A series of soil slurry experiments were performed in a carefully conceived
reactor set-up to investigate the characteristics of the catalytic decompo
sition of ozone on a sand and iron surface. Real time on-line monitoring of
ozone in the reaction module was possible using flow injection analysis co
upled with a computer-cont rolled UV detector and data acquisition system.
The effects of the soil media and size, ozone dosage, pH and p-CBA as a pro
be compound were examined at the given experimental conditions. Two apparen
t phases existed, and ozone instantaneously decomposed within one second in
the first phase. These were defined as the instantaneous ozone demand (ID)
phase, and the relatively slow decay stage. The interactions of ozone with
the soil organic matter (SOM) and metal oxides were attributed mostly to t
he instantaneous decomposition of ozone. From the probe (p-CBA) experiments
, 60-68% of total p-CBA removal occurred during the ID phase. The generatio
n of hydroxyl radicals (OH .) was demonstrated and was closely related with
metal oxides as well as SOM. Metal oxides in soil surface were considered
to have relatively faster reaction rate with ozone and provide more favorab
le reactive sites to generate higher amount of OH . than SOM. Even at one-t
enth concentration of the sands, a goethite-induced catalytic reaction outf
itted the removal rate of p-CBA among all the soils tested. More than 40% o
f total p-CBA removal occurred on the soil surface. It was inferred that th
e radical reaction with the probe compound seemed to take place not only on
the soil surface but also in the solid-liquid interface. Ozone decompositi
on and the reaction between OH . and p-CBA appeared to be independent of an
y change in pH. (C) 2001 Elsevier Science Ltd. All rights reserved.