Page 84 - Mirjam-Theelen-Degradation-of-CIGS-solar-cells
P. 84
Chapter 2
2.5.2 Degradation of conductive grids
Generally, metal grids are used to overcome the limitations in the conductivity of
TCOs. By providing an additional low-resistance path for current collection, they re-
duce the overall series resistance of the cell and thus ohmic power losses. As any other
part of the solar cell, grids are subject to degradation but, at the same time, a strategy
to deal with a degrading solar cell.
In a solar cell or module with a grid, the ohmic losses in the TCO can be controlled
by varying the spacing between the grid fingers. When decreasing the grid finger
distance, the shading increases and thus some of the initial power of the solar cell is
lost due to a reduction of the current density [112]. However, the module efficiency
can be kept at a higher level after damp heat exposure because additional grid fingers
compensate for the increasing sheet resistance of the TCO. From that point of view,
the gridded module design is preferable from a long-term performance point of view
[39]. Wennerberg et al. pointed out further advantages of a gridded module such as:
• Less inactive area (the cell stripes in a monolithically interconnected module can
be wider)
• Higher tolerance for variations
• A cheaper TCO process, since a thinner TCO can be used
• Increased degrees of freedom in the module geometry [113].
Nevertheless, the authors noted that the benefits of a gridded module have to over-
weight the cost of an additional grid deposition in order to pay off in production.
A typical grid for solar cells in many research labs consists of a thin nickel layer (e.g. of
50 nm thickness), followed by a thick aluminium layer (e.g. 2-3 µm thickness) that are
typically applied by an evapouration method. Wennerberg et al. reported that this
structure corroded considerably and led to uncertainty in the interpretation of chang -
es in the IV characteristics [38]. A modified grid structure, however, with an additional
nickel capping (50 nm), which resulted in a Ni/Al/Ni stack, improved the stability. No
visual degradation was observed after 1000 hours of damp heat exposure and the
statistics showed better reproducibility.
Pern et al. reported that a standard metallic grid of a 3 µm aluminium layer on top
of a 50 nm nickel layer was instable under damp heat [30]. The stack was subject to
hydrolytic corrosion within 50 to 100 hours of damp heat exposure thus becoming
highly resistive. Since nickel is known to be more stable against oxidation and hydro-
lysis than aluminium, Pern et al. prepared solar cells with only nickel grids. At a grid
thickness of 0.2 to 0.3 µm, the solar cells with nickel-only grids showed a lower initial
efficiency (compared to those with standard Ni/Al grid) due to the higher resistance of
the nickel grid. At nickel grid thicknesses of 1.6 to 1.8 µm, the devices demonstrated
82
2.5.2 Degradation of conductive grids
Generally, metal grids are used to overcome the limitations in the conductivity of
TCOs. By providing an additional low-resistance path for current collection, they re-
duce the overall series resistance of the cell and thus ohmic power losses. As any other
part of the solar cell, grids are subject to degradation but, at the same time, a strategy
to deal with a degrading solar cell.
In a solar cell or module with a grid, the ohmic losses in the TCO can be controlled
by varying the spacing between the grid fingers. When decreasing the grid finger
distance, the shading increases and thus some of the initial power of the solar cell is
lost due to a reduction of the current density [112]. However, the module efficiency
can be kept at a higher level after damp heat exposure because additional grid fingers
compensate for the increasing sheet resistance of the TCO. From that point of view,
the gridded module design is preferable from a long-term performance point of view
[39]. Wennerberg et al. pointed out further advantages of a gridded module such as:
• Less inactive area (the cell stripes in a monolithically interconnected module can
be wider)
• Higher tolerance for variations
• A cheaper TCO process, since a thinner TCO can be used
• Increased degrees of freedom in the module geometry [113].
Nevertheless, the authors noted that the benefits of a gridded module have to over-
weight the cost of an additional grid deposition in order to pay off in production.
A typical grid for solar cells in many research labs consists of a thin nickel layer (e.g. of
50 nm thickness), followed by a thick aluminium layer (e.g. 2-3 µm thickness) that are
typically applied by an evapouration method. Wennerberg et al. reported that this
structure corroded considerably and led to uncertainty in the interpretation of chang -
es in the IV characteristics [38]. A modified grid structure, however, with an additional
nickel capping (50 nm), which resulted in a Ni/Al/Ni stack, improved the stability. No
visual degradation was observed after 1000 hours of damp heat exposure and the
statistics showed better reproducibility.
Pern et al. reported that a standard metallic grid of a 3 µm aluminium layer on top
of a 50 nm nickel layer was instable under damp heat [30]. The stack was subject to
hydrolytic corrosion within 50 to 100 hours of damp heat exposure thus becoming
highly resistive. Since nickel is known to be more stable against oxidation and hydro-
lysis than aluminium, Pern et al. prepared solar cells with only nickel grids. At a grid
thickness of 0.2 to 0.3 µm, the solar cells with nickel-only grids showed a lower initial
efficiency (compared to those with standard Ni/Al grid) due to the higher resistance of
the nickel grid. At nickel grid thicknesses of 1.6 to 1.8 µm, the devices demonstrated
82
