Of course I will take all of this information in to consideration when I’m building my experiment with a mirror (or sheet of glass) and I’m not so interested in a dynamic collection of data showing how condensation collects on the surface, and this has proved very useful, but I had to rectify the problems so I could carry on designing a preliminary experiment.
Preliminary experiment 2
To increase the volume of condensation on the surface I decided that increasing the surface area was my best option, a way of getting around using a large sheet of glass (which I’m trying to avoid) I decided that using a round cylinder shape would maximise the surface area without necessarily increasing the area to subject to steam. A measuring jug would be easy to heat using a nicrome wire, simply I ran a preliminary test just applying a measuring jug steam (by boiling water with a Bunsen burner) and I had a significant mass of condensation collected.
This brought up the problem of getting the steam inside the jug, obviously I would not be able to do this using a kettle and this also tied in with the reasons stating above for building my own steam generator. In the chemistry labs for experiments using kwikfit apparatus, we often use heating mantles to heat up samples of chemicals ready for refluxing or distillation, this is perfect as the heating mantles are totally adjustable to the temperature of the water and I’m fairly certain will produce a more constant jet of steam. The round bottomed flask is easy to refill and it is very easy to attach a bung and tube to the top of it to direct the steam flow. I will have to take into consideration that steam will be condensing inside of the tube but its no serious problem so long as the tubing is replaced on each attempt to keep the initial temperature of the tubing the same.
Preliminary experiment 3
While in a chemistry lesson I came upon a every day piece of apparatus which would serve the purpose of condensing water perfectly. A kwikfit condenser, using a condenser I would be able to directly link it to the heating mantle to channel the steam into the condenser with minimum heat loss to condense the steam, and use the condenser in reverse (where in reflux experiments the condenser is used to condense fumes to return to the original boiling sample) I will be condensing the steam and channelling it out into a beaker on a balance, this balance will be connected to a laptop to give me a idea of how much condensation is being collected each minute. The difficulty came with controlling the temperature of the condenser, in normal practise, pressurised cold water from the tap is run through the condenser to condense almost 100% of the vapours passing through, I will need to run water of a regulated temperature through the condenser, the best solution I found for this is by siphoning it from a water bath (which I can control the temperature of directly) Using the condenser also has another advantage, as the water is constantly passing through the condenser, it will keep the condenser at a constant temperature and stop any fluctuations due to the steam. I will be able to monitor these fluctuations using a thermocouple, which I intend to build and put up the inside of the condenser. The experiment will look like this.
Once I have gained the data I need from this experiment I will be able to proceed with the next experiment.
Risk assessment:
- As I increase the temperature of the waterbath, the condenser and the surrounding apparatus will conduct the heat and become warmer, I will have to be careful around handling this equipment, and for any temperatures above 40 degrees I will have to be sure the waterbath has stopped pumping water through the condenser to allow it to cool down to room temperature.
- As the kwikfit i’m using is glass, if any of it is dropped it will shatter, in this case I will need to have protective gloves and a dustpan and brush handy to avoid cutting myself.
- I will need to be extra careful around the computer with the water, because if any of the water sprays over the computer I will be in danger of short circuiting the lab and in danger of getting a electric shock. (and of the physics department when they find out ive fried their computer!) To avoid this I will place the computer on a separate table.
- I will need to be aware of any spilt water, so I will be careful when I’m walking around the apparatus, I will keep some paper towels nearby so that if any water is spilled on the floor it can be cleaned up to avoid anybody slipping.
- Due to the siphoning of the water, the tubing and the condenser will be pressurised, I will need to make sure that the tubes will not blow off due to the pressure, or due to any blocked tubes (this will involve a rather messy preliminary experiment)
Day 1
A problem I identified before I started the experiment was that the water siphoning from the waterbath would eventually run out, and working at higher temperatures it would not make sense to keep adding large quantities of cold water (which would offset the balance of temperature diversely affecting the rate of condensation in the condenser) I decided upon recycling the hot water, and somehow getting it back into the waterbath. After exploring the possibilities, the solution I arrived at was to collect it in a basin at a lower level to the water bath, then use a pump to pump it back up into the waterbath. Ideally I think that siphoning the water from a water bath, down into another water bath is more suitable for keeping the temperature of the water constant, but as the water is not in the basin for long periods of time, it does not have long enough to cool significantly before being pumped back up to the waterbath. I obtained a bilge pump for emptying dinghies; this proves very efficient and is heat tolerant.
After testing my equipment and making sure everything was working properly, I set around running a preliminary experiment and devising a suitable method for working out the percentage yield of water collected.
The basic experiment was evaporating as much of the 50 ml3 of water as I could in the space of 15 minutes (and the condenser at a low 30 0 ) and measuring the final volume of the condensed water collected.
In the space of 15 minutes, I boiled off 31.16ml of the original 50 ml3 and 28.02 ml of this condensed and ran in to the beaker. This results in a 98.8% yield.
As the amount of steam being ejected from the outlet past the condenser was negligible at 300, its same to assume that the rest of the boiled off water is condensed inside the rest of the apparatus (the condenser etc.) I worked this out to be 3.14g remaining inside the apparatus. This was not going to be easy to get out, and would waste valuable time drying the insides of all the condensing apparatus for the repeats.
I decided that checking that the flow of steam would be important, if the flow of steam is constant then I think its safe to assume that the amount of condensed water inside the condenser is constant due to a constant rate of the condensed water leaving the condensed apparatus.
This is a simple experiment to measure the rate of steam flow, it simply involved having the round bottomed flask filled to 100ml and put into the heating mantle, then every 5 minutes taking the round bottomed flask out to measure the change in mass due to the water evaporating.
The results are as follows.
The heating mantle can clearly be seen spending the first 10 minutes heating up the water to boiling point, in which the rate of water evaporating was non linear, but past 10 minutes the rate of evaporation can clearly be seen as linear and has the equation
y = -2.189x + 115.27
As the last 30 minutes are practically completely linear, I think its safe to assume that to take the complete mass of water evaporated between 10 and 40 minutes, I can work out a fairly accurate average of mass of water boiled off each minute.
93.28 – 27.59 = 65.59 grams of water boiled off in 30 minutes
65.59 / 30 = 2 559/3000, which is approximately 2 1/5 g of water boiled off every minute. As this figure is a average, it means I can conduct my experiment any time between the first 10 and 40 minutes, and I can be fairly sure that 2 1/5 g of water is being boiled off every minutes.
I decided to work out the figure for every 10 minutes (which is the length of each of my experiments) 2 559/3000 x 10 = 21 13/15 grams of water boiled off during each experiment.
Day 2+3
As I was able to assume that the first 30 minutes of steam is constant, I can set around doing my experiments, but repeating them back to back. I see no problem doing this if I collect my samples after the first 10 minutes when the rate of steam being produced is constant. If I collect my samples of condensed water for 10 minutes, I will be able to repeat the experiment 3 times within the first 40 minutes, and using the graph above (or using a constant) I will be able to calculate the mass of water boiled off in each repeat. I think this method is also beneficial because it saves a lot of time, as it takes about half an hour to setup each experiment, and this also reduces the problem of the water that condenses inside the condenser and does not channel out into the beaker on the balance. At least the water inside the condenser will be constant through the repeats.
On days two and three I set around gathering data for the two temperatures 300c and 400c
I discovered that the temperature the waterbath is set at, is actually 50c out, so I had to adjust the temperature ranges to 350c and 450c
The results are as follows
I plotted the error bars on the graph by working out the spread of the results then averaging them, but the error is generally so small it is indistinguishable on this graph, the general trend of the individual strands of data are what I would have expected, the graph is clearly linear showing that a constant mass of water was being collected but, although this is only a 10 degrees temperature difference, the difference in gradients is much smaller than I had hoped, I have identified a flaw in my experiment and I’ll try to show it in the drawing below.
This is a drawing of my kwikfit setup, I set it up with the intention that the steam would travel through the condenser, some of it would condense (dependant on the temperature of the condenser) and the rest of it would channel out of the steam out valve. At 30 degrees, the condenser was so efficient it condensed pratically 100% of the steam, and so of course no steam was being ejected from the steam valve. This was no problem, but at 40 degrees, I noticed that the steam that had not condensed, instead of escaping through the steam out valve, it was actually condensing in the rest of the kwik fit apparatus. This is highlighted in the picture above, and the reason why this is happening is because the kiwkfit on the left is at room temperature, so that makes it exponentially more efficient at condensing steam because room temperature is around 21 degrees. In normal cases I’d expect the apparatus to conduct heat through and to develop a equilibrium with the condenser, but this obviously isn’t happening, most likely due to the fact that glass is a bad conductor of heat (and that the kiwkfit parts are separate bodies) This raises the question to whether the temperature of the tube that the steam runs through inside the condenser is actually reaching 300c, or even while steam is running through the condenser, the surface of the tube is forming an equilibrium between the temperature of the steam and the controlled temperature of the water running through the condenser.
I think the obvious solution to this is discard the steam out valve and the drop tube for the condensed water, and to just allow the condensed water to leave the condenser and be collected straight away in the beaker.
Day 4
I ran the modified experiment once again at 45 degrees, with the hope that more steam would be emitted due to it not being able to air condenser, the results actually indicated that the entire system, now simplified, has become more efficient, which is not what I had expected.
Series 2 is with the modified apparatus, and series 1 is with the original apparatus. I have a feeling that there is such a large difference in the results due to air pressure affecting the boiling water, but I cannot be sure until I have conducted the experiment with the modified apparatus at several different temperatures.
At this point, I think its nessecary to do the calculations to work out the % yield of water condensed at 45 degrees, in the 10 minutes, using my figure that I calculated earlier (21 13/15 g/10 minutes) I worked out the % yield of water vapour condensed and collected as below.
(23.07/[21 13/15])*100 = 105.5 %
This result is, odd, to say the least, only two reasonable explanations fit this result.
- Excess water has funneled out on to the balance, most probably the water inside the condenser at the begging of the experiment added to that of the water that was trickling out on to the balance, this of course does not agree with my theory above stating that there is most likely an equilibrium of water within the kwikfit apparatus.
And after working out the % yield of the experiment with the original apparatus..
(20.44/[21 13/15])*100 = 93%
- The air pressure could have diversely affected the rate of evaporation and condensation, enough to give a 7% increase in water vapor.
I believe that the second bullet point is the most reasonable, but as I did not foresee this problem to be as a large an obstacle as it is, I did not account for pressure when working out the steam rate, so I cannot be sure if this is the exact reason for the inconsistency in my results. Given the time, and a barometer, (depending on whether the air pressure changes over night) I would run a steam rate test at the beginning of each day so I could be certain with the % yields of water collecting.
I decided to run my experiment at one last time to see if I was going to be successful in having any serious rift in the % yield of water vapor condensed. I decided to run water of 60 degrees through my kwikfit apparatus, this really is the top end that the leibig condenser will be able to handle, and any higher than this, I predict that the rubber tubing will expand and weaken so much, that the connections to the condenser will not stay taught and leak.
The results of this experiment are as follows.
Above are the results from trail one plotted against the results I gained when I modified my experiment, they fitted the trend I expected, but with only a 2 g difference of water collected, this difference is too minimal to analyse, and I decided that as I was collecting practically 100% of the water evaporated, changing the temperature to 0 degrees using ice would have no effect on the water collected, because I cannot (in theory – as disproven earlier) collect more water than 100% yield which I have already established at the highest temperature I can go which is 60 degrees. I feel now that any temperatures that would make a significant difference in the amount of water condensed would be 100 degrees plus.
Day 6
I decided to explore my other original idea of using light to detect a change in water vapor content on the surface. I set around devising a suitable experiment by means of which I could evaporate steam onto a glass surface then shine light through the glass surface and use a LDR to monitor the effect on which the condensed water vapor had on the light.
The solution I came up to this experiment with is shown in the diagram below.
- It was necessary to use a closed container because so long as the background light in the room is constant, I can be fairly sure that the same amount of light is entering the container via the exposed glass at the bottom of the round bottomed flask (I light sealed the rest of the flask using insulating tape)
- I decided to abandon the steam generator I used in my last experiment, because I would need to suspend my glass surface above the steam source, so as to collect the most steam possible, but when the surface becomes saturated with water vapor, it would start to drip, and if water got inside the heating mantle, it would short circuit.
- As controlling the temperature of the round bottomed flask would be near impossible, I decided to just monitor the surface temperature because it would inevitable rise as the experiment proceeds, I noticed that because the flask is sealed, the hot air inside it will have nowhere to expand, so I fitted a small bit of tubing down the neck of the flask, and sealed the rest of it up with bluetac.
- I also felt it important to use another temperature probe inside the boiling water to see how it affects the temperature of the probe inside the round bottomed flask
This experiment is entirely dynamic, and the only problem I have is that the heating of the water with the Bunsen is not uniform, but if I keep the Bunsen at the same settings when I repeat the experiment it should not be a problem.
As I was using data loggers connected to a logic live kit, it took 1800 samples, so I decided just to display the results as a graph. The results were not at all what I would have expected, although very small, there is a distinct dip in the light consistency around 200 seconds, this cannot be due to background light changing (such as people walking around) because that is more than a blip, and towards the end of the experiment, at 600 seconds there is a definite increase in light intensity.
What I believe is happening at 200 seconds, is the water is finely condensed on the bottom of the flask, evenly scattering light in all directions, but as more water condenses, I believe that it forms droplets which are otherwise, miniature convex lenses which are helping to reflect light inside of the flask.
I decided to improve on this, by reduce the surface area for the steam to condense on, I figured that if there is less light getting in, then the light probe inside will be more sensitive to changes in light, the bottom of the flask now looked like this.
