614 Stephen P. Nighswander-Rempel et al.
The spectrograph was equipped with a liquid N2 cooled CCD
detector array (Series 200; Photometrics Ltd., Tuscon, AZ based on
EEV CCD15-11 1024·256 CCD chip [Marconi Applied Technologies,
Chelmsford, Essex, UK]). The Holospex grating we used had a range
from 35 to 3556 cm)1. A holographic notch filter was placed in the
spectrometer to reject the laser line (bandwidth of 300 cm)1, optical
density of >6 at the excitation wavelength and >80% transmission
for scattered light). Acquired spectra were corrected for variations in
detector sensitivity and a baseline correction was applied.
Calculated Raman spectra. Minimum energy geometries, harmonic
vibrational frequencies and Raman intensities for the MHMI mono-
mer and dimer, DHICA, indole and ICA were computed using B3LYP
hybrid density functional theory (23) and a 6-31(d) basis set (24) within
the Gaussian03 software suite (25). Frequencies were scaled by a factor
of 0.97, which is within the range of scaling factors suitable for the
methods used (26). A uniform Lorentzian broadening of 5 cm)1 was
applied to generate the simulated spectra, which were then normalized
by visual inspection against the experimental data. Cartesian coordi-
nates, B3LYP ⁄ 6-31(d) ground state energies and complete lists of
frequencies and Raman intensities are provided as Supplemental
Material.
possibly even the 3 positions, yielding several structurally
and possibly optically distinct compounds (9). Therefore, we
have synthesized a new compound, N-methyl-5-hydroxy-6-
methoxy-indole (MHMI), which is ideally suited to such
studies (22). Functional groups in MHMI have been
strategically placed on the indole framework, sterically
hindering binding to the 2 and 7 positions and allowing
dimerization via 4-4¢ coupling alone. We examine the near-
IR Raman spectra of MHMI and MHMI dimer, as well as
DHICA, indole and ICA for comparison. Comparing spectra
with calculations of the vibrational modes will provide
insight into the effect of dimerization on the vibrational
spectra of indolic compounds.
MATERIALS AND METHODS
Compound synthesis. 5-Benzyloxy-6-methoxyindole purchased from
Sigma Aldrich was N-methylated in dry DMF using sodium hydride as
the base and methyl iodide as the alkylating agent. The reaction was
carried out under an atmosphere of argon for 2 h. The N-methyl-5-
benzyloxy-6-methoxyindole thus obtained was subjected to hydrogen-
olysis for 3 h in ethyl acetate-acetic acid solution, using 5% palladium
on carbon as catalyst to furnish N-methyl-5-hydroxy-6-methoxyindole
(MHMI) monomer. The crude product thus obtained was purified
RESULTS
Validation of the spectrometer used here was performed by
comparison of indole and ICA Raman spectra (Fig. 1a,b) with
those available from the literature (27,28). The frequency
positions of peaks shown in the spectra acquired by the current
system matched those of previously acquired Raman spectra
precisely (data not shown). Comparison of the Raman spectra
with previous neutron scattering studies of indole (21) showed
that only the in-plane vibrations were evident in the Raman
spectra. This is to be expected, as only those vibrations
corresponding to a change in the molecular polarizability are
Raman active. Out-of-plane vibrations would require a large
displacement from the molecular plane in order to exhibit a
detectable change in the polarizability.
Addition of the carboxylic acid group to the 2 position of
the indole results in significant shifting of vibrational peaks.
Recent inelastic neutron scattering studies together with ab
initio calculations have shown that virtually all peaks between
400 and 1000 cm)1 in both indole and ICA are combination
bands incorporating several vibrational modes, and that the
carboxylic acid group primarily adjusts the bands correspond-
ing to vibrations in the pyrrole ring (21). Similarly, addition of
hydroxy and methoxy groups to the 5 and 6 positions can be
expected to affect vibrations in the benzene ring.
over a column of silica using chloroform as eluant, and then
crystallized from dichloromethane-hexane.
To synthesize the MHMI dimer, 0.65 g of MHMI was dissolved
in ethanol (45 mL) and stirred under argon. Potassium carbonate
(1 g in 6 mL H2O and 30 mL EtOH) was added and the contents left
stirring overnight under argon, then cooled in ice. The red-black
solution was acidified with HCl and extracted with chloroform. The
organic layer was washed with brine, dried and the solvent removed
to obtain 0.66 g dark solid. The solid was washed with toluene, and
dried under vacuum to furnish 0.6 g dark solid. This solid was
treated with ꢀ40 mL warm ethyl acetate, stirred under argon for
0.5 h, filtered and the solvent removed to obtain 0.52 g of a red
foam. Column purification (SiO2, CHCl3) afforded the pure mono-
coupled dimer as an off-white solid in 13% yield. Recrystallization
from ethanol-chloroform afforded yellow micro crystals: mp 200ꢁC
(decomposes); 1H NMR (500 MHz, CD3OD) d 3.77 (6 H, s, NCH3),
3.98 (6 H, s, OCH3), 5.82 (2H, m, H-3,3¢, 6.86 (2H, d, J = 3.0,
H-2,2¢), and 6.99 (2H, s, H-7,7¢). 13C NMR (75 MHz, CD3OD) d
33.05, 56.94, 93.19, 102.32, 115.74, 123.36, 127.74, 132.39, 139.24,
and 147.28; m ⁄ z [ESI+] 353 (MH+, 40%), 375 (MNa+, 100%).
C20H20N2O4Na requires 375.1321. Found: 375.1323; Anal. Calcd. for
C20H20N2O4: C 68.17%, N 7.95%, H 5.72%. Found: C 67.99%, N
7.86%, H 5.75%. This dimer is a monocoupled species from MHMI,
coupled in the 4,4¢ positions, and the two rings are staggered because
of steric interactions. The full crystal structure and other character-
ization details will be published separately.
Raman spectra for the MHMI monomer and dimer are
shown in Fig. 1c,d. In order to attribute these peaks to
molecular vibrations, B3LYP density functional theory
calculations were performed for all compounds, and only
the dominant contributing modes are listed in Tables 1–3.
Calculated spectra to which a broadening term has been
applied are shown in the figures above the experimental data,
and correspond excellently with the experimentally acquired
data. The dominant peak in indole at 762 cm)1 is due to
stretching of the benzene ring primarily and its position is
largely unaffected by the addition of the hydroxy and
methoxy groups. The peaks between 1300 and 1400 cm)1
become much more intense in MHMI relative to indole, and
are well fit by the calculations. The calculations here
attribute them to benzene ring stretches primarily, con-
tradicting older data attributing them to the pyrrole ring
(29). The peak at 1520 cm)1, which is also largely unmoved
Experimental Raman spectroscopy. Raman spectra were acquired
for indole, ICA, MHMI monomer and dimer, and DHICA in the
solid state. Raman spectra were acquired for the compounds in solid
form using a Raman system built in-house (Ontario Cancer Institute,
Toronto). A titanium:sapphire laser (Spectra Physics model 3900S)
pumped by 5 W CW DPSS green laser at 532 nm (diode-pumped
solid-state Verdi V5 laser; Coherent) was used as an excitation source
tuned at 785 nm. Excitation light was passed through a holographic
bandpass filter (Kaiser Optical Systems, Ann Arbor, MI) with greater
than 90% transmission of s-polarized light at 785 nm and
a
bandwidth of less than 2 nm. The rejected light was used to monitor
the wavelength of the excitation laser using a wavemeter WA-1600
(Burleigh Instruments, Inc., Fishers, NY). Scattered light was
collected through a collection system designed in-house and coupled
into a custom-made fiber optic ‘‘Spot to Slit line’’ type bundle
assembly consisting of 43 CeramOptec P ⁄ N UV105 ⁄ 115 ⁄ 140P28
fibers with NA of 0.28 ending in
a line of approximately
0.14 mm · 6.02 mm in the active area (CeramOptec Industries,
Inc., East Longmeadow, MA). The scattered light was analyzed with
a high-throughput holographic spectrograph (Holospec f ⁄ 1.8; Kaiser
Optical Systems).