Modelling water vapour permeability through atomic layer deposition coated photovoltaic barrier defects

Abstract

Transparent barrier films such as Al 2 O 3 used for prevention of oxygen and/or water vapour permeation are the subject of increasing research interest when used for the encapsulation of flexible photovoltaic modules. However, the existence of micro-scale defects in the barrier surface topography has been shown to have the potential to facilitate water vapour ingress, thereby reducing cell efficiency and causing internal electrical shorts. Previous work has shown that small defects (≤3 μm lateral dimension) were less significant in determining water vapour ingress. In contrast, larger defects (≥3 μm lateral dimension) seem to be more detrimental to the barrier functionality. Experimental results based on surface topography segmentation analysis and a model presented in this paper will be used to test the hypothesis that the major contributing defects to water vapour transmission rate are small numbers of large defects. The model highlighted in this study has the potential to be used for gaining a better understanding of photovoltaic module efficiency and performance.

Implementation of in Process Surface Metrology for R2R Flexible PV Barrier Films

“…Following that, the WSI was used with 5X lens objective in a clean room (class 10000), to measure the same measured area that was previously measured by the CCI, and covering 13% of the total sample area. The collected data from both the CCI and WSI instruments were analyzed through a segmentation analysis [12] using a Surfstand software package, and the criteria used to segment and count only significant defects were (3 × Sq vertical ≈ 20% Sz) and 3 μm laterally, which is in accordance with previous studies conducted by the authors [13]. The results of the analysis, shown in Fig.…”

“…22. Comparing the parameter values to an experimentally determined threshold value, obtained from extensive lab-based measurements for an ALD coating that were previously published in [4] and [13], the average roughness of a Al 2 O 3 ALD coating without defects using the WSI was found to be Sq ≈ 7 nm, as shown in Fig. 9.…”

“…Previous studies [4,13] have demonstrated that significant defects will increase the roughness Sq (with defect) larger than 3 × Sq (no defect) i.e., Sq (with defect > 21 nm). Moreover, segmentation analysis results [5,13] appear to indicate that the major contributing factor for determining the WVTR is the total number of larger defects, where a sample with a higher density of defects of greater than 3 μm in lateral dimension and a 3 × Sq vertical dimension exhibits inferior barrier properties.…”

“…Moreover, segmentation analysis results [5,13] appear to indicate that the major contributing factor for determining the WVTR is the total number of larger defects, where a sample with a higher density of defects of greater than 3 μm in lateral dimension and a 3 × Sq vertical dimension exhibits inferior barrier properties. These criteria are considered to be the basis for developing the segmentation "toolbox" shown in Fig.…”

“…The edge data were then "pruned" using Wolf pruning [12]. Following the Wolf pruning, an "area pruning" was applied, where the area was found to have a lateral diameter of up to 3 μm (which calculated based on a mathematical model published in [13] was deemed insignificant). Fig.…”