Flame Tests
Summary
Different metal ions burn with different flame colours — sodium’s yellow, copper’s blue-green, lithium’s crimson — because heated electrons emit light at each element’s characteristic wavelengths. You’ll be able to explain why the colours are sharp and element-specific, why the anion doesn’t matter, and how this became atomic spectroscopy.
History
Flame tests are among the oldest analytical techniques in chemistry. Alchemists noted that different substances produced different flame colors, though they couldn’t explain why. The systematic study began in the 1750s when Andreas Marggraf used flame color to distinguish sodium and potassium compounds.
The breakthrough came with Robert Bunsen1 and Gustav Kirchhoff2 in 1859-1860. Using a spectroscope, they showed that each element produces unique spectral lines - essentially a “fingerprint.” This work led to the discovery of cesium (sky-blue lines) and rubidium (deep red lines), and laid the foundation for atomic spectroscopy.
Today we understand that flame colors arise from electronic transitions. Heat excites electrons to higher energy levels; when they fall back, they emit photons of specific wavelengths characteristic of that element.
Hazards & preparation
PPE: safety glasses; tie back hair and sleeves; an adult runs the flame.
- Open flame. Keep flammables away and have a way to extinguish it.
- The boron test uses alcohol, which is flammable — only wet the loop, keep the bottle capped and well away from the flame. Use ethanol, not methanol: methanol is toxic by ingestion and skin absorption, and its flame is only marginally greener. Alcohol flames are nearly invisible in daylight — work somewhere dim.
- Dilute hydrochloric acid for cleaning is corrosive.
- Copper salts are toxic and harmful to aquatic life — keep off skin and out of drains.
Disposal: the alkali-metal and alkaline-earth salt solutions (Na, K, Li, Ca) are dilute — flush with water. The copper solution must not go down the drain: treat it with washing soda (sodium carbonate) to precipitate the copper, filter, and bin/hazmat the solid. See the Safety page.
Materials
- Metal salt solutions (1g salt in 10mL water each):
- Sodium chloride (Na): yellow
- Potassium chloride (K): lilac/violet
- Potassium sulfate (K): lilac/violet — same as KCl!
- Cupric chloride (Cu): blue-green (intense!) — the azure-blue is partly from CuCl; copper sulfate gives a similar but slightly greener flame (see the anion note below)
- Calcium chloride (Ca): orange-red
- Lithium chloride (Li): crimson red
- Boric acid (B): bright green — dissolve 1g in a few mL of ethanol
- Cobalt is deliberately left out: it has no characteristic flame colour (any “silvery-white” is just glowing incandescence, not emission) and cobalt chloride is a Category 1B carcinogen and reproductive toxin — not worth handling for no result.
- Wire loop (nichrome or platinum) - 5cm wire with 3mm loop
- Bunsen burner or alcohol lamp
- Dilute hydrochloric acid - 10mL (5–10% solution)
- Small containers for each salt solution
Procedure
- Dip the wire loop in HCl, then heat it in the flame until it adds no colour (10-15 seconds).
- Dip the clean loop in a salt solution.
- Hold it in the hottest part of the flame (just above the blue cone).
- Observe the characteristic colour for 2-3 seconds.
- Clean the loop in HCl between each test.
What you should see
Each salt paints the flame its own colour: an intense, hard-to-miss yellow for sodium; a soft lilac for potassium (easily drowned by any sodium); a vivid blue-green for copper; orange-red (brick) for calcium; a pink-tinged crimson for lithium; and a striking bright green for boron in alcohol. Test KCl and K₂SO₄ back to back — identical lilac.
Seeing the colours if you’re colour-blind. Flame tests are colour-only by nature, so photograph each flame with a phone and compare the images side by side, or view the flame through a cheap diffraction grating — this spreads each colour into its distinct bright lines (position, not hue), which reads the same regardless of colour vision.
| Symptom | Likely cause | Fix |
|---|---|---|
| Everything looks yellow | Sodium contamination on the loop | Clean the loop in HCl and reheat until colourless before each test |
| Colour too brief to see | Solution too dilute or loop barely wet | Use a more concentrated solution; load the loop well |
| Potassium’s lilac invisible | Masked by sodium’s yellow | View through blue cobalt glass to filter the yellow |
| Weak boron green | Not enough alcohol / too little heat | Use boric acid dissolved in ethanol; hold in the hottest flame |
Flame Colors
| Element | Color | Notes |
|---|---|---|
| Sodium (Na⁺) | Intense yellow | True atomic emission; masks other colors |
| Potassium (K⁺) | Lilac/violet | True atomic emission; view through blue glass to filter sodium |
| Copper (Cu²⁺) | Blue-green | Very intense; the blue is largely molecular CuCl/CuOH emission (anion-sensitive) |
| Calcium (Ca²⁺) | Orange-red | Brick red — molecular CaO/CaOH bands, not free-atom lines |
| Lithium (Li⁺) | Crimson red | True atomic emission |
| Boron (B) | Bright green | Boron is a metalloid; the green is molecular; use boric acid in ethanol |
| Barium (Ba²⁺) (reference) | Yellow-green | Apple green — molecular BaO/BaOH bands; not in the materials list above |
| Strontium (Sr²⁺) (reference) | Crimson red | Deeper than lithium; molecular SrO/SrOH bands; not in the materials list above |
The Science
Heat excites electrons to higher energy levels. When they fall back, they emit specific wavelengths of light unique to each element:
\[E = h\nu = \frac{hc}{\lambda}\]
Each element has different energy level spacings, producing different photon energies (colors). This is the basis of atomic emission spectroscopy and how we know the composition of distant stars.
Not every colour here is pure atomic emission, though. The clean atomic-line picture holds well for the alkali metals (Na, K, Li). But calcium, barium, and strontium glow mainly from short-lived molecules formed in the flame (CaOH, BaO, SrOH…), whose broad bands give those brick-red and apple-green hues — and boron, a metalloid, likewise emits as a molecule. So “each element, one fingerprint” is the right first idea, but several of these colours are molecular rather than free-atom.
Mostly, the flame color depends on the metal, not the anion: potassium chloride and potassium sulfate give identical lilac, because the anion doesn’t survive intact in the flame. Testing KCl and K₂SO₄ back-to-back demonstrates this directly. Copper is the classic exception — much of copper’s blue comes from molecular CuCl, so cupric chloride flames a purer azure-blue while halide-free copper sulfate leans greener. When the anion itself forms an emitting molecule, it can shift the colour.
Questions to Explore
Why do electrons emit specific colours? Why is each colour tied to a particular element rather than a continuous range?
Hint / answer
An atom’s electrons can only occupy fixed energy levels, so when an excited electron drops back it releases a photon of one exact energy — one colour. Because every element has its own unique set of level spacings, each emits its own set of colours: an atomic fingerprint.
Why is sodium’s yellow so overwhelming? Trace sodium drowns everything. What does intensity say about the transition?
Hint / answer
Sodium’s yellow transition is extremely probable (a strongly “allowed” transition) and easily excited at flame temperatures, so even a few sodium atoms emit a lot of yellow light. That’s why the tiniest contamination swamps fainter colours.
Why doesn’t the anion affect the colour? KCl and K₂SO₄ both give lilac. Why don’t chloride or sulfate contribute?
Hint / answer
In the flame the salt breaks apart and it’s the metal atoms that get excited and emit visible light. The anions either don’t emit in the visible range or are broken up, so only the metal’s colour shows — the same metal gives the same colour regardless of its partner.
How did this discover new elements? Bunsen and Kirchhoff found unfamiliar lines. Why does that mean a new element?
Hint / answer
Each element’s line pattern is unique and unchanging, so a set of lines matching no known element must come from a new one. A new compound would still only show the lines of its known constituent elements — new lines mean a new element.
How do fireworks engineers use this? Same salts (Li red, Ba green, Sr crimson). What extra challenges do they face?
Hint / answer
They must make bright, pure colour at altitude while also burning fuel/oxidizer and launching the shell — sodium impurities (yellow) must be kept out, and the metal salts, binders, and temperature carefully balanced so the intended colour dominates.
Going further
- Test the anion rule. Run KCl and K₂SO₄ (and two sodium salts) side by side to confirm the colour comes from the metal, not the anion.
- Identify an unknown. Have someone hand you unlabeled salt solutions and name the metal from the flame colour alone.
- See the yellow filtered out. View potassium’s lilac through blue cobalt glass to cut sodium contamination — the trick 19th-century chemists used.
Footnotes
Robert Bunsen — German chemist (1811–1899) who, with Kirchhoff, developed flame spectroscopy and the Bunsen burner.↩︎
Gustav Kirchhoff — German physicist (1824–1887) who co-founded spectroscopy with Bunsen.↩︎