Concentration of Oxygen in the Air

This is an attempt to use a microscale method to measure the concentration of oxygen in air. The method is still not perfected. I would welcome feedback from anyone who would like to try out the method: how successful do you find it? Do you find it consistent? What difficulties do you find? Any suggestions for improvement? Comments and suggestions to Mike Clark,


For monitoring of the temperature of the experiment, I have been using a CBL temperature probe rather than a cheaper digital thermometer or a liquid-in-glass thermometer.



AIM To measure the proportion of oxygen, by volume, in a sample of air.


MATERIALS REQUIRED One graduated 1 mL pipette, pipette filler bulb, digital thermometer to read to 0.1 degrees C, watch or clock, retort stand with clamp, about 0.6 mL of 1.0M solution (1.9 g dissolved in 10 mL of solution) of sodium pyrosulfite, Na2S2O5, (also known as sodium metabisulfite).




Some chemical substances in aqueous solution react with oxygen gas quite readily. Examples include sulfites, iron(II) ions and tin(II) ions. If such a substance were used to remove all the oxygen from a measured volume of air, then the loss of volume of the sample would indicate the proportion of oxygen in the sample.


A volume of air of about 0.7 mL can be held as a bubble in a horizontal 1 mL graduated pipette between two volumes of solution that can absorb oxygen. The volume of the bubble can be read to 0.001 mL (one tenth of a graduation). If the pipette is left for a length of time sufficient that no further change occurs in the volume of the bubble, then loss in the bubbleís volume should represent the volume of oxygen that was absorbed.


What solution should be chosen to absorb the oxygen? Sodium pyrosulfite produces fairly consistent results within a reasonable length of time. Solutions of tin(II) ions are difficult to prepare; they absorb oxygen well when the oxygen concentration of the air is high, but tend to react slowly and unreliably as the concentration of oxygen becomes low. Iron(II) sulfate solutions react slowly at low pH, and at higher pH they may leave rusty deposits in glassware.


A large excess of sodium pyrosulfite should be used. Oxygen is being absorbed at the air-liquid interfaces, the two menisci. A large excess of solute should reduce the slowing of the reaction caused by depletion of pyrosulfite reagent and accumulation of reaction products at the interfaces. Assuming a bubble size of 0.700 mL, then the content of oxygen should be a little over 0.140 mL, which represents approx 6 micromoles of oxygen gas. This amount would be absorbed by the same amount of pyrosulfite ions. This amount would be contained by about 6 microlitres of 1.0M solution. About 0.60 mL of 1.0M sodium pyrosulfite represents, therefore, an excess of approximately one hundredfold.


The volume of a gas changes with temperature, so it is necessary that temperature be recorded as well as volume. It will be assumed that the temperature of air in the bubble is the same as the temperature of surrounding air. The recorded gas volumes can be corrected for temperature, so that the initial and final volumes of gas in the bubble will be at equal temperatures. A thermometer that can read to 0.1 degrees C should be used.

The reactions involved are slow, so preparations should be made to gather observations over an extended period of time.




Using a filler bulb, carefully draw sodium pyrosulfite solution into the one millilitre graduated pipette, up to about the 0.70 to 0.65 mL graduation.


Withdraw the pipette from the sodium pyrosulfite solution, and draw in about 0.7 mL of air. Draw in a further volume of 0.25 to 0.3 mL of solution, so that there is a clear bubble of air trapped between two volumes of solution, and the two menisci at the ends of the bubble lie both within the range of the graduations.


Hold the pipette vertically for about a minute to allow any film of liquid around the bubble to run down. Then holding the pipette in a horizontal position, work quickly to remove the filler bulb, and tilt the pipette slightly and gently so that solution runs into the tip without dripping out. Clamp the pipette with the tip tilted downwards at a very small angle, and promptly read the positions of the two menisci to the nearest tenth of a graduation, that is, to 0.001 mL.


Record the time, and the room temperature to 0.1 degrees C.


The pipette should be monitored for about forty-eight hours: the positions of the meniscus and the volume of the bubble should be recorded periodically. As the rate of shrinkage of the bubble slows down, the temperature should also be recorded. This must continue until the volume of the bubble ceases to shrink.


The volume of the bubble may vary as the room temperature changes. Towards the end of the reaction, if the temperature changes between two readings, it will be necessary to correct the volumes read to the initial temperature of the experiment, using the Charlesí Law equation.

During the reaction, if a small volume of solution flows down into the space occupied by the bubble of air, the pipette should be rocked carefully before a volume is read, allowing the liquid in the pipette to move slowly to and fro until the liquid is swept up, and the gas space is free of droplets of solution. During this process, the bubble must remain intact and no solution may be spilt from the pipette.




Records should be kept of the following observations:

**At each observation: time of observation, the positions of the menisci and thus the volume of the bubble.

**At setting-up, and at observations made towards the end of the reaction time: temperature.



The "volume of bubble" is the difference between the values at the two menisci bounding the air bubble. All pipette readings should have three digits: for example, a reading that is exactly 0.71 should be recorded as 0.710, while a reading that is exactly 0.1 should be recorded as 0.100.


Temperatures should also be recorded to the nearest 0.1oC, so a reading that is exactly 18oC should be recorded as 18.0oC.


As the reaction approaches its end, values for the "volume of bubble" is the volume of the bubble should be corrected to the initial temperature of the experiment, using the Charlesí Law equation.


The "volume of oxygen absorbed" is the difference between the initial volume of the air bubble and the corrected final volume of the bubble.


It may be observed that the rate of shrinkage of the bubble is quite rapid at first, but becomes slow after twelve to fifteen hours, as the partial pressure of oxygen drops.




The percentage of oxygen in the sample can be calculated from the volume of oxygen absorbed compared with the initial volume of the bubble.


The final estimate of oxygen percentage in the air should be presented, therefore, to one decimal place, as, for example, 21.2%, or 21.0%.





There are some sources of error in this measurement:

**an error of 0.001mL in estimating the position of meniscus may have a small effect on final estimates of oxygen percentage.

**errors of as little as ±0.1oC in reading temperature may also affect volume readings, and therefore estimates of oxygen percentage.

**it is assumed in this procedure that oxygen is the only component of air that can vary and so affect the volume of the bubble. No account is taken of the water-vapour content of the initial air sample or of the final bubble, nor of the small partial pressure of sulfur dioxide in the airspace over the pyrosulfite solution.

**accuracy of results may be affected slightly by changes in atmospheric pressure over the time of the experiment.



This procedure should still be able to yield a proportion of oxygen in an air sample of between 20.0% and 22.0%. If the procedure is replicated several times, an average estimate could be made for percentage oxygen content of air.

Questions? Comments?

Mike Clark