Chromatography Fundamentals: Paper, Column and TLC

Drop a mixture at one end of a system that has a stationary phase and a flowing mobile phase, and each component partitions between the two. Spend more time in the mobile phase and you travel fast; spend more time stuck to the stationary phase and you travel slow. Polarity is the lever in adsorption chromatography; solubility ratio in partition; charge in ion-exchange; size in gel filtration; specific binding (antibody, lectin, biotin) in affinity. Same idea, five different chemical handles.

The five modes split cleanly in the lab. Adsorption on silica gel or alumina is what TLC and classical column work use; polar analytes stick harder to the polar silica surface and elute later. Partition on cellulose paper or on a bonded C18 reverse-phase column splits analytes by relative solubility. Ion-exchange resins (Amberlite IR-120, Dowex) separate amino acids, metal ions or peptides by net charge. Size-exclusion gels (Sephadex G-25, Sepharose CL-6B) sort by molecular size; large molecules cannot enter the gel pores and elute first. Affinity chromatography uses a specific biological recognition element bonded to the stationary phase, which is why immunoaffinity columns are the cleanup of choice for aflatoxin in the FSSAI reference labs.

The choice of mode decides which question the chromatogram can answer. Adsorption-mode TLC on silica is the right tool for "what alkaloid is in this datura seed extract". Reverse-phase HPLC is the right tool for "how much paracetamol is in this serum". Ion exchange is the right tool for "is this seized powder a quaternary ammonium herbicide salt". Picking the wrong mode sends a sample on a four-day round trip to the wrong central facility.

Paper chromatography uses Whatman No. 1 filter paper. The stationary phase is not the cellulose itself but water bound inside the fibres; the mobile phase is an organic solvent that creeps up by capillary action. The sample is spotted on a pencil-drawn baseline 2 cm from one end, and the strip is dipped into a chamber with the lower edge in the mobile phase but the spot above the solvent line.

The classic forensic use is ink and dye separation in questioned-document examination. A suspected forged signature, a backdated will or an altered cheque amount reduces to the same operational question: are the inks identical. The examiner cuts a strip from the questioned ink stroke, extracts in pyridine or methanol-water, spots it alongside extracts from suspect pens, and runs an n-butanol–water–acetic acid mobile phase. Each ink dye separates into its component coloured spots, and the Rf pattern is compared against the reference inks. The Government Examiner of Questioned Documents (GEQD) regional offices at Shimla, Kolkata and Hyderabad still use this as a presumptive comparison before video spectral comparator work. Resolution on paper is poor (plate count 200 to 500), which is why TLC replaced it for everything except this niche.

Column chromatography is the preparative format. A vertical glass column, 1 cm internal diameter for teaching scale or 10 cm for natural-product work, is packed with silica gel 60 (40 to 63 µm for flash, 63 to 200 µm for gravity). The sample loads on top, the mobile phase flows down under gravity or modest pressure, and the eluate is collected in numbered fractions. Each fraction is checked by TLC; fractions with the same single spot are pooled.

Forensic uses of column chromatography are mostly upstream cleanup, not identification. A toxicological extract from autopsy liver tissue contains lipids, pigments and protein fragments that swamp any HPLC or LC-MS run. A short silica cleanup column strips the worst of the matrix before the sample sees the analytical instrument. The Solid Phase Extraction (SPE) cartridges that have largely replaced classical column cleanup are themselves miniature silica or C18 columns; the physics has not changed, only the form factor.

Take an aluminium-backed plate coated with a 250 µm layer of silica gel 60 with a fluorescent indicator. The F254 marker means the silica fluoresces under 254 nm UV, so any UV-absorbing analyte appears as a dark spot against a green background. Cut the plate to 5 by 10 cm. Pencil a baseline 1 cm from the bottom and mark five spotting positions. Spot 1 to 5 µL of sample and each reference standard with a 5 µL glass capillary, keeping each spot under 3 mm in diameter. Pre-saturate the development chamber for fifteen minutes by lining it with filter paper soaked in mobile phase. Stand the plate in the chamber, let the solvent climb to about 1 cm from the top, mark the front, dry the plate, and read it.

Mobile-phase selection is the working art. A handful of mixtures cover most forensic drug-screening work at Indian state SFSL benches. Chloroform–methanol–ammonia 90:10:1 is the general drug system for amphetamines, cocaine, codeine and benzodiazepines on the same plate. Ethyl acetate–methanol–ammonia 80:10:10 is the dedicated opioid system, sharp on morphine, codeine, heroin and 6-monoacetylmorphine. n-Butanol–water–acetic acid 4:1:1 is the dye and ink system used in document sections. Toluene–ethyl acetate–diethylamine 70:20:10 is the classical alkaloid system for plant material (datura, opium poppy, strychnine, areca nut). Compositions are quoted by volume and prepared fresh, because aged mobile phases drift through evaporation.

A typical drug-screen workflow at FSL Madhuban or the SFSL at Sagar runs like this. The recovered powder is dissolved in methanol at roughly 1 mg/ml. The plate is spotted in five lanes: a methamphetamine standard, a heroin standard, a cocaine standard, the unknown, and a methanol blank. The plate is developed in chloroform–methanol–ammonia 90:10:1, dried, viewed under 254 nm UV (any spot that quenches the fluorescence shows dark), then sprayed sequentially with Dragendorff for the alkaloid screen and Marquis on a duplicate plate for the drug-class screen. A matching Rf in two different mobile phases plus a matching Marquis colour is treated as a presumptive positive that justifies booking LC-MS/MS or GC-MS confirmation. Detection limit on a routine plate is 1 to 5 µg per spot, which is why concentration of the extract before spotting matters.

Plate resolution is set by silica particle size, layer thickness and development distance. Standard 250 µm plates give an effective plate count around 5,000 over a 10 cm development, which separates four to six spots cleanly if their Rf values differ by 0.1 or more. HPTLC plates with 5 to 7 µm particles and a thinner layer resolve eight to twelve spots in the same distance, which is why NIPER Mohali uses HPTLC for herbal-drug fingerprinting under the Indian Pharmacopoeia 2022 monographs. HPTLC is a separate Module 4 topic.

A column or TLC plate is rated by how narrow its peaks (or spots) are relative to how far they have travelled. The plate count N = 16(tR/W)² treats the column as a cascade of equilibrium stages; a higher N means more equilibrium steps redistributing the analyte between phases, which produces a narrower band. A 25 cm reverse-phase HPLC column at 5 µm particle size delivers N around 25,000 to 50,000 for a well-retained peak. A gravity silica column delivers 500 to 2,000. A TLC plate around 5,000. A Whatman paper strip 200 to 500. The numbers explain why each format does the job it does.

Plate height H = L / N gives the average length per equilibrium stage. Smaller H means a more efficient column. The Van Deemter equation H = A + B/u + Cu describes how H depends on linear flow velocity u, with three terms representing eddy diffusion, longitudinal diffusion and resistance to mass transfer. The minimum of H against u is the optimum operating velocity. Too slow a flow and the band broadens by longitudinal diffusion; too fast and it broadens by mass-transfer lag; the sweet spot is the flow where Van Deemter sits at its minimum, typically 1 ml/min on a 4.6 mm i.d. column.

Resolution Rs = 2(tR2 − tR1) / (W1 + W2) decides whether two adjacent peaks are usable for quantitation. Rs of 1.5 or higher is baseline separation; each peak can be integrated independently. Rs of 1.0 is a 4 % overlap, acceptable for qualitative work but bad for quantitation. Rs below 0.8 means the two peaks cannot be reliably split. The three levers for improving Rs are: increase plate count (longer column, smaller particles, slower flow at Van Deemter's minimum); change the retention factor difference (different mobile-phase polarity, pH change for ionisable analytes); or increase selectivity α by switching stationary-phase chemistry. On TLC the same logic runs by changing mobile-phase composition, lengthening development distance, or switching from standard plates to HPTLC.