NMR Quantification of Hydrogen-Bond-Activating Effects for Organocatalysts including Boronic Acids

Article abstract of DOI:10.1021/acs.joc.8b02389

The hydrogen-bonding activation for 66 organocatalysts has been quantified using a ³¹P NMR binding experiment with triethylphosphine oxide (TEPO). Diverse structural classes, including phenols, diols, silanols, carboxylic acids, boronic acids, and phosphoric acids, were examined with a variety of steric and electronic modifications to understand how the structure and secondary effects contribute to hydrogen-bonding ability and catalysis. Hammett plots demonstrate high correlation for the Δδ ³¹P NMR shift to Hammett parameters, establishing the ability of TEPO binding to predict electronic trends. Upon correlation to catalytic activity in a Friedel-Crafts addition reaction, data demonstrate that ³¹P NMR shifts correlate to catalytic activity better than pK_a values. Boronic acids were investigated, and ³¹P NMR binding experiments predicted strong hydrogen-bonding ability, for which catalytic activity was confirmed, resulting in the greatest rate enhancement observed in the Friedel-Crafts addition of all organocatalysts studied. A detailed investigation supports that boronic acid activation proceeds through hydrogen-bonding interactions and not coordination with the Lewis acidic boron center. Using ³¹P NMR spectroscopy offers a simple and rapid tool to quantify and predict hydrogen-bonding abilities for the design and applications of new organocatalysts and supramolecular synthons.

Full text of DOI:10.1021/acs.joc.8b02389

Subscriber access provided by YORK UNIV
Featured Article
NMR Quantification of Hydrogen-bond Activating
Effects for Organocatalysts including Boronic Acids
Kayla M Diemoz, and Annaliese K. Franz
J. Org. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.joc.8b02389 • Publication Date (Web): 05 Dec 2018
Downloaded from http://pubs.acs.org on December 5, 2018
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 37
The Journal of Organic Chemistry
1
2
3
4
5
NMR Quantification of Hydrogen-Bond Activating Effects for Organocatalysts Including
Boronic Acids
6
7
Kayla M. Diemoz, Annaliese K. Franz*
8
9
Department of Chemistry, One Shields Avenue, University of California, Davis CA, 95616
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
ABSTRACT: The hydrogen-bonding activation for 66 organocatalysts has been quantified using a 31P
NMR binding experiment with triethylphosphine oxide (TEPO). Diverse structural classes, including
phenols, diols, silanols, carboxylic acids, boronic acids, and phosphoric acids, were examined with a variety
of steric and electronic modifications to understand how the structure and secondary effects contribute to
hydrogen-bonding ability and catalysis. Hammett plots demonstrate high correlation for the Δδ 31P NMR
shift to Hammett parameters establishing the ability of TEPO binding to predict electronic trends. Upon
correlation to catalytic activity in a Friedel-Crafts addition reaction, data demonstrate that 31P NMR shifts
correlate to catalytic activity better than pKa values. Boronic acids were investigated and 31P NMR binding
experiments predicted strong H-bonding ability, for which catalytic activity was confirmed, resulting in the
greatest rate enhancement observed in the Friedel-Crafts addition of all organocatalysts studied. A detailed
investigation supports that boronic acid activation proceeds through hydrogen-bonding interactions and not
31
coordination with the Lewis acidic boron center. Using P NMR spectroscopy offers a simple and rapid
tool to quantify and predict H-bonding abilities for the design and applications of new organocatalysts and
supramolecular synthons.
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 2 of 37
1
2
3
4
5
6
INTRODUCTION
7
8
9
Hydrogen-bond donors continue to be of interest for supramolecular synthons and catalyst
alternatives to metal-based Lewis acid catalysts.1 Compared to metal-based catalysts, organocatalysts do
not require sensitive air-free techniques and are often recoverable for a more environmentally friendly
option. The steric and electronic properties of organocatalysts can be tunable, however, quantifying H-
bonding abilities based on properties such as steric and electronic effects is far from complete. While
computational and experimental studies have been used to model and predict hydrogen-bond strength,
hydrogen bonds can be difficult to model due to their covalent character and considerable orientation
preference.2 Further understanding of hydrogen-bonded complexes affords insight into self-assembly and
molecular recognition with applications in both natural3 and synthetic systems.4 Systematic quantitative
methods to predict and understand hydrogen-bonding can contribute to greater understanding of self-
assembly and catalysis for many new classes of H-bond donors.
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
Current experimental methods for quantifying hydrogen-bonding ability have had success in
predicting catalytic activity better than pKa.5 Kozlowski reported a method for quantifying the hydrogen-
bonding ability of diverse catalysts based on detection of hydrogen-bonding strengths using a colorimetric
sensor by UV-vis spectroscopy (Figure 1).5a, 5b This study features a comparison of several classes of
hydrogen-bond donors and demonstrates strong correlation of catalytic activity with 1max measured using
UV-Vis spectra. More recently, this method has also been applied in drug design to quantify hydrogen-
bonding ability of carboxylic isosteres to understand the effect when the carboxylic functional group is
replaced.6
ACS Paragon Plus Environment

Page 3 of 37
The Journal of Organic Chemistry
1
2
3
4
5
6
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
Figure 1. Quantification of hydrogen-bonding strength.
We envisioned using 31P NMR spectroscopy with a commercially available phosphine oxide probe
as an alternative method to quantify hydrogen-bonding ability for a wide range of functional groups. Using
phosphine oxides as a 31P NMR probe was originally developed by Gutmann to quantify the Lewis acidity
of a variety of solvents.7 Beckett further demonstrated the method to measure the Lewis acidity of boron
Lewis acids and correlate the binding data to catalytic activity in an epoxide polymerization reaction.8 Since
then, the Gutmann-Beckett method using 31P NMR spectroscopy has shown to be a rapid and valuable
method to quantify and compare Lewis acid properties.9 More recently, Schreiner and Hilt reported a
limited study using tri-n-butylphosphine oxide as a 31P NMR probe to evaluate the hydrogen-bonding
capabilities of thioureas.5c
The systematic quantification of binding properties for a wide range of H-bond donor and BrØnsted
acidic species and the ability to predict catalytic activity is important for understanding and developing new
catalysts. We envisioned being able to study new catalysts quickly to gain insight into hydrogen-bonding
ability of non-traditional functional groups and predicting potential for catalytic activity. For example,
silanols and boronic acids have found use in organic synthesis as well as medicinal chemistry10 but remain
relatively unexplored for their H-bonding ability.11-12 Both silanols and boronic acids are known to self-
associate through hydrogen-bonding networks in the solid state and have been utilized in crystal
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 4 of 37
1
2
3
4
5
6
7
8
9
engineering.13 However, the use of boronic acids for H-bonding in molecular assembly14 and H-bonding
catalysis11 has only been preliminary explored. Silanols are a new class of H-bonding catalysts15 and are
relevant for medicinal chemistry because sila-substitution in biologically active molecules can often lead
to improved pharmacokinetic properties.16
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
Here we demonstrate the ability of 31P NMR shifts to quantify and predict hydrogen-bonding
strength for a wide variety of molecules with functional groups relevant for organocatalysis and
supramolecular assembly (Figures 1, 2). Changes in 31P NMR shifts upon binding to triethylphosphine
oxide (TEPO) have been examined to quantify trends in electronic and steric effects that are significant for
H-bonding interactions. Hydrogen-bond donors studied include phenols (1-6), other alcohols (7-9), benzoic
acids (10-12), silanols (13-15), thioureas and ureas (16-19), squaramide 20, tetrazole 21, phosphoric acids
(22) and boronic acids (23-31). To correlate the downfield 31P NMR shift to catalytic activity, rate constants
were measured for the addition of N-methylindole to 4-trifluoromethyl-trans-β-nitrostyrene using in situ
19F NMR spectroscopy. We have also demonstrated the use of TEPO as a probe to predict hydrogen-
bonding ability of various boronic acids and have compared boron species to elucidate the activation mode
for catalysis with boronic acid.
ACS Paragon Plus Environment

Page 5 of 37
The Journal of Organic Chemistry
1
2
3
4
5
6
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
Figure 2. Organocatalysts studied.
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 6 of 37
1
2
3
4
RESULTS AND DISCUSSION
5
6
Method Development and Validation
7
8
9
Based on the wide variety of functional groups with hydrogen-bond donors, it was important to
develop a method that would quantify H-bonding abilities for functional groups with both weak and strong
hydrogen bonding.5c Initial experiments revealed that upon complexing TEPO with a series of diverse
hydrogen-bond donors (i.e. hexafluoro-2-propanol (7), 2-chlorobenzoic acid (10c), and 4-nitrophenol (1a)
only one distinct peak is present in each 31P NMR spectrum indicating a rapid equilibrium between the
bound and unbound TEPO species (Figure 3A). Measurements were taken by titrating 0.5-8 equivalents of
catalyst to discover where saturation occurs, demonstrated by no additional shifting of the TEPO peak
(Figure 3B). Saturation consistently occurs at or before four equivalents of H-bond donor (Figure 3B),
therefore, using three equivalents of HBD was identified as an effective approximation of the saturation
point that also minimizes the material required. Titration experiments were also performed with
triphenylsilanol (13b) and diphenylthiourea (16a) where 3.0 equivalents was consistently identified as an
approximation of the saturation point (see Supporting Information), demonstrating the estimation holds
across a variety of catalysts classes. The concentration of HBD was shown to influence the binding with
TEPO, with higher shifts being observed at higher concentrations. The binding stoichiometry of TEPO and
the HBD was investigated through a Job plot analysis and a 1:1 binding evening was observed (see
Supporting Information).
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
(A)
ACS Paragon Plus Environment

Page 7 of 37
The Journal of Organic Chemistry
1
2
3
4
(B)
12
10
8
5
6
7
8
9
6
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-nitrophenol (1a)
4
2-chlorobenzoic acid (10c)
hexafluoro-2-propanol (7)
2
0
0
2
4
6
8
Catalyst Equivalents
Figure 3. A) Examples of 31P NMR spectra showing chemical shift changes upon TEPO binding to
HBDs. B) Titration experiments for three different HBDs with TEPO to assess binding equilibrium and
optimum equivalents of HBD.
A Hammett plot was constructed with the Δδ 31P NMR shifts to validate the ability of TEPO to
accurately assess electronic factors that contribute to hydrogen-bond strength.17 A series of phenols (1a-d),
benzoic acids (10a-e), and boronic acids (23a-e) with varying electronic effects were studied to measure
the correlation of the Δδ 31P NMR shift to the Hammet values. For all three classes of molecules, the Δδ
31P NMR shifts afforded a linear relationship (R2 > 0.99) with the Hammett values (Figure 4). The high
31
linear correlation for all three classes validates the use of P NMR shifts upon binding TEPO as a rapid
method to predict electronic trends that effect the strength of the HBD. The high correlation of ortho-
substituted benzoic acids is especially promising as potential steric effects at the ortho position can be
difficult to predict.18,19
2
1.5
1
0.5
0
0
10
20
30
-0.5
Δδ 31P NMR (ppm)
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 8 of 37
1
2
3
4
5
6
7
Figue 4. Correlation of σ parameters for para-phenols (orange), ortho-
benzoic acids (grey),20 and para-benzene boronic acids (blue) to ∆ 31P
NMR (ppm) upon addition to TEPO.
8
9
Discussion of Catalyst Classes
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
The linear correlation of the Δδ 31P NMR shifts to Hammet  values for phenols, benzoic acids and
boronic acids was encouraging for utilizing TEPO as a probe to quantify hydrogen-bonding ability across
a wide range of organocatalysts. We proceeded to measure the Δδ 31P NMR shift for TEPO complexation
with 66 different hydrogen-bond donors containing diverse functional groups and various steric and
electronic effects to quantify trends within a catalyst class (Table 1 and Figure 5). 31P NMR shifts can be
used to calculate an acceptor number (AN) that can be correlated with the strength of hydrogen-bonding
ability.7a In this study, acceptor numbers of HBDs and BrØnsted acids range from 15 to 75 (in CD2Cl2
solvent), compared to those previously observed by Gutmann from 20 (CH2Cl2) – 126 (CH3SO3H) (in
toluene).7b
ACS Paragon Plus Environment

Page 9 of 37
The Journal of Organic Chemistry
1
2
3
4
Table 1. Measured 31P NMR shifts and krel for HBDs studied
5
6


(ppm)c
pKa
H2O
pKa
DMSO
catalysta
ANd
krel
(ppm)b
7
8
9
1a
1b
59.8
57.5
9.5
7.2
41.5
36.5
7.1523
9.3625
10.824
--
6.36
3.69
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
1c
55.7
56.1
54.0
55.5
59.9
61.3
56.3
56.5
55.2
53.5
53.7
50.6
57.5
51.3
52.4
61.8
60.2
59.3
58.2
57.5
60.5
63.1
51.5
53.1
53.4
53.0
57.4
58.3
55.7
59.3
59.8
5.4
5.8
3.7
5.2
9.6
11.0
6.0
6.2
4.9
3.2
3.4
0.3
7.2
1.0
2.1
11.5
9.9
9.0
7.9
7.2
10.2
12.8
1.2
2.8
3.1
2.7
7.1
8.0
5.4
9.0
9.5
32.5
33.4
28.7
32.0
41.8
44.9
33.8
34.3
31.4
27.6
28.1
21.2
36.5
22.8
25.2
46.0
42.4
40.4
38.0
36.5
43.1
48.8
23.2
26.7
27.4
26.5
36.2
38.2
32.5
40.5
41.5
10.2126
10.2326
--
--
1.86
nd
1d
1e
1824
--
nd
2a
8.4425
9.3426
5.5327
9.3428
9.6326
10.2829
--
--
0.74
6.05
6.88
1.83
2.19
1.36
0.06
0.36
0
2b
3
--
--
4
--
5
--
6a
12.9830
6b
6c
--
--
9.6830
6d
7
--
--
9.331
--
--
2.47
nd
8
17.032
9
--
--
0
10a
10b
10c
10d
10e
11e
12
2.1726
--
8.6620
7.1
--
4.89
2.56
nd
2.9026
4.226
3.9126
2.833
1.7526
~1234
--
9.720
1120
11.0720
nd
--
nd
--
8.68
nd
13a
13b
13c
14a
14b
14c
15a
15b
15c
--
16.632
--
0.03
0.09
nd
--
--
--
--
--
2.64
3.64
0.90
4.06
5.00
--
--
--
--
--
--
--
--
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 10 of 37
1
2
3
4


(ppm)c
pKa
H2O
pKa
DMSO
catalyst
ANd
krel
(ppm)b
5
6
7
8
15d
15e
15f
15g
16a
16bf
17
59.5
56.5
60.4
58.4
54.5
58.1
55.1
53.6
56.6
59.4
52.8
57.4
65.0
68.9
64.2
72.8
63.5
61.2
56.5
55.3
70.8
59.0
58.6
55.9
53.5
54.7
55.7
50.4
74.9
9.2
40.9
34.3
42.9
38.5
29.8
37.8
31.2
27.8
34.5
40.7
26.1
36.2
53.0
61.7
51.3
70.3
50.6
44.6
34.3
31.6
65.9
39.8
38.9
32.9
37.6
30.3
32.5
20.8
74.9
--
--
nd
6.2
--
--
nd
10.1
8.1
--
--
5.8
3.74
0.26
nd
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
--
--
4.2
--
13.435
7.8
--
8.535
4.8
--
13.6635
nd
18ae
18be
19e
3.3
--
18.735
nd
6.3
--
16.136
nd
9.1
--
7.537
nd
20e
2.5
--
11.8738
nd
21e
7.1
4.3839
--
--
nd
22a
22b
22c
23a
23b
23c
23d
23e
24
14.7
18.6
13.9
22.5
13. 2
10.9
6.2
3.3740
nd
--
2.42140
1.83
0.61
15.66
9.74
6.98
nd
--
4.22140
--
7.3941
8.642
8.943
9.343
9.343
7.5041
7.8341
--
--
--
--
5.0
--
3.79
14.96
nd
20.5
8.7
--
25a
25b
25b
26a
26bf
26c
26d
27
--
8.3
--
4.06
nd
5.6
9.743
--
--
3.2
--
nd
4.4
--
--
2.49
nd
5.4
--
--
0.1
--
--
nd
24.6
7.0841
--
--
18.79
nd
28g
29f
no 31P NMR signal
--
59.7
57.7
54.5
9.4
7.4
4.2
41.3
36.9
29.8
--
--
7.24
nd
30
7.8944
10.5645
--
31
--
nd
ACS Paragon Plus Environment

Page 11 of 37
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
a Performed with catalyst concentration of 0.1 M in CD2Cl2 and three equiv. of TEPO (unless otherwise
indicated for rows with italicized font and footnotes e and f). b 31P NMR shifts measured after forming a
complex with three equiv. of TEPO; most reported as average of triplicate experiments. c values reported
relative to unbound TEPO (50.3 ppm in CD2Cl2). dAcceptor number (AN) calculated based on the formula
AN = 2.21 * (   ref 7a) eCatalyst insoluble at both 0.1 M and 0.05 M; shift reported using 0.1 M.
fInsoluble at 0.1 M; performed with catalyst concentration of 0.05 M where all material was soluble.
gCatalyst was soluble and sample remained homogeneous, but no signal was observed upon addition of
TEPO (performed in triplicate); 19F NMR signals broadened upon binding and no change was observed
using 11B NMR spectroscopy (i.e. peaks remain broad).
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
Figure 5. Scale comparing classes of hydrogen-bond donors.
Phenols
The hydrogen-bonding ability of a wide variety of structurally diverse phenols (1a-4d) was
quantified using 31P NMR spectroscopy. Phenols were the first reported example of hydrogen-bonding
catalysis where the addition of protic additives, including phenols, was observed to accelerate the rate of a
Diels Alder reaction.21 Since then phenols have been demonstrated to catalyze a variety of reactions.22 The
high correlation (R2 = 0.998) of the Δδ 31P NMR shift upon TEPO binding for para-substituted phenols 1a-
1d to Hammett values demonstrates the ability of TEPO as a 31P NMR probe to quantify electronic effects.
A Δδ range of 3.7 to 11.0 was observed for all phenols evaluated, with pentafluorophenol (3) inducing the
largest downfield 31P NMR shift (Δδ = 11.0 ppm, AN = 44.9).
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 12 of 37
1
2
3
4
5
6
7
8
9
To further examine the utility of 31P NMR shifts to quantify H-bonding ability, phenols with
different steric effects were studied. Ortho-substituted bromophenol 2a resulted in a smaller Δδ 31P NMR
upon binding to TEPO (Δδ = 5.2 ppm) compared to the para-substituted bromophenol 1b (Δδ = 7.0 ppm)
confirming that steric effects reduce hydrogen-bonding interactions. A similar decrease in H-bond strength
was observed when comparing 1-naphthol (4, Δδ = 6.0 ppm) and 2-naphthol (5, Δδ = 6.2 ppm). The
hydrogen-bonding ability of catechol (2b) was also quantified due to its performance as an active hydrogen-
bonding catalyst in the formation of carbonates from CO2.22 The 31P NMR shifts accurately predict an
increased H-bonding ability of catechol (Δδ = 9.3 ppm), relative to other substituted phenols.
Measuring the 31P NMR shift of silyl-substituted phenols provides a direct method to evaluate the
electronic effects of silicon vs carbon substituents. Several examples of silicon-containing bioactive
compounds and utilizing a Si/C switch strategy have been reported to improve the pharmacokinetic
properties of medicinal compounds.16 Therefore, directly comparing the electronic effect of a silyl vs
carbon group is an intriguing opportunity with only limited examples reported previously.46 The hydrogen-
bonding ability of 4-tert-butylphenol (1d) and 4-(trimethylsilyl)phenol (1e) were compared using 31P NMR
spectroscopy as a method to quantify the electron-donating effect of a silicon vs carbon substituent on a
benzene ring (Figure 6). 4-(Trimethylsilyl)phenol (1e) induced a smaller 31P NMR shift (Δδ = 3.7 ppm)
upon binding to TEPO compared to 4-(tert-butyl)phenol (1d, Δδ = 5.8 ppm). A decreased shift observed
with the silyl substitution indicates an increased electron-donating ability compared to the corresponding
carbon substituent. Based on 31P NMR shifts upon TEPO binding, we also observe that the trimethylsilyl
substituent in 1e exhibits a stronger overall electron-donating effect than the 4-methoxy substituent of 4-
methoxyphenol (1c, Δδ = 5.4 ppm).
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
Figure 6. Effect of Si vs C substitution on phenol H-bonding.
ACS Paragon Plus Environment

Page 13 of 37
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
Next, the hydrogen-bonding ability of several BINOL substrates was quantified as the binaphthol
motif has been utilized for hydrogen-bonding catalysis.47 (R)-(+)-1,1’-Bi(2-naphthol) (6a) showed the
largest 31P NMR shift (Δδ = 4.9 ppm) for the BINOL series, falling at the lower end of the range for mono-
substituted phenols. For BINOLs with substituents in the 3- and 3’-positions (6b-d), a weaker hydrogen-
binding interaction was observed compared to 6a. Notably, even within the series of more sterically
crowded BINOLs, electronic effects can still be quantified, with 6c (Ar = 3,5-bis(trifluoromethyl)benzene)
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
affording the largest P NMR shift (Δδ = 3.4 ppm), suggesting a stronger hydrogen-bonding interaction
with TEPO compared to 6b (Br, Δδ = 3.2 ppm) and 6d (2,4,6-triisopropylbenzene, Δδ = 0.3 ppm).
Other Alcohols and Diols
Quantification of hydrogen-bonding ability of other alcohol-containing HBDs was also
performed for comparison to phenols. Hexafluoro-2-propanol (7) invoked a larger 31P NMR shift upon
binding to TEPO (Δδ = 7.8 ppm) compared to most phenols, while triphenylmethanol (8) afforded only a
small shift (Δδ = 1.0 ppm). (4-S-trans)-2,2-Dimethyl-,,’,’-tetral(1-naphthyl)-1,3-dioxolane-4,5-
dimethanol (9) was also evaluated, showing a weak hydrogen-bonding interaction with TEPO (Δδ = 2.1
ppm), consistent with previous reports using a chromophore probe.5a
Benzoic acids
Benzoic acids have applications for catalysis as well as supramolecular assembly due to their
propensity to self-associate into higher-order species.48 Catalytic activity of benzoic acids has been reported
in a variety of transformations including Mannich reactions49 and the alkenylations of imines.50 Benzoic
acids 10-12 invoke large 31P NMR shifts (Δδ = 7.2–12.8 ppm, AN = 36.5–48.8) upon TEPO binding, with
greater downfield shifts compared to phenols that correlate to the lower pKa values of benzoic acids. The
electronic effects of ortho-substituted benzoic acids 10a-e gives high correlation (R2 = 0.992) in a Hammett
plot (Figure 4). 3,5-Dinitrobenzoic acid (11) indicates a stronger binding interaction based on the larger 31P
NMR shift (Δδ = 10.2 ppm), however, reduced solubility was observed in CD2Cl2 such that the magnitude
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 14 of 37
1
2
3
4
5
6
7
8
of the shift may not reach the saturation point favoring the bound species. Pentafluorobenzoic acid (12), the
most acidic benzoic acid investigated, exhibited the greatest 31P NMR downfield shift (Δδ = 12.8 ppm, AN
= 48.8).
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
Silanols
Silanol-containing compounds such as silanediols (14) and 1,3-disiloxanediols (15) have recently
gained attention as both hydrogen-bonding and anion-binding catalysts.15 We measured 31P NMR shifts for
various silanols upon binding TEPO to quantify their H-bonding ability relative to alcohols and other
organocatalysts, and to determine the effect of different SiOH arrangements. For direct comparison to a
carbon analog, we first measured triphenylsilanol (13b) which afforded a greater shift (Δδ = 2.8 ppm)
compared to triphenylmethanol (8, Δδ = 1.0 ppm). This result supports stronger H-bonding interactions and
increased acidity of silanols consistent with the reduced electronegativity of silicon compared to carbon
that enhances -type back-bonding interactions.32, 51 Trialkyl-substituted silanol 13a exhibits a reduced shift
(Δδ = 1.2 ppm) compared to the triphenyl silanol variant, but similar to triphenylmethanol. The H-bonding
ability of naphthyl substrate 13c was also quantified and the shift (Δδ = 3.1 ppm) remains comparable to
the phenyl substrate 13b (Δδ = 2.8 ppm).
Next, we proceeded to quantify H-bonding interactions for silanediols and 1,3-disiloxaanediols,
classes of silanols exhibiting hydroxyl motifs that do not have stable carbon analogs. These silanols also
have opportunities for cooperative H-bonding that can provide more significant enhancements to H-
bonding ability.15d,15i Dinaphthylsilanediol 14b exhibits a significant increase in the P NMR shift (Δδ =
31
7.1 ppm, AN = 36.2) compared to monosilanols. As shown in the case of 14c, electron-withdrawing groups
can be added to further increased the H-bonding ability of silanediols (Δδ = 8.0 ppm, AN = 38.2). In
comparison, the bulky electron-rich dimesitylsilanediol 14a exhibited a significant decrease in the 31P NMR
shift (Δδ = 2.7 ppm).
1,3-Disiloxanediols 15 afforded the largest 31P NMR shifts of the silanol classes investigated,
indicating the strongest H-bonding interactions upon binding with TEPO. The Δδ 31P NMR shifts for 1,3-
ACS Paragon Plus Environment

Page 15 of 37
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
disiloxanediols ranged from 5.4 to 10.1 depending on the substituents (iPr, Ph, Np or 4-fluoro-naphthyl).
Incorporating naphthyl substituents (15c, Δδ = 9.5 ppm; 15d, Δδ = 9.2 ppm) increased shifting slightly
relative to phenyl-substitution (15b, Δδ = 9.0 ppm), a trend also observed for the monosilanol. Electron-
withdrawing substituents can be added, e.g., 1,1-bis(4-fluoronaphthalen-1-yl)-3,3-diphenyldisiloxane-1,3-
diol (15f) to increase the hydrogen-bonding ability, providing the strongest hydrogen-bonding interaction
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
for all silanols studied (Δδ = 10.1 ppm, AN = 42.9). Overall, the P NMR quantification of H-bonding
abilities aligns with catalytic activity trends reported in the literature with monosilanols (13a-c) <
silanediols (14a-c) < 1,3-disiloxanediols (15a-g).15h
Thioureas, Ureas and Squaramide
A variety of functional groups with NHs as HBDs, such as thioureas and squaramides, have also
demonstrated value as organocatalysts. In order to quantify the hydrogen-bonding ability of thioureas,
relative to other HBD functional groups we choose diphenylthiourea (16a) and 1,3-bis(3,5-
bis(trifluoromethyl)phenyl)thiourea (16b).1d Hilt and coworkers have previously evaluated hydrogen-
bonding interactions for a series of thioureas, however no other HBDs were measured and the comparison
relative to other organocatalysts has not been made using this method.5c Here we observed that
31
diphenylthiourea (16a) induces a small change in P NMR shift (Δδ = 4.2 ppm at 0.1 M). Thiourea 16b
with electron-withdrawing groups results in a larger change in 31P NMR shift (Δδ = 7.8 ppm at 0.05 M due
to reduced solubility). Direct quantification of the hydrogen-bonding abilities between thioureas and ureas
(18a-18b and 19) was also performed.35 Diphenylurea 18a can be used to provide a direct comparison to
thiourea 16a and a weaker hydrogen-bonding interaction was observed for urea 16a (Δδ = 3.3 ppm vs 4.2
ppm for 16a). As expected, the incorporation of electron-withdrawing groups on urea 16b increases the
change in 31P NMR shift (Δδ = 6.3 ppm), however, decreased solubility was observed for this urea. Despite
reduced solubility in CD2Cl2, the internally activated urea (19) afforded the largest 31P NMR shift (Δδ = 9.1
ppm, AN = 40.7) compared to other ureas and thioureas investigated, which supports stronger H-bonding
abilities consistent with catalytic activity data reported by Mattson.37
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 16 of 37
1
2
3
4
5
6
7
8
9
To expand the scope of NH hydrogen-bond donors we also investigated other functional groups
that have been studied for their hydrogen-bonding ability: squaramides52 and tetrazoles.53 Squaramide 20
exhibited low solubility in CD2Cl2, which likely accounts for the small 31P NMR shift observed (Δδ = 2.5
ppm). 5-Phenyl-1H-tetrazole (21) afforded a larger 31P NMR shift (Δδ = 7.1 ppm) compared to several
ureas, thioures and the squaramide; however, low solubility was also observed suggesting that the
measurement may not accurately quantify of the bound species.
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
Phosphoric acids
Phosphoric acids were investigated to characterize Brønsted acid organocatalysts relative to H-
bonding organocatalysts. BINOL-based chiral phosphoric acids have been highly utilized as chiral
Brønstead acid catalysts for reactions such as Diels-Alder and Friedel-Crafts additions with high
enantioselectivity.54 Modifications of the BINOL backbone at the 3- and 3’-positions allows for a variety
of catalysts with different steric and electronic properties. In order to quantify the binding interactions for
BINOL-based phosphoric acids, three acids (23a-c) were studied. All three phosphoric acids showed strong
interactions with TEPO based on large 31P NMR downfield shifts. The Δδ 31P NMR shifts for phosphoric
acids 23a-c ranged from 13.9 to 18.6 depending on substitution. The impact of adding electron-
withdrawing substituents was measured for phosphoric acid 22b (Δδ = 18.6 ppm, AN = 61.7) compared to
unsubstituted phosphoric acid 22a (Δδ = 14.6 ppm). The bulky tri-isopropylaryl substituents in 22c only
slightly decrease the 31P NMR shift hydrogen-bonding ability (Δδ = 13.9 ppm).
Boronic acids
Boronic acids have applications for organocatalysis,55-56 recognition and supramolecular
assembly.14 Boronic acids are unique relative to other organocatalysts because they can exhibit both
hydrogen bonding and Lewis acid properties. It has been demonstrated in solution that the nature of the
binding partner (e.g. anion) can dictate whether H-bonding or LA interactions occur.12 However, limited
examples using organoboronic acids as hydrogen-bonding catalysts have appeared in literature.11 Using 31P
ACS Paragon Plus Environment

Page 17 of 37
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
NMR spectroscopy, phenylboronic acid (23c) demonstrates a large shift (Δδ = 10.9 ppm) indicating a strong
binding interaction with TEPO.57 Adding a CF3 group (23a) further enhances the binding interaction with
TEPO and exhibits one of the largest 31P NMR shifts (Δδ = 22.3 ppm, AN = 70.3) observed in this study.58
Across all structurally varied boronic acid catalysts studied there was a wide range in values for the
observed Δδ 31P NMR shifts (from 0.2 to 24.6 ppm), which translates to acceptor numbers of 21-75.
We envisioned that both H-bond donor and coordination to the Lewis acidic boron center are
possible modes for TEPO binding with boronic acids (Figure 7). These results inspired us to quantify the
hydrogen-bonding ability of organoboronic acids and understand the nature of activation (i.e. LA vs HBD).
Experiments were also performed to determine if a boroxine (e.g 33), present in equilibrium with the
boronic acid,59 or cyclic H-bonded species (e.g. 32)13 contributes to the observed effect.
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
Figure 7. Possible binding modes for boronic acid species.
To distinguish between the possibilities of LA or HBD interactions, the boroxine (33) and boronic
ester (34) were both synthesized and compared, resulting in smaller changes in 31P NMR shifts (Figure 8).
The synthetic boroxine 33, which still contained 26% phenylboronic acid 23c, induced a Δδ 31P NMR shift
of 7.1 ppm upon binding to TEPO.60 A commercial source of triphenylboroxine 33 (containing up to 40%
boronic acid)59 also induced a smaller shift upon binding to TEPO (Δδ = 8.2 ppm) compared to purified
phenylboronic acid (Δδ = 10.9 ppm).61 The correlation of the 31P NMR shift with the amount of boronic
acid in the boroxine suggests that the presence of boronic acid accounts for the observed shift. Furthermore,
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 18 of 37
1
2
3
4
5
6
7
8
9
only small shifts were observed for boronic esters (Δδ = 0.7 ppm for 34a and Δδ = 0.6 ppm for 34c) upon
binding with TEPO. These data support that Lewis acid activation of TEPO is only a small contributor and
instead the large shifts observed with boronic acids can be attributed to a hydrogen-bonding interaction.62
Although boronic acids are known to associate via H-bonding as a dimeric species, the current
concentrations support that the 31P NMR shifts are attributed to H-bonding with the monomer.63-64
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
Figure 8. Comparing 31P NMR ∆for TEPO binding with boroxine 33 and
boronic esters 34a and 34c. Based on NMR analysis, commercial boroxine 33
contained 36% of 23c, and boroxine 33 synthesized immediately prior to use
contained 26% of 23c.
1
In addition to 31P NMR experiments, both 11B and H NMR spectroscopy also support that the
interaction between boronic acids and TEPO is through H-bonding. For phenylboronic acid (23c), only
small downfield shifts < 1 ppm were observed by 11B NMR spectroscopy (see Supporting Information)65
upon adding TEPO, indicative of a hydrogen-bonding interaction rather than complexation with boron.12
Using 1H NMR spectroscopy, the hydroxyl peaks for phenylboronic acid shifted downfield by
approximately 1.0 ppm in the presence of TEPO, providing further evidence for a H-bonding interaction.
After determining that hydrogen bonding is the predominant mode of interaction for boronic acids
with TEPO, the binding for a series of boronic acids 23-31 was investigated using 31P NMR spectroscopy.
31
31
As demonstrated by the high correlation of Δδ P NMR shift to Hammett values (R2 = 0.99), P NMR
shifts upon TEPO binding accurately accounted for electronic trends.
Recent literature reports have cited that ortho-substituted phenylboronic acids, specifically 2,6-
disubstituted phenyl boronic acids, have increased catalytic activity relative to other substitution patterns.11
The hydrogen-bonding ability of 2-fluorophenylboronic acid (25a), 2-bromophenylboronic acid (25b), and
2-methylphenylboronic acid (25c) was quantified and all three exhibited weaker binding to TEPO
compared to phenylboronic acid (Δδ = 8.7, 8.3, and 5.6 ppm respectively). Reduced chemical shifts were
ACS Paragon Plus Environment

Page 19 of 37
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
observed when 2,6-disubstituted phenylboronic acids were tested as hydrogen-bond donors: 2,6-
dimethylphenyl boronic acid 26c (Δδ = 5.2 ppm) and 2,6-dimethoxyphenylboronic acid 26d (Δδ = 0.2
ppm). In the case of naphthalen-1-ylboronic acid (29) a stronger hydrogen-bonding interaction was
observed with TEPO (Δδ = 9.4 ppm at 0.05 M). 2,6-Difluorophenylboronic acid 26b also resulted in a
smaller NMR shift (Δδ = 4.43 ppm at 0.05 M).
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
The decreased hydrogen-bonding ability of ortho-substituted phenylboronic acids is attributed to
steric effects based on a quantitative comparison of the Δδ 31P NMR shifts for various positional isomers.
When the substituent is moved to the meta-position the hydrogen-bonding ability is increased, as
demonstrated by 3,5-difluorophenylboronic (27) acid, which caused the largest downfield TEPO shift (Δδ
= 24.6 ppm) of all hydrogen-bond donors studied. 3-Flourophenylboronic acid (24, Δδ = 20.5 ppm) formed
a stronger hydrogen-bond to TEPO compared to 2-fluorophenylboronic acid (25a, Δδ = 8.7 ppm).
Correlation of pKa and Δδ 31P NMR Shift
The Δδ for 31P NMR shifts across all classes of catalysts studied (and the corresponding AN) can
be compared to pKa values. For hydrogen-bond donors that have reported pKa values in water, strong
correlation between pKa and change in 31P NMR shifts is observed within catalyst classes with similar
substituents, similar to what Kozlowski observed using the UV-Vis probe.5a For example, para-substituted
phenylboronic acids (R2 = 0.984) and ortho-substituted carboxylic acids (R2 = 0.874) have high correlations
(Figure 9).
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 20 of 37
1
2
3
4
5
6
7
8
12
10
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
6
4
2
0
0
10
20
Δδ 31P NMR (ppm)
Figure 9. Comparison of Δδ 31P NMR shifts (upon binding TEPO) with pKa (H2O). High linear
correlations are observed within catalyst classes with similar substitution.
When looking at just para-substituted phenols once again there is a high linear correlation observed
(R2 = 0.974), but when other substitution patterns are accounted for by including 2-bromophenol (2a),
deviation from the linear trend is observed. The deviation from linear behavior can be attributed to the
hydrogen-bonding interaction with TEPO reflecting steric effects that are not accounted for in pKa
measurements. These steric effects are relevant for catalysis, suggesting that the Δδ 31P NMR measurement
is more relevant for understanding and predicting hydrogen-bond strength compared to pKa.66 Furthermore
there is a poor correlation of the 31P NMR shifts with pKa across all classes of H-bonding catalysts (R2 =
0.259, see SI).
Measurement of Catalytic Activity
To determine how well scales such as pKa and Δδ 31P NMR shifts correlate to catalytic activity, the
HBDs were evaluated as catalysts using the addition of N-methylindole (36) to 4-trifluoromethyl-trans-β-
nitrostyrene (35) as a representative reaction (Figure 10). Variations of this reaction have been used
frequently to correlate the acidity of different classes of hydrogen-bonding catalysts.67 By incorporating a
trifluoromethyl group on the nitrostyrene the reaction rates can be easily monitored using 19F NMR
spectroscopy. Only three peaks are present in the 19F NMR spectra for this reaction (staring material,
ACS Paragon Plus Environment

Page 21 of 37
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
product and internal standard) with no peak overlap, thus making 19F NMR spectroscopy a reliable method.
The indole addition reaction was studied under pseudo-first order conditions and rate constants for a variety
of organocatalysts were determined (Table 1).
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
Figure 10. 19F NMR monitoring for addition of N-methylindole
to 4-trifluoromethyl-trans-β-nitrostyrene.
It is notable that of the hydrogen-bonding catalysts studied, boronic acids showed the largest rate
enhancement in the indole addition reaction with 3,5-difluorophenylboronic (27) and (4-
(trifluoromethyl)phenyl)boronic acid (23a) being the most active catalysts studied (krel = 17.89 and 15.66
respectively, Table 1). In order to further support the H-bonding activation mode of boronic acids, boronic
esters 34a and 34c and boroxine 33 were also evaluated as catalysts for the addition of indole to nitrostyrene.
Neither boronic ester showed any substantial rate enhancement for the indole addition, verifying that the
activation of nitrostyrene by boronic acids results from a hydrogen-bonding interaction rather than a Lewis
acid activation (Figure 11). Triphenylboroxine 33 also exhibited a decreased catalytic activity (krel = 4.60
vs 6.96) indicating the importance of the hydroxy groups for catalytic activity. When unpurified
phenylboronic acid containing phenylboroxine is utilized as a catalyst, the rate of the reaction is also
decreased (see Supporting Information).
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 22 of 37
1
2
3
4
5
6
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
Figure 11. Catalytic activity of boronic esters and boroxine
in the addition of N-methylindole to -nitrostyrene; 33 contains
36% of 23c.
Correlation of pKa Values and Catalytic Activity
As pKa values have previously been used to predict the hydrogen-bonding ability of
organocatalysts2b we wanted to investigate the correlation of pKa and rates for the indole addition reaction.
When looking at known pKa values in either water or DMSO, there is a poor linear correlation between
reaction rate and pKa of catalysts (R2 = 0.076 in H2O and R2 = 0.379 in DMSO) (Figure 12). The poor
ability of pKa values to predict hydrogen-bond strength may be attributed to secondary factors that can
contribute to catalysis including dual-hydrogen bonding interaction,68 sterics and binding geometry.69 The
poor ability of pKa to predict reactivity is demonstrated by boronic acid 27 which has a pKa (7.09) higher
than several catalysts including phenol 3 (5.53) and carboxylic acid 12 (1.75), though was the most activate
catalyst in the indole addition reaction, which was accurately predicted by 31P NMR shifts.
20
16
H₂O, R² = 0.0762
DMSO, R² = 0.379
12
8
4
0
0
5
10
15
20
pKa
Figure 12. Correlation of pKa (in both H2O and DMSO) to
relative rate of catalyzed Friedel-Crafts reaction.
ACS Paragon Plus Environment

Page 23 of 37
The Journal of Organic Chemistry
1
2
3
4
5
Correlation of Δδ 31P NMR Shift and Catalytic Activity
6
7
The measured 31P NMR shifts for a variety of hydrogen-bond donors were found to correlate well
8
9
to reactivity for the addition N-methylindole (36) to 4-trifluoromethyl-trans-β-nitrostyrene (35). The 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
31
for relative rate and Δδ P NMR shift exhibit a linear relationship across all classes of catalysts with an
excellent correlation of R2 = 0.965 (Figure 13A). When focusing on individual classes of functional groups,
the correlation is similar or stronger: phenols, R2 = 0.965; carboxylic acids, R2 = 0.974; silanols, R2 = 0.963;
and boronic acids, R2 = 0.963 (Figure 13B). Based on the observed correlation the x-intercept of the best-
fit line corresponds to a Δδ 31P NMR shift of 3.3 ppm. This indicates that when the Δδ 31P NMR shift upon
binding to TEPO is < 3.3 ppm the hydrogen-bond donor is not strong enough to catalyze the indole addition
reaction. This observation was supported when both TADDOL 9 (Δδ = 2.1 ppm) BINOL 6d (Δδ = 0.32
ppm) showed no rate enhancement from the background reaction. Furthermore, silanols 13b (Δδ = 2.8
ppm) and 13c (Δδ = 3.1 ppm) both had negligible rate enhancements of less than 0.1 relative to the
background.
A)
27
20
24
16
23a
12
29
12
8
1b
4
0
26b
2a
10c
0
5
10
15
20
25
Δδ 31P NMR (ppm)
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 24 of 37
1
2
3
4
B)
phenols, R² = 0.965
5
6
7
8
20
16
12
8
carboxylic acids, R² = 0.974
silanols, R² = 0.963
boronic acids, R² = 0.963
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
4
0
0
5
10
15
20
25
Δδ 31P NMR (ppm)
Figure 13. A) Comparison of Δδ 31P NMR shift based on TEPO sensor to rate
constant of indole addition reaction, a linear correlation was found (R2 = 0.965)
for all catalyst studied. Selected compound numbers are included for reference.
B) Correlation of Δδ 31P NMR shifts and rate constants improves within catalyst
classes. 31P NMR shifts based on TEPO binding and rate constants for the
addition of N-methylindole to -nitrostyrene.
The high linear correlation observed between 31P NMR shifts for TEPO binding and relative rate
demonstrates that the TEPO NMR probe is a better predictor of catalytic ability than pKa.. The enhanced
correlation is attributed to the fact that TEPO binding is measured under catalytically relevant conditions
and can account for secondary factors such as steric effects.
The ability of 31P NMR shifts to predict and measure steric effects that are relevant to
catalysis was demonstrated with BINOL catalysts 6a-6d. pKa values would predict the 3,3-bis-CF3 phenyl
substituted BINOL 6c would have the strongest hydrogen bond (pKa = 9.68 in DMSO). In the Friedel-Crafts
addition reaction, the observed krel for unsubstituted BINOL 6a is greater than 3,3-bis-CF3 phenyl
substituted BINOL 6c (1.36 vs. 0.64). The Δδ 31P NMR shifts accurately predicted this steric effect, with
unsubstituted BINOL 6a forming a stronger hydrogen bond to TEPO than 6c (Δδ 31P NMR = 4.9 ppm vs
3.4 ppm). Other cases where the TEPO probe was able to predict reactivity that pKa wasn’t includes bromo-
substituted phenols 1b (Δδ = 7.2 ppm, pKa = 9.36, krel = 3.69) vs 2a (Δδ = 5.2 ppm, pKa = 8.44, krel = 0.74),
and naphthols 4 (Δδ = 6.0 ppm, pKa = 9.34, krel = 1.83) vs 5 (Δδ = 6.2 ppm, pKa = 9.63, krel = 2.19); in both
ACS Paragon Plus Environment

Page 25 of 37
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
31
cases the less acidic hydrogen-bond donor was a better catalyst, as predicted by P NMR shifts. TEPO
was also able to predict the increased catalytic activity of naphthalen-1-ylboronic acid 29, which did not
match the steric trends observed with other ortho-substituted phenylboronic acids. Not only is the 31P NMR
shift a better predictor of reactivity, it is also much easier to obtain than pKa values when new
organocatalysts are synthesized.
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
The H-bonding ability and catalytic activity of boronic acids was accurately predicted using 31P
NMR spectroscopy and further studied to provide evidence of a H-bonding interaction. The reduced
catalytic activity for the boronic esters and cyclic boroxine was accurately predicted based on TEPO binding
measurements using 31P NMR spectroscopy, pointing to the importance of the two hydroxyl groups of the
boronic acids for the observed catalytic activity. The ability of 31P NMR spectroscopy to identify the nature
of boronic acid activation demonstrates the value of this method for discovering and investigating new
organocatalysts.
Conclusions
In conclusion, we have quantified the hydrogen-bonding activation for a wide range of
organocatalysts using a commercially available phosphine oxide probe for rapid Δδ 31P NMR measurements.
The measured Δδ 31P NMR shifts and acceptor numbers strongly correlate to catalytic activity based on the
relative rates for a Friedel-Crafts indole addition reaction with nitrostyrene. High correlation was found
within each class of hydrogen-bonding catalyst, strongly correlating with electronic effects. When
compared to pKa values, the measured 31P NMR shifts are a better predictor of catalytic activity. We
examined emerging classes of organocatalysts including organosilanols and aryl boronic acids where the
31P NMR shifts accurately predicted relative catalytic activity. Aryl boronic acids containing EWGs
demonstrated the strongest H-bonding abilities of all organocatalysts studied. A detailed investigation
confirmed that hydrogen-bonding interactions, and not binding to the Lewis acidic boron center, is the
primary activation mode that accounts for the observed binding and catalytic activity for boronic acids. The
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 26 of 37
1
2
3
4
5
6
ease of 31P NMR measurements with TEPO has future implications for designing new organocatalysts and
quantifying hydrogen-bonding ability for supramolecular synthons and medicinal compounds.
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
Experimental Section
General Methods and Materials: All nuclear magnetic resonance (NMR) spectra were obtained on Bruker
Nanobay AVIIIHD 400 (400 MHz for H; 162 MHz for 31P; 376 MHz for 19F) equipped with an
1
autosampler, at room temperature unless noted otherwise. Chemicals shifts were reported in parts per
million (δ scale), and referenced according the following standards: tetramethylsilane internal standard for
1H signals in chloroform, fluorobenzene external standard in CD2Cl2 for 19F(1H) signals, BF3•Et2O external
standard in CD3CN for 11B NMR signals, and triethylphosphine oxide external standard in CD2Cl2 for
31P(1H) signals. Coupling constants were reported in Hertz (Hz) and multiplicities were reported as follows:
singlet (s), doublet (d), triplet (t), quartet (q), and multiplet (m), broadened (b).
Commercially available reagents were purchased and used without further purification unless otherwise
indicated. Dichloromethane was obtained from Fisher Scientific. Fluorobenzene and triethylphosphine
oxide (TEPO) were obtained from Sigma Aldrich. Diphenylsilanediol was purchased from Acros. N-
methylindole was purchased from TCI. 1,1,3,3-tetraisopropyldisiloxane and 1,3-dichloro-1,1,3,3-
tetraphenyldisiloxane were purchased from Gelest. 4-Trifluoromethyl-trans-β-nitrostyrene was synthesized
in one step following literature procedure and matched previously reported spectra.70 Triphenylsilanol
was prepared in one step using Pd/C hydrolysis of triphenylsilane and the spectral data was confirmed to
match previously reported spectra in literature.15c Dimesitlysilanediol were made according to literature
procedure.15c Di(naphthalen-1-yl)silanediol was synthesized according to literature procedure using a Pd/C-
mediated hydrolysis of dinaphthylsilane.15d 1,1,3,3-tetraisopropyldisiloxane-1,3-diol was made in one step
using a Pd/C-mediated hydrolysis of 1,1,3,3-tetraisopropyldisiloxane. 1,1,3,3-Tetraphenyldisiloxane-1,3-
diol was synthesized in one step from 1,3-dichloro-1,1,3,3-tetraphenyldisiloxane.15c Disiloxanediols 15c –
15g were synthesized according to literature procedures starting from the addition of the corresponding aryl
ACS Paragon Plus Environment

Page 27 of 37
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
Grignard to trichlorosilane.15g Para-substituted phenols were recrystallized from petroleum ether or
chloroform. Ortho-substituted benzoic acids were purchased from Acros and used without further
purification. Boronic acids were recrystallized from water prior to use. (R)-3,3'-Bis(2,4,6-
9
triisopropylphenyl)-[1,1'-binaphthalene]-2,2'-diol
and
(R)-3,3′-bis(2,4,6-triisopropylphenyl)-1,1′-
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
binaphthyl-2,2′-diyl hydrogenphosphate were synthesized according to literature procedure.71 (S)-(+)-3,3′-
Bis(3,5-bis(trifluoromethyl)phenyl)-1,1′-binaphthyl-2,2′-diyl hydrogenphosphate and (S)-(-)-3-3’-bis(3,5-
bis(trifluoromethyl)phenyl)-1,1’-bi-2-naphthol were purchased from Sigma Aldrich. 1,3-Bis(3,5-
bis(trifluoromethyl)phenyl)thiourea was synthesized according to literature procedures.72 2-Phenyl-1,3,2-
dioxaborolane and 2-(4-(trifluoromethyl)phenyl)-1,3,2-dioxaborolane were synthesized according to
literature procedure from the corresponding phenylboronic acids.73
Reactions were analyzed by thin layer chromatography (TLC) on EMD glass plates that were pre-
coated with silica gel 60 F254. The following abbreviations are used throughout: triethylphosphine oxide
(TEPO), 4-trifluoromethyl-trans-β-nitrostyrene (NS).
Procedure for 31P NMR titration experiments with TEPO. Catalyst concentration was held constant for
titration experiments and varying amounts of TEPO was added to give the desired equivalents of catalyst.
A stock solution was made with the hydrogen-bond catalyst in CD2Cl2 (0.2 M). A second stock solution of
TEPO in CD2Cl2 was also made (0.4 M). Into an oven-dried, argon-purged NMR tube was added 0.25 mL
of the catalyst stock solution, and then the necessary amount of the TEPO stock solution to give the desired
31
equivalents of catalyst, followed by the required volume of CD2Cl2 to give a volume of 0.5 mL. The P
NMR spectrum of each sample was collected at room temperature with 32 scans and compared to a TEPO
standard.
Procedure for Job plot analysis of hydrogen-bond donor and TEPO. To determine the binding
stoichiometry of TEPO and a hydrogen-bond donor, Job plots was constructed with 4-nitrophenol and
phenylboronic acid. A stock solution of the hydrogen-bond donor (0.1 M) in CDCl3 was made as well as a
ACS Paragon Plus Environment

The Journal of Organic Chemistry
Page 28 of 37
1
2
3
4
5
6
7
8
9
sock solution of TEPO in CDCl3 (0.1 M). To an oven-dried, argon purged NMR tubes was added the
necessary amounts of each stock solution to vary the mole fraction of TEPO between 0.1 and 1.0 with a
final solution volume of 0.6 mL. The 31P NMR spectrum of each sample was collected at room temperature
with 32 scans and compared to a standard of triethylphosphine oxide.
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 experimental procedure for TEPO measurement. The general procedure for measuring change
in 31P NMR TEPO shifts was adapted from the reported Gutmann-Beckett method.5c, 7b, 8a Based on the
desired concentration of 0.1 M in catalyst, hydrogen-bond donor (3.0 equiv, 0.050 mmol) was dissolved in
0.4 mL of CD2Cl2 and transferred to an oven-dried, argon-purged NMR tube. Then 0.1 mL (1.0 equiv, 0.017
mmol) of a stock solution of triethylphosphine oxide (0.17 M in CD2Cl2) was added to the NMR tube which
31
was inverted several times to allow for mixing. The P NMR spectrum of each sample was collected at
room temperature with 32 scans and compared to a standard of triethylphosphine oxide to measure the
change in chemical shift. Acceptor numbers were calculated based on the equation, AN = (2.21*(δsample
–
41.0)). Only catalysts that were completely soluble in CD2Cl2 were tested to assure that peak being
measured was the bound species. If no 1H NMR spectra are acquired, then the procedure could be modified
to prepare the samples using 0.4 mL of CH2Cl2 and only 0.1 mL of CD2Cl2 from the TEPO stock solution.
This allowed for a decreased use of expensive deuterated NMR solvent.
General Procedure for N-methylindole addition to 4-trifluoromethyl-trans-β-nitrostyrene.15h A stock
solution of 4-trifluoromethyl-trans-β-nitrostyrene (0.665 M) and fluorobenzene (0.333 M) in CD2Cl2 was
made. The catalyst (0.028 mmol) was weighed directly into an oven-dried and argon-purged NMR tube
followed by the addition of 0.42 mL of the 4-trifluoromethyl-trans-β-nitrostyrene stock solution. An initial
19F NMR scan was taken before the addition of N-methylindole (0.28 mL, 8.0 equiv) and then the reaction
was monitored by taking a spectrum every 30–60 minutes. 4 scans with a 25-second relaxation delay were
taken to assure complete relaxation for accurate integrations. Fluorobenzene was used as an internal
standard (–113 ppm) and integrations compared to 4-trifluoromethyl-trans-β-nitrostyrene and product.
ACS Paragon Plus Environment

Page 29 of 37
The Journal of Organic Chemistry
1
2
3
4
Integration ranges for 4-trifluoromethyl-trans-β-nitrostyrene = – 62.9 to –63.0 ppm; product = –62.1 to –
5
6
62.3 ppm; and internal standard = –113.0 to –113.3 ppm.
7
8
9
ASSOCIATED CONTENT
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
Supporting Information
The Supporting Information is available free of charge via the Internet at http://pubs.acs.org and includes
31P NMR shifts and kinetic data.
AUTHOR INFORMATION
Corresponding Author
Telephone: 530-752-9820. E-mail: akfranz@ucdavis.edu
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENT
We acknowledge the National Science Foundation (CHE-0847358 and CHE-1363375) for support of this
research. Julia Jennings (UCD) is acknowledged for assistance with NMR spectroscopy and manuscript
editing.
REFERENCES
(1) For recent reviews of hydrogen-bond catalysis, see: (a) Auvil, T. J.; Schafer, A. G.; Mattson, A. E.
Design Strategies for Enhanced Hydrogen-Bond Donor Catalysts. Eur. J. Org. Chem. 2014, 2014, 2633-
2646; (b) Doyle, A. G.; Jacobsen, E. N. Small-Molecule H-Bond Donors in Asymmetric Catalysis. Chem.
Rev. 2007, 107, 5713-5743; (c) Taylor, M. S.; Jacobsen, E. N. Asymmetric Catalysis by Chiral Hydrogen-
Bond Donors. Angew. Chem. Int. Ed. 2006, 45, 1520-1543; (d) Schreiner, P. R. Metal-free
organocatalysis through explicit hydrogen bonding interactions. Chem. Soc. Rev. 2003, 32, 289-296.
(2) For selected examples of modeling and predicting hydrogen-bond strength, see: (a) O’Meara, M. J.;
Leaver-Fay, A.; Tyka, M. D.; Stein, A.; Houlihan, K.; DiMaio, F.; Bradley, P.; Kortemme, T.; Baker, D.;
Snoeyink, J.; Kuhlman, B. Combined Covalent-Electrostatic Model of Hydrogen Bonding Improves
Structure Prediction with Rosetta. J. Chem. Theory Comput. 2015, 11, 609-622; (b) Gilli, P.; Pretto, L.;
ACS Paragon Plus Environment