Send the link below via email or IMCopy
Present to your audienceStart remote presentation
- Invited audience members will follow you as you navigate and present
- People invited to a presentation do not need a Prezi account
- This link expires 10 minutes after you close the presentation
- A maximum of 30 users can follow your presentation
- Learn more about this feature in our knowledge base article
Do you really want to delete this prezi?
Neither you, nor the coeditors you shared it with will be able to recover it again.
Make your likes visible on Facebook?
You can change this under Settings & Account at any time.
Introduction to Owlstone ultraFAIMS
Transcript of Introduction to Owlstone ultraFAIMS
Owlstone ultraFAIMS What is FAIMS? FAIMS is a technique derived from ion mobility spectrometry that separates ions based on changes in size & shape Applying an electric field to an ion in a buffer gas causes it to drift in the direction of the field Ion velocity Electric field strength v = K E At low fields, the drift velocity is proportional to the electric field strength and the constant of proportionality is known as the mobility (K) [Purves R W, Guevremont R, Anal. Chem. 1999, 71, 2346-2357] At higher electric fields, the ion mobility is not constant but varies with electric field
Separation speed is determined by the ion residence time
Typically sized FAIMS devices have residence time of ~10ms to ~100ms - hardly compatible with UHPLC and data-dependent FAIMS filtering Ultra-fast FAIMS fundamentals What is different about ultraFAIMS? The Owlstone ultraFAIMS device Each device consists of a set of parallel gaps in a metal substrate that forms the electrodes Higher fields ensure good resolution despite the big reduction in ion residence time In a FAIMS device, ions pass between electrodes, across which a transverse alternating electric field (the dispersion field, DF) is applied A DC field (compensation field, CF) can be superimposed onto the DF waveform to oppose a specific drift velocity, allowing selected ions to pass through the filter electrodes Principles of FAIMS separation FAIMS separation depends on all terms Higher fields also increase separation flexibility Higher terms are more important at greater E, and ions with close α1 often have dissimilar α2 (J. Phys. Chem. A 110, 2663, 2006) Assuming the MS can acquire multiple samples during the CF sweep, the FAIMS device provides pre-separation of the components in the sample In a FAIMS-MS hybrid system, ions emerging from the FAIMS device pass into the inlet of the Mass Spectrometer for analysis m/z m/z m/z intensity compensation field FAIMS-MS Analysis UltraFAIMS is designed to provide HIGH SPEED separation How do we achieve residence time < 1 ms and still achieve good separation and sensitivity? The solution is to reduce drift length and hence residence time... but increase dispersion field to compensate The field is increased by going to narrower channels
(the physics of electrical breakdown in vapour is such that narrower channels can sustain higher electric fields) The Owlstone ultraFAIMS devices are produced using micromanufacturing techniques to produce gaps of 100 microns and below
This allows maximum electric fields of more than double those in millimetre-scale devices at constant t Or at constant R 75% DF ultraFAIMS-MS Interface Schematics The device has been interfaced with an Agilent Jetstream ionisation source Transmission and Resolution Overall peak capacity (without modifiers) now up to 15-20 Variable chip sizes allows easy performance adjustment Compensation Field Open area 5mm 2 Open area 7.5mm 2 Open area 9.5mm 2 Residence time: 120 microseconds*
Higher sensitivity Residence time: 190 microseconds*
Best overall performance Residence time: 235 microseconds*
Highest peak capacity CF stability & sensitivity to source/sample CF drift due to solvent/matrix composition is known to be a problem for many FAIMS designs, particularly for higher flow applications Compensation field Infusion expt - low flow, high DF Morphine in ACN: water
Flow rate = 0.05ml/min
DF = 270Td Peak position stable with solvent composition Preliminary experiments with drug spiked into human plasma shows similar behaviour at similar flow rates. LC injection expt - high flow, lower DF Flow rate = 0.4ml/min
DF = 220Td
0.25CV scans/sec No appreciable sensitivity to solvent/matrix composition with ultraFAIMS & Agilent Jet Stream What can we do with high speed? Fast stepping of CF & DF (2ms & 10ms) allows use with ultra-high speed chromatography Within the LC peak duration, several FAIMS scans can be performed During each FAIMS scan, many MS spectra can be acquired 1-5s per peak
Up to 10 FAIMS scans/sec 50-100 steps/scan (2-100ms each)
Single MRM data point or TOF spectrum per step Real-time CF/DF scanning & hopping Two approaches for real time data acquisition and method optimization:
Hopping between multiple MRM transitions during QQQ MS acquisition
Real time optimization of CF & DF during injection 150ms 15 complete CF scans during 2s FWHM of an LC peak What can you do with ultraFAIMS? Separation of pharmaceutical excipients Protonated 2-hydroxy-4-octyloxybenzophenone (HOBP, m/z 327.1955) and the n=7 oligomer of PEG 400 (m/z 327.2013) are sufficiently close in mass (17.7ppm mass difference) that they can not be resolved by the TOF mass analyzer (required resolution ~ 130K) An ultraFAIMS sweep at 48kV/cm produces good separation – allowing more accurate mass measurement of each ion and a clean product ion spectrum Separation of peptide charge states Different charge states tend to separate well in ultraFAIMS - the high DF exposes the differences in high-field mobility of the different charge states The behaviour is consist across different peptides (both standards and tryptic digests) Singly charged ions can be selected as a group Ion trap mass spectrum of bradykinin without FAIMS separation Selective enhancement of the [M+H]+ (m/z 1060) ion at a CV of +1.4V , DF 60kV/cm Selective enhancement of the [M+2H]2+ (m/z 531) ion at a CV of +3.5V, DF 60kV/cm Ion trap mass spectra alpha-1-acid glycoprotein (AAG) tryptic digest (73pmol µl-1) – without FAIMS both 1+ and 2+ charge states are present Using appropriate FAIMS CV, the doubly charged ions can be filtered out, leaving just a spectrum of singly charged ions Assisting protein identification The top 20 ions in the spectra list generated from the FAIMS-filtered spectra was searched against the SwissProt protein PMF database using the MASCOT31 search engine Similar approach to that used in MALDI-TOFMS Alternatively, multiply charged peptide ions can be individually selected and fragmented and the resulting spectrum used to identify the peptide Optimum CV for transmission of doubly charged m/z 877 ion applied (+3V), excludes singly charged species and other doubly charged peptide ions Selection of parent ions prior to in-source CID The ultraFAIMS device can be used to select a subset of ions which can then be fragmented using in-source CID - the pre-selection produces simplified fragment ion spectra that can be used for identification or quantification without the need for tandem MS FISCID analysis of tryptic peptides in plasma TIC of tryptic peptides from LC analysis of human plasma The complex nature of the sample results in co-elution of peptides e.g. here around RT=3.5mins, 3 peptides co-elute SICs of 3 of the tryptic peptide ions LC-MS spectrum at 3.5mins without FAIMS separation LC-MS spectrum at 3.5mins with FAIMS at DV=47kV/cm and CV=2.5-2.6V - showing preferential transmission of the m/z 480 peptide LC-in source CID-MS spectrum at 3.5mins without FAIMS separation UltraFAIMS allows pre-selection of peptides prior to fragmentation... LC-in source CID-MS spectrum at 3.5mins with FAIMS at DV=47kV/cm and CV=2.5-2.6V - showing a far simpler fragment pattern This results in a simpler fragment spectrum, which increases the probability of identification TOF fragmentor voltage at 75V TOF fragmentor voltage at 340V Peptide identification was carried out via the MASCOT search engine, searched against the SwissProt protein database. All ions with greater than 10% of the base peak intensity were included in the search peak list. Peptide quantification by LC-FISCID-MS Gramcidin S (m/z 571) spiked into human plasma sample at different concentrations
FAIMS used to select the Gramcidin S ion for transmission or fragmentation (fragments at m/z 311,424, 685, 798) The FISCID technique can also be used for quantitative analysis Improved metabolite quantification Direct infusion of an IAG standard was used to optimise FAIMS transmission of IAG The quantitative performance of the prototype LC-FAIMS-MS system was evaluated by the analysis of (R/S) Ibuprofen 1-β-O acylglucuronide (IAG) metabolite spiked into urine Higher LDR observed with FAIMS separation, by over 3 orders of magnitude Improved intra-day reproducibility was observed with FAIMS pre-selection of the [IAG-H]- ion The absolute intensity of the [IAG-H]- peak is reduced ~50% because of lower FAIMS transmission, but is compensated by an improvement in signal to noise ratio, which lowers the LOQ from 0.018 to 0.010 µg/ml The incorporation of a FAIMS separation in the LC-MS analysis significantly reduced chemical interference from urine Urine blank IAG spiked urine sample Metabolite identification The ability to select the metabolite based on differential mobility results in the enhancement of IAG derived fragment ion peaks in mass spectrum (bottom) compared with LC-CID-MS without FAIMS (top) LC-FISCID-MS analysis of IAG in urine, monitoring the m/z 205 fragment of the FAIMS-selected [IAG-H]- ion (bottom), shows enhanced selectivity compared to FAIMS off LC-CID-MS (top) (V) aglycone m/z 205 Four fragments are produced from [IAG – H]- by CID (II) decarboxylated di-dehydrated glucuronate m/z 113 (III) dehydrated glucuronate m/z 175 (IV) glucuronate m/z 193 No FAIMS - IAG fragments are difficult to locate due to the complexity of the mass spectrum and poor signal to noise as a result of other ions in the matrix With FAIMS - signal-to-noise ratios for the fragment ions increased by approximately 2-fold, enhancing the response of the IAG fragments relative to other interfering peaks to aid identification Selection of large proteins from a complex matrix Large (>30kDa) proteins such as BSA tend to behave as type A ions in the extreme ultraFAIMS dispersion fields, whereas smaller proteins and peptides behave as type C ions - this means they can be readily separated by selection of appropriate CF Ubiquitin BSA
With FAIMS selection at CF = 0 - the non-BSA ions over the whole mass range are suppressed by 10000 times, allowing detection of nearly the full charge state distribution of BSA ions Mass spectra for a solution of ubiquitin and trace of BSA in a complex matrix with no FAIMS -2 0 3 Compensation field /Td 1 2 -1 Improved accurate mass measurement We selected an analyte (2-hydroxy-4-(octyloxy)benzophenone with the same nominal mass – 327 – as one of the PEG400 ions - and applied a FAIMS DF of 48kV/cm Extracted chromatogram for 326.3-330.3 Sample: 2-hydroxy-4-(octyloxy)benzophenone (0.01mg/ml) + PEG400 (0.1mg/ml) in 50:50 ACN:water + >0.1% formic acid
2-hydroxy-4-(octyloxy)benzophenone has m/z 327.1960 The two ions were well resolved in the FAIMS spectrum, allowing each to be individually selected for accurate mass measurement - reducing the mass error in each case Separation & quantification of isobaric PGIs Use of modifiers to enhance selectivity ESI-FAIMS-MS spectrum of a mixture of the three PGI’s overlaid with individual components UltraFAIMS can be used to enable rapid and direct analysis of PGIs using thermal desorption in conjunction FAIMS-MS or FISCID-MS. The 3 PGIs were partially resolved at a DF of 228Td The chip was set to constantly transmit at a compensation field of 1.3 Td to preselect NNDMT. A thermal desorption profile for the molecular ion of NNDMT (m/z 136.1121) was obtained 50 ng of NNDMT was spiked into 50 mg of starch to simulate a PGI concentration of 1 ppm. This was placed into a TD tube and thermally desorbed. This profile shows a peak-to-peak signal to noise ratio of 11.8:1 indicating that this response is in the quantifiable range. The calculated limit of detection for NNDMT is 250 ppb. Since the threshold for toxicological concern equates to a control limit of ~1.5ppm, this LOD is sufficient to make TD-FAIMS-MS suitable for fast toxicological profiling of APIs. J.Reynolds et al, ISIMS poster presentation 2012 The addition of solvent vapours and modifier gases to the carrier gas can enhance peak separation in ultraFAIMS, as well as in macro-scale FAIMS devices - this is likely to open up a wide range of further applications Effect of adding increasing concentration of carbon dioxide (top row) and methanol (bottom row) on separation of the [M-H]- ions of three positional isomers of phthalic acid. The figures on the right illustrate the best separation achieved. C. Beekman et al, IMSC 2012 poster presentation By ramping the CF over a range of values, all the ions in the sample can be transmitted through the filter in sequence, producing a CF spectrum MS inlet The use of multiple channels ensures total ion capacity is not reduced Batch micro-fabrication techniques allow extremely precise control over ion channel dimensions on a micron scale. The device is small, stable and mechanically robust Transmission as shown in table (blue fields indicate typical operating condition) *for 1.7L/min MS inlet flow Ugarov et al., ASMS presentation, 2012 Ugarov et al., ASMS presentation, 2012 L.J. Brown et al, Anal. Chem., 2012, 84 (9), pp 4095–4103 L.J. Brown et al, Anal. Chem., 2012, 84 (9), pp 4095–4103 Ugarov et al, ASMS presentation, 2012. L.J. Brown et al., Anal. Chem. 2010, 82, 9827–9834 L.J. Brown et al., Anal. Chem. 2010, 82, 9827–9834 S/N ratio 100:1 (m/z1060); 1000:1 (m/z 531)) S/N ratio 1000:1 S/N ratio 10000:1 Multiply charged ions can be selected as a group or individually L.J. Brown et al., Anal. Chem. 2010, 82, 9827–9834 AAG was identified as the top hit with a significant confidence score of 61 (where 56 or above is deemed statistically significant at a 95% confidence interval) Ion trap MS/MS spectrum of m/z 877 ion L.J. Brown et al., Anal. Chem. 2010, 82, 9827–9834 L.J. Brown et al., Anal. Chem. 2010, 82, 9827–9834 With the CV of 2.5-2.6 V applied, human serum albumin (HSA) was identified as the top hit, the only significant match, with a confidence score of 34 (where 27 or above was deemed statistically significant at a 95% confidence interval). With no FAIMS separation, LC-in source CID-MS yielded no significant hits on the database. L.J. Brown et al., Anal. Chem., 2012, 84 (9), pp 4095–4103 L.J. Brown et al., Anal. Chem., 2012, 84 (9), pp 4095–4103 L.J. Brown et al., Anal. Chem., 2012, 84 (9), pp 4095–4103 RSDs (n=6, at 0.45ug/ml) were 5.1% for precursor, and <15% for fragments Peak areas for all ions produce a linear response (R2>0.99) between 0.45-9.0ug/ml R. Smith et al, J. Chromatogr. A, 2012 (under review) R. Smith et al, J. Chromatogr. A, 2012 (under review) R. Smith et al, J. Chromatogr. A, 2012 (under review) A. Shvartsburg et al., Anal Chem. 2012 Sep 4;84(17):7297-300 J.Reynolds et al, ISIMS poster presentation 2012 J.Reynolds et al, ISIMS poster presentation 2012 For more information - or to discuss potential applications testing, please contact us at www.ultrafaims.com The mobility is related to the size and shape of the ion, and different ions drift at different speeds for a given field The alternating asymmetric high/low electric field also creates a net drift, but now the drift velocity depends on the difference between the high- and low-field mobility of the ion Separation speed is inversely proportional to residence time, so a doubling of enables 16 to 64 times faster separation without loss of resolution No FAIMS With FAIMS L.J. Brown et al., Anal. Chem., 2012, 84 (9), pp 4095–4103 500um gap - max field ~5V/um 100um gap - max field ~10V/um Large protein Small protein