Molecular Details of Hydrocracking Feedstocks

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 decline2. Therefore, the worlds demand of

heavy and sour crude oil is continuously increasing1, 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 EU5. 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.


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 (CxHyNzOvSw). 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 magnet15

. 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

Cru

de O

il API

Gra

vity

(Wei

ghte

d Ave

rage

)

0

0.2

0.4

0.6

0.8

1

1.2

1.4

1.6

Cru

de O

il Su

lfur C

onte

nt (W

eigh

ted

Ave

rage

)

API Gravity

Sulfur


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 C17H19S to C48H74S 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

S1 S2

S3 S2 S3 S1


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 C14H23S2 to C43H64S2 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.

DB

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.

DB

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.

DB

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.

DB

E


Trisulfur Species Only 59 trisulfur species with a combined relative abundance of 0.5% extending from C20H19S3 to C41H61S3, 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 C20H15 to C46H68 and from C25H21 to C48H53 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.

DB

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.

DB

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.

DBE

DMO-APPI-HC

0

3

6

9

12

15

18

21

24

27

30

33

0 10 20 30 40 50 60

Carbon No.

DBE


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

DBE

DMO-APPI-N1

0

3

6

9

12

15

18

21

24

27

30

33

20 30 40 50 60

Carbon No.

DBE

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


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 C23H18N1 to C49H76N1, 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 C25H24N1 up to C56H92N1, 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 C33H46N2 to C36H53N2 with four (e.g., benzocarbazole) to seven aromatic rings and side alkyl carbons. The DMO has, however, larger dinitrogen species, extending from C19H11N2 to C44H23N2 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 C28H31N1 to C44H63N1. 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 C29H25N1 to C45H57N1. 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


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

DBE

DMO-posion-N1

0

3

6

9

12

15

18

21

24

27

30

0 10 20 30 40 50 60 70

C No

DBE

HVGO-Posion-N1

0

5

10

15

20

25

30

0 10 20 30 40 50 60

Carbon No

DBE

DMO-poion-N1

0

5

10

15

20

25

30

0 10 20 30 40 50 60

Carbon No

DBE


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.

DB

E

DMO-negion-N1

0

5

10

15

20

25

30

0 10 20 30 40 50 60 70 80

C No.

DB

E

HVGO-negion-N1-N2

X Data

0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07

Y Da

ta

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


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,


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.


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.

Molecular Details of Hydrocracking Feedstocks

Documents

Transcript of Molecular Details of Hydrocracking Feedstocks