How will the incident sunlight intensity and the frequency

spectrum, affect the efficiency of silicon and perovskite solar

cells under the exposure of such light?

Word Count: ​3991

IB student code:
 1

Table of Contents

Content Page

Table of contents 1

Introduction 2

Investigation 17

Materials 19

Variables: 20

- Independent Variable

- Dependent Variable

- Control of Variables

Experimental Design 21

Experimental Procedures 24

- Method

Data 26

- Frequency spectrum experiment, IPCE

- Incident sunlight intensity experiment, IV

Analysis 31

Conclusion 32

Evaluation 33

Acknowledgements 34

Works Cited 35

Appendix 38
 2

Introduction

According to the report “World Energy Outlook 2018” from the International Energy

Agency, the tremendous worldwide energy demand along 2017 was around 14.000 Mtoe

(Million tons of oil equivalent) which equals 160.000 TWh approximately (​“World Energy

Outlook”).

In such report an assumption was made, namely the Sustainable Development Scenario.

This scenario “outlines an integrated approach to achieving internationally agreed objectives on

climate change, air quality and universal access to modern energy​” (​“World Energy Outlook”).

Figure 1: Total Primary Energy Demand (TPED), World.

(​“World Energy Outlook”)

From those 160.000 TWh constituting the energy demand, roughly 25.000TWh where

generated in the form of electricity.

 3

Figure 2: Electricity Generation by Technology, World.

(​“World Energy Outlook”)

According to Michael Grätzel, one of the most relevant modern scientists in new solar

photovoltaic devices, “The Sun provides approximately 100.000 terawatts to the Earth which is

about 10.000 times more than the present rate of the world’s present consumption” (Grätzel).

Updating those figures to the present demand; the Sun generates 100.000TWh of energy in 1

hour, and the world yearly demand of energy is 160.000TWh. This means that in 1.6 hours the

Earth receives all the energy demanded in a year like 2017. As the year has 8,760 hours, it can be

stated than the Sun is providing 5,500 times the energy demanded. ​Seeing how oil, coal and gas

dominate the majority of the energy mix, but will end up being consumed, is the main reason

why renewable technologies must replace them.

 4

Solar Energy

The Sun’s light is known as solar renewable energy, that when it hits the Earth, some is

reflected away, but the rest can be used. Out of the useful energy penetrating the Earth’s

atmosphere, two main systems convert the Sun’s energy into electricity. One system converts

energy into heat, while the other uses photons coming from the Sun to activate electrons into

movement through certain materials, generating direct current (National Geographic Society).

Photovoltaic Solar Energy

Photovoltaic (PV) solar energy, is the electrical energy produced when sunlight strikes a

solar cell. Such solar cell, by means photovoltaic effect, produces movement in the electrons,

when the photons (light from the Sun) incide in their surface. This effect was discovered by the

physicist Edmond Becquerel in 1839 since he demonstrating how photons excited electrons, and

when connected through a circuit, generated an electric current. This was the origin of

photovoltaic solar energy. Nevertheless, this technology wasn’t able to fully expand since “early

solar cells were too inefficient to be of much use” which made this discovery remain as a simple

curiosity (“What Is the Photovoltaic Effect?”). This technology has been investigated worldwide

due to three main historical crises; the origin of the space age, the first oil crisis, and climate

change. (Estela)

Since PV solar energy is a reliable source of energy that can power devices constantly,

space agencies started investing on the development of solar cells to power their equipments

around the 50’s; searching for the most efficient possible PV cells. This culminated by launching

the solar satellite Vanguard I on 1958, which included photovoltaic cells.

 5

Figure 3: Vanguard I. Vanguard Project - U.S.

Naval Research Laboratory, 17 Mar. 2008,

www.nrl.navy.mil/vanguard50/legacy.php.

Accessed 25 Nov. 2018

Although high efficient cells were used in space, the petrol crisis from the 70’s and the

urging climate change, made PV cells start to emerge. At this point, due to its necessity, solar

energy started to be heavily investigated. Below a graph reflects the exponential growth of the

PV generation, in which the early years can be neglected due to their limited scope.
 6

Figure 4: Solar PV power generation. Historical development and targets.

(​“World Energy Outlook”)

Photovoltaic Solar Cells

Since their early years, the efficiencies of solar cells have been recorded by The National

Renewable Energy Laboratory (NREL). This organization annually plots efficiencies into a

graph, showing a clear improvement through time. This investigation will focus on two different

technologies. One of them, crystalline silicon solar cells (CSSC), have been investigated since

the origin of photovoltaics, dominating the current solar market by “more than 90 percent”

(Martin), which reflects the investigation behind this technology. At the same time, research is

being done to discover technologies with new properties. Through this line of research,

perovskite solar cells (PSC) were discovered recently.

 7

Figure 5: Best Research-Cell Efficiencies. (“Research Cell Efficiency Records.”)

CSSCs, with efficiencies around 22%, are the best quality-price cells of the market, but

they have hit a plateau, but silicon’s efficiency is by combining them with other cells, becoming

part of the category known as multijunction cells. Fulfilling the necessities of CSSCs, recently

discovered PSC seem to be the perfect match. These cells are rapidly increasing their efficiency,

and in the last 5 years, they have increased about 10% of their efficiency, which is a striking

improvement. This fact, added to the interesting extraordinary behaviours of PSC under different

frequencies, have scientists around the world hopeful for future discoveries.
 8

Limitations

Seeing as the sun produces much more energy than we need, why don’t we use it?

Historically, the main limitation has always been linked to the cost of such devices.

Nevertheless, as the Boston Consulting Group developed in their famous theory known as the

“Experience Curve”, the cost per unit is highly reduced as more units are produced (Henderson).

In regards to PV energy, this amount is measured as cumulative production with units of power.

Figure 6: “Experience Curve” in Photovoltaics. (Henderson)

In order to reduce the unit cost, manufacturers have been forced to improve efficiencies,

and produce more units. In order to improve efficiencies of PV cells, there are two main aspects

to be considered; improving the behaviour from the incoming sunlight intensity, and taking

advantage of a greater frequency spectrum of the sunlight.

Numerous studies have established the pathway to reduce the cost of photovoltaic devices

by means of combining technologies like silicon and perovskites, known as “multi-junction

 9

concepts, in which perovskite films are combined with silicon or other materials to expand the

absorption spectral range, and convert the solar photons into electrical potential energy at a

higher voltage” (Snaith).

This means that that the ideal would be to combine the properties of a silicon solar cell

with the fact that perovskite solar cells can be created especially to convert energy in a specific

frequency. Selecting the appropriate range of solar spectrum wavelength absorption for the

perovskite, so that it’s complementary to that of silicon we can improve the efficiency of the cell

by combining both.

Figure 7: Tandem Solar Cell. Perovskite combined with Silicon. Jeangros, Q. “Perovskite

Cells for Tandem Applications | EPFL.” Transparent Conductive Oxides (TCOs) | PV-LAB, 7

Dec. 2017, www.pvlab.epfl.ch/page-124775-en.html. Accessed 25 Nov. 2018.

The challenge is to find the best materials that according to the band gap theory will

absorb the most wavelength spectrum, in order to produce more energy. The band gap theory
 10

defines that cells are separated in two bands; the valence band, where the electrons are linked to

the nucleus, and the conduction band, where they are free to move. No electrons can be in

between both bands; such space is called band gap. In order for electrons to overcome the

separation between the bands, the band gap, a specific energy is needed, which is measured in

electron-volts (eV). This energy is the one given by the Sun through the photons.

Therefore, to improve the characteristics of solar cells and save the limitations, research

is based to find inexpensive materials which complements the silicon with new band gaps that

allows to take advantage of new wavelengths ranges of the solar spectrum.

Frequency

As explained by Einstein in 1905, a new approach to the photoelectric effect was

considered. Basing his thoughts on Becquerel's previous hypothesis, he proposed that “light was

made up of discrete particles or quanta energy” (Wenham et al. 1). This idea can be summarized

in the equation: “where light, of frequency ​f o​ r wavelength ​λ,​ comes in

“packets” or photons of energy E​ph ​; ​h ​is Planck’s constant (​6.62607004 × 10​-34​ Js) and ​c i​ s the

velocity of light ​(3.00 × 10​8 ​)” ​(Wenham et al. 1).

Therefore the incoming sunlight, being composed of photons of light, can be referred as

energy. When such light incides on the solar cell, only those photons with an energy E​ph greater
​

than the band gap energy E​G​ will be able to excite electrons, and generate electric current. Those

photons having less energy than the band gap energy, will pass “through it as if they were

transparent” (Wenham et al. 29).

 11

Hence photons of light with an energy E​ph =

​ hf, will only excite electrons when E​ph >
​ E​G​.

In order for E​ph >

​ E​G​ to occur, the frequency must be equal or higher to the threshold frequency;

which is “the minimum frequency of incident light which can cause photoelectric emission”

(“Photoelectric Effect.”).

From the materials that will be investigated; silicon has a band gap of 1.11 eV, while

perovskites are known for having band gaps ranging from 2 eV to 1.2 eV, depending on their

molecular composition (“Semiconductor Band Gaps.”). Therefore, depending in the molecular

composition of the perovskite cell, cells could be adapted in order to work with certain

frequencies, opening the path to the combination of cells.

Figure 8: Frequency vs Wavelength. Lee, John W. “Bioluminescence, the Nature of the

Light.” ResearchGate, Jan. 2015, www.researchgate.net/figure/The-wavelengths-and-

 12

frequencies-of-electromagnetic-radiation-Human-visual-sensitivity_fig5_281828176.

Accessed 25 Nov. 2018.

Furthermore, out of the energy that actually gets to the solar devices, seeing how it comes

in the form of visible light, not all of that energy can be processed. This is due to the fact that

solar radiation has a wavelength spectrum that can’t be fully used by the solar devices, since not

all wavelengths of the visible light can be converted into energy.

Figure 9: Solar Radiation Spectrum in terms of wavelength. Majka,

Tomasz M. “Healthy Light Source.” ResearchGate, Feb. 2013,

www.researchgate.net/figure/Spectral-distribution-of-solar-radiation-

observed-at-the-earths-surface-4_fig3_255989402. Accessed 23 Mar. 2018.

This figure illustrates is the amount of spectral irradiance given by the Sun, showing the

variation of radiation per wavelength.

 13

Intensity

When a photon of light incides in the solar cell, an electron is released, and a hole is

formed. The released electron searches for another hole to fill, and the energy provided by the

photon will dissipate. This is known as the process of recombination.

Figure 10: Effect of solar irradiance on a silicon

solar cell. The process of recombination. (Estela)

The functioning of the photovoltaic cell is to force the electrons and holes to move along

opposite sides of the material when they recombine, and provide a difference in potential.

Figure 11: Movement of electrons,

e- and holes, h+. (Estela)

 14

As mentioned before, in order to generate current, the energy of the incident photon must

be greater than the energy of the band gap of a cell, E​ph >
​ E​G​. The energy of a photon is defined

as:

, and the energy of the whole incident sunlight, also known as intensity, is

defined as: , where ​n ​is the number of photons. Therefore, the energy is directly

proportional to the number of photons incident over a solar cell, . Therefore, a linear

relationship should be later found, when plotting the irradiance reaching the cell, against the

maximum power point. Furthermore, since the proportionality between ​E a​ nd ​n ​isn’t specific to a

single cell, but rather all cells, silicon and PSC should have identical graphs.

Crystalline Silicon Solar Cells (CSSCs)

Silicon solar cell are composed by two layers, one is called P-type (positive type),

composed by silicon and boron, and the other one is N-type (negative type), composed by silicon

and phosphorus. This is called a PN junction. The P-type includes boron since it’s missing

electrons to complete its outermost valence shell, and it creates more holes; while the N-type

includes phosphorus since it has only one electron on its outermost shell, which is easy to

remove. When put together and given a direction of flow, an electrical field occurs due to the

different density of charges between the materials.

 15

Figure 12: Illustration of a Silicon cell doped with Phosphorus and Boron. (Estela)

Perovskite Solar Cells (PSCs)

Perovskites englobe all materials with the crystalline structure CaTiO​3​.

Figure 13: Structure of a Perovskite Cell. (Ferreiro)

Following the organization above, perovskites have immense amounts of structural

possibilities. The general formula states that perovskites are all of those that form compounds in

the form of ABX​3​ (Ferreiro). Regarding the many possibilities of a perovskite solar cell,

scientists have discovered that depending on the compound that forms a perovskite solar cell,

such cell will be able to process different wavelengths of light (Snaith).

 16

Investigation

Research question and hypothesis

How will the incident sunlight intensity and the frequency spectrum, affect the efficiency

of silicon and perovskite solar cells under the exposure of such light?

To compare the two solar devices investigated, silicon and perovskite solar cells will undergo

two different experiments that will investigate their behaviour. This are incident sunlight

intensity and the frequency spectrum of the incident light.

The purpose of the first experiment is to investigate how will the variation of the

wavelength affect the conversion of photons into electrons, also known as Incident

Photon-to-Current Efficiency (IPCE). In this experiment the two cells will receive an spectrum

of wavelengths between 400 and 800 nm, in intervals of 10 nm. This experiment uses a

monochromator to generate the specific wavelength measured, and a computer to record the

current and power simultaneously. Using current and power, the IPCE will be calculated.

The second question researched consists in investigating how the variation of the

intensity over the solar cell’s surface will affect the electricity generated. This experiment uses a

solar simulator that will simulate the power given off by the Sun (exactly 1 Sun), with the use of

filters. The data will produce readings of current and voltage that the cell has under each filter.

Then, current and voltage graphs, IV graphs will be plotted for each cell and each filter, to find

the maximum power point of each cell under each filter, which is found by multiplying current

and voltage. Power will be plotted against solar irradiance, to investigate if they are proportional.
 17

My hypothesis is that in the experiment where the cells will be exposed to different

wavelengths, PSCs will be more efficient in some of the wavelengths, which will be different to

those of silicon. However, in the 2nd experiment, where we are investigating intensity, the

CSSC will show a higher current compared to the PSC. This is because the new technology is

not as developed yet, and in a situation where they both receive the same amount of energy,

without playing with the wavelengths, CSSC will be more efficient.

 18

Materials

● Sun 2000 Solar Simulator, from Abet Technologies, model 11000 and Monochromator

● Voltmeter and Ammeter

● Computer

● Crystalline Silicon Solar Cell (CSSC) of ​1 cm​2 ​dimension; use duct tape so that it’s ​1 cm​2

Figure 14: CSSC of 1cq dimension.

● Perovskite Solar Cell (PSC) from the ICIQ laboratories, of ​1 cm​2 ​dimension.

Figure 15: PSCs on a container, with the PSC

of ​1 cm​2 ​used in the experiment; top left corner.

● Filters of 0.813, 0.515, 0.252, 0.126, and 0.051 Sun.

Figure 16: Filters for the Solar Simulator.

 19

Variables

Independent Variable: (exp. 1) Wavelength of the incident sunlight intensity. (exp. 2) Voltage

across the cell.

Dependent Variable: (exp.1) The amount of current and power given by the cell. (exp.2) Voltage

(V) and Current (mA).

Control of Variables

● Devices and measuring equipment. In order to compare the CSSC and the PSC, both

might be the same through the duration of the trials. Cells may have impurities, and

especially PSC, due to their structural composition that can widely vary, it is critical to

ensure both cells are the same always (Snaith). Also, the monochromator and the solar

simulator must be the same to reduce uncertainties (“Solar Simulators.”).

● Same filters. As referenced before, E∝n, where ​E i​ s the power measured by the cell

and ​n i​ s the number of photons that is given to a cell. If a filter was to break, or was

damaged in the middle of testing, dispose of the filter appropriately, and measure all of

the readings from such filter again. This is because even superficial damage to a filter,

could cause more or less photons to go through the filter, resulting in inaccurate readings.

● Incident sunlight intensity. One of the experiments is based on the intensity reaching the

cell, but the solar simulator gives an approximate intensity that may not be the one

reaching the actual cell. In order to make sure that the intensity desired is the one hitting

the cell, a sensor will be placed next to the cell being tested, that will ensure the intensity

is the one needed.

 20

Experimental designs

Frequency spectrum experiment

The first experiment will compare the CSSC and the PSC in regards to different

wavelengths. PSC can use a different frequency spectrums depending on their composition,

therefore this experiment will investigate the frequencies that can be received by each cell in

order to conclude if they are complementary. The device that will modify the wavelength of light

inciding the cells, is called monochromator. Such light will come from a xenon lamp, and will be

reflected upon a collimating mirror, that will then reflect the light on a diffraction grating. That

will cause the light to diffract into different colours, and through a focusing mirror, the light will

fall right into a detector. The equipment will produce beams ranging from 400 nm to 800 nm,

taking measurements of the current and power of each cell, every 10 nm.

Figure 17: Mechanism of a Monochromator. Atomic

Absorption Spectroscopy Learning Module, Maryville

University, goo.gl/NcW2Ak. Accessed 23 Mar. 2018.

Incident Sunlight Intensity experiment

 21

The plan is to measure the efficiencies of both the CSSC, and the PSC. In turns, the cells

will be placed under a solar simulator, the Sun 2000 Solar Simulator, from Abet Technologies.

Figure 18 and figure 19: Different angles of the Solar Simulator.

With the use of filters, different amounts of irradiance can be given to each cell. The

irradiances tested will be​: 1, ​0.813, 0.515, 0.252, 0.126, and 0.051 Sun (1 Sun trial take place

without filter). With the use of an ammeter in series and a voltmeter in parallel, current (I) and

voltage (V) will be measured on each cell, for each of the trials. The solar simulator can be better

understood with the figure below. The solar simulator is connected to the variable resistance, of

value R, and the resistor goes from causing zero resistance, hence zero voltage and maximum

current, until the maximum voltage and zero current; according to the formula V= IR. The

equipment is connected to a computer that will record the current and voltage of each cell, under

each different filter.

 22

Figure 20: Mechanism of a Solar Simulator. Green, Martin

A., Solar Cells. The University of New South Wales, [1998].

 23

Experimental procedures

Method

1. Prepare a PSC and a CSSC, with dimensions of ​1 cm​2​. The PSC is already in such

dimension, but since the CSSC is not designed for this devices, use white duct tape to

cover the excess surface cell, leaving an operative surface of ​1 cm​2​.

2. Prepare the first experiment by connecting the monochromator to the computer that will

record the current and power from the frequency spectrum experiment. Place the ​1 cm​2

PSC inside the monochromator, and set the reading between 400 and 800 nm, which will

be the range being investigated. Adjust the monochromator to take 1 reading each 10 nm,

and start the trial. Once the trial finishes, load the raw data into the computer, and place it

in an excel for later processing. Then, substitute the PSC for the CSSC and repeat the

process. Save both sets of data in the same excel, and send them by email to yourself, so

that they can later be accessed from any computer.

3. Once the first experiment is completed, move towards the solar simulator. Place the PSC

underneath the simulator, and create a circuit with the solar simulator by connecting a

cable to it. Place the ammeter in series, and the voltmeter in parallel. Connect the cable to

the computer that was previously used in the same manner as before.

a. Before starting the trial, ensure with the use of the sensor, that the irradiance is the

one that you would like to investigate. Start the solar simulator, and once the trial

finishes, load the raw data into the computer and place it on a new excel for later

processing. The, after doing the 1st trial without a filter, to investigate the

irradiance of 1 Sun, repeat the trials with the different filters (0.813, 0.515, 0.252,
 24

0.126, and 0.051 Sun). Once all of the filters tested for the PSC, save all the raw

data into a single excel file and send it to yourself through email.

4. Once the readings with the PSC are taken, repeat steps 3, and 3.a with the use of the

CSSC.

Security risks

Caution, the solar simulator and the monochromator are very potent. Avoid placing

anything under the source of light as it is testing cells, since it can severely burn many materials.
 25

Data: Frequency spectrum experiment

(see processing of data in the appendix)

Graph 1. Wavelength VS Power (PSC)

Graph 2. Wavelength VS Power (CSSC)

 26

Graph 3. Wavelength VS IPCE for both Silicon (CSSC) and Perovskite (PSC)

Data: Incident Sunlight Intensity (ISI) experiment

 27

(see processing of data in the appendix)

Graph 4. Current VS Voltage ​(PSC)

Graph 5. Current VS Voltage ​(CSSC)

 28

Table 5. Finding the efficiency of the PSC by dividing maximum power point

by the irradiance given to a cell.

Filter label Irradiance Maximum power point - MPP Efficiency (%)

(W/m2) (W/m2) (MPP/Irradiance)

0.051_f_PK 51 7.959388889 16%

0.126_f_PK 126 15.78256 13%

0.252_f_PK 252 30.94266667 12%

0.515_f_PK 515 65.40447333 13%

0.813_f_PK 813 107.0417778 13%

1.000_f_PK 1000 133.2035333 13%

Graph 6. Irradiance against maximum power point (Perovskite)

 29

Table 6. Finding the efficiency of the CSSC by dividing maximum power

point by the irradiance given to a cell.

Filter label Irradiance Maximum power point - MPP Efficiency (%)

(W/m2) (W/m2) (MPP/Irradiance)

0.051_f_PK 51 14.585 29%

0.126_f_PK 126 14.143 11%

0.252_f_PK 252 25.868 10%

0.515_f_PK 515 58.605 11%

0.813_f_PK 813 80.672 10%

1.000_f_PK 1000 94.75 9%

Graph 7. Irradiance against maximum power point (CSSC)

 30

Analysis

In regards to the frequency spectrum experiment, the results are really surprising. It can be seen

by comparing graphs 1 and 2, that both cells are extremely similar to each other; they even look

like the same graph. Nevertheless, by taking a look at tables 1 and 2, it can be seen that the data

for such tables are just shockingly close (see appendix). This can be only due to the fact that the

PSC had a structure that tried to improve its efficiency as much as possible, hence copying the

CSSC range of 400-800 nm, since it’s the most efficient range.

On the other hand, the incident sunlight intensity experiment shows the graph that was expected

for PSC (Graph 4), but an pretty anomalous graph in regards to CSSC (Graph 5). The

equipments used were developed for PSC of 1 cm​2​, hence not prepared for a CSSC that was

covered with duct tape in order to fit the measuring equipment. For this reason, Graph 5 presents

25 times as much current as graph 4, and half its voltage. Emphasizing again how the CSSC was

not created for the measuring devices used, the high efficiency it has, makes it generate results

that escape the small scale experiments that the PSC are used to performing. This can be also

seen by comparing the smooth curves that Graph 4 presents, compared to those of Graph 5. In

Graph 5, lines curves such as the one by F_1.000 demonstrate through uneven curves, the

limitations of using this CSSC.

 31

Conclusion

As the results show, it can be concluded that the IPCE for both PSC and CSSC are very similar.

According to the research previously stated, it can be stated that the PSC that was investigated

was developed to have the greatest efficiency possible. Because of this, since solar cells normally

find the greatest efficiencies at 400-800 nm, this PSC was developed structurally to fit into such

frequency spectrum.

Therefore as it can be seen in the Graph 3, with the PSC lying under the CSSC during all the

wavelength investigated, it can be concluded that this two cells would not perform optimally if

they were put into a tandem cell. Since creating a tandem cell between them would be inefficient,

this opens new doors for investigation.

Furthermore, the incident sunlight intensity experiment showed that by changing the

amount of photons that are being given to a cell, following the proportion E∝n. This

means that the power measured by a cell is directly proportional to the amount of

photons that are given to such cell. This relationship is corroborated both by the silicon

and perovskite solar cells. When the irradiance of each filter is plotted against the

maximum power point of the cell at that time, both cells show a strong positive

correlation. The value of R^2 given in the graph, can be simplified to R, which is the

correlation coefficient. This number simply gives the correlation between the two

variable, where a value of R=1 reflects the maximum correlation possible. Their values

for R rounded to three significant figures are: 0.999 and 0.994 subsequently, proving the

maximum correlation possible between power and irradiance.

 32

Evaluation

This work effectively investigated the question that was proposed. The frequency spectrum

investigation shows how the emerging technology of PSCs have been rapidly increasing their

efficiencies, since they can produce similar amounts of power than the CSSC that was

investigated, as it can be seen in graphs 1 and 2. Nonetheless, as seen in graph 3, the PSC

investigated followed the same frequency spectrum as the CSSC, hence not being an efficient

tandem if both cells were to be combined.

On the other hand, graphs 6 and 7 shows the proportionality between the power measured by the

cell and the amount of photons given to the cell. This encourages investigations to take place in

increasing the individual efficiencies of each cell separately, since efficiency are constant

through changes in irradiance.

Therefore, in order to make the best PSC-CSSC tandem cell, the PSC should be organically

structured to work on the 0-400 nm wavelength range, while the silicon should focus on the

current range it has of 400-800 nm. If additionally, efficiencies are improved in both cells before

combining them, this would create the most efficient tandem cell according to my investigation.
33
 34

Acknowledgements

To my family,

Thank you dad for your tireless patience, thank you mum taking care of me when I got stressed,

and thank you brother for your curiosity on my topic and all your fun questions,

To all the staff at the ASV,

Thank you for helping me through every step of the way, and for allowing me to work on a lab

outside our city, for it has shown me to a world I’m now passionate of,

To all of the ICIQ members,

Thank you for opening the doors of your laboratories and allowing me to learn about how

photovoltaics are used in the real world, allowing me to expand my thoughts outside textbooks,

To my girlfriend,

Thank you for smiling everyday, and telling me that I will do something meaningful for society

in the future; thank you for making me hope,

 35

Works Cited

Estela, Francisco. Gestor De Proyectos e Instalaciones Energéticas. Energías Renovables,

Eficiencia y Auditorías Energéticas, Diseño y Explotación De Instalaciones Energéticas .

Master Ejecutivo ed., vol. 1 1, Instituto De Tecnología Eléctrica, 2008.

Ferreiro, Borja la Fuente. “Máster Universitario En Investigación Química y Química

Industrial.” Síntesis, Caracterización y Propiedades Dieléctricas De Perovskitas Híbridas

De Interés Fotovoltaico, Sept. 2017, pp. 1–49.

Grätzel, Michael. “Philosophical Transactions: Mathematical, Physical and Engineering

Sciences, Vol. 365, No. 1853, Energy for the Future (Apr. 15, 2007), Pp. 993-1005.”

Photovoltaic and Photoelectrochemical Conversion of Solar Energy, vol. 365, no. Energy

for the Future, ser. 1853, 15 Apr. 2007. 1853.

Henderson, Bruce. “The Experience Curve.” Https://Www.bcg.com, 1 Jan. 1968,

www.bcg.com/publications/1968/business-unit-strategy-growth-experience-curve.aspx​.

Accessed 24 Nov. 2018.

Martin, Richard. “Why the Future of Solar May Not Be Silicon-Based.” MIT Technology

Review, MIT Technology Review, 3 Mar. 2016,

www.technologyreview.com/s/600922/first-solars-cells-break-efficiency-record/​.

Accessed 25 Nov. 2018.

 36

National Geographic Society. “Solar Energy.” National Geographic Society, 14 Nov. 2012,

www.nationalgeographic.org/encyclopedia/solar-energy/​. Accessed 23 Nov. 2018.

“Photoelectric effect.” Factors affecting rate of a chemical reaction-catalyst-positive

catalyst-negative catalyst-inhibitor, ​www.citycollegiate.com/physicsXII_17a.htm​.

Accessed 25 Nov. 2018.

“Research Cell Efficiency Records.” Energy.Gov, National Renewable Energy Laboratory -

National Center for Photovoltaics, Sept. 2015,

www.energy.gov/eere/solar/downloads/research-cell-efficiency-records​. Accessed 26

May 2018.

“Semiconductor Band Gaps.” Total Internal Reflection,

www.hyperphysics.phy-astr.gsu.edu/hbase/Tables/Semgap.html​. Accessed 25 Nov. 2018.

Snaith, Henry J. “Present Status and Future Prospects of Perovskite Photovoltaics.” Nature

News, Nature Publishing Group, 23 Apr. 2018,

www.nature.com/articles/s41563-018-0071-z​. Accessed 9 Nov. 2018.

“Solar Simulators.” Abet Technologies, ​www.abet-technologies.com/solar-simulators/​. Accessed

30 May 2018.
 37

Wenham, Stuart R., et al. ​Applied Photovoltaics.​ Centre for Photovoltaic Devices and Systems,

[1995].

“What Is the Photovoltaic Effect?” MrSolar.com • Online Solar, LLC,

www.mrsolar.com/photovoltaic-effect/​. Accessed May 30 2018.

“World Energy Outlook.” October, ​www.iea.org/weo/​. Accessed 17 Oct 2018.

 38

Appendix

Using the raw data from Table 1 and Table 2; Graphs 1, 2 and 3 were produced. Graphs 1 and 2

were produced by plotting the wavelengths in nanometers measured, against the power in

miliwatts recorded.

Regarding Graph 3, some data processing was needed. The data collected on Table 1 at 400 nm

of wavelength, will be used to prove how the data points were obtained for Graph 3. This graph

displays IPCE as a percentage against wavelength in nm. Since the wavelength is the

independent variable which appears in the raw data, only the IPCE (Incident Photon to Current

Efficiency) needs to be calculated from the raw data in Table 1.

1. The wavelength in meters being tested must be converted into frequency in hertzs. To do

this, 400 nm will be converted into 400*10^-9 meters of wavelength. Then, using the

following equation: frequency = speed of light (2.998*10^8 ms^-1) / wavelength in

meters (400*10^-9). Frequency = 7.495*10^14 hertzs.

2. Using the frequency in hertzs, the energy in joules produced on the different wavelengths

can be calculated. Using: energy = Planck's constant (6.63*10^-34) * frequency in hertzs

(7.495*10^14). Energy = 4.97*10^-19 joules.

3. The current recorded in microamperes (uA) will be then converted into amperes (A). At

400 nm, the current recorded in uA was -0.559489, which will converted into

-0.559489*10^-6 amperes of current. In the case of a solar device with a dimension of

0.09 cm^2, current density will be found by dividing the current in amperes by 0.09,

giving a value of -6.22*10^-6 amperes per cm^2.

 39

4. Now, dividing the current density by the elementary charge of an electron, -1.6*10^-19

coulombs, we obtain the actual amount of electrons that are being excited per second,

which is 3.89*10^13 in this case.

5. Now, in order to find the amount of photons being given to the cell at different

wavelengths, the first step is to convert the power in miliwatts into watts. In this case,

57.915504 miliwatts, are 57. 915504*10^-6 watts. Now, dividing the power measured by

the solar device by the supposed energy in joules given to the cell, the amount of photons

that are incident on the solar device per second, is given, which is: 1.17*10^14.

6. Finally, dividing the amount of electrons that are emitted per second, by the amount of

photons that could potentially emit electrons per second, and multiply times 100, we

obtain the IPCE as a percentage. In this case: (3.89*10^13) / (1.17*10^14) = 0.33 * 100 =

33% of efficiency.

Regarding the second experiment, processing the raw data in tables 3 and 4, Graphs 4 and 5 were

obtained. Depending on the filter used, when receiving different solar irradiances, the solar

devices tested produced different currents and voltages. These are the ones that appear in the raw

data tables and were latter extrapolated into two graphs, each representing a solar device.
 40

Table 1. Wavelength VS Current and Power for the 1 cm​2 ​PSC

Wavelength Current (uA) Power (uW) Wavelength Current (uA) Power (uW)
(nm) (nm)

400 -0.559489 57.915504 610 -3.494 111.673416

410 -0.784787 65.85886 620 -3.465919 107.548384

420 -1.047363 73.373328 630 -3.329554 102.697016

430 -1.280835 80.097768 640 -3.189072 98.091896

440 -1.533129 89.106512 650 -3.092515 94.747208

450 -1.856883 101.004416 660 -3.032487 91.601344

460 -2.29062 118.848912 670 -3.047112 91.273368

470 -2.574734 132.906864 680 -3.139992 94.222696

480 -2.64082 127.71624 690 -3.056608 93.413304

490 -2.673263 124.6078 700 -2.811664 85.26624

500 -2.78101 122.685128 710 -2.719798 83.71592

510 -2.951717 123.61872 720 -2.666707 83.247072

520 -3.150532 126.97892 730 -2.653826 83.071904

530 -3.303927 131.233432 740 -2.608707 83.112064

540 -3.437068 135.5996 750 -2.433861 78.542424

550 -3.5255 138.006064 760 -2.235957 84.958496

560 -3.577869 139.343632 770 -1.542154 79.5316

570 -3.632995 138.107696 780 -0.723464 60.083984

580 -3.664707 135.569264 790 -0.357819 60.619244

590 -3.636243 130.149096 800 -0.202668 61.141204

600 -3.539905 119.093616

 41

Table 2. Wavelength VS Current and Power for the 1 cm​2 ​CSSC

Wavelength Current (uA) Power (uW) Wavelength Current (uA) Power (uW)
(nm) (nm)

400 -8.467896 59.925436 610 -46.12084 113.226688

410 -11.534867 67.933384 620 -46.546804 109.376296

420 -14.566216 75.267904 630 -45.471792 104.368384

430 -17.494847 81.641672 640 -44.313085 98.686

440 -20.939791 89.788648 650 -43.530446 95.516656

450 -25.095866 103.169336 660 -42.998923 92.58012

460 -31.098429 121.05592 670 -43.712549 92.262696

470 -35.239747 134.51216 680 -45.700428 95.527944

480 -35.941302 130.331088 690 -45.140478 95.209376

490 -36.427744 126.962112 700 -41.998155 86.09096

500 -37.234935 125.416656 710 -41.636955 84.77696

510 -38.485228 125.562816 720 -41.449002 83.929088

520 -40.227518 129.482856 730 -41.639247 83.91828

530 -41.602671 132.547208 740 -41.233947 82.924248

540 -42.92763 137.183984 750 -40.118262 78.981632

550 -44.144759 140.578144 760 -43.589247 86.07976

560 -44.874483 141.326512 770 -38.789836 79.905392

570 -45.588782 140.312944 780 -30.505615 60.24474

580 -46.556419 138.345488 790 -30.751497 61.185128

590 -46.689143 131.636592 800 -31.390329 61.738216

600 -45.960809 120.636464

 42

Table 3. ISI VS Voltage and Current for the 1 cm​2 ​PSC

Filter name F_0.051 F_0.126 F_0.252 F_0.515 F_0.813 F_1.000

ISI (Suns) 0.051 0.126 0.252 0.515 0.813 1.000

Voltage (V) Current (mA) Current (mA) Current (mA) Current (mA) Current (mA) Current (mA)

0 0.1285 0.2531 0.4736 0.9451 1.4822 1.8331

0.01 0.1281 0.2527 0.4726 0.945 1.4834 1.8279

0.02 0.1279 0.2519 0.4724 0.9454 1.4803 1.829

0.03 0.1277 0.2515 0.471 0.9425 1.4779 1.8233

0.04 0.1276 0.2512 0.4702 0.9396 1.4749 1.8225

0.05 0.1274 0.2509 0.4693 0.939 1.4733 1.8221

0.06 0.1269 0.2501 0.4691 0.938 1.4713 1.8223

0.07 0.1271 0.2499 0.4679 0.9355 1.4681 1.8152

0.08 0.1267 0.249 0.4676 0.935 1.4688 1.8142

0.09 0.1266 0.2491 0.4673 0.9329 1.4652 1.8133

0.1 0.1263 0.2489 0.4663 0.9327 1.4634 1.8086

0.11 0.1263 0.2483 0.4645 0.9316 1.463 1.81

0.12 0.1261 0.2482 0.4649 0.9295 1.46 1.8076

0.13 0.1257 0.248 0.4636 0.9289 1.457 1.8053

0.14 0.1258 0.2475 0.4632 0.9268 1.4568 1.8032

0.15 0.1254 0.2464 0.4634 0.9243 1.4539 1.8025

0.16 0.125 0.247 0.4622 0.9252 1.4533 1.8

0.17 0.125 0.2462 0.4615 0.9238 1.4494 1.7991

0.18 0.125 0.2462 0.4616 0.922 1.4497 1.7953

0.19 0.1249 0.2456 0.4605 0.9208 1.4491 1.7923

0.2 0.1246 0.2452 0.4591 0.9206 1.4463 1.7898

0.21 0.1245 0.2445 0.4586 0.9186 1.4434 1.794

0.22 0.1241 0.2444 0.4583 0.9165 1.4405 1.7871

0.23 0.1242 0.244 0.4574 0.9158 1.4429 1.7867

 43

0.24 0.124 0.2436 0.4571 0.9144 1.4381 1.7828

0.25 0.1237 0.2433 0.4559 0.914 1.4375 1.7824

0.26 0.1236 0.2427 0.4555 0.9116 1.4355 1.7834

0.27 0.1233 0.2424 0.4542 0.9125 1.4337 1.7788

0.28 0.123 0.2425 0.4536 0.9092 1.4343 1.7751

0.29 0.123 0.2414 0.4538 0.9079 1.4298 1.7775

0.3 0.1226 0.2411 0.4525 0.9078 1.428 1.7726

0.31 0.1225 0.241 0.452 0.9053 1.4297 1.7726

0.32 0.1221 0.2403 0.4513 0.9035 1.4248 1.7674

0.33 0.1222 0.2401 0.4497 0.904 1.4243 1.7624

0.34 0.1218 0.2394 0.4494 0.9016 1.4211 1.7637

0.35 0.1216 0.2391 0.4481 0.9014 1.4206 1.758

0.36 0.1215 0.2383 0.4477 0.9002 1.4189 1.7576

0.37 0.121 0.2377 0.4477 0.8969 1.4161 1.7575

0.38 0.1208 0.2376 0.446 0.8981 1.415 1.7545

0.39 0.1207 0.237 0.4457 0.8951 1.4123 1.7507

0.4 0.1204 0.236 0.4446 0.8938 1.4094 1.7499

0.41 0.1203 0.236 0.4436 0.8917 1.4089 1.746

0.42 0.1199 0.2352 0.443 0.8893 1.4042 1.7475

0.43 0.1197 0.2346 0.4416 0.8885 1.4036 1.7406

0.44 0.1195 0.2339 0.4403 0.8864 1.4017 1.7397

0.45 0.119 0.2336 0.44 0.884 1.3976 1.7345

0.46 0.1188 0.2329 0.4377 0.8829 1.3956 1.7329

0.47 0.1183 0.232 0.4368 0.8807 1.3942 1.7306

0.48 0.1181 0.2319 0.4371 0.8799 1.3903 1.7252

0.49 0.1175 0.2302 0.4348 0.8772 1.3885 1.7263

0.5 0.1173 0.2299 0.4334 0.876 1.3847 1.7198

0.51 0.1168 0.2288 0.4325 0.8743 1.3845 1.7159

0.52 0.1165 0.2282 0.4311 0.8714 1.3823 1.7116

 44

0.53 0.1161 0.2271 0.4295 0.8694 1.3785 1.7083

0.54 0.1155 0.2264 0.4279 0.866 1.3749 1.7047

0.55 0.115 0.2253 0.4261 0.8659 1.3699 1.7023

0.56 0.1146 0.2241 0.4253 0.8613 1.3684 1.6945

0.57 0.1142 0.2233 0.423 0.8585 1.364 1.6922

0.58 0.1134 0.2221 0.4223 0.8573 1.3615 1.69

0.59 0.1129 0.221 0.4195 0.8543 1.3568 1.6774

0.6 0.1121 0.2196 0.4184 0.8511 1.3533 1.6753

0.61 0.1115 0.2185 0.4157 0.8474 1.3499 1.6686

0.62 0.1107 0.2168 0.4142 0.8443 1.3443 1.6637

0.63 0.1099 0.2153 0.4124 0.8419 1.3397 1.6574

0.64 0.1091 0.2137 0.4094 0.8379 1.3371 1.6507

0.65 0.1082 0.2125 0.4069 0.8326 1.3321 1.6479

0.66 0.1072 0.2108 0.4042 0.8305 1.3251 1.6371

0.67 0.106 0.2091 0.4017 0.8256 1.319 1.6319

0.68 0.1049 0.207 0.3987 0.8221 1.3155 1.6263

0.69 0.1037 0.2047 0.3958 0.8172 1.3103 1.616

0.7 0.1023 0.2022 0.3923 0.8104 1.3018 1.6054

0.71 0.1007 0.1999 0.3882 0.8061 1.2946 1.5976

0.72 0.099 0.1973 0.3845 0.8003 1.2876 1.5883

0.73 0.0974 0.1944 0.3808 0.7931 1.2812 1.5812

0.74 0.0953 0.1914 0.3755 0.7878 1.2713 1.5694

0.75 0.0929 0.1876 0.3713 0.781 1.2646 1.5617

0.76 0.0901 0.1838 0.3661 0.7722 1.2519 1.5482

0.77 0.0873 0.1797 0.3598 0.7629 1.2433 1.5355

0.78 0.0838 0.1748 0.3534 0.7547 1.2305 1.5239

0.79 0.0802 0.17 0.3458 0.7437 1.2176 1.5107

0.8 0.0757 0.1639 0.3377 0.7326 1.2042 1.4949

0.81 0.0707 0.1573 0.3295 0.721 1.1883 1.4791

 45

0.82 0.0649 0.15 0.3189 0.7067 1.1726 1.462

0.83 0.0582 0.1415 0.3081 0.6915 1.1535 1.4434

0.84 0.0505 0.1319 0.2957 0.6733 1.1323 1.422

0.85 0.0417 0.1211 0.2812 0.6533 1.1074 1.3997

0.86 0.0311 0.1086 0.2642 0.631 1.0813 1.3729

0.87 0.0187 0.0941 0.2463 0.6062 1.0479 1.3419

0.88 0.0044 0.0772 0.2248 0.576 1.0145 1.3108

0.89 0.0582 0.1997 0.5422 0.977 1.274

0.9 0.0358 0.1714 0.5036 0.9318 1.2347

0.91 0.0094 0.1384 0.4596 0.8812 1.1859

0.92 0.1001 0.4085 0.8211 1.1332

0.93 0.0563 0.349 0.7535 1.0734

0.94 0.0052 0.2812 0.6767 1.0032

0.95 0.2009 0.5874 0.9227

0.96 0.1085 0.4814 0.8328

0.97 0.0027 0.3622 0.7315

0.98 0.2259 0.6154

0.99 0.0685 0.4839

1 0.3336

1.01 0.1681
 46

Table 4. ISI VS Voltage and Current for ​the 1 cm​2 ​CSSC

Filters F_0.051 F_0.126 F_0.252 F_0.515 F_0.813 F_1.000

ISI (Suns) 0.051 0.126 0.252 0.515 0.813 1.000

Voltage (V) Current (mA) Current (mA) Current (mA) Current (mA) Current (mA) Current (mA)

0 6.7501 6.6045 11.171 26.32 40.787 47.585

0.01 6.7356 6.6023 11.14 26.282 40.527 48.137

0.02 6.7025 6.5821 11.115 26.208 40.277 47.611

0.03 6.7049 6.5608 11.098 26.105 39.996 47.208

0.04 6.6783 6.5493 11.072 26.03 39.739 46.726

0.05 6.6707 6.5268 11.074 25.98 39.461 46.348

0.06 6.6366 6.5155 11.034 25.893 39.129 45.851

0.07 6.6375 6.4791 10.993 25.79 38.885 45.383

0.08 6.5998 6.4548 10.992 25.667 38.553 44.924

0.09 6.5704 6.4245 10.95 25.501 38.235 45.775

0.1 6.5591 6.4193 10.906 25.363 37.852 45.312

0.11 6.5136 6.39 10.897 25.198 37.468 44.764

0.12 6.4865 6.3492 10.837 25.045 37.051 44.233

0.13 6.4638 6.3254 10.772 24.895 36.644 43.712

0.14 6.4228 6.2763 10.735 24.676 36.256 43.158

0.15 6.3748 6.2475 10.669 24.534 35.825 42.594

0.16 6.3375 6.196 10.618 24.289 35.319 42.009

0.17 6.2914 6.1422 10.553 24.073 34.888 41.39

0.18 6.2308 6.0921 10.476 23.835 34.456 40.765

0.19 6.1663 6.022 10.383 23.578 33.899 40.105

0.2 6.0962 5.9649 10.3 23.317 33.383 39.433

0.21 6.0216 5.882 10.201 23.046 32.819 38.748

0.22 5.931 5.7985 10.083 22.728 32.236 38.05

0.23 5.8388 5.7092 9.949 22.387 31.655 37.331

 47

0.24 5.7418 5.6061 9.8096 22.064 31.034 36.587

0.25 5.6381 5.4771 9.6497 21.659 30.404 35.794

0.26 5.4973 5.3632 9.4823 21.281 29.726 35.006

0.27 5.361 5.2186 9.3002 20.825 29.039 34.154

0.28 5.2051 5.0511 9.0829 20.395 28.316 33.307

0.29 5.0292 4.8763 8.8893 19.918 27.584 32.424

0.3 4.8294 4.6784 8.6201 19.408 26.816 31.509

0.31 4.6331 4.4661 8.3444 18.876 26.023 30.565

0.32 4.3913 4.2348 8.0727 18.314 25.201 29.608

0.33 4.147 4.0022 7.7395 17.702 24.323 28.566

0.34 3.8737 3.7214 7.3985 17.069 23.439 27.538

0.35 3.5667 3.4098 7.0262 16.392 22.494 26.447

0.36 3.2449 3.0845 6.6371 15.707 21.536 25.341

0.37 2.8867 2.7321 6.1972 14.928 20.537 24.236

0.38 2.504 2.3415 5.7265 14.12 19.521 23.043

0.39 2.0874 1.9296 5.2289 13.304 18.447 21.848

0.4 1.6311 1.4688 4.6845 12.435 17.344 20.598

0.41 1.1554 0.9864 4.088 11.506 16.203 19.326

0.42 0.6371 0.4558 3.4531 10.541 15.034 17.997

0.43 0.0687 2.7739 9.5071 13.799 16.623

0.44 2.0534 8.4653 12.533 15.234

0.45 1.2777 7.3634 11.216 13.804

0.46 0.4682 6.2065 9.8293 12.318

0.47 5.0059 8.413 10.784

0.48 3.749 6.9226 9.161

0.49 2.4377 5.3904 7.5085

0.5 1.0692 3.8 5.7865

0.51 2.1439 4.0343