Pressurized sample infusion: An easily calibrated, low volume pumping system for ESI-MS analysis of reactions

Article abstract of DOI:10.1016/j.ijms.2012.03.007

Pressurized sample infusion (PSI) continuously delivers solution from a flask through capillary tubing at a flow rate that makes it suitable as a low internal volume pumping method for electrospray ionization mass spectrometry (ESI-MS). The flow rate can be predicted from the applied pressure by using the Hagen-Poiseuille equation, but variability in internal diameter of the PEEK tubing used is such that each individual length should be calibrated if accurate results are sought. Once calibrated, the values hold well for different solvents across a range of pressures. The technique has been exemplified in two reactions: the deprotection of a protected amino acid using trifluoroacetic acid, and in the platinum-catalyzed redistribution of two silanes. In both cases, the abundances of starting material, product(s) and intermediates were tracked in real time as the reaction proceeded.

Full text of DOI:10.1016/j.ijms.2012.03.007

International Journal of Mass Spectrometry 323–324 (2012) 8–13
Contents lists available at SciVerse ScienceDirect
International Journal of Mass Spectrometry
journal homepage: www.elsevier.com/locate/ijms
Classroom Article
Pressurized sample infusion: An easily calibrated, low volume pumping system
for ESI-MS analysis of reactions
Krista L. Vikse, Zohrab Ahmadi, Jingwei Luo, Nicole van der Wal, Kevin Daze, Nichole Taylor,
∗
J. Scott McIndoe
Department of Chemistry, University of Victoria, P.O. Box 3065, Victoria, BC, Canada V8W3V6
a r t i c l e i n f o
a b s t r a c t
Article history:
Pressurized sample infusion (PSI) continuously delivers solution from a flask through capillary tubing at
a flow rate that makes it suitable as a low internal volume pumping method for electrospray ionization
mass spectrometry (ESI-MS). The flow rate can be predicted from the applied pressure by using the
Hagen–Poiseuille equation, but variability in internal diameter of the PEEK tubing used is such that
each individual length should be calibrated if accurate results are sought. Once calibrated, the values
hold well for different solvents across a range of pressures. The technique has been exemplified in two
reactions: the deprotection of a protected amino acid using trifluoroacetic acid, and in the platinum-
catalyzed redistribution of two silanes. In both cases, the abundances of starting material, product(s) and
intermediates were tracked in real time as the reaction proceeded.
Received 21 January 2012
Received in revised form 22 March 2012
Accepted 23 March 2012
Available online 9 April 2012
Keywords:
Electrospray ionization
Sample introduction
Reaction monitoring
©
2012 Elsevier B.V. All rights reserved.
1
. Introduction
mass spectrometer source conventionally through means of the LC
pump [6]. Pressurized chambers have been used for capillary elec-
trophoresis/electrospray ionization, in which the solution is driven
through the separation capillary and electrospray capillary by an
electropneumatically regulated supply of nitrogen. [7] A sample
infusion bomb apparatus using 5–20 psi of nitrogen to drive a sam-
ple through 50 ␮m i.d. fused silica tubing into an ESI-MS, capable of
on-line sample concentration and clean-up, has also been demon-
strated [8].
Pressurized sample infusion (PSI) is a simple method for infusing
a dilute solution directly into a mass spectrometer without fur-
ther treatment [9]. PSI involves no mechanical pumping; only what
is essentially a cannula transfer from a flask into the instrument
through a length of chromatography tubing, and we developed it
with the intention of applying to the online monitoring of catalytic
reactions. It works well in this application, and with simple nor-
malization of the data to total ion current or an internal standard
it provides very high data density along with good point-to-point
reproducibility of quantifiable data [10]. The total solvent volume
of the pumping system is simply the volume of the tubing. While
the experimental set-up inherently allows the analysis of com-
pounds under anaerobic conditions, it works equally well for air-
and moisture-stable materials, so its application is much broader
than just organometallic catalysis, and can be used in all situations
where a simple, low internal volume pumping system is desirable.
All required materials for a PSI system are easily accessible: a
flask with two ground glass joints (or one joint and an arm), rub-
ber septum, rubber hose, a short length of PEEK tubing (∼0.5 m),
a PEEK chromatography fitting (nut and ferrule) and a source of
There are numerous methods of introducing sample solutions
to an electrospray ionization mass spectrometer (ESI-MS) [1]. Most
commonly, an HPLC pump delivers the analyte through a column
in a flow of mobile phase, and some of the eluant from the col-
umn passes into the source of the spectrometer, hence LC–MS
[
2]. Syringe pumps expel solutions directly from a syringe con-
taining the sample by means of a worm-drive [3], and these are
robust and tolerant to a wide range of solvents, but cannot be
easily adapted to operation at different temperatures. Peristaltic
pumps produce a pulsating flow and are therefore not suitable
for ESI-MS, in which a steady flow is desirable for spray consis-
tency (though these are used in, for example, inductively coupled
plasmas mass spectrometry, for sample introduction at a rate of
−1
∼
1 ml min ) [4]. Microreactors can be coupled to ESI-MS instru-
ments, but the temporal element can only be controlled by flow
rate or tubing length, so collection of each data point involves a
new experiment [5]. Solutions can also be pressurized with an
inert gas and delivered into the source of the mass spectrome-
ter. Agilent’s calibrant delivery system pushes calibrant solution
into the solvent selection valve, whereupon it is pumped into the
Abbreviations: PSI, pressurized sample infusion; Pbf, 2,2,4,6,7-
pentamethyldihydrobenzofuran-5-sulfonyl; Pmf,
2
,2,5,7,8-pentamethylchroman-6-sulfonyl; TFA, trifluoroacetic acid.
∗ Corresponding author. Tel.: +1 250 721 7181; fax: +1 250 721 7147.
E-mail address: mcindoe@uvic.ca (J.S. McIndoe).
1
387-3806/$ – see front matter © 2012 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.ijms.2012.03.007

K.L. Vikse et al. / International Journal of Mass Spectrometry 323–324 (2012) 8–13
9
Scholarly implementation. The practical part of this work
was done within an undergraduate lab course, CHEM 361, at
the University of Victoria under the supervision of NT in the
spring of 2010 and the spring of 2011. CHEM 361 is a required
laboratory course for chemistry majors students and focuses
on practical analytical chemistry lab skills and learning instru-
mental techniques of analysis. This course has been involved in
several collaborative projects within the chemistry department
as well as with other departments at the University of Victo-
ria. For this project, students individually performed flow rate
measurements for a particular solvent and tubing length with
varying pressure of nitrogen in 1 psi steps from 1 to 5 psi. From
this data collection they were to calculate an optimal pressure
to achieve a certain flow rate. The students used the ESI-MS
to do some measurements using the same system (solvent,
tubing length, and pressure). They were introduced to the ESI-
MS instrument with hands-on instruction. Overall there were
Fig. 1. Pressurized sample infusion (PSI), set-up with on-line sample dilution via a
syringe pump. This step is not required if the sample in the flask is sufficiently dilute.
71 undergraduates involved in collecting the data. KLV devel-
oped the initial technique as part of her research, and was also a
teaching assistant for CHEM 361. She was therefore involved in
all aspects of development, implementation, as well as check-
ing all of the student data submitted.
pressurized and regulated gas (e.g. N ). The flask is positioned as
2
close as is practical to the ESI-MS source and connected to a source
of pressurized, carefully regulated inert gas via a short length of
rubber tubing. One end of a piece of PEEK capillary tubing is then
inserted through a punctured rubber septum (which is wired to the
flask to ensure it does not pop out) into the flask, while the other
end is connected to the electrospray ionization inlet (Fig. 1). Stan-
dard glassware with ground glass joints are shown; we generally
use Schlenk flasks, including those with integrated condensers, to
minimize the chances of leakage. A slight overpressure (0.5–5 psi) is
applied to the contents of the flask to facilitate continuous introduc-
tion of the sample into the mass spectrometer. It is important that a
regulator capable of accurately measuring these small pressures is
used as over-pressurizing the flask is dangerous. Any ground glass
joints should be secured to prevent fittings from popping out under
pressure. For monitoring reactions the flask may be equipped with
a stir bar and placed in a temperature controlled bath to allow
for stirring and temperature control. The sample solution must
be homogeneous to ensure the PEEK tubing stays free of block-
age. If dilution of the sample solution is required, this may be
accomplished on-line by incorporating a tee in the PEEK tubing
immediately outside the flask as shown in Fig. 1. Up to a 100:1
dilution can be obtained in this fashion with careful adjustment of
pressure and flow rate. Dilution with large volumes of room tem-
perature or cold solvent also acts to effectively quench any reaction
that may be occurring, which may be advantageous in some sce-
narios. PEEK tubing with a larger inner diameter should be used
after the T-piece to ensure backflow into the flask does not occur.
The PSI method has no mechanism for measuring flow rate,
so we were interested in using the pressure – the most easily
altered experimental parameter – to control flow rate in a pre-
dictable way, without having to actually measure it for a given
combination of solvent, pressure and tube diameter and length.
Drawing on well-established relationships from fluid mechanics,
the Hagen–Poiseuille equation [11] predicts the change in pressure
to be related to the volumetric flow rate as follows:
2
. Materials and methods
Deprotection: 2.1 mg Fmoc–Arg(Pbf)–OH (Merck Schuardt
OHG) was dissolved in 1 ml acetonitrile (99.5%, CALEDON) in a
0 ml flask attached to a well-regulated supply of nitrogen (at 5 psi)
and equipped with a septum and PEEK tubing leading to the mass
spectrometer. 4 ml of trifluoroacetic acid (99.9%, CALEDON) was
injected in the reaction flask.
Silane redistribution: 6 mg of (3-(methylsilyl)propyl)
triphenylphosphonium hexafluorophosphate(V) (0.0121 mmol, 1
equivalent), 93 mg of Karstedt’s catalyst (Pt 3.91% 0.000954 mmol,
5
0
.8 equivalent), 6 mg of triphenylphosphine (0.0229 mmol, 1.9
equivalent) were dissolved in 20 mL of freshly distilled fluoroben-
zene in a 50 mL Schlenk flask in an inert atmosphere glovebox. The
flask was equipped with a septum, removed from the glovebox and
installed with PEEK tubing leading to the mass spectrometer. The
flask was pressurized with N2 and monitored by ESI-MS. Once a
steady signal was obtained, 0.15 g of freshly distilled phenylsilane
1.39 mmol, 115 equivalents) was injected into the reaction flask.
The reaction was initiated at room temperature and then was
heated to reflux after 20 min to drive it to completion.
(
Nitrogen gas (5 psi) was used as the driving force for sam-
ple infusion in both experiments. Mass spectra were collected
on a Micromass Q-ToF micro mass spectrometer in positive-ion
mode using pneumatically-assisted electrospray ionization. Capil-
lary voltage: 2900 V; cone voltage: 10 V; extraction voltage: 0.5 V;
◦
◦
source temperature: 80 C; desolvation temperature: 150 C; cone
gas flow: 100 L/h; desolvation gas flow: 200 L/h; scan time was 3 s
and the inter scan time was 0.1 s. The reaction was continued until
the amount of starting material dropped to <5% of the original value.
P = 1
28ꢁLQ
(1)
3. Results
ꢀ
ꢁ
d4
To demonstrate the utility of PSI, we monitored the acid-
catalyzed hydrolysis of a protected amino acid, Fmoc–Arg(Pbf)–OH
Scheme 1), where trifluoroacetic acid cleaves off the Pbf
group selectively and under mild conditions. The 2,2,4,6,7-
pentamethyldihydrobenzofuran-5-sulfonyl (Pbf) protecting group
was developed as a TFA-sensitive arginine side-chain protecting
group for use in Fmoc-based peptide synthesis [12], because when
compared to the 2,2,5,7,8-pentamethylchroman-6-sulfonyl (Pmf)
protecting group, Pbf removal is faster when monitored by HPLC
where ꢀP is the pressure drop, ꢂ is the dynamic viscosity, L is length
of tube, Q is the volumetric flow rate and d is the inner diameter of
the tube. Rearranging to predict a flow rate from a given overpres-
sure, for water (ꢂ = 0.001 Pa.s), 1 psi (6900 Pa) of pressure applied
to a tube of length 0.5 m and diameter 127 ␮m should generate
(
−4 4
a flow rate of (6900 × 3.14 × {1.27 × 10 } )/(128 × 0.001 × 0.5) =
−
11
3
−1
−2
−1
−1
8
.8 × 10
m s = 8.8 × 10 ␮L s = 5.3 ␮L min . This flow rate
is entirely compatible with ESI-MS experiments.

1
0
K.L. Vikse et al. / International Journal of Mass Spectrometry 323–324 (2012) 8–13
O
O
Fmoc-Arg-OH
M + H] = m/z 397
+
[
[
H]+
O
O
O S
O
O S
O
O
O
O
O
NH
NH
NH
NH
HN
HN
HN
N
H
NH2+
N
H
NH2
N
H
O
OH
O
OH
O
O
OH
Fmoc-Arg(Pbf)-OH
+
+
[M + H] = m/z 649
+
O
O
O
F C
OH
O
H
3
H+
CF3
H
O
+
O S
O
O S
O S
O
O
O
O
+
CF3
[C13H17O S]
3
O
Pbf
H O
2
+
[
M] = m/z 253
+
[
M + H] = m/z 191
Scheme 1.
[
13]. The mechanism has been adapted from that proposed by Ram-
bond), whereas the production of the ion at m/z 191 (cleavage of
age and co-workers for the cleavage of Pmc [14]. Similar reactions
have been studied using FAB-MS via periodic sampling of the reac-
tion mixture [15].
the S C bond) through several additional steps is somewhat slower.
+
Trace amounts of a key intermediate, the cationic [C13H17O S] at
3
m/z 253 (see Scheme 1), were detected, and while the abundance
was low (a maximum of ∼0.3%), it was observed to be roughly
proportional to the overall rate of reaction. Overall, the data is con-
Here we observe Pbf cleavage from Fmoc-protected arginine
under typical conditions, which proceeds as a (pseudo) first order
process. Mass spectra from t = 0 and t = 10 min appear in Fig. 2. The
reaction took ∼20 min to go to completion.
PSI allows the constant monitoring of all charged species at a
time resolution equal to the scan rate, and the abundances of the
various species of interest can be plotted against time. Reaction
times are those the reactants spend in the flask plus the time they
spend in the tube. The additional reaction time can be calculated
m/z 649
100
8
0
(
vide supra) but fast reactions may be somewhat difficult to follow.
60
Here, making the tube as short as possible is beneficial. In the case
of the hydrolysis reaction, we see the protonated starting material,
(a) time = 0 minutes
4
2
0
0
0
[
Fmoc–Arg(Pbf)–OH + H]+ at m/z 649, disappear to be replaced by
+
protonated forms of both fragments, [Fmoc–Arg–OH + H] at m/z
3
+
97 and [Pbf + H] at m/z 191. We present the traces in 2 forms: (a)
raw intensity data and (b) normalized to the total intensity of all
ions of interest (i.e. those at m/z 649, 397 and 191).
Note that the traces are somewhat ragged (individual spectra
have different intensities due to fluctuations in spray quality). Addi-
100
200
300
400
500
600
700
800 m/z
m/z 397
1
00
tion of trifluoroacetic acid caused a huge boost in ion current at
+
∼
30 s, due to its ability to protonate neutrals to form [M + H] ions
8
0
(
as well as catalyze the hydrolysis)—the ion count per data point
jumped from 3000 to a momentary value of 48,000 (Fig. 3 shows
only 0–15,000 counts, to keep the rest of the data on scale). To
smooth out these fluctuations and to focus in on the relative abun-
dances of the ions of interest, we normalized the data to the total
intensity of the three key ions (Fig. 4); the product traces disre-
gard the other product, to take account of the different ionization
efficiencies of these two ions.
60
(
b) time = 10 minutes
m/z 191
4
0
0
0
2
m/z 649
The disappearance of m/z 649 is approximately first order, with
2
a slope of −0.21 and R = 0.96 for a plot of ln[intensity] vs. time, with
1
00
200
300
400
500
600
700
800 m/z
t
½
= 3.3 min. This number is consistent with the assumption that the
hydrolysis reaction is essentially complete within 20 min. It is clear
from the traces that the two hydrolysis products are generated at
Fig. 2. Individual positive ion electrospray ionization mass spectra taken at time (a)
and (b) 10 min into the reaction. The disappearance of the peak at m/z 649 and the
growth of the products at m/z 191 and 397 are clearly evident.
0
different rates; m/z 397 is produced rapidly (cleavage of the S
N

K.L. Vikse et al. / International Journal of Mass Spectrometry 323–324 (2012) 8–13
11
1
1
1
4000
2000
0000
m/z 397
8
6
4
2
000
000
000
000
0
m/z 191
m/z 649
Fig. 5. Normalized intensity vs. time trace for charged silane (m/z 349) and appear-
ance of the product of redistribution (m/z 425). The phosphonium silane has
hexafluorophosphate as counterion; the catalyst was added at t = 2 min and the
solution heated to reflux at ∼20 min.
0
5
10
15
time (minutes)
Fig. 3. Raw intensity vs. time traces for the protected amino acid (m/z 649) and its
hydrolyzed products (m/z 397 and 191). ESI-MS spectra were accumulated for 3 s to
generate each set of data, so each trace consists of ∼380 points.
as this diameter strikes a good balance between low volume and
appropriate flow rates. It is also an LCMS standard, and is reason-
ably resistant to blocking. Its internal volume is approximately
sistent with Ramage’s postulated mechanism for TFA deprotection
1
3 ␮L per meter (the tubing volume can be easily estimated by
[
14]. This experiment demonstrates the capability of PSI-ESI-MS
2
using the equation for the volume of a cylinder, ꢁr h, so for
PEEK tubing of inner diameter 127 ␮m and length 1 m, the vol-
to extract the type of data necessary for a mechanistic analysis,
especially the high dynamic range that allows it to simultaneously
monitor species present in high (starting material, products) and
very low concentrations (intermediates).
Another example of the data provided by our approach is shown
in Fig. 4, which tracks the platinum-catalyzed redistribution of two
silanes [16], one charge-tagged with a phosphonium group. This
demonstrates the ability of the technique to monitor air-sensitive
reactions in non-routine solvents; in this case, fluorobenzene. The
temperature of the reaction was increased from room tempera-
ture to reflux at 20 min; the catalyst used was Karstedt’s catalyst,
−6
2
−8
3
ume is 3.14 × (0.5 × 127 × 10 ) × 1 = 5.06 × 10 m = 12.65 ␮L).
To determine the pressure required to obtain an appropriate flow
rate, a series of measurements at different pressures were taken
for a variety of solvents and using a range of PEEK tubing lengths.
PEEK tubing for LCMS applications has an outer diameter of 1/16
of an inch (1.5875 mm) and an inner diameter (typically) of 0.005
inches (0.0127 mm). The solvents investigated were those com-
monly used for ESI-MS experiments: water, acetonitrile, a 50:50
water:acetonitrile mixture, methanol and dichloromethane. 0.45 m
is the minimum required PEEK tubing length for our PSI apparatus,
and PEEK tube lengths of 0.45, 0.50, 0.55 and 0.60 m were examined.
Pressures between 0.5 psi and 3 psi were examined at increments
of 0.5 psi. The mass of solvent forced through a length of PEEK tub-
ing over time was recorded and each flow rate was measured a
minimum of 10 times.
Pt[(H C = CHMe Si) O] (Fig. 5).
2
2
2
2
We were also interested in how the parameter space (tube
dimensions, overpressure, solvent) would affect the flow rate.
−
1
For ESI-MS, flow rates of 1–50 ␮L min cover the normal range,
with 5 ␮L min
−1
ꢀꢀ
being typical. We used 0.005 i.d. PEEK tubing,
The intercepts in plots of flow rate vs. pressure were at slightly
positive values of overpressure, and as expected, the least viscous
solvents showed the highest flow rates at a given pressure (Fig. 6).
Vertical error bars are included, but they are almost too small to
m/z 397
m/z 191
1
0
0
0
0
.8
.6
.4
.2
0
0
0
.003
.002
m/z 253
0.001
0
0
5
10
15
20
m/z 649
0
5
10
15
time (minutes)
Fig. 4. Normalized intensity vs. time trace for the protected amino acid (m/z 649)
and its hydrolyzed products (m/z 397 and 191). The inset shows the abundance of
+
the [C13H17O3S] intermediate on a different intensity scale; note that its abundance
Fig. 6. Data from flow rate vs. overpressure experiments with different solvents and
approximates the instantaneous rate of reaction.
using PEEK tubing of length 0.6 m.

1
2
K.L. Vikse et al. / International Journal of Mass Spectrometry 323–324 (2012) 8–13
4. Conclusions
The Hagen–Poiseuille equation can be used to reliably estimate
flow rates using pressurized sample infusion, but calibration of
flow rate for a given piece of tubing is recommended in order to
estimate the internal diameter accurately. Once this simple, one-
off experiment is performed, the analyst has a reliable pumping
system for ESI-MS with a low internal volume and robustness
towards a wide variety of solvents. The high dynamic range of mass
spectrometer in conjunction with continuous monitoring allows
the simultaneous measurement of all charged species in solution,
including low abundance intermediates, as exemplified by the acid-
catalyzed deprotection of an amino acid. Air-sensitive systems
in unconventional solvents at temperatures up to reflux are also
readily accessible to this method, as shown by the silane redistri-
bution experiment. We anticipate our approach to be useful for
all chemists interested in reproducible, continuous monitoring of
reactions in which the relative abundance of ions is dynamic and
varies across several orders of magnitude.
Fig. 7. Experimental flow rates plotted vs. theoretical flow rates calculated using
the Hagen–Poiseuille equation. The 100 data points include flow rates calculated
using five different solvents, four different tube lengths and five different values
of overpressure (see Supporting information for data used to generate this plot).
The dashed lines indicate the variation expected if the inner diameter was 25% less
(
1:0.32) or 25% more (1:2.44) than the claimed value.
Acknowledgements
JSM thanks NSERC for operational funding (Discovery and Dis-
covery Accelerator Supplement), and CFI, BCKDF and the University
of Victoria for instrumental support. Thanks to Fraser Hof and
Ori Granot for useful discussions. We extend our thanks to the
two Chemistry 361 Analytical Chemistry classes that performed
some of the experimental work for this paper. From 2010: Bryan
Boots, Chelsea Burns, Alana Chester, Haeun Chung, Rebecca Courte-
manche, Sean Davidoff, Rebecca Dixon, Rowan Fox, Jenna Frerot,
Jesse Gallop, Michael Hamilton, Lisa Hart, Ian Holley, Marc-Andre
Hoyle, Christy Hui, Megan Kilduff, Philip Klein, Aiko Kurimoto,
Jessamyn Logan, Cara Manning, Rebeccah McLean, Kirsten Medd,
Krista Morrow, Joshua Nero, Dean Neville, Emma Nicholls-Allison,
Kevin Nikelski, Jaimie Parton, Gillian Reay, Travis Schwantje, Shawn
Slavin, Dirk Slot, Sinan Soykut, Liana Stammers, Christine Tough,
Derek Waghray, Sophie Waterman and Kevin Weinreich. From
2011: Qinqi Chen, James Chircoski, Kaitlin Desilets, Morgan Ehman,
Emily Eng, Nicholas Erb, Patrick Fergusson, Heather Fitzpatrick,
Graham Garnett, Damon Gilmour, Ryan Hanson, Bochao Huang,
Chih-Hao Huang, Eric Janusson, Talon Jones, Manuel Ma, Angus
MacKay, Ashley March, James McFarlane, Jamie McGuire, Simon
McPhedran, Tanya Murray, Brett Nesmoe, Elisabeth Pharo, Sarah
Polkinghorne, Lauren Rainsford, Ryan Roberts, Andrew Rosenberg,
Hollis Roth, Colin Stacey, John Warren, Elaine Wu.
be seen indicating high precision in the flow rate measurements.
Some slight deviations from linearity are observed and this can
be attributed to the precision with which the pressure was set
for each experiment—the errors in the pressure reading were esti-
mated to be of the order of ± 0.1 psi (gauge markings are at 0.2 psi
increments). The relative (to water) viscosities of the solvents at
◦
2
0 C are 0.36 for acetonitrile, 0.44 for dichloromethane, 0.59 for
methanol, 0.68 for 50:50 acetonitrile:water (approximated from
◦
data collected at 25 C) [17], and 1 for water [18].
The large amount of data collected is best summarized by
plotting the actual flow rate at a given combination of solvent,
overpressure and tubing length and plot it against the calculated
value using the Hagen–Poiseuille equation (Fig. 7). In all cases, the
flow rates vs. overpressure graphs were almost perfectly linear
2
(
R > 0.99).
What is immediately clear from the plot is that different pieces
of tubing produce markedly different results. Because the over-
pressure, tubing length and solvent viscosity can be measured
accurately, the likely source of error is in the inner diameter of
the tube, for which any discrepancy in the listed value is magni-
fied to the fourth power. A ± 25% error in diameter is enough to
4
4
account for most of the variation (1.25 = 2.44, 0.75 = 0.32). The
ꢀ
ꢀ
ꢀꢀ
manufacturer’s listed tolerance for 1/16 o.d. 0.005 i.d. PEEK tub-
ꢀ
ꢀ
ing is ± 0.001 , i.e. ± 20% error, and this tolerance matches well with
our observed experimental variability of ± 25%. Nonetheless, it is
very easy to calibrate a given length of tubing by following the pro-
cedure we describe here (i.e. using a stopwatch and a balance) and
apply the appropriate correction to the Hagen–Poiseuille equation
for all samples run using the PSI method. Once this correction is
obtained, the equation simplifies to Q = kꢀP/ꢂ, where k combines
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.ijms.2012.03.007.
References
4
the experimental correction with ꢁd /128L, which is a constant for
a given piece of tubing. We regard this approach to be more reliable
than direct measurement of the diameter, because not only is it dif-
ficult to measure inner diameters of PEEK tubing accurately on the
sub-millimeter scale, there is no guarantee that that diameter at
the end(s) is representative of the entire length of tubing. Knowing
the flow rate at a given pressure allows for quick optimization of
the experimental set-up, helping ensure that good spray conditions
are obtained and that consumption of the solution is appropriate
for the length of the experiment. Reproducibility of conditions is a
key consideration when attempting to extract kinetic information
from an online monitoring experiment.
[
1] R.B. Cole, Electrospray and MALDI Mass Spectrometry: Fundamentals, Instru-
mentation, Practicalities and Biological Applications, John Wiley
Hoboken, 2010.
& Sons,
[
2] M.C. McMaster, LC/MS: A Practical User’s Guide, John Wiley & Sons, Hoboken,
005.
2
[3] B. Boyd, R.K. Boyd, C. Basic, R. Bethem, Trace Quantitative Analysis by Mass
Spectrometry, John Wiley & Sons, Chichester, 2008.
[
4] E. de Hoffmann, V. Stroobant, Mass Spectrometry: Principles and Applications,
John Wiley & Sons, Chichester, 2007.
[5] (a) L.S. Santos, Online mechanistic investigations of catalyzed reactions by elec-
trospray ionization mass spectrometry: a tool to intercept transient species in
solution, European Journal of Organic Chemistry 23 (2008) 5–25, 3;
(
b) L.S. Santos, Reactive Intermediates: MS Investigations in Solution, Wiley
VCH, Weinheim, 2010.

K.L. Vikse et al. / International Journal of Mass Spectrometry 323–324 (2012) 8–13
13
[
[
6] As described in the Agilent 6500 Series Q-TOF LC/MS Maintenance Guide,
http://www.chem.agilent.com/Library/usermanuals/Public/G2581-90005 Q-
TOF Maintenance v2.pdf (last accessed 20.01.11).
7] (a) M. Wetterhall, M. Palmblad, P. Håkansson, K.E. Markides, J. Bergquist,
Rapid analysis of typtically digested cerebrospinal fluid using capil-
lary electrophoresis–electrospray ionization–Fourier transform ion cyclotron
reactant, products and intermediates, Angewandte Chemie International Edi-
tion 50 (2011) 8304–8306.
[11] S.P. Sutera, R. Skalak, The history of Poiseuille’s law, Annual Review of Fluid
Mechanics 25 (1993) 1–19.
[12] C.G. Fields, G.B. Fields, Minimization of tryptophan alkylation following 9-
fluorenylmethoxycarbonyl solid-phase peptide synthesis, Tetrahedron Letters
34 (1993) 6661–6664.
resonance-mass spectrometry, Journal of Proteome Research
61–366;
b) P. Kusy, K. Klepárník, Z. Aturki, S. Fanali, F. Foret, Optimization of a pressur-
1 (2002)
3
(
[13] L.A. Carpino, H. Shroff, S.A. Triolo, E.-S.M.E. Mansour, H. Wenschuh, F. Albericio,
The 2,2,4,6,7-pentamethyldihydrobenzofuran-5-sulfonyl group (Pbf) as argi-
nine side cahin protectant, Tetrahedron Letters 34 (1993) 7829–7832.
[14] R. Ramage, J. Green, A.J. Blake, An acid labile arginine derivative for peptide
ized liquid junction nanoelectrospray interface between CE and MS for reliable
proteomic analysis, Electrophoresis 28 (2007) 1964–1969.
G
[
8] K.L. Johnson, T.D. Veenstra, J.M. Londowski, A.J. Tomlinson, R. Kumar, S. Naylor,
On-line sample clean-up and chromatography coupled with electrospray ion-
ization mass spectrometry to characterize the primary sequence and disulfide
bond content of recombinant calcium binding proteins, Biomedical Chromatog-
raphy 13 (1999) 37–45.
synthesis: N -2,2,5,7,8-pentamethylchroman-6-sulphonyl-l-arginine, Tetra-
hedron 47 (1991) 6353–6370.
[15] K. Eckart, H. Schwarz, G. Becker, H. Kessler, Journal of Organic Chemistry 51
(1986) 483–486.
[16] R. Shankar, A. Joshi, S. Upreti, Polyhedron 25 (2006) 2183–2189.
[17] G.P. Cunningham, G.A. Vidulich, R.L. Kay, Several properties of
acetonitrile–water, acetonitrile–methanol, and ethylene carbonate–water
[
9] K.L. Vikse, M.P. Woods, J.S. McIndoe, Pressurized sample infusion for the con-
tinuous analysis of air- and moisture-sensitive reactions using electrospray
ionization mass spectrometry, Organometallics 29 (2010) 6615–6618.
systems, Journal of Chemical and Engineering Data
336–337.
1
(2) (1967)
[
10] K.L. Vikse, Z. Ahmadi, C.C. Manning, D.M. Harrington, J.S. McIndoe, Power-
ful insight into catalytic mechanisms through simultaneous monitoring of
[18] CRC Handbook of Chemistry and Physics, 73rd ed., 1992–1993.