Title: A Novel Method for Chemical Fingerprinting of Oil and Petroleum based products based on Electrospray Mass Spectrometry and Chemometrics
1A Novel Method for Chemical Fingerprinting of Oil
and Petroleum based products based on
Electrospray Mass Spectrometry and Chemometrics
- Ingvar Eide and Kolbjorn Zale
- Energy and Fuels, 2006
- Statoil Research Centre, Norway
- Revised January 4, 2004
2Overview
- Intro
- Petroleum
- Mass Spectrometry
- ESI-MS
- Chemometrics
- Methods
- Results
- Conclusions
- Future Implications
3Background Info.
- Aim of Study Develop a quick and simple method
for fingerprinting of oil products. - IR and GC-MS were studied.
- IR Spectra cannot be resolved into individual
mass numbers. - GC-MS
- Interpretation of spectra was very complicated.
- Large molecules with high boiling points.
- Combined Platform of ESI-MS Methodology and
Chemometrics.
4Petroleum
- 50-80 of Petroleum is Hydrocarbons
- N-alkanes
- Cycloalkanes
- Aromatics
- The amount of branched
alkanes is equal to or greater
than the n-alkane content in petroleum. - Branches are typically methyl groups
- Cycloalkanes Largest group in Petroleum
- Typically cyclopentanes and cylohexanes.
- Seven and Eight member rings exist but
in small amounts. - Aromatics 20 or less
- Benzene, PAH, and Alkyl Substituted
- Benzenes.
- Sulfur, Nitrogen, Oxygen
5Petroleum
6Petroleum (Rock Oil)
- Crude Oil Unrefined Petroleum
- Petroleum is produced by deposition of organic
matter on sedimentary rocks and burial and
compaction of that matter over of millions of
years - Petroleum is introduced in to the environment
via - Dispersion
- Evaporation
- Dissolution
- For low molecular weight, volatile,
compounds evaporation is the
most important. - API (American Petroleum Institute Gravity)
- Measure of relative density
between a petroleum liquid
and water. - Crude Oil
- Light - gt 31.1 API
- Medium - 22.3 API - 31.1 API
- Heavy - lt 22.3 API
- gt 10---Floats on Water
- lt 10---Heavier than Water
7Mass Spectrometry(MS)
- Invented by J.J. Thomson in the early 20th
Century - Used to identify and quantify compounds present
in complex organic mixtures. - A Substance is bombarded with an electron with
sufficient energy to fragment the molecule. - Represent parent compounds that have been ionized
into positive gaseous ions. - Compounds are separated by their
mass to charge ratio or (m/z) - Example CO2 z1 so 44/1 44.
8Mass Spectrometry
- Quadrupole MS A Spectrum is achieved by a
combination of radio frequency(RF) and Direct
Current(DC) fields with opposite polarities on
the opposing rods. - Oscillations by the RF field cause lighter ions
to strike the positive poles and heavier ions to
strike the negative poles.
9Mass Spectrometry Detector
- Ions pass into the detector and impact a
dynode(High Negative Voltage). This release
electrons which are drawn towards a more positive
dynode. - Electon multipliers produce millions of electrons
from one ion.
10CO2 Mass Spectrum
11Electrospray Mass Spectrometry
- Electrospray Mass Spectrometry (ESI-MS)
- Liquid Analyte
- Capillary ? Metal Plate with a High Potential ?
Charged droplets on Plate ? Evaporation with N2 ?
Pin Size Hole in Plate leads to a Mass Analyzer
in a high vacuum. - Can run from 65-3000 m/z range.
12MS Spectrum
Figure 2. Expanded image of a portion of the mass
spectrum depicted in Figure 1.
Figure 1. Illustration of the full-scan mass
spectrometry without chromatographic separation.
13ESI-MS Spectrum
14Chemometrics
- Chemometrics application of mathematical and
statistical methods to chemical data - Projections to Latent Structures or Partial Least
Squares (PLS) - Regression Modeling
- Principal Component Analysis (PCA)
- Statistical method of Identifying patterns in
data in such a way as to highlight their
differences and similarities. - Data Reduction Technique
- Useful for large sets of interrelated variables
- Good for MS because each compound has a
characteristic mass and peaks are well defined. - PCA will find eigenvectors or eigenvalues
relevant to the data using a covariance matrix. - Eigenvectors can be thought of as preferential
directions or main patterns in a data set. - Higher Value Better representation of the data.
-
-
15Principal Component Analysis
- Steps
- Creation of a Correlation/Covariance Matrix
- Covariance is a linear relationship between two
variables - C (1/(N-1)) (X-mean) )(XT-mean)
- X Data Matrix
- XT Transpose of Data Matrix
- N Number of Columns
- Eigenanalyis is performed
- Eigenvalues are ordered in decreasing variance.
- 2 Largest values are considered the principal
components - Project both PCs onto a plane Score Plot
16Principal Component Analysis
- Example
- Spectra of 15 Peaks.
- Suspected to contain three components.
- Reduced to two principal components
with three
constituents.
17Methods
- Four Crude Oils
- A - Norwegian Sector in the North Sea (Light,
API of 36.8) - B - Norwegian Sector in the North Sea (Medium,
API of 31.9) - C - South America
- D - Central Asia
- ESI-MS
- 65-1000 m/z range. Atmospheric pressure positive
ionization. - Mobile Phase Acetonitrile and Ammonium Acetate
9010 - Two Experiments
- Discriminate four crude oils
- Discriminate the two Norwegian oils
- Five Combinations of Oils
- 0,25,50,75, and 100 of each
- Dissolved in Dichloromethane(2mg/ml)
- 1 ul samples were injected into a single
quadrupole mass spectrometer (x10). - Spectra were tabulated to a row with numbers
where each number represents the height of each
spectral line.
18Methods
- Analysis
- Matrix X 40 Rows x 931 Columns(1 per integer
mass number). - Matrix X 50 Rows x 934 Columns
- Matrix Y Combinations of A and B 50 rows and 2
Columns - (Percentages are
given in each column) - PCA was performed on Matrix X.
- PLS was performed on Matrix Y.
19Results - ESI-MS Spectra
Figure 3. Spectra of the four different crude
oils (A, B, C, and D).
20Results - (PCA Score Plot)
Figure 4. Score plot obtained after
fingerprinting and multivariate pattern
recognition of the four crude oils (each sample
was analyzed 10 times).
21Results - PLS
- PLS Projections to Latent Structures.
- R2 0.996
- Q2(Goodness Prediction) 0.985
Figure 5. Observed versus predicted percentage of
oil A in a binary mixture of oils A and B (each
mixture was analyzed 10 times).
22Conclusions
- Confirm several specific spectral characteristics
of hydrocarbon-containing mixtures. - Repetitive Spacing of 14 and 2 Da.
- ESI-MS produces singly charged ions.
- ESI-MS
- Captures most of the information needed for
pattern recognition - Commercially available
- Number are increasing worldwide.
- With the combination of ESI-MS and Chemometrics
we can distinguish between oils and mixtures of
oils with impressive repeatability. - Can be analyzes without pretreatment, except for
dissolution in Dichloromethane.
23Future Implications
- (1) Characterization of crude oils, including
heavy - crudes and petroleum products
- (2) Estimation of the contribution of oil from
various - reservoirs in commingled production scenarios
- (3) Detection of the effects of well treatment
chemicals - on crudes
- (4) Source identification (e.g., of oil spills)
by comparing - samples of spilled oil and candidate oils
- (5) Identification of compounds that make samples
- different (e.g., contaminants and additives)
- (6) Time-based studies (hydrocarbon upgrading,
degradation, - oil-in-water, oil weathering, etc.) and process
control - (7) Correlation of chemical fingerprints with
measured physicochemical or environmental oil
properties
24References/Thank You
- Ingvar, Eide and Kolbjorn Zahlsen. A Novel Method
for Chemical Fingerprinting of oil and Petroleum
Products Based on Electrospray Mass Spectrometry
and Chemometrics. Energy Fuels 2005, 19,
964-967 2005. 1 Oct 2004. - Standardizing the Novel Method for Chemical
Fingerprinting of Oil and Petroleum Products
Based on Positive Electrospray Mass Spectrometry
and Chemometrics. Energy Fuels 2006, 20,
265-270. 1 Nov 2005. - Williamson, Keenneth L. Minard, Robert D., and
Masters, Katherine M. Microscale and Marcroscale
Organic Experiments Fifth Edition. Houghton
Milfin. 2007 - McLafferty, Fred W and Turecek, Frantisek.
Interpretation of Mass Spectra Fourth Edition.
University Science Books. 2003. - March and Hughes. Quadrupole Storage Mass
Spectrometry. John Wiley and Sons. 1989. - Connel, Des W. Basic Concepts of Environmental
Chemistry Second Edition. Taylor and Francis
Group. 2005 - Brereton, Richard G. Chemometrics Application of
mathematics and statistics to laboratory systems. - Prof. Powers
- President Haas
25Questions
26- San Francisco Oil Spill - Nov 7th 2007