This document describes using ion mobility and high-resolution GC/MS to analyze sulfur compounds in diesel fuel. Ion mobility provides an additional dimension of separation based on size, charge and shape of ions. This allows separation of isobaric compounds that co-elute during GC separation. The document compares untreated and treated diesel samples, identifying sulfur-containing compounds like dibenzothiophenes in the untreated sample that are removed by the desulfurization treatment process. Kendrick mass defect analysis is used to identify homologous series of compounds to better characterize the samples.

1. GC-APCI IMS of Diesel
Application Note
Authors
Sheher Bano Mohsin, David Wong,
and F. Robert Ley
Agilent Technologies, Inc.
Energy and Chemicals
Abstract
This Application Note describes the use of ion mobility and high-resolution
GC/MS for profiling sulfur compounds in a very complex sample such as diesel
fuel. The analysis of sulfur compounds in diesel can be challenging because of this
matrix complexity. Gas chromatography/high-resolution mass spectrometry is a
useful tool when identifying peaks, but, in the case of petroleum, such attempts
are usually frustrated by isobaric coeluting hydrocarbons.
Ion mobility, coupled to high-resolution GC/MS, provides an extra dimension
of separation based on size, charge, and shape. This study demonstrates the
advantage of coupling GC-APCI to ion mobility high-resolution mass spectrometry
for the study of aromatic sulfur compounds in diesel.

2. 2
Software
• Agilent MassHunter Data
Acquisition Software for Agilent
Ion Mobility Q-TOF Mass
Spectrometer, Version B.07.00
• Agilent MassHunter Qualitative
Software, Version B.07.00
GC-APCI
Figure 1 shows a schematic of the
GC-APCI IMS Q-TOF. The setup and
operation of the GC-APCI is described in
detail elsewhere1
. The source parameters
for GC-APCI are listed in Table 1.
Experimental
Sample preparation
Samples of processed and unprocessed
diesel were diluted 1:10 with isooctane,
and analyzed without any additional
sample preparation.
Instrumentation
• Agilent 7890B GC with the
GC-APCI interface
• Agilent 6560 Ion Mobility Q-TOF
GC/MS
Introduction
The economic value of diesel depends
on its chemical composition. Molecules
containing heteroatoms such as sulfur
lower the value of fuel, as they burn in air
and produce smog. The amount of sulfur
in diesel is being decreased because of
environmental regulations.
The distribution of sulfur-containing
compounds in various feedstocks is
important, as the refining process can be
modified to meet regulations. Refineries
use a process called hydrodesulfurization
to remove sulfur from sulfur-containing
compounds. For hydrodesulfurization,
sulfur bound in aromatic structures
such as benzothiophenes and
dibenzothiophenes and their
alkyl-substituted analogs is the most
difficult to remove catalytically. Therefore,
it is important to study these species.
The profiling of sulfur compounds in
diesel can be challenging because of its
very complex matrix. There are thousands
of different compounds in diesel, ranging
from alkanes, alkenes, aromatics, and
heteroatomic sulfur, nitrogen, and oxygen
compounds. The probability of completely
resolving a single class of compound from
all others in the sample is low.
Gas chromatography/high-resolution
mass spectrometry is a useful tool
when identifying peaks but, in
the case of diesel, such attempts
are usually frustrated by isobaric
coeluting hydrocarbons. Ion mobility,
coupled to GC/MS, provides an extra
dimension of separation based on size,
charge, and shape. After chromatographic
separation, the ions formed in the source
are separated again in the gas phase in
the mobility cell. Larger ions traverse
the drift tube at a slower rate than
smaller ions resulting in gas-phase ion
separation.
This Application Note demonstrates
the use of ion mobility for separating
aromatic sulfur-containing compounds
from the matrix. Two samples of
diesel are compared, before and after
hydrodesulfurization.
Table 1. GC and MS conditions.
Parameter Value
Column Agilent J&W DB-5ms, 30 m × 0.25, 0.25 µm
Injection port Split/splitless
Injection volume 1 µL
Mode Constant flow, 1.3 mL/min, 15 psi at 50 °C
Split ratio 1:10
Flow rate 1.3 mL/min
Oven program 50 °C (2 minutes) to 250 °C at 10 °C/min (10 minutes),
then 300 °C at 15 °C/min (10 minutes)
Table 2. Agilent 6560 Ion Mobility Q-TOF LC/MS source parameters for positive ion mode method.
Parameter Value
Ion mode Positive
Drying gas temperature 250
Drying gas flow 11 L
Capillary voltage 1,000 V
Skimmer 65 V
Corona 1 µA
Fragmentor 150 V
Resolution at 121 m/z
Figure 1. Schematic of the Agilent GC-APCI IMS Q-TOF system.
Ion drift tube
Rear ion funnelTrapping funnel
OctopoleHexapole
Quad Mass Filter (Q1)
Collision cell
Hi-res slicer
assembly
Hi speed
detector
assembly
Lenses
1 and 2
Ion pulser
High pressure
ion funnel
Drift tube entrance Drift tube exit
Drift IMSGC MS and MS/MS

3. 3
One of the most important attributes of
ion mobility separation is its specificity. In
particular, ion mobility effectively cleans
up background chemical noise due to
column bleed. The mobility heat map
shows hundreds of components in the
sample with overlapping compounds at
nearly every m/z value. The highlighted
polygon region in the heat map shows
the fingerprint due to column bleed.
The ion mobility of polysiloxanes is very
different from the hydrocarbon matrix,
and separate easily from sample-related
compounds. Figure 2 shows the screen
capture of these compounds.
The bottom graphic shows the mobility
filtered mass spectrum that removes
many of the overlapping chemical
background ions. This eliminates the
need to do any background subtraction.
resulting in gas-phase ion separation. The
ability to separate ions at the millisecond
time scale is particularly advantageous
when chromatographic separation of
complex mixtures is not possible. This
powerful combination of separation,
resolution, and structural information
is key for solving the most challenging
applications.
Results and Discussion
By coupling the GC to the 6560 IMS
Q-TOF, we created an analytical system
with much greater separation power
for these types of complex compound
analyses. The additional dimension of
orthogonal separation greatly adds to
the peak capacity/separation power of
the overall system. As we demonstrate,
ion mobility is effectively used to pull
apart overlapped and smaller peaks in a
complex mixture.
IMS Q-TOF
The Agilent 6560 IMS Q-TOF is a Q-TOF
LC/MS system that can be run in either
ion mobility mode or Q-TOF only mode. It
consists of a trapping funnel, drift tube,
and rear funnel between the front funnel
and ion optics leading to the quadrupole
mass filter. This design builds on the
well-established Q-TOF system in the
back end, and proven high-sensitivity
ion-funnel technology. The fast duty
cycle of ion mobility (milliseconds) is well
matched with the fast acquisition of TOF
(microseconds). The fast acquisition of
the TOF is well-suited to sampling the
very narrow peaks generated by the GC.
IMS is a gas-phase ion separation
technique that is carried out in a
drift tube filled with an inert drift gas
(nitrogen), and energized with a low DC
electric field. Larger ions traverse the drift
tube at a slower rate than smaller ions,
Figure 2. Compounds separated according to carbon number in the drift time domain. The circled area is column bleed, which
has shorter drift times than the hydrocarbons, and separates out

4. 4
The treated sample (Figure 4B) lacks
these peaks. Figure 4C shows the
mass spectrum of a region in the
chromatogram of the untreated sample
around 11.56 minutes retention time.
The peak at 213 m/z, with its isotopic
fit (superimposed in red), is indicative of
dibenzothiophene species.
The advantage of using high-resolution
mass spectrometry is apparent in
Figure 4, which shows a plot of an
extracted ion chromatogram with a mass
defect window set to mass defects below
0.1 amu. Figure 4A is the untreated
sample, which shows an abundance
of high-intensity peaks with the mass
defects below 0.1 amu that are indicative
of sulfur compounds.
Comparison of pre- and
post-treatment diesel samples
Figure 3 shows a comparison of the two
diesel samples pre- and post-treatment.
A visual inspection of the graph shows
differences in the distribution of
molecular weights in the two samples.
Sample C, which is the sample after
hydrodesulfurization, shows a molecular
weight distribution towards the lower
molecular weights expected from this
treatment.
600
500
400
300
200
100
0
0 2 4 6 8
Retention time (min)
Mass(Da)
10 12 14 16 18
Figure 3. Comparison of diesel samples before and after hydrodesulfurization. The features in blue are
from the diesel sample after treatment.
Figure 4. Extracting the dibenzothiophenes (mass defect below 0.1 amu)
×105
9.0 9.4 9.8
147.0267
155.0858
161.0421
169.1015
175.0586
183.1172
189.0739
192.0938
198.0504
213.0741
([C14
H12
S]+H)+
255.1636
231.1210
240.1870 247.1517 255.2105
10.2 10.6 11.0 11.4 11.8 12.2 12.6 13.0 13.4 13.8 14.2 14.6 15.0 15.4 15.8 16.2 16.6 17.0 17.4
0
1
2
3
4
A
B
C
Acquisition time (min)
Diesel before hyrdodesulfurization
Diesel after hyrdodesulfurization
Counts
×105
0
1
2
3
4
Counts
145 150 155 160 165 170 175 180 185 190 195 200 205 210 215 220 225 230 235 240 245 250 255 260
×105
0
1
2
3
4
Counts
Mass-to-charge (m/z)

5. 5
Using the Kendrick mass defect
plot allowed us to identify related
dibenzothiophenes. The drift spectra of a
series of related dibenzothiophenes are
shown in Figure 7.
We used mMass, a program available on
the Internet that automatically calculates
and plots Kendrick mass defect from raw
mass spectral data (Figure 6).
Analysis of dibenzothiophenes
Figure 5 shows possible structures for
dibenzothiophenes.
The use of Kendrick mass defect to
identify series of related compounds is
well established for complex samples3,4
.
The Kendrick mass scale, which defines
the mass of CH2
as exactly 14.0000, is
used to identify families of compounds
that differ only by the length of alkyl
substituents.
If we consider a homologous series of
molecules of the same class and DBE,
but varying degree of alkylation, members
of the series will vary in composition
by multiples of -CH2
, corresponding to
multiples in a mass of 14.01565 Da. A
particularly useful way to identify such
a series is to convert the IUPAC mass
to Kendrick mass. The Kendrick mass is
calculated using Equation 1.
Kendrick mass = IUPAC mass ×
(14.00000/14.01565)
Equation 1
Successive members of an alkylation
series will differ by 14.00000 in Kendrick
mass and, therefore, each will have the
same Kendrick mass defect (KMD). The
Kendrick mass defect is the difference
between nominal Kendrick mass and
exact Kendrick mass (Equation 2).
Kendrick mass defect (KMD) = Kendrick
mass – nominal Kendrick mass
Equation 2
Using this scheme, members of a given
alkylation series are easily identified,
and the assignments can be extended to
higher masses by extrapolation.
In a KMD plot, the Kendrick mass is used
for the X-axis and KMD is used for the
Y-axis. From left-to-right, the numbers of
CH2
units in the observed ions increase.
From top-to-bottom, the differences in
chemical formula, such as the degree
of unsaturation, oxidation, and so on,
increase.
Figure 5. Possible structures of dibenzothiophenes. Note: several structural isomers are possible for each
of these formulae.
C13
H10
S
CH3
S
C15
H14
S
CH3
S
C16
H16
S
CH3
S
H3
C
C14
H12
S
CH3
H3
C
S
C12
H8
S
S
Increasing CH2
184 198 212 226 240
254
Figure 6. Kendrick mass defect plot showing series of related dibenzothiophenes with the same
Kendrick mass defect.
Figure 7. Drift spectra of a series of related dibenzothiophenes.
CH3
H3
C
S
CH3
S
×104
10.5 11.0
11.13
11.77
12.8512.15
11.5 12.0 12.5 13.0 13.5
0
0.4
0.8
1.0
1.6
2.0
2.4
2.8
3.2
3.6
4.0
4.4
4.8
Counts
Drift time (ms)

6. 6
These unrelated interfering compounds
can be separated with ion mobility
(Figure 9).
Each dibenzothiophene peak has
interferences from other hydrocarbon
species. This interference is revealed
in the Kendrick mass defect plot, which
shows unrelated compounds with
the same nominal mass but different
KMD on the vertical axis (Figure 8).
Figure 8. Kendrick mass defect plot showing background and dibenzothiophene with their respective different
mass defects.
Figure 9. Ion mobility spectrum of dibenzothiophenes and other interfering hydrocarbons.

7. 7
Analysis of ASTM standards
In this study, we analyzed the ASTM
standard compound mixture. In a very
narrow GC peak at ~4.4 minutes, we
found over 10 compounds, including
many isomers in the mass range of 100
to 125 Da. The early eluters in the GC
chromatogram have more interference
and coelution of compounds in the low
molecular weight range. Ion mobility
provides higher peak capacity and much
more detailed information about the
individual species.
The Kendrick plot can also be used
to observe complex patterns in the
mass spectra of hydrocarbon species
(Figure 10). The horizontal line in
Figure 11 shows hydrocarbon species
differing by a CH2
unit but with the same
DBE value. The next series of compounds
on a horizontal line differ by 0.013, or 2H.
These patterns are easily separated in the
mobility domain, as shown in Figure 12.
Analysis of hydrocarbons
Generally, diesel samples contain many
series of hydrocarbon molecules that
overlap in the same mass range, which
can bring great challenges for data
interpretation using conventional MS
systems. By adding the ion mobility
dimension of separation and using the
data extraction feature in the Agilent IM
browser, we easily obtain information on
the ion series of interest for further study.
Figure 10. Hydrocarbon series differing by a CH2
unit. The finer pattern within the series differs by 2H.
×105
135 140 145
141.0700
145.1012
C12
H8
C13
H10
C14
H12
C15
H14
C16
H16
C17
H18
155.0856
159.1170
173.1326
183.1169
195.1170
199.1482 213.1638
227.1793
241.1949
150 155 160 165 170 175 180 185 190 195 200 205 210 215 220 225 230 235 240 245 250 255
0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
Counts
Mass-to-charge (m/z)
C15
H18
C15
H16
C15
H20
C15
H22
Figure 11. Kendrick mass defect plot for a series of related compounds with KMD of 0.1008.

8. 8
Figure 12. Ion mobility separation map for molecular weights from m/z 178 to 232.
Figure 13. Possible structural isomers for C8
H10
.
CH3
CH3
CH3
CH3
CH3
CH3
CH3
Possible isomers
1,2-Dimethylbenzene
-ortho-xylene
1,3-Dimethylbenzene
-meta-xylene
1,4-Dimethylbenzene
-para-xylene
Ethylbenzene

9. 9
References
1. Goodley, P.; Mohsin, S. B. The
GC-APCI Interface for the Agilent
Q-TOF LC/MS System Improves
Sensitivity, Mass Accuracy, and Speed
for a Wide Range of GC Applications.
Agilent Technologies, Inc. Technical
Overview, publication number
5991-2210EN, 2013.
2. Jody C. M.; et al. Conformational
Ordering of Biomolecules in the
Gas Phase: Nitrogen Collision
Cross Sections Measured on a
Prototype High Resolution Drift Tube
Ion Mobility-Mass Spectrometer.
Anal. Chem. 2014, 86 (4),
pp 2107-2116.
3. Kendrick, E. A mass scale based on
CH2
= 14.0000 for high-resolution
mass spectrometry of organic
compounds. Anal. Chem. 1963, 35,
pp 2146-2154.
4. Hughey, C. A.; et al. Kendrick mass
defect spectroscopy: A compact
visual analysis for ultrahigh-resolution
broadband mass spectra. Anal. Chem.
2001, 73, pp 4676-4681.
Acknowledgements
The authors would like to thank Ken
Imatani for his help and support.
Conclusions
Two samples of diesel fuel, before
and after hydrodesulfurization, were
compared. Aromatic sulfur-containing
compounds were differentiated from
hydrocarbons using high-resolution
mass spectrometry by filtering the data
on mass defect. Ion mobility offers an
extra dimension in separating isobaric
compounds based on drift time in this
very complex mixture.
Ion series corresponding to alkylated and
nonalkylated dibenzothiophenes were
clearly separated, and the individual
compounds were identified.
Kendrick mass defect plots were used
to recognize families of homologous
compounds. These plots allowed us to
extract the ion series of interest for the
analysis. Differential analysis of the two
samples quickly revealed differences in
the degree of sulfur removal between the
two samples.
Various structural isomers of aromatic
compounds coeluting early in the GC
chromatogram were also identified.

10. www.agilent.com/chem
This information is subject to change without notice.
© Agilent Technologies, Inc., 2015
Published in the USA, July 10, 2015
5991-5856EN