Peak broadening in HPLC: causes, dead volume and the Van Deemter equation
Why your peaks become broader and which measures actually help
Peak broadening is the phenomenon in which an initially narrow sample band becomes increasingly broad during the analysis. The result: lower peaks, poorer resolution, less sensitivity and less reliable quantification. Peak broadening is therefore the most direct threat to the quality of your chromatographic data.
The causes can be divided into two groups: processes within the column, described by the Van Deemter equation, and extra-column effects caused by dead volume in the system. This article covers both and shows which measures have the greatest effect in practice.
⏱ approx. 8 minutes reading time
What is dead volume and why does it cause peak broadening?
Dead volume, also called extra-column volume, is all the volume in the system outside the column in which the sample can spread out. Every millilitre of it contributes to peak broadening without contributing to the separation.
Dead volume is found in these places:
- The sample loop of the injection valve
- All capillaries and tubing between injector, column and detector
- The flow cell of the detector
- Couplings, fittings and unions where the connection does not fit exactly
The effect proceeds in four steps: the sample spreads out before or after the column, the peak becomes broader, the peak height decreases because the same amount of substance is distributed over a longer time, and the resolution between neighbouring peaks falls.
In short
More dead volume gives more peak broadening and therefore poorer separation. With narrow columns and UHPLC applications, this effect is dramatically greater than with classic 4.6 mm columns.
The Van Deemter equation: peak broadening within the column
The Van Deemter equation describes how the flow velocity of the mobile phase affects the efficiency of the column. The equation reads: H = A + B/u + Cu, where H is the plate height (the smaller, the more efficient) and u is the linear flow velocity.
| Term | Name | What it describes | Dominant at |
| A | Eddy diffusion | Molecules follow different flow paths through the packed bed | All flow velocities; depends on packing quality |
| B/u | Longitudinal diffusion | Molecules diffuse from high to low concentration | Low flow velocities |
| Cu | Mass transfer | Analytes need time to reach equilibrium between phases | High flow velocities |
Mnemonic
A stands for Alternative paths. B for Brownian diffusion (longitudinal diffusion). C for Contact between mobile and stationary phase (mass transfer). Remember: too slow gives diffusion, too fast gives poor mass transfer. The optimum lies in between.
In practical terms, this means that for every column there is an optimal flow velocity at which the plate height is minimal and the separation maximal. Smaller particles shift this optimum to higher flow velocities and flatten the curve, which explains why UHPLC still works efficiently at higher flows.
Recognising and diagnosing peak broadening
Before taking measures, it is important to establish where the broadening comes from. This table helps with the diagnosis.
| Observation | Probable cause | First action |
| All peaks broad, including early peaks | Extra-column dead volume | Check fittings, shorten capillaries, smaller detector cell |
| Only late peaks broad | Normal diffusion effect or flow too low | Optimise the gradient or increase the flow towards the optimum |
| Peaks broad and tailing | Dead volume or secondary interactions on the column | Check the column inlet connection; consider adjusting the pH |
| Sudden broadening after maintenance | Incorrectly mounted fitting or ferrule | Disassemble and check that the ferrule is at the right depth |
| Gradual broadening over weeks | Column degradation or blocked frit | Flush the column; replace if there is no recovery |
| Broadening with a narrow column (2.1 mm) | System not optimised for low volumes | Minimise the internal volume of the entire flow path |
Seven measures to reduce peak broadening
- Use capillaries with the smallest possible internal diameter between injector, column and detector. The difference between 0.25 mm and 0.12 mm ID is immediately visible in the peak shape with UHPLC.
- Keep capillaries as short as practically possible. Every excess centimetre adds volume.
- At every fitting, check that the capillary is fully inserted up to the sealing surface. A gap of half a millimetre creates a measurable dead volume.
- Choose a detector cell with a volume that matches your column diameter. A large cell on a narrow column destroys the gain in column efficiency.
- Dissolve your sample in the mobile phase, not in a stronger eluent. Injecting in a stronger solvent causes band broadening at the column head.
- Limit the injection volume. Too large an injection volume overloads the column head and broadens the starting band.
- Work around the optimal flow velocity of your column according to Van Deemter, not at a standard 1.0 ml/min.
Never use Teflon tape on HPLC connections
Teflon tape or other sealants on HPLC fittings create dead volume and introduce particles into the system. HPLC fittings seal via the sealing surface or the ferrule, not via the thread.
Frequently asked questions about peak broadening
What is the difference between peak broadening and tailing?
Peak broadening makes the peak symmetrically broader: the peak retains its Gaussian shape but is less sharp. Tailing is an asymmetrical distortion in which the back of the peak trails off. Tailing more often indicates secondary interactions with the stationary phase or dead volume at the column inlet, while general broadening is usually system-wide.
How much dead volume is acceptable?
That depends on your column diameter. Rule of thumb: the total extra-column volume may be no more than 10 percent of the peak volume of your narrowest peak. With a 4.6 mm column, a few tens of microlitres is often acceptable; with a 2.1 mm UHPLC column, you must stay below five microlitres.
Why do my peaks become broader at lower flow?
At low flow velocities, the B term from the Van Deemter equation dominates: longitudinal diffusion. Molecules simply get more time to diffuse out into the mobile phase. Increase the flow towards the optimum of your column to limit this effect.
