HPLC Troubleshooting | Reproducibility Tips

You have a reproducibility problem, and reproducibility issues almost always point to the instrument, not the chemistry. Before you rewrite a method, change columns, or blame the detector, you need to slow down and ask a simpler question: Why is the system behaving differently injection to injection?

Dr. Lee Polite breaks this down in a way that is both practical and fast to execute in the lab. The goal is not theory. The goal is to isolate where the variability is coming from and eliminate it systematically.

Same Sample, Same Vial, Same Result. Always.

If you inject the same sample, especially from the same vial, you should see the exact same chromatogram every time. That means:

• Same number of peaks
• Same peak heights
• Same peak areas
• Same retention times

If that is not happening, the instrument is introducing variability. At that point, it does not matter which chromatogram looks “right.” The fact that two injections do not match is the real problem.

Different Number of Peaks? They’re Already in the System.

One of the biggest red flags is when the number of peaks changes between injections. Two peaks on one run, five peaks on the next. Those extra peaks did not magically appear in your sample. They are coming from somewhere in the HPLC system.

That means the first step is not troubleshooting the method. The first step is to clean the column and flow path.

Step 1: Clean the Column the Right Way

Before evaluating any hardware, you must eliminate contamination, salts, and retained material from the column.

Start with High Water

Even though it sounds backward, start by washing with high water content:

• 90% water / 10% organic (methanol or acetonitrile)
• Use pure water with no buffers or additives
• Flow rate: ~1 mL/min
• Time: ~20 minutes

This step removes salts and buffer residues that can cause ghost peaks and baseline instability.

Then Go to 100% Organic

Next, wash the column with:

• 100% acetonitrile or methanol
• Flow rate: ~1 mL/min
• Time: 20–30 minutes

At the end of this step, the column should be completely clean. You should see a flat, stable baseline with no unexpected peaks.

If you do not, the problem is still contamination or system-related.

Step 2: The Six-Injection Test That Tells You Everything

Once the system is clean, this is where the real diagnostic work begins.

Take one vial and tell the autosampler to inject it six times in a row. Then evaluate three things:

Peak Area and Height = Autosampler Health

If peak areas and heights vary from injection to injection, you almost certainly have an autosampler problem. One of the most common culprits is a worn rotor seal, which leads to inconsistent injection volumes.

This is not a detector issue. This is not a method issue.

The problem is mechanical wear.

Retention Time = Pump Health

Retention time stability tells you about the pump. If retention times are drifting or inconsistent while peak areas are stable, you are looking at a pump-related problem, not sampling.

Together, these two measurements let you pinpoint the problem quickly without guessing.

Why Rewriting the Method Won’t Fix This

A method cannot compensate for:

• Dirty columns
• Residual peaks in the system
• Worn autosampler components
• Unstable pumping

Changing gradients, solvents, or column chemistry only masks the real issue. Worse, it creates a method that appears fragile when the true issue is hardware maintenance.

The Simple Diagnostic Workflow

When reproducibility disappears, follow this order every time:

1. Clean the column thoroughly
2. Run repeated injections from one vial
3. Use area and height to assess the autosampler
4. Use retention time to assess the pump

Do this before touching the method.

Final Takeaway

If your HPLC chromatograms are not reproducible, it is not your method. The instrument is telling you something is wrong, and if you listen carefully, it will also tell you where the problem is.

Clean the system. Control the variables. Let the data guide you.