Adaptive femtosecond laser pulse shaping is employed to achieve active quan
tum control in selected gas phase photodissociation and photoionization pro
cesses of CpFe(CO)(2)Cl. An evolutionary optimization algorithm uses experi
mental feedback from the photofragment mass spectra in a learning loop setu
p to improve the applied spectral laser pulse phases iteratively. The signi
ficance of the resulting complex temporal electric-field profiles is invest
igated by artificial reduction of the optimization parameter space. The deg
ree of problem complexity is further studied by correlation measurements of
the molecular control objective versus second-harmonic generation (SHG). I
n the case of direct ionization, the previously unknown molecular control p
roblem can be mapped one-to-one onto the search space topology of SHG, wher
eas in the case of photoproduct ratio optimization, no clear correlation is
observed. The control mechanism in the latter example is therefore much mo
re complex and cannot be explained by a simple laser pulse intensity variat
ion. Correlation techniques in a general sense are considered to be very he
lpful in solving the so-called problem of inversion, i.e., obtaining inform
ation about the underlying photophysical and photochemical processes. (C) 2
001 Elsevier Science B.V. All rights reserved.