ELSD vs CAD: Detecting Compounds a UV Detector Cannot See
Most peptide purity work runs on a UV detector, and most of the time that is the right tool. The exception is a compound, or an impurity, that simply does not absorb at the wavelengths being monitored. For those, the ELSD vs CAD question comes up quickly: both detectors respond to the amount of non-volatile material leaving the column rather than to light absorption, and both can show a QC lab things a UV trace never will. This article explains how each one works, where they differ, and what a receiving lab should make of a purity figure produced by either.
The blind spot in a UV chromatogram
A UV detector measures how much light a compound takes up at a set wavelength. Peptide methods usually watch two regions:
- Around 214 nm, where the amide bonds of the backbone absorb. Any peptide gives some signal here, and longer chains give more.
- Around 280 nm, where aromatic side chains such as tryptophan and tyrosine absorb. A sequence without them is close to silent at this wavelength.
The trouble starts with material that has neither feature. Carbohydrates, lipids, many inorganic salts, counter-ions and a fair number of small non-peptide research compounds produce little or no useful response at either setting. On a UV trace they can be present in real quantity and still look like baseline. A very short peptide with only one or two amide bonds sits in a grey zone: detectable, but weakly, and easy to underestimate next to a longer impurity.
That is not a flaw in the instrument. It is simply what absorbance measures. The practical point for anyone reviewing data on a large order is that a clean UV chromatogram proves only that nothing absorbing was left behind.
How an evaporative light scattering detector works
An evaporative light scattering detector (ELSD) turns the column effluent into droplets, drives off the solvent, and looks at whatever solid is left. The sequence is:
- A nebuliser sprays the eluent with a carrier gas, usually nitrogen, into a fine mist.
- The mist travels through a heated drift tube where the mobile phase evaporates.
- Any dried particles of analyte pass through a light beam, scatter part of it, and a photodetector records the scattered light.
Because the signal depends on particles existing at all, rather than on any optical property of the molecule, ELSD responds to nearly anything that is less volatile than the solvent. That universality is why it is used. It comes with limits that matter when you interpret the numbers:
- Curved calibration. Light scattering depends heavily on particle size, and particle size changes with concentration. The response follows a power-law relationship rather than a straight line, so quantitation needs a fitted curve, often on log axes.
- Semi-volatile analytes disappear. If a compound evaporates in the drift tube, it produces no particle and no peak.
- The mobile phase must be volatile. Phosphate or other salt buffers leave their own particles and swamp the signal. Formic acid, acetic acid, TFA, ammonium formate and ammonium acetate are the usual choices.
- Lower sensitivity than UV for compounds that do absorb well.
How a charged aerosol detector differs
A charged aerosol detector (CAD) shares the front end: nebulise, evaporate, leave dry particles. The difference is the measurement. Instead of shining light through the particles, CAD collides them with a stream of positively charged nitrogen, the particles pick up charge, and an electrometer measures the total charge collected.
Charge uptake is less sensitive to particle size than light scattering is. In practice that gives CAD three advantages over ELSD:
- A response that is more consistent from one non-volatile compound to the next, so relative amounts are closer to the truth without a separate calibration for each species.
- A wider working range before the curve bends.
- Generally better sensitivity at low levels.
The trade-offs are cost and the gradient effect. Both detectors are affected by solvent composition, because more organic solvent nebulises more efficiently and changes the particle yield. Over a gradient run the baseline and response drift. Some systems correct for it in software; others use a second pump delivering an inverse gradient after the column, so the liquid reaching the detector has a constant composition throughout.
ELSD vs CAD at a glance
| Feature | UV | ELSD | CAD |
|---|---|---|---|
| What it responds to | Light absorption by the molecule | Light scattered by dried particles | Charge carried by dried particles |
| Needs a chromophore | Yes | No | No |
| Calibration shape | Linear over a wide range | Strongly curved | Curved, but flatter and wider |
| Response across different compounds | Varies with structure | Varies with particle behaviour | More uniform |
| Buffer restrictions | Few | Volatile only | Volatile only |
| Identifies the peak | No | No | No |
When a mass-response detector is worth asking for
For a QC lead deciding what analysis to request on a volume order, three situations justify an ELSD or CAD trace alongside, or instead of, UV.
The compound has no usable chromophore
If the molecule barely absorbs, a UV purity figure is measuring something other than the main component’s share of the sample. A mass-response detector is the only practical way to see it by liquid chromatography.
Components absorb very differently
A blend, or a sample with impurities that lack the aromatic residues of the parent, will give a UV area ratio distorted by the difference in absorbance. A CAD trace reports something much nearer the actual mass proportions.
You suspect something invisible
Excess counter-ion, residual reagents or non-absorbing excipients are exactly the materials a UV trace ignores. Running CAD in series with UV on the same injection is a common way to check that nothing substantial is hiding.
What neither detector can tell you
ELSD and CAD answer “how much non-volatile material eluted here”, not “what was it”. A peak on either trace has no identity attached. Confirming that the main peak is the intended compound still requires mass spectrometry or co-elution with a characterised reference standard.
They are also not upgrades for ordinary peptide work. When a sequence absorbs normally, UV is more sensitive, linear over a wider range, cheaper to run and better understood by every lab you might compare results with. The mass-response detectors earn their place where UV cannot see; they do not replace it where it can.
Interpreting a certificate that used ELSD or CAD
An area percentage is only meaningful alongside the detector that produced it. A figure of 98 percent from an ELSD run, a CAD run and a UV run are three different measurements. The ELSD value in particular is shaped by its curved response, which tends to under-report small peaks relative to large ones, and it should not be compared directly with a UV value on another lot or from another supplier.
When logging results for a multi-vial order, record the detector with every purity number in your inventory system. Treat a purity figure with no detector stated as incomplete, and query a UV-based figure for a compound that has no chromophore, because it raises the question of what the detector was actually recording. The broader issue of unequal detector response is covered in our article on UV response factors and peptide purity.
Bulk Peptides products are sent for independent HPLC purity testing, certificates are posted for a portion of the range on the certificates of analysis page, and each vial’s cap and crimp colour ties it to the matching report.
All material covered here is supplied for analytical and in-vitro laboratory research only. It is not for human or veterinary use, and nothing in this article is advice on using any compound.
