Propagation of polar substituent effects in 1-(substituted phenyl)-6,7-dimethoxy-3,4-dihydro- and -1,2,3,4-tetrahydroisoquinolines as explained by resonance polarization concept

Article abstract of DOI:10.1021/jo0507946

Propagation of inductive and resonance effects of phenyl substituents within 1-(substituted phenyl)-6,7-dimethoxy-3,4-dihydro- and -1,2,3,4-tetrahydroisoquinolines were studied with the aid of ¹³C and ¹⁵N NMR chemical shifts and ab i

Full text of DOI:10.1021/jo0507946

Propagation of Polar Substituent Effects in 1-(Substituted
phenyl)-6,7-dimethoxy-3,4-dihydro- and
-1,2,3,4-tetrahydroisoquinolines As Explained by Resonance
Polarization Concept
Kari Neuvonen,*,† Ferenc Fu¨lo¨p,‡ Helmi Neuvonen,† Andreas Koch,§ Erich Kleinpeter,§ and
Kalevi Pihlaja†
Department of Chemistry, University of Turku, FIN-20014 Turku, Finland, Institute of Pharmaceutical
Chemistry, University of Szeged, H-6701 Szeged, POB 121, Hungary, and Department of Chemistry,
University of Potsdam, P.O. Box 691553, D-14415 Potsdam, Germany
kari.neuvonen@utu.fi
Received April 19, 2005
Propagation of inductive and resonance effects of phenyl substituents within 1-(substituted phenyl)-
6,7-dimethoxy-3,4-dihydro- and -1,2,3,4-tetrahydroisoquinolines were studied with the aid of 13C
and 15N NMR chemical shifts and ab initio calculations. The substituent-induced changes in the
chemical shift (SCS) were correlated with a dual substituent parameter equation. The contributions
of conjugative (FR) and nonconjugative effects (FF) were analyzed, and mapping of the substituent-
induced changes is given over the entire isoquinoline moiety for both series. The experimental
results can be rationalized with the aid of the resonance polarization concept. This means the
consideration of the substituent-sensitive balance of different resonance structures, i.e., electron
delocalization, and the effect of the aromatic ring substituents on their relative contributions. With
tetrahydroisoquinolines, the delocalization of the nitrogen lone pair (stereoelectronic effect)
particularly contributes. Correlation analysis of the Mulliken atomic charges for the dihydroiso-
quinoline derivatives was also performed. The results support the concept of the substituent-
sensitive polarization of the isoquinoline moiety even if the polarization pattern achieved via the
NMR approach is not quite the same as that predicted by the computational charges. Previously
the concepts of localized π-polarization and extended polarization have been used to explain polar
substituent effects within aromatic side-chain derivatives. We consider that the resonance
polarization model effectively contributes to the understanding of the polar substituent effects.
Introduction
analysis of the systematic substituent effects on the 13
C
NMR chemical shifts propagating all along the molecule
is an effective way to study the changes in electronic
states promoted by different phenyl substituents.1-5 The
substituent effects can follow the normal pattern, for
instance, electron-withdrawing substituents affecting
When substitution in one part of an organic molecule
is varied, reorganization of the electronic framework all
through the molecule occurs. This can affect the mol-
ecule’s reactivity, conformation, solubility, equilibrium
between the different potential tautomeric forms, acid-
base properties or stacking tendency, and so forth. The
(1) Craik, D. J.; Brownlee, R. T. C. Prog. Phys. Org. Chem. 1983,
14, 1-73.
(2) Reynolds, W. F. Prog. Phys. Org. Chem. 1983, 14, 165-203.
(3) (a) Halton, B.; Dixon, G. M. Org. Biomol. Chem. 2004, 2, 3139-
3149. (b) Bonesi, S. M.; Ponce, M. A.; Erra-Balsells, R. J. Heterocycl.
Chem. 2004, 41, 161-170. (c) Mezzina, E.; Spinelli, D.; Lamartina,
L.; Buscemi, S.; Frenna, V.; Macaluso, G. Eur. J. Org. Chem. 2002,
203-208.
* To whom correspondence should be addressed. Fax: +358-2-
3336700.
† University of Turku.
‡ University of Szeged.
§ University of Potsdam.
10.1021/jo0507946 CCC: $30.25 © 2005 American Chemical Society
Published on Web 11/30/2005
10670
J. Org. Chem. 2005, 70, 10670-10678

Propagation of Polar Substituent Effects
low-field (high-frequency) 13C NMR chemical shifts. The
occurrence of the so-called reverse substituent effect
(RSE) is, however, well-established at some sites of, for
example, styrenes, chalcones, carboxylic acid derivatives,
imines, and hydrazones.1-5 RSE means that EW substit-
uents affect upfield 13C NMR chemical shifts (low-
frequency shift values), whereas the ED substituents
affect low-field shifts. Extensive data shows that upfield
chemical shifts generally mean increase in the electron
density at the probe atom.4c,d,5b,6-10
FIGURE 1.
SCHEME 1
The substituent-induced changes in the chemical shift,
SCS, can be correlated with a single parameter equation
that uses substituent parameters such as the Hammett
σ values (eq 1) or with a dual substituent parameter
equation such as eq 2, which allows the blend of the
inductive (σF or σI) and resonance (σR) contributions to
vary.
is, however, often observed for the FI values (cf. eq 2) at
the two ends of a localized dipole. A typical case (3, Figure
1) is discussed by Reynolds et al.2 The FI and ∆δ values
for C(R) and C(â) differ in sign (being opposite) and
markedly also in magnitude [larger for C(â)]. It was
thought by Reynolds et al. that the net polarization
observed is a sum of two different contributing mecha-
nisms, the localized π-polarization and the π-polarization
due to the resonance structures 3.2 Reynolds et al.
suggest the use of the term extended π-polarization for
the latter phenomenon.2 As an alternative the term field-
induced resonance was suggested by Reynolds et al.,2 but
this term was first used by Topsom et al. in a different
meaning.13 The relationship between the localized and
extended polarization has been comprehensively dis-
cussed by Craik et al.12b,14
Even if convenient in predicting propagation of sub-
stituent effects, the π-polarization mechanism exhibits
limitations in quantitative treatments. This prompted us
to consider the importance of the resonance concept.4c,d
Both neutral and charged resonance structures contrib-
ute to the stability or reactivity of organic molecules. The
significance of the charged resonance structures and the
polarizability of a site in the molecule seem to be
connected. According to our resonance polarization con-
cept the substitutions (EW or ED) can either increase or
decrease contributions of charged resonance structures
as compared with the situation when X ) H (cf. Scheme
1). This alteration is reflected in the observed sensitivity
of the measurable property. A decrease in the electron
density at the probe atom by the ED substituents can be
due to the fact that the ED substituents stabilize those
resonance structures where a positive charge is oriented
SCS ) Fσ
(1)
(2)
BA
SCS ) FFσF (or FIσI) + FRσR
In eq 2, different resonance scales (usually σR , σR0, σR
,
+
σR-) can be tested, and the one that yields the best fit to
the experimental data is selected. SCS is the 13C NMR
chemical shift for a substituted compound relative to that
for the unsubstituted one. Even in those cases where a
good correlation with eq 1 is obtained, the use of eq 2
gives more information because it shows the contribu-
tions of conjugative (FR) and nonconjugative effects (FF
or FI). An observed positive F value in eqs 1 and 2 means
a normal substituent effect, and a negative one indicates
a reverse substituent effect.
Qualitatively RSE can be understood on the basis of
the so-called π-polarization mechanism, first proposed by
Reynolds et al.2,11 and later characterized in a more
detailed way by Bromilow et al.1,12 Each π-unit (a double
bond, triple bond, or aromatic ring system) is thought to
be polarized separately, the polarization being induced
by the substituent dipole in another part of the molecule
(see 1and 2). This interaction can be transmitted either
(6) Hehre, W. J.; Taft, R. W.; Topsom, R. D. Prog. Phys. Org. Chem.
1976, 12, 159-187.
(7) Krawczyk, H.; Szatylowicz, H.; Gryff-Keller, A. Magn. Reson.
Chem. 1995, 33, 349-354.
(8) Lin, S.-T.; Lee, C.-C.; Liang, D. W. Tetrahedron 2000, 56, 9619-
9623.
through the molecular framework or through the solvent
continuum. The above interaction is exactly called local-
ized or direct polarization because each π-unit in the side
chain is polarized separately. Considerable nonsymmetry
(9) Matsumoto, K.; Katsura, H.; Uchida, T.; Aoyama, K.; Machiguchi,
T. Heterocycles 1997, 45, 2443-2448.
(10) Alvarez-Ibarra, C.; Quiroga-Feijo´o, M. L.; Toledano, E. J. Chem.
Soc, Perkin Trans. 2 1998, 679-689.
(11) (a) Hamer, G. K.; Peat, I. R.; Reynolds, W. F. Can. J. Chem.
1973, 51, 897-914. (b) Hamer, G. K.; Peat, I. R.; Reynolds, W. F. Can.
J. Chem. 1973, 51, 915-926.
(12) (a) Bromilow, J.; Brownlee, R. T. C.; Craik, D. J.; Fiske, P. R.;
Rowe, J. E.; Sadek, M. J. Chem. Soc., Perkin Trans. 2 1981, 753-759.
(b) Brownlee, R. T. C.; Craik, D. J. J. Chem. Soc., Perkin Trans 2 1981,
760-764. (c) Brownlee, R. T. C.; Craik, D. J. Org. Magn. Reson. 1981,
15, 248-256.
(13) Broxton, T. J.; Butt, G.; Liu, R.; Lay, H. T.; Topsom, R. D.;
Katritzky, A. R. J. Chem. Soc., Perkin Trans. 2 1974, 463-466.
(14) Craik, D. J.; Brownlee, R. T. C.; Sadek, M. J. Org. Chem. 1982,
47, 657-661.
(4) (a) Neuvonen, K.; Fu¨lo¨p, F.; Neuvonen, H.; Pihlaja, K. J. Org.
Chem. 1994, 59, 5895-5900. (b) Neuvonen, K.; Fu¨lo¨p, F.; Neuvonen,
H.; Simeonov, M.; Pihlaja, K. J. Phys. Org. Chem. 1997, 10, 55-66.
(c) Neuvonen, K.; Fu¨lo¨p, F.; Neuvonen, H.; Koch, A.; Kleinpeter, E.;
Pihlaja, K. J. Org. Chem. 2001, 66, 4132-4140. (d) Neuvonen, K.;
Fu¨lo¨p, F.; Neuvonen, H.; Koch, A.; Kleinpeter, E.; Pihlaja, K. J. Org.
Chem. 2003, 68, 2151-2160.
(5) (a) Neuvonen, H.; Neuvonen, K. J. Chem. Soc., Perkin Trans. 2
1999, 1497. (b) Neuvonen, H.; Neuvonen, K.; Koch, A.; Kleinpeter, E.;
Pasanen, P. J. Org. Chem. 2002, 67, 6995-7003.
J. Org. Chem, Vol. 70, No. 26, 2005 10671

Neuvonen et al.
SCHEME 2
Results
Tables S3 and S4 (Supporting Information) give the
NMR chemical shifts measured for the DHIQ and THIQ
sets (Scheme 2), respectively. The measurements were
performed in CDCl3 with a low and constant sample
concentration (0.1 M) to diminish intermolecular associa-
tions. The SCS (substituent-induced chemical shift) data
were analyzed by both eq 1 and eq 2. Equation 2 gave a
better correlation than the single parameter eq 1, with
one exception only (N-2 in meta THIQ series). Therefore,
the correlation parameters obtained by eq 2 were adopted
for the discussion. The resonance parameter scale se-
+
-
lected in each case is the one from σR , σR0, σRBA, and σR
,
toward the ED substituent (C-R in 4, Scheme 1). On the
other hand, the EW substituents can stabilize those
resonance structures where a negative charge is oriented
toward the EW substituent and therefore increase the
electron density at the probe site (C-R in 5, Scheme 1).
In both cases an RSE is observed. The extent of the
inductive stabilization/destabilization depends on the
distance between X and the charged atom. This reso-
nance polarization model in general predicts correctly the
appearance of the local π-polarization. Further, it was
possible to evaluate by this model the importance of
electronic effects governing the position of the various
ring-chain equilibria4c and the origin of the large negative
FR values (-1.6 to -5.0) observed for hydrazones as
compared with the much smaller values for imines (0.2
to -0.7).4d Further, by using the resonance polarization
concept the substituent interaction between X and Y in
X-C6H4-CHdN-Y could be explained.4d
that gave the best fit to the experimental data. The
values are given in Tables 1 and 2. For comparison, in
Supporting Information in Tables S5 and S6 are also
given the FF and FR values obtained by eq 2 when a
constant set of substituent parameters (σF, σR) is used in
each case. In addition, in Tables S7 and S8 are given the
FF and FR values obtained by eq 2 using an alternative
set of substituent parameters (σF, σR0). The trends of the
substituent effects are similar as those shown in Tables
1 and 2. The excellent to satisfactory fits seen in Tables
1 and 2 indicate that the substituent effects are electronic
in origin. The main conclusions discussed below are based
on the shift ranges ∆δ, which are about 0.5-1 ppm or
higher for the carbon atoms and about 3 ppm or higher
for the nitrogen atoms. This means the substituent effects
on C-1, C-6, C-8a, and C-8 of both the DHIQ and THIQ
sets and N-2 of the DHIQ sets and the para THIQ set.
For other carbons the ∆δ values are smaller and conse-
quently the F values are small, too. It is interesting to
see, however, that also in these cases mainly from good
to excellent correlations are observed (Tables 1 and 2).
This means that systematic electronic effects are expe-
rienced over the whole molecule. The patterns of the sub-
stituent-sensitive polar character of the compounds
studied in terms of 13C and 15N NMR chemical shifts
In this work both the 13C and 15N NMR shifts were
recorded for a set of 6,7-dimethoxy-1-(para- or meta-
substituted phenyl)-3,4-dihydroisoquinolines and -1,2,3,4-
tetrahydroisoquinolines (DHIQ and THIQ, respectively;
Scheme 2) to investigate the substituent effects on their
polar character. This structural study is also due to the
fact that the isoquinoline system is present in many
medicinal compounds and alkaloids.
TABLE 1. Correlation Parameters G from Equation 1 and GF and GR and from DSP Equation 2 for NMR Shift Data of
1-(para- or meta-Substituted phenyl)-6,7-dimethoxy-3,4-dihydroisoquinolinesa
F ( s
r
FF ( s
FR ( s
scale for σR
r
FF/FR
para-Substituted Series
+
C1
N2
C3
C4
C4a
C5
C6
C7
C8
C8a
-0.70 ( 0.25
10.7 ( 0.5
0.7008
0.9897
0.9773
0.9888
0.8602
0.9557
0.9834
0.9904
0.9638
0.9542
-2.20 ( 0.07
0.14 ( 0.04
11.6 ( 0.6
σR
0.9861
0.9943
0.9927
0.9965
0.9708
0.9930
0.9892
0.9926
0.9993
0.9955
-15.7
BA
9.5 ( 0.5
0.27 ( 0.02
-0.27 ( 0.01
-0.02 ( 0.02
0.34 ( 0.01
0.72 ( 0.04
0.37 ( 0.02
-1.16 ( 0.01
-1.22 ( 0.04
σR
0.82
0
0.33 ( 0.03
-0.26 ( 0.01
-0.19 ( 0.04
0.24 ( 0.03
0.61 ( 0.04
0.31 ( 0.02
-0.76 ( 0.07
-0.82 ( 0.09
0.53 ( 0.03
-0.14 ( 0.01
-0.17( 0.01
0.27 ( 0.02
0.29 ( 0.02
0.28 ( 0.02
-0.53 ( 0.01
-0.81 ( 0.05
σR+
0.51
1.93
0.12
1.26
2.48
1.32
2.19
1.51
σR+
σR
0
σR+
σR
BA
σR
BA
σR
0
σR
meta-Substituted Series
+
C1
N2
C3
C4
C4a
C5
C6
C7
C8
C8a
-2.90 ( 0.20
10.2 ( 1.1
0.9830
0.9617
0.9276
0.9751
0.7646
0.9254
0.9657
0.9637
0.9784
0.9835
-2.87 ( 0.10
8.7 ( 0.5
-0.47 ( 0.09
2.9 ( 0.5
σR+
0.9914
0.9781
0.9664
0.9923
0.9570
0.9907
0.9955
0.9972
0.9989
0.9981
5.94
3.00
2.11
6.25
0.64
1.38
2.23
1.94
3.67
2.20
σR+
σR+
σR-
σR-
σR-
σR-
σR+
σR
0.24 ( 0.04
-0.24 ( 0.02
0.18 ( 0.06
0.48 ( 0.07
0.83 ( 0.08
0.45 ( 0.05
-1.30 ( 0.10
-1.52 ( 0.11
0.19 ( 0.02
-0.25 ( 0.01
0.09 ( 0.02
0.33 ( 0.02
0.67 ( 0.02
0.35 ( 0.01
-1.21 ( 0.02
-1.41 ( 0.02
0.09 ( 0.02
-0.04 ( 0.01
0.14 ( 0.02
0.24 ( 0.02
0.30 ( 0.03
0.18 ( 0.01
-0.33 ( 0.02
-0.64 ( 0.04
0
σR
-
a s, standard deviation; r, correlation coefficient; σF values are from ref 22a except that for Br, which is from ref 21; σR+, σR0, σRBA, σR
are from ref 22b.
10672 J. Org. Chem., Vol. 70, No. 26, 2005

Propagation of Polar Substituent Effects
TABLE 2. Correlation Parameters G from Equation 1 and GF and GR and from DSP Equation 2 for NMR Shift Data of
1-(para- or meta-Substituted phenyl)-6,7-dimethoxy-1,2,3,4-tetrahydroisoquinolinesa
F ( s
r
FF ( s
FR ( s
scale for σR
r
FF/FF
para-Substituted Series
+
C1
N2
C3
C4
C4a
C5
C6
C7
C8
C8a
-0.10 ( 0.16
-1.88 ( 0.23
-0.20 ( 0.07
-0.21 ( 0.02
0.10 ( 0.02
0.25 ( 0.03
0.41 ( 0.04
0.23 ( 0.02
-0.31 ( 0.04
-1.66 (0.16
0.2147
0.9453
0.7026
0.9546
0.8466
0.9508
0.9677
0.9550
0.9256
0.9664
-1.17 ( 0.09
0.31 ( 0.05
-1.69 ( 0.18
-0.28 ( 0.07
-0.22 ( 0.02
0.20 ( 0.01
0.28 ( 0.02
0.45 ( 0.01
0.25 ( 0.02
-0.18 ( 0.02
-2.03 ( 0.08
σR
0.8928
0.9689
0.8857
0.9945
0.9969
0.9924
0.9995
0.9948
0.9953
0.9972
-1.67
0.55
1.43
1.50
0.77
1.27
1.25
1.30
2.17
1.04
BA
-1.68 ( 0.32
-0.26 ( 0.05
-0.31 ( 0.01
0.19 ( 0.01
0.34 ( 0.01
0.56 ( 0.01
0.32 ( 0.01
-0.53 ( 0.01
-2.12 ( 0.08
σR-
σR
0
σR-
σR
0
σR
0
σR
0
σR
BA
σR
0
σR
meta-Substituted Series
+
C1
N2
C3
C4
C4a
C5
C6
C7
C8
C8a
-1.26 ( 0.16
-1.97 ( 0.24
-0.47( 0.09
-0.35 ( 0.03
0.29 ( 0.07
0.52 ( 0.07
0.71 ( 0.06
0.41 ( 0.04
-0.51 ( 0.04
-2.47( 0.16
0.9465
0.9531
0.9029
0.9734
0.8581
0.9456
0.9741
0.9648
0.9837
0.9858
-1.27 ( 0.08
-1.70 ( 0.16
-0.43 ( 0.05
-0.35 (0.01
0.19 ( 0.01
0.39 ( 0.01
0.65 ( 0.01
0.37 ( 0.01
-0.47 ( 0.02
-2.37( 0.04
-0.20 (0.08
-0.40 ( 0.16
-0.05 (0.05
-0.05 (0.01
0.20 (0.01
0.23 ( 0.02
0.34 ( 0.02
0.22 ( 0.01
-0.10 (0.02
-0.88 (0.07
σR+
0.9675
0.9514
0.9070
0.9883
0.9969
0.9944
0.9979
0.9992
0.9881
0.9980
3.24
2.01
5.25
3.09
0.73
1.22
1.68
1.44
2.05
2.36
σR+
σR+
σR-
σR-
σR
0
σR
0
σR+
σR
σR
0
-
a s, standard deviation; r, correlation coefficient; σF values are from ref 22a except that for Br which is from ref 21; σR+, σR0, σRBA, σR
are from ref 22b.
SCHEME 3
reverse effect is observed. As regards the other carbons,
at C-4a there is practically no effect, and the effects at
C-3 and C-4 are also small, the effect at C-4 being reverse
in all cases while those at C-3 are normal for the DHIQ
sets but reverse for the THIQ sets. According to the
localized π-polarization model the extent of the effects
at C-6 and C-8a should be equal but of opposite signs.
However, the size of the FF and FR values in Tables 1 and
2 reveal (i) a significant unbalance between the magni-
tude of the F values for C-6 and C-8a for both the DHIQ
and THIQ sets and (ii) a significant difference in the
mutual magnitudes of the F values for C-8a and C-8 when
compared the DHIQ and THIQ sets. The numerical F
values at C-8a are about twice as large for the DHIQ sets
and about four times as large for the THIQ sets as the
corresponding values at their C-6 carbons, with the
exception of the meta THIQ series for which the numer-
ical value of FR at C-8a is only 2.6 times as large as FR at
C-6. For the DHIQ sets the F values at C-8 and C-8a are
close to each other, whereas for the THIQ sets the F
values at C-8a are 4-11 times higher than those at C-8.
The polarization pattern depicted with the aid of the
signs of the FF and FR values is closely similar with the
para and meta DHIQ sets (Tables 1 and 2; Scheme 3).
At C-1, C-8a, and C-8 the meta derivatives, however,
show slightly larger inductive effects. In that respect the
THIQ sets follow closely similar patterns. The surpris-
ingly large reverse inductive effects (-1.17 and -1.27 for
the para and meta sets, respectively) at the sp3 hybridized
C-1 of the THIQ sets are intriguing, as are also the
resonance effects at C-1 of the DHIQ (0.14 and -0.47,
respectively, for para and meta) and THIQ (0.31 and
-0.20, respectively, for para and meta) sets.
SCHEME 4
are given in Scheme 3. This kind of variation can mean,
for instance, different ability to bind to a medicinally
important target receptor for one thing for a nitro-
substituted derivative and for another for a methoxy-
substituted derivative because of different electrostatic
interactions. Polarization of the molecules was studied
with the aid of ab initio atomic charge calculations too.
The atomic charges are given Tables S11 and S12.
When comparing Schemes 3 and 4 it is seen that the
concept of the localized π-polarization (Scheme 4) quali-
tatively predicts the NMR shift behavior at C-1, C-5, C-6,
C-7, C-8, and C-8a for both the DHIQ and THIQ sets and
also that at N-2 for the DHIQ set: a marked reverse
effect is observed at C-1, C-8, and C-8a while a normal
effect is observed at C-5, C-6, C-7, and at N-2 of the DHIQ
sets. Surprisingly, at N-2 of the THIQ series a small
Discussion
Inductive Effects. The signs of the FF values observed
for the DHIQ sets (Table 1; Scheme 3) follow a pattern
predicted by the π-polarization concept for both π units,
J. Org. Chem, Vol. 70, No. 26, 2005 10673

Neuvonen et al.
SCHEME 5
SCHEME 6
that not only π-polarization but also polarization result-
ing from the electron-donating resonance of 7-OMe (12)
and that of 6-OMe group (13) contribute (cf. Scheme 5).
The resonance structures 12 and 13 are inductively
destabilized by the ED substituents. The decrease in their
contribution and the following decrease in electron
density at C-8 and C-8a, respectively, result in low-field
shift at these carbons when the substituent becomes more
electron-donating. Consequently, the reverse substituent
effects are observed (FF < 0). The value of FF at C-8 or
C-8a is higher than expected on the basis of the localized
π-polarization because the observed value is a sum of the
two effects. The lack of the substituent effect at C-4a and
a small effect only at C-5 suggest that substitution does
not significantly change the contribution of resonance
structures 10 and 11. Diminished effects have been
observed for quaternary aromatic carbons (ring junction)
such as C-4a with other condensed aromatic systems,
too.3a The values of FF for C-1, C-8, and C-8a are with
meta-substituted series somewhat larger than with para-
substituted derivatives. This obviously is connected to the
through space contribution in the propagation of the
substituent effects and to the shorter distance between
the meta-substituents and the probe site.
The trends of the normal/reverse substituent effects
are surprisingly similar with the DHIQ and THIQ sets
(Tables 1 and 2). One explanation for the reverse effect
at C-1 of the THIQ sets (FF < 0) is that ED substituents
can inductively stabilize the partial positive charge at
that carbon (15, Scheme 6) resulting in the high-
frequency 13C chemical shifts. C-1 is now less positively
charged than the CdN carbon and the magnitude of the
reverse inductive effects at C-1 for THIQ [FF ) -1.17
(para), -1.27 (meta)] is about half of that observed for
the sp2 hybridized C-1 carbon of DHIQ. Surprisingly,
however, a reverse effect is observed at N-2 (Table 2).
The FF (-1.17) and FR (0.31) values now observed for the
sp3 hybridized C-1 carbon of THIQ para series are close
to the FF(Y) (-1.12) and FR(Y) (0.75) values recently
observed by Szatma´ri et al. for the sp3 hybridized C-1
carbon of the trans isomer of 1-(p-Y-phenyl)-3-(p-X-
phenyl)-2,3-dihydro-1H-naphth[1,2-e][1,3]oxazines and
not far from those observed for the cis isomer [FF(Y) )
-1.51, FR(Y) ) 0.60].16 When the same set of substituent
the CdN bond, and the phenyl ring (cf. 1, 2, Scheme
4).1,4c,d C-6, C-8, and C-8a exhibit marked effects, but
C-4a, C-5, and C-7 show much smaller effects. The
magnitude of the FF values at C-6 is 0.67-0.72. Much
larger effects are observed at C-8 and C-8a (from -1.16
to -1.41). The former values can be considered as normal
for a polarized phenyl unit and are in accordance with
π-polarization concept. To the contrary, the marked
asymmetry in the FF values at C-6 and C-8a is not
explainable by the direct π-polarization concept only. The
extended polarization model used by Reynolds et al.2 does
not explain the increased sensitivity to substitution at
C-8 and C-8a.
The degree of the (partial) charge development at a
certain atom depends on the balance of different reso-
nance structures. Electron-donating resonance from the
methoxy group (Scheme 5) can be effective at ortho/para
positions in respect to a methoxy substitution, and
negative charge can be induced at C-4a, C-5, C-8, or C-8a
(cf. 10-13). The electron donation from the neighboring
oxygen atom can be concluded to diminish the signifi-
cance of those resonance structures where negative
charge is situated at C-6 or C-7 (structures not shown).
With the DHIQ sets, C-1 carbon exhibits the largest
inductive effect, which is reverse for both para and meta
series. To the contrary, N-2 exhibits an extensive normal
effect. An ED substituent X inductively stabilizes the
resonance structure 9. According to the resonance polar-
ization concept, the contribution of 9 exhibiting a positive
charge at the C-1 carbon and a negative charge at N-2 is
increased by ED substituents resulting in diminished
electron density and a low-field 13C NMR chemical shift
at C-1, an RSE as the consequence (FF < 0).4c,d To the
contrary, a normal substituent effect (FF > 0) is observed
at N-2 (Table 1). With the EW substituents the situation
is opposite. This behavior of the chemical shifts at C-1
and N-2 is in accordance with the π-polarization concept,
too. When the phenyl π unit is considered, the opposite
behavior of C-8a (FF < 0) and C-6 (FF > 0) is that predicted
by the π-polarization. Further, the positive and closely
similar FF values for C-5 and C-7, which are smaller than
that for C-6, agree with that model. The reverse inductive
effects observed at C-8 and C-8a are, however, exception-
ally large, the effect at C-8a being about two times as
large as that at C-6. This is not in accordance with the
localized π-polarization concept and is not explained by
the general extended polarization model either.2 The
enhanced sensitivity at C-8 and at C-8a reflects the fact
(15) Jennings, W. B.; Wilson, V. E.; Boyd, D. R.; Coulter, P. B. Org.
Magn. Reson. 1983, 21, 279-286.
(16) Szatma´ri, I.; Martinek, T.; La´za´r, L.; Koch, A.; Kleinpeter, E.;
Neuvonen, K.; Fu¨lo¨p, F. J. Org. Chem. 2004, 69, 3645-3653.
10674 J. Org. Chem., Vol. 70, No. 26, 2005

Propagation of Polar Substituent Effects
SCHEME 7
parameters (σF, σR) is used for eq 2 in our case as was
used by Szatma´ri et al., our F values at C-1 (FR ) -1.22
and FR ) 0.73; Table S6) are in the limits of error quite
the same as those obtained for the trans isomer of the
naphthoxazine derivatives.16 Another analogous case of
an RSE of a saturated ring carbon has been observed.4c
We conclude that part of the substituent dependence of
the chemical shift of the C-1 carbon and N-2 nitrogen
can be attributed to the polar character of the CH-NH
bond (15) and part to the stereoelectronic effects (16,
Scheme 6).16 Resonance structures 16a and 16b are
stabilized by the ED substituents. As compared with the
structure 15 hybridization of both C-1 and that of N-2
are changed from sp3 to sp2, resulting in more high-
frequency (low-field) 13C/15N NMR chemical shifts. Con-
sequently, an RSE is observed at C-1 and N-2 when the
contribution of the structures 16 is increased. The effects
of the two concurrent interaction mechanisms on the
chemical shift, the polarization of the CH-NH bond and
the stereolectronic effect, are parallel at C-1 whereas they
are opposite at N-2. The reverse effect at C-1 of the THIQ
sets is about half of that observed at C-1 of the DHIQ
sets, while the extent of the reverse effect at N-2 of the
THIQ series is only ca. 20% of the normal effect observed
at N-2 of the DHIQ series. The reverse effect at N-2
suggests that the stereoelectronic effect predominates.
C-8a exhibits the largest reverse inductive effect [FF
) -2.12 (para), -2.37 (meta)] observed for the THIQ sets.
The effects are markedly larger than those observed at
C-8a for the DHIQ sets [FF ) -1.22 (para), -1.41 (meta)]
and their numerical values are 3-4 times as large as the
FF values at C-6 of the DHIQ or THIQ sets (from 0.56 to
0.72; Tables 1 and 2). This fact can be explained by
contribution of resonance structure 17 (Scheme 6; the
analogous structure can be drawn for the meta series).
ED substituents inductively destabilize it, with a de-
crease of the electron density at C-8a and low-field
chemical shifts as a consequence. The shorter distance
between the substituent and the probe site again explains
the higher effect at the meta series. With the DHIQ sets
the contribution of the analogous resonance structures
(cf. 13, Scheme 5) is contradicted by the appearance of
the resonance structure 14 possessing an elongated
conjugative system. In other words, in the DHIQ deriva-
tives the negative charge is more delocalized than in the
THIQ compounds, and this explains the smaller effects
at C-8a. Interestingly, at C-8 the DHIQ set exhibits
stronger inductive effects [-1.16 (para), -1.21 (meta)]
than the THIQ set [-0.53 (para), -0.47 (meta)]. Obvi-
ously the localization of significant partial negative
charge at C-8a for THIQ diminishes the probability of
the negative charge at C-8. In other words, the relative
contribution of resonance structure 18 is smaller than
that of 12 (Scheme 5).
induced polar effect (9 T 20) resulting in positive and
negative FR values, respectively (Scheme 7).4,12 The former
interaction where electron donation by conjugation in-
creases electron density at the imine carbon is possible
only for the para-substituted derivatives. When the
resonance induced polar effect is effective, contribution
of resonance structure 20 is significant (for ED substit-
uents) and electron density at the probe carbon is
decreased, resulting in low-field 13C NMR chemical shifts
at C-1 (FR < 0). With EW substituents the situation is
opposite. With the para series both the direct resonance
mechanism (FR > 0) (19) and the resonance induced polar
effect (FR < 0) (20) can occur concurrently. In the present
case the former one dominates resulting in FR > 0 at C-1.
For the meta derivatives of the DHIQ set the direct
resonance effect is excluded and the reverse effect (FR <
0) observed at C-1 can nicely be explained by the
resonance induced polar effect (21) alone. For the para
derivatives the influence of the resonance induced polar
mechanism can be more prominent than for the meta
derivatives because the negative charge induced by ED
substituents is situated at the ipso carbon adjacent to
the CdN bond. At N-2 a positive FR value is observed for
both the para and meta DHIQ series, the effect being for
the para series four times as large as that for the meta
series, indicating the appearance of the direct conjugation
(19, Scheme 7) for the para series.
For C-8 and C-8a a reverse resonance effect (FR < 0) is
observed for both DHIQ sets, and their values are close
to each others [at C-8 -0.53 (para), -0.33 (meta) and at
C-8a -0.81 (para), -0.64 (meta)]. These observations can
be rationalized by inspecting the contribution of reso-
nance structures such as 22 [shown for C-8a (para),
Scheme 7]. By this resonance induced polar mechanism
the EW substituents stabilize the induced negative
charge at C-8 and C-8a (cf. 12 and 13 in Scheme 5)
resulting in the upfield (low-frequency) shift values and
a reverse substituent effect (FR < 0).
The normal resonance effect at C-1 (FR ) 0.31) observed
for the para THIQ set suggests that conjugation between
polarized CH-NH bond and the substituted phenyl ring
is possible as a result of the stereoelectronic effects, 23
(Scheme 8). As compared to structure 15, in structures
23 electron density is transferred from the phenyl sub-
stitutent to C-1, resulting in a normal substituent effect.
Competition of the normal (23) and reverse (24) reso-
nance effects occurs as with the para DHIQ set, but the
transmission mechanism of the direct resonance is dif-
ferent. Increased contribution of resonance structures 23
(by ED substituents) decreases the partial negative
charge at N-2 (15). The small FR (<0) at N-2 possibly
reflects that character. A small reverse resonance effect
for C-1, analogously with meta-substituted dihydroiso-
quinolines, is observed with the meta derivatives due to
the resonance induced polar effect (25).
Resonance Effects. The positive FR values are ob-
served for the para DHIQ and THIQ sets (0.14, 0.31,
respectively), but the negative ones are seen for the
corresponding meta sets (-0.47, -0.20, respectively) at
C-1. As compared with the inductive effects prevailing
at the same sites the FR values are clearly smaller. The
interpretation of the resonance effect is more complicated
than that of the inductive effect.4a-d,12 Generally, two
different transmission mechanisms have been consid-
ered: the (direct) resonance 9 T 19 and the resonance
J. Org. Chem, Vol. 70, No. 26, 2005 10675

Neuvonen et al.
SCHEME 8
SCHEME 9
For the THIQ sets negative FR values observed at both
C-8a and C-8 suggest contribution of resonance struc-
tures such as 26 and 27 (Scheme 8; shown for C-8a, para
and meta) analogously with the DHIQ sets. There is,
however, an interesting difference. With the DHIQ sets
the observed resonance F values at C-8 are close to those
at C-8a, but with the THIQ sets the behavior of C-8 and
C-8a differ considerably from each other. The resonance
effect at C-8a is markedly higher for the para THIQ (FR
) -2.03) than for the meta (FR ) -0.88). In addition, the
resonance effects at C-8 are small [FR ) -0.18 (para),
-0.10 (meta)]. So, the resonance effects verify the conclu-
sion drawn on the basis of the inductive effects: with the
THIQ sets the partial negative charge seems to be
localized into C-8a because of the absence of the conjuga-
tive effects with the CdN unit, which are possible for
the DHIQ set (cf. 14). The higher negative charge devel-
opment at C-8a for THIQ as compared with DHIQ is a
clear difference between these two sets. The observed re-
verse resonance effects with THIQ reflect the contribu-
tion of the resonance induced polar effect only. These FR
values [meta (-0.88) and para (-2.03)] allow the evalu-
ation of their relative strengths, giving the value of 2.3
for the para/meta effect, which reflects the shorter dis-
tance between the probe site and X with the para series.
Atomic Charges. In many cases good correlations
between the NMR chemical shifts and the computational
atomic charges have been observed.4c,d,5b,6-10 Therefore we
also probed the ability of the Mulliken atomic charges to
predict the observed effects of substituents at the iso-
quinoline moiety by calculating on the ab initio level the
atomic charges for the DHIQ sets (Tables S11 and S12).
A good correlation with a slope of -1 for qN(N-2) vs
qC(C-1) of the para DHIQ set (slope ) -1.06 ( 0.10, r )
0.9670) is in accordance with the substituent-sensitive
polarization of the CdN bond. The atomic charges of the
DHIQ sets were correlated with Hammett substituent
constants σ (Table S1; cf. entries 4-6). Satisfactory
correlations were observed in most cases for the para
series, and the trends (normal/reverse) of the substituent
effects are similar with those observed with the NMR
chemical shifts (Scheme 9) with exception of C-3, C-8,
and C-8a. This similarity supports the idea that the
substituent effects both on the atomic charges and on the
chemical shifts reflect electronic substituent effects.
Satisfactory correlations in accordance with the chemical
shift behavior (Scheme 9) were observed also for the meta
series for q vs σ correlations, with exception of C-1, C-3,
C-8, and C-8a. For the meta DHIQ set the situation is
more complicated. According to the calculations the meta-
DHIQ derivatives can exhibit two conformations due to
the orientation of the 1-phenyl group [syn (specified by
us as 1 or in-conformation) or anti (specified by us as 2 or
out-conformation) in respect to the benzo moiety; see Pic-
tures S1 and S2], the energies of which are relatively
close to each other. Because the results obtained with
the minimum energy structures were in some cases un-
satisfactory, calculations were extended also to the sets
where the orientation is not varied with C-1, N-2, and
C-8. It was interesting to see that improvement was ob-
served. With both N-2 and C-8 the charge correlations
gave satisfactory results in the case of orientation 2 (out).
This fact shows that at least partly the unsatisfactory
results observed (δ vs q) are connected to appearance in
solution of a different conformational equilibrium as com-
pared with the gas phase. The fact that the slopes for
the q vs σ correlations (Table S1; entries 4-6) in most
case are higher for the meta series than for the para series
is in accordance with the fact that substituent effects on
the NMR chemical shift are stronger in the meta series
than in the para series (cf. Table 1; especially FF values).
Correlations between the NMR chemical shifts and the
Mulliken atomic charges (Table S1; entries 2 and 3) are
in most cases satisfactory, although in some cases they
fail. The poor correlation for C-1 and the negative sign
for the correlation of C-8a are intriguing. To investigate
in more detail the origin of this fact, we performed a dual
+
substituent parameter (DSP) analysis (with σF and σR
,
σR σRBA, or σR-) also for the atomic charge data. The
0
,
results are interesting. In most cases from excellent to
satisfactory correlations are observed, especially for the
para series (Tables S1, entries 7-10). This verifies the
electronic substituent effects on the atomic charges. The
varying signs of the F values support the idea of the
substituent-sensitive polarization of the molecule. The
discussion is given in Supporting Information.
10676 J. Org. Chem., Vol. 70, No. 26, 2005

Propagation of Polar Substituent Effects
(1H decoupled)/0.46 Hz/point (1H coupled), pulse width 4.35
µs (45°), acquisition time 1.09 s (1H decoupled)/2.18 s (1H
coupled), number of transients 1000-12000, pulse delay 3 s
(1H decoupled)/5 s (1H coupled), pulse sequence (JEOL)
SGBCM (1H decoupled)/SGNOE (1H coupled). Exponential
windowing with a line-broadening term of 2 Hz (1H de-
coupled)/1 Hz (1H coupled) was applied prior to Fourier
transformation. 15N NMR spectra were acquired with 2D 1H-
15N heteronuclear FG HMBC correlation experiments. 15N
Conclusions
The effect of the phenyl substituent X extends all over
the isoquinoline moiety. The negative charge develops at
C-8 and C-8a as a result of electron-donating resonance
of the methoxy substituents, and a positive one develops
at C-1 as a result of contribution of resonance structures
9 (Scheme 5) and 15 (Scheme 6), resulting to a strongly
substituent-sensitive polarization. Negative charge de-
localization via conjugation for dihydro but not for
tetrahydro derivatives causes a characteristic difference
in behavior of the dipolar character at C-8/C-8a. For
DHIQs the negative charge is situated evenly at C-8 and
C-8a, whereas for THIQs it is concentrated at C-8a. With
the THIQ set, delocalization of the nitrogen lone pair
(stereoelectronic effect) contributes to the observed sub-
stituent effects at C-1. The appearance of the reverse/
normal substituent effects at different π-units can be
explained by either the π-polarization mechanism or by
the resonance polarization model. Even if the π-polariza-
tion theory predicts the trends of the 13C NMR chemical
shift behavior, it explains neither the appearance of the
unbalanced FF values observed at C-6 and C-8a nor the
characteristic differences observed between the DHIQ
and THIQ sets. Polarization induced by a substituted
phenyl group is until now discussed as a combination of
two different substituent-sensitive effects: the localized
π-polarization and the extended π-polarization. We sug-
gest the resonance polarization concept for dealing with
this phenomenon. This approach is convenient in predic-
tion of the enhanced sensitivity of measurable properties,
such as 13C NMR shifts, to substitution. The resonance
polarization concept can successfully be used to explain
the localized π-polarization of the CdN and CdO units.
As a summary, instead of two overlapping and sometimes
tedious concepts, in many cases a single procedure can
be used to explain propagation of polar substituent
effects. Illustrative examples of the availability of the
resonance polarization concept are given in Supporting
Information.1,2,4d,17
1
NMR spectra utilized a JNH coupling of 95 Hz and were
optimized for a long-range nJNH coupling (n ) 2 or 3) of 8 Hz.
FG HMBC experiments were acquired in magnitude mode
with spectral widths appropriately optimized from the 1D
spectra and processed with zero-filling (×2, ×4), a 2Π/3-shifted
sinebell function, and exponential weighting applied in both
dimensions prior to Fourier transformation.
Theoretical Calculations. The ab initio program package
Gaussian 0318 was used for all calculations. The calculations
were carried out at the Hartree-Fock level by means of 6-31G*
split-valence basis set.19 The geometry optimizations were
performed without restrictions.
The Mulliken atomic charges were calculated. The Tripos
modeling software SYBYL 7.020 was used for preparation of
input data and visualization of results. All quantum-chemical
calculations were processed on SGI Octane and a Linux cluster
at Potsdam University.
Statistical Calculations. The sources of the substituent
constants used are as follows: σ, σ+, ref 21; σF, σR, ref 22a,
except σF and σR for Br, ref 21; σR+, σR0, σRBA, and σR- are from
0
ref 22b. Also the suitability of an alternative set of σF and σR
(from ref 23) was tested. The statistical correlations were
calculated with Fig.P Version 2.98 or FigSys Version 2.4.3
(BIOSOFT, Cambridge).
General Method for the Synthesis. The synthesis of
1-aryl-6,7-dimethoxy-3,4-dihydro- and 1,2,3,4-tetrahydroiso-
quinolines were performed by standard procedures previously
reported.24 First homoveratrylamine was reacted with substi-
tuted benzoyl chlorides by the Schotten-Baumann procedure,
followed by Bischler-Napieralski cyclization by POCl3 result-
(18) Frisch, M. J.; Trucks, G. W.; Schlegel, H. B.; Scuseria, G. E.;
Robb, M. A.; Cheeseman, J. R.; Montgomery, J. A., Jr.; Vreven, T.;
Kudin, K. N.; Burant, J. C.; Millam, J. M.; Iyengar, S. S.; Tomasi, J.;
Barone, V.; Mennucci, B.; Cossi, M.; Scalmani, G.; Rega, N.; Petersson,
G. A.; Nakatsuji, H.; Hada, M.; Ehara, M.; Toyota, K.; Fukuda, R.;
Hasegawa, J.; Ishida, M.; Nakajima, T.; Honda, Y.; Kitao, O.; Nakai,
H.; Klene, M.; Li, X.; Knox, J. E.; Hratchian, H. P.; Cross, J. B.; Adamo,
C.; Jaramillo, J.; Gomperts, R.; Stratmann, R. E.; Yazyev, O.; Austin,
A. J.; Cammi, R.; Pomelli, C.; Ochterski, J. W.; Ayala, P. Y.; Morokuma,
K.; Voth, G. A.; Salvador, P.; Dannenberg, J. J.; Zakrzewski, V. G.;
Dapprich, S.; Daniels, A. D.; Strain, M. C.; Farkas, O.; Malick, D. K.;
Rabuck, A. D.; Raghavachari, K.; Foresman, J. B.; Ortiz, J. V.; Cui,
Q.; Baboul, A. G.; Clifford, S.; Cioslowski, J.; Stefanov, B. B.; Liu, G.;
Liashenko, A.; Piskorz, P.; Komaromi, I.; Martin, R. L.; Fox, D. J.;
Keith, T.; Al-Laham, M. A.; Peng, C. Y.; Nanayakkara, A.; Challa-
combe, M.; Gill, P. M. W.; Johnson, B.; Chen, W.; Wong, M. W.;
Gonzalez, C.; Pople, J. A. Gaussian 03, Revision C.02; Gaussian, Inc.:
Wallingford CT, 2004.
The Mulliken atomic charges estimate the substituent-
sensitive polarization of the dihydroisoquinoline moiety.
Linear correlations between the NMR chemical shifts and
the atomic charges are, however, not obtained for all
atoms. The substituent sensitivities of the atomic charges
(gas phase) show different blends of the inductive and
resonance effects than the substituent sensitivities of the
NMR chemical shifts (in solvent).
Experimental Section
(19) Hehre, W. J.; Radom, L.; Schleyer, P.v. R.; Pople, J. A. Ab Initio
Molecular Orbital Theory; Wiley: New York, 1986.
(20) SYBYL 7.0; Tripos Inc.: 1699 South Hanley Road, St. Louis,
MO 63144, 2004.
NMR Measurements. NMR spectra were recorded at 25
°C on an NMR spectrometer operating at 125.78 MHz for 13
C
(21) Hansch, C.; Leo, A.; Taft, R. W. Chem. Rev. 1991, 91, 165-
195.
and 50.69 MHz for 15N on 0.1 M solutions in CDCl3. 13C spectra
were referenced internally to tetramethylsilane (0.00 ppm),
and 15N spectra were referenced externally to CH3NO2 (0.00
ppm) containing 10% w/w CD3NO2 for locking purposes. The
signal of the deuterium of the solvent was used as a lock signal
for 13C spectra. 13C NMR spectra were acquired with 1H
broadband decoupling and NOE 1H nondecoupling techniques.
13C Spectra were acquired with the following conditions:
spectral width of 30 kHz, 32 K data points (1H decoupled)/64
K data points (1H coupled), digital resolution 0.92 Hz/point
(22) (a) Taft, R. W.; Topsom, R. D. Prog. Phys. Org. Chem. 1987,
16, 1-84. (b) Ehrenson, S.; Brownlee, R. T. C.; Taft, R. W. Prog. Phys.
Org. Chem. 1973, 10, 1-80.
(23) Reynolds, W. F.; Gomes, A.; Maron, A.; MacIntyre, D. W.;
Tanon, A. Can. J. Chem. 1983, 61, 2376-2384.
(24) (a) Minor, D. L.; Wyrick, S. D.; Charifson, P. S.; Watts, V. J.;
Nichols, D. E.; Mailman, R. B. J. Med. Chem. 1994, 37, 43174328. (b)
Georgiev, V. St.; Carlson, R. P.; Van Inwegen, R. G.; Khandwala, A.
J. Med. Chem. 1979, 22, 348-352. (c) Corte´s, E. C.; Romero, E. C.;
Ram´ırez, F. G. J. Heterocycl. Chem. 1994, 31, 1425-1427.
(25) (a) Viel, C.; Dorme, R.; Rumpf, P. Bull. Soc. Chim. Fr. 1966,
1956-1966. (b) Barker, J. M.; Huddleston, P. R.; Clephane, J.; Wood,
M. L.; Holmes, D. J. Chem. Soc., Perkin Trans. 1 1985, 275-281. (c)
Bishop, D. C.; Tucker, M. J. J. Chem. Soc. C 1970, 2184-2186.
(17) Corne´lis, A.; Lambert, S.; Laszlo, P.; Schaus, P. J. Org. Chem.
1981, 46, 2130-2134.
J. Org. Chem, Vol. 70, No. 26, 2005 10677

Neuvonen et al.
ing in the 3,4-dihydroisoquinolines, which were later reduced
to 1,2,3,4-tetrahydroisoquinolines. Melting points are uncor-
rected. All new compounds (see Tables S9 and S10) gave
satisfactory microanalyses (C, H, N).25,26
Correlation parameters obtained from eq 2 using alternative
sets of substituent parameters (σF and σR or σF and σR0).
Discussion concerning the DSP analysis of atomic charges.
Examples of the usefulness of the resonance polarization
concept. Energies, atomic charges, and Cartesian coordinates
for the optimized structures of the DHIQ sets. Pictures S1 and
S2 describing the X-in and X-out conformations for the meta
DHIQ sets, respectively. Analytical data for the prepared
compounds. This material is available free of charge via the
Internet at http://pubs.acs.org.
Supporting Information Available: Statistical data for
the correlations of the atomic charges of the DHIQ sets. 13C
and 15N NMR chemical shifts for the DHIQ and THIQ sets.
(26) (a) Letcher, R. M.; Sammes, M. P. J. Chem. Educ. 1985, 62,
262-4. (b) Sarges, R. J. Heterocycl. Chem. 1974, 11, 599-601. (c) Singh,
H.; Sarin. R. J. Chem. Res., Miniprint 1988, 2623. (d) Brzezinska, E.
Acta Pol. Pharm. 1966, 53, 365-371.
JO0507946
10678 J. Org. Chem., Vol. 70, No. 26, 2005