Results >>
Bottom

Trial 1:

Bottle #	Temperature (°C)	Time (hrs)	Comments
1	-3(high), -5(low)	12	Half full, froze immediately when shaken
2	-3(high), -5(low)	24	Quarter full, did not freeze when shaken or when ice was added
3	-3(high), -5(low)	12	Three quarters full, did not freeze when shaken but froze when ice was added
4	-3(high), -5(low)	24	Was unopened, did not freeze when shaken but froze when ice was added
Test	-3(high), -5(low)	24	Froze three quarters

The first day/trial was not successful in supercooling water, since the temperature was only a few degrees below zero. Only three out of four water bottles supercooled. Two of the bottles required a seed crystal to begin the crystallization process.


This is the positioning of the four bottles during trial one. The temperature is only -2°C because the thermometer did not have time to cool down yet at the moment when I took the picture.

Trial 2:

Bottle #	Temperature (°C)	Time (hrs)	Comments
1	-2(high), -5(low)	24	Half full, did not freeze when shaken but froze when ice was added
2	-2(high), -5(low)	24	Quarter full, did not freeze when shaken or when ice was added
3	-2(high), -5(low)	12	Three quarters full, did not freeze when shaken but froze when ice was added
4	-2(high), -5(low)	12	Full, did not freeze when shaken but froze when ice was added
5	-2(high), -5(low)	12	Half full, did not freeze when shaken but froze when ice was added
6	-2(high), -5(low)	12	Quarter full, did not freeze when shaken but froze when ice was added
7	-2(high), -5(low)	24	Three quarters full, did not freeze when shaken but froze when ice was added
8	-2(high), -5(low)	24	Distilled, full, froze immediately when shaken
Test	-2(high), -5(low)	24	Froze two thirds

The second day/trial had been even warmer, and many bottles needed seed crystals to start the crystallization process. Seven out of eight water bottles supercooled. Despite the high temperature, this is a better ratio than the previous trial.


This is the positioning of the eight bottles during trial two. One of the bottles is not in the shot.

Trial 3

Bottle #	Temperature (°C)	Time (hrs)	Comments
1	-4(high), -9(low)	24	Half full, froze immediately when shaken
2	-4(high), -9(low)	24	Quarter full, froze into ice overnight
3	-4(high), -9(low)	24	Three quarters full, froze immediately when shaken
4	-4(high), -9(low)	24	Full, froze into ice overnight
5	-4(high), -9(low)	24	Half full, froze immediately when shaken
6	-4(high), -9(low)	24	Quarter full, froze into ice overnight
7	-4(high), -9(low)	24	Three quarters full, froze immediately when shaken
8	-4(high), -9(low)	24	Distilled, full, froze into ice overnight
Test	-4(high), -9(low)	24	Froze fully

The third day/trial was by far the best. The temperature had ranged from -4°C in the afternoon to -9°C at night. However, only half of the water bottles supercooled. The reason for that is that I had used the same water for all the trials in order to control the experiment better, and seed crystals may have gotten into the bottles overnight.


This is the positioning of the eight bottles during trial three. The picture was taken during early evening. One of the bottles is again not in the shot.

After the water had been left overnight to supercool, it was shaken to test if it was really supercooled. If ice formed when the water was disturbed, then the process had been a success. If no ice formed, then the water in the bottle did not supercool. A higher amount of ice in the water meant that the temperature was very low and the supercooling process was a great success. Lower ice percentage meant higher temperature and a poorer supercooling.
The above graph shows the average percentage of ice in water per trial. On the first trial, the temperature was not very great, thus the average percentage of ice was 23%. The second trial had an average of 16%, a worse result even though there were twice as many bottles. The lowest temperature was achieved on trial three; therefore there was more ice in the water (not counting the fully iced bottles which were failed attempts).
The above chart shows the success rate of supercooling per trial. In the first trial 75% of the bottles supercooled, in the second trial 88%, and in the third only 50%. If I would have used a different batch of water, the last trial would have had better results. If the temperature was significantly lower however, it would be harder to control.

Top