ICP-OES allows for sensitive analysis of organic samples

Dirk Ardelt, George Glavin, Sergej Leikin and Henk Visser look at the use of inductively coupled plasma-optical emission spectroscopy and the potential difficulties in its use in analysing organic samples.

Inductively coupled plasma-optical emission spectroscopy (ICP-OES) has become widely accepted as the laboratory aworkhorse' for inorganic elemental analysis both for traces and major components.

In contrast, although the analysis of organic sample material has been investigated ever since ICP-OES made its debut as an analytical technique in the early seventies1,2, today's routine user still might find himself confronted with unexpected difficulties when attempting to analyse organic sample material directly by ICP-OES.

A common solution to such problems is the complete mineralisation, ie the aconversion' of the organic material into an inorganic sample, eg by microwave assisted digestion. However, this additional sample preparation step is time-consuming and bears contamination and error risks, ultimately rendering the methodology impractical in today's high sample throughput world.

With that said, the goal is therefore to be able to analyse organic samples by ICP-OES directly, either as liquids, or dissolved in suitable organic solvents. Typical problems encountered with the direct introduction of organic material into the ICP can be classified into three categories: u aPractical' problems: Often both the most annoying and time-consuming, these include the choice of stable operation parameters for the ICP, as well as the selection of suitable sample introduction system components, eg a chilled spray chamber or peristaltic pump tubing that is not destroyed by the sample within minutes. u Non-spectral matrix effects: This term summarises all effects leading to a matrix-dependent change in analyte signal, which are not caused by direct overlap or interference of analyte and matrix spectral lines. Typical examples include viscosity effects during nebulization, signal depression by plasma loading, etc. u Spectral matrix effects: Under this expression, all effects are subsummised which result in a matrix-dependent change in analyte signal, caused by direct overlap or interference between analyte and matrix spectral lines, eg by emission lines of small carbon-containing molecules resulting from the organic sample matrix decomposition in the ICP.

In most cases, this order also reflects the typical approach to be capable of analysing organic samples by ICP-OES.

The first priority is to achieve a stable plasma with constant and reproducible sample introduction. Sample characteristics like high volatility or excellent dissolving power for pump tubing pose typical challenges to be overcome on this stage. After choosing nebulizer and spray-chamber, suitable excitation conditions in the plasma should be chosen along the ideal of the arobust' plasma as described by Mermet3, ie stable under varying sample composition, while achieving the necessary analytical figures of merit as detection power or stability over time. Very often, a compromise has to be found here, eg via a simplex optimisation of plasma power or plasma gasflows, resulting in an ICP not or only little affected by non-spectral matrix effects.

After solving practical problems and the reduction or removal of non-spectral matrix effects, overlaps between analyte and matrix emission lines, ie spectral matrix effects, remain as major obstacle. Among other possible solutions like the use of mixed-gas plasmas (see, eg the use of an Ar-O2 mixed gas ICP to reduce hydrocarbon interferences on the NaI589.592nm line in the analysis of biodiesel, in4), modern multichannel semiconductor detector based, simultaneous ICP-OES systems (eg using CCDs) provide an elegant way to remedy spectral interferences.

Since such instruments always measure the complete spectrum between the vacuum UV and the near IR (contemporary state of the art is a wavelength range 120-800nm) with exceptional wavelength stability and speed due to their simultaneous nature (a complete spectrum may take as little as 3s), mathematical correction methods like background subtraction over complete spectral ranges can easily be applied.

This method shall be demonstrated for the ICP-OES determination of sulphur in a strong organic solvent, 1-Methoxy-2-propylacetate (PMA), using a system with axial plasma viewing and simultaneous CCD detectors covering a spectral range between 120-800nm (SPECTRO CIROSCCD, SPECTRO Analytical Instruments, Germany). PMA is a versatile solvent, used eg for photoresists, adhesives, or solid inks. Due to the high volatility of PMA, a self-aspirating concentric nebulizer was used to introduce the sample into the ICP, operated at 1500W RF-power and with the addition of 0.01l/min O2 to the auxiliary gas to prevent carbon deposits on the axial plasma interface. Silicon tubing was used for the spray-chamber drain to prevent premature tubing destruction by PMA.

Fig. 1 displays the ICP emission spectrum in a region around the SI180.731nm emission line for different sulphur concentrations between 0.5 and 5mg/kg in PMA. It is clearly visible that the structured background spectrum from the PMA matrix is very stable, but also, due to its specific shape at the sulphur emission line position, both the sensitive determination via the SI180.731nm emission line or a reasonable positioning of background measurement points in the vicinity of this line is made difficult or even impossible.

Using now the advantages of the complete spectrum registration during each measurement, the stable PMA background spectrum can simply be subtracted, yielding the situation depicted in Fig. 2. The S emission line is now clearly discernible form the matrix spectrum, and background measurement positions can be chosen easily. Under these conditions, a limit of detection of 75mg/kg S in PMA could be achieved using the SI180.731nm emission line, an excellent value regarding the difficulties encountered with this matrix.

In conclusion, modern ICP-OES with simultaneous complete spectrum recording allows for an efficient and sensitive analysis of organic materials. The use of the complete spectral information available through this approach very often helps in reducing or removing problems caused by spectral matrix effects, which are encountered more frequently in the analysis of organic samples by axially viewed ICP-OES, compared to the inorganic case.

Enquiry No 45

Dirk Ardelt, George Glavin, Sergej Leikin, Henk Visser are with SPECTRO Analytical Instruments GmbH & Co KG, Kleve, Germany. www.spectro-ai.com

Recent Issues