This little story illustrates how a large group of scientists may go terribly wrong, even with a simple thing such density measurement of water. Water has a density maximum at 4°C. This fact is well established scientifically [1], and is taught in most introductory chemistry and physics courses, together with the amazing consequence that fish can survive winters at the bottom of lakes. To reproduce this fact in a general chemistry laboratory has its educational value, and this is precisely what we and our predecessors thought when organizing such a laboratory course at our university.quit()
During a major overhaul of the general chemistry laboratory, which happened around 2006-2009, we have replaced many of the previous experiments, as they did not properly function, were of poor educational value, or too dangerous for first year students. However, one existing experiment quickly survived the selection, and was immediately carried over into the new program, namely, the measurement of the density of water. The setup is simple. A round bottom flask is filled with water, and closed with an adjustable stopper to which a graduated pipette is fitted into, see photo. One must pay attention that the water meniscus rises into the pipette and that no bubbles are trapped. The flask in then inserted into a larger beaker filled with water, to which one adds ice to adjust the temperature to 0°C. Once the ice melts, the temperature rises slowly, one can accurately measure the change in the enclosed water volume though the height of the meniscus. The density maximum manifests itself by an initial decrease of the height of the meniscus to 4°C and its subsequent rise when the temperature increases further.
We initially used a 50 mL round bottom flask and a 2 mL graduated pipette. Each year, this experiment was performed in parallel by about 20 student groups, the students turned in written reports, and these reports were graded by the teaching assistants. Moreover, two teaching assistants were assigned to each experiment, including this one. They had to carry out the experiment prior to the student laboratory themselves and present their results to us. All students and assistants reported the density maximum.
For about 6 years, everybody was happy with this experiment. The catastrophe happened around 2013. A new assistant was assigned to this experiment, and he claimed that the experiment does not work. Our first reaction was patiently to explain that he must have made a mistake, since this experiment was working perfectly fine for a long time. He insisted by presenting a simple calculation to us. From tables [1] one finds that the relative difference in the densities of liquid water at 0°C and 4°C is actually only 0.013% [2]. For the given setup, this makes just a volume difference of 6.5 µL. This value is quite a bit below the resolution of the pipette used, which is 20 µL at best. So how do you want to measure the density maximum? For that you would need about a 10-times better resolution. Again we said you must be making a mistake somewhere. Nevertheless, we repeated the calculation ourselves. The guy was right!
In the meantime, the experiment was modified and is now finally working properly. It turned out to be quite an effort to adapt this experiment for the students. We could not find commercial graduated pipettes, which would be appropriate, but long glass capillaries would finally do the job. To their back, we attached a graduated paper strip. We further had to use a larger round bottom flask and a larger beaker. With this setup, the density minimum shows up quantitatively as a dip in the meniscus height of about 2 cm. You must indeed have a sensitive setup, as one is looking for a really small effect. We should have been skeptical from the very beginning.
How to properly measure the density minimum in a student laboratory is not the interesting part of this story. The interesting part is how on earth it was possible that hundreds of people seemed to have observed an effect, which was impossible to observe with the equipment available? Was this mass hypnosis? But all these people are scientists, and they should be immune to such errors. Apparently they were not. Scientists are humans, and they can be misled. Yes, there was the established fact that the density maximum exists, and who would ever stand up against that? Yes, there were generations of students and assistants, which have (apparently) all observed this effect with the same equipment. And yes, when I do not report the effect as a student, I could get a poor grade, or when I complain about the experiment as an assistant, I might have to fix it. When a student actually criticized the experiment, maybe he or she was told to look more carefully. Possibly, one could qualitatively observe a minute dip in the meniscus involving a fraction of mm. But the assistants had to do the experiments every year and to present their results. Apparently, we have not looked carefully enough. Maybe, most assistants did not have a maximum among their data points, and we took it for such. Or did some tweak their data a bit?
What is amazing is that the problem was there over years, but neither an assistant nor a student reported it. Hundreds of people must have seen the blunder, but no one dared to stand up. Maybe few did, but they got silenced underway. Good, the problem was finally found and fixed.
Nevertheless, the story leaves an uncomfortable feeling. This issue was so terribly trivial, totally wrong, but still, hundreds of people did let it slip by. Are there possibly other problems in our general chemistry laboratory 50that are going by unnoticed? Could there be similar problems in the lecture material? In more advanced courses and laboratories, such issues are surely much harder to track. What about scientific results obtained within research groups? Could similar mechanisms be operational as well? Nobody bothers to reproduce an experiment hundreds of times. But still, even reproducing it so many times may not help. When expectations are there, one has the tendency to fulfill them. One likes to please your colleagues, supervisors, reviewers, or wants to stay out of trouble. When you contradict the accepted wisdom, there is the tendency to get silenced. In fact, finding a correct answer might be a huge challenge.
Michal Borkovec, December 30, 2014
Updated, October 26, 2023
References
[1] Properties of water - Wikipedia, the free encyclopedia, http://en.wikipedia.org/wiki/properties_of_water
[2] The this number was wrong in a previous version of this post, and following an anonymous commentator is now corrected. Many thanks for pointing this out! The respective comment is now deleted.
There are no comments on this article yet. You may enter one with the button below.