Certificates for Chromatography Standards: Sense and Nonsense

March 1, 2010
Ruud J.P. Goedknegt

LCGC Europe

LCGC Europe, LCGC Europe-03-01-2010, Volume 23, Issue 3
Page Number: 148–150

Columns | <b>Column: Questions of Quality</b>

There is more to using and interpreting certificates for chromatography standards than meets the eye

Standards are as important in chromatography as columns, mobile phases, injectors and detectors.1 There's hardly an analytical laboratory to be found that does not stress the need for the certification of all components involved in an analysis. Quite remarkably though, standards are not always thought of as being an essential part of a certified analysis because many analysts prepare their own standards, while the other attributes are almost always purchased ready-made from a certified supplier. And while instrumental errors are often indicated by the instruments themselves and small deviations from the settings of, for example, temperature and pressure do not in most cases affect the results in a serious way, the user cannot equally easily recognize problems with the standards.

Problems with Manufacturer Instrument Certification

Before we look at the chromatography standards themselves, a small detour along the certification of the instruments may help to gain some insight into the pitfalls regarding certified analysis. All of the GC and HPLC manufacturers are ISO 9000 certified. Unfortunately, an ISO 9000 certificate does not at all guarantee that the instrument is suitable for the job that the analyst has in mind, although it does provide more certainty that any additional devices that are purchased will be of equal quality — or will have the same deviation.

To cite the manager of a large industrial analytical laboratory: "There is no obstacle to acquire an ISO 9000 certificate for producing life vests made from concrete."

The ISO Guide 17025 was designed for that reason.2 It states the criteria that the analyser meets and how this was tested. Still, this does not guarantee that the instrument will provide the correct answer to the analysis. A well-known example concerns the certification of the total organohalogen analyser (EOX, TOX). In this instrument, the sample is burnt to convert all organic halogen to yield HX, which is then automatically washed, dried and quantified via coulometric titration. One certified instrument that is currently on the market sometimes measures a mere 30% of the correct value, for example, for high-boiling PCB's, while at the same time other instruments, including a non-certified one, do provide reliable values for high boiling compounds that are difficult to burn completely. Why does the certificate not mean very much in this case? It is stated that the certified EOX-analyser has been tested using Aldrin (in accordance with the Dutch NEN 5735 Guide3). It so happens that Aldrin yields a result of 100% — even in the worst available analyser and is, therefore, not an adequate measure for the quality of an EOX-device. And so this was a poor choice on the part of the NEN-committee concerned. (The building sector, however, is pleased with this choice as it appreciates analysis reports that state low EOX-values in soil samples).

Another example of a certified analysis that gives a false sense of security concerns the test procedure that resulted in the heparin scandal. The Chinese manufacturer of this globally distributed anticoagulant illegally mixed this drug with over sulphated Chondroitin sulphate (OSCS) a much cheaper anticoagulant, but one with serious side effects. This was not recognized because of the prescribed test procedure.4

Both examples demonstrate that it is more important to do business with a bona fide and competent manufacturer than to overrate the value of a certificate that states that the product concerned passed the final inspection during the manufacturing or production. It remains a fine thing to doubt the reliability of a certified product, be it an instrument or a standard.

Standards for Certified Reference Materials (CRMs)

It is clear for these reasons that ISO Guide 9000 does not suffice for the manufacturing of CRMs (certified reference materials, more commonly described as standards and test mixtures). It is for that reason that additional Guides such as the ISO Guide 17025 referred to above and ISO Guides 31–35 have been set up.5–9

These ISO Guides require that a comparative chromatographic analysis be conducted as a means to inspect the preparation. And that is not all. The corresponding certificate also provides information regarding the shelf life of the standard. The producer must exclude the possibility of a reaction with dissolved air during storage or a reaction between the components themselves. This increases the reliability of the product.

Of course, a professional manufacturer of a mix-standard would not combine any components of which one can expect an inter-reaction based on theoretical grounds. Despite this, it is important to always check the deterioration of the composition in practice. ISO Guide 35 describes how the CRM producer comes to an expiration date for the standard via initial and on-going stability studies.

If an ISO Guide 31 certificate states that the shelf life of a product is three years, then one can conclude that an equal standard was made at least three years earlier and that it was recently subjected to a chromatographic analysis that showed that the composition had not changed.

Just how easy it is to handle CRMs in the wrong way is evident from the following case:

A 9000-2005 certified lab purchases a certified standard with 15 components, which is manufactured in series and is, therefore, much less expensive than a custom-made standard. However, a few months later the client actually needed a standard with one additional component and with the same (one year) shelf life. The client, therefore, bought a certified standard solution containing only this sixteenth component and then prepared by himself a new 16-component standard by mixing both original standards. The new standard is then filled in multiple ampoules, intended as a supply for the next year to come. He properly describes these actions in his management report and demonstrates such during the next audit.

Then the following happens: the accreditation inspector rejects the newly mixed standard because the remaining shelf life of the purchased 15-component CRM is less than the intended shelf life (one year) of the self-prepared mixture of the certified standards. The client asks the supplier for an updated certificate showing a longer shelf life for the 15-component mixture. This is a fine example of handling certified standards in the wrong way. The inspector should have said that the certified shelf life of any standard only applies up until the date when the ampoule is opened. If the composition is altered by him at a later stage, then he must store part of the newly mixed products to ascertain later whether the composition of the standards remains as expected in the course of the period in which it is used. It makes no sense to establish the expiration date of a custom-made standard until the product is manufactured. In other words, one cannot claim that the shelf life of a custom-made standard is one year until a period of one year has indeed passed and the appropriate checks have been made. When these (chromatographic) checks are successfully repeated after another year, the standard can be assigned a two-year shelf life, etc.

ISO Guide 34 extensively describes the stringent requirements that apply to a reference material producer. He is to store part of the manufactured product and inspect the quality after a certain time. This provides insight into the shelf life. In addition, the producer can then even re-certify previously supplied standards and, for example, specify a longer shelf life than initially specified on the original certificates.

Sometimes a supplier cannot guarantee the shelf life of a standard right away, because the standard was custom-made and had not been produced before.

The certificate supplied with the standard concerned stated "on-going stability programme" under the heading "shelf life". Some of these clients informed the supplier that this was not satisfactory to the inspector of the RvA (Dutch Accreditation Council) who demanded that the certificate state an exact shelf life. It appears that the inspector concerned was not aware of the fact that a standard that has been manufactured according to ISO Guide 348) for the first time is to be certified in this way. In view of the Accreditation Council's powerful position, the client has no idea how to cope with its demands. It would be to the credit of the Accreditation Council if it were to provide its staff with more adequate instructions for this situation.

Taking all of the above into consideration, one cannot but conclude that it is much safer to purchase standards from specialist suppliers, rather than produce the standards on one's own accord. The specialist sells the CRMs in large quantities, often worldwide. Should the product contain an error as a result of the delivery or for some other reason, then there will be a more than considerable chance that someone somewhere will discover the error. ISO Guide 33 even states that the CRM will be less expensive than compiling a standard on one's own, even if a CRM is very expensive — if one does the calculations right!

Employing and interpreting a certificate is not an easy thing to do. The subject should find a place in the chemistry curriculum at university and for QC-training for newly employed staff should pay ample attention to this issue.

Ruud Goedknegt was one of the first scientists in The Netherlands to research gas chromatography for food analysis in the 1950s. From 1980 until 2008 he was the owner and managing director of a wholesaler of standards and other GC/LC supplies. He was frequently contacted by customers because they encountered problems with accreditation inspectors when using custom-made standards. He is currently an independent consultant at Rugolution. E-mail: ruud@nessersluis.net "Questions of Quality" editor Bob McDowall is principal at McDowall Consulting, Bromley, Kent, UK. He is also a member of the LCGC Europe Editorial Board.

References

1. R. Boqué, A. Maroto and Y. Vander Heyden, LCGC Europe, 21(5), 264–267 (2008).

2. ISO/IEC Guide 17025 ( 2005): General Requirements for the Competence of Testing and Calibrating Laboratories. ISO, International Organization for Standardization.

3. NEN 5735, Soil quality — Quantitative determination of the content of extractable organo halogen compounds (EOX).

4. G.Somsen, CE of Contaminated Heparin, CASSS CE Pharm (Oct 2008).

5. ISO Guide 31 (2000), Reference materials — Contents of certificates and labels.

6. ISO Guide 32 (1997): Calibration in analytical chemistry and use of certified reference materials.

7. ISO Guide 33 (2000): Use of certified reference materials.

8. ISO Guide 34 (2009): General requirements for the competence of reference material producers.

9. ISO Guide 35 (2006): Reference materials, general and statistical principles for certification.