The structure, photochemical reactivity, and photophysical properties of adamantyl X-substituted aryl ethers and a comparison with the alkyl groups, methyl, tert-butyl, and allyl

Article abstract of DOI:10.1139/v05-117

The structure, photophysical properties, and photochemistry of the adamantyl aryl ethers 1 in both methanol and cyclohexane have been examined. UV absorption spectra, ¹³C NMR chemical shifts, X-ray structures, and Gaussian calculations (B3LYP/6-31G(d)) indicate that these ethers adopt a 90° conformer in the ground state. In contrast, fluorescence spectra, excited singlet state lifetimes, and calculations (TDDFT) indicated a 0° conformer is preferred in the first excited singlet state S₁. Irradiation in either solvent results in the formation of adamantane and the corresponding phenol as the major products, both derived from radical intermediates generated by homolytic cleavage of the ether bond. The 4-cyano substituted ether 1j was the only one to form the ion-derived product, 1-methoxyadamantane (16% yield), on irradiation in methanol. Rate constants of bond cleavage for these ethers from S1 were estimated by two different methods by comparison with the unreactive anisoles 2, but the effect of substituents was too small to determine structure-reactivity correlations. The temperature dependence of the quantum yields of the fluorescence of the unsubstituted, 4-methoxy and 4-cyano derivatives of 1 and 2 were also determined. These results indicated that the activated process for 1 was mainly bond cleavage for the 4-cyano substrate whereas for 2, it was internal conversion and intersystem crossing.

Full text of DOI:10.1139/v05-117

1237
The structure, photochemical reactivity, and
photophysical properties of adamantyl X-
substituted aryl ethers and a comparison with
the alkyl groups, methyl, tert-butyl, and allyl1
A.L. Pincock and J.A. Pincock
Abstract: The structure, photophysical properties, and photochemistry of the adamantyl aryl ethers 1 in both methanol
and cyclohexane have been examined. UV absorption spectra, 13C NMR chemical shifts, X-ray structures, and Gaussian
calculations (B3LYP/6-31G(d)) indicate that these ethers adopt a 90° conformer in the ground state. In contrast, fluo-
rescence spectra, excited singlet state lifetimes, and calculations (TDDFT) indicated a 0° conformer is preferred in the
first excited singlet state S1. Irradiation in either solvent results in the formation of adamantane and the corresponding
phenol as the major products, both derived from radical intermediates generated by homolytic cleavage of the ether
bond. The 4-cyano substituted ether 1j was the only one to form the ion-derived product, 1-methoxyadamantane (16%
yield), on irradiation in methanol. Rate constants of bond cleavage for these ethers from S1 were estimated by two dif-
ferent methods by comparison with the unreactive anisoles 2, but the effect of substituents was too small to determine
structure–reactivity correlations. The temperature dependence of the quantum yields of the fluorescence of the unsub-
stituted, 4-methoxy and 4-cyano derivatives of 1 and 2 were also determined. These results indicated that the activated
process for 1 was mainly bond cleavage for the 4-cyano substrate whereas for 2, it was internal conversion and
intersystem crossing.
Key words: aryl ether photochemistry, fluorescence, excited-state rate constants, excited-state temperature effects.
Résumé : On a étudié la structure, les propriétés photophysiques et la photochimie des oxydes d’adamantyle et
d’aryles 1, tant dans le méthanol que dans le cyclohexane. Les spectres d’absorption UV, les déplacements chimiques
en RMN du 13C, les structures obtenues par diffraction des rayons X et les calculs gaussiens (B3LPYP/6-31G(d)) indi-
quent que dans leur état fondement ces éthers adoptent un conformère à 90°. Par opposition, les spectres de fluores-
cence, les temps de vie de l’état singulet excité et les calculs (TDDFT) indiquent qu’un conformère à 0° est privilégié
dans le premier état singulet excité, S1. Les produits principaux d’une irradiation dans l’un ou l’autre des solvants sont
l’adamantane et le phénol correspondant qui dérivent tous les deux d’intermédiaires radicalaires générés par un clivage
homolytique de la liaison éther. L’irradiation dans le méthanol du dérivé 1j portant un substituant 4-cyano est le seul à
conduire à la formation du 1-méthoxyadamantane (rendement de 16 %), le produit dérivé de l’ion. On a utilisé deux
méthodes différentes pour évaluer les constantes de vitesse de clivage de la liaison de ces éthers à partir de S1 par
comparaison avec des anisoles non réactifs (2); l’effet des substituants est toutefois trop faible pour déterminer des cor-
rélations structure–activité. On a aussi établi la relation entre la température et le rendement quantique de fluorescence
des dérivés des composés 1 et 2, non substitués ou portant des substituants 4-méthoxy ou 4-cyano. Ces résultats indi-
quent que le processus activé pour le composé 1 est le clivage de la liaison pour le substrat 4-cyano alors que pour le
composé 2 il s’agit de la conversion interne et d’un passage intersystème.
Mots clés : photochimie d’éthers aromatiques, fluorescence, constantes de vitesse d’états excités, effets de température
sur l’état excité.
[Traduit par la Rédaction] Pincock and Pincock 1252
Introduction
glet states of substrates of the general class ArY–Z, where
the Y—Z bond is the reactive one, continue to attract inter-
est. The objective is to develop excited state structure–
reactivity relationships, like the Hammett equation, that are
well-established for ground-state reactions. Related to this
The effect that substituents (X) on aromatic rings (Ar)
have on the quantum yields and rate constants of bond cleav-
age reactions from photochemically generated excited sin-
Received 9 December 2004. Published on the NRC Research Press Web site at http://canjchem.nrc.ca on 12 October 2005.
A.L. Pincock and J.A. Pincock.2 Department of Chemistry, Dalhousie University, Halifax, NS B3H 4J3, Canada.
1This article is part of a Special Issue dedicated to organic reaction mechanisms.
2Corresponding author (e-mail: james.pincock@dal.ca).
Can. J. Chem. 83: 1237–1252 (2005)
doi: 10.1139/V05-117
© 2005 NRC Canada

1238
Can. J. Chem. Vol. 83, 2005
Scheme 1. Mechanism and products formed from the homolytic cleavage of alkyl aryl ethers and rate constants of reaction for S1.
T1
No reaction
O
O
H
O H
O
R
O
R
kisc
R
R
hν
kp
kr
H
+
S1
kf + kic
X
X
X
X
X
+ Other isomers
+ Other isomers
kir
O
C(CH3)3
O
CH2CH=CH2
O
CH3
X
X
X
O
2
3
4
a: X = H
f: X = 4-F
g: X = 3-F
h: X = 4-CF3
i: X = 3-CF3
j: X = 4-CN
k: X = 3-CN
X
b: X = 4-OCH3
c: X = 3-OCH3
d: X = 4-CH3
e: X = 3-CH3
1
goal has been one of understanding, in the excited-state
cases, the competition between the formation of the products
derived from radical pairs (from homolytic cleavage of the
Y—Z bond) and ion pairs (from heterolytic cleavage of the
Y—Z or from rapid, exergonic redox electron transfer in the
first formed radical pair). A general review of the mechanis-
tic principles involved and a survey of the observed reactiv-
ity of the Z groups for arylmethyl cases (Y = CH2) have
been published (1).
We now summarize results for four series of the class of
compounds ArO-C (1–4), i.e., phenolic ethers where Y = O
and Z = C. New results will be given for the adamantyl aryl
ethers 1, in comparison with the anisoles 2. Previously, we
have reported on the photochemistry and photophysics of the
tert-butyl (3) (2) and allyl ethers (4) (3). For these latter two
cases, the products of photolysis in either methanol or cyclo-
hexane were, in general, those expected from homolytic
cleavage of the carbon–oxygen bond to form a radical pair
followed by in-cage coupling (photo-Fries for 3, photo-
Claisen for 4) or disproportionation (Scheme 1). For both,
the reactions were shown to occur from the excited singlet
state (S1) by selective triplet quenching studies.
where σhν values were available. This correlation was sur-
prising because the σhν scale was developed from results for
the photoprotonation of styrenes and phenylacetylenes, reac-
tions that result in positive charge increases at the benzylic
carbon in the rate-determining transition state, 5 for
styrenes. A similar transition state (6) for the photo-Fries re-
action of the tert-butyl ethers 3 seems less likely than one of
opposite polarity (7). However, a very poor (nonexistent)
correlation of the rate constants for 3 with the values of σex
(5–7), which are derived from the excited state pKas of phe-
nols 8, was obtained.
δ+
H
δ_
δ+
CH CH2
δ+
O
R
X
X
X
X
5
6
8
δ_
O
δ_
O
δ+
δ+
R
H
Rate constants kr for the reaction from the excited singlet
state for the tert-butyl compounds 3 were determined by the
7
classic quantum yield (Φ) of the reaction method (eq. [1])
r
kr
kf + kic + kisc + kr
kr
(2). The rate constants for S1 of the reaction (kr), fluores-
cence (kf), internal conversion (kic), and intersystem crossing
(kisc) are shown in Scheme 1. The sum of these is kt, the re-
ciprocal of τs, the excited singlet state lifetime. Measurement
[1]
Φr =
=
= kr τs
k
t
While studying the tert-butyl ethers 3 (2), we also exam-
ined the 4-cyano compound 1j from the adamantyl deriva-
tives. An interesting result was that on photolysis of 1j in
methanol, the ion-derived product 1-methoxyadamantane
was obtained in 16% yield. This compound was synthesized
of Φ by the disappearance of starting material and τs by flu-
r
orescence decay gives kr. The rate constants obtained in this
way gave a reasonable correlation with σhν (4) and ρ = –0.77
(r = 0.975) for seven of the 10 compounds of set 3 studied
© 2005 NRC Canada

Pincock and Pincock
1239
Scheme 2. Products of the photolysis of adamantyl ethers 1 in methanol and cyclohexane.
OH
+
+
hν
R
X
H
O
CH3OH or
C6H6
11
12
13a: R = CH2OH
13b: R = C6H11
O C6H11
X
1a - 1k
+
+
H11C6 C6H11
X
CH3O
10
14
15
by reaction of the lithium salt of 1-adamantanol with 4-
fluorobenzonitrile (8), a nucleophilic aromatic substitution
reaction that relies on the electron-withdrawing nitrile group
in the para position. This procedure was unsuccessful for 3-
fluorobenzonitrile. Therefore, the other substituted deriva-
tives of 1 were not previously studied.
We now report on the synthesis, photochemistry, photo-
physical properties, and excited-state rate constants for the
complete set of the adamantyl ethers 1. These compounds
have an advantage that the fate of the non-aryl fragment of
the ether cleavage reaction can be easily observed. This was
not so for the very volatile fragments (CH3OC(CH3)3,
CH3OCH2CH=CH2, etc.) expected from sets 3 and 4.
of the ethers 1, their GC retention times were useful. They
could easily be removed by flash chromatography on silica
gel and the ethers purified by crystallization. The yields after
purification ranged from 20% to 35%. Characterization data
are given in the Experimental section; 13C NMR data, along
with those of sets 2, 3, and 4, which will be discussed in the
following, are given in Table S1 (Supporting information).3
Photochemistry of 1-adamantyl aryl ethers 4a–4k in
methanol and cyclohexane
The ethers 1 were photolyzed in either methanol or cyclo-
hexane with 254 nm lamps in a Rayonet reactor at 25 °C.
The reactions were reasonably efficient with little depend-
ence on the substituent. For instance, for approximately
100 mg samples, 83% of the unsubstituted compound 1a and
80% of the 4-methoxy compound 1b in 100 mL of methanol
had reacted after 30 min. The reactions were a factor of two
to three times slower in cyclohexane. In both solvents in all
cases, except for the 4-cyano compound 1j in methanol, the
products were those expected from the formation of radicals
by homolytic cleavage of the C—O bond (Scheme 2). As re-
ported previously (2), 16% of the methyl ether 10 was
formed from 1j in methanol indicating a minor intervention
of a pathway involving ion pairs. In agreement with this ob-
servation, 4-cyanophenoxide has been used as a leaving
group for the generation of diphenylmethyl cations (observ-
able by nanosecond LFP) from diphenylmethyl ethers (10),
Results and discussion
Synthesis of 1-adamantyl aryl ethers (1a–1k)
The ethers were synthesized by reaction of 1-bromoada-
mantane (by heating in pyridine) with the corresponding
substituted phenols according to a method described previ-
ously (9) (eq. [2]). The reaction mixtures were not homoge-
neous. The reactions were usually monitored by GC-FID
until the 1-bromoadamantane (the limiting reagent) had dis-
appeared. The crude product always contained varying
amounts of the isomeric (GC–MS) 1-adamantylphenol deriv-
atives 9, which were not characterized. However, because
these isomers were possible products in the photochemistry
OH
[2]
+
+
Pyridine
Br
O
X
OH
X
X
1a - 1k
9
3 Supplementary data for this article are available on the Web site or may be purchased from the Depository of Unpublished Data, Document
Delivery, CISTI, National Research Council Canada, Ottawa, ON K1A 0S2, Canada. DUD 4020.
© 2005 NRC Canada

1240
Can. J. Chem. Vol. 83, 2005
and in the pioneering study by Zimmerman and Somasekhara
(11) on photochemically generated triphenylmethyl cations.
The major product from the adamantyl fragment of the
ethers 1 for all substrates in both solvents was adamantane
11, derived from the 1-adamantyl radical and accounting for
70%–90% of the mass balance. This is reasonable on the ba-
sis of the fact that hydrogen atom abstraction by the 1-
adamantyl radical from both methanol and cyclohexane is
somewhat more favourable than by the tert-butyl radical
(C–H BDE = 99, 96, 99, and 96 kcal/mol for 1-adamantane,
methanol, cyclohexane, and 2-methylpropane, respectively)
(1 cal = 4.184 J) (12). The major product from the aryl frag-
ment was the corresponding phenol 12, but these gave very
broad and tailing peaks in the GC traces that could not be re-
liably quantified. Photo-Fries products (GC–MS, 15% yield
overall, presumably 2- and 4-(1-adamantyl)phenol) were
only observed for 1a in methanol. Low yields of products
13a, 13b, 14, and 15 (Scheme 2) were also formed from the
solvent-derived radicals.
basis sets but not with others (13). For instance, at the
RHF/6-31G* level, 17 is 1.39 kcal/mol above 16, and the
transition state between the two (Φ = ~65°) is 1.59 kcal/mol
above 16. These same calculations give C-O-C1 bond angles
of 119.8° and 115.5° for 16 and 17, respectively. A con-
formational minimum similar to 16 has been found for allyl
phenyl ether (4a: X–H) (15).
In terms of the photochemical reactivity of ethers 1, 3,
and 4 (as discussed later, the anisoles are photochemically
unreactive), 16 has the reactive C—O bond perpendicular to
the π system of the aromatic chromophore whereas 17 has
the same C—O bond parallel to it. Therefore, 17 should be
more reactive. As discussed now, we have used a variety of
probes to assess the conformational geometry of aryl ethers
in the ground state and conclude that the tertiary alkyl deriv-
atives 1 and 3, in contrast to the anisoles 2 and the allyl
ethers 4, have preferred geometries analogous to 17.
UV absorption spectra of ethers 1–4 in methanol and
cyclohexane
The long wavelength absorption band λmax (Lb transition)
and molar absorptivity (⑀) for the 1-adamantyl ethers 1a–1k,
and for comparison, the corresponding anisoles 2a–2k in
both methanol and cyclohexane as the solvent are reported
in Table 1. Values for the tert-butyl ethers 3a–3i in methanol
and 3a–3c in cyclohexane are also included. The Lb band for
the anisoles has a significantly higher ⑀, and the λmax tends
to occur at longer wavelengths than the tertiary alkyl cases 1
and 3. These trends are shown graphically in Fig. 1 for X =
H, 4-OCH3, and CN4 in both methanol (left in each case)
and cyclohexane (right in each case). The spectra for the
corresponding allyl derivatives, which have been reported
previously (3), are also included in these figures. The spectra
for X = 4-CN and 4-CF3 are not typical of those for the
other substitutents because these electron-withdrawing
groups interact strongly with the ether oxygen and shift the
La band to longer wavelengths, partially obscuring the Lb
band. The primary ethers (allyl (4) and methyl (2)) have
very similar spectra that differ significantly (higher ⑀ values,
longer wavelengths) from those of the tertiary alkyl ethers 1
and 3.
Geometry of aryl ethers 1–4 in the ground state
The geometry and rotational barriers for the methoxy
group of anisole have been extensively studied by both ex-
perimental and computational methods. Two limiting struc-
tures have been considered: 16 with a C(H3)-O-C1-C2
dihedral angle of Φ = 0° and 17 with the same angle equal to
90°. For 16, an sp2 hybridization model predicts conjugative
overlap between the oxygen centered lone pair in a π orbital
and the π system of the aromatic ring as indicated by the
resonance structures 16a and 16b. For 17, either an sp2 or
sp3 hybridization model predicts a weaker interaction.
Clearly, a vast body of experimental evidence demonstrates
that the methoxy group in anisole is a π donor and therefore
16 must be the minimum energy structure on the rotational
surface. A more difficult question to answer is the position
of 17 on this surface.
Most experimental studies have indicated a twofold rota-
tional surface with 17 as a maximum at ~4 kcal/mol above
16 (13). Therefore, rotation between the two minima (Φ = 0°
and 180°) would be rapid on the NMR timescale at room
temperature. However, long-range C–H NMR coupling con-
stants demonstrate that 17 is in a shallow minimum and
therefore the rotational surface is fourfold (14). Calculations
by ab initio methods have located this minimum with some
Similar observations have been made previously for
anisoles that are sterically crowded by ortho substitution at
C-2 and C-6 (16). For instance, the λmax of the Lb band in
H
C1
H
C1
O
O
C2
H
CH3
C2
H
CH3
16 (Φ = 0ο)
17 (Φ = 90ο)
O
O
O
CH3
CH3
CH3
16b
16a
4 These three were chosen to span the complete range of electron-donating and electron-withdrawing abilities of all the substitutents studied.
© 2005 NRC Canada

Pincock and Pincock
1241
Table 1. Absorbance and fluorescence data and excited-state rate constants for ethers 1–3 in methanol and cyclohexane.
b
⑀
kf
b
c
λmax
Solventa (nm)
((mol/L)–1 λ0,0
E(S1)
τsf
kt
kf(exp)
(calcd)
kf(exp)/kf k Mt – kt
Φfe
X
cm–1)
(nm) (kJ/mol)
(ns) (107s–1) (107s–1) (107s–1)
(calcd)g
(107s–1)
2.9
328
1680
519
505
1930
574
16.1
13.2
1a
2a
3a
1a
2a
3a
H
H
H
H
H
H
M
M
M
C
C
C
256
271
269
270
271
269
280
280
281
427
427
0.23
6.21
7.58
3.7
3.2
0.9
2.4
3.97
1.33
0.24h
279
281
431
427
0.21
0.29i
7.00
7.80
14.3
12.8
3.0
3.7
1.1
3.6
2.63
1.03
1.5
2.7
1b 4-OCH3
2b 4-OCH3
3b 4-OCH3
1b 4-OCH3
2b 4-OCH3
3b 4-OCH3
1c 3-OCH3
2c 3-OCH3
3c 3-OCH3
1c 3-OCH3
2c 3-OCH3
3c 3-OCH3
1d 4-CH3
2d 4-CH3
1d 4-CH3
2d 4-CH3
1e 3-CH3
2e 3-CH3
1e 3-CH3
2e 3-CH3
M
1894
299
305
402
393
0.085
0.11
2.50
2.68
40.0
37.3
3.4
4.1
2.7
4.3
1.24
0.93
M
M
C
290
280
282
289
280
272
274
272
272
274
272
269
279
270
279
266
273
266
273
2820
2000
2190
3200
2140
1930
2200
1723
1910
2220
1777
649
298
304
402
383
0.090
0.12
3.12
3.02
32.0
33.1
2.6
4.0
3.6
5.5
0.72
0.71
–1.1
–2.0
–2.3
C
C
M
M
M
C
281
285
427
419
0.084
0.076
2.54
2.42
39.3
41.3
3.8
3.1
2.9
3.1
1.32
1.00
283
284
423
423
0.106
0.093
2.63
2.48
38.0
40.3
4.0
3.7
3.0
3.5
1.32
1.09
C
C
M
M
C
288
289
288
290
283
283
283
282
414
414
414
414
423
423
423
423
0.162
0.22
0.226
0.33
0.17
0.27
0.23
0.37
5.91
6.37
6.62
7.48
6.90
8.0
16.9
15.7
15.1
13.3
14.4
12.4
13.0
11.3
4.4
3.5
3.4
4.4
2.4
3.4
3.0
4.2
1.0
3.1
1.3
3.9
0.8
2.8
0.9
3.4
4.36
1.12
2.58
1.13
3.08
1.21
3.26
1.24
1.2
1.8
2.0
1.7
2.2
1880
759
C
2170
452
M
M
C
1660
486
7.69
8.79
C
1810
1f
2f
3f
1f
2f
4-F
4-F
4-F
4-F
4-F
M
M
M
C
264
280
269
271
281
263
269
263
263
269
260
1240
2670
1200
1550
2900
1344
1590
947
284
291
423
410
0.14
0.15
3.44
3.72
29.1
26.9
4.0
4.0
1.9
4.1
2.13
0.98
289
290
277
279
414
414
431
431
0.14
0.18
0.0047
0.0055
3.54
3.86
—
29.2
25.9
—
4.0
4.0
—
2.6
4.9
1.9
2.3
1.56
0.94
—
3.3
C
1g 3-F
2g 3-F
3g 3-F
1g 3-F
2g 3-F
1h 4-CF3
M
M
M
C
C
M
—
—
—
—
905
1650
434
276
278
282
435
431
423
0.014
0.022
0.073
—
0.8
3.76
—
125.0
26.5
—
—
1.9
1.7
2.9
—
—
1.0
—
4.1
2h 4-CF3
3h 4-CF3
1h 4-CF3
2h 4-CF3
M
M
C
270
260
262
270
267
1060
551
279
431
0.093
4.47
22.4
2.1
—
—
433
281
280
289
288
427
427
414
414
0.10
0.13
0.037
0.11
4.35
5.05
2.90
3.16
22.9
19.8
34.4
31.6
2.3
2.6
1.3
3.5
—
—
3.1
2.8
C
1100
1180
—
—
1i
3-CF3
M
2.0
3.7
0.65
0.95
2i
3i
1i
3-CF3
3-CF3
3-CF3
M
M
C
277
273
268
2310
1163
1290
288
289
414
414
0.056
0.16
3.17
3.41
31.5
29.3
1.8
3.2
2.5
4.4
0.72
0.73
2.2
2i
3-CF3
C
267
2440
1j
2j
1j
2j
4-CN
4-CN
4-CN
4-CN
M
M
C
C
M
274 (sh) 1230
271 (sh) 2030
267 (sh) 1150
286
286
287
286
305
419
419
419
419
393
0.051
0.075
0.083
0.11
2.30
3.37
4.41
5.88
2.97
43.5
29.6
22.6
17.0
33.6
2.2
2.2
1.9
1.9
3.1
—
—
—
—
2.2
—
—
—
—
13.9
5.6
274
286
1310
1440
1k 3-CN
0.094
1.41
7.5
© 2005 NRC Canada

1242
Can. J. Chem. Vol. 83, 2005
Table 1 (concluded).
b
⑀
kf
(calcd)
(ns) (107s–1) (107s–1) (107s–1)
((mol/L)–1 λ0,0
E(S1)
(nm) (kJ/mol)
τsf
kt
kf(exp)
kf(exp)/kf k Mt – kt
b
c
λmax
Solventa (nm)
Φfe
X
cm–1)
(calcd)g
(107s–1)
2k 3-CN
1k 3-CN
2k 3-CN
M
C
C
290
286
288
2960
1400
2980
305
300
300
393
398
398
0.14
0.14
0.20
3.60
4.09
4.83
27.7
24.4
20.7
3.9
3.5
4.1
4.5
2.4
5.1
0.87
1.46
0.80
3.7
aM is methanol, C is cyclohexane.
bObtained from absorption spectra.
cObtained from the overlap of the absorption and fluorescence spectra.
dCalculated from Es = 12.0 × 104/λ0,0
.
l
eQuantum yield of fluorescence relative to anisole in methanol (0.24) and cyclohexane (0.29), estimated error 10%.
fSinglet lifetime by nanosecond single photon counting. The standard deviations of the fits to the experimental counts are <2%. No value indicates that
the lifetime was less than 0.5 ns.
gThe values of kf (calcd) could not be obtained for X = 4-CF3 and 4-CN because the La band was at a long enough wavelength so as to obscure the
short wavelength portion of the Lb band, preventing the use of eq. [5].
hG. Kohler, G. Kittel, and N. Getoff. J. Photochem. 18, 19 (1985).
iI.B. Berlman. Handbook of fluorescence spectra of aromatic molecules. Academic Press, New York. 1971. p. 139.
2,6-dimethylanisole shifts to shorter wavelengths (266 nm)
and lower molar absorptivity (620 (mol/L)–1 cm–1) compared
to anisole itself (269 nm, 2100 (mol/L)–1 cm–1 (16); 271 nm,
1980 (mol/L)–1 cm–1 (Table 1),3 both in cyclohexane). The
proposed explanation is that ortho substitution forces the
methoxy group into the 90° geometry of 17, reducing the
importance of structures 16a and 16b that allow conjugation
of the oxygen lone pair with the aryl ring.
bulky groups have at C-2 and C-6 is even larger (changes in
δ values ranging from –10.5 to –11.5 ppm for the adamantyl
ethers relative to the anisoles), but these changes probably
reflect to some extent a direct steric interaction rather than
only the importance of resonance effects.
Molecular orbital calculations
Two views of each of the two stable conformers for 1-
adamantyl phenyl ether (1a: X = H) are shown in Figs. 2 and
3, the former corresponding to the 0° structure 16 and the
latter to the 90° structure 17. These were obtained as geome-
try optimized minima by GAUSSIAN 98 (20) at the
B3LYP/6-31G(d) level of theory.5 Similar calculations were
performed for the para-substituted cases shown in Table 2
with OH replacing OCH3 for computational convenience.
Frequency analyses indicated that all were minima with all
frequencies positive except for small negative values corre-
sponding to methyl rotations for X = CH3. The lowest en-
ergy positive frequencies (~35 cm–1 for 16 and ~15 cm–1 for
17) correspond, by visualization in GaussView, to the rota-
tion interconverting 16 and 17.
13C NMR spectra of ethers 1–4 in CDCl3
Complete 13C NMR spectral data for the adamantyl ethers
1 along with those for the anisoles 2, tert-butyl ethers 3, and
the allyl ethers 4, are given in Table S1 (Supporting informa-
tion).3 The anisole data were taken from the Aldrich compi-
lation (17) and these values were confirmed for the para-
substituted compounds by another literature source (18). The
allyl ether data, which differ insignificantly from the anisole
data, have been reported previously (3). Relative to the
anisole derivatives, the 13C chemical shifts at C-4 for both
the adamantyl and tert-butyl derivatives have higher δ values
(deshielded) by an average of 2.57
0.52 ppm (11 com-
pounds) and 2.97 0.38 (nine compounds), respectively.
After correction for zero point energy, the 90° conformer
is calculated to be the more stable for all cases although the
difference in energy between the two decreases systemati-
cally as the substituent changes from the resonance electron-
donating 4-hydroxy group to the electron-withdrawing
4-cyano group. Clearly, this difference is dominated by the
steric preference for the 90° geometry, which moves the
ortho hydrogens on the benzene ring away from the
adamantyl hydrogens. This steric crowding in the 0° struc-
ture is indicated by the large calculated O-C1(aryl)-C2(aryl)
and C1(adamantyl)-O-C1(aryl) bond angles of 127.5° and
126.8°, respectively, at the ether functional group in contrast
to 120.1° and 102.7° for the corresponding angles in the 90°
structure. The energetic preference for the 90° conformer is
decreased by conjugation of the oxygen lone pair of the
ether with the substituent so that the 0° and 90° conformers
are calculated to have almost identical stabilities for X =
4-CN. A related observation has been made previously for
diphenyl ethers, monosubstituted on one of the two rings
(21). The more stable conformer changes from that ring hav-
Moreover, these changes vary around these averages in a
random way over the range of X groups from electron-
donating ones (CH3O, CH3) to electron-withdrawing ones
(CF3, CN) in either the meta or para position. Therefore, the
contribution that resonance structure 16b makes is not more
important for a resonance electron-withdrawing group like
CN.
For ortho-substituted anisoles, the same effect has been
observed previously (19) with changes in 13C δ values at C-4
relative to anisole of –0.3 ppm (2-methyl), +3.9 ppm (2,6-
dimethyl), +7.2 ppm (2,6-diisopropyl), and +9.7 ppm (2,6-
di-tert-butyl). These deshielding effects are a consequence
of the lowering of the π-electron density at C-4 and reflect
the decreasing importance of the resonance contributor 16b
as conformation 17 becomes more dominant. Therefore, the
bulky tert-butyl and 1-adamantyl alkyl groups are providing
a similar steric inhibition to a resonance interaction of the
ether oxygen with the benzene ring as the case for two ortho
methyl groups on the benzene ring. The effect that these
5 The optimized geometries in C1 symmetry have actual dihedral angles of 0.1° for the “0°” structure and 92.7° for the “90°” structure.
© 2005 NRC Canada

Pincock and Pincock
1243
Fig. 1. UV absorption spectra of the ethers 1–4 for X = H, 4-OCH3, and 4-CN in methanol (right) and cyclohexane (left).
1a
2a
3a
4a
1a
2a
3a
4a
1600
1200
800
400
0
2000
1500
1000
500
0
240
250
260
270
280
240
250
260
270
280
Wavelength (nm)
Wavelength (nm)
3
3000
2000
1000
1b
2b
3b
4b
1b
2b
3b
4b
3000
2000
1000
0
0
250
260
270
280
290
300
310
250
270
290
310
Wavelength (nm)
Wavelength (nm)
1500
1000
500
2000
1j
2j
4j
1j
2j
4j
1000
0
0
275
280
285
290
295
275
280
285
290
295
300
Wavelength (nm)
Wavelength (nm)
ing conjugation to the ether oxygen for 4-CN (0°) to that
conjugation being reduced for 4-OH (90°).
A scan of the dihedral angle from –105° to 105° (optimiz-
ing all other variables) for the unsubstituted case 1a gives
the energy surface shown in Fig. 4. A very shallow
minimum is observed at the 0° geometry with a barrier of
approximately 0.2 kJ/mol that separates it from the two
identical minima at 92°. This plot is quite similar to that
© 2005 NRC Canada

1244
Can. J. Chem. Vol. 83, 2005
Table 2. Calculated (GAUSSIAN 98) values for aryl adamantyl ethers 1 at the two minimum energy conformers.
Dihedral
angle (°)
E(SCF)a
(Hartree)
∆Eb
(kJ/mol)
E(SCF + ZPE)c
(Hartree)
∆E(ZPE)b
(kJ/mol)
E(Ex)d
(kJ/mol)
Osc.
strength
∆E(S1)b
(kJ/mol)
4-X
0
90
0
90
0
OH
OH
F
F
CH3
–772.19879
–772.20166
–796.21732
–796.21948
–736.30263
–771.86533
–771.86863
–795.89594
–795.89863
–735.94560
535.8
552.5
552.5
570.1
552.5
0.083
0.059
0.069
0.039
0.053
–8.0
–10.8
–8.4
–7.5
–5.7
–8.7
–7.1
CH3
90
–736.30448
–4.8
–735.94797
–6.2
567.0
0.016
H
H
CN
CN
0
90
0
–696.98521
–696.98691
–789.23183
–789.23141
–696.65781
–696.65781
–788.90365
–788.90378
565.4
577.4
536.3
539.4
0.035
0.009
0.287
0.234
–6.3
–2.1
–4.5
–1.1
–6.1
–0.4
90
aCalculated at optimized geometries by B3LYP/6-31G(d).
bThe negative values indicate the more stable structure of the pair for each X substituent.
cAfter zero point energy correction at optimized geometry.
dCalculated by TDDFT/6-31G(d) at the optimized ground-state geometry.
Fig. 3. Two views of the B3LYP/6-31G(d) optimized geometry
with a C1(adamantyl)-O-C1(aryl)-C2(aryl) dihedral angle of
92.7°. The O-C1(aryl)-C2(aryl) and the C1(adamantyl)-O-
C1(aryl) bond angles are 120.1° and 102.7°, respectively.
Fig. 2. Two views of the B3LYP/6-31G(d) optimized geometry
with a C1(adamantyl)-O-C1(aryl)-C2(aryl) dihedral angle of 0.1°.
The O-C1(aryl)-C2(aryl) and the C1(adamantyl)-O-C1(aryl) bond
angles are 127.5° and 126.6°, respectively.
calculated for anisole with the contrast that the minimum en-
ergy conformation (90° for 1 and 0° for anisole) is compli-
mentary for the two cases.
using both excited singlet state lifetimes and steady-state
spectra, and MO calculations. The conclusion is that the 0°
structure is energetically preferred for S1.
X-ray structures of adamantyl ethers 1b (X = 4-OCH3)
and 1h (X = 4-CF3)
Suitable crystals were obtained for X-ray crystallograph
structures6 for these two compounds and that for 1b is
shown in Fig. 5. In both cases, a geometry similar to the 90°
structure is preferred; the observed dihedral angles are 99°
for 1b and 107° for 1h. These derivations from the symmet-
rical 90° structures are likely due to intermolecular crystal
packing forces.
Excited singlet state lifetimes of ethers 1 and 2
The total rate of decay (kt = 1/τs) of the excited singlet
state (τs from nanosecond single photon counting) along
with the fluorescence quantum yields are reported in Table 1
in both methanol and cyclohexane solvents for the
adamantyl 1 and methyl ethers 2. The rate constants of fluo-
rescence (kf = Φ kt) are also given. Averaged over all substi-
f
tuted compounds, the ratios of kf(2)/kf(1) = 1.19 0.35 and
1.26 0.24 (10 compounds) in methanol and cyclohexane,
respectively, are equal to unity, within experimental error. In
fact, these ratios are improved to 1.02 0.15 and 1.20
Geometry of the aryl ethers 1–4 in the excited singlet
state
Information about the geometry of aryl ethers in the ex-
cited state can be obtained from fluorescence measurements
6 T.S. Cameron. Department of Chemistry, Dalhousie University. Unpublished results. 2004.
© 2005 NRC Canada

Pincock and Pincock
1245
Fig. 4. Relative energy vs. the C1(adamantyl)-O-C1(aryl)-
C2(aryl) dihedral angle for adamantyl phenyl ether by B3LYP/6-
31G(d).
Fig. 5. X-ray crystal structure of 1h (X = CF3).
5
4
3
2
1
0
-1
-120
-80
-40
0
40
80
120
Dihedral angle (°)
0.23, respectively, if the values for X = 3-CF3, which seem
unusually large, are not included in the two averages.
The general theoretical relationship (22, 23) between the
radiative lifetime (τf = 1/kf) and the absorbance spectrum is
given in eq. [3] where n is the refractive index of the solvent,
υf is the expectation value for the frequency of the fluores-
cence spectrum, and the final term is the integrated
absorbance spectrum. The constant gives kf in units of s–1 if
the frequency values are expressed in wavenumber (cm–1),
and molar absorptivity ⑀ in the usual units ((mol/L)–1 cm–1).
This equation is derived on the basis of the oscillator
strength model and the important assumption that the excited
singlet state and the ground state have similar geometries. It
has been used successfully in the past for simple aromatic
compounds like benzene, toluene, and ortho-xylene (24),
and by us for the anisoles 2 and the allyl ethers 4 (3).
values obtained experimentally and from eq. [3] for the
anisoles indicates that the ground-state and excited-state ge-
ometries are similar, presumably the 0° structure 16. In con-
trast, for the adamantyl ethers, the experimental kf values
compare well with the anisoles, whereas those calculated
from eq. [3] do not. This implies that the geometry of the
adamantyl ethers changes on excitation from the 90° geome-
try in S0 to the 0° geometry in S1.
Fluorescence spectra of ethers 1 and 2
kf = 2.88 ×10−9n2 < νf > ⑀ ln ν
Values for the λ0,0 (from the overlap of emission and exci-
tation spectra), derived singlet state energies (ES1 (kJ/mol) =
[1.197 × 105]/λ0,0 (nm)), and fluorescence quantum yields
for the anisoles 2 and adamantyl ethers 1 are reported in Ta-
ble 1. Sample fluorescence spectra for the ethers 1 and 2
with X = H, 4-OCH3, and 4-CN4 are shown in Fig. 6, for
both methanol (left in each case) and cyclohexane (right in
each case). The spectra for the corresponding allyl ethers 4
are also included.
[3]
∫
Equation [3] works very well for the anisoles 2 in both
methanol and cyclohexane, kf (exp)/kf (calcd) = 1.05 0.10
(seven compounds) and kf (exp)/kf (calcd) = 0.96 0.13 (10
compounds), respectively. These values are in Table 1.
In contrast, as shown in Table 1, eq. [3] does not work as
well for the adamantyl ethers 1, with kf (exp)/kf (calcd) =
2.27
1.14 (eight compounds) and kf (exp)/kf (calcd) =
1.98 0.83 (eight compounds) in methanol and cyclohex-
ane, respectively. The larger scatter and the average signifi-
cantly greater than unity is a consequence of the decreases
in ⑀ for the adamantyl ethers. In general, the lower ⑀ values
decrease kf (calcd) values and consequently increase ratios of
kf (exp)/kf (calcd). As discussed above, these lower values of
⑀ for the adamantyl ethers 1 were ascribed to these ethers
preferring the 90° geometry rather than the 0° geometry.
The similarity in the experimental kf values for the ani-
soles 2 and the adamantyl ethers 1 suggest similar excited-
state geometries for both. The good agreement among the kf
Two features are obvious from the spectra. The first is the
very similar wavelength dependence unrelated to the struc-
tural change in the alkyl group (from adamantyl to methyl to
allyl) in the ether. The only significant exception, a shift to a
longer wavelength of about 5 nm for the adamantyl ether,
occurs when
X = 4-CN in Fig. 6, a case where
intramolecular charge transfer in the excited state may be
important. Other than this case, the spectra imply that the
excited-state geometries of the ethers 1, 2, and 4 are similar.
This suggestion is reinforced by the similar values of ES1 ob-
tained for the anisoles and the adamantyl ethers.
© 2005 NRC Canada

1246
Can. J. Chem. Vol. 83, 2005
Fig. 6. Fluorescence spectra of the ethers 1, 2, and 4 for X = H, 4-OCH3, and 4-CN in methanol (left) and cyclohexane (right).
0.3
0.2
0.1
0.3
0.2
0.1
1a
1a
2a
2a
4a
4a
4a x 10
4a x 10
4a x
10
4a x 10
0
0
275
295
315
335
355
375
270
290
310
330
350
370
Wavelength (nm)
Wavelength (nm)
3
0.15
0.15
0.12
0.09
0.06
0.03
1b
1b
2b
2b
0.12
0.09
0.06
0.03
4b
4b
4b x 10
4b x 10
4b x 10
4b x 10
0
0
280
330
380
280
330
380
Wavelength (nm)
Wavelength (nm)
0.12
0.08
0.04
0.08
0.06
0.04
0.02
1j
2j
4j
1j
2j
4j
0
0
280
330
380
280
300
320
340
360
380
Wavelength (nm)
Wavelength (nm)
The second feature of these spectra is the difference in
quantum yields of fluorescence. In general, the adamantyl
ethers 1 and the anisoles 2 have similar Φfs in either solvent,
the adamantyl ether values being somewhat lower. In con-
trast, the allyl ethers 4 have significantly lower Φf values
than both 1 and 2. These decreases for compounds 4 reflect
the very large increases in reactivity of the excited singlet
state relative to the anisoles 2. As will be described in the
© 2005 NRC Canada

Pincock and Pincock
1247
following, according to eq. [4], the higher values of kr for
the allyl ethers result in higher values of kt, and, conse-
quently, lower values for Φf = kf /kt, kf being essentially con-
stant for any given substituent X but independent of the
static (Φ ) or dynamic (τs) measurements of fluorescence can
f
be used although the latter is preferred because more reliable
experimental values can be obtained. The assumption used is
given in eqs. [4] and [5], where the changes in kt of the reac-
tive series only differ from those (ktM) of the model series 2
by kr . Therefore, subtraction of experimentally determinable
values (kt – ktM = kr) should reliably give kr. Because these
are exclusively properties of the excited singlet state, the kr
values obtained are independent of kir , the internal return
process.
alkyl group of the ether. The similar values of Φ for the
f
adamantyl ethers and the anisoles indicate that the kr values
for the adamantyl compounds do not make a large contribu-
tion to the kt of S1.
[4]
kt = kf + kic + kisc + kr
This method was applied quite successfully to obtain kr
values for the allyl ethers 4 in both cyclohexane and metha-
nol (3). From absorbance spectra (λmax, ⑀), fluorescence
spectra (λ0,0, ES1), and calculated kf values, good evidence
was provided that the sets 2 and 4 have very similar excited-
state properties for any given equivalently substituted pair in
either methanol or cyclohexane. The resulting kr values as a
function of substituents spanned almost two orders of mag-
nitude (108–1010 s–1), and for some cases (3- and 4-methoxy
and 3- and 4-methyl), were the dominant mode of decay of
S1. A plot of kr in methanol vs. kr in cyclohexane was linear
and the slope (0.96) indicated a parallel substituent effect in
the two solvents. This observation demonstrates that only
radical pair intermediates are being generated, ion pairs not
being possible in cyclohexane.
An examination of the kt (adamantyl ethers 1) and ktM
(anisoles 2) values in Table 1 indicates that the difference re-
quired to usefully apply eqs. [4] and [5] is hardly outside ex-
perimental error for most substitutents. Small positive values
are obtained in most cases, although a few are actually nega-
tive. This observation quantitatively demonstrates that the
adamantyl ethers have very small values of kr, the rate con-
stant of reaction, relative to the other modes of decay of the
excited singlet state (kic, kf, and kisc). The only noticeable
trend in the kr values is towards higher values for those com-
pounds 1j and 1k with cyano substitutents at C-3 and C-4,
respectively, more so in methanol than in cyclohexane, but
particularly for the 4-CN compound in methanol. This cor-
rectly accounts for the observation that the 4-cyano com-
pound gives a significant yield of the ion-derived product 10
in methanol. Unfortunately, overall, the changes in kr are too
small to attempt a quantitative substituent effect correlation.
Molecular orbital calculations
Excitation energies to S1 were calculated by TDDFT/6-
31G1(d) (20) using vertical excitations from the optimized
ground-state geometries calculated previously for both the
Φ = 0° and 90° conformer, again for the substituents shown
in Table 2. In all cases, as expected, the calculated oscillator
strength was lower for the 90° than the 0° conformer. More-
over, also in agreement with the UV absorption spectra, the
excitation energy was higher for the Φ = 90° geometry than
for the 0° geometry. Combined with the ground-state ener-
gies, the relative energies for the excited singlet state
(∆E(S1)) in Table 1 can be calculated indicating that, for all
substitutents, S1 for the 0° conformer is more stable than S1
for the 90° conformer. This observation supports the sugges-
tion that excitation of compounds 1 from S0 at Φ = 90° gives
S1 at 90°, which then relaxes by rapid rotation to give the
more stable conformer S1 at 0°. Fluorescence emission and
reaction should then occur from the 0° conformer.
Rate constants of reaction (kr) from the excited singlet
state of the adamantyl ethers 1
The quantum yield method, as described by eq. [1], for
obtaining rate constants for excited-state bond cleavage reac-
tions has a major defect. In all cases where experiments have
been designed to detect it, internal return of radical pair in-
termediates (kir in Scheme 1) has been observed (1). There-
fore, the efficiency of product formation that a quantum
yield measures does not give a reliable measure of kr . In ef-
fect, kir is a form of internal conversion, although not occur-
ring directly from S1. A knowledge of the fraction of this
internal return [kir /(kp + kir)], which is undoubtably depend-
ent on the substituent X, could be used to obtain correct val-
ues of kr, but this fraction is a difficult number to obtain
experimentally without a suitable probe. For instance, in the
cases examined for the allyl ethers (4a–4d) in both methanol
and cyclohexane, the importance of kir was demonstrated by
the fact that the quantum yield method (eq. [1]) gave signifi-
cantly lower values of kr than the unreactive model method
(eqs. [4] and [5]) (3), described in the following. With no
suitable probe for the adamantyl ethers 1, we decided not to
measure quantum yields of reaction.
The temperature dependence of the fluorescence method
The effect of temperature on photochemical reactions has
received far less attention than it deserves. However, recent
studies from the laboratories of Lewis (activation barriers to
E to Z isomerization in phenyl substituted alkenes) (25) and
Zachariasse (thermally activated internal conversion in amino
substitued aromatics) (26) have shown that the measurement
of fluorescence quantum yields and lifetimes and inter-
system crossing quantum yields over a wide temperature
range provides a valuable tool to obtain rate constants for
excited singlet state processes. Again, because the properties
measured are exclusively those of S1, the rate constants kr
obtained are independent of processes like kir or reaction
from T1. We have recently applied this approach to the
phototransposition reactions of substituted benzenes (27).
The simplest assumption used is presented in eq. [6],
which is the same as eq. [4] except that the reaction process
kr is considered to follow the Arrhenius expression. For
[5]
ktM = kf + kic + kisc
The unreactive model method
For the allyl ethers 4, we recently described an alternate
method of measuring kr for reactions of S1 that relies on
comparing a photochemically reactive compound with an
unreactive one that has identical excited-state properties (3).
We chose to compare compounds 4 with the anisoles 2,
which are, in comparison, photochemically inert. Either
© 2005 NRC Canada

1248
Can. J. Chem. Vol. 83, 2005
substituted benzenes, the S0 to S1 energy gap is large so that
kic ꢀ 0 (kic ꢀ 103 s–1 for benzene with E(S1) = 444 kJ/mol).7
Because S1 → T1 is an exergonic process and considered to
be unactivated, the rate of intersystem crossing should be
temperature independent. However, Zachariasse and co-
workers (26) have clearly shown that for amino substituted
aromatics, S1 → Tn intersystem crossing can be an important
process that will be activated. Therefore, the total rate of
intersystem crossing should be expressed by eq. [7], where
k 0isc is the intersystem crossing rate constant at 0 K (temper-
ature independent), and the Arrhenius activation term is for
the activated kisc . The activation parameters for this process
are characterized by low values of Aisc ꢀ 108 to 109 s–1 and
Ea/isc ꢀ 4–6 kJ/mol.
of temperature. As shown in Table 3, this results in low A
values (2.6–8.6 × 109 s–1) that are characteristic of an acti-
vated intersystem crossing process and indicate that the
dominant temperature-dependent process for these substrates
even at room temperature is kisc, according to eq. [7]. This
conclusion was also reached by a previous study on the tem-
perature dependence of kt (29), in 3-methylpentane and etha-
nol–methanol (1:1), although the A values (~2.5 × 108)
reported are even lower than ours. The values of kisc ob-
tained by us for anisole at 25 °C of (10.0 1.2) × 10–7 s–1
(methanol) and (9.1 0.9) × 10–7 s–1 (methylcyclohexane)
are in good agreement with literature values of 8.0 × 10–7 s–1
(cyclohexane) (30) and 9.4 × 10–7 s–1 (isooctane) (31), both
obtained from directly measured Φisc /τs values. Moreover,
the contribution that kr makes to the total rate of decay of S1
[6]
[7]
kt = kf + kic + kisc + A exp(–Ea /RT)
kisc = k 0isc + Aisc exp(–Ea/isc /RT)
for anisole is small because Φ + Φ = 0.24 + 0.74 (29)
isc
(Ermolaev’s rule) in agreementfwith the observation that ir-
radiation of anisoles does not result in product formation.
Very likely, the dialkoxy compounds 2, with low A and Ea
Values for kt as a function of temperature have normally
been obtained by measuring S1 lifetimes (τs), but this is a
very time-consuming process. We have chosen a more rapid
method using τs and Φf at 25 °C (Table 1) to obtain kf =
Φf /τs. Then measurement of relative fluorescence quantum
yields over a wide temperature range (–25 to 65 °C in meth-
anol and –25 to 95 °C in methylcyclohexane)8 were done
(see Experimental section), and kt(T) = Φf (T) /Φf (25 °C)
values obtained assuming kf is constant, independent of tem-
perature. In fact, although k 0f (the superscript refers to the
refractive index (n) of vacuum) is temperature independent,
the usual experimental observation for solution measurements
is given in eq. [8] where the refractive index of the solvent is
temperature dependent. However, over the temperature range
studied here, the change in n2 for the two solvents used is in-
significant. From literature values9 of the variation in refrac-
tive index with temperature (28), the kf values for anisole
would only change from 3.1 × 10–7 to 3.3 × 10–7 s–1 from
–25 to +65 °C in methanol, and from 3.5 × 10–7 to 3.9 ×
10–7 s–1 from –25 to 95 °C in methylcyclohexane. These
ranges do not exceed the estimated error of 10% in kf (as-
suming an error of 10% in Φf), and consequently, were ig-
nored.
values, also have Φ + Φ ꢀ 1, and therefore they have Φ
ꢀ
isc
isc
0.9 in both solventfs, on the basis of Φ ꢀ 0.1 (Table 1).
f
As shown in Fig. 7, 4-cyanoanisole has a larger change in
kt as a function of temperature than does anisole even though
both are photochemically unreactive in methanol. This sug-
gests that both kisc and kr are contributing to the temperature
dependence, but that the kr process does not lead to prod-
ucts. As has been shown previously (26), to separate the ac-
tivated kisc and kr rate constants reliably would require a
combination of eqs. [6] and [7] and measurements over a
much wider range of temperatures so that the process with
lower A and Ea values (kisc) would be dominant at low tem-
peratures and kr (higher A and Ea values) would be dominant
at higher temperatures. Quantum yields of intersystem cross-
ing as a function of temperature would also be required.
However, we have recently shown (27) that substituted
aromatics undergo an activated process to form prefulvene
biradicals. These intermediates (18–20), as in Scheme 3 for
4-methylbenzontrile (32), provide a pathway for both photo-
transposition reactions (ortho, meta, para) and also a path-
way for kir, i.e., a chemical rather than photophysical route
for internal conversion. We have also estimated (27) a quan-
tum yield of Φ = 0.29 for the formation of 18, and measured
[8]
kf = k 0f n2
r
a quantum yield of formation of the meta isomer of Φp =
0.025 from the para one. Therefore, less than 10% of the in-
termediate 18 formed undergoes rearrangement to 19. We
suggest that formation of a prefulvene biradical is also the
activated kr process for 4-cyanoanisole. A similar activated
internal conversion has been observed (26) for 4-
cyanoaniline (and other aromatic amines in hydrocarbon sol-
vents) with Ea = 35 kJ/mol and A ꢀ 5 × 1012 s–1. The authors
noted that an activated, but undefined, geometric deforma-
tion in S1 must be providing a pathway for S1 to S0 decay.
If this proposal is correct, then the difference in kr be-
tween the anisoles and the adamantyl ethers obtained from
the variable temperature studies reflects the increased reac-
tivity of the adamantyl ethers because of C—O bond cleav-
age. This suggestion assumes that the activated kisc and the
Plots of kt vs. T are nonlinear and examples are shown in
Fig. 7 for anisole, 4-cyanoanisole, and 4-cyanophenyl
adamantyl ether (1j) in methanol. Other results are compiled
in Table 3 for the adamantyl ethers 1 and the corresponding
anisoles 2 for X = H, 4-OCH3, and 4-CN.4 Excellent fits of
this data to eq. [6] were obtained (r > 0.997 in all cases) giv-
ing an intercept equal to kf + kisc (assuming kic = 0) and
Arrhenius A and Ea values. Using the independently deter-
mined experimental values of kf gives kisc. The consistency
of the results is indicated by the similar values of kisc ob-
tained for any given pair of 1 and 2 with the same sub-
stituent and in either solvent.
Many of the compounds studied behave in a similar way
to anisole in Fig. 7, with very little change in kt as a function
7 See, in particular, the Introduction section of ref. 26.
8 For the variable temperature studies, methylcyclohexane, rather than cyclohexane, was used as the solvent to expand the range of experi-
mentally accessible temperatures.
9 This reference gives values of dn/dt = –0.000 39/°C for methanol and –0.000 55/°C for cyclohexane.
© 2005 NRC Canada

Pincock and Pincock
1249
Fig. 7. Temperature dependence of kt and the fits to eq. [6] for the compounds 1j (X = 4-N), 2j (X = 4-CN), and 2a (X = H) in methanol.
8 × 108
4-Cyanophenyl adamantyl ether 1j
7 × 108
4-Cyanoanisole 2j
Anisole 2a
6 × 108
5 × 108
4 × 108
3 × 108
2 × 108
1× 108
0
240
260
280
300
320
340
T (K)
4-Cyanophenyl adamantyl ether
4-Cyanoanisole
Anisole
y = m1+m2*exp(-m3/(8.314e-3*T))
y = m1+m2*exp(-m3/(8.314e-3*T))
Value Error
1.7267e+08 4.6846e+06
y = m1+m2*exp(-m3/(8.314e-3*T))
Value Error
Value
Error
m1
m2
1.9576e+08
8.318e+06
m1
m2
m3
m1
m2
m3
6.8121e+07 6.2294e+06
1.8841e+09 4.9421e+08
4.4908e+11 1.0285e+11
2.0355e+11
18.377
5.105e+10
0.73277
NA
m3
18.657
2.107e+14
0.99965
0.66811
NA
8.4225
0.90553
NA
Chisq
R
Chisq 6.3681e+13
0.99957
Chisq 6.8794e+12
0.99914
NA
R
NA
R
NA
other kr processes (for instance, formation of the prefulvene
biradical) are similar in rate for the equivalently substituted
adamantyl ether and anisole. The difference in the tempera-
ture dependence between 4-cyanoanisole (2j) and adamantyl
4-cyanophenyl ether (1j) is clearly shown in Fig. 7, and the
values of ∆kr (25 °C) in the last column of Table 2 support
this suggestion. The values are larger in methanol than
methylcyclohexane for all substrates studied, in agreement
with the observation that the efficiency of photochemical
conversion of the adamantyl ethers is greater in methanol.
Moreover, both the substituted adamantyl ethers (4-OCH3,
4-CN) react faster than the unsubstituted one. Most notable
is the high kr value for the 4-CN substrate in methanol, pre-
sumably as a result of the kr process leading to ion pairs
from S1 and the resulting product 1-methoxyadamantane.
ygen with the aryl ring. Upon excitation to S1, the geometry
relaxes to a 0° conformer that increases the conjugation but
decreases the overlap between the aromatic π system and the
bond that breaks in the excited state. This observation pro-
vides an explanation for the low photochemical reactivity of
these ethers and the lack of any large substituent effect on
the rate constant of bond cleavage from S1. For substrates
that do not react, this technique provides a simple method of
determining kisc and Φ . The method also allows a compari-
son between unreactiviescmodel compounds (the anisoles) and
reactive substrates (the adamantyl ethers) so that the kr pro-
cess that leads to products or reactive intermediates can be
separated from that which leads back to starting material
(kir), a form of internal conversion.
Experimental
Conclusions
General procedures
1
A variety of techniques have demonstrated that the
adamantyl ethers 1 have a ground-state geometry with a di-
hedral angle of 90° that reduces conjugation of the ether ox-
Routine H and 13C NMR spectra were obtained in CDCl3
on a Bruker AC 250 F NMR spectrometer in automation
mode. The 13C NMR spectra for 1a–1j in Table S1 (Sup-
© 2005 NRC Canada

1250
Can. J. Chem. Vol. 83, 2005
Table 3. Fits to eq. [6] and derived rate constants for S1 of adamantyl ethers 1 and anisoles 2.
Ab
(109s–1)
Ea
kf + kisc
kfd
(107s–1)
kisc
ktf
(107s–1)
krg
(107s–1)
∆krh
b
c
e
X
Solventa
(kJ/mol)
(107s–1)
(107s–1)
(107s–1)
3.0 (1.3)
1.8 (0.4)
24 (1.0)
9.2 (1.4)
8.4 (0.9)
16.0 (1.0)
8.7 (1.0)
6.8 (0.2)
10.6 (0.4)
9.3 (0.2)
17.7 (0.1)
5.0 (1.4)
3.6 (0.9)
7.6 (0.7)
5.6 (0.6)
7.4 (1.2)
6.4 (0.3)
3.7 (0.5)
3.5 (0.3)
25.7 (0.8)
1.0 (1.5)
0.2 (0.8)
8.9 (1.3)
1a
2a
1a
2a
1b
H
H
H
H
M
M
MC
MC
M
3.7 (0.4)
3.2 (0.3)
3.0 (0.3)
3.7 (0.4)
16.1 (0.2)
13.2 (0.1)
14.3 (0.1)
12.8 (0.1)
40.0 (0.4)
3.8 (0.6) 11.6 (0.6)
4-OCH3
4.9 (0.3)
3.6 (0.5)
4.4 (0.4)
7.7 (0.2)
7.6 (0.6)
8.5 (0.4)
3.4 (0.3) 14.3 (0.4)
4.1 (0.4) 16.4 (0.5)
2.6 (0.3) 14.8 (0.4)
4.0 (0.4) 17.6 (0.6)
2b
1b
2b
4-OCH3
4-OCH3
4-OCH3
M
20.5 (0.1)
17.4 (0.1)
21.6 (0.3)
37.3 (0.4)
32.0 (0.3)
31.3 (0.3)
16.8 (0.5)
14.6 (0.4)
9.7 (0.5)
MC
MC
4.9 (0.9)
8.6 (0.6) 10.7 (0.2)
1j
2j
1j
2j
4-CN
4-CN
4-CN
4-CN
M
M
MC
MC
450 (10)
18.7 (0.7)
18.4 (0.7)
21.0 (1.0)
14.1 (0.9)
19.6 (0.1)
17.3 (0.5)
17.3 (0.4)
13.0 (0.3)
2.2 (0.2) 17.4 (0.3)
2.2 (0.2) 15.1 (0.7)
1.9 (0.2) 15.4 (0.6)
1.9 (0.2) 11.1 (0.5)
43.5 (0.4)
29.6 (0.3)
22.6 (0.2)
17.0 (0.2)
23.9 (0.5)
12.3 (0.8)
4.3 (0.6)
4.0 (0.5)
11.6 (1.3)
0.3 (1.1)
200 (50)
240 (7)
12.3 (6)
Note: Values in brackets are estimated errors: from fits to eq. [6] for A, Ea, and (kf + kisc); from 10% for φf; from 1% for τs; from accumulated er-
rors for calculated quantities.
aSolvent: M (methanol), MC (methylcyclohexane).
bArrhenius parameters from eq. [6].
cIntercept from eq. [6].
dkf = φf kt.
e(kf + kisc) – kf .
fkt = 1/τs from nanosecond single photon counting at 25 °C.
gkr(25 °C) = kt(25 °C) – (kf + kisc).
hkr(1) – kr(2) at 25 °C.
Scheme 3. Mechanism for the phototransposition of para-methylbenzonitrile to the meta and ortho isomers and associated rate con-
stants of reaction.
T1
No reaction
NC
CN
CN
CN
kisc
hν
kr
S1
CH3
CH3
kf + kic
CH3
18
20
19
CH3
kir
CN
CN
CH3
CH3
porting information)3 were obtained on a Bruker AVANCE
500 MHz spectrometer with chemical shifts (δ) reported rel-
ative to the central line of CDCl3 at 77.16 ppm. GC–MS
analyses were performed on a PerkinElmer Autosystem XL
instrument with a mass selective detector. The column used
was a Supelco 30 m × 0.25 mm MDN-5S 5% phenyl
methylsiloxane, film thickness 0.50 µm; temperature pro-
gram: 60 °C for 1 min; 20 °C/min to 240 °C; 240 °C for
10 min. GC-FID analyses were done in a similar way except
on a Supelco DB200 column. Mass spectral data are re-
ported in units of mass over charge (m/z) for all values be-
tween 50 and the molecular ion if greater than 10% of the
base peak. Intensities are reported as a percent of the base
peak. HR-MS were obtained on a CEC 21-110 instrument
with a resolution of 5–10 × 103. The solvents used for pho-
tochemical reactions and fluorescence measurements were
used as received: methanol (EM, HPLC), cyclohexane
(Aldrich, HPLC), and methylcyclohexane (Aldrich, spectro-
photometric grade).
Preparation of the adamantyl aryl ethers 1a–1k
The ethers were synthesized by a modified literature
method (9) from 1-bromoadamantane and the substituted
phenol (Aldrich). In a typical procedure, a mixture of 8.40 g
(75 mmol) of 4-fluorophenol and 3.24 g (15 mmol) of 1-
bromoadamantane and 4.5 mL of pyridine was refluxed for
5 h. The reaction mixture was cooled, diluted with 50 mL of
water – ethylene glycol (8:1), and extracted with 2 × 15 mL
© 2005 NRC Canada

Pincock and Pincock
1251
of hexanes. The extract was washed with 5% NaOH
(25 mL), water (2 × 25 mL), dried (MgSO4), and the hex-
anes evaporated. The crude solid (2.26 g) was purified by
flash chromatography using 2.5% ethyl acetate in hexanes to
give 1.14 g (31%) of a colourless solid that was recrystall-
ized from methanol.
sity measurements were done using a PTI-210M fluores-
cence spectrometer with dual model 101 monochromators, a
75 W Xenon lamp, and a model 814 photomultiplier detec-
tor. Corrected spectra were obtained. Selected examples are
shown in Fig. 6. Fluorescence quantum yields were deter-
mined using solutions of matched absorbance by comparison
with the known fluorescence quantum yield of 0.24 (30) for
anisole in methanol and 0.29 (34) in cyclohexane. The ex-
pectation values necessary for calculating the kf values from
eq. [3] were obtained by summing the 1 nm incremental ar-
eas over the complete emission band. For temperature-
dependent quantum yield measurements, a 1 cm cuvette was
thermostatted with stirring in a cuvette holder equipped with
a Quantum Northwest TLC 50 temperature controller, which
allows relative fluorescence measurements at 10 °C intervals
from –25 to 95 °C in methylcyclohexane and –25 to 60 °C
in methanol over a 2 h time span. The stability of the lamp
over this time period was normally confirmed by rerunning
the fluorescence spectrum at 25 °C at the end.
Characterization of the adamantyl aryl ethers 1a–1k
1H NMR spectra were not particularly diagnostic: the sig-
nals expected for the substituted benzene ring, the
substituent for X = OCH3 and CH3, and the 1-adamantyl
ring system (δ 2.17, 1.89, and 1.61, all broad singlets) were
all present. 13C NMR spectral data for the ethers 1, along
with those for 2–4 are reported in Table S1 (Supporting in-
formation).3 Mass spectral data (GC–MS) were also not par-
ticularly diagnostic: a molecular ion (but not for X = 3-CF3,
4-CF3, 3-CN, and 4-CN) and C10H15, the adamantyl cation
as the base peak (100) and its fragment ions at 107 (12), 93
(22), 79 (25), and 77 (13) were observed. Other character-
ization results: 1a (X = H): mp 112 to 113 °C (lit. value (33)
mp 99 to 100 °C). 1b (X = 4-OCH3): mp 48 to 49 °C (lit.
value (33) mp 49–51 °C). 1c (X = 3-OCH3): mp 44–46 °C.
HR-MS calcd. for C17H22O2: 258.1620; found: 258.1628. 1d
(X = 4-CH3): mp 53.5–54.0 °C (lit. value (33) mp 41–
43 °C). 1e (X = 3-CH3): mp 69.0–70.0 °C. HR-MS calcd.
for C17H22O: 246.1671; found: 246.1668. 1f (X = 4-F): mp
110 to 111 °C. HR-MS calcd. for C16H19OF: 246.1420;
found: 246.1423. 1g (X = 3-F): mp 38 to 39 °C. HR-MS
calcd. for C16H19OF: 246.1420; found: 246.1419. 1h (X = 4-
CF3): mp 57.5–58.5 °C. 1j (X = 4-CN): characterized previ-
ously (2). 1k (X = 3-CN): mp 86 to 87 °C.
Singlet lifetimes were measured at 25 °C by monitoring
fluorescence decay using a PRA time-correlated, single-
photon-counting apparatus with a hydrogen flash lamp of
with a pulse width of about 1.8 ns.
Acknowledgments
We thank the Natural Sciences and Engineering Research
Council of Canada (NSERC) for financial support, Sepracor
Canada Ltd., Windsor, Nova Scotia for the donation of
chemicals, and Lei Zhang, Department of Chemistry and
Biochemistry, Concordia University, Montreal, for assistance
with the Gaussian calculations.
Irradiation of ethers
A solution of ~100 mg of the appropriate ether in 100 mL
of methanol or cyclohexane was purged with nitrogen and
then irradiated in a Rayonet photochemical reactor using 16
lamps (75 W, 254 nm). The temperature was kept at 25 °C
by circulating water in an immersion tube. Reaction progress
was monitored by GC.
References
1. S.A. Fleming and J.A. Pincock. Organic molecular photo-
chemistry. Vol. 3. Marcel Dekker, New York. 1999. p. 211.
2. D.P. DeCosta, A. Bennett, A.L. Pincock, J.A. Pincock, and R.
Stefanova. J. Org. Chem. 65, 4162 (2000).
Absorbance measurements
3. A.L. Pincock, J.A. Pincock, and R. Stefanova. J. Am. Chem.
Soc. 124, 9768 (2002).
Absorbance spectra were recorded at 1 nm resolution us-
ing a Cary 100 UV–vis spectrometer thermostatted at 25 °C.
Selected examples are shown in Fig. 1. The integrated spec-
tra needed to obtain kf values from eq. [3] were obtained by
summing the 1 nm incremental areas (the absorbance values
converted to ⑀ by dividing by the molar concentration) over
the complete absorption band. These areas are somewhat un-
certain at the short wavelength end because of overlap with
the long wavelength tail of the S2 band so that the
absorbance does not reach zero. The wavelength for the min-
imum in absorbance was used as the cut-off point for the S1
band. For the 4-cyano and 4-trifluoromethyl derivatives of 1
and 2, this overlap was too extensive to allow integration of
the S1 band and kf values were not calculated.
4. J. McEwen and K. Yates. J. Phys. Org. Chem. 4, 193 (1991).
5. E.L. Wehry and L.B. Rogers. J. Am. Chem. Soc. 87, 4234
(1965).
6. P.J. Baldry. J. Chem. Soc. 951 (1979).
7. (a) S.C. Shim, J.W. Park, and H.-S. Ham. Bull. Korean Chem.
Soc. 3, 13 (1982); (b) S.C. Shim, J.W. Park, H.-S. Ham, and
J.-S. Chung. Bull. Korean Chem. Soc. 4, 45 (1983).
8. T.F. Wolwode, C. Rose, and T.J. Wandless. J. Org. Chem. 63,
9594 (1998).
9. H. Masada and Y. Oishi. Chem. Lett. 57 (1978).
10. J. Bartl, H. Mayr, and R.A. McClelland. J. Am. Chem. Soc.
112, 6918 (1990).
11. H.E. Zimmerman and S. Somasekhara. J. Am. Chem. Soc. 85,
922 (1963).
12. Y.-R. Luo. Handbook of bond dissociation energies in organic
compounds. CRC Press, Boca Raton, Fla. 2003.
13. D.C. Spellmeyer, P.D.J. Grootenhuis, M.D. Miller, L.F.
Kuyper, and P.A. Kollman. J. Phys. Chem. 94, 4491 (1990).
14. (a) T. Schaefer and G.H. Penner. Can. J. Chem. 66, 1635
(1988); (b) T. Schaefer and G.H. Penner. Can. J. Chem. 66,
Fluorescence measurements
All samples were degassed by three freeze–pump–thaw
cycles and then thermostatted at 25 °C. The substituted
anisoles used for comparison were all commercial samples
(Aldrich) and were distilled before use. Fluorescence inten-
© 2005 NRC Canada

1252
1641 (1988); (c) T. Schaefer and R. Sebastian. Can. J. Chem.
Can. J. Chem. Vol. 83, 2005
22. S.J. Strickler and R.A. Berg. J. Chem. Phys. 37, 814 (1962).
23. J.B. Birks and D.J. Dyson. Proc. R. Soc. London Ser. A, 275,
135 (1963).
67, 1148 (1989).
15. E. Taskinen. J. Chem. Soc. Perkin Trans. 2, 1824 (2001).
16. J.C. Dearden and W.F. Forbes. Can. J. Chem. 37, 1305 (1959).
17. C.J. Pouchert and J. Behnke. The Aldrich library of 13C and
1H NMR spectra. Vol. 2. Aldrich Chemical Company, Milwau-
kee, Wis. 1993. pp. 175–241.
24. R.B. Cundall and L.C. Pereira. J. Chem. Soc. Faraday Trans.
68, 1152 (1972).
25. F.D. Lewis and X. Zuo. Spectrum (Bowling Green, OH, U.S.),
16, 8 (2003). Available from www.bgsu.edu/departments/
photochem/research/summer2003/spectrum.html.
26. S.I. Druzhinin, A. Demeter, V.A. Galievsky, T. Yoshihara, and
K.A. Zachariasse. J. Phys. Chem. 107, 8075 (2003), and
refs. cited therein.
27. C.M. Gonzalez and J.A. Pincock. J. Am. Chem. Soc. 126,
8870 (2004).
28. J. Timmermans. Physico-chemical constants of pure organic
liquids. Vol. 1. Elsevier, New York. 1950.
29. S. Dellonte and G. Marconi. J. Photochem. 30, 37 (1985).
30. G. Kohler, G. Kittel, and N. Getoff. J. Photochem. 18, 19
(1982).
31. F.A. Carroll and F.H. Quina. J. Am. Chem. Soc. 94, 6246
(1972).
32. (a) J.A. Pincock. In CRC handbook of organic photochemistry
and photobiology. 2nd ed. Edited by W. Horspool and F. Lenci.
CRC Press, Boca Raton, Fla. 2003. pp. 46–1 to 46–19; (b) P.J.
MacLeod, A.L. Pincock, J.A. Pincock, and K.A. Thompson. J.
Am. Chem. Soc. 120, 13 354 (1998).
33. U. Kraatz. Chem. Ber. 106, 3095 (1973).
34. I.B. Berlman. Handbook of fluorescence spectra of aromatic
molecules. Academic Press, New York. 1971. p. 139.
18. J. Bromilow, R.T.C. Brownlee, D.J. Craik, M. Sadek and R.W.
Taft. J. Org. Chem. 45, 2429 (1980).
19. K.S. Dhami and J.B. Stothers. Can. J. Chem. 44, 2855 (1966).
20. M.J. Frisch, G.W. Trucks, H.B. Schlegel, G.E. Scuseria, M.A.
Robb, J.R. Cheeseman, V.G. Zakrzewski, J.A. Montgomery,
Jr., R.E. Stratmann, J.C. Burant, S. Dapprich, J.M. Millam,
A.D. Daniels, K.N. Kudin, M.C. Strain, O. Farkas, J. Tomasi,
V. Barone, M. Cossi, R. Cammi, B. Mennucci, C. Pomelli, C.
Adamo, S. Clifford, J. Ochterski, G.A. Petersson, P.Y. Ayala,
Q. Cui, K. Morokuma, N. Rega, P. Salvador, J.J. Dannenberg,
D.K. Malick, A.D. Rabuck, K. Raghavachari, J.B. Foresman,
J. Cioslowski, J.V. Ortiz, A.G. Baboul, B.B. Stefanov, G. Liu,
A. Liashenko, P. Piskorz, I. Komaromi, R. Gomperts, R.L.
Martin, D.J. Fox, T. Keith, M.A. Al-Laham, C.Y. Peng, A.
Nanayakkara, M. Challacombe, P.M.W. Gill, B. Johnson, W.
Chen, M.W. Wong, J.L. Andres, C. Gonzalez, M. Head-
Gordon, E.S. Replogle, and J.A. Pople. GAUSSIAN 98. Revi-
sion A.11.2 [computer program]. Gaussian Inc., Pittsburgh,
Penn. 2001.
21. B. Uno, T. Iwamoto, and N. Okumura. J. Org. Chem. 63, 9794
(1998).
© 2005 NRC Canada