CHARACTERISTICS OF A SOLAR CELL (MP1)

Laboratory Course II. Group BA-C-09

Cormac Meagher

Elena Bazo González

Carmen Sánchez Pérez

January 2019
 Contents
1. Abstract 02
2. Theoretical Fundamentals 02
2.1 Conductivity and Resistivity 02
2.2 Semiconductors 03
2.2.1 Fundamentals of Semiconductors Physics 04
2.3 Physics of Solar cells 04
2.3.1 The pn-junction 05
2.3.2 The solar cell 07
2.4 Four-point Method 09
2.5 Differential layer conductivity 10
3. Experimental Procedure 12
3.1 Dopant 12
3.2 Solar cell 12
4. Analysis of Data 13
4.1 Dopant 13
4.2 Solar cell 17
5. Conclusion 23
6. References 24

1
 1. Abstract

The performance of this experiment could be divided in two main parts. During the
first part, the dopant concentrations of several semiconductor materials were found by
measuring their thickness and resistance. Furthermore, the dopant profile pn-junction was
determined. In the second part of the lab, the electrical properties of various types of solar
cells (polycrystalline and tandem) and their dependence with temperature and type of
irradiation were studied and compared within them.

2. Theoretical Fundaments
2.1 Conductivity and Resistivity

For electric current 𝐼 through a rod-shaped conductor with cross section 𝐴 and length
𝐿 holds the relationship:
𝑈
𝐼 = −𝑒𝐴𝑛𝑣𝑑 = 𝑒𝐴𝑛µ𝑛 (1)
𝐿

Where 𝑒 is the charge of the electron, 𝑛 is the electronic density, 𝑣𝑑 is the drift velocity of
𝑈
the electrons, which is proportional to the electric field strength 𝐸 = 𝐿 , µ𝑛 is the electron
mobility and 𝑈 is the potential in the conductor.

Fig 1: Representation of a conductor.

From Equation (1) it follows that the electrical resistance R is given by:
𝑈 1 𝐿
𝑅= = (2)
𝐼 𝑒𝑛µ𝑛 𝐴

The resistivity ρ is defined by the relationship:

𝐿
𝑅=𝜌 (3)
𝐴

Thus, for electron conductors where σ is referred as conductivity:

1
𝜌= = 𝑒𝑛µ𝑛 (4)
𝜎

2
 2.2 Semiconductors

Semiconductors are those elements or compounds whose resistivity is between the

resistivity of metals and insulators, being this range between 10−4 and 1012 𝛺𝑐𝑚.

The resistivity of semiconductors is strongly temperature-dependent. It decreases (i.e.

the conductivity increases) when the temperature increases, at least at very high and very
low temperatures. This property allows a more precise distinction between metals and
semiconductors than the one referring to its resistivity arbitrary limits. Metals become
more conductive when approaching the absolute zero point to finally become
superconductors, while semiconductors with sufficient purity become less conductive when
the temperature increases. The resistivity of semiconductors can be changed by the addition
of impurities within a wide range, which is indicated by the shaded areas in Fig 2.

Fig 2: Specific resistance of different materials at room temperature.

In semiconductors, the fact that two types of electronic (non-ionic) charge carriers
exist is unique. One them are the conduction electrons with negative charge and the others
are the gaps in the electronic structure of the valence bonds, the so-called holes, which can
be described by particles with positive charge. Therefore, for semiconductors the
conductivity follows:
1
= 𝑒𝑛µ𝑛 + 𝑒𝑝µ𝑝 (5)
𝜎
3
Where 𝑝 is the hole concentration [𝑐𝑚−3], and µ𝑝 is the hole mobility [𝑐𝑚2 𝑉 −1 𝑠 −1 ].

In most cases, however, one deals with doped semiconductor crystals in which one type of
charge carriers at room temperature predominates. In such cases, the dopant concentration
can be determined from resistance measurements when the mobilities are known.

Fig 3: Mobilities of Si and Ge at T=300K.

2.2.1 Fundamentals of Semiconductors in Physics

The properties of a semiconductor (SC) are crucial in order to determine its band gap. At
the absolute 0, the lower valence band is completely filled with charge carriers, while the
upper conduction band is completely empty. In an ideal SC, the energy range that separates
these two bands does not contain any allowed states and is referred as band gap. For
crystalline silicon, this band is 1.1 eV at T=300 K. For amorphous silicon it is
approximately 1.7 eV.

Semiconductors are doped for most technical applications. Impurities are introduced into
the crystal lattice of the semiconductor to modify their conductivity. Depending on the
impurities that are introduced, one can distinguish between n-type and p-type
semiconductors. In Silicon, an IV-valent atom is replaced by either an III-valent atom (p-
type dopant) or by a V-valent atom (n-type dopant). The III-valent element acts as an
acceptor and takes and electron, creating a hole in the valence band, which causes holes to
be the majority charge carriers (p-type SC). A V-valent element acts as a donor and
donates an electron to the conduction band (n-type SC).

For doped SC, one can determine electron and hole densities. For crystalline silicon,
𝑛 · 𝑝 = 2.1 · 1019 𝑐𝑚−3 (𝑇 = 300 𝐾).

2.3 Physics of Solar cells

Solar cells are semiconductor devices used to absorb energy from photons and convert
it into electrical voltage. The main principle behind the working of a solar cell is the
photovoltaic effect. When the photovoltaic effect occurs within a semiconductor a voltage
is created, thus a solar cell is a semiconductor whose physical properties are specific for
that type of solar cell.

4
 2.3.1 The pn-junction

A PN-junction is the combination of p-type and n-type semiconductors. The two types
of semiconductor materials are differently doped with acceptor dopants in the p-type and
donor dopants in the n-type materials. When these two materials are combined, they create
a PN-junction. In this PN-junction the electrons from the n-type material diffuse into the
p-type material where there is a low density of electrons and then the p-type electron hole
regions diffuse into the n-type material. A space charge occurs due to atomic hulls, positive
in the n-type region, negative in the p-type region. This space charge creates an electric
field which causes further diffusion to cease. The space charge zone is the name of the area
enclosed in the electric field. The net current flow in the pn-junction is zero when the Fermi
levels are equal and thermal equilibrium is reached. A diffusion current physically describes
the movement of electrons and electron holes when there is a difference in concentration
on either side of the junction.

Fig 4: Depiction of diffusion between n-type and p-type junctions with regards to potential.

The electron current density is described by the equation [2]:

𝑑𝑛(𝑥)
𝐽𝑑𝑖𝑓𝑓 = 𝑞 ∙ 𝐷𝑛 ∙ (6)
𝑑𝑥

where, 𝐷𝑛 is the diffusion constant and 𝑞 is the electron charge.

The electric field potential which is inherently negative, −𝑈𝑑𝑖𝑓𝑓 , creates a drift current
in the opposite direction to the diffusion. The charges involved in the drift process are
significantly less dense than the majority charge carriers in the diffusion process.
The drift current is given by [3]:

𝐽𝐷𝑟𝑖𝑓𝑡 = 𝑛 ∙ 𝑒 ∙ 𝜇 ∙ 𝐸 (7)

5
 Where, µ is the mobility of the material.

When thermal equilibrium is reached the sum of the drift and diffusion currents is
zero. When this condition is fulfilled it is possible to calculate the diffusion voltage ΔU
(potential difference). This can be calculated using the standard diode characteristic
equation:
𝑒𝑈
𝐼 = 𝐼0 [𝑒𝑥𝑝 ( )] (8)
𝑛𝑘𝐵 𝑇

Where, 𝐼0 is the saturation current, which is material parameter and temperature

dependent. 𝐾𝐵 is the Boltzmann constant, 𝑇 is temperature and 𝑛 is the ideality factor.
The ideality factor describes the extent of the space charge zone and is set within the
region of (1 ≤ n ≤ 2). In this example n = 1 is a good approximation of the ideality factor.

The current-voltage characteristic curve can be explained by the relation between

majority charge movement and directional polarity. The diffusion voltage (−𝑈𝑑𝑖𝑓𝑓 ) is
amplified when n- region is positive caused by polarity being reversed (U < 0). This shows
that the attenuation of the potential is dependent of the majority charge carriers. The
exception to this is when the minority charge carriers of the drift current determine the
current at the extreme of high polarity reversion which supresses the majority charge
carriers.

The saturation current can be described as the drift current purely of minority charge
carriers, this can be found by taking the potential as it approaches infinity (U→∞).

When the potential is in the forward direction (𝑈 → ∞) this causes a reduction in the
diffusion current which then allows the majority charge carriers to occur. The basic
characteristic diode equation is used to describe a theoretically idea diode, this is not the
same case for solar cells as it is for electrical components such as diodes and transistors. In
terms of cost-effective production, the solar cell is greatly simplified. There is a parallel
“shunt” and a series resistance introduced into the circuit diagram used to depict the solar
cell electrically as ohmic losses can no longer be omitted from the diode equation. The
“shunt”, 𝑅𝑠ℎ is used to describe a current created by impurities or crystal defects in the
depletion zone where the n-type and p-type materials meet.

6
 Fig 5: Solar cell equivalent circuit.

The solar cell equivalent circuit depicted in Fig. shows the circuit equivalent of a solar
cell with ohmic loss taken into account. At low currents through the pn-junction the
parallel resistance becomes a factor. In an ideal diode the “shunt”, 𝑅𝑠ℎ , approaches infinity.
A realistic solar cell can have an 𝑅𝑠ℎ value of > 1000 Ωcm-2. The series resistance, 𝑅𝑠 ,
becomes a factor when current is low through the pn-junction. In an ideal solar cell, the
series resistance is 0. In a good solar cell 𝑅𝑠 ≪ 𝑅𝑠ℎ , and the characteristic curve of the
diode exponent can be described with the equation:
𝑒𝑈 𝑈𝐼 − 𝑅𝑠
𝐼 = 𝐼0 [[𝑒𝑥𝑝 ( )] + (9)
𝑛𝑘𝐵 𝑇 𝑅𝑠ℎ

The greater the ohmic losses of the cell the more linear the characteristic curve
becomes.

2.3.2 The Solar cell

When no light is incident on a solar cell the only way for minority charge carriers to
move from the valence band to the conduction band is by thermal excitation. When light
is incident upon the solar cell electron-hole pairs are generated by the absorption of
photons. The number of minority charge carriers is increased by a substantial amount,
several orders of magnitude. The majority charge carriers don’t increase to any substantial
number. The space charge zone of crystalline silicon is ~1μm, light penetrates around 100
times deeper into the material. As the light penetrates outside the electric field the charge
carriers generated move through electric field-free space. Charge separation takes place
when the charge carriers diffuse into the space charge zone. Diffusion current stays mostly
the same while drift current of minority charge carriers increases largely.

This causes two charge zones to be formed, the p-zone is positively charged while the
n-zone is negatively charged. The voltage generated can be dissipated with contact.
The photocurrent (𝐼𝑝ℎ ) flows in the opposite direction to the diode current and thus has a
negative sign. The voltage current follows the same characteristic curve of the diode and
corresponds to the diode current sum. The drift current of the optically generated minority
carriers is described by the photocurrent.
eU
𝐼 = 𝐼0 [𝑒𝑥𝑝 − 1] − 𝐼𝑝ℎ (10)
𝑛𝑘𝐵 𝑇

7
 Fig 6: The characteristic curve of a lighted and unlighted solar cell for current-voltage
[6]
relationship.

A current of minority charge carriers is created when the solar cell is illuminated, this
current is in the reverse direction (I < 0) in the cell where the diode poles itself in the flow
direction (U > 0), as positive charge (holes) accumulate in the p-region and negative charge
(electrons) accumulates in the n-region. This means the entire curve is not of value, just
the area in the quadrant where (U > 0, I < 0). When an external voltage is applied the
other areas of the curve become relevant but otherwise are not of interest. Solar cell
characteristics are determined by three factors, Fill factor (𝐹𝐹), Open circuit voltage (𝑈𝑜𝑐 )
and short circuit current (𝐼𝑠𝑐 ). 𝐼𝑠𝑐 is defined as the maximum current that can be taken
from a short-circuited solar cell. At U = 0 the diode equation gives:

𝐼𝑠𝑐 = −𝐼𝑝ℎ (11)

As light output increases 𝐼𝑠𝑐 increases linearly with the photocurrent due to the
correspondence observed in the equation. The maximum short-circuit current is calculable
from the band gap energy. Large band gap energies correspond to lower maximums for
short-circuit current as only high energy photons of more energy than the band gap
contribute to the photocurrent utilising only a small percentage of solar energy incident
upon the cell. The maximum short-circuit current for a standardised solar spectrum is not
reached by a crystalline silicon solar cell due to recombination defects in the space charge
zone lead to not all generated charges being separated. The open circuit voltage is what
accumulates on the diode when no current is being taken, I=0 and n=1.
𝐾𝑇 𝐼𝑝ℎ 𝐾𝑇 𝐼𝑝ℎ
𝑈𝑜𝑐 = ln [ + 1] ≈ ln [ ] 𝑓𝑜𝑟 𝐼𝑝ℎ ≫ 𝐼0 (12)
𝑒 𝐼0 𝑒 𝐼0

This equation shows the logarithmic increase in 𝑈𝑜𝑐 with respect to photocurrent
increase, therefore with light output. The open circuit voltage can be imagined as the

8
maximum voltage reached when the charge separation of the diffusion potential occurs
until the potential is compensated by the voltage of the separated charges.

This diffusion voltage can be seen as the limit for the open circuit voltage. The
maximum possible open circuit voltage increases with the band gap due to the diffusion
potential being limited by the band gap energy. For an optimised cell being controlled the
maximum extractable power is given by the equation:

𝑃𝑚𝑎𝑥 = 𝐼𝑠𝑐 ∙ 𝑈𝑜𝑐 ∙ 𝐹𝐹 (13)

FF is used to describe the fill factor (~60 – 80%). Band gap energy affects 𝐼𝑠𝑐 and 𝑈𝑜𝑐
but in opposite directions. Optimal band gap energy for a photovoltaic semiconductor is
1.4 eV. Theoretical efficiencies for silicon solar cells are 28%, production of 24% efficient
solar cells are in production in laboratories while industrially produced solar cells have an
efficiency of ~15%. The fill factor can be calculated using the equations below. The terms
𝐼𝑚𝑝𝑝 and 𝑈𝑚𝑝𝑝 refer to current at maximum power point (MPP) and voltage at maximum
power point respectively. Using these terms, the fill factor and efficiency are calculated in
the equations below.
𝐼𝑚𝑝𝑝 𝑈𝑚𝑝𝑝 𝐼𝑚𝑝𝑝 𝑈𝑚𝑝𝑝 𝐹𝐹𝐼𝑠𝑐 𝑈𝑜𝑐
𝐹𝐹 = 𝜂= = (14)
𝐼𝑠𝑐 𝑈𝑜𝑐 𝑃𝑙𝑖𝑔ℎ𝑡 𝑃𝑙𝑖𝑔ℎ𝑡

2.4 Four-point method

For the determination of the resistivity of Si and Ga disks, the four-point method is
commonly used. The principle of this method is shown in Fig. It is common practice to
arrange the probes linearly and give them an approximately equal spacing. The current is
supplied and discharged via the two outer contacts. For thin and homogeneous disks, the
relationship applies:
𝑈 𝜋
𝜌= 𝑑 𝐹(𝑑/𝑠) (15)
𝐼 𝑙𝑛2

Where d is the thickness of the slice and s the distance of the measuring tips.

The factor F(d/s), which includes a correction for finite slice thickness, is for (d/s >
0.5) no longer negligible. If necessary, further corrections may be required for small samples
or at measurements at temperatures significantly different from the standard (23ºC).

9
 Fig 7: Basic structure of the four-point method.

2.5 Differential layer conductivity

For very then slices (d/s<<1) or thin films on insulating substrates, 𝐹(𝑑/𝑠) ≈ 1 is a
very good approximation. Thus, the resistivity is given by:
𝑈 𝜋
𝜌= 𝑑 (16)
𝐼 𝑙𝑛2

While the quantity 𝑈/𝐼 = 𝑅 characterizes the electrical resistance, its reciprocal 𝐼/𝑈 =
1/𝑅 represents a measurement of the “conductivity” at the measuring point of the sample.
The so-called layer conductivity is defined as:
𝑙𝑛2 𝐼
𝐺≡ (17)
𝜋 𝑈

One can obtain a particularly simple expression for the layer conductivity in thin
homogeneous slices by combining Eq and Eq:
𝑑
𝐺= (18)
𝜌

If the resistivity of a SC wafer changes perpendicular to the plane of the slab

(transversely), it can theoretically be decomposed into many homogeneous sub-slices of
thickness 𝛥𝑥𝑘 .

Fig 8: Representation of a semiconductor layer structure and the corresponding

replacement circuit diagram.

10
From the expression for the effective resistance of a parallel circuit:
1 1 1 1
= + + +⋯ (19)
𝑅 𝑅1 𝑅2 𝑅3

One can derive the following relationship:

𝑛
1 𝛥𝑥𝑘
=∑ (20)
𝜌̅ 𝜌𝑘
𝑘=1

Here, 𝜌̅ represents the average resistivity of the entire layer package.

Fig 9: Approximation of an inhomogeneous semiconductor by means of thin sublayers.

From the layer conductivity of the homogeneous structure, Eq. (18) and Eq. (20) give:
𝑛
𝑑 𝛥𝑥𝑘
𝐺𝑜 = = ∑
𝜌̅ 𝜌𝑘 (21)
𝑘=1

Where the sub-index of 𝐺𝑜 is intended to indicate the intact initial state.

After the first layer has been removed, the result is a second four-point measurement:
𝑛
𝛥𝑥𝑘
𝐺1 = ∑
𝜌𝑘 (22)
𝑘=2

For further layer removal and repeated resistance measurements one can obtain:

11
 1 −(𝐺𝑘 − 𝐺𝑘−1 )
=
𝜌𝑘 𝛥𝑥𝑘 (23)

In the limit,

1 𝑑𝐺(𝑥) 𝑙𝑛2 𝑑𝐼/𝑈

lim =− =−
𝛥𝑥𝑘 →0 𝜌 𝑑𝑥 𝜋 𝑑𝑥 (24)

Eq. (23) and Eq. (24) form the basis for determining a dopant depth profile.

3. Experimental Procedure
3.1 Dopant

The properties of semiconductors were explored by analysing the concentration of

dopants in several samples of different materials.

The samples analysed were 4 differently doped silicon samples, a phosphor doped
silicon disc that had undergone a gallium diffusion process previously on either side and an
intrinsic germanium sample.

Semiconductor resistivity is dependent on the concentration of charge carriers present

in the sample. By measuring the resistivity of the samples, it is possible to determine the
dopant concentration in the silicon sample as well as the germanium sample.
The Gallium diffused phosphor doped silicon sample was used to measure the penetration
depth of gallium atoms from diffusion. Layers of the sample were removed from the sample
to measure at different sample depths and resistance was measured at each new layer. The
concentration of dopant in terms of depth is calculable from this measurement.

3.2 Solar cell

To measure the characteristics of the solar cell in relation to the voltage-current

characteristic curve of the polycrystalline solar cell. The measurements were taken under
different temperature and incident light conditions. As a comparative sample a tandem cell
(Amorphous silicon and microcrystalline sheet).

The polycrystalline sample was measure with no light and with the solar spectrum at
different temperatures, those being 20oC, 35oC and 50oC.

The tandem cell was measure with no incident light, under the solar spectrum at
different temperatures being 20oC and 50oC. A filter set to mimic the atmosphere was also
used and the solar spectrum was measured at 20oC.

12
 4. Analysis of Data

As we stated before, the performance of this lab could be divided in two parts. In the
first part, the dopants of various semiconductor materials are determined by means of
resistance measurements. Furthermore, the dopant profile of a pn-junction is determined.
In the second part, the electrical properties dependent on the temperature and type of
insolation of two types of solar cells are determined and compared.

4.1 Dopant

The four-point probe method was used to measure the resistivity of the samples in the
first part. The four-point probe is preferable over a two-point because the contact and
spreading resistances associated with the two-point probe are large and the true resistivity
can't be actually separated from the measured resistivity. In a four-point, very little contact
and spreading resistance is associated with the voltage probes and hence one can obtain a
[7]
fairly accurate calculation of the resistivity.

For each sample, we took 5 measurements of the resistance with the 4-point method
and 3 of the thickness with a caliper. The average of each was obtained and their
uncertainties were calculated by error propagation. All the values are gathered in Table 1.

Sample Type R (𝛀) Thickness d (mm)

1 Si -type (46.8±1.5) (0.448±0.001)
2 Si n-type (9.80±0.57) (0.4537±0.0065)
3 Si n-type (6.83±0.22)· 10−2 (0.684±0.002)
4 Si p-type (1.908±0.067)· 10−3 (2.0023±0.0035)
5 Ge intrinsic (24.5±6.0) (2.114±0.003)
Table 1: Measured resistance and thickness of each sample.

The resistivity was obtained using equation (15) and its uncertainty again by error
propagation.

In Table 2 it can be seen, the quotient d/s is smaller than 0.5, so the approximation
𝐹(𝑑/𝑠) ≈ 1 can be made. For the sample 4 and 5 we got the values of the correction factor
from Figure 10[A]. The red dot corresponds to the value for sample 4 and the blue one to
sample 5. It should be noted as well that the value between the indentation tips is 𝑠 =
1.59 𝑚𝑚. Table 2 also shows the values calculated for the resistivities and the correction
factor of each sample, as well as their errors.

13
 Fig 10: Correction factor 𝐹(𝑑/𝑠) and 𝐺(𝑑/𝑠) for different values of 𝑑/𝑠

Sample Type d/s F(d/s) Resistivity 𝝆 (𝜴 · 𝒄𝒎)

1 Si -type 0.28176101 1 (9.5±1.5)
2 Si n-type 0.28532495 1 (2.02±0.66)
3 Si n-type 0.43018868 1 (2.11±0.34)· 10−2
4 Si p-type 1.25932914 0.85 (1.47±0.21)· 10−3
5 Ge intrinsic 1.3293501 0.82 (19.26±0.92)
Table 2: Calculated values of the d/s quotient, the factor F(d/s) and the resistivity ρ.

To obtain the dopant concentration for each sample, equation (5) was used. For silicon
p-type the hole concentration can be neglected. Analogously, for n-type, the donators
concentration can also be neglected. The reason behind this approach is that in doped SC,
one type charge carrier predominates at room temperature. Table 3 shows the dopant
concentration with its uncertainty and for comparison, the theoretical values found in the
[1]
literature given.

Sample Type Dopant concentration (𝑐𝑚−3 ) Theoretical (𝑐𝑚−3 )

1 Si -type (1.37095±0.00062)· 1015 1.41· 1015
2 Si n-type (2.29752±0.00066)· 1015 2.34· 1015
3 Si n-type (2.18648±0.00036)· 1017 1.20· 1018
4 Si p-type (8.8476±0.0038)· 1018 7.68· 1019
Table 3: Calculated values of the d/s quotient, the factor F(d/s) and the resistivity ρ.

14
 As we can observe in Table 3, our calculated concentrations are quite similar to the
expected ones. Concentration of sample 1 and 2 are nearly equal to the theoretical values.
Nevertheless, for samples 3 and 4 the measures differ in one order of magnitude with the
theoretical values. This can be due to the fact that the device we used for measuring
resistances was old and sometimes it malfunctioned. Another reason could be a possible
contamination of the samples. As the measurement of the resistance is done by contact, if
there are any impurities in the surface, they could affect the value.

For sample 5 the approach is slightly different. Sample 5 corresponds to the intrinsic
germanium, and thus we assume that the concentration of acceptor and donors is the same
(𝑝 = 𝑛). The dopant concentration obtained for sample 5 is presented in Table 4.

Sample Type Dopant concentration (𝑐𝑚−3 )

5 Ge-intrinsic (5.59353±0.00095)· 1013
Table 4: Dopant concentration for sample 5.

For the sample 6, the gallium doped silicon disc, we took 3 measurements of resistance
and of thickness after each layer removal. Resistance and thickness are obtained the same
way as for the other samples. For resistivity we used equation (15) from differential layer
conductivity, and its uncertainty is calculated from addition of relative errors while the
concentration of dopants is calculated as above. These values are all gathered in Table 5.
It should be noted as well that, since we didn’t exert a strongly enough force on the sample
when removing each layer, we end up with 30 series of measurements.

N R (𝛺) d (mm) 𝝆 (𝛺 · 𝑐𝑚) p (𝑐𝑚−3 )

1 (7.79±0.02) (0.765±0.001) (2.702±0.039) (4.8187±0.0008) · 1015
2 (7.86±0.01) (0.768±0.001) (2.734±0.032) (4.7592±0.0007) · 1015
3 (8.54±0.02) (0.759±0.001) (2.939±0.049) (4.4.04±0.0007) · 1015
4 (8.78±0.025) (0.767±0.006) (3.05±0.14) (4.266±0.001) · 1015
5 (12.73±0.08) (0.758±0.004) (4.38±0.23) (2.9760±0.0007) · 1015
6 (11.75±0.01) (0.753±0.004) (4.01±0.10) (3.2475±0.0006) · 1015
7 (14.5367±0.0008) (0.751±0.002) (4.95±0.14) (2.6315±0.0005) · 1015
8 (16.5767±0.0007) (0.745±0.003) (5.59±0.16) (2.3273±0.0004) · 1015
9 (21.1333±0.0007) (0.746±0.006) (7.15±0.34) (1.8223±0.0004) · 1015
10 (25.1333±0.0007) (0.738±0.003) (8.41±0.23) (1.5482±0.0003) · 1015
11 (32.2000±0.0007) (0.737±0.002) (10.76±0.36) (1.2100±0.0002) · 1015
12 (47.3000±0.0007) (0.739±0.008) (15.84±0.93) (8.219±0.002) · 1014
13 (67.3333±0.0007) (0.732±0.004) (22.54±0.64) (5.777±0.001) · 1014
14 (126.6667±0.0007) (0.729±0.002) (41.9±1.5) (3109.75±0.06) · 1011
15 (113.0000±0.0007) (0.734±0.003) (37.6±1.2) (3465.27±0.07) · 1011
16 (136.5667±0.0007) (0.730±0.003) (45.2±1.07) (28881,69±0.05) · 1011

15
 17 (181.9333±0.0007) (0.7303±0.005) (60.2±1.09) (2162.13±0.04) · 1011
18 (167.2333±0.0007) (0.728±0.002) (55.2±2.1) (2358.64±0.05) · 1011
19 (166.3333±0.0007) (0.730±0.002) (55.1±1.2) (2364.91±0.04) · 1011
20 (215.3333±0.0007) (0.727±0.002) (70.9±1.6) (1835.98±0.03) · 1011
21 (294.0000±0.0007) (0.724±0.006) (96.5±5.1) (1349.05±0.03) · 1011
22 (1609.6667±0.0007) (0.714±0.007) (521±28) (250.08±0.06) · 1011
23 (2216.6667±0.0007) (0.718±0.003) (721±19) (180.51±0.03) · 1011
24 (4973.3333±0.0007) (0.708±0.07) (1595±74) (81.628±0.018) · 1011
25 (113.1000±0.0007) (0.702±0.005) (36.0±3.2) (3618.4±1.1) · 1011
26 (127.8667±0.0007) (0.673±0.026) (39.0±7.7) (3336.8±1.8) · 1011
27 (1244.6667±0.0007) (0.691±0.003) (389±15) (334.19±0.07) · 1011
28 (1145.3333±0.0007) (0.682±0.004) (354±24) (367.8±0.1) · 1011
29 (1155.3333±0.0007) (0.681±0.004) (357±18) (364.96±0.09) · 1011
30 (1211.3333±0.0007) (0.672±0.004) (369±16) (353.10±0.08) · 1011
Table 5: Calculated values of the d/s quotient, the factor F(d/s) and the resistivity ρ.

To asses the Ga-diffusion in the silicon disc, we represent the concentration of

atoms as a function of the thickness of the sample in Figure 11.

Fig 11: Concentration of Ga atoms vs. depth.

As expected, the behaviour of the data is exponential. The dots were fitted using the
software Origin. The physical meaning of this is that we will find a greater concentration
of doping atoms near the surfaces of the Ga-diffused silicon and as we go deeper that
concentration will decrease exponentially.

16
 As we end the area with the diffused Ga atoms, we arrive to the original silicon sample.
This silicon has a homogeneous natural doping of phosphor atoms (n-doped). The value of
that donator concentration can be obtained as an average of the last values we took in the
laboratory, as they are nearly constant, which means that we were already in the original
silicon zone. The uncertainty was given by error propagation.

𝑝 = (3.6195 ± 0.0001) · 1013 𝑐𝑚−3

4.2 Solar cell

In this second part, we measured the voltage and current for the polycrystalline and
tandem cells in different configurations. With the recorded data, we are able to plot the
Current-Voltage graphs for each setting. These graphs are quite important because from
them very meaningful parameters such as the fill factor 𝐹𝐹 or the maximum power 𝑃𝑚𝑎𝑥
can be obtained, which are critical in determining the device’s output performance and
solar efficiency.

Figures from 12 to 19 show polycrystalline and tandem characteristics’ graphs, each

of them fitted using an exponential function in Origin.

Fig 12: Polycrystalline cell without light.

17
Fig 13: Polycrystalline cell with solar spectrum at 20°C.

Fig 14: Polycrystalline cell with solar spectrum at 35°C.

18
Fig 15: Polycrystalline cell with solar spectrum at 50°C.

Fig 16: Tandem cell without light.

19
 Fig 17: Tandem cell with solar spectrum at 20°C.

Fig 18: Tandem cell with filtered solar spectrum at 20°C.

20
 Fig 19: Tandem cell with solar spectrum at 50°C.

The values for 𝑈𝑜𝑐 and 𝐼𝑠𝑐 are obtained as figure 6 shows. 𝐼𝑠𝑐 is the point where the
characteristic crosses the Y axis (x=0), and 𝑈𝑜𝑐 where it crosses the X axis (y=0).
Therefore, we obtained the values gathered in Table 6 from our data and from its fittings.

Polycrystalline
Configuration 𝐼𝑠𝑐 (𝐴) 𝑈𝑜𝑐 (𝑉)
Without light -0.0001 0.0843
20°C -0.151 0.6273
35°C -0.149 0.6111
50°C -0.147 0.6136
Tandem cell
Configuration 𝐼𝑠𝑐 (𝑚𝐴) 𝑈𝑜𝑐 (𝑉)
Without light -0.015 0.5003
20°C -7.90 0.8760
20°C filtered -0.94 0.7052
50°C -5.42 0.7761
Table 6: Short-circuit current and open circuit voltages values for both cells.

It is known that solar cells are sensitive to temperature. An increase in the temperature
reduces the band gap of a semiconductor, thereby affecting most of the semiconductor
material parameters. The decrease in the band gap of a semiconductor with increasing
temperature can be viewed as increasing the energy of the electrons in the material.

21
 The band gap of the semiconductor constituting the photovoltaic cell determines what
portion of the solar spectrum it absorbs. A semiconductor will not absorb photons of energy
less than the band gap, and the energy of the electron-hole pair produced by a photon is
[8][9]
equal to the bandgap energy.

In a solar cell, the parameter most affected by an increase in temperature is the open-
circuit voltage. The open-circuit voltage decreases with temperature because of the
temperature dependence of the saturation current 𝐼0 . In Figures 13-15, 17 and 19, we can
barely notice that difference graphically, but if we look closely to our numerical values in
Table 6, we can mainly see a decrease in 𝑈𝑜𝑐 with temperature for both cells, except for
the polycrystalline at 35°C which is slightly bigger than for the 50°C. This could be due to
some errors while fitting the curve.

The short-circuit current 𝐼𝑠𝑐 increases slightly with temperature, since the band gap
energy decreases and more photons have enough energy to create electron-holes pairs.
However, this is a small effect and the temperature dependence of the short-circuit current
from a silicon solar cell is approximately 1/𝐼𝑠𝑐 · 𝑑𝐼𝑠𝑐 /𝑑𝑇 ≈ 0.0006 per °C for Si. That
might be the reason why we do not appreciate that growth in our experimental values.

The Maximum Power Points (MPP) were chosen from the 4th quadrant of the
different characteristics as explained in the theoretical fundamentals. It should be
highlighted that for the polycrystalline cell, despite of not affecting the value of the FF,
we had 2 points (in 35°C and 50°C plots) with the same maximum power but different
values of current and voltage.

Moreover, for the tandem cell in the configuration without light, there are no points
in the 4th quadrant so an MPP or either its corresponding 𝐼𝑚𝑝𝑝 and 𝑈𝑚𝑝𝑝 cannot be found.

Polycrystalline
Configuration 𝐼𝑚𝑝𝑝 (𝐴) 𝑈𝑚𝑝𝑝 (𝑉) 𝑃𝑚𝑎𝑥 (𝑊)
Without light -0.000016 0.10 -0.0000016
20°C -0.108 0.31 -0.03348
35°C -0.099 0.32 -0.03168
35°C -0.096 0.33 -0.03168
50°C -0.096 0.31 -0.02976
50°C -0.93 0.32 -0.02976
Tandem cell
Configuration 𝐼𝑚𝑝𝑝 (𝑚𝐴) 𝑈𝑚𝑝𝑝 (𝑉) 𝑃𝑚𝑎𝑥 (𝑚𝑊)
Without light - - -
20°C -6.84 0.64 -4.3776
20°C filtered -0.70 0.54 -0.3780
50°C -4.24 0.60 -2.5440
Table 7: Maximum, power values for both cells.

22
 The Fill Factor was calculated with the short circuit, open circuit and maximum power
values, as explained in the theoretical fundamentals and using equation (14). Its values are
framed in Table 8.

Polycrystalline
Configuration 𝐹𝐹 %
Without light 0.0190 1.90
20°C 0.3535 35.35
35°C 0.3479 34.79
50°C 0.3299 32.99
Tandem cell
Configuration 𝐹𝐹 %
Without light - -
20°C 0.6326 63.26
20°C filtered 0.5702 57.03
50°C 0.6048 60.48
Table 8: Fill Factor values

In 2009, typical commercial solar cells had a fill factor > 0.70. Grade B cells were
usually between 0.4 and 0.7. Our measured values for the Tandem cell are in this range,
but the Polycrystalline ones are slightly below 0.4. Cells with a high fill factor have a low
equivalent series resistance and a high equivalent shunt resistance, so less of the current
[10]
produced by the cell is dissipated in internal losses.

5. Conclusion
The aim of this experience was to study the properties of solar cells. In order to do
that, they were tested along with analyzing some semiconductors properties, as they are
the main component of photovoltaic devices.

First of all, for different semiconductor samples, such as n-type and p-type Silicon or
intrinsic Germanium, their thickness and resistance were measured. The data collected
allowed to determine for their charge carrier concentrations, taking into account some
correction factors. Although in some cases the experimental concentrations differed slightly
from the theoretical values, they were accurate enough to be considered correct.

Secondly, the dependence of the concentration of charge carriers with the depth for a
gallium-diffused silicon sample was observed, which for that kind of diffusion processes,
resulted in being exponential; and also, its natural doping was found.

After our analysis on semiconductors, the characteristic lines of a polycrystalline and

a tandem solar cell were investigated with different configurations. As seen in the results,
the temperature plays an important role on the performance of a solar cell, due to the fact
that it also affects semiconductor properties and therefore it has an important impact on

23
the parameters that have consequently an influence in the efficiency (𝐼𝑠𝑐 , 𝑈𝑜𝑐 y MPP).
The dependence of these values with temperature were observed and finally concuded that
the efficiency of the tandem cell, is greater than the polycrystalline one, and also that for
both cells the performance decreases with an arise of temperature, and so does the
efficiency.

6. References
[1] MP1 Anleitung, Institut für Material Physik.
[2] “Diffusion | PVEducation.” Pveducation.Org, 1 Jan. 2019, Accessed 26 Jan. 2019,
https://www.pveducation.org/pvcdrom/pn-junctions/diffusion.
[3] “Multi-Junction Solar Cell ▪ The Complete Information and Online Sale with Free
Shipping.” Maria-Online.Com, 1 Jan. 2019, Accessed 26 Jan. 2019, https://www.maria-
online.com/electronics/article.php?lg=en&q=Multi-junction_solar_cell.
[4] Shaik, Asif. “Drift Current.” Physics-and-Radio-Electronics.Com, 1 Jan. 2013, Accessed
26 Jan. 2019 https://www.physics-and-radio-electronics.com/electronic-devices-and-
circuits/semiconductor/drift-current.html.
[5] “Diffusion | PVEducation.” Pveducation.Org, 1 Jan. 2019, Accessed 26 Jan. 2019
https://www.pveducation.org/pvcdrom/pn-junctions/diffusion.
[6] Progress in Organic Photovoltaic Fibers Research - Scientific Figure on ResearchGate.
Available from: https://www.researchgate.net/figure/Current-voltage-I-V-curves-of-an-
organic-solar-cell-dark---illuminated-The_fig3_221918802 [accessed 27 Jan, 2019]
[7] “What Is the Advantage of the Four Probe Method over Others to Find Resistivity? -
Quora.” Quora.Com, 1 Jan. 2014, https://www.quora.com/What-is-the-advantage-of-the-
four-probe-method-over-others-to-find-resistivity. Accessed 26 Jan. 2019.
[8] Wikipedia Contributors. “Band Gap.” Wikipedia, Wikimedia Foundation, 10 Dec. 2018,
https://en.wikipedia.org/wiki/Band_gap. Accessed 26 Jan. 2019.
[9] “Effect of Temperature | PVEducation.” Pveducation.Org, 1 Jan. 2019,
https://www.pveducation.org/pvcdrom/solar-cell-operation/effect-of-temperature.
Accessed 26 Jan. 2019
[10] “Solar Cell Efficiency.” Wikipedia, Wikimedia Foundation, 9 Jan. 2019,
https://en.wikipedia.org/wiki/Solar_cell_efficiency. Accessed 26 Jan. 2019.

24