You're describing a really interesting (and frustrating) situation — let's break it down carefully:

From your description:

First sample injection: Fine.

Second sample injection: New unexpected peak at 23 minutes.

Blank injection: Same unknown peak at 23 min.

Repeated blanks: Peak disappears gradually.

Reference standard: No such peak ever.

Possible reasons for this behavior:

System Carryover/Contamination:

Most likely cause. Some residue from your sample (not the reference) is sticking to the injector needle, column head, or tubing, causing it to appear again in the blank.

Since repeated blanks gradually wash it out, this points strongly toward carryover rather than a chemical reaction.

Sample-Specific Instability/Interaction:

Something in your formulated cream sample (e.g., excipients, surfactants) might degrade or react slowly after injection, generating a secondary impurity.

Injection Needle/Loop Memory Effect:

Cream formulations, being viscous or oily, sometimes "stick" to injector surfaces, even if washed. If your system’s wash program isn't strong enough, traces stay behind.

Column Overloading or Fouling:

First injection loads something onto the column that only clears after repeated strong washing.

What you should check/try:

Increase your needle and loop wash cycles between injections (especially with strong solvents like acetonitrile or methanol).

Use a stronger blank solvent (instead of just your mobile phase, try something a little more aggressive if possible — check BP guidelines if they allow).

Flush the system longer after sample injections.

Verify your sample preparation method — maybe a filtration step (0.22 µm) or dilution helps minimize unknowns.

If possible, run a system suitability sequence: Blank → Standard → Blank → Sample → Blank → Sample again → Blank → see where the peak appears/disappears.

Important:

The fact that this doesn’t happen with the reference standard but only your sample strongly suggests it's sample-specific contamination or adsorption, not an instrument defect.

All the best 👍

El pico fantasma a 23 min en tu análisis de ketoconazol probablemente se deba a contaminación cruzada (carryover) del sistema de inyección, donde residuos de la primera muestra persisten en el autosampler o columna, liberándose en la segunda inyección. Esto explica su aparición en el blanco posterior y su desaparición tras lavados repetidos. Para resolverlo: