The loss of ozone on carbon soot aerosol was investigated in the temperatur
e range 238-330 K for reaction times up to 72 h: 200 mu g m(-3) ozone in dr
y synthetic air is depleted with an initial reaction probability gamma = 1.
2 x 10(-6) at 296 K. The reaction probability decreases with a characterist
ic time tau of about 12 h (1/tau = k(3) = (2.3 +/- 0.6) x 10(-5) s(-1)) due
to surface passivation. The reaction rate has a positive temperature depen
dence, and a complex negative dependence on the ozone concentration. A mini
mum set of four quasi-elementary reactions is required to model the observe
d concentration and time dependencies of the ozone loss rate: rapid destruc
tion of one monolayer equivalent of ozone on pristine surface sites (SS): S
S + O-3 --> SSO + O-2 (I), ozone-induced recovery of reactive sites: SSO O-3 --> SS + 2O(2) (or SS' + O-2 + CO2) (IIa, b), spontaneous recovery of r
eactive sites: SSO --> SS' + CO, and spontaneous site passivation: SSO -->
SSp (III). The kinetic parameters gamma(2ab), k(2c), and k(3), which were d
etermined in 11 independent experiments, lead to the following analytical e
xpression for the effective reaction probability gamma(eff) which applies w
hen the fast initial surface oxidation has reached a steady state:
gamma(eff) = [2 gamma(2ab) + 2.6 x 10(15) k(2c)/([O-3][c])]exp (-k(3)t).
This equation is valid over a wide range of atmospheric conditions. It impl
ies that ozone depletion on dry soot aerosol is negligible both in the trop
osphere and lower stratosphere. (C) 1999 Elsevier Science Ltd. All rights r
eserved.