A B S T R A C T It is well known and has been shown that the gain performance of Gaseous Electron Multipliers (GEM) depends on the size of the holes. With an optical scanner it is possible to measure the dimensions of the holes, and to predict the performance of GEMs. However, the gain prediction of GEMs that are manufactured with a double mask etching technique is not straightforward. With the hole size information alone, it is not possible to make precise prediction of the gain. We show that

more »

... he alignment of the photo-masks between the two sides of the GEM foils plays a crucial role. A misalignment of a few microns can lower the gain substantially. The study is performed by using the Helsinki high definition optical scanner for quality control of GEM foils, and this will show its true potential. * Corresponding authors. E-mail addresses: erik.brucken@iki.fi (E. Brücken), timo.hilden@helsinki.fi (T. Hildén). GEM based detectors show good cost efficiency for covering large areas/volumes. A recent example is the upgrade of the Time Projection Chamber of the ALICE experiment at CERN [4]. In such experiment it is important to know the quality of the GEM foils before the detector is assembled. Therefore a thorough quality assurance (QA) effort was undertaken [5] . One important QA method was the optical scanning and gain prediction of the foils, originally initiated by the studies done at the Helsinki Institute of Physics (HIP) [6, 7] . A high definition optical scanner and an advanced image analysis software, developed at HIP, is capable of measuring parameters of individual holes, such as the diameter of the polyimide hole, and the diameters of the copper holes from both sides. It was shown that gain prediction is possible based on the knowledge of those geometrical hole properties [7] . The test case back then was based on 10 × 10 cm 2 double-mask GEM foils. However, only one of the test foils was measured with a position sensitive system with high enough granularity to measure local gain variations. In particular, the foil was placed on top of a Micro-Mesh Gaseous Structure (MICROMEGAS) detector [8] with x-y strip readout. The active area was 89.6 mm × 89.6 mm covered by 358 strips per dimension. The pitch of the readout strips was 250 μm with the y-strips being 80 μm wide and the x-strips, located below, 200 μm for equal charge sharing [9] . The gain maps were binned with a bin size of 4.5 × 4.5 mm 2 . The published results were encouraging but a refined study was deemed necessary for a quantitative conclusion. In the follow up study, which is the subject of this paper, we tested three new GEM foils produced with double-mask technique with different setups. However, at first we did not find similar inverse dependency https://doi.