Guidelines for Achieving High Accuracy in Isotope Dilution Mass Spectronometry (IDMS)

1 Introduction

The technique of isotope dilution mass spectrometry (IDMS) was initially developed during the 1950s for elemental analysis. With readily available and user-friendly instrumentation, the number of applications for which elemental IDMS was used developed rapidly. IDMS was extended into the field of organic compounds in the 1970s and the range of applications for which isotopically labelled organic compounds are available continues to grow. IDMS applications range from routine survey work, such as residue analysis of dioxins, to use as a reference technique.

In mass spectrometry, unlike spectrophotometric techniques, there is no fixed relationship between the amount, and concentration, of a particular substance and the instrument response. Sensitivity for a particular compound or ion varies with, for example, time and instrumental tune parameters. These variations are in addition to those caused by, for example, sample introduction and chromatography. To achieve even moderately accurate quantification requires the use of an internal standard. An advantage of mass spectrometry lies in its ability to use isotopically enriched analogues (inorganic mass spectrometry) or isotopically labelled analogues (organic mass spectrometry) as internal standards. Indeed, in some instances IDMS is referred to as 'stable isotope internal standardisation'. Provided the isotopic analogue is added to the sample at the very beginning of the analytical method and it comes into equilibrium with the analyte without losses or isotopic fractionation, it enables exact compensation to be made for errors at all stages of the analysis, from sample digestion/preparation through to the final instrumental measurement.

Initially, IDMS of inorganic analytes was most frequently performed using thermal ionisation mass spectrometry (TIMS). More recently, inductively coupled plasma mass spectrometry (ICP-MS) based IDMS has become more prevalent, because it requires much less sample preparation prior to analysis, and can provide results of the required accuracy and precision. The inorganic sections of this guide are concerned only with the use of ICP-MS for high accuracy IDMS measurements. IDMS of organic analytes can equally well be carried out using the full range of sample inlet systems available, such as GC-MS, LC-MS and the newer technique of capillary zone electrophoresis-mass spectrometry (CZE-MS). Tandem mass spectrometers (MS-MS) can also be used. In some instances, however, the inlet system/mass spectrometer type can have an influence on IDMS analysis. The use of a GC split injector, for example, can cause isotopic fractionation.

The advent of compact and economic instrumentation, such as quadrupole and ion trap mass spectrometers, has led to the increasing use of mass spectrometry in the field of analytical chemistry. As a result, IDMS is playing an increasingly important role in trace analysis. This is also due to its greater accuracy than other calibration methods and its ability to compensate for matrix effects. Notwithstanding this, there are several reasons why, in spite of the unrivalled accuracy and precision possible with the IDMS technique, it is not more widely used. In principle, the method is simple and allows for knowledge or control of all the variables that can lead to error. In practice, achieving accurate results requires careful design of the experiment and considerable attention to detail and is hence quite time consuming. Such factors have led to a slow acceptance of the technique. It has become popular, however, in some analytical fields where the sample matrix makes sufficiently accurate quantification difficult, e.g. clinical analysis. A theoretical discussion on the influence of some instrumental parameters on the precision of IDMS measurements has been published.

1.1 Advantages and Disadvantages of IDMS

The use of IDMS has a number of advantages and disadvantages, which the prospective user should consider.

1.1.1 Advantages

Providing the sample is homogeneous and isotopic equilibrium has been reached, the advantages of IDMS include:

1. It is a definitive method because of its precision, accuracy and provision of definable uncertainty values.

2. Once equilibration of the spike and analyte isotopes has been achieved, the total recovery of the analyte is not required, because the determined value is based on measuring the ratio between the analyte and the isotopic analogue (spike).

3. The accuracy of the method is determined by the precision of this ratio measurement.

4. Analyte transformation (e.g. breakdown) during sample preparation (most applicable to organic analysis) is compensated for by use of the isotopic analogue as internal standard.

5. The methodology is less time consuming and can provide greater accuracy than standard additions.

This guide aims to provide a structured approach to the use of IDMS to achieve high accuracy analytical measurements in both organic and inorganic IDMS. The approach outlined in Sections 4 and 5 has the advantage over normal IDMS of:

1. Negating the need to accurately characterise the isotopic analogue in terms of isotopic abundance and concentration (most applicable to inorganic analysis).

2. Providing a logical framework of understanding, which allows the analyst to identify specific problems, related to their application.

3. Providing a lower uncertainty value.

4. Simplifying the equations used for organic analytes.

1.1.2 Disadvantages of IDMS

1. The cost and availability of suitable isotopic materials.

2. The cost of the mass spectrometry instrumentation required.

3. Training of the analyst is of the utmost importance; otherwise less accurate results are often achieved.

4. Isotopic equilibration needs to be shown to have been achieved.

5. Differences in the physical (e.g. solvation) and chemical properties (e.g. pKa value) between the analyte and the isotopic analogue, can affect the ions generated in the mass spectrometer (most applicable to organic analysis).

2 The Principles of IDMS

The underlying principle of IDMS5 is that an isotopically enriched analogue (inorganic MS) or an isotopically labelled analogue (organic MS) of the analyte compound is used as an internal standard in quantification by mass spectrometry. An accurately known amount of the isotopic analogue is added to the sample. The consequent ratio of the amounts of the two isotopes (one resulting from the analyte and the other from the isotopic 'spike') is measured on a portion of the sample using a mass spectrometer, so enabling the analyte concentration to be calculated. In summary the stages are:

1. Characterisation of the isotopic analogue (see Section 3.5) using a traceable natural standard by 'reverse IDMS'. If a certified isotopic analogue is used this information is provided in the certificate. This step applies primarily to inorganic analytes.

2. An accurately known amount of the isotopic analogue is then added to an accurately measured portion of the sample (see Section 3.4). This step is widely referred to as 'spiking' the sample.

3. The mixture is equilibrated for an appropriate time. For inorganic analysis destructive digestion is necessary for solid samples (see Section 3.2.1). With organic analysis sample preparation often involves extraction and purification steps. In some cases a chemical reaction is necessary to free any bound form of the analyte (e.g. conjugates) present in the sample. For analysis by gas chromatography a derivatisation step may be necessary to confer thermal stability to the analyte and spike compounds.

4. An aliquot of the spiked sample is introduced into the mass spectrometer either directly (inorganic MS) or after chromatographic separation from other compounds (organic MS) (see Section 3.7).

5. The ratio of the signal responses for the ions resulting from the mixture of the analyte and isotopic analogue is accurately measured in the spiked sample using a mass spectrometer (see Section 3.8). From this ratio the concentration of analyte is calculated by comparison with the same ratio measured for the same ions in a standard calibration mixture. The calibration mixture contains the spike used in the sample and a certified standard of the analyte.

6. Corrections need to be made for instrumental effects, such as mass bias and detector dead-time (see Section 3.10). Using the signal matching procedures outlined in this guide this can be achieved by running an alternating sequence of standards and samples.

7. Blank analysis is still required when using IDMS because any isotopic contribution to the mixture (from reagents, contamination etc.) will affect the isotopic abundance ratio and ultimately lead to a systematic bias (see Section 3.6).

3 Critical Stages and Sources of Error

3.1 Introduction

IDMS analysis can provide results of great accuracy, but to achieve that accuracy it is necessary to consider the critical stages in the experimental procedure and to understand the potential sources of error. All the general precautions associated with mass spectrometry and trace analysis apply equally well to IDMS based procedures. Issues such as sampling, sample homogeneity, contamination and losses prior to analysis will play their part in influencing the accuracy of the reported result. The purpose of this guide is to highlight the specific points pertinent to the use of IDMS for accurate measurement. Therefore, issues related to sampling etc. will not be directly addressed in this guide. For guidance on general trace analysis the following references may serve as a starting point. A useful summary of critical parameters in ICP-MS instrumentation has been published.

There are a number of critical stages in the IDMS procedure where errors may occur including:

1. Sample preparation

(a) Sample digestion and isotopic equilibration (inorganic analysis)

(b) Extraction, clean up, derivatisation (GC-MS) and isotopic equilibration (organic analysis)

2. Selection of the most appropriate isotopic internal standard (often termed the 'spike' and in this guide generically referred to as the isotopic analogue)

(a) An isotopically enriched analogue (inorganic analysis)

(b) An isotopically labelled analogue (organic analysis)

3. Addition of the isotopic analogue

4. Characterisation of the isotopic analogue

5. Blank correction

6. Instrumental analysis

7. Calibration procedure

8. Calculation of the result

9. Estimation of the uncertainty

There are a number of important sources of error in the IDMS procedure including:

1. Less than complete isotopic equilibration will lead to significant systematic errors 2. Presence of interfering ions

(a) Isobaric and polyatomic interferences (inorganic MS)

(b) Natural analyte ion signal overlap with isotopic analogue ion (organic MS)

3. Isotopic discrimination (mass fractionation, detector dead-time, mass bias)

It should be noted that errors occurring during sub-sampling, addition of the isotopic analogue, sample preparation and isotopic equilibration are not compensated for by the use of isotope dilution analysis. However, these errors can be assessed by analysis of a suitable certified reference material or an 'in-house' prepared standard, analysed alongside the sample.

The critical stages and sources of error summarised above are dealt with in greater detail in the following sections.

3.2 Sample Preparation

With IDMS it is particularly important that full equilibration between the analyte and the isotopic analogue is achieved. This will ensure identical behaviour during the analytical procedure. Sufficient time for equilibration must be allowed but, even so, it is often difficult to ensure that full equilibration has taken place. Particular attention must also be paid, for example, to the chemical forms of the analyte and the isotopic analogue, e.g. oxidation state (inorganic MS), analyte form present (organic MS) etc.

3.2.1 Inorganic IDMS

It is advantageous to add the isotopically enriched analogue at the earliest point of the analysis. For solid samples, there will be a period during the digestion phase when the analyte is in a different form from the added enriched isotope. An example of this is the analysis of trace metals in glass. The natural metal has to be released from its silicate matrix to achieve dissolution. During this period the analyte isotope may behave differently from the added analogue isotope before true equilibration is achieved. Thus, losses due to incomplete dissolution or volatile species generation can result in errors, which are unaccounted for by the IDMS procedure. Also, any contamination during the sample digestion stage will add to the final error in an IDMS analysis (this is easily corrected for by analysis of a suitable blank solution). It is essential to ensure that both isotopes are in the same chemical form. This can be achieved, for example, by an oxidation or reduction step.

3.2.2 Organic IDMS

Analyte extraction, sample clean-up and derivatisation steps may also introduce errors prior to the benefits of the isotopic spike being realised. A good example of this is the analysis of cholesterol in serum using IDMS with GC-MS measurements. Even though the serum can be spiked at an early stage with the isotopically labelled analogue this does not guarantee equilibration. In fact, the natural cholesterol is mainly bound to proteins or other lipid materials. Thus, a saponification step is required before the two forms of cholesterol are equivalent and approach equilibration. Even vigorous shaking of the unspiked sample is to be avoided, as there is evidence that foaming is linked to unwanted conversion of the indigenous cholesterol. However, it is essential that all the natural cholesterol is released, otherwise the subsequent derivatisation step for GC-MS analysis will introduce a bias between the isotopically labelled analogue and the natural form.

3.3 Selection of the Isotopic Analogue

For organic analytes, selection of the isotopically labelled analogue may have an effect on the analysis. If deuterium (D) is labelled onto an OH group, for example, then isotopic exchange may take place. Consequently the isotopic label should be introduced into a stable position of the analogue. To avoid discrimination during the various steps of the analytical procedure the label should not be in a position that affects solvation properties, pKa values or derivatisation kinetics (where applicable). When using chemical ionisation GC-MS or atmospheric pressure chemical ionisation LC-MS, if ion formation differs between the isotopic analogue and the analyte, a significant error could occur.

For inorganic analytes there are usually a limited number of options when selecting an isotopically enriched analogue for use in a given IDMS analysis. Where there is a choice, consideration should be given to possible isobaric or polyatomic interferences during the ICP-MS measurement that might be linked with a particular sample (e.g. cadmium in the presence of tin because of the isobaric interferences at m/z values of 112, 114 and 116).

3.4 Addition of the Isotopic Analogue

In general, to achieve an isotopic ion abundance ratio measurement with the best precision, the ratio of analyte concentration to isotopic analogue concentration should be such as to produce equal ion abundances. The solution for analysis should, therefore, contain similar amounts of the two isotopes. However, if measurements are to be made near the detection limit of the analyte, it is often better to use more isotopic analogue so that this response at least can be measured with greater precision.

This approach may be inappropriate for inorganic IDMS, where other situations may apply. In the case of silver, its two natural isotopes actually exist in proportions close to 1:1 (51.8% and 48.2%). Thus, the optimum spiking regime requires a final concentration ratio of ~1:4 to minimise the propagation of errors. A fuller explanation of this particular issue is given in Section 5.

In the case of organic analytes there may be some circumstances where the isotopic analogue has only a small mass difference from the natural analyte, for example, if only one carbon or hydrogen atom has been labelled. In such cases, where the isotopic analogue has a mass of only (M + 1), the analyte itself may have a significant (M + 1) abundance due to the natural isotopic composition of the analyte (contribution from 13C or 2H in the analyte). Under these circumstances a large addition (say 5 to 1) of the isotopic analogue compared to the analyte will reduce the relative concentration of the natural (M + 1) abundance compared to the (M + 1) abundance arising from the isotopic analogue. This will reduce non-linear effects arising from the natural (M + 1) abundance, and can also be particularly important where the bracketing or single point method of calibration is used (see Section 3.8). The relative molecular mass of the analogue should, wherever possible, be increased by at least three mass units, which should avoid interference of the natural isotopes of the analyte on the isotope being measured for the labelled analogue.