Ammonia Fountain

Ammonia, Upside Down

A completely dry 500 mL round-⁠bottom flask is filled with ammonia and clamped mouth down on a stand. A rubber stopper with two holes closes it.

Through one hole runs a glass tube, drawn at its top to a fine jet that reaches the centre of the flask; its lower end comes about 20 cm out of the flask and down into a beaker of water. Through the other is a dropper holding a little water. The Royal Society of Chemistry’s version uses a 10 mL syringe; Rutgers University’s demo uses a rubber bulb.

The water in the beaker has a few drops of phenolphthalein and is colourless. A teacher does this demo in a fume cupboard, with eye protection, and only in a thick round-⁠bottom flask of borosilicate glass: other flasks may implode.

500 mL round-⁠bottom flask · two-⁠hole rubber stopper · jet tube · dropper · water with phenolphthalein

Lighter Than Air

Ammonia is lighter than air: a volume of it weighs about six tenths as much as the same volume of air. So the flask is filled mouth down, the ammonia led in from below, pushing the air out of the mouth.

The flask must be dry. Ammonia dissolves in water so readily that a little water on the glass would take up the gas before the demo starts, and it would fail.

In the lab, ammonia can be made by warming concentrated ammonia solution; how factories make it is on the page “Nitrogen from the Air”. Hydrogen chloride is heavier than air, so for the same demo the flask is filled mouth up, from the bottom.

Ammonia about 0.6 · air 1 · hydrogen chloride about 1.27 (the weight of equal volumes, air = 1)

A Squeeze of Water

A squeeze of the dropper puts a few drops of water into the flask. Ammonia dissolves in water in startling amounts: at 20 °C, 100 g of water takes up 52.6 g of it. Counting the gas at 20 °C, one volume of water takes up about 740 volumes of ammonia.

The same volume of water takes up only about 0.94 volumes of carbon dioxide and about 0.03 of oxygen. Hydrogen chloride dissolves very well too, about 480 volumes; sulphur dioxide about 42.

At that rate, half a millilitre of water at one atmosphere could hold about 370 mL of ammonia, nearly three quarters of the flask’s gas.

At 20 °C, volumes of gas taken up by one volume of water: ammonia about 740 · hydrogen chloride about 480 · sulphur dioxide about 42 · carbon dioxide about 0.94 · oxygen about 0.03

The Pressure Falls

When gas dissolves, fewer gas molecules are left in the same space, so the pressure falls: p·V = n·R·T.

The flask is closed, and as the pressure falls the water can hold less ammonia. Worked out from the vapour pressures of ammonia solutions that Perman measured in 1903, half a millilitre of water left to itself would bring a 500 mL flask of ammonia down to about half an atmosphere, with about half the ammonia dissolved in those few drops.

Far less is needed. In this set-⁠up the jet is about 22 cm above the beaker’s surface, so the flask only has to fall about 2.2 kPa below the outside, about 2% of an atmosphere, for the water to reach the jet.

The page’s model: the pressure in the flask and the column in the tube · column height = pressure difference ÷ (ρg)

Pushed, Not Sucked

The flask does not suck the water up. The air presses on the beaker’s surface with about 10 newtons on every square centimetre; inside the flask the pressure is now lower, and the difference pushes the water up the tube.

In 1644 Torricelli wrote that the force holding up a column of mercury “comes from without”: the weight of the air outside. A drinking straw works the same way: lower the pressure in your mouth a little, and the air pushes the drink up.

There is a limit. Even with nothing at all left in the flask, the air could hold water up only about 10.35 m at 20 °C. In 1638 Galileo recorded that no pump could lift water above eighteen cubits, about 10.5 m.

One atmosphere holds up about 10.35 m of water (20 °C) · here only 22 cm is needed

A Fountain That Feeds Itself

Once the water reaches the jet it sprays into the flask in fine drops. Every drop meets ammonia and takes it up at once, so the more water comes in, the lower the pressure falls.

In the page’s model, a second or two after the fountain starts, less than a tenth of an atmosphere is left in the flask; the water leaves the 1.5 mm jet at about 10 m/s, strikes the top of the flask and runs down its walls.

When the ammonia is nearly all dissolved, what is left is the air that was mixed in. As the flask fills, that air is squeezed smaller, the pressure rises again and the fountain slows. How long it runs depends on the flask and the width of the jet: the Royal Society of Chemistry says some minutes; the page’s model takes about half a minute.

The page’s model: the water’s speed at the jet and the water in the flask · a 1.5 mm jet

Where the Pink Comes From

Ammonia is not pink. The beaker’s water has phenolphthalein in it, an indicator: colourless below pH 8.3, turning pink above it, and deep pink by about 10.

Ammonia in water is a weak base: NH₃ + H₂O ⇌ NH₄⁺ + OH⁻. Only a few of its molecules take a hydrogen ion, about 1.3% at 0.1 mol/L, yet the OH⁻ they give is enough to bring the pH to about 11.

So the water in the tube is still colourless, and turns pink only as it leaves the jet and meets the ammonia. With the whole flask of ammonia in about half a litre of water, the pH is about 10.9. Phenolphthalein fades again only in bases stronger than pH 14, beyond ammonia’s reach.

Phenolphthalein: colourless below pH 8.3, pink over 8.3⁠–⁠10 · the fountain’s water about pH 10.9

The Other Fountain

Hydrogen chloride makes a fountain too. At 20 °C one volume of water takes up about 480 volumes of it. In water it is a strong acid, almost wholly broken into ions, so once dissolved it stays dissolved.

In Newth’s book of lecture experiments of 1892 the water is coloured with blue litmus, and the fountain comes out red; the Royal Society of Chemistry now uses universal indicator, which turns red too. With the whole flask of hydrogen chloride in about half a litre of water, the pH is about 1.4.

The earliest account we found is in Karl Heumann’s demonstration handbook of 1876: a dry bottle of 2 to 3 litres full of ammonia, hung upside down, its tube drawn inside to a 2 mm jet; the water was coloured with litmus reddened by acid, and it rose as a blue fountain.

Hydrogen chloride about 480 volumes · blue litmus turns red in acid · pH about 1.4

The Bubble That Stays

When the fountain stops, a bubble always stays at the top: it is air. The Royal Society of Chemistry notes that its size shows how well the flask was filled. In the page’s model, a flask holding 5% air ends about 95% full; one holding 30% air, only about seven tenths.

A damp flask fails: the ammonia dissolves while it is being filled. A leak at the stopper or the tube fails too: air comes in, and the pressure never falls. That is why cork is not used.

Warmer water takes up less ammonia: at 50 °C, 100 g of water takes up 22.9 g, less than half the 20 °C figure. In the page’s model, though, a few drops at 50 °C still bring the flask down to about three quarters of an atmosphere, plenty to start the fountain. We found no source saying warm water stops it.

The page’s model: the air left in the flask and how full it ends

Taken Apart

Seven things: the flask, the stopper, the jet tube, the dropper, the clamp, the stand and the beaker. What matters most is that the stopper and the tubes do not leak, and that the flask is round-⁠bottomed, thick-⁠walled and dry.

Flask · stopper · jet tube · dropper · clamp · stand · beaker