6

6.1 Apparatus

Standard and sample preparation

Standard and sample preparation

Analysts must recognize that the ICP-AES technique demands a particularly high standard of care in all activities which affect the accuracy and precision of the analytical result. Scrupulous cleanliness is essential in all laboratory procedures. Standards and samples must be meticulously prepared and carefully handled. Ideally, the laboratory should be air-conditioned, dust free and draught free. Unless the laboratory has been correctly prepared, satisfactory results may never be obtained with sensitive analytical techniques. Strict precautions must be taken to avoid contamination of apparatus, and even though laboratory ware is stored under ideal conditions, it should be thoroughly rewashed before use (see section 6.1). Strict care should be taken to avoid contamination of all reagents and distilled water. Ideally, reagents should be entirely free of the element of interest, but this is not possible for all analyte elements in all reagents (see sections 6.2, 6.3 and 6.4).

All laboratory apparatus should be scrupulously clean, including all beakers, watchglasses, pipettes and volumetric flasks(131-138). Even when clean glassware has been stored under ideal conditions it should be thoroughly rewashed before use. Use the following procedures for cleaning glass and plastic apparatus. Fill the vessel with 20% v/v hydrochloric acid and leave for about 2 days. Include the plastic lid in the leaching and cleaning process. Discard the acid solution, and fill the vessel with 20% v/v nitric acid and leave for a further 2 days. Again, include the plastic lid in this leaching and cleaning process. Discard the nitric acid, and rinse the vessel thoroughly with distilled water. The vessel can then be filled with distilled water and left until required, or it can be thoroughly air dried. Always ensure that airborne particles cannot fall into the vessel prior to use.

Publication date: 10/99

6-1

Standard and sample preparation

6.2

Water
Water is the most commonly used reagent for both sample dissolution and dilution, and steps must be taken to ensure it is contamination free. Distilled water is usually contaminated with dissolved gases or organic material. Many investigators(131,132) have found that when metallic or Pyrex vessels are used in the distillation of water, the water produced generally contains ionic impurities. A mixed bed ion exchange system must be used to remove these impurities. The American Society for Testing and Materials (ASTM)(133,134) has graded various types of reagent water. Combinations of ion-exchange, reverse osmosis, distillation, filtering or electrodialysis techniques are used to produce each type of water. Type 1 is suitable for trace analysis and is defined as water prepared by the distillation of feed water (having a maximum conductivity of 20 m/cm at 25 oC) followed by polishing with a mixed bed of ion-exchange materials and a 0.2 m membrane filter. Ideally, any hard waters should be pretreated by a water softening process to remove any calcium and magnesium that may form scale in the distillation unit, then passed through a carbon filtration system to remove any organics. After slow distillation the water should be slowly passed through a mixed bed ion exchange system. Some of the commercially available purification systems that can be used for the production of high purity water are the Millipore Super-Q System(135,139) and a system produced by the Continental Water Conditioning Corporation. The Millipore system consists of disposable cartridges for each purification step. Frequent monitoring to ensure optimum performance is required regardless of the system chosen. A conductivity meter placed between the still and the storage vessel will indicate whether or not the system is working at maximum effectiveness.

6.3

Reagents
Always use the purest grade reagents available for all trace analyses. Reagent manufacturers usually indicate the contaminants and specify the levels. Use these as a guide, but always check the levels prior to the analysis. Even the highest grade available may not be pure enough for particularly stringent analyses, and further purification will then be necessary. Kuehner et al(140) have developed procedures for purifying mineral acids including hydrofluoric acid. A quartz sub-boiling still is used for the production of HCl, HNO3, HClO4, H2SO4 and H2O. Infra-red radiators heat the liquid being distilled, evaporating it without boiling.

6-2

Publication date: 10/99

Standard and sample preparation

The condenser is angled so that the distillate can drip to the outlet more efficiently. A liquid level control maintains the liquid to just below the overflow height, ensuring a constant volume of 400-500 mL of liquid in the still. Hydrofluoric acid is purified by distillation from a PTFE sub-boiling still. The PTFE still operates in a very similar manner to the quartz still, but modifications must be made to allow for the different surface tension of the PTFE. The operating temperature is also restricted by the softening properties of PTFE. Other reagents used must also be checked for impurities and purified if necessary(141). This could involve recrystallization if the reagent is a solid, or distillation if the reagent is a liquid. The vacuum sublimation of various metal chelates has been studied by Honjo et al(142). Many reagents can be purified using this technique. Even after reagents have been purified, it is still essential to prepare blank solutions for each analysis. This is to ensure that the results obtained for the analysis are in no way affected by residual contamination.

6.4

Blank
A blank solution is defined as a solution containing all the reagents to be added to the sample but excluding the sample itself. Contamination of the water and reagents used in the preparation of standards and samples can cause a high blank signal intensity. Always check for contamination in the blank solution. A consistent blank reading must be obtained before analytical data can be taken.

6.5

Standard preparation
Three sources of calibration standards are available: (a) Commercially-prepared ICP standards, usually with a concentration of 10 000 g/mL of the element of interest (b) Commercially-prepared multi-element synthetic standards (c) Standards prepared in the laboratory by the analyst from the pure metal or pure metallic salts. Each one is suitable for the ICP-AES technique. These standards are usually stable for many months, especially when a stabilizer has been included in the initial preparation. For most elements the stabilizer is a mineral acid, usually hydrochloric or nitric acid. For some analyses, matrix matching of the standard and sample is required in order to minimize interferences.

Publication date: 10/99

6-3

Standard and sample preparation

Fresh working standards should be prepared daily to avoid a loss of the analyte on the container wall or leaching out of the element from the container. It is good practice to perform serial dilutions rather than make one large dilution. For example, to perform a dilution of 1 in 10,000 make two separate dilutions: 1 in 100 and a further 1 in 100. The errors in the latter procedure will be much smaller than dilutions of, for example, 0.1 mL in 1 L to make a 1 in 10,000 dilution. To further decrease these errors use only grade A glassware. Check the volume delivered by weighing an aliquot of water. If, for example, a 10 mL volumetric pipette is being used, the mass delivered should be 10.0 g (since the density of water is 1.0 g/mL). If the mass delivered is incorrect, then appropriate volumetric corrections will be needed. Ambient temperature must be taken into account in this calculation, because the density of water varies with temperature. By taking these precautions the error involved in the dilution steps should be minimized. For information about calibration and selection of standards, refer to section 6.8.

6.6

Sample preparation
Each type of sample to be analyzed by ICP-AES requires a specific preparation according to its matrix. Sample treatments for specific types of samples are given in chapter 7. It is essential to have a homogenous sample solution. Nonhomogeneous samples lead to inaccuracies in the analysis. ICP-AES is primarily a solution technique, and liquid samples may require filtration, centrifugation etc. Dry or wet ashing procedures are commonly used to bring solid samples into solution. Although solutions are preferred, different sample forms such as slurries or solids can also be introduced into the ICP. For slurries, the samples must be finely ground (<10 m size) and suspended in an appropriate aqueous or organic solvent. Matrix matched standards must be used. Direct introduction of solid samples into the plasma is feasible with the techniques of direct insertion, electrothermal vaporization, spark ablation and laser ablation.

6.7

Storage
6.7.1 Samples

Samples should be analyzed as soon as possible after collection and preparation. This will minimize any errors due to contamination from the container or loss of analyte on to the container walls. Avoid using glassware to store samples since some metals may be adsorbed on to the glass surface. Transfer the sample to polyethylene containers as soon as the dissolution procedure has been completed.

6-4

Publication date: 10/99

Standard and sample preparation

To minimize any contamination or loss of analyte: (i) Clean containers carefully (as previously described). (ii) Analyze the samples on the same day as they are prepared. (iii) Where possible, acidify the sample to about pH 2 or less.

6.7.2

Standards

Always ensure that the containers used for storage have been thoroughly cleaned using the procedures previously described. As a general rule standards having a concentration of 1000 g/mL are stable for months when acidified to pH 2 or less. These can be stored in polyethylene bottles, tightly capped. Always make up fresh standards for each analysis to ensure accuracy. When the analyte concentration is less than 1 g/mL, the solution will usually be much less stable. The useful life of the solution will depend on the element and the standard matrix. Where possible, acidify with a mineral acid to pH 2 or less. Note that for safety reasons acids must not be added to cyanide standards. Photographic waste products often contain cyanide complexes, and acids should not be added to stabilize the standards. Remember: (i) Standards should be prepared immediately prior to the analysis. (ii) The analyst should be aware of the standard matrix before adding any chemical designed to enhance the stability of the element in solution. (iii) Containers must be meticulously cleaned.

6.8

Calibration
This is the process by which the relationship between the net signal intensity and elemental concentration is determined for a specific element determination. Calibration standards can be prepared either for a single element or for several elements. Multi-element standards are particularly convenient and are prepared by combining appropriate volumes of the stock solutions of the elements of interest.

Publication date: 10/99

6-5

Standard and sample preparation

6.8.1

Multi-element standards

Prior to the preparation of multi-element standards, each stock solution should be analyzed separately to determine possible spectral interferences or the presence of impurities. It is necessary to consider the type of salt (e.g. chloride, nitrate, sulfate, phosphate etc.) and mineral acid in each individual standard due to the possibility of formation of insoluble compounds upon mixing. However, the formation of many insoluble compounds is pH dependent and, with the exception of some sulfates and chromates, an acidic medium generally enhances stability. Therefore, for multi-element calibration standards, it is essential to group elements that are chemically compatible so that they will remain in solution and hence provide stable calibration standards. Calibration standards should be initially verified using the quality control standards and monitored weekly for stability. A complete set of calibration standards should be used to define the calibration accurately over the specified concentration range. According to the USEPA standard preparation specification(155), some elements are required to be grouped together to ensure chemical compatibility. They are: Group 1: Mn, Be, Cd, Pb, Zn Group 2: Ba, Cu, Fe, V, Co Group 3: Mo, Si, As, Se Group 4: Ca, Na, K, Al, Cr, Ni Group 5: Sb, B, Mg, Si, Tl

6.8.2

Calibration graph

An ICP-AES calibration has the feature of linearity over a wide concentration range, typically 4 to 6 orders of magnitude above the detection limit. Earlier workers(143-145) have reported that the calibration graphs are slightly curved in the upper concentration range, and that substantial differences in the extent of line curvature are obtained for different elements. They examined the relationship between the concentration and the standard deviation of the resulting net line intensity by applying least squares procedures for calibration. The results showed that the standard deviation of the emission signal is constant for the first three orders of magnitude, after which it increases in proportion to the signal intensity. Best results were obtained by use of a linear scale for the lower concentration and logarithmic scale for the upper concentration. Possible explanations given for the curvature include

6-6

Publication date: 10/99

Standard and sample preparation

self absorption, erroneous background correction, or non-linear response of detector and/or readout system(146). Most users are aware that ICP-AES calibration graphs are not completely linear. In practice, calibrations are close to linearity up to a concentration of 1000 g/mL, while above that concentration, curvature is a significant factor. Therefore, dilution of the sample could be required for analytical accuracy. On the other hand, modern computer-based procedures can easily accommodate the slight nonlinearity of the calibration.

6.9

Internal standards
Several workers(147-154) have studied the use of internal standards in ICP systems. Internal standardization is a method of improving analytical precision and accuracy by correcting and compensating for the effects of noise and intensity changes on the measurement. The procedure involves calculating the ratio of the analyte emission intensity to that of another element present in fixed or known concentration. Therefore, when an internal standard is used, the calculation of concentration is based on the ratio of emission intensity of the analyte to that of the internal standard. It is important to add a fixed volume of internal standard solution to every analytical solution to ensure the ratio of the original mass of sample to the mass of reference element is a constant. The internal standardization procedure has a number of advantages over direct measurement of the emission intensity whereby fluctuations of the excitation conditions within the source are compensated by proportional fluctuations in the analyte and internal standard emission intensity. Internal standardization also compensates for variations in the quantity of sample being presented for analysis or for variations in the rate of analyte introduction to the source. Several criteria should be considered when selecting the internal standard element. They are: 1. 2. 3. 4. The internal standard material must be pure. The internal standard must not be present in the analyzed sample. There must be no spectral interferences by the internal standard element on the available analyte wavelength. The internal standard solution must be chemically compatible with the sample solution.

Publication date: 10/99

6-7

Standard and sample preparation

5.

The excitation potential of the analyte emission line should be matched to that of the internal standard element emission line. An atomic internal standard line is best for atomic analytical lines, and an ionic internal standard line is best for ionic analytical lines. For example, the Fe II 259.940 nm line is a good selection as an internal standard for Mn II 257.610 nm; however, it is not appropriate as an internal standard for Cu I 324.754 nm because of the mismatch of excitation potentials.

6-8

Publication date: 10/99