Characterization of CZT crystals grown by the Boron Oxide Encapsulated Vertical Bridgman technique for the preparation of X-ray imaging detectors Laura Marchini*, Andrea Zappettini*, Nicola Zambelli*, Mingzheng Zha*, Davide Calestani*, Eduard Belas** *IMEM-CNR, Parco Area delle Scienze 37/A, 43100 Parma, Italy **Institute of Physics, Charles University, Prague 2, CZ-121 16, Czech Republic INTRODUCTION In the last years CdZnTe has emerged as a very promising material for the fabrication of room temperature X- and gamma- rays detectors. However, the large-scale production of these detectors is limited by the low yield of material with the required material properties, i. e. high resistivity, high carrier mobility lifetime product, and low inclusion density. In this frame, a new technique for the growth of CZT crystals has been developed based on a modification of the vertical Bridgman technique consisting in the encapsulation of the molten charge by a layer of boron oxide [1]. In this work we present an extensive characterization of these crystals with different techniques, ranging from energy dispersion x-ray analysis, photoluminescence mapping, contactless resistivity mapping, and infrared transmission. CONCLUSIONS We have analysed the Zn fraction along a CdZnTe ingot and on a wafer. We have found that Zn fraction along the growth axis follows the normal freeze equation, except for the first to freeze part, where a supercooling effect brings to a deviation from the theoretical slope . The best fit suggest a value of 1.58 for the effective segregation coefficient. Zn fraction in the wafer is quite homogeneous and also the resistivity map shows a good homogeneity and a high resistivity value all along the crystal. IR Transmission showed the presence of Te precipitates. Fig.5: Resistivity map of a two wafer taken from the top(Fig.5a) and the bottom (fig.5b) DISCUSSION Zn fraction calculation was carried out , on the same sample, with two different techniques: PL emission and microanalysis. In Fig. 6a, it’s possible to see a comparison between the two results, the values obtained with the two different techniques are in a very good agreement The green line in Fig.5a represents the normal freeze equation , plotted with k=1.35 and x 0 =0.1. The main deviation from the theoretical behavior is in the first solidified fraction region were is present a sharp increase. This increase was explained as effect of the supercooling in the early stages of solidification [4]. The PL measure was fitted with the normal freeze equation and the result is shown in Fig. 6b. The obtained values for the segregation coefficient and the starting Zn fractions are respectively: k=1.58 and x 0 =0.05. Fig.4 Fig.6a Fig.6b Fig.2: PL map of a CdZnTe ingot in: a) a 3D plot and b) a 2D plot Fig.3: Zn fraction map of a CdZnTe ingot in: a)3D plot and b) 2D plot EXPERIMENTAL RESULTS on the wafer The ingot was also cut perpendicularly with respect to the growth axis., in order to check the Zn fraction gradient on a wafer. Almost no segregation was found in the wafer area. The Zn fraction gradient that is still present in Fig.4b could probably be attributed to a not correct alignment between the cutt of the waferand the growth axis. The Zn fraction along the axis is shown in fig.4a and Fig.4b. The homogeneity is also confirmed by the contactless resistivity mapping, reported in fig 5a and b . The measure was performed on two half wafer, one from the top (Fig.5a) and one from the bottom (fig.5b) of the ingot. The resistivity is high and homogeneous in both samples. The resistivity measured with this technique is usually lower by a factor 2 with respect to the number calculated from the current-voltage characteristic. T his difference could be due to the surface treatments ,done during the contact preparation, like surface passivation, that actually reduce the surface currents. [3] Fig.4: Zn Fraction map of a CdZnTe wafer in: a)3D plot and b) 2D plot a) a) a) b) b) b) IR Transmittance Te precipitates are present in the CdZnTe ingot growth by the modified vertical Bridgman technique. Both large (20 m) and small (2 μm) Te precipitates are present. REFERENCES [1] A. Zappettini, M. Zha, M. Pavesi, F. Bissoli, L. Zanotti, N. Auricchio, E. Caroli, IEEE Trans. Nuc. Science, 54 , (2007,) p. 782 [2] M. Prokesch, C. Szeles, J. Appl. Phys. 100 (2006), 14503 [3] L.Marchini, A. Zappettini, E. Gombia, R.Mosca, M.Pavesi , in press on IEEE Trans. Nuc. Science [4]M.Muhlberg, P.Rudolph, C.Genzel, B.Wermke, U.Becker, J.Cryst.Growth 101 (1990) p. 275 Fig.7 : IR Transmittance images showing Te precipitates EXPERIMENTAL RESULTS along the growth axis PL mapping is able to give detailed information about the Zinc segregation along the ingot., fig 1 At 77 k the E g peak position is related to the Zn fraction via the : were E g is in eV, and x is the Zn molar fraction. The plot of the Zn fraction percentage is reported in Fig.3a and in Fig.3b , respectively the 3D and 2D plot. The values range from 13.9% , in the top, to 2% , in the last to freeze part. The Zn concentration follows the normal freeze equation: Where x 0 is the starting concentration , k is the segregation coefficient and g the solidified molar fraction. 2 2969 . 0 5006 . 0 586 . 1 ) ( x x x E g    1 0 ) 1 ( ) (    k g k x g x Fig.1 [2] X axis Z axis PL scanned area b) a)