. Caruso, Appl. Spectros., 39 (1985) 719. 5 S. E. Long, R. D. Snook and R. F. Browner, Spectrochim. Actu, 40B (1985) 553. 6 W. M. Blakemore, P. H. Casey and W. R. Collie, Anal. Chem., 56 (1984) 1376. 7 H. Matusiewicz, Z. Horvath and R. M. Barnes, Appl. Spectros., 39 (1985) 558. 8 C. Camera Rica. G. F. Kirkbright and R. D. Snook, Atom. Spectros., 2 (19Si) 172. 9 K. C. Ng and J. A. Caruso. Anal. Chem., 55 (1983) 2032. 10 M. W. Fikkanen and T. ‘M. Niemczyk, Anal. Chem., 57 (1985) 2896. Joseph Sneddon received his B.Sc. in chemistry in 1976, M.Sc. in instrumental methods of analysis in 1978 and Ph.D. in atomic spectroscopy in 1980 at the University of Struthclyde, Glusgow, U.K. He was a postdoctoral Research Fellow in 1980-81 at the University of Struthclyde and an Assistant Professor from 1981 to 1985 in the Chemistry Department at New Mexico State Vniversity in Las Cruces, New Mexico. He is currently an Associate Professor in the Chemistry Department at California State Polytechnic University. His research interests are in the general urea of atomic absorption, emission and fluorescence in electrothermal atomizers, flames and plasmas and the application of the above techniques to biological and environmental samples. Fredrick Bet-Peru received his B.Sc. in chemistry in 1972 at the National University of Iran in Tehran and Ph.D. in analytical chemistry in 1981 at Loyolu University in Chicago. He was an Assistant Professor in the Chemistry Department at Loyolu Vniversity before joining the Chemistry Department at California State Polytechnic University where he is currently an Associate Professor. His research interests are in electroanalytical and atomic absorption and emission spectrometry applied to environmental samples. Their address is Department of Chemistry, California State Polytechnic University, 3801 West Temple Avenue, Pomona, CA 91768-4016, U.S.A.

For advertising information please contact one of our advertising representatives

General Advertising

U.S.A./CANADA

GREAT BRITAIN

MichaelBaer

T.G.Scott& Sonltd.

50 East 42nd St, Suite 504, NEW YORK, NY 10017 Tel.: (212) 682-2200

Mr. M. L. White 30-32 Southampton Street LONDON WC2E 7HR Tel.: (01) 379 - 7264

Department

Elsevier/ ExcerptaMedical North-Holland MS W. van Cattenburch P.O. Box 211 1000 AE AMSTERDAM The Netherlands Tel.: (020) 5803.714/715 Telex: 18582 ESPA NL