Changes in Technical Innovation and Customer Expectations Associated with the Use of LC-MS to Solve Challenges Encountered whilst Ensuring Environmental Safety Dr Simon Hird
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Changes in Technical Innovation and
Customer Expectations Associated with
the Use of LC-MS to Solve Challenges
Encountered whilst Ensuring
Environmental Safety
Dr Simon Hird
Overview
• Introduction
• Implementing the EU Water Framework
Directive
• Analytical approaches
• Targeted analysis
• Non-targeted screening
• Identification of unknowns
• Metabolite workflows
• Summary
Solving the problems of air,
water and soil pollution within
UK government
• Department of Environment, Food and Rural
Affairs (Defra)
• Food and Environment Research Agency
• Environment Agency
• Centre for Environment, Fisheries and
Aquaculture Science
• Health and Safety Executive
• Health and Safety Laboratory
• Independent bodies
• Drinking Water Inspectorate
The Food and Environment
Research Agency
• Executive agency of The Department for
Food, Environment and Rural Affairs (Defra)
• Turnover 2011/12 ca. £70 million
• 950 staff
• Main laboratories located near York
• £135 M laboratory of 300,000 ft2 within a secure site of
80 acres
• Customers are UK Government Departments, EU
but also many commercial organisations
Laboratory facilities near York
Plant and crop
protection
Food chain safety
Environmental risk
assessment
Crisis response
Diverse interests…
Analysis for chemical safety
and quality of our food
• Environmental contaminants
• Dioxins, PCBs, PFCs, BFRs, metals …
• Natural toxicants
• Mycotoxins, phycotoxins
• Residues
• Pesticides
• Veterinary medicines
• Processing contaminants
• Acrylamide, furan, 3-MCPD…
Analysis for chemical safety
and quality of our food
• Additives
• Colours, sweeteners…
• Authenticity
• Origin of food
• Deliberate adulteration • Melamine, “Sudan“ dyes…
• Packaging migration • Bisphenol-A…
• Nanotechnology
Environmental applications
• Monitoring
• Ecotoxicology studies
• Fate and behaviour of chemicals
• Field studies
• Pesticides in crops, soils etc.
• Transport of chemicals
• Degradation studies
• Veterinary medicines in manure
• Linked to QSAR modelling
• Nanoparticles
Legislation within the EU
• EU has adopted a substantial and diverse
range of environmental measures aimed at
improving the quality of the environment
• Member States have to implement the legislation
• One recent challenge has been the
implementation of the EU Water Framework
Directive (WFD)
• Resulted in the updating of national water quality
monitoring programmes across Europe
Water Framework Directive
(Directive 2000/60/EC)
• EU water policy aims for good quality water
for all
• Water Framework Directive (WFD)
established a legal basis to protect and
restore clean water across Europe
• Places special emphasis on the ecological
quality of waters.
• Objective of the WFD is to get all water into a
healthy state by 2015
Analysis required to support the
Water Framework Directive
• Monitoring of priority substances and other
pollutants, including new emerging
substances, in surface waters
• It requires the implementation of temporal
and spatial trend monitoring programs
• The use of integrative matrices (biota and
sediments) is strongly recommended to
achieve such objectives for hydrophobic
substances
• Not just about water analysis
Analysis required to support the
Water Framework Directive
• The WFD lists Priority Substances as
chemical pollutants that pose a significant risk
to (or via) the aquatic environment at EU level
• The most dangerous are listed as Priority
Hazardous Substances
• Member States have to monitor their
concentrations in target matrices and meet
the Environmental Quality Standards (EQS)
• Equivalent to the maximum allowable
concentration but determined by ecological impact
Water Framework Directive
• Establishes minimum performance criteria for
methods of analysis to be applied across the
EU when monitoring water status, sediment
and biota
• Demonstrating the quality of analytical results
• Validated, implemented and documented in
accordance with the ISO/IEC-17025 standard
• LOQ must be ≤30% of the EQS with a maximum uncertainty of ± 50
• Often challenging analytically!
• e.g. EQS for cypermethrin in surface water is 80 pg/l
New substances recently added
to WFD priority list
• Plant protection product substances:
• Aclonifen, Bifenox, Cypermethrin, Dicofol,
Heptachlor/Heptachlorepoxide, Quinoxyfen
• Substances used in biocidal products:
• Cybutryne, Dichlorvos, Terbutryn Industrial
chemicals: Perfluorooctane sulfonic acid (PFOS),
Hexabromocyclo-dodecane (HBCDD)
• Combustion by-products:
• Dioxins and dioxin-like PCBs
• Pharmaceutical substances:
• 17-a-ethinylestradiol, 17-b-estradiol, Diclofenac
Analytical requirements
• Concern that not all analytical methods in
current use are fit for the purpose of
monitoring priority substances for compliance
with EQS
• Need for innovative technologies with which
to develop methodologies to facilitate rapid
testing for organic contaminants to low levels
of detection
• Various analytical approaches possible
Analytical trends
The Challenge
Targeted analysis
Untargeted
analysis and
profiling
Identification
of unknowns
Added value,
more
information &
differentiation
Lists, limits
&
legislation
A tough
challenge!
Generic workflow
Aqueous sample
Preparation
Filtration (and acidification?)
Extraction
SPE (disks), SPME, LLE
Cleanup
Partition, SPE, none
LC-MS/MS or LC-HRMS(/MS)
Filtered aqueous
sample
Direct
injection
Targeted analysis using LC-
MS/MS
• Having two analysers increases the selectivity
that ensures interfering peaks from other
analytes or matrix are rarely observed
• Less isobaric interferences
• Lower limits of detection become achievable
• Direct injection of aqueous samples
• Provides a greater degree of confidence for
identification
• Most common variant is the triple quadrupole
MS/MS selected reaction
monitoring (SRM)
321>152
321>257
min1.380 1.400 1.420 1.440 1.460 1.480 1.500 1.520 1.540 1.560
%
0
100
34381a_data002
VM12-04732_5 MES (2058060)
min
%
0
100
34381a_data002
VM12-04732_5 MES (2058060)
min
%
0
100
34381a_data002
VM12-04732_5 MES (2058060)
321>152
321>257
Identification
through
comparing ion
ratios with
those from
standards
m/z 321
LC-MS/MS analysis of
emerging contaminants
• Not all new substances!
• Previously undetected or not considered a risk
• Pharmaceuticals
• Personal care products
• Pesticides
• Endocrine disruptors including
• Perfluorinated compounds (PFCs)
• Fire retardants
• Transformation products of all of the above
32 pesticides from water using
Agilent 1290-6460)
After acidification,
extraction by SPE
disk: UHPLC using
C18 with ES+/-
switching (LOQs
low ppb to ppt)
240 pesticides by UHPLC-
MS/MS (SRM): Agilent 1290-
6490)
After QuEChERs
extraction, no
cleanup: UHPLC
using C18 with
ES+/- switching
(LOQs low ppb to
ppt)
Perfluorinated compounds from
mink using Agilent 1290-6490
Sulphonic Acids Carboxylic Acids
After homogenisation with methanol, partition with KOH
and WAX SPE: UHPLC using a fluorinated column with
ES- (LOQs low ppb to ppt)
Brominated fire retardants from
fish using Agilent 1290-6460
α, δ, β, ε and γ HCBDs
TBPPA
After homogenisation,
with anhydrous
Na2SO4, hexane/DCM
and sulphuric acid
modified silica: UHPLC
using C18 with ES-
(LOQs ppt)
Pharmaceuticals and veterinary
medicines in farmed fish using
Waters Acquity and Xevo TQ-S
Time0.60 0.80 1.00 1.20 1.40 1.60 1.80 2.00 2.20 2.40 2.60 2.80 3.00 3.20 3.40 3.60 3.80 4.00 4.20 4.40
%
0
100
Job34085_2_006 26: MRM of 2 Channels ES+ TIC (Difloxacin)
8.98e6
After homogenisation with acidic
acetonitrile, dSPE (C18+NH2): UHPLC
using C18 column with ES+/- switching
(LOQs low ppb to ppt)
Trends in targeted analysis
• Transfer of methods from conventional to
highly selective mass spectrometers
• HPLC with UV and FD to LC-MS/MS
• Achieve required LODs/LOQs even in
complex matrices
• Simplifies sample extraction and clean-up
• Opens up scope for multi component analysis
• Reduce analysis times
• Must also provide unambiguous identification
• Move away from ion ratios to MS/MS and mass
spectral compound libraries?
Limitations with QqQ devices
• Scope of analysis limited to pre-programmed
transitions selected
• Product ion scans (MS/MS)
• An precursor ion of a particular m/z is selected in
Q1, fragmented in the collision cell (q) and
products scanned out from Q2
• Slow scan speed limits the sensitivity of QqQ
• Searching spectral libraries so no longer have to
acquire data from a standard
• Quadrupole-linear ion trap hybrid (QqLIT)
MS/MS product ion scan
m/z 321
MS/MS product ion spectrum
MS/MS product ion spectra
Limitation of targeted approach
• Need reference standards
• Need to program methods with RTs of
analytes and specific transitions to monitor
• The targeted approach will fail to detect other
contaminants present in the sample
• Unable to go back and “mine” the data later
Trends in non-targeted analysis
• Transfer of methods from specific
methodologies to those providing data for
comparison with databases
• An alternative so-called “non-targeted”
approach
• LC-HRMS
• Database searching via mass measurements
• LC-HRMS/MS
• Also provides spectral library searching
• Non-targeted acquisition but initial data processing
tends to be still targeted…
Non-targeted acquisition
• Use of “high resolution” instruments
• Time of flight (ToF) or orbitrap mass analysers
• Full spectral information
• High mass resolving powers and mass resolution
• Specifications vary significantly
• Good mass accuracy
• Good sensitivity through improved ion optics
• Variable acquisition speeds
Comparison of ToF with orbitrap
Mass
analyser
Resolving
power
(x103)
Mass accuracy
(ppm)
Acquisition
speed (Hz)
Q 3-5 Low 2-10
IT 4-20 Low 2-10
Tof 10-60 1-5 10-100
Orbitrap 100-240 1-3 1-5
Q, ToF and orbitrap also include common hybrid configurations with Q or LIT
as the first mass analyser providing MS/MS or MSn capabilities
Holcapek et al. (2012). J.
Chromatogr. A 1259: 3
Data processing for screening
• Peak detection by extracting those ions
matched with entries in a database
• Can be psuedo molecular ions and fragments
• Recognition is based upon measurement of:
• Accurate mass
• Isotope pattern
• Retention time (if available)
• A response threshold
• Results are reported as a “hit list” with or
without creating chromatographic peaks
Non-targeted screening
• In food safety data are rarely interrogated for
analytes other than those in the database
• Emphasis has been on screening
• The aim of validation for pesticides in food is to
have no more than a 5% false-negative rate
• Methods and software parameters tend to be
optimised to find a practical balance between
reported false positives and false negatives
Non-targeted screening
• To be effective data processing must be
automated and quick
• Minimise false negatives whilst generating a
manageable number of false detects
• Apply tolerances on response threshold, retention
time and isotopic fit and the presence of a second
diagnostic ion
• It requires more computing power and data
management/storage than that traditionally
associated with LC-MS analyses using QqQ
instruments
Validation using samples that
previously tested positive using
LC-MS/MS
Commodity
(sample
numbers)
Incurred samples
Automated - no intervention (± 5 mDa)
Data processing +
Analyst intervention
Lettuce (10) 93% 97%
Grape (20) 93% 99%
Pear (20) 93% 94%
35 different pesticides, 169 residues, between 0.01 – 0.78 mg/kg
Agilent 1290-6230 ToF
Non-targeted screening
• The scope of this approach is limited by the
content and size of the database
• Screening for other compounds not listed the
database is by deconvolution for detection
and then assignment of a scored list of
possible elemental formulae based upon
measured mass and isotopic fit
• The most likely candidates are further
scrutinised by employing additional data or by
extra analyses
Trying to identify unknowns…
• Assignment of a structure for a peak detected
by accurate mass alone is unlikely unless:
• Compound detected has a significant mass defect
between the monoisotopic mass of an element
and the mass of its isotopic cluster
• Is known in the chemical literature, a reference
database or an internet resource
• The first crucial step is to obtain correct
elemental compositions
Trying to identify unknowns…
• Accurate mass measurement < 1 ppm
• Information about isotope pattern
• Fragmentation data from MS/MS or MSn
experiment
• Accurate mass measurement of precursor and/or
product
• In silico fragmentation or fragmentation trees
to aid de novo characterisation
Work flows for investigating
transformation products
• Use of work flows more associated with
analysis of metabolites
• Looking for differences between treated and
control samples
• Profiling as used for metabolomics
• Looking for specific transformations
• As used for metabolite identification
O
O
O
O
O
O
O O
OH
O
OHO
N
OH
OH
OH
O
O
O
O
O
O
O
O
O O
OH
O
OHO
N
OH
OH
OH
OH
O
Loss of CH2
(Demethylation)
Control samples taken,
Sampling time-frame
Jan 2010
Honey
collected
- Honey Flow May/June 2010
Chemical
treatment
Jan/Feb 2011
Week from
honey flow 1 2 3 4 5 7 8 6 9 10 18 Over winter
• Other matrices – brood honey, brood, bees, wax, propolis
Investigating residues in honey
after treatment of bee hives with
a veterinary medicine
Possible workflows
• LC-MS/MS analysis to determine depletion
rate of the parent compound
• Could also look for transformation products
reported in the literature and those predicted
by chemical modelling
• But are there any other relevant markers of
the residue present in the treated samples
that this approach would have missed?
• Use of LC-Tof and LC-QqTof
Looking for differences between
treated and control samples
• From 1288 entities detected, only 4 components were
associated with hive treatment
• Proposed a number of possible empirical formulae
• 2 tentatively identified by comparison of RT with standards
Compound Responses Mass Retention Time
Control Dosed /min
C10H14O9S 1 29501 310.0364 7.78
C29H39N15 1 48106 597.3501 9.36
C36H63N3O16 1 156943 793.4211 9.36
Desmycosin 1 920521 771.4407 9.36
C44H76N4O19 1 63375 964.5091 9.79
C47H78N2O17 1 78075 942.5303 9.79
[email protected] 1 87355 1099.5529 9.80
C52H87NO22 1 56072 1077.5734 9.80
C38H63NO13 1 73486 741.4294 9.99
Tylosin A 1 2690789 915.5209 9.99
C38H63NO12 1 222141 725.4347 9.99
C44H71N7O15 1 330487 937.5005 10.00
Also called
Tylosin B
Looking for specific
transformations
• Tentative identification of:
• Parent (tylosin A)
• Demethylation (tylosin C)
• Glycosylation (glycosylated tylosin A)
• Missed
• Tylosin B
• Loss of mycarose not included as a transformation
• Need to have a good understanding of likely
transformations in various environments
Summary of the two metabolite
approaches • Profiling the whole “fringerprint” vs. looking
for specific transformations
• Assign molecular formula to accurate mass
• Assign likely structure to formula
• Complimentary use of MS/MS
• Need standards to generate a library
• Expert knowledge of the application optimises
use of the software and need decent
computing power
• Can be applied to environmental analyses
LC-MS does have an Achilles
heel…
• Electrospray suffers from ion suppression
from co-elution of co-extractive component
• Impacts upon:
• “Detectability” and hence true reporting limit
• Possible reporting of false negative results
• Accuracy of quantification
• Applies to targeted, non-targeted and profiling
• Matrix-matching of calibrants and/or use of
stable isotope analogues for correction
Future challenges
• Targeted analysis
• More analytes in more samples, quicker, for less
and with lower reporting limits…
• Untargeted analysis
• Try to identify as much of the analytical data as
possible from which the relevance and
significance is determined by informatics
• Directed analysis?
• Only identity the significant (e.g. “active”, “toxic” or
estrogenic) peaks/components…
Directed analysis
Genotoxic
Summary
• Development of target compound analysis
• Focus on multi residue approaches
• Continued use of LC-MS/MS
• Increased use of GC-MS/MS for non-polars
• Implementation of non-targeted approaches
• Increased use of LC-HRMS
• Database searching
• Complimentary use of MS/MS libraries
• Profiling and metabolite identification
• Customer expectations?
More for less…
0
50
100
150
200
250
300
350
400
2002 2003 2004 2005 2006 2007 2008 2009 2010
Number of results per sample
Training is essential!
Acknowledgements
• Targeted analysis: LC-MS/MS team at Fera
• Non-targeted analysis: Richard Fussell and
Tony Lloyd
• Transformation work: Richard Fussell, Danny
Chan and Jonathan Tarbin
• CRD, VMD, FSA and various commercial
organisations for funding
• Agilent Technologies, Waters and Thermo
Scientific for support
Related Documents
