N-Vinyl and N-Aryl Hydroxypyridinium Ions: Charge-Activated Catalysts with Electron-Withdrawing Groups

Article abstract of DOI:10.1021/acs.joc.0c00498

Charge-enhanced Br?nsted acid organocatalysts with electron-withdrawing substituents were synthesized, and their relative acidities were characterized by computations, 1:1 binding equilibrium constants (K_1:1) with a UV-vis active sensor, ³¹P NMR shifts upon coordination with triethylphosphine oxide, and in one case by infrared spectroscopy. Pseudo-first-order rate constants were determined for the Friedel-Crafts alkylations of N-methylindole with trans-β-nitrostyrene and 2,2,2-trifluoroacetophenone and the Diels-Alder reaction of cyclopentadiene with methyl vinyl ketone. These results along with kinetic isotope effect determinations revealed that the rate-determining step in the Friedel-Crafts transformations can shift from carbon-carbon bond formation to proton transfer to the catalyst's conjugate base. This leads to an inverted parabolic reaction rate profile and slower reactions with more acidic catalysts in some cases. Electron-withdrawing groups placed on the N-vinyl and N-aryl substituents of hydroxypyridinium ion salts lead to enhanced acidities, more acidic catalysts than trifluoroacetic acid, and a linear correlation between the logarithms of the Diels-Alder rate constants and measured K_1:1 values.

Full text of DOI:10.1021/acs.joc.0c00498

Subscriber access provided by University of Birmingham
Article
N-Vinyl and N-Aryl Hydroxypyridinium Ions: Charge-
Activated Catalysts with Electron Withdrawing Groups
George F Riegel, and Steven Robert Kass
J. Org. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.joc.0c00498 • Publication Date (Web): 08 Apr 2020
Downloaded from pubs.acs.org on April 16, 2020
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a service to the research community to expedite the dissemination
of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in
full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully
peer reviewed, but should not be considered the official version of record. They are citable by the
Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore,
the “Just Accepted” Web site may not include all articles that will be published in the journal. After
a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web
site and published as an ASAP article. Note that technical editing may introduce minor changes
to the manuscript text and/or graphics which could affect content, and all legal disclaimers and
ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or
consequences arising from the use of information contained in these “Just Accepted” manuscripts.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 36
The Journal of Organic Chemistry
1
2
3
4
5
6
7
N-Vinyl and N-Aryl Hydroxypyridinium Ions: Charge-Activated
8
9
Catalysts with Electron Withdrawing Groups
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
George F. Riegel and Steven R. Kass*
Department of Chemistry, University of Minnesota
207 Pleasant Street, SE, Minneapolis, Minnesota 55455, United States
E-mail: kass@umn.edu
1
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 2 of 36
1
2
3
4
ABSTRACT
5
6
7
8
9
Charge-enhanced Brønsted acid organocatalysts with electron withdrawing substituents were
synthesized and their relative acidities were characterized by computations, 1:1 binding
equilibrium constants (K1:1) with a UV-vis active sensor, 31P NMR shifts upon coordination with
triethylphosphine oxide, and in one case by IR spectroscopy. Pseudo-first order rate constants were
determined for the Friedel–Crafts alkylation of N-methylindole with trans-β-nitrostyrene and
2,2,2-trifluoroacetophenone, and the Diels–Alder reaction of cyclopentadiene with methyl vinyl
ketone. These results along with kinetic isotope effect determinations revealed that the rate
determining step in the Friedel–Crafts transformations can shift from carbon-carbon bond
formation to proton transfer to the catalyst’s conjugate base. This leads to an inverted parabolic
reaction rate profile and slower reactions with more acidic catalysts in some cases. Electron
withdrawing groups placed on the N-vinyl and N-aryl substituents of hydroxypyridinium ion salts
lead to enhanced acidities, more acidic catalysts than trifluoroacetic acid, and a linear correlation
between the logarithms of the Diels–Alder rate constants and measured K1:1 values.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
ACS Paragon Plus Environment

Page 3 of 36
The Journal of Organic Chemistry
1
2
3
4
INTRODUCTION
5
6
7
8
9
Acid-catalyzed reactions are commonplace and are utilized in a wide variety of transformations
spanning from the synthesis of small molecules to biocompounds and polymers. The development
of Brønsted acids and hydrogen bond donating organocatalysts, consequently, is of significant
interest and a subject of considerable research efforts.1,2 Incorporation of electron withdrawing
groups into these species not only increases their acidities and hydrogen bond donor abilities, but
also typically leads to more active catalysts and enhanced reaction rates.3,4 Extension of this
strategy with the use of one or more positively charged centers was recently reported to be
particularly effective in non-polar solvents leading to rate accelerations corresponding to orders of
magnitude compared to non-charged substituents.5 Even more active catalysts can be envisioned
by introducing electron withdrawing groups into positively charged substrates. To explore this
idea, a series of six model hydroxypyridinium salts were examined (1–6, Figure 1). In this report,
we describe the syntheses of these six compounds, companion density functional theory
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
31
computations, IR, UV-vis, and P NMR spectroscopic studies, along with kinetic isotope effect
and rate constant determinations to assess catalyst acidities, activities, and reaction mechanisms.
R1
Me
R2
OH
OH
R3
H
H
1
2
3
4
5
6
R2
(E)-CH=CHCO2Et
R3
(E)-CH=CHCO2Oct OH
H
H
H
OH
BArF
4
2,4-(NO2)2Ph
(E)-CH=CHCN
2,4-(NO2)2Ph
OH
OH
H
+
N
R1
Figure 1. Catalysts used in this work.
3
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 4 of 36
1
2
3
4
RESULTS AND DISCUSSION
5
6
7
The tetrakis[3,5-bis(trifluoromethyl)phenyl]borate (BArF –) salt of the N-n-octyl-3-
4
8
9
hydroxypyridinium ion has previously been reported,5a and its N-methyl analog (1) was made in a
similar way.5d Various strategies were employed for the preparation of the halide precursors of the
other model species. Compound 2-Cl was synthesized via a modification of the procedure from
Jung and Buszek6 whereas 3-Cl was formed using an analogous method to Katrizky et al.7 The
preparation of 5-Br was accomplished using a similar procedure to that of Eicher-Lorka et al,8 and
the 2,4-dinitrophenyl-substituted catalyst precursors, 4-Cl and 6-Cl, were obtained via a similar
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
route to the one reported by Eda, Kurth, and Nantz.9 Conversions of the halide salts to their BArF –
4
analogs were carried out largely in the same way and took advantage of the differential solubilities
of NaCl and NaBr compared to NaBArF in organic solvents.
4
Catalyst acidities were first evaluated at 298 K in the gas phase without the BArF –
4
counteranion using B3LYP/6-31G(d,p) computations (Table 1). This methodology was previously
used to reproduce experimental Brønsted acidities and successfully predict catalyst activities in
non-polar media.5a As anticipated, the presence of vinyl and aryl substituents with electron
Table 1. Computed B3LYP/6-31G(d,p) acidities for a series of catalysts without their
counteranions.a
catalyst ΔG˚acid
ΔΔG˚acid
1
2
4
5
6
232.9
230.9
226.6
224.2
219.3
0.0
–2.0
–6.3
–8.7
–13.6
aAll values are given in kcal mol–1 at 298 K. Smaller numbers for DG˚acid correspond to stronger
acids.
4
ACS Paragon Plus Environment

Page 5 of 36
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
withdrawing groups at the nitrogen ring center lead to more acidic species than their alkyl-
substituted counterpart (i.e., 1). In addition, 6, the para-substituted analog of 4, is the strongest
acid in this series due to enhanced electron delocalization in its conjugate base.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Several spectroscopic approaches were carried out to experimentally evaluate the relative
acidities and hydrogen bond donating abilities of 1–6. Infrared spectroscopy was initially explored
as previously reported by observing the change between the O–H stretching frequency of a phenol
in carbon tetrachloride and upon coordination with 1% v/v CD3CN in CCl4.5a,5d A reduction in this
mode from 3541 cm–1 to 3128 cm–1 was observed for 3, and this difference of 413 cm–1 is larger
than the 370 cm–1 reported for N-n-octyl-3-hydroxypyridinium BArF .5a This result is consistent
4
with an acidity enhancement due to the presence of an electron withdrawing group, but 2 and 4–6
are not soluble in carbon tetrachloride or in CD3CN/CCl4 mixtures with up to 5% acetonitrile-d3.
Consequently, we decided to focus on alternative approaches.
A UV-vis titration method for evaluating hydrogen bond donor ability utilizing 7-methyl-2-
phenylimidazo[1,2-a]pyrazine-3(7H)-one (7) as a colorimetric sensor was first reported by
Kozlowski and coworkers and recently updated by Payne and Kass.5d,10 In the latter approach, a
constant concentration of 7 is maintained as it is titrated with a Brønsted acid of interest (HOAr in
this case). This leads to a blue-shift of the λmax of 7 due to its coordination with the hydrogen bond
donor. When a 1:1 association occurs (Scheme 1), an isosbestic point is observed (Figures 2 and
HOAr
Ph
HOAr
Ph
O
O
O
HOAr
HOAr
N
N
N
Ph
N
N
K1:1
K1:2
N
N
N
N
7
1:1 binding
HOAr
1:2 binding
Scheme 1. Coordination of 7 to afford 1:1 and 1:2 binding complexes.
5
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 6 of 36
1
2
3
4
5
6
7
8
9
0.4
0.3
0.2
0.1
0
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
350
400
450
500
550
600
wavelength (nm)
Figure 2. UV-vis spectra from 350 to 600 nm for the titration of 7 with 4. Only some of the spectra
are provided for clarity.
S8–S11) and a non-linear fit of the data affords the equilibrium binding constant (K1:1, Table 2).
This was the case for 1, 2, 4, and 5, but not 6. An isosbestic point is not observed for this last
compound (Figure 3) which is indicative of the formation of at least two intermediate species, and
Table 2. Non-linear binding isotherm titration results for a series of catalysts with 7.a
catalyst λmax(1:1) (nm)b K1:1
K1:2
1c
2
472.1 (471.7) 2.0 × 104
469.4 (468.8) 5.3 × 104
469.1 (468.8) 8.5 × 104
466.3 (466.0) 1.7 × 105
4
5
6
434.7
3.5 × 105 7.2 × 103
aConstant sensor concentrations of 22–34 µM in CH2Cl2 were used. Calculated values obtained
from a plot of χ/λmax at each titration point, where χ is the 1:1 mole fraction resulting from the fit
of the data. See ref. 5d for further details. Parenthetical numbers are the directly observed λmax
values from the titrations. cData from ref. 5d.
b
6
ACS Paragon Plus Environment

Page 7 of 36
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
0.30
0.20
0.10
0.00
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
350
400
450
500
550
600
wavelength (nm)
Figure 3. UV-vis spectra from 350 to 600 nm for the titration of 7 with 6. Only some of the spectra
are provided for clarity.
is consistent with the resulting 1:2 binding model fit of the titration results. A plot of the logarithm
of K1:1 versus ΔG˚acid (Figure 4) is linear for all five hydroxypyridinium ion salts examined in this
work, indicating that the 1:1 binding equilibrium constants provide a good measure of the
catalysts’ acidities and hydrogen bond donating abilities.11
14
13
12
11
10
9
215
220
225
230
235
ΔG˚acid (kcal mol–1)
Figure 4. Logarithm of 1:1 equilibrium binding constants versus computed gas-phase acidities. A
linear least-squares analysis affords ln K1:1 = –0.198 × ΔG˚acid + 56.2, r2 = 0.959.
7
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 8 of 36
1
2
3
4
5
6
7
8
9
An alternative spectroscopic approach for evaluating Brønsted acidities and hydrogen bond
donating abilities in nonpolar media was reported by Diemoz and Franz this past year.12 In this
methodology, the 31P NMR spectrum of triethylphosphine oxide (Et3PO) is recorded in
dichloromethane and subsequently titrated with a Brønsted acid of interest. Rapid formation of a
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
31
hydrogen bond complex leads to an averaged spectrum with a single downfield P signal. Its
chemical shift progressively increases until it levels off to a constant value as more of the acid is
added and the proportion of free (non-complexed) Et3PO diminishes to negligible amounts. The
magnitude of this shift (Δδ) was found to correlate with the activity of a variety of organocatalysts.
For 1 it was found that only 1.5 equivalents are needed to bind to Et3PO to achieve the maximum
downfield shift (Table S7). Since all of our other catalysts are predicted to be more acidic than 1,
1.5 equivalents of these acids were considered to be sufficient to obtain the chemical shifts of the
bound complexes (i.e. δ1.5). A summary of the results is given in Table 3 and Δδ for the meta-
substituted isomers (1, 2, 4, and 5) were found to linearly correlate with ln K1:1 from the UV-vis
data (Figure 5) and ΔG˚acid (i.e., the computed gas-phase acidities, Figure S14). The para-
Table 3. Titration results for a series of catalysts with Et3PO monitored by 31P NMR.a
catalyst δ1.5 (ppm)b Δδ (ppm)
1
2
4
5
6
60.83
61.67
61.96
62.49
65.44
10.33
11.17
11.46
11.99
14.94
aTriethylphosphine oxide (5.0 mM) in CD2Cl2 has a chemical shift of 50.50 ppm. Observed
b
chemical shifts upon addition of 1.5 equivalents of the indicated acid.
8
ACS Paragon Plus Environment

Page 9 of 36
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
16
15
14
13
12
11
10
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
9
10
11
12
13
14
ln K1:1
31
Figure 5. Triethylphosphine oxide P NMR shift changes versus ln K1:1 for meta-substituted
catalysts (filled circles) and 6 (open circle). A linear least-squares fit for the meta compounds gives
the following equation: Δδ (ppm) = 0.781 × ln K1:1 + 2.62, r2 = 0.997.
substituted derivative shows positive deviations from the lines in both plots, presumably because
of its enhanced acidity and hydrogen bond donating ability due to the conjugation of the acidic site
with the formally charged nitrogen atom center.
Computations, 1:1 equilibrium binding constants from UV-vis titrations, and the
triethylphosphine oxide 31P NMR data all predict a catalyst activity order of 6 > 5 > 4 > 2 > 1. To
assess this prognostication, the Friedel–Crafts alkylation of N-methylindole with trans-β-
nitrostyrene (eq. 1) was examined. This reaction was chosen because it is known to progress more
rapidly with stronger acids in a related series and has been widely used to assess
organocatalysts.5,12,13 Pseudo-first order rate constants were obtained using H NMR to monitor
1
the reaction progress (Table 4). The reaction rates increased as expected for the meta-substituted
Ph
NO2
cat.
NO2
+
(1)
Ph
CD2Cl2
N
N
9
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 10 of 36
1
2
3
4
5
6
Table 4. Kinetic data for the Friedel–Crafts alkylation of N-methylindole with trans-β-
nitrostyrene.a
entry catalyst k (h–1)
t1/2 (h) krel
7
8
9
1
2
3
4
5
6
7
1.9 × 10–4 3700
0.0023
1.0
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
0.082
0.17
8.5
4.2
3.6
2.4
3.3
79
2
2.0
2.4
3.6
2.6
0.11
4
0.19
5
0.29
6
0.21
TfOH
0.0088
aReactions were carried out at 27 ˚C in CD2Cl2 with 10 mol% of the indicated catalyst, 500 mM
of trans-β-nitrostyrene, and 50 mM of N-methylindole. Rate constants are corrected for the non-
catalyzed background process.
pyridinium salts (i.e., 5 > 4 > 2 > 1), but surprisingly, 6 afforded a smaller rate acceleration than 5
by a factor of 1.4, despite the former’s greater acidity, larger K1:1, and bigger Δδ. Triflic acid was
also examined, and it was found to be less active than our phenol derivatives and a poor catalyst
for this transformation. This led us to hypothesize that for the stronger acids the carbon-carbon
bond forming step in the Friedel–Crafts alkylation may no longer be rate determining, and that
deprotonation at C3 to aromatize the indole and reform the catalyst could become the slow step
(Scheme 2). Such variations with changing pH are well-known and have been reported with
organocatalysts,14 but not for the Friedel–Crafts reaction of N-methylindole with trans-β-
nitrostyrene even though it has been extensively used for benchmarking purposes.
10
ACS Paragon Plus Environment

Page 11 of 36
The Journal of Organic Chemistry
1
2
3
4
5
6
Ph
Ph
Ph
Ph
O–
O–
+
+
N
N
NO2
NO2
+ HOAr
C3
C3
C3
H
slow
3
fast
1
fast
2
O–
O
HOAr
HOAr
–OAr
N
N+
N+
N
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 2. Proposed reaction mechanism for the Friedel–Crafts alkylation of N-methylindole with
trans-β-nitrostyrene where step 3 becomes rate determining with strong acid catalyts.
To probe a possible change in the rate determining step of the mechanism, isotopically labeled
3-deuterio-N-methylindole15 was prepared and kinetic isotope effects (KIEs) for the Friedel–Crafts
alkylation with different catalysts were determined (Table 5). An inverse KIE was obtained with
4, and this is consistent with a rate determining carbon-carbon bond forming step (i.e., a change in
hybridization from sp2 to sp3 in the nucleophilic attack of N-methylindole on trans-β-nitrostyrene).
Conversely, normal KIEs were found with 6 and triflic acid. This is in accord with the
deprotonation at C3 becoming the slow step in the reaction mechanism when more acidic catalysts
are used. No isotope effect was found with 5, suggesting that nucleophilic attack and
Table 5. Kinetic data for the Friedel–Crafts alkylation of 3-deuterio-N-methylindole with
trans-β-nitrostyrene.a
entry catalyst kD (h–1) t1/2 (h) K.I.E.
1
2
3
4
4
0.24
2.8
2.4
5.2
138
0.80
1.0
1.6
1.7
5
0.29
6
0.13
TfOH
0.0050
aBackground corrected initial rates corresponding to reaction conversions of ≤ 10% were obtained
at 27 ˚C in CDCl2 using 50 mM of 3-deuterio-N-methylindole, 500 mM trans-β-nitrostyrene, and
10 mol% of the indicated catalyst.
11
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 12 of 36
1
2
3
4
5
6
7
8
9
deprotonation are competitive in this case. In accord with the proposed mechanism in Scheme 2,
2H NMR experiments showed that deuterium was only incorporated into the product at the α-
position to the nitro group and was not observed until greater than 10% conversion to the product
was reached. These findings reveal that the reaction rate for the Friedel–Crafts alkylation is not a
reliable indicator of relative acidities for acidic catalysts.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
It was anticipated that a more acidic intermediate in the Friedel–Crafts reaction is formed when
2,2,2-trifluoroacetophenone is used as the electrophile in lieu of trans-b-nitrostyrene. Two electron
withdrawing groups are present in the former case rather than one (i.e., OH and CF3 vs NO2) and
they both are one bond closer to the key C3 position. This should facilitate the re-aromatization
step (3 in Scheme 2) making it more likely that carbon-carbon bond formation will be rate
determining, and increasing the acidity range of catalysts that can be reliably probed. Mono- and
bis-coupled products are commonly observed for this transformation (eq. 2)16 and so the
selectivities along with the reaction rate constants are given in Table 6 for the least (1) and most
(5, 6, and TfOH) acidic catalysts that were explored in this study. As expected, the KIEs are smaller
for each catalyst, and this transformation proceeds 3.5 times more rapidly with 6 than with 5.
Formation of the bis-coupled product increases with the catalytic activity suggesting that the mono
: bis ratio can also serve as an indicator of the catalysts' relative acidities. Tunable catalysts with
modular syntheses such as our hydroxypyridinium salts are useful for this type of reaction since
the proper balance between reaction rate and selectivity must be struck.
N
OH
Ph
Ph
CF3
CF3
O
cat.
+
(2)
+
CD2Cl2
Ph
CF3
N
N
N
12
ACS Paragon Plus Environment

Page 13 of 36
The Journal of Organic Chemistry
1
2
3
4
5
6
Table 6. Kinetic and selectivity data for the Friedel–Crafts reaction of N-methylindole with
2,2,2-trifluoroacetophenone.a
7
8
9
entry cat.
1
kH (h–1)
t1/2 (h) krel
mono:bisb kD (h–1)c t1/2 (h) kH/kD
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
7.4 × 10–5 9400
0.0010 100:0
2
3
4
5
1
5
6
0.072
0.34
1.1
10
1
100:0
98:2
2.1
0.61
13
4.6
16
0.38
0.95
0.037
1.8
0.73
19
0.90
1.2
91:9
TfOH 0.053
0.73
86:14
1.4
aReactions were carried out at 27 °C in CD2Cl2 with 500 mM of 2,2,2-trifluoroacetophenone,
50 mM of N-methylindole or 3-deuterio-N-methylindole, and 10 mol% of the indicated catalyst.
bThe product selectivities were determined using N-methylindole. cInitial rates were used.
To avoid multistep mechanisms where the rate determining step can change with the acidity of
the catalyst, and to examine a model transformation that better reflects the Brønsted acidity and
hydrogen bond donating ability of the catalyst, the Diels–Alder reaction between cyclopentadiene
and methyl vinyl ketone was examined (eq. 3). Pseudo-first order rate constants and the endo:exo
O
O
cat.
(3)
+
CD2Cl2
product ratios were measured at 25 °C in dichloromethane-d2 with a 1 mol% catalyst loading
(Table 7). As expected based upon the computed acidities, UV-vis and 31P NMR data, the reactivity
order is now 6 > 5 > 4 > 1 and the selectivities follow a similar trend. Excellent correlations
between ln k and Δδ, ln K1:1, and ΔG˚acid are observed (Figures 6, S15 and S16), indicating that all
three approaches can serve as useful guides for catalytic activities.
13
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 14 of 36
1
2
3
4
5
6
Table 7. Kinetic and selectivity data for the Diels–Alder reaction of cyclopentadiene with
methyl vinyl ketone.a
7
8
9
entry catalyst k (h–1) t1/2 (h) krel
endo:exo
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
2
3
4
5
0.11
0.15
0.46
0.81
3.4
6.5
0.72 86:14
1
4
5
6
4.7
1
89:11
91:9
91:9
92:8
1.5
3.1
5.5
23
0.85
0.20
aReactions were carried out at 25 °C in CD2Cl2 with 1 mol% of the indicated catalyst, 250 mM
cyclopentadiene, and 25 mM methyl vinyl ketone. Background corrected rate constants for the
non-catalyzed process are given.
-2
-3
-4
-5
-6
-7
9
10
11
12
13
14
15
16
Δδ (ppm)
Figure 6. A plot of the logarithm of the rate constant for the Diels–Alder reaction of
cyclopentadiene with methyl vinyl ketone vs Δδ of the Et3PO 31P NMR shift for meta-substituted
catalysts (filled circles) and the para-isomer 6 (open circle). Linear least squares fits of the data
give ln k = 1.02 × Δδ – 16.6, r2 = 1.00 (meta isomers only) and ln k = 0.650 × Δδ – 12.4, r2 = 0.955
(all compounds).
14
ACS Paragon Plus Environment

Page 15 of 36
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
A comparison of the charge-activated catalysts with non-charged species would be useful to
obtain a better sense of their qualitative acidities. Pentafluorophenol (C6F5OH) and
pentafluorobenzoic acid (C6F5CO2H) were chosen for this purpose because they have known
31
equilibrium association constants with 710 and Δδ P NMR shifts with Et3PO,12 and the latter
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
values for pentafluorophenol and pentafluorobenzoic acid are between those for 1 and 4, and 5 and
6, respectively (Table 8). The UV-vis method, however, suggests that the two non-charged
compounds are significantly weaker acids than the hydroxypyridinium salts, and should be much
less effective catalysts. Kinetic data for both the Friedel–Crafts alkylation of N-methylindole with
trans-β-nitrostyrene and the Diels–Alder reaction of cyclopentadiene with methyl vinyl ketone
were obtained using both C6F5OH and C6F5CO2H. The resulting rate constants for the latter
reaction are in accord with the K1:1 values from the UV-vis titrations in that the two non-charged
Table 8. Spectroscopic and kinetic data for the Friedel–Crafts and Diels–Alder reactions
with all catalysts.
catalyst
K1:1
Δδ (ppm) kFC (h–1)a t1/2 (h) kDA (h–1)b t1/2 (h)
1
2.0 × 104
8.5 × 104
1.7 × 105
3.5 × 105
492c
10.33
11.46
11.99
14.94
11.0d
12.8d
0.082
0.19
8.5
0.15
0.46
0.81
3.4
4.7
4
3.6
1.5
5
0.29
2.4
0.85
0.20
160
43
6
0.21
3.3
C6F5OH
0.00043
0.0042
1600
160
0.0043
0.016
C6F5CO2H 1410c
aReactions were carried out at 27 °C in CD2Cl2 with 10 mol% of the indicated catalyst, 500 mM
trans-β-nitrostyrene, and 50 mM N-methylindole. bReactions were carried out at 25 °C in CD2Cl2
with 1 mol% of the indicated catalyst, 250 mM cyclopentadiene, and 25 mM methyl vinyl ketone.
Background corrected rate constants for the non-catalyzed processes are given. cThis value comes
from ref. 10. dThis value was taken from ref. 12.
15
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 16 of 36
1
2
3
4
5
6
7
8
9
catalysts are much less effective than 1, 4, 5, and 6.17 Both of these quantities (k and K1:1) span
approximately three orders of magnitude and their log-log plot is linear for all six catalysts that
were examined (i.e., charged and non-charged, Figure 7). An analogous correlation with the 31P
NMR Δδ shifts is not observed indicating that binding to Et3PO is not as universal a predictor of
reactivity, and depends more on the structure of the species being investigated (i.e., compounds
with the same Δδ values can have different acidities and catalytic activities). This is in keeping
with the report by Diemoz and Franz, who found separate correlations for different types of
compounds.12
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4
2
6
0
5
4
-2
1
-4
C6F5CO2H
C6F5OH
-6
-8
5
6
7
8
9
10
11
12
13
14
ln K1:1
Figure 7. A plot of the logarithm of the rate constant for the Diels–Alder reaction of
cyclopentadiene with methyl vinyl ketone vs the logarithm of the 1:1 equilibrium binding constants
with colorimetric sensor 7 for all six catalysts in Table 8. The equation for the line obtained from
a linear least squares fit is ln kDA = 0.934 × ln K1:1 – 11.1, r2 = 0.987.
16
ACS Paragon Plus Environment

Page 17 of 36
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
If one uses the Friedel–Crafts rate constants for the transformation of N-methylindole with
trans-b-nitrostyrene rather than those for the Diels–Alder reaction of cyclopentadiene with methyl
vinyl ketone in the log-log plot displayed in Figure 7, then a parabolic relationship results rather
than a straight line (Figure 8). The maximum rate is obtained with 5 and falls off with more or less
acidic catalysts as given by bigger or smaller K1:1 values, respectively. This indicates that 5
represents an inflection point in the mechanism and the rate determining step shifts from C–C bond
formation with weaker acids to proton abstraction with stronger ones (i.e., steps 1 vs 3 in Scheme
2). In accord with this view, the kinetic isotope effects are 1.0 for 5, < 1.0 for weaker acids and >
1.0 for stronger ones. It also appears that the charged hydroxypyridinium ion salts are all more
acidic than trifluoroacetic acid (CF3CO2H) given that K1:1 = 5,490 and ln K1:1 = 8.61 for this
compound.10,18
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5
2
4
6
1
kH/kD = 1.0
TfOH
C6F5CO2H
kH/kD < 1.0
C6F5CO2H
kH/kD > 1.0
Figure 8. A plot of the logarithm of the rate constant for the Friedel–Crafts reaction of N-
methylindole with trans-β-nitrostyrene versus the logarithm of the 1:1 equilibrium binding
constants with colorimetric sensor 7 (filled circles). A least squares analysis of the data gives
ln kDA = –0.169 × (ln K1:1)2 + 4.15 × ln K1:1 – 26.8, r2 = 0.996, and this quadratic equation was used
to obtain an estimated value for ln K1:1 for triflic acid (open circle) of 16.6.
17
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 18 of 36
1
2
3
4
CONCLUSION
5
6
7
8
9
Incorporation of electron withdrawing substituents into charge-containing catalysts proved to
be an effective strategy for increasing their acidities in a nonpolar solvent (CH2Cl2) as measured
by 1:1 equilibrium association constants with 7, 31P NMR shifts with triethylphosphine oxide, and
gas-phase acidities. Of these quantities, the K1:1 values were found to be more broadly suitable for
correlating catalyst activities. Kinetic isotope effects and an inverted parabolic rate profile for the
Friedel–Crafts alkylation of N-methylindole with trans-b-nitrostyrene revealed that the rate
determining step can change from carbon-carbon bond formation to proton transfer to the
conjugate base of the catalyst. This leads to a region where reaction rates decrease with increasing
acidity making this a less useful model transformation for evaluating relative catalyst acidities than
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
previously considered. All of the hydroxypyridinium BArF – salts reported in this work were also
4
found to be more acidic in dichloromethane than trifluoroacetic acid.
EXPERIMENTAL
General. All reaction vessels and NMR tubes were stored at 120 °C for a minimum of 12 h and
allowed to cool to room temperature in a vacuum desiccator over Drierite before use. Syringes and
volumetric flasks were kept in the same desiccator but were not heated. Alumina, molecular sieves,
and calcium sulfate were activated in a kiln at 300 ˚C for at least 24 h.
Acetone, chloroform, and carbon tetrachloride were supplied by Fisher Scientific whereas
ethyl propiolate, 2,2,2-trifluoroacetophenone, and pentafluorobenzoic acid were obtained from
Oakwood Chemical. 1-Chloro-2,4-dinitrobenzene, triethylphosphine oxide, sodium tetrakis[3,5-
bis(trifluoro-methyl)phenyl]borate (NaBArF ), pentafluorophenol, and HCl were purchased from
4
18
ACS Paragon Plus Environment

Page 19 of 36
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
Acros Organics, Alfa Aesar, ArkPharm, Fluka, and VWR Chemical, respectively. Deuterated
solvents were acquired from Cambridge Isotope Laboratories and all other chemicals came from
Sigma-Aldrich. Reagents were used as supplied except for N-methylindole, which was passed neat
through a pipet containing activated alumina and stored under argon. Anhydrous CH2Cl2 was used
as purchased, CD2Cl2 was dried by the addition of activated molecular sieves to the reagent bottle
at least 24 h before use, and all other solvents were dried in the same manner as N-methylindole
and stored over molecular sieves. Charged catalysts were kept under vacuum (0.05–0.1 Torr) for
at least 12 h before use.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
NMR spectra were obtained with 400 and 500 MHz spectrometers. The 1H and 13C chemical
shifts were referenced at δ 5.32 and 54.0 (CD2Cl2), δ 7.26 and 77.2 (CDCl3), δ 2.50 and 39.5
(DMSO-d6), and δ 1.94 and 1.3 (CD3CN), respectively. 2H chemical shifts were referenced relative
to benzene-d6 (7.36 ppm). Triethylphosphine oxide 31P NMR shifts were acquired in CD2Cl2 and
referenced at δ 50.5 through the deuterium lock channel according to the IUPAC unified scale.19
Medium pressure liquid chromatography was performed using an automated flash chromatography
system with silica gel columns. Melting point data were collected in unsealed capillaries and were
uncorrected. Solution and ATR FT-IR spectra were recorded with a liquid cell and a laminated
diamond crystal, respectively. High resolution mass spectra were obtained with an electrospray
ionization - time of flight mass spectrometer using PEG-200 as an internal mass calibrant. UV
spectra were taken using 10 mm quartz cells sealed with PTFE septum caps.
N-Methyl-3-hydroxypyridinium (1). The iodide and BArF – salts of this ion were prepared
4
according to the literature procedure.5d
(E)-N-(2-Carboethoxyvinyl)-3-hydroxypyridinium BArF (2). Concentrated HCl (1.75 mL,
4
21.0 mmol) was added dropwise with stirring to a 50 mL round-bottomed flask containing 3-
19
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 20 of 36
1
2
3
4
5
6
7
8
9
hydroxypyridine (1.00 g, 10.5 mmol) and 25 mL of methanol. After 10 min, the solvent was
removed in vacuo and the resulting solid was dissolved in ethanol (30 mL). Ethyl propiolate (1.60
mL, 1.55 g, 15.8 mmol) was added and this mixture was heated with an oil bath to reflux overnight.
Concentration of the resulting material with a rotary evaporator afforded a residue which was
dissolved in 25 mL of acetone. This solution was stirred overnight at room temperature and the
resulting precipitate was isolated by vacuum filtration to give 845 mg (35%) of crude 2-Cl. Some
of this material (64.7 mg, 0.282 mmol), along with 1.5 mL of CH2Cl2 and 250 mg (0.28 mmol) of
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
NaBArF , were added to a 6 dram vial and stirred overnight at room temperature. The resulting
4
precipitate was removed with a 0.45 µm PTFE membrane and then hexanes were added until the
solution just became cloudy. This mixture was subsequently dried over activated CaSO4 and
filtered through another 0.45 µm PTFE membrane. Removal of the solvent in vacuo gave 253 mg
(85%) of an off-white powder (mp 175–178 ˚C) which was stored in a dry-box under nitrogen. 1H
NMR (400 MHz, CD2Cl2) δ 8.35 (dd, J = 2.2, 1.7 Hz, 1H), 8.28 (ddd, J = 6.0, 1.2, 1.2 Hz, 1H),
8.10 (ddd, J = 8.8, 2.5, 1.0 Hz, 1H), 8.05 (d, J = 13.9 Hz, 1H), 7.99 (dd, J = 8.8, 6.0 Hz, 1H), 7.72
(s, 8H), 7.56 (s, 4H), 6.74 (d, J = 14.0 Hz, 1H), 4.37 (q, J = 7.1 Hz, 2H), 1.36 (t, J = 7.1 Hz, 3H).
13C{1H} NMR (125 MHz, CD2Cl2) δ 162.5, 162.3 (q, 1JB-C = 62 Hz), 158.4, 141.5, 136.0, 135.4,
133.8, 130.2, 130.1, 129.4 (qq, 2JF-C = 39 Hz and 2JB-C = 3.5 Hz), 125.2 (q, 1JF-C = 339 Hz), 123.8,
3
118.1 (septet, JF-C = 4.6 Hz), 63.6, 14.3. IR-ATR 3103, 1704, 1358, 1280, 1121 cm–1. HRMS
(ESI) m/z [M – BArF ]+calcd for C10H12NO3 194.0812, found 194.0829.
4
n-Octyl (E)-3-chloroacrylate. Propiolic acid (3.0 mL, 3.41 g, 48 mmol) was added over 1 min to
a 25 mL round-bottomed flask containing 5 mL (60 mmol) of concentrated HCl. This solution was
then heated with an oil bath to reflux with stirring for 18 h before being diluted with 5 mL of water.
The product was extracted with chloroform (3 × 5 mL) and the combined organic material was
20
ACS Paragon Plus Environment

Page 21 of 36
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
washed with brine (2 × 5 mL) and dried over MgSO4. Concentration in vacuo gave 4.72 g (90%)
of crude (E)-3-chloroacrylic acid as an odorous, lustrous brown solid (flakes). A portion of this
material (450 mg, 4.1 mmol), along with 700 µL (580 mg, 4.5 mmol) of n-octanol, and 160 mg
(0.84 mmol) of p-toluenesulfonic acid monohydrate, were placed in a 25 mL round-bottomed flask
equipped with a Dean-Stark trap and dissolved in 7.5 mL of toluene. This solution was heated with
an oil bath to 130 ˚C for 24 h, during which time water was removed from the system.
Concentration of the reaction mixture with a rotary evaporator was followed by addition of 15 mL
of hexanes and filtration of the resulting material. Removal of the solvent in vacuo afforded an oil
that was purified by MPLC with 5% ethyl acetate in hexanes to give 829 mg (92%) of a clear, pale
yellow liquid. The 1H NMR spectrum of this previously reported compound matched the literature
data.20 1H NMR (400 MHz, CDCl3) δ 7.35 (d, J = 13.5 Hz, 1H), 6.24 (d, J = 13.5 Hz, 1H), 4.15 (t,
J = 6.7 Hz, 2H), 1.65 (p, J = 7.6 Hz, 2H), 1.39–1.22 (m, 10H), 0.88 (t, J = 6.7 Hz, 3H).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(E)-N-(2-Carbooctoxyvinyl)-3-hydroxypyridinium chloride (3-Cl). In a 25 mL round-
bottomed flask, 779 mg (3.56 mmol) of n-octyl (E)-3-chloroacrylate and 271 mg (2.85 mmol) of
3-hydroxypyridine were dissolved in 5 mL of acetonitrile. This solution was heated to reflux with
an oil bath with stirring for 24 h, and then an additional 162 mg (0.741 mmol) of n-octyl (E)-3-
chloroacrylate was added. The reaction was heated for another 12 h and was subsequently
concentrated with a rotary evaporator. Ethyl acetate (10 mL) was added to the resulting dark brown
oil and the mixture was stirred at room temperature for 4 h. This resulted in precipitation of the
product, which was collected by vacuum filtration to afford 385 mg (43%) of a pale tan powder
1
(mp 132–134 ˚C) that was stored in a dry-box under nitrogen. H NMR (400 MHz, CDCl3)
δ 9.31 (s, 1H), 8.59 (s, 1H), 8.33-8.20 (m, 2H), 7.95 (s, 1H), 6.95 (d, J = 13.4 Hz, 1H),
4.23 (t, J = 6.7 Hz, 2H), 1.70 (p, J = 6.8 Hz, 2H), 1.41-1.20 (m, 10H), 0.87 (t, J = 6.8 Hz, 3H).
21
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 22 of 36
1
2
3
4
5
6
7
8
9
13C{1H} NMR (156 MHz, CDCl3) δ 163.3, 159.4, 142.3, 136.1, 132.8, 130.2, 128.9, 121.5, 66.7,
31.9, 29.30, 29.28, 28.6, 25.9, 22.8, 14.2. IR-ATR 2958, 2928, 2855, 2473 (OH···Cl–), 1710, 1582,
1317, 1289, 1185 cm–1. HRMS (ESI) m/z [M – Cl]+ calcd for C16H24NO3 278.1751, found
278.1768.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(E)-N-(2-Carbooctoxyvinyl)-3-hydroxypyridinium BArF (3). Dichloromethane (1 mL) was
4
added to a 6 dram vial containing 565 mg (0.638 mmol) of NaBArF and 200 mg (0.637 mmol) of
4
3-Cl and the solution was stirred at room temperature overnight. Filtration of the resulting mixture
through a 0.45 µm PTFE membrane and removal of the solvent under reduced pressure gave 720
mg (99%) of the product as an orange-brown semisolid. This material degraded with a half-life of
1
approximately 5 days when stored in a dry-box under nitrogen. H NMR (400 MHz, CD2Cl2) δ
8.34 (t, J = 2.0 Hz, 1H), 8.21 (dt, J = 6.0 and 2.0 Hz, 1H), 8.08 (ddd, J = 8.8, 2.5, and 1.0 Hz, 1H),
8.03 (d, J = 14.0 Hz, 1H), 7.89 (dd, J = 8.8 and 6.0 Hz, 1H), 7.75–7.70 (m, 8H), 7.57 (s, 4H), 6.73
(d, J = 14.0 Hz, 1H), 4.30 (t, J = 6.8 Hz, 2H), 1.73 (p, J = 7.4 Hz, 2H), 1.45–1.25 (m, 10H), 0.88
(t, J = 7.0 Hz, 3H). 13C{1H} NMR (125 MHz, CDCl3) 162.1, 161.8 (q, 1JB-C = 62 Hz), 158.1, 140.9,
2
135.1, 134.9, 132.5, 129.4, 129.2 (qq, JF-C = 39 Hz and 2JB-C = 3.8 Hz), 129.3, 124.6 (q, 1JF-C
=
338 Hz), 123.3, 117.7, 67.6, 31.8, 29.21, 29.17, 28.4, 25.8, 22.7, 14.1. IR-ATR 3084, 2933, 2861,
1706, 1582, 1355, 1277, 1127 cm–1. HRMS (ESI) m/z [M – BArF ]+ calcd for C16H24NO3
4
278.1751, found 278.1759.
N-(2,4-Dinitrophenyl)-3-hydroxypyridinium BArF (4). In a 50 mL round-bottomed flask, 320
4
mg (3.4 mmol) of 3-hydroxypyridine and 690 mg (3.4 mmol) of 1-chloro-2,4-dinitrobenzene were
dissolved in 20 mL of acetonitrile and heated to reflux with an oil bath with stirring overnight.
After cooling to room temperature, the resulting precipitate was collected by vacuum filtration and
washed with acetonitrile to give 959 mg (96%) of crude 4-Cl as a light tan powder. A portion of
22
ACS Paragon Plus Environment

Page 23 of 36
The Journal of Organic Chemistry
1
2
3
4
this material (112 mg, 0.377 mmol) was dissolved in 5 mL of water with 300 mg (0.34 mmol) of
5
6
NaBArF in a 6 dram vial. Dichloromethane (5 mL) was added and the mixture was stirred
4
7
8
9
vigorously for 3 h. The clear yellow organic layer was subsequently washed with brine (2 × 5 mL)
and dried over MgSO4. It was then triturated with hexanes until the solution just became cloudy
whereupon it was passed through a 0.45 µm PTFE membrane. Removal of the solvent in vacuo
gave 328 mg (86%) of a light tan powder (mp 54–56 ˚C) which was stored in a dry-box under
nitrogen. 1H NMR (400 MHz, CD2Cl2) δ 9.30 (d, J = 2.4 Hz, 1H), 8.85 (dd, J = 8.6 and 2.4 Hz,
1H), 8.25–8.19 (m, 2H), 8.14 (ddd, J = 6.0, 1.4, and 1.3 Hz, 1H), 8.07 (dd, J = 8.6 and 5.9 Hz,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
13
1H), 7.93 (d, J = 8.6 Hz, 1H), 7.72 (s, 8H), 7.55 (s, 4H). C{1H} NMR (125 MHz, CD2Cl2) δ
1
162.3 (q, JB-C = 62 Hz), 159.7, 150.7, 143.7, 138.6, 136.7, 135.6, 135.4, 133.4, 131.1, 130.7,
130.0, 129.4 (qq, 2JF-C = 39 Hz and 2JB-C = 3.7 Hz), 125.2 (q, 1JF-C = 339 Hz), 123.6, 118.1 (septet,
3JF-C = 5.0 Hz). IR-ATR 3114, 1694, 1611, 1553, 1481, 1354, 1279 cm–1. HRMS (ESI) m/z [M –
BArF ]+ calcd for C11H8N3O5 262.0458, found 262.0467.
4
(E)-N-(2-Cyanovinyl)-3-hydroxypyridinium BArF (5). A solution of 5.0 mL (4.1 g, 76 mmol)
4
of acrylonitrile in 5 mL of chloroform was prepared in a 25 mL round-bottomed flask fitted with
a condenser topped with an addition funnel. This setup was covered in aluminum foil to avoid
exposure to light and heated to reflux with an oil bath. Bromine (9.0 mL, 28g, 153 mmol) was then
added dropwise over 1 h and the reaction mixture was refluxed for an additional 3 h. It was
subsequently washed with 10% aqueous Na2S2O3 (3 × 5 mL) and brine (2 × 5 mL), and then dried
with Na2SO4. Rotary evaporation of the organic material afforded 16.1 g (98%) of crude 2,3-
dibromopropionitrile as a golden oil. A portion of this material (470 mg, 2.2 mmol) was dissolved
in 7 mL of methanol along with 420 mg (4.4 mmol) of 3-hydroxypyridine in a 25 mL round-
bottomed flask. This mixture was heated to reflux with an oil bath for 3 h and concentrated with a
23
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 24 of 36
1
2
3
4
5
6
7
8
9
rotary evaporator to give a dark brown viscous oil, which was combined with 7 mL of isopropanol
and stirred vigorously at room temperature overnight. The precipitate that formed was isolated by
vacuum filtration, dissolved in 2 mL of methanol, and triturated into toluene (20 mL) with stirring.
Vacuum filtration of the resulting solid gave 288 mg (57%) of crude 5-Br as a brown powder.
Some of this material (67.2 mg, 0.290 mmol) was placed in a 6 dram vial along with 1 mL of water
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
with 210 mg (0.24 mmol) of NaBArF . Dichloromethane (1 mL) was added and the mixture was
4
stirred overnight at room temperature. The organic fraction was subsequently washed with brine
(2 × 2 mL) and dried over Na2SO4. At a later date, it was triturated with hexanes until the solution
just became cloudy, dried over activated CaSO4, and passed through a 0.45 µm PTFE membrane.
Removal of the solvent under reduced pressure afforded 221 mg (91%) of a brown, slightly
1
lustrous powder (mp 157–160 ˚C) which was stored in a dry-box under nitrogen. H NMR (400
MHz, CD2Cl2) δ 8.29 (dd, J = 2.0 and 1.4 Hz, 1H), 8.23 (d, J = 6.0 Hz, 1H), 8.17 (ddd, J = 9.1,
2.7, and 0.9 Hz, 1H), 8.04 (dd, J = 8.8 and 6.0 Hz, 1H), 7.87 (d, J = 14.4 Hz, 1H), 7.72 (s, 8H),
7.56 (s, 4H), 6.38 (d, J = 14.3 Hz, 1H). 13C{1H} NMR (125 MHz, CD2Cl2) δ 162.3 (q, 1JB-C = 62
Hz), 158.8, 145.5, 137.1, 135.4, 133.4, 130.6, 129.7, 129.4 (qq, 2JF-C = 39 Hz and 2JB-C = 3.6 Hz)
125.2 (q, 1JF-C = 339 Hz), 118.1 (septet, 3JF-C = 5.0 Hz), 112.1, 104.8. IR-ATR 3208, 1355, 1278,
1123 cm–1. HRMS (ESI) m/z [M – BArF ]+ calcd for C8H7N2O 147.0553, found 147.0540.
4
N-(2,4-Dinitrophenyl)-4-hydroxypyridinium BArF (6). A 50 mL round-bottomed flask was
4
loaded with 200 mg (2.1 mmol) of 4-hydroxypyridine, 492 mg (2.4 mmol) of 1-chloro-2,4-
dinitrobenzene, and 10 mL of acetonitrile. This solution was heated to reflux with an oil bath with
stirring overnight, and the resulting precipitate was collected by vacuum filtration and washed with
acetonitrile to give 588 mg (94%) of 6-Cl as a pale pink powder. In a 6 dram vial, 100 mg (0.336
mmol) of this material was dissolved in 3 mL of methanol along with 298 mg (0.336 mmol) of
24
ACS Paragon Plus Environment

Page 25 of 36
The Journal of Organic Chemistry
1
2
3
4
NaBArF and the mixture was stirred vigorously. After 30 min, 3 mL of CH2Cl2 was added to the
4
5
6
7
8
vial and stirring continued overnight. The resulting precipitate was removed with a 0.45 µm PTFE
syringe filter and hexanes were added dropwise to the solution until it just became cloudy. Drying
with activated CaSO4, followed by a second filtration and concentration with a rotary evaporator,
afforded 340 mg (90%) of an off-white powder (mp 125–127 ˚C). 1H NMR (400 MHz, CD2Cl2)
δ 9.27 (d, J = 2.5 Hz, 1H), 8.82 (dd, J = 8.6 and 2.5 Hz, 1H), 8.27–8.23 (m, 2H), 7.88 (d, J = 8.6
Hz, 1H), 7.72 (s, 8H), 7.56 (s, 4H), 7.50–7.46 (m, 2H). 13C{1H} NMR (125 MHz, CD2Cl2) δ 174.4,
162.3 (q, 1JB-C = 62 Hz), 160.0, 150.5, 145.6, 138.3, 135.4, 131.3, 131.1, 129.4 (qq, 2JF-C = 39 Hz
and 2JB-C = 3.6 Hz), 125.2 (q, 1JF-C = 338 Hz), 123.5, 118.1 (septet, 3JF-C = 4.7 Hz), 116.8. IR-ATR
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3676, 3086, 1643, 1610, 1552, 1355, 1278, 1119 cm–1. HRMS (ESI) m/z [M – BArF ]+ calcd for
4
C11H8N3O5 262.0458, found 262.0468.
3-Deuterio-N-methylindole.15 Using CH2Cl2 as an eluent, 1 mL (8.0 mmol) of N-methylindole
was passed through a small column of activated alumina. This solution was then concentrated with
a rotary evaporator and the resulting material was stored in a 6 dram vial under argon.
Concurrently, a 50 mL two-necked round-bottomed flask equipped with a reflex condenser was
filled with D2O (10 mL) and heated to 110 ˚C with an oil bath for 1 h while the system was flushed
with argon. The solvent was then removed via syringe and the N-methylindole was added along
with 15 mL of D2O. This mixture was heated to 80 ˚C with vigorous stirring under argon for 24 h,
then stirred for an additional 96 h at room temperature. The product was subsequently extracted
with CH2Cl2 (1 × 15 mL) and the organic layer was dried over MgSO4. The solution was
concentrated in vacuo and the resulting neat liquid was passed through a plug of activated alumina
to give 657 mg (62%) of a pale mint green oil (99% D by 1H NMR) which was stored in a glovebox
25
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 26 of 36
1
2
3
4
5
6
7
8
9
under nitrogen. 1H NMR (400 MHz, CD2Cl2) δ 7.60 (d, J = 7.9 Hz, 1H), 7.35 (d, J = 8.2 Hz, 1H),
7.21 (t, J = 7.2 Hz, 1H), 7.11–7.06 (m, 2H), 6.47 (d, J = 3.1 Hz, 0.01 H), 3.79 (s, 3H).
General Friedel–Crafts kinetics procedures. To a 1 mL volumetric flask, 74.6 mg (0.500 mmol)
of trans-β-nitrostyrene or 70.2 µL (87.0 mg, 0.500 mmol) of 2,2,2-trifluoroacetophenone was
added along with 6.2 µL (6.5 mg, 0.050 mmol) of N-methylindole and 0.005 mmol of a charged
catalyst or 10 µL of a 0.5 M solution of a non-charged catalyst in CH2Cl2. This mixture was diluted
with CD2Cl2 to the mark and mixed by inverting the flask twice. It was then transferred to an NMR
tube which was capped and sealed with PTFE tape. The first NMR spectrum was collected within
5 min of the preparation of the solvent (t0) and the NMR tube was maintained at 27 ˚C in the NMR
probe (for KIE experiments) or a water bath (for all other experiments) between acquisition of
subsequent spectra. When trans-β-nitrostyrene was used as the electrophile, reaction progress was
monitored using signals at δ 6.47 (N-methylindole) and δ 5.17, 5.08 and 4.98 (Friedel–Crafts
product) or δ 3.79 (3-d-N-methylindole) and δ 3.74 (deuterated product) for the KIE experiments.
With PhCOCF3 as the electrophile, the methyl signals at 3.79 (N-methylindole), 3.83 (mono
addition product), and 3.74 ppm (bis addition product) were used. A pseudo-first-order kinetic
model was employed to determine half-lives and rate constants (Tables S1–S4, Figures S1 and
S2).
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
General procedure for the reaction of cyclopentadiene with methyl vinyl ketone.
Cyclopentadiene was generated by distillation from its dimer and stored at –78 ˚C under argon for
up to 3 h. Methyl vinyl ketone (MVK) was distilled within 1 week of when it was used, and was
kept at 4 ˚C under argon. A 1 mL volumetric flask was filled with MVK (20.3 µL, 17.5 mg, 0.250
mmol), 0.0025 mmol of a charged catalyst or 10 µL of a 0.25 M solution of a non-charged catalyst
in CH2Cl2, and sufficient CH2Cl2 before being mixed by inverting the stoppered flask twice. A
26
ACS Paragon Plus Environment

Page 27 of 36
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
portion of this solution (100 µL) was added to a separate 1 mL volumetric flask along with 21.0
µL (16.5 mg, 0.250 mmol) of cyclopentadiene and filled with CD2Cl2 before mixing this mixture
by inverting the flask twice. This solution was then transferred to an NMR tube which was capped
and sealed with PTFE tape. The first NMR spectrum was collected within 5 min of solvent addition
(t0) and the NMR tube was maintained at 25 ˚C in the NMR probe between acquisition of
subsequent spectra. Reaction progress was monitored using signals at 2.24 (MVK), 2.08 (endo),
and 2.17 ppm (exo) (Table S5, Figure S3). A pseudo-first-order kinetic model was used to
determine half-lives and rate constants.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Solution IR procedure. A 6 dram vial was loaded with 3 (29.2 mg, 25.6 µmol), CCl4 (5 mL), and
CD3CN (50 µL) and this mixture was stirred until all of the catalyst dissolved. A separate solution
of 5 mL of CCl4 and 50 µL of CD3CN with no catalyst was also prepared and a background
spectrum of this mixture was recorded in a 10 mm liquid cell that was stored in a dry-box under
nitrogen. This cell was subsequently emptied and purged with nitrogen for 1 min before being
reloaded with the sample solution. A new spectrum was obtained and background corrected.
General procedure for UV-vis titrations. 7-Methyl-2-phenylimidazo[1,2-a]pyrazin-3(7H)-one
(7), was kept under vacuum (0.05–0.1 torr) for at least 12 h before use. Three volumetric flasks
were employed to prepare the required solutions as follows: A 1 mL volumetric flask was filled
with 1.0–1.5 mg (4.4–6.7 µmol) of 7 and sufficient CH2Cl2 to afford solution A. In a 5 mL
volumetric flask, 25 µL of solution A was diluted with CH2Cl2 to the line to give solution B. A
separate 5 mL volumetric flask was used to prepare solution C, which consisted of 25 µL of
solution A, 25–30 molar equivalents of catalyst relative to 7, and sufficient CH2Cl2 to fill the flask
to the line. Background UV-vis spectra of CH2Cl2 from 300–950 nm were collected before each
sample spectrum in a 10 mm quartz cuvette with a screw-on PTFE septum cap. For the titrations,
27
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 28 of 36
1
2
3
4
5
6
7
8
9
2 mL of solution B were placed in a separate cuvette and the UV-vis spectrum was recorded.
Subsequently, 10 µL of solution C was added via syringe and the solution was swirled for 15 s and
a new spectrum was obtained. Further mixing of the solution for an additional 10 s was carried out
and the spectrum was re-recorded. This was repeated, if necessary, until consecutive spectra
differed by no more than 1 nm in λmax and 0.010 absorbance units. Additional aliquots of solution
C were added and the spectra were recorded in the same way until λmax did not change over three
consecutive additions (1, 2, 4, and 5) or until at least 20 molar equivalents of catalyst relative to
the 7 had been added (6). Equilibrium binding constants (K) were determined using absorbance
values at 440, 470, 500, and 530 nm by carrying out non-linear fits of the resulting binding
isotherms using the Supramolecular BindFit app (app.supramolecular.org/bindfit).21 In each case,
1:1. 2:1, and 1:2 associations were explored, though all of the data are in accord with 1:1 binding
except for the results with 6, which are consistent with 1:2 binding (Table S6). While λmax for the
1:1 complex can be obtained directly in the former instances, it cannot in the latter case.
Consequently, λmax(1:1) was determined by plotting χ1:1/λmax versus χ1:1, where χ1:1 is the mole
fraction of the 1:1 complex as obtained from the BindFit output file.22 The reciprocal of the sum
of the slope and y-intercept from the resulting linear least squares fit afforded λmax(1:1). These values
are all within 1.0 nm of the directly-observed, leveled-off determinations with 1, 2, 4, and 5.
Catalyst titration procedures with triethylphosphine oxide. Triethylphosphine oxide was
stored and weighed in a glovebox under nitrogen. For the titration with 1, 2.1–2.4 mg (15.7–17.9
µmol) of Et3PO was dissolved in minimal CH2Cl2, transferred to a 1 mL volumetric flask, and
diluted to the mark (solution A). In a separate 1 mL volumetric flask, 1 (9.4–10.0 mg,
9.7–10.3 µmol) was added and diluted to the line with CH2Cl2 (solution B). An NMR tube was
then filled with 100 µL of solution A and 400 µL of CD2Cl2. An initial NMR spectrum was
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
28
ACS Paragon Plus Environment

Page 29 of 36
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
obtained to establish the 31P chemical shift of Et3PO. Subsequently, 80 µL of solution B was added
via microsyringe to the NMR tube, which was then inverted twice before obtaining another
spectrum. This process was repeated until the 31P signal shifted by no more than 0.1 ppm in three
consecutive spectra (Table S7, Figure S15). For the other catalysts, full titrations were not carried
out. In these cases, 5.0 mg (37.3 µmol) of Et3PO were placed in a 1 mL volumetric flask which
was filled to the line with CH2Cl2. An NMR tube was loaded with 135 µL of this solution and 315
µL of CD2Cl2 to obtain the initial 31P NMR signal of Et3PO. Subsequently, 7.5 µmol of the selected
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
31
catalyst and 500 µL of CH2Cl2 were added with mixing and a second P NMR spectrum was
recorded to obtain the change in the chemical shift.
Computations. Calculations were performed at the Minnesota Supercomputing Institute using
Gaussian 1623 with the B3LYP functional and the 6-31G+(d,p) basis set.24 Full geometry
optimizations and subsequent vibrational frequency determinations were carried out and all of the
given stationary points correspond to energy minima with no negative eigenvalues (Table S8).
Unscaled vibrational frequencies were used to obtain zero-point energies, thermal corrections to
the enthalpies at 298 K, and entropies. Low frequency modes that contribute more than ½RT to
the enthalpy correction were replaced with ½RT.
Supporting Information Available: Kinetic, UV-vis titration, and Ph3PO 31P chemical shift data
are provided along with plots of ln kDiels-Alder vs ln K1:1 and DG˚acid, NMR spectra, computed
structures and energies, the complete citation to ref. 23, and a reference. This material is available
free of charge via the internet at http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*E-mail: kass@umn.edu.
29
ACS Paragon Plus Environment