PTR-TOF-MS: A New Instrument For Real-Time Analysis Of Multi-Component Systems: Applications To Food...
Transcript of PTR-TOF-MS: A New Instrument For Real-Time Analysis Of Multi-Component Systems: Applications To Food...
PTR-TOF-MS: A New Instrument For Real-Time Analysis Of Multi-Component
Systems: Applications To Food Analysis
Rob Linforth, Annie Blisset, Andy Taylor Dept. of Food Science
University of Nottingham
Caroline Lamont-Smith, Steve Mullock & Fraser Reich
Kore Technology Ltd. Ely, Cambs
Talk: Outline
• Overview of the Proton Transfer Reactor method
• Why a PTR-TOF-MS?• The Instrument• Some Example Data• Summary
First a confession:
This talk is heavy on the instrument description, its properties and possibilities of the instrument, because the speaker is one of the team that designed the instrument.
PTR: Proton Transfer Reactions
Essentially, PTR is a subset of Chemical Ionisation. It was defined and refined by Werner Lindinger in the 1990’s, and resulted in the first commercially available instrument (Ionicon)
• Aim is to achieve ‘Soft’ ionisation so as not to fragment the molecule(s) of interest: less fragments = less mass spectral ‘noise’.
• In positive ion mode, most frequently used reaction is:
H3O+ + R H2O + RH+
• For this protonation reaction to occur, the proton affinity of R must be greater than that of water
• The ionisation probability is almost the same as the collision cross-section, so the aim is to induce sufficient collisions to ensure efficient analyte ionisation
PTR: Proton Transfer Reactions With H3O+
Compound PA (kJ/mol)
Oxygen 421
Nitrogen 493
Carbon Dioxide 541
Sulphur Dioxide 672
Water 691
Hydrogen Sulphide 709
Benzene 750
Methanol 754
Toluene 784
Naphthalene 803
Ethyl Acetate 836
Tri ethyl phosphate 909
Compounds above water in the proton affinity table will not be ionised, whereas compounds below will
The good news is that most VOCs have higher proton affinities than water and will receive a proton from H3O+
Generally, larger molecules have larger proton affinities
Linearity of Response
Plot of H2S Peak Yield vs. H2S Concentration (H2s in Air)
0
50
100
150
200
250
300
350
400
450
500
0 10 20 30 40 50 60
Calibrated H2S Concentration(ppm)
H2S
Peak Y
ield
(x50)
Series1
Linear (Series1)
Provided that [RH+]<< [H3O+], the H3O+ signal does not change with analyte concentration, and the detected analyte intensity is linear with concentration:
[RH+] = [H3O+]0(1-e-k(R)t) [H3O+]0 [R] kt
0-50ppm H2S calibration:
Raw data plot
(k is rate constant for the reaction, and t is time to travel through the drift tube)
H2S proton affinity= 709kJ/mol
Soft Ionisation vs. Electron Impact Ionisation
70eV EI spectrum of Ethyl Butyrate
Molecular Ion appears at mass 116 m/z
Significant fragmentation to lower masses
Soft Ionisation vs. Electron Impact Ionisation
0
1000
2000
3000
4000
5000
60 70 80 90 100 110 120 130 140
Counts
Mass (m/z)ce6.1, 15ml.min, ethylbuty, 2.0mbptr
PTR-TOF-MS Spectrum of 5ppm Ethyl Butyrate
Molecular Ion at 117 m/z
(molecular ion + proton)
Very little fragmentation
Charge Transfer In The PTR Reactor
• Need to induce sufficient collisions: H3O+ and analyte mixed in a reactor tube (drift cell)
• Voltage gradient across reactor
• Pressure ~ 1 mbar, approximately 2000 collisions down 10cm length reactor
• Aim is to have dilute analyte, so that analyte molecules do not collide with anything except H3O+ . This keeps the ionisation scheme and calibration simple, unlike in Ion Mobility Spectrometry (complex inter-analyte charge transfer reactions)
The PTR Reactor and the E/n Ratio
• E is the voltage gradient down the reactor in V/cm
• n is the gas density in molecules / cc (1mbar in reactor, typically)
• E/n in Townsend; 1 Townsend = 10-17 cm2 / Vs
• Lindinger et al. (1998) showed that 120-140 T is ideal for non-fragmentation of organic molecules
• If E/n too high, fragmentation of molecule occurs
• If E/n too low, clustering of water molecules occurs and can present a problem for certain species
Reagent gas in
Analyte in
GD
SD
PTR
GD Cathode
GD Anode
Transfer Optics
The PTR Reactor and the E/n Ratio
0
10000
20000
30000
40000
50000
60000
100 120 140 160 180
Counts
Mass (m/z)
0
10000
20000
30000
40000
50000
60000
100 120 140 160 180 200
Counts
Mass (m/z)
0
20000
40000
60000
80000
100000
120000
100 120 140 160 180
Counts
Mass (m/z)TEP mass 182
Tri-ethyl phosphate, protonated mass 182
130 Townsend
195 Townsend – loss of the ethyl groups (28 mass units)
260 Townsend
PO4 + proton
Why Use A TOF-MS?
• Parallel detection means no analytical price to pay for monitoring many species, unlike a sequential analyser such as a quadrupole or magnetic sector
• TOF Cycle frequency typically 20-50 kHz = high data rate
• Full mass spectra to several hundred masses with 4-5 orders dynamic range in one second
• Possibility of ‘real time’ data analysis with < 100ms resolution, multiple species and sensible counting statistics
• Ability to interrogate a data set ‘retrospectively’: intensity of any species as a function of time, mass spectra for a variety of ‘time slices’
• Parallel detection and full mass spectra more suitable to software data reduction techniques, e.g. principle component analysis
Why Use A TOF-MS?
This mass spectral plot, acquired with a PTR-Quad-MS (a sequential scanning device) took 40 seconds
A full mass spectrum to several hundred m/z can be acquired in less than a second with a PTR-TOF-MS
Mass spectrum from breath (U. Nijmegan)
TOF-MS: Accurate Mass Capability
0
500
1000
1500
29 30 31 32 33 34 35 36
Counts
Daltons
30.00 m/zNO+
31.99 m/zO2
+
33.04 m/zCH3OH2
+ 36.04 m/z(H2O)2
+
Downside Of A TOF-MS?
• Main drawback is analysis of a continuous analyte stream: When the TOF cycle has started, no further ions can be injected without resulting in multiple overlapping spectra
• ‘Duty cycle’ describes the % of the analyte stream that is sampled
• In an orthogonal TOF (analyte stream enters TOF source at 90°), the duty cycle can reach 10%: when the Kore TOF source pulses, it ‘empties out’ ~5-10 microseconds worth of ions. Usage is therefore 5µs in 50µs, I.e. 10%
• Currently we are observing up to 500kc/s of Hydronium at the detector, approximately half that reported by the PTR-Quad (single ion mode).
• Prospect of improving duty cycle using ‘Hadamard’ methods – high frequency random pulsing of source with mathematical deconvolution of spectra - still not really proven as a workable solution
TOF MS: Orthogonal Pulsing TOF Source
ContinuousSample
DetectorIonsExtracted
Light IonsHeavy
Ions
Ions of different masses within a single ‘cycle’ arrive at the detector at different times according to the relation:
K.E. = mv2/2
Ions with m/z = 1000 has flight time ~20-50s, therefore ‘cycle time’ = 20-50s, so typical pulsing frequency = 20-50 kHz
Volts Source Off
t
Source On
0
-2kV
Extractor pulses to –400v
TOF Source
Variant on ‘Wiley – Maclaren’ source. Compensation for ion position within source
TOF MS: Orthogonal Pulsing TOF Source
Overview schematic of the instrument: TOF source to detector
Reflectron
Blanker
Blanking
Field Free Region
Detector
Pre-
DeflectorsTOF Source
Transfer Optics
PTR Source
Extract
Pulser
Pulser
Amplifier
TOF MS: Orthogonal Pulsing TOF Source
Overview schematic of the instrument
GD Source2mbar
PTR Reactor 1mbar
Transfer Optics 10-4 mbar
Mass spectrometer and detector 10-6 - 10-7 mbar
Mass Spectral Analysis: Applications to Food Analysis
University of NottinghamDepartment of Food Science
Professor Andy TaylorDr. Rob LinforthAnnie Blisset
• “Breath-by-breath” Research
• Main technique in lab: APCI
• Wanted to add PTR-TOF-MS
• Instrument recently delivered; has both PTR and APCI functionality
Data acquired at Nottingham by one of us (FR) with Rob Linforth and Annie Blisset
Breath-by-Breath Analysis Of Juicy Fruit Gum
The principle compound released during chewing of Juicy Fruit gum is ethyl butyrate, molecular mass 116
• Soft tube inserted lightly into the nostril of subject
• Breathing normally, the mass spectrometer begins acquiring data.
• Most of the exhaled breath passes out into open space, but a side-mounted capillary pipe samples the breath into the proton transfer reactor.
Breath-by-Breath Analysis: Release of Ethyl Butyrate
Idealised appearance of breath markers, such as acetone, as exhalation occurs
Release of flavours during mastication
Sig
nal i
nten
sity
Time
• First data set: take ‘raw data set’ (all ion flight times recorded as a function of elapsed experiment time)
• Integrate the ethyl butyrate protonated ion (mass 117) every second
Breath-by-Breath Analysis At Different Data ‘Granularity’
0
2000
4000
6000
8000
10000
0 10 20 30 40 50
Signal for Mass 117
Time in Seconds1 second integration
Reconstruction of ethyl butyrate signal with 1 second integration time
500
1000
1500
2000
2500
0 10 20 30 40 50
Intensity
Time
Reconstruction of Acetone signal with 1 second integration time
Breathing rate ~5/min
Breath-by-Breath Analysis At Different Data ‘Granularity’
0
1000
2000
3000
0 10 20 30 40 50 60
Signal for Mass 117
Time in Seconds0.25 second integration
Reconstruction of ethyl butyrate signal with 0.25 second integration time- further structure emerging still
0
1000
2000
3000
4000
5000
6000
0 10 20 30 40 50 60
Signal for Mass 117
Time in Seconds0.5 second integration
Reconstruction of ethyl butyrate signal with 0.5 second integration time- further structure emerging
Breath-by-Breath Analysis At Different Data ‘Granularity’
0
500
1000
1500
2000
0 10 20 30 40 50 60
Signal for Mass 117
Time in Seconds0.125 second integration
0
200
400
600
800
1000
1200
0 10 20 30 40 50 60
Signal for Mass 117
Time in Seconds0.0625 second integration
Reconstruction of ethyl butyrate signal with 0.125 second integration time
Reconstruction of ethyl butyrate signal with 0.0625 second integration time: no further structure emerging
Breath-by-Breath Analysis: Different Person
Reconstruction of Acetone signal with 0.125 second integration time
Note different breathing rate ~10/min for different person
Overlay of ethyl butyrate signal with acetone
0
200
400
600
800
1000
1200
1400
0 10 20 30 40 50
Signal
Time (seconds)Mass 59
0
500
1000
1500
2000
0 10 20 30 40 50
Signal
Time (seconds)Mass 59 and 117
Full Mass Spectra From 125ms Time Slices
0
1
2
3
4
20 40 60 80 100 120 140
Counts/0.125s
Daltons
0
1
2
3
4
20 40 60 80 100 120 140
Counts/0.125s
Daltons
34.00 –34.125 seconds
34.50 –34.625 seconds
Logs
cale
dat
a
Acetone and Ethyl butyrate
Mass Spectrum From Headspace of Freshly Macerated Tomatoes
0
1000
2000
3000
4000
60 80 100 120 140
Counts
Daltons
3 hexenol 3 hexenal
3-Hexenol vs. Hexanal, Both C6H12O
Hexanal: protonated masses at 101, 83, 55
3-Hexenol: protonated masses at 101, 83, 59 and 55
In previous slide, masses 101 and 83 were identified as 3-Hexenol, but in truth there is the possibility that 101 and 83 can be due to Hexanal.
3-Hexenol vs. Hexanal, Both C6H12O
Is it possible to find an E/n value that will give a mass 101 for one compound but not the other?
0
200
400
600
800
1000
80 85 90 95 100 105
Counts
Daltons
0
200
400
600
800
1000
1200
1400
80 85 90 95 100 105
Counts
Daltons
Hexanal 3 Hexenol83
83101
101
No ‘threshold phenomenon’ observedAlso, no E/n found for 3-Hexenol at which intensity of 101>83
Possibility of modulating PTR voltage and using software tools for compound differentiation? More work required, clearly!
Summary
• PTR-MS instruments based on quadrupole mass spectrometers are now widely used in studies of environmental and atmospheric chemistry as well as food and medical applications
• A TOF-MS increases the possibilities of real-time analysis down to <100ms “data granularity”, with full mass spectral acquisition in each time slice
• A TOF-MS permits any ion chromatogram to be re-constructed from the data set
• A TOF-MS can be operated at higher mass resolutions than a quadrupole mass spectrometer. Even at relatively low mass resolution the mass accuracy is better than 30 millimass units
• Greater mass range capability, with no discrimination against high masses up to several hundred Daltons