affecting the microstructure. Initial crystal formation occurred uniformly in the tissue at time of the expansion, but crystal growth took place at atmospheric pressure. Since the number of nuclei was large, crystals did not attain large sizes. HPSF pro- duced no appreciable damage at the surface of the sample (see Fig. 4 for the surface zone of peach frozen by high pressure). Also, damage in the inner zones was negligible (see Fig. 5 for the cen- tral zone of the peach). In HPSF, expansion to atmospheric pres- sure is followed by traditional freezing, controlled by thermal gradients (characteristic times are negligible at the surface and
57 and 80 min at the center for peach and mango, respectively) . Given that about 65% of the remaining liquid water has to be fro- zen, ice crystals in the central zone may be larger than at the sur- face depending on the power of the refrigerating system. Howev- er, this effect is not as critical as in liquid nitrogen freezing. Freeze-cracking was not observed in any case of HPSF because nucleation was uniform and massive throughout the fruit during the pressure release stage, which prevents internal stress accu- mulation due to ice expansion on freezing.
These results agree with other findings for muscle tissue and
other vegetable systems (Otero and others 1997; Otero and oth- ers 1998; Martino and others 1998; Fuchigami and others 1997b; Koch and others 1996).
Fruits that have been frozen by high-pressure-shift freezing
method maintained an acceptable microstructure, also shown by SEM micrographs. Figure 6 shows the central zone of peach frozen by HPSF procedure. Figure 7 is a magnification of Fig. 6, and Fig. 8 shows the central zone of frozen mango.