SAUDI ARAMCO JOURNAL OF TECHNOLOGY SPRING 2010

Molecular Details of Hydrocracking Feedstocks

Authors: Adnan A. Hajji, Dr. Hendrik Muller and Dr. Omer R. Koseoglu

ABSTRACT

The fine molecular details of the two hydrocracking unit
feedstock streams, vacuum gas oil (VGO) and de-
metalized oil (DMO), from Riyadh Refinery were
explored by High Resolution Fourier Transform Ion
Cyclotron Resonance Mass Spectrometry (FT-ICR MS).
The objective of this advanced characterization study
was to gain in-depth molecular information about
aromatic hydrocarbons and their heteroatom derivatives
in these two feedstocks to understand the impact of their
compositions on process performance and catalyst
deactivation. It has been found that the DMO sub-
stantially differs from the VGO in the distribution of
heteroatom content and aromatic species. The DMO
contains heavier and more condensed aromatic species
than the heavy vacuum gas oil (HVGO). Both feedstocks
were found to contain multiple heteroatom species,
which have not been reported previously in such
feedstocks. Although both feedstocks were found to
contain species with up to three sulfur- and two nitrogen
atoms per molecule, the DMO is more concentrated with
nitrogen and sulfur containing heavy polyaromatic
molecules. Nitrogen is present in both feedstocks in
pyrrole- and pyridine-structures. Besides, numerous
multi-heteroatom containing species of sulfur and
nitrogen were found in both feedstocks. It is apparent
that the heavy nature and heteroatom content of DMO
are the main causes of the severe catalyst deactivation
observed in the hydrocracking unit, which needs to be
quantified.

INTRODUCTION

While specifications on transportation fuels tighten
worldwide
1
, the availability of light and sweet crude oils
continues to decline
2
. Therefore, the worlds demand of
heavy and sour crude oil is continuously increasing
1, 3, 4
,

Fig. 1.
Tighter sulfur specifications are mandated by
legislations worldwide, e.g., 15 parts per million in water
(ppmw) for transportation fuels in the USA
3
or 10 ppmw
in the EU
5
. To meet lower sulfur specifications in refined
products, highly efficient downstream sulfur removal
processes are needed. Therefore, desulfurization is a
widely used process in the refining of crude oil fractions
(naphtha, gas oil, vacuum gas oil (VGO), etc.) to
ultimately reduce emissions from combustion engines.
Hydrocracking is also a major process in the refining
industry for the conversion of VGOs to produce higher
yields of high quality mid distillates to meet the current
and future market demands. There is a strong interest
today in the processing of low value refractory feed-
stocks to improve the refining process economics. The
feedstocks refractory nature has been described by
several bulk properties, including final boiling point,
Micro-Carbon Residue (MCR) content, total sulfur, and
total nitrogen contents. The composition of feedstock is
an important parameter impacting the process conditions
(i.e., temperature, pressure, catalyst type and volume),
more useful in optimizing the operational parameters for
achieving the desired optimum product specifications.
Different analytical tools have been employed for a
detailed characterization of petroleum feedstocks. For
example, gas chromatography (GC) with sulfur specific
detectors is routinely applied
6, 7
. Unfortunately, GC is
restricted to the hydrocarbon fractions boiling below
370 C, whereas fractions boiling above 370 C exceed
the peak capacity of the GC
8
.
Ultrahigh resolution Fourier Transform Ion Cyclotron
Resonance (FT-ICR) Mass Spectrometry is rising
recently as a powerful technique for the analysis of
heavy petroleum fractions and whole crude oils
9-11
.
Particularly, two ionization modes have been success-
fully employed for the analysis of sulfur aromatic and
polar petroleum components: electrospray ionization
(ESI) and atmospheric pressure photo ionization
(APPI)
12, 13
.
FT-ICR has been used in this study to carry out
detailed characterization of the Riyadh Refinery
hydrocracking unit feedstocks, straight run heavy
vacuum gas oil (HVGO) and de-metalized oil (DMO),
which is derived from vacuum residue within a butane
solvent deasphalting unit.

EXPERIMENTAL

Samples

The HVGO and DMO were sampled from the vacuum
distillation tower and solvent deasphalting unit at the
Riyadh Refinery. The properties of the HVGO and DMO
feedstock components are given in Table 1. DMO, which
is derived from vacuum residues by solvent deas-
phalting, contains high amounts of impurities compared
to the VGO fractions. The metal content of DMO is at
acceptable levels that can be handled by utilizing
catalyst grading material at the top of the reactor.



SAUDI ARAMCO JOURNAL OF TECHNOLOGY SPRING 2010



Fig. 1. Sulfur content and API gravity for crude oil input to U.S. refineries from 1985 to 2007
4
.

Property Unit DMO HVGO
Density Kg/Lt 0.9716 0.916
Nitrogen ppmw 1655 725
Sulfur W% 3.17 2.19
Hydrogen W% 11.4 12.2
MCR W% 8.95 0.81
C5-Insolubles W% 0.48 0.039
Nickel ppmw 2 < 1
Vanadium ppmw 8 < 1

Table 1. Bulk properties of the hydrocracking feedstocks

Mass Spectrometry and Ionization Modes

Crude oils contain a mixture of tens of thousands of
acidic, basic and neutral heteroatom species (oxygen-,
nitrogen-, and sulfur-containing molecules). To study
these species, ultrahigh resolution mass spectrometry
was utilized with three different ionization modes:
negative and positive electrospray ionization (ESI+,
ESI-) and APPI for ionizing acidic, basic, and neutral
species, respectively
14
.
This is because most of these species have very
similar masses and, therefore, ultrahigh mass resolution
(R > 300,000) is essentially needed for the correct
assignment of their elemental compositions
(C
x
H
y
N
z
O
v
S
w
). The National High Magnetic Field
Laboratory at Florida State University, Florida, was used
to obtain the raw mass spectrometric data. The mass
spectrometry is a custom built FT-ICR
13
, which was
equipped with a 9.4 Tesla superconducting magnet
15
.
The FT-ICR experimental conditions were reported
elsewhere
12, 15-17
.

The high resolution mass data is recorded with
sufficient precision and mass accuracy to assign an
elemental composition to each mass signal in the
spectrum. The identified species are then grouped into
classes according to the heteroatoms in the elemental
composition, e.g., none = pure hydrocarbons, one sulfur
(or nitrogen) atom per molecule = mono-sulfur (or
nitrogen) species, or molecules with one nitrogen and
one sulfur atom = sulfur nitrogen species.

Double Bond Equivalent and Carbon Number
Diagram

We have used the double bond equivalent (DBE)
18
as a
convenient measure for the aromatic characteristics of a
molecule. The DBE is defined as the number of
hydrogen atoms lacking from an elemental composition
in comparison to the corresponding saturated molecule
with an otherwise identical number of carbon and
heteroatoms. Every double bond or ring in a molecule
reduces the number of hydrogen atoms by two. For
example, hexane has a DBE of zero as it is fully
saturated and contains no ring. Cyclohexane contains
one ring, and therefore features a DBE of one. Benzene
features one ring and three double bonds, and therefore
has a DBE value of four. The obtained mass spectro-
metric data are plotted as carbon numbers vs. DBE for
all species of each heteroatom class and for pure
hydrocarbon species. The size of each dot in these
diagrams reflects the underlying mass spectral signals
relative intensity. The mass spectrometric intensities
cannot be compared between different species and
ionization modes due to largely differing response
factors.

30
31
32
33
34
35
36
37
38
39
40
1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007
C
r
u
d
e

O
i
l

A
P
I

G
r
a
v
i
t
y

(
W
e
i
g
h
t
e
d

A
v
e
r
a
g
e
)
0
0.2
0.4
0.6
0.8
1
1.2
1.4
1.6
C
r
u
d
e

O
i
l

S
u
l
f
u
r

C
o
n
t
e
n
t

(
W
e
i
g
h
t
e
d

A
v
e
r
a
g
e
)
API Gravity
Sulfur
SAUDI ARAMCO JOURNAL OF TECHNOLOGY SPRING 2010



























Fig. 4. Modified van-Krevelen plot showing the sulfur species found in the HVGO (left) and DMO (right). The axes of the plot are the
C/H ratio, representing an increasing aromaticity of the sulfur species as they appear higher in the graph, vs. the sulfur over carbon
atoms ratio per molecule.

RESULTS AND DISCUSSION

Sulfur Species

The modified van Krevelen plots
16
, Fig. 4, present all the
data obtained for all sulfur containing species in the
HVGO (on the left) and DMO (on the right), rendering it
easy to distinguish species with one (mono), two (di), or
three (tri) sulfur atoms per molecule. The ratio of carbon
to hydrogen atoms as an indicator for the aromaticity of
each individual elemental composition vs. the ratio of
sulfur atoms to carbon atoms in each identified ele-
mental composition are plotted along the x and y axis,
respectively.
The mono-sulfur species are relatively more
concentrated than di- and trisulfur species in the HVGO.
Similar findings were reported for the Arabian Light (AL)
crude oil from which HVGO was distilled
18
. A significant
amount of disulfur species and some trisulfur species
have been identified in this work as well.
Mono-, di- and trisulfur species were also identified
in DMO. Unlike HVGO where mono-sulfur species
dominate, DMO is relatively more enriched in disulfur
and trisulfur species. The difference in composition is
well explained by the nature of these streams as DMO is
heavier than HVGO in terms of boiling point.
The aromaticity and number of sulfur atoms per
molecule apparently increase with an increasing boiling
point of the analyzed stream. Significant amounts of di-
and trisulfur species may explain the high total sulfur
content observed in the DMO. Whether the sulfur atoms
are present in two or three isolated thiophenic aromatic


rings, or as part of one larger condensed aromatic
system has yet to be investigated for Arabian oils
19
. This
molecular feature may be related to the observed
difficulties in the hydrocracking of the DMO. The sulfur
compounds are detailed as follows.

Mono-sulfur Species

There are 778 individual mono-sulfur species in HVGO
with combined intensities accounting for 24% of all
measured signals identified. These compounds range
from C
17
H
19
S

to C
48
H
74
S

with DBE extending from two to
24, corresponding to a thiophene with up to seven
aromatic rings per molecule with the remaining carbon
atoms as alkyl substituents on the ring.
In the DMO, 764 individual mono-sulfur species
were identified as shown in Fig. 5. Carbon atom
numbers range from 22 to 50 with DBE values extending
from seven to 32, corresponding to molecules with up to
11 aromatic rings (benzonapthothiophenes) with large
alkyl substituent that account for the remaining carbon
atoms.
Generally, it can be concluded that the DMO has
larger sulfur species with more aromatic rings (higher
DBE) than those in the HVGO.
It should be noted that the linear slanted edge
observed in the upper left side of the DBE vs. carbon
number graph of the DMO can be explained by a
maximal number of aromatic rings for a given number of
carbon and hydrogen atoms. The species at this C/H-
limit have no alkyl groups; all carbon atoms are part of
the aromatic ring system.

HVGO-APPI-S1-S3
S/C
0.00 0.02 0.04 0.06 0.08 0.10 0.12 0.14
C
/
H
0.0
0.5
1.0
1.5
2.0
DMO-APPI-S1-S3
S/C
0.02 0.04 0.06 0.08 0.10 0.12 0.14
C
/
H
0.0
0.5
1.0
1.5
2.0
S
1

S
2

S
3

S
2
S
3

S
1


SAUDI ARAMCO JOURNAL OF TECHNOLOGY SPRING 2010


Fig. 5. Carbon number vs. DBE values for the signals identified
as monosulfur compounds. Data is shown for the HVGO (top)
and DMO (bottom).

Disulfur Species

HVGO has 494 individual disulfur species identified with
combined intensity accounting for 9% of all measured
signals. The disulfur compounds range from C
14
H
23
S
2
to
C
43
H
64
S
2
with DBE values extending from seven to 24. A
DBE value of eight is consistent with structures
containing two thiophene rings and one benzene ring,
e.g., thiophenobenzothiophene.

The DMO contains 606 disulfur species with
combined intensity of 16% of all measured signals. The
carbon atoms and DBE ranges are consistent with four
to 11 aromatic rings per molecule, Fig. 6. The structures
of the S2 species are closer to the maximum limit of
aromaticity than the structures observed for the mono-
sulfur species. This implies that disulfur species are
more aromatic (higher DBE values) than the mono-sulfur
species.



Fig. 6. Carbon number vs. DBE values for the species
identified as disulfur compounds. Data is shown for the HVGO
(top) and DMO (bottom).


HVGO-APPI-S1
0
3
6
9
12
15
18
21
24
27
30
33
0 10 20 30 40 50 60
Carbon No.
D
B
E
DMO-APPI-S1
0
3
6
9
12
15
18
21
24
27
30
33
0 10 20 30 40 50 60
Carbon No.
D
B
E
HVGO-APPI-S2
0
3
6
9
12
15
18
21
24
27
30
33
0 10 20 30 40 50 60
Carbon No.
D
B
E
DMO-APPI-S2
0
3
6
9
12
15
18
21
24
27
30
33
0 10 20 30 40 50 60
Carbon No.
D
B
E
SAUDI ARAMCO JOURNAL OF TECHNOLOGY SPRING 2010


Trisulfur Species

Only 59 trisulfur species with a combined relative
abundance of 0.5% extending from C
20
H
19
S
3
to
C
41
H
61
S
3
, were identified in the HVGO, Fig. 7. Data on
trisulfur species is too sparse to draw further conclusions
for the HVGO. On the contrary, the DMO features a
significant number of trisulfur species (315), accounting
for 4.5% of all the measured signals. Their carbon
numbers and DBE of DMO trisulfur species correspond
to five (e.g., dithiophenodibenzothio-phene) to 10
aromatic rings per molecule.



Fig. 7. Carbon number vs. DBE values for the trisulfur
compounds. Data is shown for the HVGO (top) and DMO
(bottom).
Aromatic Hydrocarbons

The MCR of the HVGO and the DMO are 0.81 wt% and
8.95 wt%, respectively. This is reflected in the overall
characteristics of the individual aromatic hydrocarbon
molecular species in both fractions, Fig. 8.
Hydrocarbon molecules ranging from C
20
H
15
to
C
46
H
68
and from C
25
H
21
to C
48
H
53
and accounting for
about 6% each of all measured signals were identified in
HGVO and DMO, respectively, confirming that the DMO
is slightly heavier than the HVGO as expected. The size
of the aromatic hydrocarbons range from single ring to
seven aromatic rings for HVGO, whereas this number
increases from two rings to 10 aromatic rings for DMO.
These findings confirm that DMO is generally more
aromatic and therefore demands more severe
hydrocracking conditions.


Fig. 8. Carbon number vs. DBE values for the identified
aromatic hydrocarbon species. Data is shown for the HVGO
(top) and DMO (bottom).
HVGO-APPI-S3
0
3
6
9
12
15
18
21
24
27
30
33
0 10 20 30 40 50 60
Carbon No.
D
B
E
DMO-APPI-S3
0
3
6
9
12
15
18
21
24
27
30
33
0 10 20 30 40 50 60
Carbon No.
D
B
E
HVGO-APPI-HC
0
3
6
9
12
15
18
21
24
27
30
33
0 10 20 30 40 50 60
Carbon No.
D
B
E
DMO-APPI-HC
0
3
6
9
12
15
18
21
24
27
30
33
0 10 20 30 40 50 60
Carbon No.
D
B
E
SAUDI ARAMCO JOURNAL OF TECHNOLOGY SPRING 2010































Fig. 9. Modified van-Krevelen graph showing the nitrogen species found in the HVGO and DMO. The axes of the plot are the C/H ratio,
representing the aromaticity of the depicted species vs. N/C ratio.

Nitrogen Species

Nitrogen species of significantly varying polarities have
been described in petroleum
20
. Three different mass
spectrometric ionization techniques have been proven
powerful for ionizing these different types of nitrogen
molecules according to their polarities
21
. Negative
electrospray ionization (ESI negative) is selective for
molecules susceptible to deprotonation, e.g., acids and
pyrroles. ESI in positive mode has been used for
enhanced ionization of basic molecules through
protonation. Atmospheric pressure photoionization
(APPI) in positive mode produces both radical molecular
ions and protonated ions, e.g., neutral hydrocarbons,
thiophenic sulfur compounds and pyrrolic or pyridinic
nitrogen compounds, respectively.

Nitrogen Compounds as Detected by APPI

A large number of nitrogen containing species were
found in the HVGO and DMO. Figure 9 shows a
modified van-Krevelen contour plot of the ratio of carbon
atoms over hydrogen atoms (C/H) vs. the ratio of
nitrogen atoms over carbon atoms (N/C) for the nitrogen
species identified in the HVGO on the left and for the
DMO on the right. A higher C/H ratio indicates an overall
higher aromaticity per molecule. The distribution of N1
and N2 compounds in both feedstock streams can be
readily identified.
The mono- and dinitrogen compounds in DMO have,
on average, a higher aromaticity than their corres-
ponding species in the HVGO. Additionally, the DMO
seems to be relatively more enriched with dinitrogen
species. The overall molecular mass range of mono-
nitrogen species is slightly higher in the DMO. The mass
spectral data are in agreement with the bulk analytical
data that the DMO has an overall higher aromaticity
(higher DBE values) than that of the HVGO.
The total relative abundance of the N1 compounds
accounts for 14% of all species identified using APPI.
The measured molecular masses of the mono-nitrogen
species (N1) extend from 308 Da to 684 Da in the

Fig. 10. DBE vs. carbon numbers for mononitrogen species in
the HVGO and DMO as determined by APPI. The size of the
dots represents the relative intensity of the underlying mass
signal. The blue dots denote molecules that contain a nitrogen
atom in an assumed pyridine substructure and red dots denote
nitrogen compounds with a pyrrole substructure.
HVGO-APPI-N1
0
3
6
9
12
15
18
21
24
27
30
33
20 30 40 50 60
Carbon No
D
B
E
DMO-APPI-N1
0
3
6
9
12
15
18
21
24
27
30
33
20 30 40 50 60
Carbon No.
D
B
E
N/C
0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
N/C
0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08
C
/
H
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
SAUDI ARAMCO JOURNAL OF TECHNOLOGY SPRING 2010























Fig. 11. Modified van-Krevelen graph showing all nitrogen species found by ESI in the positive mode in the HVGO (left) and DMO
(right). The axes of the plots denote the C/H ratio which represents the aromaticity of the identified species vs. the N/C ratio.

HVGO. The 476 mono-nitrogen species were detected in
HVGO with a carbon number extending from 23 to 49.
DBE values range from 4 to 23, which corresponds to
nitrogen species with one aromatic ring (pyridine) to six
aromatic rings (e.g., benzonaphthoacridine) with a large
number of carbon atoms as a side alkyl group.
There were 476 mono-nitrogen species detected in
HVGO with empirical formula ranging from C
23
H
18
N
1
to
C
49
H
76
N
1
, corresponding to one aromatic nitrogen ring
(pyridine) to six aromatic nitrogen rings (e.g.,
benzonaphthoacridine), with different alkyl substituent on
the aromatic system. On the other hand, 597 mono (N1)
species were found in the DMO with a total relative
abundance of 11.5%. These N1 species correspond to
C
25
H
24
N
1
up to C
56
H
92
N
1
, with two to 10 aromatic rings
per molecule, e.g., dibenzoacridine with large alkyl
substituents on the aromatic system.
Both HVGO and DMO contain a significant number
of aromatic species with two nitrogen atoms per
molecule here referred to as dinitrogen species (N2).
The abundance of the N2 species in the HVGO is
relatively lower than those in the DMO (0.11% vs.
0.37%) and includes mostly pyridine-type species. The
dinitrogen hydrocarbons in the HVGO range from
C
33
H
46
N
2
to C
36
H
53
N
2
with four (e.g., benzocarbazole) to
seven aromatic rings and side alkyl carbons. The DMO
has, however, larger dinitrogen species, extending from
C
19
H
11
N
2
to C
44
H
23
N
2
with seven to 10 aromatic rings per
molecule.

Basic and Nonbasic Nitrogen Species

The mono-nitrogen compounds in both HVGO and DMO
can be grouped into two main categories of different
polarity: pyridine and pyrrole type nitrogen species
20
.
The first type is basic in nature, and is a known poison to
hydrotreating/hydrocracking catalysts, while the latter is
neutral, less reactive and hard to remove by
hydrotreating
22
.

The difference in polarities of these two nitrogen
types allows us to distinguish between them using APPI
mass spectra. Pyridine type nitrogen heterocycles are
susceptible to protonation during the ionization process
and form protonated ions. Therefore, pyridine type
nitrogen species feature integer DBE values, while
pyrroles form radical cations. Pyrrole type species
feature no integer DBE values in APPI mass spectra.
Both HVGO and DMO are found to be dominated by
basic pyridine-based aromatic species, Fig. 10. The
abundance of the pyridinic species accounts for 13.2%
and 10.0%, while that of the neutral pyrrolic species
account for 1.2% and 1.5% of all measured signals in
HVGO and DMO, respectively.
The pyrrole-based species in the HVGO correspond
to three (benzocarbazole) to seven aromatic rings
(pyrroles to carbazoles with further aromatic rings in the
molecule) with the elemental formula ranging from
C
28
H
31
N
1
to C
44
H
63
N
1
. The pyrrole-based species in the
DMO are relatively more abundant than in the HVGO
with five (e.g., dibenzocarbazole) to 11 aromatic rings
(carbazole benzologues) and side alky substituents of 29
to 45 carbon atoms with empirical chemical formula
extending from C
29
H
25
N
1
to C
45
H
57
N
1
. Generally, it can
be concluded that the DMO contains overall larger
pyrrole-type nitrogen structures with higher aromatic
characteristics (higher DBE values) than the HVGO.

Nitrogen Compounds as Identified by ESI Positive
Mass Spectral Mode

Pyridine is a common basic substructure in nitrogen
containing aromatic molecules in petroleum. Molecules
that contain a pyridine ring are selectively detected by
ESI positive mode due to their high polarity.
Figure 11 shows that the N2 species are relatively
more concentrated in the DMO, but N1 and N2 species
have similar aromaticity. No N3 species were detected in
both samples.
HVGO-posion-N1-N2
N/C
0.00 0.01 0.02 0.03 0.04 0.05 0.06
C
/
H
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
N2 N1
DMO-posion-N1-N2
N/C
0.00 0.01 0.02 0.03 0.04 0.05 0.06
C
/
H
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
N2
N1
SAUDI ARAMCO JOURNAL OF TECHNOLOGY SPRING 2010


Fig. 12. DBE vs. carbon numbers for mononitrogen species in
the HVGO and DMO detected by ESI in the positive mode. The
size of the dots represents the relative intensity of the
underlying mass signal.

Mono-nitrogen Species

The total relative abundance of the N1 compounds in the
HVGO accounts for 43.3% of all observed species, Fig.
12. The HVGO has 404 individual mono-nitrogen con-
taining species (likely acridines) with three (acridine) to
seven aromatic rings per molecule (dinaphthoacridine)
with side alkyl substituents on the aromatic systems.
The data obtained from the DMO reveal overall
higher carbon numbers and more aromatic N1 species
(higher DBE values) than from the HVGO. There have
been 490 mono-nitrogen species identified in the DMO,
corresponding to acridines with up to six benzyl rings in
the same molecule (e.g., dinaphthoacridine) with
different alkyl substituents on the ring system. It has to
be noted that all the detected N1 compounds in both
HVGO and DMO are found as protonated ions and it can
be, therefore, concluded that the nitrogen atom is
located in the pyridinic substructures.

Dinitrogen Species

There were 103 different dinitrogen compounds
identified in the HVGO with a total relative abundance of
1% of all signals. Three to six aromatic rings per
molecule with different alky carbons attached to the ring
system were identified, Fig. 13.


Fig. 13. DBE vs. carbon numbers for di-nitrogen species in the
HVGO and DMO detected by ESI in the positive mode. The
size of the dots represents the relative intensity of the
corresponding mass signals.
HVGO-Posion-N1
0
3
6
9
12
15
18
21
24
27
30
0 10 20 30 40 50 60 70
Carbon No
D
B
E
DMO-posion-N1
0
3
6
9
12
15
18
21
24
27
30
0 10 20 30 40 50 60 70
C No
D
B
E
HVGO-Posion-N1
0
5
10
15
20
25
30
0 10 20 30 40 50 60
Carbon No
D
B
E
DMO-poion-N1
0
5
10
15
20
25
30
0 10 20 30 40 50 60
Carbon No
D
B
E
SAUDI ARAMCO JOURNAL OF TECHNOLOGY SPRING 2010



































Fig. 14. Modified van-Krevelen plot showing the nitrogen species found in the HVGO (left) and DMO (right) by positive mode
electrospray ionization. The axes of the plot are the C/H ratio, representing an increasing aromaticity of the depicted species the higher
they appear in the graph, vs. the nitrogen over carbon atoms per molecule ratio.


Fig. 15. DBE vs. carbon numbers for mononitrogen species in the HVGO and DMO detected by ESI in the negative mode. The size of
the dots represents the relative intensity of the corresponding mass signals. The apparent gaps in the DMO data are due to measuring
artifacts.

There were 168 different N2 species identified in the
DMO, adding up to a total relative abundance of 2% of
all identified signals. The structures indicate three to
eight aromatic rings per molecule. From the data it is
apparent that the heavier DMO contains more N2
species with overall higher aromatic characteristics than
the HVGO.

Nitrogen Compounds Identified by ESI Negative
Mass Spectral Mode

ESI in the negative mode largely detects ions formed by
deprotonation. The negative mode, therefore, selectively

detects neutral pyrrole type nitrogen species over basic
pyridine type species, which are much less susceptible
to deprotonation. An aromaticity map of the neutral
nitrogen components in HVGO and DMO is displayed in
the modified van-Krevelen plots, Fig. 14.
There were 292 individual N1 species identified in
the HVGO, Fig. 15. The total relative abundance for all
mass signals is 41% with three (carbazole) to seven
aromatic rings, including a pyrrole substructure in each
molecule (e.g., dinaphthocarbazole).
There were 335 different N1 species with relative
abundance of 9% of all mass signals detected in the
DMO. Three to seven condensed aromatic rings per
HVGO-negion-N1
0
5
10
15
20
25
30
0 10 20 30 40 50 60 70 80
Carbon No.
D
B
E
DMO-negion-N1
0
5
10
15
20
25
30
0 10 20 30 40 50 60 70 80
C No.
D
B
E
HVGO-negion-N1-N2
X Data
0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07
Y

D
a
t
a
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
DMO-negion-N1-N2
N/C
0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07
C
/
H
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
SAUDI ARAMCO JOURNAL OF TECHNOLOGY SPRING 2010

molecule were identified, highly likely alkylated
carbazoles and more condensed analogs. These larger
aromatic systems in the DMO in comparison to the
HVGO are in agreement with the higher micro carbon
residue (MCR) value.

CONCLUSION

Hydrocracking feedstocks of HVGO and DMO were
characterized on the molecular level to understand their
behavior for the process. Detailed information on the
class and type of the hetero atom species and
implications for some chemical structures in processing
has been discussed. Mono-, di- and trisulfur species
were identified in both feeds, HVGO and DMO, with
disulfur species being the most relatively abundant sulfur
class in the DMO. The aromaticity (DBE) of the sulfur
species increases with the number of sulfur atoms in the
molecule: mono-sulfur species < disulfur species <
trisulfur species. The trisulfur species are overall close to
the maximum aromaticity: a high portion of the carbon
atoms build aromatic rings and only a much smaller
number of carbon atoms constitutes to alky substituents.
A large number of mono- and dinitrogen species have
been found in both feedstocks. The gained chemical
knowledge of their elemental composition allows us to
draw conclusions on the chemical structures surrounding
the nitrogen atom. Two main types of nitrogen species
have been identified in both HVGO and DMO: basic
pyridines and neutral pyrroles. Differentiation between
basic pyridine type species and neutral pyrrole type
nitrogen species on the molecular level is reported here
for the first time in these process feeds and helps us
understand the impact of the feedstock on the
hydrocracking process performance. Both investigated
feedstocks are dominated by pyridine-based aromatic
nitrogen species. The DMO contains a significant
amount of both types of nitrogen species, rendering it far
more difficult to process. Findings of this research work
will help to optimize the DMO ratio in hydrocracking
feedstock blend.

ACKNOWLEDGMENTS

The authors wish to thank Saudi Aramco management
for their support and permission to present the
information contained in this article. The authors would
also like to thank Dr. Ryan Rogers of the National High
Magnetic Field Laboratory, Florida State University,
Florida, and his group for providing the raw mass
spectral data. An extensive version of this paper was
published in Oil & Gas Science and Technology Rev.
IFP, Vol. 63, No. 1, 2008, p. 115.

REFERENCES

1. Whitehurst, D.D., Isoda, T. and Mochida, I: "Present
State-of-the-Art and Future Challenges in the
Hydrodesulfurization of Polyaromatic Sulfur

Compounds," Adv. Catalyst, Vol. 42, 1998, pp. 345-
471.

2. Ng, S., Zhu, Y., Humphries, A., et al.: "FCC Study of
Canadian Oil Sands Derived Vacuum Gas Oils. 2.
Effects of Feedstocks and Catalysts on Distributions
of Sulfur and Nitrogen in Liquid Products," Energy
Fuels, Vol. 16, No. 5, 2002, pp. 1,209-1,221.

3. EPA, U. (1999). Clean Air Act Tier 2.

4. Administration, E.I., 2007, Petroleum Navigator.

5. Parliament, EU, Directive of the European
Parliament and of the Council, 241 Final, European
Parliament, 2001.

6. Depauw, G.A. and Froment, G.F.: "Molecular
Analysis of the Sulfur Components in a Light Cycle
Oil of a Catalytic Cracking Unit by Gas
Chromatography with Mass Spectrometric and
Atomic Emission Detection," Journal of
Chromatography A., Vol. 761, Nos. 1-2, 1997, pp.
231-247.

7. Schade, T. and Andersson, J.T.: "Speciation of
Alkylated Dibenzothiophenes through Correlation of
Structure and Gas Chromatographic Retention
Indexes," Journal of Chromatography A., Vol. 1117,
No. 2, 2006, pp. 206-213.

8. Marshall, A.G., Hendrickson, C.L., Klein, G.F., et al.:
Fourier Transform Ion Cyclotron Resonance Mass
Spectrometry: The Platform for Petroleomics, paper
presented at the 4
th
NCUT/WRI Conference on the
Upgrading and Refining of Heavy Oil, Bitumen and
Synthetic Crude Oil Conference, Edmonton, Ontario,
Canada, September 25-27, 2006.

9. Hughey, C.A., Rodgers, R.P. and Marshall, A.G.:
Resolution of 11,000 Compositionally Distinct
Components in a Single Electrospray Ionization
Fourier Transform Ion Cyclotron Resonance Mass
Spectrum of Crude Oil, Analytical Chemistry, Vol.
74, No. 16, 2002, pp. 4,145-4,149.

10. Muller, H., Andersson, J.T. and Schrader, W.:
"Characterization of High Molecular Weight Sulfur
Containing Aromatics in Vacuum Residues Using
Fourier Transform Ion Cyclotron Resonance Mass
Spectrometry," Analytical Chemistry, Vol. 77, No.
8, 2005, pp. 2,536-2,543.

11. Choudhary, T.V., Malandra, J., Green, J., Parrott,
S. and Johnson, B.: "Towards Clean Fuels:
Molecular-level Sulfur Reactivity in Heavy Oils,"
Angew. Chemistry, Int. Ed Engl., Vol. 45, No. 20,
2006, pp. 3,299-3,303.

12. Hakansson, K., Chalmers, M.J., Quinn, J.P.,
McFarland, M.A., Hendrickson, C.L. and Marshall,
SAUDI ARAMCO JOURNAL OF TECHNOLOGY SPRING 2010


A.G.: "Combined Electron Capture and Infrared
Multiphoton Dissociation for Multistage MS/MS in
a Fourier Transform Ion Cyclotron Resonance
Mass Spectrometer," Analytical Chemistry, Vol.
75, No. 13, 2003, pp. 3,256-3,262.

13. Purcell, J.M., Hendrickson, C.L., Rodgers, R.P.
and Marshall, A.G.: "Atmospheric Pressure
Photoionization Fourier Transform Ion Cyclotron
Resonance Mass Spectrometry for Complex
Mixture Analysis," Analytical Chemistry, Vol. 78,
No. 16, 2006, pp. 5,906-5,912.

14. Robb, D.B., Covey, T.R. and Bruins, A.P.:
"Atmospheric Pressure Photoionization: An
Ionization Method for Liquid Chromatography-
Mass Spectrometry," Analytical Chemistry, Vol.
72, No. 15, 2000, pp. 3,653-3,659.

15. Billon, A., Morel, F., Morrison, M.E. and Peries,
J.P.: "Converting Residues with Ifp's Hyvahl(R)
and Solvahl(R) Processes," Oil & Gas Science
and Technology - Rev. IFP, Vol. 49, No. 5, 1994,
pp. 495-507.

16. Wu, Z., Rodgers, R.P. and Marshall, A.G.: Two
and Three Dimensional van Krevelan Diagrams: A
Graphical Analysis Complementary to the
Kendrick Mass Plot for Sorting Elemental
Compositions of complex Organic Mixtures Based
on Ultrahigh Resolution, Analytical Chemistry,
Vol. 76, No. 9, 2004, pp. 2,511-2516.

17. Koseoglu, O.R.: Process for Removal of Nitrogen
and Polynuclear Aromatics from Hydrocracker and
FCC Feedstocks, Patent application, Saudi
Arabia, Saudi Arabian Oil Company, 2006.

18. Mller, H., Hajji, A.A., et al.: Utilization of Fourier
Transform Ion Cyclotron Resonance Mass
Spectrometry to Determine the Effect of Upgrading
of Arabian Heavy Crude Oil on the Hetero Atom
Compounds, SAICS-ACS paper presented at the
2007 Chemindex, Manama, Bahrain, March 26-28,
2007.

19. Fu, J., Klein, G.C., Smith, D.F., et al.:
"Comprehensive Compositional Analysis of
Hydrotreated and Untreated Nitrogen-
Concentrated Fractions from Syncrude Oil by
Electron Ionization, Field Desorption Ionization
and Electrospray Ionization Ultrahigh Resolution
FT-ICR Mass Spectrometry," Energy Fuels, Vol.
20, No. 3, 2006, pp. 1,235-1,241.

20. Qian, K., Rodgers, R.P., Hendrikson, C.L.,
Emmett, M.R. and Marshall, A.G.: "Reading
Chemical Fine Print: Resolution and Identification
of 3,000 Nitrogen Containing Aromatic
Compounds from a Single Electrospray Ionization
Fourier Transform Ion Cyclotron Resonance Mass
Spectrum of Heavy Petroleum Crude Oil," Energy
Fuels, Vol. 15, No. 2, 2001, pp. 492-498.

21. Laredo, G.C., Leyva, S., Alvarez, R., Mares, M.T.,
Castillo, J. and Cano, J.L.: "Nitrogen Compounds
Characterization in Atmospheric Gas Oil and Light
Cycle Oil from a Blend of Mexican Crudes," Fuel,
Vol. 81, No. 10, 2002, pp. 1,341-1,350.

22. Cooper, B.H. and Knudsen, K.G.: Ultra Deep
Desulfurization of Diesel: How an Understanding
of the Underlying Kinetics Reduce Investment
Costs, Practical Advances in Petroleum
Processing by C.S. Hsu and P.R. Robinson,
Springer, New York, 2006, pp. 297-372.







































SAUDI ARAMCO JOURNAL OF TECHNOLOGY SPRING 2010

BIOGRAPHIES

Adnan A. Al-Hajji is a Research
Science Consultant at the Research
and Development Center (R&DC).
He is leading the Advanced
Hydrocarbon Characterization
Activity of the Clean Fuel project.
Adnan received both B.S. and
M.S. degree in Chemistry from the
King Fahd University of Petroleum and Minerals
(KFUPM), Dhahran, Saudi Arabia.
He has been with Saudi Aramco since 1976 and has
authored numerous publications.


Dr. Hendrik Muller is a Senior
Research Scientist at the Analytical
Services Division in the Research
and Development Center (R&DC).
He joined R&DC 4 years ago from a
post doc research position at the
National Institute of Standards and
Technology (NIST), MD.
His current role in the R&DC is as Analytical
Scientist for sulfur speciation, Petroleomics, and high
resolution mass spectrometry (FT-MS) at the Advanced
Analysis Unit. Hendrik is actively engaged in several
upstream and downstream research activities (clean
fuels, tight emulsions, oil to chemicals) as a team
member and analytical representative, and by
developing eight custom tailored analytical methods
(SALAMs) for current and future research projects.
He obtained his Ph.D. in Chemistry from Muenster,
Germany on the Molecular Characterization of Vacuum
Residues.
He has authored or coauthored 25 publications and
international conference presentations in his career; 15
of those (and one patent) are for Saudi Aramcos R&DC.




















Dr. Omer R. Koseoglu is a
Research Science Consultant at the
Research and Development Center
(R&DC).
Omer has more than 30 years of
experience in Refining Process
technologies. Prior to joining Saudi
Aramco, he worked for CONOCO Inc.
at the Technology Development Center in Ponca City,
OK, as a specialist on the upgrading/utilization of the
Fisher-Tropsch products, IFP North
America/Hydrocarbon Research Inc. (HRI) specializing
in fixed-bed and ebullated-bed
hydrotreating/hydrocracking technologies, and Shell
Canada Limited at the Oakville Research Center on
hydrocracking process research and development.
He has numerous publications and a diverse patent
portfolio and is a licensed professional engineer.
Omer has a Ph.D. degree in Chemical Engineering
from the University of Toronto, Canada, a M.Sc. degree
in Chemistry from Brock University, St. Catharines,
Ontario, Canada, and a B.Sc. degree in Chemical
Engineering from Gazi University, Ankara, Turkey.