Medium effects on Zwitterionic-biradicaloid intermediates from two phenyl-α-oxoamides. Irradiations in fluid and solid protic media, neat solid phases, and the solid, smectic and isotropic phases of a completely saturated phosphonium salt at different tem

Article abstract of DOI:10.1562/2006-07-28-RA-988

The photochemical processes of two N,N-dialkyl phenyl-α-oxoamides, N,N-diisopropyl phenyl-α-oxoamide (1) and N,N-dibenzyl phenyl-α-oxoamide (2), are investigated at different temperatures in methanol and ethylene glycol (to probe the influences of H-bondi

Full text of DOI:10.1562/2006-07-28-RA-988

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Photochemistry and Photobiology, 2007, 83: 570–583
Medium Effects on Zwitterionic-Biradicaloid Intermediates from Two
Phenyl–a-oxoamides. Irradiations in Fluid and Solid Protic Media, Neat
Solid Phases, and the Solid, Smectic and Isotropic Phases of a Completely
Saturated Phosphonium Salt at Different Temperatures†
Carlos A. Chesta‡*, Mathew George§, Chuping Luo§ and Richard G. Weiss*
Department of Chemistry, Georgetown University, Washington, DC
Received 28 July 2006; accepted 26 September 2006; published online 28 September 2006; DOI: 10.1562 ⁄ 2006-07-28-RA-988
and other a-diketones react efficiently with triplet molecular
oxygen by mechanisms that depend on the medium (7–10).
ABSTRACT
The photochemical processes of two N,N-dialkyl phenyl-a-oxo-
amides, N,N-diisopropyl phenyl-a-oxoamide (1) and N,N-dib-
enzyl phenyl-a-oxoamide (2), are investigated at different tem-
peratures in methanol and ethylene glycol (to probe the influences
of H-bonding and viscosity), in the solid phase of D-sorbitol at
room temperature (to compare with the results in the liquid
alcohols and to assess the influence of a poorly organized ‘‘stiff’’
environment), in the neat solid phase (to probe the influence of
well-ordered, ‘‘stiff’’ matrices), and in the solid, smectic A2 and
isotropic phases of methyl-tris-tetradecylphosphonium tetrafluo-
roborate (1P14BF4) (to assess the ability of the intermediates to
respond to subtle changes in the order and polarity of their local
environments). From differences between the activation param-
eters for product pathways from irradiations in methanol and in
1P14BF4, we conclude that the zwitterionic pre-product inter-
mediate from 1 is much more sensitive to the polarity, viscosity
and order of its local environment than is the isomeric pre-
product biradicaloid intermediate or either of the pre-product
intermediates from 2. A very sensitive balance among the
medium parameters, as well as internal steric and electronic
factors of 1 and 2, controls the reactive pathways of the
photochemically generated intermediates.
Aryl a-oxoamides which are constrained to cisoid confor-
mations undergo electron and rapid subsequent intermolecu-
lar proton transfer from their excited triplet states (11). Aryl
a-oxoamides capable of adopting transoid conformations
undergo photoreaction efficiently through a complicated
excited singlet-state mechanism in which products are derived
in parallel from both biradicaloid and zwitterionic interme-
diates, some of which interconvert via large conformational
changes (Scheme 1) (12–18). Because it is known that the
reaction intermediates are too short-lived to diffuse appreci-
ably before yielding products (18), the distribution of the
photoproducts derived from them should be very sensitive to
the polarity and microviscosity of the local environments
where the reactant molecules undergo excitation; the location
of the a-oxoamides at the moment of their excitation
determines primarily where they react. Some longer lived
intermediates are produced upon excitation of some aryloxy-
methyl a-oxoamides via mechanisms that are not available to
aryl a-oxoamides (19).
In this study, we report the influences of both temperature
and solvent phase (and, therefore, local polarity, viscosity and
ordering) on the photoproduct selectivities of two phenyl
a-oxoamides, 1 and 2 (Scheme 2), which have been shown to
follow rather different reaction pathways when irradiated in
isotropic media despite their close structural relationship (12–
14). Our investigations exploit the mechanistic nuances
between them to explore the reaction cavities (20) in very
different anisotropic environments. Thus, irradiations have
been conducted in the isotropic protic and polar solvents,
methanol (18) and ethylene glycol, as a function of tempera-
ture to examine the influence of relaxation rates by the
molecules constituting the reaction cavities, and in the aniso-
tropic protic and polar solid matrix afforded by D-sorbitol
(also known as D-glucitol, the alcohol derived from reduction
of D-glucose; Scheme 2) to determine the effects of irradiation
when the reaction cavities are virtually unable to relax during
the periods of the molecular transformations. Although
D-sorbitol is polymorphous (21), the environment provided
to 1 and 2 by it must be similar to that in any other
immobilized polyol, and we consider it to be a model for ‘‘solid
methanol’’ at room temperature. In addition, irradiations in
INTRODUCTION
The photophysics and photochemistry of molecules containing
a-diketo moieties have received a great deal of attention both
recently and historically (1–3) as a result of several interesting
structural (N.B. their conformationally dependent photoreac-
tivity [1–5]) and electronic (N.B. their absorption spectra
which extend into the visible region) features. The simplest a-
diketone, biacetyl, fluoresces and phosphoresces at room
temperature in fluid solutions (6) and the excited states of it
†This paper is part of a symposium-in-print dedicated to Professor Eduardo A.
Lissi on the occasion of his 70th birthday.
‡Permanent address: Departamento de Quımica, Universidad Nacional de Rıo
Cuarto, 5800 Rıo Cuarto, Argentina.
§Current address: Sepax Technologies Inc., Newark, DE 19711, USA.
*Corresponding author email: weissr@georgetown.edu,
cchesta@exa.unrc.edu.ar (Richard G. Weiss and Carlos A. Chesta)
ꢀ 2007 American Society for Photobiology 0031-8655/07
570

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Photochemistry and Photobiology, 2007, 83 571
R2
R1
R1
O
O
1: R1 = R2 = CH3
2: R1 = H, R2 = Ph
R2
N
Ph
hν
R1
R1
OH
OH
R2
R2
R1
R1
R2
R1
R2
OH
[C-A]
[C-O]
O
Ph
Ph
R2
R2
N
O
N
N
Ph
Ph
N
O
R2
R1
O
O
R2
O
R1
R1
Z2
Ds
H+, H2O
A
O
[E]
R1
R2
R1
OH
+
OH
N
H
Ph
C
Ph
CHO
R2
N
R2
O
I
Ph
K
O
R1
H2O
+
XH
MA
O
H2N
R1
OH
X = -OH (PG-A)
= -OCH3 (MM)
= -NR1R2 (MA)
R1
X
R2
R2
Ph
O
Scheme 1
High performance liquid chromatography (HPLC) analyses were
performed on an HP Series 1100 chromatograph with a UV–Vis
detector and a reverse-phase column (Kromasil C18, Lumex, 5 lm,
250 · 2.1 mm). 40 ⁄ 60 and 50 ⁄ 50 acetonitrile ⁄ water mixtures were used
as eluents for 1 and 2, respectively. Chromatograms were recorded at
five different wavelengths between 220 and 350 nm, but only the one
obtained at 220 nm was used for quantitative analyses as it provided
the overall highest detection sensitivity of the photoproducts.
X-ray diffraction (XRD) of the samples in thin, sealed capillaries
(0.5-mm diam; W. Muller, Schonwalde, Germany) was performed on a
Rigaku R ⁄ AXIS image plate system with Cu Ka X-rays
(k = 1.54056 A) generated by a Rigaku generator operating at 46 kV
and 46 mA. Data were processed and analyzed with Material Data
JADE (version 5) XRD pattern processing software. Differential
scanning calorimetry (DSC) was performed using a TA 2910 DSC cell
base interfaced to a TA Thermal Analyst 3100 controller. DSC
measurements of 1 in D-sorbitol were carried out in open aluminum
pans and all other measurements in crimped, closed pans.
the neat, solid phase of 1 at different temperatures have been
reinvestigated (17) to compare the influence of its well-ordered
and less-polar reaction cavities with those from D-sorbitol, and
in the solid, smectic liquid-crystalline and isotropic phases of
the completely saturated ionic liquid crystal (22), methyl-tri-
tetradecylphosphonium
tetrafluoroborate
(1P14BF4)
(Scheme 2), a medium which possesses regions of very low
and very high polarity (23,24). We show that the principal
intermediates, especially Z2 (from 1), respond very differently
to a common environment or temperature and that the
intermediates from 1 are much more sensitive to environ-
mental changes than those from 2. The results demonstrate
that the inter-conversions of the intermediates from 1 are
balanced delicately by several factors, such as temperature, H-
bonding and local viscosity. The effects observed on the
photochemistries of 1 and 2 cannot be attributed to any one of
these factors alone. They also indicate that the photochemis-
tries of 1 and 2, especially in tandem, are useful supplements to
and may be employed as substitutes for spectroscopic probes
of solvent polarity, viscosity and ordering.
Materials. Chloroform (Fisher, HPLC grade), methanol (EM
Science, HPLC grade), ethylene glycol (Mallinckrodt, AR grade),
acetonitrile (Fischer HPLC grade), N-isopropylamine (Aldrich,
Ph
OH OH
O
O
OH
N
N
HO
Ph
Ph
OH OH
D-Sorbitol
O
O
Ph
MATERIALS AND METHODS
1
2
Instrumentation. Absorption spectra were recorded on a Varian CARY
300 Bio UV–Vis spectrophotometer. Melting points (corrected) and
polarizing optical micrographs were obtained on a Leitz 585 SM-
LUX-POL microscope equipped with crossed polars, a Leitz 350
heating stage, a Photometrics CCD camera interfaced to a computer,
and an Omega HH503 microprocessor thermometer connected to a J-
K-T thermocouple. NMR spectra were obtained on a Varian 300 MHz
NMR spectrometer interfaced to a Sun Sparc UNIX computer using
Mercury software. Chemical shift standards were internal TMS (1H)
+
P
BF
-
4
CH
3
1P14BF4
and external H3PO4
obtained using a FISONS MD-800 GC ⁄ MS with a DB-5 GC column.
(
31P). GC ⁄ MS fragmentation patterns were
Scheme 2

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
572 Carlos A. Chesta et al.
99.5 + %), di-isopropylamine (Aldrich, 99 + %), dibenzylamine
(Aldrich, 97%), benzoylformic acid (Aldrich, 97%), methyl mandelate
(Aldrich, 97%), phenylglyoxal hydrate (Aldrich, 97%), tri-tetradecyl-
phosphine (a gift from CYTEC, Inc.), and potassium ethyl xanthate
(Aldrich, reagent grade) were used as received. Mandelic acid (Aldrich,
97%) was recrystallized twice from methanol before use.
‘Dry’ D-sorbitol (Aldrich, ‡ 99%), but invariably containing traces
of water, was used as received but dried (as explained in the text) when
used as an irradiation medium. Methyl-tri-tetradecylphosphonium
tetrafluoroborate was prepared following the reported procedure (22).
TK-SmA2 = 55ꢁC (lit. 52.6ꢁC) (22), TSmA2-i = 98ꢁC (lit. 100.6ꢁC) (22).
1H NMR (CCl3D): d 2.16 (m, 6H), 1.83 (d-JP-CH3 13.2 Hz, 3H), 1.49
(m, 12H), 1.25 (m, 60H) ppm. 31P NMR (CCl3D): d 32.2 ppm.
Oxoamides 1 (mp. 132ꢁC; lit. 126ꢁC [15]) and 2 (mp 96ꢁC; lit. 97ꢁC
[15]) were synthesized from benzoylformic acid chloride and the
appropriate dialkylamine as reported (12). See Supplemental Materials
for characterization details.
Analytical protocols. The disappearance of the n-p* absorption
bands of the oxoamides, using the UV–vis detector (at 350 nm) of the
HPLC, provided a convenient method to follow the progress of the
photoconversions. However, as noted above, the photoproducts do not
absorb at k > 300 nm, and their relative yields were determined at
220 nm. Irradiation times were adjusted to maintain conversions of the
oxoamides near 20%. After irradiations, the samples in 1P14BF4 were
dissolved in chloroform that contained a ‘‘standard.’’ The standard for 1
was oxoamide 2, and vice versa. The chloroform solution was
evaporated to residue under reduced pressure at room temperature
and the residue was heated and cooled twice from small volumes of ethyl
acetate to precipitate the 1P14BF4. The filtrate, containing unreacted
oxoamide, photoproducts, and the standard, was evaporated to residue
and dissolved in acetonitrile before chromatographic analyses by
HPLC. The efficiency of the extraction procedure was found to be
>95% by HPLC analyses of several known oxoamide ⁄ photoprod-
ucts ⁄ standard mixtures placed in 1P14BF4. The other solutions and
neat 1 after irradiation were dissolved in a 3:7 (v:v) water:acetonitrile
and aliquots were analyzed directly by HPLC. No peaks from sorbitol
or ethylene glycol appeared at any of the detection wavelengths
employed.
Oxazolidinones (1O and 2O)1 and azetidinones (1A and 2A) were
obtained from the preparative photolyses of the oxoamides in
methanol. Typically, a solution containing ca 1.0 g of oxoamide in
200 mL of methanol was irradiated for 3 h (k > 300 nm). The solvent
was removed and the photoproducts were separated by column
chromatography (alumina, 80 ⁄ 13 ⁄ 7 (v ⁄ v ⁄ v) hexane ⁄ ethyl acet-
ate ⁄ methanol as eluent). Finally, the photoproducts were recrystallized
from 4 ⁄ 1 (v ⁄ v) hexane ⁄ dioxane or hexane ⁄ ethyl acetate mixtures.
Physical and spectroscopic characterization of the photoproducts is
given in Supplemental Materials.
RESULTS AND DISCUSSION
Characterization of a-oxoamides ⁄ solvent mixtures
In D-sorbitol. Powder X-ray diffraction studies, both before
and after heating D-sorbitol to ca 145ꢁC for ca 5 min
(Fig. S1a,b), indicate that its morphology is different from
the A-form (diffractogram generated [28–31] from single
crystal data [32]; Fig. S1c) and the G-form (33,34). The
endotherms at 77.9 and 126.8ꢁC from DSC heating thermo-
grams of mixtures of 1 and D-sorbitol (Fig. 1) are assigned to
the melting of crystallites of crystalline D-sorbitol and 1,
respectively. The latter, indicating some aggregation and phase
separation, were detectable at concentrations as low as
0.3 wt % (1.3 · 10)5 moles g)1-sample) regardless of the
cooling protocol employed for the melts, and the heats of
melting require that virtually all of the molecules of 1 are in
crystallites when its concentration exceeds ca 20 wt %
(Fig. S2). The small heat of melting for the D-sorbitol matrix
in Fig. 1 is evidence that the cooling protocol (in air or
immersed in cold water in sealed vessels under dry air)
produces glassy states initially (25,27). Empirically, they persist
for several hours at room temperature before becoming
crystalline. Irradiations were performed on freshly prepared
samples. More importantly, the heats of the endotherms from
melting of crystallites of 1 at concentrations £ 3 wt % indicate
that less than one-third of the solute molecules are aggregated
Sample preparations. The concentrations of the oxoamides in
1P14BF4 were typically around 3 wt % (or ꢀ100–150 lmol per gram
of the mixture). They were placed in closed flasks and heated with
stirring at ca 90ꢁC for 30 min. to ensure complete mixing. The hot
liquid was transferred to quartz capillary cells, immediately sealed
from the atmosphere and cooled to 0ꢁC by immersion in an ice-bath.
The optical paths were 0.8 mm for irradiations and 0.2 mm for optical
microscopy. The samples were not degassed because the photoreac-
tions are not sensitive to oxygen (12,18).
Various concentrations of 1 in D-sorbitol were prepared by heating
mixtures in sealed pyrex tubes to ca 140ꢁC (where D-sorbitol, mp. 96.8
(25), 110–111ꢁC (26), is a liquid and all of the 1 dissolved) and mixing
the components mechanically for ca 5 min. The samples were cooled
rapidly to room temperature by either leaving them in the air or by
placing their containers under a stream of cold water to effect
vitrification (25) and crystallization to the morph melting at 96ꢁC (27).
The concentrations of the oxoamides in methanol were ca
150 lmol mL)1. Irradiations in methanol between 10 and 60ꢁC were
conducted in 0.8 mm quartz capillary cells. Samples that were
irradiated above the boiling point of methanol were flame-sealed in
50 mm · 50 mm glass cells.
Irradiation procedures. Irradiations (k > 300 nm) were performed
using a 450 W Hanovia medium pressure mercury lamp in a double
Pyrex glass jacket filled with water. For super-ambient temperature
irradiations, the capillary cells containing neat 1 and samples with
1P14BF4, methanol or ethylene glycol as the medium were immersed in
a thermostatted glass cell filled with water and the temperature of the
system was controlled by a Forma Scientific refrigerating ⁄ heating
circulator bath. The temperature of the thermostatted cell was varied
between 5 and 98ꢁC, and the temperature ( 0.5ꢁC) was monitored
continuously inside the glass cell by a digital thermometer. Irradiation
of neat 1 at 130ꢁC was carried out by sandwiching it in powdered form
between two glass plates using a spacer (1 mm) and heating it on a hot
10
0
stage (130
130ꢁC were accomplished by placing 0.5 mL aliquots in test tubes
(5 mm id) that were strapped to thermometer and heated
(130 3ꢁC) using a hot air blower. Prior to irradiation, 1P14BF4
2ꢁC). Irradiations of solutions of 1 in ethylene glycol at
77.9°C
–10
a
samples were melted to their isotropic liquid phase, cooled to 0ꢁC,
transferred to the thermostatted glass cell, and kept there for ca 20 min
to allow temperature equilibration. Tubes containing D-sorbitol
samples were irradiated at 25ꢁC in the air.
–20
–30
126.8°C
–40
40
80
120
160
Temperature (°C)
1The photoproducts from 1 and 2, where structurally different but of the same
generic type, will be designated with a letter (or letters) and a numerical prefix
to indicate their origin.
Figure 1. DSC thermogram of
D-sorbitol.
a
22 wt %
1
solid solution in

Evaluation Only. Created with Aspose.PDF. Copyright 2002-2021 Aspose Pty Ltd.
Photochemistry and Photobiology, 2007, 83 573
+
Me
Me
Me
P
P
BF4
Me
-
-
BF4
+
P
Me
Me
P
-
-
BF4
BF4
+
P
-
Me
P
-
BF4
BF4
+
Me
P
P
-
-
BF4
BF4
+
Me
Me
-
P
P
BF4
-
BF4
+
Me
P
Me
P
BF4
-
BF4
-
+
P
-
Me
Me
-
P
BF4
BF4
Figure 2. Cartoon representation of molecules in a 2-dimensional slice of a bilayer segment of a smectic A2 or solid phase of 1P14BF4 (22).
Alternating orientations of molecules are also projected in planes perpendicular to the page.
in crystallites. On this basis, we conclude that primarily solid
solutions are obtained at £ 3 wt % concentrations of 1 in
D-sorbitol when the samples are fresh.
UV–vis absorption spectra of 1 wt % 1 in D-sorbitol
sandwiched between two quartz discs using a 1 mm spacer
and in methanol normalized at 250 nm are shown in Fig. S3.
The broadened, red-shifted spectrum in D-sorbitol is probably
the result of the polarity difference and restricted environment
of the D-sorbitol. A variety of conformations that are not
(a)
(b)
(c)
(d)
(e)
important in solution may be frozen into the D-sorbitol matrix.
In 1P14BF4. Molecules of 1P14BF4 are arranged in the
ionic liquid-crystalline (smectic A2; SmA2) and lower tempera-
ture solid (k) phases in lipophilic layers separated by planes
defined by the ionic centers (22). These two phases differ
structurally in the degree of ordering of the long alkyl chains
and the ionic centers, as well as dynamically in their ability to
undergo intramolecular motions and self-diffusion. Figure 2 is
a cartoon representation of the molecular packing within the
smectic and solid phases. X-ray diffraction measurements
indicate some residual short-range ordering of molecules in the
isotropic (i) phase as well (23).
Figure 3a shows a DSC thermogram obtained for the first
heating scan of neat 1P14BF4 as recrystallized from ethyl
acetate. The three heating endotherms are attributed to a
solid–solid (k–k) transition (ꢀ 47ꢁC), a k-smectic A2 ionic
liquid-crystalline (SmA2) transition (ꢀ 55ꢁC) and a SmA2-
isotropic liquid (i) transition (ꢀ 98ꢁC). The heats and
temperatures of these transitions are virtually the same as
those reported (22,23). The k–k phase transition temperature
in the thermogram of the second heating scan (Fig. 3b) is
observed at ꢀ40ꢁC and differs from the value obtained from
the first heating. The thermogram of the first cooling scan
(Fig. 3c) has heat flow maxima at 96, 51 and 22ꢁC, corres-
ponding to the i–SmA2, SmA2–k and k–k phase transitions,
respectively. Successive heating scans reproduce the second
one, and the second and subsequent cooling scans are like the
first. Since the SmA2—i and k—SmA2 heats and transition
temperatures are independent of the heating (or cooling for the
0
30
60
T / °C
90
120
Figure 3. DSC thermograms of neat 1P14BF4: (a) first heating, (b)
second heating, (c) first cooling. DSC thermograms of 3 wt %
(1.3 · 10)4 mol g)1) 1 in 1P14BF4: (d) first heating (second heating
of the sample overall), (e) first heating, recorded one week after sample
preparation.
reverse transitions) cycles, these results indicate that 1P14BF4
exists in two different (solid) forms as obtained from solvent
crystallization and from cooling the molten material. When the
melt-solidified material was reanalyzed by DSC after remain-
ing at room temperature for 1 week, the thermogram corres-
ponded to the first heating of the solvent-crystallized material;
the k–k transition was observed at 47ꢁC again. These results
strongly suggest that the melt-solidified solid phase is not
stable thermodynamically and slowly rearranges with time to
the more stable morph.
The first heating thermogram for a sample of 3 wt % 1 in
1P14BF4 shows two broad transitions at ꢀ41 and ꢀ87ꢁC
(Fig. 3d). The experiment was performed immediately after
sample preparation, which involves heating the mixtures to
isotropic liquids and rapidly cooling them to 0ꢁC. Therefore,
these samples correspond to ‘‘second heating’’ types. The first
cooling scan of this sample has two heating endotherms at ꢀ86
and ꢀ28ꢁC. The first heating thermogram of 1 ⁄ 1P14BF4
sample that was aged at room temperature for 1 week (Fig. 3e)