David J. Cookson

Determination of carbon C, CH, CH2 and CH3 group abundances in liquids derived from petroleum and coal using selected multiplet 13C n.m.r. spectroscopy

Selected, multiplet C n.m.r. spectra are obtained for three test samples deriving from petroleum and coal sources, by combining gated spin echo (GASPE) and conventional spin echo 13C n.m.r. procedures. Each selected multiplet spectrum contains resonances due to only one of the following groups: aromatic C or CH or aliphatic C, CH, CH2orCH3. In general artifacts contribute only minor intensity to individual spectra, with the separation between aliphatic CH and CH3 spectra being the most difficult to achieve. Each spectrum can be integrated to yield the relative abundances of CHn groups (n = 0 to 3). Selected multiplet 13C n.m.r. spectra provide a more detailed view of the component hydrocarbon groups in fossil-fuel derived materials than can be deduced from conventional 13C n.m.r. spectra.

Optimal conditions for obtaining quantitative 13C NMR, data

Abstract Relaxation effects on quantitation in 13C NMR spectroscopy have been investigated. The objective has been to define prerequisite conditions for obtaining quantitative results efficiently. For the general case, where the inverse gated decoupling procedure is applied without the addition of a paramagnetic reagent, requirements have been summarized in graphical form. Long relaxation delays may be necessary when the ratio of 1H and 13C direct relaxation rates, R H R c , is small. It has been found that when a paramagnetic reagent, viz., Cr(AcAc)3, is used, R H R c becomes large and simple sufficient conditions for quantitation can be defined explicitly. Practical considerations in the use of Cr(AcAc)3 are discussed.

Composition-property relations for jet and diesel fuels of variable boiling range

Abstract Simple linear relations between properties and composition are developed for jet and diesel fuels of different boiling ranges. Composition is defined in terms of two parameters readily determined by 13 C n.m.r. spectroscopy, namely C n , the fraction of n -alkyl carbon, and C ar , the fraction of aromatic carbon. The boiling distribution of the fuels is defined, using gas chromatography, in terms of two parameters, T 10 and T 90 , the temperatures at which 10 and 90 wt% of the fuel has vaporized. Relations of the form P = a 1 C n + a 2 C ar + Σ b i T i + k , where P is the property value and a 1 , a 2 , b i and k are constants determined by multiple linear regression, are valid for jet fuels with initial boiling points in the range 150–190°C and final boiling points in the range 230–250°C, and for diesel fuels with initial and final boiling point ranges of 190–230°C and 320–370°C respectively. Properties investigated include smoke point, aromatics content, hydrogen content, density and freezing point for jet fuels, and cloud point, pour point, aniline point, hydrogen content, density, cetane index and cetane number for diesel fuels. Except for the low-temperature properties (jet fuel freezing point and diesel fuel cloud and pour points), an even simpler model of the form previously developed for jet and diesel fuels of fixed boiling range, viz. P = a 1 C n + a 2 C ar + k are also adequate.

High-peroformance liquid chromatographic procedure for direct determination of the compound class composition of diesel and kerosine fuels using refractive index detection

Abstract A High-performance liquid chromatography method utilizing a bonded aminosilane stationary phase, n -heptane mobile phase and refractive index detection is described for quantification of the compound class composition (saturates, single-ring aromatics, double-ring aromatics, polynuclear aromatics and polars) of olefin free kerosine and diesel fuels. Refractive index response factors for each of the compound classes have been determined using fractions obtained from preparative scale column chromatographic separation for a range of fuel samples. Variations in response factors for fractions derving from different fuel sources are described and are considered sufficiently small to allow acceptable analytical accuracy for refractive index detection.

1H and 13C n.m.r. spectroscopic methods for the analysis of fossil fuel materials: Some novel approaches☆

Abstract A number of 1 H and 13 C n.m.r. methods, apart from conventional procedures which rely on chemical shift classifications alone, can be used to identify chemical or structural characteristics of fossil fuel materials. The information available via multiplet selection in 13 C n.m.r., two-dimensional J spectroscopy, homonuclear 1 H- 1 H correlation spectroscopy, and heteronuclear 13 C- 1 H correlation procedures, is discussed and illustrated.

Property−composition relationships for diesel and kerosene fuels

Abstract A range of 18 diesel fuels and 21 kerosene fuels from mainly Australian petroleum and synthetic fuel sources, including coal, shale and peat, was investigated. Compositional details were defined as the weight per cent abundances of n -alkanes, branched plus cyclic saturates, single-ring aromatics, doublering aromatics and polynuclear aromatics, using both h.p.l.c. and 13 C n.m.r. techniques. Relationships between fuel composition and a range of fuel properties were sought. Simple linear relationships between property values and compositional data were used. Explicit correlative expressions were derived using multiple linear regression analysis, with the coefficient of multiple determination, R 2 , indicating the quality of the fit between observed and calculated property values. In most cases good correlations were achieved. For diesels the properties investigated, with R 2 values in parentheses, were: inverse specific gravity (0.99); 13 C n.m.r. aromaticity (0.99); 1 H n.m.r. aromaticity (0.88); cetane index (0.97); aniline point (0.96); diesel index (0.98); and FIA-measured aromatics content (0.77). For kerosenes the properties, with R 2 values in parentheses, were: 13 C n.m.r. aromaticity (0.98); 1 H n.m.r. aromaticity (0.97); smoke point (0.88); and FIA-measured aromatics content (0.94). The results are shown to be of value in assessing the potential and limitations of hydrotreating as a process for upgrading synfuels.

Quantitative estimation of CHn group abundances in fossil fuel materials using 13C n.m.r. methods

Abstract Comments on the use and relative merits of gated spin echo (GASPE) and spin echo broad band off-resonance decoupling (SEBBORD) selected multiplet 13 C n.m.r. methods, for the identification and quantitation of CH n ( n = Q −3) groups in fossil fuel materials, are given in response to recently published reports.

Investigation of aromatic carbon sites in materials derived from petroleum and coal using 13C n.m.r. methods

Aromatic C and CH carbon sites in a variety of petroleum and coal derived materials have been investigated using a 13C n.m.r. technique termed spin echo broad band off-resonance decoupling (SEBBORD). Only resonances due to non-protonated aromatic carbon sites are observed in SEBBORD spectra such that comparison with conventional 13C n.m.r. spectra enables differentiation between aromatic C and CH group resonances. Relative abundances of non-protonated aromatic carbon sites calculated from SEBBORD data are in good agreement with values derived from a combination of conventional1H n.m.r., 13C n.m.r. and elemental analysis data. The occurrence of significant proportions of aromatic C intensity to high field of 129–130 ppm and of aromatic CH intensity to low field of 129–130 ppm has been found to be quite common. Consequently attempts to determine aromatic C and CH group abundances by partitioning conventional 13C n.m.r. spectra in the vicinity of 129–130 ppm can lead to considerable quantitative errors. SEBBORD provides more detailed information about aromatic carbon sites than can be obtained from conventional 13C n.m.r. spectra.

A novel semi-empirical relationship between aromaticities measured from 1H and 13C n.m.r. spectra

Abstract An empirical correlation between 1H n.m.r. measured aromaticity (Har) and 13C n.m.r. measured aromaticity (Car) has been found. Over 300 samples which are diverse in origins and structural characteristics have been studied. A semi-empirical functional relationship between Har and Car has been devised. It is suggested that for distillable samples with, nominally, less than 50% boiling above 320 °C, the derived function can be used to rapidly and conveniently approximate Car from Har. The difference between observed and calculated Car values for 303 such samples shows a standard deviation of just 1.8%.

Structural characteristics of branched plus cyclic saturates from petroleum and coal derived diesel fuels

The branched plus cyclic saturates fractions from three coal-derived diesel fuels and two petroleum-diesel fuels have been characterized and their major structural features compared. Application of a multiplet selection 13C n.m.r. technique (GASPE) yields individual 13C n.m.r. subspectra for each CHn group (n = 0–3) and enables the fractional abundance of CHn groups to be measured. These data are used to elucidate a number of structural features. The coal-derived samples are composed predominantly of naphthenes consisting of 2–3 rings with approximately 1 sidechain per molecule. One of the petroleum samples (PET1) contains a high proportion of branched alkanes (≈60%) with, on average, 2 branches per molecule. The naphthene component of this sample is predominantly monocyclic with approximately 3 sidechains per molecule. The second petroleum sample (PET2) has a high naphthene content (>60%). Its structural features, which are more difficult to characterize, are found to fall between those observed for the coal derived samples and PET1.