What to consider when creating or using certified reference materials.
Some analysts will look at a list of pesticides or solvents and visualize adding small amounts of pure material of each analyte into a volumetric flask and then bringing the flask up to volume with their solvent of choice. In rare instances this technique will work, but not often. Prior to making a mixture, factors such as solubility and reactivity must be considered. Are all of the analytes soluble in the solvent that I want to use? Do I need to use an intermediate solvent? Will the analytes react with each other or will they react with the solvents? Do I need to add a preservative? Which preservative should I use and will any of the analytes react negatively to the preservative? The answers to these questions come from experience and a working knowledge of the chemistry of reference material (RM) production. Manufacturing the standard mixture is only the first step in creating a certified reference material (CRM). The CRM must be characterized using a metrologically valid procedure. A certificate of analysis is produced listing certified values for specified properties and all analytes, including a calculation of total uncertainty, homogeneity, stability under specified conditions, and metrological traceability. The certificate of analysis is delivered with the standard.
If you work in an analytical chemistry laboratory you may have heard the terms certified reference material (CRM) and reference material (RM). An RM typically comes with a certificate of composition (COC) that states the identity of the product, the purity of the product, and the details of its characterization. In addition to the information found in the RM COC, the CRM certificate of analysis (COA) declares that the material characteristics were determined by a metrologically valid procedure for the specified properties and contains a statement of metrological traceability, homogeneity data, long term stability results, and all factors that determine the CRMs total combined uncertainty, to name a few. This uncertainty includes the measured uncertainty inherently found in all measurement tools used in the manufacturing process as well as the effects of storage, transportation conditions, and possible chemical interactions affecting long term stability.
CRMs, as opposed to RMs, give labs a competitive advantage and peace-of-mind. ISO 17025 recommends that accredited labs use CRMs whenever available. As a laboratory, you can trust that everything has been done to provide you with a high quality standard that helps to ensure that your data can be generated correctly the first time, and every time. After all, the quality of your data can directly affect consumer safety. Because the use of CRMs helps to ensure data quality, your services are of greater value to your clients and prospective clients.
All CRM manufacturers strive for the lowest possible combined uncertainty during the manufacturing process. There are some variables that cannot be precisely controlled during the manufacturing process such as the limitations of Class A glassware (typically, the uncertainty at a specific temperature is written on the glassware) and syringes, precise room temperature, and the uncertainty inherent in all analytical balances despite calibration with National Institute of Standards and Technology (NIST) certified weights. The most challenging uncertainty to calculate is the uncertainty obtained from possible analyte to analyte and analyte to solvent interactions.
The first skill required when making a stock standard from a neat material is knowing which solvent is miscible or soluble with the analyte and if the analyte and solvent will quickly interact under the given storage conditions. If your working standard will be in a solvent different from your stock solvent, then the stock’s solvent is sometimes known as an intermediate. For example, if you want to spike used motor oil into water, you know that oil and water aren’t miscible. If you first dilute your oil with isopropanol at a reasonable concentration, the oil–isopropanol solution can be successfully spiked into the water matrix and you will have the used motor oil in solution with your water. The isopropanol is your intermediate solvent.
You have created all of your individual stock standards and now you want to combine portions of each stock to make your working standard mixture. Again, you need to ask, are the solvents miscible and will my analytes stay in solution? Will that solution be homogenous? Will the analytes quickly interact with each other or with the solvents? To answer these questions you need experience. If you have done this before you will know what is going to happen. If this is your first time making this mixture, you won’t know for sure until you have finished the mixture and some time has passed. I will give you a hint: All organic chemicals interact as do many metals. To minimize your interaction and increase the shelf life, you need to know the correct storage conditions (colder is not always better) and which preservative, and how much of it, to add to your mixture. Storing a standard at too low of a temperature can negatively affect homogeneity and possibly cause some analytes to precipitate out of solution. It’s usually a good idea to sonicate or shake mixtures at room temperature before using, especially if they are cloudy or if you see solids floating in the solvent.
Many of you are probably now asking yourselves, what is a preservative? Analyte–solvent interaction can impact the pH of your solution and redox potential over time. Acetonitrile can break down into acetamide as the solvent ages, which makes your solvent basic. Urea based pesticides will degrade under these conditions. A small amount of dilute formic acid can be added to your mixture to counter these reactions. Over time, halogenated hydrocarbons will interact with methanol to form acids. Low pH will degrade linear ketones very quickly. A dilute ammonium hydroxide solution will slow these reactions. Now you know why you should keep your acid sensitive and base sensitive analytes in separate mixtures. These are just a couple of examples of analyte–solvent interactions. There are hundreds of known interactions and many interactions are dependent on the number and concentration of your analytes in a mixture. Many laboratories request CRMs with all possible analytes at a high concentration in one ampule for cost savings. When your CRM expires in a couple of months, are you saving money? More organic compounds at higher concentrations mean there are greater chances for fast chemical interactions to occur. Use a stock product or, if a custom one is required, keep the concentrations reasonable and allow the manufacturer the option of separating analytes into multiple mixtures. The analytes’ shelf life will be longer and your headaches fewer.
Precision and accuracy are not the only concerns you might have about your standards; stability is also a major factor. The accidental use of a standard that has degraded will result in costly reanalysis and increased turn-around time, severely hurting your bottom line and your laboratory’s credibility. CRM manufacturers follow a set of ISO 17034 protocols to estimate a products shelf life.
There are three types of data that can be used to determine a products shelf life: historical data, classical method data, and the accelerated method data. Historical data rely on the manufacturer’s knowledge and data of the exact reference material being made to know how long that material can be expected to remain accurate under the assigned storage conditions. Using the classical method, the manufacturer makes the product and tests retained samples of the standard over time until the product fails. As you can imagine, this method is very time consuming and I doubt that you want to wait years to find out when your standard will fail. Since most laboratories prefer custom standards, the only practical way to determine shelf life is by using the accelerated method.
The accelerated method makes the assumption that the product failure (rate of degradation) increases as environmental conditions become harsher compared to the stated storage conditions. This ensures that the certified values for the analytes are within the stated uncertainties for the specified shelf life. The method is typically performed by heat stressing the standard. Several ampules of the standard are made and one is placed in a storage unit at the correct temperature. The other ampules are placed in heated environments at three or more temperatures between the stated storage temperature and 100 °C for a set time. The ampulized standards are analyzed sequentially using a metrologically validated method. If recovery of any of the analytes demonstrates a recovery of 95% or less, the standard is considered failed. This data is entered into an Arrhenius plot to provide a conservative estimate of shelf life. As the product ages, retained samples can be pulled from storage and tested using the classical method to determine if the shelf life can be lengthened or if it needs to be shortened.
As you can see, a great deal of effort is put into determining the values printed on your COAs. Even so, there can be unknowable aspects in the science of chemistry and the unforeseen can occur, but you can be assured that the best available technology was used to certify your reference materials. To reduce the possibility of the unforeseen, I recommend resisting the urge to amend your custom or catalog mixtures unless truly necessary. Also, when in doubt, let the professionals do it. You will save yourself time, money, and headaches.
Don Shelly is the Food and Environmental Product Manager, North America for LGC Standards in Manchester, New Hampshire. Direct correspondence to: firstname.lastname@example.org