Experimental and theoretical investigations into the stability of cyclic aminals

Article abstract of DOI:10.3762/bjoc.12.221

Background: Cyclic aminals are core features of natural products, drug molecules and important synthetic intermediates. Despite their relevance, systematic investigations into their stability towards hydrolysis depending on the pH value are lacking. Results: A set of cyclic aminals was synthesized and their stability quantified by kinetic measurements. Steric and electronic effects were investigated by choosing appropriate groups. Both molecular mechanics (MM) and density functional theory (DFT) based studies were applied to support and explain the results obtained. Rapid decomposition is observed in acidic aqueous media for all cyclic aminals which occurs as a reversible reaction. Electronic effects do not seem relevant with regard to stability, but the magnitude of the conformational energy of the ring system and pK_a values of the N-3 nitrogen atom. Conclusion: Cyclic aminals are stable compounds when not exposed to acidic media and their stability is mainly dependent on the conformational energy of the ring system. Therefore, for the preparation and work-up of these valuable synthetic intermediates and natural products, appropriate conditions have to be chosen and for application as drug molecules their sensitivity towards hydrolysis has to be taken into account.

Full text of DOI:10.3762/bjoc.12.221

Experimental and theoretical investigations into
the stability of cyclic aminals
Edgar Sawatzky1, Antonios Drakopoulos1, Martin Rölz1, Christoph Sotriffer1,
Bernd Engels2 and Michael Decker*1
Full Research Paper
Open Access
Address:
Beilstein J. Org. Chem. 2016, 12, 2280–2292.
1Pharmazeutische und Medizinische Chemie, Institut für Pharmazie
und Lebensmittelchemie, Julius-Maximilians-Universität Würzburg,
Am Hubland, D-97074 Würzburg, Germany and 2Institut für
Physikalische und Theoretische Chemie,
doi:10.3762/bjoc.12.221
Received: 09 August 2016
Accepted: 12 October 2016
Published: 31 October 2016
Julius-Maximilians-Universität Würzburg, Emil-Fischer-Straße 42,
D-97074 Würzburg, Germany
This article is dedicated to Professor Gerhard Bringmann on the occasion
of his 65th birthday.
Email:
Michael Decker* - Michael.Decker@uni-wuerzburg.de
Associate Editor: P. R. Schreiner
* Corresponding author
© 2016 Sawatzky et al.; licensee Beilstein-Institut.
License and terms: see end of document.
Keywords:
hydrolysis; kinetics; molecular mechanics; natural products; quantum
mechanics
Abstract
Background: Cyclic aminals are core features of natural products, drug molecules and important synthetic intermediates. Despite
their relevance, systematic investigations into their stability towards hydrolysis depending on the pH value are lacking.
Results: A set of cyclic aminals was synthesized and their stability quantified by kinetic measurements. Steric and electronic
effects were investigated by choosing appropriate groups. Both molecular mechanics (MM) and density functional theory (DFT)
based studies were applied to support and explain the results obtained. Rapid decomposition is observed in acidic aqueous media
for all cyclic aminals which occurs as a reversible reaction. Electronic effects do not seem relevant with regard to stability, but the
magnitude of the conformational energy of the ring system and pKa values of the N-3 nitrogen atom.
Conclusion: Cyclic aminals are stable compounds when not exposed to acidic media and their stability is mainly dependent on the
conformational energy of the ring system. Therefore, for the preparation and work-up of these valuable synthetic intermediates and
natural products, appropriate conditions have to be chosen and for application as drug molecules their sensitivity towards hydroly-
sis has to be taken into account.
Introduction
The aminal system (N,N-acetal) is the structurally equivalent the naturally occurring alkaloids tetraponerine T1 to T8 from
analogue of the O,O-acetal. In the literature this moiety is found the venom of the New Guinean ant Tetraponera sp. [1,2]. It is
as a core element in various important structures, for example in also present in ligands of ruthenium-based catalysts for meta-
2280

Beilstein J. Org. Chem. 2016, 12, 2280–2292.
thesis [3], in imidazolidines acting as antiprotozoal and antibac- Especially the aminal-bearing tetrahydroquinazoline system is
terial agents [4,5], in Tröger’s base derivatives with diverse ap- of great interest as a structure derived from the quinazolinone
plications [6-14] (e.g., asymmetric catalysis, supramolecular and quinazoline cores. These represent privileged structures
chemistry, DNA intercalation, etc.) or in synthetic tetrahydro- with various applications in medicinal chemistry [20,21]. The
quinazolines as cholinesterase (ChE) inhibitors [15-17] as well syntheses of tetrahydroquinazolines are well described using
as ChE inhibitors based on the scaffold of the naturally different approaches: the majority relies on the direct α-amina-
occurring alkaloid physostigmine from the calabar bean tion of o-aminobenzaldehydes with heating or microwave
physostigma venenosum [18,19] (Figure 1). This is just a small irradiation [22-24], by condensation of diamines with alde-
overview and selection of examples out of numerous com- hydes or ketones yielding bicyclic structures [25-27], or by the
pounds incorporating the aminal system as the essential struc- reduction of the corresponding dihydroquinazolinones [15-17]
tural feature.
(Scheme 1).
Figure 1: Compounds described in the literature containing an aminal core for various applications. The aminal structure is highlighted in red.
Scheme 1: Synthetic approaches for the formation of the tetrahydroquinazoline moiety. Dashed lines indicate both cyclized or non-cyclized com-
pounds and the aminal structure is highlighted in red.
2281

Beilstein J. Org. Chem. 2016, 12, 2280–2292.
The aminal templates obtained can be used as starting materials applying NaBH4 onto the tetrahydroquinazoline-based aminal
for the synthesis of a broad spectrum of diverse structures. Oxi- systems were investigated, thereby providing the possibility to
dation reactions with KMnO4 or a mixture of potassium iodide “open” the rigid aminal core gaining access to sterically more
and tert-butyl hydroperoxide (TBHP) give access to quinazoli- flexible compounds [31-33] (Scheme 2).
nones and have been reported for the synthesis of the naturally
occurring alkaloids deoxyvasicinone, mackinazolinone or rutae- Although the chemistry of aminals – especially concerning
carpine [22,28] (Scheme 2). Besides total oxidation of the tetrahydroquinazolines – is in the focus of current research and
aminal core, also a partial oxidation towards 3,4-dihydroquina- includes their preparation and modification, it is remarkable that
zolines is possible for the case that the oxidation is promoted there is only little focus on the pH stability of these compounds
either by Cu(AcO)2 or by a mixture of elemental iodine and in aqueous media. In general, the aminal moiety is known to
BuLi. These methods allow the synthesis of the partially unsatu- undergo (similar to the corresponding O,O-acetals) acidic
rated alkaloids vasicine or deoxyvasicine in good yields [22,28] hydrolysis and can be considered as stable only within a certain
(Scheme 2). Furthermore, copper-catalysed reactions or oxida- pH range. This is of enormous importance especially for
tion with sodium hypochlorite were also described to yield the aminal-functionalities-bearing compounds which might be
aromatic quinazoline core [26,29,30] (Scheme 2). Besides all exposed to an acidic environment, e.g., drugs that are orally
the oxidation reactions described, also reductive conditions applied and get into direct contact with gastric acid. The pH
Scheme 2: Oxidation and reduction reactions of tetrahydroquinazolines. Dashed lines indicate both cyclized or non-cyclized compounds and the
aminal structure is highlighted in red.
2282

Beilstein J. Org. Chem. 2016, 12, 2280–2292.
stability of the aminal system is also a key property for All synthesized compounds were exposed to aqueous media at
synthetic approaches of such compounds to prevent undesired defined pH values and the time-dependent hydrolysis into the
decomposition during reaction or work-up. Knowledge about corresponding diamines 2 and aldehydes 3 (Figure 2) was quan-
the pH stability can therefore help to improve the yield tified by reversed-phase HPLC. Based on these data, quantum
during a reaction, or to prevent complete degradation of the mechanical calculations revealed that the hydrolysis of the test
product by applying inappropriate conditions. To our compounds is a thermodynamically driven process. Interest-
knowledge, there is only little data [34,35] describing the ingly, we could show that this equilibrium is strongly depend-
stability of the aminal moiety and no study conducted a system- ent on the applied reaction conditions and that a change from an
atic investigation. The conditions for stability of aminals – espe- acidic to a neutral environment can well induce the formation of
cially in tetrahydroquinazolines – are highly important because tetrahydroquinazolines, instead of their hydrolysis. We also
of the increasing relevance of such compounds in medicinal were able to determine differences in the hydrolysis rate caused
chemistry.
by the respective substituents and found the decrease in stability
of these compounds to be a result of enthalpic effects.
In the present study, we synthesized and modified tetrahydro-
Results and Discussion
quinazolines of the general structure 1 to investigate the pH
stability of the aminal core in dependence of the steric and elec- Synthesis of test compounds
tronic properties of different groups at the N-1 and N-3 nitrogen The synthesis of tetrahydroquinazolines substituted at the
atoms, as well as at the aromatic residue in position 2 phenyl ring as well as at the 3-N nitrogen atom was achieved in
(Figure 2). For that purpose, two residues were always kept 4 steps (Scheme 3). Briefly, isatoic anhydride was methylated
constant while altering the third one. To investigate the effect of using MeI to yield compound 4, followed by formation of
decreasing electron density at the nitrogen sites onto the amides 5a–d using the corresponding free amines or their salts.
stability of the aminal core, iPr, n-Pr, Me, and Ph moieties, re- Cyclization towards dihydroquinazolinones 6a–f and 7a–c was
spectively, were chosen as substituents for both the N-1 and performed using benzaldehyde derivatives under acidic condi-
N-3 nitrogen atoms. At the same time the iPr and Ph moieties tions with moderate to excellent yields. The tetrahydroquinazo-
served as bulkier substituents for investigations into steric line target compounds 8a–f and 9a–c were finally obtained by
effects. Similar to the N-1 and N-3 nitrogen atoms, 4-t-Bu, reduction with LiAlH4.
4-Me, 4-F and 4-CF3 groups were incorporated into the aromat-
ic site at position 2 to investigate the influence of decreasing The introduction of different substitution patterns at the N-1
electron density at this position. These substituents were chosen nitrogen atom was achieved using two pathways as shown in
as their influence onto the aromatic residue is mainly deter- Scheme 4. Because the direct alkylation of isatoic anhydride in
mined by inductive effects and therefore no pronounced the first reaction step with different alkyl halides failed
mesomeric effect has to be considered. In general, to estimate (Scheme 3), amide 10 was synthesized using isatoic anhydride
the electron-donating effect of all substituents (and therefore the and methylammonium hydrochloride (Scheme 4a) followed by
change in electron density at all sides), we used the data re- cyclization under acidic conditions with benzaldehyde to yield
ported by Craig [36] for the aromatic site and the data reported dihydroquinazolinone 11. Unfortunately, the introduction of
by Topliss [37] for the side chains, respectively. A disubsti- substituents at the N-1 in 11 with alkyl halides was only suc-
tuted 2,6-dichloro compound was also synthesized to study the cessful with n-PrBr in the presence of the strong non-nucleo-
influence of steric interactions at this site of the structure, as philic base t-BuOK to give compound 12a. The target com-
2,6-disubstitution of the phenyl ring prevents coplanar orienta- pound 13a could then be obtained by reduction of 12 with
tion.
LiAlH4.
Figure 2: Hydrolysis of the aminal core of tetrahydroquinazolines 1 into the corresponding diamines 2 and aldehydes 3.
2283

Beilstein J. Org. Chem. 2016, 12, 2280–2292.
Scheme 3: Reagents and conditions: (i) MeI, DIPEA, DMAc, 40 °C, 24 h; (ii) R1-NH2 or MeNH3Cl and Et3N, DMF, 40–120 °C, 3–6 h; (iii) AcOH,
70 °C, 1–4 h; (iv) LiAlH4, THF, 70 °C, 1–3 h.
As the substitution reactions using 11 failed with iPrI, a second which after reduction with LiAlH4 afforded the desired tetra-
synthetic pathway was pursued to alter the substituents at hydroquinazolines 13b,c (Scheme 4).
the N-1 nitrogen atom of the tetrahydroquinazoline core
(Scheme 4b). Thus, anthranilic acid was alkylated by reductive Stability experiments
amination with acetone and NaBH4 in two steps to yield the iso- All compounds were exposed to phosphate buffered aqueous
propyl-substituted derivative 14. The derivative 14 and the systems with defined pH values between pH 2 and pH 12 for
commercially available N-phenylanthranilic acid were con- 1 h to investigate the hydrolysis of the aminal core. After the
verted to amides 15a,b under standard conditions. Final cycliza- treatment, reversed-phase HPLC analyses were performed to
tion with benzaldehyde yielded dihydroquinazolinones 12b,c determine the ratio of intact tetrahydroquinazoline 1 and the
Scheme 4: Reagents and conditions: (a) (i) MeNH3Cl, Et3N, DMF, 70 °C, 3 h; (ii) AcOH, 70 °C, 4 h; (iii) n-PrBr, t-BuOK, DMF, 110 °C, 16 h;
(iv) LiAlH4, THF, 70 °C, 2–3 h; (b) (v) 1) acetone, MeOH, 80 °C, 5 h; 2) NaBH4, rt, 3 h; (vi) MeNH3Cl, Et3N, EDCI, HOBt, DMF, 70 °C, 8–14 h.
2284

Beilstein J. Org. Chem. 2016, 12, 2280–2292.
corresponding aldehyde 3 as cleavage product using calibration Compounds 8a–f (Figure 3a) showed no differences in the
curves (for a detailed description see Supporting Information extent of hydrolysis depending on their substitution pattern.
File 1). The results of this study are summarized in Figure 3a–c. However, a general trend for all of the test compounds was ob-
Figure 3: pH-Stability test of the aminal core toward hydrolysis in dependency of different substitution pattern at (a) the 2-phenyl residue, (b) the N-1
position and (c) the N-3 position. Experiments were performed in triplicate (mean + SD).
2285

Beilstein J. Org. Chem. 2016, 12, 2280–2292.
served: Compounds 8a–f are stable in a basic or neutral envi- the assumption that hydrolysis of tetrahydroquinazolines 1 into
ronment down to pH 6 and slowly decomposed at pH 4–5 with the corresponding diamines 2 and aldehydes 3 is a reversible
less than 20% of hydrolysis after 1 h. Interestingly, at pH 3 the reaction (cf. Figure 2), the velocity v of the reaction can be de-
aminal system is significantly hydrolysed by more than 50%, scribed in dependency of time t as follows:
and at pH 2 decomposition of the test compounds is rapidly
taking place resulting in complete degradation of the aminal
(1)
system. These results clearly show a significant pH dependency
for hydrolysis of the aminal core which is accelerated in
increasingly acidic media. Different substitution patterns at N-1 Furthermore, 1) if the conversion of 2 and 3 into 1 is suppressed
(Figure 3b) significantly alter the sensitivity of the aminal core when [H2O] >> [2] and [3], the term k−1·[2]·[3] can be
towards hydrolysis. While a methyl group (compound 8a) and neglected and 2) if water is used as solvent and therefore
an n-Pr residue (compound 13a) did not alter the hydrolysis rate remains approximately constant during the reaction, the con-
compared to all compounds of the series 8 (Figure 3a), the iPr stant k2 = k1·[H2O] can be introduced. The velocity can then be
residue (13b) as well as the Ph residue (13c) increased hydroly- described as a pseudo-first-order kinetic:
sis of the test compounds: The least stable compound 13b was
found to be completely hydrolysed already at pH 5 and also at
(2)
pH 6 more than 50% of 13b decomposed. In contrast, com-
pound 13c (Ph moiety) completely hydrolysed at pH 4 and
showed ~50% hydrolysis at pH 5. Different substituents at the Finally, rearrangement and integration gives an exponential
N-3 position (Figure 3c) showed no pronounced effects on the function from which k2 can be calculated:
pH-dependent decomposition of n-Pr (9a), Me (8a) and a Ph
(3)
(9c) group, respectively. Only the introduction of an iPr residue
(9b) at N-3 led to an increased hydrolysis rate with complete
Kinetic analysis of reference compound 8a revealed exponen-
tial first order kinetic (Figure 4) from which k2 was calculated.
degradation of 9b already at pH 4. Finally, the 1,2-dihydro-
quinazolinone compound 6a was also tested for its stability
towards hydrolysis (data not shown). This compound did not
undergo any decomposition in the tested pH range after 1 h.
Interestingly, the comparison of the kinetic parameters showed
a ~10-fold increase of the k2 value from pH 4 to pH 3 and also
The stability analyses shown in Figure 3 have to be regarded as from pH 3 to pH 2 and therefore a 10-fold increase in velocity
snapshots after 1 h of incubation time only. Therefore, the of the hydrolysis per pH unit. Only between pH 5 and pH 4 an
kinetics for this reaction were investigated in greater detail. On increase by a factor of 2 is observed. Furthermore, it is remark-
Figure 4: Kinetic analysis of hydrolysis of reference compound 8a in dependency of different pH values and calculation of kinetic parameters. Curves
and k2 values were calculated assuming a pseudo-first-order kinetic as described in Equation 3. All experiments were performed in triplicate (mean ±
SEM).
2286

Beilstein J. Org. Chem. 2016, 12, 2280–2292.
able that the half-life time of compound 8a at pH 2 is only ation of this compound due to an acidic environment might well
2.6 min and therefore complete hydrolysis of this compound is shift the equilibrium from the non-hydrolysed compound in
expected to occur within a couple of minutes.
neutral media, to its fragments in acidic media (Figure 5a). This
implies that formation of tetrahydroquinazolines from their
fragments is thermodynamically favoured in a neutral or basic
Computational studies and discussion
With regard to the results obtained from the pH-dependent environment. In fact, this is in agreement with data from litera-
stability test, two factors for hydrolysis need to be investigated. ture reporting the synthesis of tetrahydroquinazolines to take
First, the influence of an acidic or a basic environment for the place in organic solvent [25-27] as well as in non-acidic
hydrolysis has to be determined due to different hydrolysis rates aqueous media [46]. Indeed, we tried the condensation of the
at different pH values. Second, the influence of different substi- respective diamine 16 and benzaldehyde towards compound 8a
tution patterns of the test compounds should be verified in water and in acetonitrile as solvent. In both cases, complete
regarding the altered sensitivity towards hydrolysis for some condensation was observed toward compound 8a. In addition,
compounds (9b and 13b,c).
as larger amounts of benzaldehyde and diamine 16 in water are
not soluble, also a solvent mixture of water/acetonitrile 1:1 was
Initially, we determined the protonation pattern of the test com- used to dissolve all reactants and to exclude solubility effects on
pounds, as protonation of the tetrahydroquinazoline system is the condensation reaction in water (for details cf. Supporting
expected to be an essential factor of hydrolysis induction. Theo- Information File 1). Interestingly, for the least stable compound
retically, protonation might occur at the anilinic N-1, the ali- 13b (Figure 5b), the exothermicity of the reaction of the proto-
phatic N-3 or by double protonation of N-1 and N-3. Therefore nated tetrahydroquinazoline (II in Figure 5b) into its fragments
the pKa values of both nitrogens of the test compounds were (III in Figure 5b) increased to −11 kcal/mol in aqueous medi-
predicted using the empirical algorithm of MoKa [38] and vali- um. This might explain the increased hydrolysis rate of this
dated with density functional theory (DFT) energy calculations compound in acidic media compared to compound 8a’. In
of all protonation patterns for compounds 8a and 13b (cf. Sup- contrast to 8a, the hydrolysis of the neutral tetrahydroquinazo-
porting Information File 1). We used the B3LYP-D3 functional line 13b (I in Figure 5b) into its neutral fragments (IV in
[39-41] in combination with the cc-PVTZ basis sets [42,43] and Figure 5b) is also exothermic (−4.9 kcal/mol in water), which
all computations were performed with TURBOMOLE [44]. indicates the hydrolysis of this compound to take place in acidic
Solvent effects were mimicked by the Screening Model as well as in neutral environment. These observations are in
(COSMO) for water [45]. As expected, N-3 was found to be agreement with the described stability experiments of com-
more basic for all test compounds, with the exception of 9c. pound 13b, which was found to degrade also in neutral environ-
Therefore, in the following text, only the molecular forms with ment at pH 7 (cf. Figure 3b). It should also be taken into
protonation of the aliphatic nitrogen N-3 will be discussed.
account that calculations of charged compounds in solution are
less accurate than the corresponding calculations of neutral
To explore the influence of a basic or an acidic reaction envi- ones. Takano and Houk showed that the mean absolute devia-
ronment on hydrolysis, QM single point energy calculations for tion for single point energy calculations of charged molecules in
the hydrolysis of compound 8a and the least stable compound a continuum solvent model is 3–5 kcal/mol, while 1–3 kcal/mol
13b were performed. The calculations included the neutral (8a were computed for neutral molecules [47]. Due to the fact that
and 13b) and the single protonated species 8a’ and 13b’ (-’- in- hydrolysis of tetrahydroquinazolines is an equilibrium reaction,
dicates protonation at N-3 throughout) of all reactants and prod- the energy differences between the different states are within
ucts (cf. Supporting Information File 1). The computations the range of 10 kcal/mol, which is close to the accuracy of the
include NH4+ as proton donor and one water molecule was method for charged molecules. Therefore, the calculation
taken into account for the stoichiometry of the reaction. Inter- results should be approached in a manner of showing a trend,
estingly, plotting the data in water (Figure 5) as well as in the rather than expecting exact values.
gas phase (cf. Supporting Information File 1) for all relevant
fragments revealed an exothermic reaction (−6.7 kcal/mol in Taking this data together, it is predicted and experimentally
water) for the hydrolysis of 8a’ (II in Figure 5a) into the proto- proven, that basic reaction conditions shift the equilibrium
nated fragments (III in Figure 5a), while hydrolysis of the towards the formation of tetrahydroquinazolines of the general
neutral form 8a (I in Figure 5a) undergoes an endothermic reac- structure 1 from the respective diamines and aldehydes, while
tion (3.1 kcal/mol in water) into the neutral fragments (IV in acidic conditions promote hydrolysis of the protonated tetra-
Figure 5a). This indicates that the fragments show greater hydroquinazolines 1’ (Figure 6). Therefore, compounds with an
stability when protonated, while interestingly, the aminal struc- increased basicity, like those with a branched alkyl chain at the
ture is more stable when it is non-protonated. Therefore, proton- nitrogen (e.g., compound 9b), are more sensitive towards
2287

Beilstein J. Org. Chem. 2016, 12, 2280–2292.
Figure 5: Differences in energy along the reaction coordinate using the functional B3LYP-D3 for the hydrolysis of (a) compound 8a and (b) com-
pound 13b in aqueous environment.
hydrolysis due to their accelerated protonation which induces pounds was performed to investigate possible differences in the
hydrolysis.
minimum energy conformation of both the neutral and the
protonated forms. The conformational search revealed that the
Considering the influence of different substitution patterns for majority of all compounds (neutral form: 8a–c,e, 9a, 13a–c;
an increased hydrolysis rate of some compounds (9b and protonated form: 8a’–e’, 9a’, 13a’–c’), shared the same
13b,c), a systematic conformational search [48] for all com- minimum energy conformer in which the phenyl ring in posi-
2288

Beilstein J. Org. Chem. 2016, 12, 2280–2292.
Figure 6: Reaction equilibrium between tetrahydroquinazoline 1, the corresponding diamine 2 and aldehyde 3 in basic or neutral media as well as for
the protonated tetrahydroquinazoline 1’ and the corresponding diamine 2’ and aldehyde 3 in acidic media.
tion 2 and the N-3 side group are in anti-axial orientation Furthermore, the 2,6-dichloro compound 8f exhibits the same
(exemplarily shown for compound 8a in Figure 7a) while the behaviour in terms of stability as the other compounds of the 8
other compounds (neutral form: 8d,f, 9b,c; protonated form: series, even though it adheres to the equatorial minimum energy
8f’, 9b’,c’) adhere to a conformer of minimal energy where the motif, both in the neutral and in the N-3 protonated form. In
residues are in equatorial position (exemplarily shown for com- conclusion, a comparison of the conformational search and the
pound 9b in Figure 7b). Both conformations of minimum experimental data of the hydrolysis experiments proves no
energy found in this study are in agreement with crystal struc- coherence between the decreased sensitivity on hydrolysis of 9b
tures reported in the literature [46,49-52] (CCDC reference and 13b,c due to a different conformation of minimal energy;
numbers for anti-axial motif: 177049 [46], 717617 [49], 778079 especially as 13b,c and 13b’,c’ are following anti-axial confor-
[50], 722943 [51], 820149 [52]; and for equatorial motif: mation, while 9b and 9b’ are following equatorial conforma-
177050, 177052 [46]).
tion.
Interestingly, the least stable compound 13b and the significant- To further investigate the increased hydrolysis rates of 9b and
ly more stable compound 8e were found to share the same anti- 13b,c, respectively, energetic differences within the aminal
axial conformer in the neutral and in the N-3 protonated form. system of these compounds were explored. The cyclic tetra-
Figure 7: Minimum energy conformers in their neutral form with (a) an axial orientation of the phenyl system exemplarily shown for compound 8a and
(b) an equatorial orientation of the phenyl substituent exemplarily shown for 9b.
2289

Beilstein J. Org. Chem. 2016, 12, 2280–2292.
hydroquinazoline system is higher in energy by its conforma- In addition, as a proof of concept, also the energy differences
tion in comparison to the non-cyclic diamines and bulky substit- between 8a and the modified compound 8b_mod as well as the
uents, e.g., Ph or iPr residues, might additionally increase the sterically more demanding compound 8f_mod were calculated.
energy in this system (e.g., through strain), thus contributing to As expected, in the case of 8b_mod the difference to 8a is
its enhanced sensitivity towards hydrolysis. Therefore, a com- negligible (ΔU < 1 kcal/mol). For the sterically more
parison of the energy of the aminal core of the reference com- demanding compound 8f_mod the energy difference of the
pound 8a with compounds 9b and 13b,c was performed using protonated form is 1.61 kcal/mol, which is still a moderate
Molecular Mechanics (MM). For this purpose, the conformer of value and places it well below compound 9b. Obviously, an
minimal energy for compounds 9b and 13b,c was kept frozen, exact correlation of this simple measure with the hydrolytic
while the N-side groups were changed into a methyl residue. instability is not to be expected, given the limitations of the mo-
The resulting structures were therefore identical with com- lecular mechanical approach and the fact that the hydrolysis rate
pound 8a, thus differences in potential energy are attributed to of the investigated aminal system is not only a function of in-
the conformational energy of the tetrahydroquinazoline core. creased conformational energy but also dependent on other
This procedure was followed for all the aforementioned com- thermodynamic factors, as mentioned above. Nevertheless, a
pounds in their neutral and the N-3 protonated form. The rela- clear trend can be recognized, indicating that the conformation-
tive potential energy of the modified compounds (13b_mod, al energy of the ring system is at least an important if not the
13c_mod and 9b_mod) was higher in all cases compared to major contribution to the hydrolytic instability of the investigat-
compound 8a (Table 1) suggesting a correlation of increased ed systems.
conformational energy of the ring system and increased sensi-
tivity toward hydrolysis. This increase in energy might well be Conclusion
attributed to different geometrical effects including ring strain, The aminal core is a common structural element in various me-
strain of the angles or steric repulsion of the side groups. It dicinally relevant compounds and naturally occurring alkaloids.
might be possible that an increased ring strain at the nitrogens However, this system suffers from hydrolysis and therefore
of the tetrahydroquinazoline system could increase the basicity might decompose due to inappropriately applied reaction or
of these compounds and therefore accelerate the induction of work-up conditions in synthesis, or when being exposed to an
hydrolysis, although previous studies showed that the relation- acidic environment, e.g., by gastric acid when administered
ship between the magnitude of ring strain and the resulting orally.
nitrogen basicity is not straightforward [53,54]. Probably more
interesting, the hydrolysis of cyclic geminal ethers was recently To systematically investigate the pH-dependent hydrolysis of
reported to be drastically accelerated by introduction of steri- the aminal system, tetrahydroquinazolines 8a–f, 9a–c and
cally demanding side groups through reduction of the activa- 13a–c were synthesised and exposed to buffered aqueous media
tion barrier [55]. These results are consistent with the herein re- with defined pH values. A general trend was observed for all
ported data supporting a preferred elimination of the aldehyde compounds with an accelerated hydrolysis rate as a function of
fragment from the respective tetrahydroquinazoline in com- decreasing pH value. Additional density functional calculations
pounds with increased ring energy. Nevertheless, compound revealed that protonation of the N-3 nitrogen can induce hydro-
13c_mod was computed to be the least stable compound al- lysis into the corresponding fragments. Therefore, compounds
though compound 13b was experimentally proven to be less with a higher pKa value at the N-3 nitrogen might increase the
stable. Therefore, we assume that additional enthalpic or entro- hydrolysis sensitivity of these compounds due to their faster
pic effects might also be involved in altering the stability of the protonation, like for compound 9b. Computational studies as
ring system and an increase in conformational energy might be well as experimental data revealed that the formation of tetra-
only one part of the puzzle.
hydroquinazolines is favoured when exposing their fragments
Table 1: Calculated differences in potential energy (U) of the modified compounds compared to compound 8a in water with MM. Differences were
calculated by ΔU = U(Cpd_mod or Cpd_mod’) – U(8a or 8a’).
ΔU [kcal/mol]
8b_mod
0.74
8f_mod
1.47
9b_mod
13b_mod
3.05
13c_mod
5.27
neutral
2.09
protonated
0.73
1.61
2.46
3.50
5.22
2290

Beilstein J. Org. Chem. 2016, 12, 2280–2292.
7. Baldeyrou, B.; Tardy, C.; Bailly, C.; Colson, P.; Houssier, C.;
Charmantray, F.; Demeunynck, M. Eur. J. Med. Chem. 2002, 37,
315–322. doi:10.1016/S0223-5234(02)01356-9
(diamine and aldehyde) to an environment where no proton-
ation occurs (mostly due to a basic solvent) and therefore the
reaction equilibrium can be shifted to one or the other side by
protonating or deprotonating the relevant reactants. Further-
more, different substitution patterns in position 2 (series 8) of
the aromatic system did not affect the stability because of
changes in electron density (8a–e) or bulkiness (8f) of the sub-
stituents.
8. Johnson, R. A.; Gorman, R. R.; Wnuk, R. J.; Crittenden, N. J.;
Aiken, J. W. J. Med. Chem. 1993, 36, 3202–3206.
doi:10.1021/jm00073a023
9. Goswami, S.; Ghosh, K.; Dasgupta, S. J. Org. Chem. 2000, 65,
1907–1914. doi:10.1021/jo9909204
10.Wu, H.; Chen, X.; Wan, Y.; Ye, L.; Xin, H.; Xu, H.; Yue, C.; Pang, L.;
Ma, R.; Shi, D. Tetrahedron Lett. 2009, 50, 1062–1065.
doi:10.1016/j.tetlet.2008.12.067
Interestingly, hydrolysis experiments revealed an accelerated
decomposition for compounds 9b and 13b,c. To investigate
these findings, minimum conformational energy calculations
were conducted. We found that all compounds adhere to two
groups of conformers in accordance with crystal structures re-
ported in literature. Changing the minimum energy conformers
of 9b and 13b,c into the reference structure 8a revealed an
increase in ring energy which might accelerate their hydrolysis.
However, additional entropic and enthalpic effects may also in-
fluence the stability of such compounds that should be investi-
gated in further studies.
11.Harmata, M.; Kahraman, M. Tetrahedron: Asymmetry 2000, 11,
2875–2879. doi:10.1016/S0957-4166(00)00255-X
12.Pardo, C.; Sesmilo, E.; Gutiérrez-Puebla, E.; Monge, A.; Elguero, J.;
Fruchier, A. J. Org. Chem. 2001, 66, 1607–1611.
doi:10.1021/jo0010882
13.Valik, M.; Dolensky, B.; Petřičková, H.; Vašek, P.; Král, V.
Tetrahedron Lett. 2003, 44, 2083–2086.
doi:10.1016/S0040-4039(03)00177-1
14.Paul, A.; Maji, B.; Misra, S. K.; Jain, A. K.; Muniyappa, K.;
Bhattacharya, S. J. Med. Chem. 2012, 55, 7460–7471.
doi:10.1021/jm300442r
15.Huang, G.; Kling, B.; Darras, F. H.; Heilmann, J.; Decker, M.
Eur. J. Med. Chem. 2014, 81, 15–21.
doi:10.1016/j.ejmech.2014.05.002
16.Darras, F. H.; Kling, B.; Heilmann, J.; Decker, M.
ACS Med. Chem. Lett. 2012, 3, 914–919. doi:10.1021/ml3001825
17.Sawatzky, E.; Wehle, S.; Kling, B.; Wendrich, J.; Bringmann, G.;
Sotriffer, C. A.; Heilmann, J.; Decker, M. J. Med. Chem. 2016, 59,
2067–2082. doi:10.1021/acs.jmedchem.5b01674
18.Kamal, M. A.; Qu, X.; Yu, Q.-s.; Tweedie, D.; Holloway, H. W.; Li, Y.;
Tan, Y.; Greig, N. H. J. Neural Transm. 2008, 115, 889–898.
doi:10.1007/s00702-008-0022-y
Supporting Information
Supporting Information File 1
Detailed synthetic procedures, spectral data, stability
analyses and computational investigations.
[http://www.beilstein-journals.org/bjoc/content/
supplementary/1860-5397-12-221-S1.pdf]
19.Tasso, B.; Catto, M.; Nicolotti, O.; Novelli, F.; Tonelli, M.; Giangreco, I.;
Pisani, L.; Sparatore, A.; Boido, V.; Carotti, A.; Sparatore, F.
Eur. J. Med. Chem. 2011, 46, 2170–2184.
Acknowledgements
doi:10.1016/j.ejmech.2011.02.071
M. Decker gratefully acknowledges the German Science Foun-
dation (Deutsche Forschungsgemeinschaft) for financial support
(DFG DE1546/6-1 and DE1546/6-3) and the Elite Network of
Bavaria for awarding a Ph.D. position to A. Drakopoulos within
the International Doctoral Program “Receptor Dynamics”.
20.Khan, I.; Ibrar, A.; Abbas, N.; Saeed, A. Eur. J. Med. Chem. 2014, 76,
193–244. doi:10.1016/j.ejmech.2014.02.005
21.Khan, I.; Ibrar, A.; Ahmed, W.; Saeed, A. Eur. J. Med. Chem. 2015, 90,
124–169. doi:10.1016/j.ejmech.2014.10.084
22.Richers, M. T.; Deb, I.; Platonova, A. Yu.; Zhang, C.; Seidel, D.
Synthesis 2013, 45, 1730–1748. doi:10.1055/s-0033-1338852
23.Dieckmann, A.; Richers, M. T.; Platonova, A. Yu.; Zhang, C.;
Seidel, D.; Houk, K. N. J. Org. Chem. 2013, 78, 4132–4144.
doi:10.1021/jo400483h
References
1. Kim, J. T.; Gevorgyan, V. Org. Lett. 2002, 4, 4697–4699.
24.Zhang, C.; De, C. K.; Mal, R.; Seidel, D. J. Am. Chem. Soc. 2008, 130,
416–417. doi:10.1021/ja077473r
doi:10.1021/ol027129t
2. Bosque, I.; Gonzalez-Gomez, J. C.; Loza, M. I.; Brea, J. J. Org. Chem.
2014, 79, 3982–3991. doi:10.1021/jo500446f
25.Sinkkonen, J.; Zelenin, K. N.; Potapov, A.-K. A.; Lagoda, I. V.;
Alekseyev, V. V.; Pihlaja, K. Tetrahedron 2003, 59, 1939–1950.
doi:10.1016/S0040-4020(03)00148-0
3. Scholl, M.; Trnka, T. M.; Morgan, J. P.; Grubbs, R. H. Tetrahedron Lett.
1999, 40, 2247–2250. doi:10.1016/S0040-4039(99)00217-8
4. Sharma, V.; Khan, M. S. Y. Eur. J. Med. Chem. 2001, 36, 651–658.
doi:10.1016/S0223-5234(01)01256-9
26.Fan, X.; Li, B.; Guo, S.; Wang, Y.; Zhang, X. Chem. – Asian J. 2014, 9,
739–743. doi:10.1002/asia.201301296
27.Schiedler, D. A.; Vellucci, J. K.; Beaudry, C. M. Org. Lett. 2012, 14,
6092–6095. doi:10.1021/ol3029912
5. Caterina, M. C.; Perillo, I. A.; Boiani, L.; Pezaroglo, H.; Cerecetto, H.;
González, M.; Salerno, A. Bioorg. Med. Chem. 2008, 16, 2226–2234.
doi:10.1016/j.bmc.2007.11.077
28.Richers, M. T.; Zhao, C.; Seidel, D. Beilstein J. Org. Chem. 2013, 9,
1194–1201. doi:10.3762/bjoc.9.135
6. Veale, E. B.; Frimannsson, D. O.; Lawler, M.; Gunnlaugsson, T.
Org. Lett. 2009, 11, 4040–4043. doi:10.1021/ol9013602
29.Han, B.; Yang, X.-L.; Wang, C.; Bai, Y.-W.; Pan, T.-C.; Chen, X.;
Yu, W. J. Org. Chem. 2012, 77, 1136–1142. doi:10.1021/jo2020399
2291

Beilstein J. Org. Chem. 2016, 12, 2280–2292.
30.Maheswari, C. U.; Kumar, G. S.; Venkateshwar, M.; Kumar, R. A.;
Kantam, M. L.; Reddy, K. R. Adv. Synth. Catal. 2010, 352, 341–346.
doi:10.1002/adsc.200900715
License and Terms
This is an Open Access article under the terms of the
Creative Commons Attribution License
31.Mercan, D.; Çetinkaya, E.; Sahin, E. Inorg. Chim. Acta 2013, 400,
74–81. doi:10.1016/j.ica.2013.02.005
32.Sawatzky, E.; Bukowczan, J.; Decker, M. Tetrahedron Lett. 2014, 55,
2973–2976. doi:10.1016/j.tetlet.2014.03.109
(http://creativecommons.org/licenses/by/4.0), which
permits unrestricted use, distribution, and reproduction in
any medium, provided the original work is properly cited.
33.Aga, M. A.; Kumar, B.; Rouf, A.; Shah, B. A.; Andotra, S. S.;
Taneja, S. C. Helv. Chim. Acta 2013, 96, 969–977.
doi:10.1002/hlca.201200307
34.Greenberg, A.; Molinaro, N.; Lang, M. J. Org. Chem. 1984, 49,
1127–1130. doi:10.1021/jo00180a035
The license is subject to the Beilstein Journal of Organic
Chemistry terms and conditions:
35.Taylor, P. D. Talanta 1995, 42, 243–248.
(http://www.beilstein-journals.org/bjoc)
doi:10.1016/0039-9140(94)00234-J
36.Craig, P. N. J. Med. Chem. 1971, 14, 680–684.
doi:10.1021/jm00290a004
The definitive version of this article is the electronic one
which can be found at:
37.Topliss, J. G. J. Med. Chem. 1972, 15, 1006–1011.
doi:10.1021/jm00280a002
doi:10.3762/bjoc.12.221
38.Milletti, F.; Storchi, L.; Sforna, G.; Cruciani, G. J. Chem. Inf. Model.
2007, 47, 2172–2181. doi:10.1021/ci700018y
39.Becke, A. D. J. Chem. Phys. 1993, 98, 5648. doi:10.1063/1.464913
40.Lee, C.; Yang, W.; Parr, R. G. Phys. Rev. B 1988, 37, 785–789.
doi:10.1103/PhysRevB.37.785
41.Grimme, S.; Antony, J.; Ehrlich, S.; Krieg, H. J. Chem. Phys. 2010,
132, 154104. doi:10.1063/1.3382344
42.Kendall, R. A.; Dunning, T. H., Jr.; Harrison, R. J. J. Chem. Phys. 1992,
96, 6796. doi:10.1063/1.462569
43.Schäfer, A.; Huber, C.; Ahlrichs, R. J. Chem. Phys. 1994, 100, 5829.
doi:10.1063/1.467146
44.TURBOMOLE, V6.2; Turbomole GmbH, 2010.
45.Klamt, A.; Schüürmann, G. J. Chem. Soc., Perkin Trans. 2 1993,
799–805. doi:10.1039/P29930000799
46.Correa, W. H.; Papadopoulos, S.; Radnidge, P.; Roberts, B. A.;
Scott, J. L. Green Chem. 2002, 4, 245–251. doi:10.1039/B202729C
47.Takano, Y.; Houk, K. N. J. Chem. Theory Comput. 2005, 1, 70–77.
doi:10.1021/ct049977a
48.Goodman, J. M.; Still, W. C. J. Comput. Chem. 1991, 12, 1110–1117.
doi:10.1002/jcc.540120908
49.Mori, K.; Ohshima, Y.; Ehara, K.; Akiyama, T. Chem. Lett. 2009, 38,
524–525. doi:10.1246/cl.2009.524
50.Wang, Z.-G.; Xia, Z.-L.; Wang, R.; Wang, M.-L.
Acta Crystallogr., Sect. E: Struct. Rep. Online 2010, 66, o1205–o1206.
doi:10.1107/S1600536810015023
51.Wang, Z.-G.; Wang, R.; Zhang, Y.; Zhi, F.; Yang, Y.-L.
Acta Crystallogr., Sect. E: Struct. Rep. Online 2009, 65, o550.
doi:10.1107/S1600536809005182
52.Wang, Z.-G.; Wang, R.; Zhi, F.; Wang, M.-L.
Acta Crystallogr., Sect. E: Struct. Rep. Online 2011, 67, o808–o809.
doi:10.1107/S1600536811007987
53.Ohwada, T.; Hirao, H.; Ogawa, A. J. Org. Chem. 2004, 69, 7486–7494.
doi:10.1021/jo0486589
54.Radić, N.; Despotović, I.; Vianello, R. Croat. Chem. Acta 2012, 85,
495–504. doi:10.5562/cca2121
55.Repetto, S. L.; Costello, J. F.; Butts, C. P.; Lam, J. K. W.;
Ratcliffe, N. M. Beilstein J. Org. Chem. 2016, 12, 1467–1475.
doi:10.3762/bjoc.12.143
2292