Processing Methodology to Calculate the Carbon Aromaticity (Fa), Paraffinicity
(Fp), and Naphthenicity (Fn) from 13C NMR Data
NMR spectra were obtained on a Varian VXR-300S NMR spectrometer under the following
10 mm probe
angle, with gated decoupling, 10 second delay
FID must be weighted Fourier transformed with appropriate zero filling (128k).
must be carefully phased and base line corrected. Integrals should be cleared
and the integral scale set at 1000 (the integral scale can be set to any number
the operator wishes but must be maintained at that value throughout the
are defined as follows:
Region (165-100 ppm)
integral value = Ar
Region (75to –5 ppm)
integral value = Al
Aromaticity (Fa) = Ar/Ar+Al
1: Example of integral regions used to determine aromatic/aliphatic content.
and Naphthenicity Calculation:
next step is to zoom into the 80 to –5 ppm region of the spectrum and define
integrals wherever a paraffin resonance is found. It is assumed that all narrow
resonances are paraffinic, and that any obvious broader NMR peak groups that
represent a superposition of narrow paraffinic resonances are 100% paraffinic.
There is some error in this assumption but it cannot be avoided. Due to these
assumptions and the natural variance observed in oils with different chemistries
the definition of the integral positions must be left to the judgment of the NMR
operator. Here are two examples of oils processed to remove the naphthenic “hump”.
Figure 2 shows the definition of the paraffinic integrals and Figure 3 shows the
same integrals after a bc command is executed. The integrals are then cleared
and a new single integral obtained to define the parraffinic carbon (see Figure
4). The scale value of this integral must be identical to that used for the
integral found in Figure 4 represents the paraffinic carbon only (Ip). The
difference between the original “Total Aliphatic” integral and this new
paraffinic integral is the naphthenic carbon content.
Paraffinicity (Fp) = Ip / (Ar+Al)
Naphthenicity (Fn) = (Al – Ip) / (Ar+Al)
Figure 3: Effect
of baseline correction routine after integral definition (note flat baseline and
no dispersion in spectral features).
table below represents the integral cut points used for this sample. Note that
the exact number of integrals and the exact cut positions will vary slightly
from one sample to another.
Table I – Integral Cut Points
figures below show another base oil sample which required slightly different
integral definition and is an example of a spectrum that had to be processed
several times in order to avoid having the baseline correction cause serious
spectral distortion (“baseline negative” or dispersive regions of the
Aliphatic/Aromatic Integral Definition.
Figure 7: Integral
of paraffinic intensity (Ip) obtained after baseline correction. Note baseline
negative region at 24-26 ppm, and 40-42 ppm.
demonstrates an example of a poor naphthenic “hump” removal by the baseline
correction. When this occurs it will impact the quantitation of the
paraffinic/naphthenic carbon. When this occurs the spectrum should be Fourier
transformed again and the procedure performed again with slight changes to the
integral cut points in order to prevent the phenomenon shown in Figure 7.
a single methodology can be defined where cut points for integrals are strictly
definable. In all cases the NMR operator must use his best judgment to define
the integrals and decide when the baseline corrected spectrum is free from
distortions such as peak dispersions and/or “baseline-negative” regions.
A: 13C NMR Analysis is
Unvailable to Customer
order to ensure that the 13C carbon aromaticity predicted output is
correct we can utilize the empirical relationship between 1H proton
aromaticity (a number that can be derived directly from the process NMR 1H
spectrum) and 13C carbon aromaticity.
proton aromaticity is a value that is readily obtained from a 1H NMR
spectrum. In the case of the normalized integral data that is generated by the
process NMR analyzer it is simply a matter of adding the integral values
represented by points 30 to 60 to generate the proton aromaticity value. Figure
1 shows the area of the spectrum used to generate the proton aromaticity value.
There is a well-known non-linear correlation
between 1H NMR derived aromaticity and 13C NMR
derived aromaticity (see “A Novel Semi-Empirical Relationship Between
Aromaticities Measured from 1H and 13C NMR Spectra”,
David J. Cookson, C. Paul Lloyd, and Brian E. Smith, Fuel, 65, p
1247-1253, 1986.). Figure 2 shows the relationship usually observed.
while the NMR is operational the proton aromaticity number can be determined and
outputted into the DCS where an “estimated carbon aromaticity” value can be
calculated. This value can be used to confirm the “true carbon aromaticity”
value being outputted by the on-line PLS models. When the two values disagree by
more than X atomic%C (X to be agreed upon between parties) then the predicted
values shall be considered invalid. In the initial model building stage the
relationship between 13C aromaticity and the on-line proton
aromaticity will be developed in parallel with the PLS models.
checks on this system can be made by performing the actual 13C NMR
analysis at the PNA facility on a pre-determined schedule.
8: Region of NMR Spectrum
to be summed to yield proton aromaticity.
9: Empirical Relationship Between Proton and Carbon Aromaticity
B: 13C NMR is Available to Customer
Aromaticity parameter will be validated as per standard validation protocols.
For more information on Carbon Aromaticity, Naphthenicity, and
Paraffinicity contact John Edwards