Electrophilic C(2)-functionalization of nitronyl nitroxides: A reference to N-heterocyclic carbenes

Article abstract of DOI:10.1016/S0022-328X(00)00668-9

In an attempt to generalize the recently introduced concept of 'auto-umpolung', nitronyl nitroxides are identified as precursors of a special class of electron-excess carbenes. The latter can be formally derived from nucleophilic carbenes of the Wanzlick-Arduengo type by replacing redox-inactive N-substituents by lone-pair donor functions. A first systematic access to this class of compounds is achieved by lithiating the nitronyl nitroxide 11a at C(2), the central carbon position. The resulting radical anion equivalent 13 can be trapped in solution by various types of electrophiles to give C(2)-substituted nitronyl nitroxides 14, much in the same way as are trapped nucleophilic carbenes of the Wanzlick-Arduengo type. Carboxylation, addition to an aldehyde as well as silylation at the C(2)-position lead to novel types of functional nitronyl nitroxides. Elemental sulfur and selenium yield new types of anionic nitronyl nitroxides. The first characterized C(2)-derived heavy-metal complex of a nitronyl nitroxide 14g resulted, when Hg(OAc)₂ was used as a trapping agent. More conveniently this complex could also be obtained directly from 11a and Hg(OAc)₂. The X-ray structure of 14g is presented and analyzed in terms of cluster-forming secondary Hg-O interactions.

Full text of DOI:10.1016/S0022-328X(00)00668-9

www.elsevier.nl/locate/jorganchem
Journal of Organometallic Chemistry 617–618 (2001) 473–482
Electrophilic C(2)-functionalization of nitronyl nitroxides:
a reference to N-heterocyclic carbenes
Robert Weiss *, Norbert Kraut, Frank Hampel 1
Institut fu¨r Organische Chemie der Uni6ersita¨t Erlangen-Nu¨rnberg, Henkestraße 42, D-91054 Erlangen, Germany
Received 2 August 2000; accepted 18 September 2000
Abstract
In an attempt to generalize the recently introduced concept of ‘auto-umpolung’, nitronyl nitroxides are identified as precursors
of a special class of electron-excess carbenes. The latter can be formally derived from nucleophilic carbenes of the Wanzlick–
Arduengo type by replacing redox-inactive N-substituents by lone-pair donor functions. A first systematic access to this class of
compounds is achieved by lithiating the nitronyl nitroxide 11a at C(2), the central carbon position. The resulting radical anion
equivalent 13 can be trapped in solution by various types of electrophiles to give C(2)-substituted nitronyl nitroxides 14, much in
the same way as are trapped nucleophilic carbenes of the Wanzlick–Arduengo type. Carboxylation, addition to an aldehyde as
well as silylation at the C(2)-position lead to novel types of functional nitronyl nitroxides. Elemental sulfur and selenium yield new
types of anionic nitronyl nitroxides. The first characterized C(2)-derived heavy-metal complex of a nitronyl nitroxide 14g resulted,
when Hg(OAc)2 was used as a trapping agent. More conveniently this complex could also be obtained directly from 11a and
Hg(OAc)2. The X-ray structure of 14g is presented and analyzed in terms of cluster-forming secondary HgꢀO interactions. © 2001
Elsevier Science B.V. All rights reserved.
Keywords: Electrophiles; Nitronyl nitroxides; Heterocyclic carbenes
1. Introduction
flexible unit, as evidenced by many synthesized systems
of type 2 and 3, as well as by accompanying calculations
[1–3]. We have termed the underlying electronic principle
‘auto-umpolung’ [1] (‘Selbstumpolung’ in German). This
particular quality is lacking in conventional Arduengo
carbenes [4].
Obviously it would be highly desirable to generalize the
concept of ‘auto-umpolung’ for carbenes beyond the
singularity 1. A formal approach to achieve this is shown
in Scheme 2. Introducing lone-pair donor groups as
N-substituents in Arduengo carbenes would lead to
nucleophilic carbenes of type 4.
Two directly connected lone-pair functions frequently
form the basis of two-step redox systems [5]. In 4 two
such systems are directly conjugated with the carbene
center. Depending on the electronic nature of the N-at-
tached donor group this will lead to a much higher
population of the carbene’s 2pz-orbital as compared to
that of normal Arduengo carbenes. This is indicated in
Scheme 3 for the singlet form 4 (S). A corresponding
Recently we have synthesized a novel type of nucleo-
philic carbene, in which the carbene center is conjuga-
tively coupled to a 2,2%-bipyridine-based two-step redox
system, as in 1 (Scheme 1) [1,2]. This system can be
generated from its protonated precursor at −80°C in
solution and spectroscopically characterized. It forms
stable adducts with many electrophilic agents [1–3].
Depending on the electronic nature of the electrophile
two types of adducts can be discerned:
“ When attached to a lone pair donor (+M substituent)
the former carbene center functions as an ꢀM accep-
tor, as in 2.
“ When connected with an ꢀM acceptor the former
carbene center acts as an +M donor, as in 3.
Thus, due to its integrated two-step redox system
carbene 1 can function as an electronically extremely
* Corresponding author. Fax: +49-91331-8525876.
E-mail address: rweiss@organik.uni-erlangen.de (R. Weiss).
1 X-ray analysis.
triplet form
4 (T) — interestingly a p-biradical
— would also have to be considered. For both
0022-328X/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved.
PII: S0022-328X(00)00668-9

474
R. Weiss et al. / Journal of Organometallic Chemistry 617–618 (2001) 473–482
poses hydroxylamino- and hydrazino-derived carbenes
4 would be the obvious choice (ꢀD=ꢀO−, ꢀOH,
ꢀOR, ꢀNR2, etc.). Scanning the literature we found
several feasible precursors for such compounds [6]. In
this publication we approach this topic by looking at
the situation for the simplest and simultaneously most
powerful donor group ꢀD=ꢀO−.
It turns out that nucleophilic carbenes of type 5 are
completely unknown — not surprisingly in view of
their excessive electron-surplus character. Functional-
ization of 5 at the carbene center by an electrophile
would lead to the reduced form (RED) 6 of a two-
step redox system. The latter still possesses distinct
electron-surplus character and the few representatives
of this structural type are only known in protonated
form [7h,11n]. However, monooxidation of 6 would
lead to an electronically well-balanced neutral radical
7 (SEM), many stable derivates of which are known
under the common term nitronyl nitroxides [7]. Sur-
prisingly, only one representative of the corresponding
OX-form 8 (nitronyl nitrosonium salts) has been syn-
thesized and structurally characterized [8].
As can be inferred from resonance structures (a)
and (b), the electron density at C(2) in 7 and 8 can
still be flexibly modulated as a function of the sub-
stituent type E present at this position. This propen-
sity is due to the electron reservoir present at the
heteroatom functions and the structural consequences
would be analogous to those discussed in Scheme 1.
To exploit this potential it is of obvious interest to
functionalize C(2) of these systems in quite general
ways. This could probably be best achieved via gener-
ation of the underlying nucleophilic ‘carbenes’, fol-
lowed by electrophilic derivatization. Subsequently we
report on our endeavors to achieve this for nitronyl
nitroxides. In a subsequent report [9] we shall deal
with the corresponding chemistry of nitronyl nitroso-
nium salts 8.
Scheme 1.
2. Nitronyl nitroxides: synthesis and significance
The first nitronyl nitroxides have been synthesized
by Ullman in 1969 [10] and they now represent the
most thoroughly investigated class of stable organic
radicals known. Unlike other stable radicals they can
act as mono- or bifunctional ligands in metal coordi-
nation chemistry via their Lewis basic oxygen centers
[11]. Due to this, they have recently received a lot of
attention in supramolecular magneto-structural inves-
tigations [12]. In spite of the ever-growing importance
of nitronyl nitroxides the only systematic synthesis
leading to them is still the one originally introduced
by Ullman (Eq. (1)):
Scheme 2.
electronic configurations coordination of 4 to elec-
tron-demanding centers would massively umpole the
former carbene center to an electron-surplus center,
much beyond the stage of ‘auto-umpolung’ already
present in the free carbene 4.
Can such systems be realized? For all practical pur-

R. Weiss et al. / Journal of Organometallic Chemistry 617–618 (2001) 473–482
475
Scheme 3.
3. Experimental results
3.1. Deprotonation of a nitronyl nitroxide at C(2)
The starting point of further investigations was the
nitronyl nitroxide 11a (R=CH3, R%=H), which is
readily available according to the Ullman procedure (Eq.
(1)). PM3 model calculations [14] on this radical show
that its C(2) exhibits an unusually high degree of s-char-
acter (sp1.34) in its bond to the ring proton. This is
reminiscent of the bonding situation in the protonated
precursor of 1 [1] and is a direct consequence both of ring
strain in a planar five-membered ring and of the strong
s-withdrawing effect of the N-functional groups at-
tached to C(2) (cf. ‘Bent’s rules’) [15]. As a consequence,
the calculated charge on the proton (+0.20) is compara-
ble to the one found for typical precursors of Arduengo
carbenes [16], although the latter are cationic. This points
to a considerable CH-acidity of 11a, for which also exists
experimental support from the pioneering investigations
of Ullman. The latter had already shown that 11a
undergoes H/D-exchange at C(2) in alkaline medium and
can be methoxy-methylated according to Eq. (2):
(1)
This synthesis has two principal limitations:
“ The carbon centers of the two-carbon bridge have to
be tertiary. This limitation arises from the synthesis
of 8 [13]. For all known nitronyl nitroxides RꢁCH3.
“ As the condensation/oxidation sequence leading to
11 is limited to aldehydes 9, nitronyl nitroxides with
inorganic, elementorganic and metalorganic sub-
stituents at C(2) are virtually unknown.
Although the first limitation certainly invites further
study, the second one is much more serious from a
general chemical point of view. Subsequently we turn to
the question, how the latter limitation could be
removed.
Thus, the C(2)-functionalized nitronyl nitroxide 13
was obtained, albeit in low yield (24%). Subsequently, no
further C(2)-functionalizations of 11a using this method
were reported. This is not too surprising, because both

476
R. Weiss et al. / Journal of Organometallic Chemistry 617–618 (2001) 473–482
this medium and might still co-exist with 11a in an
equilibrium. Thirdly, it is notoriously difficult to isolate
neutral products from DMSO solution and this will be
particularly true for a radical of limited stability like 13.
Having recognized this, we felt that lithiation of 11a in
THF would solve the above-mentioned problems of
electrophilic C(2)-derivatization.
First, lithiation attempts using n-BuLi in THF failed
completely. The characteristic red–blue color of the
starting radical faded rapidly on addition of the base at
−78°C. This points to an irreversible destruction of the
nitronyl nitroxide, possibly by reduction (O-transfer to
n-BuLi?). Subsequent trapping attempts using various
electrophiles gave no identifiable products. Success came
then by replacing n-BuLi by the sterically hindered strong
base Liꢀdiisopropylamide, LDA. When this base was
added to a solution of 11a in THF at −78°C a color
change from red–blue to deep-red was observed. Subse-
quent addition of electrophiles gave the desired
C(2)functionalized nitronyl nitroxides in satisfactory to
good yields (45–70%), according to Eq. (3):
(2)
base and medium used by Ullman were unfortunate
choices in view of the extremely electron-rich character
of the deprotonated species — be it monomeric 12,
containing a ‘naked’ radical anion as assumed by Ullman
(less likely), or a more highly clustered form thereof (much
more likely). This is because DMSO possesses oxidizing
character and may enter into redox reactions with 12
before an electrophile can attack at the desired position.
Secondly, DMSO is also a nucleophile and it may compete
for the electrophile employed. This is particularly true for
the anion of DMSO, which is generated with KOtBu in
(3)
Scheme 4. Trapping of 13 by various electrophiles (a=MeOTf, b=p-tolylaldehyde, c=CO2, d=TMSOTf, e=sulfur for 14e and selenium for
14f, f=Hg(OAc)2).

R. Weiss et al. / Journal of Organometallic Chemistry 617–618 (2001) 473–482
477
In view of the general tendency of organylꢀlithium
compounds to form oligomeric clusters and taking into
account results to be presented below one can safely
assume that 13 exists as an n-mer (n=4?, cf. below), due
to stabilizing intermolecular LiꢀO contacts. This inter-
mediate is sufficiently stable to be kept unchanged at low
temperature for prolonged periods.
nitroxides already synthesized by Ullman via a different
route [7a,b].
3.2.2. C(2)-functionalization of 13 by (deri6ates of )
hea6ier main group elements
Obviously, C(2)-silylated nitronyl nitroxides are of
particular interest, since they could function as versatile
precursors for various C(2)-derivatizations of nitronyl
nitroxides. Due to the above mentioned limitations of the
Ullman synthesis, this class of nitronyl nitroxides has
remained inaccessible up to date. In view of the pro-
nounced oxophilicity of silicon the site of electrophilic
attack on 13 by a silylating agent is not a priori clear.
Model calculations (PM3) predicted a 31.6 kcal mol−1
preference of C(2)-attack over O-attack for trimethylsi-
lylation of 13. This was confirmed by experiment. Under
the reaction conditions specified above one single reac-
tion product is obtained using trimethylsilyl triflate as an
electrophile. After chromatographic purification a red–
blue radical is obtained in 65% yield which analyzes
correctly for 14d (Scheme 4). This structure is confirmed
by UV, IR and MS data (cf. Section 5). C(2)-silylation
follows from the ESR spectrum, which proves coupling
of the spin to two equivalent N centers (aN=7.28 G).
Using tributyl tin triflate, an analogous tin compound
was synthesized. The latter nitronyl nitroxide however
proved extremely sensitive towards hydrolysis and could
only be obtained in an impure form.
Finally, the lithiated nitronyl nitroxide 13 could also
be trapped by elemental sulfur and selenium. Thus,
radical salts 14e and 14f were obtained in reasonable
yields as blue and green crystals, respectively. Although
chromatographic purification was not possible due to the
salt-like character of these products, the seleno derivate
14f was obtained in pure form, in contrast to the sulfur
analogue, which still contained some inorganic impurity.
Structural confirmation for both compounds can be
derived from the ESR spectra, which exhibit quintets in
an intensity ratio of 1:3:5:3:1, as expected for two
equivalent N centers (with aN=7.96 G for 14e and
aN=8.20 G for 14f). The sulfur- and selenium-contain-
ing radical anions in these salts hold great promise as
novel radical ligands in coordination chemistry.
3.2. Trapping of 13 by electrophiles
The C(2)-lithiated nitronyl nitroxide was successfully
trapped by various types of electrophiles. Scheme 4 gives
an overview of functionalized nitronyl nitroxides 14 thus
obtained. With one exception all of these systems are
novel.
3.2.1. C(2)-functionalization of 13 by carbon electrophiles
In a first test run 13 was methylated by methyl triflate.
After chromatographic workup the C(2)-methylated
nitronyl nitroxide 14a was obtained in analytically pure
form. This compound was shown to be identical in all
respects (analytical data, EIMS, UV, ESR, cf. Section 5)
with an authentic sample, synthesized by Ullman accord-
ing to Eq. (1) (R%=CH3) [7h]. After this encouraging
result other electrophiles were used to trap 13. Under the
same reaction conditions the first addition of 13 to a
carbonyl group was achieved, using p-tolylaldehyde as
an electrophile. After chromatographic work-up the
carbinol derivate 14b was obtained in pure form and
good yield (43%) as a dark red radical. The UV spectrum
was characteristic of a nitronyl nitroxide (absorption at
559, 530, 323 nm). Both EI-MS and FAB-MS exhibited
the mole peak and characteristic degradation products
(cf. Section 5).
Carboxylation of the lithiated nitronyl nitroxide also
proved possible. Due to its salt-like structure the blue–vi-
olet radical 14c, could not be chromatographically
purified and the analytical data are therefore not yet
entirely satisfactory. That carboxylation has indeed
taken place can be seen from the IR (a new intense
absorption at 1668 cm−1, in addition to characteristic
nitronyl nitroxide absorptions at 1446, 1372, 1171, 1141
cm−1) and from the FAB mass spectrum, which exhibits
a prominent MLi+-peak. Furthermore, the carboxylate
radical 14c could be iododecarboxylated by iodine and
methylated by methyl triflate, as shown in Eq. (4):
3.2.3. Mercuration
Metal complexes of nitronyl nitroxides are of particu-
lar current interest in context with magneto-structural
investigations [11]. A fair number of nitronyl nitroxide
metal complexes is known, all of which have in common
nitronyl nitroxide ligands bonded via the oxygen centers
to the metal [11a–f]. By contrast, the corresponding
C(2)-derived metalorganic chemistry is non-existent. The
availability of 13 opens up a chance to develop this new
research field.
(4)
In view of the known stability of Hg(II)-carbene
complexes [17], C(2) mercuration of 13 appeared feasible.
Both 14h and 14i proved to be identical with nitronyl

478
R. Weiss et al. / Journal of Organometallic Chemistry 617–618 (2001) 473–482
WhenHg(OAc)2 wasreactedwith1equivalentof13under
the conditions shown in Scheme 4, the Hg(II)-nitronyl
nitroxide complex 14g was produced. According to MS
analysis it is accompanied by the corresponding bis-
(nitronyl nitroxide)Hg(II) complex (the latter compound
will be dealt with in a separate publication) [9]. This
mixture proved difficult to separate. However, subse-
quently we found that the monosubstitution product 14g
could be prepared much more conveniently and quite
selectively simply by reacting 11a with 1 equivalent
Hg(OAc)2 in acetonitrile at room temperature.
Thus the first stable C(2)-metallated derivate of a
nitronyl nitroxide was obtained in pure form and 77%
yield [18]. 14g was fully characterized spectroscopically.
Both EI and FAB mass spectra exhibited a prominent
mole peak at m/z=417, accompanied by the character-
istic isotope pattern of mercury. The latter element also
has a characteristic influence on the corresponding ESR
spectrum [19], which shows a somewhat more complex
splitting pattern than the nitronyl nitroxides discussed
above. This spectrum is shown in Fig. 1 and can be
interpreted as follows.
Fig. 1. ESR spectrum of 14g.
The mercury isotopes 196Hg, 198Hg, 200Hg, 202Hg, 204Hg
with a total natural abundance of 69.8% and nuclear spin
I=0 are responsible for the five-line pattern characteristic
of nitronyl nitroxides with equivalent nitrogen centers
(herewitha(N)=7.77Gandg=2.01555). Superimposed
on this dominant feature is the splitting pattern of the
corresponding 199Hg compound. The 199Hg isotope pos-
sessesI=1/2andthereforegeneratesadoubletofquintets
(with a(199Hg)=52 G). Due to the low natural abundance
of this isotope (17%) the resulting ten line spectrum is
much less intense than the aforementioned five-line
Fig. 2. X-ray structure of 14g: monomeric unit (hydrogens omitted).
Fig. 4. X-ray structure of 14g: view down the stacking axis of the
Fig. 3. X-ray structure of 14g: tetrameric unit.
tetrameric unit.

R. Weiss et al. / Journal of Organometallic Chemistry 617–618 (2001) 473–482
479
Table 1
Selected bond distances (A) and angles (°) of compound 14g
four monomeric units via a boat-like cyclic array of
,
,
symmetry-equivalent Hgꢀ(N)O contacts, each of 2.597 A
in length. A bond length comparison shows that this
intermolecular HgꢀO coordination is definitely stronger
than the above mentioned intramolecular one. These
tetrameric subunits are stacked one on top of the other
without any recognizable interaction between them.
Looking down the stacking axis a tube-like arrangement
of the tetramers becomes apparent, whose inner diameter
Hg1ꢀC1
2.023(13) N5ꢀC1ꢀHg1
2.041(10) 02ꢀN2ꢀC1
02ꢀN3ꢀC3
2.597(10) C1ꢀN2ꢀC3
125.2(9)
124.8(10)
121.8(10)
112.9(10)
127.3(7)
119.3(6)
Hg1ꢀO100
Hg1ꢀO2c1
Hg1ꢀO2c2
C1ꢀN2
C1ꢀN5
N2ꢀO2
N2ꢀC3
C3ꢀC32
C3ꢀC31
C3ꢀC4
C4ꢀN5
C4ꢀC41
C4ꢀC42
O5ꢀN5
2.597(8)
1.349(16) N2ꢀO2ꢀHg1c1
1.369(16) N2ꢀO2ꢀHg1c3
1.267(13) Hgc1ꢀO2ꢀHg1c3 112.4(3)
1.497(15) N2ꢀC3ꢀC32
1.535(19) N2ꢀC3ꢀC31
1.535(19) C32ꢀ0ꢀ01
1.559(19) N2ꢀC3ꢀC4
1.500(14) C32ꢀC3ꢀC4
106.3(10)
109.2(11)
111.4(12)
100.4(9)
114.8(11)
113.8(11)
106.5(11)
110.0(11)
110.6(13)
100.9(9)
115.2(12)
113.7(13)
123.7(10)
123.0(10)
112.2(10)
122.2(12)
121.3(15)
116.5(16)
110.8(9)
,
approximates 3.7 A (Fig. 4). The final crystal structure
of 14g is based on a parallel arrangement of identical
tetramer tubes, in which any individual tube has four
equidistant nearest neighbors. The space between these
buildingblocksisfilledbythemethylgroupsoftheacetato
ligands.
1.51(2)
1.55(2)
C31ꢀC3ꢀC4
N5ꢀC4ꢀC41
1.273(13) N5ꢀC4ꢀC42
1.21(2) C41ꢀC4ꢀC42
1.298(19) N5ꢀC4ꢀC3
C100ꢀO101
C100ꢀO100
C100ꢀC101
C1ꢀHg1ꢀO100
C1ꢀHg1ꢀO2c1
O100ꢀHg1ꢀO2c1
C1ꢀHg1ꢀO(2)c2
O100ꢀHg1ꢀO2c2
The solid state structure of the first stable C(2)-metal-
lated nitronyl nitroxide 14g contains a novel, highly
ordered arrangement of molecular magnets. Investigation
of its collective magnetic properties is clearly indicated.
1.49(2)
169.7(5)
104.5(4)
85.4(4)
102.1(4)
81.6(3)
C41ꢀC4ꢀC3
C42ꢀC4ꢀC3
O5ꢀN5ꢀC1
O5ꢀN5ꢀC4
C1ꢀN5ꢀC4
O101ꢀC100ꢀO100
O101ꢀC100ꢀO101
4. Perspectives
O2c1ꢀHg1ꢀO2c2 82.3(3)
N2ꢀC1ꢀN5
107.1(10) O100ꢀC100ꢀC101
127.6(9) C100ꢀO100ꢀHg1
Our investigations have made a radical anion equiva-
lent of a nitronyl nitroxide readily accessible. Via the
reactive intermediate 13 it will be possible to attach many
more electrophiles — organic, inorganic and metalor-
ganic — to the central carbon of the nitronyl nitroxide.
Extensive research along these lines is being pursued in
our laboratory. It is safe to predict that intermediates of
type 13 will adopt a similar key position for structural
variations of nitronyl nitroxides as do Wanzlick–
Arduengo carbenes for the corresponding development
of imidazolium chemistry [4].
However, there is a chemical dimension characteristic
of nitronyl nitroxides that is not available to imidazolium.
systems. As can be seen from Scheme 3 nitronyl nitroxides
represent the SEM forms of a two-step redox system, and
depending on the electrophile attached to the central
carbon atom the corresponding RED and/or OX forms
may also be accessible. Finally, as also indicated in
Scheme 3 and discussed in the introduction of this
publication, all three members of this two-step redox
system have the structural prerequisites to induce an
‘auto-umpolung’ at the central carbon position, should
this be required by the nature of substituent E. This view
is fully confirmed by model calculations. The interplay
of all these factors will open up a new research field.
N2ꢀC1ꢀHg1
spectrum. Eight out of these ten lines can be seen in Fig.
1, fouronboththeleft-andright-handsidesofthequintet,
which conceals two lines of the 199Hg-induced splitting
pattern.
In addition we succeeded in preparing crystals of 14g
suitable for X-ray analysis. Structural details are shown
in Figs. 2–4 [20], characteristic bond lengths and angles
are summarized in Table 1.
Fig. 2 shows the molecular structure of the C(2)-metal-
lated nitronyl nitroxide. Bond lengths and angles do not
differ significantly from the ones found for other nitronyl
nitroxides. The two ligands form an angle of 169.7° with
the Hg(II) center. This deviation from linearity is due to
secondary interactions of the metal with several other
Lewis basic oxygen centers in the vicinity. One such
interaction takes place intramolecularly between the
carbonylꢀoxygen of the OAc ligand and the mercury(II)
,
center, leading to a HgꢀO distance of 2.813 A and an endo
arrangement of the ester carbonyl group with respect to
the central atom (cf. Fig. 2).
In the crystal this Hg(II) complex forms discrete
tetramers (Fig. 3). This specific association is brought
about by two additional intermolecular HgꢀO contacts
withtwoneighboringmolecules, involvingtheLewisbasic
oxygen of the NO-functions. Only one such function per
monomeric unit entertains two such interactions with the
Hg(II) centers of two neighboring molecules, connecting
5. Experimental
5.1. General comments
All reactions were carried out under a nitrogen atmo-
sphere with anhydrous, N2-saturated solvents (except

480
R. Weiss et al. / Journal of Organometallic Chemistry 617–618 (2001) 473–482
ethyl acetate, which was not dried before use) in
Schlenk vessels.
tetramethyl-1H-imidazole-1-oxyl-3-oxide (11a) in 20 ml
THF was cooled to −78°C and 1.05 ml of a 2 M LDA
solution (2.1 mmol) were added dropwise. After stirring
for 10 min solid CO2 was added; the color became deep
purple. The solution was allowed to reach r.t. and
concentrated under reduced pressure. A purple solid was
then precipitated with Et2O (40 ml), isolated by filtration,
washed with Et2O (10 ml) and dried under vacuum. Yield
282 mg (68%).
Mass spectra were recorded on a Varian MAT 311 A
(EI) and micromass Zab Spec E (FAB) spectrometer.
The ESR spectra were obtained on a Bruker 300 ESP
(X Band), the UV spectra on a Shimadzu 3101 PC
spectrometer. The IR spectra were recorded on a
Bruker IFS 45 spectrometer.
5.2. Synthesis
Anal. Found: C, 45.39; H, 6.09; N, 11.39. Calc.: C,
46.40; H, 5.84; N, 13.52%. EI-MS (70 eV, 200°C) m/z:
312 (2M−2CO2)+, 198 (2M−2CO2−C6H12NO)+, 182
(2M−2CO2−C6H12NO2)+, 157 (M−CO2+1)+, 84
(C6H12)+, 69 (C5H9)+. FAB-MS (NBA) m/z: 629 (3M+
Li+H)+, 421 (2M+Li)+, 214 (M+Li)+. ESR (THF):
g=2.0007, aN=7.63 G. UV (CH2Cl2): u (m) 325.5, 337.0,
540.0. IR (KBr): w (intensity) 2991 (m), 1668 (s, b), 1537
(s), 1446 (s, b), 1372 (s), 1321 (s), 1218 (m), 1171 (s), 1142
(s), 902 (w), 872 (m), 819 (w), 784 (m), 740 (m), 618 (w),
545 (m).
5.2.1. 4,5-Dihydro-2,4,4,5,5-pentamethyl-1H-imidazole-
1-oxyl-3-oxide 14a
227 mg (1.44 mmol) of 4,5-dihydro-4,4,5,5-tetra-
methyl-1H-imidazole-1-oxyl-3-oxide (11a) in 15 ml THF
were cooled to −78°C and 0.79 ml of a 2 M LDA
solution (1.58 mmol) were added. After stirring for 10
min 0.7 ml (1.58 mmol) methyltrifluoromethanesulfonate
were added dropwise and the resulting solution was
allowed to reach r.t. The solvent was removed under
vacuum and the residue was chromatographed on silica
gel with ethyl acetate as solvent. Yield 149 mg (60.5%).
Anal. Found: C, 56.08; H, 9.05; N, 16.26. Calc.: C,
56.12; H, 8.83; N, 16.36%. EI-MS (70 eV, 70°C) m/z: 171
(M)+, 84 (C6H12)+, 69 (C5H9)+, 55 (C4H7)+. UV (hex-
ane): u (nm) 306.0, 317.5, 521.5, 555.5.
5.2.4. 4,5-Dihydro-4,4,5,5-tetramethyl-2-(tri-
methylsilyl)-1-imidazole-1-oxyl-3-oxide 14d
203 mg (1.29 mmol) of 4,5-dihydro-4,4,5,5-tetra-
methyl-1H-imidazole-1-oxyl-3-oxide (in 10 ml THF were
cooled to −78°C and 0.71 ml of a 2 M LDA solution
(1.42 mmol) were added. After stirring for 10 min 0.26
ml (1.42 mmol) trimethylsilyltrifluormethansulfonate
were added dropwise and the resulting solution was
allowed to reach r.t. The solvent was removed under
vacuum and the residue was chromatographed on silica
gel with ethyl acetate as solvent. Yield 222 mg (75%).
Anal. Found: C, 52.66; H, 9.37; N, 12.32. Calc.: C,
52.36; H, 9.23; N, 12.21%. EI-MS (70 eV, 60°C) m/z: 303
(M+TMS+H)+, 230 (M+H)+, 98 (TMSCN−H)+,
84 (C6H12)+, 73 (TMS)+, 69 (C5H9)+. ESR (THF):
g=2.0527, aN=7.28 G. UV (pentane): u (m) 315.5
(7869), 327.5 (10061), 549.0 (615), 591 (471). IR (KBr):
w (intensity) 2957 (s), 1635 (m), 1466 (m), 1452 (m), 1418
(s), 1367 (s), 1342 (m), 1250 (s), 1213 (w), 1170 (m), 1138
(m), 947 (m), 935 (m), 922 (m), 872 (s), 846 (s), 754 (m),
537 (w).
This compound proved to be identical to a sample
prepared according to Ullman [7a].
5.2.2. 2-[1-Hydroxy-1-(4-methylphenyl)methyl]-4,5-
dihydro-4,4,5,5-tetramethyl-1H-imidazole-1-oxyl-
3-oxide 14b
200 mg (1.27 mmol) of 4,5-dihydro-4,4,5,5-tetra-
methyl-1H-imidazole-1-oxyl-3-oxide (11a) in 10 ml THF
were cooled to −78°C and 0.7 ml of a 2 M LDA solution
(1.4 mmol) were added. After stirring for 10 min 1.27 ml
(1.27 mmol) p-tolylaldehyde were added dropwise and
the resulting solution was allowed to reach r.t. The
solvent was removed under vacuum and the residue was
chromatographed on silica gel with ethylacetate as sol-
vent. Yield 153 mg (43.4%).
Anal. Found: C, 65.18; H, 7.65; N, 9.97. Calc.: C,
64.96; H, 7.63; N, 10.10%. EI-MS (70 eV, 50°C) m/z: 277
(M)+, 260 (M−OH)+, 244 (M−OH−O)+, 157 (M−
C8H8O)+, 119 (C8H7O)+, 91 (C7H7)+, 84 (C6H12)+, 69
(C5H9)+, 55 (C4H7)+. FAB-MS (NBA) m/z: 277 (M)+.
ESR (CH2Cl2): g=2.0000, aN=7.65 G. UV (hexane): u
(m) 261.0 (4643), 323.0 (25286), 530.0 (1000), 559.0 (1071).
IR (KBr): w (intensity) 3423 (s), 2987 (m), 2925 (m), 1606
(m), 1540 (w), 1513 (m), 1452 (s), 1410 (s), 1374 (ss), 1255
(m), 1213 (m), 1166 (m), 1138 (m), 1064 (w), 822 (w), 759
(w), 540 (m).
5.2.5. 4,5-Dihydro-4,4,5,5-tetramethyl-1H-imidazole-
1-oxyl-3-oxide-2-mercapto-lithium 14e
225 mg (1.43 mmol) of 4,5-dihydro-4,4,5,5-tetra-
methyl-1H-imidazole-1-oxyl-3-oxide (11a) in 15 ml THF
were cooled to−78°C and 0.71 ml of a 2 M LDA
solution (1.43 mmol) were added dropwise. After stirring
for 10 min 46 mg (1.43 mg) sulfur were added and the
resulting solution was allowed to reach r.t.; the pale red
color became blue. Filtration over Celite, concentration
under reduced pressure and addition of petrolether
yielded a blue solid, isolated by filtration and dried under
vacuum. Yield 208 mg (80%).
5.2.3. 4,5-Dihydro-4,4,5,5-tetramethyl-1H-imidazole-
1-oxyl-3-oxide-2-carboxylate-lithium 14c
A solution of 315 mg (2 mmol) 4,5-dihydro-4,4,5,5-

R. Weiss et al. / Journal of Organometallic Chemistry 617–618 (2001) 473–482
481
Anal. Found: C, 41.05; H, 6.48; N, 11.93; S, 16.35.
Calc.: C, 43.07; H, 6.20; N, 14.35; S, 16.42%. FAB-MS
(NBA) m/z: 397 (2M+Li)+, 202 (M+Li)+. ESR
(THF): g=1.9932, aN=8.2 G. IR (KBr): w (intensity)
2982 (s), 2934 (m), 1677 (s), 1627 (m), 1537 (s), 1483
(m), 1459 (s), 1371 (s), 1343 (s), 1273 (s, b), 1215 (s, b),
1154 (s), 1139 (s), 1053 (m), 1000 (m), 950 (m), 890 (m),
866 (m), 768 (m), 739 (w), 624 (m), 541 (s).
(s), 1153 (s), 1136 (s), 1041 (s), 992 (m), 957 (w), 894
(w), 875 (w), 854 (m), 839 (m), 777 (m), 660 (m), 542
(s).
5.2.7. (Acetato-v-O)-(4,5-dihydro-4,4,5,5-tetramethyl-
3-oxido-1-oxy-1H-imidazole-2-yl)mercury 14g
To a solution of 185 mg (1.18 mmol) 4,5-dihydro-
4,4,5,5-tetramethyl-1H-imidazole-1-oxyl-3-oxide (11a)
in 10 ml acetonitrile 375 mg (1.18 mmol) of Hg(OAc)2
were added and stirred for 24 h. The purple product
was precipitated with Et2O, isolated by filtration,
washed with Et2O and dried under vacuum. Yield 377
mg (77%).
Anal. Found: C, 25.96; H, 3.54; N, 6.50. Calc.: C,
25.99; H, 3.63; N, 6.74%. EI-MS (70 eV, 120°C) m/z:
417 (M)+, 358 (M−C2H3O2)+, 202 (Hg)+, 84
(C6H12)+, 69 (C5H9)+. FABMS (NBA) m/z: 417 (M)+,
358 (M−C2H3O2)+. ESR (CH2Cl2): g=2.0000, aN=
7.52 G, aHg=55.52 G. UV (CH2Cl2): u (m) 347.0 (6146),
527 (613), 560.5 (719). IR (KBr): w (intensity) 2991 (w),
1632 (m), 1612 (m), 1574 (s), 1452 (w), 1404 (m), 1344
(m, b), 1262 (w), 1211 (w), 1160 (w), 1137 (w), 1020 (w),
921 (w), 841 (w), 802 (w), 648 (m).
5.2.6. 4,5-Dihydro-4,4,5,5-tetramethyl-1H-imidazole-
1-oxyl-3-oxide-2-seleno-lithium 14f
184 mg (1.17 mmol) 4,5-dihydro-4,4,5,5-tetramethyl-
1H-imidazole-1-oxyl-3-oxide (11a) in 20 ml THF were
cooled to −78°C and 0.58 ml of a 2 M LDA solution
(1.17 mmol) were added dropwise. After stirring for 10
min 92 mg (1.17 mmol) selenium were added and the
resulting solution was allowed to reach r.t.; the pale red
color became green. Filtration over Celite, concentra-
tion under reduced pressure and addition of Et2O
yielded a green solid, isolated by filtration and dried
under vacuum. Yield 157 mg (55%).
Anal. Found: C, 35.28; H, 5.49; N, 10.82. Calc.: C,
34.73; H, 5.00; N, 11.57%. ESR (THF): g=2.0096,
aN=7.96 G. IR (KBr): 2984 (s), 2933 (s), 1676 (s), 1620
(s), 1576 (s), 1537 (s), 1449 (s), 1367 (s), 1283 (s), 1213
5.3. Crystallographic studies
Table 2
Crystals suitable for X-ray structure analysis were
obtained by solvent diffusion of Et2O in a saturated
solution of compound 14j in CH2Cl2. The measure-
ments were carried out on a Nonius Mach3 single-crys-
tal diffractometer with graphite-monochromated
Crystal data and structure refinement
Empirical formula
Formula weight
Temperature (K)
Wavelength (A)
Crystal system
Space group
C9H15HgN2O4
415.82
173
0.71073
Tetragonal
I-4
,
,
Mo–Ka radiation (u=0.71073 A). The hydrogen
atoms were fixed in idealized positions with a riding
model. Table 2 lists details of unit cell data, intensity
data collection and structure refinement.
Unit cell dimensions
,
a (A)
15.1008(6)
15.1008(6)
10.9154(3)
90
90
90
2489.08(16)
8
12.368
Deep purple, block
0.250×0.1×0.1
1.91–29.50
,
b (A)
,
c (A)
5.4. Supplementary material
h (°)
i (°)
k (°)
Volume (A )
Z
Crystallographic data for the structural analysis have
been deposited with the Cambridge Crystallographic
Data Center, CCDC No. CCDC 147580 for compound
14g. Copies of this information may be obtained free of
charge from: The Director, CCDC, 12 Union Road,
Cambridge, CB2 1EZ, UK (fax: +44-1223-336033;
e-mail: deposit@ccdc.cam.ac.uk or www: http://
www.ccdc.cam.ac.uk).
3
,
Absorption coefficient (mm−1
Crystal color, habit
Crystal size (mm)
q-Range for data collection (°)
Reflections collected
Independent reflections
Reflections
)
2706
2701 [Rint=0.0461]
2412
Absorption correction method
Refinement method
c-scans
Full-matrix least-squares on
F2
References
Data/restraints/parameters
Goodness-of-fit on F2
Final R indices [I\2|(I)]
R indices (all data)
2701/0/145
1.059
R1=0.441, wR2=0.10423
R1=0.0544, wR2=0. 1110
0.52(3)
[1] R. Weiss, S. Reichel, M. Handke, F. Hampel, Angew. Chem.
110 (1998) 352; Angew. Chem. Int. Ed. Engl. 37 (1998) 344.
[2] R. Weiss, S. Reichel, Eur. J. Inorg. Chem. (2000) 1935.
[3] Ph.D. Thesis, S. Reichel, University of Erlangen-Nu¨rnberg,
1998.
Absolute structure parameter
Largest difference peak and hole
1.299 and −2.079
−3
,
(e A
)
[4] D. Bourissou, O. Guerret, F.P. Gabbai, G. Bertrand, Chem.
Rev. 100 (2000) 39.

482
R. Weiss et al. / Journal of Organometallic Chemistry 617–618 (2001) 473–482
[5] K. Deuchert, S. Hu¨nig, Angew. Chem. 12 (1978) 927.
Y. Mochizuki, T. Kato, H. Inoue, S. Ohba, Chem. Lett. (1997)
251. (c) A. Izuoka, M. Hiraishi, T. Abe, T. Sugawara, J. Am.
Chem. Soc. 122 (2000) 3234. (d) J.P. Sutter, M.L. Kahn, O.
Kahn, Adv. Mater. 11 (1999) 863. (e) O. Felix, M.W. Hosseini,
A.D. Cian, J. Fischer, L. Catala, P. Turek, Tetrahedron Lett. 40
(1999) 2943. (f) T. Mitsumori, K. Inoue, N. Koga, H. Iwamura,
J. Am. Chem. Soc. 117 (1995) 2467. (g) L. Dulog, J.S. Kim,
Angew. Chem. 102 (1990) 403. (h) C. Batz, P. Amann, H.-J.
Deiseroth, L. Dulog, Liebigs Ann. Chem. (1994) 739. (i) T.
Akita, Y. Mazaki, K. Kobayashi, J. Org. Chem. 60 (1995) 2092.
(j) T. Akita, K. Kobayashi, Tetrahedron Lett. 52 (1996) 6893. (k)
T. Akita, Y. Mazaki, K. Kobayashi, Tetrahedron Lett. 36 (1995)
5543. (l) M.M. Matsushita, A. lzuoka, T. Sugawara, T.
Kobayashi, N. Wada, N. Takeda, M. Ishikawa, J. Am. Chem.
Soc. 119 (1997) 4369. (m) M. Tanaka, K. Matsuda, T. Itoh, H.
Iwamura, J. Am. Chem. Soc. 120 (1998) 7168. (n) G. Ulrich, P.
Turek, R. Ziessel, Tetrahedron Lett. 37 (1996) 8755. (o) F.M.
Romero, R. Ziessel, Tetrahedron Lett. 40 (1999) 1895. (p) C.
Stroh, R. Ziessel, Tetrahedron Lett. 40 (1999) 4543. (q) C. Stroh,
P. Turek, R. Ziessel, Chem. Commun. (1998) 2337.
[6] (a) H. Link, W. Klo¨tzer, M. Karpitschka, M. Montavon, R.
Mu¨ssner, N. Singewald, Angew. Chem. 102 (1990) 559. (b) W.
Klo¨tzer, H. Baldinger, E.M. Karpitschka, J. Knoflach, Synthesis
(1982) 592. (c) G. Laus, W. Klo¨tzer, Synthesis (1989) 269 (d) G.
Laus, J. Stadlwieser, W. Klo¨tzer, Synthesis (1989) 773. (e) G.
Laus, J. Stadelwieser, W. Klo¨tzer, Synthesis (1990) 795. (f) V.
Bock, W. Klo¨tzer, N. Singewald, G. Strieder, Synthesis (1987)
1058.
[7] (a) D.G.B. Boocock, R. Darcy, E.F. Ullman, J. Am. Chem. Soc.
21 (1968) 5945. (b) D.G.B. Boocock, E.F. Ulman, J. Am. Chem.
Soc. 24 (1968) 6874. (c) E.F. Ullman, D.G.B. Boocock, Chem.
Commun. 20 (1969) 1161. (d) P.W. Kopf, R. Kreilich, D.G.B.
Boocock, E.F. Ullman, J. Am. Chem. Soc. (1970) 4531. (e) E.F.
Ullman, L. Call, J.H. Osieckie, J. Org. Chem. 35 (1970) 3623. (f)
E.F. Ullman, L. Call, J. Am. Chem. Soc. 92 (1970) 7210. (g) L.
Call, E.F. Ullman, Tetrahedron Lett. (1973) 961. (h) E.F. Ull-
man, J.H. Osieckie, D.G.B. Boocock, R. Darcy, J. Am. Chem.
Soc. 94 (1972) 7049. (i) E.F. Ullman, L. Call, S.S. Tseng, J. Am.
Chem. Soc. 95 (1973) 1677. (j) M. Deumal J. Cirujed, J. Vaciena,
J.J. Novoa, Chem. Eur. J. 5 (1999) 1631.
[13] (a) M. Lammchen, T.W. Mittag, J. Chem. Soc. C (1966) 2300.
(b) R. Sayre, J. Am. Chem. Soc. 77 (1955) 6689.
[14] (a) J.P. Foster, F. Weinhold, J. Chem. Soc. 102 (1980) 7211. (b)
A.E. Reed, F. Weinhold, Chem. Rev. 88 (1988) 899.
[8] A. Caneschi, J. Laugier, P. Rey, J. Chem. Soc. Perkin Trans. 1
(1987) 1077.
[9] R. Weiss, N. Kraut, in preparation.
[10] E.F. Ullman, J.H. Osieckie, J. Am. Chem. Soc. 90 (1968) 1078.
[11] (a) A. Caneschi, D. Gatteschi, R. Sessoli, P. Rey, Acc. Chem.
Res. 22 (1989) 392. (b) A. Caneschi, P. Chiesi, L. David, F.
Ferraro, D. Gatteschi, R. Sessoli, Inorg. Chem. 32 (1993) 1445.
(c) A. Caneschi, F. Ferraro, D. Gatteschi, P. Rey, R. Sessoli,
Inorg. Chem. 29 (1990) 4217. (d) A. Caneschi, F. Ferraro, D.
Gatteschi, P. Rey, R. Sessoli, Inorg. Chem. 30 (1991) 3126. (e) C.
Benelli, A. Caneschi, D. Gatteschi, L. Pardi, P. Rey, Inorg.
Chem. 29 (1990) 4223. (f) A. Caneschi, D. Gatteschi, M.C.
Melandri, P. Rey, R. Sessoli, Inorg. Chem. 29 (1990) 4228. (g)
G. Ulrich, R. Ziessel, Tetrahedron Lett. (1993) 1211. (h) G.
Ulrich, R. Ziessel, Tetrahedron Lett. 35 (1993) 1215.
[15] H.A. Bent, Chem. Rev. 61 (1961) 275.
[16] A.J. Arduengo, Acc. Chem. Res. 32 (1999) 913.
[17] (a) H.-W. Wanzlick, H.-J. Scho¨nherr, Angew. Chem. 80 (1968)
154. (b) A.J. Arduengo, R.L. Harlow, W.J. Marshall, T.K.
Prakasha, Heteroat. Chem. 7 (1996) 421.
[18] After completion of this work we discovered that 14g is men-
tioned in an early patent claim by R.K. Leute, E.F. Ullman, US
Patent 3,697,535, 1969. Only the analytical composition of an
otherwise structurally uncharacterized compound of unknown
purity is cited in support of this claim.
[19] By courtesy of Professor Dr. U. Zenneck and Dipl. Chem. M.
Zeller, Institut fu¨r Anorganische Chemie der Universita¨t Erlan-
gen-Nu¨rnberg.
[12] (a) R. Kumai, M.M. Matsushita, A. Izuoka, T. Sugawara, J.
Am. Chem. Soc. 116 (1994) 4523. (b) N. Yoshioka, M. Irisawa,
[20] L.J. Farrugia, J. Appl. Crystallogr. 30 (1997) 565.
.