What about column temperature and length? They also affect peak separation.

Prarthana Gopinath, the mobile phase ratio was varied, and gradient elution was tried. Column temperature (30-40) and mobile phase flow rate (0.35, 0.6, 1.0 mL/min) were also varied, but the analytes still co-eluted. The column is a Roc C18 5um 250x4.6mm.

Cheslin Moodley, what is retention time of the analytes?

Iaroslav Kuziv, the retention time is approx. 7.6 min when a flow rate of 0.5 mL/min is used.

You can try increasing the column temperature till 45 degrees, if your analytes are thermally stable. Are you sure it is two peaks co-eluting and not a broad peak of a single analyte?

Prarthana Gopinath, Using the same method, I injected the pure analytes (separately i.e. not mixed), and they have similar retention times - so in a mixture, the one large peak is a contribution of both analytes.

I will give the temperature change a try! Thank you :)

Cheslin Moodley As for me, you must to use buffers with around neutral pH (6.5-7.5). It may be sodium phosphate, triethylammonium phosphate or even tributylammonium phosphate buffers. Ttriethylammonium and tributylammonium buffers increase R.t. of pyruvate anions but not influence on R.t. of pyruvaldehyde and its hydrated forms.

Quick search finds some papers on absorptions properties of pyruvic acid (https://doi.org/10.1039/D0CP01810D (references in this paper; also Figure 7 shows relative percentage of anion and neutral forms of both ketone and diol forms of pyruvic acid at different pHs)) and pyruvaldehyde (https://doi.org/10.1016/1010-6030(91)87017-P) that may help you to choose detection wavelength.

Iaroslav Kuziv, interesting! I will look into this and give it a try. Thank you for your contribution and assistance :)