Interference-Free Drinking Water Analysis Using ICP-OES

November 1, 2010

Special Issues

Volume 0, Issue 0

With ICP emission spectrometry, many spectral lines are emitted for each element. The fact that spectral lines for samples containing several elements can overlap is well known as spectral interference. For this reason, it is necessary to use a spectrometer with a resolution over a certain level. Even then, spectral interference might be possible. This article describes a way to measure water samples using inductively coupled plasma–optical emission spectroscopy using a simultaneous instrument with a CCD detector and a software package that incorporates the knowledge of experienced analysts as a database, simplifying the selection and confirmation of wavelengths, to allow high precision and interference-free analytical results.

With ICP emission spectrometry, many spectral lines are emitted for each element. The fact that spectral lines for samples containing several elements can overlap is well known as spectral interference. For this reason, it is necessary to use a spectrometer with a resolution over a certain level. Even then, spectral interference might be possible. This article describes a way to measure water samples using inductively coupled plasma-optical emission spectroscopy using a simultaneous instrument with a CCD detector and a software package that incorporates the knowledge of experienced analysts as a database, simplifying the selection and confirmation of wavelengths, to allow high precision and interference-free analytical results.

It is estimated that more than 100,000 chemical substances are contaminating our drinking water every day. As water is the basis for a healthy life, strict control as well as a sophisticated and reliable water purification and supply is a crucial condition for health and the prevention of epidemics all over the world.

A wealth of analytical methods exists for the quality control of drinking water and wastewater. The method and type of instrumentation employed is dictated by the composition and formulation of the samples to be measured. Spectroscopy is the preferred method if quick and simple information about the sample material is required, both qualitatively and quantitatively. The optimum spectroscopic method is determined by the sample characteristics.

Atomic spectroscopy applying atomic absorption spectrometers and inductively coupled plasma–optical emission spectrometry (ICP-OES) instruments is used widely for the accurate determination of major, minor, and trace elements in aqueous solutions. Experimental data have been generated using international reference material as well as real life samples from various locations. Performance of different system configurations has been evaluated, and optimized methods have been prepared for achieving the highest sensitivity and lowest detection limits.

Multielement analysis of aqueous solutions in compliance with international drinking water regulations is one of the main application areas of ICP-OES.

Experimental

The system used in this study (ICPE-9000, Shimadzu Europa GmbH, Germany) is suitable for axial and radial plasma observation. Its high-performance echelle optics with special "Schmidt mirror" enable the effective use of the entire 1024 × 1024 pixel CCD detector area. In this way, a resolution of higher than 0.005 nm is attained over the entire wavelength range of 167–800 nm. The detector, which has an antiblooming function, reliably acquires signal intensities, even at long exposure times. All samples can therefore be determined accurately within one single analysis sequence, including samples with very different element concentrations. The "reprocessing" function of the software (ICPEsolution, Shimadzu) enables the determination of additional elements or changing the concentration range for alternate wavelengths without the need for new measurements.

The vacuum optics combined with mini torch technology considerably reduce argon gas consumption. The torch used here reduces the argon gas consumption by half that of conventional torches without loss in sensitivity. In addition, time-consuming rinsing of the optics with ultrapure gas is no longer necessary. The system is ready for operation and stable within the shortest possible time. Equipped with the optional autosampler, the system can be fully automated for high sample throughput operation. The system status is monitored continuously and can be retrieved at any time, whenever needed.

Method development and data analysis software: The software described earlier offers users additional support with two integrated assistant functions. The "Development Assistant" creates complex calibrations, from the selection of optimal wavelengths up to the composition of standard concentrations. In combination with the "Monitor Function" for qualitative analysis, this assistant points to possible interference problems or incorrect wavelength selection before the actual calibration takes place, and displays solutions for method modification or error correction. The "Diagnosis Assistant" evaluates data already measured, and compares this with information from various databases. Data evaluation and recalculation have never been more straightforward, as the complete emission spectrum of a sample is continuously available.

Table I: ICPE-9000 system parameters for water analysis

Analysis of Drinking Water and Mineral Water Samples

Various accessories are available for sample introduction. Aqueous samples, for example drinking water, are introduced into the cyclone spray chamber via a coaxial nebulizer and transferred subsequently into the mini torch via the carrier gas flow of 0.7 mL/min. The system parameters are shown in Table I. Under these conditions, a series of different drinking water and mineral water samples has been analyzed for main and trace elements. A reference material (NIST SRM 1640) was measured along with the series as a control sample. The excellent recoveries are shown in Table II.

Table II: Comparison of the certified and measured concentrations of the NIST SRM 1640 control sample

Interferences in ICP-OES

Interferences in ICP-OES are typically cumulative and lead, in practice, to higher measurement results with neighboring emission lines of different elements typically contributing to the signal. In the wavelength range of 200–400 nm, there are more than 200,000 spectral lines, making manual selection of the most suitable element line in the desired concentration range very difficult. In this way, the savings in time offered by simultaneous ICP spectrometers for multielement analysis in comparison with sequential systems is partially reduced. The selection of suitable lines for measurement is time-consuming and, accordingly, requires time-consuming postprocessing of all analytical data after measurement.

The system places special emphasis on the actual advantages of such a system: high sample throughput over a relatively narrow time window.

Wavelength Selection

In sequential systems, measurements are usually carried out at the most sensitive wavelengths where spectral interferences are negligibly small if the analyte concentration is sufficiently high. In simultaneous ICP systems, the selection of the optimal wavelength is far more important, as the resolution of these systems is lower compared with that offered by high-resolution sequential spectrometers.

For systems with a CCD detector, one should also take into account that for high element concentrations, the most sensitive wavelength reaching the detector can lead to overexposure. This means that, at high concentrations, a less sensitive line can be more advantageous, which on the other hand, can be more susceptible to spectral interferences. Therefore, for analytical routine measurements it is vitally important to minimize spectral interferences for samples that are to be measured over a wide concentration range.

Automatic Selection of Correct Wavelengths

The system described in this article enables qualitative analysis of unknown samples. In this way, it is possible to obtain an overview of all main and trace elements of the sample and, based upon the line profile, to select wavelengths that exhibit suitable intensities and that are also free from spectral interferences.

The selection of optimal lines can be carried out manually. However, wavelength selection via the software described here is easier and faster. The software offers a function that supports method development and selects the optimal line for each element. In addition, it suggests the calibration range for the elements. In this way, a quantitative method can be created from a qualitative method that, with corresponding calibration standards, is tuned exactly to the samples to be measured.

For very high sample concentrations causing overexposure of the corresponding lines on the detector, an alternative line can be selected retroactively from the spectrum. In this way, ICP systems with CCD detectors offer the possibility to select any wavelength retroactively to be used for postcalculation of the measuring results without the need to reanalyze the sample. This is the fundamental advantage with respect to ICP spectrometers with CID or segmented CCD detectors that allow only a limited selection of wavelengths to be added postmeasurement.

Interference Correction Standard

Figure 1 shows the line profile of a typical spectral interference of two neighboring element lines of cadmium and iron. The higher the iron concentration, the stronger the cadmium signal will be affected by the additive interference.

Figure 1: Line profile of a spectral interference.

The detector records the cadmium signal at wavelength 226.502 nm as well as that of iron at wavelength 226.505 nm. The intensities of these two elements cannot be separated and an interference correction therefore needs to be carried out. The software selects an additional interference correction standard measured before the calibration standards, which compensates for spectral interferences.

When a pure iron solution is measured using the cadmium wavelength of 226.502 nm, the detector will nevertheless record a signal generated only by interference of iron. This problem can be solved easily using an interference correction standard of known concentration, whereby a correction factor using different wavelengths is calculated for the intensities as described in Figure 2 for cadmium and iron.

Figure 2: Use of an interference correction standard.

The software contains all element-specific wavelength information and correlative interferences as well as the associated intensities in a database and, therefore, enables reliable analyses of complex samples, compensating for any possible spectral interference.

Uwe Oppermann is with Shimadzu Europa GmbH, Duisburg, Germany. Jürgen Schram is with Hochschule Niederrhein, University of Applied Sciences, Krefeld, Germany.