In my experience, most people don’t like using an internal standard. This probably stems from the extra sample prep needed, and more than likely this opinion was formed when a person was at a technician level. People just want to put a sample in the vial, press start and be done with it. But in fact, an internal standard is very powerful, when I look at data that uses an internal standard, I instantly have more confidence in the data. But this was not always the case, I too was prejudiced against internal standards as well.
Many years ago, I was working in a laboratory managing GCs, and I had a fellow chemist who was over the Analytical area, he had been there longer and helped me get up to speed with the instruments in my area. One day, I came into the lab, and he was in front of the D5769 instrument looking puzzled. When I asked him what was up, he proceeded to tell me that he had just calibrated the instrument and had 0.9999 for the correlation coefficient. I said that is great, so what is the problem. He said the highest concentration standard had the lowest area counts, which of course is the opposite of what we would expect. At the time I was not familiar with D5769, so I was looking at this with fresh eyes, but it was obvious that this was strange. We had another person come into the lab who was responsible for setting up GCs for our laboratories all over the country. So, we presented the problem to him, his advice was to run the low standard 10 times, I looked at him with puzzlement and doubt, unfortunately I cannot always control the look on my face. Then I asked, what good would that do? He said you obviously lost sensitivity, so you need to check the low standard. I said to him, look at the calibration curve, it has 0.9999 correlation coefficient, you cannot get much better than that, if we lost sensitivity, this would not be possible. Then I looked at the other chemist and asked, does this method use an internal standard? To which he replied yes. Then I said, check your syringe solvent wash, which was dry. So, I told him to refill the solvent wash and rerun the high standard, which came back normal.
This experience cemented my understanding of the internal standard and trust in it, not to mention the importance of filling your solvent wash bottles. The difference between an external standard and an internal standard, is that you must add a known amount of an analyte that is similar to the analytes of interest. An external standard is running known concentrations of the analytes. During external standard calibration, we plot response (area) versus concentration. But an internal standard calibration the response is the ratio of the analyte to the internal standard (area of the analyte/area of the internal standard). The internal standard negates any injection errors, if you can see both the analyte and the internal standard, the ratio remains the same. That is why the calibration had a 0.9999 correlation coefficient. This would not be possible with an external calibration.
In my previous article “Why Sample Introduction Matters”, I mentioned that every instrument is essentially three components, sample introduction, chemistry, detectors, and the problem nearly always lies in sample introduction. By looking at any analytical instrument this way, it really removes distractions from the troubleshooting process. Allowing you to utilize your instrument more effectively.
Most laboratory personnel do not understand the analytical instruments or the methods they perform daily, so it is not surprising that those outside the laboratory understand it even less. Those of us who are over laboratories have an amazing opportunity to learn, but we don’t all take advantage of it. The instruments around us are marvels of innovation, everything about it was designed to solve a problem, you should take the time to understand how they work.
After reading this article you should have an idea of the power of an internal standard. I also mentioned previously that the problem rarely lies with the detector. In “Why Integration Parameters Matter”, I suggested using a Certified Reference Material as your QC. This is a very powerful combination with an ISTD. As I mentioned, detectors are very linear, in fact if you overlaid your last few calibration curves, you would find that they changed very slightly, and the change is more likely due to standard degradation or environmental conditions of the laboratory than it is of the detector. It is rarely necessary to recalibrate your detector, it is more important that you mitigate the variables that affect sample introduction.
If you use an ISTD combined with a Certified Reference Material for your QC, you can align precision with accuracy. Typically, when you are tracking your QC, you are only concerned with precision, you assume that there is some bias, but if your precision is good, you know the instrument is functioning properly. Using the ISTD with a CRM means that you can adjust your integration parameters to ensure that you get an accurate result on your QC. I spent a lot of time investigating QC charts, specifically on D5599, I tracked temperature, pressure, humidity, and I also tracked the area counts of the ISTD. I noticed that when the lab was hotter or colder, the area counts of the ISTD would change, which led me to investigate the chromatogram where I noticed the change in peak shape. With a CRM-ISTD combination, I could optimize my integration parameters to correct for this drift, rather than perform a calibration, once the optimized integration parameters are saved to the method all subsequent samples use the same integration parameters, ensuring that data for my samples were as accurate as possible, even when working in a challenging laboratory environment. The Internal Standard Matters, because if you know how to use it, it can improve your data quality, the extra sample prep is well worth it.