. A mixture of 0.110 g (0.51 mmol) of 9a and 0.050 g (0.95 mmol) of sodium methoxide in 7 ml of methanol was stirred at 25°. After 22 hr, tic showed no more starting material. The mixture was evaporated to 1 ml of liquid and placed on a silica gel preparative tic plate. Elution with 10:85:5 chloroform-Skellysolve B-ethanol gave a band at Ri 0.60 which yielded 30 mg (28%) of 12 as a light yellow oil: ir (neat) 1633 cm""1 (C=N); nmr (CC14) 2.7 (s, 5, aromatic),

The photolysis of N-iodoamides has been shown to provide a general route to y-lactones. The mechanism and stereochemical aspects of the reaction have been investigated. A general reagent for the preparation of N-iodoamides has been found in t-butyl hypochlorite-iodine.Lactonisation is most conveniently effected by photolysis of the amide in presence of excess of iodinating agent (t-butyl hypochlorite or lead tetra-acetate with iodine).THE synthesis of saturated lactones from saturated acids by oxidative procedures has not hitherto been accomplished in high yield or by a general meth0d.l In contrast the oxidative cyclisation of aromatic acids has been well attested.2 I n connection with the interrelationship of sesquiterpenoid lactones3 there was need for a general method for converting saturated acids into y-lactones. The present Paper reports the solution of this problem.A preliminary communication has already appeared.4Our first approach t o the problem was to attempt the generation of acyloxy-radicals by photochemical methods. Since the photolysis of alkyl hypohalites,5 especially hypoiodites,6 is a convenient and efficient procedure for inducing intramolecular attack of derived alkoxy-radicals on non-activated hydrogen, we studied the photolysis of acyl hypoiodites. However, in this system, decarboxylation is so rapid that there is little if any indication of the formation of y-iodo-acids or of y-lactones. In fact, the photolysis of acyl hypoiodites provides a good method for decarb~xylation.~ (11) --+ (111)VIII) with possible competition from the process (IV) --+ (VI). Reduction to practice involves the selection of X such that the radical -CO*X*We next turned to a general proposition which is summarised as (I)

No abstract

N-Methyl-and N-acetyl-biphenyl-2-carboxamidyls, generated from the corresponding Nbromides in photochemically initiated alkyl radical or bromine atom chain reactions, cyclise to give Ar,-5 and Ar,-6 products, indicating the accessibility of a low-energy excited C-state for these species. N-Methoxybiphenyl-2-carboxamidyl radical, from photolysis of the N-chloro derivative, yields no products from Ar,-5 or Ar,-6 cyclisation; this indicates a thermodynamically stable ll-ground state and a high-energy and thermally inaccessible excited Sstate. The Nmethyl biphenyl -2-sulphonamidyl radical from photolysis of the N-iodo or N-bromo derivative affords only the Ar,-6 product; this represents chemical evidence for a large n-C energy separation. MNDO calculations predict the n-ground states of Nalkoxy-amidyls and -aminyls to be more nucleophilic than those of amidyls, imidyls, and sulphonamidyls, in keeping with the observed reactivities. The reactivities and regioselectivities observed, together with e.s.r. data and MNDO results, support mixing between the ll-ground and Xexcited states of amidyls, whereas alkoxyamidyls, cyclic amidyls, and sulphonamidyls have pure nground states. Mixing in amidyls is facilitated by their acyclic nature and by a small energy difference between their ll-ground and %excited states.Amidyl (l), imidyl (2), and sulphonamidyl (3) radicals belong to a class which, because of the non-bonding electrons, can have two distinct electronic states where the electron resides upon the singly bonded heteroatom. For amidyls these states are designated ll (4) and C (5). and theoretical (MO) 2*8-1 ' evidence points to a ll-ground state for amidyls (1). Calculations point to a n-ground state of imidyls (2).' '-13 Both e.s.r.6*1e16 and ab initio calculations l o point to a II-ground state for sulphonamid yls.The search for chemical evidence for the ground state structures of (1) and (2) has revealed interesting results. Skell and his co-workers l 7 have suggested that both Xand n-states of the succinimidyl radical can be generated selectively from Nbromosuccinimide in chain reactions involving, respectively, nucleophilic primary or secondary alkyl radicals and electrophilic halogen atoms in abstraction of the N-bromine atom. Although these routes have been supported by MNDO calculation^,'^ recent papers by Tanner1* and by Walling1g have questioned the intermediacy of different radical states in chain bromination. Ab initio calculations show that the Cstate is higher in energy than the ll-state by 84-105 kJ mol-'; MNDO provides a difference of 61 kJ mol-'.'The participation of the Z-state (5) of the biphenyl-2carboxamidyl radical (ll), formed by photolysis of the Niodamide (6), is indicated by the formation of Ar,-5 cyclised products (14)--(16).20-22 Although the intermediate radical (U) is thermodynamically less stable than (27), the Ar,-5 mode of addition is irreversible at temperatures below 333 K.23 This mode is peculiar to the X-state since in the transition state (28) there is optimum orbital overlap betwee...

No abstract