caffeine by HPLC

[plateeny]

Hi every body

I used to determine caffeine in tea using this conditions:
30% methanol and 1% perchloric acid
c18 column
1.00 ml/min flow rate
PDA detector at 250 nm

but I got this time like two peaks not separated well

how u explain this?

[tom jupille]

That depends on the details of the sample and the column.

Was this a "sample" (i.e. an extract of tea) or a "standard" (i.e., pure caffeine)?

What was the size of the column (length and diameter)?

What were the actual retention times and baseline widths of the peaks?

[plateeny]

it happened for standard and the tea extract

the column of 25 cm long, retention time at 4 mints

[tom jupille]

the column of 25 cm long,

and what diameter?

Assuming 4.6 mm diameter, your column has a dead volume (internal volume) of about 2.5 mL. At 1 mL/min, your dead time (t0) is about 2.5 minutes. A retention time of 4 minutes means a k' value of about 0.6.

Because you are seeing the problem with (presumably) pure standard, that suggests that your column has developed a headspace or partially-plugged frit. Regardless, trying to quantitate any peak with a k' of only 0.6 is asking for trouble. In our method development courses, we recommend that k' values for quantitated peaks should be >1 (and >2 for pharmaceutical regulatory assays).

[Victor]

What is the purpose of the 1% perchloric acid in your HPLC separation?

It is certainly possible to get good peak shapes for caffeine leaving out the perchloric acid, although it is possible that the acid has some useful effect on the retention of potential interfering co-extractants.

Caffeine is an extremely weak base, and I am uncertain its retention time would be affected by the presence of the perchloric acid. Nevertheless, your mobile phase is quite acidic and may be reducing the column lifetime, resulting in split peaks etc that you are observing.