Conformational preferences of Ac-Gly-NHMe in solution

Abstract

The conformational behaviour of Ac-Gly-NHMe in nonpolar, polar and polar protic solutions was systematically studied in this work by theoretical calculations and experimental infrared and ¹H NMR spectroscopies. Ac-Gly-NHMe prefers a gauche conformer with a strong seven-membered intramolecular hydrogen bond for the isolated compound and in nonpolar solvents, but such preference changes in polar and polar protic solvents. Elucidation of Ac-Gly-NHMe preferences was also supported by studying the conformers of its CF₃-C(O)-Gly-NHMe and Ac-Gly-N(Me)₂ derivatives in solution.

---

1. Introduction

The conformational equilibrium of amino acids and small peptides is a topic of intense research in the literature, which is being studied both experimentally and theoretically¹ in order to elucidate polypeptide and protein polymeric chain structure and folding pathways.² In particular, glycine, the simplest amino acid, is by far the most studied compound. Conformational preferences of glycine are, however, far from being fully understood.³ Indeed, conformational preferences of not only glycine, but of all amino acids are indicated to be the result of a complex interplay between intramolecular hydrogen bond (IHB) formation⁴ and steric and hyperconjugative interactions for the conformations, well-known e.g. for the simplest hydrocarbons.⁵

In an effort to understand amino acid conformational preferences and the forces that govern such preferences we have been undertaking systematic studies for different amino acid compounds and some of their ester derivatives.⁶ Contrary to the common interpretations from the literature, we have found that the interplay between steric and hyperconjugative interactions and not IHBs are the main forces ruling the conformational behaviour of this important class of natural compounds.

The rationalization of the forces that govern peptide-like compounds of the general formula Ac-R-NHMe (R = amino acid) is desirable to understand the structures of natural macromolecules that contain such amino acid residues as building blocks. In the present paper we report experimental ¹H NMR and infrared (IR) conformational studies of the dipeptide model Ac-Gly-NHMe (1) and its fluorinated CF₃-C(O)-Gly-NHMe (2) and N-methylated Ac-Gly-N(Me)₂ (3) derivatives (Scheme 1). The experiments in solution are supported by theoretical calculations, in the framework of quantum topological methods as the Quantum Theory of Atoms in Molecules (QTAIM),⁷ Electron Localization Functions (ELF)⁸ and the recently developed Non-Covalent Interactions (NCI)⁹ and Density Overlap Regions Indicator¹⁰ methods and the orbital based Natural Bond Orbital (NBO) method.¹¹

Scheme 1 Ac-Gly-NHMe (1), CF₃-C(O)-Gly-NHMe (2) and Ac-Gly-N(Me)₂ (3) structure representations.

2. Experimental section

NMR spectra

Compounds 1, 2 and 3 were purchased from Ukrorgsyntez Ltd. (UORSY) and used without further purification. ¹H NMR experiments were recorded on a Bruker Avance-III spectrometer operating at 600.17 MHz for ¹H. Spectra were recorded in solutions of ca. 1 mg in 0.7 mL of CD₂Cl₂, acetone-d₆, acetonitrile-d₃, DMSO-d₆, CD₃OH and H₂O (18.2 MΩ cm from a Millipore system). An insertion tube with D₂O in the H₂O sample was used in order to maintain the field-frequency lock and avoid deuteration of the N–H bonds. Commercial solvents were referenced to internal TMS. Typical conditions used were as follows: a probe temperature of 25 °C, from 4 to 256 transients (depending on solute solubility), a spectral width of 6.0 kHz, 64 k data points, an acquisition time of 5.5 s and zero-filled to 128 k points. The WATERGATE (water suppression by gradient-tailored excitation)¹² and solvent presaturation¹³ approaches were used in order to suppress the H(O) solvent signal in the H₂O and CD₃OH solvents. ¹H NMR spectra are provided in the ESI.†

IR spectra

The IR spectra were recorded on a FTIR Shimadzu IRPrestige-21 spectrometer equipped with a CsI beamsplitter. Spectra of compounds 1–3 were obtained in CH₂Cl₂ and acetonitrile solvents by using a 0.5 mm width NaCl round cell window with a concentration of 0.02 M. The following IR spectrometer conditions were used: number of scans = 128, resolution = 2 cm⁻¹, spectral range = 650–4000 cm⁻¹. The equipment was purged with continuous dry nitrogen gas. Spectra in H₂O (18.2 MΩ cm from a Millipore system) and D₂O (99.9% from Sigma Aldrich) were obtained with a ZnSe 45° incidence angle Pike Tech ATR-8000HA horizontal attenuated total reflectance (HATR) sampling accessory. Reflectance spectra were converted to absorption spectra by the Kramers–Kronig analysis method. Experimental and predicted IR spectra are provided in the ESI.† The experimental spectrum regions corresponding to N–H stretching bands in CH₂Cl₂ and CH₃CN were deconvoluted by using the GRAMS curve fitting software.¹⁴

Theoretical calculations

Conformers of 1, 2 and 3 were initially searched by 3-dimensional potential energy surfaces (PES) constructed by scanning its ψ[N–C–C(O)–N] and ϕ[C(O)–N–C–C(O)] dihedral angles (Fig. 1) from 0° to 360° in steps of 10° at the B3LYP/cc-pVDZ level (Fig. S1 in the ESI†), using the Gaussian 09 program.¹⁵ This procedure, however, gave rise to only 2 conformers, a and b,¹⁶ for 1 and 2 and only one a conformer for 3 (Fig. 2).

Fig. 1 Dipeptide model dihedral angle representations.

Fig. 2 Most stable conformers of compounds 1, 2 and 3 optimised at the B3LYP-D3/aug-cc-pVDZ level (O red, N blue, C grey).

Additionally, a B3LYP/cc-pVDZ Monte Carlo conformational search was carried out in Spartan 14 program¹⁷ by using a 10 kcal mol⁻¹ threshold and 5000 K maximum temperature, which give rise to many more conformers, namely 11, 9 and 7 for 1, 2 and 3, respectively (Fig. S2 in the ESI†). Optimisations and frequency calculations were carried out at the B3LYP and B3LYP-D3 levels for all conformers found in the Monte Carlo calculations, and the lack of negative frequencies confirmed that all conformers are energy minima. All conformers of compound 1 were re-optimised by using the B3LYP, BLYP, BP86, B97 and M06 functionals with and without DFT-D3 (ref. 18 and 19) corrections and the MP2 ab initio method with the aug-cc-pVDZ basis set and also by using the AM1, PM3 and PM6 semi-empirical methods (energy values in Table S1 in the ESI†) as implemented in the Gaussian 09 program. The B3LYP-D3/aug-cc-pVDZ level showed the smallest mean absolute deviation (MAD) from CCSD(T)-F12a/VDZ-F12 single point calculations performed on MOLPRO program²⁰ (Table S1 in the ESI†) and, hence, it was used in all subsequent calculations. The B3LYP-D3/aug-cc-pVDZ energies were converted into enthalpies and Gibbs free energies using standard thermodynamic corrections from the B3LYP-D3/aug-cc-pVDZ frequency calculations. The enthalpies were in better agreement with experimental IR populations than Gibbs free energies (see the section on infrared spectra in the ESI†). All conformers were also optimised in the IEF-PCM [integral equation formalism variant of the polarizable continuum model]²¹ implicit solvent model at the B3LYP-D3/aug-cc-pVDZ level. NBO analysis¹¹ was performed at the B3LYP-D3/aug-cc-pVDZ level employing geometries fully optimised at the same level for the isolated compounds. NMR ³J_HH spin–spin coupling constant (SSCC) values were calculated at the BHandH/EPR-III level.^22,23 This level was used because the BHandH functional performs well for a large variety of spin–spin coupling constants (SSCCs) involving carbon, fluorine and hydrogen atoms²⁴ and the EPR-III basis set that was developed and optimised for the computation of the Fermi-contact term, which is usually the leading component of SSCCs.²⁵ The second-order polarization propagator approximation (coupled cluster singles and doubles) SOPPA(CCSD)²⁶ method was also used for comparison with the BHandH/EPR-III level. SOPPA(CCSD) calculations used the EPR-III basis set for ¹H and the cc-pVDZ basis for the remaining atoms and were ran in the Dalton 2013 program.²⁷ QTAIM, ELF, NCI and DORI topological analysis were carried out on the electron densities obtained from the B3LYP-D3/aug-cc-pVDZ optimised geometries through the AIMALL 14.06.21,²⁸ TopMod²⁹ and NCIPLOT 3.09 programs, respectively.

3. Results and discussion

Calculated conformer populations are shown in Table 1 for Ac-Gly-NHMe (1). These populations are derived from enthalpies obtained at the B3LYP-D3/aug-cc-pVDZ level for the isolated compound an in the IEF-PCM continuum media (geometrical representations of all conformers are given in Fig. S2 in the ESI†). Conformer 1a is the global minimum for the isolated compound, corresponding to 61.1% of the total population. This conformer is also called as γ¹⁶ in the literature, since it may be found in γ-turns of polypeptide and proteins;³⁰ it has also been labeled C7, because it may form a N–H⋯O 7-membered IHB.³¹ With the IEF-PCM implicit model, the population of this conformer decreases to 46.1% in the fairly unpolar solvent CH₂Cl₂ and it is even smaller in more polar solvents (Table 1) with considerable increase of conformer 1d, which may form a 5-membered N–H⋯N IHB (Table 1 and Fig. 2). Population of conformer 1b, also called C5 in the literature (because it may form a 5-membered N–H⋯O IHB),¹⁶ decreases from the isolated compound to solution, but remains almost constant in the other solvents. The dipole moments (μ) of 1a, 1b and 1d are calculated to be 3.27 D, 3.35 D and 5.22 D, respectively. Thus, it is reasonable that population of conformer 1d increases with the dielectric constant of the media due to its higher dipole moment in comparison with 1a and 1b.

Table 1 Conformer populations (in %) of compound 1 from enthalpies (ΔH) obtained at the B3LYP-D3/aug-cc-pVDZ level for the isolated compound and in different IEF-PCM solvent models

	Isolated	CH2Cl2	Acetone	CH3CN	DMSO	CH3OH	H2O
1a	61.1	46.1	36.5	32.4	31.2	33.0	29.2
1b	36.4	27.3	24.8	23.1	22.6	23.4	21.5
1c	1.5	3.6	3.1	2.9	2.8	2.9	2.6
1d	0.8	19.6	31.9	37.7	39.5	36.9	42.7
1e	0.2	2.2	2.5	2.5	2.5	2.5	2.5
1f	0.0	0.0	0.3	0.3	0.3	0.3	0.3
1g	0.0	1.0	0.5	0.6	0.6	0.6	0.6
1h	0.0	0.2	0.3	0.3	0.3	0.3	0.4
1i	0.0	0.1	0.1	0.1	0.1	0.1	0.1
1j	0.0	0.0	0.0	0.0	0.0	0.0	0.0
1k	0.0	0.0	0.0	0.0	0.0	0.0	0.0

---

Experimental and theoretical IR spectra of the amide N–H bond stretching region of Ac-Gly-NHMe are shown in Fig. 3 (in CH₂Cl₂ and CH₃CN solvents). Only conformers 1a, 1b and 1d, which account for >90% of total population at the B3LYP-D3/aug-cc-pVDZ level, were used for the theoretical. Experimental IR populations were corrected for each conformer with the calculated N–H stretching intensities in km mol⁻¹ (for the graph of calculated intensity for each conformer see Fig. S3 in the ESI†).

Fig. 3 IR N–H bond stretching region for 1. (a) Experimental deconvoluted spectrum in CH₂Cl₂. (b) Theoretical B3LYP-D3/aug-cc-pVDZ spectrum in CH₂Cl₂ (IEF-PCM). (c) Experimental spectrum in CH₃CN. (d) Theoretical B3LYP-D3/aug-cc-pVDZ spectrum in CH₃CN (IEF-PCM). Conformer populations are indicated in each spectrum. Theoretical populations obtained at [B3LYP-D3/aug-cc-pVDZ] level. Experimental IR populations were corrected by calculated conformer molar absorptivities.

The observed and calculated populations of 1a and 1b in CH₂Cl₂ are in reasonable agreement.³² From experimental IR, conformers 1a and 1b are the most populated conformers in CH₂Cl₂, with an observed population of 68.1% and 31.9%, respectively (Fig. 3). The calculated population of conformer 1a (46.1%) is smaller than the observed experimental IR result and calculated population of conformer 1b (27.3%) is in good agreement with the experimental. Conformer 1d could not be observed experimentally, which IR band could be hidden below the most abundant 1a and 1b conformers. Population of conformer 1d in CH₂Cl₂ is calculated to be 19.6% at the B3LYP-D3/aug-cc-pVDZ level with IEF-PCM implicit solvent model. The IEF-PCM calculations in acetonitrile indicate that conformer 1d becomes the global minimum with 37.7% of the total population. However, the experimental N–H band of Ac-Gly-NHMe in acetonitrile is much broader than in CH₂Cl₂ (Fig. 3), presumably due to intermolecular HB formation with the solvent, and no experimental conformer population could be derived from this spectrum. It was also not possible to obtain the experimental populations in H₂O and D₂O from the amide N–H stretching bands, since H₂O absorbs strongly in the same region range as N–H bands of 1 and also D₂O absorbs in the same region as N–D bands, which arise from proton exchange with the solvent. However, experimental regions corresponding to amide I (C=O stretchings) and amide II bands [C(O)–N–H angular deformations] could be observed. While they present many shoulders in CH₂Cl₂ and CH₃CN, corresponding to a mix of conformers 1a, 1b and 1d (Fig. 4a and c), these bands seem to be more symmetrical in H₂O and D₂O (Fig. 4e and f). Thus, one might infer that only one conformer would be present in water. However, the bands are very broad in water (presumably due to intermolecular HB formation between Ac-Gly-NHMe C=O and N–H bonds and the solvent); and bands from other conformers could just be hidden within. In fact, the IEF-PCM calculations (Table 1) indicate that all conformers 1a, 1b and 1d would be present in considerable amount in water, 1d being the global minimum.

Fig. 4 IR amide I (C=O stretching) and II (N–H stretching) regions for 1. (a) Experimental spectrum in CH₂Cl₂. (b) Theoretical spectrum in CH₂Cl₂. (c) Experimental spectrum in CH₃CN (amide I only, since CH₃CN absorbs strongly in the amide II region). (d) Theoretical spectrum in CH₃CN. (e) Experimental spectrum in H₂O. (f) Experimental spectrum in D₂O. Theoretical spectra from conformers 1a, 1b and 1d IR intensities/populations obtained at the B3LYP-D3/aug-cc-pVDZ level with the IEF-PCM model.

QTAIM, ELF, NCI, DORI and NBO methods were then applied for the isolated Ac-Gly-NHMe conformers 1a, 1b and 1d in order to understand the intramolecular interactions that stabilise each conformer. The ELF, NCI, DORI and NBO methods found an IHB for all 3 conformers, while QTAIM found it only for conformer 1a (ESI Fig. S4†). Indeed, QTAIM is being repeatedly criticised in the literature, since it may not find a HB in situations where it is expected to be formed either by other theoretical methods or by experiment.³³ The ELF, through the so-called core-valence bond index (CVBI),³⁴ indicates that conformer 1a forms the strongest IHB. The same is found with NCI and DORI, through the signal(λ₂)ρ values from RDG and DORI peaks corresponding to IHB formation, and with NBO analysis, through n → σ^*_NH interaction energies (Table 2; details in the ESI Fig. S4–S8†). These findings are consistent with the short calculated C=O⋯H–N distance in the 7-membered ring closed by the IHB (2.04 Å; Table 2). Larger distances are found in conformer 1b, which forms a weak C=O⋯H–N IHB within a 5 membered ring (2.21 Å) and in 1d, which forms the weakest N⋯H–N hydrogen bond (2.35 Å).

Table 2 IHB parameters for compounds 1 and 2 from QTAIM (ρ), ELF (CVBI), NCI and DORI [sign(λ₂)ρ in au] and NBO orbital interactions corresponding to IHBs (n → σ^*_NH) in kcal mol⁻¹. Calculated IHB distances are also shown in Å

	Ac-Gly-NHMe			CF3-C(O)-Gly-NHMe
	1a	1b	1d	2a	2b	2c	2d
a More positive CVBI values correspond to weaker IHBs.b More negative values correspond to stronger IHBs.
ρ	0.022	—	—	0.017	—	—	—
CVBI^a	+0.012	+0.032	+0.042	+0.027	+0.030	+0.030	+0.045
sign(λ2)ρ^b	−0.022	−0.019	−0.016	−0.017	−0.020	−0.020	−0.015
nO(1) → σ^*NH	2.57	0.67	—	1.65	0.79	0.83	—
nO(2) → σ^*NH	3.76	2.04	—	2.69	2.49	2.59	—
nN → σ^*NH	—	—	1.24	—	—	—	1.24
nF(2) → σ^*NH	—	—	—	1.10	0.94	0.94	1.10
IHB distance	2.04	2.21	2.35	2.14	2.19	2.18	2.37

---

The N⋯H–N hydrogen bond in 1d may be rationalised to be weak due to the low availability of the amide nitrogen lone pairs (n_N), which are expected be in resonance within the R₂N–C=O amide fragment. Indeed, the Natural Resonance Theory (NRT),³⁵ indicates that 3 from the 4 main Ac-Gly-NHMe resonance hybrids (from a total of 126) have the nitrogen lone pairs in resonance (Fig. 5). All charged resonance hybrids strengthen the IHBs in conformers 1a and 1b, but weaken the N⋯H–N IHB in 1d IHB, since it localises negative charges in the O atoms (H atom acceptors in 1a and 1b) and positive charges on the N atoms (H atom acceptor in 1d).

Fig. 5 Main resonance contributor percentages obtained from the NRT analysis to Ac-Gly-NHMe conformer 1a. Percentage values for conformers 1b and 1d are almost the same, with a maximum deviation of only 1.6%.

It is worth to mention that both NCI and DORI found other weak intramolecular interactions, such as 5-membered O⋯H–C IHBs, that could not be found by QTAIM, ELF and NBO methods, wherein DORI found the highest number of those interactions (ESI; Fig. S4–S8†). Such interactions are indicated to be stabilising by both NCI and DORI. However, both NCI and DORI use the sign of λ₂ parameter in order to differentiate stabilising and destabilising interactions. As observed in previous works: “care is recommended when interpreting the sign of λ₂ in very weak interactions, because in these cases the sign might depend on the method of calculation”.³⁶

Relative total enthalpy corrected energies [ΔH(T)], natural non-Lewis (hyperconjugative) contribution energies [ΔH(NL)] and natural Lewis structure (steric/electrostatic) contribution energies [ΔH(L)] for conformers 1a, 1b and 1d obtained from NBO analysis (deletion of all donor–acceptor interactions) at the B3LYP-D3/aug-cc-pVDZ level are collected in Table 3.

Table 3 Total relative enthalpies [ΔH(T)], energy of the hypothetical case where hyperconjugation is removed [ΔH(L)],^a and hyperconjugative energy [ΔH(NL)],^a all in kcal mol⁻¹, for Ac-Gly-NHMe conformers 1a, 1b and 1d isolated and in different media (IEF-PCM). Calculations at the B3LYP-D3/aug-cc-pVDZ level

		Isolated	CH2Cl2	Acetone	Acetonitrile	DMSO	CH3OH	H2O
a Obtained by adding the enthalpic corrections from ΔH(T).
1a	ΔH(T)	0.00	0.00	0.00	0.09	0.14	0.07	0.23
	ΔH(L)	11.00	11.93	11.91	11.90	11.87	11.90	11.90
	ΔH(NL)	13.57	12.44	11.99	11.81	11.73	11.83	11.67
1b	ΔH(T)	0.31	0.31	0.23	0.29	0.33	0.27	0.41
	ΔH(L)	2.14	3.43	3.47	3.48	3.46	3.48	3.47
	ΔH(NL)	4.40	3.63	3.32	3.19	3.13	3.21	3.06
1d	ΔH(T)	2.57	0.51	0.08	0.00	0.00	0.00	0.00
	ΔH(L)	0.00	0.00	0.00	0.00	0.00	0.00	0.00
	ΔH(NL)	0.00	0.00	0.00	0.00	0.00	0.00	0.00

---

Table 3 indicates that 1a suffers the highest steric interactions [ΔH(L) values], followed by 1b and 1d as the conformer that experiences the lowest steric interactions. Hyperconjugative stabilisation operates in the other way round [ΔH(NL) values], i.e., it is highest for 1a and lowest for 1d. Conformer 1a is the most stable for the isolated compound, and it has an approximately gauche geometry, with dihedral angles of ϕ[N–C–C(O)–N] = 68.6° and ψ[C(O)–N–C–C(O)] = 81.2°. Thus, the conformational preference in Ac-Gly-NHMe is a consequence of the well known gauche effect,³⁷ i.e., 1a is the lowest energy conformer even though it experiences the highest steric and electrostatic destabilisation. The stability of conformer 1a is assisted by its strong IHB within a 7-membered ring, which explains its high hyperconjugative stabilisation by increasing it by 6.33 kcal mol⁻¹ (n_O(1) → σ^*_NH = 2.57 kcal mol⁻¹ + n_O(2) → σ^*_NH = 3.76; orbital representations in the ESI; Fig. S8†). The Natural Steric Analysis (NSA)³⁸ is in qualitative agreement with the ΔH(L) energy parameter and indicates that 1a is more destabilised due to steric interactions (+250.27 kcal mol⁻¹) than 1b (+249.66 kcal mol⁻¹) and 1d (+245.67 kcal mol⁻¹), whose steric energy values are not due to any particular orbital–orbital interaction, but the sum of all of them.

As shown previously, theory indicates that conformer 1d becomes the most stable in acetonitrile. Indeed, the stability of 1a is highly dependent on its N–H⋯O 7-membered IHB, while that of conformer 1d is due to its minor destabilisation by steric effects. Also, 1b could have increased population in polar solvents not only due to its higher dipole moment, but also due to its smaller dependence of IHB stabilisation than 1a and 1b.

We turn now to ¹H NMR, experimental ³J_HH spin–spin coupling constant (SSCC) and chemical shift values (Table 4). The ³J_HaHb values are almost constant in the studied solvents. Based on the well known Karplus relationship,³⁹ one would expect that the ³J_HaHb values would be similar for 1a and 1d, with higher values than 1b, since the former conformers have both a cis and an anti relationship between Ha and Hb atoms, while 1b has only anticlinal relationships between these atoms (Fig. 6). Thus, the observation that the ³J_HaHb values are almost constant in different solvents would be either because the conformer populations do not change among the applied solvents or that the populations are shifting from conformer 1a to 1d, which have similar ³J_HaHb values. IEF-PCM calculations (Table 1) suggest that the second hypothesis is the correct one, i.e., the population of 1b is almost constant in the different solvents and that of 1a shifts to 1d when the solvent polarity increases.

Table 4 Experimental chemical shifts (ppm) and ³J_HH spin–spin coupling constants (SSCCs, Hz) of Ac-Gly-NHMe in solvents with different dielectric constants (ε)

Solvent	ε	δH(a)	δH(b)	δH(c)	δH(d)	δH(e)	^3JHaHb	^3JHcHe
CD2Cl2	8.9	6.18	3.83	5.92	1.99	2.78	5.34	4.86
Acetone-d6	20.7	7.34	3.77	7.15	1.92	2.69	5.76	4.74
CD3CN	37.5	6.71	3.68	6.52	1.92	2.66	5.88	4.80
DMSO-d6	46.7	8.08	3.61	7.74	1.85	2.57	5.94	4.62
CD3OH	32.7	8.25	3.79	7.90	2.00	2.73	5.82	4.74
H2O	80.1	8.30	3.85	7.85	2.05	2.74	5.76	4.80

---

Fig. 6 Newman representations of conformers of 1.

BHandH/EPR-III and SOPPA(CCSD)/EPR-III ³J_HaHb SSCCs for all 1 conformers are given in the ESI (Table S5†). Both methods indicate that the ³J_HaHb values for 1a and 1d are indeed similar (∼7 and ∼6 Hz, respectively) and higher than the corresponding values for 1b [∼3 Hz for BHandH and ∼2 Hz for SOPPA(CCSD), respectively]. Fig. 7a shows the calculated ³J_HaHb, weighted by all populations of conformer 1. In this case, BHandH/EPR-III results are in better agreement with experiment than SOPPA(CCSD)/EPR-III. Fig. 7b uses corrected IR populations (from Fig. 3) and BHandH/EPR-III calculated ³J_HaHb values. ³J_HaHb values obtained from theoretical and IR-derived populations in CH₂Cl₂ are in reasonable accordance (4.76 Hz and 5.77 Hz) with the experimental value (5.34 Hz) and theoretical is in excellent agreement in CH₃CN (theoretical = 5.87 Hz; experimental = 5.98 Hz). Thus, theory and experimental IR and ¹H NMR indicates that 1a and 1b are preferred for the isolated compound and in nonpolar solvents, but 1d is the preferential one and compete with 1b in more polar solvents. If mostly one conformer is present in water, it may not be 1b, even though it has an extended geometry and presumably smaller ΔG of solvation than the remaining conformers, because the calculated ³J_HaHb SSCC for 1b (BHandH = 2.7 Hz and SOPPA = 1.8 Hz; ESI Table S5†) is much smaller than the experimental (5.6 Hz). On the other hand, conformer 1d more closely matches the experimental value in water (BHandH = 7.2 Hz and SOPPA = 6.2 Hz). The competition between 1b and 1d in water is in agreement with previous molecular dynamics and QM/MM studies from the literature, which found both 1b and 1d depending on the level of calculation⁴⁰ and that 1d should be the preferential if increased number of water molecules are taken into account. Indeed, by simulating 11 water molecules around Ac-Gly-NHMe, Boopathi et al.⁴¹ showed, by using molecular dynamics calculations, that conformer 1d would be the preferential one in water.

Fig. 7 Sum of the weighted contributions [(η_i/η_T) × J] for the ³J_HaHb couplings of all Ac-Gly-NHMe conformers (obtained as ³J_H10H11 + ³J_H11H13/2); SSCCs were obtained in different solvents (B3LYP-D3/aug-cc-pVDZ level optimisation) by using the IEF-PCM model. (a) Conformer contributions obtained from ΔH populations and (b) conformer contributions obtained from experimental IR populations in CH₂Cl₂ (using BHandH/EPR-III calculated ³J_HH).

In order to get a deeper insight into the factors that govern conformer stability in the more polar solvents, we decided to “manipulate” the H-bond in both 1a and 1d by studying the CF₃-C(O)-Gly-NHMe (2) derivative. The electron withdrawing CF₃-group should weaken the C=O⋯H–N IHB in 2a and 2d (representations in Fig. 2) in comparison to 1a and 1d, because it withdraws electron density from the H atom acceptor groups in these conformers. Also, the CF₃-group may strengthen the IHB in conformer 2b in comparison to 1b, since it withdraws electron density from the H(N) atom participating in the IHB in this conformer. All ELF, NCI, DORI and NBO parameters indicate that this is indeed the case (Table 2). QTAIM again could find an IHB only for conformer 2a. All methods, except QTAIM, also indicate formation of a CF⋯HN IHB for conformers 2a–d, which is of similar strength for all of them (n_F(2) → σ^*_NH interaction energies; Table 2).

Theory indicates that conformer 2a is not the most stable conformer. Conformer 2b is the most stable one with more than 80% of the total population of 2 (Table 5). Such relative stability decreases in more polar solvents and 2d becomes progressively more stable as the dielectric constant increases. IR populations are not in quantitative agreement with theory. Although 2b conformer is the most stable in CH₂Cl₂ (59.6%), conformer 2d becomes the global minimum in acetonitrile (56.3%; Fig. 8). Thus, conformer 2d, which forms the weakest IHB, is the most stable in polar solvents for both 1 and 2. Amide I bands in H₂O (Fig. 8e) and D₂O (Fig. 8f) are sharper for 2 than for 1 (cf. Fig. 4e and f), but show some shoulders in the amide II band. This could be taken as indication that there is more than one conformer in water, which could be both 2b and 2d.

Table 5 Conformer populations (in %) from enthalpies (ΔH) of compounds 2 and 3, obtained at the B3LYP-D3/aug-cc-pVDZ level for the isolated compound and in different IEF-PCM solvents

	Isolated	CH2Cl2	Acetone	CH3CN	DMSO	CH3OH	H2O
a 2d is not a minimum for the isolated molecule.
2a	15.8	8.4	6.8	6.4	6.2	6.4	6.0
2b	81.1	74.5	72.2	70.7	70.2	70.9	69.4
2c	3.03	9.3	8.3	8.0	7.9	8.0	7.8
2d^a	—	7.6	12.5	14.8	15.5	14.5	16.5
3a	2.1	2.6	3.0	3.1	3.2	3.1	3.3
3b	97.7	93.6	91.0	89.6	89.1	89.8	88.3
3c	0.2	3.2	5.0	5.8	6.1	5.7	6.6
3d	0.0	0.3	0.5	0.6	0.7	0.6	0.8

---

Fig. 8 IR N–H stretchings (a–d) and C=O stretching regions (e and f) for 2. (a) Experimental deconvoluted spectrum in CH₂Cl₂. (b) Theoretical B3LYP-D3/aug-cc-pVDZ IR spectrum in CH₂Cl₂ (IEF-PCM). (c) Experimental deconvoluted spectrum in CH₃CN. (d) Theoretical B3LYP-D3/aug-cc-pVDZ IR spectrum in CH₃CN (IEF-PCM). (e) Experimental spectrum in H₂O. (f) Experimental spectrum in D₂O. Conformer populations are indicated in each spectrum. Experimental IR populations were corrected by conformer calculated molar absorptivities.

¹H NMR parameters for 2 are collected in Table 6. Unfortunately, the H(N) atom in 2 is much more acidic than in 1 and exchanges quite fast within polar protic solvents. It is thus not possible to determine ³J_HaHb SSCC in methanol and water, which could have indicated if either 2b or 2d would be the preferential one, since they have different calculated ³J_HaHb values (ESI Table S5†).

Table 6 Experimental chemical shifts (ppm) and ³J_HH spin–spin coupling constants (SSCCs, Hz) of CF₃-C(O)-Gly-NHMe in solvents with different dielectric constants (ε)

Solvent	ε	δH(a)	δH(b)	δH(c)	δH(d)	^3JHaHb	^3JHcHd
CD2Cl2	8.9	7.19	3.96	5.62	2.84	4.68	4.86
Acetone-d6	20.7	8.53	3.97	7.31	2.73	5.64	4.74
Acetonitrile-d3	37.5	7.76	3.84	6.50	2.69	4.98	4.75
DMSO-d6	46.7	9.62	3.76	7.98	2.60	5.88	4.62
CD3OH	32.7	—	3.91	8.03	2.75	—	4.63
H2O	78.5	—	4.02	7.99	2.75	—	—

---

Another way to probe if either the b or d conformer would be the preferential one in polar protic solvents, is to look at derivatives where one of them is disfavoured by design. Conformers 1d and 2d are stabilised by an IHB involving the C-terminal NHMe group (Fig. 2). Because changing this group to NMe₂ should block this interaction, we finally studied Ac-Gly-N(Me)₂ (3). Theoretical calculations indicate that 3b has ∼90% of the total population in all solvents (Table 5) and that, as expected, the geometries of conformers a and d are not the same as for compounds 1 and 2 (Fig. 2). This has consequences for the chemical shifts and SSCCs (Table 7). Thus, if 3b is the preferential conformer, with ∼90% of the total population in all solvents, one would expect that the experimental ³J_HaHb SSCC would decrease considerably for 3 in comparison to 1 and 2. However, as shown in Table 7, the ³J_HaHb SSCCs for 3 are overall only slightly smaller than those observed in 1.

Table 7 Experimental chemical shift values (ppm) and ³J_HH spin–spin coupling constant (SSCC) values (Hz) of Ac-Gly-N(Me)₂ in solvents with different dielectric constants (ε)

Solvent	ε	δH(a)	δH(b)	δH(c)	δH(d)(e)^a	δH(d)(e)^a	^3JHaHb
a H(d) and H(e) were not assigned.
CD2Cl2	8.9	6.81	4.00	1.99	2.96	2.94	4.26
Acetone-d6	20.7	7.19	3.98	1.93	2.90	3.01	4.80
Acetonitrile-d3	37.5	6.75	3.93	1.92	2.88	2.93	5.16
DMSO-d6	46.7	7.91	3.89	1.86	2.94	2.82	5.46
CD3OH	32.7	8.07	4.05	2.02	3.05	2.96	5.04
H2O	80.1	—	4.07	2.05	3.03	2.93	—

---

Experimental IR spectra of compound 3 in CD₂Cl₂, acetonitrile and water are shown in Fig. 9. In excellent agreement with theoretical calculations, experimental IR populations indicate that conformer 3b is the most prevalent in CH₂Cl₂ accounting for 93.4% of the total population (Fig. 9). The N–H band is very broad in acetonitrile and conformer populations could not be taken from it. Conformers 3c and 3d have the highest calculated dipole moment values (10.31 D and 10.08 D, respectively), while 3b has a relative small calculated dipole moment (4.62 D). Differently from compounds 1 and 2, the amide I IR band of compound 3 has a shoulder in water (Fig. 3f), hence, indicating that more than one conformer is stable in this solvent. Thus, even though 3b has an extended geometry and is more prone to be solvated by water, such conformer would not be the most stable in more polar or polar protic solvents if other conformers with higher dipole moments are present. This may also be the case for compounds 1 and 2, whose conformers 1b and 2b compete with 1d and 2d, respectively, in polar solvents. However, as shown previously, d conformers have higher dipole moments than b conformers and, consequently, should be the preferential ones in polar and polar protic solvents.

Fig. 9 IR N–H stretching (a–d) and C=O stretching regions (e and f) regions for 3. (a) Experimental deconvoluted spectrum in CH₂Cl₂. (b) Theoretical B3LYP-D3/aug-cc-pVDZ IR spectrum in CH₂Cl₂ (IEF-PCM). (c) Experimental deconvoluted spectrum in CH₃CN. (d) Theoretical B3LYP-D3/aug-cc-pVDZ IR spectrum in CH₃CN (IEF-PCM). (e) Experimental spectrum in H₂O. (f) Experimental spectrum in D₂O. Conformer populations are indicated in each spectrum. Experimental IR populations were corrected by conformer calculated molar absorptivities.

4. Conclusions

The conformational preferences of Ac-Gly-NHMe change considerably from nonpolar solvents such as CH₂Cl₂ to polar (CH₃CN) and polar protic solvents (methanol and water). Theoretical calculations and experimental IR indicate that the conformational preferences of Ac-Gly-NHMe shifts from 1a, which is stabilised by a strong N–H⋯O IHB and is prevalent for the isolated molecule and in nonpolar solvents, to conformers 1b and 1d, which are stabilised to a lesser extent by IHBs and have higher dipole moments (for 1d). These results are supported by experimental ³J_HaHb SSCC values and theoretical calculations. The IR and ¹H NMR experimental and theoretical results obtained for CF₃-C(O)-Gly-NHMe and Ac-Gly-N(Me)₂ derivatives highlight the results obtained for Ac-Gly-NHMe, indicating that conformers with higher dipole moments such as 1d in Ac-Gly-NHMe may be the preferential ones in polar protic solvents. We hope that the results of this work may help to understand the conformational behaviour of glycine residues in peptides, proteins and smaller models thereof in either nonpolar and polar environments.

Acknowledgements

The authors thank a Grant #2012/03933-5, São Paulo Research Foundation (FAPESP) for providing financial support for this research and a scholarship to RAC #2011/01170-1 (FAPESP). CNPq is also acknowledged, such as for fellowships (to RR) and a studentship (to RAC). MB thanks the School of Chemistry and EaStCHEM for support and for access to a computer cluster maintained by Dr H. Früchtl. We also thank Prof. David O'Hagan for support and fruitful discussions.

Notes and references

1. (a) C. K. Kim, B.-H. Park, H. W. Lee and C. K. Kim, Org. Biomol. Chem., 2013, 11, 1407; (b) J. Groule and F. Jensen, J. Chem. Theory Comput., 2011, 7, 1783; (c) L. Comez, L. Lupi, A. Morresi, M. Paolantoni, P. Sassi and D. Fioretto, J. Phys. Chem. Lett., 2013, 4, 1188; (d) D. Russo, M. A. Gonzalez, E. Pellegrini, J. Combet, J. Ollivier and J. Teixeira, J. Phys. Chem. B, 2013, 117, 2829; (e) V. L. Cruz, J. Ramos and J. Martinez-Salazar, J. Phys. Chem. B, 2012, 116, 469; (f) B. M. Marsh, E. M. Duffy, M. T. Soukup, J. Zhou and E. Garand, J. Phys. Chem. A, 2014, 118, 3906; (g) E. V. Dornshuld, R. A. Vergenz and G. S. Tschumper, J. Phys. Chem. B, 2014, 118, 8583; (h) R. Schweitzer-Stenner, Biophys. J., 2002, 83, 523; (i) H. Torii, J. Phys. Chem. A, 2006, 110, 4822; (j) M. Candelaresi, E. Ragnoni, C. Cappelli, A. Corozzi, M. Lima, S. Monti, B. Mennucci, F. Nuti, A. M. Papini and P. Foggi, J. Phys. Chem. B, 2013, 117, 14226.
2. (a) B. J. Tooze, Introduction to Protein Structure, Garland, New York, 1991; (b) H. Dong, M. Sharma, H.-X. Zhou and T. A. Cross, Biochemistry, 2012, 51, 4779; (c) D. Russo, J. Teixeira, L. Kneller, J. R. D. Copley, J. Ollivier, S. Perticaroli, E. Pellegrini and M. A. Gonzalez, J. Am. Chem. Soc., 2011, 133, 4882; (d) M. C. Asplund, M. T. Zanni and R. M. Hochstrasser, Proc. Natl. Acad. Sci. U. S. A., 2000, 97, 8219.
3. (a) V. Barone, M. Biczysko, J. Bloino and C. Puzzarini, Phys. Chem. Chem. Phys., 2013, 15, 10094; (b) B. M. Marsh, E. M. Duffy, M. T. Soukup, J. Zhou and E. Garand, J. Phys. Chem. A, 2014, 118, 3906; (c) V. Barone, M. Biczysko, J. Bloino and C. Puzzarini, J. Chem. Theory Comput., 2013, 9, 1533; (d) V. Barone, M. Biczysko, J. Bloino and C. Puzzarini, Phys. Chem. Chem. Phys., 2013, 15, 1358.
4. (a) J. L. Alonso, I. Peña, J. C. López and V. Vaquero, Angew. Chem., 2009, 121, 6257; (b) R. M. Balabin, Phys. Chem. Chem. Phys., 2010, 12, 5980; (c) J. Oomens, J. D. Steill and B. Redlich, J. Am. Chem. Soc., 2009, 131, 4310.
5. (a) V. Pophristic and L. Goodman, Nature, 2001, 411, 565–568; (b) R. A. Cormanich and M. P. Freitas, J. Org. Chem., 2009, 74, 8384–8387; (c) Y. Mo, J. Org. Chem., 2010, 75, 2733–2736.
6. (a) R. A. Cormanich, L. C. Ducati and R. Rittner, Chem. Phys., 2011, 387, 85; (b) R. A. Cormanich, L. C. Ducati and R. Rittner, J. Mol. Struct., 2012, 1014, 12; (c) R. A. Cormanich, L. C. Ducati, C. F. Tormena and R. Rittner, Chem. Phys., 2013, 421, 32; (d) R. A. Cormanich, L. C. Ducati, C. F. Tormena and R. Rittner, J. Phys. Org. Chem., 2013, 26, 849; (e) C. J. Duarte, R. A. Cormanich, L. C. Ducati, C. F. Tormena and R. Rittner, J. Mol. Struct., 2013, 1050, 174; (f) R. A. Cormanich, L. C. Ducati, C. F. Tormena and R. Rittner, Spectrochim. Acta, 2014, 123, 482.
7. R. F. W. Bader, Atoms in Molecules: A Quantum Theory, Clarendon, Oxford, 1990.
8. B. Silvi and A. Savin, Nature, 1994, 371, 683.
9. E. Johnson, S. Keinan, P. Mori-Sánchez, J. Contreras-García, A. Cohen and W. Yang, J. Am. Chem. Soc., 2010, 132, 6498.
10. P. de Silva and C. Corminboeuf, J. Chem. Theory Comput., 2014, 10, 3745.
11. A. E. Reed, L. A. Curtiss and F. Weinhold, Chem. Rev., 1988, 88, 889.
12. (a) M. Piotto, V. Saudek and V. Slenár, J. Biomol. NMR, 1992, 2, 661; (b) V. Sklenár, M. Piotto, R. Leppik and V. Saudek, J. Magn. Reson., Ser. A, 1993, 102, 241.
13. D. I. Hoult, J. Magn. Reson., 1976, 21, 337.
14. Grams/AI v. 9.0, ThermoFisher, Woburn, MA, USA, 2009.
15. M. J. Frisch, G. W. Trucks, H. B. Schlegel, G. E. Scuseria, M. A. Robb, J. R. Cheeseman, G. Scalmani, V. Barone, B. Mennucci, G. A. Petersson, H. Nakatsuji, M. Caricato, X. Li, H. P. Hratchian, A. F. Izmaylov, J. Bloino, G. Zheng, J. L. Sonnenberg, M. Hada, M. Ehara, K. Toyota, R. Fukuda, J. Hasegawa, M. Ishida, T. Nakajima, Y. Honda, O. Kitao, H. Nakai, T. Vreven, J. A. Montgomery Jr, J. E. Peralta, F. Ogliaro, M. Bearpark, J. J. Heyd, E. Brothers, K. N. Kudin, V. N. Staroverov, R. Kobayashi, J. Normand, K. Raghavachari, A. Rendell, J. C. Burant, S. S. Iyengar, J. Tomasi, M. Cossi, N. Rega, J. M. Millam, M. Klene, J. E. Knox, J. B. Cross, V. Bakken, C. Adamo, J. Jaramillo, R. Gomperts, R. E. Stratmann, O. Yazyev, A. J. Austin, R. Cammi, C. Pomelli, J. W. Ochterski, R. L. Martin, K. Morokuma, V. G. Zakrzewski, G. A. Voth, P. Salvador, J. J. Dannenberg, S. Dapprich, A. D. Daniels, Ö. Farkas, J. B. Foresman, J. V. Ortiz, J. Cioslowski and D. J. Fox, Gaussian 09, Revision D.01, Gaussian, Inc., Wallingford CT, 2009.
16. Because the larger quantity of conformers studied in the present work, it was not used the commonly adopted nomenclature from the literature: A. Perczel, J. G. Angyin, M. Kajtar, W. Viviani, J.-L. Rivail, J. F. Marcoccia and I. G. Csizmadia, J. Am. Chem. Soc., 1991, 113, 6256.
17. Y. Shao, L. F. Molnar, Y. Jung, J. Kussmann, C. Ochsenfeld, S. T. Brown, A. T. B. Gilbert, L. V. Slipchenko, S. V. Levchenko, D. P. O'Neill, R. A. DiStasio Jr, R. C. Lochan, T. Wang, G. J. O. Beran, N. A. Besley, J. M. Herbert, C. Y. Lin, T. Van Voorhis, S. H. Chien, A. Sodt, R. P. Steele, V. A. Rassolov, P. E. Maslen, P. P. Korambath, R. D. Adamson, B. Austin, J. Baker, E. F. C. Byrd, H. Dachsel, R. J. Doerksen, A. Dreuw, B. D. Dunietz, A. D. Dutoi, T. R. Furlani, S. R. Gwaltney, A. Heyden, S. Hirata, C.-P. Hsu, G. Kedziora, R. Z. Khalliulin, P. Klunzinger, A. M. Lee, M. S. Lee, W. Z. Liang, I. Lotan, N. Nair, B. Peters, E. I. Proynov, P. A. Pieniazek, Y. M. Rhee, J. Ritchie, E. Rosta, C. D. Sherrill, A. C. Simmonett, J. E. Subotnik, H. L. Woodcock III, W. Zhang, A. T. Bell, A. K. Chakraborty, D. M. Chipman, F. J. Keil, A. Warshel, W. J. Hehre, H. F. Schaefer, J. Kong, A. I. Krylov, P. M. W. Gill and M. Head-Gordon, Phys. Chem. Chem. Phys., 2006, 8, 3172.
18. S. Grimme, S. Ehrlich and L. Goerigk, J. Comput. Chem., 2011, 32, 1456.
19. S. Grimme, J. Antony, S. Ehrlich and H. Krieg, J. Chem. Phys., 2010, 132, 154104.
20. (a) H.-J. Werner, P. J. Knowles, G. Knizia, F. R. Manby and M. Schütz, WIREs Comput. Mol. Sci., 2012, 2, 242; (b) H.-J. Werner, P. J. Knowles, G. Knizia, F. R. Manby, M. Schütz, P. Celani, T. Korona, R. Lindh, A. Mitrushenkov, G. Rauhut, K. R. Shamasundar, T. B. Adler, R. D. Amos, A. Bernhardsson, A. Berning, D. L. Cooper, M. J. O. Deegan, A. J. Dobbyn, F. Eckert, E. Goll, C. Hampel, A. Hesselmann, G. Hetzer, T. Hrenar, G. Jansen, C. Köppl, Y. Liu, A. W. Lloyd, R. A. Mata, A. J. May, S. J. McNicholas, W. Meyer, M. E. Mura, A. Nicklass, D. P. O'Neill, P. Palmieri, D. Peng, K. Pflüger, R. Pitzer, M. Reiher, T. Shiozaki, H. Stoll, A. J. Stone, R. Tarroni, T. Thorsteinsson and M. Wang, MOLPRO, version 2012.1, a package of ab initio programs, see http://www.molpro.net.
21. G. Scalmani and M. J. Frisch, J. Chem. Phys., 2010, 132, 114110.
22. A. D. Becke, J. Chem. Phys., 1993, 98, 5648.
23. V. Barone, in Recent Advances in Density Functional Methods, Part I, ed. D. P. Chong, World Scientific Publ. Co., Singapore, 1996.
24. F. Nozirov, T. Kupka and M. Stachów, J. Chem. Phys., 2014, 140, 144303.
25. R. Suardíaz, C. Pérez, R. Crespo-Otero, J. M. García de la Vega and J. San Fabián, J. Chem. Theor. Comput., 2008, 4, 448.
26. T. Enevoldsen, J. Oddershede and S. P. A. Sauer, Theor. Chem. Acc., 1998, 100, 275.
27. Dalton, a Molecular Electronic Structure Program, Release DALTON2013.0, 2013, see http://daltonprogram.org/.
28. T. A. Keith, AIMAll (Version 14.06.21), TK Gristmill Software, Overland Park, KS, USA, 2013, http://aim.tkgristmill.com.
29. S. Noury, X. Krokidis, F. Fuster and B. Silvi, ToPMoD, Laboratoire de Chemie Théorique, Université Pierre et Marie Curie, París, http://www.lct.jussieu.fr/pagesperso/silvi/topmod_english.html 1999.
30. K.-C. Chou, Anal. Biochem., 2000, 286, 1.
31. Y. K. Kang and J. B. Byun, J. Comput. Chem., 2010, 31, 2915.
32. We note that “C5” isomers akin to 1b can have N–H stretching frequencies below 3400 cm⁻¹ e.g. for Cα-tetrasubstituted homopeptides: M. Crisma, A. Moretto, C. Peggion, L. Panella, B. Kaptein, Q. B. Broxterman, F. Formaggio and C. Toniolo, Amino Acids, 2011, 41, 629 , our assignment of frequencies above that value is in accordance with the computations.
33. (a) J. R. Lane, J. Contreras-Garcia, J.-P. Piquemal, B. J. Miller and H. G. Kjaergaard, J. Chem. Theory Comput., 2013, 9, 3263; (b) F. Weinhold, P. R. Schleyer and W. C. McKee, J. Comput. Chem., 2014, 35, 1499.
34. F. Fuster and B. Silvi, Theor. Chem. Acc., 2000, 104, 13.
35. E. D. Glendening and F. Weinhold, J. Comput. Chem., 1998, 19, 593.
36. (a) A. Otero-de-la-Roza, E. R. Johnson and J. Contreras-García, Phys. Chem. Chem. Phys., 2012, 14, 12165; (b) R. A. Cormanich, R. Rittner, M. P. Freitas and M. Bühl, Phys. Chem. Chem. Phys., 2014, 16, 19212; (c) R. A. Cormanich, R. Rittner, D. O'Hagan and M. Bühl, J. Phys. Chem. A, 2014, 118, 7901.
37. P. R. Rablen, R. W. Hoffmann, D. A. Hrovat and W. T. J. Bordenc, J. Chem. Soc., Perkin Trans. 2, 1999, 1719.
38. J. K. Badenhoop and F. Weinhold, J. Chem. Phys., 1997, 107, 5406.
39. M. Karplus, J. Chem. Phys., 1959, 30, 11.
40. H. Hu, M. Elstner and J. Hermans, Proteins: Struct., Funct., Genet., 2003, 50, 451.
41. S. Boopathi and P. Kolandaivel, J. Biomol. Struct. Dyn., 2013, 30, 158.

---

Footnote

† Electronic supplementary information (ESI) available: 1–3 compounds PES, conformer geometrical representations. QTAIM, ELF, NCI and DORI details and experimental and theoretical IR and NMR spectra. See DOI: 10.1039/c4ra16472e

---