Flame tests and discharge tubes: reading atomic fingerprints

Put metal salts into a flame and split the colour into lines with a spectroscope; compare the line spectra of hydrogen, helium, neon and mercury tubes; then identify an unknown sample from its lines alone.

Difficulty ●○○ ≈ 25 min

Aim Compare atomic lines with molecular bands, learn to identify elements by line positions rather than colour alone, and see how sodium’s glare interferes with the potassium test.
Emission spectra · the fingerprints of elements

Spectroscope

White light (incandescent): a continuous spectrum
Unknown sample #1

Compare the reference spectrum above with the unknown, line by line, until every line matches.

What the eye sees blue computed from the visible lines through the CIE eye model
Wavelength (in air)
431.40 nm
Frequency
6.947 × 10¹⁴ Hz
Photon energy
2.873 eV 277.2 kJ/mol
Emitter
CH molecular band (teaching band, FWHM 3 nm)
Colour
violet
Relative intensity
100.0 %

Intensity basis: Molecular bands sit at literature band positions but are drawn as Gaussian “teaching bands” whose widths and heights are illustrative; atomic lines are from NIST.

Line wavelengths come from the NIST Atomic Spectra Database (air wavelengths); hydrogen’s Balmer lines are computed from the Rydberg formula and converted to air. Glow colours are the visible lines pushed through the CIE 1931 observer — computed, not picked. Ca, Sr, Ba and Cu flames owe their colours mainly to band emission from molecules formed in the flame (CaOH, SrOH, SrCl, BaCl, BaOH, CuCl, CuOH), drawn as “teaching bands” at literature positions with illustrative widths and heights. Line brightness is compressed as the square root of intensity; the cobalt glass is a three-band teaching model.

Procedure

  1. Start with the blank flame: the blue inner cone of a Bunsen burner comes from molecular bands of the CH and C₂ radicals. Note their positions so you do not mistake them later.
  2. Work through LiCl, NaCl, KCl, CaCl₂, SrCl₂, BaCl₂ and CuCl₂. Describe each flame colour first, then select the strongest lines or bands on the spectroscope and record the wavelength and the emitter.
  3. Go back to potassium and tick “Sample contains a trace of sodium”: what colour is the flame now? Then tick “View through cobalt-blue glass” and watch potassium’s lilac come back.
  4. Switch to the hydrogen, helium, neon and mercury tubes. For hydrogen, match each Balmer line’s photon energy to a gap in the level diagram.
  5. Finally, the unknowns: choose a reference source and line its spectrum up against the unknown, line by line; press “The reference matches the unknown” only when every line agrees. Then press “Next sample” and try again.

What to notice

  • Lithium shows a single deep-red line (670.8 nm — a doublet too close to resolve); sodium shows the yellow 589.0/589.6 nm doublet, and even a trace of sodium can dominate the whole flame.
  • Potassium’s red doublet at 766.5/769.9 nm sits at the edge of vision and is barely seen; the lilac you perceive comes mainly from a weak violet doublet near 404.4 nm.
  • Calcium, strontium, barium and copper give broad bands rather than thin lines: their colours come mainly from molecules formed in the flame (CaOH, SrOH and SrCl, BaCl and BaOH, CuCl and CuOH), with atomic lines only a small part.
  • Similar colours do not mean the same element: lithium and strontium both burn red, but their lines sit in completely different places.

In a real lab

  • Wear eye protection. The Bunsen burner is an open flame — keep flammables away. Soluble barium salts are toxic, and copper and strontium salts must not be swallowed either; wash your hands afterwards.
  • Between samples, dip the platinum or nichrome wire in concentrated hydrochloric acid and heat it in the flame until it no longer colours it, to avoid cross-contamination; the acid also turns samples into volatile chlorides that colour the flame more strongly.
  • Sodium is everywhere (sweat, glassware, reagent impurities), so a flash of yellow does not prove a sample contains sodium; always view the potassium test through cobalt glass.
  • Discharge tubes run on several thousand volts — never touch the electrodes; mercury lamps emit intense 254 nm ultraviolet, so do not look at a bare tube.
  • The simulated atomic lines come from the NIST database; molecular bands are drawn as smooth “teaching bands” — real bands have fine vibrational–rotational structure, and intensities change with flame temperature and concentration.

All experiments