Column Chromatography Calculator (Flash, from TLC)

In short: Enter the TLC Rf values of your spots and the crude mass. You get the column volumes to elute each spot (CV = 1/Rf), the ΔCV separation check, silica mass, column size, eluent volume and fraction size. A helper suggests the next hexane/EtOAc ratio to try. Free, in your browser.

How to use it

1. Run a TLC plate in the solvent you plan to use for the column. You can draw it with the TLC diagram tool.
2. Enter each spot's Rf and mark your product as the target.
3. Enter the crude mass. The silica ratio is chosen from the ΔCV of the closest spot, or pick your own.
4. If the target Rf is outside 0.2–0.35, use the eluent helper to pick the next ratio, run a new TLC and enter the new Rf values.
5. Pack the column, load the sample in a narrow band and collect fractions of about ¼ CV. Check them by TLC.

Worked examples

• Rf to CV: Rf 0.25 → 1/0.25 = 4 CV. Rf 0.35 → 2.9 CV. Rf 0.5 → 2 CV.
• ΔCV: spots at Rf 0.5 and 0.3 → 1/0.3 − 1/0.5 = 3.33 − 2.00 = 1.33 CV. That is above 1, so the separation is routine.
• Same ΔRf, different ΔCV: Rf 0.7 vs 0.5 (ΔRf 0.2) gives ΔCV 0.57; Rf 0.3 vs 0.1 (also ΔRf 0.2) gives ΔCV 6.7. Lower Rf spreads spots further apart on the column.
• Silica: 1 g crude at 50:1 → 50 g silica. At 0.5 g/mL that is a 100 mL bed: a 3 cm column filled about 14 cm high, with one CV of roughly 65 mL.

The rules behind the numbers

Column volumes from Rf

A compound with retention factor k elutes after (1 + k) column volumes. On TLC, Rf = 1/(1 + k), so CV ≈ 1/Rf when the column uses the same silica and the same solvent as the plate. This is the basis of the CV method used by flash-system makers such as Biotage and Teledyne ISCO. Real columns elute a little differently because the TLC plate is not in equilibrium with the vapour and the binder changes the silica slightly.

Target Rf and separation

Still, Kahn and Mitra (J. Org. Chem. 1978, 43, 2923), who introduced flash chromatography, recommended a solvent that gives the product Rf ≈ 0.35, about 15 cm (6 in) of 40–63 µm silica and a flow of about 5 cm/min. Today 0.2–0.35 is the usual window. The verdicts used here are:

The 30–100:1 range is the common lab rule for silica:crude (w/w). The tiers are a rule of thumb, not a published standard. As a check, around Rf 0.35 a ΔRf of 0.2 corresponds to ΔCV ≈ 1, which is the "easy" row of the Still table below.

Still, Kahn & Mitra (1978) column table

Column size, solvent and fractions

• Bed volume = silica mass ÷ packed density (default 0.5 g/mL). Diameter is chosen for a 15 cm bed and rounded to a common glass-column size.
• One CV ≈ 1.3 mL per g of silica (typical of packed 40–63 µm silica and prepacked cartridges, which list about 1.2–1.5 mL/g).
• Eluent ≈ 2 CV to equilibrate + 1/Rf CV to elute the target (or the slowest spot) + 2 CV for the tail of the band. Fractions ≈ ¼ CV.

Eluent helper

For a strong solvent B in a weak solvent, retention on silica roughly follows the Snyder–Soczewinski relation log k = log k₀ − n·log φB. From one plate and an assumed n (default 1), the tool solves for the fraction φ_B that gives the target Rf. With two plates at different ratios it fits n. The relation is approximate, fails near 0% or 100% polar solvent, and does not handle changes in selectivity. Always confirm with a new TLC.

Limitations

• CV = 1/Rf assumes the same silica, the same solvent and isocratic elution. Column Rf values are often a little lower than on the plate.
• The silica ratio, density and CV-per-gram values are typical numbers, not measurements of your silica. Edit them under Assumptions.
• The tool does not model gradients, band broadening, dry loading versus wet loading, or reversed-phase columns.
• The eluent helper is an empirical guide for the next TLC plate. It cannot predict changes in spot order.
• Very polar or acid/base-sensitive compounds may streak or decompose on silica; no calculation predicts that.