Figure 3. An abrupt, vertical pattern of degradation within the hybrid layer, as illustrated by the radiographs and TEMs of primary molars that were retrieved from the third clinical subject (Control group, a, c, and e; Experimental chlorhexidine group, b, d, and f). C, resin composite; A, adhesive; H, hybrid layer; D, dentin. (a) Bitewing radiograph showing the Class I restoration in tooth 85 (control, circled). (b) Bitewing radiograph showing the Class I restoration in tooth 75 (experimental, circled). (c) Undemineralized, unstained, silver-impregnated section from the control tooth. The hybrid layer exhibited a pale electron density (gray), with regions containing silver deposits (pointer). In addition, there were electronlucent (white) zones within the hybrid layer that did not contain silver deposits (open arrow). Since silver impregnation was performed prior to epoxy resin infiltration, these white zones probably contained materials that were more highly penetrable by the electron beam than the original resin-infiltrated collagen fibrils. (d) The same type of section from the experimental tooth, showing similar silver deposits (pointer) within the hybrid layer, but with the lack of those electronlucent regions depicted in Fig. 3c. (e) Demineralized section from the control tooth that was stained with uranyl acetate and lead citrate. Vertical regions (asterisk) were found abruptly within the hybrid layer in which the bulk of the collagen fibrils have degraded, leaving behind strands of loosely arranged microfibrils (i.e., partially degraded gelatin; see Fig. 4e). Resin tags that were surrounded by islands of intact collagen fibrils (open arrowhead) remained within these abrupt regions of degradation. (f) The same type of section from the experimental tooth, showing the absence of degradation from the hybrid layer (between open arrowheads) with chlorhexidine as an MMP inhibitor.