Importance of the Electron Correlation and Dispersion Corrections in Calculations Involving Enamines, Hemiaminals, and Aminals. Comparison of B3LYP, M06-2X, MP2, and CCSD Results with Experimental Data

Article abstract of DOI:10.1021/acs.joc.5b01814

While B3LYP, M06-2X, and MP2 calculations predict the δG° values for exchange equilibria between enamines and ketones with similar acceptable accuracy, the M06-2X/6-311+G(d,p) and MP2/6-311+G(d,p) methods are required for enamine formation reactions (for example, for enamine 5a, arising from 3-methylbutanal and pyrrolidine). Stronger disagreement was observed when calculated energies of hemiaminals (N,O-acetals) and aminals (N,N-acetals) were compared with experimental equilibrium constants, which are reported here for the first time. Although it is known that the B3LYP method does not provide a good description of the London dispersion forces, while M06-2X and MP2 may overestimate them, it is shown here how large the gaps are and that at least single-point calculations at the CCSD(T)/6-31+G(d) level should be used for these reaction intermediates; CCSD(T)/6-31+G(d) and CCSD(T)/6-311+G(d,p) calculations afford δG° values in some cases quite close to MP2/6-311+G(d,p) while in others closer to M06-2X/6-311+G(d,p). The effect of solvents is similarly predicted by the SMD, CPCM, and IEFPCM approaches (with energy differences below 1 kcal/mol).

Full text of DOI:10.1021/acs.joc.5b01814

Subscriber access provided by UTSA Libraries
Article
Importance of the Electron Correlation and Dispersion Corrections in
Calculations Involving Enamines, Hemiaminals, and Aminals. Comparison
of B3LYP, M06-2X, MP2, and CCSD Results with Experimental Data
Alejandro Castro-Alvarez, Hector Carneros, Dani Sanchez, and Jaume Vilarrasa
J. Org. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.joc.5b01814 • Publication Date (Web): 10 Nov 2015
Downloaded from http://pubs.acs.org on November 19, 2015
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 free 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 accessible to all
readers and 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.
The Journal of Organic Chemistry 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 16
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
Importance of the Electron Correlation and Dispersion Corrections
in Calculations Involving Enamines, Hemiaminals, and Aminals.
Comparison of B3LYP, M06-2X, MP2, and CCSD Results with
Experimental Data
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
Alejandro Castro-Alvarez, Héctor Carneros, Dani Sánchez, and Jaume Vilarrasa*
Departament de Química Orgànica, Facultat de Química, Universitat de Barcelona, Diagonal 645, 08028 Barcelona, Catalonia, Spain
jvilarrasa@ub.edu
ABSTRACT: While B3LYP, M06-2X, and MP2 calculations predict the ∆G˚ values for exchange equilibria between enamines
and ketones with similar acceptable accuracy, the M06-2X/6-311+G(d,p) and MP2/6-311+G(d,p) methods are required for
enamine formation reactions (for example, for
B3LYP/6-31+G(d,p)
B3LYP/6-311+G(d,p)
B3LYP/6-31G(d)
B3LYP/6-31+G(d)
M06-2X/6-31G(d)
MP2/6-31G(d)
M06-2X/6-311+G(d,p)
M06-2X/6-31+G(d)
enamine 5a, arising from 3-methylbutanal and
pyrrolidine). Stronger disagreement was observed
when calculated energies of hemiaminals (N,O-
acetals) and aminals (N,N-acetals) were
MP2/6-31+G(d)
MP2/6-311+G(d,p)
∆Gº (5a) in kcal/mol
,
CCSD(T)/6-31+G(d)
CCSD(T)/6-311+G(d,p)
from the gas phase to DMSO
exptl
10
5
–5
0
C6D6
DMSO-d6
compared with experimental equilibrium constants, which are reported here for the first time. Although it is known that the
B3LYP method does not provide a good description of the London dispersion forces, while M06-2X and MP2 may overestimate
them, it is shown here how large the gaps are and that at least single-point calculations at the CCSD(T)/6-31+G(d) level should be
used for these reaction intermediates; CCSD(T)/6-31+G(d) and CCSD(T)/6-311+G(d,p) calculations afford ∆G˚ values in some
cases quite close to MP2/6-311+G(d,p) while in others closer to M06-2X/6-311+G(d,p). The effect of solvents is similarly
predicted by the SMD, CPCM, and IEFPCM approaches (with energy differences below 1 kcal/mol).
INTRODUCTION
The revival of the chemistry of enamines and iminium salts as a consequence of the development of the field of
organocatalysis is outstanding.1 To account for the reactions disclosed or developed, many research groups have carried out
calculations based on the density functional theory (DFT) of the species and transition states presumably involved in these
catalytic processes.2 Sometimes, DFT calculations have been used to design new organocatalysts before synthesizing them, to
try to increase the chances of success.
ACS Paragon Plus Environment
1

The Journal of Organic Chemistry
Page 2 of 16
1
2
3
4
5
6
7
8
There have been many warnings about the shortcomings of diverse functionals, especially regarding the use and abuse3 of
the popular B3LYP method4 by many organic chemists (such as ourselves). For example, Schleyer et al. and Schreiner et al.
reported the systematic errors of B3LYP and related functionals in computing the energies of hydrocarbons,3a–d Tirado-Rives
and Jorgensen studied a large set of compounds, including isomerization enthalpies for O- and N-containing molecules,3g
Houk et al. evaluated the case of Diels–Alder additions,3h and Hoffmann, Schleyer, and Schaefer asked for more caution in the
energy predictions and common sense in the terminology.3i More recently, Houk et al. estimated the sources of error in the
reaction enthalpies for aldol, Mannich and α-aminohydroxylation reactions.3j Among the handicaps of several DFT methods to
provide reliable energy values, an important one is that they do not take into account the London dispersion forces; in simple
terms, they do not consider the weak attraction between pairs of non-polar atoms and between pairs of molecules arising from the
interaction of instantaneous multi-poles. Other more modern DFT methods, such as M06-2X by Truhlar et al.,5 which give results
closer to those obtained by the Møller–Plesset theory (MP2, for example),6 are known to perform much better in this regard. More
recently, Grimme et al. have included dispersion-corrected terms in DFT methods (DFT-D, DFT-D3), achieving much more
reliable results.7 These algorithms have been implemented in the ORCA package8 and in the 2013 version of Gaussian 09.9
Assessment of dispersion corrections in DFT methods is a hot topic.3f,7 Dispersion-corrected MP2 calculations have also been
reported.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
Our interest lies in the performance of DFT methods when providing insight into enamine- and iminium-based catalytic reactions
and the design of new catalysts. In fact, in the past ten years, we have carried out B3LYP calculations with different basis sets in
connection with several master's degrees and PhD theses. Some of those results have appeared as supplementary material or in the main
text of publications by our group11 or in collaboration with other research groups.12 The present study, thus, had two goals: (a) to check
how reliable these former results were (as a self-criticism, which may also serve to rectify or reevaluate other results, if necessary); and
(b) to determine which methods, among the most popular, are appropriate for calculating equilibria involving hemiaminals (N,O-acetals)
and aminals (N,N-acetals) summarized in Scheme 1.13
Scheme 1. Equilibria among Hemiaminals, Enamines, and Aminals
from Carbonyl Compounds 1–7 and Secondary Amines a–c
N
H
– H2O
H
N
O
N
O
N
+
H/R
N
R
N
H
R
H2O
R
H/R
H/R
H/R
R
enamine
aminal
hemiaminal
H
O
N
N
H
O
N
N
N
+
HO
R3C
N
H
H
R3C
H
R3C
H2O
iminium
H
R3C
hydroxide
hemiaminal
aminal
ACS Paragon Plus Environment
2

Page 3 of 16
The Journal of Organic Chemistry
1
2
3
4
O
O
O
O
O
O
O
O
O
O
O
5
4
1
3
5
2
6
7
6
7
8
9
OTMS
O
N
N
H
Ph
Ph
N
H
O
H
a
b
c
10
11
Hemiaminals are expected to be short-lived species in amine-catalyzed processes of carbonyl compounds with α-hydrogens, such as
1–5, as they quickly dehydrate to give the active species, i.e. enamines (although significant proportions of H2O in the medium may
militate against this dehydration). Anyway, hemiaminals are crucial intermediates in the formation of either enamines or iminium
salts and might explain some exchange reactions. Moreover, when the carbonyl compounds do not have enolizable hydrogens (such
as 6 and 7), hemiaminals might be detected since enamines cannot be formed, provided that carboxylic acids are absent otherwise
iminium salts are favored. Although not productive, aminals are also plausible intermediates,11a depending on the relative amount of
amine(s) in the medium (especially toward the end of the reaction, when starting materials are partially exhausted). They may also be
involved in exchange reactions, if two amino groups or different amines are present in the medium.
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
RESULTS AND DISCUSSION
Equilibria between enamines and ketones. We first evaluated the relative thermodynamic stability of enamines. Table 1 shows the
equilibrium reaction between the enamine from pyrrolidine and cyclohexanone (1a, prepared independently and isolated) and 2,2-dimethyl-1,3-
dioxan-5-one (1,3-dihydroxyacetone isopropylidine acetal, 2). For this isodesmic reaction (not hypothetical but real), B3LYP, M06-2X, and
MP2 calculations, with geometries optimized with different basis sets, predicted similar total energy values (between –3.2 and –4.1 kcal/mol).14
When thermal and entropy corrections were included, the predicted ∆G˚ values were also quite close. There are no significant differences
between ∆E and ∆G˚, as there is the same number of similar molecules on both sides of the chemical equation.
The solvent effects were calculated by means of the SMD method,15a but when we compared this with the CPCM and IEFPCM methods the
results were similar15b (also for other equilibria15c). The effect of a non-polar solvent such as benzene, with respect to the gas phase, was predicted
to be quite small (a decrease of ∆G˚ of 0.2–0.6 kcal/mol), as expected; it was quite general or was assumed to be general and was not always
calculated. The effect of a polar solvent such as DMSO lowered the calculated ∆G˚ value more (0.8–1.3 kcal/mol). When we determined, by 1H
NMR spectroscopy, the Keq values for such an exchange reaction (1a + 2 = 1 + 2a) in C6D6 and in DMSO-d6,16 the agreement between
the predicted and experimental values was very good. Bearing in mind the inherent approximations of MO calculations, the experimental
errors (as the equilibrium constants were determined from peak integrations in the 1H NMR spectra, at relatively high concentrations),
the presence of several conformers for many of the molecules studied (from which we have only selected that of the lowest energy for
the calculation of the Gibbs free energy values, or free enthalpy values, as usual), and how the solvent effects were calculated in our
study (implicit solvent models, single-point calculations), the agreement is surprisingly excellent.
ACS Paragon Plus Environment
3

The Journal of Organic Chemistry
Page 4 of 16
1
2
3
4
5
Thus, for the equilibrium shown in Table 1, with species with similar electronic delocalization and steric hindrance on both
sides, the simplifications and errors are compensated for. Higher level calculations, such as CCSD(T) calculations also indicated
in Table 1 (bottom), are not required for these types of equilibria, as computational chemists recognize.
6
7
8
9
Table 1. Calcd vs. Exptl ∆G˚ Values for the Equilibrium
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
between Enamines 1a and 2a and Their Ketones
δH 2.88
2.44
N
O
N
O
O
3.74
+
4.43
+
O
5.93
4.23
1.98
O
O
2
1
2a
1a
∆E (kcal/mol)
B3LYP/6-31G(d)
B3LYP/6-31+G(d)
B3LYP/6-311+G(d,p)
M06-2X/6-31G(d)
MP2/6-31+G(d)//B3LYP/6-31G(d)
MP2/6-31+G(d)//B3LYP/6-31+G(d)
MP2/6-31+G(d)
–3.5
–3.5
–3.6
–3.8
–4.0
–3.7
–3.7
–4.1
–4.1
–3.2
MP2/6-311+G(d,p)//B3LYP/6-31G(d)
MP2/6-311+G(d,p)//B3LYP/6-31+G(d)
CCSD(T)/6-31+G(d)//B3LYP/6-31+G(d)
a
a
∆G˚(kcal/mol) ∆G˚(C6H6)
∆G˚(DMSO)
–3.1b
–3.3
–3.9
–4.2
–5.2
–4.7
B3LYP/6-31G(d)
M06-2X/6-31G(d)
MP2/6-31+G(d)
CCSD(T)/6-31+G(d)
//B3LYP/6-31+G(d)
–3.6
–3.4
–3.6c
–4.2
–3.7
–3.1
Keq ≈ 160 (C6D6)
Keq ≈ 540 (DMSO-d6)
∆G˚exp ≈ –3.0 kcal/mol
∆G˚exp ≈ –3.6 kcal/mol
a Solvent effects were estimated by means of the SMD method (see the main text).
b This value changes to –3.4 kcal/mol using the vibrational frequency scaling factors for
B3LYP/6-31G(d) (see SI); all DFT ∆G˚ values given henceforward are without scaling.
c The same result with MP2/6-31G(d), taking into account or not the scaling corrections.
We obtained similar excellent agreements for equilibrium reaction between enamine 1a and cyclopentanone (3) to give 1 and
(1-cyclopentenyl)pyrrolidine (3a), for which we experimentally determined Keq = 2.0±0.2 (C6D6) and Keq = 2.1±0.2 (DMSO-d6).
At various calculation levels, using geometries optimized with different basis sets, the predicted reaction energies were practically
the same; for example, single-point calculations with the MP2/6-31+G(d) and MP2/6-311+G(d,p) methods provided practically
identical results starting from different geometries, as shown in more detail in the Supporting Information (SI).
Formation of enamines from aldehydes and pyrrolidine. In sharp contrast, for the formation of enamines the use of large basis
sets and diffuse functions gave much lower free energy values. The cases of pyrrolidine enamines of propanal (from 4, Scheme 1, to its
enamine 4a) and 3-methylbutanal (from 5 to enamine 5a) were first studied. The details are given in the SI. Table 2 summarizes the
essential points for the case of 5a. Only M06-2X/6-311+G(d,p), MP2/6-311+G(d,p), CCSD(T)/6-31+G(d), and CCSD(T)/6-311+G(d,p)
predict that these equilibria are shifted to the right, close to experimental data (the M06-2X value being closer to our experiments in C6D6
and the MP2 and CCSD values closer to our experiments in DMSO-d6). Thus, when there is only one conjugate species in the chemical
equation, the better the treatment of the electron correlation, the better the agreement with the experimental value, as expected.
ACS Paragon Plus Environment
4

Page 5 of 16
The Journal of Organic Chemistry
1
2
3
4
5
It is worth noting that the CCSD(T)/6-31+G(d) and CCSD(T)/6-311+G(d,p) results are almost identical. Thus, for the
enamines examined here it is not necessary to use the CCSD(T)/6-311+G(d,p) method, which is much more expensive: each
structure with > 14 atoms in the second period required > 14 days with 6 processors working together.
6
7
8
9
Table 2. Calcd vs. Exptl ∆G˚ Values, in kcal/mol, for the
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
Formation of Enamine 5a from Aldehyde 5
H
O
N
+ H2O
5a
+
N
H
5
a
∆G˚
∆E
∆G˚
(DMSO)
B3LYP/6-31G(d)
B3LYP/6-31+G(d)
B3LYP/6-31+G(d,p)
B3LYP/6-311+G(d,p)
10.0 10.4 9.0
5.9 6.4 5.5
1.3 1.9 1.3
0.5 0.8 0.0
M06-2X/6-31G(d)
M06-2X/6-31+G(d)
M06-2X/6-311+G(d,p)
7.4 7.3 5.7
3.9 3.8 2.5
–1.6 –1.4 –2.3
MP2/6-31G(d)//B3LYP/6-31G(d)
MP2/6-31+G(d)//B3LYP/6-31G(d)
MP2/6-311+G(d,p)//B3LYP/6-311+G(d,p)
MP2/6-311+G(d,p)//M06-2X/6-311+G(d,p)
6.5 6.9 5.1
2.0 2.5 1.1
–3.3 –3.0 –4.3
–3.5 –3.3 –4.6
CCSD(T)/6-31+G(d)//B3LYP/6-31+G(d)
–2.8 –2.4 –3.2
CCSD(T)/6-311+G(d,p)//B3LYP/6-311+G(d,p) –3.1 –2.8 –3.5
Keq = 14.3 (C6D6)
Keq ≥ 500 (DMSO-d6)
∆G˚exp = –1.6 kcal/mol
∆G˚exp ≤ –3.7 kcal/mol
For the sake of comparison, the reaction yielding enamine 5b from 3-methylbutanal and the Jørgensen–Hayashi catalyst
(henceforward J–H, see b)1,17 is shown in Table 3. Whereas the B3LYP/6-311+G(d,p) numbers predict that the equilibrium is
very shifted to the left, which does not agree with the experimental fact, MP2/6-311+G(d,p) calculations predict an equilibrium too
shifted towards 5b (overestimation of the dispersion energy). M06-2X/6-311+G(d,p) gives an intermediate value. It can be
assumed again that CCSD(T) methods would provide results between M06-2X and MP2 (of Table 3), again closer to the
experimental ones, but we did not perform these calculations (they would have been too costly).
Table 3. Calcd vs. Exptl ∆Gº Values, in kcal/mol, for
the Formation of Enamine 5b from Aldehyde 5
Ph
TMS
O
H
O
N
Ph
+
+ H2O
Ph
Ph
N
H
O
TMS
5
b
5b
∆E ∆G˚
(J–H)
∆G˚
(DMSO)
B3LYP/6-31G(d)
14.8 15.8 14.3
10.2 11.9 10.5
5.1 6.1 4.9
B3LYP/6-31+G(d)
B3LYP/6-311+G(d,p)
0.5 1.6
0.2
M06-2X/6-311+G(d,p)//B3LYP/6-311+G(d,p)
MP2/6-311+G(d,p)//B3LYP/6-311+G(d,p)
–5.0 –4.0 –5.8
Keq ≈ 20 (DMSO-d6),
∆G˚exp ≈ –1.8 kcal/mol
ACS Paragon Plus Environment
5

The Journal of Organic Chemistry
Page 6 of 16
1
2
3
4
Formation of hemiaminals. With all this background, we were ready to examine related equilibria that, to the best of our
knowledge, have been never subjected to the scrutiny of high-level calculations.
5
6
7
8
9
When a simple hemiaminal, or N,O-acetal, such as HA-4a (Table 4) was calculated at different levels of theory, it was noted
that B3LYP, with different basis sets, predicted ∆G˚(DMSO) values of 11.6–13.7 kcal/mol; that is to say, when there were no
differences in conjugation (electronic delocalization) on either side of the chemical equation, small differences were obtained
between using smaller or larger basis sets. The ∆G˚(DMSO) values calculated with M06 and MP2 were 10 kcal/mol lower (1.9–
3.2 kcal/mol). No experimental data are available for comparison since, as known, aldol reactions take place quickly in this
case, via the enamine that is rapidly formed by dehydration of the hemiaminal and/or via base catalysis.
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
Due to the lack of experimental values regarding hemiaminals, we carried out additional calculations using dispersion-
corrected DFT methods,18 such as B2PLYP-D3 and wB97XD, just to check the effect of such corrections. We did not plan to
systematically compare a long series of these DFT functionals. It is worth noting that wB97XD/6-311+G(d,p) gave ∆G˚(DMSO)
= 4.9 kcal/mol, an intermediate value between the two extremes mentioned in the preceding paragraph, although closer to the
M06 and MP2 values. This value was also close to that of 5.4 kcal/mol predicted by our highest level calculation, that is, using
the CCSD(T)/6-311+G(d,p) method. In other words, B3LYP methods are not reliable in these cases, as they predict that the
attack of secondary amines on carbonyl groups are much less exothermic than expected, so that, once the entropic term is
included, the equilibria are shifted too far to the left to allow for the detection of hemiaminals. By contrast, high-level
calculations predicted that the equilibrium in Table 4 (∆Gº in DMSO) was not so shifted to the left. These observations held in
other cases in which the carbonyl groups are sterically more hindered and, consequently, the hemiaminals are more crowded
(such as the reaction of 3 with pyrrolidine to yield HA-3a, and of 6 with pyrrolidine to yield HA-6a, included as SI in an
expanded Table 4).
Table 4. Calcd ∆G˚ Values, in kcal/mol, for the
Formation of the Hemiaminal Related to 4a (HA-4a)
H
O
O
+
N
H
N
4
a
HA-4a
∆G˚
∆E ∆G˚ (DMSO)
B3LYP/6-31G(d)
B3LYP/6-31+G(d)
B3LYP/6-311+G(d,p)
–6.9 8.7 11.6
–5.7 9.9 13.7
–5.9 9.8 13.6
B2PLYP-D3/6-311+G(d,p)//B3LYP/6-31+G(d) –9.6 6.0 10.1
wB97XD/6-311+G(d,p)//B3LYP/6-31+G(d)
–14.5 1.1 4.9
M06-2X/6-31G(d)
M06-2X/6-31+G(d)
M06-2X/6-311+G(d,p)
–16.8 –1.2 1.9
–16.0 –0.4 3.4
–16.8 –0.9 2.8
MP2/6-311+G(d,p)//B3LYP/6-31+G(d)
MP2/6-311+G(d,p)//B3LYP/6-311+G(d,p)
–15.6 0.0 3.1
–15.7 0.1 3.2
CCSD(T)/6-31+G(d)//B3LYP/6-31+G(d)
CCSD(T)/6-311+G(d,p)//B3LYP/6-311+G(d,p) –13.6 2.1 5.4
–13.0 2.4 6.2
ACS Paragon Plus Environment
6

Page 7 of 16
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
To study one example of hemiaminal formation with the equilibrium shifted to the right we chose methyl glyoxylate (methyl
glyoxalate, methyl 2-oxoacetate, 7), where the presence of an EWG linked to the carbonyl may relatively stabilize19 the
corresponding hemiaminal, HA-7a. Within each level of theory (see Table 5, first equilibrium), the use of smaller or larger basis
sets almost did not affect the calculated reaction energies (as mentioned for other cases in which the differences in electronic
delocalization of the molecules involved in the equilibria are not important).
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 B3LYP calculations suggested that the first equilibrium of Table 5 was shifted slightly to the left, whereas M06-2X and MP2
indicated that it was shifted far to the right. The experimental fact, using commercially available ethyl glyoxylate (OHC–COOEt, 7'),
is that the equilibrium is shifted far to the right, in such a way that we could not measure the equilibrium constant due to its high
value (full conversion of 7' and pyrrolidine into HA-7'a in C6D6). Our most reliable calculations, CCSD(T), suggested that the error
in the B3LYP calculations can be around 7.5 kcal/mol. On the other hand, M06-2X and MP2, again with respect to CCSD(T),
overestimated the stability of the HA-7a by 3.4–4.0 kcal/mol and 1.3–1.8 kcal/mol, respectively.
It can be observed that CCSD(T)/6-31+G(d) and CCSD(T)/6-311+G(d,p) values do not differ too much. The former, although
generally less reliable, is a good approach to the latter, at least with regard to the formation of this and previous hemiaminals.
Again, wB97XD/6-311+G(d,p) gave results intermediate between MP2/6-311+G(d,p) and CCSD(T)/6-311+G(d,p), all of them
close to each other.
Table 5. Calcd ∆G˚ Values, in kcal/mol, for the Reactions of Methyl
Glyoxylate with Pyrrolidine
H
O
N
N
H
O
4.95
3.80
O
O
O
81.9 N
HA-7aa
82.8N
+
N
H
H2O
O
O
7aab
O
a
7
∆G˚
∆G˚
∆G˚ (DMSO)
∆E ∆G˚
∆E
(DMSO)
2.4 4.9
2.5 4.9
B3LYP/6-31+G(d)
B3LYP/6-311+G(d,p)
–12.9
–13.0
1.3 1.5 0.0
–2.1 0.6
–6.1 –3.7
B2PLYP-D3/6-311+G(d,p)//B3LYP/6-31+G(d)
wB97XD/6-311+G(d,p)//B3LYP/6-31+G(d)
–17.4
–21.4
–8.5 –5.9
–8.7 –6.3
–9.3 –6.8
M06-2X/6-31+G(d)
M06-2X/6-311+G(d,p)//B3LYP/6-311+G(d,p)
M06-2X/6-311+G(d,p)
–23.8
–24.3
–24.5
–6.3 –3.8
–6.6 –4.6
MP2/6-31+G(d)//B3LYP/6-31+G(d)
MP2/6-311+G(d,p)//B3LYP/6-311+G(d,p)
–21.6
–22.2
–7.6 –7.4
–8.9
–4.9 –2.5
–5.0 –2.8
CCSD(T)/6-31+G(d)//B3LYP/6-31+G(d)
CCSD(T)/6-311+G(d,p)//B3LYP/6-311+G(d,p)
–20.4
–20.5
a With EtOCOCHO (7') in C6D6 the first equilibrium—the formation of hemiaminal—was
quickly shifted to the right (to HA-7'a). b By contrast, the homologue of aminal 7aa (7'aa)
appeared slowly (a few days at rt to reach the 2nd equilibrium, unless catalytic amounts of
PhCOOH were added), though it was also fully shifted to the right.
ACS Paragon Plus Environment
7

The Journal of Organic Chemistry
Page 8 of 16
1
2
3
4
5
6
7
8
9
Formation of aminals from hemiaminals. The conversion of hemiaminal HA-7a into aminal 7aa (that is, equilibrium HA-
7a + a = 7aa + H2O, Table 5, right side) was also calculated. B3LYP indicated that it is slightly endothermic, while according to
MP2 it is highly exothermic and, as expected for an equilibrium involving two molecules on both sides—small differences in
entropy—, the ∆G˚ values indicated that the possible equilibrium is completely shifted to the right. Experimentally, we observed
that this was the case in the NMR tube, using the hemiaminal of ethyl glyoxylate (HA-7'a). The conversion was slow, even in the
presence of an excess of pyrrolidine, but it sufficed to add a catalytic amount of benzoic acid to the NMR tube to cause an almost
immediate and quantitative transformation of the hemiaminal, HA-7'a, into the aminal, 7'aa.20 Overnight treatment of 7' with 210
mol % of pyrrolidine, in the presence of 4-Å MS, also gave 7'aa quantitatively.
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 reaction of 7 with the Jørgensen–Hayashi catalyst (b) was also computationally examined and the results were parallel
(see SI, addendum to Table 5). With the commercially available ethyl glyoxylate (7'), we were able to determine the equilibrium
constant for the formation of hemiaminal HA-7'b (7,600 M–1, that is, above 7,000 L·mol–1). It was feasible since this first
equilibrium is in practice less shifted to the right (in comparison with the reaction of 7' with pyrrolidine to give HA-7'a), while
the second equilibrium (formation of aminal 7'bb) did not intervene, as it was slower and not so favorable.
Formation of aminals from enamines. We also examined the formation of aminals from the addition of amines to enamines,
that is, we calculated the first step of the general equilibria shown in Scheme 2.
Scheme 2. Plausible Equilibria between Enamines and Aminals
N
N
N
N
+
+
N
H
N
H
H
R
R
enamine
enamine
R
aminal
The energies of aminal 5aa at different levels of theory and then the free enthalpy for equilibrium 5a + a = 5aa (aminal from
3-methylbutanal and pyrrolidine) were calculated. A brief summary is shown in Table 6. Here, the gap between B3LYP/6-
311+G(d,p) and the other methods was larger than ever. The ∆E values of M06-2X, wB97XD, and CCSD(T) were practically
identical. The values predicted by M06-2X/6-311+G(d,p) agreed with the experimental ones, even more than those coming from
CCSD(T)/6-31+G(d). In the light of the corresponding results shown in Table 5, we assume that CCSD(T)/6-311+G(d,p), not
calculated, would give a value even closer to M06-2X and the experimental one. Using the experimental value in DMSO as the
reference, the error (underestimation) of B3LYP/6-311+G(d,p) was around 14.5 kcal/mol (in DMSO) while the overestimation of
the dispersion corrections by MP2/6-311+G(d,p) was only around 3 kcal/mol (also in DMSO). It is well established that B3LYP
methods are very often inappropriate for energy calculations, but here we show how large they may be when moderately crowded
molecules such as 5aa are involved.
ACS Paragon Plus Environment
8

Page 9 of 16
The Journal of Organic Chemistry
1
2
3
4
5
The effect of the size of the basis set is indicated in the SI, in an expanded Table 6. With small basis sets the ∆E values are
more negative (and thus the ∆Gº values are less unfavorable or more favorable to the formation of 5aa), due to the poorer
description of the electronic delocalization of 5a.
6
7
8
9
Table 6. Calcd vs. Exptl ∆G˚ Values, in kcal/mol, for
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 Formation of Aminal 5aa from 5a and Pyrrolidine
N
N
N
N
+
N
H
N
a
5a
5aa
∆G˚ ∆G˚
∆E ∆G˚
(C6H6) (DMSO)
B3LYP/6-311+G(d,p)
12.4 14.9 15.8
–3.1 –0.5 0.4
–2.5 0.2 1.2
–5.0 –2.6 –1.8
–2.9 –0.2 0.7
–3.5
–19.0
–19.1
–21.0
–18.9
wB97XD/6-311+G(d,p)//B3LYP/6-311+G(d)
M06-2X/6-311+G(d,p)
MP2/6-311+G(d,p)//B3LYP/6-311+G(d)
CCSD(T)/6-31+G(d)//B3LYP/6-31+G(d)
Keq ≈ 0.24 L·mol–1 (C6D6)
∆G˚exptl ≈ 0.8 kcal/mol
Keq ≈ 0.11 L·mol–1 (DMSO-d6)
∆G˚exptl ≈ 1.3 kcal/mol
Equilibria between enamines (via aminals). Table 7 compares two equilibria between enamines and secondary amines.
In the first, where the conjugation and steric hindrance on both sides are similar: (i) the size of the basis set is not significant;
(ii) the B3LYP errors are small; and (iii) all the other methods give values quite close to the experimental ones.
Experimentally, under the reaction conditions (with nearly equimolar amounts of the secondary amine, in NMR tubes at 25
ºC), we did not detect aminal 1ac, so it is indicated within brackets. The ∆Gº value (gas phase) calculated at the M06-2X/6-
311+G(d,p) level for the 1c + a = 1ac equilibrium is 3.2 kcal/mol (to be compared with ∆Gº = –2.5 kcal/mol, also in the gas
phase at the same level, for the 5a + a = 5aa equilibrium shown in Table 6).
Finally, in Table 8, where the large substituent may actually give rise to steric hindering, the size of the basis set is
insignificant, as in the preceding Table, but now the gap between B3LYP/6-311+G(d,p) and the other methods is more
important. While MP2/6-311+G(d,p) surprisingly predicts positive ∆Gº values (generally attributed to an overestimation of the
dispersion corrections), the M06-2X/6-311+G(d,p) results are close to the available experimental value in C6D6. By the way,
spectra were not registered in DMSO-d6 because of the partial cleavage of the O–Si bond, in this and other cases in which
equilibria are reached slowly. Although the exchanges of Tables 7 and 8 may also occur through partial hydrolysis of the
enamines if a trace of water was present in the NMR tubes, what matters here is the position of the equilibria.
ACS Paragon Plus Environment
9

The Journal of Organic Chemistry
Page 10 of 16
1
2
Table 7. Calcd vs. Exptl ∆G˚ Values, in kcal/mol, for the Exchange
3
4
of the 1-Cyclohexenyl Group between Secondary Amines a and c
5
6
7
8
O
O
N
N
O
N
O
O
+
+
N
H
N
H
N
O
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
a
1ac
c
1c
1a
∆G˚(DMSO)
∆E ∆G˚
B3LYP/6-31+G(d)
B3LYP/6-311+G(d,p)
–2.5
–1.7
–1.7
–2.8
–2.8
–2.6
wB97XD/6-311+G(d,p)//B3LYP/6-31+G(d)
–1.4
–0.5
–1.6
M06-2X/6-31+G(d)
M06-2X/6-311+G(d,p)
–1.7
–1.7
0.2
–0.6
–0.9
–1.6
MP2/6-311+G(d,p)//B3LYP/6-31+G(d)
MP2/6-311+G(d,p)//M06-2X/6-31+G(d)
–0.2
–0.4
0.6
1.4
–0.4
–0.5
CCSD(T)/6-31+G(d)//B3LYP/6-31+G(d)
–1.1
–0.2
–1.3
Keq = 8.4±0.8 (C6D6)
Keq = 6.7±0.9 (DMSO-d6)
∆G˚exptl ≈ –1.1 kcal/mol
∆G˚exptl ≈ –1.3 kcal/mol
Table 8. Calcd vs. Exptl ∆G˚ Values, in kcal/mol, for the Exchange
of an Alkenyl Group between Secondary Amines a and b
Ph
Ph
Ph
Ph
TMS
O
N
N
N
O
+
+
N
H
O
N
H
a
Ph
Ph
N
TMS
TMS
5b
b
5a
5ab
∆E
∆G˚
∆G˚(DMSO)
–5.3
–5.0
–4.8
–4.8 –5.4
–4.3 –5.6
–4.6 –5.3
B3LYP/6-31G(d)
B3LYP/6-31+G(d)
B3LYP/6-311+G(d,p)
–2.6
1.6
–2.2 –2.9
1.7 1.0
M06-2X/6-311+G(d,p)//B3LYP/6-311+G(d,p)
MP2/6-311+G(d,p)//B3LYP/6-311+G(d,p)
Keq ≈ 75 (C6D6) ∆G˚exptl ≈ –2.6 kcal/mol
CONCLUSIONS
Only when an identical number of similar molecules are on both sides of a chemical equation—similar electronic
delocalization, comparable steric hindrance—may the B3LYP energies be partially reliable. The B3LYP methods, which are
so efficient for geometry optimizations, give large errors for the examples shown in Tables 2, 3, 4, 5, 6, and 8. Higher-level
calculations, for example MP2 and/or M06-2X, with large basis sets, are necessary in these cases. Obviously this is well
known,3,7 but we have disclosed here how much all these methods underestimate or overestimate the electron correlation and
London dispersion forces regarding enamines, hemiaminals, and aminals. Top-level calculations cannot always be efficiently
ACS Paragon Plus Environment
10

Page 11 of 16
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
carried out, but the outcomes have been compared with experimental ∆G˚ values in C6D6 and/or DMSO-d6 solutions reported
here for the first time. Although a study of the performance of dispersion-corrected DFT methods was outside the scope of this
work, we observed that wB97XD/6-311+G(d,p) gives energy values very close to CCSD(T), for hemiaminals and aminals.
For the formation of simple enamines, the CCSD(T) free enthalpy values were almost intermediate between the M06-2X/6-
311+G(d,p) and MP2/6-311+G(d,p) values (see the case of 5a, Figure 1, left). For the equilibria of formation of hemiaminals,
the M06-2X/6-311+G(d,p) values were systematically more exoergic than the MP2/6-311+G(d,p) and CCSD(T) results, as
graphically indicated in Figure 1 (center) for the case of hemiaminal HA-7a. Finally, for the formation of aminals from
enamines and secondary amines, the MP2/6-311+G(d,p) energies were more negative than those determined experimentally
and/or predicted by CCSD(T) and M06-2X/6-311+G(d,p), as shown in Figure 1 (right) for the equilibrium between 5a and 5aa
in the gas phase. For more crowded molecules (such as enamines from the J–H catalyst), the classical MP2 fails, as the
energies are further overestimated (updated methods7,10 would be needed). Anyway, what is clear is that the energy values
predicted by B3LYP/6-311+G(d,p) for the thee equilibria of Figure 1 (corresponding to Tables 2, 5, and 6, respectively) are
wrong; those predicted by B3LYP with smaller basis sets, not included in Figure 1 for the sake of simplicity, are worse (left),
similar (center), and paradoxically some kcal/mol lower (right). We hope that these conclusions will be useful to others and
help avoid mistakes that we made years ago.
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
B3LYP/6-311+G(d,p)
12
B3LYP/6-311+G(d,p)
B3LYP/6-311+G(d,p)
0
0
∆Gº
(kcal/mol)
∆Gº
(kcal/mol)
∆Gº
(kcal/mol)
M06-2X/6-311+G(d,p)
–4
–2
CCSD(T)/6-31+G(d)
CCSD(T)/6-31+G(d)
CCSD(T)/6-311+G(d,p)
MP2/6-311+G(d,p)
CCSD(T)/6-311+G(d,p)
0
MP2/6-311+G(d,p)
M06-2X/6-311+G(d,p)
CCSD(T)/6-31+G(d)
–8
–4
M06-2X/6-311+G(d,p)
MP2/6-311+G(d,p)
O
H
N
–6
N
O
+
+ H2O
O
N
H
O
N
O
+
a
N
5
5a
N
H
O
+
O
N
a
N
H
7
HA-7a
a
5a
5aa
Figure 1. Comparison of calculated ∆G˚ values in the gas phase for the reactions of
formation of enamine 5a (left), hemiaminal HA-7a (center), and aminal 5aa (right)
from the corresponding carbonyl compounds (left, center) or enamine (right).
ACS Paragon Plus Environment
11

The Journal of Organic Chemistry
Page 12 of 16
1
2
3
EXPERIMENTAL SECTION
4
5
6
7
8
9
Computational Methods. The calculations were carried out with the Gaussian 09 package (Revision D.01, 2013)9 with
methods B3LYP4 M06-2X,5 MP2,6 and CCSD(T).21 The ORCA package was also used in a few cases for the sake of
comparison.8 The stationary points were characterized by frequency calculations. Gibbs free energies (free enthalpies) at
298.15 K for all the reactions were calculated on the basis of the rigid rotor/harmonic oscillator approximation. The solvent
effects were calculated by means of the SMD method,15a but also the CPCM and IEFPCM methods were sometimes used,
again for the sake of comparison. The effect of scaling factors was evaluated on the equilibrium of Table 1 for the B3LYP/6-
31G(d) and MP2/6-31G(d) methods;22 as no significant differences were noted (see SI), these corrections were not considered
in the remaining equilibria (Tables 2–8), for the sake of simplicity.
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
Starting Materials and General Information. All the carbonyl compounds used in this work (1–7) are known and most
of them are commercially available; they were dried over 4-Å MS before use. 1,3-Dihydroxyacetone isopropylidene acetal
(2,2-dimethyl-1,3-dioxan-5-one, 2) was prepared according to the procedure of Enders et al.,23 although we purified the
compound by flash column chromatography over silica gel (CH2Cl2) instead of by distillation. A technical grade,
commercially available mixture of ethyl glyoxylate (7') and toluene (50%) was purified by flash chromatography (95:5 to 1:1,
hexanes/ethyl acetate, to remove toluene and polymeric material), followed by fractional distillation24 at 20 mbar under N2
(and stored after dilution with C6D6). Enamine 1a is commercially available. Most NMR spectra were registered in C6D6 and
in anhydrous DMSO-d6. 1H NMR spectra were recorded on 400 MHz spectrometers, with 5 s of mixing time. Chemical shifts
are reported in ppm with the solvent resonance as the internal standard (C6D5H in C6D6 at 7.16 ppm, CD3SOCHD2 in DMSO-
d6 at 2.50 ppm); data are reported in the following order: chemical shift, multiplicity (s = singlet, d = doublet, t = triplet, q =
quartet, br = broad, m = multiplet), coupling constants in Hz, integration. 13C NMR spectra were recorded on 100.6 MHz
instruments with proton decoupling; chemical shifts are reported in ppm with the solvent as the internal standard (C6D6, δ
128.06 ppm; DMSO-d6, δ 39.52 ppm). The most relevant cross peaks of 2D NMR experiments are marked on the spectra;
HSQC cross-peaks belonging to CH and CH3 are tagged in blue and those to CH2 in red. FTIR spectra of HA-7'a and 7'aa in
their liquid form (oils) were registered; only the main absorptions, in cm–1, are given. The mass spectra were obtained by the
electrospray ionization (ESI+, TOF) technique.
Reactions of formation of enamines, hemiaminals, and aminals, as well as exchange reactions, were generally carried out and/or
followed in standard NMR tubes by mixing appropriate amounts of reactants or reagents; details for the determination of the
equilibrium constants from the 1H NMR integrations of relevant peaks are given as SI.
ACS Paragon Plus Environment
12

Page 13 of 16
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
Representative Procedure for the Preparation of Enamines. A solution of L-proline methyl ester (compound c, 2.05 g,
15.9 mmol), cyclohexanone (compound 1, 1.64 mL, 15.9 mmol), and p-toluenesulfonic acid (TsOH·H2O, 60 mg, 0.31 mmol)
in cyclohexane (40 mL) was heated at reflux temperature in a Dean–Stark apparatus for 5 h. Removal of the solid (salt of the
proline derivative and TsOH) by filtration or decantation, evaporation of the solvent under vacuum, addition of CH2Cl2 and co-
evaporation (two or three times) in the rotary evaporator, and removal of residual solvents and cyclohexanone with a vacuum
pump afforded enamine 1c as a yellow oil (practically quantitative yield),25 which was analyzed by NMR and used without
further purification (to avoid its easy hydrolysis). Enamines 1a,26 2a,27 3a,28 5a,29 and 5b30 (in this last case no TsOH was
added, to avoid partial desilylation of the OTMS group and subsequent reactions) were similarly prepared in > 95% yields.
Representative Example of Enamine Formation (as Monitored by NMR). Pyrrolidine (compound a, 4.0 mg, 0.06
mmol) was added to a solution of isovaleraldehyde (3-methylbutanal, 5, 5.0 mg, 0.06 mmol) in C6D6 (0.7 mL) or in anhydrous
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
DMSO-d6 (0.7 mL) in a vial. The mixture was directly transferred to a NMR tube and H NMR spectra were recorded until
equilibrium was attained (no additional increase of the peaks of 5a, usually within 1 h).
Representative Example of an Exchange Reaction. 2,2-Dimethyl-1,3-dioxan-5-one (2, 14.0 mg, 0.10 mmol) was added
1
to a solution of 1a in C6D6 (0.7 mL) or in anhydrous DMSO-d6 (0.7 mL) and H NMR spectra were recorded until the relative
heights of the peaks of 2, 1a, 2a, and 1 did not change further (usually within 1 h).
Hemiaminal HA-7'a. To ethyl glyoxylate (ethyl oxoacetate, 7', 7.2 mg, 0.07 mmol) in C6D6 (up to 0.7 mL) was added
pyrrolidine (5.0 mg, 0.07 mmol) and the spectra were registered: 1H NMR (C6D6) δ 4.92 (s, 1H), 3.91 (qd, J = 7.1, 3.8, 2H), 2.76
(m, 4H), 1.53 (m, 4H), 0.87 (t, J = 7.1, 3H); 13C NMR (C6D6) δ 172.3, 82.1, 61.5, 47.1, 24.6, 14.1. Removal of the solvent under
good vacuum, without heating, gave HA-7'a as a colorless oil (12.0 mg, ca. 100%); FTIR 3430 (br), 1730; HRMS (ESI+) m/z calcd
for C8H16NO3+ (M + H)+ 174.1125, found 174.1119; the base peak was the corresponding pyrrolidinium ion (C4H8N+=CHCOOC2H5),
m/z calcd for C8H14NO2+ 156.1019, found 156.1014.
Aminal 7'aa. To ethyl glyoxylate (ethyl oxoacetate, 7', 7.2 mg, 0.07 mmol) in C6D6 (up to 0.7 mL) was added pyrrolidine (10.0
mg, 0.14 mmol), and the mixture was treated overnight with 4-Å molecular sieves. Filtering and removal of the solvent under good
vacuum, without heating, afforded 17.0 mg (ca. 100%) of 7'aa as an oil; when distillation was attempted at 1 mbar with a much
larger volume of sample only decomposition was noted. 1H NMR (C6D6) δ 4.02 (q, J = 7.2, 2H), 3.80 (s, 1H), 2.89–2.72 (m, 8H),
1.60 (m, 8H), 0.99 (t, J = 7.1, 3H); 13C NMR (C6D6) δ 169.4, 82.8, 59.8, 49.7, 24.0, 14.7; FTIR 1744; HRMS (ESI+), the (M + H)+
+
peak, C12H23N2O2 , m/z 227.1754, can hardly be observed (even at low voltages and temperatures), while pyrrolidinium ion
C4H8N+=CHCOOC2H5, m/z 156.1014, was the main peak; however, registering the spectrum at 25 V, from a solution prepared in
anhydrous benzene and diluted just before injection with anhydrous acetonitrile, the (M + Na)+ peak was clearly observed, m/z calcd
for C12H22N2NaO2+ 249.1573, found 249.1566.
ACS Paragon Plus Environment
13

The Journal of Organic Chemistry
Page 14 of 16
1
2
3
ASSOCIATED CONTENT
4
5
Supporting Information
6
7
Details of calculations, copies of NMR and 2D-NMR spectra, and procedures for the determination of equilibrium constants.
8
9
AUTHOR 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
Corresponding Authors
*E-mail: (J.V.) jvilarrasa@ub.edu
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
Grant CTQ-2012-39230 (Spanish Government, FEDER) is acknowledged. A.C.-A. is a doctorate student with a fellowship from
the Chile Government; he thanks Daniel Aravena (Max Plank Institute at Mülheim) for indications about ORCA. H.C. is currently
a doctorate student financed by CTQ-2012-39230. D.S. has held a studentship of the University of Barcelona (UB, 2011–2014)
and is currently a postdoctoral fellow of the Fundació Cellex de Barcelona.
REFERENCES
(1) For very recent, representative reviews, see: (a) Rodriguez, J.; Coquerel, Y. ChemCatChem 2015, 7, 1263–1264. (b) Reboredo, S.;
Parra, A.; Alemán, J. Asymmetric Catalysis 2014, 1, 24–31. (c) Fraile, A.; Parra, A.; Tortosa, M.; Alemán, J. Tetrahedron 2014, 70, 9145–9173.
(d) Gualandi, A.; Mengozzi, L.; Wilson, C. M.; Cozzi, P. G. Chem. Asian J. 2014, 9, 984–995. (e) Deng, Y.; Kumar, S.; Wang, H. Chem. Commun.
2014, 50, 4272–4284. (f) Desmarchelier, A.; Coeffard, V.; Moreau, X.; Greck, C. Tetrahedron 2014, 70, 2491–2513. (g) Hogdson, D. M.; Charlton,
A. Tetrahedron 2014, 70, 2207–2236. (h) Mlynarski, J.; Bas, S. Chem. Soc. Rev. 2014, 43, 577–587. (i) Duan, J.; Li, P. Catal. Sci. Technol. 2014, 4,
311–320. (j) Jiang, H.; Albrecht, L.; Jørgensen, K. A. Chem. Sci. 2013, 4, 2287–2300.
(2) (a) For an excellent, exhaustive review, see: Cheong, P. H.-Y.; Legault, C. Y.; Um, J. M.; Çelebi-Ölçüm, N.; Houk, K. N. Chem. Rev. 2011, 111,
5042–5137. For very recent papers including DFT calculations of enamines, see: (b) Jindal, G.; Sunoj, R. B. J. Org. Chem. 2014, 79, 7600–7606. (c)
Hubin, P. O.; Jacquemin, D.; Leherte, L.; Vercauteren, D. P. Chem. Phys. 2014, 434, 30–36. (d) Shylesh, S.; Hanna, D.; Gomes, J.; Krishna, S.; Canlas,
C. G.; Head-Gordon, M.; Bell, A. T. ChemCatChem 2014, 6, 1283–1290. (e) Szollosi, G.; Csampai, A.; Somlai, C.; Fekete, M.; Bartok, M. J. Mol. Catal.
A: Chem. 2014, 382, 86–92. (f) Caruana, L.; Fochi, M.; Franchini, M. C.; Ranieri, S.; Mazzanti, A.; Bernardi, L. Chem. Commun. 2014, 50, 445–447.
(3) (a) Wodrich, M. D.; Corminboeuf, C.; Schleyer, P. v. R. Org. Lett. 2006, 8, 3631–3634. (b) Schreiner, P. R.; Fokin, A. A.; Pascal, R. A.; de
Meijere, A. Org. Lett. 2006, 8, 3635–3638. (c) Wodrich, M. D. Corminboeuf, C.; Schreiner, P. R.; Fokin, A. A.; Schleyer, P. v. R. Org. Lett. 2007, 9,
1851–1854. (d) Schreiner, P. R. Angew. Chem., Int. Ed. 2007, 46, 4217–4219 (review). (e) Riley, K. E.; Op't Holt, B. T.; Merz, K. M. J. Chem.
Theory Comput. 2007, 3, 407–433 (atomic and molecular properties of test sets), and references therein. (f) Grimme, S.; Steinmetz, M.; Korth,
M. J. Org. Chem. 2007, 72, 2118–2126 (isomerization energies, dispersion corrections). (g) Tirado-Rives, J.; Jorgensen, W. L. J. Chem. Theory
Comput. 2008, 4, 297–306. (h) Pieniazek, S. N.; Clemente, F. R.; Houk, K. N. Angew. Chem., Int. Ed. 2008, 47, 7746–7749 and references cited
ACS Paragon Plus Environment
14

Page 15 of 16
The Journal of Organic Chemistry
1
2
3
4
5
6
7
8
9
therein. (i) Hoffmann, R.; Schleyer, P. v. R.; Schaefer, H. F. Angew. Chem., Int. Ed. 2008, 47, 7164–7167 (also see the comments of G. Frenking, W.
Koch, M. Reiher, and F. M. Bickelhaupt that follow the essay). (j) Wheeler, S. E.; Moran, A.; Pieniazek, S. N., Houk, K. N. J. Phys. Chem. 2009,
113, 10376–10384 and references cited therein. (k) Schenker, S.; Schneider, C; Tsogoeva, S. B.; Clark, T. J. Chem. Theory Comput. 2011, 7, 3586–
3595 (relative TS energies). (l) Burke, K. J. Chem. Phys. 2012, 136, 150901/1–150901/9 (perspective) and references cited therein.
(4) (a) Becke, A. D. J. Chem. Phys. 1993, 98, 5648−5650. (b) Stephens, P. J.; Devlin, F. J.; Chabalowski, C. F.; Frisch, M. J. J. Phys. Chem.
1994, 98, 11623–11627.
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) (a) Zhao, Y.; Truhlar, D. G. Theor. Chem. Account 2008, 120, 215–241. (b) Zhao, Y.; Truhlar, D. G. Acc. Chem. Res. 2008, 41, 157–167.
(6) (a) Møller, C.; Plesset, M. S. Phys. Rev. 1934, 46, 618–622. (b) Head-Gordon, M.; Pople, J. A.; Frisch, M. J. Chem. Phys. Lett. 1988, 153, 503–506.
(7) (a) Grimme, S. J. Comput. Chem. 2006, 27, 1787–1799. Also see references contained in the following reviews: (b) Johnson, E. R.; Mackie,
I. D.; Di Labio, G. A. J. Phys. Org. Chem. 2009, 22, 1127–1135. (c) Grimme, S. WIREs Comput. Mol. Sci. 2011, 1, 211–228. (d) Ehrlich, S.;
Moellmann, J.; Grimme, S. Acc. Chem. Res. 2013, 46, 916–926. (e) Corminboeuf, C. Acc. Chem. Res. 2014, 47, 3217–3224. (f) Cho, Y.; Cho, W.
J.; Youn, I. S.; Lee, G.; Singh, N. J.; Kim, K. S. Acc. Chem. Res. 2014, 47, 3321–3330. For a general review on non-covalent interactions (WFT and
DFT), see: (g) Riley, K. E.; Pitonak, M.; Jurecka, P.; Hobza, P. Chem. Rev. 2010, 110, 5023–5063 and references therein.
(8) ORCA 3.0.2. (a) Neese, F. The ORCA program system, Wiley Interdiscip. Rev.: Comput. Mol. Sci. 2012, 2, 73–78. (b) Neese, F.;
Wennmohs, F.; et al., Max-Planck-Institut für Chemische Energiekonversion, Mülheim (www.cec.mpg.de).
(9) Gaussian 09. Revision D.01 (2013). Frisch, M. et al., Gaussian, Inc., Wallingford, CT (www. gaussian.com). See SI for the full citation.
(10) For example: (a) Cybulski, S. M.; Lytle, M. L. J. Chem. Phys. 2007, 127, 141102/1–141102/4. (b) Hesselmann, A. J. Chem. Phys. 2008,
128, 144112/1–144112/9. (c) Tkatchenko, A.; DiStasio, R. A.; Head-Gordon, M.; Scheffer, M. J. Chem. Phys. 2009, 131, 094106/1–094106/7. (d)
Grimme, S.; Goerigk, L.; Fink, R. F. WIREs Comput Mol Sci 2012, 2, 886–906. (e) Ohnishi, Y.; Ishimura, K.; Ten-no, S. J. Chem. Theory Comput.
2014, 10, 4857–4861. (f) Huang, Y.; Goldey, M.; Head-Gordon, M.; Beran, G. J. O. J. Chem. Theory Comput. 2014, 10, 2054–2063.
(11) For example: (a) Carneros, H.; Sánchez, D.; Vilarrasa, J. Org. Lett. 2014, 16, 2900–2903 (transfer of alkenyl groups between secondary
amines). (b) Sánchez, D.; Castro-Alvarez, A.; Vilarrasa, J. Tetrahedron Lett. 2013, 54, 6381–6384 (oxazolidinone/enamine ratios). (c) Sánchez, D.;
Bastida, D.; Burés, J.; Isart, C.; Pineda, O.; Vilarrasa, J. Org. Lett. 2012, 14, 536–539 (relative stability of enamines).
(12) Seebach, D.; Yoshinari, T.; Beck, A. K.; Ebert, M.-O.; Castro-Alvarez, A.; Vilarrasa, J.; Reiher, M. Helv. Chim. Acta 2014, 97, 1177–1203.
(13) For the sake of simplicity, in this Scheme the ball depicted at position 2 of the pyrrolidine ring may be a hydrogen atom or any substituent
other than a COOH group (which has a strong tendency to form oxazolidinone rings or, in very polar media, zwitterionic iminium carboxylates) or
similar groups. Moreover, R3CCHO means not only pivalaldehyde and similar branched aliphatic aldehydes but also ArCHO and RCH=CHCHO; in
the presence of PhCOOH or other suitable acids as additives, the iminium hydroxide depicted in the second row of Scheme 1, also for simplicity,
must be replaced by the corresponding iminium carboxylates (plus water), as expected.
(14) The twist-boat conformer of ketone 2 is drawn in Table 1. It is predicted to be 2–3 kcal/mol more stable than the chair conformer, at all levels.
(15) (a) Marenich, A. V; Cramer, C. J.; Truhlar, D. G. J. Phys. Chem. B 2009, 113, 6378–6396. (b) The SMD, CPCM, and IEFPCM approaches
were compared for this case (1a + 2 = 1 + 2a) at the B3LYP/6-31G(d) level, and the predicted ∆G˚(DMSO) values were –3.9, –4.6, and –4.6 kcal/mol,
respectively. (c) The SMD, CPCM, and IEFPCM approaches were also compared for other equilibria; for example, for the 5 + a = 5a + H2O
ACS Paragon Plus Environment
15

The Journal of Organic Chemistry
Page 16 of 16
1
2
3
4
5
6
7
8
9
equilibrium at the B3LYP/6-31+G(d) level (see Table 2), and the ∆G˚(DMSO) values turned out to be 5.7, 5.5, and 5.5 kcal/mol, respectively; at the
CCSD(T)/6-31+G(d)//B3LYP/6-31+G(d) level the predicted values were –3.3, –3.5, and –3.5 kcal/mol, respectively.
(16) In DMSO-d6 (moist, as acquired) the determined Keq value was 670; in DMSO-d6 dried over 4-Å MS the value is that given in Table 1 (540).
(17) For conformational analyses of enamines derived from this and similar organocatalysts, see: (a) Schmid, M. B.; Zeitler, K.; Gschwind, R.
M. Chem. Sci. 2011, 2, 1793–1803. (b) Hayashi, Y.; Okamura, D.; Yamazaki, T.; Ameda, Y.; Gotoh, H.; Tsuzuki, S.; Uchimaru, T.; Seebach, D.
Chem. Eur. J. 2014, 20, 17077–17088, and references therein.
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
(18) (a) Goerigk, L.; Grimme, S. J. Chem. Theory Comput. 2011, 7, 291–309. (b) Chai, J. D.; Head-Gordon, M. Phys. Chem. Chem. Phys. 2008,
10, 6615–6620.
(19) For precedents, see: (a) Boehme, H.; Sadanandam, Y. S. Arch. Pharm. 1973, 306, 227–236. (b) Zhang, F.; Peng, Y.; Liao, S.; Gong,
Y. Tetrahedron 2007, 63, 4636–4641. For an hemiaminal from prolinol, see: (c) Domínguez de María, P.; Bracco, P.; Castelhano, L. F.;
Bargeman, G. A. ACS Catal. 2011, 1, 70–75. For Cu- and Pd-catalyzed reactions of aliphatic hemiaminals, see: (d) Sakai, N. Hori, H.;
Yoshida, Y.; Konakahara, T.; Ogiwara, Y. Tetrahedron 2015, 71, 4722–4729 and references therein.
(20) For representative, classical studies on aliphatic aminals (a) Duhamel, L.; Siret, P. Bull. Soc. Chim. Fr. 1975, 908. (b) Aube, P.;
Christot, I.; Combret, J. C.; Klein, J. L. Bull. Soc. Chim. Fr. 1988, 1009–1014. For the preparation of aminals in water, see: (c) Jurcik, V.;
Wilhelm, R. Tetrahedron 2004, 60, 3205–3210. For reversible aminal formation (dynamic mixtures), see: (d) Buchs, B.; Godin, G.;
Trachsel, A.; de Saint Laumer, J.-Y.; Lehn, J.-M.; Herrmann, A. Eur. J. Org. Chem. 2011, 4, 681–695. For reactions of aminals with
organometallics, see: (e) Werner, V.; Ellwart, M.; Wagner, A. J.; Knochel, P. Org. Lett. 2015, 17, 2026–2029.
(21) Scuseri, G. E.; Schaefer III, H. F. J. Chem. Phys. 1998, 90, 3700–3703 and references cited therein.
(22) http://cccbdb.nist.gov/vibscalejust.asp (National Institute of Standards and Technology, Computational Chemistry Comparison and
Benchmark Database, Release 16a, August 2013, Johnson III, R. D., Ed.).
(23) Enders, D.; Voith, M.; Ince, S. J. Synthesis 2002, 1775–1779.
(24) Otley, K. D.; Ellman, J. A. J. Org. Chem. 2014, 79, 8296–8303.
(25) (a) Yamada, S.; Hiroi, K.; Achiwa, K. Tetrahedron Lett. 1969, 48, 4233–4236. (b) Kanzian, T.; Lakhdar, S.; Mayr, H. Angew. Chem., Int.
Ed. 2010, 49, 9526–9529 and references therein.
(26) (a) Seebach, D.; Beck, A. K.; Badine, D. M.; Limbach, M.; Eschenmoser, A.; Treasurywala, A. M.; Hobi, R. Helv. Chim. Acta 2007, 90,
425–471. (b) Cone, E. J.; Garner, R. H.; Hayes, A. W. J. Org. Chem. 1972, 37, 4436–4439 and references therein.
(27) Cagide-Fagín, F.; Alonso, R. Eur. J. Org. Chem. 2010, 6741–6747.
(28) (a) Zhao, X.; Liu, D.; Xie, F.; Zhang, W. Tetrahedron 2009, 65, 512–517. (b) Russell, G. A.; Wang, K. J. Org. Chem. 1991, 56, 3473–3479.
(29) Halland, N.; Lie, M. A.; Kjaersgaard, A.; Marigo, M.; Schiøtt, B.; Jørgensen, K. A., Chem. Eur. J. 2005, 11, 7083–7090.
(30) (a) Schmid, M. B.; Zeitler, K.; Gschwind, R. M. J. Am. Chem. Soc. 2011, 133, 7065–7074. (b) Schmid, M. B.; Zeitler, K.; Gschwind, R. M.
Chem. Sci. 2011, 2, 1793–1803. (c) Seebach, D.; Sun, X.; Sparr, C.; Ebert, M.-O.; Schweizer, W. B.; Beck, A. K. Helv. Chim, Acta 2012, 95, 1064–
1078. (d) Burés, J.; Dingwall, P.; Armstrong, A.; Blackmond, D. G. Angew. Chem., Int. Ed. 2014, 53, 8700–8704.
ACS Paragon Plus Environment
16