banner image
No data available.
Please log in to see this content.
You have no subscription access to this content.
No metrics data to plot.
The attempt to load metrics for this article has failed.
The attempt to plot a graph for these metrics has failed.
Diffusion barrier properties of molybdenum back contacts for Cu(In,Ga)Se2 solar cells on stainless steel foils
Rent this article for


Image of FIG. 1.
FIG. 1.

Schematic shape of the penetration profile of surface, grain boundary, and bulk diffusion (adapted from Heitjans et al. 16 ). The proposed geometries are from the Fisher model. 17

Image of FIG. 2.
FIG. 2.

Different back contact configurations used for the evaluation. The set includes six samples (S1–S6) with corresponding references (RefS1–RefS6). The references additionally have an adhesion-promoting Ti layer.

Image of FIG. 3.
FIG. 3.

Density of Mo back contacts versus their thickness and the influence of different deposition conditions (see Fig. 2 ). The density increases with the film thickness. Films with process A show higher density than with process B. The bulk density of Mo is 10.22 g/cm3 (at 20 °C). 31

Image of FIG. 4.
FIG. 4.

Residual stress in the back contact films of samples S1–S6. The bilayer designs (S1 and S4) show comparable stress values as single layer MoB (S3 and S6), which could be a result of stress relaxation at the interface.

Image of FIG. 5.
FIG. 5.

STEM image of reference contact (RefS) reproduced on glass substrate. The microstructure of the Ti/MoA interface shows an abrupt change, whereas the MoA/MoB interface shows continuous growths of the columns. STEM-EDS mapping profile indicates a lower Mo density at the interface of MoA/MoB.

Image of FIG. 6.
FIG. 6.

ToF-SIMS depth profiles of back contact RefS6 (I) and S4 (II) after CIGS processing, showing Fe diffusion towards the CIGS layer. For RefS6, Fe is blocked at the Steel/Ti (1) and Ti/MoA interface (3). Inside of the Ti layer (2), the Fe concentration has a constant value of ∼1.2 at. %. In case of the sample S4, the Fe concentration strongly decreases starting at the Steel/MoA interface (1) and is reduced in the Mo film. The SIMS profiles were sputtered from the front towards the substrate (right to left). To improve visualization, the x-axis was scaled with respect to the layer thickness.

Image of FIG. 7.
FIG. 7.

Normalized Fe distribution profile in CIGS for samples S6 and S5 measured by ToF-SIMS. For both samples, the Fe concentration linearly decreases from the back contact to the CdS buffer layer.

Image of FIG. 8.
FIG. 8.

Influence of the back contact density on the Fe amount in CIGS and loss in cell efficiency compared to the respective reference samples (X = 4…6). Increasing the Mo back contact density reduces the Fe diffusion and the loss in efficiency, whereby the bilayer back contact (S4) shows almost the same performance as the high-density MoA single-layer sample (S5).

Image of FIG. 9.
FIG. 9.

Comparison of sample S6 and RefS6. The carrier density N (I) of S6, measured with CV, is ∼3.5 times lower compared to RefS6. JV (II) and EQE (III) curves show measured (symbols) and SCAPS-simulated (lines) values, where both the linearly graded and constant Fe defect concentration are shown. Best match was achieved with a linear Fe impurity profile, whereas for a constant profile the calculated Voc is too low.


Generic image for table
Table I.

Parameters of the Mo sputtering process A and B and the Ti adhesion-promoting layer. The thickness of the layers was controlled by the speed of the substrate movement.

Generic image for table
Table II.

Summary of JV results, where the average value for each sample (av.) and the difference to the corresponding reference (Δ = SX − RefSX, X = 1…6) is shown. An ARC was applied for all cells.

Generic image for table
Table III.

Summary of the parameters used for SCAPS simulation for samples RefS6 and S6, which are layer width (W), material (ε) and vacuum permittivity (ε0), bandgap (Eg), electron (μ e) and hole mobility (μ h), shallow donor (ND) and acceptor (NA) concentration, and effective states in conduction (NC) and valance (NV) band. The bandgap of the CIGS layers is graded and was taken from ToF-SIMS measurements (not shown). Additionally, mid-gap state defects were assumed for all layers (Nmgd), which are acceptor states for CdS and donor states for the other layers. For sample S6, a FeIn,Ga defect concentration (NFe) with energy of 320 meV above the valence band was used for the simulation, which was linearly graded for CIGSL and constant for CIGSC. The RefS6 sample (CIGSRef) was assumed to be Fe-defect free.


Article metrics loading...


Full text loading...

This is a required field
Please enter a valid email address
752b84549af89a08dbdd7fdb8b9568b5 journal.articlezxybnytfddd
Scitation: Diffusion barrier properties of molybdenum back contacts for Cu(In,Ga)Se2 solar cells on stainless steel foils