Evolution of ElectroSpray Ionization for LCMS

In the 1980s, conventional ElectroSpray Ionization (ESI) revolutionized the identification and characterization of non-volatile molecules using LC/ESI-MS (50-5000 ul/min).  ESI utilizes a high voltage emitter with a high flow sheath gas to ionize and desolvate ions, but excessive gas dilutes the sample ions such that only a small percentage of the sample gets into the MS for analysis.  LC/ESI-MS is easy to use, provides robust operation, and allows high throughput analyses making it the technique of choice for pharmaceutical scientists involved in all stages of drug discovery and drug development.

In the 1990s, NanoSpray Ionization (NSI) made nanoLC/NSI-MS (10-1000 nl/min) the gold standard for proteomics research.  NSI utilizes a high voltage tapered capillary emitter without sheath gas to ionize and desolvate sample ions in a Taylor cone spray, which allows most of the sample ions into the MS.  Since NSI (like ESI) is concentration dependant, sensitivity increases as flows decrease.  NSI can be difficult to setup and maintain (generally requires X,Y,Z positioning and cameras to monitor the spray) and the tapered tips are prone to clogging and fouling, which makes robust 24/7 operation difficult to achieve.  This lack of robustness, coupled with the low throughput of nanoLC makes NSI less attractive to pharmaceutical scientists or proteomics researchers doing high throughput protein quantitation.

The Advance CaptiveSpray Ionization (CSI) source provides the next step in the evolution of LC/MS (0.1-100 ul/min), bridging the flow, sensitivity, and robust operational gap between ESI and NSI. CSI utilizes a high voltage non-tapered capillary emitter attached to the vacuum of the MS, which pulls in gas around the emitter, desolvating and funneling all the sample ions into the MS.   It’s “Plug and Play” operation requires no cameras or xyz alignment, providing ESI robustness with NSI sensitivity.  CaptiveSpray is used by proteomics researchers who want higher throughput and/or robust operation for qualitative and quantitative identification, as well as by pharmaceutical scientists who need enhanced sensitivity without sacrificing robustness or throughput.

Fig 1  Evolution of CaptiveSpray

Fig 2  CaptiveSpray Components

Gas inlet 1 provides axial gas flow around the emitter to keep liquid from rolling back onto the outer surface of the emitter.  This eliminates sample precipitation on the emitter outlet responsible for nano-flow spray instability in conventional tapered pulled tips.   Gas inlet 2 provides gas flow directed at the emitter outlet, nebulizing the liquid into a Taylor cone spray independent of the flow rate and emitter ID/OD.  This constant spray eliminates the need for carefully balanced flow, voltage, and xyz adjustment for optimum performance in NanoSpray. 

Fig 3  CaptiveSpray Emitter Tip Assembly

Fig 4  Spray Vortex Focusing

How Does an Advance CaptiveSpray Source Interface LC to MS

CaptiveSpray is simple to setup and operate, providing unattended 24/7 robust spray at flows from 250 nl/min to 20 µl/min.  To setup CaptiveSpray, mount the source housing on the MS and attach the column to the source probe.  Attach the LC flow to the column inlet and start the flow of liquid into the column.  Insert the source probe into the housing on the MS and then turn on the voltage (1000 – 3000 V).  Monitor the MS ions in the Tune Mode of the MS.  Decrease the voltage in 0.2kV steps to find the threshold of ionization.  Once the ionization threshold is found, increase the voltage by 0.2kV, and begin routine operation. 

The design of CaptiveSpray provides optimum positioning of the probe, eliminating the need for X,Y, Z positioning.  Because the vacuum of the MS pulls in gas around the emitter tip, liquid never rolls back on the tip (even with pure aqueous mobile phase), resulting in a stable spray across the LC gradient.   With a stable spray and no fouling of the untapered CaptiveSpray emitter tip, there is no need to monitor the spray with expensive video equipment.

Although the Advance CaptiveSpray source was developed initially for Thermo MS, Michrom is actively collaborating with other MS vendors to expand CaptiveSpray technology to all LCMS platforms.  The Advance CaptiveSpray source for ABI/Sciex MS systems is introduced, with shipments in the 4th quarter of 09.   Bruker is currently in development with anticipated launch in spring 2010.  We have also had interest from other vendors, including Waters, Agilent, Varian, Hitachi, Shimadzu. 

CaptiveSpray Bridges the Gap Between ElectroSpray and NanoSpray Ionization

The Axial Desolvation Vacuum Assisted Nano Capillary Electrospray (ADVANCE) source was introduced in 1996 as an alternative to NanoSpray for proteomics applications.  This source has since evolved into a new electrospray technology entitled CaptiveSpray Ionization (CSI), which bridges the gap between NanoSpray Ionization (NSI) and conventional Electrospray Ionization (ESI).   With hundreds of customers using the source over the past four years, Michrom has characterized and optimized the Advance CaptiveSpray source to work over a broad flow range for a variety of LCMS applications. 

 Fig 5  CaptiveSpray Flow and Sensitivity Bridges the gap between NSI and ESI

Qualitative Proteomics – Biomarker Discovery

For qualitative proteomics applications (i.e. biomarker discovery), where sensitivity is the most important LCMS parameter, most researchers are using flows between 200 – 1000 nl/min using 75 – 100 µ ID LC columns with NSI-MS.  CaptiveSpray works very well in this flow range and provides comparable sensitivity to a well optimized NSI-MS system, with the added value of simple, robust (24/7) operation.   Although this flow range and column size range is optimum for NSI-MS, CSI-MS has an optimum flow range between 500 and 5000 nl/min using 100 – 200 µ ID LC columns. The use of larger columns expands the loading capacity providing more robust column performance, enhancing sensitivity for proteomics applications.   

Quantitative Proteomics – Biomarker Validation

In quantitative proteomics (i.e. biomarker validation), throughput and robustness are as important as sensitivity, making CSI-MS an ideal choice for these applications.  Although most quantitative proteomics is currently being done using NSI-MS, several researchers have sacrificed sensitivity and switched to ESI-MS in order to achieve throughput and robustness required for their applications.   Current users of CSI-MS who do quantitative proteomics have found it to be ideally suited to this application, providing ESI robustness and ease of use while delivering the sensitivity of nanaospray. 

Small Molecule  - ADMET, DMPK

Since biomarker validation and screening in clinical research samples requires the same degree of throughput, robustness and sensitivity as many pharmaceutical discovery applications (ADMET, DMPK, etc), CaptiveSpray is finding increased utility with pharmaceutical researchers.  When coupled with a HT capLC, CSI-MS provides ESI throughput and robustness with NSI sensitivity and allows sample prep and MS analysis of much smaller sample sizes and reduced mobile phase demands.

CaptiveSpray Improves Throughput and Sensitivity for Complex Samples

For complex proteome analyses, such as whole cell lysate digests, LCMS runs of 1 – 4 hours are required to get enough separation and MSMS detection time to identify the thousands of peptides in these complex samples.  In the example below, 1 ug of an E. coli lysate digest (maximum loading capacity of a 75 µ ID column) was separated over 2 hours at 200 nl/min and NSI-MS/MS detection identified 867 proteins in a total run time of 200 minutes.  

10 ug of an E. coli lysate digest (maximum loading capacity of a 200 µ ID column) was then separated over 30 minutes at 2 µl/min and CSI-MS/MS detection identified 942 proteins in a total run time of 40 minutes.  The higher flow rate and concomitant larger column ID used with CaptiveSpray results in a five fold higher sample throughput and a larger number of identified proteins relative to the lower flows and smaller columns used with NanoSpray.

Fig 6  Increaed loading capacity and identification of whole cell lysate digests

CaptiveSpray Increases Mass Spec Utilization Time and Peak Capacity

For less complex samples (i.e 1D gel slice digests), attempts to increase sample throughput using NSI-MS have met with limited success, as the lower flows utilized in nanoLC limit the speed with which even simple samples can be analyzed.  The upper trace in the figure below was taken from a NanoSpray manufacturer’s application note, where 100 fmol of a tryptic digest of BSA was separated on a 75 µ ID by 100 mm long column with a 20 min gradient and 45 min total run time at 250 nl/min, with detection by NSI-MS. 

For higher throughput runs of simple samples at nanoliter flows, the impact of gradient delay and column re-equilibration becomes much more significant than for the longer runs used for more complex samples.  In the upper trace of the example below, the effective mass spec utilization time (from first to last sample component separated over the gradient run) is ~ 33% (15 min of separation and MS/MS information in the 45 min run).   For nanoflow separations, the effective peak capacity (total separation time divided by average peak width) is also limited by this time loss, where the peak capacity in the NSI example below is ~ 50 (15 min / 0.3 min).

The lower trace in the figure below was taken from a CaptiveSpray run designed to match the NSI separation, where 100 fmol of a tryptic digest of BSA was separated on a 200 µ ID by 100 mm long column with a 24 min gradient and 30 min total run time at 1 µl/min, with detection by CSI-MS. For this CaptiveSpray run, the effective mass spec utilization time is ~ 80% (24 minutes of separation and MS/MS information in the 30 minute run), and the peak capacity in the CSI example below is ~ 200 (24 min / 0.15 min).

Fig 7  Chromatographic front end increases mass spec utilization by reducing delay and re-equilibration

Using CaptiveSpray to Increase Sample Throughput for Simple Samples

Simple proteomics samples (i.e. 2D gel spot digests) can be analyzed even faster using higher flows with CaptiveSpray LCMS, as shown in the example below.  Using a high throughput nano-capillary LC system, 1 fmol of BSA digest was first analyzed on a 75 µ x 150 mm column with a 30 min gradient and 50 min total run time at 200 nl/min, with detection by CSI-MS (upper trace).   The same LCMS was then used to analyze 1 fmol of BSA digest on a 200 µ x 50 mm column with a 3 min gradient and 5 min total run time at 4 µl/min (lower trace).

For this simple protein digest, the two LC-CSI/MS runs showed similar sensitivity, resolution and overall peptide detection, but at the higher flow, sample throughput was increased 10 fold.  With a 5 min total LCMS run time, 288 gel spot digests from a 2D gel could be analyzed in 24 hours.

 Fig 8  10x throughput increase with equivalent sensitivity and coverage

CapLC-CaptiveSpray-MS Versus UHPLC-ElectroSpray-MS

For high throughput protein quantitation applications (i.e. biomarker validation), throughput and robustness become as important as resolution and sensitivity in order to insure high quality results.  Researchers with very high throughput requirements have adopted ESI-MS as the technique of choice, but the loss in sensitivity can be an issue when trying to quantitate low abundance proteins.

This example shows that CaptiveSpray MS is ideally suited for high throughput quantitative protein applications, as it can provide the throughput, resolution and robustness of ESI without sacrificing sensitivity for low abundance protein quantitation.

 Fig 9   50X sensitivity gain of  HPLC/CSI vs Ballistic UHPLC/ESI

Capillary UHPLC-CaptiveSpray-MS Versus Nano UPLC-NanoSpray-MS

Nanoflow UHPLC coupled with NSI-MS is shown in multiple applications to increase throughput, resolution and sensitivity required for low abundance protein quantitation using nano-spray ionization.   Side-by-side comparisons of the throughput gains from sub-two micron particles in published application notes illustrate the advantages of CSI over small particles with NSI.

The upper trace shows a 20 fmol BSA digest separated on a 0.1x50mm column with a 4 min gradient and 8 min total run time at 1 µl/min, with NSI-MS detection.  For this run, the effective mass spec utilization time is ~ 20% (1.5 min of separation and MS/MS information in the 8 min run), and the peak capacity in this run is ~ 50 (1.5 min / 0.03 min).  A 20 fmol BSA digest separated on a 0.2x50mm column with a 4 min gradient and 5 min total run time at 4 µl/min, with CSI-MS detection.  For this run, the effective mass spec utilization time is ~ 60% (3 min of separation and MS/MS information in the 5 min run), and the peak capacity in this run is ~ 100 ( 3 min / 0.03 min).

 Fig 10  CSI advantage of higher flow/lower delay provides more MS utilization

Evolution of ElectroSpray Ionization for LCMS

In the 1980s, conventional ElectroSpray Ionization (ESI) revolutionized the identification and characterization of non-volatile molecules using LC/ESI-MS (50-5000 ul/min).  ESI utilizes a high voltage emitter with a high flow sheath gas to ionize and desolvate ions, but excessive gas dilutes the sample ions such that only a small percentage of the sample gets into the MS for analysis.  LC/ESI-MS is easy to use, provides robust operation, and allows high throughput analyses making it the technique of choice for pharmaceutical scientists involved in all stages of drug discovery and drug development.

In the 1990s, NanoSpray Ionization (NSI) made nanoLC/NSI-MS (10-1000 nl/min) the gold standard for proteomics research.  NSI utilizes a high voltage tapered capillary emitter without sheath gas to ionize and desolvate sample ions in a Taylor cone spray, which allows most of the sample ions into the MS.  Since NSI (like ESI) is concentration dependant, sensitivity increases as flows decrease.  NSI can be difficult to setup and maintain (generally requires X,Y,Z positioning and cameras to monitor the spray) and the tapered tips are prone to clogging and fouling, which makes robust 24/7 operation difficult to achieve.  This lack of robustness, coupled with the low throughput of nanoLC makes NSI less attractive to pharmaceutical scientists or proteomics researchers doing high throughput protein quantitation.

The Advance CaptiveSpray Ionization (CSI) source provides the next step in the evolution of LC/MS (0.1-100 ul/min), bridging the flow, sensitivity, and robust operational gap between ESI and NSI. CSI utilizes a high voltage non-tapered capillary emitter attached to the vacuum of the MS, which pulls in gas around the emitter, desolvating and funneling all the sample ions into the MS.   It’s “Plug and Play” operation requires no cameras or xyz alignment, providing ESI robustness with NSI sensitivity.  CaptiveSpray is used by proteomics researchers who want higher throughput and/or robust operation for qualitative and quantitative identification, as well as by pharmaceutical scientists who need enhanced sensitivity without sacrificing robustness or throughput.

Fig 1  Evolution of CaptiveSpray

Fig 2  CaptiveSpray Components

Gas inlet 1 provides axial gas flow around the emitter to keep liquid from rolling back onto the outer surface of the emitter.  This eliminates sample precipitation on the emitter outlet responsible for nano-flow spray instability in conventional tapered pulled tips.   Gas inlet 2 provides gas flow directed at the emitter outlet, nebulizing the liquid into a Taylor cone spray independent of the flow rate and emitter ID/OD.  This constant spray eliminates the need for carefully balanced flow, voltage, and xyz adjustment for optimum performance in NanoSpray. 

Fig 3  CaptiveSpray Emitter Tip Assembly

Fig 4  Spray Vortex Focusing

How Does an Advance CaptiveSpray Source Interface LC to MS

CaptiveSpray is simple to setup and operate, providing unattended 24/7 robust spray at flows from 250 nl/min to 20 µl/min.  To setup CaptiveSpray, mount the source housing on the MS and attach the column to the source probe.  Attach the LC flow to the column inlet and start the flow of liquid into the column.  Insert the source probe into the housing on the MS and then turn on the voltage (1000 – 3000 V).  Monitor the MS ions in the Tune Mode of the MS.  Decrease the voltage in 0.2kV steps to find the threshold of ionization.  Once the ionization threshold is found, increase the voltage by 0.2kV, and begin routine operation. 

The design of CaptiveSpray provides optimum positioning of the probe, eliminating the need for X,Y, Z positioning.  Because the vacuum of the MS pulls in gas around the emitter tip, liquid never rolls back on the tip (even with pure aqueous mobile phase), resulting in a stable spray across the LC gradient.   With a stable spray and no fouling of the untapered CaptiveSpray emitter tip, there is no need to monitor the spray with expensive video equipment.

Although the Advance CaptiveSpray source was developed initially for Thermo MS, Michrom is actively collaborating with other MS vendors to expand CaptiveSpray technology to all LCMS platforms.  The Advance CaptiveSpray source for ABI/Sciex MS systems is introduced, with shipments in the 4th quarter of 09.   Bruker is currently in development with anticipated launch in spring 2010.  We have also had interest from other vendors, including Waters, Agilent, Varian, Hitachi, Shimadzu. 

CaptiveSpray Bridges the Gap Between ElectroSpray and NanoSpray Ionization

The Axial Desolvation Vacuum Assisted Nano Capillary Electrospray (ADVANCE) source was introduced in 1996 as an alternative to NanoSpray for proteomics applications.  This source has since evolved into a new electrospray technology entitled CaptiveSpray Ionization (CSI), which bridges the gap between NanoSpray Ionization (NSI) and conventional Electrospray Ionization (ESI).   With hundreds of customers using the source over the past four years, Michrom has characterized and optimized the Advance CaptiveSpray source to work over a broad flow range for a variety of LCMS applications. 

 Fig 5  CaptiveSpray Flow and Sensitivity Bridges the gap between NSI and ESI

Qualitative Proteomics – Biomarker Discovery

For qualitative proteomics applications (i.e. biomarker discovery), where sensitivity is the most important LCMS parameter, most researchers are using flows between 200 – 1000 nl/min using 75 – 100 µ ID LC columns with NSI-MS.  CaptiveSpray works very well in this flow range and provides comparable sensitivity to a well optimized NSI-MS system, with the added value of simple, robust (24/7) operation.   Although this flow range and column size range is optimum for NSI-MS, CSI-MS has an optimum flow range between 500 and 5000 nl/min using 100 – 200 µ ID LC columns. The use of larger columns expands the loading capacity providing more robust column performance, enhancing sensitivity for proteomics applications.   

Quantitative Proteomics – Biomarker Validation

In quantitative proteomics (i.e. biomarker validation), throughput and robustness are as important as sensitivity, making CSI-MS an ideal choice for these applications.  Although most quantitative proteomics is currently being done using NSI-MS, several researchers have sacrificed sensitivity and switched to ESI-MS in order to achieve throughput and robustness required for their applications.   Current users of CSI-MS who do quantitative proteomics have found it to be ideally suited to this application, providing ESI robustness and ease of use while delivering the sensitivity of nanaospray. 

Small Molecule  - ADMET, DMPK

Since biomarker validation and screening in clinical research samples requires the same degree of throughput, robustness and sensitivity as many pharmaceutical discovery applications (ADMET, DMPK, etc), CaptiveSpray is finding increased utility with pharmaceutical researchers.  When coupled with a HT capLC, CSI-MS provides ESI throughput and robustness with NSI sensitivity and allows sample prep and MS analysis of much smaller sample sizes and reduced mobile phase demands.

CaptiveSpray Improves Throughput and Sensitivity for Complex Samples

For complex proteome analyses, such as whole cell lysate digests, LCMS runs of 1 – 4 hours are required to get enough separation and MSMS detection time to identify the thousands of peptides in these complex samples.  In the example below, 1 ug of an E. coli lysate digest (maximum loading capacity of a 75 µ ID column) was separated over 2 hours at 200 nl/min and NSI-MS/MS detection identified 867 proteins in a total run time of 200 minutes.  

10 ug of an E. coli lysate digest (maximum loading capacity of a 200 µ ID column) was then separated over 30 minutes at 2 µl/min and CSI-MS/MS detection identified 942 proteins in a total run time of 40 minutes.  The higher flow rate and concomitant larger column ID used with CaptiveSpray results in a five fold higher sample throughput and a larger number of identified proteins relative to the lower flows and smaller columns used with NanoSpray.

Fig 6  Increaed loading capacity and identification of whole cell lysate digests

CaptiveSpray Increases Mass Spec Utilization Time and Peak Capacity

For less complex samples (i.e 1D gel slice digests), attempts to increase sample throughput using NSI-MS have met with limited success, as the lower flows utilized in nanoLC limit the speed with which even simple samples can be analyzed.  The upper trace in the figure below was taken from a NanoSpray manufacturer’s application note, where 100 fmol of a tryptic digest of BSA was separated on a 75 µ ID by 100 mm long column with a 20 min gradient and 45 min total run time at 250 nl/min, with detection by NSI-MS. 

For higher throughput runs of simple samples at nanoliter flows, the impact of gradient delay and column re-equilibration becomes much more significant than for the longer runs used for more complex samples.  In the upper trace of the example below, the effective mass spec utilization time (from first to last sample component separated over the gradient run) is ~ 33% (15 min of separation and MS/MS information in the 45 min run).   For nanoflow separations, the effective peak capacity (total separation time divided by average peak width) is also limited by this time loss, where the peak capacity in the NSI example below is ~ 50 (15 min / 0.3 min).

The lower trace in the figure below was taken from a CaptiveSpray run designed to match the NSI separation, where 100 fmol of a tryptic digest of BSA was separated on a 200 µ ID by 100 mm long column with a 24 min gradient and 30 min total run time at 1 µl/min, with detection by CSI-MS. For this CaptiveSpray run, the effective mass spec utilization time is ~ 80% (24 minutes of separation and MS/MS information in the 30 minute run), and the peak capacity in the CSI example below is ~ 200 (24 min / 0.15 min).

Fig 7  Chromatographic front end increases mass spec utilization by reducing delay and re-equilibration

Using CaptiveSpray to Increase Sample Throughput for Simple Samples

Simple proteomics samples (i.e. 2D gel spot digests) can be analyzed even faster using higher flows with CaptiveSpray LCMS, as shown in the example below.  Using a high throughput nano-capillary LC system, 1 fmol of BSA digest was first analyzed on a 75 µ x 150 mm column with a 30 min gradient and 50 min total run time at 200 nl/min, with detection by CSI-MS (upper trace).   The same LCMS was then used to analyze 1 fmol of BSA digest on a 200 µ x 50 mm column with a 3 min gradient and 5 min total run time at 4 µl/min (lower trace).

For this simple protein digest, the two LC-CSI/MS runs showed similar sensitivity, resolution and overall peptide detection, but at the higher flow, sample throughput was increased 10 fold.  With a 5 min total LCMS run time, 288 gel spot digests from a 2D gel could be analyzed in 24 hours.

 Fig 8  10x throughput increase with equivalent sensitivity and coverage

CapLC-CaptiveSpray-MS Versus UHPLC-ElectroSpray-MS

For high throughput protein quantitation applications (i.e. biomarker validation), throughput and robustness become as important as resolution and sensitivity in order to insure high quality results.  Researchers with very high throughput requirements have adopted ESI-MS as the technique of choice, but the loss in sensitivity can be an issue when trying to quantitate low abundance proteins.

This example shows that CaptiveSpray MS is ideally suited for high throughput quantitative protein applications, as it can provide the throughput, resolution and robustness of ESI without sacrificing sensitivity for low abundance protein quantitation.

 Fig 9   50X sensitivity gain of  HPLC/CSI vs Ballistic UHPLC/ESI

Capillary UHPLC-CaptiveSpray-MS Versus Nano UPLC-NanoSpray-MS

Nanoflow UHPLC coupled with NSI-MS is shown in multiple applications to increase throughput, resolution and sensitivity required for low abundance protein quantitation using nano-spray ionization.   Side-by-side comparisons of the throughput gains from sub-two micron particles in published application notes illustrate the advantages of CSI over small particles with NSI.

The upper trace shows a 20 fmol BSA digest separated on a 0.1x50mm column with a 4 min gradient and 8 min total run time at 1 µl/min, with NSI-MS detection.  For this run, the effective mass spec utilization time is ~ 20% (1.5 min of separation and MS/MS information in the 8 min run), and the peak capacity in this run is ~ 50 (1.5 min / 0.03 min).  A 20 fmol BSA digest separated on a 0.2x50mm column with a 4 min gradient and 5 min total run time at 4 µl/min, with CSI-MS detection.  For this run, the effective mass spec utilization time is ~ 60% (3 min of separation and MS/MS information in the 5 min run), and the peak capacity in this run is ~ 100 ( 3 min / 0.03 min).

 Fig 10  CSI advantage of higher flow/lower delay provides more MS utilization