The concept of the limit of detection (LOD) has been, and still is, one of the most debated in analytical chemistry. Determination of a limit of detection seems like it should be a simple process, but it is one of the most misunderstood concepts with regards to chemical analysis and measurement. When low detection limits are needed to quantify the composition of a chemical mixture (such as proving that a reaction has reached its end state, or establishing that trace contaminants are absent to some required level of certainty), the equipment being used to monitor the process must be able to demonstrate its capability to meet those requirements.
The need for statistically robust limits of detection is particularly relevant in flow chemistry scenarios. Chemists and chemical engineers responsible for process monitoring have worked with chemical reactions in continuous-flow for decades because of the economy of scale, chemical process safety, and the push for efficiency. Flow chemistry touches many aspects of industrial processing and can be used in continuous reaction chemistry, for monitoring feedstocks, and protein and other chemical purification processes. Effective process monitoring in flow chemistry requires both a low limit of detection and fast sampling times.
The Concept of Detection
The limit of detection (LOD) is usually defined as the lowest quantity or concentration of a component that can be reliably detected with a given analytical method. Intuitively, the LOD would be the lowest concentration obtained from the measurement of a sample (containing the component) that we would be able to reliably discriminate from the concentration obtained from the measurement of a blank sample (a sample not containing the component). In the case of Raman spectroscopy, the challenge lies in detecting a Raman peak or band specific to the analyte of interest, and having confidence that it is a true Raman signal detection, not just a statistical fluctuation that “tricks” the chemometric model into thinking that the specific analyte has been detected.
To achieve a 99.7% confidence level for a normal distribution, the limit of detection would be set to 3 times the root-mean-square (RMS) error of measurements of a detected analyte. Using this “3 sigma” threshold, false detections of the analyte (i.e., false detections that are just random variations in the Raman spectrum) would happen in only 3 out of every 1000 measurements. If even higher confidence is required regarding the detection of a particular analyte, the LOD threshold can be raised to 4 times or 5 times the measurement RMS to lower false detections to 1 out of every 16,000 measurements or 1 out of every 1.7 million measurements, respectively.
So what does this have to do with Raman spectroscopy? Raman methods are becoming increasingly popular in process monitoring (and flow chemistry in particular) because they provide excellent specificity, the ability to detect and quantify many different constituents of a mixture. Unfortunately, conventional Raman analyzers are forced to base their analysis results on a weak and inherently noisy spectroscopic signal, so the RMS error of the measurements is high, and the calculated limits of detection are proportionately high as well. But with Tornado’s revolutionary and proprietary advancements in Raman technology, the RMS error can be significantly reduced, and the Raman analysis can achieve much lower (and thus better) limits of detection.
Examples of Raman as a flow monitoring tool
Let us look at some specific examples of Raman flow cell scenarios. In a flow protein purification flow chemistry scenario, you can look at:
- The protein of interest (API)
- Impurities from protein purification
- Aggregated proteins
- Concentration of components in feedstocks
- To ensure proper control of reaction processes in continuous streams
- Raman responds to all physical phases with as is sampling: gases, liquids, and solids.
- Raman can be used in conjunction with fiber optics. The measurements can be direct or using optimized flow cells to provide maximum sampling flexibility.
- Raman provides molecular information to characterize the specific chemistry and provide actionable information to facilitate maximum control.
High Throughput Virtual Slit (HTVS™) technology
Traditional Raman spectrometers utilize slits to improve resolution, but this significantly reduces the light throughput. A typical spectrometer design uses a narrower entrance slit to achieve higher resolution at the cost of throughput. Essentially, you are throwing away 90% of the light coming into your instrument. Why is Tornado Spectral Systems able to provide a 10x photon throughput and show these low limits of detection with short acquisition times? Bolstered by its patented High Throughput Virtual Slit (HTVS) technology, Tornado’s HyperFlux spectrometers deliver significantly enhanced photon collection power. Raman, as we know, is a low energy technique. We are essentially looking for that one photon in a million. The percentage of shifted photons will differ based on the interaction of each bond or laser frequency since flow chemistry scenarios contain low levels of components of interest or transient “plugs” of material that need to be measured quickly. The sensitivity of the Raman device used for this purpose is critical.
If flow chemistry is part of your analytical portfolio, we encourage you to view the webinar, “Exploring the Limits of Detection of a Model Compound in a Flow Scenario using High-Throughput Raman Spectroscopy“. This presentation provides a flow chemistry example illustrating the potential for high sensitivity with Raman measurements. Viewers will be presented with an accessible high-level overview of how Tornado’s HTVS™ technology allows higher-throughput measurements which lead to better time resolution, sensitivity, and quantitative resolution/accuracy.
For more information about our chemical analysis and measurement solutions, please contact us at firstname.lastname@example.org