[en] Using the Fast On-line Reaction Apparatus (FORA), the influence of various gas-purification columns onto the formation of metal carbonyl complexes (MCCs) under single-atom chemistry conditions was investigated. MCCs were synthesized from single atoms of Mo, Tc, Ru and Rh being produced by the spontaneous fission of ${}^{252}$Cf and recoiling into a CO-gas containing carrier gas atmosphere. The in-situ synthesized MCCs were volatile enough to be transported by the carrier gas to a charcoal trap where they were adsorbed and their subsequent decay was registered by γ-spectrometry. It was found that the type and combination of purification columns used to clean the applied CO-gas strongly influences the obtained formation and transport yields for all MCCs. With the exception of Rh-carbonyl, intense gas-purification strategies resulted in reduced formation and transport yields for MCCs in comparison with less efficient or even completely missing purification setups. It was postulated that the observed reduction in yield might depend on the content of Fe(CO)${}_5$ and Ni(CO)${}_4$, as well as potentially other MCCs, in the CO-gas, being formed by the interaction between CO and the steel-surfaces of FORA as well as from impurities in the used charcoal traps. Subsequently, it was shown that macro amounts of Fe(CO)${}_5$, Ni(CO)${}_4$, Mo(CO)${}_6$ and Re${}_2$(CO)${}_{10}$ added to the used process gas indeed increase significantly the overall yields for MCCs produced by ${}^{252}$Cf fission products. Ni(CO)${}_4$ appeared the most potent to increase the yield. Therefore, it was used in more detailed investigations. Using isothermal chromatography, it was shown that Ni(CO)${}_4$ does not affect the speciation of carbonyl species produced by the ${}^{252}$Cf fission product ${}^{104}$Mo. For ${}^{107}$Tc, ${}^{110}$Ru and ${}^{111}$Rh a speciation change cannot be excluded. For ${}^{111}$Rh a speciation change cannot be excluded. An inter-carbonyl transfer mechanism is suggested boosting the formation of MCCs. The current discovery might allow for new opportunities in various research fields, which are currently restricted by the low overall yields for MCCs produced under single-atom chemistry conditions. Examples are the chemical investigation of transactinides or the generation of radioactive ion beams from refractory metals at accelerators.