Advanced Lab Course 50:

Photovoltaics
Team 122 :
Ahmed Krifa 03698507, Hamza Ouerfelli 03698446, Pascal Sedlmeier 03693939
Technical University of Munich
5 July 2019

Contents
1 Introduction 2

2 Basics 2
2.1 Reflection measurement . . . . . . . . . . . . . . . . . . . . . 2
2.2 Voltage-current characteristic of solar cells . . . . . . . . . . . 2
2.2.1 Derivation of the formula . . . . . . . . . . . . . . . . . 2
2.2.2 Other important parameters . . . . . . . . . . . . . . . 3
2.3 Quantum efficiency . . . . . . . . . . . . . . . . . . . . . . . . 4

3 Results/evaluation 5
3.1 Reflection measurement . . . . . . . . . . . . . . . . . . . . . 5
3.2 Voltage-Current characteristics . . . . . . . . . . . . . . . . . 6
3.3 Quantum efficiency . . . . . . . . . . . . . . . . . . . . . . . . 8

4 Questions 8
4.1 Measurement of the reflexion . . . . . . . . . . . . . . . . . . . 8
4.2 Current-voltage curve . . . . . . . . . . . . . . . . . . . . . . . 9
4.3 Lock-in amplifier . . . . . . . . . . . . . . . . . . . . . . . . . 10
4.4 Spectrally resolved quantum efficiency . . . . . . . . . . . . . 10

5 Appendix 11

1
1 Introduction
A solar cell, or photovoltaic cell, is an electrical device which directly converts
the energy of light into electricity by the photovoltaic effect. In this paper,
the essential properties of different types of solar cells will be investigated
and characterized.

2 Basics
2.1 Reflection measurement
Reflection is when light bounces off an object. If a light beam of intensity I0
propagating in air (with a refractive index n0 ) strikes a medium with different
refractive index n1 , only a part of the beam can penetrate into the body and
another part is reflected:

I(hν) = I0 (1 − R(hν)) (1)

I(hν) corresponds to the intensity of the beam coupled into the solid, R(hν)
describes the fraction of the reflected intensity dependent on the refractive
indices:
n0 − n1 2
R(hν) = ( ) (2)
n0 + n1

2.2 Voltage-current characteristic of solar cells

The voltage-current dependency is elementary in order to make statements
about the efficiency, and indirectly about the age or the quality of a solar cell.

2.2.1 Derivation of the formula

Theoretically, this relation can be derived by means of the energy-dependent
photocurrent IL , which impinges on the solar cell. For this purpose, the
product of the number of charge carriers generated per incident photon, also
dn
called quantum yield, and the energy-dependent photon density dEph (E = hf )
is integrated over the entire energy range:
Z
dnph
IL = e QE(E) dE (3)
dE

2
Combined with the general current equations of a pn-transition
eU
I = Is (exp( ) − 1) (4)
Akb T
A is the diode factor (A = 1 or 2, depending on the recombination strength
within the space charge zone), T the absolute temperature and IS the reverse
current :
Lp Ln
Is = e(pn0 + np0 ) (5)
Tp Tn
where L represents the diffusion length and τ the lifetime of the holes/electrons.
Therefore, it applies for an illuminated solar cell:
eU
I = Is (exp( ) − 1) − IL
Akb T
(6)

The formula clearly shows that the current generated by the solar cell flows
contrary to the typical forward direction of a diode (cf. sign of IS and IL ).

2.2.2 Other important parameters

Important characteristics can be extracted from extreme cases of Equation
(6): The short-circuit current ISC (SC for ’short circuit’), which results for U
= 0 V and therefore ISC = −IL and the open circuit voltage UOC (OC for
’open circuit’) for I = 0 A.
Assuming that IL is much larger than IS and that IL ≈ ISC , the open circuit
voltage can be described as follows:
Akb T ISC
UOC = ln( ) (7)
q IS

The maximum power Pm (see Fig. 1), indicated as MPP (maximum power
point) can be calculated with the two quantities Um and Im .
The now calculable full factor FF is a good measure of the cell and shown
graphically in Figure 1. As a consequence of the current-voltage characteristic,

3
only a part (blue) of the theoretical maximum power (yellow) is reached (the
yellow area includes the blue one).
Um Im
FF = (8)
UOC ISC

Figure 1: Graphical representation of the full factor [wikipedia]

The efficiency is then given by:

Pm UOC ISC
η= = F F. (9)
PLicht PLicht
For further electrotechnical considerations, the series and parallel resistance
(Rs and Rp ) of the solar cell can also be determined by means of the U − I
characteristic. For this purpose, the derivative of the characteristic curve is
formed at I = 0 or U = 0, correspondingly :
∂U ∂U
Rs = |I=0 , Rp = |U =0 (10)
∂I ∂I

2.3 Quantum efficiency

The so-called quantum efficiency is crucial to evaluate the efficiency of pho-
tovoltaic devices. It corresponds to the ratio of the released charge carriers

4
to the incident photons. This can be expressed by means of the measured
current I, the charge of a single charge carrier q, the wavelength of the incident
radiation λ and the radiation power φL (λ):
Ne h.c.I
QE(λ) = = (11)
Nν (λ) q.λ.φL (λ)

3 Results/evaluation
3.1 Reflection measurement
The characteristic curves recorded by the UV / VIS spectrometer of the differ-
ent types of solar cells can be seen in figures 2a to 2f. Except for amorphous
silicon, a sharp drop of reflection component is observed for all samples at a
wavelength of approximately 1100 nm. The reason for this is the band gap of
silicon which is 1.12 eV at 300 K, which corresponds to a wavelength of 1107
nm.
Due to the disordered crystal structure of amorphous silicon, unbound elec-
trons can be found in it that form localized states in the band gap. Thus
there is no longer a clear drop at a specific wavelength 2a.

With formula 2.2 it follows for the refractive index of crystalline silicon:
√
1± R
n1 = n0 · √ (12)
1∓ R
Consider only the solution where n1 > 1, the refractive index for the smooth
side of the crystalline silicon is as shown in figure 3.
For long wavelengths up to approximately 1200 nm there is a constant refrac-
tive index of n1 = 5.432 ± 0.013. For falling wavelengths, n1 drops sharply
because the absorption increases sharply due to the band gap. The refractive
index can be described in the range from 1050 nm to 1200 nm by a straight
line with n1 = (0.0118 ± 0.0002)λ − (8.55 ± 0.25). For smaller wavelengths,
n1 is initially constant between 550nm and 1050nm at n1 = 3.82 and then
increases exponentially. This can be justified by the fact that electrons that
are excited by very high-energy radiation relax to the upper edge of the band
gap and thereby emit part of the absorbed energy again as radiation.
The thickness of the amorphous silicon can be determined using the reflection

5
measurement in figure 2a, because the light rays are reflected both when
entering the silicon cell and when exiting and so the thickness can be deter-
mined as a result of this interference.
The difference in the optical path lengths of the two beams is:
1
∆L = n1 · 2d − λ (13)
2
the minuend represents the additional path of the beam in the plate and the
subtrahend takes into account the phase shift of the beam reflected directly
on the surface.
If this difference is equal to 12 kλ, destructive interference is observed for odd
k and constructive interference for even k, which are the minima and maxima
in figure 2a. Since the difference ∆k between neighboring extreme points is
one, d is equal to the slope of figure 4, where k was chosen so that it is zero
at the first extreme point. Thus d = (2542.87 ± 108.22)nm.

3.2 Voltage-Current characteristics

Before we can sketch the Voltage-Current characteristics, we have to mea-
sure the luminance of the halogen lamp which is used. After the working
temperature is reached, the power of the lamp is set on 100 mW/cm2 , what
corresponds to a 90◦ inclination of the sun. For the amorph solar cell a
range betweeen -0,5 and 1,5 V is measured and a range between -0,5 and
0,65 V for the cristalline solar cell, each time for the dark and the illuminate
configuration. Theses graphs are shown in Figures 5, 6, 7 and 8.
We can observe a shift of the illuminated characteristic in negative y-direction
(compare with Figure 7 and Figure 8). This shift describes the photocurrent
in equation (6). We can also see a saturation for the cristalline solar-cell
(compare with Figures 6 and 8) at approximately 0, 55 V. With the Figures
5 − 8 we can determine the open circuit voltage and the short circuit current.
The open circuit voltage UOC is the voltage of the characteristic for I = 0
A and the short circuit current ISC is the current of the characteristic for
U = 0 V. The values of UOC and ISC in the different configurations are given
in table 1.
In order to calculate the fill factor FF we need the current Im and the voltage
Um at the working point (WP). By definition the power of the solar cell
at the WP has to be maximal, so we are looking for the greatest product
Pm = Um · Im . The algebraic sign of the current and the power is negative,

6
 Table 1: Values of UOC and ISC for the different cells

UOC ISC
amorph unbeleuchtet 0.23 V 4.19 nA
amorph beleuchtet 0.83 V 0.83 mA
cristalline unbeleuchtet 5 mV 0.7 nA
cristalline beleuchtet 0.5 V 6.38 mA

Table 2: Values of voltage, current and power at the working point for the
different cells

Um Im Pm
amorph 0.64 V 0.68 mA 0.44 mW
cristalline 0.39 V 5.45 mA 2.125 mW

which shows that power is delivered. In table 2 however we get rid of the sign
for practical purposes.
To determine the efficiency we multiply the set irradiance at AM1: 90◦ ,
Plight
A
= 100mW
cm2
with the area of the illumination surface (circular with radius
2mm) and we get: Plight = 12, 56mW . Calculating now the ratio PPlightm
for the
amorph and cristalline solar cell we get ηa = 3, 5% and ηc = 16, 9% for the
efficiency of the amorph and the cristalline solar cell. If we compare these
values to reference values (amorph 7-10 % and cristalline 17-24 %), we see
that the cells we used were quite old, since the efficiency is lower.

Using the U-I characteristics and equation (10) we can determine the re-
sistances in series and in parallel. These are given in table 3.

Table 3: Values of the resistances in series and in parallel for the different
cells

Rs Rp
amorph unbeleuchtet 9.1 µΩ 9.1 µΩ
amorph beleuchtet 0.14 mΩ 13 mΩ
cristalline unbeleuchtet 5.5 mΩ 5.5 mΩ
cristalline beleuchtet 0.6 µΩ 8.6 µΩ

7
3.3 Quantum efficiency
From the measured voltages depending on the wavelength we can determine
the quantum efficiency of the cristalline and the amorph solar cell using
equation (14).
h · c · USZ
QE(λ) = 1W
(14)
q · λ · R · UP yro · 2878V
We get Figure 9 for the quantum efficiencies of the amorph and the cristalline
solar cell as a function of the wavelenght.

4 Questions
4.1 Measurement of the reflexion
• Why do we measure a smaller lost of reflexion for the rough
silicium wafer-side than for the smooth side ?

We measure a smaller lost of reflexion for the rough side, because

a fraction of the reflected light can fall back again on the wafer thanks
to the rough surface of the cell.
• Why do the cristalline silicium-samples show a local maximum
at ca. 366 nm ?

Together with the indirect bandgap of silicium at 1,1 eV there exists a

direct bandgap at 3,4 eV. The energy of 3,4 eV is exactly the energy of
a photon with a wavelengh of 366 nm. Light of this energy has therefore
the possibility to stimulate electrons several time. This phenomenon
leads to a local maximum at ca. 366 nm in the cristalline silicium-
samples.
• Why is the cristalline silicium solar-cell so black ?

The cristalline solar-cell appears darker due to an anti-reflective layer.

The thickness of the layer is responsible for a specific interference color.
• How can the reflexion be further reduced ?

8
 One way to reduce reflexion is to work on the surface. If we consider
black silicon for instance we get a material with very low reflectivity.
Black silicon is a needle-shaped surface structure. The needles are made
of single-crystal silicon and have a height above 10 µm and diameter
less than 1 µm .

• Why are the measurements at large wavelenghts so noisy ?

When we increase the wavelenght, the thermal noise of the whole

equipment also increases.

4.2 Current-voltage curve

• How can one calculate the resistance in series or in parallele
from the unilluminated curve ?

Thanks to the current-voltage curve it is possible to identify the resis-

tance in series and in parallele. Following formelas hold :
∂U
Rs = |I=0 (15)
∂I

∂U
Rp = |U =0 (16)
∂I
As the leckage current of the cell has been neglected, these formulas are
only approximations.

• The manufacturer of the cristalline solar-cell gives for the re-

sistance in parallel Rp more than 105 Ω. Why is the measured
value very smaller ?

The main reason for this difference is due to the degradation of the cell
with time. As the cell is certainly several years old, there are leckage
currents at the pn-junction and imperfections such that the measured
resistance is smaller.

9
• What are the most important sources of error to determine
the efficiency ?

The efficiency being calculated from the light power and the electrical
power of the cell, it is logical that the errors of both powers affect the
error of the efficiency. As we can neglect the resistance on the supply
to the shunt, the photo current has only the relative uncertainty of the
shunt’s resistance. Other important sources of error are the adjustment
of the halogen lamp and the determination of the light power by the
photodiode.

• Why are the off-load voltages and the short-circuit currents

of the different solar-cells at constant illumination so different
?

The two solar-cells are made mostly of silicium, but the difference can
be seen particularly at the band structure. Cristalline silicium has the
band gap Eg at ca. 1,1 eV. For amorph silicium the band gap is at
ca. 1,7 eV. If no photo current flows, then the no-load voltage is equal
to the potential difference, which appears at the pn-junction. We can
write : eUOC ≈ Eg .

4.3 Lock-in amplifier

• Assume the output voltage U is maximal and you change the
phase position now of 90 or 180 degrees. Which output volt-
ages do you expect ?
Since the output voltage U is linear with the cosines function, a 90
degree change in the phase position will then result into a zero output
voltage and a 180 degree into a −U output voltage.

4.4 Spectrally resolved quantum efficiency

• Why does the quantum efficiency of the different cells at 300
nm drop to zero (different reasons for the different cells) ?

10
 The amorph silicium used is located into the cell behind a glass pane,
which is almost non-permeable for UV-light. When we compare it
to the cristalline solar-cell, we observe that this cell allows almost a
complete reflexion of UV-light (300 nm) thanks to the thickness of the
anti-reflective layer. If we get interested in smaller wavelenghts, then
we note that the efficiency is better. Photons of higher energy cannot
be completely absorbed.

• Is there a correlation between the dropping of the quantum

efficiency close to the band edge and the type of the bandgap
of the absorbing material ?

Yes, we can observe further steps in the quantum efficiency due to

additional valence band edges.

5 Appendix
[1] TU Munich, Lab Course Manual for Capacitive properties of a gold/electrolyte
interface
https://www.ph.tum.de/academics/org/labs/fopra/docs/userguide-09.en.pdf

11
 (a) Amorphous silicon (b) Antireflective solar cell

(c) black silicon (d) Crystalline silicon-smooth side

(e) Crystalline silicon-rough side (f) Pyramidal solar cell

Figure 2: Reflection component different types of solar cells

12
Figure 3: Refractive index of crystalline silicon depending on the wavelength

13
Figure 4: Order of the extreme point as a function of the inverse wavelength
multiplied by 2n1

14
Figure 5: U-I characteristic for the dark amorph solar cell

15
Figure 6: U-I characteristic for the dark cristalline solar cell

16
Figure 7: U-I characteristic for the illuminated amorph solar cell

17
Figure 8: U-I characteristic for the illuminated cristalline solar cell

(a) amorphous solar cell (b) crystalline solar cell

Figure 9: Quantum efficiency depending on the wavelength

18