Article
pubs.acs.org/ac
Electrostatic-Spray Ionization Mass Spectrometry
Liang Qiao,†,§ Romain Sartor,†,§ Natalia Gasilova,† Yu Lu,† Elena Tobolkina,† Baohong Liu,‡
and Hubert H. Girault*,†
†
Laboratoire d’Electrochimie Physique et Analytique, Ecole Polytechnique Fédérale de Lausanne, Station 6, CH-1015 Lausanne,
Switzerland
‡
Department of Chemistry, Institute of Biomedical Sciences, Fudan University, Shanghai, 200433, P.R. China
S Supporting Information
*
ABSTRACT: An electrostatic-spray ionization (ESTASI) method has been
used for mass spectrometry (MS) analysis of samples deposited in or on an
insulating substrate. The ionization is induced by a capacitive coupling
between an electrode and the sample. In practice, a metallic electrode is
placed close to but not in direct contact with the sample. Upon application
of a high voltage pulse to the electrode, an electrostatic charging of the
sample occurs leading to a bipolar spray pulse. When the voltage is positive,
the bipolar spray pulse consists first of cations and then of anions. This
method has been applied to a wide range of geometries to emit ions from
samples in a silica capillary, in a disposable pipet tip, in a polymer
microchannel, or from samples deposited as droplets on a polymer plate. Fractions from capillary electrophoresis were collected
on a polymer plate for ESTASI MS analysis.
E
emitter to induce voltage inside the emitter. Similarly, Zhang et
al. have used alternating current (AC) high voltages to perform
a contactless spray, where the charge states of peptides were
controllable.18 In comparison with classical ESI, the high
voltage was not directly applied to the sample solution, and no
electrode reaction could occur. This is an advantage when
electrochemical reactions during conventional ESI need to be
avoided. Even in a contact mode, pulsed spray sources were
found to be useful when mass spectrometry was coupled with
microseparation strategies for analyzing trace amounts of a
sample mixture.19−22 The former publications about contactless
spray were mainly focused on applications. A comprehensive
investigation on the working mechanisms is of interest for
development of new techniques and for opening a wide range
of applications for mass spectrometry analysis.
Herein, we employed a constant high voltage power supply
and an electrical circuit comprising two synchronized switches
to generate high voltage pulses for onsetting a spray. The
principle of this contactless spray was thoroughly investigated
and shown to stem from a capacitive coupling effect. Under the
application of a high voltage on the electrode isolated from
sample solution with respect to the MS at ground, two
capacitors are formed in series: electrode−insulator−sample
solution and sample solution−gas−MS inlet. During electrostatic charging of the capacitors under high voltage, charges
accumulate at the solution−gas interface. When the electrostatic pressure is larger than the Laplace pressure, spray
lectrospray is a phenomenon studied as early as 1749 when
Nollet described the spray from a metallic orifice that was
electrostatically electrified.1 Since the 1980s, electrospray
ionization (ESI) has been widely used as a powerful technique
to softly ionize large compounds from solution for mass
spectrometry (MS) analysis.2−4 Nowadays, ESI is a general
ionization technique that has been applied to a wide range of
biomolecules and coupled with various types of mass analyzers,
including ion traps (IT), time-of-flight (TOF), quadrupole,
Fourier-transform ion cyclotron resonance (FTICR), and ITorbitrap.5−8
ESI is based on the ejection of charged droplets from the tip
of a capillary or a microchannel. The droplets reduce in size
during the flight by solvent evaporation and/or by Coulomb
explosion to form gas phase ions representative of the species in
solution.9 In traditional ESI-MS, a direct current (DC) high
voltage is applied between an electrode contacting the solution
in a microchannel or a capillary and the mass spectrometer
acting as the counter electrode. When a current flows through
the electrospray emitter, electrochemical reactions occur at the
electrode|solution interface to ensure the electroneutrality of
the solution and to form charged droplets for ESI.9−11
Contactless electrochemical techniques based on capacitive
coupling have been developed either to monitor the adsorption
of molecules on a substrate by measuring variations of
interfacial capacitances12 or to measure the conductivity of an
electrolyte solution within an insulating container. The latter is
routinely used to detect chemicals fractionated by capillary
electrophoresis (CE).13−15
Recently, Cooks et al. have reported a contactless method to
generate a pulsed spray.16,17 A pulsed high voltage waveform
was applied on an electrode 2 mm away from a nanospray
© 2012 American Chemical Society
Received: May 16, 2012
Accepted: August 9, 2012
Published: August 9, 2012
7422
dx.doi.org/10.1021/ac301332k | Anal. Chem. 2012, 84, 7422−7430
Analytical Chemistry
Article
Scheme 1. Schematic Representation of the Setups for Electrostatic-Spray Ionization of Solutions from (a) a Microfluidic Chip
and (b) a Micropipet Tip; (c) Schematic Representation of the Electrostatic-Spray Ionization of Samples in a Micro-Droplet on
an Insulating Platea
a
The counter electrode can be either a mass spectrometer or simply a metallic plate. HV: high voltage; A: ammeter.
ionization happens. By grounding the electrode, capacitors are
discharged and spray of ions with opposite charges happens.
On the basis of this electrostatic capacitor charging−
discharging principle, we named the contactless spray technique
as electrostatic-spray ionization (ESTASI). We have tested this
ESTASI approach on different systems such as microfluidic
chips, micropipet tips, and silica capillaries. Furthermore, the
spray has been directly performed from microdroplets on an
insulating plate, making spray ionization of the samples possible
from any substrates. We have collected fractions from capillary
electrophoresis (CE) as microdroplets on an insulating plate for
capacitive ESTASI MS analysis. ESTASI should open a wide
range of applications for mass spectrometry analysis.
■
placed around the tip. By applying a high potential pulse on the
electrode using the two switches, ESTASI was realized.
To spray from microdroplets, droplets of the sample solution
(1 μL) were deposited on an poly(methylmethacrylate)
(PMMA) plate patterned with wells or hydrophilic patches.
The electrode was placed behind the plate close to the droplet
to induce the spray. The droplet was put in front of the mass
spectrometer inlet. By applying high potential pulses to the
electrode using the two switches, ESTASI was generated.
CE Coupled with ESTASI MS. A mixture of peptides
generated from the tryptic digestion of myoglobin was used as
sample for CE separation coupled with ESTASI. The digestion
was performed by incubating 5 mg/mL myoglobin (horse
heart) with 0.17 mg/mL trypsin (bovine pancreas) in a buffer
of 25 mM NH4HCO3 (1 mL) under 37 °C for 18 h. The
protein was denatured by heating at 100 °C for 5 min prior to
tryptic digestion.
Standard CE separation of the myoglobin tryptic digest (150
μM, 21 nL per sample injection) followed with UV detection
was first performed on an Agilent 7100 CE system (Agilent,
Waldbronn, Germany). An untreated fused silica capillary (50
μm inner diameter, 375 μm outside diameter, 51.5 cm effective
length, 60 cm total length) obtained from BGB analytik AG
(Böckten, Switzerland) was used for separation. Solution of
10% acetic acid, pH = 2, has been employed as a background
electrolyte. Sample was injected during 20 s at pressure of 42
mbar. Separation was performed at a constant voltage of 30 kV.
Afterward, the capillary was cut at the point of detection
window and then coated with a conductive silver ink (Ercon,
Wareham, MA, USA) over a length of 10 cm from the outlet
that was then fixed outside the CE apparatus. The same CE
separation was performed with the same sample, while the
fractions were directly collected on an insulating polymer plate
by a homemade robotic system that was previously designed for
EXPERIMENTAL SECTION
Information on chemicals and fabrication of microchip is given
in Supporting Information SI-1.
Electrostatic-Spray Ionization. When spraying from a
microchannel in a microfluidic chip or from a polyimide coated
silica capillary, an electrode was placed close to the microchannel or capillary. A high potential (±6 kV) was applied to
the electrode by closing switch 1, switch 2 being open as shown
in Scheme 1. The electrode was then grounded after opening
switch 1 and closing switch 2. A LabView program was written
to control the two switches to synchronize their operation. The
end of the microchannel was placed just in front of the mass
spectrometer inlet. A syringe pump was used to infuse a sample
solution into the microchannel at adjustable flow rates.
Warning: It is important to use proper isolation measures when
pulsing high voltages as electromagnetic spikes are generated.
To spray from a micropipet tip, 1 μL of sample solution was
previously loaded into the disposable tip. An electrode was
7423
dx.doi.org/10.1021/ac301332k | Anal. Chem. 2012, 84, 7422−7430
Analytical Chemistry
Article
Scheme 2. Equivalent Circuit of the Electrostatic-Spray Ionization during Capacitors (a) Charging and (b) Discharginga
a
The diode symbolizes the direction of the spray current: spray of cations in (a) and spray of anions in (b).
surface tension at the tip is not sufficient to prevent the
emission of charged droplets, and this second capacitor can be
considered as a leaky capacitor with a diode in parallel, Scheme
2a.
Indeed, electrospray occurs when the electrostatic pressure at
the tip of the emitter is larger than the Laplace pressure. The
electrostatic pressure is a function of the surface charge
accumulation
collecting CE fractions on a matrix-assisted laser desorption/
ionization (MALDI) sample plate.23 Silver ink coating was
connected to the ground during CE separation.
After drying all the droplets, the polymer plate was placed
between the electrode and MS inlet. One microliter of an acidic
buffer (1% acetic acid in 49% water and 50% methanol
(MeOH)) was deposited on each sample spot to dissolve the
peptides for ESTASI-MS detection.
For comparison, a parallel experiment was performed to
collect fractions from CE separation on a MALDI sample plate.
The fractions were then analyzed by a Bruker microflex
MALDI-TOF mass spectrometer (Bruker Daltonics, Bremen,
Germany).
Mass Spectrometer. The ESI and ESTASI-MS experiments were performed on a Thermo LTQ Velos mass
spectrometer. An enhanced trap scan rate (10 000 Th/s) was
selected to obtain a good spectral resolution. During ESTASI,
the spray voltage of the internal power source of the LTQ
Velos was set to 0. A battery that can provide a high voltage up
to 10 kV was used as the external power source to induce the
spray.
The MALDI-MS experiments were performed on a Bruker
microflex MALDI-TOF mass spectrometer. After drying the
fractions from CE, 1 μL of 10 mg/mL 2,5-dihydroxybenzoic
acid in 50% water and 50% acetonitrile was deposited on each
sample spot as matrix. The mass spectrometer worked under
the reflectron-scanning mode for all experiments.
pE = σ 2/2ε0
where σ is the surface charge density and ε0 is the permittivity
of vacuum. If the tip of the emitter, e.g., at the Taylor cone, is
assumed to be hemispherical, the Laplace pressure is simply
given by
PL = 2γ /r
where γ is the surface tension of the liquid and r is the radius of
the hemisphere from where droplets are formed. To obtain
electrospray, it is important to keep the surface of the solution
small enough to increase σ for a given interfacial charge stored
on the capacitor C2.
The equivalent capacitance of the whole system is given as:
Ceq = C1C2/(C1 + C2), where C is the capacitance of the
corresponding capacitor C. Upon application of a high voltage
U, the maximum charge on the second capacitor representing
the tip of the emitter is UCeq. By placing the emitter far from
the MS inlet, C2 becomes very small. Ceq is then equal to C2,
and the maximum charge at the tip of the emitter is UC2. Since
C2 is very small, this charge is too small to induce electrospray.
By placing the emitter close enough to MS inlet, C2 cannot be
neglected, and the charge at the tip is UCeq. When this charge is
large enough, the emitter will spray.
In the case of the microchip shown in Scheme 1a, the first
capacitor can be approximately viewed as a parallel-plate
capacitor, and the capacitance is
RESULTS AND DISCUSSION
ESTASI Principle. The ESTASI principle is demonstrated
here using a microfluidic chip consisting of a microchannel
(100 μm × 50 μm) and an electrode as shown in Scheme 1a.
The electrode was made with carbon ink and placed 2 mm
away from the microchannel as a band having a lateral surface
area of 50 μm × 5 mm. Either the mass spectrometer or a
metallic plate was used as the counter electrode. The metallic
plate was used for measuring the current transients.
When the electrode is connected to the high potential source
(switch 1 closed, switch 2 open), the system behaves as two
capacitors in series before the electrospray happens. The first
capacitor (C1) is a conductor (the electrode)−insulator−liquid
solution capacitor, and no current can flow through it; its
capacitance depends on the geometry of the device. The second
capacitor (C2) is at the spray emitter and is a liquid solution−
gas−metal (the counter electrode or mass spectrometer)
capacitor; its capacitance varies as a function of the position
of the emitter with respect to the mass spectrometer. When the
charge accumulated on the second capacitor is too large, the
■
C1 = ε0εrA /d
where A is the lateral surface area of the electrode (50 μm × 5
mm), d is the distance between the band electrode and the
microchannel (2 mm), and εr is the relative permittivity of the
insulator, here 3.5 for PI. Therefore, to facilitate ESTASI, one
should minimize d and maximize A. Specifically, with the
geometry of the microchip illustrated in Scheme 1a, the
capacitance of the first capacitor is around 4 fF, which could
supply a charge around 20 pC when applying a high voltage of
6 kV, which is enough to onset the spray as detailed in the
Supporting Information SI-2. These data should be compared
with ESI from a similar microchip but where the electrode is in
7424
dx.doi.org/10.1021/ac301332k | Anal. Chem. 2012, 84, 7422−7430
Analytical Chemistry
Article
acetic acid) was infused by a syringe pump into the
microchannel of the microfluidic chip to be sprayed and
analyzed by MS. The switches were operated by LabView to
perform a series of pulses.
ESTASI-MS results shown in Figure 2 were obtained under a
flow rate of 20 μL/h and the following time sequence
parameters: t1 = t4 = 0.05 s, t2 = 0.25 s, and t3 = 0.15 s. The
definitions of t1, t2, t3, and t4 are shown in Scheme SI-3,
Supporting Information. Figure 2a shows the total cation
current (TCC) recorded by the MS as a function of time, which
indicates a very stable pulsed ESTASI process. Zooming in at
the current signal, we found clearly in Figure 2b that the peaks
of TCC appeared periodically every 1.5 s. In each cycle, four
mass spectra were collected with similar time intervals,
indicating that the generation of each spectrum took 375 ms
and that one spectrum included 1.25 charging−discharging
cycles. Mass spectra collected from the peak (Figure 2c) of
TCC showed peaks of the doubly and triply protonated
angiotensin I. In contrast, a weak signal was generated from the
baseline of the TCC, Figure 2d. Each cycle of the MS detection
is formed by several procedures, including injection, ion
transfer, ion scanning, and ion detection. Only when the
electrostatic-spray ionization happens during the injection, an
intensive spectrum is obtained, corresponding to the cycle of
the TCC as a function of time. Here, the maximum injection
time was set to 50 ms.
The analytical figures of merit of ESTASI are further
characterized and compared with ESI. We found that peptides
or small proteins with a concentration of 50 nM could be
efficiently ionized by ESTASI and detected by a LTQ Velos
mass spectrometer (Figure 3a,c). The microchip emitter and
experimental conditions were the same as those in Figure 2.
The limit of detection (LOD) of ESTASI-MS was found to be
at 5 to 10 nM for standard peptides and at 10 to 15 nM for
standard small proteins (Figure 3b,d). For comparison, ESI was
performed with a similar microchip emitter, where the
geometry of microchannel is completely the same as the one
used for ESTASI while the electrode contacts directly the
solution inside the microchannel, shown as Scheme SI-2.1,
Supporting Information. The LOD was found to be at 20 to 25
nM and 45 to 50 nM for the same peptides and proteins,
respectively, under the same experimental conditions, Figure
SI-4.1, Supporting Information. Actually, the microchip ESI
shows slightly better LOD than ESI performed with the
standard commercial ionization source. The latter holds LOD
for the same peptide and protein of 45 to 50 nM and 65 to 70
nM, respectively (Figure SI-4.2, Supporting Information).
Detailed information on LOD characterization is in Supporting
Information SI-4. The better LOD of ESTASI compared to ESI
may stem from its pulsed spray nature or the higher voltages
used.
Large proteins can also be analyzed by ESTASI-MS. Bovine
serum albumin with a molecule weight of 66 kDa was well
detected by microchip ESTASI-MS, and the peaks for BSA with
various charge states were clearly observed, Figure SI-5a,
Supporting Information. When bovine lactoferrin (76 kDa) was
used, the protein could be observed by ESTASI-MS. However,
it was hard to distinguish the ions with different charge states,
Figure SI-5b, Supporting Information. Similar results were
obtained with ESI-MS under the same experimental conditions
for the two large proteins, Figure SI-5c,d, Supporting
Information. Despite the ionization method, the mass analyzer
(linear ion trap) and detector can also limit the detection of
contact with the solution. Typically, during microchip ESI, we
operate with a high voltage of 4 kV and a total ion current of
100 nA, which yields a resistance value for the air gap of 40 GΩ.
As the spray occurs, the second capacitor leaks and is
bypassed by a resistance (R) and the system behaves as a RC
circuit, with a time constant τ = RC1, i.e., around 0.2 ms for the
microchip. The experimental time constant is in the order of
0.1 s and stems mainly from the rise time of the switch box
system. Similar calculations for the other emitters are given in
the Supporting Information SI-2.
When part of the positive charge at the tip of the solution is
sprayed, the positive charge stays on the electrode, meaning
that an excess negative charge builds up in the solution during
the spray. To alleviate this problem, the electrode is then
disconnected from the power supply and grounded to discharge
both capacitors. Since the charge stored in the first capacitor is
large enough to induce electrospray, part of the negative charge
built up at the surface of the solution can be released by spray
of anions, Scheme 2b. By repeating this cycle, alternating
electrostatic-spray of cations and anions is realized during the
charging−discharging of the two capacitors.
Ion Current during ESTASI. Current transients generated
during ESTASI were measured using a metallic plate counter
electrode. ESTASI was realized by pushing a solution of 50%
MeOH, 49% H2O, and 1% acetic acid through the microchannel in the microchip, shown as Scheme 1a, and applying a
high potential pulse on the electrode integrated in the
microchip. The metallic plate was placed 0.5 cm away from
the microchip emitter. As shown in Figure 1, when a positive
Figure 1. Current transients measured (blue line) using a counter
electrode during ESTASI. The application of a high potential pulse is
shown as the dashed green line. A solution of 50% MeOH/49% H2O/
1% acetic acid was used for the electrostatic-spray. A high potential of
6 kV was applied on the electrode to induce the electrospray.
high voltage was applied, a positive spray current with a peak
value of around 100 nA was observed. A negative spray current
was detected as soon as the electrode was disconnected from
the power supply and grounded. The current transients
observed corroborate directly the capacitor charging−discharging principle of ESTASI.
ESTASI-MS from a Microchip: Standard Samples,
Limit of Detection, and Large Proteins. When a mass
spectrometer was used as the counter electrode, mass-to-charge
(m/z) ratios of the ions generated from ESTASI were analyzed.
Angiotensin I solution (0.1 mM in 50% MeOH/49% H2O/1%
7425
dx.doi.org/10.1021/ac301332k | Anal. Chem. 2012, 84, 7422−7430
Analytical Chemistry
Article
Figure 2. (a, b) Total cation current (TCC) as a function of time and (c, d) the mass spectra under positive MS mode. The ions were generated by
electrostatic-spray ionization with a microfluidic chip as emitter when a high potential (6 kV) was applied. The mass spectrometer was scanning
under a range of 150 Th to 2000 Th and a scan rate of 10 000 Th/s. Only a range from 200 Th to 1000 Th is shown here since the rest is blank
without any peaks.
Figure 3. The mass spectra of angiotensin I ((a) 50 nM and (b) 5 nM) and cytochrome C ((c) 50 nM and (d) 10 nM) under positive MS mode.
Samples were all prepared in a buffer of 50% MeOH/49% H2O/1% acetic acid. The ions were generated by ESTASI with a microchip emitter when
a high potential (6 kV) was applied. The sample solutions were infused into the microchip by a syringe pump at a flow rate of 20 μL/h. Pulse
sequence parameters were set to t1 = t4 = 0.05 s, t2 = 0.25 s, and t3 = 0.15 s.
large proteins. These results indicate that ESTASI does not
limit the ionization of large proteins up to 76 kDa compared to
ESI.
ESTASI-MS from a Silica Capillary: Sequential Spray of
Cations and Anions. ESTASI-MS was also tested with a
commercial polyimide coated silica capillary (50 μm inner
diameter, 360 μm outside diameter), the electrode being merely
a steel crocodile clip surrounding the outer surface of the
capillary, which can be simply viewed as a ring electrode to
generate a cylindrical capacitor with the electrolyte solution
7426
dx.doi.org/10.1021/ac301332k | Anal. Chem. 2012, 84, 7422−7430
Analytical Chemistry
Article
Figure 4. (a) The TCC as a function of time and (b) the mass spectrum of angiotensin I under the positive MS mode; (c) the total anion current
(TAC) as a function of time and (d) the mass spectrum of acetate anions under the negative MS mode. A silica capillary was used as the emitter for
ESTASI. The mass spectrometer was scanning under a range of 150 Th to 2000 Th (a and b) or 50 Th to 70 Th (c and d) and a scan rate of 10 000
Th/s. Only a range from 200 Th to 1000 Th is shown in (b) since the rest is blank without any peaks.
and 4b, the scan ranges were set to 150 Th to 2000 Th.
Therefore, the ion scanning procedure would take by itself 185
ms, already more than the half period of one charging−
discharging cycle. In Figure 4c, the scan range was 50 Th to 70
Th, corresponding to only 2 ms spent for ion scanning.
In addition to ESTASI-MS analysis of peptides, the analysis
of different proteins was also realized using the silica capillary as
the emitter. Two proteins, myoglobin and beta-lactoglobulin (5
μM in 50% MeOH/49% H2O/1% acetic acid, respectively),
were used to illustrate the principle, Figure SI-7.1, Supporting
Information. This shows that ESTASI is possible to be
conveniently performed after separation techniques, such as
liquid chromatography (LC) and CE, where silica capillaries are
usually used as the solution outlet. It should be mentioned that
both negative and positive high potentials can be used to
induce the ESTASI and produce cations for MS analysis
benefiting from the capacitor charging−discharging principle.
In contrast, MS should detect only negative ions during ESI
when a negative high potential is applied on the emitter and
when the MS is grounded. Even if the cations are formed
during negative ESI, they will not enter the MS because of the
electric field direction. In ESTASI, spray of anions happens
during capacitor charging, while spray of cations happens
during capacitor discharging when a negative high potential is
applied on the electrode and when the MS is grounded. These
ions can always enter the MS because the electric field direction
is also changing during capacitor charging−discharging.
Specifically, we detected myoglobin cations when a negative
high potential (−6 kV) was used to induce ESTASI and when
the ion trap was scanning cations, Figure SI-7.2, Supporting
Information.
ESTASI-MS from Disposable Pipet Tips. ESTASI-MS
was also performed with disposable plastic pipet tips (∼300 μm
inside the capillary, where the polymer coated silica worked as
the insulator of the capacitor.
Shown as Figure 4, ESTASI was performed by infusing the
angiotensin I solution (0.1 mM in 50% MeOH/49% H2O/1%
acetic acid) through the capillary at a flow rate of 20 μL/h and
applying a high potential (6 kV) sequence to the electrode (t1 =
t4 = 0.05 s, t2 = 0.25 s, and t3 = 0.15 s). When the mass
spectrometer was scanned in the positive mode, angiotensin I
cations were observed (Figure 4b) and the pulsed spray was
found to be quite stable from the TCC response in Figure 4a.
The TCC response showed a frequency of 1.1 Hz, indicating
that the generation of each spectrum took 450 ms and included
1.5 charging−discharging cycles.
When the mass spectrometer was scanned in the negative
mode, anions generated during the capacitor discharging were
detected. Under the experimental conditions, the anions were
mainly acetate ions ([Ac]−, 59 Th). Therefore, a m/z scan
range of 50−70 Th was selected. Figure 4c shows the zoomed
total anion current (TAC) as a function of time detected by the
MS. The peaks of TAC appeared with a frequency of 3.33 Hz,
synchronizing the LabView program (t1 = t4 = 0.05 s, t2 = 0.25
s, and t3 = 0.15 s). Since two spectra were collected in each
cycle, the generation of each spectrum took 150 ms and
included 0.5 charging−discharging cycles.
Figure 4d shows the acetate ions detected. Deprotonated
angiotensin I anions could also be observed during capacitor
discharging when the angiotensin I was dissolved in a basic
buffer of 20 mM NH4HCO3 in 50% water/50% methanol,
Figure SI-6, Supporting Information. The observation of both
cations and anions during ESTASI induced by a positive high
voltage also supports the charging−discharging principle.
During all MS experiments, an enhanced scan rate was set for
the LTQ Velos, corresponding to 10 000 Th/s. In Figures 2
7427
dx.doi.org/10.1021/ac301332k | Anal. Chem. 2012, 84, 7422−7430
Analytical Chemistry
Article
inner diameter at the end), with an electrode placed around the
tip to induce the spray, as shown in Scheme 1b. According to
the calculation in Supporting Information SI-2, the capacitance
of the first capacitor is large enough to induce electrospray. The
tip was loaded with 1 μL of angiotensin I solution (0.1 mM in
50% MeOH/49% H2O/1% acetic acid) and placed in front of
the mass spectrometer inlet. When a high voltage (6 kV) was
applied on the electrode, cations were produced by ESTASI
and detected by the mass spectrometer, Figure 5. Because of
Figure 5. TCC as a function of time and mass spectrum of angiotensin
I under positive MS mode. The ions for MS detection were generated
by ESTASI directly from a disposable micropipet tip as illustrated in
Scheme 1b. A positive potential of 6 kV was used and the time
sequence parameters were set to: t1 = t4 = 0.05 s, t2 = 0.25 s, and t3 =
0.15 s.
Figure 6. Mass spectra of (a) angiotensin I (100 μM in 99% H2O/1%
acetic acid) and (b) myoglobin (50 μM in 99% H2O/1% acetic acid)
under positive MS mode. The ions were generated by ESTASI when a
pulsed positive high potential (6 kV) was applied to the electrode and
when 1 μL of the sample solution was deposited on an insulating plate.
The LabView time sequence parameters were set to: t1 = t4 = 0.05 s, t2
= 0.25 s, and t3 = 0.15 s.
the lack of the flow and low amount of the sample, single
ESTASI was generated. With the micropipet tip as emitter, it is
possible to run high throughput analysis of biological samples
taken directly from a microtiter plate.
ESTASI-MS from Microdroplets on an Insulating Plate.
As shown in Supporting Information SI-2, microdroplets can
also be sprayed by ESTASI, and the molecules in the droplets
can be ionized for MS analysis. Arrays of microdroplets were
deposited on an insulating plate, which was a 2 mm-thick
PMMA plate patterned with wells (2 mm in diameter, 100 μm
deep, Scheme 1c and Figure SI-8, Supporting Information).
The electrode (platinum column with the diameter of 1 mm)
was placed behind the PMMA plate opposite the inlet of mass
spectrometer. By moving the insulating plate to bring the
microdroplet close to the mass spectrometer inlet and applying
pulsed high potential (6 kV), spray of the droplet was realized
and the generated ions were detected by MS.
One microliter of 0.1 mM angiotensin I solution in 99%
water and 1% acetic acid was deposited on the PMMA
substrate. Singly, doubly, and triply protonated angiotensin I
cations were generated by ESTASI and detected by the linear
ion trap mass spectrometer, Figure 6a. Changing the droplet to
1 μL of 0.05 mM myoglobin in 99% water and 1% acetic acid,
we obtained the mass spectrum of multiprotonated myoglobin
ions, Figure 6b. Furthermore, ESTASI-MS was used to analyze
a myoglobin tryptic digest (1 μL of 30 μM in 50% MeOH/49%
H2O/1% acetic acid) deposited on the polymer substrate.
Shown as Figure SI-9 and Table SI-9, Supporting Information,
12 peaks were observed corresponding to 11 peptides.
This droplet spray ionization method is quite novel and has a
great application potential. Since samples are deposited on a
substrate as droplets for ionization, droplet ESTASI holds some
similarity to matrix-assisted laser desorption/ionization
(MALDI), but of course, the present approach does not
require the use of a matrix nor of a laser. The present ESTASI
method can be easily synchronized with pulsed time-of-flight
analysis for more economical sample consumption. Combined
with automatic deposition techniques,23 droplet ESTASI is able
to work as an interface between separation methods, such as
CE and LC, and MS.
As a proof-of-concept, we have employed CE to separate a
mixture of peptides obtained from myoglobin digestion by
trypsin. During the separation, the fractions were collected on
an insulating polymer substrate for ESTASI-MS analysis with
the help of a homemade robotic system that was previously
designed for collecting CE fractions on a MALDI sample
plate.23 The CE separation with sample collection on a plate is
schematically illustrated in Scheme 3. Figure 7a shows the CEUV result of separated peptides. The peptides with the
migration time between 3.5 and 8.5 min were collected on
the polymer plate as 18 spots shown as Figure 7b. For
comparison, the same fractions were also collected on a
MALDI sample plate for MALDI MS analysis. Figure 7c,d
7428
dx.doi.org/10.1021/ac301332k | Anal. Chem. 2012, 84, 7422−7430
Analytical Chemistry
Article
peptide) when emitters with similar geometries were employed,
Figure SI-11, Supporting Information, Figure 2c, and Figure SI7.1a, Supporting Information. As shown in Figure 3 and Figures
SI-4 and SI-5, Supporting Information, ESTASI-MS shows
better LOD than ESI-MS and does not limit the detection of
large proteins. During all experiments, no distinct fragmentation of peptides or proteins was observed by ESTASI compared
to ESI, indicating the soft ionization nature of ESTASI.
Considering that the electrode doest not contact the sample
solution during ESTASI, the oxidation of samples on the
electrode can be avoided; therefore, ESTASI should be better
than ESI to avoid modification/fragmentation of the sample
during the spray.
In addition, the ESTASI can be more economical in the view
of sample consumption and MS scanning because of its pulsed
spray nature. By modifying the time sequence controlling the
switches and synchronizing the pulses with MS, ESTASI-MS
works only when it is requested but not continuously as in
conventional ESI-MS.
Scheme 3. CE Separation with Sample Collection on a Plate
shows the mass spectra of ESTASI-MS of fractions 9 and 10,
where one peptide was clearly found from each spectrum. This
result is confirmed by MALDI-MS, shown as Figure SI10.1,10.2, Supporting Information. The peptides observed by
ESTASI-MS and MALDI-MS from the CE fractions are
compared in Table SI-10, Supporting Information. Seventeen
peptides were detected by combining the two strategies from
the 18 fractions. Each method identified 15 peptides, where 13
peptides were observed by both ESTASI-MS and MALDI-MS.
These results indicated that the CE-ESTASI-MS strategy is as
powerful as the reported CE-MALDI-MS strategy and that the
two ionization strategies can be complementary.
ESTASI vs ESI. By taking advantage of the ESTASI
principle, different geometries and different applications can
be envisaged such as spraying from a droplet on an insulating
plate. The latter opens many possibilities in analytical
chemistry. With this technique, molecules on a surface, such
as a piece of tissue or an electrophoresis gel, can be directly
ionized for MS detection by simply adding a droplet of buffer
that can dissolve the target molecule. This provides an
alternative strategy to ionize a sample from a surface as
compared to MALDI and desorption electrospray ionization
(DESI).24 Interestingly, ESTASI can be operated on a single
pulse mode, meaning that analyses can be performed
consuming only a few attoliters of samples.
ESTASI-MS gave ion currents more or less as strong as ESIMS for the same concentrations of samples (protein or
■
CONCLUSION
This work clearly shows that ESTASI is a versatile technique
that offers many potential applications in analytical chemistry.
The principle is based on the charging of a liquid surface with
the geometry of the system that allows the focusing of the
electric field to onset the formation of a Taylor cone at the
surface of the liquid. From an experimental viewpoint, pulsed
voltages can be applied using a DC high voltage source and a
synchronized switch system. ESTASI was demonstrated on
different geometries including polymer coated silica capillaries,
disposable plastic micropipet tips, and polyimide microchips.
The technique was further applied to spray microdroplets
deposited on an insulating plate and to produce ions of the
chemicals dissolved in the droplet. ESTASI is ideally suited for
bioanalytical studies because of its very convenient connection
with many separation techniques.
Figure 7. (a, b) CE-UV of the myoglobin tryptic digestion. (c) ESTASI-MS of fraction 9. (d) ESTASI-MS of fraction 10.
7429
dx.doi.org/10.1021/ac301332k | Anal. Chem. 2012, 84, 7422−7430
Analytical Chemistry
■
Article
(18) Peng, Y.; Zhang, S.; Gong, X.; Ma, X.; Yang, C.; Zhang, X. Anal.
Chem. 2011, 83, 8863−8866.
(19) Lu, Y.; Zhou, F.; Shui, W. Q.; Bian, L. P.; Yang, P. Y. Anal.
Chem. 2001, 73, 4748−4753.
(20) Chao, B. F.; Chen, C. J.; Li, F. A.; Her, G. R. Electrophoresis
2006, 27, 2083−2090.
(21) Berggren, W. T.; Westphall, M. S.; Smith, L. M. Anal. Chem.
2002, 74, 3443−3448.
(22) Wei, J. F.; Shui, W. Q.; Zhou, F.; Lu, Y.; Chen, K. K.; Xu, G. B.;
Yang, P. Y. Mass Spectrom. Rev. 2002, 21, 148−162.
(23) Busnel, J.-M.; Josserand, J.; Lion, N.; Girault, H. H. Anal. Chem.
2009, 81, 3867−3872.
(24) Takats, Z.; Wiseman, J. M.; Gologan, B.; Cooks, R. G. Science
2004, 306, 471−473.
ASSOCIATED CONTENT
S Supporting Information
*
(1) Chemicals and microchip fabrication, (2) ESTASI
mechanism, (3) schematic illustration of the employed
LabView program, (4) limit of detection of native ESI, (5)
ESTASI-MS and ESI-MS analysis of large proteins, (6) ESTASI
of angiotensin anions under positive voltage, (7) analysis of
proteins by ESTASI-MS using a commercial silica capillary as
emitter, (8) pictures of the PMMA plate patterned with wells,
(9) ESTASI of a myoglobin digest deposited as droplets on the
polymer plate, (10) CE-ESTASI-MS and CE-MALDI-MS, and
(11) conventional ESI-MS of angiotenisn I and myoglobin.
This material is available free of charge via the Internet at
http://pubs.acs.org.
■
AUTHOR INFORMATION
Corresponding Author
*Fax: +41 (0)21 693 36 67. Tel: +41 (0)21 693 31 51. E-mail:
hubert.girault@epfl.ch.
Author Contributions
§
L.Q. and R.S. contributed equally. The manuscript was written
through contributions of all authors. All authors have given
approval to the final version of the manuscript.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
We thank the Swiss National Science Foundation for
supporting the project ″Analytical tools for proteome analysis
and redoxomics (200020−127142)″.
■
■
REFERENCES
(1) Nollet, J. A. Recherches sur les causes particulières des phénomènes
électriques, 1ère ed.; Chez les frères Guerin: Paris, 1749.
(2) Yamashita, M.; Fenn, J. B. J. Phys. Chem. 1984, 88, 4451−4459.
(3) Alexandrov, M. L.; Gall, L. N.; Krasnov, N. V.; Nikolaev, V. I.;
Pavlenko, V. A.; Shkurov, V. A. Int. J. Mass Spectrom. Ion Processes
1983, 54, 231−235.
(4) Fenn, J. B.; Mann, M.; Meng, C. K.; Wong, S. F.; Whitehouse, C.
M. Science 1989, 246, 64−71.
(5) Wang, Y.; Schubert, M.; Ingendoh, A.; Franzen, J. Rapid Commun.
Mass Spectrom. 2000, 14, 12−17.
(6) Lu, W. Y.; Callahan, J. H.; Fry, F. S.; Andrzejewski, D.; Musser, S.
M.; Harrington, P. D. Talanta 2011, 84, 1180−1187.
(7) Williams, K.; Hawkridge, A. M.; Muddiman, D. C. J. Am. Soc.
Mass Spectrom. 2007, 18, 1−7.
(8) Hardman, M.; Makarov, A. A. Anal. Chem. 2003, 75, 1699−1705.
(9) Abonnenc, M.; Qiao, L. A.; Liu, B. H.; Girault, H. H. In Annual
Review of Analytical Chemistry; Annual Reviews: Palo Alto, 2010; Vol.
3, pp 231−254.
(10) Rohner, T. C.; Lion, N.; Girault, H. H. Phys. Chem. Chem. Phys.
2004, 6, 3056−3068.
(11) Van Berkel, G. J.; Kertesz, V. Anal. Chem. 2007, 79, 5510−5520.
(12) Gamby, J.; Abid, J. P.; Girault, H. H. J. Am. Chem. Soc. 2005,
127, 13300−13304.
(13) Fercher, G.; Haller, A.; Smetana, W.; Vellekoopt, M. J. Anal.
Chem. 2010, 82, 3270−3275.
(14) Tuma, P.; Samcova, E.; Stulik, K. Electroanalysis 2011, 23,
1870−1874.
(15) Wang, J.; Chen, G.; Muck, A. Anal. Chem. 2003, 75, 4475−4479.
(16) Huang, G.; Li, G.; Ducan, J.; Ouyang, Z.; Cooks, R. G. Angew.
Chem., Int. Ed. 2011, 50, 2503−2506.
(17) Huang, G.; Li, G.; Cooks, R. G. Angew. Chem., Int. Ed. 2011, 50,
9907−9910.
7430
dx.doi.org/10.1021/ac301332k | Anal. Chem. 2012, 84, 7422−7430