The photochemistry of 1-(3,5-dimethoxyphenyl)-2-(4-methoxyphenyl)ethyl ethanoate in alcohol solvents: A search for carbocation rearrangements

Article abstract of DOI:10.1139/v03-072

The photochemistry of the title compound 1 in methanol and 2,2,2-trifluoroethanol has been examined. In both solvents two ether products were obtained: one (18) resulting from trapping of the carbocation 2 (expected from photosolvolysis of 1), and the other (19) from the carbocation 3 (expected after rearrangement by hydride migration of cation 2). The substituted trans- and cis-stilbene derivatives 20 and 21 were also primary pbotoproducts. Analysis of product yields as a function of time revealed that the ether product 19 was formed by secondary photolysis of the stilbene derivatives, presumably by a pathway involving excited state protonation. Nanosecond laser flash photolysis results demonstrated that substituted trans-stilbene 20 was produced on the same time scale as the laser pulse.

Full text of DOI:10.1139/v03-072

709
The photochemistry of 1-(3,5-dimethoxyphenyl)-2-
(4-methoxyphenyl)ethyl ethanoate in alcohol
solvents: A search for carbocation
rearrangements
J.C. Roberts and J.A. Pincock
Abstract: The photochemistry of the title compound 1 in methanol and 2,2,2-trifluoroethanol has been examined. In
both solvents two ether products were obtained: one (18) resulting from trapping of the carbocation 2 (expected from
photosolvolysis of 1), and the other (19) from the carbocation 3 (expected after rearrangement by hydride migration of
cation 2). The substituted trans- and cis-stilbene derivatives 20 and 21 were also primary photoproducts. Analysis of
product yields as a function of time revealed that the ether product 19 was formed by secondary photolysis of the
stilbene derivatives, presumably by a pathway involving excited state protonation. Nanosecond laser flash photolysis re-
sults demonstrated that substituted trans-stilbene 20 was produced on the same time scale as the laser pulse.
Key words: ester photochemistry, stilbene photoadditions, carbocation rearrangements.
Résumé : On a étudié la photochimie du composé mentionné dans le titre 1 dans le méthanol et dans le 2,2,2-
trifluoroéthanol. Dans les deux solvants on obtient deux éthers : L’un deux (18) résulte du piégeage du carbocation 2
(produit attendu de la photolyse du composé 1), et l’autre (19) provient du carbocation 3 (produit attendu d’un réarran-
gement par migration d’un ion hydrure du cation 2). Les dérivés trans et cis des stilbènes substitués 20 et 21 sont éga-
lement des photoproduits primaires. L’analyse des produits obtenus en fonction du temps, révèle que l’éther 19 résulte
d’une photolyse secondaire des dérivés du stilbène, vraisemblablement selon un chemin impliquant la protonation de
l’état excité. Les résultats de la photolyse éclair au laser, avec des impulsions de l’ordre de la nanoseconde, démontrent
que le trans-stilbène substitué 20 est produit dans la même échelle de temps que l’impulsion laser.
Mots clés : photochimie d’un ester, photoadditions du stilbène, réarrangements de carbocations.
[Traduit par la Rédaction] Roberts and Pincock 722
their rates of reaction with various nucleophiles and
solvents.
Introduction
The photochemical generation of carbocations by cleav-
age of arylmethyl leaving group σ bonds (ArCR2-LG) is cur-
rently of considerable interest (1). One focus has been on the
mechanism for formation of the arylmethyl cation with the
two possibilities being either (i) direct heterolytic cleavage
to an ion pair from the excited singlet state, or (ii) excited
state homolytic cleavage followed by rapid redox electron
transfer converting the first formed radical pair to an ion pair
(2). Product studies have been extensively used to probe
these two possible pathways. Another focus has been the di-
rect observation of arylmethyl cations by laser flash
photolysis (LFP) techniques (3), which has provided infor-
mation regarding the reactivity of these species, in particular
The observation of products resulting from rearrange-
ments of carbocations is a well-established phenomenon in
ground state chemistry and appears in early chapters of all
modern text books on introductory organic chemistry. The
driving force for these rearrangements is formation of a
more stable cation from a less stable one, with both 1,2-
carbon and hydrogen migrations as well-known examples.
We hoped to apply the above two photochemical techniques
to observe a hydride shift by both product studies and di-
rectly by LFP. We chose the substrate 1 and the cations 2
and 3 as targets (eq. [1]).
The idea behind this choice is that the 3,5-dimethoxy-
phenyl chromophore efficiently promotes the photochemical
Received 13 January 2003. Published on the NRC Research Press Web site at http://canjchem.nrc.ca on 23 June 2003.
Dedicated to Professor Don Arnold for his contributions to chemistry. We have both enjoyed and benefited from our many
interactions with him as a colleague, friend, and research supervisor (JAP, 1972–1973; JCR, 1997–1998).
J.C. Roberts and J.A. Pincock.1 Department of Chemistry, Dalhousie University, Halifax, NS B3H 4J3, Canada.
1Corresponding author (e-mail: james.pincock@dal.ca).
Can. J. Chem. 81: 709–722 (2003)
doi: 10.1139/V03-072
© 2003 NRC Canada

710
Can. J. Chem. Vol. 81, 2003
rate constant of 5 × 105 s–1 in 1,1,1,3,3,3-hexafluoroiso-
propanol (HFIP). In a second example (11), the irradiation
of the 9-fluorenol derivative 9 in methanol gave the ether 10
(28%) and the elimination product 11 (7%). However, the
ether 12 (which would have been formed after a 1,2-hydride
shift converting 13 to 14) was not detected in the reaction
mixture (Scheme 2). When 9 was submitted to strong acid
conditions at –78°C, 13 was detected spectroscopically, but
14 was not observed until the solution was warmed to room
temperature. Quenching of the cold or warmed solution with
methanol gave 10 or 12, respectively. The authors rational-
ized these observations by suggesting that cation 14 is not
formed by a 1,2-hydride shift, but rather by a deprotonation–
protonation sequence involving the by-product 11 (a process
which is only possible in the strong acid experiments). This
suggests that the activation barrier for rearrangement of 13
to 14 is high enough to prevent the process from competing
with rapid quenching by methanol when 13 is generated
photochemically.
[1]
Results
Synthesis of ester 1
Arylmethyl esters such as 1 are easily accessed by
esterification of the corresponding alcohols, and therefore,
alcohol 15 was identified as the direct precursor that was re-
quired for the preparation of 1. For the synthesis of 15, a
Grignard reaction between 3,5-dimethoxy-bromobenzene 16
and 2-(4-methoxyphenyl)ethanal 17 was chosen. As shown
in Scheme 3, the bromide 16 was synthesized by a
Sandmeyer reaction (12) in a low yield (33%) that is consis-
tent with literature precedent (13). The aldehyde 17 was pre-
pared in 74% yield over three steps by Darzens’ sequence
from 4-methoxybenzaldehyde as reported by Macchia and
co-workers (14). Grignard coupling of 16 with 17 gave 15
(96% yield), and esterification by the procedure of Steglich
and Neises (15) gave the desired ester 1 (70% yield). Details
of the procedures are in the Experimental section.
A methanolic solution of ester 1 had absorption maxima at
274 nm (ε = 3600 M–1 cm–1) and 280 nm (ε = 3300 M–1 cm–1).
Excitation of 1 at 274 nm results in fluorescence with λmax
at 298 nm, and the 0,0 band is at 288 nm (415 kJ mol–1).
Since 3,5-dimethoxybenzyl acetate itself does not fluoresce,
this observation is attributed to the 4-methoxyphenyl chro-
mophore, which will absorb competitively with the more re-
active portion of the molecule. To check for any interaction
between the two chromophores, the absorption spectra of 4-
methylanisole and 1-(3,5-dimethoxyphenyl)ethyl ethanoate
were recorded and normalized. Summation of the normal-
ized spectra resulted in a slightly higher absorption (8%)
than was observed at the λmax of 1 (274 nm). No increased
absorption was observed at longer wavelengths. Therefore,
the competitive absorption between the two chromophores
may decrease the quantum yield of the reaction, but there
does not appear to be any strong ground-state interaction be-
tween them.
generation of carbocations via C—O bond cleavage (4, 5).
For instance, the yield of ion-derived products from the
photolysis of 3,5-dimethoxybenzyl acetate in methanol is
60% with a quantum yield of 0.37 (6). A rate constant for
this excited state bond cleavage has not been determined be-
cause the lack of fluorescence precludes determination of
the excited singlet state lifetime, but we estimate that the
rate constant is greater than 1 × 109 s–1. If it is formed,
carbocation 2 is expected to rearrange rapidly to the more
stable isomer 3 by a hydride shift. The difference in stability
for the two cations can be demonstrated by the relative rates
of solvolysis for 4-methoxybenzyl tosylate and 3,5-
dimethoxybenzyl tosylate (1:1 × 105 in 80% aqueous ace-
tone at 25°C) (7). Finally, the 4-methoxybenzyl cation has a
characteristic absorption band at 340 nm that has been used
previously to study its reactivity by LFP (8). Therefore, if 3
is formed from 2, the time evolution of its growth could be
observable.
A similar approach has been reported recently by Lee-
Ruff and co-workers (9). In the first example (9), photolysis
of the vicinal diol 4 in acetonitrile, methanol, and 2,2,2-
trifluoroethanol (TFE) gave the radical-derived product 5,
along with the pinacol rearrangement product 6 (Scheme 1).
The proposed mechanism proceeds through the cations 7 and
8. The efficient formation of 7 results from the previously
reported high photochemical reactivity of 9-hydroxyfluorene
derivatives to give fluorenyl cations which are not stabilized
by aromatic delocalization because they have 4n π electrons
(10). The cation 7 was observed by LFP, and its unimole-
cular rearrangement to 8 (not observed) was found to have a
Photolysis of ester 1 in methanol
Photolysis of a nitrogen-saturated solution of 1 in metha-
nol using a Rayonet reactor and 254 nm lamps afforded a
© 2003 NRC Canada

Roberts and Pincock
711
Scheme 1. Formation of ketone 6 via the photochemical pinacol rearrangement of cation 7.
Scheme 2. The photochemistry of fluorenol 9 in methanol.
mixture of seven products after irradiation for 1 h
(Scheme 4). Figure 1 shows the normalized product yields
as a function of time. Preliminary identification of the prod-
ucts was made on the basis of their mass spectra (GC–MS).
Pure samples were obtained either by independent syntheses
or by isolation following photochemical reactions (vide in-
fra). The reaction was also monitored by gas chromatogra-
phy with a flame ionization detector (GC-FID). The GC-FID
response was calibrated using each of the pure samples, al-
lowing the accurate calculation of product yields (Table 1).
Further details on the methods used to obtain accurate val-
ues for the product yields can be found in the Experimental
section.
As shown in Fig. 1 and Table 1, almost 90% of 1 is con-
sumed after 1 h of irradiation, and by this time there is very
little change in the relative percentages of the products over
time. For the purposes of the current project, the methyl
ethers 18a and 19a (R = CH3) are of the greatest interest,
since they are the products that would be formed from
nucleophilic attack of the solvent on carbocations 2 and 3.
Although the yield of 19a is difficult to measure at low con-
versions, the ratio of 18a to 19a appeared to decrease with
© 2003 NRC Canada

712
Can. J. Chem. Vol. 81, 2003
Scheme 3. The synthesis of 1-(3,5-dimethoxyphenyl)-2-(4-methoxyphenyl)ethyl ethanoate 1.
Scheme 4. Products detected following the photolysis of ester 1 in methanol and TFE.
time until a steady state ratio of 21:1 was reached; after only
10 min, however, 18a:19a = 33:1.
stilbene derivatives (16). Finally, products 23 and 24 are
radical-derived by-products of the reaction.
The substituted stilbenes 20 and 21 are clearly primary
photoproducts, but at low conversions the trans isomer is
formed in higher yield than the cis (trans:cis = 2:1 at 2%
conversion). The two isomers then photoequilibrate to a ratio
of 1:1 at high conversion of 1. The substituted phenanthrene
22 is produced by secondary photochemistry — a photo-
chemically allowed conrotatory electrocyclic reaction of the
cis-stilbene 21, followed by oxidation of the dihydrophen-
anthrene intermediate. This reaction has been well-
established for the photochemistry of other electron-rich
As noted above, the identification of most products was
made by comparison with authentic samples. Ether 18a was
synthesized by reaction of alcohol 15 with sodium hydride
and iodomethane. To produce 19a in a similar alkylation
method, the required alcohol 25 was synthesized by
Grignard coupling of 4-bromoanisole and 2-(3,5-dimethoxy-
phenyl)ethanal 26 (the aldehyde, in turn, was prepared by
the Darzens’ sequence from 3,5-dimethoxy-benzaldehyde).
The substituted trans-stilbene 20 was produced by dehydra-
tion of 15 with p-toluenesulfonic acid in benzene. Photo-
© 2003 NRC Canada

Roberts and Pincock
713
Fig. 1. Product yields as a function of time for the photolysis of ester 1 in methanol (Note: compounds 23 and 24 have been removed
for improved clarity).
Table 1. Product yields after photolysis of ester 1 for 1 h in methanol and TFE.
Solvent
% Conversiona
18b
19b
2
20
21
22
4
5
23
5
4
24
Methanol (R = CH3)
TFE (R = CH2CF3)
89
82
42
30
18
13
19
13
10
7
27
aProduct yields are normalized to 100% because mass balances were essentially quantitiative; see the Experimental section
for complete analysis details.
bFor R = CH3, products are designated 18a and 19a. For R = CH2CF3, products are designated 18b and 19b.
chemical cis–trans isomerization of 20 in acetonitrile
provided access to the substituted cis-stilbene 21 in a steady-
state ratio of trans:cis = 1:2 (Pyrex filter, 280 nm cut-off).
Finally, irradiation of 20 in an aerated solution of methanol
gave the substituted phenanthrene 22 after purification by
column chromatography. Complete synthetic procedures for
all of these compounds are included in the Experimental sec-
tion.
An issue of some concern was whether or not 18a could
react by secondary photochemistry, and thereby revert to
carbocation 2. To this end, a nitrogen-saturated solution of
18a in methanol was irradiated. Although the ether was con-
sumed (25% conversion after 1 h, 47% after 5 h), only the
reduction product 23 was detected. This is clearly a result of
radical, not cation, intermediates. Column chromatography
of the reaction mixture after a high-conversion photolysis of
18a provided a convenient method for isolating larger
amounts of pure 23 than were obtained following photolysis
of 1 in methanol.
photochemistry of 1 in TFE was examined under the same
conditions as the experiments in methanol. Figure 2 shows
the normalized product yields as a function of time. As indi-
cated in Scheme 4 and Table 1, the same photoproducts
were formed, keeping in mind that the addition of TFE to
cations 2 and 3 leads to the ethers 18b and 19b (i.e., with
R = CH2CF3). The synthesis of both trifluoroethyl ethers
was accomplished from alcohols 15 and 25 using a modifi-
cation of the Mitsonobu reaction developed by Falck and co-
workers (17). Attempts to alkylate either 15 or 25 using
NaH–DMSO and 2-bromo-1,1,1-trifluoroethane gave no de-
tectable amounts of the desired ethers. Just as in the case of
the ether 18a (R = CH3), 18b (R = CH2CF3) was also
checked for its photochemical reactivity (this time in TFE).
Consumption of the ether was observed (29% conversion af-
ter 1 h), and compound 23 was again the only product de-
tected. None of the isomeric ether 19b was detected,
confirming that 18b is not sufficiently reactive to give
carbocation intermediates.
Although the % conversion of 1 over 1 h is essentially the
same in methanol and TFE (89 vs. 82%) the product yields
do show some significant differences. As discussed earlier
for the reaction in methanol (R = CH3), the ratio of the ether
Photolysis of ester 1 in TFE
With hopes of prolonging the lifetime of cation 2, and
thereby allowing more time for the rearrangement to 3, the
© 2003 NRC Canada

714
Can. J. Chem. Vol. 81, 2003
Fig. 2. Product yields as a function of time for the photolysis of ester 1 in TFE (Note: compounds 23 and 24 have been removed for
improved clarity).
product derived from cation 2 (18a) to the product derived
from cation 3 (19a) is 21:1 after 1 h. The analogous ratio for
18b:19b (the reaction in TFE, R = CH2CF3) is 30:27 (ap-
proximately 1:1), supporting the possibility that the less
nucleophilic solvent (TFE) perhaps allows more time for the
1,2-hydride shift to occur. However, examination of Fig. 2
reveals that the ratio of 18b:19b is not constant. In fact, the
changes in the relative amounts of the ether products during
the photolysis of 1 are much more pronounced in TFE
(18b:19b = 16:1 after 2 min; 1:1 after 1 h) than in methanol
(18a:19a = 33:1 after 10 min, 21:1 after 1 h).
With the intention of observing the growth of cation 3 di-
rectly, a solution of 1 in TFE was subjected to LFP at
266 nm. As shown in the representative spectrum (Fig. 3),
an intense absorption band was observed from 280 (the low-
est wavelength used) to 350 nm, with a maximum of approx-
imately 300 nm. Importantly, the signal did not decay over
any time window investigated with the laser system (10 ns to
50 µs). The same signal was observed in solutions that were
purged with either oxygen or nitrogen. To explore the idea
that the signal might be due to a transient species that de-
cays over a much longer time period, a solution of 1 in TFE
was submitted to 50 laser pulses, and monitored using a con-
ventional UV–vis spectrometer. The signal did not decay af-
ter 30 min, indicating that this signal is not due to a reactive
intermediate, but rather to a strongly absorbing photo-
product. Comparison of the laser spectrum with the UV ab-
sorption spectra of the isolated products 18–23 suggests that
trans-stilbene 20 (λmax = 303 nm, εmax = 29 000 M–1 cm–1)
is the compound responsible for the absorption (Fig. 4).
cient of 20 (ε254 = 3030 M–1 cm–1) is more than three times
larger than that of 1 (ε254 = 800 M–1 cm–1). Furthermore, 20
absorbs much more strongly than 1 at longer wavelengths.
This means that when the photolysis of 1 has proceeded to
the point that [1] < 3[20], the majority of the light will be
absorbed by 20. The results from the photolysis of 1 in the
two solvents suggest that this condition would be met be-
tween 10 and 15 min after starting the photolysis.
At this point, an important issue must be addressed. Be-
cause the formation of 20 during LFP of the ester 1 makes
the direct observation of cations 2 and 3 virtually impossi-
ble, the use of product yields in assessing the importance of
rearrangement becomes critical. If the excited state of 20 is
sufficiently reactive to photochemical addition of the alcohol
solvent to give the ethers 18 and 19, then any conclusions
about cation rearrangements will be incorrect. To address
this point, the photochemistry of 20 in methanol and TFE
was examined.
Photolysis of substituted trans-stilbene 20 in methanol
and TFE
The formation of 20 during the photolysis of 1 corre-
sponds to the loss of acetic acid from the parent ester. For
this reason, 1 equiv of acetic acid was added to the solution
of 20 in either solvent before irradiation, so as to reproduce
the reaction conditions as accurately as possible. No product
formation was observed after stirring the solutions of 20 in
methanol or TFE with added acetic acid for 48 h in the dark,
thus, ensuring that 20 does not react by ground-state chemis-
try during the time frame of the photolysis experiment.
Photolysis of 20 in the two solvents gave a mixture of
four products (Scheme 5). All of the products have already
been observed during the photolysis of 1, and so calibrated
Also shown in Fig. 4 is the absorption spectrum of 1. At
the maximum light output of the lamp used for the
photolysis experiments (λ = 254 nm), the extinction coeffi-
© 2003 NRC Canada

Roberts and Pincock
715
Scheme 5. Products detected following the photolysis of substituted trans-stilbene 20 in methanol and TFE.
Fig. 4. Comparison of the absorption spectra of ester 1 and sub-
stituted trans-stilbene 20 (both 3.45 × 10–5 M in methanol).
Fig. 3. Change in optical density following LFP at 266 nm of
ester 1 in TFE.
(eq. [1]). This was based on the idea that TFE, being a less
nucleophilic solvent, would promote the formation of
carbocation 3 from carbocation 2. The difference in the final
product ratios (18:19) seemed to support this idea. However,
closer inspection of the data revealed that this ratio is not
constant over the course of the experiment. For the reaction
in methanol, 18a:19a = 33:1 after 10 min, and 20:1 after
60 min. The change is even more drastic for the photolysis
in TFE: after 2 min 18b:19b = 16:1, but after 1 h the ratio is
18b:19b = 1:1. The change in the relative amounts of the
products that result from nucleophilic attack on carbocations
2 and 3 indicates that a secondary photochemical reaction
occurs during the steady-state photolysis of 1. One possible
explanation is that the ethers 18 decay over the course of the
photolysis experiments involving 1. Although these ethers
were shown to be photochemically active in the control ex-
periments, they appear to react too slowly to account for the
observed change in the 18:19 ratio. Furthermore, the pres-
ence of several strongly absorbing compounds in the same
reaction mixture will make absorption by the ethers 18 even
less likely. Indeed, the only experiment in which decay of
either 18a or 18b was detected was during the photolysis of
20 in TFE (Fig. 4). The yield of 18b does decrease over the
final 2 h of the experiment (4 to 2%), but only after most of
the substituted stilbenes have been converted to products.
A much more satisfying explanation of the observed
change in the 18:19 ratio is that secondary photochemistry
provides an alternative pathway for the formation of the
product yields were calculated (Table 2). The normalized
product yields as a function of time are shown in Figs. 5
(methanol) and 6 (TFE). In striking contrast to the ester 1,
which exhibits very similar reactivity in either methanol or
TFE, 20 reacts much faster in TFE than in methanol. After
1 h, the % conversion of 20 in methanol is 48%, whereas the
% conversion is 80% after the same time in TFE. In both
solvents, the substituted stilbene isomers approach the
steady state composition of 1:1 and then disappear simulta-
neously, although this process occurs much more rapidly in
TFE. Another important observation is that the ether prod-
ucts observed require the formation of both carbocations 2
and 3 by photochemical protonation of 20. Furthermore, in
contrast to the results for 1, the ethers produced from
nucleophilic attack on the rearranged carbocation 3 (i.e., 19a
and 19b) are formed in greater yield than their isomeric
counterparts (18a and 18b). In fact, 18a was not detected
even after 5 h of irradiation. These results strongly suggest
that cation 3 is formed rapidly upon photolysis of 20 in
methanol and TFE.
Discussion and conclusions
The general mechanism that has been developed for ester
photochemistry seems to hold for ester 1. Both ion- and
radical-derived products are formed. The results from the
photolysis of 1 in methanol and TFE initially suggested that
the desired 1,2-hydride shift occurred as was predicted
© 2003 NRC Canada

716
Can. J. Chem. Vol. 81, 2003
Table 2. Product yields after photolysis of substituted trans-stilbene 20 in methanol and TFE.
Solvent
Time (h)
% Conversiona
18b
19b
21
22
4
10
60
87
8
26
9
Methanol (R = CH3)
Methanol (R = CH3)
TFE (R = CH2CF3)
TFE (R = CH2CF3)
1
5
1
4
48
63
80
95
0
0
4
2
88
64
27
3
8
aProduct yields are normalized to 100% because mass balances were essentially quantitiative; see the Experimental section
for complete analysis details.
bFor R = CH3, products are designated 18a and 19a. For R = CH2CF3, products are designated 18b and 19b.
Fig. 5. Product yields as a function of time for the photolysis of substituted trans-stilbene 20 in methanol.
products derived from the cation 3, specifically the ethers
19. As shown in the control experiments involving 18a and
18b, there is no pathway for converting 18 to 19. Rather, the
secondary photochemical reaction is addition of the solvent
to the excited state of the substituted trans-stilbene 20. As
clearly shown by the steady-state photolysis results, cation-
derived products (i.e., 18 and 19) can be formed in reason-
able efficiency from 20 under the same reaction conditions
used for the photolysis of ester 1. Furthermore, this reaction
favours the ethers 19 over 18, which is the reverse of the
regiochemistry observed for the photolysis of 1. The LFP
experiments demonstrate that 20 is formed rapidly upon irra-
diation of 1, and will be available to absorb competitively
with 1 very early in the photolysis experiments involving the
ester. Therefore, as the substituted trans-stilbene 20 accumu-
lates during the photolysis of 1, the photochemistry of 20
will become more important, and 19 (formed primarily from
20) will eventually be formed more rapidly than 18 (formed
primarily from 1). Indeed, even if the formation of 20 did
not make the observation of benzylic carbocations by LFP
impossible (by obscuring the wavelengths of interest), the
observation of a signal corresponding to cation 3 would not
be conclusive evidence for the rearrangement of interest.
Unfortunately, the rapid formation of 20 upon photolysis of
1 makes the ester a poor substrate for the investigation of
cation rearrangements. Our results are similar to those of
Lee-Ruff and co-workers (11) (Scheme 2), where the desired
cation rearrangement is too unfavourable to compete with
rapid quenching by the alcohol solvent (either by nucleo-
philic attack or by deprotonation). However, the photo-
chemistry of 1 is further complicated by the fact that the
by-products themselves react by secondary photochemistry,
making reliable assessment of the rearrangement extremely
difficult.
There are several other points that need to be addressed
for a thorough understanding of the photochemistry of 1 and
its photoproducts. Firstly, although deprotonation of cation 2
or 3 appears to be the most likely pathway for formation of
20, more study is needed to rule out other mechanisms.
Work is in progress to examine the possibility that substi-
tuted stilbenes 20 and 21 are produced by either a radical
pathway or by concerted photochemical elimination of acetic
acid from 1. In addition, the 1,2,2-trimethyl derivative of 1 is
currently being synthesized. This substrate may allow the
observation of a rearrangement without the complications
present for 1 because rearrangement and nucleophilic trap-
© 2003 NRC Canada

Roberts and Pincock
717
Fig. 6. Product yields as a function of time for the photolysis of substituted trans-stilbene 20 in TFE.
ping may be the only reactions available for the initial car-
bocation.
mediate, which then inserts directly to the methanol O—
H(D) bond. However, the influence of electron-rich substitu-
ents in the meta position was never addressed (the only com-
pound examined with meta substitutents was the symmetric
3,3′-dimethylstilbene). In more recent work, Lewis et al. (21,
22) have made several detailed studies of the effect of
electron-donating substituents (mostly amino) on the fluo-
rescence behaviour of substituted stilbenes and related
compounds. These results indicate that stilbenes with
electron-donating meta-substituents have longer singlet life-
times and higher fluorescence quantum yields than the para-
substituted analogues. Furthermore, the authors attribute this
observation to a higher barrier for bond torsion in the meta-
substituted cases. Addition of methanol was only observed
in one case (21) (eq. [2]). However, methanol was not em-
ployed as a solvent for many of the photophysical studies
that were of interest to the authors, so the reaction may in
fact be more general.
Secondly, the photochemical addition of alcohols to sub-
stituted stilbenes needs to be investigated in much greater
detail. At first, the reaction appears to be analogous to the
photohydration of substituted styrenes that has been studied
extensively by McEwen and Yates (18), who demonstrated
that excitation of 3-methoxystyrene in water gave the
Markovnikov addition product faster than 4-methoxystyrene
(another example of the photochemical meta effect). Addi-
tional support for the presence of a carbocation interme-
diate in such photochemical additions was provided by
McClelland and co-workers (19), who found that laser flash
photolysis of 4-methoxystyrene in TFE produced a transient
absorption signal at 340 nm that was attributed to the corre-
sponding cation. However, the presence of a second aro-
matic ring in the single chromophore of the substituted
stilbenes makes the assessment of substituent effects much
more complicated than for substituted styrenes (with regards
to both the rate of the reaction and the regiochemistry of the
addition).
A survey of the literature indicates that the photochemical
addition of methanol to a variety of substituted stilbenes was
first observed by Laarhoven and co-workers (20). In contrast
to the results for substituted styrenes, the yield of stilbene–
solvent adduct was found to be independent of acid concen-
tration. This suggested that protonation of the singlet excited
state is not the rate-limiting step in the photochemical addi-
tion of methanol to substituted stilbenes. With the aid of ex-
periments conducted in deuterated methanol, the authors
concluded that the addition of methanol to stilbenoid sys-
tems occurs via two competing mechanisms: (i) direct addi-
tion across the central bond; and (ii) rearrangement of the
excited state via a 1,2-hydride shift to give a carbene inter-
[2]
© 2003 NRC Canada

718
Can. J. Chem. Vol. 81, 2003
1
From the results of the current study, just how the photo-
yield 33%: mp 64–65°C, lit. (13) mp 64–66°C. H NMR δ:
chemistry of 20 fits with the examples in the literature is not
yet clear. The rapid addition of the solvent (TFE in particu-
lar) appears to be because of the two meta substituents (the
3,5-dimethoxyphenyl ring), while the preferred regiochem-
istry is governed by the para substituent (4-methoxyphenyl
ring). We had originally believed that the increased reactiv-
ity of 20 in TFE compared to methanol could be attributed
to the greater effective acidity of the former solvent, which
might lead to more rapid protonation of the excited state.
However, the study by Laarhoven and co-workers (20)
seems to indicate that the mechanism of solvent addition
may be even more complicated. Clearly, a larger set of stil-
bene derivatives is required to confidently assess the factors
controlling the reaction. This work is currently in progress.
6.66 (d, 2H, J = 2.4 Hz), 6.37 (t, 3H, J = 2.4 Hz), 3.76 (s,
6H). 13C NMR δ: 161.2, 123.0, 109.8, 99.8, 55.5. GC–MS
m/z: 218 (82.6), 216 (100.0), 108 (69.7), 79 (41.6), 77
(56.2), 63 (48.9), 51 (26.3).
2-(4-Methoxyphenyl)ethanal (17)
This compound was prepared from 4-methoxybenzalde-
hyde by glycidic ester condensation as described (14), yield
of 74% over three steps: 1H NMR δ: 9.66 (d, 1H, J =
2.4 Hz), 7.09 (d, 2H, J = 8.5 Hz), 6.88 (d, 2H, J = 8.5 Hz),
3.75 (s, 3H), 3.57 (d, 2H, J = 2.4 Hz). 13C NMR δ: 199.8,
159.0, 130.8, 123.9, 144.4, 55.2, 49.6. GC–MS m/z: 121
(100), 91 (20.5), 78 (23.9), 77 (31.2), 51 (12.2).
1-(3,5-Dimethoxyphenyl-2-(4-methoxyphenyl)ethan-1-ol
(15)
Experimental
For this reaction, all glassware was dried in an oven and
purged with nitrogen gas prior to use. Liquid transfers were
performed using a cannula needle under positive pressure. A
solution of 3,5-dimethoxybromobenzene 16 (4.00 g,
18.4 mmol) in THF (20 mL) was prepared, and then trans-
ferred to a dropping funnel atop a three-necked 100 mL
round-bottomed flask containing magnesium turnings
(2.68 g, 110 mmol) in THF (5 mL). Approximately 10% of
the aryl halide solution was run into the flask, along with a
small crystal of iodine. After 15 min of stirring, the yellow
colour of the iodine disappeared, and the mixture began to
reflux. The remaining solution in the dropping funnel was
added to the magnesium over 15 min, and the mixture was
refluxed for 30 min after the addition was complete. After
cooling to room temperature, the resulting orange solution
was transferred under nitrogen to a clean three-necked flask.
A solution of 2-(4-methoxyphenyl)ethanal 17 (2.76 g,
18.4 mmol, distilled under vacuum prior to use) in THF
(20 mL) was prepared. The solution was then added
dropwise to the solution containing the Grignard reagent,
and the resulting mixture was refluxed for 30 min after the
addition was complete. After cooling to room temperature,
the yellow solution was poured into a separatory funnel con-
taining saturated ammonium chloride solution (100 mL) and
dichloromethane (50 mL). The layers were separated, and
the aqueous portion was extracted with dichloromethane
(2 × 50 mL). The combined organic extracts were washed
with distilled water and saturated sodium chloride solution
(2 × 75 mL each). After drying over anhydrous magnesium
sulfate and filtering, the solvent was removed under reduced
pressure to give 5.10 g of the product alcohol (96%). Further
reactions were performed using the crude material, although
characterization was performed using a sample recrystallized
General procedures
Melting points were determined on a Fisher–Johns appa-
ratus and are uncorrected. H and 13C NMR spectra were re-
1
corded in deuterated chloroform on a Bruker AC 250F
instrument, and chemical shifts are reported as parts per mil-
lion (ppm) relative to tetramethylsilane internal standard.
The coupling constants of the ABX systems observed in sev-
eral compounds were analysed by matching the line posi-
tions with those of simulated spectra. Gas chromatography
was performed on a Perkin-Elmer Autosystem XL instru-
ment (controlled by a computer with TurboMass and
TurboChrom software) with one Turbomass detector and one
FID (both columns: Supelco 30 m/0.25 mm MDN-5S 5%
phenyl methylsiloxane, film thickness 0.50 µm). For GC–
MS, the injection volume was 1 µL; mass spectral data are
reported in units of mass over charge (m/z) with intensities
relative to the base peak (in brackets). For GC-FID, an injec-
tion volume of 2.5 µL was used. The same temperature pro-
gram was used for both detectors: 60°C for 1 min, ramp at
20°C/min to 240°C, hold at 240°C for 20 min. Ultraviolet
spectra were recorded on a Varian–Cary Bio 100 spectrome-
ter.
Methanol, water, ethyl acetate, and hexanes were all dis-
tilled prior to use. Dichloromethane, benzene, dimethyl-
sulfoxide (DMSO), TFE, pentane, and the deuterated
solvents were all reagent grade, while acetonitrile was
HPLC grade (these solvents were used without extra purifi-
cation). Tetrahydrofuran (THF) was distilled from sodium–
benzophenone, and then a second time from lithium alu-
minium hydride before being kept under nitrogen. All start-
ing materials were supplied by Aldrich Chemicals, with the
exception of 3,5-dimethoxyaniline and 3,5-dimethoxybenza-
ldehyde (Avocado Chemicals). Thin layer chromatography
was performed using plates from Eastman-Kodak. Prepara-
tive chromatography was performed using 60–250 mesh sil-
ica gel from Silicycle.
o
1
from ethyl acetate – hexanes, mp: 97–99 C. H NMR δ:
7.09 (d, 2H, J = 8.6 Hz), 6.83 (d, 2H, J = 8.6 Hz), 6.49 (d,
2H, 2.4 Hz), 6.36 (t, 1H, J = 2.4 Hz), 4.76 (m, 1H, J1 =
8.5 Hz, J2 = 4.9 Hz, J3 = 2.4 Hz), 3.77 (s, 9H), 2.95 (m, 1H,
J1 = 13.7 Hz, J2 = 4.9 Hz), 2.89 (m, 1H, J1 = 13.7 Hz, J2 =
8.5 Hz), 2.07 (d, 1H, J = 2.4 Hz). 13C NMR δ: 160.8, 158.4,
146.5, 130.5, 130.0, 113.9, 103.8, 99.6, 75.4, 55.4, 55.3,
45.1. GC–MS m/z: 271 (5.6), 270 (35.8), 167 (13.1), 139
(35.8), 122 (100.0), 121 (49.9), 77 (16.0). HR-MS calcd.:
288.1361; found: 288.1357 ± 0.0008.
Synthetic procedures
3,5-Dimethoxybromobenzene (16)
This compound was prepared from the diazonium ion
made from 3,5-dimethoxyaniline using the procedure
described (12) for the synthesis of o-bromochlorobenzene,
© 2003 NRC Canada

Roberts and Pincock
719
158.0, 144.4, 130.4, 113.5, 104.6, 99.6, 85.4, 56.9, 55.3,
55.2, 43.8. GC–MS m/z: 303 (0.8), 302 (4.4), 271 (1.5), 270
(6.3), 182 (10.4), 181 (100.0), 121 (12.1). HR-MS calcd.:
302.1518; found: 302.1522 ± 0.0008.
1-(3,5-Dimethoxyphenyl)-2-(4-methoxyphenyl)ethyl
ethanoate (1)
This compound was prepared using the method of
Steglich and Neises (15). A solution of N,N-dimethyl-4-
aminopyridine (63 mg, 0.52 mmol) and acetic acid (312 mg,
5.2 mmol) in dichloromethane (25 mL) was prepared, and
then added to a solution of 1-(3,5-dimethoxyphenyl)-2-(4-
methoxyphenyl)ethan-1-ol 15 (1.5 g, 5.2 mmol) in dichloro-
methane (30 mL). After the resulting solution was cooled in
an ice bath, 1,3-dicyclohexylcarbodiimide (1.17 g, 5.67 mmol)
was added in one portion. The resulting mixture was stirred
at 0°C for 5 min, and then allowed to stir at room tempera-
ture for 3 h. The urea precipitate was filtered off, the solvent
was removed under reduced pressure, and the residue was
taken up in dichloromethane (50 mL) and filtered again. The
clear solution was then washed with 0.5 M hydrochloric acid
and distilled water (2 × 25 mL each). The organic material
was then dried with anhydrous magnesium sulfate, filtered,
and the solvent removed under reduced pressure to give a
clear oil. The oil was adsorbed onto silica for column chro-
matography. Using 5% ethyl acetate – hexanes as the eluant
provided 1.5 g of a solid product, which was recrystallized
from the same solvent mixture to give 1.2 g of colourless
1-(3,5-Dimethoxypheny)-2-(4-methoxyphenyl)-1-(trifluoro-
ethoxy)ethane (18b)
This compound was prepared using the method of Falck
et al. (17). To a solution of 1-(3,5-dimethoxyphenyl)-2-(4-
methoxyphenyl)ethan-1-ol 15 (0.300 g, 1.04 mmol) in ben-
zene (20 mL) was added 1,1′-(azodicarbonyl)dipiperidine
(0.525 g, 2.08 mmol). The flask was purged with nitrogen
for 10 min, and tri(n-butyl)phosphine (0.421 g, 2.08 mmol)
was added. After stirring the reaction mixture for another
15 min, TFE (1.04 g, 10.4 mmol) was added. The mixture
was stirred at room temperature for 1 h, and the solvent was
then removed under reduced pressure. The residue was taken
up in dichloromethane, filtered, and the liquid then adsorbed
onto silica gel for column chromatography. Elution with 5%
ethyl acetate – hexanes gave the desired product (0.154 g,
1
40% yield). H NMR δ: 7.03 (2H, d, J = 8.6 Hz), 6.78 (d,
2H, J = 8.6 Hz), 6.38 (s, 3H), 4.43 (dd, 1H, J1 = 5.5 Hz, J2 =
7.3 Hz), 3.77 (s, 3H), 3.75, (s, 6H), 3.64 (m, 2H), 3.09 (dd,
1H, J1 = 14.0 Hz, J2 = 7.3 Hz), 2.86 (dd, 1H, J1 = 14.0 Hz,
J2 = 5.5 Hz). 13C NMR δ: 161.0, 158.2, 142.6, 130.5, 129.7,
124.0 (q, J = 278.6 Hz), 113.5, 104.6, 100.1, 84.9, 66.0 (q,
J = 34.3 Hz), 55.3, 55.2, 43.5. GC–MS (m/z): 371 (1.5), 370
(7.9), 270 (6.8), 250 (11.6), 249 (100.0), 166 (22.1), 121
(81.1). HR-MS calcd.: 370.1392; found: 370.1392 ± 0.0008.
1
crystals (70%), mp: 68–69°C. H NMR δ: 7.03 (d, 2H, J =
8.6 Hz), 6.78 (d, 2H, J = 8.6 Hz), 6.40 (d, 2H, J = 2.4 Hz),
6.37 (t, 1H, J = 2.4 Hz), 5.81 (m, 1H, J1 = 7.9 Hz, J2 =
6.1 Hz), 3.77 (s, 3H), 3.75 (s, 6H), 3.08 (m, 1H, J1
14.0 Hz, J2 = 7.9 Hz), 2.97 (m, 1H, J1 = 14.0 Hz, J2 =
6.1 Hz), 2.02 (s, 3H). 13C NMR δ: 170.1, 160.7, 158.3,
142.6, 130.5, 129.1, 113.6, 104.6, 99.7, 76.7, 55.3, 55.2,
42.1, 21.2. GC–MS m/z: 270 (41.5), 167 (49.1), 139 (20.3),
121 (100), 77 (14.9). HR-MS calcd.: 330.1467; found:
330.1461 ± 0.0008.
=
2-(3,5-Dimethoxyphenyl)ethanal (26)
The sodium salt of 3-(3,5-dimethoxyphenyl)glycidic acid
was prepared in two steps (62% yield) using the same proce-
dures as described for the synthesis of 2-(4-methoxy-
phenyl)ethanal 17 (14). Decarboxylation of the sodium salt
was accomplished using the method of Bullimore et al. (23),
giving the desired aldehyde in 20% yield (12% over three
1-(3,5-Dimethoxypheny)-2-(4-methoxyphenyl)-1-
methoxyethane (18a)
1
A 60% sodium hydride – oil suspension (0.08 g of sus-
pension, 0.002 mmol NaH) was washed with hexane to re-
move the oil. After decanting the washes, the residue was
taken up in DMSO (2 mL) and added to a solution of 1-(3,5-
dimethoxyphenyl)-2-(4-methoxyphenyl)ethan-1-ol 15 (0.288 g,
1.00 mmol) in DMSO (5 mL). The mixture was stirred at
room temperature for 30 min, and then a solution of methyl
iodide (0.284 g, 2.00 mmol) in DMSO (5 mL) was added
dropwise. After stirring the solution for 3.5 h, distilled water
(12 mL) was added slowly to quench the reaction. The or-
ganic layer was drawn off, and the aqueous portion was ex-
tracted with dichloromethane (3 × 10 mL). The combined
organic material was washed with distilled water (3 ×
25 mL), dried with anhydrous magnesium sulfate, and fil-
tered. Removal of the solvent under reduced pressure gave
0.25 g of material, which was adsorbed onto silica for col-
umn chromatography. Elution using 2.5% ethyl acetate –
hexanes gave the pure ether (0.098 g, 32%). Characterization
was performed on a sample that was further purified by
bulb-to-bulb distillation. 1H NMR δ: 7.03 (d, 2H, J =
6.7 Hz), 6.78 (d, 2H, J = 6.7 Hz), 6.38 (m, 3H), 4.21 (dd,
1H, J1 = 5.5 Hz, J2 = 7.6 Hz), 3.77 (s, 3H), 3.75 (s, 6H),
3.20 (s, 3H), 3.01 (dd, 1H, J1 = 7.6 Hz, J2 = 13.7 Hz), 2.82
(dd, 1H, J1 = 5.5 Hz, J2 = 13.7 Hz). 13C NMR δ: 160.8,
steps). H NMR δ: 9.69 (t, 1H, J = 2.4 Hz), 6.40 (t, 1H, J =
1.8 Hz), 6.35 (d, 2H, J = 1.8 Hz), 3.77 (s, 6H), 3.58 (d, 2H,
J = 2.4 Hz). 13C NMR δ: 199.2, 161.3, 134.0, 107.6, 99.3,
55.3, 50.7.
1-(4-Methoxyphenyl)-2-(3,5-dimethoxyphenyl)ethan-1-ol
(25)
A solution of 4-bromoanisole (1.04 g, 5.55 mmol) in THF
(10 mL) was prepared under nitrogen, and then transferred
to a dropping funnel atop a three-necked 100 mL round-
bottomed flask containing magnesium turnings (0.81 g,
33.3 mmol). Approximately 10% of the aryl halide solution
was added to the magnesium along with an iodine crystal,
and the reaction began within 10 min. The remaining aryl
halide solution was added over 5 min, and the resulting mix-
ture was heated to reflux for 30 min after the addition was
complete. After cooling to room temperature, the resulting
orange solution was transferred under nitrogen to a clean
three-necked flask.
A solution of 2-(3,5-dimethoxyphenyl)ethanal 26 (1.00 g,
5.55 mmol) in THF (10 mL) was added dropwise to the
Grignard reagent, and the resulting mixture was refluxed
gently for 30 min. The mixture was then cooled, and added
to a separatory funnel containing saturated ammonium chlo-
© 2003 NRC Canada

720
Can. J. Chem. Vol. 81, 2003
1
ride (50 mL) and dichloromethane (25 mL). The layers were
separated, and the aqueous portion was extracted with di-
chloromethane (2 × 50 mL). The combined organic extracts
were washed with water and saturated sodium chloride (2 ×
25 mL each), and then dried with anhydrous magnesium sul-
fate. Removal of solvent under reduced pressure gave a
crude oil, which was purified by column chromatography
(10% ethyl acetate – hexanes, eluant) to give the desired
55% yield). H NMR δ: 7.16 (d, 2H, J = 9.2 Hz), 6.30 (d,
2H, J = 9.2 Hz), 6.31 (t, 1H, J = 1.8 Hz), 6.29 (d, 2H, J =
1.8 Hz), 4.50 (dd, 1H, J1 = 5.5 Hz, J2 = 7.3 Hz), 3.80 (s,
3H), 3.72 (s, 6H), 3.62 (m, 2H), 3.14 (dd, 1H, J1 = 14.0 Hz,
J2 = 7.3 Hz), 2.83 (dd, 1H, J1 = 14.0 Hz, J2 = 5.5 Hz). 13C
NMR δ: 160.5, 159.6, 140.1, 132.0, 128.1, 124.0 (q, J =
278.6 Hz), 114.0, 107.4, 98.8, 84.1, 65.8 (q, J = 34.3 Hz),
55.3, 55.2, 44.8. GC–MS (m/z): 370 (1.1), 271 (5.8), 270
(27.6), 220 (10.6), 219 (100.0), 135 (27.3). HR-MS calcd.:
370.1392; found: 370.1389 ± 0.0008.
1
product (1.00 g, 63% yield). H NMR δ: 7.26 (d, 2H, J =
8.5 Hz), 6.87 (d, 2H, J = 8.5 Hz), 6.33 (s, 3H), 4.81 (dd, 1H,
J1 = 8.5 Hz, J2 = 4.9 Hz), 3.78 (s, 3H), 3.73 (s, 6H), 2.92
(m, 2H, J1 = 8.5 Hz, J2 = 4.9 Hz, J3 = 13.7 Hz), 2.12 (s,
1H). 13C NMR δ: 160.7, 159.0, 140.6, 136.1, 127.2, 113.7,
107.4, 98.6, 74.7, 55.3, 55.2, 46.3.
trans-1-(3,5-Dimethoxyphenyl)-2-(4-methoxyphenyl)ethene
(20)
A
solution of 1-(3,5-dimethoxyphenyl)-2-(4-methoxy-
phenyl)ethan-1-ol 15 (4.00 g, 13.9 mmol) in benzene
(600 mL) was prepared in a three-necked round-bottomed
flask. A portion of p-toluenesulfonic acid (0.29 g, 1.5 mmol)
was added, and the solution was heated to reflux with stir-
ring. Water was removed from the mixture by way of a
Dean–Stark trap. After 6 h, analysis by GC–MS indicated
that 97% of the starting material had reacted, and the reac-
tion mixture was allowed to cool to room temperature. The
solution was then washed with distilled water, and saturated
sodium chloride solution (2 × 200 mL each). A portion of
benzene (100 mL) was used to reextract the aqueous washes.
The combined organic material was then dried with anhy-
drous magnesium sulfate, filtered, and the solvent was re-
moved under reduced pressure to give an orange oil.
Isolation of the desired product was achieved using column
chromatography with 20% ethyl acetate – hexanes (eluant).
Recrystallization of the resulting product from pentane gave
white crystals (1.27 g, 34% yield), mp 55–57°C, lit. (24) mp
2-(3,5-Dimethoxypheny)-1-(4-methoxyphenyl)-1-methoxy-
ethane (19a)
A 60% sodium hydride – oil suspension (0.083 g of sus-
pension, 2.08 mmol NaH) was washed with hexane to re-
move the oil. After decanting the washes, the residue was
taken up in DMSO (5 mL) and added to a solution of 2-(3,5-
dimethoxyphenyl)-1-(4-methoxyphenyl)ethan-1-ol 25 (0.300 g,
1.00 mmol) in DMSO (10 mL). The mixture was stirred at
room temperature for 30 min, and then a solution of methyl
iodide (0.295 g, 2.08 mmol) in DMSO (5 mL) was added
dropwise. After stirring the solution for 5 h, distilled water
(15 mL) was added slowly to quench the reaction. The or-
ganic layer was drawn off, and the aqueous portion was ex-
tracted with dichloromethane (3 × 20 mL). The combined
organic material was washed with distilled water and satu-
rated sodium chloride solution (2 × 20 mL each), dried with
anhydrous magnesium sulfate, and filtered. Removal of the
solvent under reduced pressure gave 0.34 g of material,
which was adsorbed onto silica for column chromatography.
Elution using 5% ethyl acetate – hexanes gave the pure ether
(0.24 g, 32%). Characterization was performed on a sample
that was further purified by bulb-to-bulb distillation. 1H
NMR δ: 7.15 (d, 2H, J = 8.5 Hz), 6.87 (d, 2H, J = 8.5 Hz),
6.29 (t, 1H, J = 1.8 Hz), 6.26 (d, 2H, J = 1.8 Hz), 4.27 (dd,
1H, J1 = 6.2 Hz, J2 = 7.4 Hz), 3.80 (s, 3H), 3.71 (s, 6H),
3.17 (s, 3H), 3.06 (dd, 1H, J1 = 13.4 Hz, J2 = 7.4 Hz), 2.80
(dd, 1H, J1 = 13.4 Hz, J2 = 6.2 Hz). 13C NMR δ: 160.4,
159.1, 140.9, 133.6, 128.0, 113.7, 107.4, 98.3, 84.4, 56.5,
55.2 (two signals), 45.0. GC–MS (m/z): 302 (not observed),
286 (9.3), 270 (6.7), 165 (11.5), 151 (29.8), 122 (9.2), 121
(100.0). HR-MS calcd.: 302.1518; found: 302.1526 ± 0.0008.
1
53–54°C. H NMR δ: 7.43 (d, 2H, J = 8.6 Hz), 7.05 (d, 1H,
J = 16.5 Hz), 6.91 (d, 1H, J = 16.5 Hz), 6.90 (d, 2H, J =
8.6 Hz), 6.65 (d, 2H, J = 2.4 Hz), 6.38 (t, 1H, J = 2.4 Hz),
3.83 (s, 9H). 13C NMR δ: 161.0, 159.4, 139.7, 129.9, 128.7,
127.8, 126.6, 114.1, 104.3, 99.6, 55.4, 55.3. GC–MS m/z:
271 (17.6), 270 (100.0), 269 (15.5), 239 (19.7), 224 (13.0),
196 (12.7), 195 (12.33), 165 (10.3), 153 (12.4), 152 (17.6),
141 (10.5). HR-MS calcd.: 270.1256; found: 270.1263 ±
0.0008.
cis-1-(3,5-Dimethoxyphenyl)-2-(4-methoxyphenyl)ethene
(21)
A solution of trans-1-(3,5-dimethoxyphenyl)-2-(4-meth-
oxyphenyl)ethene 20 (0.40 g, 1.5 mmol) in acetonitrile
(340 mL) was prepared in a large photolysis reaction vessel,
and purged with nitrogen for 30 min. A 450 W, medium-
pressure Hanovia mercury lamp with a Pyrex filter (300 nm
cut-off) was employed to irradiate the solution for 30 min.
Analysis by GC-FID indicated that by this time the mixture
had achieved a photostationary state consisting of a 2:1 ratio
of the cis and trans isomers. The solvent was removed under
reduced pressure, and the residue was prepared for column
chromatography. Separation of the isomers was achieved us-
ing 2.5% ethyl acetate – hexanes (eluant), and the pure cis
2-(3,5-Dimethoxypheny)-1-(4-methoxyphenyl)-1-(trifluoro-
ethoxy)ethane (19b)
This compound was prepared using the method of Falck
et al. (17). To a solution of 2-(3,5-dimethoxyphenyl)-1-(4-
methoxyphenyl)ethan-1-ol 25 (0.300 g, 1.04 mmol) in ben-
zene (20 mL) was added 1,1′-(azodicarbonyl)dipiperidine
(0.525 g, 2.08 mmol). The flask was purged with nitrogen
for 10 min, and tri(n-butyl)phosphine (0.421 g, 2.08 mmol)
was added. After stirring the reaction mixture for another
15 min, TFE (1.04 g, 10.4 mmol) was added. The mixture
was stirred at room temperature for 1 h, and the solvent was
then removed under reduced pressure. The residue was taken
up in dichloromethane, filtered, and the liquid then adsorbed
onto silica gel for column chromatography. Elution with 5%
ethyl acetate – hexanes gave the desired product (0.210 g,
1
isomer was isolated as a clear oil (0.15 g, 38% yield). H
NMR δ: 7.19 (d, 2H, J = 8.5 Hz), 6.78 (d, 2H, J = 8.5 Hz),
6.52 (d, 1H, J = 12.2 Hz), 6.44 (d, 1H, J = 12.2 Hz), 6.43 (d,
2H, J = 2.4 Hz), 6.32 (t, 1H, J = 2.4 Hz), 3.78 (s, 3H), 3.67
(s, 6H). 13C NMR δ: 160.6, 158.7, 139.5, 130.3, 130.2,
129.6, 128.7, 113.5, 106.6, 99.7, 55.2. GC–MS m/z: 271
© 2003 NRC Canada

Roberts and Pincock
721
(16.44), 270 (100.0), 269 (17.5), 239 (22.5), 224 (14.5), 165
(10.6), 153 (12.7), 152 (18.6), 141 (10.6), 127 (11.2), 115
(14.6). HR-MS calcd.: 270.1256; found: 270.1254 ± 0.0008.
amount of acetic acid (11 µL, 0.20 mmol) was added to the
solution, and the solution was stirred overnight to check for
the presence of ground-state reactions (none were found).
The photolyses were performed using a Rayonet reactor
with 10 low-pressure mercury lamps (254 nm emission).
While the reaction was in progress, 1 mL samples were ana-
lyzed using the GC conditions outlined above.
The purified photoproducts were used to obtain calibra-
tions for the GC-FID response of each compound as a func-
tion of concentration. These calibrations were then used to
convert the peak areas from the reaction chromatograms to
concentrations. The concentrations of the components were
then converted to percentages based on the initial concentra-
tion of the starting material and the amount of material con-
sumed during the reaction.
2,4,6-Trimethoxyphenanthrene (22)
A solution of trans-1-(3,5-dimethoxyphenyl)-2-(4-meth-
oxyphenyl)ethene 20 (0.30 mg, 1.1 mmol) in methanol
(340 mL) was prepared in a large photolysis reaction vessel.
To promote the formation of the desired product, no nitrogen
was used. A 450 W, medium-pressure Hanovia mercury
lamp with a Pyrex filter (280 nm cut-off) was employed to
irradiate the solution for 7 h. The solvent was removed un-
der reduced pressure, and the residue was adsorbed onto sil-
ica and placed at the top of a dry-flash column (2 cm
diameter, 7.5 cm length). Elution with 2.5% ethyl acetate –
hexanes gave 0.10 g of a white solid. Recrystallization from
methanol gave the desired product as clear crystals (0.076 g,
Laser flash photolysis of 1 was performed using a 2.28 ×
10–4 M solution in TFE, which had an absorbance of 0.45 at
266 nm. A portion of this solution was excited using a
Continuum Nd:Yag NY-61 laser (266 nm, <8 ns/pulse,
1
25% yield), mp 111–113°C. H NMR δ: 9.08 (d, 1H, J =
2.8 Hz), 7.76 (d, 1H, J = 8.8 Hz), 7.65 (d, 1H, J = 8.6 Hz),
7.49 (d, 1H, J = 8.8 Hz), 7.17 (dd, 1H, J1 = 2.8 Hz, J2 =
8.6 Hz), 6.88 (d, 1H, J = 2.7 Hz), 6.74 (d, 1H, J = 2.7 Hz),
4.10 (s, 3H), 3.99 (s, 3H), 3.95 (s, 3H). 13C NMR δ: 161.9,
160.0, 158.2, 158.1, 136.0, 131.6, 129.4, 128.1, 124.5,
115.4, 114.7, 109.6, 101.3, 99.0, 55.8, 55.4, 55.3. GC–MS
(m/z): 269 (18.1), 268 (100.0), 225 (16.4), 210 (22.7), 152
(15.6), 139 (18.2). HR-MS calcd.: 268.1099; found:
268.1112 ± 0.0008.
≤
15 mJ/pulse). Several time domains (10 ns to 50 µs) were
used, as well as three different conditions: air-saturated, ni-
trogen-saturated, and oxygen-saturated.
Acknowledgments
We thank the Natural Sciences and Engineering Research
Council of Canada (NSERC) for financial support and
Sepracor Canada Ltd., Windsor, Nova Scotia, for the dona-
tion of chemicals. J.C.R. also thanks NSERC for a post-
graduate scholarship.
1-(3,5-Dimethoxypheny)-2-(4-methoxyphenyl)ethane (23)
A
solution of 1-(3,5-dimethoxyphenyl)-2-(4-methoxy-
phenyl)-1-methoxyethane 18a (0.0523 g, 0.173 mmol) in
methanol (50 mL) was prepared and poured into a quartz re-
action vessel. After purging with nitrogen gas for 30 min,
the stirred solution was irradiated for 5 h at 25°C. The sol-
vent was removed under reduced pressure, and the residue
was purified by column chromatography (2.5% ethyl ace-
tate– hexanes, eluant) to give the desired product (0.020 g,
References
1. S.A. Fleming and J.A. Pincock. Organic molecular
photochemistry. Vol. 3. Marcel Dekker, New York. 1999.
p. 211.
2. J.A. Pincock. Acc. Chem. Res. 30, 43 (1997).
3. R.A. McClelland. Tetrahedron, 52, 6823 (1996).
4. J.A. Pincock and P.J. Wedge. J. Org. Chem. 59, 5587 (1994).
5. H.E. Zimmerman and V.R. Sandel. J. Am. Chem. Soc. 85, 915
(1963).
1
42% yield). H NMR δ: 7.10 (d, 2H, J = 8.5 Hz), 6.83 (d,
2H, J = 8.5 Hz), 6.33 (s, 3H), 3.79 (s, 3H), 3.77 (s, 6H),
2.84 (s, 2H). 13C NMR δ: 160.7, 157.87, 144.3, 133.8,
129.4, 113.8, 106.5, 97.9, 55.3, 55.2, 38.5, 36.8. GC–MS
(m/z): 273 (2.1), 272 (11.4), 151 (2.0), 122 (9.1), 121 (100.0).
6. D.P. DeCosta, N. Howell, A.L. Pincock, J.A. Pincock, and S.
Rifai. J. Org. Chem. 65, 4698 (2000).
2-(3,5-Dimethoxypheny)-1-(4-methoxyphenyl)propane (24)
This photoproduct was not synthesized or isolated; thus,
the identification rests solely on the GC–MS spectrum (m/z):
287 (2.3), 286 (12.0), 165 (15.0), 122 (10.2), 121 (100.0).
The molecular ion of m/z 286 and the major fragment ions
of m/z 165 and m/z 122 (the two possible benzylic
carbocations that may be formed by cleavage of the molecu-
lar ion) strongly support the assignment.
7. M. Fujio, M. Goto, T. Susuki, I. Akasaka, M. Mishima, and Y.
Tsuno. Bull. Chem. Soc. Jpn. 63, 1146 (1990).
8. R.A. McClelland, C. Chan, F. Cozens, A. Modro, and S.
Steenken. Angew. Chem. Int. Ed. Engl. 30, 1337 (1991).
9. M. Hoang, T. Gadosy, G. Hedieh, D.F. Hou, A.C.
Hopkinson, L.J. Johnston, and E. Lee-Ruff. J. Org. Chem. 63,
7168 (1998).
10. P. Wan and E. Krogh. J. Am. Chem. Soc. 111, 4887 (1989).
11. G. Mladenova, L. Chen, C.F. Rodriquez, K.W.M. Sui, L.J.
Johnston, A.C. Hopkinson, and E. Lee-Ruff. J. Org. Chem. 66,
1109 (2001).
12. A. Vogel. Vogel’s textbook of organic chemistry. Longman
Scientific and Technical, London. 1989. p. 935.
13. M.R. Detty and B.J. Murray. J. Am. Chem. Soc. 105, 883
(1983).
14. P. Crotti, M. Ferreti, F. Macchia, and A. Stoppioni. J. Org.
Chem. 51, 2759 (1986).
Photolysis procedures
In all four photolysis reactions, a similar procedure was
used. The solid starting material (1: 100 mg, 0.30 mmol; 20:
50 mg, 0.20 mmol) was placed in a 100 mL volumetric
flask, and the flask was filled to the mark with the solvent of
interest. After the substrate was fully dissolved, the solution
was poured into a quartz reaction vessel and purged with
nitrogen for 30 min. The solution was thermostated at 25°C
with an immersion circulating water tube, and mixed with a
magnetic stirrer. In the reactions involving 20, an equimolar
15. W. Steglich and B. Neises. Angew. Chem. Int. Ed. Engl. 17,
522 (1978).
© 2003 NRC Canada

722
Can. J. Chem. Vol. 81, 2003
16. K. Noller, F. Dosteyn, and H. Meier. Chem. Ber. 121, 1609
(1988).
21. F.D. Lewis and J.S. Yang. J. Am. Chem. Soc. 119, 3834
(1997).
17. J.R. Falck, J. Yu, and H.S. Cho. Tetrahedron Lett. 35, 5997
(1994).
22. F.D. Lewis, W. Weigel, and X. Zuo. J. Phys. Chem. A, 105,
4691 (2001).
18. J. McEwen and K. Yates. J. Phys. Org. Chem. 4, 193 (1991).
19. F.L. Cozens, V.M. Kanagasabapathy, R.A. McClelland, and S.
Steenken. Can. J. Chem. 77, 2069 (1999).
20. J. Woning, A. Oudenampsen, and W.H. Laarhoven. J. Chem.
Soc. Perkin Trans. II, 2147 (1989).
23. B.K. Bullimore, J.F.W. McOmie, A.B. Turner, M.N. Galbraith,
and W.B. Whalley. J. Chem. Soc. C, 1289 (1967).
24. E. Mannila, A. Talvitie, and E. Kolehmainen. Phytochemistry,
33, 813 (1993).
© 2003 NRC Canada