Atmospheric aerosol is ubiquitous and its impact on climate and health has not been fully understood yet. Multiple diverse processes and sources lead to atmospheric aerosol formation and emission and this reflects on its heterogeneity in chemical composition and physical properties. This chemical complexity does not allow full characterization on a molecular level by mean of any singular analytical technique. In this work we describe the chemical composition of the organic fraction of atmospheric and simulated aerosol through the functional groups (FGs) of which is formed. FGs allow us to describe an organic mixture in a simplified way nevertheless retaining some chemical specificity. To measure the chemical bonds that can be associated to different FGs we use Fourier transform infra-red (FTIR) spectroscopy. This technique can detect and quantify chemical bonds that undergo a change in the dipole moment in their vibrational modes. Here we introduce a new chemoinformatic tool that allow us to extract FGs from sets of compounds. This method guarantees no overlap between different FGs that contain similar substructures and has been tested for completeness and uniqueness using the full set of compounds included in the oxidation mechanism of alpha-pinene and 1,3,5-trimethylbenzene in the Master Chemical Mechanism (MCMv3.2). We then use it to study the FG evolution of aerosol generated using the MCMv3.2 (a semi explicit chemical model) coupled with an absorptive partitioning module. In this way we are able to compare the composition of the simulated aerosol with the composition of aerosol generated in laboratory studies in which FTIR spectroscopy was used as analytical technique. By analyzing the discrepancies between model and measurements we suggest possible missing reactions mechanisms in the model that could improve the comparison with laboratory results. FTIR was then used to analyze ambient organic aerosol collected in Alabama during summer 2013, as we participated to SOAS (Southern Oxidant Aerosol Study) field campaign. We show that the FGs distribution that we measured can be partly explained as being formed from isoprene photooxidation in the atmosphere. Moreover, our results are consistent with results obtained with on-line analytical techniques such as aerosol mass spectrometry and aerosol chemical speciation monitor. Comparisons with on-line techniques suggest that we were able to measure only the non-volatile fraction of the organic aerosol. In the last part of this work we introduce a technique that can allow the measurement of the FG distribution of the fraction that is lost during sampling onto a Teflon filter. By desorbing and cryofocusing molecules previously adsorbed on a medium, such as a filter, onto a Silicon window, we are able to measure the FG distribution without the interference of the filter itself. Here we show a proof of concept study using a laboratory standard.