Quantitative and Qualitative Analysis of Three Classes of Sulfur Compounds in Crude Oil

Article abstract of DOI:10.1002/anie.201703205

Owing to the environmental hazards arising from sulfur-containing combustion products, strong legal regulations exist to reduce the sulfur content of transportation fuels down to a few ppm. With the ongoing depletion of low-sulfur crude oil reservoirs, increased technological efforts are needed for crude oil refining to meet these requirements. The desulfurization step is a critical part of the refining process but partly suffers from the recalcitrance of certain species to sulfur removal and the inability to quantitatively understand the behavior of individual compound classes during the process. We herein present a new and simple approach for the parallel quantification of three different classes of sulfur species present in crude oils by LC separation and on-line detection and quantification by ICP-MS/MS. This approach will help to estimate the amount of recalcitrant species and thus facilitate the optimization of desulfurization conditions during fuel production.

Full text of DOI:10.1002/anie.201703205

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Quantitative and qualitative analysis of three classes of sulfur
compounds in crude oil
Authors: Alessandro Vetere, Daniel Pröfrock, and Wolfgang Schrader
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.201703205
Angew. Chem. 10.1002/ange.201703205
Link to VoR: http://dx.doi.org/10.1002/anie.201703205
http://dx.doi.org/10.1002/ange.201703205

10.1002/anie.201703205
Angewandte Chemie International Edition
COMMUNICATION
Quantitative and qualitative analysis of three classes of sulfur
compounds in crude oil
Alessandro Vetere[a], Daniel Pröfrock[b] and Wolfgang Schrader[a]*
Dedicated to Professor Dr. Paul D. Schettler on the occasion of his 80th birthday
Abstract: Due to environmental hazards arising from sulfur
containing combustion products, strong legal regulations exist to
reduce the sulfur content of transportation fuels down to a few ppm.
With the ongoing depletion of low-sulfur crude oil reservoirs,
increased technological efforts are needed for crude oil refining to
meet these requirements. The desulfurization step is a critical part of
the refining process but partly suffers from recalcitrance of certain
species against sulfur removal and the inability to quantitatively
understand the behaviour of individual classes during the process.
Here a new and simple approach for the parallel quantification of
three different classes of sulfur species present in crude oils using
LC-separation and an online detection and quantification by ICP-
MS/MS is shown. This new and simple approach will help to
estimate the amount of recalcitrant species and thus to allow a better
optimization of desulfurization conditions during fuel production.
needs to be optimized and fine-tuned for any given feedstock
that is, in turn, characterized by the large variety of different
compounds present therein. This optimization, however, can
only be achieved when quantitative information on the total
amount of different sulfur species and their distribution into
relevant compound classes is available.
Scheme 1. Products of the hydrodesulfurization (HDS) process with
dibenzothiophene as an example.
To the best of our knowledge no single method has yet been
reported for the group selective quantification of sulfur or any
other group of compounds within a full crude oil. Total sulfur
content is commonly determined as a bulk parameter through
elemental analysis as part of feedstock characterization.
However, no molecular information regarding its chemical
surrounding is gained with this method.
Albeit strong efforts are undertaken to move from fossil to
renewable material based transportation fuels crude oil is
estimated to remain a major energy resource throughout the
next decades. Reservoirs of light and sweet (low-sulfur) crudes
are meanwhile mostly depleted and heavier and sourer (high-
sulfur) oils with sulfur contents of up to 14 wt.-% are shifting into
focus.[1] Sulfur containing compounds form SO2-3 on combustion
and have thus been mainly responsible for the formation of acid
rain.[2] Nowadays, strong legal requirements force oil refineries
to virtually remove sulfur from any transportation fuel. Limits in
the European Union as well as in the United States are as low
as 10 ppm.[3] The removal of sulfur by hydrodesulfurization
(HDS) arguably is one of the most widely used catalytic
reactions in industry. It uses hydrogen at elevated pressures and
temperatures on heterogeneous catalysts to yield gaseous H2S
and sulfur-free products.[4] The process, however, is rather
unselective and also results in the undesired hydrogenation of
aromatic moieties.[5] An example of the reaction and possible
byproducts is shown in Scheme 1.
Numerous studies have dealt with the characterization and
quantification of sulfur containing compounds in fossil materials.
These studies used mostly
a gas chromatographic (GC)
approach, some relying on molecular mass spectrometry.[6]
However, a full quantification of sulfur in a given crude oil
sample is not possible by this approach. Response factors of
single compounds differ largely from each other and cannot – as
often done – be assumed to be unity. Unfortunately also no
reference standards are available for most compounds found in
a crude oil since the number of unique compounds is expected
to exceed one million.[7] This problem can be avoided by sulfur
selective detection with uniform response because in such a
case the response is independent from the chemical
surrounding of a sulfur atom and only one reference standard is
needed. For this purpose the sulfur chemiluminescence detector
(SCD), the atomic emission detector (AED) or inductively
coupled plasma mass spectrometry (ICP-MS) have been used
after one- and two-dimensional GC separations.[8] However, only
volatile, thermally stable compounds are amenable to GC, while
a considerable portion of a crude oil is non-volatile with
molecular weights being well above 500 Da.[9]
Unfortunately, a number of sulfur containing compounds
withstand the process. Among these recalcitrant species are
mostly condensed thiophenic compounds, while non-condensed
thiophenes and aliphatic thiols or thioethers (sulfides) are
relatively easily converted.[4-5] To gain high quality fuels in a
cost-efficient and resource saving manner the HDS process
As a result, a method for the quantification of sulfur from a
whole crude oil that includes information on the chemical nature
of the analytes requires both a separation that ideally takes
place in the liquid phase and a selective detection that gives a
uniform response based on the amount of sulfur reaching the
detector.
[a]
[b]
Dipl.-Chem. A. Vetere, Prof. Dr. W. Schrader
Max-Planck-Institut für Kohlenforschung
Kaiser-Wilhelm-Platz 1, D-45470 Mülheim an der Ruhr, Germany
E-mail: wschrader@kofo.mpg.de
Dr. D. Pröfrock
Helmholtz-Zentrum Geesthacht
Institut für Küstenforschung
Here a simple, one-step alternative based on an HPLC
Max-Planck-Straße 1, D-21502 Geesthacht, Germany
separation
co-workers [9b, 10] is presented. They reported Ligand Exchange
Chromatography (LEC) on Pd(II) coated stationary
approach
developed
by
Andersson
and
Supporting information for this article is given via a link at the end of
the document.
a
This article is protected by copyright. All rights reserved.

10.1002/anie.201703205
Angewandte Chemie International Edition
COMMUNICATION
chromatographic phase for the separation of the aromatic
fraction of crude oils into sulfur-free, thiophenic and sulfidic
procedures with a total of seven single steps over the course of
about two weeks, each of which bearing the risk of errors. This
is a rather tedious and time consuming procedure, and therefore,
here a method is introduced that allows the analysis of different
sulfur containing classes directly after online separation by
HPLC. The separation performance of our protocol was verified
by online coupling of the LC system to an ultrahigh resolving FT
Orbitrap Elite mass spectrometer (Thermo Fisher Scientific,
Bremen, Germany). The chromatographic behavior of the sulfur
containing classes was monitored by piecewise summation of
the mass spectra obtained and assignment of elemental
compositions to the resulting peaks. This allows a thorough
characterization of the eluting compounds. At the mass resolving
power given, hydrocarbons are easily distinguished from sulfur
containing compounds, also in minute amounts. With the change
of mobile phase three main fractions were eluted, as expected
(Figure 2, top panel). The first fraction (F1), eluted between
2-15.5 min, mainly consists of sulfur-free hydrocarbons, with
sulfur containing compounds being only of minor abundance as
is shown in the class distribution in the bottom panel of Figure 2.
compounds
without
chemical
conversion.
Thiophenic
compounds are further separated, as non-condensed (1-ring)
thiophenes are reported to elute into the sulfur-free fraction,
hence the three groups of sulfur containing compounds that
differ most in recalcitrance towards HDS – thioethers, 1-ring
thiophenes and condensed thiophenes – can be analyzed
separately with one method.
For sulfur selective detection and quantification in the liquid
phase inductively coupled plasma mass spectrometry (ICP-MS)
is the method of choice as it provides a uniform response due to
all analytes being atomized within the plasma. Nonetheless,
sulfur quantification by ICP-MS has long been problematic due
to low ionization efficiency and molecular interferences on all
sulfur isotopes. Most problematic is the interference between the
+ [11]
main isotope 32S+ and 16O2 . A novel ICP-MS instrument that
includes a collision/reaction cell in a triple-quadrupole (QqQ)
setup allows avoiding such interferences.[12] Further details on
the principle can be found in the supporting information.
The method presented here combines for the first time the
liquid chromatographic (LEC) separation of sulfur containing
compound classes from a full crude oil with powerful online
detection by different types of mass spectrometry. This allows
qualitative assignments of the molecular species by ultrahigh
resolution mass spectrometry using direct coupling to an
Orbitrap analyzer[13] and additionally, online sulfur selective
detection and quantification by ICP-MS/MS. Prior to
quantification by ICP-MS/MS the separation conditions were
altered from the original protocol and verified by online HPLC-
FTMS using photoionization in a separate experiment. The
general experimental setup is depicted in Figure 1. Details on
alterations of the original LC-protocol can be found in the
supporting information.
Figure 1. Experimental setup for the course of this study, solid and dashed
routes are alternatives. Qualitative analyses were performed by hyphenation
of the HPLC separation to APPI Orbitrap MS, quantitative analyses were run
using ICP-MS/MS. Calibration for ICP-MS/MS and loop injections of the full
samples were done after replacing the LC column with a Zero-Dead-Volume
(ZDV) union. In all cases the mobile phase flow was split 1:20 and the larger
flow routed to a UV detector for alignment of the mass spectrometric datasets.
Figure 2. Qualitative evaluation of the LC separation. The top panel shows the
UV-chromatograms (dotted and dashed) and the Total Ion Chromatogram
(TIC) obtained for Orbitrap-MS (solid). Arrows indicate fraction boundaries
containing the distinct compound groups, the amount of THF in the mobile
phase is indicated by the dash-dotted line (right axis). The bottom panel shows
intensity based class distributions throughout the given fractions. Note that
with the given setup, sulfidic species in fractions 3 and 4 are mostly detected
as oxidized, protonated ions ([M+OH]+ and [M+O2H]+).
Wang and co-workers recently introduced an alternative
approach to separate thiophenic from sulfidic compounds.[14]
Their protocol, however, included several derivatization
This article is protected by copyright. All rights reserved.

10.1002/anie.201703205
Angewandte Chemie International Edition
COMMUNICATION
These species include non-condensed thiophenic species with
double bond equivalents (DBE) of 3 or higher (see also top
separation. The calibration curve spanned a range between 2
and 34 ng sulfur per injection, with the linear correlation
coefficient being R = 0.998. Loop injections were performed in
duplicate. The overall sulfur content of the whole crude oil was
found to be 0.475±0.007 wt.-%. The result was confirmed by
panel in Figure 3). The DBE
4 series corresponds to
tetrahydrobenzothiophenes, the next high abundance DBE
series 7 corresponds to thiophenes with an additional non-
condensed phenyl-group or four naphthenic rings in a side chain. standard addition of DBT to the sample, resulting in a content of
The second fraction (F2, 15.5-27.5 min), eluted by the
addition of 4 % THF, contains condensed thiophenic species
(DBE 6 and higher), with local maxima for benzothiophenes
0.48±0.03 wt.-%.
(DBE
6),
dibenzothiophenes
(DBE
9)
and
benzonaphthothiophenes (DBE 12).
For the remainder of the chromatographic run sulfidic
species are eluted, as can be seen from the changed DBE
distribution (Figure 3). The last fraction (27.5-39.5 min) can be
divided into two sub-fractions. Compounds with DBE 8 or 9 are
eluting first (F3, 27.5-37 min). These correspond to thioethers
with two phenyl substituents on the sulfur that are more easily
released due to a lower nucleophilicity compared to thioethers
with purely aliphatic (non-aryl) substituents. In fact, the most
abundant signal here represents diphenyl sulfide (C12H10S).
Figure 4. LC-ICP-MS/MS chromatograms obtained after separation of the
crude oil sample. The dotted trace represents the UV-chromatogram at
287 nm used for alignment of the datasets. Additionally, the S (solid) and Pd
(dashed) selective chromatograms obtained by ICP-MS/MS are shown.
Arrows indicate fraction boundaries containing the distinct compound groups.
For quantification of the separated groups of sulfur containing
species the LC column was introduced into the system as shown
in Figure 1 and the crude oil sample (1,000 µg mL-1) was
injected onto the column. As depicted in Figure 4 the sulfur
selective chromatogram of the ICP-MS/MS follows the UV trace
(and therefore also the Orbitrap MS data) as expected. The area
of the first peak, representing 1-ring thiophenes, is relatively
small and therefore magnified for visibility. This is expected as
non-condensed thiophenes are typically not very abundant
Figure 3. Intensity based DBE distributions of the sulfur containing classes
throughout the different fractions.
Sulfides with no direct attachment of aromatic groups are
predominantly eluted by very high concentrations of competitive
ligand (THF) during minutes 37-39.5 (F4). Here, DBE series 1-3
dominate, indicating the aliphatic nature of the corresponding
compounds. Common among the later fractions is that sulfur
containing compounds are mostly detected as oxidized,
protonated species. Sulfoxides are not typically abundant in
crude oils and also data from measurements using electrospray
ionization do not suggest that such species are present in high
amounts in this sample. It has previously been shown that easily
oxidizable compounds (such as sulfides) are susceptible to in-
source oxidation when using Atmospheric Pressure Photo
Ionization (APPI).[15]
Focus here was the group-type resolved quantification of
sulfur in a whole crude oil in a one-step analysis, using liquid
chromatographic separation. For external calibration the
chromatographic column was replaced by a ZDV union, such
that injection loop fillings of dibenzo[b,d]thiophene (DBT) were
directly introduced into a continuous flow of mobile phase
(100 % A). The same setup was used for direct loop injection of
the full crude oil sample (25 µg mL-1) without chromatographic
compared with (di-)benzothiophenes.
A
major difference
between the sulfur selective trace in ICP-MS/MS and the
chromatograms obtained by HPLC-UV and HPLC-FTMS is an
additional sharp increase in intensity at 39.5 min with the mobile
phase being 100 % THF. This signal coincides with the
appearance of significant amounts of palladium (see dashed
trace in Figure 4) in the effluent. It seems that small amounts of
Pd are released from the stationary phase when using such high
amounts of competitive ligand. However, the sulfur eluted along
with the Pd cannot be attributed to the injected sample. No
relevant amounts of corresponding species are found during
qualitative FTMS analysis. Also is the overall amount of sulfur
detected after 39.5 min by far exceeding expectations. This
portion alone would account for a sulfur recovery of 342 %. One
suggestion is that a considerable portion of the sulfur detected
after 39.5 min originates from sulfur directly bound to the Pd-ion,
such as the used mercaptopropano linker. For these reasons
This article is protected by copyright. All rights reserved.

10.1002/anie.201703205
Angewandte Chemie International Edition
COMMUNICATION
integration of the sulfur trace was only performed until the valley
before significant Pd-elution begins (39.5 min), giving a total
sulfur recovery of 95.8 %.
used for a determination of the total sulfur content of the sample,
the chromatographic approach followed here can be used to
easily and quantitatively distinguish between different groups of
sulfur containing species. Still, further investigations should be
undertaken to avoid the release of palladium from the stationary
phase, which should also lead to a prolonged lifetime of the
chromatographic columns.
Overall integration results, given in Table 1, show that the
total amount of sulfur eluted from this sample is mainly present
in sulfidic species (79.8 %), followed by condensed thiophenes
(19.1 %). Only a minor amount of sulfur is present as non-
condensed thiophenes (1.1 %). Compared with loop injection
results 95.8 % recovery of sulfur compounds was achieved. A
few, though typically minor compound classes cannot be
addressed by this method. Thiols and disulfides have been
reported to be irreversibly retained by the column and will
therefore contribute to the non-recovered portion of the crude.
However, both classes as well as the non-eluting sulfides can be
considered as easily desulfurizable.
Experimental Section
Experimental details can be found in the supplementing information.
Acknowledgements
The authors thank Dr. David Stranz (Sierra Analytics, Modesto,
CA, USA) for access to new Software for the calculation of
complex mass spectra.
Table 1. Quantitative results of sulfur content after direct loop injection (total
sulfur) and after LC fractionation into different compounds classes (F1-4). LC
recoveries are compared with loop injection results. Values are wt.-% sulfur
within the sample.
Keywords: Analytical Methods • Crude oil • LC-MS • Sulfur
Total
F1[a]
F2[b]
F3[c]
F4[c]
Recovery Recovery
Quantification
sulfur
(wt-%)
wt-% S
wt-% S
wt-% S
wt-% S
(F1-3)
(F1-4)
[1]
[2]
J. Andersson, Chem. Unserer Zeit 2005, 39, 116-120.
a) G. E. Likens, F. H. Bormann, Science 1974, 184, 1176-1179; b) J. N.
Galloway, Z. Dianwu, X. Jiling, G. E. Likens, Science 1987, 236, 1559-
1562.
0.475
±0.007
0.005
±0,001
0.087
±0,001
0.063
±0,001
0.300
±0,001
32.7%
±0.6%
95.8%
±0.8 %
[3]
a) Vol. 98/70/EC, L ed. (Ed.: E. Parliament), Official Journal of the
European Communities, 1998; b) Vol. CFR, Title 40, Chapter I ,
Subchapter C, Part 80, Subpart O, Section 80.1603.
[a] Non-condensed (1-ring) thiophenes. [b] Condensed thiophenes. [c] Sulfides.
[4]
[5]
[6]
R. Javadli, A. de Klerk, Appl. Petrochem. Res. 2012, 1, 3-19.
Y. Sun, R. Prins, Angew. Chem., Int. Ed. Engl. 2008, 47, 8478-8481.
a) M. Li, T. G. Wang, B. R. T. Simoneit, S. Shi, L. Zhang, F. Yang, J.
Chromatogr. A 2012, 1233, 126-136; b) H. Muller, F. M. Adam, S. K.
Panda, H. H. Al-Jawad, A. A. Al-Hajji, J. Am. Soc. Mass Spectrom.
2012, 23, 806-815.
The direct quantification of individual species in crude oil suffers
from the complexity given by more than one million distinct
chemical compounds present in such samples. In general,
quantification requires
a pure signal and an individually
[7]
[8]
J. Blomberg, P. J. Schoenmakers, J. Beens, R. Tijssen, J. High Resolut.
Chromatogr. 1997, 20, 539-544.
determined response factor for each compound of interest. Here,
for the first time the quantitative analysis of compound groups
present in a complex crude oil is presented by combining
uniform response detection via ICP-MS/MS with molecular
detection using APPI-Orbitrap MS, both utilizing separation
capabilities from liquid chromatography.
The online HPLC-FTMS data gained by FT Orbitrap mass
spectrometry clearly show a successful separation of sulfur
containing species into three main groups of compounds namely
non-condensed thiophenes, condensed thiophenes and
thioethers. Overall, the results shown here present a new fast
and group selective quantification of sulfur present in a whole
crude oil. The method does not require any sample pretreatment,
other than dilution, and is additionally capable of analyzing also
non-volatile and/or thermally labile compounds. The information
gained by this technique should enable processing refineries to
optimize their desulfurization workflows and thus support the
manufacturing of high quality fuels at economical costs with
lower environmental contamination.
a) H. Behbehani, M. K. Andari, Pet. Sci. Technol. 2000, 18, 51-61; b) R.
Hua, Y. Li, W. Liu, J. Zheng, H. Wei, J. Wang, X. Lu, H. Kong, G. Xu, J.
Chromatogr. A 2003, 1019, 101-109; c) R. Hua, J. Wang, H. Kong, J.
Liu, X. Lu, G. Xu, J. Sep. Sci. 2004, 27, 691-698; d) C. López García, M.
Becchi, M. F. Grenier-Loustalot, O. Païsse, R. Szymanski, Anal. Chem.
2002, 74, 3849-3857; e) S. G. Mössner, S. A. Wise, Anal. Chem. 1999,
71, 58-69; f) B. Bouyssiere, P. Leonhard, D. Profrock, F. Baco, C.
Lopez Garcia, S. Wilbur, A. Prange, J. Anal. At. Spectrom. 2004, 19,
700-702.
[9]
a) S. K. Panda, J. T. Andersson, W. Schrader, Angew. Chem., Int. Ed.
2009, 48, 1788-1791; b) H. Muller, J. T. Andersson, W. Schrader, Anal.
Chem. 2005, 77, 2536-2543.
[10] a) A. Japes, M. Penassa, J. T. Andersson, Energy Fuels 2009, 23,
2143-2148; b) T. Nolte, T. N. Posch, C. Huhn, J. T. Andersson, Energy
Fuels 2013, 27, 97-107.
[11] J. Giner Martínez-Sierra, O. Galilea San Blas, J. M. Marchante Gayón,
J. I. García Alonso, Spectrochim. Acta, Part B 2015, 108, 35-52.
[12] O. Hanousek, M. Brunner, D. Profrock, J. Irrgeher, T. Prohaska, Anal.
Methods 2016, 8, 7661-7672.
[13] A. Vetere, W. Schrader, ChemistrySelect 2017, 2, 849-853.
[14] M. Wang, S. Zhao, K. H. Chung, C. Xu, Q. Shi, Anal. Chem. 2015, 87,
1083-1088.
Despite the high signal intensity observed when washing the
column with 100 % THF the integration results using the
chromatographic data still agree well with the results from the
loop injection measurements. Thus, when loop injections are
[15] S. K. Panda, K.-J. Brockmann, T. Benter, W. Schrader, Rapid Commun.
Mass Spectrom. 2011, 25, 2317-2326.
This article is protected by copyright. All rights reserved.

10.1002/anie.201703205
Angewandte Chemie International Edition
COMMUNICATION
Entry for the Table of Contents (Please choose one layout)
Layout 2:
COMMUNICATION
Alessandro Vetere[a], Daniel Pröfrock[b]
and Wolfgang Schrader [a]
*
Page No. – Page No.
Quantitative and qualitative analysis
of three classes of sulfur compounds
in crude oil
Sulfur in oil: Desulfurization of crude oil is probably the mostly used catalytic
reaction with the highest output in industry. Still, the optimization of such a reaction
is not based on molecular information but on bulk parameters avoiding any
structural knowledge. Here, a detailed quantitative and qualitative analysis is
reported that allows optimizing such important processes based on molecular
information.
This article is protected by copyright. All rights reserved.