Conformation-Dependent Intramolecular Electron Transfer in N-(Aminoalkyl)-9-phenanthrenecarboxamides

Article abstract of DOI:10.1021/jp9526758

The molecular structure and photophysical behavior of several secondary and tertiary N-(aminoalkyl)phenanthrenecarboxamides have been investigated.Secondary (aminoalkyl)amides exist predominantly in the Z conformation, whereas tertiary amides exist as mixtures of Z and E conformers and semirigid piperazines as mixtures of chair conformers.Rate constants for endergonic intramolecular electron transfer are found to be highly dependent upon molecular structure.The aromatic and amide groups of the tertiary amides are essentially orthogonal, and thus, an E aminoalkyl group can adopt low-energy conformations in which there is spatial overlap between the aromatic and amine groups, whereas such overlap is not possible for either a Z aminoalkyl group or the piperazines.The observation of more rapid intramolecular electron transfer quenching of the phenanthrene singlet by an appended trialkylamine in the E vs Z conformation is attributed to this difference in overlap.An increase in the phenanthrene-amide dihedral angle is also found to result in a decrease in the rate constant for intramolecular electron transfer quenching by a Z aminoalkyl group.In the case of appended tertiary anilines, efficient electron transfer quenching occurs for both Z and E conformers.The Z conformers form fluorescent exciplexes, providing a new example of exciplex-type emission in the absence of direct ?-? overlap.Exciplexes formed by the E conformers are nonfluorescent and apparently undergo rapid intersystem crossing.The strong exciplex emission observed at low temperatures both in solution and in frozen glasses is attributed to ground state dimers or aggregates.

Full text of DOI:10.1021/jp9526758

J. Phys. Chem. 1996, 100, 4055-4063
4055
Conformation-Dependent Intramolecular Electron Transfer in
N-(Aminoalkyl)-9-phenanthrenecarboxamides
Frederick D. Lewis* and Eric L. Burch
Department of Chemistry, Northwestern UniVersity, 2145 Sheridan Road, EVanston, Illinois 60208-3113
ReceiVed: September 13, 1995; In Final Form: NoVember 12, 1995X
The molecular structure and photophysical behavior of several secondary and tertiary N-(aminoalkyl)-
phenanthrenecarboxamides have been investigated. Secondary (aminoalkyl)amides exist predominantly in
the Z conformation, whereas tertiary amides exist as mixtures of Z and E conformers and semirigid piperazines
as mixtures of chair conformers. Rate constants for endergonic intramolecular electron transfer are found to
be highly dependent upon molecular structure. The aromatic and amide groups of the tertiary amides are
essentially orthogonal, and thus, an E aminoalkyl group can adopt low-energy conformations in which there
is spatial overlap between the aromatic and amine groups, whereas such overlap is not possible for either a
Z aminoalkyl group or the piperazines. The observation of more rapid intramolecular electron transfer
quenching of the phenanthrene singlet by an appended trialkylamine in the E vs Z conformation is attributed
to this difference in overlap. An increase in the phenanthrene-amide dihedral angle is also found to result
in a decrease in the rate constant for intramolecular electron transfer quenching by a Z aminoalkyl group. In
the case of appended tertiary anilines, efficient electron transfer quenching occurs for both Z and E conformers.
The Z conformers form fluorescent exciplexes, providing a new example of exciplex-type emission in the
absence of direct π-π overlap. Exciplexes formed by the E conformers are nonfluorescent and apparently
undergo rapid intersystem crossing. The strong exciplex emission observed at low temperatures both in
solution and in frozen glasses is attributed to ground state dimers or aggregates.
Introduction
CHART 1
Formation of an amide linkage provides a simple method of
connecting small molecular building blocks to form larger
molecules. In addition to its function as a linker, the amide
group can also serve as a conformational control element.1
Amides with two different substituents on nitrogen exist as
mixtures of Z and E conformational isomers. Most secondary
amides (R2 ) H) exist predominantly in the Z conformation
(eq 1); whereas tertiary amides exist as mixtures of Z and E
isomers, with rotational barriers generally in excess of 15 kcal/
mol. In the case of aromatic carboxamides, the arene-carbonyl
dihedral angle, æ1, is dependent upon the steric demands of the
arene. Amide-arene resonance favors a small value of æ1 for
benzamides (<30°), whereas nonbonded repulsion can result
large values of æ1 (>60°), as in the case of 9-phenanthrenecar-
boxamides.1,2 The amide group remains planar (æ2 ∼ 0°) even
in the tertiary amides of sterically demanding aromatic car-
boxamides.
amines has also been investigated. The aminoalkyl groups in
the secondary amides 2 and 3 exist predominantly in the Z
configuration, whereas the tertiary amides 5-10 exist as
mixtures of E and Z isomers that interconvert slowly on the
excited state time scale. This permits comparison of intermo-
lecular quenching by Z vs E isomers as functions of aminoalkyl
chain length (6 vs 7) and amine oxidation potential (6 vs 8 and
6 vs 9).5 The effect of the dihedral angle æ1 upon intramolecular
electron transfer quenching can be investigated by comparison
of the results for the lactam 14, secondary amide 2, and tertiary
amides 6. Finally, the effects of restricted conformational
mobility have been probed for the semirigid piperazines 11 and
12.
(1)
The amide group has been employed in investigations of
photoinduced electron transfer as a linker in the construction
of diad and triad molecules.3,4 However, the effect of the
amide’s distinct conformational preferences upon electron
transfer processes has not been investigated. We report here
the results of our investigations of intramolecular quenching of
the fluorescent states of the singlet N-(aminoalkyl)-9-phenan-
threnecarboxamides. For purposes of comparison, intermo-
lecular quenching of the model compounds 1, 4, and 13 by
Results and Discussion
Synthesis and Structure of Linked Amides and Esters.
The synthesis of the model amides 1 and 4 and lactam 13 have
X Abstract published in AdVance ACS Abstracts, February 15, 1996.
0022-3654/96/20100-4055$12.00/0 © 1996 American Chemical Society

4056 J. Phys. Chem., Vol. 100, No. 10, 1996
Lewis and Burch
previously been reported.2,6 The linked amine-amides (2, 6,
7, and 9-11) were prepared via the reaction of phenanthrene-
9-carboxylic acid chloride with excess diamine. The aniline-
amides (3, 8, and 12) were prepared by dicyclohexylcarbodi-
imide (DCC) coupling of phenanthrene-9-carboxylic acid with
the appropriate diamine. The amine-lactam 14 was prepared
via a modification of the synthesis of 13. All products were
purified by column chromatography followed by recrystalliza-
tion and characterized by 1H NMR, IR, absorption, and
fluorescence spectral data. The crystal structure of 8 and
solution conformational analysis of 5-8 will be reported
elsewhere.7
The secondary amides 2 and 3, piperazines 11 and 12, and
1
lactam 14 exhibit a single set of H NMR resonances at room
temperature for each type of proton. Thus, they exist either in
a single conformation or in rapidly interconverting conforma-
tions. The chemical shifts of the aminoalkyl protons for 2 and
14 are similar, indicating that the secondary amide 2 exists in
the Z conformation, as is normally the case for the secondary
amides.1 The piperazines 11 and 12 display separate signals
for the amide R-methylene protons but a single amine N-methyl
or N-phenyl signal.
Figure 1. Folded conformations of (a) Z-6 and E-6 and (b) the two
chair conformations of 12.
TABLE 1: Absorption Maxima, Extinction Coefficients, and
Fluorescence Maxima for Phenanthrene-9-carboxamides in
Hexane Solution
amide S2, nm ꢀA (M-1cm-1
)
S1, nm ꢀA (M-1cm-1
)
fl, nm
1
2
3
4
6
7
8
9
10
11
12
13
14
298
299
299
298
300
300
300
300
300
298
300
316
12900
11900
9370
10600
12700
9230
10600
11500
12700
11400
13900
9700
350
350
350
348
348
350
349
350
350
348
350
353
180
160
130
120
130
110
130
110
210
120
150
700
372
372
372
368
368
367
366
368
368
373
375
374
374
Tertiary amides with inequivalent N-substituents usually exist
as mixtures of Z and E conformers (Chart 1).1 The spectra of
5-10 display two sets of resonances that can be assigned to
the E and Z conformers on the basis of chemical shift data and
solvent-induced shifts. R-Substituents Z to the carbonyl oxygen
have larger chemical shifts than R-substituents E to the carbonyl,
and E substituents display larger upfield shifts in benzene than
Z substituents.8 In the case of 9 and 10 these assignments were
confirmed by spin-decoupling studies. The ratio of E/Z isomers
for the tertiary amides can be determined by integration of the
two N-methyl resonances (N-isopropyl in the case of 10). The
Z isomer content determined in this manner is 51-59% for 5-9
and 85% for 10. The larger Z isomer content for 10 is consistent
with the reported preference of bulky substituents for the E
position in benzamides.1,8 The Z/E ratios of 6, 8, and 10 are
independent of temperature (-40 to 20 °C) in CDCl3 solution
and are also independent of solvent polarity.
1
energy absorption bands similar in appearance to the La and
1Lb bands of phenanthrene and a more intense band at a shorter
wavelength (ca. 225 nm). Absorption maxima and extinction
coefficients in hexane solution are summarized in Table 1. The
shorter wavelength maxima for lactams 13 and 14 are red-shifted
with respect to those of the secondary and tertiary amides,
presumably reflecting enhanced conjugation in the planar lactam
vs nonplanar amides. Aminoalkyl substituents are observed to
have only minimal effects on the position of the absorption
maxima, indicating the absence of strong interaction between
the phenanthrene and amine functional groups. Since the
absorption maximum for N,N-dimethylaniline is at 298 nm (ꢀA
) 2200), an increase in the intensity of the short wavelength
bands of 3, 8, and 12 was expected but not observed. We found
no evidence for charge-transfer absorption bands such as those
observed by Barlow et al.10 for N-(aminoalkyl)phthalimides.
Room temperature fluorescence is observed for all of the
phenanthrene derivatives investigated. Fluorescence maxima
in hexane solution are reported in Table 1 and, like the long
wavelength absorption maxima, display only minor variation
with structure or solvent polarity. The fluorescence spectra of
the tertiary amides and lactams (4-13) are highly structured,
like the fluorescence of phenanthrene, whereas the fluorescence
of the secondary amides 1-3 is broader. This difference is
attributed to a change in geometry for the secondary amides (a
decrease in æ1, eq 1, in the singlet vs ground state) but not for
the tertiary amides or lactams.2 In addition to the locally excited
fluorescence observed for all of the phenanthrene derivatives,
structureless, long-wavelength emission attributed to intramo-
lecular exciplexes (Vide infra) is observed for the linked anilines
3 and 8.
The ground state conformations of the (aminoalkyl)amides
2, 6, and 11 were investigated using a modified MM2 force
field as supplied in the Chem 3D Plus software package.9 The
phenanthrene-carbonyl dihedral angles (æ1) were assumed to
be the same as those calculated for the model compounds 1
and 4. The potential energy surfaces for 1 and 4 are fairly flat
for 50° < æ1 < 120° with minima at 115° for 1 and 90° for 4
(relative to the planar s-cis enone conformation (æ1 ) 0°)). In
the global minima for 2 and Z-6, the â-carbon of the aminoalkyl
group is perpendicular to the amide plane and the amine and
amide nitrogens are anti. The rotamer in which the nitrogens
are gauche is calculated to be of slightly higher energy (1-2
kcal/mol). Minimized, folded conformations of Z-6 and E-6
are shown in Figure 1a. Overlap of the amine lone pair and
phenanthrene π-orbitals is possible for the latter, but not the
former, isomer. There are two global minima for 11 and 12,
corresponding to the two diastereomeric flattened piperazine
chair conformers (Figure 1b). The twist boat is calculated to
be higher in energy by ca. 5 kcal/mol. Thus, the two chair
conformers are expected to interconvert rapidly at room tem-
perature on the NMR time scale, in accord with the observation
of a single set of proton resonances.
Absorption and Fluorescence Spectra. The electronic
absorption and fluorescence spectra of the model compounds
1, 4, and 13 have been previously described.2,6 All of the
phenanthrene derivatives in the present study display two low-
Fluorescence quantum yields (Φf), decay times (τ), and
preexponentials (A) in hexane and acetonitrile solution are

N-(Aminoalkyl)-9-phenanthrenecarboxamides
J. Phys. Chem., Vol. 100, No. 10, 1996 4057
TABLE 2: Fluorescence Quantum Yields and Lifetimes for
Phenanthrene-9-carboxamides in Hexane and Acetonitrile
Solution
TABLE 3: Energetics and Kinetics of Intermolecular
Electron Transfer Quenching
amide
hexane
CH3CN
τ, ns
27.4
5.3
0.2
solvent
1
4
13
amide
ΦF
0.09
0.04
<0.005
0.03
τ, nsa
23.7
14.1
0.2
13.9
15.8
15.0 (0.69)
3.1 (0.31)
14.7 (0.71)
8.7 (0.21)
0.4, 0.40b
15.6 (0.65)
6.9 (0.35)
15.6 (0.73)
5.0 (0.25)
5.3 (0.30)
3.8 (0.70)
0.7
Es, eVa
hexane
MeCN
hexane
hexane
MeCN
MeCN
hexane
3.42
3.52
3.56
EAred, Vb
-1.90 -2.04 -1.87
0.43
1.8
-0.17 -0.12 -0.28
5.4
0.18
2.3
1
2
3
4
∆Get, eV (Et3N)c
10-9ksv, M-1 s-1 (Et3N)d
∆Get, eV (Et3N)c
10-9ksv, M-1 s-1 (Et3N)d
∆Get, eV (PhNMe2)c
0.48
2.8
0.32
3.0
22.4
22.5
2.6
0.22
1.7
4.7
0.07
1.2
5
0.04
0.03
6-l
6-s
7-l
7-s
8
9-l
9-s
10-l
10-s
11-l
11-s
12
13
14
12.8 (0.27)
0.2 (0.79
18.5 (0.49)
1.2 (0.51)
0.2, 0.22b
19.0 (0.055)
1.4 (0.45)
19.7 (0.75)
10-9ksv, M-1 s-1 (PhNMe2)d hexane
0.03
a Singlet energy estimated from the highest energy fluorescence
maximum. b Reduction potential vs Ag/AgI. c Calculated using eq 3.
See text. d Rate constant for intermolecular quenching determined from
the slope of linear Stern-Volmer plots (eq 2) and the measured lifetime
(Table 3).
<0.005
0.03
0.04
6.9 (0.27)
0.01
the appearance of intermolecular exciplex fluorescence in
nonpolar and moderately polar solvents. The slopes of linear
Stern-Volmer plots (eq 2) for quenching by amine provide
values of the quenching constant kqτ, where kq is the rate
constant for intermolecular quenching and τ is the singlet
lifetime in the absence of quenching.11b,c Values of kq calculated
from the quenching constant and the measured singlet lifetime
(Table 2) are reported in Table 3 along with the singlet energies
and reduction potentials of the amides. The oxidation potentials
of Et2NH, Et3N, and PhNMe2 are 1.65, 1.42, and 1.16 V vs
Ag/AgI.
8.8 (0.85)
5.0 (0.15)
0.3
20.2
0.2
<0.005
0.09
0.01
19.4
3.4
a Lifetime data determined with a time resolution of ca. 0.2 ns except
as noted (preexponentials in parentheses). b Data obtained with a time
resolution of 50 ps.
reported in Table 2. Decay profiles were obtained using a single
photon counting apparatus, and the decays were deconvoluted
using a single or multiple exponential least-squares analysis as
described by James et al.11a Samples were excited near the
maximum of the 255 nm absorption band and monitored near
the emission maximum (Table 1). The time resolution of our
apparatus is ca. 0.2 ns, and thus, decay times <0.5 ns should
be regarded as approximate. The decay times for 8 are similar
to those obtained using an apparatus with 50 ps time resolution.
The decay times >1 ns reported as single exponentials in Table
2 are best fit by a single exponential decay (A ) 1.0), and those
reported as double exponentials are best fit as double exponen-
tials. For decay times <1 ns, improved fits could be obtained
with double or triple exponentials, but in all cases (including 3
and 8 for which exciplex fluorescence is observed) the reported
decays account for >99% of the total locally excited fluores-
cence intensity and the goodness of fit, as judged by the
randomness of residuals, autocorrelation function, and reduced
ø2 values < 1.2, was considered acceptable.
I°/I ) 1 + kqτ[amine]
(2)
The observation of exciplex fluorescence upon quenching of
1, 4, and 13 by PhNMe2 and the larger values of kq for
quenching by PhNMe2 vs Et3N are indicative of an electron
transfer mechanism for fluorescence quenching. The free energy
change for electron transfer quenching can be calculated from
the singlet energies and redox potentials for the amides reported
in Table 3 and the amine oxidation potentials using Weller’s
equation (eq 3), where C is an empirical solvent constant (0.54
eV in hexane and -0.06 eV) in acetonitrile).12 The values of
kq for quenching by PhNMe2 in hexane solution are near the
rate of diffusion (ca. 2 × 1010 M-1 s-1). The values of kq for
quenching by Et2NH and Et3N are smaller and display surpris-
ingly little dependence on amide singlet state electron affinity
red
(EA - ES), amine oxidation potential, or solvent polarity.
The secondary amides 1-3, model tertiary amides 4 and 5,
and lactams 13 and 14 all display single exponential fluorescence
decay. The tertiary amides display dual exponential fluores-
cence decay, except in the case of the aniline-amides 8 and
12, which have decay times similar to the time resolution of
our instrument. The decay times of 4, 6, and 11 were measured
in several additional solvents (benzene, dichloromethane, and
benzonitrile) and have values intermediate between those
reported for hexane and acetonitrile.
Similar results were reported for quenching of E-stilbene and
several arenes by secondary and tertiary aliphatic amines and
attributed to the formation of an exciplex or heteroexcimer with
only partial electron transfer character.13 In such cases, exciplex
formation is highly reversible and the rate constant kq is
determined by the rates of exciplex formation, dissociation, and
decay.11b,c
∆Get ) EDox - EAred - ES + C
(3)
The fluorescence lifetimes and quantum yields for the linked
compounds are all shorter than those of the model compounds
(Table 2). However, the fluorescence rate constants calculated
The linked amine-amides 2 and 3 and amine-lactam 14
display single exponential decay with fluorescence quantum
yields and lifetimes smaller than those of the model compounds
1 and 13 (Table 2). Since the values of kf for the linked amines
are similar to those for the model compounds, we assume that
the decrease in lifetime is a consequence of intramolecular
electron transfer quenching. Rate constants for intramolecular
quenching (kq) calculated from the lifetimes of the linked amine
(τ) and model compound (τ0) using eq 4 are reported in Table
4. Intramolecular photoinduced electron transfer between arenes
and amines separated by a flexible polymethylene chain has
from the measured lifetimes and quantum yields (kf ) Φfτ-1
)
are similar for the linked compounds and the appropriate model
compound. Thus, the shorter lifetimes of the linked compounds
are attributed to the occurrence of intramolecular quenching
rather than perturbation of the phenanthrene chromophore.
Inter- and Intramolecular Quenching. The fluorescence
intensity (Ι) of the model compounds 1, 4, and 13 is quenched
by added amines. Quenching of the fluorescence of 1, 4, and
13 by PhNMe2 (but not by Et3N or Et2NH) is accompanied by

4058 J. Phys. Chem., Vol. 100, No. 10, 1996
Lewis and Burch
TABLE 4: Calculated Rate Constants for Intramolecular
Quenchinga
The values of kq assigned to the Z isomers of the (N,N-
dimethylamino)ethylamides increase in the order Z-6 < 2 <
14 (Table 4). The structure dependence of intramolecular
quenching stands in marked contrast to intermolecular quench-
ing, which is essentially independent of the amide structure or
amine oxidation potential (Table 3). One possible reason for
this difference is that overlap of the singlet phenanthrene and
amine orbitals is possible for intermolecular quenching but not
for intramolecular quenching of the Z isomers. The amide group
serves as an intermediary in the intramolecular quenching of
singlet phenanthrene by the aminoalkyl group. Interaction of
the phenanthrene and amide π-orbitals should be most effective
when these groups are coplanar, as in the lactams 13 and 14.17
The planar lactam 13 also has a lower reduction potential than
the tertiary amide 4 (Table 3). Thus, the increase in rate constant
for Z-6 < 2 < 14 may reflect a decrease in the dihedral angle
æ1 (eq 1) in this series of molecules.
amide
10-6kq, M-1 (hexane)
10-6kq, M-1 (MeCN)
2
3
29
4900
3.4
260
4.7
52
2400
0.8
82
0.8
82
150
>5000
34
>5000
9.6
790
4500
8.2
670
6.3
160
69160
3300
>5000
Z-6
E-6
Z-7
E-7
8
Z-9
E-9
Z-10
E-10
11
130200
1400
260
12
14
a Values calculated using eq 4 and lifetime data from Table 3.
been extensively investigated.14 In nonpolar solvents, arene-
trialkylamine exciplexes are proposed to adopt specific folded
conformations that maximize orbital overlap and Coulombic
attraction while avoiding high-energy conformations of the
polymethylene chain. The relatively slow values of kq for
intramolecular quenching by N,N-dimethylamines 2 and 14 may
reflect conformational restrictions on overlap between the arene
and amine as well as the endergonic nature of electron transfer.
Changing the aminoalkyl linker from ethyl to propyl in 6 vs
7 has little effect on the kq values of the Z isomer but results in
a decrease in kq for the E isomer (Table 4). The extended linker
in Z-7 still does not permit direct phenanthrene-amine orbital
overlap. The extended linker in E-7 is expected to increase
the entropic requirement for intramolecular exciplex formation
without improving the orbital overlap, resulting in a decrease
in kq. Changing the quencher from a tertiary amine to a
secondary amine in 6 vs 9 or 10 results in a decrease in kq for
both Z and E isomers, providing another instance of greater
structural sensitivity for intra- vs intermolecular quenching.
Secondary amines are expected to be much poorer quenchers
than tertiary amines on the basis of their higher oxidation
potentials. However, N-H-arene hydrogen bonding may help
stabilize secondary amine exciplexes.13b Such hydrogen bond-
ing is not possible for the Z isomers of the (aminoalkyl)amides.
kq ) τ-1 - τ°-1
(4)
Dual exponential fluorescence decay is observed for all the
flexibly linked amine-tertiary amides 6-10 except 8 (Table
3). The lifetimes of both components are shorter than those of
the model tertiary amides 4 and 5. The two decays are assigned
to the E and Z conformers based on the reasonable assumption
that interconversion of these conformers is slow on the time
scale of fluorescence decay.15 The observation of dual expo-
nential decay requires the presence of a covalently attached
amine quencher, as evidenced by the observation of single
exponential decay from the model amide 5, which exists as a
1:1 mixture of E and Z conformers. The rate constants for
intramolecular quenching by the two conformers can be
calculated using eq 4 and the measured lifetimes. The assign-
ment of the larger values of kq to the E isomers is based upon
the anticipated differences in phenanthrene-amine orbital
overlap in the E vs Z isomer (Figure 1a). This assignment is
supported by the effects of replacing the amide N-methyl of 9
with the N-isopropyl of 10, which results in an increase in the
ground state population of the Z conformer and an increase in
the preexponential for the longer-lived fluorescence decay (Table
2).
The N-methylpiperazine 11 provides an interesting example
of dual exponential decay in a molecule with a semirigid
linker.18 We tentatively assign the two fluorescence decays
(Table 2) to the two diastereomeric chair conformations, which
have similar MM2-minimized energies (Figure 1b). The barrier
for ring inversion of piperazines is too small to permit
observation of the two diasteromers by NMR but should be large
enough to prevent singlet state equilibration. Both of the values
of kq for 11 are more similar to those for the E vs Z isomers of
6 and 7, even though phenanthrene-amine orbital overlap is
not possible for 11. The larger values of kq for 11 vs Z-6 might
reflect either a lower amine oxidation potential19 or involvement
of through-bond interactions in the quenching process. Brouwer
et al.20a have observed extensive through-bond interaction in
the cation radical of N,N′-dimethylpiperazine.
Single exponential decay is observed for the aniline-amides
8 and 12 (Table 2), even though the former exists as a mixture
of Z and E isomers and the latter as a mixture of chair
conformers. The observed fluorescence decay constants are near
the time resolution of our instrumentation. Hence, a more rapid
decay would not be resolved from the lamp profile and two
similar decay constants most likely would not be resolved.
Assuming the observed decay for 8 is that of the Z isomer, the
calculated quenching rate constant for Z-8 is 700 times that of
Z-6 (Table 4). This large rate enhancement can be attributed
to the lower ionization potential and smaller nuclear reorganiza-
tion energy for oxidation of the aromatic vs aliphatic amines.
π-Donors such as aniline are also reported to be more effective
long-range donors than are n-donors such as trialkylamines.20b
A related example of isomer-dependent intramolecular elec-
tron transfer quenching has recently been reported by de Silva
et al.16 The observation of efficient intramolecular electron
transfer quenching for the naphthalimide 15, but not for 16,
was attributed to the intramolecular charge-transfer character
of the naphthalimide singlet state. The flexibly linked amine
in 15 can directly overlap the electron-deficient region of the
singlet state, whereas the linked amine in 16 is connected to
the electron-rich region of the singlet state.
Solvent Dependence of Electron Transfer Quenching.
Rate constants for intermolecular quenching of 1, 4, and 13 by

N-(Aminoalkyl)-9-phenanthrenecarboxamides
J. Phys. Chem., Vol. 100, No. 10, 1996 4059
TABLE 5: Exciplex Fluorescence Maxima and Decay Times
exciplex fluorescence intensities of 1:aniline, 3, and 4:aniline
are strong in hexane and very weak in acetonitrile.
exciplex
1:PhNMe2 hexane
diethyl ether
solvent
λ
max, nm
τex, nsa
The observation of fluorescence from aniline-amide exci-
plexes, but not from trialkylamine-amide exciplexes, is analo-
gous to the behavior reported by Barlow et al.10 for the
N-(aminoalkyl)phthalimides 17 and 18. The lower oxidation
potential of anilines vs trialkylamines should result in the
formation of more stable exciplexes (eq 3). In addition, the
Franck-Condon factor for exciplex fluorescence may be
unfavorable for trialkylamine exciplexes because of the large
change in geometry between the pyramidal neutral and planar
cation radical.14a Intramolecular exciplex fluorescence has been
observed for the linked (aminoalkyl)phenanthrenes 19-21, even
though the thermodynamics of electron transfer are even less
favorable than for the amine-amides investigated in this
study.22,23 In the case of 19 exciplex fluorescence is observed
in hexane,22 whereas moderately polar solvents are necessary
for the observation of exciplex fluorescence from 20 and 21.23
The differences in linked (aminoalkyl)phenanthrenes can be
correlated to geometric restrictions upon overlap of the phenan-
threne and amine. Overlap requires little nuclear motion in the
case of 19, chain folding in the case of 20 or 21, and is
geometrically impossible in the case of 2 or the Z isomers of 6
and 7.
467
489
512
450
525
dichloromethane
hexane
3
0.6 (0.51), 2.0 (0.49)
0.8 (0.69), 2.5 (0.31)
0.6 (0.61), 2.3 (0.38)b
THF
MeCNc
4:PhNMe2 hexane
diethyl ether
560
440
469
495
418
470
510
2.8
5.4
3.2
dichloromethane
8
hexanec
THF
MeCN
0.6 (0.69), 3.5 (0.31)
0.2 (0.68) , 1.2 (0.32)
a Decay times in nitrogen-purged solutions determined at emission
maximum except as noted. Values in parentheses are decay preexpo-
nentials. b Decays monitored at 450 nm. c Very weak exciplex fluo-
rescence.
PhNMe2 are near the diffusion limit in hexane solution (Table
3) and have similar values in acetonitrile solution. Singlet decay
times for the linked anilines 3, 8, and 12 are near the time
resolution of our instrumentation in hexane and are even shorter
in acetonitrile (Table 2). Rate constants for intermolecular
quenching of 1, 4, and 13 by Et3N are approximately a factor
of 10 below the diffusion limit in hexane solution and show
small, variable increases in acetonitrile solution (Table 3). The
small solvent effect, like the weak dependence on ∆Get, is
consistent with a relatively small extent of electron transfer in
the quenching process. The solvent dependence of the value
of kq for intramolecular quenching by linked N,N-dimethyl-
alkylamines displays an interesting structural dependence. Most
of the Z isomers (2, 6, 9, and 14) and E isomers (6, 7, and 9)
display moderate rate increases (factors of 5 to 20) in acetonitrile
vs hexane solution (Table 4). Smaller increases in kq are
observed for Z- and E-10, plausibly reflecting steric hindrance
of solvation of the bulky isopropyl amine.
In the case of the piperazine 11, a decrease in kq is observed
in acetonitrile vs hexane solution. A decrease in the value of
kq as the solvent dielectric increases beyond ca. 20 is fairly
common for rigidly linked donor-acceptor systems.3b,21 This
behavior has been explained in terms of Marcus theory as
resulting from a greater increase in the solvent reorganization
energy relative to ∆GET as the solvent dielectric constant
increases. In physical terms, at the transition state a charge is
produced before solvent reorganization can occur. The solvent
shell initially corresponds to a neutral molecule and thus will
be destabilizing to the charged state. This solvent-induced
barrier increases as the solvent dielectric constant increases,
resulting in a decrease in the probability of charge transfer and
in the observed value of kq. Systems with flexible linkers can
adopt folded conformations, decreasing their dipole moments
and thus reducing the solvent-induced barrier.
Exciplex fluorescence is observed for the aniline-amide 3
even though the aminoalkyl group is Z with respect to the amide
carbonyl and thus cannot achieve direct overlap with the
phenanthrene π-orbitals. Arene-aniline exciplexes adopt a
sandwich-type geometry when possible. However, exciplex
fluorescence has been observed for intramolecular arene-aniline
exciplexes that cannot adopt sandwich geometries.24 The
absence of exciplex fluorescence from the piperazine 12
indicates that some conformational mobility is required for
exciplex formation, even in the case of the stronger aniline
donors. We presume that the fluorescent exciplex of 3 has a
folded geometry similar to that shown for Z-6 in Figure 1a. In
this geometry there is no direct overlap of phenanthrene
(acceptor) and amine (donor) orbitals and the amide acts as an
intermediary in the electron transfer process.
Since 8 exists as a mixture of Z and E isomers (1.3:1 ratio),
its exciplex fluorescence (Table 5) might be attributed to either
one or both of these isomers. Similarities between the behavior
of the exciplexes formed by 3 and 8 suggest that the exciplex
fluorescence of 8 arises only from the Z isomer. The slopes of
Lippert-Mataga plots26 (Figure 2) of the exciplex energy vs
the solvent polarity parameter ∆f (eqs 5 and 6) are similar for
3 and 8. The slopes of the plots for the intramolecular
exciplexes are larger than those for the intermolecular exci-
plexes. This difference in slopes is indicative of a larger
exciplex dipole moment for the intra- vs intermolecular exci-
plexes, which could result from a larger separation of arene
acceptor and amine donor in the intramolecular exciplex. The
identical slopes of the Lippert-Mataga plots for 3 and 8 indicate
that they most likely have similar geometries. The more
compact, folded geometry expected from E-8 should have a
Exciplex Fluorescence. As previously noted, intermolecular
fluorescence quenching of 1 and 4 by N,N-dimethylaniline is
accompanied by the appearance of exciplex fluorescence.
Exciplex fluorescence is also observed upon intramolecular
fluorescence quenching in the aniline-amides 3 and 8 but not
12. Exciplex fluorescence maxima in several solvents are
reported in Table 5. The maxima for the secondary amide
exciplexes (1:aniline and 3) are at lower energy than those for
the tertiary amide exciplexes (4:aniline and 8) in hexane solution
the difference in energy being comparable to the difference in
reduction potentials for 1 vs 4 (Table 3). The exciplex
fluorescence intensity for 8 is very weak in hexane and stronger
in more polar solvents, including acetonitrile, whereas the

4060 J. Phys. Chem., Vol. 100, No. 10, 1996
Lewis and Burch
Figure 2. Solvent polarity dependence of the exciplex fluorescence
maxima for 1:PhNMe2 (O), 3 (b), 4:PhNMe2 (4), and 8 (2).
Figure 3. Temperature dependence of the fluorescence excitation and
emission spectra of 3 in methylcyclohexane solution.
dipole moment similar to that of the intermolecular exciplexes
and smaller than that of 3.
group. In order to test this possibility, the fluorescence behavior
of the corresponding 9-anthracenecarboxamide 22 was inves-
tigated. Dual exponential intramolecular exciplex decay was
observed for 22 even though the two amide faces are equivalent.
Two exciplex geometries could also result from folding of the
aminoalkyl group over the amide with either the N-phenyl or
the N-methyl occupying the position cis to the carbonyl oxygen
in the folded conformation.
νex ) νex° - (2µ2/hcF3)∆f
(5)
(6)
∆f ) (ꢀ - 1)/(2ꢀ + 1) - (n2 - 1)/(4n2 + 2)
Our assignment of the exciplex fluorescence from 8 to the Z
isomer leaves to be answered the question of why the exciplexes
of E-6-8 are nonfluorescent even though electron transfer
quenching is rapid and arene-amine orbital overlap can readily
be achieved. The absence of exciplex fluorescence from an
intramolecular exciplex in nonpolar solvents requires that return
electron transfer and/or intersystem crossing be very rapid. As
discussed in the following section, the observation of phospho-
rescence from E-8 in low-temperature glasses suggests that rapid
intersystem crossing may be responsible for the absence of
exciplex fluorescence from E-8.
Single exponential exciplex decay is observed for the
intermolecular exciplex 4:PhNMe2, whereas more complex
decay is observed for the intramolecular exciplexes (Table 5).
The exciplex decays reported for 3 and 8 were obtained from
triexponential fits (ø2 < 1.25) in which the two reported
components account for >99% of the fluorescence intensity.
Both of the reported exciplex decay times are longer than the
locally excited fluorescence decays reported in Table 2, indicat-
ing that exciplex formation is largely irreversible. Our inability
to detect a rising component in the exciplex decays is consistent
with locally excited decay times, which are comparable to the
time resolution of our apparatus. In the case of 3 the two
exciplex decays at different emission wavelengths in THF
solution have slightly different preexponentials.
One possible explanation for the observation of dual expo-
nential exciplex decay with similar contributions to the total
fluorescence intensity for 3 and 8 (Table 5) is that there are
two noninterconverting exciplex geometries. A similar explana-
tion for dual exponential exciplex decay has been proposed by
Helsen et al.25 for some intramolecular pyrene-indole exci-
plexes. Two exciplex geometries could result from folding of
the aminoalkyl group over the nonequivalent faces of the amide
Temperature Dependence of Exciplex Fluorescence. The
temperature dependence of the fluorescence excitation and
emission spectra of the linked aniline-amides 3 in methylcy-
clohexane (MC) solution are shown in Figure 3. Decreasing
temperature results in a marked increase in exciplex fluorescence
intensity at temperatures below 250 K. Similar behavior is
observed for 8. There is little change in the appearance of the
excitation or emission spectra of 3 or 8 below the glass transition
temperature of MC other than the appearance of phenanthrene-
like phosphorescence (λmax ) 470 nm) from 8. Phosphores-
cence is also observed for the piperazine 12 at 77 K in a MC
glass. Weak, structureless emission (λmax ∼ 420 nm) is also
observed from 12 in the glass but not in fluid solution at or
above 155 K. To summarize, at 77 K we observe only exciplex
fluorescence from 3, which exists exclusively in the Z config-
uration, a mixture of exciplex fluorescence and locally excited
phosphorescence from 8, which exists as a mixture of E and Z
isomers, and mainly phosphorescence from 12. These observa-
tions support the proposal that only Z isomers form fluorescent
exciplexes.
The increase in exciplex fluorescence intensity with decreas-
ing temperature for 3 and 8 is accompanied by a change in the

N-(Aminoalkyl)-9-phenanthrenecarboxamides
J. Phys. Chem., Vol. 100, No. 10, 1996 4061
dominant decay from the short-lived components observed at
room temperature (Table 5) to long-lived components at 205 K
and below. At 155 K the decay of 3 is dominated by two
components with decay times of 16 and 52 ns and the decay of
8 by components with decay times of 22 and 66 ns. Weakly
bound intramolecular arene-amine exciplexes often display a
modest increase in exciplex fluorescence intensity and lifetime
with decreasing temperature as a consequence of enhanced
through-bond interactions14a or decreased rate constants for
exciplex dissociation.23,27 However, intramolecular exciplex
fluorescence usually is very weak below the glass transition
temperature because of restricted conformational mobility.
Exceptions to this generalization have been reported. Okada
and Mataga28 observed a high ratio of exciplex:locally excited
fluorescence for the linked (aminoalkyl)anthracene 23 and
suggest that a loosely bound “molecular pair” might be formed
in the matrix. Barlow et al.10 also observed exciplex fluores-
cence from 18 at 77 K in several glassy matrices. The ratio of
exciplex:locally excited fluorescence intensity was found to
increase with increasing concentration in an ethanol glass,
leading to the proposal that exciplex fluorescence might be
attributed to molecular aggregates.
(acceptor) and amine (donor) orbitals (Figure 1a). Rate
constants for quenching by Z conformers increase as the
phenanthrene-amide dihedral angle æ1 (eq 1) decreases, owing
to increased mixing of the arene and amide frontier orbitals.
Dual exponential decay observed for the piperazines 11 and 12
is attributed to slightly different rates of intramolecular electron
transfer in the two chair conformers (Figure 1b). Increasing
solvent polarity results in an increase in the rates of intramo-
lecular electron transfer for the flexible (aminoalkyl)amides but
a decrease in rates for the semirigid piperazines. Intramolecular
quenching is far more sensitive than intermolecular quenching
to changes in the solvent, the phenanthrene singlet state
reduction potential, and the amine oxidation potential.
Exciplex fluorescence is only observed for the Z conformers
of aniline donors. A pronounced increase in exciplex fluores-
cence intensity and lifetime is observed below 250 K in
methylcyclohexane solution or the frozen glass. These changes
are attributed to the formation of ground state dimers or
aggregates at low temperatures. The exciplexes of E conformers
apparently decay via intersystem crossing. The less stable
exciplexes formed by the Z and E conformers of trialkylamine
donors are nonfluorescent.
The aminoalkyl-9-phenanthrenecarboxamides are moderately
stable upon exposure to ultraviolet light in deoxygenated
solution. Intramolecular arene-amine photoaddition reactions
have been observed for secondary, but not tertiary, (aminoalkyl)-
phenanthrenes.22,23 Intramolecular cycloaddition reactions have
also been observed for ester-linked styrene-phenanthrenecar-
boxylates.31 The photostability of the (aminoalkyl)phenan-
threnes plausibly reflects the short lifetimes of the exciplex
intermediates or the absence of least-motion pathways for
exciplex proton transfer processes.14g,h
Several lines of evidence support the assignment of the
intense, long-lived exciplex fluorescence observed for 3 and
Z-8 below 250 K (Figure 3) to molecular aggregates. Among
these are the increase in the exciplex fluorescence intensity and
broadening of the fluorescence excitation spectra at low tem-
peratures and the similar appearance of the total emission
spectrum above and below the glass transition temperature.
Moreover, increasing the concentration of 3 from 10-5 to 10-4
M has little effect on the ratio of exciplex:locally excited
fluorescence at 295 K or in the 77 K glass but results in an
increase in this ratio from 4:1 to 12:1 at 205 K. Both 3 and
Z-8 can pack as head-to-tail dimers (or extended aggregates),
whereas the geometrical constraints of the piperazine ring
prevent 12 from forming head-to-tail dimers. The geometry of
the intermolecular phenanthrene-aniline ground state complex
is presumably similar to that for the van der Waals complexes
of anthracene with anilines.29 Folded ground state conformers
have been implicated in some intramolecular electron transfer
processes and exciplex formation and might be responsible, in
part, for the strong low-temperature exciplex emission for 3 and
8.30
Experimental Section
General Methods. Melting points were determined on a
Fisher-Johns apparatus and are uncorrected. 1H NMR spectra
were recorded on a Varian Gemini 300 or Varian XLA 400
spectrometer with TMS as an internal standard. UV-visible
absorption spectra were measured with a Hewlett-Packard 8452
A diode-array spectrometer. Fluorescence spectra were recorded
using a Spex Fluoromax. Samples of 10-5 M phenanthrene
derivative in dried solvent contained in a square quartz cell
equipped with a side arm and serum cap were deoxygenated
by purging with nitrogen for 10 min. For low-temperature
fluorescence spectra, a nitrogen-cooled Oxford Instruments
optical cryostat (DN 1704) connected to an ITC 4 temperature
controller ((0.5 K) was mounted in the sample compartment
of the fluorescence spectrophotometer. IR spectra were recorded
using a Mattson FT-IR. Fluorescence quantum yields were
determined relative to phenanthrene in hexane (ΦF ) 0.13)32
for solutions of matched absorbance (ca. 0.1 OD).
Fluorescence decays were acquired for solutions prepared as
described above using a PTI-LS1 single photon counting
instrument with a gated hydrogen arc lamp (time resolution ca.
0.2 ns). Samples were excited at 260 nm (A ∼ 0.5) and decays
measured at the emission maximum with the acquisition of
12 000 counts in the peak channel. The decays were decon-
voluted using a single or multiexponential least-squares analysis
as described by James et al.11a The goodness of the fit was
judged by the reduced ø2 value as well as the randomness of
the residuals and the autocorrelation function. The ø2 values
were less than 1.2.
Conclusion
The ease of construction of the amide linkage has permitted
the investigation of intramolecular electron transfer for a large
number of aminoalkyl-9-phenanthrenecarboxamides. Tertiary
(aminoalkyl)amides 6-10 exist as mixtures of Z and E
conformers in which the planar amide group is essentially
orthogonal to the phenanthrene plane in solution. The piper-
azines 11 and 12 exist as mixtures of chair conformers.
Secondary (aminoalkyl)amides exist predominantly as Z con-
formers and have smaller excited state phenanthrene-amide
dihedral angles, whereas the lactam ring in 14 is planar.
These conformational properties have a pronounced effect
upon the intramolecular electron transfer quenching of amide-
linked phenanthrene-amines. More rapid quenching is ob-
served for the E vs Z conformers of the tertiary amides as a
consequence of more effective overlap of the phenanthrene
Cyclic voltammetric experiments were performed with a
three-electrode Bioanalytical Systems Inc. (BAS) apparatus in
nitrogen-purged acetonitrile solutions using Et4NBF4 as the

4062 J. Phys. Chem., Vol. 100, No. 10, 1996
Lewis and Burch
electrolyte.33 The reference electrode consisted of Ag/AgI
separated from the solution by a fitted glass bridge with a
VYCOR membrane and filled with a saturated Et4NI solution
in CH3CN. The working electrode (BAS) consisted of a 1.5
mm diameter polished platinum disk embedded in a cobalt glass
seal. The counterelectrode was a platinum wire (BAS). The
scan rate was 0.10 Vs-1 and the step size was 2.44 mV. The
halfwave potential of ferrocene was 0.668 ( 0.005 V under
these conditions.
TABLE 6: Synthesis and Characterization of
9-Phenanthrenecarboxamides
method, mp
compound yield
IR
1H NMR 1H NMR
(oC) (CdO, cm-1
) amide Me amine Me
2
3
5
6
7
8
9
10
11
12
14
A, 79% 124-5
1651.2
2.27
3.02
B, 7% 178-9
1658.9
1620.3
1624.2
1624.2
1628.0
1624.2
1614.5
1616.5
1630.0
1672.4
A, 81% oil
A, 16% oil
A, 54% oil
B, 11% 143-4
A, 89% 76-9
A, 34% oil
A, 75% 102-4
B, 9% 188-9
C, 4% 144-6
2.83, 3.25a
2.87, 3.29a 2.41, 1.90a
2.85, 3.27a 2.32, 1.98a
3.10, 3.35a 2.84, 2.64a
2.88, 3.28a 2.59, 2.14a
1.11, 1.43b 1.30, 0.69b
3.29, 4.0d 2.32
3.43, 4.1d
The force-field-type calculations were performed on a Mac-
intosh IIsi computer using an MM2-type field as supplied in
the Chem 3D software package.9
4.85e
2.35
A series of alkyl-chain conformations were examined. All
bonds and angles were allowed to relax to a minimum
convergent energy on the force field surface for a range of linker
conformations to determine local and global energy minima.
In most cases this was done for more than one very similar
geometry to ensure a true local minimum.
a Chemical shifts (ppm) in CDCl3 solution. First value is for the Z
isomer, and the second is for the E isomer. b Chemical shift of isopropyl
methyl group. First value is for the Z isomer. d Chemical shifts of
R-methylenes. e Chemical shift of lactam methylene.
(100%, 27% overall). 1H NMR (CDCl3): d ) 2.90 (t, 2H),
2.94 (s, 3H), 3.37 (t, 2H), 6.70 (t, 1H), 6.75 (d, 2H), 7.22 (t,
2H).
Materials. Dichloromethane (Aldrich) was distilled over
calcium hydride prior to use. Benzene (Aldrich) was distilled
over sodium metal prior to use. Tetrahydrofuran (Aldrich) was
distilled over molecular sieves prior to use. All other solvents
were freshly opened spectral grade or HPLC grade (Aldrich and
Fisher) and were used without further purification. Diethyl-
amine (Aldrich) and triethylamine (Aldrich) were distilled prior
to use. N-methylaniline (Aldrich) and N,N-dimethylaniline-
(Aldrich) were distilled under reduced pressure prior to use.
N,N-dimethylethylenediamine, N,N,N′-trimethylethylenediamine,
N,N′-dimethylethylenediamine, N,N′-diisopropylethylenedi-
amine, N,N,N′-trimethylpropylenediamine, N-methylpiperizine,
N-phenylpiperizine, N-methyl-N-isopropylamine, and N,N-di-
methylethylamine (Aldrich) were used as received. The syn-
thesis of phenanthrene-9-carboxylic acid chloride, 1, 4, and 13
have been previously described.2,6
N,N′-Dimethyl-N-phenylethylenediamine. N,N′-dimethyl-N-
phenylethylenediamine was synthesized in an analogous manner
using methylamine gas (Fluka). N-methyl-2-(N′-methylaniline)-
acetamide was isolated as a yellow solid (41%). 1H NMR
(CDCl3): d ) 2.84 (d, 3H), 3.01 (s, 3H), 3.86 (s, 2H), 6.60 (b,
1H), 6.73 (d, 2H), 6.85 (t, 1H), 7.28 (t, 2H). N,N′-dimethyl-
N-phenylethylenediamine was isolated as a brown oil (40%
overall). 1H NMR (CDCl3): d ) 2.29 (s, 2H), 2.47 (s and t,
2H), 2.81 (t, 1H), 2.95 (s, 3H), 3.46 (m, 2H), 6.70 (t, 1H), 6.77
(d, 2H), 7.23 (t, 2H).
The phenanthrene-9-carboxamide derivatives were synthe-
sized by the three general methods described below. The
1
method, yield, mp, and IR and H NMR data are summarized
in Table 6. Absorption and fluorescence spectral data are
summarized in Table 1.
N-Methyl-N-phenylethylenediamine. N-methylaniline (0.09
mol) and sodium methoxide (Aldrich, 0.095 mol) were mixed
in methanol (50 mL) and cooled to 0 °C in an ice bath. To this
mixture, R-bromoethylacetate (Aldrich, 0.09 moles) was added
by addition funnel over 1 h while stirring. After addition was
completed, the mixture was stirred 2 more hours at 0 °C and
then overnight at room temperature. The solvent was removed
under vacuum and the residual oil dissolved in ethyl ether (50
mL). The organic layer was washed with 5% aqueous NaOH
and then H2O. The organic layer was dried with CaCl2 and
the solvent removed. TLC (CHCl3/MeOH, 3/2) showed one
product. The crude 2-(N-methylaniline)ethylacetate (0.09 mol)
and sodium methoxide (0.095 mol) were dissolved in methanol
(50 mL) and bubbled with ammonia (Fluka) for 15 min. The
mixture was sealed and stirred overnight at 40 °C and then
refluxed for 8 h. The solvent was removed under vacuum and
the residual oil dissolved in ethylacetate, washed with 5%
aqueous NaOH and then with H2O. The organic layer was dried
with CaCl2 and the solvent removed. The crude solid was
dissolved in hot ethylacetate and cooled to yield 4 g (27%) of
2-(N-methylaniline)acetamide as gray plates. mp 168-9 °C.
1H NMR (CDCl3): d ) 3.04 (s, 3H), 3.87 (s, 2H), 5.53 (b,
1H), 6.46 (b, 1H), 6.76 (d, 2H), 6.85 (t, 1H), 7.27 (t, 2H).
Method A.34 Phenanthrene-9-carboxylic acid chloride (0.0025
mol) was dissolved in benzene (5 mL), and an excess of the
amine (ca. 10 equiv) was added. The mixture was stirred
overnight at 35 °C. The solution was dissolved in ethylacetate
(50 mL) and washed 2 times with 5% aqueous HCl (50 mL), 2
times with 5% aqueous KOH (50 mL), and 2 times with H2O.
The organic layer was dried with CaCl2 and the solvent removed
under vacuum. The product was then purified by column
chromatography (silica, CHCl3, or CHCl3/5%MeOH), and solid
products were recrystallized 2 times from either ethylacetate,
ethylacetate/ hexane, or EtOH.
Method B.35 This method was used for the aniline derivatives
that were not stable in the presence of the acid chloride.
Phenanthrene-9-carboxylic acid (0.0045 mol) was mixed with
the amine (0.006 mol) and 1,3-dicyclohexylcarbodiimide (DCC,
Aldrich, 0.005 mol) in benzene (30 mL) under N2. The mixture
was stirred overnight at room temperature. The solvent was
evaporated and the DCC salt was removed by successive
suspension of the crude mixture in ethyl ether and filtration.
The product was then purified by column chromatography
(silica, CHCl3) 2 times and then recrystallized 2 times from
ethylacetate/hexane.
Method C.6 The lactam 14 was prepared by the method
employed for the preparation of 13 using N,N-dimethylethlene-
diamine instead of methylamine.
The 2-(N-methylaniline)acetamide (0.011 mol) was dissolved
in tetrahydrofuran (25 mL) and slowly added to a stirred mixture
of LiAlH4 (0.044 mol) in refluxing tetrahydrofuran (50 mL).
The mixture was refluxed for 48 h. After cooling, the excess
LiAlH4 was reacted with H2O (ca. 2 mL), the solution was
filtered, dried with CaCl2, and the solvent removed to yield a
1.6 g of N-methyl-N-phenylethylenediamine as a brown oil
Acknowledgment. Financial support for this project was
provided by the National Science Foundation. We thank F. G.
Bordwell for the use of the electrochemical equipment, M. Gahr

N-(Aminoalkyl)-9-phenanthrenecarboxamides
J. Phys. Chem., Vol. 100, No. 10, 1996 4063
(16) de Silva, A. P.; Gunaratne, H. Q. N.; Habib-Jiwan, J.-L.; McCoy,
C. P.; Rice, T. E.; Soumillion, J.-P. Angew. Chem., Int. Ed. Engl. 1995, 34,
1728.
(17) INDO/S calculations indicate that the LUMO of 14 has delocalized
enonecharacter but that the carbonyl π-orbial coefficients for the other
frontier orbitals of 1, 4, and 14 are insignificant.6
(18) For the use of piperidines as semirigid spacers see the following.
Schuddeboom, W.; Scherer, T.; Warman, J. M.; Verhoeven, J. W. J. Phys.
Chem. 1993, 97, 13092.
for the subnanosecond lifetime measurements, A. M. Miller for
some of the steady state absorption and fluorescence data, and
B. A. Yoon for some of the NMR spectral data.
References and Notes
(1) Stewart, W. E.; Siddall, J. H. Chem. ReV. 1970, 70, 517.
(2) Lewis, F. D.; Barancyk, S. V.; Burch, E. L. J. Am. Chem Soc. 1992,
114, 3866.
(3) (a) Schmidt, J. A.; McIntosh, A. R.; Weedon, A. C.; Bolton, J. R.;
Connolly, J. S.; Hurley, J. K.; Wasielewski, M. R. J. Am. Chem Soc. 1988,
110, 1733. (b) Liu, J. Y.; Bolton, J. R. J. Phys. Chem. 1991, 95, 6924.
(4) Wasielewski, M. R. Chem. ReV. 1992, 92, 435.
(5) For a preliminary account of this aspect of our results see the
following. Lewis, F. D.; Barancyk, S. V.; Burch, E. L. J. Am. Chem Soc.
1994, 116, 1159.
(6) Lewis, F. D.; Burch, E. L. J. Photochem. Photobiol. A, in press.
(7) Lewis, F. D.; Burch, E. L. Manuscript in preparation.
(8) Lewin, A. H.; Frucht, M. Org. Magn. Reson. 1975, 7, 206.
(9) Chem3D Plus; Cambridge Scientific Computing, Inc.: Cambridge,
MA, 1990.
(10) Barlow, J. H.; Davidson, R. S.; Lewis, A.; Russell, D. R. J. Chem.
Soc., Perkin Trans. 2 1979, 1103.
(11) (a) James, D. R.; Siemiarczuk, A.; Ware, W. R. ReV. Sci. Instrum.
1992, 63, 1710. (b) Ware, W. R.; Watt, D.; Holmes, J. D. J. Am. Chem
Soc. 1974, 96, 7853. (c) Lewis, F. D.; Bassani, D. M. J. Photochem.
Photobiol. A 1992, 66, 43.
(12) Weller, A. Z. Phys. Chem. (Wiesbaden) 1982, 133, 93.
(13) (a) Lewis, F. D.; Ho, T.-I. J. Am. Chem. Soc. 1977, 99, 7991. (b)
Lewis, F. D.; Correa, P. C. J. Am. Chem. Soc. 1984, 106, 194.
(14) (a) Van der Auweraer, M.; Grabowski, Z. R.; Rettig, W. J. Phys.
Chem. 1991, 95, 2083. (b) Van Haver, Ph.; Helsen, N.; Depaemelaere, S.;
Van der Auweraer, M.; De Schryver, F. C. J. Am. Chem. Soc. 1991, 113,
6849. (c) Van der Auweraer, M.; Viaene, L.; Van Haver, Ph.; De Schryver,
F. C. J. Phys. Chem. 1993, 97, 7178. (d) Pasman, P.; Mes, G. F.; Koper,
N. W.; Verhoeven, J. W. J. Am. Chem. Soc. 1985, 107, 5839. (e) Wegewijs,
B.; Hermant, R. M.; Verhoeven, J. W.; Kunst, A. G. M.; Rettschnick, R. P.
H. Chem. Phys. Lett. 1987, 140, 587. (f) Verhoeven, J. W. Pure Appl. Chem.
1990, 62, 1585. (g) Lewis, F. D.; Bassani, D. M.; Burch, E. L.; Cohen, B.
E.; Engelman, J. A.; Reddy, G. D.; Schneider, S.; Gahr, M. J. Am. Chem.
Soc. 1994, 116, 597. (h) Lewis, F. D.; Reddy, G. D.; Bassani, D. M.;
Schneider, S.; Jaeger, W.; Gedeck, P.; Gahr, M. J. Am. Chem. Soc. 1994,
116, 597.
(19) The oxidation potentials of Et3N and N-methylpiperidine are 1.42
and 1.28 V, respectively.
(20) (a) Brouwer, A. M.; Langkilde, F. W.; Bajdor, K.; Wilbrandt, R.
Chem. Phys. Lett. 1994, 225, 386. (b) Jacques, P.; Allonas, X. Chem. Phys.
Lett. 1995, 233, 533.
(21) (a) Oliver, A. M.; Craig, D. C.; Paddon-Row, M. N.; Kroon, J.;
Verhoeven, J W. Chem. Phys. Lett. 1988, 150, 366. (b) Borovkov, V. V.;
Ishida, A.; Takamuku, S.; Sakata, Y. Chem. Lett. 1993, 737.
(22) Sugimoto, A.; Sumida, R.; Tamai, N.; Inoue, H.; Otsuji, Y. Bull.
Chem. Soc. Jpn. 1981, 54, 3500.
(23) Lewis, F. D.; Cohen, B. E. J. Phys. Chem. 1994, 98, 10591.
(24) (a) Wasielewski, M. R.; Minsek, D. W.; Niemczyk, M. P.; Svec,
W. A.; Yang, N. C. J. Am. Chem Soc. 1990, 112, 2823. (b) Rettig, W.;
Haag, R.; Wirz, J. Chem. Phys. Lett. 1991, 180, 216.
(25) Helsen, N.; Viaene, L.; Van der Auweraer, M.; De Schryver, F. C.
J. Phys. Chem. 1994, 98, 1532.
(26) Lippert, E. Z. Naturforscher 1955, 109, 541. (b) Mataga, N.; Kaifu,
Y.; Koizumi, M. Bull. Chem. Soc. Jpn. 1955, 28, 690.
(27) Swinnen, A. M.; Van der Auweraer, M.; De Schryver, F. C.;
Nakatani, K.; Okada, T.; Mataga, N. J. Am. Chem. Soc. 1987, 109, 321.
(28) Okada, T.; Mataga, N. Bull. Chem. Soc. Jpn 1976, 49, 2190.
(29) Piuzzi, F.; Brenner, V.; Millie´, P.; Tramer, A. J. Photochem.
Photobiol. A 1994, 80, 95.
(30) (a) Reynders, P.; Ku¨hnle, W.; Zachariasse, K. J. Am. Chem. Soc.
1990, 112, 3928. (b) Kawakami, J.; Iwamura, M.; Nakamura, J. Chem. Lett.
1992, 1013. (c) Yu, H.-T.; Vela, M. A.; Fronczek, F. R.; McLaughlin, M.
L.; Barkley, M. D. J. Am. Chem. Soc. 1995, 117, 348.
(31) Itoh, H.; Takada, M.; Asano, F.; Senda, Y.; Sakuragi, H.; Tokumaru,
K. Bull. Chem. Soc. Jpn. 1993, 66, 224.
(32) Berlman, I. B. Fluorescence Spectra of Aromatic Molecules;
Academic Press: New York, 1971.
(33) For a full description see the following. Bordwell, F. G.; Cheng,
J.-P.; Bausch, M. J. J. Am. Chem. Soc. 1988, 110, 2867.
(34) Wilson, W. D.; Wang, Y.-H.; Kusuma, S.; Chandrasekaran, S.;
Boykin, D. W. Biophys. Chem. 1986, 24, 101.
(15) Amide E,Z photoisomerization has been observed for nonconjugated
secondary amides. See the following. (a) Wang, Y.; Purrello, R.; Spiro, T.
G. J. Am. Chem. Soc. 1989, 111, 8274. (b) Song, S.; Asher, S. A.; Krimm,
S.; Shaw, K. D. J. Am. Chem. Soc. 1991, 113, 1155.
(35) Sheehan, J. C.; Hess, G. P. J. Am. Chem. Soc. 1955, 77, 1067.
JP9526758