Anirban
Maity
,
Kuruva
Balanna
,
Constantin G.
Daniliuc
and
Armido
Studer
*
Organisch-Chemisches Institut, Universität Münster, Corrensstraße 40, 48149 Münster, Germany. E-mail: studer@uni-muenster.de
First published on 25th March 2025
Cyclobutanes are strained carbocycles found in many drugs and natural products. Herein, we report a diastereoselective 1,3-nitrooxygenation of bicyclo[1.1.0]butanes with tert-butylnitrite and TEMPO to access 1,1,3-trisubstituted cyclobutanes. Various bicyclo[1.1.0]butanes effectively participated in the radical reaction yielding the substituted cyclobutane scaffolds with excellent yields and diastereoselectivity. The reaction is catalyst-free, easy to perform, and scalable and can be conducted in open air. Products obtained serve as substrates for the synthesis of 1,1,3,3-tetrasubstituted cyclobutanes with good yields and diastereoselectivity.
![]() | ||
Scheme 1 (a) Cyclobutane ring containing drugs. (b) Photochemical [2 + 2] cycloaddition. (c) Accessing cyclobutanes from BCBs. (d) This work: diastereoselective N,O-1,3-difunctionalization of BCBs. |
Moreover, the Baran group9 developed amination of BCBs using amines as nucleophiles. Biju et al.10 and Feng, Qu, Yu et al.11 independently developed diastereoselective arylation of disubstituted-BCBs. Along with nucleophilic ring-openings, several radical additions1,12 to BCBs were reported, predominantly for mono-functionalization. Considering difunctionalizing ring-opening of BCBs, except for few cases, the majority of these transformations suffer from low diastereoselectivity. Along these lines, diastereoisomerically enriched 1,1,3-trisubstituted cyclobutanes were successfully prepared by the Fox group13 through organocuprate addition and subsequent trapping using an external electrophile. The Glorius group14 developed a 1,3-carbothiolation of BCBs through an intramolecular group transfer process to secure high syn-selectivity. The Hari group15 presented a 1,3-carboheterodifunctionalization of BCBs for the synthesis of spirocyclobutyl lactones and lactams, albeit with moderate diastereocontrol. Boronate-BCBs were investigated by the Aggarwal group16 who achieved highly diastereoselective 1,3-carbodifunctionalization by employing various electrophiles or electrophilic radicals to induce the ring-opening process. The preinstalled boronate functionality controls both regio- and diastereoselectivity of the products, 1,1,3-trisubstituted cyclobutanes.
In addition to mono and 1,3-difunctionalization, cycloaddition17 and insertion reactions1 of BCBs have offered synthetic chemists direct and straightforward routes to access cyclobutane rings containing complex 3D-enriched scaffolds. However, only carbon–carbon (C–C) and carbon–heteroatom (C–S) 1,3-difunctionalization was presented in a diastereoselective manner to date. The diastereoselective heteroatom–heteroatom (X,Y) 1,3-difunctionalization of bicyclo[1.1.0]butanes is rare.18 In order to achieve a heteroatom–heteroatom, particularly (O–N) 1,3-difunctionalization, we considered using tert-butyl nitrite (tBuONO)19 as a nitrogen atom source in combination with the TEMPO radical20 as an oxygen atom source (Scheme 1d). tBuONO upon thermal homolysis will generate a persistent NO-radical, which in the presence of molecular oxygen will be oxidized to a transient NO2-radical. The NO2-radical should then add along the central C–C bond of the BCB to generate a transient adduct cyclobutyl radical, which can engage in a selective cross-coupling with the persistent TEMPO radical to afford the desired 1,3-difunctionalized cyclobutane product. Interestingly, in this reaction two persistent radicals (NO and TEMPO) are present and the persistent NO-radical is selectively converted to a transient NO2-radical during the course of the reaction, while TEMPO remains unchanged. The suggested difunctionalizing BCB ring-opening reaction presents challenges that are product stability21 and diastereoselectivity control, as radical/radical cross couplings generally occur with very low barriers.
# | Solvent | Temperature | dr | Yielda,b (%) |
---|---|---|---|---|
a Reaction conditions: 1 (0.1 mmol), tBuONO (0.20 mmol), TEMPO (0.15 mmol), solvent (0.1 M), air, 18 h. b Determined by 1H NMR using 1,3,5-trimethoxy benzene as an internal standard. c After removal of solvent the crude was re-dissolved in CHCl3 (0.1 M) in the presence of the given base (20 mol%) and stirred for 18 h at rt. d Isolated yield and diastereomeric ratio after silica gel column chromatography of the reaction mixture. | ||||
1 | DMSO | 70 °C | 2.2![]() ![]() |
38 |
2 | CHCl 3 | 70 °C |
1.4![]() ![]() |
82 |
3 | DCE | 70 °C | 1.4![]() ![]() |
76 |
4 | Toluene | 70 °C | 1.2![]() ![]() |
48 |
5 | 1,4-Dioxane | 70 °C | 1.1![]() ![]() |
64 |
6 | CH3CN | 70 °C | 1.0![]() ![]() |
74 |
7 | CHCl3 | 25 °C | 1.7![]() ![]() |
63 |
8 | CHCl3 | 100 °C | 3.9![]() ![]() |
81 |
9c | CHCl3 (DMAP) | 70 °C | 4.4![]() ![]() |
82 |
10c | CHCl3 (DBU) | 70 °C | 5.8![]() ![]() |
61 |
11c | CHCl3 (Cs2CO3) | 70 °C | 5.3![]() ![]() |
81 |
12 | CHCl 3 | 70 °C |
>20![]() ![]() |
74d |
The optimization campaign was continued by varying reaction temperature using CHCl3 as the solvent. At room temperature a significantly lower yield (63%) resulted, possibly due to the slow decomposition of tBuONO (Table 1, entry 7). Upon increasing temperature from 70 °C to 100 °C the yield of the reaction remained unchanged (81%); however, an improved diastereoselectivity was noted. This surprising stereochemical outcome indicated that isomerization likely occurred after initial BCB-nitrooxygenation, probably through deprotonation/re-protonation of the acidic proton next to the nitro group. Despite the fact that reversible alkoxyamine C–O-bond homolysis in 2 could not be excluded as the isomerization process at that temperature, we tested whether the poor diastereomeric ratio can be improved by applying a deprotonation/re-protonation strategy through simple addition of a base. To this end, several bases were tested as catalysts for the isomerization of 2. After radical nitrooxygenation, the solvent was removed and the crude nitro-cyclobutane 2 was then re-dissolved in CHCl3 in the presence of DMAP (20 mol%) and stirred for 18 h at room temperature. The diastereoselectivity improved without compromising the yield (Table 1, entry 9, 82%, dr 4.4:
1). With DBU (20 mol%) as the base, the diastereomeric ratio further improved, albeit at the expense of the NMR yield (Table 1, entry 10, 61%, dr 5.8
:
1). In the presence of Cs2CO3 (20 mol%), product 2 was obtained with a diastereomeric ratio of 5.3
:
1 without scarifying the yield (Table 1, entry 11). Pleasingly, upon purification of the crude product 2 through silica gel column chromatography without any prior base treatment, we found complete isomerization and 2 (syn-isomer) was isolated in 74% yield with an excellent diastereomeric ratio (dr > 20
:
1, Table 1, entry 12). The relative configuration was unambiguously assigned by X-ray analysis and other compounds prepared in this study were assigned in analogy. Obviously, the additional isomerization step using an external base is not required, as isomerization to the thermodynamic syn-product efficiently occurs during silica gel chromatography. With these optimized conditions in hand, we then investigated the scope of the reaction.
First, the aryl group of the keto BCB was varied. 1-Naphthyl and 6-methoxy-substituted 2-naphthyl-keto-BCBs afforded the desired products 3 and 4 in 73% and 86% yield with high diastereoselectivity (dr 20:
1 and 12
:
1) (Scheme 2). The unsubstituted phenyl keto-BCB provided the desired cyclobutane 5 in 72% yield and high stereoselectivity (dr 13
:
1). Aryl keto-BCBs carrying an electron donating or withdrawing group at the aryl ring were eligible substrates. Thus, BCBs carrying an aryl group or an alkyl group at the 4-position of the aryl substituent such as 4-phenyl, 4-methyl, or 4-tert-butyl, afforded the desired products 6–8 with good to excellent yields (75–89%) and good to excellent diastereoselectivity (dr 8
:
1 to > 20
:
1). The reason for the varying selectivity as a function of the para-substituent is not fully understood. Along the same lines, electron-rich 4-MeOC6H4 and electron-poor 4-CF3C6H4-substituted BCBs afforded the desired products 9 and 10 in 76% and 83% yield, with good diastereoselectivity (dr 8
:
1 and 10
:
1). Renewed chromatography of 9 led to a slightly improved diastereoselectivity (11
:
1 versus 8
:
1). 4-Halo-substituted aryl keto-BCBs delivered the targeted 1,3-nitrooxygenated products 11–13 in good to excellent yields (80–94%) and good diastereoselectivity (dr 8
:
1 to 16
:
1). Installing a substituent at the meta position of the aryl ring such as in 3,5-dimethyl-C6H3, 3,5-dimethoxy-C6H3, and 3-fluoro-C6H4-substituted keto-BCBs afforded the desired products 14–16 in good to excellent yields (74–92%) and good to excellent diastereoselectivity (dr 8
:
1 to dr > 20
:
1). ortho-Methyl substituted phenyl keto-BCB delivered the 1,3-nitrooxygenated product 17 in 76% yield with high diastereoselectivity (dr 17
:
1). A 2,4-disubstituted phenyl keto-BCB worked well to afford 18 in 80% yield with good diastereoselectivity (dr 10
:
1). Heterocyclic rings such as thiophene and furan are tolerated and the corresponding 1,3-nitrooxygenated products 19 and 20 were isolated in 79% and 88% yields with high stereoselectivity (dr 16
:
1 and 11
:
1). However, a 1,3-disubstituted-BCB did not afford the desired 1,3-nitrooxygenated product 21, possibly because of steric reasons: the 3-phenyl substituent in the BCB hinders NO2-radical attack and ring opening does not occur.
Moreover, phenylsulfonyl-activated BCBs also engage in the radical cascade reaction, as shown by the successful synthesis of sulfone 23 that was isolated in 70% yield with excellent diastereoselectivity (dr > 20:
1). Notably, this methodology is not restricted to aromatic keto-BCBs. It is also effective for the ring-opening di-functionalization of ester-BCBs, amide-BCBs, and alkyl keto-BCBs, yielding trisubstituted cyclobutane derivatives with good to excellent yields and low to moderate diastereoselectivity (22, 87%, dr 7
:
1; 24, 81%, dr 1
:
1; 25, 74%, dr 5
:
1; 26, 57%, dr 5
:
1).
The synthetic value of the nitrocyclobutane products was documented by conducting several follow-up reactions (Scheme 3). Michael reaction of 2 with methyl acrylate and acrylonitrile afforded the 1,1,3,3-tetrasubstituted cyclobutanes 27 and 28 in 99% and 76% yields with good to excellent diastereoselectivity (dr 7:
1 and dr > 20
:
1). Palladium catalysed allylation of 2 with allyl alcohol was achieved to afford 29 in 68% yield with moderate diastereoselectivity (dr 5.2
:
1). The relative configuration of the separable isomers was assigned by NOE experiments (see the ESI†). Selectivity of the other products derived from 2 was assigned in analogy. para-Quinomethide as an electrophilic coupling partner delivered cyclobutane 30 in 78% yield, albeit with poor diastereoselectivity. Thermal homolysis of the weak C–O bond of alkoxyamine 2 in the presence of γ-terpene as a hydrogen atom transfer (HAT)22 reagent afforded the 1,3-disubstituted cyclobutane 31 in 44% yield (88%, brsm) and moderate diastereoselectivity. Hydrazination of nitrocyclobutane 2 under basic conditions gave cyclobutene 32 in 48% isolated yield. Hence, the targeted nitrocyclobutane obtained after initial hydrazination eliminates HNO2 under the applied conditions.
In order to study the effect of O2, the reaction of 1 was repeated under an argon atmosphere and only traces of product 2 were detected (Scheme 4a). Based on this observation and literature reports19 we propose a plausible mechanism that is depicted in Scheme 4b. Thermal homolysis of the N–O bond in tBuONO leads to the formation of the tert-butoxyl radical along with the persistent nitroso (NO) radical, which in the presence of molecular oxygen is directly oxidized to the NO2-radical. The NO2-radical then engages in a homolytic substitution at carbon cleaving the central C–C bond of the bicyclo[1.1.0]butane to generate the adduct radical 33. This adduct radical is then trapped by the persistent TEMPO radical23 to deliver the isolated product. Of note, the TEMPO-trapping does not occur with a high stereoselectivity and isomerization through a deprotonation/re-protonation sequence forming a thermodynamically more stable isomer during silica gel chromatography and ensures the very high diastereoselectivity observed in many of our transformations.
Footnote |
† Electronic supplementary information (ESI) available. CCDC 2411255. For ESI and crystallographic data in CIF or other electronic format see DOI: https://doi.org/10.1039/d4sc08753d |
This journal is © The Royal Society of Chemistry 2025 |