valence isomerization between diazo compounds and …szolcsanyi/education/files/organicka chemia...

MICROREVIEW

DOI: 10.1002/ejoc.201100224

Valence Isomerization between Diazo Compounds and Diazirines

Sergei M. Korneev*[a][‡]

Keywords: Diazirines / Small ring systems / Diazo compounds / Isomerization / Photochemistry / Thermochemistry

Diazo compounds and diazirines have been separatelystudied in detail and are widely used in research and for ap-plied purposes. Although the mutual isomerization of thesecompounds was discovered 50 years ago, it has still beenonly modestly investigated. Meanwhile, this type of mutualtransformation has to be taken into account on every oc-

1. Introduction

For a long time, diazo compounds of type 1 and di-azirines 2 (Figure 1) served as an interesting example of achemical misinterpretation, because chemists, while drawingthe structure of one, implied the other, not knowing thatboth types of compounds can exist simultaneously.

After the discovery of the diazo compounds their struc-tures became the subject of prolonged discussion. Origi-nally cyclic structures of type 2 were attributed to diazoace-tic acid derivatives by Curtius[1] and to diazomethane byPechmann.[2] Later, Angeli,[3] Thiele,[4] and Langmuir[5] pro-vided evidence supporting the linear form 1. As a result ofthe scientific discussion the two classes of compounds be-came connected. As was discovered later, this was not acci-dental.

[a] Department of Organic Chemistry, St. Petersburg StateUniversity,University prosp. 26, St. Petersburg, 198504, RussiaE-mail: [email protected]

[‡] Current address: Institute of Chemistry, University ofOsnabrück,Barbarastr. 7, 49076 Osnabrück, GermanyFax: +49-541-969-2370E-mail: [email protected]

Sergei M. Korneev was born in Leningrad (currently St. Petersburg), Russia. He gained his Ph.D. in Organic Chemistryfrom Leningrad State University, Russia, under the supervision of Prof. I. Korobitsyna. He served as a postdoctoral fellowin the laboratory of Prof. D. Kaufmann at the Institute of Chemistry, Technical University of Clausthal-Zellerfeld, Ger-many. After working in the Chemistry Departments of St. Petersburg State University under Prof. V. Nikolaev, Kaiserslau-tern University (Prof. M. Regitz), and Göttingen University (Prof. A. de Meijere), he continued his research at theUniversity of Osnabrück. Current interests include synthetic application of diazo compounds, photochemistry, and molecu-lar design.

Eur. J. Org. Chem. 2011, 6153–6175 © 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 6153

casion when one of these compounds is used or investigated.This review discusses all main cases of mutual isomerizationbetween diazo compounds and diazirines with the goal ofproviding an impulse for its rational application, as well asdescribing the isomerism.

Figure 1. General structures of diazo compounds 1 and diazirines2.

Further investigations drew an exact line between the lin-ear and the cyclic compounds. The linear structures of di-azo compounds were confirmed in different ways: by elec-tronic diffraction in 1934,[6] by IR spectroscopy in 1949,[7]

by isotope labeling in 1957,[8] and by microwave spec-troscopy in 1958.[9]

The first diazirines were synthesized in 1960–1961 byPaulsen[10] and Schmitz[11] and their cyclic structures wereunambiguously confirmed in 1962 by both physical[12] andchemical methods.[13] The last point of doubt relating to thestructures and the existence of these types of compoundshad thus been laid to rest. It was shown that the linearisomers 1[14] and the cyclic isomers 2[15] not only differ intheir physical characteristics, but also have little or nothingin common in their chemical properties, except the tendencyto abstract nitrogen and mutual isomerization. The sim-plicity of thermal or photochemical elimination of nitrogen

S. M. KorneevMICROREVIEWboth from diazo compounds and from diazirines allows fora limited possibility of reversible valence isomerization.

2. Isomerization of Diazirines into DiazoCompounds

The chemistry of diazirines has a relatively short history.The time of their discovery coincided with the period ofintensive research into carbenes, so it is not surprising thatthe influence of light on diazirines, with the goal of generat-ing carbenes, was among the first topics to be studied. Thepresence of diazo isomers among other products after pho-tolysis was thus established practically at once. Several pro-cesses compete with possible isomerization in the course ofdiazirine decay (on photo- or thermoactivation): namely,fluorescence, internal conversion, intersystem crossing, car-bon–nitrogen bond cleavage followed by intramolecularprocesses (e.g., coupling with loss of nitrogen, formationof alkenes through 1,2-hydride migration), intermolecularreactions (e.g., with solvent or other diazirines), and simplenitrogen extrusion to generate the carbene intermediates. Ithas been shown that the substituents have a significant rolein determining whether or not linear diazo isomers areformed.

2.1. Diazirine

Both diazirine and diazomethane are poorly stable gasesunder ambient conditions and are therefore used in dilutedform. Explosions of sufficiently concentrated compoundson warming or on contact with metal surfaces have beenrepeatedly encountered.[16] Photolysis of diazirine gave riseto some diazomethane, but thermolysis has not been re-ported.

Primary experiments with photochemical decompositionof diazirine (3, Scheme 1) in the gas phase (in admixturewith N2, λ = 320 nm) showed that one of the initial prod-ucts of the reaction is diazomethane (8, confirmed by UVabsorption at 213, 217, 229, 420 nm), with a quantum yieldof formation close to 0.2.[17] Its levels increased at the begin-ning of the reaction and decreased along with the consump-tion of diazirine, due to secondary photolysis. It is interest-ing that the nearest homologues of diazirine – methyl-,ethyl-, and dimethyldiazirine – do not form stable diazo

Scheme 1. Photolysis of labeled and unlabeled diazomethanes anddiazirine in matrixes.

www.eurjoc.org © 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim Eur. J. Org. Chem. 2011, 6153–61756154

compounds in the gas phase, but instead lose their excessenergy through intramolecular reactions, with nitrogen eli-mination and rearrangements.[18] Irradiation of the same di-azirines in solution resulted in the formation of diazo com-pounds in yields of up to 40%.

It has thus been shown that both the structure of thestarting diazirine and the environment influence the stabili-zation of excited molecules and hence the reaction products.

Irradiation of diazirine (3) in an isotopically labeled ni-trogen matrix at 20 K (Scheme 1) is a first-order reactionthat leads to the formation of the diazomethanes 6–8, lab-eled in the cases of 6 and 7 with the nitrogen from the ma-trix, as shown by IR spectroscopy.[19] 15N-Labeled diazirine(5) is not formed. It was concluded that the formation ofdiazomethane in the matrix occurs in two steps: i. moleculeexcitation causes nitrogen elimination, forming carbene (4),which ii. reacts with a neighboring nitrogen molecule to af-ford the diazomethanes 6–8.

The photolysis of diazomethane (8) under the same con-ditions did not result in the formation of diazirine (5). How-ever, upon irradiation of the 15N-labeled diazomethane 6,redistribution of the label with the formation of approxi-mately equal amounts of the two labeled diazomethanes 6and 7 was observed, along with the formation of unlabeleddiazomethane (8), through incorporation of nitrogen fromthe unlabeled matrix. Such a product distribution confirmsthe intermediacy of carbene (4), formed in the course ofthe transformations of the labeled diazomethane 6 in thematrix.

Two thermal reactions could occur in the ground state(S0) of diazirine (3): 1) isomerization to diazomethane, and2) decomposition to give singlet carbene and molecular ni-trogen. The computed barrier energy heights clearly testifythat methylene formation should be the priority processwith respect to isomerization,[20] however this predictionneeds to be checked.

2.2. Alkyldiazirines

The alkyldiazirines 2 (Scheme 2) are stable and readilyavailable chemicals, whereas the isomeric diazoalkanes 1 areusually volatile and labile compounds, the occurrence ofwhich during the courses of reactions of alkyldiazirines hasthus been confirmed only by use of special methods. Spec-troscopy (IR, UV) has served as a powerful method to es-tablish the fact of isomerization during low-temperaturephotolysis. During thermolysis, however, no diazo com-pounds have been observed. Alternative trapping experi-

Scheme 2. Photolysis of alkyldiazirines.

Isomerization between Diazo Compounds and Diazirines

ments have been suggested and proved highly efficient indemonstrating high yields of diazo isomers both under pho-tolysis and under thermolysis.

Photolysis of the diazirine 2a in a N2 matrix (312 nm,10 K, 12 h) led to the formation of the carbene 9a and insig-nificant quantities of the diazoalkane 1a (IR: 2020 cm–1).[21]

Similarly, photolysis of the diazirine 2b in an Ar matrix(λ � 200 nm, 12 K) gave the diazomethane 1b (IR:2135 cm–1), which was stable under the irradiation condi-tions, together with the carbene 9b as primary products.[22]

Dependence of the product distribution on the wave-length of the initiating light was observed in the case of thediazirine 2c in a N2 matrix. Short-wavelength irradiation(334 nm, 6 K, 28 h) produced two main products: the diazo-methane 1c (IR: 2053 cm–1) and the carbene 9c (IR:802 cm–1). Use of longer-wavelength irradiation (475 nm,1 h) resulted in complete consumption of the carbene 9cand gave stable products, but the diazomethane 1c persistedunder irradiation with this light. It is interesting that irradi-ation of the diazirine 2c in an Ar matrix – unlike in N2 –at 20 K drove the reaction to the side of the carbene 9c onlyand that no diazomethane 1c was observed.[23]

Upon partial photolysis of the diazirines 2d[24]

(Scheme 2) and 10a (Scheme 3, IR: 1582 cm–1)[25] with long-wavelength light (350 nm, CH2Cl2, 30 min) an intense pinkcolor (λmax ≈ 500 nm) attributed to the corresponding diazocompounds 1d and 11a (IR: 2040 cm–1) appeared. Qualita-tive treatment with acrylonitrile (0.1 m) showed that the di-azoalkane 1d (half-life under ambient conditions was about1 h) disappeared immediately on addition of the reagent.One might expect that trapping of 1d in such a way couldallow estimation of the quantity present.

Scheme 3. Photolysis and thermolysis of dialkyldiazirines.

An attempt to obtain the diazoalkane 11a (Scheme 3) bythermolysis (118 °C, 1.5 h, sealed tube) from the startingdiazirine 10a failed,[25] not entirely surprisingly, because thehalf-life of 11a at ambient temperature is only about 2 min.

The products both of photolysis (350 nm) and of ther-molysis (118 °C) of the diazirine 10b indicate that some di-azo compound 11b must be formed in both processes, butno 11b was itself detected (Scheme 3).[25]

Eur. J. Org. Chem. 2011, 6153–6175 © 2011 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim www.eurjoc.org 6155

Photolysis of the diazirines 10c and 10d (Scheme 3) in thepresence of N-phenylmaleimide resulted in the formation ofthe bicyclic pyrazolines 12c and 12d. The same results wereobtained on thermal decomposition of the diazirines 10cand 10d.[26] The intermediacy of the diazo isomers 11c and11d was thus revealed by means of the trap and N-phenyl-maleimide appears to be a suitable partner for linkage ofdiazo compounds through 1,3-dipolar cycloaddition reac-tions.[27]

Unexpected results were obtained in attempts to designphotoactivatable phospholipids to undergo cross-linkingwith proteins on photoexcitation.[28,29] The (fluoroalkyl)di-azirines 13a–c were tested for this purpose (Scheme 4). Itwas proposed that the carbene intermediates could not bestabilized by intramolecular F– shifts analogously with 3-(trifluoromethyl)-3-phenyldiazirine[30] and should give notolefins, as in the case of H– shifts, but products of intermo-lecular cross-linking.

Scheme 4. Photolysis of (fluoroalkyl)diazirines.

On irradiation of the diazirine 13a (IR: 1600 cm–1) withshort-wavelength light in solution (cyclohexane, cyclohex-ene, or MeOH), the isomeric diazoalkane 14a (IR:2120 cm–1) was obtained in a high yield together with prod-ucts of nitrogen abstraction: the cyclopropane 15a (intra-molecular C–H insertion) and the olefin 16a (radical mi-gration).[28,29] No intermolecular carbene reactions weretherefore observed. The presence of a small amount ofAcOH in cyclohexane solution did not change the productyields (14a was observed to be stable in the presence of lowconcentrations of the acid), but in pure AcOH the diazo-alkane was not detected.

Similarly, irradiation of the monosubstituted diazirine13b (Scheme 4) gave the products both of isomerization (thediazoalkane 14b, IR: 2120 cm–1) and also of nitrogen elimi-nation (the cyclopropane 15b and the olefin 16b). The per-halodiazirine 13c isomerized into the corresponding diazo-methane 14c (IR: 2120 cm–1).

It was thus established that: i. the α,α-perfluorodiazirines13 and the diazoalkanes 14 are stable under ambient condi-tions, and ii. all the corresponding carbenes are stabilizedwith respect to intramolecular reactions (migration of aradical or insertion in β-C–H bond) in the absence of α-protons.

In the thermolysis of the vinyldiazirine 17 (Scheme 5) at200 °C, nitrogen elimination levels of only 7 % were ob-served, together with the formation of the pyrazole 18(about 90%).[16c] The same pyrazole 18 was also obtainedafter stirring of 17 at room temperature for a week. The

S. M. KorneevMICROREVIEWvinyl diazomethane 19 was proposed as one of the possibleintermediates of this reaction, because it forms the samepyrazole 18 on standing at room temperature for 2 days.

Scheme 5. Thermolysis of the vinyldiazirine 17.

2.3. Spirodiazirines

Spirodiazirines are close analogues of alkyldiazirineswith very similar stabilities, availabilities, and properties. Inaddition, though, they also display some subtle features dueto steric effects that allow them to be considered separatelyfrom the alkyldiazirine family. Like their open-chain ana-logues, cyclic diazoalkanes are labile and chemically activecompounds. Spirodiazirines undergo isomerization to diazocompounds in moderate yields (30–60%) under photolysisor thermolysis conditions. The colors of the reaction mix-tures, UV and IR spectroscopy, and chemical traps wereused to visualize the formation of diazo isomers, with fuller-ene serving as the most powerful agent for detection. Isom-erization was not observed on photolysis of diazirines in thegas phase, nor in complexes with zeolite and cyclodextrins.

Photoactivation of the spirodiazirines 20c and 20d(Scheme 6), with attached cycles of various sizes (n = 6, 7),in the forms of liquid films resulted in a change in the initialyellow color of the films to a reddish one that remainedafter termination of photolysis (Pyrex filter).[31] IR spectrain each case showed a new absorption at 2060 cm–1 thatcorresponded to the diazoalkane absorption region[32] andsuggested the conversion of the diazirines 20c and 20d intothe diazo compounds 21c and 21d.

Scheme 6. Photolysis of spirodiazirines.

Qualitatively, the band corresponding to the diazo com-pound 21b disappeared within several minutes after ter-mination of the photolysis of the diazirine 20b (n = 5). Forthe larger-ring compounds 20c and 20d (n = 6, 7), the life-times of the diazo compounds 21c and 21d were severaltimes longer. UV spectra showed that 30–50% of the diazir-ines 20a–d were converted into the diazoalkanes 21a–d.Furthermore, in the course of photolysis of the diazirines20a–d the formation of the azines 22a–d, which might beformed from the corresponding diazo compounds, was ob-


served. In view of these two factors, the levels of conversionof the diazirines 20 into the diazoalkanes 21 are thus (ap-proximately) 20 % for 21a (n = 4), 36% for 21b (n = 5), 59%for 21c (n = 6), and 70 % for 21d (n = 7).[18,31] However,there is an opinion that the formation of azines might notbe due solely to diazo precursors.[33–35] It should be notedthat photolysis of the diazirine 20b (n = 5) in the gas phaseat pressures above 20 mm gave cyclohexene (25b) and nitro-gen as main products, without the diazoalkane 21b beingdetected.[18]

A quantitative estimation of the isomerization processwas obtained by photolysis of the diazirines 20a–d in thepresence of acetic acid. It is known that diazirines are inertto acids whereas diazoalkanes react rapidly with them. Theyields of the acetates 24, interaction products of the diazo-alkanes 21 and AcOH, on photolysis were 40% for 24a (n= 4), 39 % 24b (n = 5), 34% for 24c (n = 6), and 32% for 24d(n = 7), together with the hydrocarbons 25a–d and otherproducts. Photolysis of the diazirine 20b (n = 5) in deuter-ioacetic acid showed that cyclohexene (25b, 20 %) with adeuterium label in the vinylic position was formed alongwith the acetate 24b (39%). This means that both productshad their origins in the interaction between the diazoalkane21b and AcOD. The degree of isomerization of the diazirine20b to the diazoalkane 21b on photolysis is thus about59%.[31,33]

No diazo isomers were detected in the course of photoly-sis of the diazirines 20c (n = 6) and 20d (n = 7), entrappedwithin α-, β- and γ-cyclodextrin (not shown) in the solidstate, and only the formation of nitrogen elimination prod-ucts (main) and azines (minor) was noted. Whereas the for-mation of azines was almost completely prevented in the α-and γ-cyclodextrin complexes, in β-cyclodextrin complexesthey were produced in moderate yields (18–42 %).[35,36] Co-ordination of the diazirines with the cyclodextrins thus sup-presses their ability to isomerize to their diazo counterparts,with nitrogen elimination becoming the main process.

The diazo group is a classical dipole, and so Huisgenreactions (1,3-dipolar cycloadditions) have been used todemonstrate the potential for visualization of labile diazoisomers by trapping with dipolarophiles.[37] Photolysis ofthe diazirine 20b with dimethyl acetylenedicarboxylate intoluene solution (Scheme 7) gave the diazaspirodecadiene26 (18–22 %), provided that the photolysis of 26 itself wassuppressed with a suitable optical filter, such as aq.NiCl2.[38] If photolysis was carried out without the filter thespiro ester 27 was formed in a high yield, because 26 is

Scheme 7. Photolysis of spirodiazirines.


unstable under these conditions. Photoirradiation of the di-azirine 20b in the presence of the spiro ester 27 in combina-tion with the NiCl2 filter also induced trapping of the inter-mediate 21b in high yield because 27 is also a good di-polarophile.

The yield of diazirine-to-diazo isomerization for 20b ob-tained in the cycloaddition reaction (22%) thus appearedto be considerably lower than those calculated from the re-action by IR monitoring (36%) or observed in the reactionwith acetic acid as a trap (59 %). The significant differencein yields between cycloaddition and AcOH decompositionis obviously the result of a high thermal instability of 21brelative to its speed of cycloaddition.

The formation of a diazo compound was not affectedsignificantly by the introduction of a methyl group α to thediazirine moiety. Qualitatively it was demonstrated that apentane solution of the diazirine 20e (IR: 2939 cm–1) turnedpeach-colored during photolysis for a few hours, which wasascribed to the formation of the linear diazo compound 21e(IR: 2025 cm–1).[39]

Deuterium labels are widely used for investigation of re-action mechanisms. Analysis of the distribution of labels inproducts obtained from the photolysis of the spirobicyclicdiazirine 28a (Scheme 8) in MeOD thus identified about23% deuteration in the nortricyclene 33.[40] This result re-flects the proportion of diazoalkane participation in theformation of 33, because the carbene 30 will also form 33by intramolecular insertion and in this case no deuteriumlabels will be incorporated. Protonation of the intermediatediazoalkane 29a by the solvent gave rise to diazonium ions(not shown), followed by reaction with MeOD resulting inthe formation of the ethers 31 and 32 with C1 and C2 deute-rium labels. Trapping experiments showed that the etherswere formed from the diazo compound. Addition of diethylfumarate as a scavenger to 29a in MeOH resulted in a dropin the relative yield of 31 + 32 from 42 % to 13%, whereasin the presence of fumaronitrile only traces of ethers wereobtained, which unequivocally identifies their origin.

Scheme 8. Photolysis of spirodiazirines in methanol.

Moreover, photolysis of 28a in a MeOD/fumaronitrilemixture afforded 33 virtually free of deuterium (�0.02 D),so in this case only the carbene 30a was the precursor for33. The azine 34 was obtained as an additional product ofthe photolysis. Its quantity increased (from 6% to 32%)with the concentration of 28a (from 70 to 490 mmolL–1)whereas no significant changes in the product distribution


of 31–33 were observed. It appears that the immediate pre-cursors of 31–33 are not involved in the formation of 34.

As a result, these experiments testify that the unobserv-able diazoalkanes 29a and 29b participated in the photoly-sis of diazirines 28a and 28b to large extents [29a (R = H):62 %, 29b (R = Me): 80%].

A low yield of the scavenger product 37 (Scheme 9) wasobtained upon irradiation of the diazirine 35 in the presenceof diethyl fumarate.[41] On one hand this could point to theformation of the diazo isomer 36. On the other hand, how-ever, the low yield of 37 could be an indication both of aninsignificant isomerization step and of inefficient activity ofdiethyl fumarate in the diazo dipole trapping.

Scheme 9. Photolysis of a spirodiazirine in the presence of diethylfumarate.

The initial concentration of the diazirine solution wasshown to influence the product ratio.[42] Photolysis of a di-lute solution of the diazirine 38 with a few laser shots(Scheme 10) gave two new absorption bands: the first oneat 234 nm is due to the diazo compound 39 and the secondone at 214 nm mainly due to the azine 40. In a more con-centrated solution of the diazirine 38 the yield of the diazocompound 39 was unchanged, but that of the azine 40 hadsignificantly increased. Incomplete correlation between az-ine formation and disappearance of diazo compound is ob-vious in this case and attests that the mechanism of thediazirine formation does not only involve simple dimeriza-tion of diazo molecules.[35]

Scheme 10. Laser photolysis of a spirodiazirine.

The rearrangement of the spiro[adamantane-2,3�-diazir-ine] derivative 41a to the diazoadamantane 42a (Scheme 11)could be monitored by FT-IR spectroscopy both in solution(n-heptane) and in KBr (pellet).[43] After only a few secondsof photolysis (Hanovia, 450 W) at room temperature astrong band corresponding to the diazo compound 42a (IR:2033 cm–1) became observable.[44] The primary photoprod-ucts of the diazirine 41a – diazoadamantane (42a) and 2-adamantylidene (43a) – were formed with quantum yieldsof about 0.5 each. It is likely that the excited singlet stateof 41a, which has a lifetime of 0.24 ns, is their direct precur-sor.[45]

S. M. KorneevMICROREVIEW

Scheme 11. Photolysis of adamantane-derived diazirines.

Interestingly, when the diazirine 41a was incorporated inzeolite inclusion complexes the formation of the diazo com-pound 42a was in several cases totally suppressed and onlythe azine 44a (up to 12%) could be detected upon photoly-sis.[46] Nitrogen elimination had thus become the main pro-cess, as a result of coordination between the diazirine moi-ety and the internal surface of a zeolite hole, the same asin the case of cyclodextrins.[36]

Fumaronitrile served as an effective scavenging agent forthe intermediate diazoadamantanes 42a and 42b (λmax =330 nm) formed on photodecomposition of the diazirines41a and 41b (Scheme 11) in aprotic solvents, driving thereactions to the corresponding pyrazolines 45a and 45b.[47]

The yields of the scavenging products agree well with otherresults of photolysis in which 60 % of the diazirine 41a re-acted through the intermediacy of the diazo compound 42aand the other 40% decomposed directly to 2-adamantanyl-idene (43a).[48,49]

In contrast, a thermal experiment (140 °C, 1 h) showedthe preferential decomposition of the diazirine 41a to thecarbene 43a and products of its further transformations,with no diazoadamantane (42a) or trapping product 45abeing detectable.[47] Most likely this negative observation isthe result of insufficient stability of both 42a and 45a underthe reaction conditions.

Thermolysis of the diazirine 41a in the presence of C60

at 150 °C (dichlorobenzene, Pyrex tube in darkness) re-sulted in the formation of the corresponding adducts C60Adas a mixture of two isomers – the fulleroid 46 and the meth-anofullerene 47 (ratio 35:65, Scheme 12) – in 95% yieldwith no azine being detectable.[50] Irradiation of a benzene

Scheme 12. Decomposition of the adamantane-derived diazirine41a in the presence of C60.


solution of 41a and C60 (Pyrex tube, 15 °C) gave the sameadducts C60Ad in 80% yield with the ratio of the two iso-mers (the fulleroid 46 and the methanofullerene 47) as51:49.[49] This reflects the formation ratio of the diazo com-pound and the carbene during the reaction[51,52] and is con-sistent with the results obtained under other condi-tions.[48,49] Fullerene can thus serve as more powerful trapthan fumaronitrile for thermolysis study.

2.4. Halodiazirines

Diazirines containing halogen atoms (F, Cl, Br) are rea-sonably stable and accessible compounds.[15m] In contrast,α-halo-substituted diazoalkanes are extremely unstableeven at low temperatures, which makes their detection veryhard.[53] There are only a few positive examples in which thecorresponding diazo compounds have been detected eitherspectroscopically or chemically in the courses of photo-chemical reactions. Sometimes specific stabilization of diazocompounds by one particular halogen and not by the othersoccurs. Thermal isomerization is currently unknown.

Irradiation of the diazirine 2e (Scheme 13), matrix-iso-lated in argon (320–380 nm, 10 K), led to a small IR peakat 2050 cm–1, which might be due to chloro(phenyl)diazo-methane (1e).[54] However, generation of the α-halogenodi-azo compounds 1f from the 3-aryl-3-halodiazirines 2f wasnot confirmed by kinetic measurements or trapping experi-ments in photolysis at room temperature,[55] nor on ther-molysis.[56,57] Modern techniques allow the observation ofthe formation of short-lived molecules even under ambientconditions, so in ultrafast laser flash photolysis (UFP,270 nm) of the diazirine 2e at room temperature in CHCl3two broad bands in the IR spectrum were observed withina few picoseconds of the laser pulse.[58] These did not decaywithin the detection window (3 ns) and were assigned to thediazomethane 1e (2040 cm–1) and to the singlet carbene 9e(1583 cm–1).

Scheme 13. Reactions of halodiazirines.

Irradiation both of 3-benzyl-3-chlorodiazirine (2g) andof 3-benzyl-3-bromodiazirine (2h), matrix-isolated in argon(λ = 380 nm, 10 K), produced traces of (benzyl)chlorodi-azomethane (1g, IR: 2048 cm–1) and (benzyl)bromodiazo-methane (1h, IR: 2045 cm–1), respectively. No diazo bandfor 1i (at 2100–2000 cm–1) was detected on photolysis ofmethylchlorodiazirine (2i) in the matrix.[59] The same is truefor 3-chloro-3-methoxydiazirine (2j), which shows no ab-sorption due to the diazo compound 1j at 1900–2300 cm–1

on irradiation (λ � 340 nm, Ar, 10 K).[56]


Laser-flash photolysis of the diazirines 2k and 2m(Scheme 13) in isooctane at 25 °C produced transient UVabsorptions in the 230–250 nm range, decaying with “life-times” of ca. 0.5 s and assigned to the corresponding diazo-methanes 1k and 1m, due to their spectral similarity withthe absorption of diazomethane (8, λmax = 220 nm).[48] Atlower temperature (–30 °C) the lifetimes increased to severalminutes and quantum yields of formation were measuredas approximately 10–13%.

The fluorodiazomethane derivative 49e (Scheme 14) wasidentified by its IR absorption (2029 cm–1) as the primaryproduct of photoreaction of the fluorodiazirine 48a in anAr matrix.[61] Further irradiation of the matrix showed thetotal disappearance of diazo and azido groups and theproduct was identified from its IR spectrum and with theaid of calculations as the carbenonitrene 50a in a basic trip-let state. This was followed by stabilization to 51a. It isinteresting that under similar conditions the chlorodiazirine48b and the bromodiazirine 48c showed no tendency to iso-merize to linear diazo compounds of type 49 and that theazidocarbenes 53be and 53ce were instead identified as in-termediates of these reactions, followed by the carbenonitr-enes 50b and 50c.

Scheme 14. Photolysis of halodiazirines in an Ar matrix.

Remarkably, the monosubstituted diazirine 48d (X = H)was converted under the same conditions into the diazo-nitrene 55, and with further stabilization into the product51d.[61]

A similar pattern was observed for the p-isomers 52a–c (Scheme 14).[62] It appeared that the fluorodiazomethanederivative 49f (IR: 2023 cm–1) was formed in the earlystages of irradiation of the fluorodiazirine 52a in an Ar ma-trix (13 K, 1 min), followed by fast (9 min) decompositionto the diradical 54a. In the cases of the 3-chlorodiazirine52b and the 3-bromodiazirine 52c, no diazo isomers of type49 were observed, but the azidocarbenes 53bf and 53cf wereformed. These eliminated nitrogen and transformed into thediradicals 54b and 54c and further stable molecules.

The gas-phase thermolysis of difluorodiazirine (56,Scheme 15) yielded the azine 59, tetrafluoroethylene, hexa-fluorocyclopropane, and other products.[33] The same azine59 was formed on photolysis of 56 in an Ar matrix (4 K).It was suggested that nitrogen elimination takes place notfrom the diazirine 56, but after its isomerization to difluoro-diazomethane (57). A bimolecular reaction between the


electrophilic difluorocarbene (58) and either difluorodiazir-ine (56) or its linear isomer 57 formed by photoisomeriza-tion has been postulated as a possible mechanism.[63] How-ever, no clear confirmation of the generation of the diazoproduct 57 has been obtained.

Scheme 15. Thermolysis of 3,3-difluorodiazirine in the gas phase.

The chlorostyrene 63a, formed on photolysis of the 3-chlorodiazirine 60a (Scheme 16), could be generated eitherfrom the intermediate carbene 62a or from the isomericchlorodiazomethane 61a.[64] On one hand, the carbene 62acould easily be transformed into the chlorostyrene 63athrough an intramolecular hydride shift. On the other hand,the diazoalkane 61a could also form 63a through acid-cata-lyzed decomposition. To clarify whether or not the diazirine60a would undergo isomerization to the diazo compound61a its photolysis (λ � 350 nm) in 3-methylpentane at–175 °C was studied. The IR spectrum of the reaction mix-ture indeed showed the absorption due to the diazoalkane61a (2040 cm–1). The intensity of this band essentially didnot vary over several hours at –175 °C, but it was signifi-cantly weakened on heating to –70 °C. These results do notnecessarily mean that the diazirine 60a will also photoiso-merize to the diazo compound 61a at normal temperatures,but they do at least attest to the possibility.

Scheme 16. Photolysis of halodiazirines.

IR spectroscopy confirmed that irradiation of the halodi-azirines 60a (Scheme 16, 13C NMR: 48.4 ppm), 60b (13CNMR: 48.3 ppm), or 60c (13C NMR: 39.8 ppm), matrix-isolated in argon (ca. 380 nm, 10 K), produced only tracesof the benzyldiazomethanes 61a (IR: 2048 cm–1), 61b (IR:2047 cm–1), or 61c (IR: 2045 cm–1), in contrast with signifi-cant yields of the corresponding carbenes 62a–c and thestyrenes 63a–c.[59]

No isomerization of the bromodiazirine 64a (Scheme 17),the chlorodiazirine 64b, or the fluorodiazirine 64c to thediazo isomers 65a–c was detected under monochromatic ir-radiation (64a: 330 nm/1 h, 64b: 318 nm/1 h, 64c: 316 nm/2 h) in an Ar matrix at 12 K.[65] Similarly, photolysis of thechlorodiazirine 64b and the fluorodiazirine 64c in N2 ma-trixes did not lead to the diazoalkanes 65b and 65c. How-ever, photolysis of the bromodiazirine 64a with monochro-matic light in a N2 matrix (6 h) showed the total disappear-ance of the diazirine 64a together with the detection of ab-

S. M. KorneevMICROREVIEWsorption bands corresponding to the diazoalkane 65a (IR:2095 cm–1), the carbene 66a, and the olefin 67a. Continua-tion of photolysis with short-wavelength light (265 nm) re-duced the yield of the carbene 66a (10%) and increasedthose of the olefin 67a (50 %) and the diazoalkane 65a(40%). Irradiation of the same mixture with long-wave-length light (460 nm) reduced the amount of the diazoal-kane 65a (5 %) and increased the formation of the carbene66a (25%) and the olefin 67a (70%).[65]

Scheme 17. Photolysis of trifluoromethyldiazirines in N2 matrixes.

Wide-range photolysis (λ � 280 nm) led to the completedisappearance of the diazoalkane 65a and the carbene 66awith the formation of the olefin 67a (100%). From theknown facts[19,21] it was concluded that the formation of thediazo compound in this case occurs through the additionof N2 from the matrix to the bromocarbene.

Treatment of the 3-alkoxy-3-chlorodiazirine 68(Scheme 18) with AlCl3 in dry benzene (dark, 20–30 min)afforded the alkylbenzene 71 in 95 % yield.[66a] As computedin the gas phase, the AlCl3 complex with the diazomethane69 (the linear isomer of 68) is one of the possible intermedi-ates. (The function of the AlCl3 in this case is simply tocatalyze the loss of nitrogen from 68 through a ring-openingstep to afford the alkoxy-chlorocarbene 70.[66b])

Scheme 18. Decomposition of the chlorodiazirine 68 in the pres-ence of AlCl3.

C60 has recently been suggested as a probe for distin-guishing carbene/diazo compound partition in the photoly-sis of diazirines (Scheme 19).[49,51] Photolysis of the diazir-ine 2k in toluene/dichlorobenzene solution (Pyrex filter,30 min) together with an equimolar amount of C60 showedthe yield of the corresponding fulleroid 73k (mixture of twodiastereomers in 4:1 ratio; the main isomer is shown) fromthe diazomethane 1k to be 4%, in contrast with the meth-anofullerene 74k (14%) from the carbene 9k and 1-chloro-2-methylprop-1-ene (72e) from an unspecified precursor.[67]

The primary product of addition of 1k of to C60 – the corre-sponding 3H-pyrazoline (not shown) – was detected but ap-peared to be stable only at –40 °C.


Scheme 19. Photolysis of chlorodiazirines in the presence of C60.

Analogously, irradiation of 3-chloro-3-phenyldiazirine(2e) and C60 (15 °C, Pyrex tube, benzene) afforded the cor-responding adduct C60(CPhCl), identified as the methano-fullerene 74e, in 99 % yield.[49] No adduct 73e was observed,so no diazomethane 1e was formed.

2.5. Silyldiazirines

Silyldiazirines obtained from the corresponding diazocompounds have shown the capability for back-isomeriza-tion into linear partners only in some cases in Ar matrixesat low temperature. No data on their thermal isomerizationhave yet been reported.

The first silyldiazirine, compound 75a (Scheme 20), wasobtained on irradiation of the diazomethane 76a, matrix-isolated in argon at 8 K, as a component of a photostation-ary mixture.[68] It was later established that the diazirine isalso reasonably stable at room temperature and could evenbe distilled. Moreover, the photochemical approach becamea convenient route to 75a at room temperature.[69]

Scheme 20. Photolysis of silyldiazirines in Ar matrixes.

Short-wavelength irradiation of pure diazirine 75acaused decomposition with loss of nitrogen to give the sil-ene 77a very cleanly.

No diazo compound 76a was detected. The less substi-tuted Si derivatives 75b and 75c showed similar behavior.[70]

It was mentioned in one case that 75a forms a photo-stationary mixture with its linear isomer 76a on irradiationin an Ar matrix (8 K), finally giving the 1,1,2-trimethylsila-ethylene 77a. Details of the experiment and product distri-bution were not reported, however.[68]

A tendency to isomerize was also demonstrated bythe bis(diazirinyl)silane 78a (Scheme 21, IR: 1642,1639 cm–1).[71,72] Short-wavelength irradiation in an Ar ma-trix gave an insignificant quantity of a diazo compound,most likely 79a, together with the silirene 80a. The IR data,however, do not allow distinction between a mixture of 79a


with the bisdiazirine 78a and the diazodiazirine 81a. Twofurther bis(diazirines) – 78b and 78c – demonstrated similarisomerization ability.

Scheme 21. Photolysis of bis(diazirinyl)silanes in Ar matrixes.

Incorporated of silicon in a crowded unsaturated cycleand attachment to a diazirine moiety (Scheme 22) pre-vented its isomerization to a linear diazo structure underboth photolysis and thermolysis conditions.[73,74] Interrup-tion of photolysis of the silyldiazirine 82a (IR: 1620 cm–1,Pyrex filter, benzene/tBuOH or benzene/MeOH, 12 h) at in-complete conversion showed only nitrogen eliminationproducts: the silole 84 and the 1,2-dihydrosiline 85. Noisomerization to 83a was observed, but a certain amount ofthe initial diazirine 82a was recovered (32%). Photolysis of82a in MeOH gave similar products.

Scheme 22. Decomposition of the (l-silacyclopentadieny)diazirine82a.

Thermolysis of the diazirine 82a (benzene/tBuOH,150 °C, 10 min) also did not lead to the formation of thediazomethane 83a. If stable enough under the thermolysisconditions it should be detectable, if formed.[73,74] However,diazo-to-diazirine isomerization occurred.

2.6. Aryldiazirines

Like halodiazirines, aryldiazirines are fairly readily ac-cessible and are therefore well studied. Their linear isomers,the corresponding aryldiazomethanes, are moderately stableand can be observed under ambient conditions. Both pho-tolysis and thermolysis of aryldiazirines tend towards theformation of the corresponding diazomethanes at the ex-pense of nitrogen extrusion. The more polar the solventused for thermal isomerization, the higher is the yield ofdiazo compound obtained. The C60 probe technique hasbeen used to determine the quantitative partitioning be-tween carbene production, diazoalkane production, andother reactions.

One of the first publications on diazirine synthesis notedthat the color of colorless 3-methyl-3-phenyldiazirine turnsto pink-red and shows an absorption band correspondingto the diazo group at 2042 cm–1 in its IR spectrum onstanding in sunlight for two days.[75] Similarly, photolysis of


the diazirines 86a and 86b (Scheme 23) in hexane (Pyrexfilter) showed new photosensitive products (UV: 275,480 nm) that at first accumulated and then disappeared.[55]

The kinetics of the disappearances of the diazirines 86a and86b in the presence of AcOH were first order, with one ofthe reaction products being identified in each case as theacetate 88a or 88b, the product of a reaction between aceticacid and the diazo compound 87a or 87b. In the course ofthe photolysis of the diazirines 86a and 86b either in liquidfilms or in CCl4 solutions the absorptions corresponding tothe diazirine group at 1600 cm–l in the IR spectra decreasedtogether with growth of bands corresponding to the diazogroup at 2000 cm–l. All these experiments showed that the3-aryl-3H-diazirines 86a and 86b isomerized at least par-tially on photolysis to afford the linear diazoalkanes 87aand 87b .[56]

Scheme 23. Photolysis of monoaryldiazirines.

Ultrafast IR spectroscopic studies of ultrafast laser flashphotolysis (LFP) of the diazirine 86c (CHCl3, room tem-perature) with 270 and 375 nm excitation showed the pro-duction of the transient isomeric diazo compound 87c (IR:2030 cm–1).[76] The development of the integrated diazoband over time reveals its formation with two time con-stants of 4.4 and 240 ps, with an intensity ratio of 75:25. Itwas concluded that the diazo compound 87c was formedby two pathways: a fast pathway (75%) from the S2 state ofthe diazirine 86c and a slow pathway (25 %) from the S1

state of the diazirine. Furthermore, a short-wavelength laserpulse (270 nm) excited the diazirine to the S2 electronicstate, in which the direct formation of the diazo compoundand the singlet carbene were so rapid that they could evencompete with the internal conversion leading to the S1 state.A long-wavelength laser pulse (355 nm) pumped the diazir-ine 86c only to the S1 excited state. As a result, the diazirineisomerizes into the diazo compound as the major decaypathway, as well as forming the singlet carbene as a minordecay pathway.

Rearrangements of alkyl-aryldiazirines to diazo com-pounds appear to dominate over direct nitrogen extrusionand can take place either photochemically or thermally(Scheme 24). Interruption of the thermolysis of the diazir-ines 89a and 89b (DMSO, 100 °C) at incomplete conversionallowed the isolation of the corresponding diazopentane90a (UV: 520 nm) and diazoethane 90b (IR: 2040 cm–1), to-gether with the azines 92a and 92b, the ketones 93a and93b, and other products.[77] This result has shown that themechanism of the decomposition of the diazirines 89a and89b is not unidirectional. It was also demonstrated that thepolarity of the solvent, but not its nature, would influencethe product distribution. Neat decomposition of the diazir-

S. M. KorneevMICROREVIEWines mainly produces the azines.[35] It is interesting to notethat the complex of the diazirine 89b with β-cyclodextrinproduces mainly cyclopropanes or styrene on thermal andphotolytic decomposition and that no diazo compound orazine are formed.[78]

Scheme 24. Reactions of alkylphenyldiazirines.

Photolysis of the diazirine 89c (Scheme 24) in the nearUV (353 nm) led to the intermediate isomeric diazometh-ane 90c (about 35% by spectroscopy) together with the di-rect formation of carbene products.[30] After total decompo-sition both of the diazirine 89c and of the diazomethane90c (90 min) the yields of insertion products were 95% (in-sertion into the O–H bond of MeOH) or 50% (insertioninto the C–H bond of cyclohexane). Moreover, the diazoproduct appeared to be reasonably stable and did not de-compose in AcOH in cyclohexane (0.1 m) in an attemptedtrapping experiment.

Being themselves obtained by oxidation reactions, diazir-ines are generally inert to oxidizing agents. When reactionwith an oxidizing agent does take place it does so with thediazirine decomposition products. Thermal oxidation of thediazirine 89a with m-chloroperoxybenzoic acid (MCPBA)showed that the oxidation proceeded to form the decompo-sition products of the carbenic and diazo pathways(Scheme 24).[79] The ketone 93a could be formed as a resultof oxidation of the diazopentane 90a with MCPBA whereasthe cis- and trans-pentene oxides 94a are derived from thecarbene intermediate 91a, which rearranges to the cis- andtrans-pentenes before oxidation.

It was shown that the decomposition of 89a in chloro-benzene proceeds 50 % in carbenic fashion and 50% by thediazo process. In a more polar solvent (nitrobenzene) the


decomposition occurs 75% by the diazo pathway and in aneven more polar solvent (DMSO) this pathway accounts foras much as 90 % of the product. The diazirine moiety isknown to be less polar than the diazo group, so the morepolar the solvent used for thermal isomerization, the betterit stabilizes an increase in polarity in the transition stateand the higher the yield of diazo compound obtained.[80]

Both thermal and photochemical decomposition of thediazirine 89a in CD3CO2D (Scheme 24) gave the acetate 96and a mixture of the cis- and trans-pentenes 97.[79b,79c] Theidentical yields and product ratios in the two reactions ledto the conclusion that both processes pass through the sameintermediate: the diazoalkane 90a. The similarity of the re-sults implies that both processes occur in the ground stateof 90a irrespective of the activation method and the charac-ter of the initial activated state. Protonation of the diazo-alkane 90a leads to the diazonium ion 95, which gives theknown thermal reaction products 96 and 97. If a carbenehad been the initial product an intramolecular migration ofa hydride ion should have driven a reaction to afford theundeuterated olefin 97, which was not found.

An attempt to trap the diazomethane intermediate 90bin the form of a cycloadduct with a dipolarophile was notsuccessful (Scheme 24).[81] Photo1ytic decomposition of 89bin the presence of N-methylmaleimide in ethereal solutiongave no [2+3] cycloaddition product 98 despite the evidentformation of the diazo compound 90b (development of ared color on photolysis of the diazirine 89b in solution anda quantitative yield of an insertion product of type 88 onphotolysis in the presence of acetic acid). Only the azine92b (8%) and the cyclopropane 99 (43%) were isolated.However, the corresponding pyrazoline 98 (23 %, mixtureof diastereomers), the direct cycloadduct of the diazo com-pound and the dipolarophile, was obtained from a solvent-free photolysis, together with the azine 92b (4%) and thecyclopropane 99 (41%, mixture of diastereomers). Onceagain it was demonstrated that the outcome of trapping ofintermediates depends, among other factors, on the natureof the reaction medium.

Thermal rearrangement of the diazirine 89a to the diazocompound 90a was also investigated in the presence offullerene (C60, Scheme 25).[50–52] The fulleroid C60CPh(nBu)(101), with the nBu group located above the five-memberedring, was confirmed by 1H NMR. Yields were high: 70% inan o-dichlorobenzene/C6D6 mixture and 76% in DMSO. Inaddition, the minor fulleroid 102, the methanofullerene 103,and (E)-1-phenylpent-1-ene (100) were detected. These re-sults indicate that the major decomposition pathway for thediazirine 89a is the formation of the diazomethane 90a(about 80 %) and agree with other thermal experiments car-ried out with the compound.[79]

Spectral methods are a powerful tool for the detection ofunstable aryldiazomethanes at low temperatures(Scheme 26). Narrow-band photolysis of the diazirine 104(IR: 1639, 1631 cm–1) in an Ar matrix resulted in almostquantitative isomerization to the diazoalkane 105 (IR:2042 cm–1, unstable under usual conditions). Further pho-tolysis of 105 with long-wavelength light transformed it into


Scheme 25. Thermolysis of the aryldiazirine 89a in the presence ofC60.

the also unstable alkyne 106. Broad-band irradiation(λ � 328 nm) initiates both the primary photolysis of thediazirine 104 and the secondary photolysis of its product,resulting in a mixture of the diazoalkane 105 and the alkyne106.[82]

Scheme 26. Photolysis of the bicyclodiazirine 104 in an Ar matrix.

2.7. Diaryldiazirines

Data for diaryldiazirines are not numerous, because ofhigh instability of diazirines with two bulky substituents allappearance. An attempt to synthesize the diphenyldiazirine108 (Scheme 27) by a general synthetic scheme accordinglyfailed.[83] As the product of oxidation of the intermediatediaziridine 107 either with oxygen in the air or with HgOonly diphenyldiazomethane (109) was obtained. Its forma-tion was explained by steric factors making the diazirine108 unstable and leading to its thermal (room temperature)isomerization to the diazoalkane 109.

Scheme 27. Reactions of diaryldiazirines.


The spirodiazirine 110a (Scheme 27) in fact appears to bethe only stable diaryldiazirine (isolated by chromatography)currently known, and its properties have been studied tosome degree.[84] Thermolysis of 110a in MeOH led to itscomplete isomerization to the diazo compound 111a. Pho-tolysis of the diazirine 110a in MeOH gave products ofcarbene reactions with the solvent, although partial forma-tion of the diazo compound 111a was noted. From kineticmeasurements the assumption was made that the formationof the carbene products 112a and 113 occurs as a resultof the decomposition of the diazirine 110a and not of theintermediate diazofluorene 111a.

2.8. α-Keto-Diazirines

Even the first attempts to synthesize α-keto-diazirineshowed that such compounds, unlike their linear diazo-ketone isomers, are much less stable and less accessible thanaliphatic diazirines. The diazirine 20b (Scheme 28) is thusstable at room temperature[85] whereas the diazirine 114loses nitrogen with a half-life of 1 h.[86] The stable bicyclicdiazirines 115[87] and 116a[88] were identified later. Theirstabilities were explained in terms of the positive influenceof strained factors in the molecule skeletons, consistentlywith the behavior of other cyclic azo compounds.[89] Theperfluorinated keto-diazirine 117c also appeared to bestable, which was explained in terms of the distinctive prop-erties of fluorine atoms.[90,91] In general, data for α-keto-diazirines are fragmentary and poor. Several known exam-ples testify that irradiation leads mainly to the nitrogen eli-mination products and that isomeric diazoketones areformed in minor quantities. The same is true for thermoly-sis.

Scheme 28. Structures and photolysis products of diazirines.

The diazirine 116a, obtained serendipitously in a synthe-sis of strained molecules, was the first α-keto-diazirinestable under conventional conditions.[88] Photolysis of 116aled to the formation of the acid 119a and small amount ofthe diazoketone 118a. Because the photolysis of the diazo-ketone 118a also gave the isomeric diazirine 116a the as-sumption was made that these two compounds undergophotoreversible interconversion. Further conversion ofthese isomers involving Wolff rearrangements can proceedwith each partner of this pair, but probably with differentrates.

S. M. KorneevMICROREVIEWPhotolysis of the diazirines 117a and 117b (Scheme 29)

in an Ar matrix with short-wavelength light led to nitrogenabstraction and rearrangement to the ketene 121a (IR:2192 cm–1), but formation of the diazoketones 120a and120b was not confirmed.[91] Irradiation of a close ana-logue – the diazirine 117c – with long-wavelength lighttransformed it into the ketene 121c and the diazoketone120c, but in this case no yields or product ratio were re-ported.[90]

Scheme 29. Photolysis of (perfluoroalkyl)diazirines in Ar matrixes.

Thermolysis of the diazirine 117c in CCl4 (80 °C) ledonly to the formation of insertion products with the sol-vent.[90] Heating in the gas phase (80 °C, 3.5 h, 300 mm,diluted with Ar) resulted in quantitative elimination of ni-trogen with the formation of the ketene 121c. The isomericdiazoketone 120c is stable under these conditions, so it isassumed that thermal isomerization of the diazirine 117cdoes not occur and that 121c is formed from it directlywithout stepwise opening of the diazirine cycle.

The diazirine 115 (Scheme 30) appeared to be inert tothe action of long-wavelength light (λ � 395 nm). In thecourse of short-wavelength photolysis, however, it disap-peared quickly, giving the intermediate diazocamphor 122(UV: 250, 300 nm) with a quantum yield close to 0.7 andas a result of subsequent nitrogen elimination afforded theester 124 (formation of the carbene 123 is thus not ruledout).[87] Obviously, isomerization does not reach a photo-stationary stage, because the isomeric diazocamphor 122formed is also light-sensitive. Thermolysis of the diazirine115 proceeded effectively even at 50 °C (in EtOH the half-life was around 33 h, in isooctane about 12.5 days) and theyield of the diazocamphor 122 was increased to 20 % atabout 50% conversion, but the ester 124 again appeared asthe end product.

Scheme 30. Photolysis of the spiroketo-diazirine 115.

The pair made up by the diazocamphor 122 and the di-azirine 115 is unique because of the thermal stabilities ofboth cyclic and linear isomers. As a rule, only one of thetwo isomers displays this property.[87]

2.9. Carbamidodiazirines

The monocarbamidodiazirines 125a–c (Scheme 31) werethe first diazirines with α-carbonyl groups that were found


to be stable under conventional conditions and could beisolated and investigated. Both the diazirines and the iso-meric diazoamide are fairly stable, although the data forthem are not numerous. The formation of diazo compoundsfrom diazirines in low yields is observed both on photolysisand on thermolysis.

Scheme 31. Isomerization of carbamidodiazirines.

Irradiation of the diazirine 125a with monochromaticlight (312 nm, 1 h) led to a mixture of compounds. One wasidentified as the product of isomerization: the linear N-methyldiazoacetamide 126a (IR: 2140 cm–1), quite a stablecompound.[92] Similar irradiation of the diphenylamide125b (334 nm, 2.25 h) also led to the formation of a com-plex mixture of products contained the isomeric diazoamide126b (IR: 2100 cm–1). Photolysis of 125b with short-wave-length light (312 nm) no longer led to formation of the di-azoamide 126b, possibly because of absorption of this radi-ation by the diphenylamide group. Thermolysis of the diaz-irine 125c in boiling toluene quantitatively yielded N-meth-yloxindole and did not show the presence of the isomericdiazoamide 126c.

Irradiation of the diazirine 127a (Scheme 32) and subse-quent chromatography allowed the isolation of the meth-oxyindole 129a and the oxindole 130a (the same productsas obtained on photolysis of the diazoamide 128a).[93] For-mation of the isomeric diazoamide 128a was thus not ob-served. The almost colorless diazirine 127a is reasonablystable (m.p. 169 °C with decomposition), but graduallytransforms into the initial red diazoindoline 128a, almostquantitatively in solution or in the crystalline form evenat room temperature (conditions of this reaction are notreported).

Scheme 32. Reactions of spirocyclic carbamidodiazirines.

Kinetic measurements showed that the half-life of the di-azirine 127b was around 14 h at 18 °C and about 12 min at50 °C and that the energy of activation for thermal isomer-ization to the diazoindole 128b was about 27 kcalmol–1.


Low stabilities of keto-diazirines had been observed earlier:114, for example, eliminates nitrogen within an hour atroom temperature.[86,94]

The bis(carbamido)diazirine 131, unlike the monocarb-amidodiazirines 125a, 125b, and 127a or the dialkyl diazir-ine-3,3-dicarboxylates 135, 139a, and 139b, did not showa propensity to isomerize to the diazo compound 132 onphotolysis (Scheme 32).[95] Only products of nitrogen ab-straction – the barbituric acid 133 and the ester 134 – wereidentified, although no data were reported either for thediazirine 131 or for the products of photolysis (the samecompounds were also obtained from photoirradiation ofthe isomeric diazodiamide 132).

2.10. Diethyl Diazirine-3,3-dicarboxylate

In spite of the fact that the diazomalonic ester 136(Scheme 33) has been known for more than 40 years[96] andthat its cyclic isomer – the diazirine 135 – and analogueshave been synthesized in the form of various derivatives ofmalonic acid,[97] data for their mutual transformation haveappeared only more recently. The dependence of the isom-erization yields on the wavelength of the irradiating lighthas been reported, although they are certainly not high.Thermolysis gave only traces of diazo compound, togetherwith formation of other products.

Scheme 33. Decomposition of diethyl diazirine-3,3-dicarboxylate(135).

Irradiation of the diazirine 135 with monochromaticlight at room temperature led to the formation of two mainreaction products: the methoxymalonate 137 (main, car-bene insertion) and the diazomalonate 136 (minor, isomer-ization), the ratio of which varied depending on the wave-length of the irradiating light.[98]

It is significant that the maximum yield of isomerizationwas observed on irradiation in a part of the spectrum(300 nm) in which the formed diazomalonate 136 has anabsorption minimum between two maximums (UV: 251 and344 nm). Apparently, irradiation close to the absorptionmaxima of this photosensitive reaction product causes sec-ondary photolysis and completely destroys 136.

Thermolysis of the diazirine 135 (IR: 1755, 1730 cm–1)results in the formation of insignificant quantities of thediazoester 136 (IR: 2123, 1758 cm–1) along with the malon-ate 137. As in the case of the cyclic analogue 139a(Scheme 34), the rates of decomposition of the diazirine 135


in H2O appeared to be higher than those in dioxane over a30–70 °C temperature range, which clearly points to a simi-larity of their mechanisms of isomerization.[98]

Scheme 34. Structures of diacyldiazirines.

2.11. Spirocyclic Dialkyl Diazirine-3,3-dicarboxylates

The diacyldiazirines 138 (Scheme 34), unlike the isomericdiacyldiazomethanes (linear and cyclic), appear to be com-pletely unstable and no any reports of their observation un-der any conditions have yet been made, though ideas ontheir involvement in the photochemical decomposition ofthe corresponding diazodiketones have been put forward.[99]

Diazirines bearing dicarboxy and dicarbamide groups aremore favored. The diazirine 139a, obtained from the diazo-Meldrum’s acid derivative 140a, is known to be reasonablystable over several weeks at room temperature.[100] The di-azirines 131 and 135, originating from the diazobarbituricacid 132[95] and the diazomalonic ester 136,[98] respectively,have also been reported. It was disclosed that thermolysisof dialkyl diazirine-3,3-dicarboxylates (Scheme 35) tendstowards isomerization into the diazo partners, whereas inphotoreactions both isomerization and nitrogen eliminationcompete.

Scheme 35. Thermolysis of spirocyclic dialkyl diazirinedicarboxyl-ates.

Kinetic measurements on thermolysis of the diazirines139a and 139b in mixtures (CDCl3/MeOH, CDCl3/H2O, orC6D6/MeOH) at 50 °C showed that their half-decay periodswere 1.3–4.1 h.[99] Two competitive processes were thus ob-served: isomerization with the formation of the diazo com-pounds 140a and 140b (22–47 %) and elimination of nitro-gen with formation of the Wolff rearrangement products141 and 142 (40–78%). The elimination products shouldoriginate from the diazirines 139a and 139b, the diazo com-pounds 140a and 140b being stable under the thermolysisconditions. It is interesting that thermolysis carried out inmore polar reaction media (MeOH, 40 °C) in practice gavethe diazo compound 140a quantitatively.[101] Moreover, thediazirine 139a rapidly turns into the diazo isomer 140a inthermal fashion only in solution, remaining unchangedover several months in crystalline form.[102]

S. M. KorneevMICROREVIEWThe results obtained were confirmed by calculations re-

lating to the potential energy barrier to the thermal isomer-ization of the diazirine to the diazo compound in the gasphase (about 30–37 kcalmol–1).[16] At the same time thebarrier for the return isomerization of the diazo to the di-azirine compound is about 53–60 kcalmol–1, considerablyhigher than the energy of activation for Wolff rearrange-ment (47–48 kcalmol–1).[80] The thermolysis of the diazirinetherefore should not lead to loss of nitrogen.

The polarity of a reaction medium essentially influencesnot only the reaction pathway, but also the rate of isomer-ization of the diazirines to the diazo compounds 140a or140b (Scheme 36). Their isomerization in water thus pro-ceeds approximately 10 times more rapidly than in dioxane,due to the fact that the diazirine moiety is considerably lesspolar than the diazo group and that a polar solvent (H2O)stabilizes an increase in polarity in a transition state betterthan a less polar one (dioxane).[80]

Scheme 36. Photolysis of spirocyclic diazirines in various solutions.

The diazirines 139a and 139b are almost inert with re-spect to long-wavelength light (λ � 320 nm).[103] In con-trast, short-wavelength irradiation (λ � 210 nm) appearedto result in fast and almost unidirectional processes, withonly slight isomerization to the diazo compounds 140a and140b (0–8 %), against nitrogen abstraction to afford 141–144 (49–97%). The yields of the diazo compounds 140a and140b calculated by IR spectroscopy did not exceed 1–2 %over the courses of the reactions. The lack of the diazo com-pounds among photolysis products is not in fact surprisingbecause they are also sensitive to the irradiation. It is neces-sary to note that the lifetimes of the diazirines 139a and139b in a solution are not long, varying from 3.5 h (CDCl3)to 4 h (C6D6) at 50 °C.[102,103]

In general, a decrease in the wavelength of initiating light(Table 1) considerably shifted the photolysis of the diazirine139a (UV: 244, 284 nm) towards the nitrogen abstraction

Table 1. Conditions and products of reactions of the diazirine 139ain MeOH.

Reaction conditions Ratio of the reaction products140a 142 144a

hν, 254 nm 4 93 3hν, 300 nm 44 56 0hν, 350 nm 36 63 1Thermolysis, 40 °C 100 0 0


products 142 and 144a relative to isomerization into thediazo compound 140a (UV: 248, 329 nm). Similar, but morestrongly expressed dependence was observed for 140a.[101]

3. Isomerization of Diazo Compounds toDiazirines

In spite of the fact that photochemical isomerization ofdiazirines to diazo compounds became known almostsimultaneously with their discovery, the first example of re-turn isomerization was found only in 1971.[104] The numberof discovered isomerizations of diazo compounds remainsrather limited to date. This might be due to a variety ofdifficulties. Namely: i. several types of diazo compounds,such as diazoalkanes or halogenodiazomethanes, are too la-bile to be investigated on this subject, ii. a number of diazir-ines, such as carbonyldiazirines, appear to be of low sta-bility, iii. attempts to obtain disubstituted or hindered diaz-irines, such as diphenyl- or Boc-carbamidodiazirine, havefailed, and iv. study of fast isomerization processes has be-came possible only recently, thanks to the development ofpowerful and high-speed techniques.

According to calculations, the ΔE values for isomeriza-tions of diazomethanes 1 to diazirines 2 range from 22.2(R = BH2) to –12.4 kcalmol–1 (R = F).[105,106] Diazometh-anes substituted with the groups –CH=CH2, –CH=O,–C�CH, –C�N, and –CH3 appear to be more stable thanthe corresponding diazirines by 5.6, 12.6, 7.1, 9.5 and1.9 kcalmol–1, respectively. The first four groups would beexpected to provide π-acceptor stabilization of the diazo-methanes. These substituents thus define which of the twoisomers is more stable, which is largely attributable to theinfluences of electronegativity and conjugation on the sta-bilities of the diazomethanes.

3.1. Diazomethanes

Because of the low thermal and photostabilities of simplealkyl diazomethanes, data for their possible isomerizationinto diazirines are not known. Only fluorinated alkylgroups or Si derivatives significantly increase the stabilitiesof the corresponding diazomethanes and make it possibleto observe their conversion into the isomeric diazirines.Long-wavelength irradiation of these diazo compounds re-sults in low-to-medium yields of diazirines, whereas short-wavelength excitation leads to nitrogen abstraction. No dataon their thermal isomerization have been reported.

Irradiation of the diazomethane 14a (UV: 260, 400 nm)[29] with long-wavelength light (Scheme 37) led to its totaldisappearance and the partial formation of the diazirine13a (UV: 310 nm) along with the carbene reaction products15a and 16a.[28] No diazirine 13a was observed on short-wavelength irradiation of 14a and only products of photo-degradation were obtained.


Scheme 37. Photolysis of (fluoroalkyl)diazomethanes.

The monosubstituted diazomethane 14b (UV: 260,400 nm)[28,29] gave trace amounts of the isomeric 13b (UV:310 nm) only on irradiation with long-wavelength light. Useof the perhalogenodiazomethane 14c under similar condi-tions also resulted in the formation of the isomeric diazirine13c (UV: 310 nm) in low yield.[28]

Photolysis of (trimethylsilyl)diazomethane (76a,Scheme 38) in an Ar matrix (λ � 360 nm) produced aphotostationary mixture of the diazomethane 76a (IR:2080 cm–1) and the diazirine 75a (IR: 1640 cm–1).[68] Pro-longed irradiation of this equilibrium mixture in an Ar ma-trix led to elimination of nitrogen from both isomers andthe formation of the triplet carbene 145a (ESR spectrum),with subsequent stabilization as the silene 77a on heatingabove 45 K.[107] It was found that irradiation of the diazo-methane 76a in solution also led to the formation of thediazirine 75a at room temperature.[68,69] Newly formed 75awas stable enough to be isolated and characterized underthese conditions. Photolysis of 76a at 4 °C with short-wave-length light (�300 nm) led to fast disappearance of the di-azomethane and the very clean formation of the silene77a.[107] The less substituted diazomethanes 76b and 76cshowed similar behavior.[70]

Scheme 38. Photolysis of silyldiazomethanes in Ar matrixes.

Irradiation of the bis(diazomethyl)silanes 79a–c (IR:2082, 2073 cm–1) with blue light in argon matrixes initiallygave the bis(diazirinyl)silanes 78a–c (IR: 1642, 1639,1636 cm–1) and small amounts of the silirenes 80a–c(Scheme 39).[71,72] Short-wavelength irradiation of 79a–c(305 nm) rapidly resulted in the disappearance of the diazocompounds and the silirenes 80a–c were the only products.

Photolysis of the silyldiazomethane 83a (IR: 2070 cm–1,Scheme 40) in a benzene/MeOH solution (Pyrex filter, 6 h)drove the reaction towards the silyldiazirine 82a (IR:1620 cm–1, isolated by chromatography on SiO2) along withproducts of nitrogen abstraction (146, 147, andothers).[73,74] Continued irradiation of 83a in MeOD (18 h)as well as photolysis in tBuOH decreased the amounts ofthe diazirine 82a (9% and 38%, respectively).


Scheme 39. Photolysis of bis(diazomethyl)silanes in Ar matrixes.

Scheme 40. Photolysis of silyldiazomethanes in benzene/ROH.

In contrast, use of an optical filter (phenanthrene inMeOH) resulted in an increase in the yield of the diazirine82a after 3 h of irradiation in both alcohols (54% in MeOHand 68 % in tBuOH).

A low yield of the disubstituted diazirine 82b (IR:1610 cm–1) was obtained on irradiation of the disubstituteddiazomethane 83b (IR: 2030 cm–1) in a benzene/tBuOHmixture (7 h) (Scheme 40). This diazirine was also isolatedby chromatography (SiO2) and appeared to be stable underambient conditions (m.p. 142 °C).

The presence of the second substituent attached to a lin-ear diazo group thus complicates, but does not exclude, itsisomerization into the diazirine ring, although the oppositeopinion had originally been expressed.[73]

3.2. Diaryldiazomethanes

Diazomethanes of this type are reasonably stable,whereas the diazirines are almost unknown. Formation ofdiazirines with two aromatic rings on photoexcitation of thecorresponding diazomethanes has been observed only fortwo similar structures and only when the π�π* absorptionband was activated.[84] Thermal diazo-to-diazirine isomer-ization is not known, though the reverse reaction has beenreported.

One initial product of the photolysis of 9-diazo-1,8-di-azafluorene (111a, IR: 2091 cm–1) with long-wavelengthlight (350 nm) in MeOH, cyclohexane, or benzene is, appar-ently, the diazirine 110a (Scheme 41). Analysis of the reac-tion mixture (NMR) showed that the diazirine 110a grad-ually accumulated during photolysis and reached its maxi-mum yield after 2 h (about 20%). Further irradiation re-duced the yield to almost zero and replaced 110a with prod-ucts of carbene reactions [112a (42%), 113 (16 %), andothers]. It should be noted that although the diazirine 110ahas been isolated by chromatography (SiO2), only the NMRspectrum in the form of a set of multiplets has been re-ported.

S. M. KorneevMICROREVIEW

Scheme 41. Photolysis of diaryldiazomethanes.

The dependence of the diazirine formation on the wave-length of the irradiating light was studied. On photolysisat ca. 310 nm, into a π�π* absorption band, the exciteddiazofluorene 111a rearranged to the diazirine 110a (8.8%at 18 % conversion) and lost nitrogen to form the carbeneproducts. However, when 111a was photolyzed at ca.420 nm, into the n�π* band, no diazirine was formed.

The sensitized photolysis of 111a in MeOH (λ � 380 nm,benzil as the sensitizer) led to the elimination of nitrogenwith the quantitative formation of fluorene 113 and no di-azirine 110a being detected.

Irradiation of the isomeric 3,6-diazafluorene 111b (IR:2078 cm–1, Scheme 41) in cyclohexane (350 nm) gave onlysmall amounts of the diazirine 110b along with the inser-tion product 112b (85 %).[108] The same reaction in benzenesolution (350 nm, 10 min) also led to a low yield of the di-azirine 110b (8.5%, detected by NMR). It is interesting thatreduction of the concentration of the initial diazo com-pound 111b resulted in reduction of the relative yield of thediazirine 110b: for a 5.5� 10–3 m solution of 111b in MeOHthe yield of 110b was 20% whereas for a 1.2 �10–3 m solu-tion of 111b it was 10% (no other characteristics of thediazirine 110b were reported).

3.3. Diazoketones

Diazoketones are much more stable than diazoalkanesand could be used under ambient conditions and in theirpure forms. Their cyclic isomers are of low stability, as al-ready became clear at the first attempt at their synthesis.[86]

Data for isomerizations of diazoketones to α-keto-diazir-ines are not numerous. Only long-wavelength light initiatedthese isomerizations, with low-to-medium yields. Thermalisomerization is unknown.

The first stable α-keto-spirodiazirines – 116a and 116b(Scheme 42) – were obtained during the course of the pho-tolysis of the strained tricyclic diazoketone 118a.[88] Theacid 119a was identified as the main product, together withthe minor α-keto-diazirine 116a. It has appeared muchmore stable (m.p. 65 °C) than the previously obtained α-keto-diazirine 114, which decomposed even at room tem-perature.[86] For better understanding of the role of thedouble bonds in stabilization of the diazirine cycle the satu-rated analogue – the diazoketone 118b – was synthesizedand subjected to photolysis. Under similar conditions the


saturated analogues – the diazirine 116b (isolated by TLC/SiO2) and the acid 119b – were obtained with approximatelythe same yields.

Scheme 42. Photolysis of cyclic diazoketones.

It was thus postulated that α-keto-diazirines can be stabi-lized not only by conjugation[109] or by a neighboring amidemoiety,[92,104] but also by some other still unknown fac-tors.[88]

Another stable spirodiazirine – compound 115 – was ob-tained from the diazocamphor 122 (Scheme 43).[87] On irra-diation of 122 with visible light two parallel processes werediscovered: isomerization to the diazirine 115 (IR:1750 cm–1) and elimination of nitrogen followed by re-arrangement to form either the ester 124 (reaction in EtOH)or other products (reaction in hexane). The new α-keto-diazirine 115 was found to be stable under ambient condi-tions and to be tolerant of isolation by chromatography(SiO2) and recrystallization (m.p. 95–96 °C). It should benoted that the diazirine 115, like many other camphor de-rivatives, is a very volatile substance and could be easilysublimed.[87]

Scheme 43. Photolysis of the bicyclic diazoketone 122.

A photoisomerization to a stable diazirine was observedin the cyclic diazoketone 148 (IR: 2075, 1660 cm–1,Scheme 44), containing two activated aromatic rings, whichcould not in any way be termed strained.[110] Long-wave-length irradiation (406 nm) of 148 produced the α-keto-di-azirine 149 (IR: 1700 cm–1) in a high yield. It appearedstable and was isolated from the reaction mixture by TLC(m.p. 147 °C). Application of shorter-wavelength light(λ � 350 nm) in MeOH (5 h, 25 °C) led to the formationonly of nitrogen elimination products: the ester 150 (Wolffrearrangement), the methoxyketone 151 (O–H insertion),and the corresponding indanone (8%, diazo group re-duction).

Irradiation of the diazoketone 120a (UV: 235, 263,357 nm, Scheme 45) with long-wavelength light(λ � 335 nm) led to the formation of an intermediate prod-uct first wrongly identified as the ketocarbene 152.[91b] Ac-tually it was the diazirine 117a (IR: 1761 cm–1),[90] whichwas detected on irradiation of 120a in an Ar matrix(λ � 320 nm) along with the ketene 121 (IR: 2192 cm–1), the


Scheme 44. Photolysis of the diazoindanone.

expected product of Wolff rearrangement. Photolysis with aPyrex filter (λ � 280 nm) both in a matrix and in the gasphase, as well as short-wavelength irradiation (λ � 210 nm)in a matrix, yielded only the ketene 121.

Scheme 45. Photolysis of (perfluoroalkyl)diazoketones.

It is interesting that room-temperature photolysis of 120awith long-wavelength light (λ � 320 nm) is now used forpreparative-scale synthesis of 117a in 45% yields.[90,91]

Photolysis of close analogues – the diazoketones 120band 120c (Scheme 45) – in a polycrystalline phase at 12 K(λ � 335 nm) also gave the reasonably stable diazirines 117b(IR: 1752 cm–1) and 117c (IR: 1762 cm–1), which surviveda short period of warming to room temperature withoutchanges.[91]

3.4. Diazoamides

Both diazoamides and carbamidodiazirines are stableunder ambient conditions, being not only detectable, butalso isolable and characterizable. Visible light irradiation ofdiazoamides has given a few diazirines, but UV photolysishas resulted in the formation only of nitrogen eliminationproducts. The presence of a second substituent attached toan amide moiety blocks diazirine formation when a linearderivative is formed, but does not preclude the formationof the diazirine if two susbtituents attached to an amidemoiety are closed into a cycle. Thermal diazo-to-diazirineisomerization for these pairs has not been described, butthe reverse reaction occurs.

Diazoamides are known as the first class of compoundsto have demonstrated isomerization to stable diazirinesupon light activation.[104] Irradiation of N-diazoacetylpiper-idine (153a, Scheme 46) with visible light gave N-(diazirinyl-carbonyl)piperidine (154a, IR: 1655 cm–1) in low yield. UVlight caused nitrogen elimination and the formation of theazabicycle 155. Isomerization of the diazoamide to the di-azirine appeared to be successful only in the case of themonosubstituted derivative 153a. When a second substitu-ent was attached to a diazo moiety – in 153b – no diazirine154b was observed. The assumption was made that bulky


substituents hinder the formation of diazirine cycles orcause rapid reverse reactions. This was confirmed by thefact that attempts to prepare diphenyldiazirine (108) in thesame way as the spirofluorenediazirine 110a by a “dark”pathway failed,[83,84] whereas 3-alkyl-3-phenyldiazirines areknown and stable.[75]

Scheme 46. Photolysis of diazoamides.

Irradiation of diazoacetylphenylalanine (156, Scheme 47)and diazoacetylproline (158a) with visible light(λ � 400 nm) gave the stable diazirines 157 (UV: 310 nm,m.p. 104 °C) and 159a (IR: 1658 cm–1, m.p. 98 °C), eachwith only one substituent attached to a diazirine moiety.[104]

Close analogues such as ethyl diazoacetate (160), diazoacet-one (161), and diazoacetophenone (162) are also monosub-stituted diazocarbonyls and might also have given mono-substituted diazirines, but did not.[92] In this they resemblethe disubstituted diazoamides 153b, 158b, and 163 whichgave no detectable diazirines upon irradiation.

Scheme 47. Propensities of diazocarbonyls to isomerization.

The occurrence of isomerization is affected not only bythe number of substituents, but also by the light wavelength.Irradiation of the diazoamide 126e with visible light(Scheme 48) thus drove the reaction towards the diazirine125e whereas use of UV light led to the formation of thecyclic lactams 165e and 166e.[104,111]

Scheme 48. Photolysis of diazomonoamides.

S. M. KorneevMICROREVIEWThe lightly substituted diazoamides 126a and 126d

formed the stable diazirines 125a and 125d on irradiationwith visible light. The methylphenylamide 126c and the di-phenylamide 126b were also converted into the correspond-ing diazirines 125c and 125b, so aromatic substituents at-tached to the amide moiety did not essentially affect thecourse of visible light photolysis. Irradiation of the di-azoamides 126b and 126c with UV light, however, led tothe formation of N-methyl and N-phenyloxindoles as themain identified products, with the diazirines 125b and 125cthus being bypassed.

These experiments demonstrated that visible light gen-erally initiated isomerization of α-diazoamides to diazirinesirrespective of substitution on the amide group. It was pro-posed that the irradiation generates the diazoamide exitedstates 167 (Scheme 49) with oscillatory energy low enoughto allow the isomerization process to compete successfullywith the formation of carbenes.[92] In fact, diazo-to-diazir-ine photoisomerization is observed just in cases of amidesof diazoacids and not of acids themselves or their esters. Itappears to be the case that the amide nitrogen plays themain role in the isomerization. In the structure 168a (X =NR�) it might participate in delocalization of a positivecharge and thus promote the formation of the diazirine cy-cle in 125. The structure 168b (X = O), an ester case, shouldexperience a minimal contribution to charge stabilization inthe intermediate, promoting diazirine ring cloure onlyslightly. The formation of structure 168c (X = CH2) for sta-bilization of ketones is difficult to imagine at all.

Scheme 49. Proposed mechanism of isomerization of diazoamidesto diazirines.

This mechanism also explains why no diazirine has beenformed from a diazoamide of a malonic acid semi-ester: theC atom bearing a diazo group is quaternary in this case anda structure of type 168a would become unstable.

It is interesting that cyclic diazoamides (Scheme 50) alsotend to isomerize into stable diazirines in spite of the factthat the diazirine rings are doubly substituted in suchcases.[93,104] Irradiation of the diazoindoline 128a (IR: 2085,1670 cm–1) in MeOH (290 nm) in the presence of oxygenled to the insertion of the formed carbene into the O–Hbond of MeOH to produce 129a (30%) and alternativelydemonstrated reduction to 130a (25%). Interruption ofphotolysis at approximately 50 % conversion of 128a, how-ever, allowed the isolation of the intermediate keto-diazirine127a (IR: 1740, 1690, 1630 cm–1) by chromatography(SiO2).

Similarly, the diazoindoline homologue 128b (IR: 2110,1680 cm–1, Scheme 50) also gave the diazirine 127b (IR:1730, 1620 cm–1) on irradiation in MeOH.


Scheme 50. Photolysis of cyclic diazoamides.

The diazobarbituric acid 132, another cyclic diazodi-amide, formed the stable diazirine 131 on irradiation underambient conditions (Scheme 51).[95]

Scheme 51. Decomposition of the diazobarbituric acid 132.

Long-wavelength excitation of the diazo compound 132in alcohol led to the formation of the diazirine 131, whichappeared stable at room temperature. Short-wavelengthlight caused nitrogen elimination and gave 133 (reduction)and 134 (Wolff rearrangement), although yields and charac-teristics of the photolysis products were not reported.

3.5. Diazodiesters

Diazodicarbonyl compounds, diazodiesters in particular,are quite stable and storable even at room temperature indarkness for a long time without significant changes. Diazir-ines related to diazodiesters, namely diazo-Meldrum’s acidand diazomalonic esters, are the only known compounds ofthis type. They have proven to be reasonably stable underambient conditions and have been studied in some detail.The degree of isomerization on photolysis depends stronglyon the wavelength of irradiation: the shorter the wave-length, the lower the yield of isomerization. Thermal diazo-to-diazirine isomerization has not been described for thesepairs, but return reactions occur in solutions, as in the di-azoamide case.

The first isomerization of diazo-Meldrum’s acid (140a,Scheme 52) to the corresponding diazirine was observed in1978, when it was used in syntheses of cyclopropanes.[100]

Direct photolysis of 140a (UV: 332, 249 nm) in the presenceof an excess of cyclohexene (Pyrex filter, λ � 350 nm)caused excitation of the n�π* transition in 140a (λmax =332 nm) and led to its isomerization to the diazirine 139awith a yield of 35% (m.p. 84 °C, IR: 1769 cm–l).

The target product of carbene decomposition, the corre-sponding cyclopropane 169, appeared to be a minor one.Use of a quartz filter (λ = 253.7 nm), however, resulted inexcitation of the π�π* transition in 140a (λmax = 249 nm)and the formation of diazirine 139a decreased to 4% whilethe corresponding cyclopropane 169 became the main prod-uct. Back “dark” isomerization of the diazirine 139a to140a proceeds at room temperature within a few weeks.[100]


Scheme 52. Photolysis of diazo-Meldrum’s acid in cyclohexene.

This experiment started a new phase of study of interme-diates in the course of photolytic decompositions of diazocompounds and also made diazo-Meldrum’s acid the mainobject of these investigations.

Testing of photodecomposition of 140a in protic media(H2O, MeOH) became the next step of the study(Scheme 53).[104] It was shown that the diazo compound140a is inclined to Wolff rearrangement, like its carbocyclic“relatives” and unlike its acyclic analogue, the diazomalonicester 136.

Scheme 53. Photolysis of diazo-Meldrum’s acid in protic media.

On irradiation of 140a with short-wavelength light(λ � 210 nm) in H2O or MeOH, two processes are ob-served: 1) isomerization to the diazirine 139a (minor), and2) loss of nitrogen (main) with formation of 141a or 142a(Wolff rearrangement) and of 144a (reduction).

Photolysis of 140a in a poly(methyl methacrylate)(PMMA) film (254 nm, 22 °C) showed some differencesfrom a solution reaction: 140a disappeared (IR: 2170 cm–1)and the ketoacid 141a (IR: 1810 and 1770 cm–1) was formed(Scheme 54).[112–114] The formation of the diazirine 139awas not observed and the absence of the ketene 170 wasexplained in terms of its fast reaction with water adsorbedin the polymeric film. It was found that the quantum yieldof decomposition of diazo-Meldrum’s acid (140a) had in-creased from 0.37 in MeOH (0.27 in CHCl3) to 1.0 in aPMMA film.

Scheme 54. Photolysis of diazo-Meldrum’s acid in PMMA.

One of the deactivation pathways of exited 140a* in-volves interconversion to the corresponding diazirine 139aformed on long-wavelength irradiation (350 nm) of diazo-Meldrum’s acid (140a) in solution[100] as the main pho-


toprocess for diazo compounds in general.[14d,108] If the ra-diationless decay channels in excited 140a* are coupled toC=N=N bond bending of this kind, the rigid matrix couldsuppress this motion and enhance the probability of nitro-gen extrusion and subsequent rearrangement.[112]

Detailed investigation of the mechanism of photodecom-position of 140a has shown that the direction of photolysisis critically influenced by the wavelength of the excitinglight (Table 2, Scheme 55).[80]

Table 2. Reaction conditions and yields of products (at low levelsof conversion) on photolysis and thermolysis of the diazo com-pound 140a.

Reaction conditions Yields of reaction products [%]139a 142a 144a

hν, 254 nm 7 91hν, 254 nm, sens. 0 0 100hν, 300 nm 17 81hν, 350 nm 55 42hν, 355 nm 94 6Δ, 80 °C, MeOH 0 100 0

Scheme 55. Mechanism of photodegradation of diazo-Meldrum’sacid.

Short-wavelength irradiation (254 nm) of 140a (UV: 248,329 nm) in solution (MeOH) at room temperature thusshowed that the main pathway of its decomposition (at 10 %conversion) is nitrogen elimination and formation of theketo ester 142a. Long-wavelength irradiation (355 nm) leadsmainly to isomerization to the diazirine 139a (UV: 244,284 nm). Triplet-sensitized photolysis gave Meldrum’s acid(144a) quantitatively.

The following mechanism was suggested for the reaction(Scheme 55).[80] Short-wavelength irradiation of the diazocompound 140a excites its molecules to the S2 level (alter-natively, a set of highly excited singlet states – S5, S7 – couldbe populated by molecules, which could relax to the dissoci-ative S2 state by ultrafast internal conversion[115]). A pro-portion of them (about 34 %, quantum yield of photolysisat λ = 254 nm around 0.34) has time to lose nitrogen and totransform into the ketene 170 (Wolff rearrangement). The

S. M. KorneevMICROREVIEWothers, due to internal conversion, pass to the S1 level, atwhich a small proportion of the molecules (about 3%) canstill transform into the diazirine 139a. Another small pro-portion falls from the S1 to the T1 level, at which they losenitrogen and form Meldrum’s acid (144a). All remainingmolecules, which had no opportunity to react, reverted tothe ground state S0 through dark deactivation processes.

Excitation with long-wavelength light (355 nm) allowsmolecules of 140a to reach only the lowest excited level S1,at which isomerization to diazirine 139a takes place onlywith a tiny yield. The higher excited level S2, at which mole-cules can lose nitrogen, is not accessible now, so the reac-tion does not go this direction. The absence of phosphores-cence or fluorescence at deactivation testifies in favor of afast thermal rather than a photochemical transition fromthe level S1 to the ground state S0 (this conclusion confirmsa low quantum yield at λ = 350 nm, around 0.025). Triplet-sensitized excitation transforms the molecules to the T1

level, at which only the formation of the reduced product144a is possible.

A detailed study of excitation processes of diazo-Meld-rum’s acid (140a) by high-speed IR spectroscopy has con-firmed the suggested decomposition scheme.[115]

From the practical point of view, the longer the wave-length of light acting on the diazo compounds 140a and140b (Table 3), the higher are the yields of the diazirines139a and 139b obtained [5–12% at λ � 210 nm (quartz fil-ter) and 43–44% at λ � 310 nm (glass filter)].[103] A similarpattern was observed on irradiation of 140a in cyclohex-ene.[100] It should be noted that photochemical formation isstill the only accessible method for synthesis of the diazir-ines 139a and 139b. It is interesting that solvent is also cap-able of influencing the direction of the photochemical pro-cess (Table 3). Irradiation of the diazo compounds 140a and140b in pure MeOH with short-wavelength light(λ � 210 nm) leads to the Wolff rearrangement products141 or 142 in high yields, whereas use of H2O/THF orMeOH/THF mixtures reduces the total yield of nitrogenelimination products 141 or 142 and 144a or 144b and leadsto the formation of small amounts of the isomeric diazirine139a or 139b (IR: 139a: 1790 cm–1, 139b: 1770 cm–1).[103]

Photolysis of 140b in a Me2S/MeOH mixture with long-wavelength light (λ � 310 nm) led to a high yield of the di-azirine 139b (43%) together with small amounts of nitrogenabstraction products: the keto ester 142b (18 %) and the sul-fonium ylide 143b (2%).[116]

Numerous data on the behavior of cyclic diazodicarbonylcompounds have given new impetus to a revision of theproperties of their linear analogues (Scheme 56).[80,101]

It has been reported that irradiation of the diazomalonicester 136 (UV: 251, 344 nm) at room temperature leads tothe formation of two main products: the methoxymalonate137 (elimination of nitrogen) and the diazirine 135 (isomer-ization, UV: 244, 291 nm).

Indeed, both the diazoester 136 and its cyclic analogue140a display proportional dependence of the product ratiosfor these two processes on the wavelength of the irradiatinglight: the greater the wavelength, the higher the yield of


Table 3. Dependence of photolysis of the diazo compounds 140aand 140b (product yields) on the irradiating light wavelength andon the solvents.

λ [nm] Time Solvent Yields of reaction products [%][h] 139a,b 141 or 142 144a,b

�210 1–2 H2O or 5–12 56 64 3–6MeOH[a]

�210 1–2 MeOH 0 �95 0�280 6 H2O or 25 30 31 7–9

MeOH[a]

�310 9 H2O or 43 13 18 5–13MeOH[a]

[a] As a solution in THF.

Scheme 56. Photolysis of the diazomalonic ester 136.

isomerization. For the linear compounds, however, the dif-ferences in the ratios of reaction pathways as a function ofthe light wavelength are not as pronounced as for their cy-clic isomers.

Triplet-sensitized photolysis of 136, as in the case of itscyclic analogue 140a, led to diazo group reduction, in thiscase to form the malonic ester 171.[98] The mechanism ofphotolysis product formation in this case, as a whole, alsocorresponds with that for the cyclic diazo-Meldrum’s acid(140a).[80,101]

4. Conclusion

The valence isomerism between diazo compounds anddiazirines is a unique example of induction of profoundchanges by shifting of single chemical bonds that act likeswitches between linear and cyclic forms, producing pairsof compounds amazingly different from each other in theirphysical and chemical properties. The collected observa-tions on these isomerizations are still not large in numbereven for the two classes of compounds. The isomerizationscan be initiated thermally and/or photochemically, theyields of diazirines or diazo compounds obtained in thisway not being high as a rule. Shorter-wavelength lightmainly initiates nitrogen elimination, whereas longer-wave-length irradiation leads to isomerization products to somedegree. Clearly, thermal isomerization is effective only incases of thermodynamically stable products.

In certain cases diazo-to-diazirine isomerization (barbit-uric acid or Si derivatives) or diazirine-to-diazo compoundevolution (H, alkyl, α-halo, monoaryl derivatives) are infact unidirectional processes. In other cases isomerizationcan be proceeded reversibly, although the yields of original


compound resulting from two consecutive isomerizationswill not be high [about 3% for the α-ketone family, 6 % forthe malonic ester family, 9% for the α-amide family, 14%for the α,α-difluoroalkyl family, 20 % for the diazafluorenefamily, and 39% for the Meldrum’s acid family (very roughyields are given)].

Interestingly, several functional isomers other than diaz-uirines and diazo compounds exist for CH2N2 sets of atoms(cyanamides, isocyanamides, nitrylimines and carbodi-uimides) and can also be involved in mutual valence isomer-ization.[94]

Certainly, diazo–diazirine isomerism has a great effect onmany aspects of chemistry and ranks with other electro-cyclic reactions involving isomerism of heterocyclic three-membered rings, including the oxaziridine-to-nitrone,[117]

aziridine-to-azomethine ylide,[118] epoxide-to-carbonylylide,[119] and azirine-to-nitrile ylide[120] rearrangements.

Success in experimental chemistry has initiated the devel-opment of a theory. A significant number of calculationsutilizing different approaches have been performed to ex-plain the formation or lack thereof of these or other inter-mediates during thermal and photochemical isomeriza-tion.[20,87,98,102,106,115,121] However, the following phrasesstill apply often enough:

“... there are still controversies about the photolytic andthermolytic reaction mechanisms of diazirines ...”,[121e]

“... experimental information has never been successfullyrationalized from a mechanistic point of view”,[121j]

“... which is an artefact of the calculations…”,[20]

“These complexities have made a comprehensive under-standing of diazirine photochemistry an elusive goal”,[122]

“... no rules exist to predict the relative efficiency of theformation of the diazo intermediate competes with the pro-duction of carbene”.[67]

Nevertheless, considerable progress in chemical theoryshould form a good basis for further discoveries of newfundamental principles of mutual isomerization between di-azirines and the diazo compounds as well as their use.

[1] T. Curtius, J. Prakt. Chem. 1889, 39, 107–139.[2] H. v. Pechmann, Ber. Dtsch. Chem. Ges. 1894, 27, 1888–1891.[3] a) A. Angeli, Atti. Reale Accad. Lincei 1907, 16, 790; b) A.

Angeli, Atti. Reale Accad. Lincei 1911, 20, 626.[4] J. Thiele, Ber. Dtsch. Chem. Ges. 1911, 44, 2522–2525.[5] a) I. Langmuir, J. Am. Chem. Soc. 1919, 41, 1543–1559; b) I.

Langmuir, J. Am. Chem. Soc. 1920, 42, 274–292.[6] H. Boersch, Monatsh. Chem. 1934, 65, 311–337.[7] a) D. A. Ramsay, J. Chem. Phys. 1949, 17, 666–667; b) B. L.

Crawford, W. H. Fletcher, D. A. Ramsay, J. Chem. Phys. 1951,19, 406–412.

[8] K. Clusius, U. Lüthi, Helv. Chim. Acta 1957, 40, 445–456.[9] A. P. Cox, L. F. Thomas, J. Sheridan, Nature 1958, 181, 1000–

1001.[10] S. R. Paulsen, Angew. Chem. 1960, 72, 781–782.[11] a) E. Schmitz, R. Ohme, Chem. Ber. 1961, 94, 2166–2173; b)

E. Schmitz, R. Ohme, Tetrahedron Lett. 1961, 2, 612–614.[12] L. Pierce, V. Dobyns, J. Am. Chem. Soc. 1962, 84, 2651–2652.[13] E. Schmitz, R. Ohme, R.-D. Schmidt, Chem. Ber. 1962, 95,

2714–2717.[14] For general reviews on diazo compounds, see: a) R. Huisgen,

Angew. Chem. 1955, 67, 439–463; b) O. P. Studzinsky, I. K. Ko-robitsyna, Russ. Chem. Rev. 1970, 39, 834–843; c) The Chemis-


try of Diazonium and Diazo Groups (Ed.: S. Patai), Wiley, NewYork, 1978, pp. 1–1045; d) M. Regitz, G. Maas, Diazo Com-pounds. Properties and Synthesis, Academic Press, Orlando,Fla, 1986, pp. 1–596; e) G. Maas, in: Methoden der OrganischenChemie (Houben-Weyl) (Ed.: M. Regitz), Thieme, Stuttgart,Germany, 1989, vol. 19b/2, pp. 1022–1409; f) T. Ye, A. M.McKervey, Chem. Rev. 1994, 94, 1091–1160; g) H. Zollinger,Diazo Chemistry II: Aliphatic, Inorganic and OrganometallicCompounds, VCH, Weinheim, Germany, 1995, pp. 1–522; h) H.Heydt, in: Science of Synthesis (Ed.: A. Padwa), Thieme, Stutt-gart, Germany, 2004, vol. 27, pp. 843–935; i) A. F. Noels, An-gew. Chem. 2007, 119, 1228; Angew. Chem. Int. Ed. 2007, 46,1208–1210; j) Y. Sadakane, Yakugaku Zasshi 2007, 127, 1693–1699; k) Y. Zhang, J. Wang, Chem. Commun. 2009, 5350–5361;l) N. R. Candeias, C. A. M. Afonso, Curr. Org. Chem. 2009, 13,763–787; m) G. Maas, Angew. Chem. 2009, 121, 8332; Angew.Chem. Int. Ed. 2009, 48, 8186–8195; n) J. N. Johnston, H. Mu-chalski, T. L. Troyer, Angew. Chem. Int. Ed. 2010, 49, 2290–2298; o) G. Burdzinski, M. S. Platz, J. Phys. Org. Chem. 2010,23, 308–314.

[15] For general reviews on diazirines, see: a) E. Schmitz, Angew.Chem. 1964, 76, 197–206; Angew. Chem. Int. Ed. Engl. 1964, 3,333–341; b) M. Bauer, E. Müller, in: Methoden der OrganischenChemie (Houben-Weyl) (Ed.: E. Müller), Thieme, Stuttgart,Germany, 1968, vol. 10/4, pp. 895–922; c) W. Rundel, in: Meth-oden der Organischen Chemie (Houben-Weyl) (Ed.: E. Müller),Thieme, Stuttgart, Germany, 1975, vol. 4/5b, pp. 1158–1257; d)J. B. Moffat, in: Chemistry of Diazonium and Diazo Groups(Ed.: S. Patai), Wiley Interscience, New York, 1978, pp. 1–70;e) E. Schmitz, Adv. Heterocycl. Chem. 1979, 24, 63–107; f)M. T. H. Liu, Chem. Soc. Rev. 1982, 11, 127–140; g) H.-W.Heine, in: The Chemistry of Heterocyclic Compounds. SmallRing Heterocycles (Ed.: A. Hassner), Wiley, New York, 1983,vol. 42/2, pp. 547–656; h) Chemistry of Diazirines (Ed.:M. T. H. Liu), CRC Press, Inc., Boca Raton, Fla, 1987, vol. 1–2, pp. 1–160; i) Methoden der Organischen Chemie (Houben-Weyl) (Eds.: M. Regitz, B. Giese), Thieme, Stuttgart, Germany,1989, vol. E19b, pp. 1701–2276; j) R. A. Moss, Acc. Chem. Res.1989, 22, 15–21; k) X. Creary, Acc. Chem. Res. 1992, 25, 31–38;l) G. Bertrand, B. Bertrand, Carbene Chemistry: From FleetingIntermediates to Powerful Reagents, CRC Press, New York,2002, pp. 1–320; m) R. A. Moss, Acc. Chem. Res. 2006, 39,267–272.

[16] a) G. S. Paulett, R. Ettinger, J. Chem. Phys. 1963, 39, 825–827;b) M. T. H. Liu, Chem. Eng. News 1974, 52, 3; c) E. Schmitz,C. Hörig, C. Gründemann, Chem. Ber. 1967, 100, 2093–2100.

[17] M. J. Amrich, J. A. Bell, J. Am. Chem. Soc. 1964, 86, 292–293.[18] a) H. M. Frey, Pure Appl. Chem. 1964, 9, 527–537; b) H. M.

Frey, Adv. Photochem. 1966, 4, 225–256.[19] C. B. Moore, G. C. Pimentel, J. Chem. Phys. 1964, 41, 3504–

3509.[20] J. F. Arenas, I. López-Tocón, J. C. Otero, J. Soto, J. Am. Chem.

Soc. 2002, 124, 1728–1735.[21] R. S. Sheridan, R. A. Moss, B. K. Wilk, S. Shen, M. Wlostow-

ski, M. A. Kesselmayer, R. Subramanian, G. Kmiecik-Lawry-nowicz, K. Krogh-Jespersen, J. Am. Chem. Soc. 1988, 110,7563–7564.

[22] A. K. Mal’tsev, P. S. Zuev, O. M. Nefedov, Izv. Akad. NaukSSSR, Ser. Khim. 1985, 957–958.

[23] J. R. Ammann, R. Subramanian, R. S. Sheridan, J. Am. Chem.Soc. 1992, 114, 7592–7594.

[24] E. L. Tae, M. S. Platz, Tetrahedron Lett. 1999, 40, 2875–2878.[25] I. D. R. Stevens, M. T. H. Liu, N. Soundararajan, N. Paike, J.

Chem. Soc. Perkin Trans. 2 1990, 661–667.[26] D. J. Faitfax, D. J. Austin, S. L. Xu, A. Padwa, J. Chem. Soc.

Perkin Trans. 1 1992, 2837–2844.[27] A. P. Molchanov, V. V. Diev, R. R. Kostikov, Russ. J. Org.

Chem. 2002, 38, 259–263; (translation of the article in Russian:Zh. Org. Khim. 2002, 38, 281–285).

S. M. KorneevMICROREVIEW[28] B. Erni, H. G. Khorana, J. Am. Chem. Soc. 1980, 102, 3888–

3896.[29] Diazo compounds 14a, 14b are poorly stable and were used

after photogeneration from the corresponding diazirines 13aand 13b in the form of crude reaction mixtures.

[30] J. Brunner, H. Senn, F. M. Richards, J. Biol. Chem. 1980, 255,3313–3318.

[31] G. F. Bradley, W. B. L. Evans, I. D. R. Stevens, J. Chem. Soc.Perkin Trans. 2 1977, 1214–1220.

[32] G. W. Cowell, A. Ledwith, Quart. Rev. 1970, 24, 119–167.[33] E. W. Neuvar, R. A. Mitsch, J. Phys. Chem. 1967, 71, 1229–

1232.[34] A. Padwa, D. Eastman, J. Org. Chem. 1969, 34, 2728–2732.[35] For the formation of azines several mechanisms involving par-

ticipation not only of diazo compounds but also of carbenesand diazirines have been suggested. Their occurrence in a reac-tion thus cannot unequivocally specify their precursor. All inall they could appear only as a result of some bimolecular reac-tion. See, for example, ref.[15h,34,42,46].

[36] a) R. Kupfer, U. H. Brinker, Liebigs Ann. 1995, 1721–1725; b)J.-L. Mieusset, G. Wagner, K.-J. Su, M. Steurer, M. Pacar, M.Abraham, U. H. Brinker, Eur. J. Org. Chem. 2009, 34, 5907–5912.

[37] See, for example: a) G. Maas, in: Synthetic applications of 1,3-dipolar cycloaddition chemistry towards heterocycles and naturalproducts (Eds.: A. Padwa, W. H. Pearson), Wiley, New York,2002, pp. 539–622; b) H. Pellissier, Tetrahedron 2007, 63, 3235–3285.

[38] H. Gstach, H. Kisch, Chem. Ber. 1982, 115, 2586–2591.[39] a) M. G. Rosenberg, S. M. Kam, U. H. Brinker, Tetrahedron

Lett. 1996, 37, 3235–3238; b) S. D. Isaev, G. G. Zhalnina, Z. N.Murzinova, S. I. Lastovenko, A. G. Yurchenko, Zh. Org. Khim.1988, 24, 142–148.

[40] W. Kirmse, T. Meinert, D. A. Modarelli, M. S. Platz, J. Am.Chem. Soc. 1993, 115, 8918–8927.

[41] a) M. G. Rosenberg, U. H. Brinker, J. Org. Chem. 2001, 66,1517–1522; b) The scavenger product 37 of unspecified stereo-chemistry is obviously formed from the initial cycloadditionproduct (the corresponding 4,5-dihydro-3H-pyrazole is notshown) of the diazo compound 36 to diethyl fumarate; c) Fum-arates usually react with diazo compounds with high yields.See, for example ref.[41d,e]; d) U. M. Dzhemilev, V. A. Dokichev,S. Z. Sultanov, R. I. Khusnutdinov, Yu. V. Tomilov, O. M. Ne-fedov, G. A. Tolstikov, Izv. AN SSSR, Ser. Khim. 1989, 1861–1869; e) Y. Nakano, M. Hamaguchi, T. Nagai, J. Org. Chem.1989, 54, 1135–44.

[42] R. Bonneau, M. T. H. Liu, J. Phys. Chem. A 2000, 104, 4115–4120.

[43] H. Bayley, J. R. Knowles, Biochemistry 1978, 17, 2420–2423.[44] T. Bally, S. Matzinger, L. Truttmann, M. S. Platz, S. Morgan,

Angew. Chem. 1994, 106, 2048; Angew. Chem. Int. Ed. Engl.1994, 33, 1964–1966.

[45] R. Bonneau, B. Hellrung, M. T. H. Liu, J. Wirz, J. Photochem.Photobiol. A: Chem. 1998, 116, 9–19.

[46] a) R. Kupfer, M. D. Poliks, U. H. Brinker, J. Am. Chem. Soc.1994, 116, 7393–7398; b) M. G. Rosenberg, U. H. Brinker, J.Org. Chem. 2003, 68, 4819–4832.

[47] W. Knoll, J.-L. Mieusset, V. B. Arion, L. Brecker, U. H.Brinker, Org. Lett. 2010, 12, 2366–2369.

[48] R. Bonneau, M. T. H. Liu, J. Am. Chem. Soc. 1996, 118, 7229–7230.

[49] a) A. R. Tuktarov, U. M. Dzhemilev, Russ. Chem. Rev. 2010,79, 585–610; b) T. Akasaka, M. T. H. Liu, Y. Niino, Y. Maeda,T. Wakahara, M. Okamura, K. Kobayashi, S. Nagase, J. Am.Chem. Soc. 2000, 122, 7134–7135 and references cited therein

[50] M. T. H. Liu, Y.-K. Choe, M. Kimura, K. Kobayashi, S. Na-gase, T. Wakahara, Y. Niino, M. O. Ishitsuka, Y. Maeda, T.Akasaka, J. Org. Chem. 2003, 68, 7471–7478.

[51] C60 reacts with diazomethane to yield the fulleroid,[52a,b]

whereas carbene adds to C60 to afford the methanofuller-


ene.[52c] These distinctions might be useful in differentiatingwhether carbene or diazo compound is involved as the reactiveintermediate.

[52] a) F. Diederich, L. Isaacs, D. Philp, Chem. Soc. Rev. 1994, 23,243–255; b) A. B. Smith II, R. M. Strongin, L. Brard, G. T.Furst, W. J. Romanow, K. G. Owens, R. C. King, J. Am. Chem.Soc. 1993, 115, 5829–5830; c) A. Hirsh, Y. Rubin, in: Topics inCurrent Chemistry: Fullerenes and Related Structures (Ed.: A.Hirsh), Springer, Berlin, 1999, vol. 199, pp. 1–92.

[53] For example, see: a) K. Afarinkia, M. V. Vinader, in: Compre-hensive Organic Functional Group Transformations (Eds.: A. R.Katritzky, O. Meth-Cohn, C. W. Rees), Elsevier, New York,1995, vol. 5, pp. 865–867; b) G. L. Closs, J. J. Coyle, J. Am.Chem. Soc. 1965, 87, 4270–4279; c) G. L. Closs, J. J. Coyle, J.Am. Chem. Soc. 1962, 84, 4350.

[54] G. A. Ganzer, R. S. Sheridan, M. T. H. Liu, J. Am. Chem. Soc.1986, 108, 1517–1520.

[55] R. A. G. Smith, J. R. Knowles, J. Chem. Soc. Perkin Trans. 21975, 686–694.

[56] a) M. T. H. Liu, D. H. T. Chien, J. Chem. Soc. Perkin Trans. 21974, 937–941; b) In the later publication[57] on the thermolysisof the diazirine 2e in chlorobenzene at 130 °C it was declaredthat the colorless reaction mixture turned yellow for severalseconds before the nitrogen evolution and formation of the az-ine product. This color was assigned to the chloro-phenyldiazo-methane intermediate 1e, but no clear evidence in support ofthis statement was reported.

[57] M. T. H. Liu, K. Ramakrishnan, Tetrahedron Lett. 1977, 18,3139–3142.

[58] Y. Zhang, G. Burdzinski, J. Kubicki, M. S. Platz, J. Am. Chem.Soc. 2008, 130, 16134–16135.

[59] S. Wierlacher, W. Sander, M. T. H. Liu, J. Am. Chem. Soc.1993, 115, 8943–8953.

[60] R. S. Sheridan, M. A. Kesselmayer, J. Am. Chem. Soc. 1984,106, 436–437.

[61] T. Enyo, A. Nicolaides, H. Tomioka, J. Org. Chem. 2002, 67,5578–5587.

[62] A. Nicolaides, T. Enyo, D. Miura, H. Tomioka, J. Am. Chem.Soc. 2001, 123, 2628–2636.

[63] D. E. Milligan, D. E. Mann, M. E. Jacox, R. A. Mitsch, J.Chem. Phys. 1964, 41, 1199–1203.

[64] W. R. White III, M. S. Platz, J. Org. Chem. 1992, 57, 2841–2846.

[65] J. E. O’Gara, W. P. Dailey, J. Am. Chem. Soc. 1992, 114, 3581–3590.

[66] a) R. A. Moss, J.-M. Fedé, S. Yan, J. Am. Chem. Soc. 2000,122, 9878–9879; b) When R provides a moderately stable R+

ion, ROCCl readily fragments to R+, CO, and Cl–. See, forexample ref.[66c]; c) R. A. Moss, Acc. Chem. Res. 1999, 32, 969–974.

[67] T. Wakahara, Y. Niino, T. Kato, Y. Maeda, T. Akasaka,M. T. H. Liu, K. Kobayashi, S. Nagase, J. Am. Chem. Soc.2002, 124, 9465–9468.

[68] O. L. Chapman, C.-C. Chang, J. Kolc, M. E. Jung, J. A. Lowe,T. J. Barton, M. L. Tumey, J. Am. Chem. Soc. 1976, 98, 7844–7846.

[69] W. J. Leigh, C. Kerst, R. Boukherroub, T. L. Morkin, S. I. Jen-kins, K. Sung, T. T. Tidwell, J. Am. Chem. Soc. 1999, 121,4744–4753.

[70] M. Trommer, W. Sander, A. Patyk, J. Am. Chem. Soc. 1993,115, 11775–11783.

[71] M. Trommer, W. Sander, C. Marquard, Angew. Chem. 1994,106, 816–818.

[72] M. Zielinski, M. Trommer, W. Sander, Organometallics 1999,18, 2791–2800.

[73] A. Sekiguchi, H. Tanikawa, W. Ando, Organometallics 1985, 4,584–590.

[74] W. Ando, H. Tanikawa, A. Sekiguchi, Tetrahedron Lett. 1983,24, 4245–4248.


[75] C. G. Overberger, J.-P. Anselme, J. Org. Chem. 1964, 29, 1188–1190.

[76] Y. Zhang, G. Burdzinski, J. Kubicki, S. Vyas, C. M. Hadad, M.Sliwa, O. Poizat, G. Buntinx, M. S. Platz, J. Am. Chem. Soc.2009, 131, 13784–13790.

[77] a) M. T. H. Liu, B. M. Jennings, Can. J. Chem. 1977, 55, 3596–3601; b) M. P. Doyle, K. G. High, S.-M. Oon, A. K. Osborn,Tetrahedron Lett. 1989, 30, 3049–3052.

[78] C. J. Abelt, J. M. Pleier, J. Org. Chem. 1988, 53, 2159–2162.[79] a) M. T. H. Liu, I. Yamamoto, Can. J. Chem. 1979, 57, 1299–

1303; b) M. T. H. Liu, G. E. Palmer, N. H. Chishti, J. Chem.Soc. Perkin Trans. 2 1981, 53–57; c) B. M. Jennings, M. T. H.Liu, J. Am. Chem. Soc. 1976, 98, 6416–6417.

[80] A. Bogdanova, V. V. Popik, J. Am. Chem. Soc. 2003, 125,14153–14162.

[81] P. Bennett, L. M. Harwood, J. Macro, A. McGregor, J. Chem.Soc. Chem. Commun. 1994, 961–962.

[82] A. Nicolaides, T. Matsushita, K. Yonezawa, S. Sawai, H. To-mioka, L. L. Stracener, J. A. Hodges, R. J. McMahon, J. Am.Chem. Soc. 2001, 123, 2870–2876.

[83] C. G. Overberger, J.-P. Anselme, Tetrahedron Lett. 1963, 4,1405–1408.

[84] a) Y.-Z. Li, G. B. Schuster, J. Org. Chem. 1986, 51, 3804–3811;b) Y.-Z. Li, G. B. Schuster, J. Org. Chem. 1987, 52, 4460–4464.

[85] H. M. Frey, A. W. Scaplehorn, J. Chem. Soc. A 1966, 968–970.[86] E. Schmitz, A. Stark, C. Hörig, Chem. Ber. 1965, 98, 2509–

2515.[87] a) H. Rau, M. Bokel, J. Photochem. Photobiol. A: Chem. 1990,

53, 311–322; b) S. P. Walch, J. Chem. Phys. 1995, 103, 4930–4936.

[88] T. Miyashi, T. Nakajo, T. Mukai, J. Chem. Soc. Chem. Com-mun. 1978, 442–443.

[89] P. S. Engel, Chem. Rev. 1980, 80, 99–150.[90] E. D. Laganis, D. S. Janik, T. J. Curphey, D. M. Lemal, J. Am.

Chem. Soc. 1983, 105, 7457–7459.[91] a) M. Torres, P. Raghunathan, J. L. Bourdelande, A. Clement,

G. Toth, O. P. Strausz, Chem. Phys. Lett. 1986, 127, 205–209;b) M. Torres, J. L. Bourdelande, A. Clement, O. P. Strausz, J.Am. Chem. Soc. 1983, 105, 1698–1700.

[92] R. A. Franich, G. Lowe, J. Parker, J. Chem. Soc. Perkin Trans.1 1972, 2034–2041.

[93] a) E. Voigt, H. Meier, Angew. Chem. 1975, 87, 109–110; b) E.Voigt, H. Meier, Chem. Ber. 1975, 108, 3326–3335.

[94] a) E. Schmitz, Dreiringe mit zwei Heteroatomen, Springer Ver-lag, Berlin, 1967, pp. 1–114; b) E. Schmitz, Three-memberedcycles with two heteroatoms, Mir, Moscow, 1970, pp. 1–181.

[95] M. Ulbricht, J.-U. Thurner, M. Siegmund, G. Tomaschewski,Z. Chem. 1988, 28, 102–103.

[96] M. Regitz, A. Liedhegener, Chem. Ber. 1966, 99, 3128–3147.[97] G. V. Shustov, N. B. Tavakalyan, R. G. Kostyanovsky, Angew.

Chem. 1981, 93, 206–207.[98] A. Bogdanova, V. V. Popik, J. Am. Chem. Soc. 2004, 126,

11293–11302.[99] V. V. Shevchenko, N. N. Khimich, M. S. Platz, V. A. Nikolaev,

Russ. J. Org. Chem. 2006, 42, 1213–1219 (translation of thearticle in Russian: Zh. Org. Khim. 2006, 42, 1232–1238).

[100] T. Livinghouse, R. V. Stevens, J. Am. Chem. Soc. 1978, 100,6479–6482.


[101] A. Bogdanova, V. V. Popik, J. Am. Chem. Soc. 2003, 125,1456–1457.

[102] V. A. Nikolaev, N. N. Khimich, I. K. Korobitsyna, Khim. Get-erotsikl. Soedin. 1985, 321–325.

[103] V. V. Shevchenko, N. N. Khimich, M. S. Platz, V. A. Nikolaev,Tetrahedron Lett. 2005, 46, 435–438.

[104] G. Lowe, J. Parker, J. Chem. Soc. Chem. Commun. 1971,1135–1136.

[105] M. A. McAllister, T. T. Tidwell, J. Am. Chem. Soc. 1992, 114,6553–6555.

[106] M. A. McAllister, T. T. Tidwell, J. Org. Chem. 1994, 59, 4506–4515.

[107] M. R. Chedekel, M. Skoglund, R. L. Kreeger, H. Shechter, J.Am. Chem. Soc. 1976, 98, 7846–7848.

[108] Y.-Z. Li, G. B. Schuster, J. Org. Chem. 1987, 52, 3975–3979.[109] M. T. H. Liu, K. Toriyama, Can. J. Chem. 1973, 51, 2393–

2396.[110] M. E. Jung, P. Y.-S. Lam, M. M. Mansuri, L. M. Speltz, J.

Org. Chem. 1985, 50, 1087–1105.[111] R. R. Rando, J. Am. Chem. Soc. 1970, 92, 6706–6707.[112] M. A. Winnik, F. Wang, T. Nivaggioli, Z. Hruska, J. Am.

Chem. Soc. 1991, 113, 9702–9704.[113] T. Lippert, A. Koskelo, P. O. Stoutland, J. Am. Chem. Soc.

1996, 118, 1551–1152.[114] T. Lippert, P. O. Stoutland, Appl. Surf. Sci. 1997, 109, 43–47.[115] G. Burdzinski, J. Réhault, J. Wang, M. S. Platz, J. Phys.

Chem. A 2008, 112, 10108–10112.[116] V. A. Nikolaev, V. V. Shevchenko, M. S. Platz, N. N. Khimich,

Russ. J. Org. Chem. 2006, 42, 815–827 (translation of the arti-cle in Russian in: Zh. Org. Khim. 2006, 42, 840–852).

[117] D. Christensen, K. A. Jørgensen, R. G. Hazell, J. Chem. Soc.,Perkin Trans. 1 1990, 2391–2397.

[118] E. Vedejs, J. W. Grissom, J. Am. Chem. Soc. 1988, 110, 3238–3246.

[119] G. Bentabed-Ababsa, A. Derdour, T. Roisnel, J. A. Sáez, P.Pérez, E. Chamorro, L. R. Domingo, F. Mongin, J. Org.Chem. 2009, 74, 2120–2133.

[120] A. Padwa, J. Org. Chem. 2009, 74, 6421–6441.[121] For example, see: a) R. Hoffmann, Tetrahedron 1966, 22, 539–

545; b) B. T. Hart, Aust. J. Chem. 1973, 26, 461–476; c) B. T.Hart, Aust. J. Chem. 1973, 26, 477–488; d) R. Becerra, J. M.Figuera, V. Menendez, A. Tobar, R. Martinez-Utrilla, An.Quim. A 1981, 77, 63–69; e) P. L. Müller-Remmers, K. Jug, J.Am. Chem. Soc. 1985, 107, 7275–7284; f) F. Cases, F. Mora,J. M. Pérez, THEOCHEM 1988, 166, 445–448; g) S. S. Kram-arenko, A. I. Boldyrev, B. I. Stepanov, Zh. Fiz. Khim. 1990,64, 2413–2416; h) A. I. Boldyrev, P. v. R. Schleyer, D. Higgins,C. Thomson, S. S. Kramarenko, J. Comput. Chem. 1992, 13,1066–1078; i) V. Termath, D. J. Tozer, N. C. Handy, Chem.Phys. Lett. 1994, 228, 239–245; j) N. Yamamoto, F. Bernardi,A. Bottoni, M. Olivucci, M. A. Robb, S. Wilsey, J. Am. Chem.Soc. 1994, 116, 2064–2074; k) I. Fedorov, L. Koziol, A. K.Mollner, A. I. Krylov, H. Reisler, J. Phys. Chem. A 2009, 113,7412–7421.

[122] J. S. Buterbaugh, J. P. Toscano, W. L. Weaver, J. R. Gord,C. M. Hadad, T. L. Gustafson, M. S. Platz, J. Am. Chem. Soc.1997, 119, 3580–3591.

Received: February 18, 2011Published Online: August 31, 2011

valence isomerization between diazo compounds and …szolcsanyi/education/files/organicka chemia...

Documents