Intramolecular hydrogen bonds in normal and sterically compressed o-hydroxy aromatic aldehydes. Isotope effects on chemical shifts and hydrogen bond strength

Article abstract of DOI:10.3390/molecules24244533

A number of o-hydroxy aromatic aldehydes have been synthesized to illustrate the effect of steric compression and O···O distances on the intramolecular hydrogen bond and the hydrogen bond energies. Hydrogen bond energies have been calculated using the ‘hb

Full text of DOI:10.3390/molecules24244533

molecules
Article
Intramolecular Hydrogen Bonds in Normal
and Sterically Compressed o-Hydroxy Aromatic
Aldehydes. Isotope Effects on Chemical Shifts
and Hydrogen Bond Strength
Poul Erik Hansen 1,
*
, Fadhil S. Kamounah 2, Bahjat A. Saeed 3 , Mark J. MacLachlan 4
and Jens Spanget-Larsen 1
1
Department of Science and Environment, Roskilde University, Universitetsvej 1,
DK-4000 Roskilde, Denmark; spanget@ruc.dk
Department of Chemistry, University of Copenhagen, Universitetsparken 5,
2
DK-2100 Copenhagen, Denmark; fadil@chem.ku.dk
Department of Chemistry, College of Education for Pure Sciences, University of Basrah, Basrah 61004, Iraq;
3
bahjat.saeed@yahoo.com
4
Department of Chemistry, University of British Columbia, 2036 Main Mall, Vancouver, BC V6T 1Z1, Canada;
mmaclach@chem.ubc.ca
*
Correspondence: Poulerik@ruc.dk; Tel.: +45-76742432
ꢀꢁꢂꢀꢃꢄꢅꢆꢇ
Academic Editor: Goar Sánchez
Received: 24 October 2019; Accepted: 3 December 2019; Published: 11 December 2019
ꢀꢁꢂꢃꢄꢅꢆ
Abstract: A number of o-hydroxy aromatic aldehydes have been synthesized to illustrate the effect
of steric compression and O···O distances on the intramolecular hydrogen bond and the hydrogen
bond energies. Hydrogen bond energies have been calculated using the ‘hb and out’ method using
either the MP2 method or the B3LYP functional with the basis set 6-311++G(d,p). However, several
compounds cannot be treated this way. Hydrogen bond energies are also determined using electron
densities at bond critical points and these results are in good agreement with the results of the
‘
hb and out’ model. Two-bond deuterium isotope effects on 13C chemical shifts are suggested as an
experimental way to obtain information on hydrogen bond energies as they easily can be measured.
Isotope effects on aldehyde proton chemical shifts have also been measured. The former show very
good correlation with the hydrogen bond energies and the latter are related to short O···O distances.
Short O···O distances can be obtained as the result of short C=C bond lengths, conjugative effects,
and steric compression of the aldehyde group. Short O···O distances are in general related to high
hydrogen bond energies in these intramolecularly hydrogen-bonded systems of resonance assisted
hydrogen bond (RAHB) type.
Keywords: isotope effects on chemical shifts; Steric compression; hydrogen bond strength; MP2 and
B3LYP calculations; o-hydroxy aromatic aldehydes; atoms-in-molecules
1. Introduction
Hydrogen bonding, hydrogen bond strength and hydrogen bond energy are still subjects that
attract much attention both in chemistry and in biology [
of the resonance assisted type is important in DNA bases themselves [
pyridoxal-50-phosphatate and the related aldimines [
], and also in other small molecule such as
1
–
9]. Intramolecular hydrogen bonding
8], in co-factors such as
9
gossypol treated in this study. Intramolecular hydrogen bonding also has influence on uptake over
membranes. In this investigation we consider the intramolecular hydrogen bonding in a number
of o-hydroxy aromatic aldehydes. In a qualitative fashion, the strength can be judged from OH
Molecules 2019, 24, 4533; doi:10.3390/molecules24244533
www.mdpi.com/journal/molecules

Molecules 2019, 24, 4533
2 of 15
stretching frequencies [10], OH chemical shifts, two-bond deuterium isotope effects on 13C chemical
shifts [11] or O···O distances. In a more quantitative fashion, the energy difference between the
hydrogen-bonded and the open form (OH group turned 180◦ about the C-O axis) (‘hb and out’) has
been suggested as a theoretical measure of the hydrogen bond energy [12]. This method has mainly
been used in intermolecular systems but also in a few cases in intramolecularly hydrogen-bonded
systems [7,13]. It requires that the OH group of the open form is not involved in steric or other
disturbing interactions. Other methods, such as the molecular tailoring approach [14] has recently been
studied by Rusinska-Roszak [15]. Hydrogen bond energies have also been estimated from Bader’s
atoms–in-molecules (AIM) analysis [16] using electron densities at the bond critical points. Only a
few experimental estimates of hydrogen bond energies have been published. Reuben [17] suggested a
correlation between hydrogen bond energies and two-bond deuterium isotope effects on 13C chemical
shifts based on the work of Schaefer [18]. This led to the equation ln(2
is the hydrogen bond energy in kcal/mol and the two-bond isotope effect is in ppb. This equation was
later used in proteins [19,20].
∆
) = 2.783 + 0.34 E, in which E
·
o-Hydroxy aromatic aldehydes are very suitable molecules for investigating hydrogen bond
energies estimated by the ‘hb and out’ method as the CH proton has no special interactions in the “out”
molecule (see later). Previously, related systems such as o-hydroxyacyl aromatics have been studied
and showed rather strong hydrogen bonds, partly due to steric compression, but it was also found that
steric compression could lead to non-planar hydrogen-bonded systems [21,22].
The compounds explored in this paper cover a broad range of steric interactions and bond lengths
(bond orders) of the “double” bond (double in quotation marks as it in most cases refers to a bond in
an aromatic system) linking the hydrogen bond donor and acceptor. Therefore, they provide a number
of scenarios: (i) steric compression of the hydrogen bond partners; (ii) conjugation of the aldehyde
group with substituents other than the o-hydroxy group; (iii) variation of the length of the CC bond
connecting the hydrogen bond donor and acceptor; and (iv) variation of the O···O distance between
the oxygen atoms of the hydrogen bond. Researchers have judged the importance of the O···O distance
for the hydrogen bond strength very differently [2,23]. Perrin has shown that no obvious correlation
exists for charged systems [2].
Here we present the results (structures and hydrogen bond energies) of an experimental and a
theoretical study of intramolecular hydrogen bonding for a series of non-charged systems. It should
be noted that all bond lengths and distances are calculated. Figure 1; Figure 2 show the compounds
explored. The compounds in Figure 1 have been selected to have more steric strain and lower or higher
π
-bond order of the CC bond linking the donor and acceptor group than in most systems previously
investigated. The ‘hb and out’ method was chosen for molecules with no disturbing interactions taking
place in the open form. The MP2 method was chosen as this had shown good results previously [ ].
7
Electron densities at the bond critical points [16] have also been related to two-bond deuterium isotope
effects on 13C-NMR chemical shifts.

Molecules 2019, 24, 4533
3 of 15
Figure 1. Experimental two-bond deuterium isotope effects in ppb on 13C chemical shifts (normal font)
and aldehyde and OH proton chemical shifts in ppm (in italics). The structure used in the calculation
for 11 is a short version in which the second half is replaced by a phenyl ring. Data for 11 from Ref. 24
(Reproduced with permission from [24]). For a full set of isotope effects on chemical 13C chemical shifts
see Supplementary Materials Scheme S1.
Figure 2. Other compounds investigated. Isotope effects for compounds 13–23 can be found in [13]
(Reproduced with permission from [13]).

Molecules 2019, 24, 4533
4 of 15
2. Results
2.1. Structures and Energies
Structures were optimized and energies calculated in two different ways: MP2 with the basis set
6-311++G(d,p), or using the B3LYP functional with the same basis set. The structure of gossypol (11) is
extraordinary with a very short O···O distance (2.46 Å) [25]. In 5-nitrosalicylaldehyde (17), the nitro
group is twisted slightly out of the benzene ring plane in the MP2 calculations, whereas it is in the ring
plane in the B3LYP calculations.
The hydrogen-bond energies were estimated by subtracting the computed energies of the
non-hydrogen-bonded forms (OH group turned 180◦) from the energies of the hydrogen-bonded forms
as originally demonstrated for this type of molecules by Cuma, Scheiner, and Kar [12]. Unless otherwise
indicated hydrogen bond energies quoted in the following were computed by this ‘hb and out’ method.
The hydrogen bond energy estimate was computed only for compounds with no disturbing interference
in the rotated forms, meaning that compounds like
1, 2, 5, 7, 9, 11, and 12 are not considered. 3 is also
excluded as the C=O group turns out of the plane in the MP2 calculations in the open form, and
7
is excluded because the structure did not converge in the MP2 calculations. However, a number of
other compounds are included, such as 6-methylsalicylaldehyde (18), 4- and 5-hydroxysalicylaldehyde
(14) and (15), 4-methoxysalicylaldehyde (16) (see Figure 3) and 5-nitrosalicylaldehyde (17),
2-hydroxynaphthaldehyde (22) and 3,6-dimethoxy-1-hydroxy-2-naphthaldehyde (24) (see Figure 2).
6-isopropylsalicylaldehyde (19) and 6-t-butylsalicylaldehyde (20) have also been calculated, although
the compounds have not been synthesized.
Figure 3. Conformers of 14–16.
A possible complication in the application of the ‘hb and out’ method as discussed above is
interactions in the open form. The OH bond lengths of the open forms of the calculated benzene
derivatives are similar for all compounds indicating that no unusual interactions take place. As seen in
Figure S2, a good correlation is observed between the ‘hb and out’ hydrogen bond energies calculated
by the MP2 method and by the B3LYP functional. In this paper, we mostly show plots with MP2 values,
but the trends are the same with B3LYP values.
Energies can also be estimated on the basis of electron densities at the bond critical points,
as discussed for example by Afonin et al. [5]. A good correlation is found with ‘hb and out’ calculated

Molecules 2019, 24, 4533
5 of 15
energies as seen in Figure 4. Steric compression clearly influences the OH···O=C distances as seen in
Figure 5 and energies as demonstrated later.
17
y = 281.2x - 2.0943
R² = 0.93
15
13
11
9
7
5
0.03
0.035
0.04
0.045
0.05
0.055
0.06
0.065
ρ
Figure 4. Hydrogen bond energies (MP2, ‘hb and out’) in kcal/mol vs. electron density at the bond critical
points ( ) in atomic units. Compounds investigated are 10 12 15 17 21, 5-chlorosalicylaldehyde,
and 5-methyl-1,3-benzenedialdehyde.
ρ
5,
7,
8,
,
–
,
,
Figure 5. Experimental deuterium isotope effects on aldehyde proton 1H-NMR chemical shifts
(ppb, in black) and OH···O=C calculated distances (Å, in red).

Molecules 2019, 24, 4533
6 of 15
For the compounds 4-hydroxy- (14), 4-methoxy- (16) and 5-hydroxysalicylaldehyde (15), the ‘hb
and out’ hydrogen bond energies depended on the orientation of the substituent in 4- or 5-position as
discussed in the following. The a-form of Figure 3 has the lowest energy for the hydrogen bonded
ones and the d-form for the open forms. The characteristics of the two forms are that in these forms the
OH- and OR-bonds are more or less perpendicular. Hence, the calculated ‘hb and out’ hydrogen bond
energies depend significantly on the orientation of the substituent, and for 4-methoxysalicylaldehyde
(16) the difference is as large as 1.4 kcal/mol.
For
4 the “doubly” open form has also been calculated. In this case the calculated energy is
19.9 kcal/mol compared to the value of 9.4 kcal/mol for one OH group turned.
2.2. Deuterium Isotope Effects on Chemical Shifts
2.2.1. Deuterium Isotope Effects on 13C Chemical Shifts
Deuterium isotopeeffectsaredefinedas n
the deuterium and the atom in question and exemplified with 13C-NMR. Data for salicylaldehydes
13 17) and hydroxynaphthaldehydes (22 24) are taken from [17 21]. Both types of data have been
∆C=δC(OH)-δC(OD), nbeingthenumberof bondsbetween
(
–
–
,
previously used in the discussion of hydrogen bond strength. The discussion is centered around two-bond
deuterium isotope effects on 13C chemical shifts (TBDIE). A comparison of 4,6-dimethylsalicyaldehyde-
and salicylaldehyde shows larger two-bond isotope effects in the former. In a similar way, for
5,6-dimethyl-2,3-dihydroxy-1,4-benzenedialdehyde and 2,3-dihydroxy-1,4-benzenedialdehyde, the former
shows larger two-bond isotope effects. Gossypol (11) shows a very large TBDIE of 0.85 ppm [22]. This
value is much larger than found in 2-hydroxynaphtaldehyde (21) (0.411 ppm) [21] or in
9 (Figure 2).
Another interesting observation is the almost identical two-bond isotope effects of 5-nitrosalicylaldehyde
(17), 3,5-dinitrosalicylaldehyde (7), and salicylaldehyde (13).
OH chemical shifts have been used to gauge hydrogen bond strength. As seen in Figure 6 a
reasonable correlation is found between OH chemical shifts and TBDIE.
17
16
15
14
13
12
11
10
y = 7.3798x + 9.7354
R² = 0.84
0
0.2
0.4
0.6
0.8
1
TBDIE
Figure 6. OH chemical shifts in ppm vs. two-bond deuterium isotope effects on 13C chemical shifts
(TBDIE) in ppm.
2.2.2. Deuterium Isotope Effects on 1H Chemical Shifts
As seen from Figure 5, deuterium isotope effects on aldehyde proton chemical shifts can be found
in some of the compounds. Although these are small, because the chemical shift range of 1H is small,
they easily can be measured. Isotope effects on aldehyde proton chemical shifts were observed in all
compounds with OH···O=C distances shorter than 1.729 Å (Figure 5).

Molecules 2019, 24, 4533
7 of 15
2.3. Correlations
In Figure 4 energies estimated from the ‘hb and out’ method (MP2) are plotted vs. electron
densities at the bond critical points (ρ). A rather satifactory linear correlation is found; the plot using
B3LYP results is very similar (not shown). A plot of the ‘hb and out’ energies calculated with the MP2
method and with the B3LYP functional vs. observed two-bond deuterium isotope effect (TBDIE) on
13C chemical shifts is shown in Figure 7. Again, good correlations are observed.
22
y = 14.445x + 8.0541
20
R² = 0.94
18
16
14
y = 13.554x + 6.5939
12
R² = 0.96
10
Gossypol
8
0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
TBDIE
Figure 7. Plot of ‘hb and out’ hydrogen bond energies in kcal/mol calculated either with
MP2/6-311++G(d,p) or with B3LYP/6-311++G(d,p) vs. observed two-bond deuterium isotope effect
(TBDIE) on 13C chemical shifts in ppm. Top correlation line is B3LYP, bottom one is MP2. The observed
TBDIE for gossypol is marked by the vertical blue line.
Two-bond deuterium isotope effects on 13C-NMR chemical shifts (TBDIE) are claimed to be good
measures for hydrogen bond strength [11]. This claim is supported by the results shown in Figure 7.
In Figure 8 TBDIE’s are plotted vs. electron densities at the bond critical point. A much larger array of
compounds can then be included. It is seen, that especially data for 5 and 10 are falling off the line.
0.7
0.6
y = 15.666x - 0.4223
0.5
R² = 0,83
0.4
0.3
0.2
0.1
0
0.03
0.035
0.04
0.045
0.05
0.055
0.06
0.065
Electron density at bond critical points
Figure 8. Experimental TBDIE’s vs. electron densities at bond critical points.

Molecules 2019, 24, 4533
8 of 15
In Figure 9, OH bond lengths show excellent correlation with hydrogen bond energies. The main
outlier is , a compound with two hydrogen-bonded systems. A plot of the hydrogen bond energies vs.
4
the O···O distances is given in Figure 10. The correlations are reasonable.
0.994
0.992
y = 0.0023x + 0.9563
R² = 0.95
0.99
0.988
0.986
0.984
0.982
0.98
0.978
0.976
0.974
8
9
10
11
12
13
14
15
16
Hydrogen bond energy
Figure 9. Calculated OH bond lengths in Å vs. ‘hb and out’ hydrogen bond energies in kcal/mol (MP2).
The plot of Figure 10 shows that in general a shorter the O···O distance is related to a higher
hydrogen bond energy, but also that it is necessary to divide the compounds into two groups, sterically
hindered and non-sterically hindered. The two lower outliers among the sterically hindered ones
are 6-t-butyl- (20) and 6-isopropylsalicylaldehyde (19), whereas the one above is the phenanthrene
derivative (10).
18
17
y = -59.578x + 168.46
R² = 0.85
16
15
14
13
12
11
10
9
10
Zone 3
Zone 2
20
19
Zone 1
y = -39.17x + 113.03
R² = 0.67
8
8
2.5
2.52
2.54
2.56
2.58
2.6
2.62
2.64
2.66
2.68
2.7
R(O...O)
Figure 10. Hydrogen bond ‘hb and out’ energies (kcal/mol) vs. O···O distances (Å) computed with MP2.
The data are divided into non-sterically hindered compounds (black diamonds, upper correlation line)
with compounds like 4, 8, 13–17 and sterically hindered ones (grey squares, lower correlation line).

Molecules 2019, 24, 4533
9 of 15
Finally, a plot of experimental TBDIE vs. calculated O···O distances is shown in Figure 11. This
allows inclusion of compounds like 11 and 12. The corresponding hydrogen bond energies
1, 2, 5, 7, 9,
in the plots in Figures 6–9 and 11 are for Boltzmann averaged values, whereas the structural parameters
are for the a-form.
0.9
0.7
0.5
y = -1.9149x + 5.2785
R² = 0.49
O
H
0.3
0.1
H
O
y = -4.5776x + 12.12
R² = 0.92
2.45
2.5
2.55
2.6
2.65
2.7
-0.1
-0.3
R(O···O)
Figure 11. Plot of observed TBDIE vs. O···O distances (Å) computed with MP2. As in Figure 10, black
diamonds refer to non-sterically hindered, grey squares to sterically hindered cases. The former include
compounds like 4 and 8, as well as 13–17.
3. Discussion
One of the aims of the present paper is to relate deuterium isotope effects on chemical shifts
to hydrogen bond strength and to describe some of the parameters contributing to hydrogen bond
strength in intramolecularly hydrogen-bonded systems. No experimental hydrogen bond energies are
available for intramolecularly hydrogen-bonded systems of the kind studied here. In the present study,
the hydrogen bond strength is estimated using: (i) the hydrogen bond minus the OH out method
(‘hb and out’) [12] and (ii) electron density at the bond critical point. The compounds included in the
former type of calculations are only those with no special interactions of the OH group when this is
turned 180◦ around the C-O axis. The main question is whether this method is accurate. In the case of
4
both the conformer with one group turned out and the one with two groups turned out have been
studied. As seen in the Results section, the energies are not strictly additive; but it must in this context
be remembered, that different orientation of substituents could lead to a variation. It was found (Figure
S2) that the MP2/6-311++G(d,p) and the B3LYP/6-311++G(d,p) procedures gave linearly related results,
but the B3LYP energies are approximately 1.4 kcal/mol higher (Figure 7).
Rusinska-Roszak [15] has used the molecular tailoring approach (MTA) for a series of o-hydroxy
aromatic aldehydes, but mostly in a narrow range of hydrogen bond energies as seen in Figure 12.
Rusinska-Roszak did not include PAH systems in the MTA analysis; the reason is stated as “polycyclic
compounds are not included due to their dependence on the position of the -OH and C=O substituents”.
Translated this could mean “because of different degrees of conjugation”. But this is true for most
of the compounds investigated by Rusinska-Roszak and certainly affects compound
5 (the point at
0.43 ppm in Figure 12). The molecular tailoring approach method gave mostly smaller hydrogen bond
energies than MP2 and B3LYP, but the trends are clearly the same for all three types of calculations
(MP2, B3LYP and MTA). The range of energies in our investigation is ~4 kcal/mol. Despite the variation
for some compounds due to the orientation of substituents it can be seen that in general, a shorter
O···O distance is related to a higher hydrogen bond energy (Figure 10).

Molecules 2019, 24, 4533
10 of 15
16
15
14
13
12
11
10
9
y = 17.419x + 3.907
R² = 0.96
8
7
0
0.1
0.2
0.3
0.4
0.5
0.6
0.7
TBDIE
Figure 12. Plot of hydrogen bond energies in kcal/mol calculated by the molecular tailoring approach
(MTA) vs. observed TBDIE in ppb. Energies from [15] (Reproduced with permission from [15]).
From the plot if Figure 7 it can also be seen that two-bond deuterium isotope effects (TBDIE)
on 13C chemical shifts can be used to predict hydrogen bond strengths and the same is true for OH
chemical shifts (Figure S4). Obviously, these two parameters show a decent correlation (Figure S3).
Two-bond deuterium isotope effects on 13C chemical shifts depend to a large extent on the sum of bond
orders between the OD deuterium and the carbon in question. The sums of bond orders will in the
present compounds increase if the importance of the resonance form c of Figure 13 increases. A short
C2-O distance (for an example see Figure 13) will be favored by a short C1-C2 distance due to increased
RAHB, but also by a short O···O distance as coulomb interactions will be maximized. Therefore, it is
expected, as seen in Figure 7, that the TBDIE is a reasonable measure of hydrogen bond strength.
Figure 13. Resonance structures of salicylaldehyde, 5-nitrosalicyladehyde and 4-alkoxysalicylaldehydes.
1
The observation of deuterium isotope effects on the aldehyde H resonance (see Figure 5) in
compounds with short OH···O=C distances is indicative of orbital overlap. This overlap will likewise
strengthen the hydrogen bond.

Molecules 2019, 24, 4533
11 of 15
Factors relevant to intramolecular hydrogen bonding are the interaction between the donor and
the acceptor referred to as RAHB, the resonance from electron withdrawing or electron donating
substituents [26], coulomb interactions, and orbital overlap. In this respect the O···O distance is an
important factor. The significance of the O···O distance has been discussed intensely; some think it is
very important [23], others that it is not a good measuring stick for hydrogen bond strength [2].
To a large extent, the O···O distance is determined by the length of the “double” bond connecting the
donor and acceptor. In the corresponding hydrocarbons we find the following bond lengths: benzene
1.40 Å, C1-C2 bond of naphthalene 1.36 Å, C2-C3 bond of naphthalene 1.42, and C9-C10 of phenanthrene
1.34 Å. We have divided the data into two groups: non-steric (such as 3-hydroxy-2-naphthaldehyde,
salicylaldehyde, 4- and 5-substituted salicylaldehydes and 2,3-dihydroxy-1,4-benzenedialdehyde
and 1-hydroxy-2-naphtaldehyde) and steric (4,6-dimethylsalicylaldehyde, 6-methoxysalicylaldehyde,
2-hydroxy-1-naphtaldehyde and 10-hydroxy-9-phenanthrenealdehyde). The trend from Figure 10
is rather clear: a short O···O distance is generally found in compounds with large hydrogen bond
energy. Leading to a short distance is steric pressure as in
it can be described in more detail. In the plot of Figure 10 three “zones” can be recognized for the
non-steric ones: compound (first zone), followed by the benzene derivatives (zone 2), and at higher
2, 3, 5, 6, 9, and 10, as well as in 21. However,
8
energies the naphthalene derivative 1-hydroxy-2-naphthalene (zone 3). The sterically hindered benzene
derivatives are shifted to an O···O distance of ~2.61Å, the sterically hindered naphthalenes to ~2.56 Å,
the phenanthrene (10) to ~2.52 Å and finally, at the highest energy, is found 11. The bond length also
correlates with the
therefore with the resonance assistance.
The hydrogen bond energy of is 1.25 kcal/mol higher than that of salicylaldehyde. The hydrogen
bond energy of is similar to that of 6-methylsalicylaldehyde. For 6-isopropyl-salicylaldehyde the
energy is only 0.2 kcal/mol higher than for , but in the case of 6-t-butyl-salicylaldehyde (20) the
energy is 1 kcal/mol higher than for
, and the O···O distance is 0.05 Å shorter. Another example
is gossypol (11). This compound has a very large TBDIE of 0.85 ppm (Figure 11). It has also
been reasoned that the compound does not show tautomeric behavior despite the very short O···
π-bond order of the bond connecting the hydrogen bond donor and acceptor and
6
6
6
6
O
distance calculated as 2.47 Å (B3LYP) compared to the X-ray distance of 2.46 Å [25]. From the plot of
Figure 7, a hydrogen bond energy of 18-20 kcal/mol can be predicted. It is obvious that part of the
reason for the short O···O distance is the short C1-C2 bond, but in the similar naphthalene derivative
2-hydroxynaphtaldehyde the O···O distance is 2.597 Å. Can the extra substituents be responsible for
the short distance? In 2,3-dihydroxybenzaldehyde (12) the O···O distance is actually longer than in
salicylaldehyde. The hydroxyl group in position 8 cannot contribute mesomerically. Gossypol is
correctly a ‘dimer’. This is in the present situation mimicked by a phenyl group in position 7. This
phenyl group is twisted very heavily out of the naphthalene ring plane (76◦) and does not contribute
mesomerically, so the short O···O distance has to come from steric compression of the OH group in
position 8. Again it is seen that steric compression leads both to a shorter O···O distance and to a higher
hydrogen bond energy, but the relationship only holds on a coarse scale.
4. Materials and Methods
4.1. General Information
3,5-Dimethylphenol, 9-bromophenanthrene, 2-methoxynaphthalene, 3,5-dinitrosalicylaldehyde,
and anhydrous magnesium chloride were purchased from TCI Research Chemicals, Haven, Belgium.
All other reagents and solvents were analytical grades purchased from Sigma-Aldrich Chemical Co.,
and used as received unless otherwise stated. Fluka silica gel/TLC-cards 60,778 with fluorescence
indicator 254 nm were used for TLC chromatography. Merck silica gel 60 (0.040–0.063 mm, Merck,
1
Darmstadt, Germany was used for flash chromatography purification of the products. H-NMR and
13C-NMR spectra were recorded at 500 MHz and 126 MHz or 400 MHz and 100 MHz on an Ultrashield
Plus 500 or 400 spectrometer, Bruker, Fallaenden, Germany using CDCl3 or DMSO-d6 as a solvent

Molecules 2019, 24, 4533
12 of 15
and TMS as internal standard. LC/MS (ESI) was carried out on a Bruker MicrOTOF-QIII-system with
ESI-source with nebulizer 1.2 bar, dry gas 10.0 L/min, dry temperature 220 ◦C, capillary 4500 V, end
plate offset −500 V, Bruker, Hamburg, Germany.
4.2. Synthesis
The synthesis of the compounds 2,3-dihydroxy-1,4-benzenedicarboxaldehyde (
-5,6-dimethyl-1,4-benzenedicarboxaldehyde ( ) [28], 2,4-dihydroxy-1,3-benzene- dicarboxaldehyde
) [29], 4,6-dihydroxy-1,3-benzenedicarboxaldehyde ( ) [29], 2,4,6-trihydroxy-1,3,5-benzenetricarboxal
dehyde ( ) [29], and 2,3-dihydroxy-1,4-naphthalenedicarboxaldehyde ( ) [28] are described previously.
1) [27], 2,3-dihydroxy
2
(3
4
5
9
4,6-Dimethylsalicylaldehyde ( ): Compound was prepared according to the procedure described by
6
6
Knight et al. [30]. A solution of 3,5-dimethylphenol (1.22 g, 10 mmol) in dry acetonitrile (50 mL) was
treated with dry triethylamine (5.22 mL, 37.5 mmol) and anhydrous magnesium chloride (1.43 g,
15 mmol) under nitrogen atmosphere. The mixture was stirred for 15 min. Dry paraformaldehyde
(2.03 g, 6.75 mmol) was added under nitrogen atmosphere and the mixture refluxed for 3 h. The solution
cooled to room temperature and poured into a cold 5% aqueous HCl solution and left stirred for 45 min.
This was extracted with diethyl ether (4
× 40 mL), the ether washed with brine (2 × 30 mL) and dried over
anhydrous magnesium sulphate. The solvent was evaporated under reduced pressure and the crude
product was purified by flash column chromatography on silica gel using dichloromethane/n-hexane
(4:1) as eluent, affording the pure product as white crystals (1.00 g, 66%). The NMR data are
consistent with those reported in literature [22]. HRMS-ESI calculated for C9H10O2 [M + H]+ 150.0680,
found 150.0658.
3-Hydroxy-2-naphthaldehyde (8): The starting material 3-methoxy-2-naphthaldehyde was prepared according
to the procedure described by Legouin et al. [31]. The crude product was purified by flash column
chromatography on silica gel using dichloromethane to afford pure product as a pale yellow solid (0.59 g,
56%). The NMR data are consistent with that reported in literature [31]. HRMS-ESI calculated for
C12H10O2 [M + H]+ 186.0681, found 186.0645. Next a flame-dried 50 mL round-bottom flask equipped
with a magnetic stirrer, nitrogen inlet was charged with a solution of 3-methoxy-2-naphthaldehyde
(0.21 g, 1.5 mmol) in dry dichloromethane (20 mL) and cooled to 0 ◦C. BBr3 (6.0 mL, 1.0 M solution
◦
in dichloromethane) was added under a nitrogen atmosphere and was stirred at 0 C for 15 min,
and at room temperature for 20 h. The reaction mixture was poured into ice and 1.0 M HCl (10 mL).
The mixture was stirred for 20 min and extracted with dichloromethane (2 × 30 mL), dried over
anhydrous magnesium sulphate and evaporation of solvent under reduced pressure gave a crude
product purified by flash column chromatography on silica gel, using dichloromethane to affords
pure product as a pale yellow solid (0.14 g, 73%). The NMR data are consistent with that reported in
literature except that the two signals at 10.32 ppm must be assigned as the OH resonance and that
at 10.08 ppm as the resonance of the aldehyde proton. HRMS-ESI calculated for C11H8O2 [M + H]+
172.0524, found 172.0538.
9-Methoxyphenanthrene: This compound was synthesized according to previously reported procedure [31].
The light amber solid was purified by flash column chromatography on silica gel, using dichloromethane/
n-hexane (2:1) to afford the pure product as off white solid (1.2 g, 91%). The NMR data are consistent with
that reported in the literature [31]. HRMS-ESI calculated for C15H12O [M + H]+ 208.0888, found 208.0858.
10-Methoxy-9-phenanthrenealdehyde: This compound was synthesized by modification of literature
methods [31]. The resulting light brown solid residue was purified by flash column chromatography
on silica gel, using an eluent gradient of n-hexane/dichloromethane/(2:1), then (1:1) and finally (1:4) to
give a white solid (0.83 g, 67%). The NMR data are in agreement with those of Ref. 31. HRMS-ESI
calculated for C16H12O2 [M + H]+ 236.0837, found 236.0816.

Molecules 2019, 24, 4533
13 of 15
10-Hydroxy-9-phenanthrenealdehyde (10): This compound was synthesized by modification of the
procedure of [31]. A flame dried 50 mL round-bottom flask equipped with a magnetic stirrer and
a nitrogen inlet was charged with a solution of 10-methoxy-9-phenanthrenecarboxaldehyde (0.61 g,
2.6 mmol) in dry dichloromethane (25 mL) and cooled to
10 ◦C. Under nitrogen atmosphere was
−
added a solution of BBr3 (6.0 mL, 1.0 M solution in dichloromethane) and stirred at 0 ◦C for 15 min
and at room temperature for 1.5 h. The reaction mixture was poured onto ice in 1.0 M HCl (30 mL).
The mixture was stirred for 30 min and extracted with dichloromethane (2
×
50 mL), dried over
anhydrous magnesium sulphate and evaporation of solvent under reduced pressure gave a crude
product purified by flash column chromatography on silica gel, using dichloromethane/n-hexane (2:1)
as eluent to afford the pure product as a pale yellow solid (0.54 g, 96%). The NMR results are similar to
those reported [31]. HRMS-ESI calculated for C15H10O2 [M + H]+ 222.0681, found 222.0652.
4.3. Deuteration
The compounds are deuterated in CDCl3 by stirring the dissolved compound with a mixture of
D2O:H2O. After stirring over night the water phase is removed and the CDCl3 phase is dried with
anhydrous sodium sulfate.
4.4. Calculations
Quantum chemical calculations were performed using the Gaussian09 software package
(Gaussian.com, Wallingford, CT, USA) [32]. Gas phase equilibrium geometries and model hydrogen bond
energies of the investigated molecules were computed with either B3LYP [33,34] density functional theory
or MP2 Møller-Plesset second order perturbation theory [35 36] using the 6-311++G(d,p) basis set [37
,
–39]
and default options [32]. Calculated vibrational frequencies were checked for imaginary values. For a
number of species, the geometry optimization failed to converge with MP2/6-311++G(d,p) in spite of
several attempts, a problem apparently associated with the inclusion of diffuse functions in the basis set
(no problems were observed with MP2/6-311G(d,p)). Theoretical estimates of the intramolecular hydrogen
bond energies were obtained by the ‘hb and out’ model: For each species, two geometry optimizations
were performed, one for the hydrogen-bonded form, and one for the open form where the OH group is
rotated by 180◦ around the C-O axis. The energy difference between the two forms is taken as an estimate
of the energy of the hydrogen bond. The AIM analysis was based on B3LYP/6-311++G(d,p) calculations
using the AIMALL17.11 version (AIMAll, Overland Park, KS, USA) [40].
5. Conclusions
It is seen that two-bond isotope effects on 13C chemical shifts are a good measure of hydrogen
bond strength for this type of system, whether the hydrogen bond energy is calculated by the ‘hb and
out’ method or estimated from electron densities at bond critical points. Two-bond deuterium isotope
effects on 13C chemical shifts is an experimental parameter that can easily be measured with very few
exceptions. In general, strong hydrogen bonds tend to have short O···O distances. The O···O distance
can be modified either by shortening the bond connecting the hydrogen bond donor and the acceptor,
by conjugation, or by steric compression of the aldehyde group. Compression of the OH group has no
significant effect. Judging from Figures 10 and 11, steric compression seems slightly less effective than
short “double” bonds and conjugation.
Supplementary Materials: The following are available online, Scheme S1: Deuterium isotope effects on 13C
chemical shifts. Figure S2: Plot of hydrogen bond energies, B3LYP vs. MP; Figure S3: Observed OH chemical
shifts in ppm vs. hydrogen bond energies; Table S1: Calculated hydrogen bond energies hb and out method.
Author Contributions: P.E.H. have proposed the idea and conducted the experiments; J.S.-L. and B.A.S. have
primarily done calculations; F.S.K. and M.J.M. are responsible for the synthesis of compounds. All authors have in
the end contributed to the manuscript.
Funding: This work has no external fundings.

Molecules 2019, 24, 4533
14 of 15
Acknowledgments: The authors wish to thank Annette Christensen and Britt Willer Clemmensen for their expert
help in recording of the NMR spectra.
Conflicts of Interest: The authors declare no conflicts of interest.
References
1.
2.
3.
Perrin, C.L.; Nielson, J.B. “Strong” Hydrogen Bonds in Chemistry and Biology. Annu. Rev. Phys. Chem. 1997
48, 511–544. [CrossRef] [PubMed]
Perrin, C.L. Are Short, Low-Barrier Hydrogen Bonds unusually Strong? Acc. Chem. Res. 2010, 43, 1550–1557.
[CrossRef] [PubMed]
Hansen, P.E.; Spanget-Larsen, J. NMR and IR investigations of strong intramolecular hydrogen bonds.
Molecules 2017, 22, 552. [CrossRef] [PubMed]
,
4.
5.
Scheiner, S. Special issue: Intramolecular Hydrogen Bonding 2017. Molecules 2017, 22, 1521. [CrossRef]
Afonin, A.V.; Vashenko, A.V.; Sigalov, M.V. Estimating the energy of intramolecular hydrogen bonds from
1H-NMR and QTAIM calculations. Org. Biomol. Chem. 2016, 14, 11199–11211. [CrossRef]
Gilli, P.; Pretto, L.; Bertolasi, V.; Gilli, G. Predicting Hydrogen-Bond Strength from Acid-Base Molecular
properties. The pKa slide rule: Toward the Solution of a Long-Lasting problem. Acc. Chem. Res. 2009, 42,
33–44. [CrossRef]
6.
7.
8.
Grabowski, S.J. An estimation of strength of intramolecular hydrogen bonds-ab initio and AIM studies.
J. Mol. Struct. 2001, 562, 137–143. [CrossRef]
Guillaumes, L.; Simon, S.; Fonseca Guerra, C. The Role of Aromaticity, Hybridization, Electrostatics,
and Covalency in Resonance-Assisted Hydrogen Bonds of Adenine-Thymine (AT) Base Pairs and Their
Mimics. ChemistryOpen 2015, 4, 318–327. [CrossRef]
9.
Sharif, S.; Huot, M.C.; Tolstoy, P.M.; Toney, M.D.; Jonsson, K.H.M.; Limbach, H.-H. 15N Nuclear magnetic
Resonance Studies of Acid-Base properties of Pyridoxal-50-phospate Aldimines in Aqeous solution. J. Phys.
Chem. B 2007, 111, 3869–3876. [CrossRef]
10. Arunan, E.; Desiraju, G.R.; Klein, R.A.; Sadlej, J.; Scheiner, S.; Alkorta, I.; Clary, D.C.; Crabtree, R.H.; Dannenberg, J.J.;
Hobza, P.; et al. Definition of the hydrogen bond (IUPAC Recommendation 2011). Pure Appl. Chem. 2011, 83,
1537–1641. [CrossRef]
11. Hansen, P.E. NMR Studies of Compounds with Intramolecular Hydrogen bonds. In Isotope Effects in Chemistry
and Biology; Kohen, A., Limbach, H.-H., CRC, Eds.; Taylor and Francis: Milton Park, UK, 2006; pp. 253–280.
12. Cuma, M.; Scheiner, S.; Kar, T. Competitionbetweenrotamerizationandprotontransferino-hydroxybenzaldehyde.
J. Am. Chem. Soc. 1998, 120, 10497–10503. [CrossRef]
13. Grabowski, J.J. Theoretical studies of strong hydrogen bonds. Annu. Rep. Prog. Chem. Sect. C 2006, 102,
131–165. [CrossRef]
14. Gadre, S.R.; Shirsat, R.N.; Limaye, A.C. Molecular Tailoring approach for Simulation of electrostatic Properties.
J. Phys. Chem. 1994, 98, 9165–9169. [CrossRef]
15. Rusinska-Roszak, D. Energy of Intramolecular Hydrogen Bonding in ortho-Hydroxybenzaldehydes, Phenones
and Quinones. Transfer of Aromaticity from ipso-Benzene ring to the Enol System(s). Molecules 2017, 22, 481.
[CrossRef] [PubMed]
16. Bader, R.F.W. Atoms in Molecules. A Quantum Theory; Oxford University Press: New York, NY, USA, 1990.
17. Reuben, J. Intramolecular Hydrogen Bonding as Reflected in the Deuterium Isotope effects on Carbon-13
Chemical shifts. Correlation with Hydrogen bond Energies. J. Am. Chem. Soc. 1986, 108, 1735–1738. [CrossRef]
18. Schaefer, T. A Relationship between Hydroxyl proton Chemical shifts and torsional Frequencies in some
Ortho-Substituted phenol Derivatives. J. Phys. Chem. 1975, 79, 1888–1890. [CrossRef]
19. Hansen, P.E.; Tüchsen, E. Deuterium Isotope Effects on Carbonyl Chemical Shifts of BPTI. Hydrogen-bonding
and Structure Determination in Proteins. Acta Chem. Scand. 1989, 43, 710–712. [CrossRef]
20. Tüchsen, E.; Hansen, P.E. Hydrogen Bonding Monitored by Deuterium Isotope Effects on Carbonyl 13
C
Chemical Shifts in Basic Pancreatic Trypsin Inhibitor. Intra Residue Hydrogen Bonds in Antiparallel ß-sheet.
Int. J. Biol. Macromol. 1991, 13, 2–8.
21. Hansen, P.E. Isotope Effects on Nuclear Shielding. Intra-molecular Hydrogen- bonded Ketones, Aldehydes
and Esters. Magn. Reson. Chem. 1993, 31, 23–37. [CrossRef]

Molecules 2019, 24, 4533
15 of 15
22. Hansen, P.E.; Bolvig, S.; Wozniak, K. Steric compression and twist in o-hydroxy acyl aromatics with
intramolecular hydrogen bonding. J. Mol. Struct. 2005, 749, 155–168. [CrossRef]
23. Sanz, P.; Mó, O.; Yanez, M.; Elguero, J. Resonance-assisted hydrogen bonds: A critical examination. Structure and
stability of the enols of beta-diketones and beta-enaminones. J. Phys. Chem. A 2007, 11, 3585–3591. [CrossRef]
24. That, Q.T.; Phung, K.P.; Hansen, P.E. Schiff Bases of Gossypol. A NMR and DFT Study. Magn. Reson. Chem.
2005, 43, 302–308. [CrossRef]
25. Gdaniec, M.; Ibragimov, B.T.; Talipov, S.A. Lattice inclusion-compounds of Gossypol-Structure of the
2,3-Gossypol-Benzaldehyde Coordinatoclathrate. Acta Crystallogr. 1991, C47, 573–577.
26. Pareras, G.; Palusiak, M.; Duran, M.; Solà, M.; Simon, S. Tuning of strength of the Resonance-assisted
Hydrogen Bond in o-Hydroxybenzaldehyde by Substitution in the Aromatic Ring. J. Phys. Chem. A 2018
122, 2279–2287. [CrossRef] [PubMed]
,
27. Gallant, A.J.; Yun, M.; Sauer, M.; Yeung, C.S.; Maclachlan, M. Tautomerization in Naphthalenediimines:
A Keto-Enamine Schiff Base Macromolecule. Org. Lett. 2005, 7, 4827–4830. [CrossRef]
28. Shopsowitz, K.E.; Edwards, D.; Gallant, A.J.; MacLachlan, M.J. Highly substituted Schiff base microcycles via
hexasubstituted benzene: A convenient double Duff formylation of catechol derivatives. Tetrahedron 2009, 65,
8113–8119. [CrossRef]
29. Sauer, M.; Yeung, C.; Chong, J.H.; Patrick, B.O.; Maclachlan, M.J. N-Salidylidenes: Tautomers for Formation
of Hydrogen-Bonded Capsules, clefts and Chains. J. Org. Chem. 2005, 71, 775–788. [CrossRef]
30. Knight, P.D.; Clarkson, G.; Hammond, M.L.; Kimberley, B.S.; Scott, P. Radical and migratory insertion
reaction mechanisms in Schiff base zirconium alkyls. J. Organomet. Chem. 2005, 690, 2125–5144. [CrossRef]
31. Legouin, B.; Gayral, M.; Uriac, P.; Cupifia, J.-F.; Levoin, N.; Toupet, L.; van de Weghe, P. Molecular Tweezers:
Synthesis and Formation of Host-Guest complexes. Eur. J. Org. Chem. 2010, 28, 5503–5508. [CrossRef]
32. Frisch, M.J.; Trucks, G.W.; Schlegel, H.B.; Scuseria, G.E.; Robb, M.A.; Cheeseman, J.R.; Scalmani, G.; Barone, V.;
Mennucci, B.; Petersson, G.A.; et al. Gaussian09, Revision D.01; Gaussian, Inc.: Wallingford, CT, USA, 2009.
33. Becke, A.D. Density-Functional Thermochemistry. III. The Role of Exact Exchange. J. Chem. Phys. 1993, 98,
5648–5652.
34. Lee, C.; Yang, W.; Parr, R.G. Development of the Colle-Salvetti Correlation-Energy Formula into a Functional
of the Electron Density. Phys. Rev. B 1988, 37, 785–789. [CrossRef] [PubMed]
35. Head-Gordon, M.; Pople, J.A.; Frisch, M.J. Development of the Colle-Salvetti Correlation-Energy Formula
into a Functional of the Electron Density. Chem. Phys. Lett. 1988, 153, 503–506. [CrossRef]
36. Møller, C.; Plesset, M. Note on an Approximation Treatment for Many-Electron Systems. Phys. Rev. 1934, 46,
618–622. [CrossRef]
37. Ditchfield, R.; Hehre, W.J.; Pople, J.A. Self-consistent Molecular-orbital methods. 9. Extended Gaussian-type
Basis for Molecular-Orbital Studies of Organic Molecules. J. Chem. Phys. 1971, 54, 724–728. [CrossRef]
38. Clark, T.; Chandrasekhar, J.; Spitznagel, G.W.; Schleyer, P.V.R. Efficient diffuse function-augmented basis-sets
for anion calculations. 3. The 3-21+G basis set for 1st-row elements, Li-F. J. Comp. Chem. 1983, 4, 294–301.
[CrossRef]
39. Frisch, M.J.; Pople, J.A.; Binkley, J.S. Self-consistent Molecular-orbital methods. 25. Supplementary functions
for Gaussian-Basis sets. J. Chem. Phys. 1984, 80, 3265–3269. [CrossRef]
40. AIMAll (17.11.14), Todd, A. Keneth, TK Gristmill Software. Available online: aim.tkgristmill.com (accessed on
4 December 2019).
Sample Availability: Samples of the compounds may be available from the authors. Please inquire.
©
2019 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).