Switchable polarity solvent (SPS) systems: Probing solvatoswitching with a spiropyran (SP)-merocyanine (MC) photoswitch

Article abstract of DOI:10.1039/c3ob41204k

The switchable polarity solvent (SPS) of 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) and an alcohol (e.g. 1-propanol) reversibly switches to a higher polarity ionic liquid, [DBUH⁺][RCO₃^-], when treated with CO₂. A long-lived species with unique properties was detected in an investigation into the use of SPS to control the lifetime of the merocyanine (MC) form in a spiropyran (SP)-MC molecular photoswitch. Irradiation of SP in 1-propanol (PrOH) in the presence of DBU generates a new species (λ_max = 420 nm). This species converts to MC upon bubbling with CO₂, which produces [DBUH⁺][PrOCOO^-]. It is proposed that a mixture of 1,2 and 1,4 alkoxide addition products form as a result of nucleophilic attack on the conjugated diene system of MC, where alkoxide formation arises from equilibration of highly basic DBU and the alcohol. These adducts revert to MC upon application of CO₂ or addition of acid. Determination of the overall equilibrium constant for alkoxide adduct formation involving DBU was afforded through Benesi-Hildebrand analysis. The Royal Society of Chemistry.

Full text of DOI:10.1039/c3ob41204k

Organic &
Biomolecular Chemistry
View Article Online
View Journal
PAPER
Switchable polarity solvent (SPS) systems:
probing solvatoswitching with a spiropyran
Cite this: DOI: 10.1039/c3ob41204k
(SP)–merocyanine (MC) photoswitch†
Alaina R. Boyd,a Philip G. Jessop,*a Julian M. Dustb and Erwin Buncel*a
The switchable polarity solvent (SPS) of 1,8-diazabicyclo[5.4.0]undec-7-ene (DBU) and an alcohol (e.g.
1-propanol) reversibly switches to a higher polarity ionic liquid, [DBUH+][RCO3−], when treated with CO2.
A long-lived species with unique properties was detected in an investigation into the use of SPS to
control the lifetime of the merocyanine (MC) form in a spiropyran (SP)–MC molecular photoswitch.
Irradiation of SP in 1-propanol (PrOH) in the presence of DBU generates a new species (λmax = 420 nm).
This species converts to MC upon bubbling with CO2, which produces [DBUH+][PrOCOO−]. It is proposed
that a mixture of 1,2 and 1,4 alkoxide addition products form as a result of nucleophilic attack on the
conjugated diene system of MC, where alkoxide formation arises from equilibration of highly basic DBU
and the alcohol. These adducts revert to MC upon application of CO2 or addition of acid. Determination
of the overall equilibrium constant for alkoxide adduct formation involving DBU was afforded through
Benesi–Hildebrand analysis.
Received 11th June 2013,
Accepted 23rd July 2013
DOI: 10.1039/c3ob41204k
www.rsc.org/obc
development has been hindered by the ready thermal rever-
sion of the open MC form to the closed SP.
Introduction
Establishment of two (or more) stable molecular states, inter-
Solvent modification,8 metal ion complexation to MC,7,10–12
convertible upon application of mild stimuli without degra- acidification,14,15 and structural modulation of the SP16 have
dation of the material, is a key requirement of practical all been explored as approaches to control reversion of MC to
molecular switching.1–5 Among our materials science research SP i.e. to extend the MC lifetime.
efforts we have been engaged in the study of spiropyran (SP)–
Our objective in the study described herein (and briefly
merocyanine (MC) interconversion as a model photoswitch.6–15 described earlier17) was to use switchable solvents18–20 to
SP–MC photoswitching (Scheme 1) involves absorption of UV control the MC lifetime. The utility of these solvents has been
light (ca. 365 nm) by the ring-closed SP (λsp = 200–300 nm in recently reviewed.21 Switchable polarity solvents (SPS), which
solution) with a concomitant C–O bond scission and ring were invented in our lab,18 are solvents that can be made to
opening to the MC (λMC = 500–610 nm).8 SP–MC photoswitch switch polarity with a mild stimulus such as the introduction
or removal of carbon dioxide. Consequently, these solvents
have been found useful in reactions such as Heck couplings22
and the polymerization of styrene in post-polymerization
removal of catalyst from polymer19 and for the extraction of oil
from algae.23 Further applications of these versatile media are
wide ranging.21
Because we have previously shown that MC lifetime is
enhanced in polar solvents,8 we wondered whether our switch-
Scheme 1 Spiropyran–merocyanine photoswitch.
able solvents might allow better control of the MC lifetime. For
example, the rate constant for MC→SP thermal reversion
drops from 1.22 × 10−1 s−1 in less-polar toluene (π* = 0.5424) to
2.85 × 10−3 s−1 in more polar dimethylformamide (DMF; π* =
0.8824), i.e. t1/2 increases from about 6 s in toluene to over
4 min in DMF. The Kamlet–Taft parameter, π*, is a polarity/
polarizability scale from 0 to 1 referenced to cyclohexane and
dimethyl sulfoxide, respectively.25 As an example of a SPS, an
equimolar mixture of 1-propanol and DBU has medium
aDepartment of Chemistry, Queen’s University, Kingston, ON, Canada K7L 3N6.
E-mail: jessop@chem.queensu.ca; Fax: +1-613-533-6669; Tel: +1-613-533-3212
bGrenfell Campus-Memorial University of Newfoundland, Corner Brook, NL, Canada
A2H 5G4. E-mail: jmdust@grenfell.mun.ca; Fax: +1-709-839-8125;
Tel: +1-709-620-6330
†Electronic supplementary information (ESI) available: Supplementary tables
and figures. See DOI: 10.1039/c3ob41204k
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
Scheme 2 DBU/alcohol switchable polarity solvent.
polarity (π* = 0.7120). Introduction of CO2 (0.1 MPa; 1 atm) con-
verts the liquid into a more polar liquid: an ionic liquid (π* =
0.9820) having protonated DBU as the cation and propylcarbo-
nate as the anion [DBUH+][PrCO3−] (Scheme 2).18
The system acts as a polarity switch; removal of carbon
dioxide via nitrogen purging restores the original DBU–PrOH
medium with its lower polarity. In terms of π*, the initial
DBU–PrOH medium is found to be slightly less polar than
DMF, but after CO2 treatment the polarity approaches that of
water. It was the aim of the current study to utilise this polarity
switching to introduce a new level of control over the lifetime
of the MC species. Therefore, enhanced merocyanine lifetime
could arise from the solvatoswitching from lower to higher
polarity media upon introduction of CO2 to DBU–PrOH and
comparable systems.
The present study sought to control the lifetime of the MC
species, formed by irradiation of 1′,3′,3′,-trimethylspiro[2H-
1-benzopyran-2,2-indoline] (6-nitroBIPS), with the use of
switchable polarity solvents. However, an unexpected and
stable product was observed when the photochromic switch
was triggered in the switchable solvent,17 as was also found by
Darwish et al.26–28 We now report further the interesting
effects of SPS switching on SP–MC interconversion and the
nature of the long-lived species formed in DBU–PrOH upon
irradiation of the SP.
Scheme 3 Schematic illustration of the SP–MC and SPS system interconver-
sions. Processes are described in text.
The overlaid spectra shown in Fig. 1 depict the combination
of photoswitching (1 → 2) and CO2 controlled SPS switching
displayed in Scheme 3. The initial spectrum in 1-propanol
alone is that of SP 1 with its major band at 350 and no signifi-
cant absorbance in the 450–650 nm region. Irradiation at
365 nm photogenerates MC 2 in PrOH with its dominant band
at 545 nm; with addition of DBU the ca. 545 nm peak
diminishes in intensity almost to baseline (Absorbance < 0.05)
along with the appearance of the 420 nm species (B, Scheme 3).
Bubbling with CO2 triggers the solvent switch from DBU–PrOH
to [DBUH+][PrCO3−]. As shown in Fig. 1, the MC 2 spectrum is
effectively restored. The presence of DBU in the system indefi-
nitely delays the ring closure of the system back to 1 by the for-
mation of the new species responsible for the 420 nm band.
Results and discussion
Overview of solvatoswitching and photoswitching
Photoswitching in PrOH, without DBU present, is accom-
plished by irradiation at 365 nm, which causes the maximum
absorbance (λmax) of the spectrum to shift from 350 nm for SP
1 to a new position at 545 nm, attributed to MC 2.
Scheme 3 depicts the four states in which this system
exists. The UV-vis spectrum of 1 in a DBU–PrOH mixture (1 : 1
v/v or 0.01 : 1 v/v) was found to have a λmax at 350 nm (A). Sol-
vatoswitching to the [DBUH+][PrCO3−] system did not affect
the UV-vis spectrum of 1 (C). However, irradiation at 365 nm
in DBU–PrOH caused a shift to 420 nm (B), as observed by
Darwish et al.,26 which did not revert to the 350 nm peak of 1
over the period of monitoring (more than one month). The
UV-vis spectrum after the introduction of CO2 to this system
has the same λmax (545 nm) as 2 in PrOH alone, and 2 reverts Fig. 1 SP, 1, in PrOH medium prior to irradiation has no significant absorbance
in 450–650 nm region (blue trace). Irradiation at 365 nm photogenerates the
back to 1 at the same rate as it would in PrOH alone (D). After
removing CO2 from the system by bubbling N2 with mild
heating (60 °C), the irradiated system is observed to absorb pri-
marily at 420 nm (B), with a small peak at 545 nm which
decays quickly leaving only the absorbance at 420 nm.
MC form (red trace, λmax = 545 nm). Addition of DBU reduces the absorbance of
the band for MC to <0.05 and generates the species that absorbs at 420 nm in
this DBU–PrOH solvent mixture (green trace). Saturation with carbon dioxide (to
switch to the DBUH+ PrOCOO− medium) results in almost complete restoration
of the band for MC, 2 (purple trace).
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
This meets, therefore, the criterion of enhanced stabilization
of one of the bi-stable states needed in molecular switching.
DBU is well-recognised both as a base30,31 and a nucleo-
phile.32 As a nucleophile, DBU could, in principle, attack the
conjugated indolium moiety in a 1,2 fashion (across the Me–
N+vC– double bond) or in a 1,4 Michael mode involving the
exocyclic double bond of 2. Hence, the MC is considered an
ambident electrophile.32 On steric grounds alone, the 1,4
addition should be preferred and would give rise to 5 as a
possibility for the 420 nm species.
Candidates for the 420 nm species
The crux of the foregoing interpretation based on Scheme 3
and illustrated by the spectra in Fig. 1 rests on the identifi-
cation of the 420 nm species.
There are a number of candidates for the 420 nm species.
In a study of a set of spiropyrans, 3, in a series of room temp-
erature ionic liquids (RTILs), Wu and coworkers29 found that
dissolution and irradiation of 3 in a family of N-alkyl-N′-methyl-
imidazolium salts (where the alkyl group was (CH2)3CH3 or
(CH2)2OH and the anions were typically alkylcarboxylates or
alkyldicarboxylates) gave a UV-vis spectrum dominated by a
420–450 nm absorbance band with a small 350–360 nm peak
for unmodified SP forms, 3. Further irradiation caused no
spectroscopic change. Here the 420–450 nm species was
ascribed to the cis-form of the relevant MC, 4, stabilized by
electrostatic and hydrogen-bonding interactions with the RTIL
including the carboxylate counterions. The 420–450 nm band
This species would derive resonance stability through delocali-
sation of the negative charge over the nitrophenoxide ring as
well as delocalization of the positive charge inside the DBU
group. A large hypsochromic shift from 545 nm of 2 to give the
observed 420 nm λmax is concordant with adduct 5, which has
curtailed conjugation compared to MC 2.
was not observed upon dissolution of the SP 3 in RTILs having
−
non-nucleophilic anions such as BF4−, PF6
.
Further support for assignment of structure 5 as the
420 nm species is afforded by the study of Baidya and Mayr
into the nucleophilicity of DBU among other bicyclic ami-
dines.32 DBU was assigned a Mayr nucleophilicity value, N, of
15.29, making it a relatively strong nucleophile. More to the
point, with a set of seven Michael acceptors DBU added in the
1,4 manner, requisite for the formation of adduct 5.
Alternative MC adducts could result from DBU acting
preferentially as a base rather than as a nucleophile. In this
case, alkoxide, RO−, would be produced by equilibration
of DBU and the alcoholic co-solvent. The position of the
equilibrium depends on the pKa of the relevant ROH. In turn,
the efficacy of trapping free MC as an adduct (e.g. 6a) via
nucleophilic addition (1,4) is tied to the concentration of alk-
oxide present.
In contrast with these results29 where the 420–450 nm species
is formed upon dissolution of 1 in the carboxylate RTILs, 1
absorbs at 350 nm in [DBUH+][PrCO3−]. Instead, the 420 nm
species in our study forms upon irradiation of 1 in the low
polarity DBU–PrOH medium or upon removing CO2 from the
[DBUH+][PrCO3−] containing 2. Noting that our system does
not reproduce the conditions found by Wu et al., it appears
unlikely that a cis-2: solvent complex comparable to 4 accounts
for our observed 420 nm band. Furthermore, spectroscopic evi-
dence in previous related work supports 2 as a more stable
trans conformer rather than a cis form;10 DFT computational
studies also support the trans form of MC as the most stable
MC conformer in general.12
As a nucleophile that is not as sterically hindered as DBU, RO−
Clearly, DBU plays an essential role in these transformations, could add either 1,4 (to give 6a) or 1,2 (to give 6b). These possi-
not just as a solvent system component.
bilities will be discussed further below.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
Acidification of 420 nm species
species (ESI, Fig. S1†). Over time the small peak for free 2 con-
tinues to decline in intensity in favour of the 420 nm peak.
With N-tert-butyl-N′,N′-dimethylformamidine (BDF; 0.01 : 1
v/v) the intensity of the 545 nm band for 2 is not suppressed to
the same extent as in the TMG–PrOH or DBU–PrOH and the
420 nm peak recorded shortly after mixing is also not as
intense (ESI, Fig. S2†).
Addition of trifluoroacetic acid (TFA) to the solution contain-
ing the 420 nm species and free MC leads to significant spec-
troscopic changes (Fig. 2). First, the 420 nm peak declines as
the intensity of the 545 nm band of free MC 2 rises. The spec-
trum is more complex, however, as a result of overlap of bands
in the 360–280 nm region.
The fact that both amidines and guanidines result in for-
mation of the 420 nm band suggests a common species, the
propoxide addition products 6a/b proposed above. It is appar-
ent that 2 can be intercepted by PrO− to form 6a/b, effectively
locking up 2 and preventing its re-cyclization to 1.
A useful comparison to the foregoing results with the ami-
dines is provided by the amines pyridine and N-ethylbutyl-
amine (ESI, Fig. S3 and S4†). The 420 nm band is not present
in the UV-vis spectrum when either N-ethylbutylamine or pyri-
dine are added to a solution containing 2 in propanol. If the
420 nm band could be assigned to a N-centred adduct compar-
able to 5 from DBU and 2, then one would expect to observe it
in the presence of less sterically hindered bases as well. The
absence of the 420 nm species here argues against such an
amine addition product as the identity for the 420 nm species.
The two sets, the amidines on one hand, and pyridine and
N-ethylbutylamine on the other, have distinctly different basi-
city ranges. In acetonitrile, DBU has been reported to have a
pKaH value of 23.0 0.1,30 24.13,33 and 23.9334 while TMG is
similarly basic in acetonitrile with a pKaH reported as 23.3.35
BDF would be expected to have a similar basicity.36 The other
amines are notably weaker bases. For pyridine, the pKaH in
As a control experiment, SP 1 was irradiated with 365 nm light
in pure EtOH to give free MC 2. Upon TFA acidification free 2
was protonated to give 2-H+, i.e., the MC form protonated on
the phenoxyl oxygen which has been previously observed.14
Comparison of the spectra from the control experiment and
those of Fig. 2 show that when TFA is added to the DBU–EtOH
system containing the 420 nm species and free 2 results in
partial reversion of the 420 nm species to 2 but also protona-
tion of some of the free MC to give 2-H+.13–15
Extension to other bases
The behaviour of MC 2 in a number of base–PrOH mixtures acetonitrile is 12.13–12.33.34,37 Although the corresponding
provide support for the structural assignment of the propoxide pKaH value for N-ethylbutylamine in acetonitrile has not been
adducts, 6a/b, and rules out an N-centred adduct comparable reported, Coetzee and Padmanabhan37 have determined the
to the DBU addition product, 5. Injection of 1,1,3,3-tetra- comparable pKaH value for diethylamine (18.75) and di-n-butyl-
methylguanidine (TMG) into a cuvette containing a 5 × 10−5
M
amine (18.31) and it is reasonable to suggest that N-ethylbutyl-
PrOH solution of 1 (0.01 : 1 v/v), which was irradiated for 90 s amine would have a pKaH value falling within this range. Thus,
to form MC 2, results in almost complete disappearance of the the amidines are significantly more basic (pKa 24–23) than
545 nm band for 2 with immediate formation of the 420 nm pyridine and N-ethylbutylamine (12.3–18.8). While the current
Fig. 2 (a) Control experiment: SP in EtOH prior to irradiation (light blue trace). The spectrum of SP does not change with the introduction of DBU into the system (red
trace). Irradiation gives both free MC at 545 nm and the species that absorbs at ca. 420 nm (green trace). Addition of trifluoroacetic acid (TFA) results in an increase in
intensity (concentration) of MC 2 with concomitant decline (and wavelength shift) for the species at about 420 nm in this medium. The 280–360 nm region reflects a
combination of absorbances for free 2 and protonated MC, i.e. 2-H+. (b) Overlay spectrum of the change over time of the SP–MC–DBU spectrum after the addition of TFA.
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
data precludes a definitive demarcation line it is clear that very coworkers have reported the use of SP 1 as a sensitive, selective
strong neutral bases are required to yield significant concen- and reversible cyanide receptor in aqueous media.38 Hence,
trations of propoxide (or alkoxide generally) leading to adduct UV irradiation (334 nm) of 1 generates the merocyanine 2 that
formation with 2 (or related MC).
is efficiently trapped by CN− as a 1,2 adduct (421 nm). Calcu-
lations support the assignment of the UV-vis spectrum as the
1,2 adduct in this case,38 as do H NMR spectra, determined
in acetonitrile-d3 solvent, that will be considered further
below.
1
Identification of the 420 nm species
The work by Darwish et al. with spiropyrans in DBU–MeOH
resulted in a species that absorbed at 432 nm, comparable to
the 420 nm species observed in the current work; the 432 nm
species formed upon irradiation of a spiropyran related to 1 in
DBU–MeOH.26 1H NMR spectra of the system (DBU–MeOH-d4)
were complex and could not be assigned definitively. However,
in the Darwish work the 432 nm peak was attributed to the
methoxide adduct comparable to 6b, the product of 1,2
addition to the indolium MeN+vC–bond.
At this stage there are two competitive candidates for the
420 nm species found in our study, formed upon irradiation of
1 in the presence of DBU–ROH: (a) a DBU adduct—likely on
steric grounds to be a 1,4 adduct, 5, from addition of DBU to
the 4-position of the open MC, 2, or, (b) 1,4- and/or 1,2-alkox-
ide adduct(s), 6a and/or 6b, respectively, formed by the conju-
gate base of the alcohol component of the relevant DBU–ROH
medium. Both could revert to the original open form 2 upon
addition of carbon dioxide to the system.
Among the evidence in support of this assignment of the
432 nm species,26 the Darwish group prepared the SP, 7, that
Saturation of the medium with carbon dioxide would both
remove DBU from the system by converting it to its conjugate
acid, DBUH+, and remove alkoxide and ROH through addition
of these to carbon dioxide to yield ROCOO−, the other com-
ponent of the RTIL medium. The decrease in concentration of
free DBU from the equilibrium between DBU and 2 and the
DBU addition product, 5, would lead to dissociation of the
adduct back to DBU and 2. Hence saturation of the DBU–ROH
system containing 5 with CO2 should regenerate the spectrum
of the MC 2 and this is observed. However, if the adducts are
alkoxide addition products, 6a and/or 6b, CO2 saturation
would give DBUH+ that could protonate the RO moiety and
convert it into a good leaving group, ROH. Alternatively,
removal of alkoxide from the system via addition to carbon
dioxide would again shift the equilibrium back to free MC, 2.
Effectively CO2 saturation of the DBU–PrOH medium is a
mechanism of formation of acid, DBUH+, that causes reversion
of the adduct(s) to free MC.
photoisomerises upon irradiation to the merocyanine
8
(530 nm in MeOH). Addition of equimolar (or greater concen-
tration) DBU gave a 413 nm band comparable to the 432 nm
peak observed in the original system by Darwish et al. Deproto-
nation of the OH of the pendant N-3-hydroxypropyl group of 8
accompanied by 1,2 addition to the indolium moiety gave the
cyclic adduct, 9 (413 nm). As in our system, addition of CO2
restored the free MC form (530 nm band for compound 8 in
methanol) in the case of the system studied by the Darwish
group (Scheme 4).26
The regiochemistry in formation of the 1,2 adduct 9 from 8
in the presence of DBU–MeOH is controlled by the length of
the pendant N-(CH2)nOH group. In this regard, the 4-position
is difficult to access by the relatively short N-propyl alkoxide
formed from the MC, 8, and would yield a less stable eight-
member ring adduct if formed.
In our system with MC, 2, in DBU–ROH either 1,2 or 1,4
addition could occur. It is pertinent that Shiraishi and
Similarly, the TFA experiments show that the 420 nm peak
declines and the 545 nm band of 2 increases in intensity with
added acid. Again, protonation of the DBU group of 5, or of
the RO function(s) of 6a/b converts these into good leaving
groups and the corresponding complexes would be expected to
dissociate to give free MC, as observed. These control experi-
ments do not differentiate between the two possibilities.
In a study of an N-ethyl carboxylic acid analog of 1,
irradiation in the presence of excess NaOCH3 gave the
deprotonated form of the MC, i.e. 10, and both 1,4 and
1,2 methoxide adducts, 11a, b (Scheme 5).39 1H NMR spectra
confirmed the structures shown. Simulated spectra (Chem-
Draw) are in major agreement with the recorded spectra; some
peaks of 11a, b are not present as a result of H/D exchange or
are buried under similar signals, e.g. OCH3 of 11a, b addition
products and the CH3O−/CH3OH of the reagent, present in
excess39 (see ESI, Tables S1 and S2†). Comparison with
spectra of the 1,2 cyanide adduct reported by Shiraishi et al.38
and the 1,2 methoxide adduct here are also in reasonable
agreement, within the differences imposed by the different
nucleophiles.
Scheme 4 SP–MC photoswitch modified with an N-3-hydroxypropyl group
(7/8). Addition of DBU causes deprotonation of the hydroxy group causing a
cyclic adduct to form (9).
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
Scheme 5 N-Ethyl carboxylic acid analog of MC (10) and the two adducts observed after the addition of NaOMe (11a/b).
The results found with the MC 10, as well as the results the double bond is more highly substituted by groups other
reported by the Darwish group,26–28 argue in favour of the than hydrogen; the 1,2 adduct contains a double bond that is
420 nm species being assigned to an alkoxide adduct or not so highly substituted. The 1,4 adduct presumably is the
adducts, 6a/b. More recently, Darwish and coworkers have more stable by analogy to common alkene systems. The struc-
reported that when a spirooxazine, structurally analogous to 1, tural reorganisation inherent in 1,4 addition, however, raises
is irradiated to photogenerate the merooxazine in DBU– the activation barrier for this mode of reaction and makes the
toluene the merooxazine forms cleanly27 without the appear- process kinetically disfavoured. Drawing the analogy to the
ance of spectroscopic bands comparable to the 420 nm absor- merocyanine systems, then, the cyanide adduct characterised
bance of the present work.
by Shiraishi et al.38 is likely also a kinetic product. With alkox-
Other evidence that argues against formation of a DBU- ides and by appeal to the results concerning 10 it seems likely
adduct, 5a, comes from the report of Muthyala et al.40 who pre- that both adducts are formed in the current case, i.e. 6a/b. The
pared an ionic liquid of the N-methylimidazolium type substi- distribution of the two regioisomers could change depending
tuted with
a
1,5,7-triazabicyclo[4.4.0]decen-5-ene (TBD) on the steric bulk of the alkoxide nucleophile.
A final point should be made concerning the UV-vis spec-
moiety, which would be expected to have basicity and nucleo-
philicity comparable to DBU. When a wide-range of substi- trum. The 420 nm band in DBU–PrOH does not change over
tuted chalcones were added to a solution of this RTIL time (>1 month observation). If the 1,2 adduct gives way to the
containing an acidic carbon acid, such as malononitrile, the more stable 1,4 adduct over time then the spectrum of both
carbanion generated by equilibration of the TBD component adducts must be virtually identical, since no change in the
of the RTIL with the active methylene compound added to the spectrum over this time can be detected.
chalcones in Michael fashion. No addition of the TBD moiety
On the basis of this assignment of the 420 nm UV-vis band
(similar to the 1,2 and/or 1,4 propoxide adducts, 6a/b) is in our system to a mixture of the two alkoxide addition pro-
reported.40
ducts, a Benesi–Hildebrand analysis44 was used, as discussed
A further issue is whether a single alkoxide adduct is next.
formed or both 1,2 and 1,4 adducts form. The evidence from
the study of the merocyanine 10 suggests that both 1,2 and 1,4
adducts could be expected to form in the present DBU–ROH
systems.39 In fact, with the model ambident electrophile
methyl vinyl ketone (3-buten-2-one) cyanide addition in the
1,2 manner is known to be kinetically controlled while 1,4
addition is thermodynamically controlled.41,42 As pointed out
by Mayr, Breugst and Ofial,43 1,4 additions are essentially
additions to a CvC and are, generally, more exergonic than
the corresponding 1,2 additions to CvO. Another way to
frame this is that 1,4 addition results in an adduct in which
Benesi–Hildebrand analysis of alkoxide adduct formation
from MC, 2
A series of spectra were recorded, starting with pure PrOH
solvent and 2 formed by irradiation of 1 (Fig. 3). Successively
increasing concentrations of DBU were introduced and each
spectrum was determined for the system that now consisted of
DBU and 2 in 1-propanol. Ratios of concentrations of DBU : 2
ranged from 5 : 1 to 150 : 1. The sequence of spectra (Fig. 3)
display the progressive decline in the 545 nm peak for 2
accompanied by the appearance and step-wise growth in the
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
Clearly, the equilibrium constant, K, determined here rep-
resents formation of the alkoxide addition products, 6a/b as a
function of DBU concentration. It is plausible that the constant
K arises from two coupled equilibria. The first generates alkox-
ide (PrO− in the current case) via equilibration of PrOH with
DBU:
K1
DBU þ PrOH
DBUHþ þ PrOꢂ
ð2Þ
ꢂ!
ꢂ
In the second equilibrium, PrO− adds reversibly to MC 2 to
give the mixture of 1,2 and 1,4 adducts, 6a/b:
K2
PrOꢂ þ MC
6a=b
ð3Þ
ꢂ!
ꢂ
Fig. 3 Series of UV-vis spectra commencing with MC (from irradiation of SP) in
pure PrOH (red spectrum), overlaid by spectra with increasing concentration
ratios of 2 : DBU (e.g. 1 : 5 orange spectrum and 1 : 100 indigo spectrum). Note
that as the ratio of DBU increases the peak at 545 nm for free MC declines in
intensity as that at 420 nm increases correspondingly. These absorbance and
concentration data were used to construct the Benesi–Hildebrand plot (eqn (5)
and figure).
Therefore, K, the equilibrium constant measured via the
Benesi–Hildebrand treatment, is an overall equilibrium con-
stant as given in eqn (4):
K
ꢂ!
DBU þ PrOH þ MC
DBUHþ þ 6a=b
ð4Þ
ꢂ
Eqn (4) is equal to the product of K1 and K2, i.e., K = K1 ×
K2. Note, as well, that the adducts 6a/b bear an overall negative
charge relative to the starting MC, 2, and eqn (4) is charge
balanced.
band at 420 nm assigned to the 1,2 and 1,4 alkoxide addition
products, 6a/b. From the absorbances for MC 2 at 545 nm and
the concentrations of 2 and DBU a Benesi–Hildebrand plot44
was constructed as per eqn (1):
Examination of K1 shows that it may be estimated from the
acid dissociation constants, Ka, for 1-propanol (Ka-PrOH) and
for the conjugate acid of DBU (Ka-DBUH+):
½MCꢀ0
1
1
1
¼
ꢁ
þ
ð1Þ
A420
Kε420 ½DBUꢀ0 ε420
KaPrOH
KaDBUHþ
K1
¼
ð5Þ
where A420 is the 420 nm absorbance of 2, K represents the
overall equilibrium constant for formation of the adducts 6a/b,
and ε420 corresponds to the molar absorptivity of the addition
product(s) band at 420 nm.
The pKa for DBUH+ has been measured by Leffek and co-
workers,30 who assigned a value of 23.9 0.1 in acetonitrile
solvent. This value is in good agreement with that more
recently determined in the same solvent by the Koppel
group.34 The Leffek group, through a correlation between the
pKa values in acetonitrile with those in water for a series of
guanidines and amidines, calculated a value for DBU in water:
14.3.30 If the pKa value for 1-propanol, anchored to water as
the standard state, is taken to be approximately 16, in line with
other common low molar mass alcohols,45 then K1 could be
The resulting Benesi–Hildebrand plot is shown in Fig. 4
and, as can be seen, shows excellent linearity (R2 = 0.999). The
equilibrium constant, K, determined this way has a value of
4.1 × 103 M−1 with a molar absorptivity for the alkoxide
adducts of 1.8 × 104 M−1 cm−1
.
estimated to be about 10−2, i.e., K1 = Ka-PrOH/Ka-DBUH+
=
10−16/10−14 = 10−2. It follows that K2, the equilibrium constant
associated with propoxide attack on 2 to give the 1,2 and 1,4
addition products, has an approximate value of 105 M−1, i.e.
K/K1 = K2. The second addition step in the overall process is
very favourable and accounts for the ready formation of the
adducts in the DBU–PrOH system upon photogeneration of 2
in this medium.
This is in accord with the experimental results where
irradiation of SP 1 gives the 420 nm species as the only band
in the DBU–PrOH system. The free MC 2 that was anticipated
to be formed was not seen. Instead, 2 is efficiently trapped as the
thermodynamically more stable propoxide adducts, 6a/b. It has
been thus shown that the lifetime of the merocyanine form 2
is extended at will.
Fig. 4 Benesi–Hildebrand plot constructed using the data shown in figure. The
equilibrium constant determined from the plot is K = 4.1 × 103 M−1 and the
molar absorptivity of the peak at 420 nm is 1.8 × 104 M−1 cm−1
.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem.

View Article Online
Paper
Organic & Biomolecular Chemistry
flasks. Samples were prepared using 2.5 mL of the desired
solvent (e.g. 1-propanol) added to the quartz cuvette using a
Conclusions
The current study probing SPS using the SP–MC molecular 2.5 mL syringe, followed by addition of 25 µL of the SP stock
photoswitch has highlighted the following conclusions: solution using a 25 µL microsyringe. Samples were irradiated
1. A long-lived species (420 nm) is formed in lower polarity for 90 s at 365 nm. For studying the effects of DBU or another
DBU–PrOH upon irradiation of SP, 1. base, 25 µL of the base was added to the cuvette, unless other-
2. The 420 nm species is proposed to comprise a mixture of wise specified. Spectra were recorded for a period of 3.5 h to
1,2 and 1,4 alkoxide adducts formed by reversible addition to obtain kinetic data.
the conjugated diene system of MC, 2.
DBU–PrOH (1 : 1), SP. To a quartz cuvette that contained
3. The adducts revert to free MC, 2, when the SPS switches 2.5 mL of DBU–PrOH (1 : 1 v/v), 25.0 μL of a 0.005 M solution
to higher polarity, DBUH+/PrOCOO− upon application of CO2.
of SP, 1, in toluene (stock solution) was injected (final concen-
4. UV irradiation (365 nm) of SP, 1, in the room temperature tration in the cell: 5.00 × 10−5 M). The spectrum was scanned
ionic liquid-like DBUH+/PrOCOO− medium gives MC 2 in and recorded (350 or 275 nm to 700 nm).
photoswitching which metamorphoses into the adducts at
DBU/PrOH with 1 irradiated to give 2. The above-mentioned
420 nm upon discharge of CO2 by N2 purging. Thus, MC 2 may cuvette was then placed in the UV-vis/fluorescence cabinet
be indefinitely locked up in the form of the long-lived adducts and irradiated at 365 nm for 20 min, at which point it was
and MC lifetime thereby extended at will.
returned to the spectrophotometer cell compartment and a
5. Thus, the present work combines two distinct areas of spectrum of the resulting yellow solution (λmax = 420 nm) was
switching: SPS18,19 and molecular photoswitching with the obtained.
SP–MC system.1–16
DBU–PrOH with 1 → DBUH+/PrOCOO−. The same cuvette
was then bubbled with carbon dioxide (2 bubbles per s) for 1 h
via a narrow gauge stainless steel needle and the cell was
then returned to the spectrophotometer. Bubbling this
solution with CO2 regenerated MC, 2, as shown by the UV-vis
spectrum.
Experimental
Materials and instrumentation
DBU (Aldrich 98%) was refluxed over CaH2, distilled in vacuo,
bubbled with CO2 for 1 h, filtered to remove the resulting
bicarbonate salt and stored in a solvent bomb under N2. 1-Pro-
panol (PrOH; Aldrich) was stored over 4 Å molecular sieves.
Trifluoroacetic acid (TFA; Aldrich) was used as purchased.
Other common solvents and bases were obtained commer-
cially and used without further purification. Supercritical CO2
(99.999%, <0.5 ppm H2O) and N2 (99.998%, <3 ppm H2O) were
used as received from Praxair. DBU–PrOH (1 : 1 v/v) was pre-
pared in a nitrogen atmosphere glove box from the appropriate
volumes of the components in a volumetric flask that was then
septum-sealed; the medium had a viscosity of 5.5 cP. The spiro-
pyran (1, 6-nitroBIPs = 6′-nitro-1,3,3-trimethylspiro(indolino-
2,2′benzopyran) was prepared by condensation of 2-methylene-
1,3,3-trimethylindoline (Aldrich) with 2-hydroxy-5-nitrobenz-
aldehyde (Aldrich), according to literature methods;6,46 after
recrystallization from isopropanol (70% yield), spectroscopic
and other physical properties were in good agreement with lit-
erature.47 NMR spectra of samples of 1 dissolved in (CD3)2SO
DBU–PrOH with 1 irradiated then treated with TFA. Two
experiments were conducted to determine whether TFA would
disrupt the MC–DBU system. First, a cuvette was prepared as
stated previously of 5 × 10−5 M SP in ethanol and the spectrum
was acquired. An equal molar amount of DBU was added and
new spectra were acquired both before and after irradiation.
An equal molar amount of TFA was then added and the sub-
sequent spectrum acquired.
To show the changes over time, overlay spectra were
acquired over a period of 88 min. Two stock solutions were pre-
pared; a stock solution of 0.1 M DBU was prepared in propa-
nol, and one of 0.1 M TFA was prepared in toluene. A cuvette
was prepared of 5 × 10−5 M SP in propanol with 25 µL of the
DBU stock solution. The sample was irradiated and the spec-
trum was acquired, after which 25 µL of the TFA stock solution
was introduced into the sample and the spectrum was
recorded as shown in Fig. 2.
Benesi–Hildebrand experiments
were recorded using
a Bruker-Avance 400 spectrometer
A series of spectra were recorded (separate cuvettes) starting
with SP, 1 (5.00 × 10−5 M final concentration) in pure PrOH.
Generally, after mixing, the sample was irradiated at 365 nm
for 90 s and the spectrum of MC and the 420 nm species (if
present) then recorded. Solvent media were prepared by
mixing appropriate volumes of SP-containing PrOH and DBU
such that the following ratios of 2 (=SP initial) : DBU were
obtained: 1 : 5, 1 : 10, 1 : 25, 1 : 50, 1 : 75, 1 : 100 and 1 : 150.
Each sample was scanned from 350 to 750 nm. These spectra
are compiled sequentially in Fig. 4. Data from Fig. 3 (i.e. absor-
(400.0 mHz, 1H; 100.0 mHz, 13C).
UV-vis spectra (scans and single wavelength) were recorded
using a HP 8542 photodiode array spectrophotometer, thermo-
statted to 25 °C ( 0.5 °C). UV-vis irradiation of samples was
performed in a Spectroline CX Series ultraviolet fluorescence
analysis cabinet, equipped with a Spectroline XX-15N UV lamp
(dual 15 W, 365 nm) and the sample was irradiated for 90 s.
Solvent switching UV-vis experiments
General spectrophotometric method. Stock solutions of SP bance for MC, 2, at 545 nm) were used to construct the
(5 × 10−3 M) were prepared in toluene in 5 mL volumetric Benesi–Hildebrand graph (Fig. 4) according to eqn (1).
Org. Biomol. Chem.
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Paper
23 A. R. Boyd, P. Champagne, P. J. McGinn, K. M. MacDougall,
J. E. Melanson and P. G. Jessop, Bioresour. Technol., 2012,
118, 628–632.
Acknowledgements
The authors gratefully acknowledge funding from NSERC (EB,
PJ), The Canada Research Chairs Program (PJ), Xerox Research 24 M. J. Kamlet, J. L. M. Abboud, M. H. Abraham and
Centre Canada (EB, PJ), Grenfell Campus Vice President’s
Research Fund (JD), Walter C. Sumner Foundation (AB).
R. W. Taft, J. Org. Chem., 1983, 48, 2877–2887.
25 M. J. Kamlet, J. L. Abboud and R. W. Taft, J. Am. Chem.
Soc., 1977, 99, 6027–6038.
26 T. A. Darwish, R. A. Evans, M. James, N. Malic, G. Triani
and T. L. Hanley, J. Am. Chem. Soc., 2010, 132, 10748–
10755.
Notes and references
1 S. Z. Janicki and G. B. Schuster, J. Am. Chem. Soc., 1995, 27 T. A. Darwish, R. A. Evans, M. James and T. L. Hanley,
117, 8524–8527. Chem.–Eur. J., 2011, 17, 11399–11404.
2 B. L. Feringa, Molecular switches, Wiley-VCH, Weinheim, 28 T. A. Darwish, R. A. Evans and T. L. Hanley, Dyes Pigm.,
2001. 2012, 92, 817–824.
3 F. Pina, M. J. Melo, M. Maestri, R. Ballardini and 29 Y. Wu, T. Sasaki, K. Kazushi, T. Seo and K. Sakurai, J. Phys.
V. Balzani, J. Am. Chem. Soc., 1997, 119, 5556–5561.
4 B. Champin, Chem. Soc. Rev., 2007, 36, 358–366.
5 R. S. Stoll, Angew. Chem., Int. Ed., 2010, 49, 5054–5075.
Chem. B, 2008, 112, 7530–7536.
30 K. T. Leffek, P. Pruszynski and K. Thanapaalasingham,
Can. J. Chem., 1989, 67, 590–595.
6 S.-R. Keum, K.-B. Lee, P. M. Kazmaier and E. Buncel, Tetra- 31 D. Basavaiah, A. J. Rao and T. Satyanarayana, Chem. Rev.,
hedron Lett., 1994, 35, 1015–1018. 2003, 103, 811–892.
7 J. T. C. Wojtyk, E. Buncel and P. M. Kazmaier, Chem. 32 M. Baidya and H. Mayr, Chem. Commun., 2008, 1792–
Commun., 1998, 1703–1704. 1794.
8 J. T. C. Wojtyk, A. Wasey, P. M. Kazmaier, S. Hoz and 33 I. Kaljurand, T. Rodima, I. Leito, I. A. Koppel and R.
E. Buncel, J. Phys. Chem., 2000, 104, 9046–9055. Schwesinger, J. Org. Chem., 2000, 65, 6202–6208.
9 S. Swansburg, E. Buncel and R. P. Lemieux, J. Am. Chem. 34 I. Kaljurand, A. Kütt, L. Sooväli, T. Rodima, V. Mäemets,
Soc., 2000, 122, 6594–6600.
10 J. T. C. Wojtyk, P. M. Kazmaier and E. Buncel, Chem.
Mater., 2001, 13, 2547–2551.
I. Leito and I. A. Koppel, J. Org. Chem., 2005, 70, 1019–
1028.
35 I. M. Kolthoff, M. K. Chantooni and S. Bhowmik, J. Am.
Chem. Soc., 1968, 90, 23–28.
11 Z. Tian, R. A. Stairs, M. Wyer, N. Mosey, J. M. Dust,
T. M. Kraft and E. Buncel, J. Phys. Chem., 2010, 114, 11900– 36 H. Junge, A. Boddien, F. Capitta, B. Loges, J. R. Noyes,
11909.
S. Gladiali and M. Beller, Tetrahedron Lett., 2009, 50, 1603–
12 G. Cottone, Chem. Phys. Lett., 2004, 388, 218–222.
1606.
13 N. Xiao, Y. Chen, R. Lemieux, E. Buncel, G. Iftime and 37 J. F. Coetzee and G. R. Padmanabhan, J. Am. Chem. Soc.,
P. M. Kazmaier, Mol. Cryst. Liq. Cryst., 2005, 431, 337–344. 1965, 87, 5005–5010.
14 J. T. C. Wojtyk, A. Wasey, N.-N. Xiao, P. M. Kazmaier, 38 Y. Shiraishi, K. Adachi, M. Itoh and T. Hirai, Org. Lett.,
S. Hoz, C. Yu, R. P. Lemieux and E. Buncel, J. Phys. Chem.
A, 2007, 111, 2511–2516.
15 J. Hobley, Phys. Chem. Chem. Phys., 1999, 1, 3259–3267.
16 A. Kellmann, F. Tfibel, E. Pottier, R. Guglielmetti, A. Samat
2009, 11, 3482–3485.
39 A. Wasey, PhD Thesis, Queen’s University, 2003.
40 M. K. Muthyala, B. S. Chhikara, K. Parang and A. Kumar,
Can. J. Chem., 2012, 90, 290–297.
and M. Rajzmann, J. Photochem. Photobiol. Chem., 1993, 76, 41 J. Clayden, N. Greeves, S. Warren and P. Wothers,
77–82.
Organic chemistry, Oxford University Press, New York, NY,
17 A. R. Boyd, E. Buncel and P. G. Jessop, Proceedings of the
2001.
92nd Canadian Chemistry Conference and Exhibition, 42 W. Nagata and M. Yoshioka, in Organic Reactions, John
Hamilton, 2009.
18 P. G. Jessop, D. J. Heldebrant, X. Li, C. A. Eckert and 43 H. Mayr, M. Breugst and A. R. Ofial, Angew. Chem., Int. Ed.,
C. L. Liotta, Nature, 2005, 436, 1102–1102. 2011, 50, 6470–6505.
19 L. Phan, D. Chiu, D. J. Heldebrant, H. Huttenhower, 44 H. A. Benesi and J. H. Hildebrand, J. Am. Chem. Soc., 1949,
Wiley & Sons, Inc., 2004.
E. John, X. Li, P. Pollet, R. Wang, C. A. Eckert, C. L. Liotta
and P. G. Jessop, Ind. Eng. Chem. Res., 2008, 47, 539–545.
20 P. G. Jessop, D. A. Jessop, D. Fu and L. Phan, Green Chem.,
2012, 14, 1245.
71, 2703–2707.
45 F. A. Carroll, Perspectives on Structure and Mechanism in
Organic Chemistry, Solutions Manual, John Wiley and Sons,
2011.
21 P. G. Jessop, S. M. Mercer and D. J. Heldebrant, Energy 46 S.-R. Keum, M.-S. Hur, P. M. Kazmaier and E. Buncel,
Environ. Sci., 2012, 5, 7240–7253. Can. J. Chem., 1991, 69, 1940–1947.
22 R. Hart, P. Pollet, D. J. Hahne, E. John, V. Llopis-Mestre, 47 S.-R. Keum, K.-B. Lee, P. M. Kazmaier, R. A.
V. Blasucci, H. Huttenhower, W. Leitner, C. A. Eckert and
C. L. Liotta, Tetrahedron, 2010, 66, 1082–1090.
Manderville and E. Buncel, Magn. Reson. Chem., 1992,
30, 1128–1131.
This journal is © The Royal Society of Chemistry 2013
Org. Biomol. Chem.