[Plumbum]

Its a flash photolysis. Its the decomposition of a compound. We detected the decomposition visually. Oh and we measured in different wave lenghts: from 540 to 640 nm

[Corribus]

Still not much information to go on, but it helps.

Flash photolysis can be thought of a bimolecular reaction between two reactants: your chemical (call it A) and a photon.  Typically the "concentration" of the photon is high enough to be considered "constant", in the sense that removing one photon does not really change the "concentration" of photons appreciably. From the point of view of the photon, the reaction is zero order (such that zero order reactions exist, which they don't, not really). That is, the photon concentration doesn't show up in the rate expression.

This is generally not true for the other reactant. In the initial stages of the reaction, where concentration of A is high, the probability of a photon striking A are high, and thus the rate of subsequent chemistry is concomitantly high. As A is depleted, the probability of an absorption event diminishes, and so does the reaction rate. Therefore the rate does typically change as a function of the concentration of A. In some concentration ranges the reaction may be zero order (or pseudo-zero order), but typically there should be some critical concentration at which point the reaction rate begins to change with the concentration - the collision  probability becomes important, as in any bimolecular reaction. In your case, it appears to me that the fall off of your signal is exponential with respect to the reaction time, and the natural logarithm of the signal is almost certainly fittable to a linear function, suggesting the decay is a single exponential term. This is not uncommon for photophysical/photochemical events with a single decay pathway.

(From a photophysical/chemical perspective, there are also numerous "reaction channels" that can determine whether the absorption event results in chemistry or results in no reaction. Some of these processes are concentration independent, but not all of them are. Often these will determine if another reactant is involved - e.g., oxygen - or whether multiple A's are involved. All of these influence the overall kinetics and therefore order of the reaction. In the case of some monitoring techniques, e.g., fluorescence, you will often see multi-exponential decay of the signal, but some of the exponential terms may be due to solvent relaxation, structural relaxation, reaction with oxygen, etc., although whether or not you can see all these will depend on the pulsewidth of your light source. Most flash photolysis experiments are done with pulses on the order of 1 ns or longer, meaning that internal relaxation processes (that occur prior to chemistry) are not observed.)