The amount of data from geochemical analysis using samples collected in oil wells grows simultaneously to the investment in the exploration and production sector. On the other hand, the treatment and interpretation of these results are still very dependent on experts and demand time. With the generation of extensive databases, data mining presents itself as a good alternative to explore them through statistical methods and computational algorithms, providing technological differential and agility to the system. In an experimental way, with data from 200 oils from the Potiguar Basin, these tools were implemented, with the consequent suggestion of a workflow that would, in the end, return a reasonable accuracy in predicting their genetic classification. Using multidimensional scaling (MDS) and clustering (dendrogram and k-means types) from 60 initial attributes, the optimal set was reduced to 26. Applying Machine Learning, 92.50% of median accuracy were obtained in the Decision Tree algorithm, 95.00% in Random Forest and 87.50% in Artificial Neural Network. Comparing to an analysis previously presented at the pertinent literature, the benefits in terms of efficiency can be realized with the adoption of the methodology herein proposed. Keywords: Organic geochemistry; Data Mining; Multivariate Statistics; Workflow.

Polar compounds from three Brazilian crude oils were isolated and separated into seven fractions according to the polarity (low polarity-LP, low medium polarity-LMP, high medium polarity-HMP, high polarity-HP, basic-BAS, acidic-AC) by using hetero-medium pressure liquid chromatography (H-MPLC). Chemical characterization of all fractions was performed by mass spectrometry using an electrospray ionization source. In positive ion mode, the N-1 compound class was detected with high relative abundance in all samples. Interestingly, the unsaturation degree of this heteroatom class decreased from fractions LP, LMP, HMP to the HP, BAS and AC fractions. On the other hand, the type of compound classes found in negative ion mode was more diverse, revealing that nitrogen species (N-1) corresponding to carbazoles, were eluted in the less polar fraction (LP). Conversely, NxOy (x = 1, y = 1-4) heteroatom classes were found with a higher relative abundance in LMP, HMP, HP and BAS fractions. Overall, the H-MPLC reduced the complexity of the polar fraction contained in whole crude oils and allowed to extend the chemical characterization since the number of molecular formulas assigned in all fractions exceeds significantly to those detected in the whole crude oils. This separation methodology is an interesting alternative to other techniques such as gas chromatography (GC, GCxGC), liquid/liquid extraction, extrography, solid phase extraction and SARA (saturates, aromatics and resins) separation. MPLC not only isolate polar compounds but also fractionate them by the degree of polarity without previous de-asphalting steps. This separation approach, along with the molecular characterization of polar compounds by ESI-Orbitrap MS is also a tempting option to perform geochemical assessment of crude oils and source rocks.

The ionizable molecular composition of 36 lacustrine oil samples and 29 marine oil samples from different off-shore Brazilian basins have been screened by ultrahigh-resolution and accuracy Fourier transform ion cyclotron resonance mass spectrometry with electrospray ionization in the negative-ion mode [ESI(-)-FT-ICR MS]. The negative mode has been demonstrated successfully to characterize source rock because it favors the detection of heteroatomic compounds such as nitrogen-, sulfur-, and oxygen-containing (NSO), which reveal information about geochemical environmental. ESI(-)-FT-ICR MS data was processed with a new approach that unifies information from the chemometrics tool and petroleomics conventional data processing. Novel diagrams were proposed that use variable important to projection (VIP), such as molecular formula obtained, via partial least-square discriminant analysis (PLS-DA) to plot in diagrams, such as those that display double bond equivalence (DBE) versus number carbon atom (NCA) and H/C ratio versus molecular mass (MM) trends. This new approach is advantageous since it is able to verify and compare all samples at once, obtaining more selective variables, such as molecular formula, and with more confidence because the figures of merit are acquired by statistical tests from PLS-DA, whereas it also provides the usual information from petroleomics tools such DBE, NCA, H/C ratio, MM, and molecular classes. This new approach revealed significant differences in the molecular distribution of heteroatomic components for N-containing and O-containing molecules (N class and O class) allowing a clear separation of all samples in two groups of marine or lacustrine crude oils.

Ultrahigh-resolution Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR MS) allows the molecular-level characterization of ultracomplex mixtures such as crude oil samples. By combining FT-ICR MS with electrospray ionization, a myriad of polar compounds can routinely be identified and assigned with unmatched mass resolution and accuracy. The profile of polar compounds containing NSO atoms can be used to track and evaluate important parameters of crude oil. Recently, using ESI FT-ICR MS in the negative mode (https://doi.org/10.1016/j.orggeochem.2017.10.004), we investigated changes in the profile of polar compounds as a function of thermal maturity. Sulfur-containing compounds are completely destroyed during maturation, and the relatively high content of O-containing compounds decreases via decarboxylation and dehydration. The number of double bond equivalents (DBE) increased, indicating the occurrence of aromatization and condensation. However, the carbon number distribution shifted to lower values as a function of thermal maturity. Six regressions based on changes in the contents of O-2 compounds were proposed as maturity parameters. Here, by focusing on tracking and evaluating the impact of thermal maturity on basic polar compounds, a set of hydrous pyrolysis (HP) products was analyzed using an ESI(+) 7.2 T LTQ FT-ICR MS system. The samples at various stages of maturity consist of one immature bitumen (original sample), eleven expelled oil samples and eight residual bitumen samples. Determining the changes in the contents of N- and NO-containing compounds with maturity allows us to propose new parameters for assessing thermal maturity that cover the full window of oil generation.

Three crude oils recovered from three different wells of the Mero Field in the pre-salt of the Santos Basin were selected for detailed geochemical analysis. The samples were analyzed using a 7.2 T LTQ FT-ICR MS instrument by negative electrospray ionization (-ESI), focusing on the polar compounds, i.e., nitrogen-, sulfur-, and oxygen-containing compounds (NSO). Additionally, a combination of traditional geochemical methods including GC-FID, GC-MS, and carbon isotopic composition (whole oil and n-alkanes) were used to assess the samples. Through this work, it was demonstrated that -ESI FT-ICR MS is a reliable method for assessing crude oil composition and providing information about the origin and thermal maturity of the samples. Results showed that the dominant heteroatom classes are N-1,O-2,N1O1, and O-1. Due to the similarity of double bound equivalent (DBE), which means number of unsaturation present in an organic molecule, and carbon number distributions for the N-1 class species it is possible to suggest that Mero's filling history had an oil charge representing the peak of the oil window (0.7-0.9 %Ro) and that the oils were generated by a source rock deposited in a lacustrine environment. (C) 2019 Elsevier Ltd. All rights reserved.

The thermal maturity level impact on polar compounds was analyzed for a suite of hydrous pyrolysis (HP) products using a 7.2 Tesla LTQFTICR-MS instrument. The sample suite was analyzed via electrospray ionization in the negative ion mode focusing on the polar compounds, i.e., nitrogen-, sulfur-, and oxygen-containing (NSO) compounds. The HP experiments were performed under isothermal conditions for 2 h at eleven different temperatures (300, 310, 320, 325, 330, 340, 345, 350, 355, 360 and 365 degrees C) to simulate the full range of thermal levels starting at early bitumen generation to maximum oil generation. The maturity suite samples consist of one immature bitumen (original sample), eleven expelled oil samples and eight residual bitumen samples. In general, O-x compounds are destroyed by decarboxylation and dehydration with increasing maturity; sulfur compounds decrease; the aromaticity and degree of condensation increase; and steranoic and hopanoic acids decrease. Additionally, we investigated fluid retention properties comparing the carbon number distribution between the expelled oil and residual bitumen samples from each experiment. The maturity related changes of the acidic O-2 compounds in the expelled oil samples led us to create several regressions that we propose as new maturity parameters covering the full range of oil generation. (c) 2017 Elsevier Ltd. All rights reserved.

The performance of the high-field MegaOrbitrap Fourier transform mass spectrometer (FT-MS) with electrospray ionization (ESI) was evaluated to perform petroleum sample characterization via classical petroleomics approaches. Pertinent parameters that underpin the main figures of merit, that is, signal to noise ratios, dynamic range, spectral error, scan speed, mass accuracy and mass resolving power = R p, and provide subsidies to develop these analyzers were tested. Comparisons are made with data obtained using the most common petroleomics instrument, which is a Fourier transform ion cyclotron resonance mass spectrometer (FT-ICR MS), that has been used in the last decade in our laboratory for crude oil analysis providing R p of 340 000 at m/z 400 with transients of 3 s duration, and has been extensively demonstrated to fulfill all major requirements for precise petroleomics investigations. The high-field compact MegaOrbitrap mass analyzer, when operated at an R p = 840 000 at m/z 400 (R p > 1 000 000 at m/z 200) with a detection time of 3 s, was found to be well suited for adequate characterization of crude oil. Accurate class classification and mass accuracy below 1 ppm was obtained leading to proper, comprehensive petroleomics characterization.

Lacustrine and marine crude oils from different off-shore Brazilian basins were analyzed using a 7.2 Tesla LTQ FTICR-MS instrument. The samples were analyzed via electrospray ionization in the negative ion mode focusing on the polar compounds, i.e., nitrogen-, sulfur-, and oxygen-containing (NSO) compounds. We also employed a combination of other geochemical methods, such as GC-FID and GC-MS analyses, to characterize and assess the depositional environments of the different oil families. The results indicate that lacustrine oils tend to be enriched in Nx compounds, while marine oils show preference for Ox compounds. The dominant heteroatomic classes in crude oils are N1, followed by O1, O2, and N1O1 with remarkable differences in their distributions between marine and lacustrine, strongly suggesting the control by the kerogen type of the heteroatomic compounds found in these crude oils. Considerable differences in the DBE distribution of the main classes analyzed between the crude oils allowed an efficient geochemical characterization regarding their origin. The use of negative ESI FTICR-MS as a geochemistry tool can provide additional information beyond that obtained with currently employed geochemical methods, resulting in the full comprehension of crude oil composition. (C) 2018 Elsevier Ltd. All rights reserved.