Test of a dry objective with correction collar

Test of a dry objective with correction collar G. Sirri for the Bologna group Dry objectives having an high numerical aperture give aberration when imaging through coverslips that deviate from design thickness and refractive index. To prevent artifacts, many objectives are equipped with correction collars that help compensate for coverslip thickness variations. By means of the Cover Glass Correction slider the position of a movable lens group is adjusted inside the objective. In this way, it is possible to observe the specimen through variable glass thickness from 0 to 2 mm. This can be useful in emulsion scanning because thickness of the intermediate medium between the front lens and the object plane ranges from 0 (when we scan the top) to about 0. 3 mm (bottom). Note that when the collar is adjusted, focus tends to shift and the magnification can slightly change. LNGS 25 May 2005 G. Sirri – INFN BO 1

Characteristics of the objective This type of objective can be found normally in the market (from Leica, Nikon, Olympus or Zeiss) and are largely used by biologists. We have used for this test the ZEISS LD Achroplan (code 440864) magnification: 40 x numerical aperture: 0. 60 working distance: 1. 8 mm correction range: 0 -2 mm We installed this objective in our prototypal microscope equipped with Zeiss tube lens and our custom CMOS camera. The Field of View is about 395 x 310 µm² The micron/pixel ratio is 0. 310 µm (top) and 0. 305 (bottom). LNGS 25 May 2005 G. Sirri – INFN BO 2

Test conditions In our case the usage of this objective requires a continuum collar rotation synchronously with the movement of the vertical stage. However, we used only 2 correction positions d~0 for the top and d~0. 25 for the bottom. 1. 6 cm² has been scanned in the plates 34, 35, 36, 37, 38, 39 of the reference brick (CERN June 2004, Brick 5 without lead) and one-by-one compared with the same measurements with the standard oil objective in the standard microscope. 20 levels (4 inactive) for view were grabbed. The minimum area of the clusters was 3 pixels. Since we do not have any automatic rotating system, we scanned all the fragments of the top and then of the bottom. After the merging the two raw data files into one as in an ordinary acquisition, we analyze the data with FEDRA. LNGS 25 May 2005 G. Sirri – INFN BO 3

Grains analysis Level-to-level distance ~ 0. 5 µm z cluster e. Z Vertical clusters chains obtained using the lib. EGA of FEDRA z 0 grains OIL Cluster area (pixels) DRY cluster-to-grain_center vertical distance (µm) LNGS 25 May 2005 G. Sirri – INFN BO 4

Base Tracks (pl 35, 1. 6 cm²) QUALITY CUT BACKGROUND SIGNAL LNGS 25 May 2005 G. Sirri – INFN BO 5

Slope Y Base Tracks (pl 34, 1. 6 cm², chi_puls cut) After the quality cut DRY: ~ 110 tracks/mm² OIL: ~ 92 tracks/mm² Slope X LNGS 25 May 2005 G. Sirri – INFN BO 6

Base Tracks [microtracks – base tracks agreement] DRY Plate 35 1. 6 cm² (0. 0 , 0. 0) Plate 35 (0. 4 , 0. 0) σ = 6 mrad σ = 23 mrad Plate 35 OIL 4. 8 cm² (0. 0 , 0. 0) σ = 9 mrad LNGS 25 May 2005 (0. 4 , 0. 0) σ = 21 mrad G. Sirri – INFN BO 7

Volume Tracks (6 plates – no holes – 1. 6 cm²) LNGS 25 May 2005 G. Sirri – INFN BO 8

Volume tracks : angular residuals DRY OIL Linear fits: LNGS 25 May 2005 G. Sirri – INFN BO 9

Volume tracks : position residuals DRY OIL LNGS 25 May 2005 G. Sirri – INFN BO 10

Volume Tracks : efficiencies DRY OIL LNGS 25 May 2005 G. Sirri – INFN BO 11

Conclusions o o These are preliminary results because the on-line parameters were not well tuned. However, the use of this technique seems to give results comparable to our oil results on the same conditions. LNGS 25 May 2005 G. Sirri – INFN BO 12