Evaporative HPLC Detectors; CAD (Charged Aerosol Detector) and ELSD (Evaporative Light-Scattering Detector)

• If you wish to read about their development and/or operating principles, then please review the early published patents and many articles available through the web. Be cautious when reviewing any "sales" brochures or articles on these detectors as a great deal of misinformation may be found.

E.L.S.D. modules for HPLC applications were first developed and commercialized in early 1980. CAD units were first described ~ 2001 (US patent 6,568,24) and commercialized after 2004. Both types of evaporative detectors have undergone many updates over the years. They are complementary and focused on the same application areas where conventional UV/VIS detectors do not provide for or allow detection of specific compounds. While claimed limits of detection vary by manufacturer, both designs are highly sample and method dependent so fair comparisons are rare (sales literature is often very biased to make one system fail). Significant differences in cost between the two detectors are noteworthy, with CAD units currently costing several times as much as ELSD units. Let us take a look at some of the characteristics and uses of these very unusual niche detectors.

Applications: CAD and ELSD are both used with a wide range of non-volatile sample types. Targeted at compounds which have weak or no UV chromophore (e.g. Carbohydrates, fats, lipids, triglycerides, polymers, surfactants, oils).
 

Thousands of application notes and journal articles are available for both types of detectors (esp. for ELSD with almost 40 years of use) and a keyword search on the web is the best way to find them. As someone who was involved in the early development and design of these detectors, I have used them successfully to develop several hundred different types of methods. They have proven to be useful for a number of difficult applications, but their higher cost and even higher training and skill requirements still place them outside of most users labs. As with LC/MS detectors, CAD/ELSD modules may require far more maintenance and advanced training to use than most chromatographers have received. As such, it is my opinion that you consider their potential use in your projects only after other more conventional methods have failed to provide results. Due to the high level of training needed, difficulty to operate and maintain, high cost of operation and poor reproducibility, IMHO they should be a "last choice".

Detection: NOT “Universal” detectors (sourced to marketing misinformation from vendors and early academic reviews which over simplified their 'operation', not of the actual commercial instruments). While detection is partially based on the analyte’s chemical or physical properties, the actual output observed is in fact also based on the properties of the mobile phase (volatility and purity), sample volatility/stability and especially the many different custom detection settings chosen by the user (gas flow, heater temperatures, flow rate, specific detector used, level of contamination inside the detector). As such, their output is very subjective since it is based on both the specific chromatography method selected, the condition of the detector, the lab environment used-in, and the detailed operational settings chosen by the operator. They can detect everything from dirt, buffers, undissolved chemicals or particulate matter in your mobile phase. Even pressure changes on the detector's exhaust line can effect the output.

“Destructive” Detectors: As with an LC/MS system, the mobile phase is evaporated away from the sample and sample collection is not possible at the exhaust. They are best used as a secondary detector, with a primary detector sch as a UV/VIS module placed in front of the CAD/ELSD (to increase your chances of detecting something that the CAD or ELSD may miss). ELSD and CAD units will NOT detect all samples. If sample collection is required, a low volume, micrometer valve flow-splitter can be fitted to the evaporative detector’s inlet port. Note: Depending on the flow-splitter's split ratio, the detector’s signal output may be reduced.

Mobile Phase Requirements: Evaporative detectors require a fully volatile mobile phase (similar to LC/MS requirements). The use of non-volatile additives can contaminate or damage them (no phosphate buffers!). Use of non-volatile buffers or additives, low purity materials, contamination of the gas, mobile phase or by samples may result in excessive noise levels limiting detection. Use high-purity grades of mobile phase and additives. Examples of Mobile phases used: "Popular LC/MS and HPLC Volatile Mobile Phase Buffers"

Isocratic and Gradient Capable: Unlike RID or EC, CAD/ELSD allows the use of gradients and the use of UV obscuring solvents. Because the mobile phase is evaporated away, little to no baseline drift occurs during gradient analysis (often improving integration results). Sample types which dissolve best in solvents such as methylene chloride, acetone, chloroform or other strong UV absorbing solvents may find that these detectors assist in developing better quality methods. Reduced gradient baseline drift plus the option of using UV absorbing solvents are two characteristics which make them well suited to application areas such as lipids, polymers and oils. Flatter baselines allows for better quality peak integration.

Gas Requirements: Similar to the requirements of an electrospray LC/MS system, both CAD and ELSD modules use very large volumes of high-purity gas (i.e. Nitrogen) to safely evaporate the mobile phase away. Exhausting these large volumes of solvent vapor and gas into a fume hood is just as important during site prep. Be sure and factor these costs and the required space into any site-prep plan.

Operational Reproducibility and Method Transfer:  Recording the exact detector settings used in the method may not provide any guarantee of being able to duplicate the results obtained. No two instrument models are the same so results may vary (similar to LC/MS). Results obtained for each sample are relative to the specific instrument, the chosen settings & method used (again, much like LC/MS) and the internal condition of the detector used. Compare the many critical heat, gas flow and atomization related CAD/ELSD settings to the more common UV/VIS detector where only the wavelength, bandwidth and flow cell dimensions need to be specified to easily duplicate the detector setup. CAD/ELSD internal contamination levels, nebulizer spray patterns, gas flow rates, quality of the mobile phase and operator training may all contribute to variations. *As with all methods and detection systems, proper training and good method design will insure success.

Quantitation: Can be used for quantitative analysis across a wide dynamic range spanning multiple orders of magnitude with some success. High quality reproducible methods are achievable with both types of detectors, but will require calibration tables with many additional standards per order of magnitude.

Linearity and Output Characteristics: Except in the most narrow concentration ranges, neither detector is likely to provide a linear response. Different samples will need their own full calibration table and curve fit, per method. Quantitation can be improved through the use of larger numbers of calibration levels (more than normal) plus a high quality chromatography data analysis software package which includes many available non-linear curve fit options (polynomial, quadratic, sigmoidal, exponential, log…etc). Output often changes across orders of magnitude so be sure to optimize the curve fit for each sample type. Different sample types will often have different response outputs at different retention times. This is most easily observed during a gradient analysis. As the mobile phase composition changes, so does the response for EACH sample (this is NOT a UV/VIS detector).

Optimization Process:Unlike a UV/VIS or RID system which simply needs to warm up and stabilize, CAD and ELSD systems may require a methodical optimization process of adjusting the flow rate, gas flow and heating temperatures to optimize the measured S/N peak ratios for each sample and each method used (yes, ever one of them). Optimization of detection conditions usually involves making multiple measurements (Peak and Baseline S/N ratios) to find the best settings to use with each sample type and method. This optimization process is time consuming and changes may need to be made to the method over time as the detector fills up with baked-on sample material (changing the spray pattern via nebulization changes).

Operational Complexity: Methods which utilize CAD/ELSD systems may be more complicated and time consuming to learn, use and validate then conventional detectors. Specialized detector cleaning procedures are often needed. The detectors may become internally contaminated during use (sample builds up inside the unit). Failure to clean and maintain them may lead to high noise levels and/or inaccurate results. Due to the additional maintenance needs, lack of traditional linearity, and overall complexity, we recommend their use only when: (1) Conventional detectors or methods of analysis are not possible or unsatisfactory and (2) where the operator has demonstrated a high level of practical hands-on training through use of the detector and/or has sufficient experience (advanced level) in chromatography.

For more information:

•  “Advanced Evaporative Light-Scattering Detector (ELSD) For A Variety of HPLC Applications”;
• “Evaporative Light-Scattering Detector (ELSD) Optimization Procedure. Varex ELSD IIA, Technical Note # 0024”;