·Home ·Table of Contents ·General

Temperature dependence of 1/F noise and transport characteristics as a non-destructive testing of monocrystalline silicon solar cells A.Ibrahim and Z.Chobola Technical University of Brno, Faculty of Civil Engineering, Physics department, Zizkova 17, 602 00 Brno, CZ., Fax: +420541147666, Tel:+420541147659. E-Mail: ali@dp.fce.vutbr.cz Contact

ABSTRACT

Dependence of the noise spectral voltage density S_V(f) in the frequency range 1Hz to 10⁵ Hz and transport characteristic for a monocrystalline silicon solar cells have been investigated. The magnitude of the noise spectra for the Si solar cell shows a decrease of noise magnitude with increasing temperature between 300K to 400K. Also for I-V curves, both recombination-generation and diffusion current components are increases with temperature.

INTRODUCTION

The monocrystalline silicon solar cells of large area (10x10 cm²) with structure n⁺pp⁺⁺ fabricated on p-type Czochralski Cz silicon wafers of efficiency 14% and more are important in terrestrial applications. As a result of this, the measurements of the both transport characteristics (I-V) and low frequency noise at a temperature range from 300K to 400K are so important. The magnitude of this noise is a sensitive function of the technology and is a measure of the quality and reliability of the solar cell under test, where by determining the source of noise ,i.e, the defect on the construction, an improvement of the product can be obtained by eliminating the imperfections.
Usually the solar cell junction device degradation and operation failure result generally from uncontrolled technology procedures or steps in the semiconductors crystal growth, junction formation, surface treatment and packaging. These can produce imperfections such as additional energy states in the forbidden gap, impurity distribution irregularities can degenerate local overtopped microregions in the junction, clusters of the impurity atoms, dislocations, deviation from the desired junction geometry,-----,etc. Some of the imperfections bring about additional charge carriers transport "excess current". Also these imperfections act as sources of noise. The excess noise is relatively much higher than the corresponding excess current. Also these imperfections can initiate irreversible processes, which are usually processes of the device degradation [1-4].
For low frequency noise in semiconductors there are some physical which generate it such as, generation-recombination, trap related diffusion, mobility fluctuation, and carrier number fluctuation. In this paper the measurements of the I-V curves and S_V(f) for the Si solar cells at different temperatures from 300K to 400Kare presented.

EXPERIMENTAL SET-UP

The noise measurements were made using the system shown in Fig.1. The Selective Nanovoltmeter type 233(UNIPAN) has an equivalent noise referred to its input by connecting to a preamplifier type 233-7(-20dB) and the Si solar cell sample is connected with the UNIPAN through an electronic circuit by using special holder (suitable for increasing temperature). The UNIPAN type 233 is fixed at 10³Hz and the sample is biased by stabilized d.c power source to obtain 10 mA through the sample. The sample under test, low noise preamplifier 233-7 and the bias voltage source all are operated in a shielded suitable box from lead for protection . A computer was used to control the measurements and also has the ability to do some data manipulation and plotting

Fig 1: Block diagram of the noise voltage measurements under temperature

Fig 2: I-V-T characteristics of a Si solar cell of structure n+pp++ of area 100 cm2

EXPERIMENTAL RESULTS AND DISCUSSIONS

(A) TEMPERATURE DEPENDENCE OF THE DARK I-V CHARACTERISTICS
The temperature dependence of the dark I-V characteristics is often useful indicator of the principal mechanism responsible for current conduction in a device. Over the range of bias, where the Si solar cells operate in the photovoltaic modes, the mechanism of the I-V characteristics should be diffusion, the same process as in a Shockley p-n junctions. Measurements on a monocrystalline Si solar cell of n⁺pp⁺⁺ structure using a 0.7 W.cm p-type silicon wafer over the 300K to 400Kare shown in Fig.2. The experimental data can be described to a high degree of precision by an equations of the form[5]:

(1)

(2)

where D_p and D_n are the diffusion coefficients for the holes and electrons, respectively. W is the width of the depletion layer and t_d is the diffusion lifetime of the charge carriers in the junction and finally, L_n and L_p are the diffusion lengths for electrons and holes respectively. For logarithmic segments at high voltage end of each characteristics prior to the series resistance region, the n₁ close values are close to be 1 indicating a diffusion process. The low voltage portions have n₂ close to 2 indicating recombination in the depletion layer of the junction, as given by the second term in Eq.(1).

The most striking feature of these curves is that they are parallel straight lines over nearly three decades of the current, with a positive temperature coefficient over the whole range of bias voltage.

(B) 1/F NOISE MEASUREMENTS WITH TEMPERATURE
Hooge and Hoppenbrouwers[6] have measured the 1/f noise voltage generated in continuous gold films (with physical properties close to bulk values) in the presence of a steady current. They found that the noise power spectrum S_V(f) for samples at room temperature could be expressed by the empirical formula:

(3)

where N is the number of the charge carriers in the sample, f is the frequency, and V is the voltage across the sample. And the 1/f noise is often considered as arising from resistance fluctuations that generate a fluctuation voltage in the presence of a steady current. Equation(3) is valid also for semiconductors, where 1/f noise in semiconductors is a bulk and surface effects[7].
The 1/f noise in metal films could be caused by temperature fluctuations that modulate sample resistance R and generate voltage fluctuations in the presence of a steady current I [8]. Thus:

(4)

<(DT)²> is the mean-square temperature fluctuation. At thermal equilibrium where C_v is the heat capacity of the sample. At room temperature, C_v=3NK, and S_v(f)aV²b²T²/3N. The temperature fluctuation DT in the sample of a resistance R and temperature coefficient of resistivity will be observed as a voltage fluctuation DV = IRbDT in the presence of a constant current I. The voltage fluctuation spectrum S_V(f) is then related to the temperature fluctuation spectrum S_T(f) by S_v(f) = V²b²S_T(f). If the temperature fluctuations are due to equilibrium exchange of energy between the sample and its environment, then in this case Dependencies of the S_V(f) with frequency for a Si solar cell of a n⁺pp⁺⁺ structure at a temperatures ranges from 300K to 400K are shown in Fig.3.
The sample is under a forward bias voltages of values 200 mV, 150 mV and 100 mV for 300K, 350K and 400K, respectively. As the temperature increases S_V(f) decreases and S_V(f) is proportional to f^-1. The equilibrium temperature fluctuations modulating the resistance of the sample and this is the physical origin of the 1/f noise under temperature effect or at thermal equilibrium. [9]

The dependence of the noise voltage with 1/T at 10³ Hz of the UNIPAN measuring unit and the current through the sample is fixed at 8 mA is shown in Fig.4. The temperature fluctuations obey the diffusion equation. The activation energy from this Arrhenius plot in forward direction is 0.44 eV for the Si solar cell.

Fig 3: The SV(f) vs frequency for a Si solar cell at of structure n+pp++ in a temperature range of 300-400K.

Fig 4: Noise voltage as a function of 1/T for the Si solar cell of structure n+pp++.

The results of the voltage noise for the Si solar cells of a structure n⁺pp⁺⁺ in the temperature range from 25-125 ⁰C are useful for calculating the activation energy of the sample under test and also can be used as a tool for evaluating the quality of the product of technology.

CONCLUSIONS

The experimental results for both I-V and 1/f noise spectroscopy in this study for a monocrystalline Si solar cells with a n⁺pp⁺⁺ structure were investigated, and the following conclusions were obtained:
1. in a small applied voltage the recombination-generation current flows through the solar cells,
2. by increasing the applied voltage the diffusion current dominates and
3. the spectral voltage density decreases with increasing temperature of the cell as a result of equilibrium resistance fluctuations.

REFERENCES

1. R.Kolarova. and P.Vasina, 5^th Electronic Devices and Systems, EDS'98 Int. Conf. June 11-12, in Brno, CZ.,P.177,1998.
2. A.Ibrahim, and Z.Chobola, 5^th Electronic Devices and Systems Int. Conf. EDS'98,June 11-12, in Brno, CZ.,P.105,1998.
3. P.Hurska, and R.Kolarova,5^th Electronic Devices and Systems, EDS'98 Int. Conf..June 11-12, in Brno, CZ.,P.173,1998.
4. A.Ibrahim and Z.Chobola, Int. Workshop 97, in Prague,20-22 January, CZ.,P.63,1997.
5. J.Shewchun, D.Burk, and M.B.Spitzer, IEEE Tran.Electron Devices, Vol.ED-27, No.4, P.705,1980.
6. F.N.Hooge, and A.M.H.Hoppenbrouwers, Physica 45, P.386, 1969.
7. F.N.Hooge, Phys. Lett. A29, P.139, 1969.
8. F.R.Voss, and J. Clarcke, Physical Rev. B, Vol.13(2), No.2,P.556,1976.
9. Pazdera L. Brablec A. - Non-destructive Testing of Justicion in Material by Noise Diagnostics, In: International conference on advanced materials and technologies, Plzen, June13-16,1995

© AIPnD , created by NDT.net

|Home|    |Top|