Modulation of Conformational Preferences of Heteroaromatic Ethers and Amides through Protonation and Ionization: Charge Effect.

Multiple approaches reveal the strong effects of a positive charge introduced by protonation or ionization on the conformation of o-heteroaromatic ethers and amides. The ethers and amides containing an ortho-N heteroatom are syn-preferring while those containing an ortho-O or ortho-S heteroatom are mostly anti-preferring. However, for all the monocyclic o-heteroaromatic ethers and amides, the protonated ones are all anti-preferring while the ionized ones are all syn-preferring. Interestingly, although both the protonation and ionization introduce a positive charge, they have such different effects on molecular conformation, very informative for understanding the origin of conformational preferences. Detailed analysis shows that the population of the introduced positive charge dictates the conformational preferences via electrostatic and orbital interactions. Compared to ortho-heteroatoms, meta-heteroatoms have weaker effect on conformational preference. Achieved by complete inductive method, the regularity of conformational preferences and switching provides easy ways to modulate conformers (by pH or redox), and makes this kind of ether or amide bond a conformational hinge applicable to design of functional molecules (drugs and materials) and modulation of molecular biological processes.

Introduction

Conformational preference and switching are of theoretical1, 2, 3, 4, 5, 6 and practical7, 8, 9, 10 importance. The origin of conformational preference and internally rotational barrier has long been related to the steric,2 quantum3 and electrostatic effects.5, 6, 11 And there are still controversies about which factor is more important.5 Heterocycles are usually important components of natural compounds12, 13 and functional molecules14, 15 and can influence molecular conformation and properties. Since the heterocyclic ring adjacent to an ether oxygen11 or amide6 can determine molecular conformation, they can be used as heterocyclic hinges to orient a chain in conformational space.7, 16 In drug design, this kind of heterocyclic hinge has been used as switch between agonist and antagonist ligand conformations.7 Strong conformational preferences of 2‐methoxyheteroarenes induced by electron pair repulsion make the heteroaromatic ether bond a potential conformational switch.11 However, information and knowledges on the conformation of heteroaromatic molecules are still scattering. A generalized regularity of conformational preferences and switching can be used for molecular design, especially for the functional molecules with special conformational structure.

For the water‐ and oxygen‐rich system of atmosphere, hydrosphere, geosphere and biosphere on earth, the pH‐ and redox‐related reactions are of great importance, which are taught in early stage of chemical education. The pH‐ and redox‐triggered conformational switching or stabilization may play crucial roles. For example, the pH‐triggered conformational switch of heteroaromatic lipids can control the channels on membrane.17 The ionization/oxidation‐induced conformational switching of some aromatic compounds is also found.18, 19 If the protonation or ionization can modulate the conformation of heteroaromatics, the heterocyclic hinges can operate by pH or redox. Since both the protonation and ionization introduce one positive charge into the molecule, the effect of the positive charge on stability of molecular conformers will provide information helpful for understanding the origin of conformational preferences.

To explore the effect of charge on the molecular conformation, we investigated the effect of protonation and ionization on conformational preferences of all the o‐heteroaromatic ethers and amides by calculating energy differences between syn and anti conformers. The conformational switching caused by protonation and ionization are discussed in detail.

Results

Conformational Preferences of o‐Heteroaromatic Ethers: Effects of Protonation and Ionization

Figure 1 depicts the syn conformers of all kinds of the monocyclic o‐methoxyheteroarenes having ortho‐heteroatom effect (compounds 1–36), which are calculated at ωB97XD/cc‐pVTZ20, 21 level with Gaussian 09.22 Note that the syn or anti conformers are defined according to the orientation of methyl group with respect to the in‐plane lone pair electrons on the ortho‐heteroatom. For those containing more than one ortho‐heteroatoms, according to the conventional rules, the priority has the largest atom (higher atomic number) in determining the substituent position. However, since the effects of ortho‐heteroatom on conformational preferences have the order N>O or S, the ortho‐position of each compound is re‐defined accordingly. Table 1 lists the calculated energy differences between syn and anti conformers of o‐methoxyheteroarenes having ortho‐heteroatom effect. It can be seen that, for all the ethers having an ortho‐N heteroatom 1–28, the syn conformers have lower potential energy and are more stable than the corresponding anti ones, suggesting the syn preferences.11, 23 This means that the ortho‐N heteroatom has a strong syn‐preferring effect, which is also verified by performing resonance enhanced two‐photon ionization experiments on 2‐[6]‐1N 1 (Figure 2) and the cationic spectroscopy.18 On the other hand, for ethers having an ortho‐O or ortho‐S but no ortho‐N (ethers 29–36), most have anti preferences (ethers 29–34, 36), except for the 2‐[5]1O4N5N 35 (with very weak syn conformational preference). This means that the ortho‐O or ortho‐S heteroatom has a relatively weak anti‐preferring effect, which may be affected by other factors. The o‐heteroatoms play an imperative role in determining the conformational preferences, while the heteroatoms on other positions have slight effects. So, based on the conformational preferences, all the monocyclic o‐heteroaromatic ethers can be divided into two groups. The ethers 1–28 having one ortho‐N heteroatom belong to group I, which prefer syn conformers. The ethers 29–36 having only one heteroatom (O or S) at the double ortho‐positions belong to group II, which mainly prefer anti conformers. Chein and Corey11 observed the syn preferring ortho‐N and the anti preferring ortho‐O or ortho‐S heteroatom effects in studying ethers 1, 3, 12, 13, 29 and 32, but, by checking all the o‐methoxyheteroarenes, we found an exception, the syn preference of 2‐[5]1O4N5N 35. For the o‐methoxyheteroarenes, the ortho‐heteroatom effects on conformational preferences derived by complete induction method here can be considered as a kind of propensity regularity, which is convenient for application.

Figure 1

The syn conformers of o‐methoxyheteroarenes. The standard nomenclature is given at first. For convenience, the corresponding abbreviation is employed to represent the nomenclature and structure of each compound, where the first digit marks the position of OCH3 on the ring and the second digit in square brackets, [5] or [6], marks the five‐ or six‐membered ring, followed by digits and symbols corresponding to the position of heteroatoms on the ring. Finally, the re‐defined abbreviations setting the position of OCH3 as ortho are displayed in italic type.

Table 1

The conformational preferences and rotational barriers in different states (in kcal/mol).a

Numbers	Compoundsb	ΔEc	ΔE(H+)d	ΔE(−e)e	EB f	EB(−e)g	ΔEB h
1	2‐[6]1N	4.08	−3.23	6.22	7.69	19.77	12.08
2	2‐[5]1N −	1.68	−3.70 i	3.91	2.88	10.98	8.10
3	2‐[6]1N6N	4.96	−1.65	2.84	8.04	8.35	0.31
4	2‐[6]1N5N	3.96	−3.34	3.11	8.90	11.19	2.29
5	2‐[6]1N4N	4.06	−2.77	6.21	7.53	15.91	8.38
6	2‐[5]1N5N −	2.24	−2.96	3.13	3.31	11.70	8.39
7	2‐[5]1N4N −	1.06	−4.18 i	4.43	2.68	11.24	8.56
8	2‐[5]1N5O	3.61	−1.68	1.14	5.88	15.44	9.56
9	2‐[5]1N5S	4.23	−1.81	4.61	6.79	17.86	11.07
10	2‐[5]1N4O	0.53	−4.47	5.09	3.88	16.53	12.65
11	2‐[5]1N4S	0.84	−4.59	4.80	4.46	16.80	12.34
12	2‐[5]1N3O	3.22	−3.03	2.42	5.54	13.61	8.07
13	2‐[5]1N3S	3.34	−2.78	4.29	6.09	15.14	9.05
14	2‐[6]1N5N6N	4.34	−1.70	3.53	9.05	11.21	2.16
15	2‐[6]1N4N5N	4.36	−2.47	4.40	9.28	12.55	3.27
16	2‐[6]1N4N6N	4.95	−1.02	3.44	7.95	8.52	0.57
17	2‐[5]1N4N5N −	1.65	−3.01	4.00	3.01	12.27	9.26
18	2‐[5]1N3O5N	3.77	−2.28	2.03	5.60	14.21	8.61
19	2‐[5]1N3S5N	3.97	−1.92	3.43	6.12	13.28	7.16
20	2‐[5]1N4N5O	3.68	−0.86	2.99	5.76	16.56	10.80
21	2‐[5]1N4N5S	3.98	−1.26	5.17	6.61	17.56	10.95
22	2‐[5]1N4O5N	0.84	−3.08	* j	3.25	* j	* j
23	2‐[5]1N4S5N	1.37	−3.18	3.46	4.32	14.37	10.05
24	2‐[5]1N3O4N	2.27	−4.31	1.43	6.54	14.71	8.17
25	2‐[5]1N3S4N	2.89	−3.05	4.93	7.29	16.40	9.11
26	2‐[6]1N4N5N6N	4.83	−0.75	* j	9.54	* j	* j
27	2‐[5]1N3O4N5N	2.45	−4.04	0.08	6.99	14.54	7.55
28	2‐[5]1N3S4N5N	3.25	−2.31	3.62	7.44	8.27	0.83
29	2‐[5]1O	−1.14	−6.82 i	1.58	1.26	12.25	10.99
30	2‐[5]1S	−0.92	−4.45 i	0.62	1.21	12.51	11.30
31	2‐[5]1O5N	−0.16	−5.49 i	3.47	3.40	15.35	11.95
32	2‐[5]1S5N	−0.47	−3.55	0.75	3.15	14.75	11.60
33	2‐[5]1O4N	−1.37	−6.85 i	2.03	0.91	12.99	12.08
34	2‐[5]1S4N	−1.13	−4.3 i	1.64	1.01	13.42	12.41
35	2‐[5]1O4N5N	0.15	* j	* j	4.20	* j	* j
36	2‐[5]1S4N5N	−0.36	* j	2.40	3.31	* j	* j

a Calculated at ωB97XD/cc‐pVTZ level. b The protonated and ionized ethers are identified by alphanumeric numbering, which are not displayed in this table. For example, the protonated and ionized 2‐[6]1N are defined as 1 a and 1 b, respectively. c Energy differences for the ethers before ionization, ΔE=E−E, the values marked in bold and italic represent syn and anti preferences, respectively. d Mono‐protonation at the ortho‐N heteroatom of the ethers in group I, while the ortho‐O or S atom is protonated for the ethers in group II. e Energy differences for the ionized ethers, ΔE(−e)=E(−e)−E(−e). f The rotational barriers of OCH3 group for ethers before ionization. g The rotational barriers of OCH3 group for ethers after ionization. h ΔEB=EB(−e)−EB. i The values estimated by single point energy. j Unstable. Data are unavailable due to the failed convergence during self‐consistent field (SCF) or optimization steps.

Figure 2

The one‐color resonance enhanced two‐photon ionization (1C−R2PI) spectrum of 2‐[6]‐1N 1. The S1←S0 electronic transition energy (E1) of syn conformer is set as 0 cm−1. The E1 of anti conformer is predicted by CIS/cc‐pVTZ calculations (the electronic energy corrected by EOM‐CCSD/cc‐pVTZ) at about 107 cm−1 lower than that of the syn conformer, which is not observed in the 1C‐R2PI spectrum, indicating too low population to be detected. The observed bands are assigned based on the cationic spectroscopy.18

For ethers having both ortho‐N (preferring syn conformer) and ortho‐O or S heteroatoms (preferring anti conformer), due to the pull‐push mechanism, there might be an additive effect on conformational preferences. This kind of cooperating effect is confirmed by comparing the data for ethers containing both ortho‐N and ortho‐O or S heteroatoms with that of ethers containing only one ortho‐heteroatom (N, O or S). For example, the conformational preference of 2‐[5]1N3O 12 is determined to be 3.22 kcal/mol, remarkably higher than the 1.68 kcal/mol for 2‐[5]1N− 2 and 0.53 kcal/mol for 2‐[5]1N4O 10, showing the existence of additive effect, although the effects of charge and meta‐heteroatom can't be excluded. It can be seen that, not only the 2‐[5]1N4O 10, but also 2‐[5]1N5N 6 and 2‐[5]1N4S 11 have weaker syn preferences compared to the molecules 1, 3, 4, 5 containing six‐membered ring. Probably the meta‐heteroatom effects play an important role. For ethers having two ortho‐N heteroatoms 37–49 (Figure S1), effects of meta‐heteroatoms on conformational preferences should be considered, which seems irregular (Table S1).

Conformational modulation is usually based on conformational switching. It is reported24, 25 that the conformational preferences can be affected by protonation. Figure 3 (a) displays the potential curves of 2‐[6]1N 1 and its conjugate acid, 2‐[6]1NH 1 a. For the neutral 2‐[6]1N 1, the energy of anti conformer is 4.08 kcal/mol higher than that of the syn one, indicating a syn preference. While for 2‐[6]1NH 1 a, the energy difference between the two conformers is determined to be −3.23 kcal/mol and a strong anti preference is observed. Even for the ether having only one stable conformer (ethers 50–53, Figure S2), the anti preferences can also be found after ortho‐N protonation (Table S2). The potential curves for other ethers 2–28 in group I are given in the Figures S3 and S4. It is demonstrated that, although the neutral or anionic ortho‐N remarkably prefer syn conformers, the protonated ortho‐N prefers the anti ones. This means that, for the ethers containing ortho‐N heteroatom, the conformational preferences can be modulated through protonation or deprotonation on the ortho‐N heteroatom. For the ethers of group II, as shown in Table 1, the anti preferences of ethers only having ortho‐O or S heteroatom can be observed before and after the protonation, and the syn conformers of the protonated species are nonplanar26 and relatively unstable. So, it can be concluded that, for all the monocyclic o‐heteroaromatic ethers, the protonation on o‐heteroatom gives anti preferences. Additionally, the anti preferences of ethers in group II are enhanced by protonation (data for compounds 2‐[5]1OH4N5N 35 a and 2‐[5]1SH4N5N 36 a are unavailable due to the failed convergence in optimization steps), and most of protonated ethers in group I & II have strong anti preferences (>2.0 kcal/mol), which might be useful for molecular design.

Figure 3

The switching of conformational preference of (a) 2‐[6]1N 1 upon protonation, (b) 2‐[5]1O 29 upon ionization, and (c) 2‐[5]1S 30 upon ionization.

For the ethers containing two or more heteroatoms, the ortho‐heteroatom is not necessarily the one with strongest basicity, and practically the protonation of all heteroatoms could be considered. Additional calculations show that, for the ethers containing two heteroatoms, the protonation of all the heteroatoms also give the anti preferences, suggesting that the protonation on the non‐ortho heteroatoms has weaker effect than that of o‐heteroatoms. However, for the ethers containing more than two heteroatoms, application of this conformational modulation through multi‐protonation may be difficult. As shown in Table S3, among the fifteen compounds with only one ortho‐N heteroatom and heteroatoms at other position, ten of them have a conformational switching upon complete protonation, except for 2‐[6]1N4N5N 15, 2‐[6]1N4N6N 16, 2‐[5]1N4N5O 20, 2‐[5]1N4N5S 21 and 2‐[6]1N4N5N6N 26. A possible remedy for the application of mono‐protonation triggered conformational switching of those five ethers might be related to the coordination reaction,27 which can remove the preferred conformer from the chemical equilibrium.

As demonstrated in our previous study,28 ionization can trigger the conformational switching of some aromatic compounds. The ionization/oxidation effect on the conformational preferences of o‐methoxyheteroarenes is investigated, as tabulated in Table 1. Firstly, for the compounds of group I, syn preferences exist before and after the ionization (data for the ionized state of compounds 2‐[5]1N4O5N 22 and 2‐[6]1N4N5N6N 26 are unavailable due to the failed convergence of optimization steps). Ionization can either enhance or decrease the conformational preferences, but, for application in molecular design, we can find the strong syn preference (>2 kcal/mol) at least in one of two states (before and after ionization) for most ethers of group I, except for the 2‐[5]1N4O5N 22, which has a weak syn preference (0.84 kcal/mol) and is unstable in ionized state. For the ethers with only one ortho‐O or S heteroatom 29–36 (goup II), the anti preferences can be changed to syn preferences upon ionization, suggesting an ionization‐induced conformational switching (as shown in Figure S5). It is to say, although these ethers have anti preferences before ionization, they, similar to ethers of group I, have syn preferences after ionization. For example, in the ionized state, the syn 2‐[5]1O 29 b and syn 2‐[5]1S 30 b are more stable than the anti conformers (Figures 3(b) and (c)). Marked with conformational distributions (data for compounds 2‐[5]1O4N5N 35 b and 2‐[5]1S4N5N 36 b are not available), the ionization reactions are exhibited in Figure S6. So, it can be concluded that, for all the studied ethers, the ionized or oxidized ones have syn preferences. Additionally, enhancement of conformational preferences upon ionization (Figure S7) is observed for those ethers having weak syn preferences (<1 kcal/mol), except for 2‐[5]1N4O5N 22 that is unstable in the ionized state.

It must be noted that, for the effect of protonation and ionization/oxidation as shown in Table 1, the exceptions which are unstable in the protonated or ionized states only occur for some ethers containing more than two heteroatoms. These ethers are rare compounds and cannot be seen so often in literature, and they are usually not available for use. Having typical effect of protonation and ionization/oxidation, the strong conformational preferences of ethers having one or two heteroatoms are easy applicable and recommended.

The syn‐anti interconversion barriers are summarized in the last three columns of Table 1, which are remarkably increased upon ionization, due to the increased p‐π* conjugation between OCH3 and ring in the ionized state. For the ionization‐triggered conformational switching, excess energy is needed to overcome the higher barrier. Otherwise the conformation will keep as before the ionization. Although it might be a difficulty for the application of the conformational modulation by ionization, it seems that we can take advantage of the higher barriers for the preparation of non‐classical conformers. Detailed discussions are shown in the Supporting Information as Figure S8.

Conformational Preferences of o‐Heteroaromatic Amides: Effect of Protonation and Ionization

For other heteroaromatic compounds like amides, some of which have recently been explored as drugs,6, 7 the effect of protonation and oxidation on their conformational preferences are similar to that of heteroaromatic ethers. For the amides containing one or two heteroatoms (Figure 4), all the protonated ones are anti‐preferring, while all the ionized ones are syn‐preferring, as shown in Table 2. Interestingly, although both protonation and ionization introduce a positive charge, they have such different effects on conformational preferences. Note that, for the amides, the syn‐preference of ortho‐N and anti‐ preference of ortho‐S are similar to those of ethers, however, unlike the anti‐preference of ethers containing ortho‐O, the syn‐preference is found for amides containing ortho‐O. Additionally, similar to the ethers, there are also several anomalies among the data for the amides in Table 2, such as the small values (<1 kcal/mol) for energy differences, and the unavailable data due to the failed convergence during the self‐consistent field (SCF) or geometrical optimization processes. To confirm the anomalies in Tables 1 and 2 obtained at ωB97XD/cc‐pVTZ level, two more computational methods are employed and the results are tabulated in Table 3. Firstly, the calculations are performed at ωB97XD/6‐311++g(d,p) level to investigate whether the anomalies are induced by the basis set. As seen from Table 3, it is found that most of the values are similar to the results calculated with cc‐pVTZ basis set. For compounds 27 b and 31, although the sign of these values are opposite to those results calculated at ωB97XD/cc‐pVTZ level, all these values are very small and the conformational populations N : N approximate 1 : 1. The conformational energy differences are also calculated by using mp2/6‐311++g(d,p),29 which reveals that the values obtained from mp2 calculations are very similar to those from DFT calculations.

Figure 4

The syn conformers of o‐heteroaromatic amides containing one or two heteroatoms.

Table 2

The conformational preferences of heteroaromatic amides in different states (kcal/mol).a

Groups	Numbersb	Compoundsc	ΔEd	ΔE(H+)e	ΔE(−e)f
Group I	1′	2′‐[6]1N	8.84	−7.39	3.77 g
(ortho‐N)	2′	2′‐[5]1N −	7.44	−5.29	8.05
	3′	2′‐[6]1N6N	9.90	−5.03	4.30
	4′	2′‐[6]1N5N	8.51	−7.18	6.73 g,h
	5′	2′‐[6]1N4N	7.73	−8.29	9.40 g,h
	6′	2′‐[5]1N5N −	7.84	−4.54	8.30
	7′	2′‐[5]1N4N −	6.78	−5.68	7.88
	8′	2′‐[5]1N5O	6.59	−4.38	0.82
	9′	2′‐[5]1N5S	7.37	−5.40	*i
	10′	2′‐[5]1N4O	6.94	−6.84	11.78 g,h
	11′	2′‐[5]1N4S	7.43	−7.02	9.83
	12′	2′‐[5]1N3O	3.50	−7.60	4.31 g,h
	13′	2′‐[5]1N3S	7.77	−5.28	*i
Group II	29′	2′‐[5]1O	3.44	*i	4.55 g,h
(ortho‐O or S)	30′	2′‐[5]1S	−0.07	−6.07	0.72
	31′	2′‐[5]1O5N	4.36	*i	10.85 g,h
	32′	2′‐[5]1S5N	−0.14	−5.84	0.45
	33′	2′‐[5]1O4N	2.79	*i	*i
	34′	2′‐[5]1S4N	−0.48	−5.35	5.46 g,h

a Calculated at ωB97XD/cc‐pVTZ level. b The serial number of amides are assigned according to the same way with those of ethers in Table 1. Each number followed by the symbol “ ′  ”. c The symbol “ ′ ” following the substituent position number represents an amide group instead of the ether group. d Energy differences for the amides, ΔE=E−E, the values marked in bold and italic represent syn and anti preferences, respectively. e Mono‐protonation at the ortho‐N heteroatom of the compounds in group I, but the ortho‐O or S atom of the compounds in group II. f Energy differences for the ionized amides, ΔE(−e)=E(−e)−E(−e). g Partial optimization results are obtained by freezing the rotational coordinate of the acylamino at 0°. h The values are estimated by single point energy. i Unstable. Data are unavailable due to the failed convergence during self‐consistent field (SCF) or optimization steps.

Table 3

The anomalies in conformational preferences of ethers and amides calculated at different levels.a

Number	Compounds	ΔE			ΔE(+H+)			ΔE(−e−)
		Method 1b	Method 2c	Method 3d	Method 1	Method 2	Method 3	Method 1	Method 2	Method 3
	Ethers
10	2‐[5]1N4O	0.53	0.80	0.66
11	2‐[5]1N4S	0.84	1.03	1.28
20	2‐[5]1N4N5O				−0.86	−0.93	−1.10
22	2‐[5]1N4O5N	0.84	1.19	1.58				Opt failure	Opt failure	Opt failure
26	2‐[6]1N4N5N6N				−0.75	−0.83	−0.91	Opt failure	Opt failure	SCF failureh
27	2‐[5]1N3O4N5N							0.08	−0.01	Opt failure
31	2‐[5]1O5N	−0.16	0.16	Ime
35	5‐[5]1O4N5N	0.15	0.53	0.21	Opt failuref	Opt failure	Opt failure	Opt failure	Opt failure	Opt failure
36	5‐[5]1S4N5N	−0.36	−0.59	Im	Opt failure	Opt failure	Opt failure	2.40	3.04	4.36
	Amides
9′	2′‐[5]1N5S							syn i	syn	SCF failure
13′	2′‐[5]1N3S							syn	syn	syn
29′	2′‐[5]1O	3.44	3.74	2.87	NH2−COH+ g	NH2−COH+	NH2−COH+
30′	2′‐[5]1S	−0.07	−0.06	−0.74
31′	2′‐[5]1O5N	4.36	4.74	3.82	NH2−COH+	NH2−COH+	NH2−COH+
32′	2′‐[5]1S5N	−0.14	−0.19	−0.49
33′	2′‐[5]1O4N	2.79	3.01	2.25	NH2−COH+	NH2−COH+	NH2−COH+	syn	syn	SCF failure
34′	2′‐[5]1S4N	−0.48	−0.49	−1.00

a Anomalies are the small values and failures shown in Tables 1 and 2. b ωB97XD/cc‐pVTZ method. c ωB97XD/6‐311++g(d,p) method. d mp2/6‐311++g(d,p) method. e One imaginary frequency is found, which involves out‐of‐plane motion of OCH3 group. f The failed convergence in the optimization process. g The protonation occurred on the O atom of the acylamino (CONH2) instead of the o‐heteroatom on the ring. h Failed SCF convergence. i Only syn conformer is obtained even the input structure is anti.

As listed in Table 3, the similar results calculated with different methods provide a substantiation for the anomalies observed for the conformational preferences. Particularly, in the neutral ground state, conformational preferences are moderate for the ethers 10, 11, 22, 31, 35 and 36, similar to those for the amides 30′, 32′ and 34′. For the protonated or ionized molecules, mainly due to the failure of convergence, some data are still unavailable, indicating that these molecules might become unstable upon protonation on o‐heteroatom or ionization.

In summary, theoretical approaches show that, for o‐heteroaromatic ethers or amides, the ortho‐heteroatom dictates conformational preference, with ortho‐N preferring syn but ortho‐S preferring anti, however, the protonation or ionization gives only anti or syn preference, respectively.

Potential Applicatons of the Conformation Modulation by Protonation and Ionization

This simple finding enables easy prediction of conformational preferences under acidification or oxidizing conditions. The pH‐ and redox‐triggered conformational switch can alter the biological function of drugs by dramatically influencing complementarity between drug and receptor. Understanding key factors determining conformation will help to select heteroaromatic templates for elaboration in drug discovery and performance. For the recently developed C3a agonist/antagonist7 (compounds 54 and 55) from different conformers or several kinds of inhibitors30, 31, 32 (EGFR 56, Hsp70 58, factor Xa 59 and MAP kinase inhibitor 60) containing structures similar to o‐heteroaromatic ethers or amides, the pH or redox effect should be taken into account in their design or function (Figure 5). Additionally, the conformational fixation of amide has been used as a new synthetic method, in the ring‐closing metathesis to yield medium‐sized lactams.33

Figure 5

The potential effects of protonation and ionization on the conformations of drugs. (a) the C3a agonist 54 and the protonation effect on conformational preference of 1,3‐thiazole‐4‐carboxamide 11′; (b) the C3a antagonist 55 and the ionization induced conformational switching of 2‐thiophenylcarboxamide 30′; (c) the EGFR inhibitor 56 and the pH‐triggered conformational switching of ether 57 (substructure of compound 56), (d) Hsp70, (e) factor Xa, (f) MAP kinase inhibitor.

The protonation induced conformational switching can also be applied in the experiments of molecular recognition.27 The combination of 2‐[6]1N 1 and adenine 61 was found to occur at low pH, as the electrostatic surfaces of 2‐[6]1N 1/2‐[6]1NH 1 a conformers and adenine shown in Figure 6, the preferred anti conformer of 2‐[6]1NH 1 a is more suitable for combination with adenine 61 through two hydrogen bonds to simulate the nucleic acid base pair (A−T) than the preferred syn conformer of 2‐[6]1N 1.

Figure 6

Electrostatic surfaces of 2‐[6]1N 1/2‐[6]1NH+ 1 a conformers and adenine 61.

Discussion

About the Origin of Conformational Preferences

The anti‐preferring protonated states and syn‐preferring ionized states may provide critical information for understanding the origin of conformational preferences. The steric,2 quantum3 and electrostatic effects5, 6, 11 have long been related to the origin of conformational preferences and bond rotational barrier, although there are still controversies about which factor is more important. For the ethane‐like molecules, such as ethane, methylamine, methanol and hydrazine, the two energy partition schemes5 revealed that the electrostatic interaction determines the barrier height (conformational stability) even though the steric effects and orbital interactions play indispensable roles. For the o‐heteroaromatic ethers and amides which are much more complicated than ethane‐like1, 2, 3, 5 small molecules, Chien and Corey considered the electronic repulsion between lone‐pairs of electrons as the determining factor of conformational preferences of o‐heteroaromatic ethers,11 while Reid et al. reported that the stereoelectronic effects exerted by heteroatom dictate the conformation of heterocyclic amides.6 Although both the two reactions involve introducing a positive charge into a molecule, the protonated and ionized molecules are not isoelectronic, and their structures are different. For a molecule containing N electron and M nucleus, the ionized molecule has N−1 electron and M nucleus, while the protonated one has N electron and M+1 nucleus. According to the equilibrium thermodynamics, the protonated form has one more neutral hydrogen atom than the corresponding ionized one (ground state), which may account for the different effects of protonation or ionization on the conformational preference of o‐heteroaromatic ethers and amides. In this section, the discussions starting from the charge population analysis will provide some unambiguous physical images on the origin of conformational preferences. A qualitative analysis of electrostatic effect is firstly conducted with some typical compounds, then followed by quantitative analyses of steric and quantum effects.

Electrostatic Interactions Between OCH3 and Ring

Figure 7 displays the partial natural population analysis (NPA) charges on atoms of OCH3 group and two ortho positions. As shown in Figure 7(a), the positive charges on the Hmethyl atoms in compounds 2‐[6]1N 1 and 2‐[6]1NH 1 a are +0.179 and +0.189, respectively. Correspondingly, the ortho‐Haryl atoms are also positively charged (+0.221 for 2‐[6]1N 1 and +0.262 for 2‐[6]1NH 1 a). On the other side of OCH3, the charge on ortho‐N heteroatom is determined to be −0.509 in 2‐[6]1N 1 and −0.485 in 2‐[6]1NH 1 a, while the charge on the introduced proton is +0.438 in 2‐[6]1NH 1 a, which shows that, upon protonation, the introduced positive charge is mainly populated on the proton connecting to the ortho‐N heteroatom. For the protonated ether 2‐[6]1NH 1 a, the electrostatic repulsion6, 11 between the introduced positive proton (+0.438) and the positive Hmethyl atoms of OCH3 group (+0.189) will significantly make the syn conformers unstable. Similar results are also found in Figures 7(b)–(c). For 2‐[5]1SH 30 a shown in Figure 7(d), the ortho‐S heteroatom becomes more positive upon protonation, indicating that the introduced positive charge is mainly populated on ortho‐S instead of the proton itself, which will also destabilize the syn conformer. Additionally, another indispensable factor supporting the anti preferences of protonated ethers is the spatial or steric effects, due to the increased crowdedness caused by the protonation on the o‐heteroatoms. However, the quantitative analysis below indicates that the steric effect seems not the determinant.

Figure 7

Partial NPA charges on OCH3 and ortho‐positions of the syn conformers of (a) 2‐[6]1N 1, (b) 2‐[5]1N− 2, (c) 2‐[5]1O 29 and (d) 2‐[5]1S 30 together with their protonated and ionized states.

The other reaction introducing a positive charge is ionization. The positive charges on the Hmethyl (+0.217 for 2‐[6]1N 1 b) and ortho‐Haryl (+0.255 for 2‐[6]1N 1 b) atoms are increased upon ionization, and the increased electrostatic repulsion between them will destabilize the anti conformers. While the negative charge on ortho‐C is remarkably decreased upon ionization, and the electrostatic attraction between it and the positive Hmethyl will be weakened, which also destabilize the anti conformers. For the ortho‐heteroatoms, as shown in Figures 7(a)‐(c), these o‐heteroatoms become slightly less negative upon ionization (−0.458 for 2‐[6]1N 1 b, −0.583 for 2‐[5]1N . 2 b and −0.418 for 2‐[5]1O 29 b), and the decreased attractive interactions with Hmethyl are still the supporting factor for the syn preferences observed in these ionized compounds. However, as shown in Figure 7(d), the positive charge on ortho‐S is increased upon ionization, and the repulsive interaction with Hmethyl is against the syn preferences of 2‐[5]1S 30 b. The quantitative analysis below shows that, due to the decreased negative charge on ether‐O atom, the delocalization effects dominate the syn preferences of 2‐[5]1S 30 b. As shown in Figure 7(a)‐(d), the negative charge on ether‐O atom is remarkably decreased upon ionization.

It seems that the population of the introduced positive charge in a molecule can to some extent determine the conformational preferences. For the protonated compounds, the introduced positive charge is mainly located on the proton or the ortho‐heteroatom, due to the increased repulsion between Hmethyl and proton, the protonated molecule will be anti‐preferring. While for the ionized compounds, the population of the introduced positive charge on the ether‐O, Hmethyl and ortho‐Haryl may play important roles in determining the syn preference. These findings rationalize the effects of protonation and ionization on conformational preferences. However, it seems that, especially for few neutral compounds containing ortho‐O heteroatom, the conformational preferences cannot be interpreted only by the electrostatic interactions. For instance, the attraction between the positive Hmethyl (+0.174) and the negative hetero‐O atom (‐0.446) does not support the anti preference of 2‐[5]1O 29. While the electronic repulsion between lone‐pairs of electrons of the ether‐O atom (‐0.484) and hetero‐O atom6 also destabilize the anti conformer of 2‐[5]1O 29. Similarly, as shown in Figure 7(d), the strong repulsion between the positive Hmethyl (+0.208) and the positive hetero‐S atom (+0.439) is contradictory to the syn preference of 2‐[5]1S+ 30 b. In fact, besides of the electrostatic factor, it has been found that the repulsion and steric effects within molecules34, 35 act in similar way but cause quite opposite effects. So, the steric and quantum factors also need to be considered.

Steric and Orbital Interactions

The labeling for each atom of 2‐[6]1N 1, 2‐[5]1N 2, 2‐[5]1O 29 and 2‐[5]1S 30 are shown in Figure S9. The pairwise steric exchange energy approximates the steric exchange interaction between two semi‐localized natural localized molecular orbitals (NLMOs). And the steric exchange energy (ΔEex) between OCH3 group and the ring are approximately described by the summation of the pairwise energy terms, which are related to the orbitals nearby OCH3 group. For the conjugation (ΔEp‐π*) and hyperconjugation (ΔEHyper) effects, energies are obtained by examining all possible interactions between “filled” Lewis‐type NBOs (donor) and “empty” non‐Lewis NBOs (acceptor) with the aid of second‐order perturbation theory. The detailed energetic components for exchange repulsion, p‐π* conjugation and hyperconjugation are tabulated in Tables S5–S10. Here, only the summarized results of energy differences are shown in Table 4.

Table 4

Effects of exchange repulsion and delocalization.a

Numbers	Compounds	ΔEex	ΔEp‐π*	ΔEHyper	ΔEdeloc	ΔEsum b	ΔE c
1	2‐[6]1N	−0.47	4.92	1.65	6.57	6.10	4.08
2	2‐[5]1N−	−0.20	−0.89	1.44	0.55	0.35	1.68
29	2‐[5]1O	0.55	−2.09	0.05	−2.04	−1.49	−1.14
30	2‐[5]1S	0.88	−3.56	0.27	−3.29	−2.41	−0.92
1 b	2‐[6]1N+	−1.02	3.68	2.58	6.26	5.24	6.22
2 b	2‐[5]1N.	−0.15	2.34	2.23	4.57	4.42	3.91
29 b	2‐[5]1O+	−0.05	0.19	0.93	1.12	1.07	1.58
30 b	2‐[5]1S+	−0.10	−0.33	0.86	0.53	0.43	0.62

a Calculated at ωB97XD/cc‐pVTZ level. b ΔEsum=ΔEex+ΔEp‐π*+ΔEHyper. c Energy differences between syn and anti conformer.

The calculated differences in steric exchange repulsion36 and delocalization between the syn and anti conformers (Table 3) show that, for the ethers and ionized ethers, the ΔEdeloc values have the same tendency with respect to those of ΔEsum and conformational energy differences, ΔE. While the steric exchange repulsion term, ΔEex, are mostly smaller than 1.0 kcal/mol and the tendencies are not in agreement with ΔE. Particularly, for compound 30 b, the conformational energy differences, ΔE (+0.62 kcal/mol), can be approximately described by ΔEdeloc (+0.53 kcal/mol) instead of the ΔEex (−0.10 kcal/mol). Similar results can be found for compounds 29 and 30 where the determining term is ΔEdeloc (−2.04 and −3.29 kcal/mol for 2‐[5]1O 29 and 2‐[5]1S 30, respectively). While the sign of ΔEex (0.55 kcal/mol for compound 29 and 0.88 kcal/mol for compound 30) is opposite to those of ΔE (−1.14 and −0.92 kcal/mol for compounds 29 and 30, respectively). These findings show that the delocalization effects3, 4 are the supporting and determining factor of conformational preferences of neutral compounds containing ortho‐O or S atom (29–36). While the steric exchange repulsion seems relatively unimportant, which can be either a supporting factor or not. So, in combination with the electrostatic analysis, it shows that both the orbital interactions and the electrostatic attractions are main stabilizing factors to the syn or anti preferences. On the other hand, the destabilizing factors are determined by the electrostatic repulsion instead of the steric exchange repulsion between OCH3 group and the ring. Certainly, there are still some questionable disadvantages in these natural bond orbital analyses leading to some deviations, such as the incomplete orbital selection, the omitted electrostatic contributions, etc. Undoubtedly, more detailed theoretical investigations are helpful for clarifying the origin of conformational preferences of these ortho‐substituted heteroarenes.

Weak Interactions – an IGM Analysis

To provide a more unambiguous physical image on the origin of conformational preferences and switching, the interfragment interactions between the substituent and heteroaromatic ring were investigated by performing IGM analysis. The IGM proposed by Lefebvre et al.,37 is a very useful methodology to study interfragment and intrafragment interactions by using pro‐molecular density. Briefly, only some essential idea of this method is outlined in the most general forms. To study the interfragment interactions, two important functions, gIGM,inter and ginter, are defined in the equations (1) and (2), respectively (Scheme 1). The ginter represents the sum of density gradient of each atom in their free states, where the i and A are index of atoms and fragments, respectively. While gIGM,inter is the IGM type of gradient, which is calculated as sum of the absolute value of . Noting that the gIGM,inter is the upper limit of ginter, and a new descriptor δginter uniquely defining the intermolecular interaction regions can be derived as the difference between gIGM,inter and ginter (equation (3)).

Scheme 1

Definitions of the gIGM,inter and ginter functions.

The term sign(λ2)ρ(r) represents the product of electron density ρ(r) and sign of the second eigenvalue of electron‐density Hessian matrix, and can be mapped to δginter isosurfaces by different colors to detect the regions as well as the characters (repulsive or attractive) of the weak interactions.

Figure 8 displays the results of the IGM analysis for some typical compounds 1, 30, 1′ and 30′ together with their corresponding protonated and ionized forms. As shown in graph (a) of Figure 8, for the neutral anti conformers, the isosurfaces are filled by green to red, revealing that the interactions in these regions are mainly repulsive. While for the syn conformers, the properties of the interactions in neutral compounds 1 and 1′ are remarkably different from those in the compounds 30 and 30′. Particularly, the interfragment interactions between substituent (OCH3 or CONH2) and pyridine are explicitly attractive. While in the case of compounds 30 and 30′, the isosurfaces in these regions are mainly colored by green, revealing a very weak attractions. These differences support the results listed in Tables 1 and 2, i. e., the compounds 1 and 1′ are syn preferred while the two thiophene derivatives (compounds 30 and 30′) are slightly anti preferred.

Figure 8

The plotting for the δginter isosurfaces of interfragment interactions between the OCH3 and the heteroaromatic ring of (a) the neutral compounds, (b) the protonated forms and (c) the ionized forms. The isosurfaces for each conformers are color‐filled by the blue‐green‐red (BGR) color scale method.

For protonation, since our attention are mainly focused on the protonation on ortho‐N heteroatom and only some estimated values (without ZPCs) are available for compound 30 a, here only the results of compounds 1 a and 1′  a are exhibited in Figure 8(b). For the anti conformers, there are still moderate repulsions between the substituent and the ring. Comparing with the unprotonated syn conformers in Figure 8(a), the blue regions in syn 2‐[6]1NH 1 a and 2′‐[6]1NH 1′ a are remarkably smaller than those in the neutral 2‐[6]1N 1 and 2′‐[6]1N 1′, indicating the weaker attractions in these protonated syn conformers. While the red areas are increased in these two protonated syn conformers, which coincides with the protonation‐induced conformational switching as shown in Tables 1 and 2.

As shown in Figure 8(c), the interactions in anti 2‐[6]1N 1 b are slightly more repulsive than that in anti 2‐[6]1N 1, while the strength of attractive interactions in syn 2‐[6]1N 1 b is similar to the neutral form, revealing the enhanced syn preference upon ionization. In the case of compound 30, the red or orange isosurface between OCH3 and ortho‐S heteroatom in syn 2‐[5]1S 30 is slightly decreased upon ionization, indicating a decreased repulsion. These changes are not remarkable in the amide 30′, which coincides with the weak conformational preferences listed in Table 2. For the 2′‐[6]1N 1′  b cation, it is found that the attractive interaction region (blue region between NH2 and ortho‐N heteroatom) shrinks remarkably. This shrinking of attraction area may be attributed to the decreasing of negative charge on the ortho‐N heteroatom upon ionization, which will weaken the N−H⋅⋅⋅N hydrogen bond. By using Espinosa's model (based on QTAIM data),38, 39 the energy of NH⋅⋅⋅N hydrogen bond in the molecule 1′ and 1′  b are determined to be −1.89 kcal/mol and −0.57 kcal/mol, respectively, revealing that upon ionization, the NH…N hydrogen bond has been weakened. This supports the above IGM analysis shown in Figure 8.

Investigators usually suggested some factor as the origin of conformational preferences.5 However, it seems to us that, due to the diversities of molecular structures, the determining factor for conformational preference might be different for different molecular systems. Considering the analysis of NPA charge, steric and orbital interactions, the repulsion between substituent (OCH3 or CONH2) and heteroaromatic ring is probably dominated by electrostatic repulsion instead of the relatively unimportant steric repulsion. While both the electrostatic and orbital interactions (via conjugation and hyperconjugation) might be responsible for the interfragment attractions. As for the protonated and ionized o‐heteroaromatic ethers and amides, the electrostatic effect can mainly account for the conformational preferences, which is in agreement with the earlier studies on several neutral ethane‐like molecules. Intrinsically, it is the difference in reactive sites between protonation (introducing a positive H+) and ionization (removing an electron) that is responsible for the origin of different effects on conformational preferences. Upon protonation or ionization, the resultant conformational preferences are mainly determined by charge distribution in the molecules, which might simultaneously affect the electrostatic and orbital interactions.

Conclusions

The positive charge introduced by protonation or ionization has strong effects on the conformaional preferences of o‐heteroaromatic ethers and amides, providing potential methods for conformation control. Calculations reveal that, although these ethers and amides have not only syn preferences for those containing an ortho‐N heteroatom but also anti preferences for those containing ortho‐O or S heteroatom, all the protonated ethers or amides are anti preferring but all the ionized ones are syn preferring. Difference in population of the positive charge introduced by protonation or ionization can mainly account for the different effect on conformation. The determining factor of conformational preference is the electrostatic interaction for the protonated compounds and most of the ionized compounds, but the orbital interactions for those ionized compounds containing ortho‐S heteroatom. Since the conclusions are based on the complete inductive method, they could be considered as rules for conformational preferences of these ethers and amides. The protonation and ionization could be easy and applicable ways to modulate the conformers of monocyclic o‐heteroaromatic ethers or amides, by pH control and oxidation. In virtue of its two conformation, the aromatic ether oxygen or amide bonds can operate to orient a chain in conformational space, and the strong conformational preferences and switching can be harnessed for the design of functional molecules and modification of molecular biological process.

Experimental Section

Computational Methods

The geometrical structures of all the studied ethers and amides are optimized at ωB97XD/cc‐pVTZ20,21 level, which is followed by frequency calculations to obtain zero point corrections (ZPCs). All the values of the conformational energy differences are derived from the expression, ΔE= E−E. Some data are marked with alphanumeric numbering, which represent the single point energy differences between syn and anti conformers, due to the failures of convergence during geometry optimizations or self‐consistent field (SCF) calculations. All the calculated energy values are confirmed by a high precision method Gaussian‐4 (G4),40 which are shown in Table S4. The potential curves are calculated by relaxed scan at ωB97XD/cc‐pVTZ level. Without ZPCs, the conversion barriers between syn and anti conformers are obtained from these potential curves. The exchange repulsion and delocalization (Here, the delocalization effect (ΔEdeloc) includes p‐π* conjugation (ΔEp‐π*) and σ‐σ* hyperconjugation (ΔEHyper) nearby the OCH3 group are calculated by using the stand‐alone NBO 5.0 program.41 The Independent Gradient Model (IGM)37 is employed to visually explore the interfragment interactions, which is performed by using Multiwfn package42 and the Visual Molecular Dynamics (VMD) program.43

R2PI Spectroscopic Experiment

The 2‐[6]1N (2‐methoxypyridine, b.p. 142–143 °C) sample with a purity of 99 % was purchased from Acros Corporation, which was used without further purification and seeded in the Ar carrier gas (ca. 2.5 atm) at room temperature (ca. 298 K). To generate a supersonic molecular beam, the sample seeded Ar gas was expanded into the vacuum chamber (source chamber, ca. 3.0×10−3 Pa) through a pulsed valve (General valve series 9, diam. 0.25 mm). Then the molecular beam was collimated by a skimmer (diam. 1 mm) before entered the ionization chamber (ca. 3.0×10−5 Pa).

The home‐built R2PI experimental system has been described in our previous publication.44 Briefly, in the 1C‐R2PI spectroscopic experiment, the molecular beam interacted perpendicularly with one UV laser beam to obtain the ionized sample molecules, which can be successively accelerated by two DC electric fields of 200 and 3500 V ⋅ cm−1. After being focused by einzel lens, these accelerated cations entered into a 1.0 meter‐long field‐free tube.

The cations were detected by a dual‐stacked micro‐channel plate (MCP) detector. Then the signals were amplified through a pre‐amplifier (Stanford Research System, SR445A) and collected by a multi‐channel scaler (MCS, Stanford Research system, SR430) and transported to the data collection terminal. By using a pulse delay generator (DG535, Stanford Research System), the sequential control of the whole system was achieved. Usually, the excitation energies of the conformers coexisting in a sample may only differ by a few hundreds of wavenumbers. To obtain a vibronic spectrum with high resolution, the step size of scanning job was set to 0.04 nm/step (ca. 2.5 cm−1/step) and each point was accumulated for 300 laser shots to counteract the fluctuation of laser power.

Pumped by a Nd:YAG laser at the frequency of 10 Hz, a tunable dye laser (Sirah Dye Laser‐CSTR) provided the UV laser radiation in the R2PI experiment. In the present work, Coumarin 153 and Pyrromethene 597 dyes were used.

Conflict of interest

The authors declare no conflict of interest.

As a service to our authors and readers, this journal provides supporting information supplied by the authors. Such materials are peer reviewed and may be re‐organized for online delivery, but are not copy‐edited or typeset. Technical support issues arising from supporting information (other than missing files) should be addressed to the authors.

Supplementary

Click here for additional data file.