Skip to content
Publicly Available Published by De Gruyter February 28, 2017

Mechanistic study of stereoselectivity in azoalkane denitrogenations

Manabu Abe EMAIL logo and Sayaka Hatano

Abstract

Since 1965, the stereoselectivity in azoalkane denitrogenation has attracted much attention in both synthetic organic chemistry and physical organic chemistry. In this paper, a short review of the recent findings on the mechanism underlying the fascinating stereoselectivity in azoalkane denitrogenation is presented. The two types of singlet diradicals, i.e. the puckered and planar conformations, were found to play important roles in the stereoselectivity in the photochemical denitrogenation of cyclic azoalkanes. The presence of the puckered singlet diradical, which is the third isomer in homolysis, resolves the mechanistic puzzle reported so far for the stereoselectivity in azoalkane denitrogenations.

Introduction: mechanism underlying the denitrogenation of cyclic azoalkanes

The denitrogenation of azoalkanes such as 1,2-diazabicyclo[2.2.1]hept-1-ene (AZ1) has attracted much attention not only for the denitrogenation mechanism, but also for the synthesis of strained compounds (Scheme 1). The C2 symmetric cyclopentane-1,3-diyl diradical DR1 is expected to produce a 1:1 mixture of bicyclo[2.1.0]pentanes ret-CP1 and inv-CP1 in the denitrogenation of AZ1 (Scheme 1). However, stereochemical labeling studies clarified that the inversion product inv-CP1 was preferentially formed over the retention product ret-CP1, both in the thermal [1], [2], [3], [4], [5] and photochemical denitrogenation reactions [6], [7].

Scheme 1: Thermal and photochemical denitrogenation mechanisms of cyclic azoalkanes AZ1.
Scheme 1:

Thermal and photochemical denitrogenation mechanisms of cyclic azoalkanes AZ1.

Computational studies [8], [9] at the CASPT2/CASSCF level of theory revealed the following in photochemical reactions (hν): (1) the stepwise C–N bond breaking is an energetically more favored process than the concerted denitrogenation in the electronically excited n,π* state, to give the axial conformer of the diazenyl diradical ax-DZ1; (2) the puckered diradical puc-DR1 formed in the denitrogenation of ax-DZ1 is a precursor of ret-CP1; (3) inv-CP1 is formed from the equatorial conformation eq-DZ1 via a homolytic substitution (SH2) reaction; and (4) the planar diradical pl-DR1 affords a 1:1 mixture of CP1. Thus, the selective formation of inv-CP1 from eq-DZ1 via an SH2 mechanism is proposed [10]. In the thermal denitrogenation reactions (ΔT) [11], the concerted denitrogenation of AZ1 was proposed to selectively afford inv-CP1 [12], [13]. Thus, the dynamics effect and the reaction trajectory during the concerted denitrogenation process play important roles in the selective formation of inv-CP1.

Stereoselectivity in the photochemical denitrogenation of cyclic azoalkane AZ2

In contrast to the selective formation of the inverted ring-closing products (Scheme 1), the retention product ret-CP2 has been quantitatively isolated in the photochemical denitrogenation of AZ2 in the last two decades (Scheme 2) [14], [15], [16]. The stereochemical outcome was different from that for AZ1. The selectivity has been believed to result from the repulsive steric effect, because the ret-CP2 isomer is more stable in energy than the inv-CP2 isomer. Our recent mechanistic studies, however, contradicted this simple explanation for the observed stereoselectivity. In this paper, the fascinating stereoselectivity in the photochemical denitrogenation reactions is summarized to clarify the mechanism of photochemical denitrogenation.

Scheme 2: Spin-state dependent change in the stereoselectivity in the photochemical denitrogenation of AZ2; ret-CP2 versus inv-CP2.
Scheme 2:

Spin-state dependent change in the stereoselectivity in the photochemical denitrogenation of AZ2; ret-CP2 versus inv-CP2.

First, laser flash photolysis (LFP) experiments were conducted for detecting the intermediates in the photochemical denitrogenation of AZ2 (Scheme 2). The relatively long-lived singlet diradical pl-1DR2 was detected at ~570 nm, which decayed by first-order kinetics with a lifetime of ~300 ns at 293 K [14]. The quantum yield for the photochemical denitrogenation was determined to be 0.94±5%. Thus, the singlet diradical pl-1DR2 was proposed to be the intermediate for the selective formation of ret-CP2.

A computational study revealed that the calculated energy of the transition state ret-TS2 for the formation of ret-CP2 from pl-1DR2 was ~20 kJ mol−1 higher than that for the formation of inv-CP2 via inv-TS2 (Fig. 1) [17]. Hence, the selective formation of inv-CP2 can be expected in the denitrogenation of AZ2. The larger repulsive interaction “a” between the phenyl ring and the cyclopentane ring in ret-TS2 than “d” in inv-TS2 might be the reason for the higher transition state energy of ret-TS2 (Fig. 1). Thus, the prediction based on the computational study apparently contradicts the experimental observation, i.e. the selective isolation of ret-CP2. Interestingly, a small difference of ~13 kJ mol−1 was predicted between the energies of pl-1DR2 and inv-CP2. Thus, thermal isomerization from inv-CP2 to ret-CP2 was thought to be possible.

Fig. 1: Reaction profiles of pl-1DR2 at the BS-UB3LYP/6-31G(d) level of theory. The values of the relative thermal enthalpy, ΔHrel, are calculated at 298 K and 1 atm; ΔEzpe is the electronic energy after the zero-point energy corrections, see Ref. [17].
Fig. 1:

Reaction profiles of pl-1DR2 at the BS-UB3LYP/6-31G(d) level of theory. The values of the relative thermal enthalpy, ΔHrel, are calculated at 298 K and 1 atm; ΔEzpe is the electronic energy after the zero-point energy corrections, see Ref. [17].

Low-temperature 1H NMR analysis of photochemical denitrogenation of cyclic azoalkane AZ2

Low-temperature 1H NMR (400 MHz) spectroscopic analyses were performed to detect the formation of inv-CP2 in the photolysis of AZ2. Indeed, the formation of inv-CP2 was detected in the photolysate at 188 K using an in situ NMR setup (Table in Scheme 2) [18]. The thermal isomerization from inv-CP2 to ret-CP2 was also observed after warming the sample to 298 K. The activation parameters, Ea and log(A/s−1), for the thermal isomerization were determined to be 60.4 kJ mol−1 and 12.0, which were consistent with the computed activation energy from pl-1DR2 to ret-CP2 (Fig. 1). The computationally predicted thermal isomerization, thus, was proved by the in situ NMR spectroscopic analyses. Surprisingly, inv-CP2 was the minor product in the direct photochemical denitrogenation of AZ2; the ratio of ret-CP2/inv-CP2 was 82/18 at 188 K (Table in Scheme 2) [19], although the selective formation of inv-CP2 from pl-1DR2 was predicted by the computations (Fig. 1).

To obtain more information about the stereoselectivity, the triplet-sensitized denitrogenation of AZ2 was conducted at 199 K in toluene-d8 (Table in Scheme 2). Interestingly, the exclusive formation of inv-CP2 was observed in the in situ NMR spectroscopic analyses, with ret-CP2/inv-CP2=7/93 [18]. Thus, direct photolysis via the singlet excited state of the azo chromophore selectively produced ret-CP2; on the other hand, inv-CP2 was exclusively formed in the triplet-sensitized denitrogenation. A mechanism that explains the observed spin-state dependent change in stereoselectivity in the photochemical denitrogenation of AZ2 is summarized in Scheme 3.

Scheme 3: Mechanism of spin-state dependent change in stereoselectivity in AZ2 denitrogenation.
Scheme 3:

Mechanism of spin-state dependent change in stereoselectivity in AZ2 denitrogenation.

In the triplet-sensitized denitrogenation of AZ2 in the presence of benzophenone (Ph2CO), the triplet diazenyl diradical ax-3DZ2 is first generated from the triplet excited state 3AZ2*, which quantitatively generates the planar triplet diradical pl-3DR2 via ax-3DZ2. After intersystem crossing (ISC) to the planar singlet diradical pl-1DR2, inv-CP2 is exclusively formed via inv-TS2. The energy barrier for inv-CP2 is calculated to be lower than that for the reaction via ret-TS2 (Fig. 1). The exclusive formation of pl-1DR2 from 3AZ2* is reasonable, since the triplet diradicals have long lifetimes that allow the geometries to stabilize before the spin-forbidden ISC occurs. The planar singlet diradical pl-1DR2 can be generated from the most stable planar conformation of the triplet diradical pl-3DR2.

The selective formation of ret-CP2 under the direct irradiation conditions results from the intervention of another diradical intermediate, puc-1DR2, which may be obtained by the denitrogenation of the diazenyl diradical ax-1DZ2. The planar singlet diradical pl-1DR2 is formed from the equatorial diazenyl diradical eq-1DZ2, as theoretically predicted for the direct denitrogenation of AZ1 (Scheme 1). The following experimental evidences support the hypothetical mechanism for the spin-state dependence of stereoselectivity.

Transient absorption spectroscopy and low-temperature spectroscopy of cyclic azoalkane denitrogenations

The singlet diradical pl-1DR2 was observed in a triplet-sensitized denitrogenation of AZ2 (Fig. 2) [18], when using xanthone (Xan) as a triplet sensitizer. Because the molar absorption coefficient (ε) of the triplet excited state 3Xan* with λmax~625 nm is 3240 M−1 cm−1, the ε value for pl-1DR2 was determined to be 4900 M−1 cm−1 at λ~560 nm by the integrated profiles method (IPM) [19], [20].

Fig. 2: Quenching of the photogenerated (λexc=355 nm) xanthone (Xan) triplet, 3Xan*, and generation of pl-1DR2 in degassed acetonitrile at 293 K ([Xan]=0.50 mM (Abs355=0.13), [AZ2]=0.33 mM (Abs355=0.03)).
Fig. 2:

Quenching of the photogenerated (λexc=355 nm) xanthone (Xan) triplet, 3Xan*, and generation of pl-1DR2 in degassed acetonitrile at 293 K ([Xan]=0.50 mM (Abs355=0.13), [AZ2]=0.33 mM (Abs355=0.03)).

From the ε value for pl-1DR2, the chemical yield of pl-1DR2 was calculated to be quite low, ~16%, in the direct photolysis of AZ2, although the quantum yield (Φ) of the denitrogenation was quite high, i.e. 0.94±5%. The chemical yield of pl-1DR2 was consistent with the yield of inv-CP2, i.e. 18% (Table in Scheme 2). Other intermediates that produce ret-CP2 would also be formed during the denitrogenation.

A plausible intermediate for ret-CP2 is the puckered singlet diradical puc-1DR2. Indeed, a similar puckered diradical puc-1DR3 having sterically hindered substituents was detected under the low-temperature matrix conditions at ~450 nm in the denitrogenation of AZ3 (Scheme 4) [21]. In the 2-methyltetrahydrofuran (MTHF) matrix, the typical absorption signal of the singlet diradical pl-1DR3 was observed at ~570 nm and 117 K. Below ~100 K, a strong absorption signal appeared at ~450 nm instead of that at ~570 nm. The signal at ~450 nm was assigned to the puckered singlet diradical puc-1DR3. The third bond-stretch isomer was also computationally found to be an equilibrium geometry at the BS-UM06-2X/6-31G(d) level of theory. The observed absorption bands for pl-1DR3 and puc-1DR3 were in good agreement with those calculated for pl-1DR2 and puc-1DR2 by a TD-DFT method (TD-UB3LYP/6-31+G(d)) and the higher level ab initio method (CAS(2,2)+DDCI3).

Scheme 4: Steady-state UV/vis absorption spectra obtained in the photochemical reaction of AZ3 irradiated at 366 nm in 2-methyltetrahydrofuran (MTHF) at 117 K and 94 K.
Scheme 4:

Steady-state UV/vis absorption spectra obtained in the photochemical reaction of AZ3 irradiated at 366 nm in 2-methyltetrahydrofuran (MTHF) at 117 K and 94 K.

Summary

In this paper, the stereoselectivity in the photochemical denitrogenation of the azoalkane AZ2 is discussed in detail. A hitherto unknown stereoselectivity was observed for the photochemical denitrogenation of AZ2. The configurationally retained ring-closed isomer (ret-CP2) was selectively formed by the direct irradiation of the azo chromophore of AZ2. Under triplet-sensitized irradiation conditions, the configurationally inverted ring-closed isomer (inv-CP2) was formed selectively. Fast isomerization of inv-CP2 into ret-CP2 was observed in the dark. Transient absorption and low-temperature spectroscopic analyses revealed that two different singlet diradicals are involved in the formation of CP2 in the direct photochemical denitrogenation, i.e. the puckered puc-1DR2 and the planar pl-1DR2. The former produces ret-CP2, while the latter affords inv-CP2. The presence of the puckered singlet diradical resolves the mechanistic puzzle of the stereoselectivity in azoalkane denitrogenations.


Article note:

A collection of invited papers based on presentations at the 23rd IUPAC Conference on Physical Organic Chemistry (ICPOC-23), Sydney, Australia, 3–8 July 2016.


Acknowledgments

M.A. gratefully acknowledges the financial support by the Grant-in-Aid for Science Research on Innovative Areas “Stimuli- responsive Chemical Species (No.2408)” (JSPS KAKENHI Grant Number JP24109008).

References

[1] P. S. Engel. Chem. Rev.80, 99 (1980).10.1021/cr60324a001Search in Google Scholar

[2] R. J. Crawford, R. J. Dummel, A. Mishra. J. Am. Chem. Soc.87, 3023 (1965).10.1021/ja01091a053Search in Google Scholar

[3] W. R. Roth, M. Martin. Justis Liebigs Ann. Chem.702, 1 (1967).10.1002/jlac.19677020102Search in Google Scholar

[4] W. R. Roth, M. Martin. Tetrahedron Lett.8, 4695 (1967).10.1016/S0040-4039(01)89583-6Search in Google Scholar

[5] E. L. Allred, R. L. Smith. J. Am. Chem. Soc.89, 7133 (1967).10.1021/ja01002a063Search in Google Scholar

[6] W. Adam, M. Diedering, A. Trofimov. J. Phys. Org. Chem.17, 643 (2004).10.1002/poc.834Search in Google Scholar

[7] W. Adam, M. Grune, M. Diedeiring, A. V. Trofimov. J. Am. Chem. Soc.123, 7109 (2001).10.1021/ja005887dSearch in Google Scholar PubMed

[8] N. Yamamoto, M. Olivucci, P. Celani, F. Bernardi, M. A. Robb. J. Am. Chem. Soc.120, 2391 (1998).10.1021/ja971733vSearch in Google Scholar

[9] A. Sinicropi, C. S. Page, W. Adam, M. Olivucci. J. Am. Chem. Soc.125, 10947 (2003).10.1021/ja0263137Search in Google Scholar PubMed

[10] W. Adam, H. Garcia, M. Diedering, V. Marti, M. Olivucci, E. Palomares. J. Am. Chem. Soc.124, 12192 (2002).10.1021/ja026321nSearch in Google Scholar PubMed

[11] M. Abe, C. Ishihara, S. Kawanami. J. Am. Chem. Soc.127, 10 (2005).10.1021/ja044269kSearch in Google Scholar

[12] M. B. Reyes,B. K. Carpenter. J. Am. Chem. Soc.122, 10163 (2000).10.1021/ja0016809Search in Google Scholar

[13] B. K. Carpenter. Chem. Rev.113, 7265 (2013).10.1021/cr300511uSearch in Google Scholar

[14] M. Abe, W. Adam, M. Hara, M. Hattori, T. Majima, M. Nojima, K. Tachibana, S. Tojo. J. Am. Chem. Soc.124, 6540 (2002).10.1021/ja026301lSearch in Google Scholar

[15] M. Abe, J. Je, M. Mishima. Chem. Soc. Rev.41, 3808 (2012).10.1039/c2cs00005aSearch in Google Scholar

[16] M. Abe. Chem. Rev.113, 7011 (2013).10.1021/cr400056aSearch in Google Scholar

[17] T. Nakagaki, T. Sakai, T. Mizuta, Y. Fujiwara, M. Abe. Chem. Eur. J.19, 10395 (2013).10.1002/chem.201300038Search in Google Scholar

[18] M. Abe, S. Tada, T. Mizuno, K. Yamasaki. J. Phys. Chem. B, 120, 7217 (2016).10.1021/acs.jpcb.6b05342Search in Google Scholar

[19] K. Yamasaki, A. Watanabe. Bull. Chem. Soc. Jpn.70, 89 (1997).10.1246/bcsj.70.89Search in Google Scholar

[20] K. Yamasaki, A. Watanabe, T. Kakuda, I. Tokue. Int. J. Chem. Kinet.30, 47 (1998).10.1002/(SICI)1097-4601(1998)30:1<47::AID-KIN6>3.0.CO;2-USearch in Google Scholar

[21] J. Ye, S. Hatano, M. Abe, R. Kishi, Y. Murata, M. Nakano, W. Adam. Chem. Eur. J.22, 2299 (2016).10.1002/chem.201503975Search in Google Scholar PubMed

Published Online: 2017-02-28
Published in Print: 2017-06-27

©2017 IUPAC & De Gruyter. This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License. For more information, please visit: http://creativecommons.org/licenses/by-nc-nd/4.0/

Downloaded on 31.1.2023 from https://www.degruyter.com/document/doi/10.1515/pac-2016-1215/html
Scroll Up Arrow