Producing liquid oxygen in the classroom

Badminton School takes no responsibility for injury or damage resulting from following any procedures or techniques described in this document or for using any materials or equipment referred to. Liquid oxygen is a very dangerous material to handle. Fires involving liquid oxygen are practically impossible to put out and some materials—particularly organics—can explode on contact with the substance. It is the reader's responsibility to produce a suitable risk assessment and use this to inform their choice of materials and a protocol for working with liquid oxygen.

The authors would draw readers' attention to safety test that they undertook to investigate the combustibility of a latex balloon containing liquid oxygen described at the end of this article to illustrate how severe a fire involving liquid oxygen can be.

Work with liquid nitrogen should only be undertaken by appropriately trained people as it presents many dangers including cryogenic tissue damage and the risks of asphyxiation and pressure related explosions as the liquid nitrogen gradually boils away. Badminton School cannot take any responsibility for injuries sustained when working with liquid nitrogen. Key factors that should be considered include:

• The use of appropriate personal protective equipment
• The use of appropriate equipment to avoid touching dangerously cold surfaces
• Working in an environment with adequate ventilation
• Working with at least one other person
• Alerting colleagues and other users of a building that liquid nitrogen is present
• Ensuring that liquid nitrogen is never stored in a sealed container

Most classroom methods observed by the author require the use of an oxygen generator such as hydrogen peroxide mixed with manganese dioxide. Using a chemical generator, it is likely that more gaseous oxygen will be produced than is needed and it can be difficult to extinguish the reaction when enough oxygen has been produced. Better equipped laboratories have access to oxygen cylinders from a supplier of lab gases, but these are generally very expensive considering the volume of oxygen that they store.

The proposed method of making available sufficient oxygen involves the use of oxygen from disposable welding gas bottles. Oxygen from such a source has been used by the author successfully and the purity of this source is usually cited as greater than 99%. It is the reader's responsibility to ensure that they obtain a suitable supply of oxygen gas.

Small gas bottles are available from suppliers such as Weld UK (http://welduk.com/) at a typical price of £25 for a cylinder that contains 110 l of oxygen and £40 for an oxygen-safe regulator with a pressure gauge (figure 1). Please note that this is not a recommendation; readers should satisfy themselves that they can obtain a suitable supply of oxygen.

Zoom In Zoom Out Reset image size

An appropriate risk assessment must be conducted before working with gas cylinders. Please note that it is critical that a pressure regulator designed for use with oxygen cylinders is used and that this regulator is used exclusively with oxygen cylinders to avoid the risk of auto ignition/explosion.

Other than the obvious risks of dealing with cryogenic substances, the main safety hazard associated with making liquid oxygen observed by the author is that usually a reasonably large quantity of oxygen escapes during the production phase of the demonstration. In particular, introducing gas into a copper condensation coil often forces droplets of liquid oxygen to spit out of the coil in an uncontrolled fashion. The oxygen produced will also need to be contained in an open vessel and will boil away quite quickly. Creating an oxygen enriched atmosphere may increase the risk of materials catching fire or intensify a fire and is therefore undesirable. To avoid creating an oxygen rich atmosphere around the demonstrator, a closed system is used.

The first step in the procedure is to fill a balloon with oxygen from a welding gas bottle using an oxygen-safe regulator (figure 2). There are many sources of balloons, but large clear sausage balloons from Just-So Occasions (www.justsooccasions.co.uk/) have been found to work well. Please note that this is not a recommendation; readers should conduct their own research as some materials may spontaneously combust when in contact with liquid oxygen.

Zoom In Zoom Out Reset image size

The balloons used hold approximately 6 l of oxygen and are made from latex that seems to be stronger than similar balloons; balloons from this source can be purchased for approximately 50p when purchased in bulk. When fully inflated, the balloons used are approximately 1.5 m long and they fit into the mouth of a 1 l glass demonstration Dewar with a few millimetres of clearance at the sides (figure 3). Filling the balloon can be accomplished well ahead of the start of the demonstration, allowing for the gas cylinder to be put back into storage.

Zoom In Zoom Out Reset image size

To conduct the demonstration a 1 l glass demonstration Dewar is three quarters filled with liquid nitrogen. If more liquid nitrogen is added it is difficult to immerse the balloon without considerable splashing of liquid nitrogen as the balloon is fed in, raising the risk of cryogenic injury to the demonstrator. Using less can result in running out of liquid nitrogen towards the end of the demonstration as the balloon needs to be re-immersed in the liquid nitrogen frequently because the oxygen inside it boils off relatively quickly.

Next the balloon is gradually fed into the Dewar (figure 3). The reasonably large surface area in contact with the liquid nitrogen combined with 13 K temperature difference between the nitrogen's boiling point and the oxygen's condensation point means that fully immersing the balloon is a reasonably quick process.

It is possible that the latex may perish either while cooling or warming the oxygen. The demonstration should be conducted away from sources of ignition. The quantity of oxygen used is relatively small and should rapidly disperse when vented, negligibly increasing the room's concentration in the event of the balloon failing. Should any liquid oxygen escape, this method only produces a small amount of liquid oxygen and most, if not all, of the liquid will have boiled away before coming into contact with any surface.

Once most of the oxygen has condensed, withdraw the balloon and, assuming that you are using a clear balloon, a pale blue liquid can clearly be seen inside it (figure 4) especially when back lit.

Zoom In Zoom Out Reset image size

While the oxygen is liquid, the balloon will be drawn to a strong rare earth magnet (figure 5) showing its paramagnetic nature (The Royal Institution 2013). The oxygen boils away reasonably quickly, so you will need to regularly re-immerse the balloon throughout the demonstration. The magnet illustrated is a '50 kg pull' neodymium-iron-boron magnet. Even with a magnet of this strength, the effect is quite weak.

Zoom In Zoom Out Reset image size

It is vital that a risk assessment of the use of strong magnets is conducted when using this type of magnet as it can exert sufficient force to cause serious injuries to fingers and produce sharp fragments if it shatters after an impact.

Many strategies can be used to dispose of the oxygen inside the balloon safely at the end of the demonstration. Venting of the balloon in the demonstration room will produce a negligible increase in the concentration of oxygen in the room's air. One possible method is to cut the end of the (still deflated) balloon off and place the now open balloon in to a glass beaker which can be left somewhere away from ignition sources. The liquid oxygen will gently boil away with any splashes being contained by the beaker.

As mentioned above when working with liquid oxygen a thorough risk assessment must be undertaken.

As part of our risk assessment heat was applied to latex balloons containing liquid oxygen to see how difficult it was to ignite them (figure 6) and how severe the resulting fire would be.

Zoom In Zoom Out Reset image size

It is left to the reader to undertake a suitable risk assessment and to decide on any tests, along with the methodology that they wish to implement and the following should not be considered as indicative of what may happen with other batches of balloons or other oxygen cylinders in circumstances other than those prevailing in the laboratory at the time the test was undertaken.

With the balloons used and under the prevailing conditions, it was found to be very difficult to ignite the balloon from the outside by applying a lit spill (figure 6). However, once the balloon was ignited, it was observed to burn extremely intensely (figure 7) creating notable warming to the skin even when stood several metres away. The intensity of the fire and the high temperature apparently generated suggest that it would be advisable to take precautions against allowing a balloon containing liquid oxygen to ignite.

Zoom In Zoom Out Reset image size

The authors would like to thank the Headmistress (Mrs R Tear) and the Board of Governors for fostering an environment that encourages research and enables science communication in the local community.

We would also like to thank the Science Faculty at Badminton School for their help, in particular Karen Noonan for all of her support in preparatory work for experiments and both Paul Foster (Head of Physics) and Victoria Goldsack for help with initial trials of the demonstration.

This work would not have been possible without help from a number of students, including Mary Cheng who helped with the preliminary research, Sarah March, Elena Bashkova, Jing Wen Hillary Wang, Natalie Tse, Julia Gay Torne and Livia Eichmeyer who helped with documenting the process.

The research has been undertaken using equipment purchased in part by a grant from the Institute of Physics' School Grants Scheme. We are extremely grateful for the Institute's support which has enabled more than 600 junior schools pupils to see the liquid oxygen demonstration in the past year.