Introduction:
The idea for this investigation came from my childhood experience. When I had fish, specifically goldfish as pets at home, I used to wonder why they needed air to breathe when they live underwater and thought it was such a pointless thing to put an air pump in the water tank. Later I came to be aware of the importance of oxygen, and as living things including human, animals and plants, it is an essential part of our water body health. I later found out that even animals underwater such as fish and bacteria need dissolved oxygen which refers to the oxygen that is dissolved in water, to live. However, in recent years, water pollution is becoming one of the main problems that are causing many fish to die because of a decrease in dissolved oxygen in the water. By investigating my topic, I am able to explore the effectiveness of dissolved oxygen in a natural environment as well as understanding the importance of providing a suitable living condition for the animals to survive in a present-day.
Hypothesis and Background Science:
I hypothesise that as the temperature increases, the dissolved oxygen in the water sample decreases. I formed this hypothesis using my knowledge from the course as well as some of the background research that I’ve done. The amount of dissolved oxygen that water can hold depends on the temperature and salinity of the water, however, since I am going to use fresh water for my investigation, I will only be focusing on temperature effect. Since oxygen is less soluble in warm water than in the cold, there will be less oxygen as the temperature rises which could cause some of the certain kind of fish to die.
The oxygen dissolved in water can be referred to as a solution which is a type of homogeneous mixture composed of two or more substances, in this case, the solute is O2 molecule that is being dissolved, and H2O is the solvent, the substance that dissolves the other substances. In water (H2O) molecule, each hydrogen atom shares a pair of an electron with the oxygen atom. Since oxygen is more electronegative, it attracts electrons towards to its nucleus, creating a polar molecule due to its uneven distributed shape. However, on the other hand, O2 is a non-polar molecule because of its symmetrically distributed electron cloud between the O2 atoms.
When O2 molecule enters the water, the slightly negative end of H2O molecule as shown in Figure 1 approaches it, the electron cloud of O2 molecule isolate itself to reduce the negative repulsion. This results in the weak intermolecular force that’s holding the O2 bond gets broken, and the large solute turns into smaller individual particles, causing O2 and H2O to become weakly attracted to each other. This is the formation of a solution which specifically is a physical process; therefore, the O2 molecules can enter the water from the atmosphere or as a byproduct of the photosynthesis of aquatic plants and they can shuttle back and forth between the atmosphere and the water.
The Winkler Method is a technique used to measure the amount of dissolved oxygen present in freshwater systems. The results of the experiments can determine the health of the water body as well as predicting the oxygen-related activities in the water.
Figure 1
The diagram taken from Water – Structure and Properties.
Variables:
Independent variable: Temperature (°C)
The temperature of the sample can be increased using water baths. I have managed to reach five temperatures which are as follows; room temperature (23 °C), 45 °C, 54 °C, 60 °C and 67 °C. To heat the water samples, put enough sample into a glass bottle to make sure that it doesn’t float in the water baths. Place the container in the water baths without a lid on to stimulate the environment of a lake so that oxygen is free to move around. Then keep checking the thermometer and when it reaches just above the target temperature (to avoid it from decreasing while I titrate the sample solution), take the bottle out from the water baths.
Dependent variable: Dissolved oxygen content (ppm)
The dissolved oxygen content can be measured by iodometric titration which is a method of volumetric chemical analysis using a redox titration as well as the iodine test where the disappearance of the dark blue colour (from the starch solution) indicates the end of the experiment. For each temperature, I did two trials to ensure reliable results and these data will be used to calculate the oxygen content.
Table 1 – Controlled variables of the experiment
Controlled Variable
How is it to be controlled/manipulated (for control – what value will be held?)
Rationale – why control is needed and why this method was chosen
Temperature control
Temperature affects the amount of oxygen that can be dissolved in fresh water.
When heating the water samples, I always checked the temperature before titrating and noted down
Volume of the water sample in a conical flask
The volume of the water sample affects the amount of sodium thiosulfate solution needed to titrate the sample.
I made sure to use a measuring cylinder (+/- 0.05 ml) to measure the volume of the sample, in order to ensure accuracy.
Volume of the solutions added
The volume of each solution added affects the amount of sodium thiosulfate solution needed to titrate the sample.
I used a gratitude pipette (+/- 0.5 ml) to ensure that I put the exact amount of solutions into the sample.
Source of water sample
Different ways of filling water sample containers affect the amount of oxygen dissolved in water.
By slowly filling the bottle with fresh water until it is overflowing, so that any oxygen flowing in the air won’t dissolve into the water.
Detection of end point of titration
Determining each point of titration such as how pale is pale yellow, etc. is affected by visual perception.
I took photos of the sample before and after titrating to make sure that I get the right colour for every trial.
Methodology
Apparatus:
• 1 x water tank with an air pump
• 350 ml glass bottle
• 4 x plastic pipettes (+/- 0.5 ml) – one for each solution
• 200 ml conical flask
• 250 ml beaker
• 100 cm3 volumetric flask
• 1 x burette clamp
• 1 x retort stand
• 1 x burette
• 1 x funnel
• 50 ml measuring cylinder (+/- 0.5 ml)
• digital pH meter
• glass thermometer
• 50 ml of distilled water bottle
• 1 x 50 ml – 1.00 mol.L-1 manganese sulfate solution
• 1 x 50 ml – sodium iodide solution
• 0.003 mol.L-1 sodium thiosulfate solution (made by diluting a 0.006 mol.L-1 solution)
• 1 x 50 ml – 9 mol.L-1 sulfuric acid
• 1 x 50 ml – starch indicator (1g of starch boiled with 100 ml of water)
Table 2 – Chemicals used in this experiment and its safety measure
Chemical used
Safety measure (hazards and environmental concerns)
1.00 mol.L-1 manganese sulfate solution
The diluted solution therefore low toxicity, however gloves were used when handling.
sodium iodide solution
Low toxicity and can be irritant to the eyes – gloves and safety glasses were worn.
9 mol.L-1 sulfuric acid
Highly corrosive and is potentially explosive due to its concentration and the strength of the chemical. It can cause severe skin burns and irritant to the eyes, etc. therefore, gloves and safety glasses were worn whenever handling it.
Sodium thiosulfate
The solution was diluted, therefore it was not as harmful – gloves were still used.
No contact with oxidants – risk of an explosion
¬ All of these chemicals can be poured down the sink as they are inorganic chemicals and are not harmful to the environment.
Experimental Procedure:
This method was adapted from educational resources, Microbial Life, entitled ‘The Winkler Method – Measuring Dissolved Oxygen’ by Monica Z. Bruckner (Montana State University, 2018).
https://serc.carleton.edu/microbelife/research_methods/environ_sampling/oxygen.html
Except where I have altered the method: I used sodium iodide solution instead of an alkali-iodide-azide as preparing sodium iodide solution is easier than alkali iodide-azide.
Before starting the experiment, ensure to put on lab coat, safety glasses and gloves when handling chemicals.
Preparation of sodium thiosulfate solution:
1. Measure 50 ml of 0.006 mol.L-1 sodium thiosulfate solution by using a 50 ml measuring cylinder (+/- 0.5 ml).
2. Add the above solution into a 100 cm3 volumetric flask, then fill it up to the blue line with distilled water to dilute it.
3. Mix the solution well.
I have diluted the sodium thiosulfate solution (from 0.006 mol.L -1 to 0.003 mol.L -1) to get a high accuracy data.
1. Slowly fill a 350 ml glass bottle with water from a water tank until it is overflowing.
2. Immediately, add 2 ml of manganese sulfate solution followed by adding 2 ml of sodium iodide solution, both using two different plastic pipettes making sure to put the tip of each pipette into the bottle so that only the original water overflows.
3. Carefully replace stopper to ensure that no air is being introduced. Mix the bottle by inverting several times. If oxygen is present, a brown precipitate will appear, then allow the sample to settle down. (Picture 1)
4. Add 4 ml of sulfuric acid into the mixture via a pipette, again by allowing the sample to overflow. Restopper the bottle and mix the contents well by inverting until the brown precipitate disappears. At this point, the sample must be titrated within 4 hours.
5. Take a 50 ml sample of the solution using a measuring cylinder (+/- 0.5 ml) and place it in a 200 ml conical flask. (Picture 2)
Since I wanted to stimulate the environment for the fish to live in such as a lake or a river, I filled the bottle from a water tank with an air pump and not straight from a water tap, to make sure that the water I collect have some oxygen dissolved in it.
(Picture 1) (Picture 2)
These photographs were taken by me using an iPhone XS, on 19/10/2018, during the experiment.
Titration of sodium thiosulfate:
1. Fill a burette with diluted (0.003 mol.L-1) sodium thiosulfate solution up to the top (0 ml). (Picture 3)
2. Titrate the sample by slowly dropping diluted titrant solution using the burette (+/- 0.5 cm3), continually swirling the mixture until it fades to pale yellow.
3. Add 0.5 ml of the starch indicator into the conical flask and swirl to mix the solution. It should turn dark blue/black. (Picture 4)
4. Continue adding sodium thiosulfate solution until the blue colour just disappears and it becomes a clear solution. (Picture 5)
5. Read the burette volume from the bottom of the meniscus and repeat the process again with different temperatures.
(Picture 3) (Picture 4) (Picture 5)
These photographs were taken by me using an iPhone XS, on 19/10/2018, during the experiment.
I repeated the process two times for each temperature in order to reduce the random errors. I have also repeated this experimental procedure as well as making diluted sodium thiosulfate when necessary.
Measuring pH of water samples using pH meter:
During the experiment, I also measured the pH of water sample at 35 °C while I titrate the solution using a pH meter. The photograph below (Picture 6) displays the set-up where clamp is being used to hold the pH meter and it was connected to my laptop to plot the graph using Logger Pro.
(Picture 6)
This photograph was taken by me using an iPhone XS, on 18/10/2019 during the experiment.
Raw Data:
Table 3 – Volume of sodium thiosulfate solution used in the titration at different temperatures
Volume of sodium thiosulfate solution used (cm3)
Temperature (°C)
Titre 1
Titre 2
Room temperature 23.0
Initial reading (+/- 0.05cm3)
0.00
22.20
Final reading (+/- 0.05cm3)
22.20
46.30
Titre (+/- 0.05cm3)
22.20
24.10
45.0
Initial reading (+/- 0.05cm3)
0.00
18.20
Final reading (+/- 0.05cm3)
18.20
35.40
Titre (+/- 0.05cm3)
18.20
17.20
54.0
Initial reading (+/- 0.05cm3)
0.00
13.30
Final reading (+/- 0.05cm3)
13.30
23.40
Titre (+/- 0.05cm3)
13.30
10.10
60.0
Initial reading (+/- 0.05cm3)
0.00
12.10
Final reading (+/- 0.05cm3)
12.10
25.10
Titre (+/- 0.05cm3)
12.10
13.00
67.0
Initial reading (+/- 0.05cm3)
0.00
13.50
Final reading (+/- 0.05cm3)
13.50
27.20
Titre (+/- 0.05cm3)
13.50
13.70
In terms of qualitative data, the process of making water samples which includes the formation of the brown precipitate, remained consistent throughout the experiment. They seemed to be the same colour with similar texture in each trial as I made sure to add the exact amount of solutions. As the temperature increased, it was fairly obvious that the light straw colour appeared at a quicker rate, showing the affect of increasing temperature. However, it also made the experiment a lot more difficult in terms of keeping the same temperature. As I repeated the experiment several times, it became easier and much faster to do each titration near the end.
Processed Data:
Table 4 – The process of calculating the dissolved oxygen (ppm) and absolute uncertainties for room temperature (23 °C)
Core calculations
Propagation of uncertainties
1. Average titre:
22.20 cm3 + 24.10 cm3 / 2 = 23.15 cm3
Uncertainty in average titre:
24.10 cm3 – 22.20 cm3 /2 = +/- 0.95 cm3
% = (0.95 cm3 / 23.15 cm3) x 100 = 4% (1sf)
2. Moles of S2O3-2:
n = c x v
where v is the average titre = 23.15 cm3
n(S2O32-) = 0.003 mol.dm-3 x 0.02315 dm3 = 6.945 x 10-5 mol
The original solution of sodium thiosulfate was diluted 1 in 2 using a 100 cm3 volumetric flask with an uncertainty of +/- 1 cm3.
% = 0.1 / 100 x 100 = 0.1%
3. Consideration of mole ratios:
n(O2): n(S2O32-) = 1 : 4
n (O2) = 6.945 x 10-5 / 4 = 1.736 x 10-5 mol
No additional error.
4. concentration of O2:
c = n / v
Where v = 0.050 dm3
n = 1.736 x 10-5 mol
c (O2) = 1.736 x 10-5 mol / 0.050 dm3 = 3.473 x 10-4 mol.dm-3
C (O2) = 3.473 x 10-4 mol.L-1
C(O2) = c x M (O2)
3.473 x 10-4 mol.L-1 x 32 g.mol-1 = 1.11 x 10-2 g.L-1
1.11 x 10-2 g.L-1 x 1000 mg.L-1
C(O2) = 11.10 ppm (mg.L-1)
% uncertainty in the volume measured:
+/- 0.5 cm3/ 50.00 cm3 x 100 = 1.0%
% = 1.0%
5. concentration of O2 with absolute/overall uncertainty:
C(O2) = 11.10 ppm (mg.L-1)
Overall uncertainty:
% added together = 4% + 0.1 % + 1% =5.1%
Therefore the absolute uncertainty in the measured concentration:
5.1 x 11.10 ppm / 100 = 0.5661 ppm
A.U. = +/- 0.6 ppm
Table 5 – a processed data table showing how dissolved oxygen content (ppm) varies with temperature (°C)
Temperature
(°C)
Dissolved oxygen
(ppm)
Absolute uncertainties (+/-)
Uncertainties
(%)
23
11
0.6
5.1
45
8.5
0.3
3.9
54
5.6
0.8
15
60
6.0
0.3
4.7
67
6.5
0.2
2.6
Graph of Trend:
Other than this, I also plotted a graph for the measure of pH of the water sample at 35 °C which turned out to be a straight line graph; pH of 1.5 – 1.6
Comparing to a Literature Value:
Table 6 – The experimental error between the experimental values and literature values of dissolved oxygen (ppm)
Temperature (°C)
Experimental value (ppm)
Literature value (ppm)
Experimental error (%)
23
11
8.6
28
45
8.5
5.9
44
54
5.6
–
–
60
6.0
–
–
67
6.5
–
–
No literature values could be found for 54 °C, 60 °C and 67 °C.
These literature values were taken from – Oxygen, Dissolved.pdf. (n.d.). Retrieved November 1, 2018 from
https://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=23&cad=rja&uact=8&ved=2ahUKEwjZhIa627HeAhUYWysKHaMhCAcQFjAWegQIAhAC&url=https%3A%2F%2Fwww.hach.com%2Fasset-get.download-en.jsa%3Fid%3D7639984251&usg=AOvVaw31s2G77SIZGe8bCKlzY4BX
An example calculation for the percentage error for room temperature = (8.6 – 11 / 8.6) x 100 = 28%
As shown in Table 6 and Graph 2, the literature values and the data that I have collected appear to have a huge difference, the experimental values being higher in comparison to what’s expected to be. Comparing both the slopes of the lines, they have reasonably the same gradient as displayed below:
gradient for literature value = – 0.1227 ppm/°C
gradient for experimental value = – 0.1200 ppm/°C
The experimental values and literature values seem to have the same trend line as they’re almost parallel to each other which suggests that the experimental values are highly precise but not entirely accurate. The fact that all my points are higher than the literature values could indicate that the fresh water that I have collected from the water tank had more dissolved oxygen present, or some of the oxygen that’s flowing around in the air may have managed to dissolved in the water as I collected it.
Analysis & Conclusion:
The gradient of the trend line for the experimental values drawn in Graph 1 being a negative number shows a clear decrease in the level of dissolved oxygen as the temperature rises. From looking at this, the level of dissolved oxygen, 5.4 ppm at 54 °C seems extremely low and is obviously an outlier as it goes far off from the trend line compared to the other plots. It is a little odd that it’s been plotted lower than the point at 67 °C, where it should have been higher than that point as the oxygen content is decreasing as you move towards to the right-hand side of the graph.
In conclusion, temperature does affect the amount of dissolved oxygen present in the water; thus there is a linear relationship between temperature and dissolved oxygen content. As the temperature increases, the amount of oxygen dissolved in the freshwater decreases. Both the experimental values and the literature values show a constant decrease in the level of oxygen in Graph 1, specifically an increase in temperature from 23 °C to 45 °C while the dissolved oxygen content goes down from 11 ppm to 8.5 ppm. As seen from the photographs I have added in the method from the above, it was evident that a reaction was taking place as the colour of the water sample turned dark blue from a pale yellow colour when the starch solution was added, then it fully disappeared at the end of the experiment. The experimental value obtained 28% for room temperature and 44% for 45 °C, higher than expected compared to the theoretical values, meaning that there were a number of systematic errors as well as random errors making the value larger than it was supposed to be. This will be discussed later in the evaluation.
From looking at Graph 2, we can calculate an uncertainty in the gradient and y-intercept as the half-range of the worst fit lines’ gradients and intercepts:
Uncertainty in gradient =
Uncertainty in y-intercept =
The calculation and the graph above shows the negative linear line of best fit:
All of the data I have collected are higher than the literature values, though the two values seem to have the same trend line being almost parallel to each other. Whether I’m getting the correct values or not is such a difficult factor to judge especially during the experiment, only by observations without any reliable data to base it on. However, after comparing my values to the literature values, it is obvious that my data is highly precise, yet not very accurate.
The conclusion seems valid for this investigation as this is the expected relationship between temperature and the dissolved oxygen content. The amount of dissolved oxygen should decrease if the temperature is high, as oxygen is less soluble in warm water than in the cold. This could mean that keeping a cold temperature would increase the level of oxygen which creates a safer and healthier environment for the animals underwater. In order to save as many animals as possible, we must first stop dumping the waste into the ocean.
Evaluation:
The difference between the theoretical (literature) and experimental values show that 44% was a systematic error including 5% as a random error due to the precision of the instruments. All random errors are always present in measurement, but they can be improved by repeating the experiment. However, systematic error is a consistent error associated with faulty of the equipment and methods of observation such as misusing the instruments, etc. Therefore, it cannot be adjusted unless identifying the source of the error.
One of the largest systematic error that caused the result to be less accurate was detecting the end point of the reaction. Since this isn’t measurable, it was relatively demanding to judge whether the blue colour of the sample solution has disappeared and is entirely clear or not. As I mentioned before, the plot at 54 °C from Graph 1 is clearly an outlier as it is extremely lower than it should have been. The only way to get a valid result as possible is to slowly add the sodium thiosulfate into the conical flask dropwise so that I know the end of the experiment as soon as the colour disappears.
Another systematic error that could have hugely affected my results is the difficulty with the temperature control. After heating the water sample in water baths, the temperature decreased as I carried out the experiment. Though I made sure that it reaches just above the target temperature, I couldn’t keep that temperature for very long which have affected my data. The amount of dissolved oxygen present in the water sample at 67 °C is clearly not on the trend line. The point is higher than it is supposed to be which could suggest that the temperature was not exactly at 67 °C and it would have been lower than what it was meant to be. In order to reduce this kind of error, I will have to heat the samples higher than the target temperature as well as checking it before I titrate samples and even afterwards. This will allow me to see how much the temperature has dropped during titration so that for the next experiment, I could possibly estimate the temperature that I have to reach.
I would say that the sources of water sample could count as a systematic error, which may have hugely affected my result being higher than the literature values. Even though I made sure to fill the bottle slowly by trying not to let oxygen that’s flowing in the air into the container. The fact that my values are relatively higher than the theoretical values indicate that perhaps I have repeatedly been collecting the wrong water way without even realising which made the experiment less accurate. To avoid this from happening, I will have to be very careful not to let oxygen in the atmosphere dissolve into the water.
The random error that could count towards is the accuracy of a graduated pipette that may not necessarily have affected my results; however, it had to be considered as I carried out the experiment. Every time I took these measurements, I made sure to do it the right and the same way, for example, when using a graduated pipette, it needed to be ensured that no air gap’s taking up some volume of it. In my opinion, using a graduated cylinder would have been much quicker and easier to measure the exact amount of solutions.
2018-11-6-1541508559