Photochemical generation, intramolecular reactions, and spectroscopic detection of oxonium ylide and carbene intermediates in a crystalline ortho-(1,3-dioxolan-2-yl)-diaryldiazomethane

Article abstract of DOI:10.1039/b814387k

Photochemical excitation of a diaryldiazomethane with an ortho-acetal both in solution and in crystals led to products that originate from the expected diarylcarbene and form an intramolecular oxonium ylide. While crystals strongly favored the formation of a benzocyclobutane by intramolecular hydrogen abstraction in the triplet carbene, reactions in benzene led exclusively to products that derive from the oxonium ylide. Studies in methyl cyclohexane glasses and in mixed crystals at 77 K led to the spectroscopic detection of triplet carbene ³C, which partially transformed into a new species that we assign as the sought-after oxonium ylide Y. The formation of a formally ionic intermediate suggests that the scope of reactions by reactive intermediates in crystalline solids may be broader than it is generally assumed. The Royal Society of Chemistry 2009.

Full text of DOI:10.1039/b814387k

View Article Online / Journal Homepage / Table of Contents for this issue
PAPER
www.rsc.org/obc | Organic & Biomolecular Chemistry
Photochemical generation, intramolecular reactions, and spectroscopic
detection of oxonium ylide and carbene intermediates in a crystalline
ortho-(1,3-dioxolan-2-yl)-diaryldiazomethane†‡
Miguel A. Garcia-Garibay* and Hung Dang
Received 19th August 2008, Accepted 5th December 2008
First published as an Advance Article on the web 3rd February 2009
DOI: 10.1039/b814387k
Photochemical excitation of a diaryldiazomethane with an ortho-acetal both in solution and in crystals
led to products that originate from the expected diarylcarbene and form an intramolecular oxonium
ylide. While crystals strongly favored the formation of a benzocyclobutane by intramolecular hydrogen
abstraction in the triplet carbene, reactions in benzene led exclusively to products that derive from the
oxonium ylide. Studies in methyl cyclohexane glasses and in mixed crystals at 77 K led to the
spectroscopic detection of triplet carbene 3C, which partially transformed into a new species that we
assign as the sought-after oxonium ylide Y. The formation of a formally ionic intermediate suggests
that the scope of reactions by reactive intermediates in crystalline solids may be broader than it is
generally assumed.
Introduction
Although reactive intermediates tend to follow numerous reaction
pathways in solution,1 our group has shown that radical pairs2
and carbenes3,4 react with remarkable selectivities and specificities
when generated in crystals of their precursors.5 We previously
reported that ruby-red crystals of 1,2-diaryldiazopropanes trans-
form into (Z)-1,2-diarylpropenes in nearly quantitative yields by
a highly selective 1,2-H shift of their photochemically generated
carbene intermediates (Scheme 1).6,7 Reactions in solution yielded
a mixture of products that included 1,2-H shifts, 1,2-Ar migrations
and reactions with solvents.8 Recognizing that crystalline solids
may control the reactivity of high-energy species while providing
valuable structural insight from X-ray diffraction data,9 we
Scheme 2 Formation of an oxonium ylide from an aryl carbene with an
ortho-dioxolane.
decided to explore the generation and reactivity of oxonium
ylides10 (Y, Scheme 2). Our interest in oxonium ylides stems from
We selected for this study samples of ortho-(1,3-dioxolan-2-yl)-
their potential for the synthesis of complex structures11 and, more
diaryldiazomethane derivative 1a (Scheme 2, R = 4-biphenylyl,
importantly, from the possibility of increasing the number of
reactions in crystalline solids.
X = N). Compound 1a is a crystalline and high-melting analog
of the monoaryldiazomethane derived from the sodium salt of
tosyl hydrazone 1b (R = H, X = N(Na)Ts) previously studied
in solution by Kirmse and Kund.12 As illustrated in Scheme 2,
oxonium ylide Y can be ideally described as a ground state species
with a positively charged hypervalent oxygen linked directly to a
carbon atom carrying an unshared pair of electrons. The oxonium
Scheme 1 Selective 1,2-H shift in 1,2-diarylpropylidenes generated in
crystals of their diazo precursors.
ylide is formed by donation of a lone pair from the adjacent
oxygen to the empty p-orbital of the sp2-hybridized singlet state
carbene (1C), which in the case of diaryl carbenes exists in small
equilibrium concentrations as compared to the more stable triplet
(3C).13 Recent studies by Moss et al. have confirmed that the
formation of oxonium ylides depends strongly on the nature of the
carbene and the ether. While it is calculated that favorable cases
give rise to typical C–O sigma bonds, others are best described
in terms of a weak solvation effect.14 The intermediacy of an
oxonium ylide in the case of sodium tosyl hydrazone 1b can be
deduced from product analysis. UV irradiation in a non-reactive
solvent primarily produced 6,8-dioxabicyclo[3.2.1]octane A (10%)
Department of Chemistry and Biochemistry, University of California, 405
Hilgard Ave, Los Angeles, CA, USA. E-mail: mgg@chem.ucla.edu; Fax: +1
310 825 0767; Tel: +1 310 825 3159
† Electronic supplementary information (ESI) available: 1H and 13C NMR
spectra of compounds 1a, 6, 8, 9 and 10. Additional details of the crystal
structures determinations of 1a and 8. CCDC reference numbers 698857–
698858. For ESI and crystallographic data in CIF or other electronic
format see DOI: 10.1039/b814387k
‡ This work was supported by NSF grants CHE0551938 and
DMR0605688
1106 | Org. Biomol. Chem., 2009, 7, 1106–1114
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
and 1,4-dioxane B (66%).12 It was suggested that A forms by formal
insertion of the carbene into a C–H bond of the neighboring
dioxolane while formation of B was assigned to a formal 1,2-alkyl
migration (Stevens-like rearrangement) of ylide Y.12
opening. Interestingly, photochemical reactions in crystals of 1a
favor a benzocyclobutane product (10), which presumably derives
from hydrogen-transfer and radical cyclization from the triplet
carbene. Small amounts of products 8 and 9 from the oxonium
ylide are also formed upon extended irradiation. Irradiation in
methyl cyclohexane matrices and in mixed crystals at 77 K led to
Expecting distinguishable products from the carbene and the
ylide, we selected compound 1a as the starting point to analyze
the formation of the latter. In this paper we wish to report the
synthesis and X-ray structure of diaryldiazo acetal 1a (Scheme 3),
its solution and solid state photochemistry, as well as spectroscopic
evidence for the detection of both the carbene and oxonium ylide.
As reported below, the solution photochemistry of diazo 1a in
benzene at ambient temperature differs from that of the diazo
derived from the tosyl hydrazone salt 1b in diglyme. Irradiation of
1a resulted in three major products that we assign as originating
from the oxonium ylide (Scheme 4). Two of these products (8
and 9) can be accounted for by 1,2-alkyl migrations (Stevens
rearrangements) and the third one (6) by an electrocyclic ring
3
the spectroscopic detection of triplet carbene C, which partially
transformed into a new species that we assign as the sought-after
oxonium ylide intermediate Y. Our results suggest that the ylide
and the carbene exist in equilibrium in the crystal, with the triplet
carbene being the dominant species.
Results and discussion
Synthesis and characterization of the diazo precursor
The diazo precursor 1a was synthesized in five steps from
the commercially available 2-bromobenzaldehyde (Scheme 3).
Bromoacetal 2 was isolated in 98% yield after treating 2-
bromobenzaldehyde with anhydrous ethylene glycol in refluxing
benzene. Treatment of 2 with n-BuLi at -78 ◦C followed by
addition of 4-biphenylcarboxaldehyde gave alcohol 3 in 88%
isolated yield. Swern oxidation of alcohol 3 gave ketone 4 in
near quantitative yield. The next step involved refluxing ketone
4 with anhydrous hydrazine in absolute ethanol to give hydrazone
5 (51%) as a thick viscous oil that solidified after several days at
room temperature. Hydrazone formation was sluggish, probably
due to the steric hindrance of the ketone by the neighboring ortho-
acetal and the biphenyl substituent. Oxidation of hydrazone 5 with
mercury(II) oxide afforded diazo 1a as deep ruby-red prisms grown
from diethyl ether solutions.
The structure of diaryldiazomethane 1a was determined by
conventional spectroscopic methods and confirmed by single
crystal X-ray diffraction analysis. Solution NMR carried out in
deuterated chloroform showed the 1H signal of the acetal proton
at 5.88 ppm and the signals of the ethylenedioxy group in the range
of 3.93–4.13 ppm. The diazo carbon signal occurs at 122.6 ppm
and the acetal carbon at 101.5 ppm.
Scheme 3 Synthesis of diaryldiazomethane 1a.
X-Ray diffraction data acquired on deep-red prisms obtained
from diethyl ether were solved in the triclinic crystal space group
¯
P1 with one molecule in the asymmetric unit (Fig. 1). The
structure of 1a has the planes of the biphenyl and 1,3-dioxolane
nearly orthogonal to that of the central ortho-phenylene, with the
diazo moiety conjugated with the biphenyl substituent. Assuming
that conformational motions will be minimal in the crystal, the
feasibility of ylide formation is suggested by the relatively close
distances between the prospective carbene carbon and the two
Scheme 4 Formation of carbene (C) and ylide (Y) intermediates and their
Fig. 1 Crystals of diaryldiazomethane 1a and ORTEP diagram with
more likely intramolecular reaction pathways.
thermal ellipsoids drawn at the 50% probability level.
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 1106–1114 | 1107
©

View Article Online
Table 1 Product distribution from photolysisa of diaryldiazomethane 1a
1
aromatic, vinylic and ethylenedioxy hydrogens in the H NMR
in solution and in crystalsb
1
spectrum, which were assigned with the help of H–1H COSY
and 1H–1H NOESY measurements. Three sets of aromatic signals
at 7.80 ppm (2H), 7.43 ppm (4H) and 7.20 ppm (3H) ppm
show the expected cross-correlations and account for the nine
hydrogens of the biphenyl group. A spin system corresponding to
the four vinylic hydrogens of the ortho-quinodimethane ring (Hb–
He) occurs between 6.73 and 5.66 ppm (Scheme 5). The isolated
enol ether hydrogen (Ha) appears as a singlet at 6.16 ppm. Signals
corresponding to the ethylenedioxy bridge occur as multiplets
centered at 3.77, 3.69, 2.07 and 1.44 ppm, with chemical shifts
that suggest shielding effects from the neighboring p-systems. The
1H–1H NOESY spectrum shows cross peaks between Ha and
Hb, which helps identify the resonances corresponding to Hc–
He. A cross peak between Hb and the most shielded signal of the
ethylenedioxy group at 1.44 ppm suggests a conformation where
the structure of the benzodioxacyclooctane is folded over the
shielding region of the ortho-quinodimethane p-system. Samples
of 6 were stable upon standing for long periods at ambient tem-
perature showing no tendency to cyclize into benzocyclobutane 9
or towards an orbital symmetry allowed trans-fused analog.
Reaction medium
6
8
9
10
C6H6
Crystal
15
0
15
10
43
10
0
75
a UV irradiations were carried out with a Hanovia medium pressure Hg
lamp using a l > 290 nm glass filter. b Isolated yields after ca. 90% of the
reactant had been consumed.
˚
acetal oxygens with O1–C13 and O2–C13 distances of 3.85 A and
˚
4.10 A, respectively.
Photochemical experiments and product analyses in solution
Photochemical experiments were carried out with argon-
exchanged 0.1 M benzene solutions at ambient temperature
filtering the output of a 400 W mercury vapor Hanovia lamp
with a l > 380 nm cutoff filter (Table 1). Analysis of the reaction
mixture by thin layer chromatography and 1H NMR revealed three
major products, which were subsequently separated on silica gel
using a 3 : 1 v/v mixture of methylene chloride and hexane. The
three compounds were isolated in ca. 43%, 15% and 15% yields,
and were later assigned the structures 9, 6 and 8 (Scheme 3),
respectively. Although the NMR spectra of compound 8 originally
suggested that it might correspond to structure 7, as expected from
the report of compound 1b, the correct structure was determined
by single crystal X-ray diffraction. As described in more detail
below, reactions in crystals produced benzocyclobutane 10 as the
major product along with small amounts of 8 and 9.
Samples of o-QDM 6 were characterized by their pale yellow
color and a UV spectrum with a lmax = 372 nm that extends
up to ca. 475 nm, which is characteristic of simple ortho-
quinodimethanes (Fig. 2).15 Key to the assignment is the fact that
there are only two carbon signals in the sp3 region at 66.22 and
39.08 ppm, respectively, which are assigned to the ethylenedioxy
bridge. The structure was also identified by the distribution of
Scheme 5 Assignments and key NOE correlation to assign of or-
tho-quinodimethane 6.
The 1H spectrum of 1,4-dioxane 9 in CDCl3 is characterized by
a set of overlapping aromatic signals between 7.23 and 7.59 ppm
(13H), a benzylic methine at 5.34 ppm and two sets of signals
between 3.81–3.88 (3H) and 4.01–4.04 (1H) ppm that correspond
to the ethylenedioxy bridge. The 13C NMR spectrum of 9 shows
the expected number of signals including 14 aromatic carbons,
a quaternary cyclobutane carbon at 83.26 ppm, a cyclobutane
methine at 80.38 ppm, and the two ether methylenes at 61.66 and
60.61 ppm. These signals are consistent with a 1,4-dioxane derived
from a formal 1,2-shift in the oxonium intermediate. Compound
9 corresponds to the simpler analog B reported by Kirmse and
Kund (Scheme 2).
The 1H NMR spectrum of 8 in CDCl3 is characterized by signals
that extend from 7.65 ppm in the aromatic region all the way up to
2.07 ppm, as it might have been expected for structure 7. However,
single crystal X-ray diffraction analysis (vide infra) revealed the
connectivity present in isomer 8. Aromatic signals between 7.65
and 7.02 ppm account for the required 13H, a sharp singlet at
6.39 ppm is assigned to the benzylic acetal, two multiplets at 3.89
and 3.40 ppm with a 2JHH = 11.8 account for a methylene attached
to the oxygen atom, and two multiplets at 2.76 and 2.07 ppm
2
with JHH = 12.6 account for the other methylene signals. One
can also detect the expected vicinal couplings. Analyses of the 13
C
NMR and a 2D heteronuclear 1H–13C HMQC spectra confirm that
proton signals at 2.07 and 2.76 ppm correlate with the same carbon
Fig. 2 UV-Vis spectrum of o-QDM 6 in methyl cyclohexane.
1108 | Org. Biomol. Chem., 2009, 7, 1106–1114
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
at 35.70 ppm, and that signals at 3.40 and 3.89 ppm correlate with
the carbon at 59.45 ppm. Likewise, the acetal proton at 6.39 ppm
correlates with the carbon signal at 101.01 ppm.
The structure of 8 was corroborated by single crystal X-ray
diffraction analysis (Fig. 3). Compound 8 crystallized in the mono-
clinic space group P21/c, with one molecule in the asymmetric unit
and four in the unit cell (Z = 4). The structure of 8 was confirmed
as a benzobicyclo[3.2.1]octane, with the 1,3-dioxane ring adopting
a chair conformation. The 4¢-biphenyl substituent on C10 and
the acetal proton on C3 at the bridgehead positions adopt
equatorial positions on the chair of the 1,3-dioxane and the ortho-
phenylene ring (C4–C9) occupies the axial position. The biphenyl
substituent adopts a nearly planar conformation with a slight twist
of the two aromatic rings (C13–C14–C17–C22 torsional angle =
ca. 5◦).
Scheme 6 Hydrogen abstraction by 3C leading to formation of benzo-
cyclobutane 10. Acid hydrolysis of the ketal in 10 leads to benzocyclobu-
tanone 11.
Spectroscopic analysis of the triplet carbene and oxonium ylide in
methylcyclohexane glasses and in crystals at 77 K
Since it is well known that diarylcarbenes possess a triplet
ground state,1,13 ylide formation must occur either immediately
after denitrogenation or from a thermally equilibrated singlet.
Knowing that triplet carbenes possess a relatively strong T1–T0
fluorescence,17 we decided to explore the detection of the carbene
and to probe the formation of the oxonium ylide (presumably
from the thermally populated singlet). A dilute solution of diazo
1a (ca. 10-4 M) in spectroscopic grade methylcyclohexane (MCH)
was cooled to 77 K and irradiated with l = 340 nm. As expected,
Fig. 3 ORTEP diagram of dioxabicyclo[3.2.1]octane 8 with thermal
ellipsoids at the 50% probability level.
3
the fluorescence excitation and emission spectra of carbene C
are essentially identical to those of diphenyldiazomethane.16 The
emission spectrum consists of two maxima at 560 and 600 nm and
a shoulder by 650 nm. The excitation spectrum is characterized
by absorption maxima of ca. 350 nm, and two weaker bands
corresponding to the lowest energy transition at ca. 420 and
480 nm (Fig. 4).18
Photochemical experiments and product analysis in the crystalline
solid state
Solid-state photolysis were carried out with powdered samples
of diaryldiazomethane 1a sandwiched between microscope slides
and covered with a l > 380 nm filter at 0 ◦C using a 400 W
mercury vapor Hanovia lamp. While photolysis to low conversion
values (ca. 10%) produced benzocyclobutenone acetal 10 as the
only detectable product, irradiation to 95% conversion led to the
formation of acetal 10 in 75% yield along with ca. 10% each of
8 and 9. The spectroscopic characterization of compound 10 was
based on the analysis of its spectral data and its transformation
to benzocyclobutanone 11 by acid hydrolysis. A highly diagnostic
benzylic cyclobutane singlet is observed at 4.97 ppm rather than
close to 6.0 ppm, as it occurs when the benzylic hydrogen is
part of an acetal carbon. The dioxolane ring is characterized
by three multiplets at 4.19–4.11 ppm (2H), 4.03–3.99 ppm (1H)
and 3.70 ppm (1H). The new quaternary ketal carbon resonates
at 111.25 ppm. In a key experiment, acid hydrolysis led to
the formation of the expected benzocyclobutanone, which was
characterized by HRMS (m/z = 270.1045 for C20H14O) and by
FT-IR spectroscopy, which included the characteristically high-
Fig. 4 Fluorescence excitation (left) and emission (right) spectra of 3C.
Irradiated samples of 1a also showed a bright and persistent
orange-red discoloration, suggesting the presence of a second
non-fluorescent chromophore. A UV-Vis absorption revealed a
very intense broad band with a lmax = 490 nm, which persists
indefinitely at 77 K but disappears upon warming (Fig. 5). A
control experiment with (4-phenyl)diphenyldiazomethane showed
an indistinguishable carbene emission, but not the bright orange
transient. Based on this observation we hypothesized that the
=
frequency stretching band of the benzocyclobutenone C O at
1763 cm-1 (lit.16 1760 cm-1). It is likely that compound 10 is
generated from the triplet carbene (3C) by abstraction of the acetal
hydrogen at C20 to give a triplet 1,4-biradical (3BR), which com-
bines to form the four-membered ring after intersystem crossing
(Scheme 6).
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 1106–1114 | 1109
©

View Article Online
Reactions of carbenes and oxonium ylides
Our interpretation of the product analysis and spectroscopic data
as a function of reaction media is based on the reaction pathways
illustrated in Scheme 4. While it is well established that carbenes
can be generated efficiently by photochemical excitation of diazo
compounds both at ambient temperature and at 77 K,20 the
formation of oxonium ylides by reaction with ethers has been
inferred mainly from product analysis studies.21 The spectroscopic
detection of oxonium ylides in solution by laser flash photolysis
has been limited to a few carbenes with a singlet ground state.22,23
A specific but reversible interaction between several free carbenes
and ethers has been documented in terms of shifts in the carbene
absorption spectrum24 and/or in terms of extension of the carbene
lifetime.14,25 It is expected that the formation of oxonium ylides
from aryl and diaryl carbenes will proceed through the thermally
populated singlet state, such that triplet carbene, singlet carbene
and oxonium ylide may coexist at concentrations that depend
on their relative energies. Taking an experimentally determined
Singlet–Triplet (S–T) gap of 2.6 kcal mol-1 for diphenyl carbene as
a guideline, we may estimate equilibrium populations at 300 K of
the order of ca. 75 : 1 in favor of 3C.26 While the relative energies of
1C and Y are not known at this time, a common feature of solution
products 6, 8 and 9 is that they all have the former diazo carbon
attached to an oxygen atom, as expected for compounds that
arise from an oxonium ylide intermediate. ortho-Quinodimethane
6 may be formulated as the result of a ring opening reaction
and compounds 8 and 9 may be viewed as arising from two
possible Stevens rearrangements. While the latter are described
as formal 1,2-migrations of the anionic carbon to either of the two
carbons flanking the oxonium center (illustrated by the blue and
red arrows in the Scheme), based on evidence from analogous
ammonium and sulfonium species it is commonly agreed that
these reactions proceed by dissociative processes involving short-
lived radical intermediates.27 It is interesting that neither 7 nor 10,
which are the carbene-derived products, are detectable in benzene
solution. This suggests that ylide formation and rearrangement
from the singlet carbene must be favorable with respect to singlet
C–H insertion. While it is likely that the triplet diaryl carbene is
the most abundant species at equilibrium, it is well known that
intramolecular hydrogen abstraction by triplet diaryl carbenes is
a relatively slow process. In fact, ortho-methyl substituents have
been used as “protecting” groups as they enhance the half-life of
diphenylcarbene by factors of 104–105, into the ms regime.28,29
Fig. 5 UV-Vis Absorption spectrum of the persistent signal to assigned
to ylide Y, lmax ª 490 nm, which is formed by irradiation of diazo 1a in
MCH at 77 K.
490 nm transient corresponds to oxonium ylide Y (Schemes 2
and 4).
When microcrystalline samples of diazo 1a were irradiated and
analyzed at 77 K there was no detectable fluorescence of the triple
carbene intermediate. Assuming that the carbene emission may
be quenched by the diazo precursor we prepared mixed crystals
with solutions containing 1% of diazo 1a and 99% of ketoacetal
4. It has been previously shown that analogous ketones and diazo
compounds are able to form solid solutions,19 a fact that can be
easily confirmed by spectroscopic methods, including the visual
reddish-orange coloration of the mixed crystals.
With time-dependent experiments carried out with excitation at
l = 350 nm and front-face fluorescence detection at 550 nm we
were able to detect the rapid growth of the carbene emission, which
was followed by slower decay (Fig. 6). A full emission spectrum
acquired with a fresh sample confirmed that the carbene was the
emitting species. A qualitative correlation was observed between
the decrease in the concentration of carbene 3C and an increase in
the color of ylide Y. A time constant of ca. 200 s was estimated
for the decay of the carbene signal, which did not go to zero,
suggesting that an equilibrium was established. We also noticed
that oxonium ylide Y is persistent for several hours at 77 K,
both in glasses and in crystals, but it disappears rapidly upon
warming.
Formation of carbenes and oxonium ylides
While the emission spectrum in Fig. 4 is identical to that obtained
from samples of 4-phenyl-substituted diphenyldiazomethane, the
spectrum in Fig. 5 was only observed for the compound bearing
an ortho-acetal (1a). It is worth noting that the position of
the absorption band is very similar to those previously re-
ported for several types of oxonium ylides,14,23 including those
formed with 4-nitrophenylchlorocarbene,14,22a carbomethoxy-2-
naphthylcarbene,22b and several alkyl-chlorocarbenes.23 In fact, a
recent computational and experimental study by Moss et al. on
4-nitrophenylchlorocarbene with several aliphatic and aromatic
ethers reports a very appealing and elegant model regarding
the spectroscopic characterization of “O-ylide”, “O-ylidic” and
Fig. 6 Emission growth and decay of carbene C in mixed crystals of 4 at
77 K.
1110 | Org. Biomol. Chem., 2009, 7, 1106–1114
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
“p-complex” interactions.14 The definitions proposed by Moss
et al. are based on the formation of a “fully (or nearly fully)
developed O–C bond”, “an ether-carbene transient with O–C dis-
tances that exceed typical O–C bond lengths”, and “charge trans-
fer complexes formed between aromatic p-donors and carbene
acceptors”. With a combination of ground [PBE/6–311+G(d)]
and excited state calculations [TD-B3LYP/6-311+G(d)//PBE/6-
311+G(d)] to correlate the O–C bond distances with the ab-
sorption spectra, the authors proposed the following model: (1)
pristine solid at low conversion values favors the formation of
a carbene-derived benzocyclobutane acetal. Since oxonium ylide
products 8 and 9 are only detected at high conversion products
they are likely to originate at defect sites. While a more detailed
characterization of the reaction mechanism in solution and in
crystals will require additional examples to establish the generality
of this process, this study demonstrates that the scope of reactions
in crystals may be much broader than is generally assumed.
˚
O-ylides are characterized by O–C bonds of ca. 1.45 A and
Experimental
an absorption band with a lmax of 460–510 nm. O-Ylides were
reported with 4-nitrophenylchlorocarbene with diethyl ether, THF
and 18-crown-6. (2) O-Ylidic complexes, which can be observed
with aromatic ethers such as anisole, 1,4-dimethoxybenzene and
1,2,3,5-tetramethoxybenzene, absorb in the same general region
as the O-ylides, but (3) they are accompanied by the absorption
of the p-complexes with a lmax~340–380 nm. Notably, as the O-
ylidic species are less stable than the p-complexes, their relative
absorbances display the changes in intensity expected for an
equilibration process.
Based on the model suggested by Moss et al., 14 the formation
of a species with a lmax ª 480 nm is consistent with an O-ylide
or an O-ylidic interaction. The fact that the band appears to be
indefinitely stable at 77 K is indicative of a species that exists
in an energy well, but one can make no suggestions regarding
its relative energy with respect to the undetected singlet state
1H and 13C NMR spectra were obtained on an AMS 360
(360 MHz), Bruker ARX400 (400 MHz), or Avance Bruker
ARX500 (500 MHz) spectrometer. 1H and 13C NMR spectra
were obtained in C6D6 or CDCl3 with TMS as an internal
standard, unless otherwise noted. IR spectra were obtained
on a Perkin-Elmer Paragon 1000 FT-IR spectrometer. UV-Vis
absorption spectra were acquired with a Hewlett-Packard 8453
spectrophotometer. Fluorescence spectra were recorded with a
Spex-Fluorolog II spectrofluorimeter and corrected for nonlin-
ear instrumental response. High-resolution electron impact (EI-
HiRes) mass spectra were obtained on a VG Autospec (Micromass,
Beverly, MA) spectrometer.
2-(2-Bromophenyl)-1,3-dioxolane 232
To a solution of 2-bromobenzaldehyde (8.0 g, 43.2 mmol) in 30 mL
of benzene in a 50 mL three-necked round-bottomed flask were
added in one portion anhydrous ethylene glycol (2.7 mL, 48 mmol)
and p-toluenesulfonic acid monohydrate (5.29 mg, 27.8 mmol).
The resulting solution mixture was refluxed until the theoretical
yield of water had been collected in a Dean–Stark trap. TLC
showed complete disappearance of the starting material. The
mixture was cooled, washed with 10% aqueous sodium hydroxide
(2 ¥ 30 mL), followed by DI water (2 ¥ 30 mL) and brine (1 ¥
30 mL), and the organic layer was dried over anhydrous MgSO4.
The solvent was removed in vacuo and pumped under high vacuum
to give 9.7 g (98%) of a clear yellow viscous oil; nmax/cm-1 (neat)
3067, 2954, 2888, 1593, 1571, 1473, 1443, 1388, 1270, 1211, 1125,
1092, 1042, 1023, 970, 944, 758. dH (500 MHz, CDCl3) 7.59 (dd,
J = 7.7, 1.8 Hz, 1H), 7.55 (dd, J = 8.0, 1.2 Hz, 1H), 7.32 (dt,
J = 7.5, 1.2 Hz, 1H), 7.20 (dt, J = 7.9, 1.8 Hz, 1H), 6.09 (s,
1H), 4.16–4.09 (m, 2H), 4.07–4.01 (m, 2H); dC (125 MHz, CDCl3)
136.50, 132.84, 130.50, 127.71, 127.30, 122.81, 102.47, 65.35; m/z
(EI) 227.9788 (C9H9BrO2 requires 227.9796).
˚
carbene. An X-ray-determined distance of 3.85 A between the
prospective carbene carbon and the closest oxygen atom in the
crystals of 1a (Ar2(N2)C–O) suggests that formation of ylide or
ylidic complexes probably requires rotation of the dioxolane group
with respect to the aromatic ring that bears the two groups. While
it is likely that both oxonium ylide and carbene are present in the
crystal, the only product observed at low (<20%) conversions is
benzocyclobutane 10, which probably originates from the triplet
carbene precursor. The structure of 1a, with a diazo carbon–acetal
˚
hydrogen C–H distance of 2.49 A, should facilitate hydrogen
abstraction by the triplet carbene followed by ring closure to
generate 10 in a topochemically allowed reaction.30 It is well known
that reactions in crystals are determined by the size and shape of
the reaction cavity rather than by the intrinsic energetics of the
participant species.31 While formation of the ylide intermediate
may be within reach at early stages of the solid-to-solid conversion,
it is likely that 1,2-migrations towards 8 and 9 may be sterically
too demanding. The fact that these compounds are formed upon
extended irradiation suggests that they form at defect sites.
(o-(1,3-Dioxolan-2-yl)phenyl)(4-biphenylyl)methanol 3
Conclusions
Bromoacetal 2 (5.2 g, 22.70 mmol) was added to a 250 mL
round-bottomed flask equipped with a magnetic stir bar. The flask
was purged with argon for several minutes. Dry tetrahydrofuran
(THF) was added to the acetal and then the solution was cooled
to -78 ◦C in a dry ice–acetone bath. To the cooled solution at
-78 ◦C, n-BuLi (15.6 mL, 1.6 M in hexanes) was added dropwise
and then the reaction mixture was stirred for 1 h. A solution
of 4-biphenylcarboxaldehyde (4.34 g, 23.82 mmol) in 50 mL of
dry THF was then added slowly and the reaction mixture stirred
overnight. Column chromatography of the crude reaction mixture
on silica gel (4 : 1 hexanes–ethyl acetates, v/v) gave 6.6 g (88%) of a
viscous yellow oil, which crystallized into a pale yellow solid over
We have shown that suitably substituted carbene precursors may
give rise to detectable, formally ionic oxonium ylide intermediates
in non-polar organic glasses and in crystalline solids. Ruby-
red crystals of diazo precursor 1a were solved in the triclinic
¯
space group P1. The generation of oxonium ylide is supported
by a strong UV absorption band with lmax= 490 nm at 77 K,
and by rearrangement products in solution with a connectivity
that derives from the expected O–C interaction, including ortho-
quinodimethane 6, dioxabicyclo[3.2.1.]octane 8, and 1,4-dioxane
9. While detection of the ylide and the carbene suggest that
reactions in crystals may occur under a pre-equilibrium, the
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 1106–1114 | 1111
©

View Article Online
two days; nmax/cm-1 (neat) 3461 s (OH), 3029, 2974, 2889, 1599,
1449, 1409, 1387, 1351, 1221, 1108, 1072, 948, 768; dH (500 MHz,
(CD3)2CO) 7.64–7.62 (m, 2H), 7.60–7.57 (m, 3H), 7.51–7.48 (m,
3H), 7.44–7.41 (m, 2H), 7.37–7.31 (m, 2H), 7.31–7.26 (m, 1H), 6.36
(d, J = 4.3 Hz, 1H), 6.08 (s, 1H), 4.84 (d, J = 4.3 Hz, 1H), 4.14–
4.08 (m, 2H), 4.05–3.97 (m, 2H); dC (125 MHz, (CD3)2CO) 144.76,
144.61, 141.60, 140.19, 135.71, 129.79, 129.62, 128.66, 128.08,
127.98, 127.73, 127.59, 127.27, 126.91, 102.07, 70.83, 65.78, 65.77;
m/z (EI) 332.1420(C22H20O3 requires 332.1412).
101.62, 65.51, 65.40; m/z (EI) 344.1523 (C22H20N2O2 requires
344.1525).
(o-(1,3-Dioxolan-2-yl)phenyl)(4-biphenylyl)diazomethane 1a
Hydrazone 5 (1 g, 2.90 mmol) was dissolved in 200 mL of anhy-
drous diethyl ether in a 500 mL round-bottomed flask equipped
with a magnetic stir bar, to which anhydrous MgSO4 (2.58 g,
21.43 mmol) and yellow mercury(II) oxide (7 g, 32.32 mmol)
were added to the stirring solution. A freshly prepared saturated
solution of potassium hydroxide in absolute ethanol (20–30 drops)
was added to initiate the oxidation of the hydrazone to the diazo
compound. After the first 10 drops of the saturate KOH–EtOH
solution, formation of a light pink color and darkening of the
mercury(II)oxide were observed. NOTE: Addition of 1 mL of
anhydrous hydrazine to the reaction was sometimes necessary to
initiate the oxidation. However, major darkening of the mercury(II)
oxide was observed, to which an additional 2–3 spatula tips of fresh
mercury(II) oxide were added. The reaction flask was wrapped with
aluminium foil and stirred vigorously under an argon atmosphere
at room temperature. After 8–10 hours of stirring under argon,
complete consumption of the starting hydrazone was observed
by TLC (2 : 1 hexanes–ethyl acetate, v/v). The reaction mixture
was suction-filtered through a Hirsch funnel fitted with a glass
microfibre (Whatman, 7.0 cm, 100 circles, Cat. No. 1820070) filter
paper. The solution was concentrated by a flow of dry argon to give
dark violet plates in near quantitative yields. nmax/cm-1 (KBr) 3058,
3029, 2953, 2887, 2043, 1666, 1601, 1519, 1487, 1449, 1394, 1315,
1206, 1106, 1072, 972, 944, 833, 763; dH (500 MHz, CDCl3) 7.78–
7.74 (m, 1H), 7.58–7.53 (m, 4H), 7.48–7.47 (m, 3H), 7.43–7.40 (m,
2H), 7.33–7.29 (m, 1H), 6.98–6.96 (m, 2H), 5.88 (s, 1H), 4.13–4.08
(m, 2H), 4.00–3.95 (m, 2H); dC (125 MHz, CDCl3) 140.55, 138.05,
136.72, 131.90, 130.79, 130.00, 129.23, 128.78, 128.74, 127.62,
127.60, 127.22, 127.03, 126.66, 122.60, 101.60, 65.49.
(o-(1,3-Dioxolan-2-yl)phenyl)(4-biphenylyl)methanone 4
Alcohol 3 (6.1 g, 18.35 mmol) was completely oxidized as reported
by Swern33 to yield 5.8 g (97%) of the ketone as a light pale yellow
crystalline solid. TLC (3 : 1 hexane–ethyl acetates, v/v) showed
complete oxidation of the starting alcohol and 1H and 13C NMR
spectra in CDCl3 showed >95% purity. A portion of the crude
ketone was purified by column chromatography on silica gel (3
: 1 hexanes–ethyl acetates, v/v). From the pure fractions after
1 day, large clear colorless plate-like needles grew. The colorless
crystals were discovered to turn bright yellow upon exposure to
365 nm light (UV lamp) and reverted to the colorless state within
less than one minute when the UV lamp was removed. nmax/cm-1
=
(KBr) 3030, 2970, 2900, 1659 (C O), 1604, 1580, 1484, 1447, 1402,
1307, 1280, 1266, 1214, 1113, 1066, 969, 932, 851, 794, 756, 702;
dH (500 MHz, CDCl3) 7.87 (d, J = 8.4 Hz, 2H), 7.71 (d, J =
7.7 Hz, 1H), 7.66 (d, J = 8.4 Hz, 2H), 7.63–7.61 (m, 2H), 7.52 (dt,
J = 7.7, 1.3 Hz, 1H), 7.47–7.34 (m, 5H), 6.05 (s, 1H), 3.93–3.86
=
(m, 4H); dC (125 MHz, CDCl3) 197.04 (C O), 145.69, 139.80,
138.52, 136.80, 136.24, 130.60, 129.89, 128.88, 128.38, 128.17,
127.95, 127.22, 127.01, 126.94, 101.45, 65.10; m/z (EI) 330.1256
(C22H18O3 requires 330.1256).
(o-(1,3-Dioxolan-2-yl)phenyl)(4-biphenylyl)methanone
hydrazone 5
General procedure and setup for photolysis in solution
Ketone 4 (2.03 g, 6.14 mmol) and 50 mL of absolute ethanol were
added to a three-necked round-bottomed flask equipped with a stir
bar, Dean–Stark trap, and a water condenser. The suspension was
heated to ca. 110 ◦C to dissolve all the ketone. To the refluxing
reaction mixture, anhydrous hydrazine (7.5 mL, 0.239 mol) was
added in one portion and the reaction was refluxed overnight
(ca. 16 h), followed by an additional 24 h of refluxing. TLC (3 :
1 hexanes–ethyl acetate, v/v) showed four new spots, two above
and two below the ketone. Excess ethanol and hydrazine were
evaporated from the crude reaction in a rotary evaporator. The
viscous organic layer was dissolved in 50 mL of diethyl ether,
washed with DI water (3 ¥ 50 mL), dried over anhydrous MgSO4,
filtered, and concentrated in vacuo to give a cloudy pale yellow
viscous oil, which was purified by column chromatography on
silica gel (3 : 1 hexanes–ethyl acetates, v/v) to yield 1.068 g (51%)
of pure hydrazone as a pale yellow viscous oil. nmax/cm-1 (KBr)
3404, 3290, 3029, 2888, 1576, 1487, 1397, 1331, 1270, 1211, 1114,
1070, 943, 846, 767, 731; dH (500 MHz, CDCl3) 7.79–7.77 (m,
1H), 7.59–7.56 (m, 2H), 7.55–7.50 (m, 6H), 7.43–7.39 (m, 2H),
7.33–7.30 (m, 1H), 7.19–7.17 (m, 1H), 5.62 (s, 1H), 5.42 (br s,
2H, NH2), 4.14–4.00 (m, 2H), 3.93–3.87 (m, 2H); dC (125 MHz,
CDCl3) 147.71, 140.67, 140.62, 137.18, 136.41, 132.71, 130.63,
129.36, 128.99, 128.71, 127.47, 127.26, 126.91, 126.83, 126.38,
Approximately 10 mg of pure diazo was dissolved in ca. 1 mL
of deuterated benzene. The solution was placed in an NMR
tube and argon was slowly bubbled through the solution for
30 minutes to deoxygenate the sample. To the NMR tube, 1,1,2,2-
tetrachloroethane (1 : 1 mole ratio of diazo to standard) was
added to monitor the extent of conversion of starting material
to products. Photolysis was carried out in a photolysis chamber
(400 W mercury-vapor lamp) at room temperature with a 380 nm
cutoff filter. The photolysis was monitored by 1H NMR every hour
(total of ca. 6 h) until complete disappearance of the diazo was
observed. The combined products were isolated by preparatory
thin layer chromatography using a mixture of methylene chloride
and hexanes (3 : 1 methylene chloride–hexanes, v/v).
Solid-state photolysis
Crystals of diazo 1a (ca. 150 mg) were ground into a fine powder
and placed into a Petri dish. A 380 nm cutoff filter was placed
on top of the sample, which was placed over an aluminium block
with an extension immersed ◦in a large ice–water reservoir to help
maintain the sample near 0 C. The sample was exposed to UV
light from a 400 W mercury vapor lamp kept in a jacket with
◦
circulating water kept at 0 C. Reaction progress was monitored
1112 | Org. Biomol. Chem., 2009, 7, 1106–1114
This journal is
The Royal Society of Chemistry 2009
©

View Article Online
by 1H NMR of samples taken at various intervals until complete
disappearance of the starting material was observed.
˚
with a Mo-target X-ray tube (l = 0.71073 A) operated at 2250 W
power. The detector was placed at a distance of 4.986 cm from
the crystal. A total of 1321 frames were collected with a scan
width of 0.3◦ in w, with an exposure time of 30 s/frame. The
total data collection time was ca. 18 h. The frames were integrated
with the Bruker SAINT software package using a narrow-frame
integration algorithm. Analysis of the data showed negligible
decay during the data collection. The structure was refined based
on the respective space groups, using the Bruker SHELXTL
(Version 5.3) Software Package.
(1Z,6Z)-1-(Biphenyl-4-yl)-3,4-dihydrobenzo[f ][1,4]dioxocine 9
(o-quinodimethane 6)
Isolated in 15% yield. nmax/cm-1 (neat) 3029, 2959, 2924, 2854,
1726 (w), 1629, 1598, 1521, 1488, 1405, 1345, 1280, 1262, 1067,
930, 844, 767; dH (500 MHz, C6D6) 7.81–7.79 (m, 2H), 7.44–7.42
(m, 4H), 7.22–7.19 (m, 3H), 6.73 (d, J = 9.3 Hz, 1H), 6.16 (s, 1H),
5.88–5.85 (m, 1H), 5.83–5.80 (m, 1H), 5.66 (d, J = 9.1 Hz, 1H),
3.77 (ddd, J = 8.6, 7.5, 1.0 Hz, 1H), 3.69 (ddd, J = 11.6, 8.6, 5.4,
1H), 2.07 (ddd, J = 12.3, 5.4, 1.0 Hz, 1H), 1.44 (ddd, J = 12.3,
11.6, 7.5 Hz, 1H); dC (125 MHz, CDCl3) 150.24, 141.49, 140.38,
129.47, 128.84, 128.55, 127.60, 127.37, 127.04, 127.02, 123.41,
123.17, 120.92, 114.56, 112.73, 66.22, 39.08 (Note: one carbon
could not be resolved); m/z (EI) : 314.1309 (314.1307 requires
C22H18O2).
Compound 1a (X-ray). Ruby-red plates were grown from
¯
diethyl ether, MW = 342.38, triclinic, space group P1, a =
˚
˚
˚
6.0742(10) A, b = 11.4003(18) A, c = 12.3098(19) A, a =
93.982(3)◦, b = 93.601(3)◦, g = 90.252(3)◦, Z = 2, rcalcd = 1.340 mg
-3
-1
˚
m , F(000) = 360, l = 0.71073 A, m(Mo Ka) = 0.087 mm ,
T = 100(2) K, crystal size = 0.4 ¥ 0.35 ¥ 0.2 mm3. Of the 5564
reflections collected (1.66 ◦ ≤ q ≤ 28.29 ◦), 3874 [R(int) = 0.0138]
were independent reflections; max./min. residual electron density
311 and -201 e nm-3, R1 = 0.0398 [I > 2s(I)], and wR2 = 0.1185
(all data).
cis-1-(Biphenyl-4-yl)-benzo[g]2,5-dioxabicyclo[4.2.0]octane 9
(cis-1,4-dioxane 9)
Isolated in 43% yield. nmax/cm-1 (KBr) 3055, 3028, 2957, 2921,
2858, 1458, 1400, 1337, 1280, 1266, 1195, 1150, 1110, 1092, 1034,
823, 756; dH (500 MHz, CDCl3) 7.59–7.57 (m, 4H), 7.49–7.40 (m,
7H), 7.37–7.32 (m, 2H), 5.34 (s, 1H), 4.05–3.99 (m, 1H), 3.90–3.80
(m, 3H); dC (125 MHz, CDCl3) 147.09, 145.49, 140.75, 140.56,
140.39, 130.41, 130.03, 128.73, 127.28, 127.11, 127.08, 126.25,
124.41, 122.90, 83.26, 80.38, 61.66, 60.61; m/z (EI) 314.1309
(C22H18O2 requires 314.1307).
Compound 8 (X-ray). Colorless plate-like needles grown from
dichloromethane, MW = 314.36, monoclinic, space group P21/c,
˚
˚
˚
a = 8.8692(18) A, b = 7.5279(15) A, c = 23.237(5) A, b =
97.064(4)◦, Z = 4, rcalcd = 1.356 mg m-3, F(000) = 664, l =
-1
˚
0.71073 A, m(Mo Ka) = 0.086 mm , T = 100(2) K, crystal size =
0.3 ¥ 0.08 ¥ 0.05 mm3. Of the 5448 reflections collected (1.77 ◦
≤
q ≤ 24.71 ◦), 2207 [R(int) = 0.0575] were independent reflections;
max./min. residual electron density 177 and -192 e nm-3, R1 =
0.0416 [I > 2s(I)], and wR2 = 0.0795 (all data).
5-(Biphenyl-4-yl)-benzo[f ]2,8-dioxabicyclo[3.2.1]octane 8
Isolated in 15% yield. nmax/cm-1 (KBr) 3062, 3018, 2965, 2921,
2850, 1560, 1489, 1463, 1401, 1108, 1086, 966, 904, 749; dH
(500 MHz, CDCl3) 7.65–7.55 (m, 6H), 7.45–7.26 (m, 5H), 7.03 (d,
J = 7.3 Hz, 2H), 6.39 (s, 1H), 3.89 (ddd, J = 11.5, 12, 1.0 Hz, 1H),
3.40 (ddd, J = 17, 12, 4 Hz, 1H), 2.77 (ddd, J = 17, 13, 12 Hz, 1H),
2.07 (ddd, J = 13, 4.0, 1.0 Hz, 1H); dC (125 MHz, CDCl3) 145.72,
140.95, 140.68, 139.62, 138.67, 128.91, 128.78, 128.13, 127.40,
127.20, 127.13, 125.88, 121.28, 120.19, 101.01, 85.74, 59.45, 35.70;
m/z (tof) 314.1264 (C22H18O2 requires 314.1307).
Notes and references
1 (a) Reviews of Reactive Intermediates, ed. M. S. Platz, R. A. Moss
and M. Jones, John Wiley and Sons, Hoboken, NJ, 2007; (b) Reactive
Intermediate Chemistry, ed. R. A. Moss, M. S. Platz and M. Jones,
John Wiley and Sons, Hoboken, NJ, 2004.
2 (a) T. Choi, K. Peterfy, S. I. Khan and M. A. Garcia-Garibay, J. Am.
Chem. Soc., 1996, 118, 12477–12478; (b) K. Peterfy and M. A. Garcia-
Garibay, J. Am. Chem. Soc., 1998, 120, 4540–4541; (c) Z. Yang and
M. A. Garcia-Garibay, Org. Lett., 2000, 2, 1963–1965; (d) Z. Yang,
D. Ng and M. A. Garcia-Garibay, J. Org. Chem., 2001, 66, 4468–4475;
(e) D. Ng, Z. Yang and M. A. Garcia-Garibay, Tetrahedron Lett., 2001,
42, 9113–9116; (f) L. M. Campos, H. Dang, D. Ng, Z. Yang and M.
A. Garcia-Garibay, J. Org. Chem., 2002, 67, 3749–3754; (g) D. Ng, Z.
Yang and M. A. Garcia-Garibay, Tetrahedron Lett., 2002, 43, 7063–
7066; (h) M. E. Ellison, D. Ng, H. Dang and M. A. Garcia-Garibay,
Org. Lett., 2003, 5, 2531–2534.
3 (a) S. H. Shin, A. E. Keating and M. A. Garcia-Garibay, J. Am. Chem
Soc., 1996, 118, 7626–7627; (b) S. H. Shin, D. Cizmeciyan, A. E.
Keating, S. I. Khan and M. A. Garcia-Garibay, J. Am. Chem. Soc.,
1997, 119, 1859–1868; (c) C. N. Sanrame, C. P. Suhrada, H. Dang and
M. A. Garcia-Garibay, J. Phys. Chem. A, 2003, 107, 3287–3294.
4 M. A. Garcia-Garibay, Acc. Chem. Res., 2003, 36, 491–498.
5 (a) M. A. Garcia-Garibay, A. E. Constable, J. Jernelius, T. Choi,
D. Cizmeciyan and S. H. Shin, Physical Supramolecular Chemistry,
Kluwer Academic Publishers, Dordrecht, 1996; (b) A. E. Keating, M.
A. Garcia-Garibay, in Molecular and Supramolecular Photochemistry,
ed. V. Ramamurthy and K. Schanze, Marcel Dekker, New York, 1998,
vol. 2, pp. 195–248.
2-(Biphenyl-4-yl)-2H-spiro[cyclobutabenzene-1,2¢-
[1,3]dioxolane] 10
Isolated in 75% yield. dH (500 MHz, CDCl3) 7.59 (d, J = 7.3 Hz,
2H), 7.55 (d, J = 8.1 Hz, 2H), 7.48–7.30 (m, 7H), 7.24 (d, J =
8.2 Hz, 2H), 4.97 (s, 1H), 4.19–4.11 (m, 2H), 4.03–3.99 (m, 1H),
3.70 (dd, J = 14.4, 7.0 Hz, 1H); dC (125 MHz, CDCl3) 145.36,
144.34, 140.95, 139.66, 137.55, 131.35, 128.84, 128.70, 128.57,
127.11, 127.02, 126.94, 123.69, 122.21, 111.25, 64.95, 64.68,
64.21; m/z (EI) 314.1337 (C22H18O2 requires 314.1307). The acetal
was converted to the corresponding benzocyclobutenone by acid
-1 16
hydrolysis. nmax/cm-1 (neat) 1763 (s C O), (lit C O 1760 cm );
=
=
m/z (EI) 270.1039 (C20H14O requires 270.1045).
6 (a) S. H. Shin, A. E. Keating and M. A. Garcia-Garibay, J. Am. Chem
Soc., 1996, 118, 7626–7627; (b) S. H. Shin, D. Cizmeciyan, A. E.
Keating, S. I. Khan and M. A. Garcia-Garibay, J. Am. Chem. Soc.,
1997, 119, 1859–1868.
7 M. A. Garcia-Garibay, S. Shin and C. N. Sanrame, Tetrahedron, 2000,
56, 6729–6737.
X-Ray diffraction
The X-ray intensity data were measured at 100 K on a Bruker
SMART 1000 CCD-based X-ray diffractometer system equipped
This journal is
The Royal Society of Chemistry 2009
Org. Biomol. Chem., 2009, 7, 1106–1114 | 1113
©

View Article Online
8 (a) H. Tomioka, N. Hayashi, Y. Izawa, V. P. Senthilnathan and M. S.
Platz, J. Am. Chem. Soc., 1983, 105, 5053–5057; (b) M. A. Garcia-
Garibay, J. Am. Chem. Soc., 1993, 115, 7011–7012; (c) M. A. Garcia-
Garibay, C. Theroff, S. H. Shin and J. Jernelius, Tetrahedron Lett., 1993,
52, 8415–8418.
9 L. M. Campos and M. A. Garcia-Garibay, Reactive Intermediates
in Crystals: Form and Function, in Reviews of Reactive Intermediate
Chemistry, ed. M. S. Platz, M. Jones, R. Moss, Wiley-Interscience,
Hoboken, NJ, 2007.
10 (a) A. Padwa and S. F. Hornbuckle, Chem. Rev., 1991, 91, 263–309;
(b) J. S. Clark, Nitrogen, Oxygen and Sulfur Ylide Chemistry: A Practical
Approach in Chemistry, Oxford University Press, Oxford, 2002.
11 (a) Oxonium ylides and synthesis: X. Guo, H. Huang, L. Yang and W.
Hu, Org. Lett., 2007, 9, 4721–4723; (b) G. K. Murphy, F. P. Marmsaeter
and F. G. West, Can. J. Chem., 2006, 84; (c) J. M. Mej´ıa-Oneto and A.
Padwa, Org. Lett., 2006, 8, 3275–3278; (d) F. P. Marmsaeter, G. K.
Murphy and F. G. West, J. Am. Chem. Soc., 2003, 126, 14724–14725;
(e) A. Oku, N. Murai and J. Baird, J. Org. Chem., 1997, 62, 2123–
2129.
20 M. S. Platz, in The Chemistry, Kinetics, and Mechanisms of Triplet
Carbene Processes in Low Temperature Glasses and Solids, ed. M. S.
Platz, Plennum Press, New York, 1990, pp. 143–212.
21 (a) Product analysis: M. Jones, Jr. and R. T. Ruck, Tetrahedron Lett.,
1998, 39, 2277–2280; (b) T. Sueda, T. Nagaoka, S. Goto and M.
Ochiai, J. Am. Chem. Soc., 1996, 118, 10141–10149; (c) H. Tomioka,
N. Kobayashi, S. Murata and Y. Ohtawa, J. Am. Chem. Soc., 1991, 113,
8771–8778.
22 (a) S. Celebi, M. -L. Tsao and M. S. Platz, J. Phys. Chem. A, 2001,
105, 1158–1162; (b) J. -L. Wang, T. Yuzawa, M. Nigam, I. Likhotvorik
and M. S. Platz, J. Phys. Chem. A, 2001, 105, 3752–3756; (c) E. M.
Tippman, M. S. Platz, I. B. Svir and O. V. Klymenko, J. Am. Chem.
Soc., 2004, 126, 5750–5762.
23 R. A. Moss, J. Tiang, R. R. Sauers and K. Krogh-Jespersen, J. Am.
Chem. Soc., 2007, 129, 10019–10028.
24 I. Naito, A. Oku, N. Otani, Y. Fujiwara and Y. Tanimoto, J. Chem.
Soc., Perkin Trans. 2, 1996, 725–729.
25 R. A. Moss, J. Tian, R. R. Sauers and K. Krogh-Jespersen, J. Am.
Chem. Soc., 2007, 129, 10019.
12 W. Kirmse and K. Kund, J. Am. Chem. Soc., 1989, 111, 1465–
26 K. B. Eisenthal, R. A. Moss and N. J. Turro, Science, 1984, 225, 1439–
1473.
1445.
13 (a) K. B. Eisenthal, N. J. Turro, E. V. Sitzmann, I. R. Gould, G.
Hefferon, J. Langan and Y. Cha, Tetrahedron, 1985, 41, 1543; (b) D.
Griller, A. S. Nazran and J. C. Scaiano, Tetrahedron, 1985, 41, 1525.
14 R. A. Moss, L. Wang, E. Weintraub and K. Krogh-Jespersen, J. Am.
Chem. Soc., 2008, 130, 4651–4659.
15 (a) E. Migirdicyan and J. Baudet, J. Am. Chem. Soc., 1975, 97, 7400–
7404; (b) C. R. Flynn and J. Michl, J. Am. Chem. Soc., 1974, 96, 3280–
3288.
16 (a) R. V. Stevens and G. S. Bisacchi, J. Org. Chem., 1982, 47,
2396–2399; (b) R. V. Stevens and G. S. Bisacchi, J. Org. Chem., 1982,
47, 2393–2396.
17 (a) A. M. Trozzolo and W. A. Gibbons, J. Am. Chem. Soc., 1967, 89,
239–243; (b) J. C. Scaiano, Solution Photochemistry of Carbenes and
Biradicals, ed. M. S. Platz, Plenum Press, New York, 1990, pp. 353–
368.
27 (a) W. D. Ollis, M. Rey, I. O. Sutherland and G. L. Closs, J. Chem. Soc.,
Chem. Commun., 1975, 543; (b) H. Iwamura, M. Iwamura, T. Nishida,
M. Yoshida and J. Nakayama, Tetrahedron Lett., 1971, 63.
28 W. Kirmse, Angew. Chem., Int. Ed, 2003, 42, 2117–2119.
29 (a) H. Tomioka, H. Okada, T. Watanabe and K. Hirai, Angew. Chem.,
Int. Ed. Engl., 1994, 33, 873–875; (b) H. Tomioka, H. Okada, T.
Watanabe, K. Banno, K. Komatsu and K. Hirai, J. Am. Chem. Soc.,
1997, 119, 1582–1593; (c) Y. -M. Hu, K. Hirai and H. Tomioka, J.
Phys. Chem. A, 1999, 103, 9280–9284; (d) K. Hirai, K. Yasuda and
H. Tomioka, Chem. Lett., 2000, 94–95; (e) Y. -M. Hu, Y. Ishikawa, K.
Hirai and H. Tomioka, Bull. Chem. Soc. Jpn., 2001, 74, 2207–2218;
(f) K. Hirai, T. Iikubo and H. Tomioka, Chem. Lett., 2002, 1226–1227;
(g) T. Iikubo, K. Hirai and H. Tomioka, Org. Lett., 2002, 4, 2261–2264;
(h) T. Koshiyama, K. Hirai and H. Tomioka, J. Phys. Chem. A, 2002,
106, 10261–10274.
18 J. C. Scaiano and D. Weir, Can. J. Chem., 1988, 66, 491–494.
19 (a) R. W. Brandon, G. L. Closs, C. E. Davoust, C. A. Hutchison, Jr., B.
E. Kohler and R. Silbey, J. Chem. Phys., 1965, 43, 2006; (b) C. Cheng,
T. -S. Lin and D. J. Sloop, Chem. Phys. Lett., 1976, 44, 576; (c) D.
C. Doetschman and C. A. Hutchison, J. Chem. Phys., 1972, 56, 3964;
(d) D. J. Graham and T. -S. Lin, Chem. Phys. Lett., 1982, 73, 411; (e) A.
M. Trozzolo, E. Wasserman and W. A. Yager, J. Chim. Phys., 1964,
61, 1663; (f) R. J. M. Anderson, B. E. Kohler and J. M. Stevenson, J.
Chem. Phys., 1979, 71, 1559; (g) H. Sixl, R. Mathes, A. Schaupp and
K. Ulrich, Chem. Phys., 1986, 107, 105.
30 (a) G. M. J. Schmidt, Solid State Photochemistry, Verlag Chemie, New
York, 1976; (b) G. J. M. Schmidt, Pure Appl. Chem., 1971, 27, 647.
31 M. D. Cohen, Angew. Chem., Int. Ed. Engl, 1975, 14, 386.
32 (a) G. D. Hartman, B. T. Phillips and W. Halczenko, J. Org. Chem., 1985,
50, 2423–2427; (b) G. D. Hartman, W. Halczenko and B. T. Phillips, J.
Org. Chem., 1985, 50, 2427–2431.
33 (a) A. J. Mancuso and D. Swern, Synthesis, 1981, 165–185; (b) A. J.
Mancuso, D. S. Brownfain and D. Swern, J. Org. Chem., 1979, 44,
4148–4150; (c) A. J. Mancuso, S. -L. Huang and D. Swern, J. Org.
Chem., 1978, 43, 2480–2482.
1114 | Org. Biomol. Chem., 2009, 7, 1106–1114
This journal is
The Royal Society of Chemistry 2009
©