Purpose: To determine the quantitative contribution of dentin hybridiz
ation to bonded assembly strength and demonstrate the micromorphology
of the interface with and without collagen present. Materials and Meth
ods: Four groups of 10 molar teeth were finished to a 320 grit dentin
smear layer. Two groups served as controls and two experimental groups
were subjected to collagenase digestion of the collagen exposed by ac
id conditioning. All-Bond 2 and Amalgambond were used to bond Bisfil a
nd Epic resin composite, respectively. Stored in water at 37 degrees C
for 24 hours the assemblies were tested in a shear mode at a crosshea
d speed of 5 mm/minute. Means and standard deviations were subjected t
o analysis for statistical significance. Twenty four teeth in four gro
ups were examined by scanning (SEM) and transmission electron microsco
py (TEM) for the relationship between resin and conditioned dentin wit
h and without the collagen network. Results: All-Bond 2 and Amalgambon
d controls were 28.41 +/- 3.9 and 19.04 +/- 5.96 MPa, collagenase-trea
ted groups scored 26.43 +/- 2.90 and 19.70 +/- 4.25 MPa respectively.
No significant difference existed between the control and experimental
groups. SEM showed an intertubular collagen network with patent tubul
es and a pronounced porous, irregular dentin topography following coll
agen digestion. A distinct hybrid zone and tubular penetration was obs
erved but the collagenase-treated specimens showed only resin in the t
ubules and their lateral extensions. TEM confirmed the absence of a di
stinct hybrid zone in the collagenase groups with a tight, gap-free ju
nction between the resin and the undemineralized dentin. An electron d
ense zone (< 50 nm) at the leading edge of conditioning was observed f
or All-Bond 2 and Amalgambond groups. It was concluded that the resin-
reinforced or hybridized, collagenous network does not detract from, n
or contribute any significant quantitative value per se to dentin bond
ing with the systems tested.