Investigation of the gas phase reactivity of the 1-adamantyl radical using a distonic radical anion approach

Article abstract of DOI:10.1039/b711156h

The gas phase reactions of the bridgehead 3-carboxylato-1-adamantyl radical anion were observed with a series of neutral reagents using a modified electrospray ionisation linear ion trap mass spectrometer. This distonic radical anion was observed to undergo processes suggestive of radical reactivity including radical-radical combination reactions, substitution reactions and addition to carbon-carbon double bonds. The rate constants for reactions of the 3-carboxylato-1-adamantyl radical anion with the following reagents were measured (in units 10^-12 cm³ molecule^-1 s ^-1): ¹⁸O₂ (85 ± 4), NO (38.4 ± 0.4), I₂ (50 ± 50), Br₂ (8 ± 2), CH ₃SSCH₃ (12 ± 2), styrene (1.20 ± 0.03), CHCl₃ (H abstraction 0.41 ± 0.06, Cl abstraction 0.65 ± 0.1), CDCl₃ (D abstraction 0.035 ± 0.01, Cl abstraction 0.723 ± 0.005), allyl bromide (Br abstraction 0.53 ± 0.04, allylation 0.25 ± 0.01). Collision rates were calculated and reaction efficiencies are also reported. This study represents the first quantitative measurement of the gas phase reactivity of a bridgehead radical and suggests that distonic radical anions are good models for the study of their elusive uncharged analogues. The Royal Society of Chemistry.

Full text of DOI:10.1039/b711156h

View Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Investigation of the gas phase reactivity of the 1-adamantyl radical using a
distonic radical anion approach
David G. Harman and Stephen J. Blanksby*
Received 20th July 2007, Accepted 10th September 2007
First published as an Advance Article on the web 19th September 2007
DOI: 10.1039/b711156h
The gas phase reactions of the bridgehead 3-carboxylato-1-adamantyl radical anion were observed with
a series of neutral reagents using a modified electrospray ionisation linear ion trap mass spectrometer.
This distonic radical anion was observed to undergo processes suggestive of radical reactivity including
radical–radical combination reactions, substitution reactions and addition to carbon–carbon double
bonds. The rate constants for reactions of the 3-carboxylato-1-adamantyl radical anion with the
−
12
3
−1 −1
18
following reagents were measured (in units 10 cm molecule s ):
50 ± 50), Br (8 ± 2), CH SSCH (12 ± 2), styrene (1.20 ± 0.03), CHCl
abstraction 0.65 ± 0.1), CDCl
O
2
(85 ± 4), NO (38.4 ± 0.4), I
2
(
2
3
3
3
(H abstraction 0.41 ± 0.06, Cl
3
(D abstraction 0.035 ± 0.01, Cl abstraction 0.723 ± 0.005), allyl
bromide (Br abstraction 0.53 ± 0.04, allylation 0.25 ± 0.01). Collision rates were calculated and
reaction efficiencies are also reported. This study represents the first quantitative measurement of the
gas phase reactivity of a bridgehead radical and suggests that distonic radical anions are good models
for the study of their elusive uncharged analogues.
Introduction
Bridgehead radicals are fascinating molecules with their rigid
structures making them exceptional probes of fundamental chem-
istry. Invariant intermolecular distances and bond angles allow
these species to be utilised in the exploration of subjects such
Scheme 1
as orbital and bond interactions, as well as conformational and
1
substituent effects upon reactivity. Bond angles for a series of
bridgehead radicals have been obtained by electron spin resonance
are supported by previous calculations by Schaefer and co-
workers and are consistent with the radical centre being slightly
1
3
3
(
ESR) spectroscopy from measurement of C hyperfine tensors
and at 77 K the 1-adamantyl radical is estimated to have a bridge-
flattened by H–Cb–Ca hyperconjugation.
◦
2
head C–C–C angle of 113.6 . This experimentally derived value
is congruent with gas phase structure calculations that predict
In solution, bridgehead radicals undergo the expected suite of
free radical reactions including: (i) addition to multiple bonds, (ii)
abstraction of hydrogen or halogen atoms, (iii) coupling with other
radicals, and (iv) b-scission promoted rearrangements, although
such isomerisations are facile only for species containing 3- or
◦
a bond angle of 113 for the 1-admantyl radical using density
3,4
◦
functional methods. Such deviation from the 118 for the near
5–7
planar tert-butyl radical, demonstrates the degree of pyrami-
dalisation around the bridgehead radical. Further computational
studies predict that as the geometry of a carbon-centred radical
deviates increasingly from its most stable planar conformation,
the energy increases since spin delocalisation mechanisms such
2
4-membered rings. Pyramidalisation of a carbon-centred radical
at a bridgehead carbon might be expected to increase its rate of
reaction relative to unconstrained tertiary radicals. Alternatively,
the strained geometry about the bridgehead radical centre might
8
9
as resonance and hyperconjugation are compromised. These
retard reactions in some instances due to frontier orbital effects.
calculations also predict an increase in the bridgehead C–H bond
dissociation energy in strained hydrocarbons as a function of
increases in the sum of the H–C–C bond angles. Such calculations
To this point, however, a paucity of empirical data has provided
a significant impediment to the rigorous exploration of such
ideas. In one of the few experimental comparisons available, rate
constants for the addition of bicyclo[1.1.1]pent-1-yl radical to a-
8
are supported by experimental measurements of spin density at
radical centres that reveal an increase from 83 to 88% upon
transition from 1-adamantyl to, the even further distorted, 1-
7
−1 −1
−14
−1
3
−1
methylstyrene (1.4 × 10 M
s
or 2.32 × 10 molecule cm s )
and hydrogen atom abstraction from 1,4-cyclohexadiene (4.6 ×
2
5
−1 −1
−16
−1
3
−1
bicyclo[1.1.1]pentyl radical. In a previous computational study
10 M
s
or 7.64 × 10 molecule cm s ) were measured at
4
◦
10
from our own laboratory we have shown that the 1-adamantyl
25 C by laser flash photolysis. These rate constants are, respec-
tively, 200 and 50 times greater than those for the same reactions
of tert-butyl, confirming that for these reactions the bridgehead
radical is more reactive than its unconstrained analogue.
radical 2 is pyramidal about C but has shorter Ca–Cb bonds,
1
longer Cb–Cc bonds, larger Cb–Ca–Cb angles and smaller Ca–
Cb–Cc angles than adamantane 1 (Scheme 1). These observations
Bridgehead radicals provide a unique model in which to
investigate the intrinsic reactive properties of carbon-centred
radicals. Gas phase radical reactions, however, have typically been
Department of Chemistry, University of Wollongong, Gwynneville, NSW
2
522, Australia. E-mail: blanksby@uow.edu.au
This journal is © The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 3495–3503 | 3495

View Online
difficult to study for several reasons: the occurrence of secondary
chemistry, high reaction rates, instability of intermediates and
difficulties with detection. Yet, important radical-mediated pro-
cesses such as combustion of hydrocarbons and the production
of photochemical smog often lack solid evidence in support of
11
postulated reaction schemes and putative reaction intermediates.
In this paper, we describe how it is now possible to study a single
step in the reactions of a bridgehead radical using a combination
of distonic radical anions and linear ion trap mass spectrometry.
Ion trap mass spectrometry allows gas phase ions to be isolated
(
purified) by mass selection, contained (stored) within electric
fields for periods of time and fragmented (energised) by collision
with an inert bath gas to yield smaller ions which provide structural
information. Product ions may then be trapped and fragmented
further, often several times. A significant recent development has
been the conversion of ion trap mass spectrometers into ‘gas
phase synthetic reactors’ by the introduction of reagents into
the trap atmosphere such that ion–molecule reactions may be
Scheme 2
resulted in the observation of carboxylate anion 5 at m/z 332
Scheme 2). This ion was trapped then subject to collision induced
dissociation (CID), which produced the desired radical ion 3 of
m/z 178 efficiently, demonstrating that Barton esters are also
effective in the generation of gas phase radicals. The intermediate
acyloxy radical 6 was not observed due to the extremely rapid
12,13
observed.
For our purposes, the technique has the following
(
strengths: the reactions of one type of ion at time may be studied,
like-charge repulsion prevents ions from reacting with each other
(
no interference with radical termination events), and reaction rate
data may be obtained.
As mass spectrometers detect only ionic species, it was necessary
to introduce a charged functional group into the 1-adamantyl
radical. However, the charge and radical centres must remain
separate—both spatially and electronically—in order not to
interfere with the reactivity of the radical. The charge must be
relatively inert, to avoid a predominance of ionic chemistry. Ac-
4
rate of decarboxylation. In the present study, reactions of 3 with
nine reagents were observed and occurred overall in a manner
similar to what would be predicted from a knowledge of solution
phase chemistry. Second order rate constants were measured and
reaction efficiencies were estimated based on theoretical collision
rates.
14
cordingly, the distonic radical anion 3-carboxylato-1-adamantyl
(Scheme 1) was a logical target of choice. The distonic approach
to radical reactivity has been employed previously by the groups
3
Results and discussion
15–17
18
19–21
of Kentt a¨ maa,
Kass and O’Hair.
Validation of kinetic measurements
The calculated structures of the 1-adamantyl radical 2 and 3-
carboxylato-1-adamantyl radical anion 3 reveal that comparable
bond lengths and angles are very similar, suggesting that distonic
As a test of the accuracy of our experimental approach, we
measured the rate constants of three ion–molecule reactions whose
kinetics had previous been determined by flowing afterglow-
selected ion flow tube (FA-SIFT) mass spectrometry (Table 1).
These reactions were: (1) the displacement of iodide from
iodomethane by Cl (eqn (1)), (2) the displacement of iodide
from iodomethane by Br (eqn (2)), and (3) the bimolecular reac-
tion of 1,1,1,3,3,3-hexafluoroisopropoxide ion with bromoethane
eqn (3)). Although this last reaction yields products resulting from
both a substitution and elimination reaction (eqn (3a) and (3b),
ion 3 makes an excellent model to study the reactivity of 2
4
(
Scheme 1). Radical anion 3 can be generated by oxidative
25,26
decarboxylation of the 1,3-adamantane dicarboxylate dianion 4
4
(
Scheme 2), in a similar manner to the production of distonic 4-
35
−
carboxylato-1-phenyl radical anions from benzene dicarboxylate
79
−
18
dianions. However, low yields from this process have been
encountered in the presence of some gas-phase reagents due
to the preponderance of ionic reactions of the dianion. A
convenient solution to this problem lay in the derivatisation of
one carboxylic acid group of 1,3-adamantane dicarboxylic acid to
its N-acyloxypyridine-2-thione—or colloquially, Barton ester—
(
−
respectively), Br (m/z 79/81) is the only charged product detected
in this experiment and thus the rate constant determined is in
fact the sum of the rate constants of two second-order processes.
As a check of accuracy with reactions of radicals, the rate
22
functional group. Della and Tsanaktsidis have successfully
adopted this function in the generation of radical intermediates for
23
the synthesis of bridgehead halides. In solution, treatment of an
N-acyloxypyridine-2-thione with a radical initiator, or subjection
to thermolysis or photolysis, is known to generate a carbon
centred radical by homolysis of the N–O bond, then subsequent
2
Table 1 Comparison of second order rate constants (k ) for the ion–
molecule reactions depicted in eqn (1)–(4), determined by the current ion
trap method and by FA-SIFT or FT-ICR techniques
3
22
Reaction
k
2
by current ion trap
equation method/cm molecule
Literature k
2
/cm
Literature
method
rapid decarboxylation. Our computational estimate of the bond
dissociation energy of the weak N–O bond of methyl Barton ester
3
−1 −1
−1 −1
s
molecule
s
−
1
−10
−10
−11
−12
25
is 65 kJ mol . Preparation of the adamantyl Barton ester acid was
accomplished in one step from 1,3-adamantane dicarboxylic acid
(1)
2)
(3)
4)
1.2 ± 0.3 × 10
3.0 ± 0.4 × 10
5.6 ± 0.6 × 10
1.66 ± 0.03 × 10
2.89 ± 0.09 × 10
7.83 ± 0.33 × 10
FA-SIFT
FA-SIFT
FA-SIFT
−
−
11
25
25
(
12
24
by adaptation of a standard method. Electrospray ionisation of a
methanolic solution of the Barton ester in the mass spectrometer
3.75 ± 0.1 × 10−
12
8 ± 4 × 10
−12
FT-ICR
15
(
3
496 | Org. Biomol. Chem., 2007, 5, 3495–3503
This journal is © The Royal Society of Chemistry 2007

View Online
constant for the reaction of the distonic 3-carboxylato-1-phenyl
radical anion (7) with dimethyl disulfide (eqn (4)) was determined
and is compared to the value previously reported from Fourier
15
transform ion cyclotron resonance (FT-ICR) mass spectrometry.
The results of kinetic measurements for the reactions shown in
eqn (1)–(4) are presented in Table 1 and represent an average
of at least three separate determinations, each under differing
neutral concentrations. The uncertainties represent one standard
deviation of the precision.
3
5
−
35
−
Cl + CH
3
–I → Cl–CH
3
+ I
(1)
(2)
7
9
−
79
−
Br + CH
3
–I → Br–CH
3
+ I
−
(
CF
3
)
2
CH–O + CH
3
CH
2
–Br → (CF
3
)
2
−
CH–O–CH
2
CH
3
+
Br
(3a)
(3b)
Scheme 3
−
(CF
3
)
2
CH–O + CH
3
CH
CH
2
–Br → (CF
3
) CH–OH
+ Br
2
−
+
2
=CH
2
buffer gas. Further experimental evidence for the bridgehead
structure of 3 is presented below.
The ion at m/z 178 reacts with adventitious dioxygen inside the
ion trap, forming an ion of m/z 210 that has been assigned as the
4
peroxyl radical 9. When the latter ion is isolated and provided with
additional energy in a CID experiment, the original progenitor ion
(
4)
at m/z 178 is reformed. Furthermore, when this new ion of m/z
5
1
78 is subsequently trapped (in an MS experiment), it reforms
m/z 210 at the same rate as initial peroxyl radical formation.
Such observations are consistent with previous reports of the
loss of dioxygen from the benzyl peroxyl radical forming the
benzyl radical and suggest that in this instance the structure
of the adamantyl radical is unchanged following addition and
The measured rate constants obtained for reactions (1)–(3) agree
favourably with the respective FA-SIFT values, considering that
26
the absolute accuracy of the latter method is estimated at ± 20%
29
and an estimation of accuracy of ion-trap derived rate constants of
1
3
±
25%. The measured rate constant for the reaction of a distonic
−
12
subsequent removal of O . In addition, CID of the ion we have
2
radical anions (eqn (4)) of 3.75 ± 0.1 × 10 also agrees well
assigned to be 9 forms a small amount of m/z 193, corresponding
to the 1,2-epoxyadamantane-3-carboxylate anion 9b which forms
by loss of hydroxyl radical (Scheme 3). If m/z 210 were the peroxyl
radical of 8, under CID one should expect also to see formation
of m/z 177 via loss of hydroperoxyl radical, but none is observed.
To ensure that the structure of the radicals of m/z 178 formed
by oxidative decarboxylation of the 1,3-adamantane dicarboxylate
dianion 4 and by homolysis of the Barton ester carboxylate anion
with a value previously obtained by FT-ICR mass spectrometry
−
12
3
−1 −1 15
of 8 ± 4 × 10 cm molecule s . The data suggest that the
linear ion trap mass spectrometer used in this study is expected
to provide highly reproducible gas phase rate measurements (high
precision) and yield rate constants of reasonable absolute accuracy
for the reactions of both closed-shell and radical ions with neutral
reagents.
5
were identical, the kinetics of their reaction with background
Verification of the structure of the radical
oxygen were measured under identical conditions. The ratio of
rate constants was kBarton/kdianion = 0.98 ± 0.05, indicating that
radical structures were identical. In addition, the kinetic plots were
Before discussion of the reactions of the 3-carboxylato-1-
adamantyl radical anion (3), it is important to establish that the
putative structure shown in Scheme 1 is the correct connectivity
of the ion observed at m/z 178 resulting from CID of the Barton
ester anion 5. We have previously addressed the possibility that
2
highly linear (R = 0.9983 and 0.9994 respectively) and reaction
to immeasurably small ion counts of m/z 178 was evident at large
reaction times, indicating that this ion population consists of the
single isomer 3 not a mixture of 3 and 8.
the radical in question may instead be the ring-opened isomer
4
8
(Scheme 3). Tertiary radical 8 is stabilised by resonance
The distonic radical ion of m/z 178 reacts with molecular iodine
to yield an ion of m/z 305, attributed to 3-iodoadamantane-1-
carboxylate anion 10 (Scheme 4). Commercial adamantanecar-
boxylic acid was converted to 3-hydroxyadamantanecarboxylic
acid 11 by treatment with alkaline, aqueous potassium
permanganate. Subsequent treatment of 11 with aqueous
HI yielded 3-iodoadamantanecarboxylic acid 12 quantitatively
(Scheme 4).
interactions with the adjacent carboxylate group and by relief
of ring strain relative to 3, thus making it the most likely
structural alternative to 3. Electronic structure calculations at
the B3LYP/6-31+G(d)//HF/6-31+G(d) level of theory suggest
−
1
30
that an activation energy of 81 kJ mol must be overcome for
4
conversion of 3 into 8. It was previously argued that this barrier
was sufficiently large to prevent isomerisation given: (i) the gentle
method of radical formation, (ii) the fact that CID does not
produce energetic secondary collisions in ion traps and (iii) the
Fig. 1 depicts comparative spectra resulting from CID of m/z
305 for the product resulting from reaction between radical 3 and
iodine (Fig. 1a) and for authentic 3-iodoadamantanecarboxylate
27,28
rapid cooling (5–10 ms)
of ions by collisions with the helium
This journal is © The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 3495–3503 | 3497

View Online
imparts greater collision energy to an ion given the same applied
excitation amplitude. It is clear that the spectra look similar
since the only product detected is iodide ion at m/z 127. In
fact, the iodide ion remained the only product over the range
of collision energies. Fig. 1c illustrates the comparison between
the CID spectra for the authentic and gas-phase synthesized
species at collision energy 15 (arbitrary units) at activation times
ranging from 0–100 ms, representing a wide internal energy
distribution. Data are expressed as iodide ion abundance as a
proportion of the total ion count (i.e., normalized ion counts)
versus activation time. Data points for the trap-synthesized
product ion and the authentic iodocarboxylate anion 10 compare
favourably at all activation times, indicating that the ions of
m/z 305 have identical structure. This evidence strongly supports
the structure assigned to 3, confirming that the adamantane
ring system has not opened upon formation of the radical
nor during subsequent reactions. Analogous experiments were
also undertaken using the reagent bromine instead of iodine.
Authentic 3-bromoadamantanecarboxylic acid was obtained by
HBr treatment of 11. Comparison of CID spectra at several
collision energies indicated that the product ion resulting from
reaction of radical 3 with bromine was entirely consistent with
authentic 3-bromo-1-adamantanecarboxylate (data not shown).
Thus, data from several experiments are consistent with the
structure of the ion at m/z 178 being that of a bridgehead radical
Scheme 4
3
and not the ring opened isomer 8.
Reactions of the bridgehead radical
Reactions of the distonic bridgehead radical ion 3 with nine
reagents were observed and their rate constants determined at
3
07 ± 1 K, the ambient temperature of the ion trap section of
the mass spectrometer. It has been demonstrated that for low
Q values (such as 0.25, used for these experiments) the effective
temperature of ions held at zero collision energy inside an ion
trap is approximately equivalent to the temperature of the buffer
gas, and hence the temperature of the gas inlet to the trap
31,32
itself.
Gronert has determined the effective temperature of ions
contained within a three dimensional ion trap (from the same
manufacturer as the linear trap employed herein) to be 310 ± 20 K,
by measurement of the equilibrium constant of a complexation
32
reaction displaying very large temperature dependence. Collision
rates z have been calculated at 307 K by average dipole orientation
33
34
(
ADO) theory using an available Fortran routine to enable re-
action efficiencies to be estimated. Collision rates for ion–molecule
reactions exceed those for comparable uncharged systems and are
variable because they depend upon the mass of the reagent and
attractive intermolecular forces between the neutral reagent and
the ion. Reaction efficiencies U are calculated by dividing the
Fig. 1 Comparative mass spectra for CID of m/z 305 at a normalised
collision energy of 15 arbitrary units and reaction time 50 ms. Spectrum
(
a) is of the authentic 3-iodoadamantanecarboxylate anion 10 and (b)
the trapped product resulting from the reaction between radical 3 and
iodine. (c) Comparison of the amount of iodide product ion formed
versus activation time for CID at normalised collision energy 15. The
solid curve represents data for authentic 3-iodoadamantanecarboxylate
empirical rate constants k by theoretical collision rates z and are
expressed as a percentage. Reaction rate constants and calculated
reaction efficiencies are presented in Table 2.
2
1
0 and the dashed curve those for the product ion resulting from reaction
Radical–radical combination reactions
of 1-adamantyl-3-carboxylate radical ion 3 with iodine inside the ion trap.
Gross and co-workers have previously observed the gas phase ad-
dition of 32 Da to an ethylpyridinium radical cation in the presence
of oxygen. This observation was attributed to the radical–radical
1
0 (Fig. 1b), generated by deprotonation of parent acid 12. Both
35
spectra were obtained with nominally identical iodine vapour
concentrations since the presence of massive neutrals in the trap
combination of the b-distonic radical cation and dioxygen to form
3
498 | Org. Biomol. Chem., 2007, 5, 3495–3503
This journal is © The Royal Society of Chemistry 2007

View Online
Table 2 Kinetic data, including second order rate constants (k
2
), for the reaction at 307 ± 1 K of 3-carboxylato-1-adamantyl radical (3) with neutral
reagents. Uncertainties represent one standard deviation
3
−1 −1
3
−1 −1
Reagent
Reaction
Rate constant k
2
/cm molecule
s
Collision rate z/cm molecule
s
Efficiency U (%)
1
8
−11
−10
O
2
radical combination
radical combination
I abstraction
Br abstraction
thiomethylation
addition
H abstraction
Cl abstraction
D abstraction
Cl abstraction
Br abstraction
allylation
8.5 ± 0.4 × 10
5.4 × 10
16
5.7
6.8
1.2
0.96
0.65
0.042
0.067
0.0036
0.074
0.055
0.026
−
−
−
11
12
13
−10
−10
−10
−9
NO
I
Br
CH
3.84 ± 0.04 × 10
6.1 × 10
−
−
11
2
5 ± 5 × 10
8 ± 2 × 10
7.3 × 10
12
2
6.7 × 10
−
11
3
SSCH
3
1.2 ± 0.2 × 10
1.2 × 10
−10
−10
−10
−10
PhCH=CH
1.20 ± 0.03 × 10
1.8 × 10
2
−
13
13
CHCl
CHCl
CDCl
CDCl
3
3
3
3
4.1 ± 0.6 × 10
9.7 × 10
−
6.5 ± 1 × 10
3.5 ± 0.9 × 10
9.7 × 10
−
14
9.7 × 10
−10
−10
−10
7.23 ± 0.05 × 10
9.7 × 10
−
13
CH
CH
2
=CH–CH
2
2
Br
Br
5.3 ± 0.4 × 10
9.7 × 10
2.5 ± 0.1 × 10−
13
9.7 × 10
=CH–CH
2
a peroxyl radical. In the present study, the bridgehead radical
was observed to undergo a radical–radical combination with
both (a) dioxygen and (b) nitric oxide (Scheme 5). Reaction of
phase experiments, this comparison presents the first evidence for
the enhanced reactivity of bridgehead radicals in the gas phase.
The reaction of 3 with nitric oxide produced an ion of m/z
208, assigned to be the 3-nitrosoadamantanecarboxylate anion
3
1
8
the radical 3 with O-labelled dioxygen yields a major product of
−
11
3
−1 −1
−11
3
−1 −1
m/z 214, with a rate constant of 8.5 × 10 cm molecule s .
The product ion is ascribed the structure of doubly O-labelled
-carboxylato-1-adamantylperoxyl radical anion 13 and is formed
14, with rate constant 3.84 ± 0.04 × 10 cm molecule
efficiency of about 6%. The addition of NO to the radical anion
rather than the displacement of CO (as previously observed by
s
and
1
8
3
2
37
with a reaction efficiency of 16%. Given that no kinetic barrier is
expected in the formation of the peroxyl radical, the relatively low
efficiency for this reaction reflects the ability of this alkyl peroxyl
Wenthold and Squires for the acetate radical anion) provides
further support to the distonic character of 3. No rate constant
could be found in the literature for the reaction of a bridgehead
radical with NO but previous studies of reactions of carbon-
centred radicals with NO have been reported and represent an
interesting comparison. The rate constant for the reaction of tert-
butyl with NO determined by the muon spin relaxation technique
−
1
radical to accommodate the 169 kJ mol of energy released by
4
the formation of the nascent carbon–oxygen bond. The kinetic
isotope effect is expected to be small between the rate constants for
1
8
16
reaction of 3 with O
2
and naturally abundant O . Consequently,
2
−
11
3
−1 −1
the measured rate constant can be used in conjunction with the
competing appearance of unlabelled peroxyl radicals at m/z 210 to
estimate the background concentration of dioxygen within the ion
is 1.0 ± 0.3 × 10 cm molecule
2.2 ± 0.6 × 10 cm molecule
0.1 × 10 cm molecule
s
and that for ethyl radical is
−
11
3
−1 −1
38
s
at 307 K. A value of 3.5 ±
−
11
3
−1
−1
s
was measured by colour centre
9
−3
trap to be 3.0 ± 0.2 × 10 molecules cm . This value is remarkably
laser kinetic spectroscopy for the reaction of ethynyl radical with
1
0
−3
39
−12
3
−1 −1
close to the crude estimate of 10 molecules cm presented in our
previous communication.
NO and 9.5 ± 1.2 × 10 cm molecule
s
for benzyl radical
4
40
with NO at 298 K by laser flash photolysis. Our value compares
favourably with these, with a rate constant ca. four times greater
than that of tert-butyl, which could be accounted for by increased
reactivity arising from bridgehead strain and from greater collision
frequency arising from the ion–dipole interactions.
Substitution reactions
Scheme 5
Radical 3 underwent substitution reactions with (a) iodine, (b)
bromine, (c) dimethyl disulfide, (d) chloroform, (e) deuterochlo-
roform and (f) allyl bromide (Scheme 6). The concentrations of
bromine and iodine in the trap were estimated by measuring the
mass decrease over time of the halogen in a sealed vial connected
to the low pressure helium inlet to the trap. Uncertainties represent
primarily those in the mass loss measurements. Formation of 3-
iodo- 10 and 3-bromoadamantanecarboxylate anion 15 occurred
Bayes and co-workers have previously measured the rate
constants for the reaction of butyl radicals generated by laser
in the gas phase at room temperature
by photoionisation mass spectrometry. Their results for n-butyl
flash photolysis with O
2
36
(
0
0.75 ± 0.14), sec-butyl (1.66 ± 0.22) and tert-butyl (2.34 ±
−
11
3
−1 −1
◦
.39 × 10 cm molecule s ) reveal the reactivity trend 3 >
◦
◦
2
> 1 and are of comparable magnitude to those obtained
in the present study using the distonic radical anion approach.
Furthermore, using simple hard sphere theory, we estimate the
collision frequency for a tert-butyl radical with dioxygen at 298 K
−
10
3
−1 −1
is 3.9 × 10 cm molecule s , resulting in a reaction efficiency
of 6.0%. This value is less than half that for the reaction of 3,
suggesting that the bridgehead radical is more reactive with oxygen
than its planar cousin. Consistent with expectation from solution
Scheme 6
This journal is © The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 3495–3503 | 3499

View Online
−
11
−12
3
with rate constants of 5 ± 5 × 10 and 8 ± 2 × 10 cm
−
1
−1
molecule
s
respectively. Given that the rate constants could
not be measured with high precision, it is difficult to establish with
certainty which of the two is faster.
Kentt a¨ maa and co-workers have previously used dimethyl
disulfide (DMDS) as an effective probe of the distonic character
Scheme 8
17
of radical cations. They have established that non-distonic
resonance stabilised) species are generally unreactive towards the
(
at gas phase free radical polymerisation—was not observed,
indicating that the resonance stabilised intermediate radical was
considerably less reactive than the bridgehead radical initiator.
However, addition of adventitious dioxygen to the benzylic radical
to give peroxyl radical 22 was observed. Once formed, subsequent
trapping of 21 gave amounts of m/z 314, the yields escalating with
increasing reaction time. This process was considerably slower
reagent while distonic radical cations undergo a substitution reac-
tion at sulfur resulting in formation of a thiomethyl ether. From
the literature, however, the case for radical anions is less clear, but
at face value one should expect comparable reactivity. Bridgehead
radical 3 underwent a relatively rapid reaction (k = 1.2 ± 0.2 ×
−
11
3
−1 −1
10
cm molecules s , U = 0.96%) with DMDS to yield an ion
of m/z 225, assigned as the 3-methylthioadamantanecarboxylate
anion 16, demonstrating the ease with which methylthio radicals
than the reaction of 3 with O
2
, an estimate for k
2
being 2 ± 1 ×
−13
3
−1 −1
10
cm molecule s .
may be displaced in an S
H
2 process. The observation of a CH S–
3
adduct and the relative efficiency of this reaction is taken as
additional evidence for the distonic structure of 3. For comparison,
distonic aryl radical anion 3-carboxylato-1-phenyl 7 reacts with
Conclusion
Electrospray ionisation ion trap mass spectrometry has enabled
generation of the distonic 3-carboxylato-1-adamantyl radical
anion 3 from a readily prepared precursor and allowed the study
of the radical reactions of this mass-selected bridgehead radical
with neutral reagents. Strong evidence has been obtained for
the intact bridgehead radical structure. Reaction with common
reagents occurs in a manner expected from classic solution phase
free radical chemistry, with radical–radical combination, addition
to double bonds and substitution reactions observed. Rate data
have been obtained for the first time for 12 reactions of 3 in
the gas phase and represent the first gas phase rate data for
−
12
3
−1 −1
DMDS with a rate constant of 8 × 10 cm molecules
s
and
15
efficiency of 0.5%.
Chlorine atom abstraction from chloroform by 3 to form
9 is sluggish compared with most other reactions, but clearly
1
consonant with Cl abstraction from CDCl
respectively) as expected. A large primary kinetic isotope effect of
/k
= 13 ± 2 was evident for hydrogen/deuterium atom removal
in the formation of 17 and 18 respectively. The ratio of abstraction
of hydrogen to chlorine in chloroform was k
3
(U = 0.067 and 0.074%
k
H
D
H
/kCl = 0.62.
Both bromination and allylation of 3 with allyl bromide,
forming 15 and 20 respectively, were relatively slow compared
to other substitution processes, with alkylation being favoured
any bridgehead radical. Rate constants spanned several orders
18
of magnitude, the fastest reaction being the addition of O
2
(8.5 ×
by a factor of 2.1. Given the steric bulk of 3, installation of the
−10
3
−1 −1
10
cm molecule s ) whose rate constant was 2400 times that
ꢀ
allyl group at the bridgehead carbon most likely occurs via an S
H
2
−14
for the abstraction of a deuterium atom from CDCl
3
(3.5 × 10 ).
41
mechanism, with 3 attacking at the methylene of the double bond
Scheme 7a). Based on these data, however, a direct S 2 reaction
A comparison of reaction efficiencies for the reactions of 3 and tert-
butyl radical with dioxygen suggest that the pyramidal bridgehead
species is more reactive than its near-planar relative.
(
H
at the allylic position cannot be rigorously excluded (Scheme 7b).
Experimental
Reagents (AR grade) and solvents were obtained commercially,
taken from freshly opened bottles and used without further
1
8
purification, unless stated otherwise.
2
O gas (95%) was obtained
˚
from Cambridge Isotopes. DMF was dried over activated 3 A
molecular sieves. Crystalline N-hydroxypyridine-2-thione was
obtained economically by treatment with concentrated HCl of a
commercial 40% aqueous solution of sodium 2-mercaptopyridine
42
N-oxide, as described previously. Chloroform was run through
silica gel to obtain a sample which was ethanol free. Allyl bromide
was treated in the same manner to afford a colourless sample.
Scheme 7
Addition to carbon–carbon double bonds
N-(1-Adamantanoyloxy-3-carboxylic acid)pyridine-2-thione
2
4
Radical 3 underwent addition to styrene yielding a product ion
of m/z 282, almost certainly the stabilised benzylic radical 21
A general method was adapted and performed under minimal
light to minimise decomposition of the photolabile product. A
stirred solution of 1,3-adamantanedicarboxylic acid (100 mg,
(
Scheme 8). This reaction appeared considerably slowed due to
◦
the decreased collision frequency resulting from the low polarity of
styrene, but the efficiency was in fact two-thirds that for reaction of
0.446 mmol) in 500 lL dry DMF was cooled to 0 C. A
solution of dicyclohexylcarbodiimide (94 mg, 0.46 mmol) and N-
3
with DMDS. A subsequent addition of styrene to 21—an attempt
hydroxypyridine-2-thione (55 mg, 0.43 mmol) in 1.5 mL dry DMF
3
500 | Org. Biomol. Chem., 2007, 5, 3495–3503
This journal is © The Royal Society of Chemistry 2007

View Online
was added dropwise over 5 min. The mixture was allowed to warm
to room temperature, stirred for an additional 8 h, then filtered to
remove the insoluble dicyclohexyl urea. A further 2 mL DMF was
used to rinse the precipitate. The filtrate was treated with 15 mL
3-Iodoadamantanecarboxylic acid 12
-Hydroxyadamantane-1-carboxylic acid (501 mg, 2.55 mmol)
3
and 55% aqueous HI (5.0 mL, 66 mmol) were placed in a capped
◦
◦
reaction vial and stirred for 16 h at 60 C. After cooling to 0 C, the
mixture was filtered and the precipitate washed with 3 × 5 mL cold
water to remove all colour from the product. After drying under
vacuum, the yield of white crystals was 0.780 g (99.8%), mp 166–
water, causing the yellow product to precipitate. After cooling to
◦
0
C, the product was collected by filtration, washed with 10 mL
cold water and dried under vacuum to give a yellow solid (110 mg).
A portion (40 mg) was purified by flash chromatography on silica,
◦
44
◦
1
68 C (lit. 165 C), and the sample was used without further
purification. d (500 MHz, CDCl ) 1.75–1.82 (2H, m), 1.95–2.05
(4H, m), 2.02–2.12 (2H, m), 2.53–2.2.60 (4H, m), 2.71–2.79 (2H,
m), 5.00 (1H, s, br, OH). d (125 MHz, CDCl ) 31.9 (2C), 34.4,
eluting with 20–40% (v/v) EtOAc in chloroform (R 0.34, 20%
f
H
3
EtOAc), which yielded a yellow, crystalline solid (18 mg, 35%), mp
◦
1
72.5–174 C. Recrystallisation from CHCl
3
–hexane gave yellow
SOCD ) d 1.76 (s,
◦
C
3
prisms of mp 176–177 C. d
H
(500 MHz, CD
3
3
3
6.8 (2C), 44.6, 45.7 (C–I), 51.0 (2C), 52.2, 181.5 (C=O). m (neat
br, 2H), 2.08–2.11 (m, 8H), 2.26 (s, br, 2H), 2.44 (s, br, 2H), 6.88
ddd, 1H), 7.43 (ddd, 1H), 7.54 (dd, 1H), 8.34 (dd, 1H), 12.20 (s,
−1
powder)/cm 2934, 2904, 1704 (vs), 1450, 1413, 1332, 1283 (s),
1261 (s), 1233, 1076, 967, 893, 820.
(
−
1
br, 1H). m (neat powder)/cm 3327, 2932, 2858, 1785, 1714 (s),
1
1
699 (vs), 1626, 1610, 1576, 1529, 1454, 1413, 1276, 1232, 1132,
053, 978, 906, 754.
Nitric oxide
Nitric oxide was generated and purified on a 50 mL scale by
45
an ingeniously simple established method. The reaction in a
60 mL capped disposable syringe between sodium nitrite (0.250 g,
3-Hydroxyadamantanecarboxylic acid 11
3
1
.62 mmol) and ferrous sulfate (3.0 mL of an aqueous solution
.2 M in FeSO and 1.6 M in H SO ) produced about 60 mL
3
0
A published method was altered slightly for convenience. KOH
4.955 g, 88.31 mmol) was dissolved in 200 mL water and 1-
adamantanecarboxylic acid (20.0 g, 111 mmol) was dissolved
portionwise in the stirred solution. KMnO (20.0 g, 127 mmol) was
4
2
4
(
of crude nitric oxide. After expulsion of the liquid, the gas was
washed by drawing 5 mL of aqueous 1 M NaOH solution into
the syringe and shaking. The base wash was expelled and the now
colourless gas was washed further with 2 × 5 mL water, and then
dried by passage into a second 60 mL syringe, through PVC tubing
4
added in several portions and the mixture was stirred overnight,
then heated at 80 C until it became brown in colour. After cooling
to RT, concentrated H
190 mmol) NaHSO to destroy the MnO
◦
2
SO
4
was added (∼30 mL), followed by 20 g
containing a plug of anhydrous MgSO .
4
(
3
2
. An aqueous suspension
of the white product resulted, which was treated with 20 g NaCl,
then cooled to 0 C, filtered and washed repeatedly with 10%
Mass spectrometry
◦
aqueous H
2
SO
4
, then with cold water. After drying the product
Mass spectra were obtained using a Finnigan (now ThermoFisher)
LTQ quadrupole linear ion trap spectrometer (San Jose, CA).
28
under vacuum (20.9 g, 96% crude), it was dissolved in 320 mL
boiling EtOAc, hot filtered and allowed to recrystallise to yield
white prisms (9.09 g, 42%) of mp 203–205.5 C (lit. 202–203 C).
A second crop (2.57 g, 12%) of mp 198–200 C was obtained by
concentration of the mother liquors. d
Solutions of analytes at concentrations below 1 mM in HPLC
grade methanol were subject to electrospray ionization (ESI) in
negative mode. Capillary potentials were typically 3–4 kV. Once
formed, anions were held in the ion trap at a manufacturer-stated
pressure of ultrahigh purity helium (BOC, Australia) and reagent
of 2.50 ± 0.20 mTorr. Pressure inside the trap was measured at
◦
30
◦
◦
H
(300 MHz, CD
3
SOCD
3
)
1
2
.47–1.52 (2H, m), 1.52–1.56 (4H, m), 1.62–1.67 (6H, m), 2.08–
.15 (2H, m), 4.49 (1H, s, br, OH), 12.01 (1H, s, br, COOH). d
SOCD
C
25
(
75 MHz, CD
3
3
) 29.7 (2CH), 34.9, 37.6 (2C), 42.9 (quat),
2.58 ± 0.13 mTorr using a known rate constant for the reaction
of bromide with methyl iodide (eqn (2)). Radical 3 was formed
by mass selection of ion 5 (2–5 Th isolation window), followed
by CID (25–35 arbitrary normalised collision energy units) for a
time of 30 ms. Once formed, the ion isolation window for radical
−
1
4
3
8
4.3 (2C), 46.5, 66.3 (C–O), 177.7 (C=O). m (neat powder)/cm
446, 2909, 1705 (vs), 1265 (vs), 1248 (vs), 1229, 1120, 1010, 941,
80, 723.
3
was set to 2.0 Th to exclude isotope peaks and the Q-value
set to 0.25 to ensure that the effective temperature inside the
trap was not significantly elevated above ambient. Ions arriving
from the ion optics were subjected to two trapping cycles prior to
kinetic analysis to ensure their effective thermalisation with buffer
gas. An acquisition of 100–200 scans was usually obtained and
reaction times varied from 10 ms to 5 s.
3
-Bromoadamantanecarboxylic acid
3
-Hydroxyadamantane-1-carboxylic acid (500 mg, 3.55 mmol)
and 48% aqueous HBr (5.0 mL, 44 mmol) were placed in a capped
◦
◦
reaction vial and stirred for 3.5 h at 90 C. After cooling to 0 C,
the mixture was filtered and the precipitate washed with 5 mL
cold water. The resulting product was dried under vacuum to
Reagent molecules were provided to the linear trap atmosphere
in known concentration in a manner similar to that reported by
◦
43
yield white crystals (0.659 g, 99.8%), mp 144.5–146.5 C (lit.
◦
12
145–146 C), used without further purification. d
H
(500 MHz,
Gronert et al. for a 3-D trap instrument. Mixtures of helium
CDCl
3
) 1.65–1.74 (2H, m), 1.89–1.96 (4H, m), 2.21–2.23 (2H, m),
and reagent enter the trap inside the mass spectrometer via a
nipple on the exit endcap electrode and exit via small holes in the
entrance and exit endcap electrodes. A mixing system allowing the
introduction of reagents into the helium stream was constructed
(Fig. 2). A reagent (liquid or gas) is injected by pump-driven
2
.28–2.35 (4H, m), 2.48–2.53 (2H, m), 5.25 (1H, s, br, OH). d
C
(
4
2
125 MHz, CDCl
3
) 31.5 (2C), 34.4, 36.8 (2C), 44.6, 48.0 (2C),
9.3, 63.2 (C-Br), 181.7 (C=O). m (neat powder)/cm− 3446, 2914,
1
858, 1689 (vs), 1293, 1280, 1090, 948, 830, 768.
This journal is © The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 3495–3503 | 3501

View Online
removed more quickly from the trap into the surrounding vacuum
chamber.
k
P
1
k
2
=
(5)
n
ꢀ
ꢁ
᭹
[
R ]
t
ln
= − k
1
+ c
(6)
᭹
[
R ]0
ꢂ
molar flow rate of neutral
molar flow rate of helium
molar mass of neutral
molar mass of helium
P
n
= P
T
×
×
(
7)
Fig. 2 A schematic drawing of the mixing system that permits the
introduction of reagents into the ion trap mass spectrometer.
It is known that for low Q values, the effective temperature
of ions at equilibrium inside the trap is equivalent to ambient
31,32
temperature.
The temperature of the case surrounding the ion
trap was measured at 307 ± 1 K, which is taken as being the
effective temperature for these experiments. In some instances,
reaction of the radical with a neutral formed two or more products.
If a secondary species was formed as a result of decomposition of
a primary product, its concentration was attributed to that of
the primary to ensure accurate kinetics. Formation of products
by separate pathways was treated by a standard parallel pseudo-
first order kinetics analysis. Collision rate coefficients for the ion–
molecule reactions were estimated by average dipole orientation
syringe into the helium gas (3–5 psi) flow. The stainless steel helium
line (2 m) is wound around a copper bar fitted with a thermostatted
◦
1
00 W electric heating element so as to allow heating (25–250 C)
of the gas flow if volatility of the reagent at room temperature
is insufficient. A variable leak valve (Granville-Phillips Model
2
03, Boulder, Colorado) in the heated zone admits a flow of the
reagent–helium mixture to the ion trap such that the pressure may
be varied in the trap. A second metering valve can be used to
vary the concentration of reagent by controlling the split flow,
similar to a GC injector port. For admission of corrosive reagents
or those possessing a very low vapour pressure, an inlet restricted
with PEEK tubing is provided. The entire heated zone can be
placed under vacuum to expedite removal of a reagent prior to the
introduction of the next.
33
(
ADO) theory for 307 K. The dipole moment, molar mass and
polarisability are required for the neutral and the charge and molar
mass are required for the distonic ion. An available Fortran routine
34
was used to facilitate these calculations.
Acknowledgements
Kinetic measurements
SJB and DGH acknowledge the Australian Research Council
(DP0452849) and the University of Wollongong for funding.
The authors acknowledge Mr Benjamin Kirk for computational
assistance and Messrs Martin Riggenbach, Steven Cooper, Larry
Hick and Peter Sarakiniotis for assistance in the design and
construction of equipment. We gratefully acknowledge Professor
Richard O’Hair, Drs George Khairallah and Tom Waters for
helpful discussions regarding kinetic analyses and the design of
instrument modifications.
4
27,28
At full capacity, the trap contains approximately 2 × 10 ions.
A practical concentration range for neutral reagents inside the
trap is 10 –10 molecules cm and the volume inside the trap is
9
13
−3
3
27,28
1
3.5 cm .
Consequently, the molar ratio of neutral reagent to
6
10
ions is of the order 10 –10 , clearly establishing the ions as the
limiting species. Thus, bimolecular reactions of ions and neutral
reagents display pseudo-first order kinetics.
3
−1 −1
True second order rate constants, k
2
(cm molecule s ) for the
reactions of 1-adamantyl-3-carboxylate radical anion with each
12
neutral reagent were obtained by Gronert’s method from the
References
−
1
pseudo-first order rate constant, k
neutral reagent in the ion trap, P
eqn (5). Pseudo-first order rate constants, k
1
(s ) and the pressure of the
−
3
(molecule cm ) according to
1 J. C. Walton, Chem. Soc. Rev., 1992, 21, 105–112.
2 C. J. Rhodes, J. C. Walton and E. W. Della, J. Chem. Soc., Perkin Trans.
n
−
1
1
(s ) were obtained
2
, 1993, 2125–2128.
from a series of mass spectra, recorded as a function of reaction
time between the radical ion (R) and neutral reagent. The reaction
time is defined as the interval between the isolation of the selected
ion and ejection of all ions from the trap for analysis. A plot of the
natural logarithm of the abundance of the radical ion at m/z 178
as a proportion of the total signal abundance (initial abundance of
R) against reaction time yielded a linear relationship, with slope
3
G. Yan, N. R. Brinkmann and H. F. Schaefer, J. Phys. Chem. A, 2003,
107, 9479–9485.
4
5
D. G. Harman and S. J. Blanksby, Chem. Commun., 2006, 859–861.
D. Griller, K. U. Ingold, P. J. Krusic and H. Fischer, J. Am. Chem. Soc.,
1
978, 100, 6750–6752.
6
7
J. Pacansky, W. Koch and M. D. Miller, J. Am. Chem. Soc., 1991, 113,
317–328.
B. Noller, R. Maksimenka, I. Fischer, M. Armone, B. Engels, C.
Alcaraz, L. Poisson and J. M. Mestdagh, J. Phys. Chem. A, 2007, 111,
of −k
extended to 4–5 half lives. Pressure of the neutral reagent (P
for each reaction was calculated from the pressure inside the ion
trap (P ) and the respective molar flows and molecular weights
1
(eqn (6)). Plots consisted usually of 8 data points and
1
771–1779.
n
)
8 Y. Feng, L. Liu, J. T. Wang, S. W. Zhao and Q. X. Guo, J. Org. Chem.,
2004, 69, 3129–3138.
9
I. Fleming, Frontier Orbitals and Organic Chemical Reactions, Wiley-
T
Interscience, Chichester, 1976.
of the helium damping gas and reagent, using eqn (7). The final
term accounts for difference in effusion, since lighter molecules are
1
0 J. T. Banks, K. U. Ingold, E. W. Della and J. C. Walton, Tetrahedron
Lett., 1996, 37, 8059–8060.
3
502 | Org. Biomol. Chem., 2007, 5, 3495–3503
This journal is © The Royal Society of Chemistry 2007

View Online
1
1 P. D. Lightfoot, R. A. Cox, J. N. Crowley, M. Destriau, G. D. Hayman,
29 F. F. Fenter, B. Noziere, F. Caralp and R. Lesclaux, Int. J. Chem. Kinet.,
1994, 26, 171–189.
30 G. L. Anderson, W. A. Burks and Harruna, II, Synth. Commun., 1988,
18, 1967–1974.
31 A. V. Tolmachev, A. N. Vilkov, B. Bogdanov, L. Pasa-Tolic, C. D.
Masselon and R. D. Smith, J. Am. Soc. Mass Spectrom., 2004, 15,
1616–1628.
32 S. Gronert, J. Am. Soc. Mass Spectrom., 1998, 9, 845–848.
33 T. Su and M. T. Bowers, Classical ion-molecule collision theory, in Gas
Phase Ion Chemistry, ed. M. T. Bowers, Academic Press, New York,
1979, vol. 1, p. 83.
34 K. F. Lim, Quantum Chemistry Program Exchange, 1994, 14, 3.
The program Colrate (1994) is available for download from the
author’s website at Deakin University, Geelong, Victoria, Australia:
http://www.deakin.edu.au/∼lim/programs/COLRATE.html.
35 S. J. Yu, C. L. Holliman, D. L. Rempel and M. L. Gross, J. Am. Chem.
Soc., 1993, 115, 9676–9682.
M. E. Jenkin, G. K. Moortgat and F. Zabel, Atmos. Environ., Part A,
1
992, 26, 1805–1961.
1
1
1
1
1
1
1
1
2
2 S. Gronert, L. M. Pratt and S. Mogali, J. Am. Chem. Soc., 2001, 123,
081–3091.
3 T. Waters, R. A. J. O’Hair and A. G. Wedd, J. Am. Chem. Soc., 2003,
25, 3384–3396.
4 B. F. Yates, W. J. Bouma and L. Radom, J. Am. Chem. Soc., 1984, 106,
805–5808.
3
1
5
5 C. J. Petzold, E. D. Nelson, H. A. Lardin and H. I. Kentt a¨ maa, J. Phys.
Chem. A, 2002, 106, 9767–9775.
6 K. K. Thoen, R. L. Smith, J. J. Nousiainen, E. D. Nelson and H. I.
Kentt a¨ maa, J. Am. Chem. Soc., 1996, 118, 8669–8676.
7 K. M. Stirk, J. C. Orlowski, D. T. Leeck and H. I. Kentt a¨ maa, J. Am.
Chem. Soc., 1992, 114, 8604–8606.
8 D. R. Reed, M. Hare and S. R. Kass, J. Am. Chem. Soc., 2000, 122,
1
0689–10696.
9 A. Karnezis, C. K. Barlow, R. A. J. O’Hair and W. D. McFadyen, Rapid
36 T. M. Lenhardt, C. E. McDade and K. D. Bayes, J. Chem. Phys., 1980,
72, 304–310.
Commun. Mass Spectrom., 2006, 20, 2865–2870.
0 S. Wee, A. Mortimer, D. Moran, A. Wright, C. K. Barlow, R. A. J.
O’Hair, L. Radom and C. J. Easton, Chem. Commun., 2006, 4233–
37 P. G. Wenthold and R. R. Squires, J. Am. Chem. Soc., 1994, 116, 11890–
11897.
4
235.
38 H. Dilger, M. Stolmar, U. Himmer, E. Roduner and I. D. Reid, J. Phys.
Chem. A, 1998, 102, 6772–6777.
39 J. W. Stephens, J. L. Hall, H. Solka, W. B. Yan, R. F. Curl and G. P.
Glass, J. Phys. Chem., 1987, 91, 5740–5743.
40 T. Ebata, K. Obi and I. Tanaka, Chem. Phys. Lett., 1981, 77, 480–
483.
2
2
1 C. K. Barlow, S. Wee, W. D. McFadyen and R. A. J. O’Hair, Dalton
Trans., 2004, 3199–3204.
2 D. H. R. Barton, D. Crich and W. B. Motherwell, J. Chem. Soc., Chem.
Commun., 1983, 939–941.
2
2
3 E. W. Della and J. Tsanaktsidis, Aust. J. Chem., 1989, 42, 61–69.
4 D. H. R. Barton, J. MacKinnon, R. N. Perchet and C. L. Tse, Efficient
synthesis of bromides from carboxylic acids containing a sensitive
functional group: dec-9-enyl bromide from 10-undecenoic acid, in
Organic Syntheses, Collective Volumes, ed. R. L. Danheiser, John Wiley
41 G. A. Russell, P. Ngoviwatchai and Y. W. Wu, J. Am. Chem. Soc., 1989,
111, 4921–4927.
42 D. H. R. Barton, D. Bridon, I. Fernandezpicot and S. Z. Zard,
Tetrahedron, 1987, 43, 2733–2740.
&
Sons, New York, 2004, vol. 10, p. 237.
5 S. Gronert, C. H. Depuy and V. M. Bierbaum, J. Am. Chem. Soc., 1991,
13, 4009–4010.
6 C. H. Depuy, S. Gronert, A. Mullin and V. M. Bierbaum, J. Am. Chem.
Soc., 1990, 112, 8650–8655.
43 I. Handa and L. Bauer, J. Chem. Eng. Data, 1984, 29, 223–225.
44 H. Stetter and J. Mayer, Chem. Ber., 1962, 95, 667–672.
45 (a) B. M. Mattson and J. Lannan, Microscale Gas Chemistry, Part
5. Experiments with Nitrogen Oxides, Chem13 News, 1997, Feb., 255;
(b) B. M. Mattson, M. P. Anderson and S. E. Mattson, Microscale Gas
Chemistry, 4th edn, Educational Innovations, 2006; (c) The author
has instructions on his website (accessed last on 27 June 2007):
http://mattson.creighton.edu/AllGases.html.
2
2
1
2
2
7 J. C. Schwartz, ThermoFisher, personal communication.
8 J. C. Schwartz, M. W. Senko and J. E. P. Syka, J. Am. Soc. Mass
Spectrom., 2002, 13, 659–669.
This journal is © The Royal Society of Chemistry 2007
Org. Biomol. Chem., 2007, 5, 3495–3503 | 3503