Hydrogen/deuterium exchange on aromatic rings during atmospheric pressure chemical ionization mass spectrometry

Article abstract of DOI:10.1002/rcm.4488

It has been demonstrated that substituted indoles fully labelled with deuterium on the aromatic ring can undergo substantial exchange back to partial and even fully protonated forms during atmos-pheric pressure chemical ionisation (APCI) liquid chromatography/mass spectrometry (LC/MS). The degree of this exchange was strongly dependent on the absolute quantity of analyte, the APCI desolvation temperature, the nature of the mobile phase, the mobile phase flow rate and the instrument used. Hydrogen/deuterium (H/D) exchange on several other aromatic ring systems during APCI LC/MS was either undetectable (nitrobenzene, aniline) or extremely small (acetanilide) compared to the effect observed for substituted indoles. This observation has major implications for quantitative assays using deuterium-labelled internal standards and for the detection of deuter-ium-labelled products from isotopically labelled feeding experiments where there is a risk of back exchange to the protonated form during the analysis.

Full text of DOI:10.1002/rcm.4488

RAPID COMMUNICATIONS IN MASS SPECTROMETRY
Rapid Commun. Mass Spectrom. 2010; 24: 1105–1110
Published online in Wiley InterScience (www.interscience.wiley.com) DOI: 10.1002/rcm.4488
Hydrogen/deuterium exchange on aromatic rings during
atmospheric pressure chemical ionization mass
spectrometry
1
2
2
3
Noel W. Davies *, Jason A. Smith , Peter P. Molesworth and John J. Ross
1
Central Science Laboratory, Bag 74, University of Tasmania, Hobart, Tasmania, Australia 7001
School of Chemistry, Bag 75, University of Tasmania, Hobart, Tasmania, Australia 7001
School of Plant Science, Bag 55, University of Tasmania, Hobart, Tasmania, Australia 7001
2
3
Received 15 December 2009; Revised 27 January 2010; Accepted 28 January 2010
It has been demonstrated that substituted indoles fully labelled with deuterium on the aromatic ring
can undergo substantial exchange back to partial and even fully protonated forms during atmos-
pheric pressure chemical ionisation (APCI) liquid chromatography/mass spectrometry (LC/MS). The
degree of this exchange was strongly dependent on the absolute quantity of analyte, the APCI
desolvation temperature, the nature of the mobile phase, the mobile phase flow rate and the
instrument used. Hydrogen/deuterium (H/D) exchange on several other aromatic ring systems during
APCI LC/MS was either undetectable (nitrobenzene, aniline) or extremely small (acetanilide)
compared to the effect observed for substituted indoles. This observation has major implications
for quantitative assays using deuterium-labelled internal standards and for the detection of deuter-
ium-labelled products from isotopically labelled feeding experiments where there is a risk of back
exchange to the protonated form during the analysis. Copyright # 2010 John Wiley & Sons, Ltd.
The use of stable isotope dilution mass spectrometry forms
the basis of many quantitative analytical methods. As a
subset of these, compounds with deuterium replacing
hydrogen on an aromatic ring have been used together with
liquid chromatography/mass spectrometry (LC/MS) inter-
faces that use heating during nebulisation and ionisation,
such as Atmospheric Pressure Chemical Ionisation (APCI)
standards. These standards are typically prepared from the
corresponding unlabelled compounds by treatment with a
solution of deuterium chloride for days at room temperature
or hours at 1008C in a sealed tube using a microwave reactor.
Several deuterated indole-ring plant hormones and their
precursors prepared this way have been successfully utilised
by our group in quantitative gas chromatography (GC)/MS
1
9
and Turbo Ion Spray. Examples include amphetamines,
determinations. In these compounds deuterium replaces all
2
–4
5
6
nicotine and its metabolites,
indoles, azo dyes and
hydrogens attached to the indole ring.
7
benzodiazepines, among others. As well as their common
use as internal standards for quantitative analyses, deuter-
ated compounds are also the target analytes in many
We recently attempted APCI LC/MS determinations at the
low to sub-nanogram range for tryptamine (1a – molecular
weight (MW) 160 Da) and Nv-acetyltryptamine (2a – MW
8
–10
biosynthetic studies (e.g.
and disposition studies using labelled pharmaceuticals.
), and in drug metabolism
202 Da) utilising the nominally 2,4,5,6,7-D
of each analyte (1b & 2b), respectively, as internal standards.
However, significant amounts of the D , D , D , D and
even D forms of these substituted indoles were unexpect-
5
-labelled versions
1
1
The synthesis of deuterium-labelled compounds for use in
both mass spectrometry and fundamental studies by
deuterium/hydrogen (D/H) exchange from the unlabelled
4
3
2
1
0
edly detected. Investigations by other techniques (GC/MS
and electrospray (ES)-MS) confirmed that the compounds
themselves had not lost any deuterium atoms, and revealed
that this effect was occurring during the APCI desolvation/
ionisation process.
1
2
versions has been comprehensively reviewed. Common
methods include acid- and base-catalysed exchange, as well
as the use of homogeneous and heterogeneous catalysts.
As part of a compound-based approach to determining
hormone pathways in plants, and the effects of specific genes
on these pathways, we have used acid-catalysed D/H
exchange to synthesise several stable-isotope-labelled plant
hormones or their putative precursors for use as internal
Many MS techniques, including APCI, have been used as
tools to investigate conventional ‘exchangeable’ protons, such
as those on hydroxyl and amino groups, by infusing the ion
13,14
source with D O or CD OD
and following the D/H
2
3
exchange occurring in the gas phase. However, we are
unaware of any reports on the acid-catalysed gas-phase
exchange of aromatic protons/deuterons during LC/MS
analysis. This has potentially significant implications for
*
Correspondence to: N. W. Davies, Central Science Laboratory,
Bag 74, University of Tasmania, Hobart, Tasmania, Australia
001.
E-mail: noel.davies@utas.edu.au
7
Copyright # 2010 John Wiley & Sons, Ltd.

1
106 N. W. Davies et al.
analyses where deuterated aromatic ring compounds are
either used as internal standards or are themselves the target
analytes.
valve for sample introduction. Organic solvents were HPLC
grade; water was Milli-Q. Solvents were pumped by a Waters
Alliance 2690 HPLC. Default conditions were: spray current
6 mA, capillary voltage 15 V, capillary temperature 1508C,
desolvation temperature 3508C, mobile phase 50:50 metha-
nol/1% acetic acid in water, flow rate 0.5 mL/min and sheath
gas pressure 40 psi. These parameters were varied for
individual experiments as described below. Tryptamine,
Parameters such as sample concentration, nature of the
mobile phase, flow rate, and APCI desolvation temperature
were examined for their influence on the magnitude of this
effect. The effect of chromatographic peak tailing on H/D
ratios was also examined.
þ
aniline and acetanilide were analysed as their [MþH] ions,
ꢁ
and nitrobenzene as the [M] ion. For all data any traces of
EXPERIMENTAL
background ions at the relevant m/z values were removed by
baseline subtraction prior to measurement of relative
abundances. Samples were dissolved in methanol and
injected via the loop into the mobile phase stream. Narrow
scan ranges were employed, including ‘Zoomscan’ higher
resolution mode where possible.
Preparation of labelled standards
The labelled indole compounds were prepared as described
9
previously.
2
,3,4,5,6-D -Aniline (3a)
5
5
D -Nitrobenzene (100 mL, 0.978 mmol; Cambridge Isotope
Laboratories) and zinc (710 mg, 10.8 mmol) were refluxed in
glacial acetic acid (1 mL) for 1 h. The residue was diluted with
water (5 mL) and poured into 2M sodium carbonate solution
Electrospray MS analysis
Electrospray analyses were carried out on the LCQ at a
capillary temperature of 2008C, capillary voltage of 15 V,
capillary temperature 2008C, needle voltage of 4 kV and
sheath gas pressure of 85 psi.
(
20 mL). The zinc salts were dissolved with conc. ammonia,
and extracted with tert-butyl methyl ether (3 ꢀ 15 mL). The
combined organic extracts were washed with water (20 mL),
brine (20 mL), dried on sodium sulphate, filtered and the
solvent removed under reduced pressure, yielding the title
compound (64.2 mg, 0.655 mmol) as a pale oil in 67% yield.
Orbitrap APCI-MS
Further APCI analyses were carried out on a Thermo
Scientific Orbitrap FTMS mass spectrometer by loop injection
(
2 mL loop) into a stream of 0.5 mL/min of 50:50 methanol/
2
,3,4,5,6-D -Acetanilide (3b)
1% acetic acid. The spray current was 3.6 mA, capillary
voltage 3 V, capillary temperature 2008C, desolvation
temperature 3508C, and sheath gas pressure 30 psi. Resol-
ution was set to 30 000. Narrow scan ranges around the ions
of interest were used.
5
Triethylamine (1.10 mL, 7.92 mmol), acetic anhydride
0.620 mL, 6.58 mmol) and dimethylaminopyridine (22.3 mg,
.183 mmol) were added sequentially to a solution of 2,3,4,5,6-
-aniline (64.0 mg, 0.654 mmol) in anhydrous dichloro-
(
0
D
5
methane (5 mL) under an atmosphere of nitrogen. After
stirring for 18 h the reaction was diluted with dichloromethane
(20 mL), and the organic phase washed with saturated
potassium hydrogen sulphate solution (2 ꢀ 20 mL), 2 M
sodium carbonate solution (2 ꢀ 20 mL) and dried with
anhydrous magnesium sulphate. The mixture was filtered
through a plug of silica gel using 60% ethyl acetate/hexanes for
elution. The solvent was removed under reduced pressure
yielding the title compound (58.7 mg, 0.419 mmol) in 64% yield.
RESULTS
Derivatisation and GC/MS analysis
2
,4,5,6,7-D
5
-Tryptamine (1b) was converted into its bis-
5
The level of deuterium incorporation into the D -tryptamine
pentafluoropropionyl derivative and analysed by GC/MS as
(1b) standard, as determined by GC/MS as its bis-penta-
fluoropropionyl derivative (after appropriate corrections for
naturally occurring isotopes and the slight inter-channel
interference between adjacent masses on the quadrupole
9
described previously using a Varian 1200 triple quadrupole
mass spectrometer in tandem mass spectrometry (MS/MS)
mode. Corrections were made for natural isotopes and the
0
.4% interference between adjacent channels on the quadru-
system used), was 94 atom%, with 74% being in the D
and no detectable D or D (see Table 1).
As noted, initial attempts at APCI LC/MS using this D
5
form
pole SRM analyses (as determined by analysis of unlabelled
tryptamine) before calculating the degree of deuterium
1
0
5
-
incorporation. D
5
-Acetanilide was analysed directly by GC/
tryptamine (1b) sample yielded mixed levels of deuteration.
þ
MS on a Kratos Concept ISQ in selected ion monitoring mode
with channels for D through to D , at a resolution of 1000,
The [MþH] ion for D -tryptamine (1b) is at m/z 166, and
5
other degrees of deuteration were clearly observed at m/z 165
(D ), 164 (D ), 163 (D ), 162 (D ) and finally the D form at m/z
0
7
and dwell time of 50 ms per ion. Appropriate corrections
were made for naturally occurring isotopic abundances.
4
3
2
1
0
161. Blank runs after tryptamine injections did result in the
observation of tryptamine ions due to the surface interactions
with the loop and transfer lines, and so several wash cycles of
the loop with 1% acetic acid were employed to remove this
‘memory’ effect. After this procedure no significant ions
APCI-MS analysis
Initial APCI-MS analyses were undertaken on a Finnigan
(
Thermo) LCQ Classic fitted with a 5 mL Cheminert loop
Copyright # 2010 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2010; 24: 1105–1110
DOI: 10.1002/rcm

H/D exchange on aromatic rings during APCI-MS 1107
5
Table 1. Percentage of each level of deuteration and total aromatic atom% deuterium of nominally ‘D ’-tryptamine (1b) observed
under different analytical conditions
Experiment
%D
5
%D
4
%D
3
%D
2
%D
1
%D
0
Atom% D
GC/MS (single analysis)
LCQ Electrospray
APCI acetonitrile
74.3
20.2
4.8
0.6
nd
nd
nd
nd
nd
93.7
77.1 ꢂ 1.1
73.4 ꢂ 1.8
61.8 ꢂ 1.4
60.3 ꢂ 2.1
47.6 ꢂ 0.7
38.6 ꢂ 0.6
32.6 ꢂ 0.8
27.0 ꢂ 0.3
28.4 ꢂ 1.7
17.8 ꢂ 0.8
17.9 ꢂ 0.4
25.4 ꢂ 1.1
20.8 ꢂ 0.1
31.7 ꢂ 0.2
26.6 ꢂ 1.0
29.5 ꢂ 1.3
29.2 ꢂ 1.3
25.4 ꢂ 2.1
3.0 ꢂ 0.6
5.0 ꢂ 0.7
7.0 ꢂ 0.7
6.0 ꢂ 0.5
11.6 ꢂ 0.2
12.9 ꢂ 0.3
15.9 ꢂ 1.3
14.2 ꢂ 0.8
13.7 ꢂ 0.8
95.0 ꢂ 0.3
91.6 ꢂ 0.5
87.8 ꢂ 1.2
82.3 ꢂ 0.5
82.3 ꢂ 0.6
72.4 ꢂ 0.2
70.9 ꢂ 1.1
65.0 ꢂ 1.3
63.2 ꢂ 1.0
1.8 ꢂ 0.2
2.9 ꢂ 1.0
3.2 ꢂ 0.4
4.8 ꢂ 0.3
8.7 ꢂ 0.8
9.5 ꢂ 0.8
11.0 ꢂ 0.3
10.5 ꢂ 1.1
0.9 ꢂ 0.1
1.5 ꢂ 0.3
2.5 ꢂ 0.5
2.6 ꢂ 0.3
6.7 ꢂ 0.1
7.1 ꢂ 0.4
8.7 ꢂ 1.0
10.2 ꢂ 1.4
1.0 ꢂ 0.1
1.4 ꢂ 0.6
7.2 ꢂ 0.6
1.8 ꢂ 0.3
6.4 ꢂ 0.7
5.4 ꢂ 0.5
10.0 ꢂ 1.1
11.8 ꢂ 1.8
APCI methanol
APCI Orbitrap FTMS 2 ng
APCI aq/MeOH 50 ng
ꢃ
APCI Aq/MeOH
APCI Aq
APCI Aq/MeOH 0.2 mL/min
APCI Aq/MeOH 1 ng
ꢃ
This experiment was repeated for each variable tested; this row relates to the testing of different mobile phases.
Averages are shown with standard deviations; n ¼ 3. APCI analyses were all 5 ng injections on the LCQ, using desolvation temperature of 3508C
and flow rate of 0.5 mL/min unless otherwise stated. ‘Aq’ is 1% acetic acid in water. ‘Aq/MeOH’ is 50:50 1% acetic acid in water/methanol. ‘nd’ is
not detected. Appropriate corrections were made for natural isotope abundances. Within experimental error, the GC/MS and electrospray data
effectively both represent the actual deuterium labelling of the sample.
were observed at any of the target m/z values for blank
injections.
injections, a strong discrimination was observed across the
peak, with dramatically increased H/D exchange observed
on the tail of the peak as the sample concentration dropped.
Figure 4 shows this effect by examining the mass spectra
averaged across the peak at half height and across the tail of a
loop injection, with markedly lower atom% deuterium seen
Figure 1 shows the marked effect of APCI desolvation
temperature, with substantial increases in H/D exchange at
3
508C (the optimum temperature for this analysis) compared
to 2508C. Due to the significant effect on yield of protonated
molecules by working significantly below or above optimal
desolvation temperatures, only these two desolvation
temperatures were examined quantitatively.
Figure 2 shows the effect of various pure and mixed
mobile phases. Very little exchange was noted with
acetonitrile, small reductions in atom% deuterium were
observed with methanol, and the most exchange was seen
with 100% aqueous mobile phase. The presence of 1% acetic
acid in the mobile phase had little apparent effect on the
þ
level of H/D exchange, since H
by the APCI discharge voltage.
3
O
is efficiently generated
The degree of H/D exchange was strongly affected by the
amount of sample injected, as noted by the difference
between 1 ng, 5 ng and 50 ng loop injections (Fig. 3) with very
marked increases in exchange at the lower levels. This was
Figure 2. Effect of APCI mobile phase on H/D exchange for
5
nominally D -tryptamine (1b). Flow rate 0.5 mL/min, APCI
most apparent in terms of percentage of D
0
, which more than
desolvation temperature 3508C, 5 ng injections. Each point
is the average of three loop injections – error bars show
standard deviations.
tripled between the 50 ng and 5 ng levels. Moreover, as
tryptamine tends to ‘tail’ due to surface binding even on loop
Figure 1. Effect of APCI desolvation temperature on H/D
5
Figure 3. Effect of amount of nominally D -tryptamine (1b)
exchange for nominally D
5
-tryptamine (1b). Flow rate
.5 mL/min, mobile phase 50:50 methanol/1% acetic acid,
ng injections. Each point is the average of three loop injec-
injected. Flow rate 0.5 mL/min, APCI desolvation temperature
3508C, mobile phase 50:50 methanol/1% acetic acid. Each
point is the average of three loop injections – error bars show
standard deviations.
0
5
tions – error bars show standard deviations.
Copyright # 2010 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2010; 24: 1105–1110
DOI: 10.1002/rcm

1
108 N. W. Davies et al.
5
Figure 4. Mass spectra in the protonated molecule region for nominally D -tryptamine (1b) under APCI
conditions for a 5 ng injection with water present in the mobile phase: (A) average across the peak at half
height and (B) average towards the tail of the loop injection peak.
in the latter. The effect can also easily be observed by
examining mass chromatograms at each level of deuterium
incorporation. Figure 5 shows the effect on the measurement
of deuterium incorporation by inclusion and exclusion
respectively of the chromatographic peak tailing.
5
degree of exchange observed. A 2 ng injection of D -
tryptamine (1b) on the Orbitrap yielded the same atom%
deuterium as a 50 ng injection on the LCQ, but the former had
0
unexpectedly elevated levels of D (7.2%) relative to the other
partially deuterated forms. All Orbitrap data were acquired
at high mass resolution (30 000) and the quantitation was
done on the exact masses for the various tryptamine [MþH]þ
ions, minimising the possibility of interferences from back-
ground ions.
There was an inverse correlation between flow rate and
H/D exchange (Fig. 6), which can be readily attributed to the
lower flux of sample through the ion source at lower flow
rates reducing the sample concentration in the ion source and
hence increasing H/D exchange.
5
Table 1 summarises the change in deuterium levels for D -
To eliminate the possibility of the observed H/D exchange
being due to some specific catalytic activity in the 11-year-old
ion source used, the same measurements were undertaken
on a Thermo Orbitrap mass spectrometer in APCI mode. This
tryptamine (1b) under some of these different conditions and
on two different instruments. Other indole compounds,
5
including D -Nv-acetyltryptamine (2b), showed very similar
levels of H/D exchange. The GC/MS and ES-MS results
indicated some D , D and a trace of D forms were present in
also resulted in H/D exchange effects for D -tryptamine (1b),
5
4
3
2
although of a substantially lower magnitude using similar
conditions to those used on the LCQ, indicating that source
design and/or history of use are significant factors in the
the original sample – these data then represent the ‘no
exchange’ case.
5
D -Nitrobenzene (99.6% atom% D; Cambridge Isotopes
laboratories Inc.) had poor ionisation efficiency in negative
APCI, and at the relatively high concentrations for which
ꢁ
the M ion could be clearly observed there was no
detectable H/D exchange. Despite excellent sensitivity in
positive ion APCI, no significant H/D exchange could be
detected for D -aniline (3a). For D -acetanilide (3b) the
5
5
effect was initially difficult to detect, but when amounts
were lowered to single and sub-nanograms a small amount
of H/D exchange was noted. However, the atom%
deuterium only changed from ꢄ94% for 50 ng injections
to ꢄ89% for 1 ng injections. Table 2 shows details for D
5
-
acetanilide (3b) at different injection levels. Statistical
analysis revealed that all differences in atom% deuterium
between the 50 ng, 5 ng, and 1 ng levels for D -acetanilide
5
Figure 5. Effect of peak tailing on deuterated tryptamine
measurements. Solid line is average across the peak at half
height, dotted line is average across the whole peak. APCI
desolvation temperature 3508C, mobile phase 50:50 metha-
nol/1% acetic acid, 5 ng injections. Each point is the average
of three loop injections – error bars show standard deviations.
were statistically significant (P < 0.001 for 50 ng vs. 5 ng,
5 ng vs. 1 ng and 50 ng vs. 1 ng).
3
Racemic D -salbutamol (4), containing no aromatic ring
deuterium atoms, unsurprisingly did not show any evidence
of H/D exchange down to sub-nanogram amounts under the
same APCI conditions as used for D -tryptamine (1b).
5
Copyright # 2010 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2010; 24: 1105–1110
DOI: 10.1002/rcm

H/D exchange on aromatic rings during APCI-MS 1109
5
Table 2. Percentage of each level of deuteration and total aromatic atom% deuterium of nominally ‘D ’-acetanilide (3b) during
APCI
Experiment
%D
5
%D
4
%D
3
%D
2
%D
1
%D
0
Atom% D
GC/MS (single analysis)
APCI Aq/MeOH 50 ng
APCI Aq/MeOH 5ng
APCI Aq/MeOH 1ng
71.6
25.5
3.0
nd
nd
nd
nd
nd
nd
nd
nd
nd
93.8
73.4 ꢂ 0.6
70.3 ꢂ 0.7
62.4 ꢂ 1.5
23.0 ꢂ 0.4
24.9 ꢂ 0.7
25.3 ꢂ 0.3
3.3 ꢂ 0.3
4.0 ꢂ 0.2
8.3 ꢂ 0.9
0.3 ꢂ 0.05
0.8 ꢂ 0.1
4.1 ꢂ 1.0
93.9 ꢂ 0.1
92.9 ꢂ 0.2
89.2 ꢂ 0.8
Averages are shown with standard deviations; n ¼ 4. APCI analyses were conducted using a desolvation temperature of 3508C and a flow rate of
.5 mL/min. ‘Aq/MeOH’ is 50:50 1% acetic acid in water/methanol. ‘nd’ is not detected. Appropriate corrections were made for natural isotope
0
abundances. Differences in atom% D between the 50 ng, 5 ng, and 1 ng injections were all statistically significant (P < 0.001). Within experimental
error, the GC/MS and 50 ng APCI data effectively both represent the actual deuterium labelling of the sample.
DISCUSSION
measurements with sufficient precision in this picogram range
on a signal distributed among five different ions. Nevertheless
this is the range at which many quantitative analyses are
required – for plant hormones the initial spike of internal
standard to the whole sample is often in the sub-nanogram
5
A nominally ‘D ’-labelled tryptamine (1b) standard that had
been shown by GC/MS analysis to be ꢄ75% D and ꢄ94
5
atom% deuterium, with no detectable D , resulted in as little
0
as 28% D and 63 atom% deuterium, and as much as 12% D
5
0
5
range. Unsurprisingly, the use of acetonitrile alone as the
upon APCI-MS analysis under conditions that strongly
favoured exchange – i.e. a low level of analyte, presence of
some aqueous mobile phase and relatively high desolvation
mobile phase resulted in negligible H/D exchange.
Strong correlations with all the parameters tested were
observed. In addition to the specific parameters investigated,
aspects such as source design, discharge current, needle
position, source cleanliness and prior history, and inlet
capillary temperature may all influence the degree of
aromatic H/D exchange.
þ
þ
temperature. Given that H
3
O
and/or CH OH were
3
2
present in abundance in the gas phase whenever significant
exchange was detected, and the relatively high desolvation
temperatures involved, H/D exchange in principle could be
claimed to be not unexpected given that acid-catalysed
exchange is the method by which deuterated aromatics can
be prepared in the first place.
We suggest that indole compounds are more prone to
exchange due to the high electron density of the aromatic ring
1
5
due to electron donation from the indole nitrogen. This is
consistent with the reactivity trend for electrophilic aromatic
substitution in which aromatics with electron-donating
groups react faster than those with electron-withdrawing
groups attached. As a nitro group is strongly electron-
withdrawing, it is strongly deactivating for this reaction and
hence no exchange is observed. Aniline might be expected to
behave the same as indole; however, under the acidic APCI
conditions the amine will be protonated and the resultant
ammonium ion acts as an electron-withdrawing group. For
tryptamine the APCI protonation is on the terminal amine,
and hence remote from the indole ring, minimising any
electron-withdrawing effect. With acetanilide, protonation of
the carbonyl oxygen rather than the nitrogen results in an
electron-withdrawing resonant hydroxylated iminium ion,
minimising H/D exchange. The result for these compounds is
consistent with the reactivity of the different aromatic
systems towards electrophilic aromatic substitution under
These solution-phase syntheses of deuterium-labelled
compounds involve the electrophilic aromatic substitution
of the aromatic ring with deuterium chloride resulting in
exchange of the hydrogens for deuterium. However, the
apparent efficiency of this back-exchange under APCI
þ
þ
conditions when H O and/or CH OH were present, given
3
3
2
that the solution methods require many hours at high
temperatures to achieve high percentage of deuterium
incorporation, was quite unexpected. The exposure time to
acidic gas-phase cations under APCI was only fractions of a
second, and yet the aromatic ring atom% deuterium was
reduced by up to one-third under certain conditions.
The exchange was even more marked at extremely low
sample concentrations, but it was difficult to take quantitative
1
5,16
acidic conditions.
Further evidence for the gas-phase
reactivity of indoles can be found from the proton affinities
calculated for the carbons at positions 2, 4, 5, 6 and 7, which
ꢁ
1 17
.
are all between 876 and 900 kJ mol
In comparison the
proton affinity of carbons 2 to 5 of nitrobenzene is
ꢁ
1 18
,
6
76kJ mol
while those of water and methanol are 696
ꢁ
1
19
and 761kJ mol , respectively.
A significant implication of these observations is that
quantitative data may be seriously skewed by the conversion
of almost fully deuterated standards of electron-rich
aromatics to lower levels of deuterium incorporation if
direct isotope dilution calculations are used, as is typically
the case in plant hormone studies. Furthermore, the detection
of a D0 peak may not indicate the presence of genuine
Figure 6. Effect of mobile phase flow rate on H/D exchange
5
for nominally D -tryptamine (1b). APCI desolvation tempera-
ture 3508C, mobile phase 50:50 methanol/1% acetic acid, 5 ng
injections. Each point is the average of three loop injections –
error bars show standard deviations.
Copyright # 2010 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2010; 24: 1105–1110
DOI: 10.1002/rcm

1
110 N. W. Davies et al.
unlabelled material in the original sample. As most targeted
quantitative determinations do not include channels for
levels of deuterium incorporation less than the nominal level,
this H/D exchange might go unnoticed unless sufficient of
the D0 form is created to be detected in negative control
samples.
REFERENCES
1
. Kelly T, Gray TR, Huestis MA. J. Chromatogr. B 2008; 867:
194.
2
. Meger M, Meger-Kossien I, Schuler-Metz A, Janket D,
Scherer G. J. Chromatogr. B 2002; 778: 251.
3. Kim I, Huestis MA. J. Mass Spectrom. 2006; 41: 815.
4. Gray TR, Shakleya DM, Huestis MA. J. Chromatogr. B 2008;
8
63: 107.
As exchange of deuterium on an aromatic ring can
potentially occur under the acidic conditions of the APCI
plasma, inclusion of channels for intermediate levels of
deuteration for all deuterated standards in initial exper-
iments may be prudent. In this small study the levels of
analytes were relatively high due to the need to measure
with precision the various partially deuterated forms
down to less than 1% of the total signal, and also due to the
age of the instrument used for the bulk of the work.
Modern mass spectrometers typically have sensitivities in
the femtogram range for many analytes, further increasing
the need for caution when using deuterated standards with
heated API ion sources, as there is a strong inverse
correlation between the amount of analyte and the degree
of H/D exchange.
5
. Sugawara S, Hishiyama S, Jikumaru Y, Hanada A, Nishi-
mura T, Koshiba T, Zhao Y, Kamiya Y, Kasahara H. Proc.
Natl. Acad. Sci. USA 2009; 106: 5430.
. Mazzotti F, di Donna L, Maiuolo L, Napoli A, Salerno R,
Sajjad A, Sindona G. J. Agric. Food Chem. 2008; 56: 63.
. Zweigenbaum J, Heinig K, Steinborner S, Wachs T, Henion J.
Anal. Chem. 1999; 71: 2294.
. Putzbach K, Krucker M, Albert K, Grusak MA, Tang G,
Dolnikowski GG. J. Agric. Food Chem. 2005; 53: 671.
. Quittenden LJ, Davies NW, Smith JA, Molesworth P,
Tivendale ND, Ross JJ. Plant Physiol. 2009; 151: 1130.
6
7
8
9
1
0. Kai K, Horita J, Wakasa K, Miyagawa H. Phytochemistry 2007;
8: 1651.
6
1
1. Daniels S, Espina R, Kao K, Yuan H, Lin J, Diamond S,
Johnson B, Rodgers J, Prakash S, Unger S, Christ D,
Miwa G, Gan L-S, Mutlib AE. Chem. Res. Toxicol. 2008; 21:
1
672.
1
2. Atzrodt J, Derdau V, Fey T, Zimmermann J. Angew. Chem.
Int. Ed. 2007; 46: 7744.
1
1
3. Liu DQ, Hop CECA. J. Pharmaceut. Biomed. 2005; 37: 1.
4. Wolff J-C, Laures AM-F. Rapid Commun. Mass Spectrom. 2006;
2
0: 3769.
15. Mayr H, Kempf B, Ofial AR. Acc. Chem. Res. 2003; 36:
6.
6. Brown HC, Okamoto Y. J. Am. Chem. Soc. 1958; 80:
979.
17. Otero N, Mandado M, Mosquera RA. J.Phys.Chem. 2007; 111:
557.
Acknowledgements
6
The authors thank the Australian Research Council for funds
for both the LCQ and Orbitrap mass spectrometers, Edwin
Lowe for the Orbitrap data and Dr Glenn Jacobson for a
1
4
5
sample of D -salbutamol. PPM is thankful to the University
3
1
8. Pol a´ sˇ ek M, Ture cˇ ek F. J. Am. Chem. Soc. 2000; 122: 9511.
19. Hunter EP, Lias SG. J. Phys. Chem. Ref. Data 1998; 27:
413.
of Tasmania and the School of Chemistry for support
through a Thomas Crawford Scholarship.
Copyright # 2010 John Wiley & Sons, Ltd.
Rapid Commun. Mass Spectrom. 2010; 24: 1105–1110
DOI: 10.1002/rcm