How intramolecular hydrogen bonding (IHB) controls the C–ON bond homolysis in alkoxyamines

Abstract

Recent amazing results (Nkolo et al., Org. Biomol. Chem., 2017, 6167) on the effect of solvents and polarity on the C–ON bond homolysis rate constants k_d of alkoxyamine R₁R₂NOR₃ led us to re-investigate the antagonistic effect of intramolecular hydrogen-bonding (IHB) on k_d. Here, IHB is investigated both in the nitroxyl fragment R₁R₂NO and in the alkyl fragment R₃, as well as between fragments, that is, the donating group on the alkyl fragment and the accepting group on the nitroxyl fragment, and conversely. It appears that IHB between fragments (inter IHB) strikingly decreases the homolysis rate constant k_d, whereas IHB within the fragment (intra IHB) moderately increases k_d. For one alkoxyamine, the simultaneous occurrence of IHB within the nitroxyl fragment and between fragments is reported. The protonation effect is weaker in the presence than in the absence of IHB. A moderate solvent effect is also observed.

---

Introduction

Since the pioneering work of Rizzardo and co-workers,^1,2 alkoxyamines have been applied in several fields such as tin-free radical organic chemistry^3,4 – as initiators for radical cyclization,^5,6 1,2-radical additions,⁷ and several others⁵ – NMP (nitroxide mediated polymerization) and its variants: in situ NMP,⁸ ESCP,⁹ NMP2,^10,11 SL-NMP,¹² and CI-NMP,¹³ materials sciences – for self-healing polymers,¹⁴ optoelectronic materials,¹⁵ and encoding systems¹⁶ – and in biology,^17–19 as agents for theranostics. For alkoxyamines to be used as agents for theranostics (Fig. 1), the concept of a “smart” alkoxyamine was proposed,¹⁷ that is, a highly stable alkoxyamine switching to a highly labile alkoxyamine through chemical reactions (Fig. 2). For several years, our group has been promoting the chemical alkoxyamine activation using protonation,²⁰ oxidation,²⁰ alkylation,²⁰ or metal cation coordination.^21,22 However, in our view, activation processes based on physico-chemical events cannot be disregarded. Indeed, very recently,²³ a dramatic solvent effect on the C–ON bond homolysis rate constant k_d (Scheme 1) has been reported for 1 (Fig. 3). Intramolecular hydrogen-bonding (IHB) displays very different trends, that is, IHB within the nitroxyl fragment affords an increase in k_d (Fig. 4a)^24–27 and IHB from the alkyl fragment to the nitroxyl fragment (interR, Fig. 4c) affords a decrease in k_d spanning from weak (2–3 kJ mol⁻¹)²⁸ to moderate²⁹ (ca. 10 kJ mol⁻¹). Interestingly, other types of IHBs have not yet been investigated – IHB from the nitroxyl fragment to the alkyl fragment (interN, Fig. 4d) and IHB within the alkyl fragment (Fig. 4b) – and are the focus of this article. Several models 2–7 (Fig. 3) were prepared and their structures were determined by X-ray and NMR analysis.

Fig. 1 Concept for the application of alkoxyamines as agents for theranostics. Reproduced with permission of the American Chemical Society (see ref. 46).

Fig. 2 Concept for “smart” alkoxyamines. Reproduced with permission of the Royal Society of Chemistry (see ref. 17).

Scheme 1 C–ON bond homolysis in alkoxyamines.

Fig. 3 Alkoxyamines and nitroxides discussed in the article.

Fig. 4 Various types of IHBs: (a) within the nitroxyl fragment (intraN), (b) within the alkyl fragment (intraR), (c) from alkyl to nitroxyl fragments (interR), and (d) from nitroxyl to alkyl fragments (interN). Dotted blue lines represent IHB.

The occurrence and the type of IHB were determined by combining ³¹P and ¹H NMR with DFT calculations. The influence of IHB on k_d was investigated in tert-butylbenzene (t-BuPh) as a non-polar solvent and water as a polar/hydrogen bond acceptor (HBA) solvent, known to suppress IHB.³⁰

Results

Preparation of alkoxyamines 2–7

Alkoxyamine 3 was prepared either as previously reported²⁶ (black route in Scheme 2) or by the hydrolysis³¹ of 2 prepared from protected amino alcohol 2a (blue and magenta routes in Scheme 2). Alkoxyamine 2 was prepared either by the protection³² of 3 (black route in Scheme 2) or from the protected phosphorylated amino alcohol 2b (magenta route in Scheme 2).

Scheme 2 Preparation of 2 and 3: (a) (1) 2a or 3a (1.1 eq.), t-Bu-CHO (1 eq.), pentane, 65 °C, 2 day, (2) (EtO)₂P(O)H (1.1 eq.), 45 °C, 7 days, 64–78%; (b) 2b or 3b (1 eq.), Oxone® (4 eq.), Na₂CO₃ (6 eq.), EtOH/H₂O (3 : 1), 5 h, rt, 51–56%; (c) (1) Cu(0) (1.1 eq.), CuBr (0.55 eq.), PMDETA (0.55 eq.), argon bubbled benzene, 30 min, rt, (2) 1-bromoethylbenzene (1.1 eq.), 2˙ or 3˙ (1 eq.), rt, overnight, 65–72%; (d) 3a, 3b, or 3 (1 eq.), t-BuMe₂SiOTf (1.5 eq.), 2,6-lutidine (2 eq.), DCM, 4 h, rt 80–90%; (e) 2 (1 eq.) TBAF (1 M in THF) (1.2 eq.), THF, rt, 92%.

Alkoxyamine 4 was prepared using the conventional Mn(salen)₂ salt procedure (Scheme 3).³³ The two diastereoisomers were separated and RR/SS-4 ‡ was recrystallized. Then, it was protected using t-BuMe₂SiCl to afford 5 in good yield (Scheme 3).³⁴ §

Scheme 3 Preparation of 4 and 5: (a) (1) salen ligand (0.05 eq.), MnCl₂ (0.05 eq.), i-PrOH, 30 min, rt; (2) 3˙ (1 eq.), 2-vinyl pyridine (1 eq.), i-PrOH; (3) NaBH₄ (5 eq.), 4 h, rt, 54%; (b) (1) 4 (1eq.), pyridine (5 eq.), AgNO₃ (1.5 eq.), THF, 5 min, rt; (2) t-BuMe₂SiCl (1.5 eq.), overnight, rt, 59–67%.

Alkoxyamine 6 was prepared from the corresponding bromide 6b coupled to 6˙ using the conventional ATRA procedure.³⁵ However, 7 was obtained as a crude mixture of diastereoisomers which cannot be separated. Then, crude 7 was hydrolyzed into crude 6 whose diastereoisomers were separated, and the diastereoisomer RS/SR‡ was crystallized. The diastereoisomers of 6 were acetylated into 7 (Scheme 4).

Scheme 4 Preparation of 6 and 7: (a) 5a (1 eq.), NBS (1.1 eq.), benzoyl peroxide (0.1 eq.), CCl₄, reflux, 69%; (b) (1) CuBr (0.5 eq.), Cu(0) (1 eq.), PMEDTA (0.5 eq.), benzene, Ar, 10 min, rt; (2) 6˙ (1.1 eq.), 6b (1 eq.), benzene, 12 h, rt, (3) MeOH, K₂CO₃, H₂O, 3 days, rt, 63%; (c) (1) 6 (1 eq.), Et₃N (4 eq.), CH₂Cl₂; (2) Ac₂O (3 eq.), 3 days, rt, 65–90%.

X-ray analysis

Alkoxyamines RR/SS-4 and RS/SR-6 were crystallized and analyzed by XRD (Fig. 5).‡ X-ray structures exhibit the conventional distances and angles observed for such types of molecules (see the ESI†). Dihedral angles <O₅C₆C₇N₈> are 45° and 90° larger (Fig. 6d) than the 90° required at TS for RR/SS-4 and RS/SR-6, respectively.³⁶ However, from the X-ray structure (Fig. 5 and 6) it is clear that no IHB occurs in RR/SS-4 in the solid state although it is a good model to observe IHB of type (a) or (d) (Fig. 4) or both. By contrast, RS/SR-6, which is a good model for IHB of type (b) or (c) or both (Fig. 4), exhibits only IHB of type (c).¶ ^37,38 The absence of IHB in the RR/SS-4 X-ray structure is in sharp contrast with its occurrence in solution (vide infra). This difference in conformation between the crystal and solution has already been observed several times for alkoxyamines^27,28 and it is ascribed to the packing effect, which forces intermolecular H-bonding at the expense of IHB.

Fig. 5 X-ray structures for RR/SS-4 (left) and RS/SR-6 (right).‡

Fig. 6 Cram (a) and Newman projections through (b) N₄–O₅, (c) O₅–C₆, (d) C₆–C₇, and (e) C₃–N₄ bonds for RR/SS-4 (top) and RS/SR-6 (bottom).

NMR analysis

In contrast with XRD, ¹H and ³¹P NMR spectra show that IHBs of types (a) and (d) occur at the same time for RR/SS-4.|| As already reported for 2 and 3,²⁹ Δδ for 4 and 5 decreases from benzene-d₆ to DMSO-d₆ (Table 1 and Fig. 7), as expected from the increase in the parameter β³⁰ – the hydrogen bond acceptor (HBA) property of the solvent – that is, the increase in the ability to suppress IHB. The occurrence of IHB between the diethylphosphoryl and the hydroxyl groups (intraN in Fig. 8) is supported by the difference in shifts δ and the signal pattern for the MeCH₂O group observed between RR/SS-4 and RR/SS-5 in benzene-d₆ (protons labelled a in Fig. 9)** and its suppression in DMSO-d₆ (very similar patterns for protons labelled a–c for RR/SS-4 and RR/SS-5 in Fig. 9). However, the chemical shifts δ and the signal pattern of pyridyl protons (Fig. 9) are very different for RR/SS-4 and RR/SS-5 in benzene-d₆, while they are similar in DMSO-d₆. As similar observations were also reported upon protonation (or other types of activation) of the pyridyl moiety in alkoxyamines,^20–22 this change in the signal is ascribed to the occurrence of IHB between the hydroxyl group and the pyridyl moiety (interN in Fig. 8). Temperature dependence shows only line broadening when the temperature is decreased down to −80 °C, meaning that the rotation around the C–N bond of the group carrying the hydroxyl function is faster than the resolution time of ¹H NMR. DFT calculations (vide infra) led us to disregard the occurrence of 3-center IHB, as shown in Fig. 8.

Fig. 7 ³¹P NMR signals in various solvents for (a) a (2 : 1) mixture of RR/SS-4 (left) and RR/SS-5 (right), and (b) a (2 : 1) mixture of RS/SR-4 (left) and RS/SR-5 (right).

Fig. 8 Possible conformations affording intraN IHB (left), interN IHB (middle), and 3-center IHB (right) for RR/SS-4.

Fig. 9 ¹H NMR spectra of RR/SS-4 (top) and RR/SS-5 (bottom) in the range 2.5–6 ppm (top row) and 6–9 ppm (pyridyl proton zone, bottom row) in benzene-d₆ (left) and DMSO-d₆ (right). Labelling of protons: a–e represent MeCH₂O, CHP, CHMe, OH, and CH₂O, respectively.

Table 1 Solvent effect on diastereoisomers of 4 and 5 investigated by ³¹P NMR

Solvent^a	β	δ (ppm)		Δδ^d (ppm)	δ (ppm)		Δδ^d (ppm)
		RR/SS-4 ^c	RR/SS-5 ^c		RS/SR-4 ^c	RS/SR-5 ^c
a 85% H3PO4 was used as an internal reference (0 ppm). b Hydrogen bond acceptor parameter β. Given in ref. 30. c Ratio 4 : 5 = 2 : 1. d Δδ = δ4 − δ5.
CDCl3	0.10	27.29	26.11	1.18	26.13	25.22	0.91
C6D6	0.10	27.35	25.86	1.49	26.29	24.95	1.34
CD3CN	0.40	27.17	26.25	0.92	26.57	25.25	1.32
Acetone-d6	0.48	27.27	26.31	0.96	26.69	25.21	1.48
CD3OD	0.66	27.32	27.04	0.28	26.42	26.02	0.40
DMSO-d6	0.76	26.46	26.15	0.31	25.61	24.95	0.66

---

The same trends are observed for Δδ (Fig. 7 and Table 1) and ¹H NMR signals of EtO protons (labelled a in Fig. 1SI†) for RS/SR diastereoisomers of 4 and 5 as those for diastereoisomers RR/SS and are ascribed to the occurrence of IHB between the diethylphosphoryl and the hydroxyl groups (Fig. 10). In sharp contrast to what was observed for the diastereoisomer RR/SS, no significant differences in chemical shifts and signal patterns in the pyridyl proton zone are observed for RS/SR-4 and RS/SR-5, whatever the solvent (Fig. 1SI†). Consequently, only intraN IHB (type (a) in Fig. 4) is observed in RS/SR-4 (Fig. 10a).

Fig. 10 Possible conformations affording intraN IHB for RS/SR-4 (a) and intraR IHB for RR/SS-6 (b).

The same trends are observed for Δδ (Fig. 11b and Table 2) and ¹H NMR signals of EtO protons (labelled a in Fig. 2SI†) for RS/SR diastereoisomers of 6 and 7 as those for 4 and 5 (vide supra) and are ascribed to the occurrence of IHB between the diethylphosphoryl and the hydroxyl groups (Fig. 6) in RS/SR-6. On the other hand, the very similar signal patterns of protons in the aromatic zone for RS/SR-6 and RS/SR-7 (Fig. 2SI†) support the non-occurrence of intraR IHB (type b in Fig. 4 and 6).

Fig. 11 ³¹P NMR signals in various solvents for (a) a (2 : 1) mixture of RR/SS-6 (left) and RR/SS-7 (right), and (b) a (2 : 1) mixture of RS/SR-6 (left) and RS/SR-7 (right).

Table 2 Solvent effect on diastereoisomers of 6 and 7 investigated by ³¹P NMR

Solvent^a	β	δ (ppm)		Δδ^d (ppm)	δ (ppm)		Δδ^d (ppm)
		RR/SS-6 ^c	RR/SS-7 ^c		RS/SR-6 ^c	RS/SR-7 ^c
a 85% H3PO4 was used as an internal reference (0 ppm). b Hydrogen bond acceptor parameter β. Given in ref. 30. c Ratio 6 : 7 = 2 : 1. d Δδ = δ6 − δ7.
CDCl3	0.10	24.48	24.48	0.0	27.27	25.30	1.97
C6D6	0.10	24.52	24.32	0.20	26.90	24.82	2.08
CD3CN	0.40	24.64	24.47	0.17	27.25	25.19	2.06
Acetone-d6	0.48	24.63	24.34	0.29	27.65	25.62	2.03
CD3OD	0.66	25.51	25.27	0.24	27.65	26.05	1.60
DMSO-d6	0.76	24.45	24.08	0.37	26.97	25.30	1.67

---

No significant changes in Δδ (Table 2 and Fig. 11b) and in the signal patterns of EtO-groups (protons labelled a in Fig. 3SI†) are observed for RR/SS-6 and RR/SS-7, pointing to the absence of interR IHB (type c in Fig. 4). By contrast, significant changes in the signal patterns of aromatics are observed between RR/SS-6 and RR/SS-7 in benzene-d₆, whereas very similar signal patterns for RS/SR-6 and RS/SR-7 are observed (Fig. 12) in DMSO-d₆, i.e., the absence of any type of IHB, supporting the occurrence of intraR IHB in RR/SS-6 in benzene-d₆ (Fig. 10b).

Fig. 12 ¹H NMR spectra of RR/SS-6 (top) and RR/SS-7 (bottom) in the range 3–6 ppm (top row) and 6–9 ppm (pyridyl proton zone, bottom row) in CDCl₃ (left) and DMSO-d₆ (right). Labelling of protons: a–e represent MeCH₂O, CHP, CHMe, OH, and CH₂O, respectively.

Protonation of alkoxyamines

Alkoxyamines 4–7 carry a pyridyl moiety on the alkyl fragment suitable for activation by protonation. Consequently, protonation in benzene-d₆ (model for tert-butylbenzene t-BuPh) in the presence of trifluoroacetic acid is confirmed by changes in ¹H NMR shifts in the aromatic zone, e.g. RR/SS-4 (Fig. 4SI†). The values of pK_a (Table 3)³⁹ are estimated from the ¹H NMR shift of the signal in the aromatic zone, e.g., RR/SS-4 (Fig. 13), and are given by the modified Hasselbach–Henderson equation (eqn (1), exemplified by 4/4H+, and Table 3).^40,41

Fig. 13 Titration curves for RR/SS-4 (0.01 M, ■, inset: signal of the aromatic proton zone) and RS/SR-4 (0.01 M, ●), D₂O/MeOH-d₄ = 1 : 1.

Table 3 Values of pK_a for 4–7

	4		5		6		7
	RR/SS	RS/SR	RR/SS	RS/SR	RR/SS	RS/SR	RR/SS	RS/SR
a All pH values measured in D2O/MeOH-d4 (1 : 1) were re-estimated using pH = 0.929·pH* + 0.42. pH* is the pH measured in D2O/MeOH-d4 solutions using a pH-meter calibrated with non-deuterated water. See ref. 39. b pKa = 5.89 for ortho-ethylpyridine. See ref 40. c pKa = 4.85 for ortho-2-hydroxyethylpyridine. See the ESI. d pKa values of 4.21 (RS/SR) and 3.99 (RR/SS) were reported for 12. See ref. 41.
pKa ^a^,^b	4.08^c	4.32^c	3.81^c	4.25^c	4.25^c^,^d	3.61^c^,^d	3.24^c^,^d	3.33^c^,^d

---

As already reported, the pK_a of 12 is ca. 1.7–1.9 units lower than that reported for the ortho-ethylpyridine. It is ascribed to the presence of the nitroxyl fragment as a strong electron withdrawing group (EWG). The one unit lower pK_a for the ortho-hydroxyethylpyridine is also ascribed to the presence of the OH group as an EWG. Therefore, the lower pK_a values for 4–7 (Fig. 13) than those for ortho-ethylpyridine and ortho-hydroxyethylpyridine are due to the presence of the nitroxyl fragment as an EWG. It is noteworthy that alkoxyamines 4–6 exhibit pK_a values very close to those for model 12, except for RR/SS-6.

The one unit lower pK_a value for 7 than that for 12 is ascribed to the acylated hydroxyl group, which exhibits a higher electron withdrawing property than the OH group, the basicity of the pyridine moiety being thus decreased.

Kinetic investigations

Homolysis rate constants k_d,T were measured using either the ³¹P NMR method²⁷ for 4, 4H+, 5, 7 and 7H+ or the EPR method²⁰ for 2, 2′, 2′′, 3′, 6 and 6H+, as previously reported. Differences in E_a observed between diastereoisomers for 2, 2′, 2′′, 3, 5 and 7 in t-BuPh as a solvent are in the range of 0–2 kJ mol⁻¹ – generally observed for diastereoisomers reported in the literature – and do not deserve more comments.³⁶ Alkoxyamines exhibiting IHB are discussed later. Except for 6, comments made for diastereoisomers in t-BuPh as a solvent hold in water, as IHB is suppressed.

DFT calculations

DFT calculations at the M062X/6-31+G(d,p) level of theory in the gas phase were performed to determine both the thermodynamics of the homolysis and the most stable conformations for the diastereoisomers of 2–4 and 6 (Table 1SI†).⁴² Important geometrical parameters – bond lengths, distances and angles – are reported in Table 1SI† and agree with X-ray data and those reported for molecules of the same family, except for the occurrence of IHB in RR/SS-4. Thermodynamics of 4 and 6 show the same trends as that experimentally reported (Table 1SI†). Only the 3 most stable conformations A, B and C †† are reported for 4 and 6 (Fig. 14). Valence angles, bond lengths and distances for IHB are reported in Table 5. Depending both on the valence angle α and on the distance d_H⋯X between the H and X atoms implied in the H-bonding, IHB is given as strong (α > 150° and d smaller than the sum of van der Waals radii of H and X atoms), as weak (α < 120° and d closer to the sum of the van der Waals radii of H and X atoms) or as medium for other combinations.¶ ^37,38 On these grounds, IHB for the most stable conformers is considered as strong for RS/SR-6 (conformer A) and RS/SR-9, and as medium for other diastereoisomers (Fig. 14 and Table 5). For all cases, IHB is smaller by ca. 1.8 Å and by 0.5 Å for OH⋯O=P and for OH⋯N interactions, respectively, than the respective sums of van der Waals radii¶ – d_OH⋯OP = 2.61 Å and d_OH⋯N = 2.64 Å – and all valence angles α are larger than 130°. Nevertheless, the occurrence of IHB is controlled mainly by the steric hindrance. As illustrated by RR/SS-9, when steric hindrance is too large, IHB does not occur, and in some cases, for example, RR/SS-4 and RR/SS-6, the most stable conformer B does not display the strongest IHB.‡‡ All other possible IHBs are disregarded due to distances d_OH⋯OP and d_OH⋯N much larger than the sums of the van der Waals radii, except for OH⋯NO for the conformer C of RS/SR-6. Nonetheless, the IHB observed between the hydroxyl group and the N atom of the nitroxyl moiety in the conformer C is not strong enough to balance the steric strain in C, which is less stable than A by 19 kJ mol⁻¹ (Fig. 14).

Fig. 14 Most stable conformations A–C†† for RR/SS-4, RS/SR-4, RR/SS-6, and RS/SR-6. All conformers are on the same energetic scale (kJ mol⁻¹). Dotted lines for the difference in energy between isomers and full lines for the difference in energy between conformers.

Discussion

Free motions in the nitroxyl fragment

Taking into account (i) NMR observations for RR/SS-4 denoting the simultaneous occurrence of intraN and interN IHBs (Fig. 4), (ii) calculations ascribing these two IHBs to two different conformers (Fig. 14 and Table 5), (iii) dramatically restricted bond rotations in the group carrying the diethylphosphoryl group attached to the nitroxyl moiety in nitroxides, and in the subsequent alkoxyamines,§§ ^43–45 the question of the free rotation around the C–N bond for the other alkyl group attached to the nitroxyl moiety is raised. Indeed, the conformations of the nitroxyl fragment are expected to be ruled by the conformations of the nitroxide, as highlighted by the occurrence of the same IHB in 3˙ ²⁶ and 3.²⁷ Whatever the route for the preparation of 2 and 3 (Scheme 2 and ESI†), ¹H NMR spectra of intermediates are the same (see the ESI†) as well as k_d values within the experimental error (Table 4), supporting that for the group carrying the hydroxyl function, the rotation of the C–N bond is free, whereas the rotation of the C–N bond for the moiety carrying the diethylphosphoryl group is, in general, strongly restricted in the nitroxide,^46,47 and likely in the alkoxyamine.

Table 4 C–ON bond homolysis rate constant k_d,T at various temperatures T for alkoxyamines under various conditions of solvents and pH and their corresponding activation energies E_a, the homolysis rate constant k_d re-estimated at 120 °C and predicted activation energies E′_a

	Solvent	T (°C)	RR/SS			RS/SR			E′a^a^,^b		References^j
			k d,T ^c^,^d	E a ^b^,^e	k d ^f^,^g	k d,T ^c^,^d	E a ^b^,^h	k d ^f^,^i	RR/SS	RS/SR
a Predicted values of Ea using the incremental scale. b In kJ mol^−1. c In 10^−4 s^−1. d Given by eqn (4). e Estimated using kd values in the 4^th column and the frequency factor A = 2.4 × 10^14 s^−1 in eqn (5). See ref. 51. f In 10^−3 s^−1. g Estimated using Ea values in the 5^th column and the frequency factor A = 2.4 × 10^14 s^−1 in eqn (5). See ref. 51. h Estimated using kd values in the 7^th column and the frequency factor A = 2.4 × 10^14 s^−1 in eqn (5). See ref. 51. i Estimated using Ea values in the 8^th column and the frequency factor A = 2.4 × 10^14 s^−1 in eqn (5). See ref. 51. j t.w.: this work. n.d.: not determined. n.m.: not measured. k Prepared via the route 3a → 3b → 3˙ → 3 → 2. l Measured by EPR. m Used as a model: Δ11→2 = +0.4 kJ mol^−1 for the diastereoisomer RR/SS and Δ11→2 = +0.9 for the diastereoisomer RS/SR. n Prepared via the route 3a → 2a → 2b → 2˙ → 2′(2) → 3′ (3). o Prepared via route 3a → 3b → 2b → 2˙ → 2′′ (2). p Used as a model: Δ11→3 = −1.5 for RR/SS and Δ11→3 = −1.2 for RS/SR. q k d determined by ^31P NMR. See ref. 20. r E′a = Ea,11 + Δ11→3 + Δ11→12. s Protonation is performed by adding 2 eq. of TFA. t Protonation expected to decrease Ea by 7.9 kJ mol^−1, as highlighted by the protonation of 12 in t-BuPh. See ref. 41. u E′a = Ea,11 + Δ11→2 + Δ11→12. v E′a = Ea,11 + Δ11→9 + Δ11→12. w E′a = Ea,11 + Δ11→10 + Δ11→10. x Used as a model: Δ11→9 = −0.6 for RR/SS and Δ11→9 = +6.9 for RS/SR. y Used as a model: Δ11→10 = −1.0 for RR/SS and Δ11→10 = +0.9 for RS/SR. z Used as a reference. aa Used as a model: Δ11→12 = 0 for RR/SS and Δ11→12 = +1.0 for RS/SR. ab H2O : MeOH (1 : 1) is used as a solvent. ac Used as a model: Δ11→3 = +1.1 for RR/SS and Δ11→3 = +0.2 for RS/SR. ad Water as a solvent. ae E′a = Ea,11 + Δ11→3 + Δ11→12. af Protonation expected to decrease Ea by 13.5 kJ mol^−1, as highlighted by the protonation of 12 in a H2O/MeOH mixture. See ref. 41. ag E′a = Ea,11 + Δ11→9 + Δ11→12. ah Used as a model: Δ11→9 = −1.8 for RR/SS and Δ11→9 = +1.4 for RS/SR. ai Used as a model: Δ11→12 = −2.0 for RR/SS and Δ11→12 = −1.0 for RS/SR.
2 ^k^,^l	t-BuPh	90	3.0	124.4	7.1	3.5	123.9	8.2	—^m	—^m	t.w.
2′ ^l^,^n	t-BuPh	100	9.4	124.3	7.3	10.5	124.0	8.0	n.d.	n.d.	t.w.
2′′ ^l^,^o	t-BuPh	90	3.2	124.2	7.6	3.5	124.0	8.0	n.d.	n.d.	t.w.
3 ^l	t-BuPh	n.m.	n.m.	122.5	12.5	n.m.	121.8	15.5	—^p	—^p	29
3′ ^l^,^n	t-BuPh	100	14.7	122.9	11.2	16.7	122.5	12.7	n.d.	n.d.	t.w.
4 ^q	t-BuPh	80	4.1	120.1	26.1	2.4	121.7	16.0	122.5^r	122.8^r	t.w.
4H+ ^q	t-BuPh^s	50	0.8	114.3	154.1	0.6	115.3	113.5	112.2^t	113.8^t	t.w.
5 ^q	t-BuPh	100	110.0	123.7	8.7	220.0	121.7	16.0	124.4^u	124.9^u	t.w.
5H+ ^q	t-BuPh	50	1.3	113.1	225.4	1.0	113.7	187.6	115.8^t	113.8^t	t.w.
6 ^l	t-BuPh	100	23.7	121.5	17.0	2.9	128.0	2.3	123.4^v	128.9^v	t.w.
6H+ ^l	t-BuPh^s	83	11.8	117.9	51.2	1.2	124.8	6.2	113.6^t	120.1^t	t.w.
7 ^q	t-BuPh	83	2.8	122.2	13.7	4.4	120.9	20.4	123.0^w	124.9^w	t.w.
7H+ ^q	t-BuPh^s	61	0.5	119.2	34.9	0.8	118.3	45.9	114.1^t	117.0^t	t.w.
8	t-BuPh	n.m.	n.m.	123.2	10.1	n.m.	122.7	11.8	n.d.	n.d.	27
9	t-BuPh	n.m.	n.m.	123.4	9.5	n.m.	129.9	1.3	—^x	—^x	29
10	t-BuPh	n.m.	n.m.	123.0	10.7	n.m.	123.9	8.1	—^y	—^y	29
11	t-BuPh	n.m.	n.m.	124.0	7.9	n.m.	123.0	10.7	—^z	—^z	29
12	t-BuPh	n.m.	n.m.	124.1	7.8	n.m.	123.8	8.4	—^aa	—^aa	41
12H+	t-BuPh	n.m.	n.m.	116.0		n.m.	116.2		n.d.	n.d.	41
3	—^ab	n.m.	n.m.	125.1	5.7	n.m.	122.8	11.4	—^ac	—^ac	27
4 ^q	pH = 7.0^ad	80	8.3	118.0	49.6	12.8	116.8	71.7	121.3^ae	123.6^ae	t.w.
4H+ ^q	pH = 1.4^ad	50	4.6	109.6	649.3	4.9	109.4	690.3	104.5^af	103.3^af	t.w.
6 ^l	pH = 7.0^ad	93	14.9	120.3	24.6	5.0	124.1	7.7	118.4^ag	124.8^ag	t.w.
6H+ ^l	pH = 1.4^ad	72	14.6	113.8	179.6	12.8	114.2	158.9	106.8^af	110.6^af	t.w.
8	—^ab	n.m.	n.m.	123.8	8.4	n.m.	123.2	10.1	n.d.	n.d.	29
9	—^ab	n.m.	n.m.	122.2	13.7	n.m.	124.4	7.0	—^ah	—^ah	29
11	—^ab	n.m.	n.m.	123.0	10.7	n.m.	124.0	7.9	—^z	—^z	29
12	—^ab	n.m.	n.m.	121.4	17.5	n.m.	122.5	12.5	—^ai	—^ai	41
12H+	—^ab	n.m.	n.m.	108.0		n.m.	109.0		n.d.	n.d.	41

---

IHB in alkoxyamines

A quick glance at 3, 4, 6 and 9 shows that several types of IHBs may occur (Fig. 4). The occurrence of interR IHB for RS/SR-9 has been reported in the literature.²⁹ It has been observed by X-ray,²⁹ NMR,²⁹ and IR²⁹ and is supported by DFT calculations and does not deserve more comments, as RS/SR-9 is used as a model. Diastereoisomer RR/SS-9 does not exhibit IHB.²⁹ The occurrence of intraN IHB for 3 was recently studied by NMR¶¶ ²⁷ and supported by DFT calculations in this work (see the ESI† and Table 5) and does not deserve more comments. X-ray data, NMR analysis, and DFT calculations support IHB of type interR, intraR, and intraN (Fig. 4, 8, 10 and 14) for RS/SR-6, RR/SS-6, and RS/SR-4, respectively (see the ESI†). In contrast with diastereoisomer RS/SR-6, no IHB is observed in diastereoisomer RR/SS-4 by X-ray analysis, although ³¹P NMR supports the occurrence of intraN IHB, ¹H NMR supports the occurrence of interN IHB. The appealing 3-center IHB displayed in Fig. 8 is disregarded by DFT calculations (too high energy conformer).|||| On the other hand, DFT calculations show a small difference in energy of 2.5 kJ mol⁻¹ between the stable conformer B and conformer A, more stable than C by 16 kJ mol⁻¹, meaning that a fast equilibrium between B and A likely accounts for the simultaneous occurrence of intraN and interN IHBs observed by NMR. This fast equilibrium requires fast bond rotation around the C–N bond of the group carrying the hydroxy function. This fast rotation is nicely supported both by the temperature dependence of RR/SS-4 (not shown) and the kinetics and NMR observations reported for 2 and 3 (vide supra). The small difference in energy of 8 kJ mol⁻¹ between diastereoisomers of 4 agrees with the small difference of 1.6 kJ mol⁻¹ in E_a.

Table 5 Valence angles α <OHO_P> and <OHN>, and distances d_OH⋯OP and d_OH⋯N for conformers A,^aB,^b and C ^c of diastereoisomers of 4 and 6

Alkoxyamine	Conformer	α (°)		Distances^d (Å)		Δ IHB ^e
		<OHOP>	<OHN>	d OH⋯OP	d OH⋯N
a Conformer exhibiting IHB between HO and P=O functions. b Conformer exhibiting IHB between OH and N functions. c The most stable conformer with no IHB. d Bold values denote the most stable conformers. e In kJ mol^−1. n.e. not eligible. f Δ IHB = Ea,3 − Ea,8. g Not given because of too long dH⋯X forbidding any IHB. See ref. 38. h Δ IHB = Ea,4 − Ea,5. See the ESI. i Δ IHB = Ea,6 − Ea,7. See the ESI. j Not available. k A medium strength IHB between the HO group and the N-atom of the nitroxyl moiety is observed: dOH⋯N = 1.99 Å and <OHN> = 140°. l Δ IHB = Ea,9 − Ea,10. See the ESI.
RR/SS-3	A	148	—^j	1.83	—^e	−0.7^f
RS/SR-3	A	150	—^j	1.83	—^e	−0.9^f
RR/SS-4	A	148	—^g	1.83	6.55	−3.6^h
	B	—^g	149	6.25	2.06	n.e.
	C	—^g	—^g	6.79	6.87	n.e.
RS/SR-4	A	146	—^g	1.85	6.61	0^h
	B	—^g	161	4.78	2.13	n.e.
	C	—^g	—^g	6.68	8.05	n.e.
RR/SS-6	A	170	—^g	1.82	5.09	n.e.
	B	—^g	136	5.16	2.11	−0.7^i
	C	—^g	—^g	6.65	4.82	n.e.
RS/SR-6	A	173	—^g	1.79	4.56	+7.1^i
	B	—^g	136	4.99	2.03	n.e.
	C	—^g	—^j^,^k	4.91	4.84^k	n.e.
RR/SS-9	C	—^g	—^j	6.00	—^j	n.e.
RS/SR-9	A	174	—^j	1.79	—^j	+6.0^l
RR/SS-4	X-ray	—^g	—^g	6.63	6.91	n.e.
RS/SR-6	X-ray	164	—^g	1.96	4.29	n.e.

---

Influence of IHB on k_d

Several linear multiparameter relationships developed over the last two decades to investigate various effects involved in the changes in k_d ^25,36 cannot account for the effect of IHB. In parallel with these relationships, the group additivity approach⁴⁸ provides E_a values predicted with a good accuracy using a scale developed a decade ago.⁴⁹ To use the group additivity approach, alkoxyamine 11 was selected as a benchmark, as it displays the nitroxyl fragment with the lowest functionalization in that series and a secondary aromatic alkyl fragment. Thus, estimated activation energies E′_a were determined (Table 4 and ESI†). Taking into account the small difference in polarity between the alkyl fragment carrying a pyridyl (σ_I = 0.06)^50,51 or a phenyl (σ_I = 0.07)⁵² moiety, no significant difference was reported,⁴¹ and the presence of the hydroxyl group capable of intraR IHB does not provide more than a 2-fold increase in k_d, as highlighted by the couples 6/9 and 4/3. To unveil the effect of IHB on k_d, the hydroxyl group was protected by silylation in 2 and 5, by methylation in 8, and by acylation in 7 and 10, to suppress the occurrence of IHB (see the ESI†). Kinetic results showed that the 4 types of IHBs can be gathered into two main families: alkyl and nitroxyl fragments exhibiting intra IHB which show no significant differences (less than a factor 2) between the alkoxyamines carrying the free hydroxyl group (3, RS/SR-4, and RR/SS-6) and those carrying the protected hydroxyl group (2, RS/SR-5, and RR/SS-7) and inter IHB between alkyl and nitroxyl fragments which shows a clear 6–9-fold decrease in k_d between the alkoxyamines carrying the free hydroxyl group (diastereoisomers RS/SR of 6 and 9) and those carrying the protected hydroxyl group (diastereoisomers RS/SR of 7 and 10). Deeper insights into the geometrical parameters – the IHB valence angle α <OHX> and the IHB distance d_OH⋯X – ruling the strength of the IHB and its influence on k_d given by Δ_IHB (see Table 5) are unveiled by the 3D plot f(Δ_IHB,d,α) in Fig. 15. Indeed, intra IHB exhibits a very weak effect (−1 < Δ_IHB < 0 kJ mol⁻¹ except for intraN in RR/SS-4) associated with short H-bonds (d_OH⋯X < 1.9 Å) and closed angles α (α < 150°) denoting medium strength IHB,*** whereas interR IHB exhibits a strong effect (Δ_IHB > 6 kJ mol⁻¹) associated with short H-bonds (d_OH⋯X ≈ 1.8 Å) and open angles α (α ≈ 175°) denoting strong IHB. As mentioned above, diastereoisomer RR/SS-4 simultaneously displays intraN and interN IHBs and, thus, the decay of RR/SS-4 is described in Scheme 5 with K₁ = k₁/k₋₁ for the equilibrium constant between interN and intraN conformers, and k₂ and k₃ the rate constants for the homolysis of interN and intraN conformers, respectively. Taking into account that interN IHB makes a bond between nitroxyl and alkyl fragments as interR IHB does, the same effect is expected, i.e., an increase in E_a in regard to the homologue with no IHB. Therefore, k₃ must be smaller than k₂ which corresponds to intraN IHB noted to increase k_d (vide supra). Hence, disregarding k₃, and assuming both a fast equilibrium – supported by very close calculated energy levels (vide supra) – between conformers B (interN) and A (intraN) and the pre-equilibrium assumption (k₋₁ ⋙ k₂ and K₁ = k₁/k₋₁), k_d (considered as the apparent rate) is given as k_d = k₁k₂/(k₋₁ + k₂) (eqn (2) or as k_d = K₁k₂ as in eqn (3), Scheme 5). As the decrease in E_a is in the expected range (see Table 4), the values of k_d should be very close to k₂, meaning that K₁ ≈ 1 is in good agreement with the value estimated by DFT calculations, i.e., K₁ = 0.42.†††

Fig. 15 A 3D plot highlighting the relationship between the energetic effect of IHB Δ_IHB (kJ mol⁻¹), IHB valence angle α (°), IHB distances d_OH⋯X (Å) for the diastereoisomers of 3, 4, 6 and 9. Black and red symbols are intra and inter fragment IHB, respectively (see Fig. 4). Blue dots are for projection on the XY plane.

Scheme 5 Equilibrium between interN and intraN IHB and the subsequent homolysis.

Solvent effect on k_d

The solvent effect in alkoxyamines does not increase too much interest despite the very amazing results recently reported, such as a striking 1500-fold increase in k_d from t-BuPh to water as solvents or a slight but clear 5–20-fold increase in k_d for homologue alkoxyamines of 12.^53,54 Solvent effects for 3,²⁷9,²⁹ and 12 ⁴¹ have been previously reported. Alkoxyamines 4, 6, 9 and 12 show a small ca. 1.5–3-fold increase in k_d, as expected from the literature,^53,54 except for RS/SR-4, RS/SR-6 and RS/SR-9. The 3-fold and 5-fold increases in k_d for diastereoisomers RS/SR of 6 and 9 are ascribed to the suppression of IHB from t-BuPh to water/MeOH as solvents. The slight 4.5-fold increase in k_d observed for RS/SR-4, from t-BuPh to water/MeOH as solvents, might be ascribed to the suppression of stabilizing IHB which balances the steric strain due to the configuration. The unexpected slight ca. 2-fold decrease in k_d for 3 is ascribed to a better solvation of the alkoxyamine, and hence a better stabilization, and higher E_a from t-BuPh to water/MeOH as solvents than for the other alkoxyamines investigated.

For protonated alkoxyamines 4H+, 6H+ and 12H+, a 3–12-fold increase in k_d is observed and is ascribed to intimate ion pair dissociation and solvation effects,⁵⁵ except for RS/SR-6, for which a 25-fold increase in k_d is observed due to the combination of IHB suppression and intimate ion pair dissociation.

Effect of protonation on k_d

A ca. 10-fold and 60-fold increase in k_d for 12 in t-BuPh and water/MeOH as solvents was reported upon protonation. Amazingly, the protonation effect for 4, i.e., a ca. 6-fold and 12-fold increase in t-BuPh and water/MeOH as solvents, respectively, and that for 6, i.e., a 3-fold and ca. 8-fold increase (for the RR/SS diastereoisomer) in t-BuPh and water/MeOH as solvents, respectively, are less strong than that for 12, although 6 and 4 exhibit structures very similar to that of 12! Suppressing IHB in water/MeOH for RS/SR-6H+ affords a 20-fold increase in k_d upon protonation. This surprising lower effect of protonation for 4 and 6 is ascribed to the better solvation of the alkoxyamine for 4 and 6 than that for 12, due to the presence of the hydroxyl group affording the better stabilization of the starting alkoxyamine, which partly balances the polar effect due to protonation.

Experimental section

Alkoxyamines 3 were prepared as previously reported,²⁶ with the modification of the first step. All solvents and reactants for the preparation of alkoxyamines were used as received. Routine reaction monitoring was performed using silica gel 60 F₂₅₄ TLC plates; spots were visualized upon exposure to UV light and a phosphomolybdic acid solution in EtOH as a stain revealed by heating. Purifications were performed on a Reveleris® X2 flash chromatography system (BUCHI, Switzerland); a solvent delivery system: high pressure HPLC pumps; pump flow rate: 1–200 mL min⁻¹; maximum pressure: 200 psi; gradients: linear (see Fig. 5SI† for the profile); UV wavelength range: 200–500 nm; flash Reveleris® and GraceResolv™ cartridges: silica 40 μm, silica weight (g): 4, 12, 24, 48, 80 and 120. ¹H, ¹³C, and ³¹P NMR spectra were recorded in CDCl₃ on a 300 or 400 MHz spectrometer. Chemical shifts (δ) in ppm were reported using residual non-deuterated solvents as the internal reference for ¹H and ¹³C-NMR spectra, and as an internal capillary filled with 85% H₃PO₄ for ³¹P-NMR spectra. High-resolution mass spectra (HRMS) were recorded on a SYNAPT G2 HDMS (Waters) spectrometer equipped with a pneumatically assisted atmospheric pressure ionization source (API). Positive mode electrospray ionization was used on samples: electrospray voltage (ISV): 2800 V; opening voltage (OR): 20 V; nebulizer gas pressure (nitrogen): 800 L h⁻¹. Low resolution mass spectra were recorded on an ion trap AB SCIEX 3200 QTRAP instrument equipped with an electrospray source. The parent ion [M + H]⁺ is quoted.

Diethyl (1-((1-hydroxy-2-methylpropan-2-yl)amino)-2,2-dimethylpropyl)phosphonate (3b)

Under N₂, 2-methyl-2-aminopropanol 3a (6.8 g, 1.1 eq., 76.7 mmol) was added dropwise to a solution of pivalaldehyde (6.0 g, 1 eq., 69.7 mmol) in pentane (50 mL). The solution was heated at 65 °C with a Dean–Stark device for 2 days. Then, pentane was evaporated and diethylphosphite (10.6 g, 1.1 eq., 76.7 mmol) was added at r.t., and the mixture was heated at 45 °C for seven days. The solution was acidified with 1 M HCl (50 mL), and washed with dichloromethane (2 × 30 ml). The aqueous layer was basified with NaHCO₃ and then extracted with dichloromethane (2 × 20 ml), the organic layer was dried (MgSO₄), and the solvent was evaporated to yield aminophosphonate 3b (16.1 g, 78%).²⁶

Diethyl (1-((1-((tert-butyldimethylsilyl)oxy)-2-methylpropan-2-yl)amino)-2,2-dimethylpropyl) phosphonate (2b)

The same procedure as that for 3b was applied to 2a. 2a (3 g, 1.1 eq., 16.23 mmol), pivalaldehyde (1.27 g, 1 eq., 14.75 mmol), pentane (20 mL) and diethylphosphite (2.2 g, 1.1 eq., 16.23 mmol). After flash-chromatography (gradient petroleum ether (PE)/AcOEt: 0% to 100% of AcOEt), aminophosphonate 2b was isolated (3.8 g, 64%, colourless oil). ¹H NMR (300 MHz, CDCl₃) δ: 4.09 (qd, J = 7.1, 1.3 Hz, 4H), 3.32 (s, 2H), 2.81 (d, J_H–P = 17.0 Hz, 1H), 1.68 (s, 1H), 1.30 (t, J = 7.1 Hz, 6H), 1.06 (s, 3H), 1.04 (s, 9H), 0.99 (s, 3H), 0.90 (s, 9H), 0.03 (s, 6H). ¹³C NMR (75 MHz, CDCl₃) δ: 72.8 (d, J_C–P = 1.8 Hz), 61.4 (d, J_C–P = 7.4 Hz), 61.0 (d, J_C–P = 7.6 Hz), 58.9 (d, J_C–P = 138.7 Hz), 54.2 (d, J_C–P = 1.6 Hz), 35.2, 35.1, 27.9 (d, J_C–P = 6.2 Hz, 2C), 25.9 (3C), 25.1 (d, J_C–P = 1.5 Hz), 24.2, 18.2, 16.5 (d, J_C–P = 4.6 Hz), 16.4 (d, J_C–P = 5.0 Hz), −5.5 (2C). ³¹P NMR (162 MHz, CDCl₃) δ: 30.09. HRMS m/z (ESI) calcd for C₁₉H₄₅NO₄PSi [M + H]⁺ 410.2850, found: 410.2852.

(1-(Diethoxyphosphoryl)-2,2-dimethylpropyl)-(2-((tert-butyldimethylsilyl)oxy)-1,1-dimethylethyl)amino-N-oxyl radical (2˙)

Aminophosphonate 2b (1.0 g, 1 eq., 2.4 mmol) was dissolved in a mixture of EtOH/H₂O (3 : 1) (15 mL) and Na₂CO₃ (1.55 g, 6 eq., 14.6 mmol). Then, Oxone® (3.0 g, 4 eq., 9.8 mmol) was added in small portions at r.t. within 2 h under vigorous stirring. After completion, Et₂O (10 ml) was added, and the precipitate was filtered off. Et₂O was removed under vacuum. The aqueous phase was extracted with dichloromethane (2 × 15 ml), and dried over MgSO₄ and the solvents were evaporated under reduced pressure. The crude was purified by flash chromatography (gradient PE/AcOEt: 0% to 100% of AcOEt) and nitroxide 2˙ was isolated (529 mg, 51%). HRMS m/z (ESI) calcd for C₁₉H₄₄NO₅PSi· [M + H]⁺ 425.2721, found: 425.2720.

Diethyl (1-((1-((tert-butyldimethylsilyl)oxy)-2-methylpropan-2-yl)(1-phenylethoxy)amino)-2,2-dimethylpropyl)phosphonate (2)

To a suspension of CuBr (55 mg, 0.55 eq., 0.388 mmol) and Cu powder (49 mg, 1.1 eq., 0.78 mmol) in degassed benzene (argon bubbling for one hour) (3 mL) was added N,N,N′,N′,N′′-pentamethyldiethylenetriamine (80 μL, 0.55 eq., 0.39 mmol). After stirring for 10 min, a solution of nitroxide 2˙ (300 mg, 1 eq., 0.71 mmol) and (1-bromoethyl)benzene (103 μL, 1.1 eq., 0.78 mmol) in degassed benzene (3 mL) was transferred to the first solution. The mixture was allowed to stir for 12 h. The solution was diluted with Et₂O, quenched with a NH₄Cl sat. solution, washed with water and brine, and dried over MgSO₄. The solvents were evaporated under reduced pressure to give the crude product as a 1 : 1 mixture of diastereoisomers (³¹P-NMR ratio). The crude product was purified by automatic flash-chromatography (gradient PE/AcOEt: 0% to 100% of AcOEt) to yield alkoxyamines RR/SS-2 (pale yellow oil, 119 mg, 32%) and RS/SR-2 (pale yellow oil 123 mg, 33%). RS/SR-2¹H NMR (300 MHz, CDCl₃) δ: 7.37 (d, J = 7.3 Hz, 2H), 7.17 (m, 3H), 5.18 (q, J = 6.4 Hz, 1H), 3.86 (m, 2H), 3.69 (d, J_H–P = 26.6 Hz, 1H), 3.58 (s, 2H), 3.38 (m, 1H), 3.21 (m, 1H), 1.48 (d, J = 6.5 Hz, 3H), 1.15 (m, 18H), 0.86 (m, 12H), 0.00 (s, 6H). ¹³C NMR (75 MHz, CDCl₃) δ: 143.2, 127.9 (2C), 127.7 (2C), 127.3, 78.5, 70.0 (d, J_C–P = 139.2 Hz), 69.0, 65.1, 61.6 (d, J_C–P = 6.4 Hz), 58.5 (d, J_C–P = 7.5 Hz), 35.4 (d, J_C–P = 5.3 Hz), 30.5 (d, J_C–P = 6.0 Hz, 2C), 25.9 (3C), 23.4, 22.6, 21.2, 18.2, 16.3 (d, J_C–P = 5.6 Hz), 16.1 (d, J_C–P = 6.9 Hz), −5.4, −5.5. ¹P NMR (121 MHz, CDCl₃) δ: 24.98. RR/SS-2¹H NMR (300 MHz, CDCl₃) δ: 7.45 (m, 5H), 5.13 (d, J = 6.8 Hz, 1H), 4.53 (m, 1H), 4.22 (m, 3H), 3.86 (d, J_H–P = 26.5 Hz, 1H), 3.24 (q, J = 9.9 Hz, 2H), 1.74 (d, J = 6.7 Hz, 3H), 1.49 (m, 6H), 1.40 (s, 9H), 1.06 (s, 3H), 1.03 (s, 3H), 0.98 (s, 9H), 0.00 (s, 6H). ¹³C NMR (75 MHz, CDCl₃) δ: 145.5, 128.2 (2C), 127.1, 126.8 (2C), 85.4, 69.7 (d, J = 138.1 Hz), 69.6 (d, J = 1.6 Hz), 65.3, 61.6 (d, J = 6.3 Hz), 58.8 (d, J = 7.5 Hz), 35.7 (d, J = 6.1 Hz), 29.9 (d, J = 5.8 Hz, 2C), 25.8 (3C), 24.5, 23.7, 22.1, 18.0, 16.8 (d, J = 5.4 Hz), 16.2 (d, J = 6.7 Hz), −5.5, −5.6. ³¹P NMR (121 MHz, CDCl₃) δ: 26.19. HRMS for RR/SS-2 and RS/SR-2 mixture m/z (ESI) calcd for C₂₇H₅₃NO₅PSi [M + H]⁺ 530.3425, found: 530.3425.

General procedure for TBS protection of alcohols 3a, 3b and 3 ^32,56

In a flask under N₂, alcohols 3a, 3b or 3 (1.0 eq.), tert-butyldimethylsilyl trifluoromethanesulfonate (1.5 eq.) and 2,6-lutidine (2 eq.) were stirred in dichloromethane at room temperature for 4 hours. A saturated NaHCO₃ solution was added and the layers were separated. The organic layer was dried over Na₂SO₄, filtered, and concentrated in vacuo. Purification was performed using flash chromatography, affording 2a, 2b, or 2 as an oil.

Hydrolysis³¹ of silylated compound 2

In a flask under N₂, RR/SS-2 or RS/SR-2 (30 mg, 1 eq., 0.06 mmol) was dissolved in THF (2 mL), and tetra-n-butylammonium fluoride (1 M in THF) was added (68 μL, 1.2 eq., 0.07 mmol). After stirring for 1 hour at room temperature, THF was evaporated under reduced pressure and the crude was purified by automatic flash-chromatography (gradient PE/AcOEt: 0% to 100% of AcOEt) to yield alkoxyamine RR/SS-3′ or RS/SR-3′ with a yield higher than 90%.

Diethyl-(1-((1-hydroxy-2-methylpropan-2-yl)(1-(pyridin-2-yl)ethoxy)amino)-2,2-dimethylpropyl) phosphonate (4)

In an open flask, MnCl₂ (70 mg, 0.05 eq., 0.35 mmol) was added to a stirred solution of salen ligand (130 mg, 0.05 eq., 0.35 mmol) in i-PrOH. After 30 minutes of stirring at room temperature, a solution of 3˙ (2.2 g, 1 eq., 7.1 mmol) and 2-vinylpyridine (1.15 mL, 1.5 eq., 10.6 mmol) in i-PrOH was added first, and then solid NaBH₄ (1.07 g, 4 eq., 28.4 mmol) was added in small portions. The resulting suspension was stirred at room temperature for 4 h. It was then diluted with EtOAc and 1 M aq. HCl was carefully added. Solid NaHCO₃ was then added until neutralization. The layers were separated, and the organic phase was washed with water and brine, and dried over Na₂SO₄. The solvent was evaporated to give the crude product as a 1 : 2 mixture of diastereoisomers (³¹P-NMR ratio). The diastereomers were separated by automatic flash column chromatography (gradient PE/AcOEt: 0% to 100% of AcOEt) to afford RR/SS-4 (pale yellow crystal 560 mg, 19%) and RS/SR-4 (pale yellow oil 1.01 g 35%). RS/SR-4¹H NMR (300 MHz, CDCl₃) δ: 8.44 (dd, J = 5.1, 1.7 Hz, 1H), 7.59 (td, J = 7.7, 1.8 Hz, 1H), 7.45 (d, J = 7.8 Hz, 1H), 7.07 (ddd, J = 7.6, 4.8, 1.2 Hz, 1H), 5.20 (q, J = 6.5 Hz, 1H), 3.89 (d, J = 12.7 Hz, 1H), 3.67 (m, 4H), 3.57 (d, J_H–P = 26.5 Hz, 1H), 3.44 (d, J = 12.6 Hz, 1H), 1.57 (d, J = 6.6 Hz, 3H), 1.12 (s, 3H), 1.10 (s, 9H), 1.09 (s, 3H), 1.04 (t, J = 7.1 Hz, 3H), 0.98 (t, J = 7.1 Hz, 3H). ¹³C NMR (75 MHz, CDCl₃) δ: 162.1, 148.6, 136.9, 122.3, 121.6, 78.7, 70.5 (d, J_C–P = 136.3 Hz), 70.0, 64.9, 61.7 (d, J_C–P = 6.6 Hz), 60.1 (d, J_C–P = 7.9 Hz), 35.3 (d, J_C–P = 4.0 Hz), 30.6 (d, J_C–P = 5.7 Hz, 3C), 26.9, 22.0, 20.4, 16.4 (d, J_C–P = 5.7 Hz), 15.9 (d, J_C–P = 6.8 Hz). ³¹P NMR (121 MHz, CDCl₃) δ: 26.13. HRMS m/z (ESI) calcd for C₂₀H₃₈N₂O₅P [M + H]⁺ 417.2513, found: 417.2513. RR/SS-4¹H NMR (400 MHz, CDCl₃) δ: 8.49 (m, 1H), 7.66 (td, J = 7.7, 1.8 Hz, 1H), 7.43 (d, J = 7.9 Hz, 1H), 7.15 (ddd, J = 7.6, 4.9, 1.2 Hz, 1H), 5.09 (q, J = 6.7 Hz, 1H), 4.64 (s, 1H), 4.17 (m, 4H), 3.62 (d, J_H–P = 26.4 Hz, 1H), 3.49 (d, J = 12.3 Hz, 1H), 3.04 (d, J = 12.4 Hz, 1H), 1.54 (d, J = 6.7 Hz, 3H), 1.33 (t, J = 7.1 Hz, 6H), 1.18 (s, 9H), 1.02 (s, 3H), 0.68 (s, 3H). ¹³C NMR (101 MHz, CDCl₃) δ: 164.2, 148.7, 136.7, 122.4, 121.7, 86.3, 69.6 (d, J_C–P = 136.1 Hz), 67.5, 65.1, 62.0 (d, J_C–P = 6.6 Hz), 60.6 (d, J_C–P = 7.8 Hz), 35.8 (d, J_C–P = 4.6 Hz), 30.1 (d, J_C–P = 5.7 Hz, 3C), 26.8, 23.6, 23.2, 16.6 (d, J_C–P = 5.7 Hz), 16.4 (d, J_C–P = 6.6 Hz). ³¹P NMR (162 MHz, CDCl₃) δ: 27.29. HRMS m/z (ESI) calcd for C₂₀H₃₈N₂O₅P [M + H]⁺ 417.2513, found: 417.2515.

Diethyl ((R)-1-((1-hydroxy-2-methylpropan-2-yl)((S)-1-(pyridin-2-yl)ethoxy)amino)-2,2-dimethylpropyl)phosphonate (RS/SR-5)

Alkoxyamine RS/SR-4 (400 mg, 1 eq., 0.96 mmol) was dissolved in 20 mL THF, and pyridine was added (388 μL, 5 eq., 4.80 mmol). Silver nitrate (244 mg, 1.5 eq., 1.44 mmol) was added and after stirring for 5 minutes, t-butyldimethylsilyl chloride (217 mg, 1.5 eq., 1.44 mmol) was added and stirring was continued at room temperature overnight. At the end of the reaction period, the solution was washed with 10% NaHCO₃ solution and extracted with DCM. The combined organic phases were dried with MgSO₄ and the solvent was evaporated. The crude product was purified by flash chromatography (petroleum ether/AcOEt 7 : 3) to afford RS/SR-5 (pale yellow oil, 341 mg, 67%).¹H NMR (400 MHz, CDCl₃) δ: 8.52 (ddd, J = 4.9, 1.8, 0.9 Hz, 1H), 7.60 (m, 2H), 7.11 (ddd, J = 7.3, 4.8, 1.5 Hz, 1H), 5.26 (q, J = 6.6 Hz, 1H), 3.87 (m, 2H), 3.77 (d, J_H–P = 26.7 Hz, 1H), 3.71 (m, 1H), 3.63 (s, 2H), 3.57 (m, 1H), 1.58 (d, J = 6.6 Hz, 3H), 1.17 (m, 18H), 1.03 (t, J = 7.1 Hz, 3H), 0.91 (s, 9H), 0.04 (2s, 6H). ¹³C NMR (101 MHz, CDCl₃) δ: 162.5, 148.6, 136.1, 122.5, 122.1, 79.9, 69.8 (d, J_C–P = 137.0 Hz), 69.2, 65.4, 61.3 (d, J_C–P = 6.4 Hz), 59.1 (d, J_C–P = 7.6 Hz), 35.5 (d, J_C–P = 5.1 Hz), 30.3 (d, J_C–P = 5.9 Hz), 25.9 (5C), 23.2, 22.8, 20.2, 18.3, 16.4 (d, J_C–P = 5.8 Hz), 16.1 (d, J_C–P = 7.0 Hz), −5.4 (2C). ³¹P NMR (162 MHz, CDCl₃) δ: 25.22. HRMS m/z (ESI) calcd for C₂₆H₅₂N₂O₅PSi [M + H]⁺ 531.3378, found: 531.3380.

Diethyl((R)-1-((1-hydroxy-2-methylpropan-2-yl)((R)-1-(pyridin-2-yl)ethoxy)amino)-2,2-dimethylpropyl)phosphonate (RR/SS-5)

The same procedure as that for RS/SR-5 was applied to RR/SS-4. RR/SS-4 (210 mg, 1 eq., 0.5 mmol), 10 mL THF, pyridine (203 μL, 5 eq., 2.52 mmol), silver nitrate (128 mg, 1.5 eq., 0.76 mmol), t-butyldimethylsilyl chloride (114 mg, 1.5 eq., 0.76 mmol). RR/SS-5 (pale yellow oil, 157 mg, 59%). ¹H NMR (400 MHz, CDCl₃) δ: 8.53 (ddd, J = 4.9, 1.8, 0.9 Hz, 1H), 7.66 (td, J = 7.7, 1.8 Hz, 1H), 7.34 (d, J = 7.8 Hz, 1H), 7.15 (ddd, J = 7.5, 4.9, 1.2 Hz, 1H), 5.13 (q, J = 6.8 Hz, 1H), 4.37 (m, 1H), 4.09 (m, 3H), 3.73 (d, J_H–P = 26.2 Hz, 1H), 3.12 (s, 2H), 1.59 (d, J = 6.8 Hz, 3H), 1.36 (t, J = 7.1 Hz, 3H), 1.30 (t, J = 7.1 Hz, 3H), 1.23 (s, 9H), 0.88 (s, 3H), 0.83 (s, 3H), 0.81 (s, 9H), −0.14 (2s, 6H). ¹³C NMR (75 MHz, CDCl₃) δ: 164.5, 148.8, 136.2, 122.1, 121.7, 86.4, 69.8 (d, J_C–P = 138.3 Hz), 69.3 (d, J_C–P = 1.4 Hz), 65.4, 61.6 (d, J_C–P = 6.4 Hz), 58.9 (d, J_C–P = 7.5 Hz), 35.8 (d, J_C–P = 6.0 Hz), 29.8 (d, J_C–P = 5.8 Hz), 25.8 (3C), 23.3, 23.0, 22.6, 18.1, 16.8 (d, J_C–P = 5.5 Hz), 16.2 (d, J_C–P = 6.8 Hz), −5.6 (2C). ³¹P NMR (162 MHz, CDCl₃) δ: 26.11. HRMS m/z (ESI) calcd for C₂₆H₅₂N₂O₅PSi [M + H]⁺ 531.3378, found: 531.3380.

2-Bromo-2-(pyridin-2-yl)ethyl acetate (6b)

2-(Pyridin-2-yl)ethyl acetate 6a (2.0 g, 1 eq., 12.0 mmol), N-bromosuccinimide (2.4 g, 1.1 eq., 13.2 mmol) and benzoyl peroxide (0.03 g, 0.01 eq., 0.12 mmol) were mixed with CCl₄ (100 mL). An argon bubbled solution was refluxed (77 °C). The reaction was monitored by thin layer chromatography (TLC) after until complete disappearance of 6a. The solution was then quenched and washed with NaHCO₃ sat., and the aqueous phases were extracted with CH₂Cl₂. The combined organic layers were washed with water, dried over MgSO₄ and concentrated under reduced pressure. The crude product was subjected to automatic flash-chromatography (gradient PE/AcOEt: 0% to 100% of AcOEt) to yield bromide 6b (2.02 g, 69%). ¹H NMR (400 MHz, CDCl₃) δ: 8.59 (d, J = 5.6 Hz, 1H), 7.70 (t, J = 8.6 Hz, 1H), 7.42 (d, J = 7.9 Hz, 1H), 7.28–7.20 (m, 1H), 5.18 (t, J = 7.1 Hz, 1H), 4.73 (d, J = 7.1 Hz, 2H), 2.02 (s, 3H). ¹³C NMR (101 MHz, CDCl₃) δ: 170.3, 156.7, 149.4, 137.3, 123.6, 123.1, 66.3, 49.1, 20.7. HRMS (ESI) calc. for C₉H₁₁NO₂Br [M + H]⁺: 243.9968; found: 243.9967.

Diethyl-(1-(tert-butyl-(2-hydroxy-1-(pyridin-2-yl)ethoxy)amino)-2,2-dimethylpropyl) phosphonate (6)

To a suspension of CuBr (660 mg, 0.55 eq., 4.6 mmol) and Cu powder (580 mg, 1.1 eq., 9.1 mmol) in degassed benzene (argon bubbling for one hour) (30 mL) was added N,N,N′,N′,N′′-pentamethyldiethylenetriamine (1 mL, 0.55 eq., 4.6 mmol). After stirring for 10 min, a solution of 6˙ (2.68 g, 1.1 eq., 9.1 mmol) and bromide 6b (2.0 g, 1 eq., 8.3 mmol) in degassed benzene (30 mL) was cannulated into the first solution. The mixture was allowed to stir for 12 h. The solution was diluted with EtOAc, quenched and washed with 50% (v/v) aq. ammonia solution and NaHCO₃ saturated solution, and dried over MgSO₄. The solvents were evaporated under reduced pressure. The crude product (1.1 g, 1 eq., 2.4 mmol) with a 1 : 4 diastereomeric ratio (³¹P NMR ratio) was dissolved in MeOH (7 mL), and a solution of K₂CO₃ (663 mg, 2 eq., 4.8 mmol) in H₂O (7 mL) was added at once to the flask. The solution was allowed to stir for 3 days and then quenched with water. The aqueous phase was extracted with CH₂Cl₂ and dried over MgSO₄, and the solvents were evaporated under reduced pressure. The crude product was subjected to automatic flash-column chromatography (PE/acetone 7 : 3) to afford RS/SR-6 (144 mg) and RR/SS-6 (490 mg) as colourless oils, corresponding to a total yield of 634 mg (63%). RS/SR-6¹H NMR (400 MHz, CDCl₃) δ: 8.53 (d, J = 4.8 Hz, 1H), 7.64 (t, J = 7.7 Hz, 1H), 7.37 (d, J = 7.7 Hz, 1H), 7.18–7.09 (m, 1H), 5.93 (t, J = 6.9 Hz, 1H), 5.11 (s, 1H), 4.44 (dd, J = 16.5, 7.7 Hz, 1H), 4.28 (dd, J = 16.7, 9.4 Hz, 1H), 4.21–4.09 (m, 2H), 4.02 (dd, J = 15.8, 7.9 Hz, 1H), 3.84 (dd, J = 15.1, 8.5 Hz, 1H), 3.43 (d, J_H–P = 27.6 Hz, 1H), 1.38 (t, J = 7.8 Hz, 3H), 1.34 (t, J = 7.6 Hz, 3H), 1.25 (s, 9H), 0.92 (s, 9H). ¹³C NMR (101 MHz, CDCl₃) δ: 160.8, 149.0, 136.0, 122.7, 122.3, 91.0, 68.8 (d, J_C–P = 139.6 Hz), 65.0, 62.1, 62.1, 59.8 (d, J = 7.6 Hz), 35.8 (d, J = 5.1 Hz), 30.9 (d, J = 5.8 Hz, 2C), 28.1 (3C), 16.7 (d, J = 5.5 Hz), 16.24 (d, J = 7.0 Hz). ³¹P NMR (162 MHz, CDCl₃) δ: 27.16. HRMS (ESI) calc for C₂₀H₃₈N₂O₅P⁺: 417.2513 [M + H]⁺; found: 417.2512. RR/SS-6¹H NMR (400 MHz, CDCl₃) δ: 8.48 (d, J = 5.5 Hz, 1H), 7.73 (s, 1H), 7.65 (t, J = 8.5 Hz, 1H), 7.20–7.13 (m, 1H), 5.29 (t, J = 5.6 Hz, 1H), 4.72 (s, 1H), 4.34 (dd, J = 11.2, 5.3 Hz, 1H), 4.01 (dd, J = 11.2, 6.0 Hz, 1H), 3.96–3.83 (m, 2H), 3.73–3.60 (m, 2H), 3.46 (d, J_H–P = 26.7 Hz, 1H), 1.25 (s, 9H), 1.21 (s, 9H), 1.13 (t, J = 7.1 Hz, 3H), 1.03 (t, J = 7.1 Hz, 3H). ¹³C NMR (101 MHz, CDCl₃) δ: 159.7, 148.2, 136.4, 123.3, 122.6, 80.5, 70.2 (d, J_C–P = 138.8 Hz), 64.7, 62.2, 61.3 (d, J = 6.7 Hz), 59.8 (d, J = 7.7 Hz), 35.4, 35.3, 30.8 (d, J = 5.9 Hz, 2C), 28.1 (3C), 16.3 (d, J = 6.0 Hz), 16.0 (d, J = 6.9 Hz). ³¹P NMR (162 MHz, CDCl₃) δ: 24.48. HRMS (ESI) calc for C₂₀H₃₈N₂O₅P⁺: 417.2513 [M + H]⁺; found: 417.2511.

(S)-2-((tert-Butyl((R)-1-(diethoxyphosphoryl)-2,2-dimethylpropyl)amino)oxy)-2-(pyridin-2-yl)ethyl acetate (RS/SR-7)

RS/SR-6 (144 mg, 1 eq., 0.35 mmol) was diluted in CH₂Cl₂ (5 mL) and triethylamine (0.2 mL, 4 eq., 1.4 mmol) was added. After 2 minutes, acetic anhydride (0.1 mL, 3 eq., 1.1 mmol) was added slowly via a syringe. The reaction was allowed to stir for 3 days and was then quenched with a saturated aqueous solution of NaHCO₃. The solution was extracted with CH₂Cl₂. The combined organic extracts were dried with MgSO₄, filtered, and concentrated under vacuum. The crude was purified by automatic flash-column chromatography (petroleum ether/acetone 9 : 1) to yield RS/SR-7 (140 mg, 90%) as a white solid. ¹H NMR (300 MHz, CDCl₃) δ: 8.54 (d, J = 4.7 Hz, 1H), 7.66 (t, J = 7.7 Hz, 1H), 7.36 (d, J = 7.8 Hz, 1H), 7.21–7.12 (m, 1H), 5.30 (dd, J = 7.7, 4.2 Hz, 1H), 4.84 (dd, J = 10.9, 4.2 Hz, 1H), 4.69–4.59 (m, 1H), 4.48–4.33 (m, 1H), 4.21 (td, J = 16.6, 8.9 Hz, 1H), 4.03 (m, J = 17.2, 12.7, 7.2 Hz, 2H), 3.35 (d, J_H–P = 26.1 Hz, 1H), 1.79 (s, 3H), 1.36 (t, J = 6.4 Hz, 3H), 1.31 (t, J = 7.0 Hz, 3H), 1.23 (s, 9H), 0.87 (s, 9H). ¹³C NMR (101 MHz, CDCl₃) δ: 170.1, 160.1, 148.8, 135.9, 123.3, 122.6, 87.5, 69.5 (d, J_C–P = 138.9 Hz), 65.7, 61.8 (2C), 61.7 (d, J = 6.4 Hz), 59.3 (d, J = 7.4 Hz), 35.8 (d, J = 5.4 Hz), 30.1 (d, J = 5.8 Hz, 2C), 28.4 (3C), 20.7, 16.8 (d, J = 5.4 Hz), 16.3 (d, J = 6.8 Hz). ³¹P NMR (121 MHz, CDCl₃) δ: 25.27. HRMS (ESI) calc for C₂₂H₄₀N₂O₆P⁺: 459.2619 [M + H]⁺; found: 459.2624.

(R)-2-((tert-Butyl((R)-1-(diethoxyphosphoryl)-2,2-dimethylpropyl)amino)oxy)-2-(pyridin-2-yl)ethyl acetate (RR/SS-7)

The same procedure as that for RS/SR-7 was applied to RR/SS-6. RR/SS-6 (490 mg, 1 eq., 1.18 mmol), CH₂Cl₂ (10 mL), triethylamine (0.5 mL, 4 eq., 3.5 mmol), acetic anhydride (0.3 mL, 3 eq., 3 mmol) flash-column chromatography (PE/acetone 9 : 1) RR/SS-7 (343 mg, 65%) as a colorless solid. ¹H NMR (300 MHz, CDCl₃) δ: 8.55 (d, J = 4.8 Hz, 1H), 7.69–7.57 (m, 2H), 7.20–7.08 (m, 1H), 5.35 (dd, J = 6.5, 3.6 Hz, 1H), 4.76 (m, 2H), 3.88 (p, J = 7.2 Hz, 2H), 3.74–3.50 (m, 2H), 3.44 (d, J_H–P = 27.1 Hz, 1H), 1.86 (s, 3H), 1.23 (s, 9H), 1.19 (t, J = 7.1 Hz, 3H), 1.15 (s, 9H), 1.00 (t, J = 7.1 Hz, 3H). ¹³C NMR (101 MHz, CDCl₃) δ: 170.6, 158.1, 148.7, 135.8, 124.6, 122.7, 80.8, 69.6 (d, J_C–P = 138.6 Hz), 64.1, 62.0, 61.4 (d, J = 6.6 Hz), 59.5 (d, J = 7.7 Hz), 35.3 (d, J = 4.4 Hz), 30.7 (d, J = 5.9 Hz, 2C), 28.0 (3C), 20.7, 16.3 (d, J = 5.9 Hz), 16.1 (d, J = 6.8 Hz). ³¹P NMR (162 MHz, CDCl₃) δ: 24.27. HRMS (ESI) calc for C₂₂H₄₀N₂O₆P⁺: 459.2619 [M + H]⁺; found: 459.2623.

Kinetic measurements

The values of the homolysis rate constant k_d were determined by monitoring either the concentration of nitroxide by EPR or the concentration of alkoxyamine by ³¹P NMR. For EPR, a sample tube filled with a solution of 10⁻⁴ M of each diastereoisomer in tert-butylbenzene or in H₂O/MeOH (1 : 1) was set in the EPR cavity. EPR signals were recorded. The temperature was controlled by using a BVT2000 temperature controlling unit. Measurements of k_d by ³¹P NMR required the use of TEMPO as an alkyl radical scavenger. The NMR tubes were filled with a stock solution of 0.02 M of alkoxyamine in tert-butylbenzene or in H₂O/MeOH with 2 equiv. of TEMPO. Buffer solutions were used for specific pH conditions instead of H₂O. k_d values were given by eqn (4). Activation energies E_a were estimated using eqn (5) and the average frequency factor A = 2.4 × 10¹⁴ s⁻¹. The values of k_d and E_a are listed in Table 4.

k_d = Ae^−E_a/RT (5)

Conclusion

Different types of IHBs exhibiting very different effects are highlighted. These effects depend on the strength of the IHB, which in turn is straightforwardly related to its geometric parameters – d_OH⋯X the distance for IHB and the IHB valence angle α. That is, strong interR IHBs are expected for α larger than 160° and d_OH⋯X smaller than 1.8 Å affording an increase in E_a (Tables 4 and 5). As geometric parameters – α ≈ 150° and d_OH⋯X ≈ 2.1 Å – underline a weak IHB, interN IHB does not exhibit the same effect as that of interR IHB. Indeed, this interN IHB is weak enough that it is in equilibrium with the intraN IHB which, in turn, favours slightly the C–ON bond homolysis, i.e., decreasing E_a. Nevertheless, a strong interN IHB is expected to afford the same effect as that of a strong interR IHB. IntraN and intraR IHBs exhibit features of IHB of intermediate strength – α < 150° and d_OH⋯X > 1.8 Å – affording a decrease in E_a. However, the effect of intraR and intraN depends a lot on the strength of these IHBs at TS, that is, on the stereoelectronic requirement for the C–ON bond homolysis and on the stabilization of the products, i.e., partly due to the strength of IHB. Hence, stronger IHBs in products than in starting materials stabilize TS and decrease E_a. This assumption is supported by the stronger IHB in 3˙ (α = 178° and d_OH⋯X = 1.57 Å) than in 3 and 4 (Table 5).‡‡‡ ⁵⁷

As a rule of thumb, one may assume that IHB between alkyl and nitroxyl fragments (interN or interR IHB) affords an increase in E_a which might be pictured as the cleavage of two bonds – C–ON bond and IHB – and that IHB inside each fragment (intraN or intraR IHB) affords a slight decrease in E_a provided the stabilization due to IHB is larger for the released radicals than that for the starting material. Otherwise, an increase in E_a might be expected.

This work highlights the combination of several effects – IHB, solvents, intimate ion pairs, and protonation – to strikingly decrease the alkoxyamine half-life time t_1/2, as highlighted by t_1/2 = 123 days for 6 in t-BuPh as a solvent at 37 °C and by t_1/2 = 14 hours for 6H+ in water/MeOH as a solvent at 37 °C, that is, a 210-fold increase in k_d. These results nicely highlight the potential of such alkoxyamines as switches for applications in biology.¹⁷

Conflicts of interest

There are no conflicts of interest to declare.

Acknowledgements

SRAM, PB, VR and GA thank Aix-Marseille Université and CNRS for support. SRAM and ME are grateful to the Russian Science Foundation (grant 15-13-20020) for supporting this work and to the Multi-Access Chemical Service Center SB RAS for spectral and analytical measurements. SRAM, VR, GA and JPJ are grateful to ANR for funding this project (ANR-14-CE16-0023-01). PN thanks the Ministère de la Recherche du Gabon.

Notes and references

1. D. H. Solomon, E. Rizzardo and P. Cacioli, Eur. Pat. Appl135280, 1985; D. H. Solomon, E. Rizzardo and P. Cacioli, US Patent4581429, 1986; Chem. Abstr., 1985, 102, 221335q.
2. Nitroxide Mediated Polymerization: From Fundamentals to Applications in Materials Sciences, ed. D. Gigmes, RSC Polymer Chemistry Series 19, Royal Society of Chemistry, London, 2016; and 513 references cited therein.
3. A. Studer, Chem. Soc. Rev., 2004, 33, 267–273.
4. L. Tebben and A. Studer, Angew. Chem., Int. Ed., 2011, 50, 5034–5068.
5. A. Studer, Angew. Chem., Int. Ed., 2000, 39, 1108–1111.
6. J. Xu, E. J. E. Caro-Diaz, L. Trzoss and E. A. Theodorakis, J. Am. Chem. Soc., 2012, 134, 5072–5075.
7. C. Wetter, K. Jantos, K. Woithe and A. Studer, Org. Lett., 2003, 5, 2899–2902.
8. V. Sciannamea, R. Jérôme and C. Detrembleur, Chem. Rev., 2008, 108, 1104–1126.
9. E. H. H. Wong, M. H. Stenzel, T. Junkers and C. Barner-Kowollik, Macromolecules, 2010, 43, 3785–3793.
10. Y. Guillaneuf, D.-L. Versace, D. Bertin, J. Lalevée, D. Gigmes and J. P. Fouassier, Macromol. Rapid Commun., 2010, 31, 1909–1913.
11. E. Yoshida, Colloid Polym. Sci., 2010, 288, 1639–1643.
12. G. Audran, E. G. Bagryanskaya, P. Brémond, M. V. Edeleva, S. R. A. Marque, D. A. Parkhomenko, O. Y. Rogozhnikova, V. M. Tormyshev, E. V. Tretyakov, D. V. Trukhin and S. I. Zhivetyeva, Polym. Chem., 2016, 7, 6490–6499.
13. G. Audran, E. G. Bagryanskaya, M. V. Edeleva, S. R. A. Marque, D. A. Parkhomenko, E. V. Tretyakov and S. I. Zhivetyeva, J. Polym. Sci.: Part A: Polym. Chem., 2017 , submitted.
14. M. Q. Zhang and M. Z. Rong, Polym. Chem., 2013, 4, 4878.
15. B. Schulte, M. Tsotsalas, M. Becker, A. Studer and L. De Cola, Angew. Chem., Int. Ed., 2010, 49, 6881–6884.
16. R. K. Roy, A. Meszynska, C. E. Laure, L. Charles, C. Verchin and J.-F. Lutz, Nat. Commun., 2015, 6, 1–8.
17. G. Audran, P. Brémond, J.-M. Franconi, S. R. A. Marque, P. Massot, P. Mellet, E. Parzy and E. Thiaudière, Org. Biomol. Chem., 2014, 12, 719–723.
18. D. Moncelet, P. Voisin, N. Koonjoo, V. Bouchaud, P. Massot, E. Parzy, G. Audran, J.-M. Franconi, E. Thiaudière, S. R. A. Marque, P. Brémond and P. Mellet, Mol. Pharmaceutics, 2014, 11, 2412–2419.
19. N. A. Popova, G. M. Sysoeva, V. P. Nicolin, V. I. Kaledin, E. V. Tretyakov, M. V. Edeleva, S. M. Balakhnin, E. L. Louschnikova, G. Audran and S. Marque, Bull. Exper. Biol. Med., 2017, 164(7), 61–65.
20. P. Brémond, A. Koïta, S. R. A. Marque, V. Pesce, V. Roubaud and D. Siri, Org. Lett., 2012, 14(1), 358–361.
21. G. Audran, E. Bagryanskaya, I. Bagryanskaya, P. Brmond, M. Edeleva, S. R. A. Marque, D. Parkhomenko, E. Tretyakov and S. Zhivetyeva, Inorg. Chem. Front., 2016, 3, 1464–1472.
22. G. Audran, E. Bagryanskaya, I. Bagryanskaya, M. Edeleva, S. R. A. Marque, D. Parkhomenko, E. Tretyakov and S. Zhivetyeva, ChemistrySelect, 2017, 2, 3584–3593.
23. P. Nkolo, G. Audran, P. Brémond, R. Bikanga, S. R. A. Marque and V. Roubaud, Org. Biomol. Chem., 2017, 15, 6167–6176.
24. K. Matyjaszewski, S. Gaynor, D. Greszta, D. Mardare and T. Shigemoto, J. Phys. Org. Chem., 1995, 8, 306–315.
25. S. Marque, H. Fischer, E. Baier and A. Studer, J. Org. Chem., 2001, 66, 1146–1156.
26. S. Acerbis, D. Bertin, B. Boutevin and D. Gigmes, Helv. Chim. Acta, 2006, 89, 2119–2132.
27. G. Audran, P. Brémond, S. R. A. Marque and T. Yamasaki, J. Org. Chem., 2016, 81, 1981–1988.
28. E. G. Bagryanskaya, P. Brémond, T. Butscher, S. R. A. Marque, D. Parkhomenko, V. Roubaud, D. Siri and S. Viel, Macromol. Chem. Phys., 2015, 216(5), 475–488.
29. P. Brémond, T. Butscher, V. Roubaud, D. Siri and S. Viel, J. Org. Chem., 2013, 78, 10524–10529.
30. C. Reichardt and T. Welton, Solvent and Solvent Effect in Organic Chemistry, Wiley-VCH, Weinheim, 2011.
31. E. J. Corey and A. Venkateswarlu, J. Am. Chem. Soc., 1972, 94, 6190.
32. E. J. Corey, H. Cho, C. Rücker and D. H. Hua, Tetrahedron Lett., 1981, 22, 3455.
33. R. Braslau, N. Naik and H. Zipse, J. Am. Chem. Soc., 2000, 122, 8421–8434.
34. K. K. Ogilvie, G. H. Hakimelahi, Z. A. Proba and D. P. C. McGee, Tetrahedron Lett., 1982, 23((19), 1997–2000.
35. K. Matyjaszewski, B. E. Woodworth, X. Zhang, S. G. Gaynor and Z. Metzner, Macromolecules, 1998, 31, 5955–5957.
36. E. G. Bagryanskaya and S. R. A. Marque, in Kinetic Aspects of Nitroxide-Mediated Polymerization, RSC Polymer Chemistry Series, n° = 19, Nitroxide Mediated Polymerization: From Fundamentals to Applications in Materials Sciences, ed. D. Gigmes, Royal Society of Chemistry, 2016, ch. 2, pp. 45–113.
37. R. S. Rowland and R. Taylor, J. Phys. Chem., 1996, 100, 7384–7391.
38. G. A. Jeffrey and W. Saenger, Hydrogen Bonding in Biological Structures, Springer, Berlin, Heidelberg, 1994.
39. A. Krȩżel and W. Bal, J. Inorg. Biochem., 2004, 98, 161–166.
40. R. J. L. Andon, J. D. Cox and E. F. G. Herington, Trans. Faraday Soc., 1954, 50, 918–910.
41. G. Audran, M. Bim Batsiandzy Ibanou, P. Brémond, S. R. A. Marque, V. Roubaud and D. Siri, Org. Biomol. Chem., 2013, 11, 7738–7750.
42. 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 A.02), Gaussian, Inc., Wallingford, CT, 2009.
43. S. Marque, J. Org. Chem., 2003, 68, 7582–7590.
44. S. Acerbis, E. Beaudoin, D. Bertin, D. Gigmes, S. Marque and P. Tordo, Macromol. Chem. Phys., 2004, 205, 973–978.
45. A. Studer, K. Harms, C. Knoop, C. Müller and T. Schulte, Macromolecules, 2004, 37, 27–34.
46. G. Audran, R. Bikanga, P. Brémond, J.-P. Joly, S. R. A. Marque and P. Nkolo, J. Org. Chem., 2017, 82, 5702–5709.
47. P. Nkolo, G. Audran, P. Brémond, R. Bikanga, S. R. A. Marque and V. Roubaud, Tetrahedron, 2017, 73, 3188–3201.
48. S. W. Benson, Thermochemical Kinetics. Methods for the Estimation of Thermochemical Data and Rate Parameters, John Wiley & Sons, Inc., New York, 1968.
49. E. Beaudoin, D. Bertin, D. Gigmes, S. R. A. Marque, D. Siri and P. Tordo, Eur. J. Org. Chem., 2006, 1755–1768.
50. C. Hansch, A. Leo and R. W. Taft, Chem. Rev., 1991, 91, 165–195.
51. D. Bertin, D. Gigmes, S. R. A. Marque and P. Tordo, Macromolecules, 2005, 38, 2638–2650.
52. M. Charton, Prog. Phys. Org. Chem., 1981, 13, 119–251.
53. G. Audran, P. Brémond, S. R. A. Marque and G. Obame, Polym. Chem., 2012, 3, 2901–2908.
54. G. Audran, P. Brémond, S. R. A. Marque and G. Obame, J. Org. Chem., 2012, 77(21), 9634–9640.
55. G. Audran, P. Brémond, S. R. A. Marque and G. Obame, J. Org. Chem., 2013, 78, 7754–7757.
56. T. J. Barker and E. R. Jarvo, Angew. Chem., Int. Ed., 2011, 50, 8325–8328.
57. M. C. Foti, R. Amorati, G. F. Pedulli, C. Daquino, D. A. Pratt and K. U. Ingold, J. Org. Chem., 2010, 75, 4434–4440.

---

Footnotes

† Electronic supplementary information (ESI) available: ¹H, ¹³C NMR and HRMS spectra of 2–7 and 2˙. DFT calculations for the most stable conformers of 4 and 6. A gradient profile for flash chromatography. The titration curve for 2-hydroxyethylpyridine, and XRD for RR/SS-4 and RS/SR-6. CCDC 1550955 and 1550956. For ESI and crystallographic data in CIF or other electronic format see DOI: 10.1039/c7ob02223a

‡ CCDC: 1550955 for RR/SS-4 and 1550956 for RS/SR-6.†

§ The procedure described in Scheme 2 was not used because it afforded the isomerization of the pure diastereoisomer.

¶ d_P=O⋯HO = 1.96 Å and <O₁H₁₅O₁₄> = 164°. The bond radii of H, N and O are r_H = 1.09 Å, r_N = 1.55 Å and r_O = 1.52 Å, respectively. See ref. 37.

|| RR/SS-5 exhibiting no IHB was chosen as a model for 4 and very similar solvent effects were expected.

** Protons from CH₂OH (labelled e in Fig. 9) are expected to be sensitive to the silylation of the hydroxyl functions, whereas protons labelled b and c are not expected to be very sensitive.

†† Conformer A exhibiting IHB between HO and P=O functions, conformer B exhibiting IHB between the OH function and N atom, and C the most stable conformer with no IHB.

‡‡ Indeed, conformer A exhibits geometrical parameters (Table 5) featuring stronger IHB than in B as highlighted by the shorter d_OH⋯OP and flatter angle α for RR/SS-6.

§§ When the t-Bu group attached to the nitroxyl moiety is replaced by a bulkier group CMe₂R, no difference in k_d is observed, meaning that the nitroxyl fragment adopts a conformation which forces the R group to be in such a position that its bulkiness is cancelled. This is due to the levelled steric effect. See ref. 43–46.

¶¶ In ref. 29, no intraN IHB is observed in X-ray data.

|||| Indeed, a 3-center IHB (Fig. 8) would require a conformation for the aromatic ring exhibiting a 1,3-syn allylic destabilizing interaction with H10, whereas such an interaction does not occur for conformers A and B.

*** The weak effect of IHB in RR/SS-6 is due to a weak IHB (the weakest of those investigated), as highlighted by its longest d_OH⋯X and its closed angle α.

††† For such molecules, a difference of 2.5 kJ mol⁻¹ in energy is in the limit of accuracy and reliability of the method.

‡‡‡ A similar effect on remote IHB has already been reported in ref. 57. The authors would like to thank the reviewer for pointing this issue at TS.

---