Further studies of the thermal and photochemical Diels-Alder reactions of N-methyl-1,2,4-triazoline-3,5-dione (MeTAD) with naphthalene and some substituted naphthalenes

Article abstract of DOI:10.1021/jo9906429

MeTAD thermally reacted with naphthalene (2) and methylated naphthalenes to give equilibrium mixtures of starting materials and [4 + 2] cycloadducts. Methyl substitution on the naphthalene ring generally increased both the amount of cycloadduct formed and the rate of cycloaddition relative to 2. The isolated cycloadducts were all thermally labile and quantitatively reverted to the parent naphthalene in the presence of 2,3-dimethyl-2-butene as a trap for liberated MeTAD. The rates of the cycloreversion reactions were affected by substitution patterns but not appreciably by solvent. A mechanism for the cycloaddition reaction is presented that proposes the involvement of a charge-transfer complex. Photochemically, MeTAD demonstrated lower regioselectivity in its reactions with substituted naphthalenes relative to the corresponding thermal reactions.

Full text of DOI:10.1021/jo9906429

J . Org. Chem. 2000, 65, 2863-2869
2863
F u r th er Stu d ies of th e Th er m a l a n d P h otoch em ica l Diels-Ald er
Rea ction s of N-Meth yl-1,2,4-tr ia zolin e-3,5-d ion e (MeTAD) w ith
Na p h th a len e a n d Som e Su bstitu ted Na p h th a len es
Gary W. Breton* and Kristy A. Newton
Berry College, Department of Chemistry, P.O. Box 495016, Mount Berry, Georgia 30149-5016
Received April 19, 1999
MeTAD thermally reacted with naphthalene (2) and methylated naphthalenes to give equilibrium
mixtures of starting materials and [4 + 2] cycloadducts. Methyl substitution on the naphthalene
ring generally increased both the amount of cycloadduct formed and the rate of cycloaddition relative
to 2. The isolated cycloadducts were all thermally labile and quantitatively reverted to the parent
naphthalene in the presence of 2,3-dimethyl-2-butene as a trap for liberated MeTAD. The rates of
the cycloreversion reactions were affected by substitution patterns but not appreciably by solvent.
A mechanism for the cycloaddition reaction is presented that proposes the involvement of a charge-
transfer complex. Photochemically, MeTAD demonstrated lower regioselectivity in its reactions
with substituted naphthalenes relative to the corresponding thermal reactions.
Several years ago Sheridan described the photochemi-
cal reaction of MeTAD (1) with a number of aromatic
compounds including benzene and naphthalene (2).1,2 The
reactions with benzene and 2 afforded novel Diels-Alder
cycloadducts (eq 1) that have served as precursors to
several theoretically interesting molecules.2-4
workers reported that equilibria were established be-
tween starting materials and Diels-Alder cycloadducts
when naphthalene and substituted naphthalenes were
heated (typically 100 °C for 24 h) in the presence of excess
maleic anhydride.6 Cycloadducts were generally formed
in poor yields with simple mono- and disubstituted
naphthalenes (1-47%),6a,b but higher conversions were
reported at high pressures (∼9000 atm) or with poly-
alkylated naphthalenes.6c,d The cycloadducts reverted to
starting materials upon heating. Singlet oxygen does not
react with 2, but it does react readily with naphthalene
compounds that are at least disubstituted with alkyl
groups (e.g., 1,4- or 1,8-dimethylnaphthalene).7
While the photochemical reactions of MeTAD with 2
and some substituted naphthalenes have already been
reported,1,3,4 the corresponding thermal reactions of tri-
azolinediones with naphthalenes have only been briefly
mentioned in the literature.4 Herein we provide a more
detailed report on the thermal and photochemical reac-
tions of MeTAD with 2 and a series of methylated
naphthalenes (compounds 3-9 in Chart 1). We also
report on our studies of the thermal cycloreversions of
the [4 + 2] adducts.
While more reactive in a Diels-Alder sense than
benzene, naphthalenes are generally poor diene compo-
nents in the classical Diels-Alder cycloaddition reaction.5
Cycloaddition has been reported, therefore, only with
strong dienophiles such as maleic anhydride and singlet
oxygen in addition to MeTAD (eq 2).6,7 Herzog and co-
Resu lts
1. Th er m a l Rea ction s of MeTAD w ith Na p h th a -
len e a n d Su bstitu ted Na p h th a len es. A. Rea ction
w ith Na p h th a len e (2). When MeTAD is mixed with 2
in CDCl3 (initially 0.40 M each) at 29 °C in the absence
(5) Biermann, D.; Schmidt, W. J . Am. Chem. Soc. 1980, 102, 3163.
(6) (a) Kloetzel, M. C.; Herzog, H. L. J . Am. Chem. Soc. 1950, 72,
1991. (b) Kloetzel, M. C.; Dayton, R. P.; Herzog, H. L. J . Am. Chem.
Soc. 1950, 72, 273. (c) Plieninger, H.; Wild, D.; Westphal, J . Tetrahe-
dron 1969, 25, 5561. (d) Oku, A.; Ohnishi, Y.; Mashio, F. J . Org. Chem.
1972, 37, 4264.
(7) (a) Wasserman, H. H.; Larsen, D. L. J . Chem. Soc., Chem.
Commun. 1972, 253. (b) Adam, W.; Prein, M. Acc. Chem. Res. 1996,
29, 275.
* To whom correspondence should be addressed. e-mail: gbreton@
berry.edu. office: (706)-290-2661. fax: (706)-238-7855.
(1) (a) Kjell, D. P.; Sheridan, R. S. J . Am. Chem. Soc. 1984, 106,
5368. (b) Kjell, D. P.; Sheridan, R. S. J . Photochem. 1985, 28, 205.
(2) Hamrock, S. J .; Sheridan, R. S. J . Am. Chem. Soc. 1989, 111,
9247.
(3) Kjell, D. P.; Sheridan, R. S. J . Am. Chem. Soc. 1986, 108, 4111.
(4) Burger, U.; Mentha, Y.; Thorel, P. J . Helv. Chim. Acta 1986, 69,
670.
(8) For comparison, Burger et al. (ref 4) reported a value of Keq )
0.4 for the reaction of N-phenyl-1,2,4-triazoline-3,5-dione with naph-
thalene at 25 °C in CDCl3.
10.1021/jo9906429 CCC: $19.00 © 2000 American Chemical Society
Published on Web 04/27/2000

2864 J . Org. Chem., Vol. 65, No. 10, 2000
Breton and Newton
Ch a r t 1
Ta ble 1. Rea ction of MeTAD w ith Sym m etr ica lly
Su bstitu ted Na p h th a len esa
b
F igu r e 1. van’t Hoff plot for the reaction of MeTAD with
naphthalene (2).
cmpd
solvent
Keq
yield of cycloadduct, %c
2
CDCl3
2.0
34 (21)
C6D6
0.3
0.1
0.1
10
4
4
40
Ta ble 2. Th er m a l Rea ction of MeTAD w ith
Un sym m etr ica lly Su bstitu ted Na p h th a len esa
CD3CN
(CD3)2CO
CDCl3
yield of cycloadduct, %c
3
4
2.8
(CD3)2CO
CDCl3
(CD3)2CO
CDCl3
0.1
17.7
0.2
3
b
naphthalene
solvent
Keq
CA
CB
69 (63)
7
89 (69)
6
CDCl3
(CD3)2CO
CDCl3
(CD3)2CO
CDCl3
(CD3)2CO
CDCl3
2.1
0.1
4.5
0.1
12.8
0.2
10
d
6
d
58 (39)
8
86 (60)
21
29
5
43 (21)
4
6
d
d
d
5
351
7
8
9
a
Conducted at 29 °C with initial concentrations of naphthalene
b
and MeTAD at 0.40 M each. Determined by relative integrations
of appropriate signals in the 1H NMR spectrum. Values are reliable
to within (5% c Determined by 1H NMR spectroscopy. Values in
parentheses are isolated yields.
110
0.8
(CD3)2CO
a
Conducted at 29 °C with initial concentrations of naphthalene
of light, an equilibrium is established within 18 h in
which a 34% yield (Keq ) 2.0) of cycloadduct 2C (where
C denotes a [4 + 2] cycloadduct with MeTAD) is formed
as the only product (eq 3 and Table 1).8 No change in
b
compound and MeTAD at 0.40 M each. Determined by relative
integrations of appropriate signals in the 1H NMR spectrum.
Values are reliable to within (5%. c Determined by 1H NMR
d
spectroscopy. Values in parentheses are isolated yields. No
cycloadduct observed in the 1H NMR spectrum.
for attainment of the maximum yield of cycloadduct to
ensure that an equilibrium had been established. Reac-
tion of MeTAD with 1,5-dimethylnaphthalene (3) afforded
an equilibrium mixture of starting materials and cy-
cloadduct 3C with Keq ) 2.8 within 10 min. Reaction with
2,6-dimethylnaphthalene (4) required approximately 8 h
to attain equilibrium although the exent of the equilib-
rium in favor of cycloadduct was much greater than that
of compound 3. When the reactions of 3 and 4 were
conducted in acetone-d6 as solvent, lesser equilibrium
amounts of cycloadduct were observed. Reaction of
MeTAD with 2,3,6,7-tetramethylnaphthalene (5) in CDCl3
afforded the greatest yield of cycloadduct (89%) observed
with all of the naphthalenes investigated requiring
approximately 5 h of reaction time to reach equilibrium.
Cycloadducts 4C and 5C were isolated by quenching
excess MeTAD in the reaction mixture with 2,3-dimethyl-
2-butene and isolating the product via preparative thin-
layer chromatography. The adducts were thermally labile
and reverted quantitatively back to starting naphthalene
in the presence of 2,3-dimethyl-2-butene as a trap for the
liberated MeTAD. Adduct 3C was especially labile,
however, and resisted all of our efforts at its isolation
employing preparative TLC as well as HPLC (see below).
Rate studies on its cycloreversion indicated a half-life of
approximately 9 min at 29 °C. A more detailed discussion
of the cycloreversion process of the adducts is presented
below.
the equilibrium ratio was observed even after an ad-
ditional 7 days of reaction time. The equilibrium constant
for the reaction was determined at a series of tempera-
tures over the range 26-50 °C. A van’t Hoff plot (Figure
1) afforded ∆H° ) -44 kJ /mol and ∆S° ) -141 J /mol.
The identity of the reaction solvent had a strong impact
on the extent of the reaction (Table 1). Reaction was
observed to readily take place in nonpolar (e.g., benzene)
and weakly polar (e.g., CDCl3) solvents, but strongly polar
solvents (e.g., acetone) discouraged adduct formation.
Adduct 2C quantitatively reverted to starting material
(as a solution in CDCl3) in the presence of 2,3-dimethyl-
2-butene as a trap for liberated MeTAD.9
B. Rea ction of MeTAD w ith Sym m etr ica lly Su b-
st it u t ed Met h yla t ed Na p h t h a len es. Reactions of
MeTAD with symmetrically substituted naphthalenes
3-5 to afford cycloadducts 3C-5C were conducted in the
dark at 29 °C in CDCl3, and were followed by periodic
1
monitoring of the reaction using H NMR spectroscopy.
The results are summarized in Table 1. The reactions
were monitored at least 24-48 h past the time required
(9) The alkene 2,3-dimethyl-2-butene reacts rapidly and quantita-
tively with MeTAD in an ene reaction to afford a relatively inert
urazole product. See: Ohashi, S.; Leong, K.-W.; Matyjaszewski, K.;
Butler, G. J . Org. Chem. 1980, 45, 3467.
C. Rea ction of MeTAD w ith Un sym m etr ica lly
Su bstitu ted Meth yla ted Na p h th a len es. Unsymmetri-

Diels-Alder Reactions of MeTAD
J . Org. Chem., Vol. 65, No. 10, 2000 2865
cally substituted naphthalenes (6-9) present two non-
equivalent aromatic rings as possible sites of attack by
MeTAD. For the purposes of discussion and comparison,
cycloadducts CA and CB will be defined as the [4 +
2]-cycloadducts resulting from reaction at the substituted
ring (1,4-addition) and the unsubstituted ring (5,8-
addition), respectively (eq 4). The results of these reac-
cycloadducts CA and CB were not separable by analytical
or preparative TLC using conventional SiO2 plates.
Although some separation was possible utilizing analyti-
cal SiO2 plates that had been impregnated with AgNO3,
attempts at the corresponding preparative method failed.
However, we were able to separate and purify the
regioisomeric cycloadducts formed from compounds 8 and
9 utilizing preparative HPLC (see the Experimental
Section). Regioisomers were readily differentiated by
their characteristic vinyl and bridgehead proton absorp-
1
tions in the H NMR spectra. The vinyl protons of the
cycloadducts consistently absorbed between 6.3 and 6.8
ppm while the bridgehead protons absorbed between 5.4
and 6.0 ppm. From the number and multiplicities of the
various absorptions, assignment of structure to that of
cycloadduct CA or CB was straightforward (see Experi-
mental Section).
Unfortunately, the 1,4-adducts of all of the 1-methyl-
substituted naphthalenes (3, 6, and 7) were too thermally
unstable for isolation and complete characterization as
they all experienced severe degradation during attempts
at preparative TLC or HPLC. However, the 1H NMR
spectra of these cycloadducts (obtained via subtraction
of the signals of the starting materials from the spectrum
of the reaction mixture [see Experimental Section]) are
completely consistent with the proposed structures.
Furthermore, each of the adducts underwent quantitative
cycloreversion to afford MeTAD and the starting naph-
thalene in the same manner as the more well-character-
ized cycloadducts. For these reasons we are quite confi-
dent in the correctness of the structures proposed.
D. Obser va tion of Ch a r ge-Tr a n sfer In ter m ed i-
a tes. Sheridan reported that UV-vis spectroscopy of
CCl4 solutions of MeTAD (0.6 mM) in the presence of an
excess (0.6 M) of naphthalene revealed a new absorption
band that was attributed to a 1:1 charge-transfer complex
between the two reactants.1 In our hands at these initial
concentrations, a broad unstructured absorption with λmax
) 378 nm was observed. Sheridan further reported that
selective irradiation of this complex (formed in only small
amounts) led directly to cycloadduct formation, suggest-
ing that the ground-state charge-transfer complex initi-
ated at least one pathway toward product formation.1
Charge-transfer complexes have been implicated as
product precursors in a number of reactions between
triazolinediones and electron-rich alkenes and aroma-
tics.1,10-12 Their intermediacy in reactions with aromatics
has been readily ascertained by the formation of char-
acteristic blood-red solutions upon mixing reactants.
tions are summarized in Table 2.
The reaction of MeTAD with 1-methylnaphthalene (6)
in the dark at 29 °C in CDCl3 was followed by periodic
1
monitoring of the reaction using H NMR spectroscopy.
Within the first 10 min of reaction a 15% yield of
cycloadduct 6CA was observed. As the reaction proceeded,
however, the yield of 6CA dropped slightly while cycload-
duct 6CB began to form. The yield of 6CB increased over
the course of 24 h to a maximum yield of 29%, while 6CA
was afforded with a final yield of 10%. No further change
was observed with extended reaction times. The com-
bined yield of cycloadduct corresponded to Keq ) 2.1
which was comparable to the equilibrium constant ob-
tained with naphthalene itself (Keq ) 2.0), but slightly
lower than that obtained with compound 3 (Keq ) 2.8). A
lesser equilibrium amount of cycloadduct was observed
when the reaction was conducted in acetone-d6.
A similar trend was observed during the course of
reaction of MeTAD with 1,4-dimethylnaphthalene (7). At
very short reaction times (<2 h), cycloadduct 7CA was
predominant (e.g., at t ) 0.5 h, there was 11% 7CA and
7% 7CB) while at longer reaction times cycloadduct 7CB
was predominant. The final ratio of 7:1 for 7CB to 7CA
after 24 h was greater than that observed for compound
6 (3:1).
Methyl substitution at the 2-position, such as in
2-methylnaphthalene (8) or 2,3-dimethylnaphthalene (9)
favored formation of cycloadduct CA. At short reaction
times (<0.5 h) the reaction of MeTAD with 8 afforded
similar amounts of cycloadducts 8CA (9%) and 8CB (6%).
The yield of 8CB increased to a maximum within 5 h
(11%) and then began to decline to a final equilibrium
value of 6%. The yield of 8CA, however, continued to
increase to a final equilibrium value of 58% over 48 h.
Similarly, compound 9 afforded a final equilibrium value
of 86% of 9CA with no observance of 9CB in the final
equilibrium mixture (after 48 h), although cycloadduct
9CB was observed in small amounts (5-6% at its
maximum) during the course of the reaction. When the
reaction of 9 was conducted in acetone-d6 a lower yield
of 9CA (21%) was observed in the final equilibrium
mixture than was observed in CDCl3. There was no
evidence (within limits of detection) for the formation of
9CB even at short reaction times.
Mixing of MeTAD with all of the substituted naphtha-
lenes in chloroform resulted in the formation of deep
blood-red solutions characteristic of MeTAD charge-
transfer complexes. New absorptions in the UV-vis
spectra attributed to charge-transfer complexes were
observed when MeTAD (0.6 mM) was mixed with com-
pounds 6 and 8 (each initially 0.6 M). The λmax values
for these absorptions (413 nM for 6 and 394 nM for 8)
were red-shifted from that observed for 2, which was
consistent with the presence of electron-donating methyl
(10) Hall, H. J . J . Org. Chem. 1983, 48, 1708.
(11) Hall, H. J .; Kaler, L.; Herring, R. J . Org. Chem. 1984, 49, 2579.
(12) Hall, H. J .; J ones, M. L. J . Org. Chem. 1983, 48, 822.
(13) The change in λmax of the charge-transfer complexes of MeTAD
with 6 and 8 relative to that of 2 (35 and 16 nm, respectively) correlate
well with the corresponding shifts reported for the TCNE complexes
of these same compounds (36 and 25 nm, respectively). See ref 18.
Many of the cycloadducts were isolated in the same
manner as was described previously for symmetrically
substituted naphthalenes. Generally the regioisomeric

2866 J . Org. Chem., Vol. 65, No. 10, 2000
Breton and Newton
Ta ble 3. Ra te Con sta n ts for th e Th er m a l Cyclor ever sion
of Cycloa d d u ct 2Ca ,b
Ta ble 4. Ra te Con sta n ts for th e Th er m a l Cyclor ever sion
of Va r iou s Cycloa d d u ctsa
solvent (ꢀ)
temp/°C
rate/10-5 s-1
cycloadduct
rate/ 10-5 s-1
rel rate
C6D6 (2.3)
CDCl3 (4.8)
(CD3)2CO (21)
CD3CN (38)
(CD3)2SO (47)
CDCl3
CDCl3
CDCl3
CDCl3
29
29
29
29
29
35
40
45
50
2.93 ( 0.26
3.24 ( 0.02
3.94 ( 0.03
5.34 ( 0.02
4.92 ( 0.05
6.01 ( 0.02
12.4 ( 0.12
23.3 ( 0.4
43 ( 2
2C
3C
4C
5C
6CA
6CB
7CB
8CA
8CB
9CA
9CB
3.24 ( 0.02
132 ( 2
2.41 ( 0.01
1.79 ( 0.02
223 ( 2
1.99 ( 0.02
1.13 ( 0.02
1.09 ( 0.01
4.83 ( 0.02
0.55 ( 0.05
10.6 ( 0.1
1
41
0.74
0.55
69
0.6
0.35
0.34
1.5
0.17
3.3
a
Conducted according to the general procedure provided in the
Experimental Section. Each independent run was performed at
least twice and the results averaged.
b
a
Conducted at 29 °C according to the general procedure
provided in the Experimental Section. b Each independent run was
performed at least twice, and the results were averaged.
groups on the rings.13 When this process was repeated
with compound 7, no charge-transfer complex and no free
MeTAD were observed. Apparently, under these condi-
tions, all of the MeTAD was quickly converted to cyclo-
adduct.
Ta ble 5. Ca lcu la ted Ra te Con sta n ts (k2 in Sch em e 1) for
th e Cycloa d d ition Rea ction s of MeTAD w ith Va r iou s
Na p h th a len esa ,b
naphthalene
solvent
rate/10-5 s-1
2. P h otoch em ica l Rea ction s of MeTAD w ith Su b-
stitu ted Na p h th a len es. Sheridan has reported the
formation of cycloadduct 2C in 40% isolated yield upon
irradiation of a solution of MeTAD and 2 in CCl4.1 We
repeated this reaction in acetone (in which only small
amounts of 2C were formed thermally) and isolated
compound 2C in 48% yield. Irradiation of an equimolar
solution of MeTAD and compound 6 in CCl4 afforded only
cycloadduct 6CB in 38% yield after purification by
preparative TLC.14 There was no indication of the pres-
2
CDCl3
C6D6
6.6 (1)
0.9 (0.14)
0.4 (0.06)
0.5 (0.08)
370
(CD3)2CO
CD3CN
CDCl3
CDCl3
CDCl3
3
4
5
4.3
630
a
Calculated using the Keq values in Table 1 and the cyclorever-
sion rates in Table 4. In parentheses are provided the rates
relative to the reaction conducted in CDCl3.
b
1
ence of cycloadduct 6CA in the H NMR spectrum of the
which is qualitatively similar to that previously reported
by Sheridan (t1/2 = 45 min at 40 °C in benzene).1
Activation parameters for the cycloreversion process were
determined to be ∆H‡ ) 23.6 kcal/mol and ∆S‡ ) -0.8
cal/deg‚mol (at 25 °C). The rate was relatively insensitive
to solvent polarity (Table 3), exhibiting a change in rate
of a factor of less than 2 from the least polar solvent
tested (benzene-d6) to the most polar (dimethyl sulfoxide-
d6).
The rates of cycloreversion of several other cyclo-
adducts in CDCl3 were conducted in the same manner
as described above, and the results are summarized in
Table 4. As with 2C, all of the cycloreversion reactions
exhibited first-order kinetics.
crude reaction mixture. Irradiation of an equimolar
solution of MeTAD and 6 in acetone also afforded
cycloadduct 6CB as the only detected product. Irradiation
of an equimolar solution of MeTAD and 8 in CCl4 afforded
a 46% isolated yield of a mixture of cycloadducts 8CA and
8CB in a 58:42 ratio, respectively, reflecting a mild
preference for reaction at the more electron-rich ring.14
Finally, photochemical reaction of MeTAD with 9 in CCl4
afforded a mixture of cycloadducts 9CA and 9CB in a ratio
of 68:32, respectively. The ratio was little affected by
solvent, affording a 60:40 ratio of 9CA to 9CB under
identical conditions in the more polar solvent acetone.
3. Th er m a l Cyclor ever sion of Cycloa d d u cts. Col-
orless solutions of purified cycloadducts began to turn
pink within seconds of dissolution in CDCl3 due to
liberation of MeTAD from the cycloadduct in a retro
Diels-Alder process. In the absence of a chemical trap
for MeTAD, an equilibrium was again established be-
tween cycloadduct, naphthalene, and MeTAD which
severely complicated attempts at measuring rates of
cycloreversion. However, by conducting the reaction in
the presence of a suitable trap for MeTAD, the forward
cycloaddition process was halted, and the rate of cyclo-
reversion was more conveniently measured. Thus, 2,3-
dimethyl-2-butene was employed as a chemical trap for
liberated MeTAD in all of the cycloreversion experiments
studied.9
Discu ssion
1. Th er m a l Rea ctivity. A. Th e Rea ction Mech a -
n ism . Rate constants for the cycloaddition reactions of
MeTAD with naphthalene in various solvents were
calculated (Table 5) by assuming Keq for the reaction to
be simply the ratio of the rate constant for the cyclo-
addition process and the rate constant for the cyclorever-
sion process. The relative rates are remarkably similar
to those previously reported for the Diels-Alder cyclo-
addition of MeTAD with 1,3-cyclooctadiene (1:0.14:0.03:
0.12 for CH2Cl2, benzene, acetone, and acetonitrile,
respectively).15 Recent theoretical studies on the cycload-
dition of MeTAD to dienes suggested that for dienes in
which an s-cis conformation is readily obtained (as is the
case with the naphthalenes) a concerted cycloaddition
pathway is favored.16a In several cases in which inter-
Rate constant determinations for the cycloreversion of
2C were made at a series of temperatures (Table 3) using
1H NMR spectroscopy. First-order kinetics over several
half-lives were observed in all cases examined. Cycload-
duct 2C exhibited a half-life of 1.5 h in CDCl3 at 40 °C,
(15) Isaksen, H.; Snyder, J . P. Tetrahedron Lett. 1977, 889.
(16) (a) Chen, J . S.; Houk, K. N.; Foote, C. S. J . Am. Chem. Soc.
1988, 120, 12303. (b) J ansen, F.; Foote, C. S. J . Am. Chem. Soc. 1987,
109, 6376.
(14) Sheridan (ref 1a) provided some initial experimental results
on this reaction.

Diels-Alder Reactions of MeTAD
J . Org. Chem., Vol. 65, No. 10, 2000 2867
Sch em e 1
formation than does substitution at the 1-position (e.g.,
compare the Keq values obtained for the compounds 3 and
6 with those of 4 and 8 in Tables 1 and 2). Monosubsti-
tution at the 2-position of a single ring of the naphthalene
nucleus is approximately equally effective at promoting
cycloadduct formation as is monosubstitution at both
rings (e.g., compare the Keq values obtained for compound
4 with that of 8). However, disubstitution on the same
ring is more effective than is disubstitution on separate
rings (i.e., compare the Keq values obtained for compound
9 with that of 4). It would therefore appear that the rings
act essentially independent of each other in their reac-
tions with MeTAD since greater electron-density at the
reacting ring is more important a factor than is electron-
density of the entire naphthalene core. This is further
borne out when comparing the reactivity of compound 6
with that of 2. Compound 6 contains an electron-donating
methyl group while 2 does not. Despite this, the Keq
values for the two reactions (Keq ) 2.0 and 2.1 for 2 and
6, respectively) are essentially equivalent because MeTAD
is directed toward reaction with the unsubstituted ring
of 6 (see discussion below for a rationalization of this
behavior) where the distant methyl group apparently has
little effect.
C. Th e Cycloa d d ition P r ocess. Table 5 summarizes
the calculated cycloaddition rates for reaction of MeTAD
with symmetrically substituted naphthalenes in CDCl3.
Methyl substitution at the 1-position of the naphthalene
nucleus as in 3 has a dramatic rate-enhancing effect and
increases the rate of cycloaddition by a factor of 50 over
that of 2. Consistent with this was the observation that
the reaction of 6 with MeTAD at short reaction times
(<30 min) revealed an initial preference for attack of
MeTAD at the substituted aromatic ring to afford adduct
6CA. Similarly, in the reaction of MeTAD with compound
7, adduct 7CA was the only product formed at short
reaction times (<20 min). In both cases the electron-
donating effect of the 1-methyl substituent activates the
rings toward cycloaddition relative to the unsubstituted
ring.
Substitution at the 1-position is more effective at
increasing the rate of cycloaddition than is substitution
at the 2-position as is observed from comparison of the
rates of cycloaddition of MeTAD with compounds 3 and
4. The poorer activating effect of the 2-methyl substituent
is further illustrated by the reactions of MeTAD with
compounds 8 and 9 where, unlike the reactions with 6
and 7, effective competition by both rings for reaction
with MeTAD occurred, and both regioisomeric adducts
CA and CB were observed throughout the reactions.
D. Th e Cyclor ever sion P r ocess. Examination of the
thermal cycloreversion data reported in Tables 3 and 4
reveals some interesting trends. For example, it is
apparent that methyl substitution at the 1- and 6-posi-
tions has a rate-accelerating effect on the reversion
process (e.g., 6CA cycloreverts 69 times faster than 2C,
and 8CB 1.5 times faster) while methyl substitution at
the 2- and 5-positions has a rate retarding effect (e.g.,
8CA cycloreverts 2.9 times slower than 2C, and 6CB 1.7
times slower).20 In addition, the substituents exert their
mediates were involved in the cycloaddition process the
intermediates proved trappable by the addition of exter-
nal nucleophiles.16 However, when the reaction between
compound 3 and MeTAD was conducted in a solvent
mixture of 3:1 CDCl3/CD3ODsa mixture that has been
successful in the past for trapping dipolar intermediates
involved in MeTAD reactionssno new adducts were
observed so that a long-lived dipolar intermediate (e.g.,
a zwitterion or an aziridinium imide) may be ruled out
as a possible intermediate.16 Given this evidence, it is
likely that the formation of cycloadducts occur via
concerted [4 + 2] cycloaddition processes. The lower
yields of cycloadduct obtained in the reaction of MeTAD
with naphthalene and substituted naphthalenes in the
more polar solvents (e.g., acetone) may be attributed to
selective solvation of the polar triazolinedione which
shifts the equilibrium toward starting materials.17
The role of charge-transfer intermediates in these
reactions has proven difficult to elucidate. Two reason-
able mechanisms for the reaction of MeTAD with naph-
thalene and substituted naphthalenes are presented in
Scheme 1. The two proposed mechanisms differ in their
description of the involvement of a charge-transfer
complex in the overall reaction pathway. On one hand,
the formation of charge-transfer complex intermediates
of naphthalene and substituted naphthalenes with strong
dienophiles such as MeTAD,1 maleic anhydride,6d
TCNE,18 and singlet oxygen19 have all been reported. In
addition, reactions of MeTAD with several electron-rich
aromatic compounds have been demonstrated to proceed
via initial formation of charge-transfer complexes.10-12
These observations would suggest the intermediacy of a
charge-transfer complex as a fast equilibrium step prior
to cycloadduct formation (mechanism A). Alternatively,
the charge-transfer complex could exist simply as a
nonproductive equilibrium process independent of the
cycloaddition process (as in mechanism B). In support of
this is the observation that the extent and rate of the
reactions have little dependence on the ionization energy
of the reacting naphthalenes. Unfortunately, therefore,
given the available evidence we are unable at this point
to satisfactorily distinguish between these two mecha-
nistic possibilities.
B. Effect of Su bstitu tion on Keq. The extent of
formation of cycloadduct in the established reaction
equilibria depends on the amount and pattern of substi-
tution on the naphthalene ring. In all cases investigated,
methyl substitution increased the amount of cycloadduct
formed at equilibrium. Substitution at the 2-position
appears to be more effective at promoting cycloadduct
(17) Burrage, M. E.; Cookson, R. C.; Gupte, S. S.; Stevens, D. R. J .
Chem. Soc., Perkin Trans. 2 1975, 1325.
(18) Frey, J . E.; Andrews, A. M.; Combs, S. D.; Edens, S. P.; Puckett,
J . J .; Seagle, R. E.; Torreano, L. A. J . Org. Chem. 1992, 57, 6460.
(19) Aubry, J .-M.; Mandard-Cazin, B.; Rougee, M.; Bensasson, R.
V. J . Am. Chem. Soc. 1995, 117, 9159.
(20) The stabilizing effect of a 2- and 5-methyl substituent toward
cycloreversion has also been observed in a series of 1,4-endoperoxides
formed via singlet oxygenation of 1,4-dimethylnaphthalene, 1,4,5-
trimethylnaphthalene, and 1,2,4-trimethylnaphthalene. See ref 7a and
also: Turro, N. J .; Chow, M.-F.; Rigaudy, J . J . Am. Chem. Soc. 1981,
103, 7218.

2868 J . Org. Chem., Vol. 65, No. 10, 2000
Breton and Newton
activating or deactivating effects in concert. For example,
the relative rate of cycloreversion of 3C was estimated
by multiplying the relative rate of 6CA by that of 6CB to
afford a value of 41 in excellent agreement with the
experimentally determined value (Table 1). In the same
manner, relative rates for the following cycloadducts were
estimated: 4C (0.51), 5C (0.56), 7CB (0.36), 9CB (2.3),
and 9CA (0.12) in reasonable agreement with the experi-
mentally determined values. Using this method, the
relative rate of cycloreversion of 7CA (a compound that
we were unable to isolate) is estimated to be 4800 (i.e.,
692).
The rate-accelerating effect of a 1-substituent in the
cycloreversion process may be accounted for by evoking
a destabilizing steric interaction between the carbonyl
group of the urazole ring with the naphthalene methyl
substituent which would be severe for cycloadducts
substituted at the 1-position but minimal for those
substituted at the 2(3)-, 5(8)-, or 6(7)-positions. In support
of this, molecular mechanics calculations (MM2) on
compound 6CA indicated an unsymmetrical structure
with a significantly longer C1-N bond (1.481 Å) than the
C4-N bond (1.472 Å).21 Modeling of 6CB, 8CA, 8CB
suggested symmetrical structures, indicating the lack of
steric interactions of the urazole ring with the methyl
groups in these positions. Interestingly, asymmetry is not
nearly so marked in the 1,4-cycloadduct of singlet oxygen
and 1-methylnaphthalene (C1-O bond ) 1.435 Å and
C4-O bond ) 1.431 Å) where a steric interaction is
expected to be considerably less. This would explain why
the reported singlet oxygenation of 1,4-dimethylnaph-
thylene afforded the 1,4-cycloadduct contrary to the
regioselectivity of MeTAD. However, it was also reported
that singlet oxygenation of a 1,4-dialkylnaphthylene
containing particularly bulky alkyl groups did direct
reaction to the 5,8-positions.22 In addition, it was ob-
served that the resulting oxygenated 1,4-cycloadduct of
this naphthalene was 10 times less stable than the 5,8-
adduct in qualitative agreement with our findings for the
relative stabilities of the corresponding MeTAD adducts.
E. Con clu sion s. The observed equilibria established
between each of the naphthalenes and MeTAD may be
rationalized in terms of kinetic versus thermodynamic
control of the reactions. For example, reaction of MeTAD
with naphthalenes substituted at the 1-position afforded
poorer equilibrium amounts of cycloadduct than naph-
thalenes substituted at the 2-position (e.g., compare the
Keq values for the reactions of compounds 3 and 4) despite
the superior rate-enhancing effect of a 1-substituent in
the cycloaddition process. This is explained by the fact
that whereas the 1-substituted methyl group kinetically
favors formation of the 1,4-cycloadducts in the cyclo-
addition reaction, the same methyl group destabilizes
these adducts in a thermodynamic sense as a result of
the steric interaction with the urazole ring. Owing to the
reversibility of the cycloaddition process, the thermody-
namics of the system eventually predominates. However,
a methyl substituent at the 2-position (e.g., 4) kinetically
favors formation of the 1,4-cycloadduct, is located far
enough away to have little destabilizing steric interaction
with the urazole ring of the product and retards cyclo-
reversion of the adduct. All of these factors favor a greater
Keq value for 4 than that of 3 as is observed. For
unsymmetrically substituted compound 8, methyl sub-
stitution at the 2-position kinetically activates the A-ring
toward cycloaddition relative to the B-ring and it also
retards cycloreversion of the resulting adduct. Thus,
reaction of 8 occurs predominantly at ring A. This effect
is even more pronounced for the 2,3-dimethylated com-
pound 9. For compound 6, however, despite the rate-
enhancing effect of the 1-methyl group in the cycloaddi-
tion reaction, the resulting 1,4-cycloadduct is thermo-
dynamically less stable that the 5,8-cycloadduct and
reactivity is ultimately directed predominantly to the
B-ring. A similar effect is observed for the 1,4-dimeth-
ylated compound 7.
The reported reactions of maleic anhydride with sub-
stituted naphthalenes are also reversible processes, and
demonstrated similar selectivities.6c Thus, reaction of
maleic anhydride with 1-methyl and 1,4-dimethylnaph-
thalene afforded 5,8-cycloadducts while reaction with
2-methyl and 2,3-diethylnaphthalene afforded 1,4-cy-
cloadducts.
2. P h otoch em ica l Rea ctivity. A viable reaction
mechanism for the photochemical reaction of MeTAD
with naphthalenes has already been proposed in which
both singlet and triplet photoexcited MeTAD cycloadd to
the naphthalene through the intermediacy of a contact
radical-ion pair.1 Unlike the behavior of MeTAD in its
thermal reactions, MeTAD displayed only mild regio-
selectivity in its photochemical reactions with substituted
naphthalenes. Reaction with compounds 8 and 9 yielded
mixtures of isomeric cycloadducts with a small preference
for reaction at the more electron-rich ring. Uncharacter-
istically, however, reaction with 6 afforded only product
from reaction at the unsubstituted ring (i.e., 6CB).
However, this result may be attributed to a difference in
cycloreversion rates of the two initially formed products.
Reaction at the substituted ring would form cycloadduct
6CA which is less stable than that of 6CB. Cycloreversion
of the less stable adduct would continually diminish the
concentration of 6CA while restocking the supply of
MeTAD that may add to form the more stable adduct
6CB. As the reaction progresses, therefore, adduct 6CB
continues to accumulate while the concentration of 6CA
is eventually depleted.
Exp er im en ta l Section
Materials were obtained commercially and used without
further purification unless otherwise specified. N-Methyl-
urazole was oxidized to MeTAD according to literature pro-
cedure, and sublimed under vacuum for final purification.23,24
Acetone used in photochemical reactions was dried over CaCl2
and distilled just prior to use. 1H NMR spectra were obtained
at 60 MHz and 13C NMR spectra at 15.1 MHz. The compound
2,3,6,7-tetramethylnaphthalene was obtained as a gift.
Gen er a l P r oced u r e for th e Th er m a l Rea ction s of
MeTAD w ith Na p h th a len e a n d Su bstitu ted Na p h th a -
len es. To a solution of MeTAD (25 mg, 2.21 mmol) in 0.50
mL of deuterated solvent (see Tables 1 and 2) away from
sources of direct light was added the naphthalene compound
(2.21 mmol). The resulting solution was mixed well and
transferred to a clean dry NMR tube which was wrapped with
aluminum foil to prevent exposure to light. A 1H NMR
spectrum was generally taken immediately afterward, and the
sample then placed in a water bath (usually set at 29 °C to
(21) MM2 calculations performed with CS Chem3D Pro, Cambridge-
Soft Corporation.
(22) Pierlot, C.; Aubry, J .-M. Chem. Commun. 1997, 2289.
(23) Mallakpour, S. E. J . Chem. Ed. 1992, 69, 238.
(24) Cookson, R. C.; Gupte, S. S.; Stevens, I. D. R.; Watts, C. T. Org.
Synth. 1971, 51, 121.

Diels-Alder Reactions of MeTAD
J . Org. Chem., Vol. 65, No. 10, 2000 2869
coincide with the NMR probe temperature). The sample was
removed at regular intervals for analysis by H NMR. If the
5.73 (d, J ) 5.3 Hz, 1H), 5.52 (d, J ) 1.9 Hz, 1H), 2.92 (s, 3H),
1.97 (d, J ) 1.8 Hz, 3H). 13C NMR (CDCl3): δ 157.3, 143.5,
137.0, 135.7, 126.9, 126.5, 124.4, 122.9, 122.3, 61.3, 57.6, 25.7,
1
cycloadduct was to be isolated, the sample was rinsed with 2
mL of CH2Cl2 into a separatory funnel containing 20 mL of
CH2Cl2 and the excess MeTAD was quenched by the addition
of a few drops of 2,3-dimethyl-2-butene. The organic layer was
washed once with 20 mL saturated aq NaHCO3, dried over
Na2SO4, filtered, and concentrated. The solid residue was then
purified by preparative TLC (20 × 20 cm SiO2 plates) using
an appropriate solvent mixture of ethyl acetate and hexane.
Regioisomeric adducts were generally purified by preparative
HPLC on a 25 cm × 10 mm 5 µm Supelcosil SPLC-Si column
eluting with 10% 2-propanol in hexane at a flow rate of 3 mL/
min (UV detector set at 255 nm). Purified products were
characterized by IR and NMR spectroscopy.
19.5. IR (KBr): 1765, 1715, 1455 cm-1
.
5,8-Cycloa d d u ct w ith 2-Meth yln a p h th a len e (8CB).1 1
H
NMR (CDCl3): δ 7.3-7.0 (m, 3H), 6.79 (dd, J ) 3.2 Hz, 2H),
5.76 (br t, J ) 3.8 Hz, 2H), 2.93 (s, 3H), 2.33 (s, 3H). IR (KBr):
1750, 1710, 1450 cm-1
.
1,4-Cycloa d d u ct w ith 2,3-Dim eth yln a p h th a len e (9CA).
1H NMR (CDCl3): δ 7.22 (m, 4H), 5.47 (s, 2H), 2.88 (s, 3H),
1.87 (s, 6H). 13C NMR (CDCl3): δ 159.5, 138.6, 135.5, 128.8,
124.4, 64.4, 27.7, 18.0. IR (KBr): 1760, 1710, 1450 cm-1
.
5,8-Cycloa d d u ct w ith 2,3-Dim eth yln a p h th a len e (9CB).
1H NMR (CDCl3): δ 7.10 (m, 2H), 6.78 (dd, J ) 3.1, 4.0 Hz,
2H), 5.72 (dd, J ) 3.1, 4.0 Hz, 2H), 2.93 (s, 3H), 2.23 (s, 6H).
1,4-Cycloa d d u ct w ith Na p h th a len e (2C).1 1H NMR
(CDCl3): δ 7.26 (m, 4H), 6.81 (dd, J ) 3.1, 4.0 Hz, 2H), 5.81
(dd, J ) 3.1, 4.0 Hz, 2H), 2.93 (s, 3H). IR (KBr): 1760, 1705,
IR (KBr): 1770, 1705, 1450 cm-1
.
Gen er a l P r oced u r e for th e P h otoch em ica l Rea ction s
of MeTAD w ith Na p h th a len e a n d Su bstitu ted Na p h th a -
len es. Solutions of MeTAD (50 mg) and the naphthalene
compound (15 mM each) in CCl4 or acetone were deoxygenated
with a stream of dry nitrogen gas and irradiated through a
sealed water-jacketed Pyrex flask using three 300 W incan-
descent bulbs until the solution became colorless (between 10
min to 1 h depending on the naphthalene compound). The
solutions were concentrated, analyzed by 1H NMR spectros-
copy, and subjected to purification by preparative TLC or
HPLC as described above.
1460 cm-1
.
1,4-Cycloa d d u ct w ith 1,5-Dim eth yln a p h th a len e (3C).
1H NMR data were obtained by difference on an equilibrium
mixture of MeTAD, 3, and cycloadduct due to the thermal
lability of the adduct which prevented its isolation (see Results
section above): 1H NMR (CDCl3): δ 7.09 (m, 3H), 6.79 (dd, J
) 5.7, 7.7 Hz, 1H), 6.44 (dd, J ) 1.7, 7.7 Hz, 1H), 6.0 (dd, J )
1.7, 5.7 Hz, 1H), 2.83 (s, 3H), 2.43 (s, 3H), 2.3 (s, 3H).
1,4-Cycloa d d u ct w ith 2,6-Dim eth yln a p h th a len e (4C).
1H NMR (CDCl3): δ 7.2 (m, 3H), 6.30 (dp, J ) 5.8, 1.8 Hz,
1H), 5.65 (d, J ) 5.8 Hz, 1H), 5.47 (d, J ) 1.8 Hz, 1H), 2.89 (s,
3H), 2.31 (s, 3H), 1.95 (d, J ) 1.9 Hz, 3H). 13C NMR (CDCl3):
δ 157.4, 144.1, 137.1, 133.7, 132.9, 127.1, 124.6, 123.5, 122.9,
Gen er a l P r oced u r e for th e Kin etics Mea su r em en ts of
th e Th er m a l Cyclor ever sion s. A solution of cycloadduct
(approximately 0.30 M), dichloroethane (1 equiv), and 2,3-
dimethyl-2-butene (2 equiv) in 0.5 mL of CDCl3 (or otherwise
indicated solvent) in an NMR tube was heated in a waterbath
maintained at the desired temperature. Periodic determina-
tions of the extent of the reaction were made by examining
the integration of the N-Me signals of the adduct (ap-
proximately δ 2.9) and the ene-product (δ 3.02) versus the
signal of dichloroethane (δ 3.70) as an internal standard. Clean
first-order kinetics was observed over at least 3 half-lives in
all cases. An average value of two or more runs was taken for
each rate constant.
61.4, 57.8, 26.0, 21.8, 19.8. IR (KBr): 1760, 1720, 1460 cm-1
.
1,4-Cycloa d d u ct w ith 2,3,6,7-Tetr a m eth yln a p h th a len e
(5C). 1H NMR (CDCl3): δ 7.07 (s, 2H), 5.40 (s, 2H), 2.90 (s,
3H), 2.22 (s, 3H), 1.86 (s, 3H). 13C NMR (CDCl3): δ 155.2,
134.0, 132.7, 123.6, 121.7, 60.0, 23.6, 17.8, 13.8. IR (KBr):
1765, 1710, 1455 cm-1
.
1,4-Cycloa d d u ct w ith 1-Meth yln a p h th a len e (6CA).1 1
H
NMR data were obtained by difference on a mixture of MeTAD,
6, and 6A (at short reaction times) due to the thermal lability
of the adduct which prevented its isolation (see Results section
above): 1H NMR (CDCl3): δ 6.78 (dd, J ) 5.8, 7.7 Hz, 1H),
6.44 (dd, J ) 1.7, 7.7 Hz, 1H), 5.73 (dd, J ) 1.7, 5.8 Hz, 1H),
2.87 (s, 3H), 2.31 (s, 3H). The aromatic protons of compound
6A were obscured by the aromatic protons of compound 6.
Ack n ow led gm en t is made to the donors of the
Petroleum Research Fund, administered by the Ameri-
can Chemical Society, for partial support of this re-
search. The NMR upgrade used in this work was made
possible by the National Science Foundation’s Division
of Undergraduate Education (ILI-IP) through DUE
Grant no. 9750684. The UV-vis spectrophotometer used
in this work was made possible by the 1998 Camille and
Henry Dreyfus Special Grant Program in the Chemical
Sciences. We also thank Dr. Yoshio Aso of the Depart-
ment of Applied Chemistry, Hiroshima University, for
the generous gift of 2,3,6,7-tetramethylnaphthalene.
5,8-Cycloa d d u ct w ith 1-Meth yln a p h th a len e (6CB).1 1
H
NMR (CDCl3): δ 7.08 (br m, 3H), 6.81 (dd, J ) 3.1, 4.0 Hz,
2H), 6.06 (dd, J ) 3.1, 4.0 Hz, 1H), 5.77 (dd, J ) 3.1, 4.0 Hz,
1H), 2.90 (s, 3H), 2.43 (s, 3H). 13C NMR (CDCl3): δ 157.4 (2
C), 136.1, 134.6, 132.9 (2 C), 131.8, 128.7, 126.6, 120.8, 57.3,
53.8, 25.9, 18.3. IR (KBr): 1765, 1715, 1460 cm-1
.
1,4-Cycloa d d u ct w ith 1,4-Dim eth yln a p h th a len e (7CA).
1H NMR data were obtained by difference on a mixture of
MeTAD, 7, and 7A (at short reaction times) due to the thermal
lability of the adduct which prevented its isolation (see Results
section above): 1H NMR (CDCl3): δ 7.25 (br s, 4H), 6.43 (s,
2H), 2.78 (s, 3H), 2.30 (s, 3H).
5,8-Cycloa d d u ct w ith 1,4-Dim eth yln a p h th a len e (7CB).
1H NMR (CDCl3): δ 6.96 (m, 2H), 6.86 (dd, J ) 4.1, 3.2 Hz,
2H), 6.06 (dd, J ) 3.2, 4.1 Hz, 1H), 2.94 (s, 3H), 2.44 (s, 6H).
Su p p or tin g In for m a tion Ava ila ble: 1H NMR spectra for
cycloadducts 2C, 4C, 5C, 6CB, 7CB, 8CA, 8CB, 9CA, and 9CB.
This material is available free of charge via the Internet at
http://pubs.acs.org.
IR (KBr): 1765, 1705, 1460 cm-1
.
1,4-Cycloa d d u ct w ith 2-Meth yln a p h th a len e (8CA).1 1
NMR (CDCl3): δ 7.24 (m, 4H), 6.33 (dp, J ) 5.3, 1.8 Hz, 1H),
H
J O9906429