MAGNETIC RESONANCE IN CHEMISTRY

Magn. Reson. Chem. 37, 767–773 (1999)

ESR of new b-phosphorus nitroxides from aldo- and

keto-nitrones

D. Lawrence Haire,1 ∗ Edward G. Janzen1 and Ivan Hrvoic2

1
Department of Clinical Studies, Ontario Veterinary College, University of Guelph, Guelph, Ontario, N1G 2W1, Canada
2
Gem Systems, 52 W. Beavercreek Road, Unit 14, Richmond Hill, Ontario, L4B 1L9, Canada

Received 15 March 1999; revised 19 May 1999; accepted 22 May 1999

ABSTRACT: This investigation deals mainly with the synthesis of new phosphorus–nitroxide free radicals. Nitroxides
with a ˇ 31 P group were synthesized from the aldo-nitrones ˛-phenyl-N-tert-butylnitrone and 5,5-dimethyl-1-
pyrroline-N-oxide (DMPO) as well as the keto-nitrones 2,5,5-trimethyl-1-pyrroline-N-oxide (2,5,5-M3 PO) and 2-
phenyl-5,5-dimethyl-1-pyrroline-N-oxide (2-Ph-DMPO). Three different routes were explored. The first two routes
involved nitronyl spin trapping of phosphorus-centred radicals generated thermally or photochemically by hydrogen
abstraction by tert-butoxyl radicals from di-tert-butyl peroxyoxalate or from photoexcited benzophenone, respectively.
The third approach used the addition of phosphorus anions (generated by the strong, hindered base lithium
diisopropylamide) to the nitrones and subsequent one-electron oxidation to the corresponding nitroxides. The most
novel feature of the work presented here may be the synthesis of ˇ-phosphoranyl [P(OR)4 ] nitroxide spin adducts. For
comparison purposes the related ˇ-phosphityl [P(O)(OR)2 ] nitroxides were also assembled and characterized by ESR
spectroscopy. Copyright  1999 John Wiley & Sons, Ltd.
KEYWORDS: phosphorus-substituted nitroxides (aminoxyls); nitrones; ESR; spin trapping; free radicals;
magnetometry

INTRODUCTION We reasoned that the phosphoranyl moiety could be

‘stabilized’ by covalent attachment ‘ˇ’ to the nitroxide
The present study is devoted mainly to the search for function (Scheme 2). This approach has been applied
new, persistent free radicals that may be applicable in to advantage, for example, in some very recent spin
weak-field magnetometers.1 To be useful in magnetome- trapping1,4 – 9 and nonspin trapping6,10 – 12 systems using
ters, a free radical should be persistent and exhibit a phosphityl [P(O)(OR)2 ] groups. Of course, the 31 P hyper-
large hyperfine splitting [i.e. in its electron spin reso- fine splitting is expected to be much less than that of phos-
nance (ESR) spectrum]. This phenomenon is related to phoranyls (ca 90 mT) because the unpaired electron is two
the dynamic nuclear polarization (DNP) effect. Owing to bonds away (ˇ) from the nitroxide nitrogen (Scheme 2).
ž
their remarkable stability, nitroxide free radicals (R2 NO ) In this study, we examined nitroxides derived from both
have become the most popular molecules for magnetom- aldo- and keto-nitrones. This nomenclature is based upon
etry. It is noteworthy that phosphorus-centred radicals the condensation reaction between the carbonyl compound
ž ž
such as phosphinyls ( PR2 ), phosphonyls [ P(O)R2 ] and [R1 C(O)R2 ] and the alkylhydroxylamine (R3 NHOH)
ž
phosphoranyls [ P(OR)4 ] display a wide range of phos- whether or not this route was used to synthesize the
phorus hyperfine splittings (from around 10 to 100 mT).2
nitrone. Thus, aldo- and keto-nitrones are from aldehydes
The type of ligands (e.g. R groups) attached to the phos-
and ketones, respectively (Scheme 3).
phorus nucleus also profoundly affects the magnitude of
A wide range of nitrone spin traps have been
the phosphorus hyperfine splitting.2 Owing to the huge
31 developed over the years13 – 18 but the two most popular
P hyperfine splittings (e.g. in excess of 90 mT) for
have been ˛-phenyl-N-tert-butylnitrone (PBN) and 5,5-
some phosphoranyl radicals,3 they too have received some
recent attention as potential alternatives to nitroxides in dimethyl-1-pyrroline-N-oxide (DMPO) (their Chemical
magnetometry. Unfortunately, the half-life of the phos- Abstracts names are 2-methyl-N-phenylmethylene-2-
ž
phoranyl radical from 1 [ P(OR)4 ] (Scheme 1) is in the propanamine-N-oxide and 3,4-dihydro-2,2-dimethyl-2H-
order of seconds2 whereas those for some silylated ana- pyrrole-1-oxide, respectively). It is interesting that
logues are only slightly longer lived at 2.7 h.3 Hence these improvements to these spin traps have been made
phosphoranyls are not as yet usable in commercial mag- recently by the incorporation of deuterium, carbon-
netometers. 13 and nitrogen-15. Deuteration generally sharpens
the spin adduct ESR lines, thus enhancing resolution
* Correspondence to: D. Lawrence Haire, Department of Clinical and sensitivity. Carbon-13 provides an extra hyperfine
Studies, Ontario Veterinary College, University of Guelph, Guelph, splitting that helps identify the spin trapped radical.
Ontario, N1G 2W1, Canada.
Contract/grant sponsor: Natural Sciences and Engineering Research Finally, nitrogen-15 (spin D 1/2) substitution provides
Council of Canada. higher sensitivity than nitrogen-14 (spin D 1)

Copyright  1999 John Wiley & Sons, Ltd. CCC 0749–1581/99/110767–07 $17.50
768 D. LAWRENCE HAIRE ET AL.

Scheme 3. The spin trapping reaction for aldo- and

ž
keto-nitrones. FR and RA refer to the reactive free radical
and radical addend, respectively.

are expected to be more persistent than the correspond-

ing adducts from aldo-nitrones because an important
decay route for the spin adducts (disproportionation)13 is
blocked owing to the absence of a ˇ-hydrogen (Scheme 4).
The enhanced persistence, however, may come at a
cost. Specifically, the structural information about the
radical addend via the ˇ-hydrogen hyperfine splitting
(Schemes 2–4) is lost. Although they were not employed
in this study, it is possible to regain some of the radical
addend structural information by carbon-13 labelling of
the keto-nitrone spin trap. Janzen’s group35 recently intro-
duced the nitronyl -13 C-labelled 2-phenyl-5,5-dimethyl-1-
pyrroline-N-oxide (2-Ph-DMPO) and Barasch et al.36 syn-
thesized the nitronyl -13 C-labelled 2-methyl-5,5-dimethyl-
ž 1-pyrroline-N-oxide (2,5,5-M3 PO). Another desirable fea-
Scheme 1. General routes to the [ˇ-31 P] nitroxides. BO is
ture of spin trapping with keto-nitrones (vs aldo-nitrones)
the tert-butoxyl radical from di-tert-butyl peroxyoxalate
(DBPO). BPŁ is photoexcited benzophenone. LDA is lithium
is that oxidation of the nitrones to the acyl nitroxides is
diisopropylamide. prevented (Scheme 5). Incidentally, for the ESR spectra
ž ž
of PBNOX and DMPOX (Scheme 5), see Haire et al.23
and Aurich and Trosken,37 respectively.

Scheme 4. Nitroxide spin adduct decay by disproportion-

Scheme 2. Structures of the [ˇ-31 P] phosphoranyl ation. R1 D H. RA refers to the radical addend.
and-phosphityl nitroxides. RA refers to the radical addend.
It is interesting that spin trapping stereochemistry has
because it yields fewer ESR lines two versus three. become a much studied topic.38 These studies parallel
Isotopically labelled PBNs have been studied by several those on chirality and nitrones in general synthetic organic
investigators and the advantages of deuterated PBN,19 – 22 chemistry, which is a very active research area.39 – 45
carbon-13 PBN23,24 and nitrogen-15 PBN25 have been
amply demonstrated. Deuterated26 – 29 and nitrogen-15-
labelled27,28 DMPO-type spin traps have also appeared in DNP magnetometers
the literature. It is worth noting that a related deuterated
and nitrogen-15 labelled cyclic nitrone (imadazolyl) spin Before switching to the main focus of this paper (ESR
trap has been synthesized recently by Haseloff et al.30 and synthesis), some mention should be made of the
Another approach to improving the spin trapping ability dynamic nuclear polarization (DNP) effect. In essence,
of nitrones is to utilize keto-nitrones.31 – 34 The spin adducts DNP magnetometers act as NMR spectrometers in which

Copyright  1999 John Wiley & Sons, Ltd. Magn. Reson. Chem. 37, 767–773 (1999)
 ESR OF NEW b-PHOSPHORUS NITROXIDES FROM ALDO- AND KETO-NITRONES 769

sintered-glass funnel is definitely contraindicated because it may catal-

yse a violent decomposition of DBPO. These crystals are washed with
6 ml of cold (around 5 ° C) hexane. The combined filtrate is cooled in a
dry-ice–acetone bath for 20 min. The DBPO crystals are collected using
a Buchner funnel. The DBPO crystals are then dissolved in a minimum
quantity (around 83 ml) of room temperature hexane. This solution is
cooled to 25 ° C using a dry-ice–acetone bath. The DBPO crystals
are collected using a Buchner funnel and washed with 6 ml of cold
(around 5 ° C) hexane. Around 4 g (17 mmol) or a 41% yield is obtained.
The product is very sensitive to heat, moisture and shock. Therefore,
it is stored in small quantities (<0.5 g) in sealed plastic (polypropy-
lene) test-tubes in plastic jars (containing calcium sulfate, Drierite) in a
Scheme 5. Oxidative decay of an aldo-nitrone spin trap refrigerator.
to the acyl nitroxide (or nitroxone). R1 D H. The acyl The aldo-nitrones ˛-phenyl-N-tert-butylnitrone (PBN) (11) and 5,5-
ž dimethyl-1-pyrroline-N-oxide (DMPO) (12) (Scheme 6) were purchased
nitroxides from PBN and DMPO are known as PBNOX
ž from Aldrich. The keto-nitrones 2,5,5-trimethyl-1-pyrroline-N-oxide
and DMPOX , respectively. In benzene solution the ESR (2,5,5-M3 PO) (13) and 2-phenyl-5,5-dimethyl-1-pyrroline-N-oxide (2-
ž
spectrum for PBNOX has aN D 0.798 mT whereas for Ph-DMPO) (14) (Scheme 6) were synthesized and purified as
ž N
DMPOX a D 0.645 mT and a 2H D 0.328 mT (see Refs 23 described.32
and 37, respectively).

there is no applied magnetic field (i.e. only that of the earth

or other object of study). DNP magnetometers generally
detect the NMR signal of 1 H solvent nuclei that has
been amplified by the presence of small amounts of
persistent free radicals (usually nitroxides). This is the
well known Abragam Overhauser effect.46 It is noteworthy
that the DNP effect in the ESR spectra of nitroxides
is well established1 and that related non-magnetometric
investigations have recently been reported. Practical47,48
(e.g. O2 and ESR imaging) and theoretical approaches,
including DNP calculations49 for 14 N vs 15 N nitroxides,
cross-relaxation determinations50 and solvent studies51,52 Scheme 6. Structures of the aldo- and keto-nitrone spin
have appeared lately. traps used in this study.

Syntheses of the ˇ-phosphorus nitroxides were accomplished as

EXPERIMENTAL follows. The [ˇ-31 P] nitroxides were formed thermally by the reaction
of 0.1 M nitrone (11–14) (Scheme 6) with 0.1 M phosphorane (1) or
Syntheses 0.1 M phosphite (2), and a few milligrams of di-tert-butyl peroxyoxalate
(DBPO) in 2 ml of benzene. The photochemical route involved
photolysis of a 2 ml solution of 0.1 M nitrone (11–14) (Scheme 6),
The phosphorane 1, known as octamethyl-2,2,3,3,7,7,8,8-tetraoxa- 0.1 M phosphorane (1) or phosphite (2) and 0.05 M benzophenone in
1,4,6,9-phosphaspiro-4,4-nonane, was synthesized from hexamethyl- benzene with a low-pressure 75 W xenon lamp for around 30 min prior
phosphorus triamide (HMPT) (Aldrich) and pinacol (Aldrich) according to ESR analysis. The base route utilized a 2 ml benzene solution of 0.1 M
to a procedure modified from the literature.53 Into a round-bottomed nitrone, 0.1 M phosphorane (1) or 0.1 M phosphite (2) and 0.1 M lithium
flask equipped with a magnetic stirrer and reflux condenser are placed diisopropylamide (LDA).
22.45 g or 25.00 ml (137 mmol) of HMPT and 32.51 g (275 mmol)
of pinacol. To an argon cylinder is connected a mineral oil bubbler
(flowmeter). From this bubbler the argon gas passes through anhydrous
calcium sulfate (indicating Drierite). The tubing is connected next to Spectra
a stainless-steel syringe needle (12 gauge, 12 in long). The needle is
positioned just above the reaction mixture. Gradually the mixture is
heated over 2 h to around 160 ° C. The reaction vessel is gently purged The 1 H, 13 C and 31 P NMR spectra were recorded at
with dry argon gas after 1 h of heating. A strong amine odour is present room temperature using a Bruker instrument at 400,
after around 30 min of heating. The Drierite drying tube turns dark blue 100 and 162 MHz, respectively. All ESR samples were
at this stage. The heating and argon purging are stopped after around
4 h of heating. There is no amine odour at this point and the drying purged with either nitrogen or argon gas for 15 min
tube remains light blue. The 1 H, 13 C and 31 P NMR spectra indicate the prior to ESR analysis. The ESR spectra were recorded
correct product in good purity (>95%) (see Table 1). The crude product in quartz round cells except for the in situ photolysis
melts at around 81 ° C (the literature53 m.p. is 89 ° C). The product may be
purified by recrystallization using light petroleum ether. It is noteworthy experiments, which used quartz flat cells. Spectra were
that this phosphorane (1) decomposes violently when heated over 160 ° C obtained using a Bruker ER 200 D spectrometer operating
at atmospheric pressure. at room temperature at around 9.8 GHz (or at X-band).
The tert-butoxyl radical source di-tert-butyl peroxyoxalate (DBPO)
was also synthesized by a modified literature procedure.54 Into a
round-bottowed flask, equipped with a magnetic stirrer, thermome-
ter, reflux condenser, drying tube and pressure-equalizing funnel, are RESULTS AND DISCUSSION
placed 15.27 ml (84.0 mmol) of 5.5 M tert-butyl hydroperoxide in nonane
(Aldrich), 6.64 g or 6.79 ml (84.0 mmol) of pyridine and 83 ml of hexane.
To this ice-cooled (around 5 ° C) vigorously stirred mixture are added The 1 H, 13 C and 31 P NMR data for phosphorane 1
dropwise 5.33 g or 3.66 ml (42.0 mmol) of oxalyl chloride in 35 ml of are collected in Table 1. It is worth mentioning that
hexane. After this addition, the mixture is allowed to warm slowly to
near room temperature. The pyridinium hydrochloride is then removed we attempted to produce the phosphorus-centred radical
ž
by filtration using a porcelain or plastic Buchner funnel. Caution: a [ P(OR)4 ] (Scheme 1) from 0.2 M phosphorane 1 and 0.2 M

Copyright  1999 John Wiley & Sons, Ltd. Magn. Reson. Chem. 37, 767–773 (1999)
770 D. LAWRENCE HAIRE ET AL.

Table 1. 1 H, 13
C and 31
P data for the phosphorane 1 Table 2. ESR hyperfine splittings for the [ˇ-31 P]phospho-
(Scheme 1)a ranyl and-phosphityl nitroxides 3–10 in benzene at room
temperaturea
1
H NMR: 7.40 [d, 1H, phosphoranyl H,
31
J.1 H, 31 P/ D 798.66 Hz], 1.33 (s, 12 H, methyls), 1.03 (s, 12 H, Generation aN aˇ H aˇ P
methyls) Nitroxide conditionsa (mT) (mT) (mT) Ref.
13
C NMR: 78.55 (s, quaternaries), 24.32 [d, methyls, ž
3 BO 1.440 0.306 2.478 This Work
J.13 C, 31 P/ D 4.54 Hz], 23.89 [d, methyls,
3 BPŁ 1.438 0.300 2.538 This work
J.13 C, 31 P/ D 6.98 Hz] 31
No P adduct LDA — — — This work
31
P NMR: 39.16 [d, J.31 P, 1 H/ D 798.54 Hz); literature:53 3
39.8 [d, J.31 P, 1 H/ D 802 Hz] 4b BO
ž
1.475 0.318 2.475 55
ž
a 1 13
The chemical shifts .υ/ for the H and C spectra are relative to internal 4 BO 1.465 0.306 2.433 23
Ł
TMS and those for 31 P are relative to external H3 PO4 . All spectra were 4 BP 1.450 0.294 2.431 This work
recorded at room temperature in C6 D6 . 4 LDA 1.460 0.300 2.430 This work
31 ž
No P adduct BO — — — This work
5
di-tert-butyl peroxyoxalate (DBPO) in benzene at room 5 BPŁ 1.488 1.700 5.138 This work
temperature. We also tried some in situ photochemical 5 LDA 1.475 1.763 5.163 This work
31 ž
experiments using a low-pressure 75 W xenon lamp. No P adduct BO — — — 56
The above solution (with DBPO), one with 0.2 M 6
benzophenone, and one with 0.2 M di-tert-butyl peroxide 6 BPŁ 1.32 1.69 4.50 56
31
(Aldrich), were all photolysed to no avail. Apparently, to No P adduct LDA — — — This work
6
observe this 31 P-centred radical by ESR, either a higher ž
7 BO 1.388 — 5.166 This work
temperature or a stronger UV lamp (e.g. 1000 W)2 is
7 BPŁ 1.381 — 5.156 This work
required. No 31
P adduct LDA — — — This work
7
ž
8c BO 1.386 — 4.985 57
8 Multi-step 1.375 — 5.000 58, 59
Routes to the b-phosphorus nitroxides route
8 LDA 1.381 — 4.975 This work
ž
Owing to the lack of persistence of this phosphoranyl 9 BO 1.350 — 3.800 This work
Ł
radical, we embarked upon some spin trapping studies 9 BP 1.353 — 3.813 This work
with four nitrones. The two aldo-nitrones were PBN No 31 P adduct LDA — — — This work
9
(11) and DMPO (12) and the keto-nitrones were 2,5,5- ž
10 BO 1.350 — 3.446 This work
M3 PO (13) and 2-Ph-DMPO (14) (Schemes 3 and Ł
10 BP 1.347 — 3.458 This work
6). Additionally, because ˇ-phosphityl nitroxides are 10d BZP 1.367 — 3.459 57
available by the spin trapping of phosphityl radicals 10 LDA 1.383 — 3.463 This work
ž
[ P(O)(OR)2 ] (Schemes 1, 2 and 3) from hydrogen
ž
abstraction by photochemically55 or thermally23 generated a
BO and BPŁ refer to hydrogen abstraction by tert-butoxyl radicals
tert-butoxyl radicals or photoexcited benzophenone,56 and photoexcited benzophenone, respectively. LDA refers to lithium
diisopropylamide. This strong base forms phosphorus anions that add
we reasoned that the hitherto unknown ˇ-phosphoranyl to nitrones, the adducts of which, are oxidized to nitroxides.
b
nitroxides might also be obtainable in these ways. The This nitroxide was formed by hydrogen abstraction of the phosphite
ESR spectral data for phosphoranyl nitroxides 3, 5, 7 and by tert-butoxyl radicals from the photolysis of di-tert-butyl peroxide.
c
This nitroxide was formed by the addition of a methyl radical
9 (Scheme 2) (Table 2) demonstrate the viability of these (from ˇ-cleavage of a tert-butoxyl radical) to 2-(diethylphosphityl)-5,5-
routes. The use of a strong, hindered base (i.e. LDA) to dimethyl-1-pyrroline-N-oxide.
d
This nitroxide was formed by the addition of a phenyl radical (from
generate ˇ-phosphorus nitroxides (e.g. phosphoranyl or benzoyl peroxide) to 2-(diethylphosphityl)-5,5-dimethyl-1-pyrroline-N-
phosphityl), to our knowledge, has not been tried before. oxide.
This approach has, however, met with varied success.
Specifically, only the phosphoranyl adduct of DMPO (5) ESR spectral patterns
(Scheme 2) could be unambiguously detected along with
the phosphityl adducts of PBN (4), 2,5,5-M3 PO (8) and The PBN (3, 4) and DMPO (5, 6) [ˇ-31 P] nitroxide
2-Ph-DMPO (10). The DMPO (12)–phosphite (2)–LDA spin adducts of these aldo-nitrones exhibited similar ESR
mixture yielded a complex, uninterpretable ESR spectrum spectra to the extent that all are 12-line triplet of doublet of
with well over 30 lines. It should be pointed out that a doublet patterns due to the nitroxide nitrogen, ˇ-hydrogen
novel ‘reverse addition’ route to nitroxides 8 and 10 is and ˇ-31 P nuclei (Scheme 2) (Table 2). Naturally, the
included in Table 2. Specifically, nitroxides 8 and 10 were corresponding ESR spectra from the keto-nitrones 2,5,5-
formed by the spin trapping of methyl and phenyl radicals, M3 PO (7, 8) and 2-Ph-DMPO (9, 10) are simpler owing
respectively, by an interesting, new ˇ-phosphityl nitrone.57 to the lack of a ˇ-hydrogen (Scheme 2) (Table 2). These
For completeness, the multi-step route to 8 devised by spin adducts display six-line triplet of doublets patterns
Tordo and co-workers58,59 is also included. due to the nitroxide nitrogen and ˇ-31 P nuclei. For each

Copyright  1999 John Wiley & Sons, Ltd. Magn. Reson. Chem. 37, 767–773 (1999)
 ESR OF NEW b-PHOSPHORUS NITROXIDES FROM ALDO- AND KETO-NITRONES 771

of the four spin traps there are small but significant detected. The photochemical route with benzophenone
changes in the spectra of the respective phosphoranyl seemed to work in all the cases in which it was tried
versus the phosphityl spin adducts (5–10). A notable (although the ESR signals were generally weaker than
exception is PBN, whose spin adducts (3 and 4) show with the other methods). This, however, was not the case
essentially the same hyperfine splittings. The nitrogen and with the strong, hindered base LDA. For three of the
ˇ-hydrogen hyperfine splittings are especially insensitive nitrones (PBN (11), 2,5,5-M3 PO (13) and 2-Ph-DMPO
to the structure of the 31 P moiety for all of the spin (14)] the phosphityl but not the phosphoranyl spin adduct
adducts (3–10) (Scheme 2) (Table 2). The ˇ-31 P coupling was detectable. The situation with DMPO (12) is more
is more sensitive, however. It is consistently larger for the complex. As mentioned previously, a complicated ESR
phosphoranyl than the phosphityl adducts. For the adducts spectrum is observed (over 30 lines). DMPO is a more
of DMPO (5, 6), 2,2,5-M3 PO (7, 8) and 2-Ph-DMPO reactive molecule towards free radicals than PBN (11),60
(9, 10) the ˇ-31 P hyperfine splitting is larger by 0.6, 0.2 2,5,5-M3 PO (13),61 or 2-Ph-DMPO (14).35 Presumably,
and 0.3 mT, respectively. The most striking differences, DMPO is also more reactive to nucleophiles. Thus, the
however, are in the magnitudes of the ˇ-31 P splittings base LDA may add to this nitrone in a nucleophillic
from spin adducts from different spin traps (i.e. 3 and fashion at a faster rate constant. LDA could also remove
4 vs 5 and 6 vs 7 and 8 vs 9 and 10). The largest a hydrogen from DMPO (e.g. from the allylic position)
ˇ-31 P couplings are from the cyclic nitrone spin traps. and this carbanion could then add to DMPO to produce
The ˇ-31 P hyperfine splittings for nitroxides from DMPO DMPO-DMPO-yl.13 These three adducts (phosphorus-,
(12), 2,2,5-M3 PO (13) and 2-Ph-DMPO (14) are in the nitrogen- and carbon-centred) could contribute to the
ranges 4.5–5.1, 5.0–5.2 and 3.4–3.8 mT, respectively. observed ESR complexity.
The smallest ˇ-31 P couplings are seen with the PBN spin
adducts (2.4–2.5 mT).
Magnetometry considerations
ž
Nitroxide persistence It is noteworthy that all three routes (BO , BPŁ and LDA)
(Scheme 1) to the ˇ-phosphorus nitroxides could produce
Although the half-lives of the various spin adducts (3–10) strong ESR signals, although the BPŁ route yielded com-
(Scheme 2) were not measured quantitatively, the expected paratively weaker signals. As found previously, however,1
trend was seen qualitatively. The persistences decreased the percentage yields were generally low (<1%). Never-
ž ž
in the order 2,5,5-M3 PO-RA ³ 2-Ph-DMPO-RA > theless, since only milligram quantities are required for
ž ž
PBN-RA > DMPO-RA . magnetometry applications, these routes are significant.
The aldo-nitrone spin adducts (3–6) were observable In low-field magnetometry a number of factors affect the
by ESR for hours to days whereas the keto-nitrone spin sensitivity. The first consideration is the magnitude of
adducts (7–10) appear to be indefinitely stable (weeks the ESR hyperfine splitting .a/; generally, the larger the
to months). The increased persistence of the keto-nitrone a value the better. Thus, the DMPO (5, 6) and 2,5,5-
spin adducts (7–10) may be due to the blockage of an M3 PO (7, 8) spin adducts, which exhibit the largest ˇ-31 P
important decay route. Disproportionation is not possible hyperfine splittings (around 5 mT), are favoured. The mag-
because these spin adducts do not possess a ˇ-hydrogen netometer sensitivity varies inversely with the number of
(Schemes 2 and 4). Another favourable feature of keto- ESR spectral lines. Therefore, nitroxides 7 and 8 are bet-
nitrone spin trapping is the enhanced stability of the spin ter than nitroxides 5 and 6 because the former exhibit six
trap itself. Specifically, a common decomposition route lines whereas the latter exhibit 12 lines. The lesser persis-
for a nitrone spin trap is oxidation to the acyl nitroxide tence, as mentioned previously, also disfavours nitroxides
(Scheme 5). Owing to the absence of a nitronyl -hydrogen 5 and 6. Thus, of all the ˇ-phosphorus nitroxides exam-
this decay route is also not possible. ined in this study, the one best suited to magnetometry
seems to be the ˇ-phosphoranyl nitroxide 7. It exhibits
31
the largest ˇ-31 P hyperfine splitting .aˇ P D 5.156 mT)
Synthesis successes and failures and seems indefinitely persistent.

The three synthetic routes (Scheme 1) to the ˇ-phosphorus

nitroxides yielded various results. The success or failure CONCLUSION
of a particular route varied widely from spin trap to
spin trap and whether it was the phosphorane 1 or the The precursor for the phosphoranyl nitroxides, the phos-
phosphite 2 that was being reacted (Table 2). The failure phorane 1 was synthesized by a modified literature proce-
of the tert-butoxyl route with DMPO (12) may be due dure and characterized by NMR spectroscopy. Its 1 H, 13 C
to the extremely fast (i.e. diffusion-controlled) addition and 31 P NMR data are collected in Table 1.
of this radical to the spin trap60 versus the hydrogen Three synthetic routes to some ˇ-phosphoranyl (3, 5, 7,
abstraction with the phosphorane 1 or the phosphite 2. 9) and ˇ-phosphityl (4, 6, 8, 10) nitroxides were explored
In fact, only the tert-butoxyl spin adduct of DMPO (Schemes 1 and 2). The ˇ-phosphoranyl nitroxides (3, 5, 7,
(aN D 1.308, aˇ H D 0.798 and a H D 0.198 mT) was 9) represent the first ˇ-phosphorus nitroxides of this class.

Copyright  1999 John Wiley & Sons, Ltd. Magn. Reson. Chem. 37, 767–773 (1999)
772 D. LAWRENCE HAIRE ET AL.

ž 14. For example, see R. Konaka, in Bioradicals Detected by ESR

All three routes (i.e. BO , BPŁ and LDA) (Scheme 1)
Spectroscopy, edited by H. Ohya-Nishiguchi and L. Packer, p. 143.
were viable routes to most of these ˇ-31 P nitroxides MCBU Press, Boston, MA (1995).
(see Table 2 for exceptions). The [ˇ-31 P] nitroxides were 15. For example, see N. Sankurati and E. G. Janzen, Tetrahedron Lett.
formed from the aldo-nitrones PBN (11) and DMPO (12) 37, 5313 (1996).
16. For example, see R. Sato, K. Ito, H. Igarashi, M. Uejima, K. Naka-
and the keto-nitrones 2,5,5-M3 PO (13) and 2-Ph-DMPO hashi, J. Morioka and M. Takeishi, Chem. Lett. 1059 (1997).
(14). The ESR spectral data for all the ˇ-phosphorus 17. For example, see C. P. Sar, E. Hideg, I. Vass and K. Hideg, Bioorg.
nitroxides in this study are collected in Table 2. PBN Med. Chem. Lett. 8, 379 (1998).
18. For example, see D. A. Becker, in Organic Synthesis: Theory and
nitroxides (3, 4) and DMPO nitroxides (5, 6) exhibited Applications, edited by T. Hudlicky, Vol. 4, p. 155. JAI Press,
12-line spectra whereas 2,5,5-M3 PO nitroxides (7, 8) London (1998).
and 2-Ph-DMPO nitroxides (9, 10) showed six lines. 19. E. G. Janzen, U. M. Oehler, D. L. Haire and Y. Kotake, J. Am.
Chem. Soc. 108, 6858 (1986).
Phosphoranyl nitroxides could be easily distinguished 20. D. L. Haire and E. G. Janzen, Magn. Reson. Chem. 32, 151 (1994).
from phosphityl nitroxides of the respective spin traps 21. G. S. Timmins, G. K. Barlow, J. A. Silvester, X. Wei and
by significant differences (0.2–0.6 mT) in their ˇ-31 P A. C. Whitwood, Redox Rep., 3, 125 (1997).
22. P. G. Mekarbane and B. J. Tabner, Magn. Reson. Chem. 36, 826
couplings. An exception to this rule is the PBN spin (1998).
adducts (3, 4), whose spectra were identical. The cyclic 23. D. L. Haire, U. M. Oehler, P. H. Krygsman and E. G. Janzen, J.
nitrone spin adducts (5, 6, 7, 8) displayed the largest ˇ-31 P Org. Chem. 53, 4535 (1988).
24. A. G. Motten, L. A. Levy and R. E. London, J. Magn. Reson. 80,
hyperfine splittings (around 5 mT). The phosphoranyl spin 112 (1988).
adduct (7) from 2,5,5-M3 PO exhibited the largest ˇ-31 P 25. Y.-K. Zhang, Z. Naturforsch., Teil B 51, 139 (1996).
coupling (5.156 mT). For this reason and its persistence, 26. D. L. Haire, U. M. Oehler, H. D. Goldman, R. L. Dudley and
E. G. Janzen, Can. J. Chem. 66, 2395 (1988).
this nitroxide may be the best candidate, of the nitroxides 27. J. F. W. Keana, L. Lex, J. S. Mann, J. M. May, J. H. Park, S. Pou,
in this study, for use in magnetometry. V. S. Prabhu, G. M. Rosen, B. J. Sweetman and Y. Wu, Pure Appl.
Chem. 62, 201 (1990).
28. S. Pou, G. M. Rosen, Y. Wu and J. F. W. Keana, J. Org. Chem.
55, 4438 (1990).
Acknowledgements 29. H. J. Halpern, S. Pou, M. Peric, C. Yu, E. Barth and G. M. Rosen,
J. Am. Chem. Soc. 115, 218 (1993).
30. R. F. Haseloff, I. A. Kirilyuk, S. I. Dikalov, V. V. Khramtsov,
This work is supported by the Natural Sciences and Engineering D. I. Utepbergenov, I. E. Blasig and I. A. Grigorev, Free Rad. Res.
Research Council of Canada (NSERC). Grateful acknowledgement is 26, 159 (1997).
hereby made. D. L. H. thanks the Natural Sciences and Engineer- 31. E. G. Janzen and Y.-K. Zhang, J. Magn. Reson. B 101, 91 (1993),
ing Research Council of Canada for an NSERC industrial research and references cited therein.
fellowship. 32. E. G. Janzen, Y.-K. Zhang and D. L. Haire, Magn. Reson. Chem.
32, 711 (1994).
33. E. G. Janzen, Y.-K. Zhang, D. L. Haire, Y. Kotake and M. Payne,
Appl. Magn. Reson. 6, 511 (1994).
REFERENCES 34. D. L. Haire, E. G. Janzen, V. J. Robinson and Y.-K. Zhang, Magn.
Reson. Chem. 33, 796 (1995).
35. E. G. Janzen, Y.-K. Zhang and D. L. Haire, J. Am. Chem. Soc. 116,
1. For some of our previous work on ˇ-phosphorus nitroxides for mag- 3738 (1994).
netometry, see D. L. Haire, E. G. Janzen, G. Chen, V. J. Robinson 36. D. Barasch, M. C. Krishna, A. Russo, J. Katzhendler and A. Samuni,
and I. Hrvoic, Magn. Reson. Chem. 37, 251 (1999), and references J. Am. Chem. Soc. 116, 7319 (1994).
cited therein. 37. H. G. Aurich and J. Trosken, Chem. Ber. 106, 3483 (1973).
2. P. Tordo, in Landolt-Bornstein, New Series Group II, Magnetic 38. J. C. Matasyoh, U. Abram, P. Schuler and H. B. Stegmann, Magn.
Properties of Free Radicals, Phosphorus-centered Organic Radi- Reson. Chem. 36, 422 (1998), and references cited therein.
cals and Related Radical Ions, edited by Hans Fischer, Vol. 17e, 39. M. Closa and R. H. Wightman, Synth. Commun. 28, 3443 (1998).
p. 254. Springer-Verlag, Berlin (1988), and previous compilations 40. S. Kobayashi and M. Kawamura, J. Am. Chem. Soc. 120, 5840
and references cited therein. (1998).
3. S. Marque, Y. Berchadsky, P. Bertrand, A. Fournel, P. Tordo, 41. P. Merino, S. Franco, N. Garces, F. L. Merchan and T. Tejero, J.
K. Lang, M. Moussavi and E. Belorizky, J. Phys. Chem. A 101, Chem. Soc., Chem. Commun. 493 (1998).
5640 (1997). 42. S.-I. Murahashi, H. Ohtake and Y. Imada, Tetrahedron Lett. 39,
4. C. Rizzi, R. Lauricella, B. Tuccio, J.-C. Bouteiller, V. Cerri and 2765 (1998).
P. Tordo, J. Chem. Soc., Perkin Trans. 2 2507 (1997). 43. H. G. Aurich, M. Soeberdt and K. Harms, Tetrahedron 55, 1249
5. C. Rizzi, S. Marque, F. Belin, J.-C. Bouteiller, R. Lauricella, (1999).
B. Tuccio, V. Cerri and P. Tordo, J. Chem. Soc., Perkin Trans. 44. U. Chiacchio, A. Rescifina, D. Iannazzo and G. Romeo, J. Org.
2 2513 (1997). Chem. 64, 28 (1999).
6. A. Rockenbauer, A. Mercier, F. LeMoigne, G. Olive and P. Tordo, 45. M. Noguchi, H. Okada, S. Nishimura, Y. Yamagata, S. Takamura,
J. Phys. Chem. A 101, 7965 (1997). M. Tanaka, A. Kakehi and H. Yamamoto, J. Chem. Soc. Perkin
7. J.-L. Clement, B. C. Gilbert, W. F. Ho, N. D. Jackson, M. S. New- Trans. 1 185 (1999).
ton, S. Silvester, G. S. Timmins, P. Tordo and A. C. Whitwood, J. 46. W. Muller-Warmuth and K. Meise-Gresch, Adv. Magn. Reson. 11,
Chem. Soc., Perkin Trans. 2 1715 (1998). 1 (1983), and references cited therein.
8. G. Olive, F. LeMoigne, A. Mercier, A. Rockenbauer and P. Tordo, 47. T. Guiberteau and D. Grucker, J. Magn. Reson. 124, 263 (1997).
J. Org. Chem. 63, 9095 (1998). 48. T. Guiberteau and D. Grucker, Phys. Med. Biol. 43, 1887 (1998).
9. S. Pietry, T. Liebgott, C. Frejaville, P. Tordo and M. Culcasi, Eur. 49. T. Guiberteau and D. Grucker, J. Magn. Reson. B 110, 47 (1996).
J. Biochem. 254, 256 (1998). 50. B. M. Odintsov, R. L. Belford and R. B. Clarkson, J. Phys. Chem.
10. C. Mathieu, A. Mercier, D. Witt, L. Dembkowski and P. Tordo, A 101, 116 (1997).
Free Rad. Biol. Chem. 22, 803 (1997). 51. N. Horasan, M. M. Sunnetcioglu, R. Sungur and G. Bingol, Z.
11. C. Mathieu, B. Tuccio, R. Lauricella, A. Mercier and P. Tordo, J. Naturforsch., Teil A 52, 485 (1997).
Chem. Soc., Perkin Trans. 2 2501 (1997). 52. A. O. Salman, M. M. Sunnetcioglu, R. Sungur and G. Bingol, J.
12. C. Chachaty, C. Mathieu, A. Mercier and P. Tordo, Magn. Reson. Magn. Reson. 134, 1 (1998).
Chem. 36, 46 (1998). 53. M. Sanchez, R. Wolf, R. Burgada and F. Mathis, Bull. Soc. Chim.
13. E. G. Janzen and D. L. Haire, in Advances in Free Radical Chem- Fr. 773 (1968).
istry, edited by D. D. Tanner, Vol. 1, p. 253. JAI Press, Greenwich, 54. P. D. Bartlett, E. P. Benzing and R. E. Pincock, J. Am. Chem. Soc.
CT (1990), and references cited therein. 82, 1762 (1960).

Copyright  1999 John Wiley & Sons, Ltd. Magn. Reson. Chem. 37, 767–773 (1999)
 ESR OF NEW b-PHOSPHORUS NITROXIDES FROM ALDO- AND KETO-NITRONES 773

55. G. Brunton, B. C. Gilbert and R. J. Mawby, J. Chem. Soc., Perkin 59. F. LeMoigne, A. Mercier and P. Tordo, Tetrahedron Lett. 32, 3841
Trans. 2 650 (1976). (1991).
56. P. Tordo, M. Boyer, A. Friedmann, O. Santero and L. Pujol, J. Phys. 60. D. L. Haire and E. G. Janzen, Can. J. Chem. 60, 1514 (1982).
Chem. 82, 1742 (1978). 61. E. Finkelstein, G. M. Rosen and E. J. Rauckman, J. Am. Chem. Soc.
57. E. G. Janzen and Y.-K. Zhang, J. Org. Chem. 60, 5441 (1995). 102, 4994 (1980).
58. A. Mercier, Y. Berchadsky, Badrudin, S. Pietry and P. Tordo,
Tetrahedron Lett. 32, 2125 (1991).

Copyright  1999 John Wiley & Sons, Ltd. Magn. Reson. Chem. 37, 767–773 (1999)