Investigating the temperature dependence of the fluorescence emission of quinine in tonic water using a homemade combined UV photometer and fluorescence spectrometer

Maximilian C Harvey

Summer 2020


1 Abstract 3
2 Introduction 3
3 Preliminary theory 3
3.1 Introduction to fluorescence . . . . . . . . . . . . . . . . . . . . . 3
3.2 Examples of fluorescence . . . . . . . . . . . . . . . . . . . . . . . 4
3.3 The design of the spectrometer . . . . . . . . . . . . . . . . . . . 5
3.4 Electron excitation . . . . . . . . . . . . . . . . . . . . . . . . . . 7
3.5 Decay Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.6 Jablonski Diagrams . . . . . . . . . . . . . . . . . . . . . . . . . . 8
3.7 Stokes Shift . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
3.8 The molecule: Quinine . . . . . . . . . . . . . . . . . . . . . . . . 10
4 Experiment 1 – Determining the temperature dependence of
the fluorescent emission 11
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
4.2 Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
4.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13
4.4 Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14
4.5 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
4.6 Evaluation and estimation of uncertainties . . . . . . . . . . . . . 16
5 Experiment 2 – Assessing the variation in absorbance intensity
with changing temperature 17
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
5.2 Method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
5.3 Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
5.4 Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
6 Conclusions 21
7 Evaluation and adaptations 22
  • Abstract

The temperature dependence of the fluorescence emission of quinine in tonic water was demonstrated using a homemade combined UV photometer and fluorescence spectrometer composed of LEDs and LEGO. The PMCC of the data was used in a hypothesis test to show the significance of the results. Further absorption data was recorded to suggest the most probable mechanism causing the dependence. This was identified as either dynamic quenching or a shift in the emission/absorption spectra. Further investigation over a greater range of wavelengths is required to clarify the proposed hypothesis.

  • Introduction

Photochemistry is the specific area within Chemistry focused on the chemical e ects that light has upon a sample in question. Within the con nes of photo-chemistry, fluorescence is a phenomenon that occurs when a substance absorbs electromagnetic radiation and consequently emits electromagnetic radiation of a lower energy, or longer wavelength than that which was incident upon it.[1] This process can be described using the theory that is contained within the topic of photochemistry. The intensity of this emission can be recorded and it can be found that this intensity is a ected by a number of factors such as the concentration of the fluorescent compound within the solution.[2]

The aim of this paper is to investigate how the variation of temperature will affect the intensity of fluorescence emission of a sample of tonic water (containing quinine), and to determine what mechanism is responsible for the dependency if any is observed.

The experiments involved in this investigation will be carried out using a homemade combined UV photometer and fluorescence spectrometer that is con-structed using lego and LEDs.[3] The reasons for the usage of such apparatus is as a consequence of the Covid-19 pandemic which has rendered access to a professional grade fluorometer and UV spectrometer impossible. Despite the required use of more rudimentary apparatus, due consideration has been made into making sure that the results obtained had as little uncertainty as possible and these considerations made are detailed at length in the paper.

  • Preliminary theory

In order to understand the subsequent phases of this investigation, it is necessary to understand some of the theory behind the phenomenon of fluorescence so that the results presented at later phases of the paper will make sense.

3.1 Introduction to fluorescence

The points requiring attention are the theory behind fluorescence emission and the consideration made when building the apparatus as a result of this theory

To understand fluorescence, one must rst look at what is happening to a sample when electromagnetic radiation is incident upon it. When this process occurs, there are a number of different outcomes that are possible. Firstly, the incident radiation can be re ected – such re ection is the reason why objects have speci c colour. Secondly, the light could be transmitted through the sample

  • An example could be water where nearly all of the light entering the sample comes out the other side. And finally, the incident radiation could be absorbed by the sample.

The process of fluorescence is primarily concerned with the second and third of the three scenarios. Looking at the last of the three, when the sample absorbs the incident radiation the question must be asked as to what happens after the absorption occurs.

The incident radiation can be thought of as a source of energy. The energy of this source is dependent upon the frequency of the radiation incident upon the sample. Using the particle theory of light and hence considering the incident radiation in terms of individual photons, the energy of an individual photon can be calculated using 1 and 2:

where f is the frequency of the light and h is the planck constant. The relationship c = f can be used to form 2.

As a result of this it can be said that for every photon of a frequency, f, that is incident upon a sample, the sample will gain an energy equal to the value of E. [4]

However this subsequently poses the question as to how this increase in energy will a ect the sample when high numbers of photons are incident upon it.

3.2 Examples of fluorescence

Quinine is a molecule that has proved e ective as a drug for use in antimalarial treatments [5] and is responsible for the sharp taste of tonic water. It is also a molecule that will absorb a portion of any UV light that is incident upon it and so is the perfect example with which to observe the effect that occurs when a sample absorbs incident electromagnetic radiation. Looking at the image below of quinine when exposed to UV radiation in comparison to water the result can easily be seen.

(a) Before UV irradiation                              (b) During UV irradiation

Figure 1: Comparison of Water (left) and Tonic water (right)

The result as seen in 1 is that blue/green visible light is emitted by the sample of tonic water. We can determine whether these results are in fact an example of fluorescence by looking at the wavelength of both the incident and emitted electromagnetic radiation:

Wavelength of UV light = 100-400nm [7]

Wavelength of blue-green light = 455-577nm [7]

As seen in the data presented above, the emitted radiation is at a greater wavelength than the incident upon the sample and hence the process is indeed an example of fluorescence. This shift in the wavelength of the emission is known as the ‘Stoke’s shift’, which will be covered subsequently in more detail.

The di erences could also be thought of in terms of energy and photons. Using 1 and 2, it can be shown that the energy of a singular emitted photon is less than that of a single incident photon.[8]

The most important aspect of the demonstration above however is to high-light which measurements must be made in order to conduct an investigation revolving around fluorescence. To properly understand what is going on, it is necessary to measure both the absorbance of the sample, that is to say, how much radiation the sample is absorbing, and also the uorescent emission of the sample.

3.3 The design of the spectrometer

The homemade apparatus is designed so that both of these quantities can be measured.[3]

LEDs are the crucial element to the design of the spectrometer. Their prop-erties enable them to act both as sources of electromagnetic radiation but also as detectors of it too. This means that they form both the incident source and the detectors for the whole build.

The source of radiation is a UV LED of wavelength 370nm and consequently the absorbance detector is a UV LED of the same wavelength as the absorbance measurement is a measurement of how much light is transmitted through the sample without being altered. The fluorescence detectors, of which there are

two in the design are at a greater wavelength than that of the source and were chosen to be green LEDs of wavelength 568nm.

The circuitry of the spectrometer is simple and is shown in 2. A 33 ohm resistor is fitted to the source LED to ensure that the LED itself is not damaged when the power supply is turned on. The resistor and other circuitry was attached using a soldering iron.

The four LEDs being used surround the sample which is denoted by the blue box in 2. The uorescent detectors are at right angles to the incident source of UV radiation so that they only measure the uorescent emission of the sample contained within the cuvette.

Figure 2: The circuity of the spectrometer

The nal outcome of the spectrometer is shown in 3

Figure 3: The spectrometer after construction

The additional LEGO sca olding around the LEDs and the contained cu-vette is to enable easy connection of the crocodile clips which have been xed in place for ease of connection to the multimeter.

Having now understood what measurements are required in order to con-duct the experiments desired, attention can be turned to the mechanism of fluorescence and the reasons for quinine possessing this characteristic.

In order to do this, one must consider what is actually happening when the sample gains energy. The law of conservation of energy states that energy can neither be created nor destroyed but only transformed between stores.[9] Because of this, the mechanism of fluorescence must be able to explain where the energy goes once it has been absorbed and also why there is a difference in energy between the incident photons and the emitted photons.

3.4 Electron excitation

The current model of the atom uses orbitals to show regions in space where an electron is most likely to be found. In each orbital can be a maximum of two electrons that are spin paired (meaning that they have the opposite spin). There are however in any given atom, a number of atomic orbitals that all di er in energy. The electrons obey a rule known as the Aufbau principle and hence ll the orbital that has the most negative energy. The reason the energy is described as being negative is because when describing energy that is not the absolute energy of something, the value of ‘0’ energy is arbitrary and can be chosen. When discussing the energy of atomic orbitals, one is only interested in the di erences in energy between two orbitals and not the absolute energy of the orbitals and hence the energies can be negative. For the description of atomic orbital energies, the value of 0 energy is chosen to be the energy when an electron is at a distance of in nity from the nucleus and hence whenever the electron is in any atomic orbital, this will always be lower in energy than when it is at in nity which consequently explain the sign of these values.[10]

Considering the single electron of a hydrogen atom which is located in the 1s orbital, the question can be asked as to what will happen when the electron absorbs a UV photon that is incident upon it? The electron now has an energy that is greater than the orbital that it is in by a value equal to the energy of the UV photon. As a result, the electron is promoted to a higher orbital that is less negative in energy, so represents an increase in energy. (it is important to note that this promotion of the electron in hydrogen by UV light may not be an allowed transition but here it shall be used as an example due to its simplicity). As a result of this promotion, we say that the hydrogen atom has been excited and is consequently in an excited state. This process can be represented by 3:

H ! H (3)

The * represents the excited state of the hydrogen atom. This excited state however is one that is not of great stability for the species and consequently it is found that the excited state must decay by some mechanism.[11] There are a

great number of mechanisms by which an excited state can decay and in most cases, for a sample, it will be a mixture of a number of mechanisms that occur. The reason for the decay is in order to return the species to the lower (or more negative in energy) ground state which is more stable.

3.5 Decay Mechanisms

Figure 4: Decay mechanisms of excited states [12]

This image above demonstrates a number of the ways in which an excited state might decay. As seen in the diagram, the mechanism of luminescence (that contains the process of fluorescence) is among them. These other mechanisms will be of greater importance later when suggesting a mechanism for any tem-perature dependence of uorescent emission found.

3.6 Jablonski Diagrams

So that we might understand how the mechanism of fluorescence occurs, it is useful to make use of a speci c type of diagram known as a Jablonski diagram. These Jablonski diagrams are the best way to visualize the process of uores-cence.

Figure 5: Example of a jablonski diagram [13]

The Jablonski diagram visually depicts a representation of why the incident radiation is at a higher frequency than the emitted radiation as a result of the process of fluorescence. As can be seen in the diagram, the di erence in energy between the emitted radiation and the incident radiation is due to a process known as vibrational relaxation. This vibrational relaxation occurs nearly in-stantaneously after the excitation occurs.[14] The process involves some of the energy of the excited electron being converted into thermal energy which causes the surrounding solution to increase in temperature slightly.[15] As a result, the electron falls down to a lower vibrational level. The process of fluorescence itself occurs from the lowest vibrational level of the rst excited state to the ground state. However, it is important to note that the fluorescence does not explicitly have to go to the lowest vibrational level of the ground state which means that a range of wavelengths of light can be emitted by the sample.[16] It is also important to note how fluorescence is not the only decay process that is depicted on this diagram. The process that is known as phosphorescence is also depicted. The main observable di erence between fluorescence and phospho-rescence is that the emission of light in fluorescence is almost instant, meaning that light will only be emitted whilst the incident source is on. This is di erent to the process of phosphorescence in which the incident energy can be stored for release at a later time. The reason for this di erence is to do with the spin of the excited electron in comparison to its ground state partner. On order for phosphorescence to occur, the excited electron must undergo internal conver-sion, which swaps its spin, meaning that its spin is in the same direction as its ground state partner.

3.7 Stokes Shift

The Stoke’s shift is the name given to the shifting of the spectra. It describes the di erences in wavelength between the incident photons and the emitted

photons. In 6, it can be seen how the emission spectra has been shifted towards a greater wavelength in comparison to the absorbance spectra. The stroke’s shift is a key characteristic of fluorescent emission.

It is as a result of the vibrational relaxation occurring that the Stoke’s shift is present, meaning that we see a reduction in the frequency between the incident and the emitted electromagnetic radiation.

Figure 6: An example of the Stokes Shift [17]

3.8 The molecule: Quinine

The nal bit of preliminary theory required is to look in a bit more depth at the molecule quinine, and the properties of this molecule that give it its uorescent properties.

Looking at the structure of quinine in 7 there are a number of interesting points to note. The most important of all of these is that within the molecule, we can see that there is a large region towards the middle that has a lot of alternating double and single bonds. We might describe the molecule in this case as having a large ‘pi conjugated system’.

Figure 7: Structure of quinine [18]

In fact it is actually the pi conjugation that is responsible for giving quinine its fluorescent properties and by looking at other fluorescent molecules such as fluorescein and Rhodamine B in 8 we can see that they too also have large conjugated pi systems.

(a) Rhodamine B [19] (b) Fluorescein [20]

Figure 8: Some other uorescent compounds with large conjugated pi systems

The molecule of quinine itself has a number of interesting properties. It has 4 chiral centres leading to a total of 16 stereoisomers. Whilst quinine is used as an antimalarial drug, one of its stereoisomers, quinidine, shown in … is occasionally used as a class 1A antiarrhythmic drug. [21]

Figure 9: The optical isomer of quinine, quinidine. [22]

Both quinine and quinidine are originally from the bark of the cinchona tree which is most commonly found in Southern and Central america. [23]

With detailed understanding of the mechanism of fluorescence and the prac-ticalities behind detecting the phenomenon, the experimental phase of the project can be commenced.

  • Experiment 1 – Determining the temperature dependence of the fluorescent emission

Practical to determine whether there is a correlation between the temperature of a sample of quinine and it’s uorescent emission – conducted on the homemade combined fluorometer and UV photometer. The significance of any results will be veri ed using a hypothesis of the data’s PMCC.

4.1 Introduction

The rst experiment detailed below was designed in order to determine whether there was a relationship between the temperature of the sample within the spec-trometer, and the intensity of the fluorescent emission. Whilst the literature already demonstrates that there is a negative correlation between the temper-ature of the sample and the intensity of the emission [24], this means that the rst experiment can also be used to determine the capabilities of the homemade apparatus and assess whether it would be t for use in future investigations.

4.2 Method

In order to determine whether there was a temperature dependency in the fluorescent emission of the chosen sample, it was necessary to test a number of samples of the same concentration of tonic water (that contained the uores-cent compound quinine sulfate) at a variety of temperatures and measure the intensity of the fluorescent emission using the homemade combined UV pho-tometer and fluorescence spectrometer.

Prior to the execution of the investigation, it was decided that the samples would be tested at 3 temperatures, 0oC, 25oC and 50oC. Although naturally, the experiment would be more accurate and reliable had a greater number of di er-ent temperatures been tested, it was decided that due to the high uncertainty in a number of the readings that were being taken as part of the experiment, the apparatus would be unable to di erentiate between temperature levels that were too close due to its intrinsic low resolution. Further details of this decision are explained in greater depth in the sources of error section for this experiment.

One notable consideration that is of particular importance revolves around measuring the temperature of the sample and how the temperature of that sample is altered once it is placed into the spectrometer. Due to the small amount of sample that was being placed into the cuvette, and the relatively large size of the thermometer being used, it was decided that a larger sample should be heated and the temperature of that sample measured before a small sample was removed to be used for the investigation. Not utilising this process would have potentially caused the temperature of the sample to be altered by a considerable amount when the thermometer was inserted to take the reading as the thermometer would have been at a di erent temperature to the sample and so a change would have been induced by the transfer of internal energy between the two. It was ascertained that the uncertainty would be comparatively less by measuring the temperature of the larger sample and then removing the test sample.

Furthermore, it was found that due to its small volume the temperature of the sample increased substantially once it had been situated in the spectrometer that was on for a period of time. This in retrospect was due to the fact that the LEDs, despite their benefit over conventional filament lamps, of being more efficient, were still dissipating a large amount of internal energy and so were heating up the sample. Further heating also occurred as a result of the internal

energy released by the decay process of vibrational relaxation. In order to ensure that this didn’t a ect the results of the experiment to any great extent, once the samples were placed into the apparatus, the device was only turned on momentarily. The samples exposure to the incident radiation was limited to being just long enough in order to record the potential difference across the spectrometer that was measuring the fluorescence emission of the sample before being quickly turned off to prevent the sample from heating up.

As a consequence to these considerations, the process formed was as follows:

  1. Firsty the stock sample of tonic water was heated to the correct temper-ature using a water bath shown in 10a.
  2. A small sample was transferred to the cuvette using a pipette.
  3. Once the cuvette was placed securely into the housing of the spectrometer, a box was placed over the spectrometer to act as a shield that would prevent any background radiation being incident upon the experiment and thus a ecting the results obtained. (Depicted in 10b)
  4. With everything set up, the spectrometer was then turned on momentarily in order to take the reading whilst preventing the sample and the rest of the apparatus from heating up to a significant extent.
  5. The process was then repeated with other samples, both of the same temperature and of different temperatures.

(a) The homemade water bath (b) Improvised shielding

Figure 10: Improvised apparatus used during the experimental process

4.3 Results

Having decided the temperature at which to test the samples, and having ac-counted for the sources of error that have been mentioned above, the practical was carried out and the results are detailed below.

Figure 11: Results from experiment 1

The power supply was set to 4.5V giving the source UV LED a terminal p.d of 3.6V. This was kept constant throughout the whole investigation so is not included within the calculations detailed below. These results can be represented graphically in the following way using graphing software [25]

Figure 12: Graph of results from experiment 1 [25]

4.4 Analysis

Using the results plotted on the graph, it is easy to suggest that there might be a negative relationship between the temperature of the sample and the intensity of the fluorescent emission.

In order to test the hypothesis that there is a relationship between these two variables, a hypothesis test was carried out using the Pearson Product Moment Correlation Coefficient of the data. The PMCC is a value that once calculated, gives a value ranging between -1 and 1 that says by how much two variables are correlated, with 1 being a perfect positive linear correlation and -1 being a perfect negative linear correlation. The value of 0 represents no correlation. The PMCC was used as the data fits a bivariate normal distribution. The calculation of this test statistic is given by 4:

r = p Sxy (4)

Fortunately this calculation need not be carried out by hand and a spread-sheet was used to calculate the value of the PMCC. The spreadsheet used for the calculation is inserted below, along with the received value of r for the sample.

Figure 13: Calculation of PMCC for experiment 1

In order to calculate the value of r this spreadsheet uses 4, but also 5, 6 and


Sxx = X(x E(X))2 (5)
Syy = X(y E(Y ))2 (6)
Sxy = X(x E(X))(y E(Y )) (7)

The value of r as calculated from the spreadsheet was -0.981 (3s.f.) which is in line with the literature of there being a negative correlation between the two variables.[26] However, it is necessary to determine whether the r value calculated from the data recorded is a signi cant result and not one that could have been obtained by random error. The process and results of the hypothesis test carried out are shown in:

H0 : = 0 where is P M CC of the population

H1: >0

r = PMCC of the sample from the population

= 0:005

r = 0:981

C:V f or n = 9 : 0:7498

jrj >

Reject H0


As can be seen from the test above, the value for r was well within the critical region in order for it to be considered extreme. The use of the signi cance level of 0.005 means that the value of r calculated from the experiment is in the most extreme 0.5% of the data, and so it can be said with con dence that the results of the experiment were signi cant. A one tailed hypothesis test was chosen as the negative correlation was already known from other sources. As a result the hypothesis test was a veri cation of this already established result. The n value of 9 refers to how many data points were involved in the calculation of the PMCC that is being tested.

4.5 Conclusions

Through the analysis of the data from this experiment, it can be concluded that there is a negative correlation between the temperature of the sample of tonic water and the intensity of the uorescent emission. This conclusion can be drawn as a result of the calculated value for the PMCC of the data, which was shown to be signi cant throught the hypothesis test, conducted at the 0.5 % level.

In addition to concluding that the uorescent emission of quinine in tonic water is temperature dependent, the results of this experiment also demonstrate the capabilities of the homemade combined UV photometer and fluorescence spectrometer.

4.6 Evaluation and estimation of uncertainties

Throughout this experiment there were a number of sources of error due to the nature of the circumstances leading to the type of apparatus upon which the practical had to be conducted. The uncertainties due to the measurements themselves were comparatively small: namely the uncertainty of 0.5oK when using the thermometer, shown in 9, which led to an absolute uncertainty of <1%, and the uncertainty due to the readings of the voltmeter, shown in 10, of

0.0005V leading to an absolute uncertainty of 1.5%.

uncertainty = uncertainty in the measurement 100 (8)
smallest measurement made
absolute uncertainty = 0:5 100 = 0:18% (9)
absolute uncertainty = 0:0005 100 = 1:5% (10)

The greatest uncertainty in this experiment however arises from the trans-fer of internal energy between the sample and the surroundings, and vice versa. This process has the potential to drastically a ect the values recorded by a much greater extent than the previously addressed factors as the temperature could vary signi cantly in the time it takes for the readings to be taken. Unfortunately there is no easy way to estimate the uncertainty as a result of this process due


to a number of reasons. Firstly, the rate of heat transfer is dependent upon the di erences in temperature between the sample and the surroundings; a greater di erence results in a greater rate of transfer between the two. Secondly, it is di cult to estimate the conductivity between the sample and the apparatus as the apparatus has many components with di erent compositions. To mit-igate this issue as much as possible, a number of considerations were made. The temperatures chosen were equally spaced around the room temperature of approximately 25oC. This meant that the rate of transfer of internal energy would have been comparable between the temperatures of 0oC and 50oC. In addition to this, the readings were taken as quickly as possible from the point of the sample being removed from the water bath in an attempt to ensure that the minimum amount of internal energy was transferred before the reading was taken.

This procedure was also necessary to ensure minimum heating from the internal energy being transferred to the sample both from the irradiating UV LED but also through the vibrational relaxation of the sample itself.

The nal consideration made throughout the investigation was concerning background levels of radiation. When the multimeter was connected across the detectors, but the source UV lED was o , background levels of electromagnetic radiation were detected. Rather than subtracting this background level from the results, a shield was placed over the apparatus whilst the measurements were made in order to remove the background radiation.

  • Experiment 2 – Assessing the variation in ab-sorbance intensity with changing temperature

An investigation to determine whether the absorbance of the tonic water samples is dependent upon their temperature in order to determine the causal mechanism of the uorescent emission temperature dependence.

5.1 Introduction

Having shown in the previous experiment that the uorescent emission of the tonic water is temperature dependent, it is useful as a next stage to try to determine the cause of this dependence.

Whilst this cannot be done with absolute certainty due to the circumstances of the investigation and the apparatus being used, a narrowing of the possible mechanisms can be achieved.

There are two main scenarios that could explain the temperature dependence of the uorescent emission and these are as follows:

  1. Fewer excited states are created meaning that the overall uorescent in-tensity is less
  2. The excited states that are created, decay by a di erent mechanism, and so do not emit visible light.


The second of these two scenarios would mean that the number of excited states decaying via the mechanism of fluorescence would decrease and hence the uorescent intensity recorded would decrease. For this to be the case, an increase in temperature would have to favour a non-radiative decay process, one which does not result in the emission of visible light, but that still dissipates the excess energy of the excited state, allowing the excited molecule to return to its more stable ground state.

In order to distinguish what the cause of the dependence is, it is useful to ex-amine how the absorption of the UV incident radiation varies with temperature. This can be done by using the lego apparatus as a UV photometer, making use of the two UV LEDs. This will give a reading as to how much UV radiation is transmitted through the sample contained within the cuvette.

The data recorded across the range of temperatures can then be tested in the same way using a hypothesis test of the PMCC. This will lead to one of two conclusions:

  1. There is no change in the absorption intensity as temperature varies – meaning that the same number of excited states are being produced, but are decaying via alternative pathways.
  2. There is a change in the absorption intensity as temperature varies (most likely a positive correlation) – inferring that there is a change in the number of excited states being produced.

These two outcome each infer a di erent set of possible causes of the temper-ature dependence. However it is also possible that the temperature dependence is caused by a combination of both fewer excited states being produced and also a proportion of the excited states that are produced, decaying by an alternate mechanism.

5.2 Method

The method is identical to the method presented for experiment 1, save for a few exceptions. Firstly, a greater number of temperatures were chosen to be tested. This decision was made as a result of the success of the apparatus’ performance in the rst experiment. Secondly, instead of measuring the terminal potential di erence across the green LED, the terminal potential di erence was measured across the UV LED that was not connected to the power supply.

The same considerations were made with regards to ensuring minimal heat transfer between the sample and its surroundings and consequently the samples were only irradiated for the minimum amount of time necessary to take the readings.


5.3 Results

Figure 14: Table of results from the absorption experiment

In 14 can be seen the results from the second experiment. Each temperature was as per the previous experiment tested 3 times.

These results that were obtained were once again plotted using the GW analysis application [25] which shows an apparent positive linear correlation between the temperature and the intensity of the terminal potential di erence across the absorbance detector.

Figure 15: Graph of the results from absorption experiment[25]

5.4 Analysis

Despite there appearing to be a positive linear correlation between the two variables, a hypothesis test was conducted once again to determine whether the relationship was signi cant.

In order to complete this hypothesis test however, the PMCC of the data was rst calculated using another excel spreadsheet


Figure 16: Calculation of PMCC for experiment 2

The value of the PMCC calculated using (4,5,6,7) in the spreadsheet is 0.928.

This is the value which has been used in the hypothesis test below:

H0 : = 0 where is P M CC of the popluation

H1: >0

r = P M CC of the sample f rom the population

= 0:01

r = 0:928

C:V f or n = 15 : 0:6411

jrj >

Reject H0

As can be seen above, the result of the hypothesis test is that the relationship between the absorbance and the temperature is a signi cant one and thus it can be concluded that there is a positive linear relationship between the two. For this hypothesis test, a two tailed test was necessary as the relationship expected was not known and so a two tailed test was conducted at the 1% signi cance level.

This means that the number of excited states being produced decreases as the temperature increases and this in part explains the result of the rst of experiment where the negative relation between the temperature and the uorescent emission intensity was shown.

However, this fact may not alone account for the whole of the explanation as to why there is a temperature dependence in the uorescent emission. Part of the dependence could also be due to the excited states that are produced decaying by an alternate pathway.

One possible explanation is that the number of excited states does not change, but it is just that the apparatus is ill-equipped to properly measure this


scenario. This would be the case if there was a shift in the absorbance spectra as a result of the changing temperature. If the absorbance spectra shifted, then this could mean that the same number of quinine molecules are being excited, but just by a di erent wavelength to the one that the spectrometer can detect. Consequently it would appear that there is a change in the number of excited states, despite this not being the case.

This would mean that the reason for the temperature dependence would be as a result of the decay of the excited states and not due to a decrease in the number of excited states being produced.

Working on this principle, there should be a non-radiative decay mechanism that is more favourable when there is a greater temperature.

It turns out that this mechanism is most probably the process of dynamic quenching. This is a process by which due to the collision between an excited species, with another species, the excited species decays back to the ground state without the emission of electromagnetic radiation. The excess energy is often converted to internal energy. It is found that an increase in temperature increases the amount of dynamic quenching that occurs as the higher tempera-ture means that more collisions occur leading to a greater number of instances where the excited states decay by this mechanism. [27]

One further possibility could be that just as the absorbance spectra shift, so too does the fluorescence spectra[28] and so the spectrometer cannot detect the emitted photons as they are of the incorrect wavelength. This would mean that despite there still being a temperature dependence of the emission, it is not concerning the intensity of the emission but the wavelengths of the emission and hence the position of the corresponding spectra.

Whilst all of these scenarios give a valid explanation as to the occurrence of the temperature dependence of the uorescent emission, it is impossible to say with any certainty which one is correct due to the high uncertainty that is present due to the use of the homemade spectrometer.

  • Conclusions

The temperature dependence of the uorescent emission of quinine found in tonic water was successfully shown using a homemade combined UV photome-ter and fluorescence spectrometer build using LEGO and LEDs. The further experiments conducted suggested possible reasons and mechanisms to explain the dependence but without conclusion due to the high uncertainty and lack of precision involved. Whilst certain mechanisms such as dynamic quenching were suggested, the spectrometer must be upgraded to measure both uorescent and absorption spectra over a greater range of wavelengths in order to give a more comprehensive answer. This would enable shifts of the absorbance and fluorescence emission spectra to be observed as well as variations in their in-tensities leading to a more thorough picture of the phenomenon’s temperature dependence as a whole.


  • Evaluation and adaptations

Having completed the project, there are a number of analytical points that should be made that would improve the experiment for the next time. These concern not only the reduction of uncertainties but also the improvement of conclusions

First and foremost, there are a number of adaptations that could be made to the spectrometer in order to achieve both of these things. A large issue with the spectrometer is that there is only one wavelength of LED to measure the ab-sorbance spectra and one wavelength of LED to measure the uorescent emission spectra. As a result, it was impossible to determine whether the results obtained were due to a temperature dependence that actually reduced the uorescent or transmitted intensities, or whether the changes simply caused the emissions to shift outside of the detectable range. The implementation of either a modular design where the LEDs could be swapped out for ones of varying wavelengths or a design that incorporated multiple detectors at once would enable this to be observed and measured. This would allow for a stronger conclusion to be drawn as to the determining of the mechanism responsible for the temperature dependence.

An alternative would be to use a webcam instead of LEDs and use a pro-gram that would generate the emission spectra of the sample without having to implement multiple modules.

In addition to this, a larger detector would also be of use. This would allow for a larger sample to be irradiated within the cuvette. Whilst this would reduce the rate of transfer of energy between the sample and the surroundings, it would also potentially allow for a thermometer to be inserted into the sample, allowing for continuous monitoring of the temperature and also the possibility of using a data logger to record the continuous change of the emission over the full range of temperature.

Whilst the use of the homemade combined UV photometer and fluorescence spectrometer can be deemed a success, there is a great load more that could be achieved with some careful alterations to the design.



  1.\_Chemistry/ Supplemental\_Modules\_(Analytical\_Chemistry)/Analytical\ _Sciences\_Digital\_Library/Active\_Learning/In\_Class\ _Activities/Molecular\_and\_Atomic\_Spectroscopy/03\ _Text\%3A\_Molecular\_and\_Atomic\_Spectroscopy/3\%3A\ _Molecular\_Luminescence/3.6\%3A\_Variables\_that\_Influence\ _fluorescence\_Measurements


  1. Eivind V. Kvittingen, Lise Kvittingen Orcid, Thor Bernt Mel , Birte Jo-hanne Sjursnes, and Richard Verley J. Chem. Educ. 2017, 94, 10, 1486{1491 August 22, 2017
  3. Achan, J., Talisuna, A.O., Erhart, A. et al. Quinine, an old anti-malarial drug in a modern world: role in the treatment of malaria. Malar J 10, 144 (2011).
  4. Personal photograph






  1. James Keeler, Peter Wothers Why Chemical Reactions Happen Oxford Uni-versity Press 2009 Chapter 4
  3. R.P.Wayne Photochemisty Butterworths 1970

  1. Michael P. Grubb, Philip M. Coulter, Hugo J. B. Marroux, Andrew J. Orr-Ewing,and Michael N. R. Ashfold Unravelling the mechanisms of vibrational relaxation in solution Chem Sci. 2017 Apr 1; 8(4): 3062{3069. Published online 2017 Feb 10
  2. Theoretical_Chemistry_Textbook_Maps/Book\%3A_Time_Dependent_ Quantum_Mechanics_and_Spectroscopy_(Tokmakoff)/16\%3A_Quantum_ Relaxation_Processes/16.01\%3A_Vibrational_Relaxation


  1. thermo

  1. structure.svg




  1. quinine_and_quinidine_/703421




  1. android.graphicalanalysis&hl=en\_GB



  1. Georgios Arnaoutakis, Dirk Nather QUENCHING OF fluorescence WITH TEMPERATURE N P27 v.1; 5 Feb. 16
  2. Edward L. Nichols The E ects of Temperature on Phosphorescence and fluorescence Proceedings of the American Philosophical Society Vol. 49, No. 196 (Aug. – Sep., 1910), pp. 267-280

About the Author

Max Harvey is a passionate scientist hoping to study Natural Sciences at university in the coming years. His passions include photochemistry and particle physics. He has interests also in conducting scientific experiments using the everyday household objects around him.

Leave a Reply

Your email address will not be published. Required fields are marked *