Obtention of solar cell parameters, through convergence of iterative cycles. Part 1: Theoretical analysis and cycles proposal.

In this Part 1 of this series of articles, two iterative cycles are proposed to accurately determine the shunt resistance ( R s h ), the series resistance ( R s ), the ideality factor (n), the light current ( I lig ), and the saturation current ( I sat ) of solar cells, within the one diode model. First, R s and n are obtained linearly fitting ∂ V ∂ l n I ' vs. I ' , where I ' is a new defined current I ' = I + I s a t + I l i g - V - I R s R s h + n k T R s h . Then, R s h and I sat are obtained using Procedure A and B proposed in [2]. Once these four solar cell parameters are obtained, a correction to I lig is deduced and applied. The deduction of these five solar cell parameters is reused to recalculate I ' and the iterative cycles are redone till some convergence criteria is achieved. The accuracy and number of cycles necessary to achieve reasonable results are tested and discussed on ideal (noiseless) current voltage (IV) curves with measured points per voltage of P V = 11 , 21, 51 and 101 measured points V . These two cycles are compared with two different common parameter extraction methods. The results given in this Part 1 are used in Part 2 to calculate the five solar cell parameters of IV curves found in the literature.

Obtention of solar cell parameters, through convergence of iterative cycles. Part 2: Application to experimental current-voltage measurements.

In this Part 2 of this series of articles, the application of the iterative cycles CycleA and CycleB proposed in Part 1, to determine the solar cell parameters (the shunt resistance ( R s h ), the series resistance ( R s ), the ideality factor (n), the light current ( I lig ), and the saturation current ( I sat )) on experimental current voltage (IV) and current density (JV) curves, is given. Several number of measured points per voltage ( P V ) are attempted, from approximately P V = 1 m e a s u r e d p o i n t s V to P V = 52 m e a s u r e d p o i n t s V . In one case, the application of the iterative cycles to IV curves showing the roll-over effect is discussed, while in another case, their application to solar panels is analysed, revealing that the iterative cycles can also be used in the case of solar panels, and not only for laboratory-made solar cells, in voltage ranges larger than [0 V,1 V]. Also, cases in darkness and under illumination are evaluated. In most cases, reasonable values are obtained for R s h , R s , n , I sat and I lig , which simulated properly the IV or JV curves.

Study of the heavily p-type doping of cubic GaN with Mg.

We have studied the Mg doping of cubic GaN grown by plasma-assisted Molecular Beam Epitaxy (PA-MBE) over GaAs (001) substrates. In particular, we concentrated on conditions to obtain heavy p-type doping to achieve low resistance films which can be used in bipolar devices. We simulated the Mg-doped GaN transport properties by density functional theory (DFT) to compare with the experimental data. Mg-doped GaN cubic epitaxial layers grown under optimized conditions show a free hole carrier concentration with a maximum value of 6 × 1019 cm-3 and mobility of 3 cm2/Vs. Deep level transient spectroscopy shows the presence of a trap with an activation energy of 114 meV presumably associated with nitrogen vacancies, which could be the cause for the observed self-compensation behavior in heavily Mg-doped GaN involving Mg-VN complexes. Furthermore, valence band analysis by X-ray photoelectron spectroscopy and photoluminescence spectroscopy revealed an Mg ionization energy of about 100 meV, which agrees quite well with the value of 99.6 meV obtained by DFT. Our results show that the cubic phase is a suitable alternative to generate a high free hole carrier concentration for GaN.

Valence band offset at the Si/SiSn interface by applying deep level transient spectroscopy.

A set of Si1-x Sn x /Si(001) quantum wells (QWs) is grown by applying molecular beam epitaxy. The activation energies of holes in these QWs are studied by deep-level transient spectroscopy. It is observed that the holes activation energies increase monotonically with the Sn fraction (x). The valence band offset between pseudomorphic Si1-x Sn x and Si obeys the dependence ΔE(v) = 1.69x eV, while the offset between the average valence bands of unstrained Si1-x Sn x /Si heterojunction was deduced and obeys the dependence ΔE(v(av)) = 1.27x eV.

Sb-mediated Ge quantum dots in Ti-oxide-Si diode: negative differential capacitance.

The negative differential capacitance (NDC) effect is observed on a titanium-oxide-silicon structure, formed on n-type silicon with embedded germanium quantum dots (QDs). The Ge QDs were grown by an Sb-mediated technique. The NDC effect was observed for temperatures below 200 K. We found that approximately six to eight electrons can be trapped in the valence band states of Ge QDs. We explain the NDC effect in terms of the emission of electrons from valence band states in the very narrow QD layer under reverse bias.

Note: Effects of several thermal glues used on temperature dependent Hall measurements.

The effects of four thermal glues (cry-con, fixogum, RS 503-357, and silicon-high vacuum-grease from Leybold vacuum) on temperature dependent Hall measurements on n-type silicon are tested. All thermal glues yielded the same results (resistivity, mobility, and charge carrier density) between 300 and 190 K. The use of RS 503-357 drastically distorts the expected results below 190 K, probably due to a phase transition and its latent heat, which affects the sample temperature during the phase transition. All the other thermal glues give reproducible results down to 100 K. Below 100 K, the use of cry-con, fixogum, and the silicon-high vacuum-grease from Leybold vacuum yield decreasing mobility and charge carrier density and increasing resistivity, as temperature decreases, but with different magnitudes. This is explained as the thermal properties of each glue start to diverge. Fixogum seems to give the best thermal conductivity, while the silicon-high vacuum grease from Leybold vacuum performs the worst below 100K. Crycon has an intermediate behavior between these two former ones. Cooling speed plays an important role at these low temperatures.