Synthesis of 1 H -quinazoline-4-ones using intramolecular aromatic nucleophilic substitution

The anions of 1-(2-bromobenzoyl)-3-phenylthiourea 1 , 1-(2-chlorobenzoyl)-3-phenylthiourea 2 and 1-(2-bromobenzoyl)-3-phenylurea 8 undergo intramolecular nucleophilic substitution (putative S N Ar mechanism), and not intramolecular S RN 1 substitution, to yield 1-phenyl-2-thioxo-2,3-dihydro-1 H -quinazolin-4-one 6 and 1-phenyl-1 H -quinazoline-2,4-dione 9 respectively. Under the same reaction conditions with the addition of copper(I) iodide, phenylthioureas 1 and 2 gave a rearrangement to the respective 2-halogeno-N -phenyl- benzamides.


Introduction
Aromatic S RN 1 substitutions have been extensively researched over the last twenty years and have become a useful synthetic protocol as well as a fascinating area of organic mechanism. 12][3] We have shown that benzothiazoles can be synthesized by intramolecular S RN 1 substitution but we were only able to initiate the single electron transfer chain (SET) reaction using the process on entrainment with enolate anions of acetone and diethyl phosphite. 2ubstitution of side chain anions onto o-halogenoarenes is a good synthetic procedure and we sought to extend our S RN 1 studies 2,4 to the synthesis of 1H-quinazolin-4-ones.
The synthesis of a range of 1H-quinazoline-4-ones is shown in Scheme 1.We quickly realized that the cyclizations were S N Ar reactions and not S RN 1 substitutions.S RN 1 substitutions are formally aromatic nucleophilic substitutions but are able to take place on non-'activated' arenes because of activation by SET to yield intermediate radical anions which are able to dissociate under the reaction conditions.In contrast, S N Ar substitutions are not chain reactions and require strong electron withdrawing groups to lower the electron density on the arenes to facilitate attack by nucleophiles.The presence of the carbonyl in the ortho-position to the halogen in our precursors obviously lowers the electron density sufficiently to allow intramolecular S N Ar substitution.This facet of S N Ar reactions has not been widely identified in the literature 5,6 although we suspect that many examples have been studied but not mechanistically identified.
Our paper is a further example of taking care in not assuming S RN 1 mechanisms because the starting materials suggest that SET will be favourable.We have changed our initial mechanistic assignment in several studies. 4,7,8For instance, the reaction between phenylthiolate anions and αhalogenonitroalkanes gives substitution to α-(phenylsulfanyl)nitroalkanes in dipolar aprotic solvent by an S RN 1 mechanism but switches to yielding disulfides in protic solvents by nucleophilic attack on the halide. 4,8In a further example of apparent S RN 1 substitution, reactions between 2,6-diiodophenols and their phenolates anions to yield dityrosines, proceed by a non-SET nucleophilic substitution. 7The diagnostic tests for the S RN 1 chain reaction are well worked out and should always be used to assign the S RN 1 mechanism. 1,2,4,7,8
The yield of 5 was not significantly altered using S RN 1 diagnostic tests.Carrying out the reaction in the dark (no light catalysis) gave an isolated yield of 29%.The use of pdinitrobenzene as a strong electron acceptor to inhibit the SET step in the chain gave an isolated yield of 46%.Both reactions gave high yields of the crude product.The nature of the intermediate anion is unknown, either a mixture of mono-anions and/or the dianion.The most acidic proton is the 'imide' hydrogen but the anion must reside on the aniline-nitrogen (i.e. 3 as shown in Scheme 1).A large excess of base was used (5 equiv.)and when the amount was cut to one equivalent, the yield dropped to 37% with unchanged starting material (29%).The dianion is unlikely to be the reactive species because the electron withdrawing effect of the carbonyl would be lost in the S N Ar reaction.When no base was used only starting material was recovered (80%) thereby eliminating the thiourea as the nucleophile.Light catalysis was not further used in the studies.A mechanism proceeding via a benzyne intermediate cannot be ruled out.Our earlier studies with the related compounds, N-(3-bromophenyl)-thiobenzamide and N-(3-chlorophenyl)thiobenzamide, under the same reactions conditions only yielded unaltered starting material, whereas the corresponding 2-halogeno compounds yielded 2-phenylbenzothiazole by an S RN 1 mechanism. 2This evidence ruled out benzyne intermediates for the latter reactions, 2 but does not necessarily rule out benzyne intermediates in the present study.
The nature of the leaving group is important in S RN 1 reactions because the lower the energy of the unpaired electron in the Ar-hal bond the faster the reaction (I > Br >> Cl >> F), i.e. very slow for X = Cl. 1,3In contrast, the order for S N Ar is F >> Cl, Br, I.When the chloro starting material 2 was used the yield was also high with an isolated yield of 5 of 58%, i.e. not significantly lower.This result again strongly indicates a S N Ar mechanism.It is possible that a S RN 1 substitution is favourable for these precursors but that the S N Ar reaction is faster.
When 1-(2-bromobenzoyl)-3-phenylurea 7 was reacted under the same conditions, 1-phenyl-1H-quinazoline-2,4-dione 8 was formed in high yield as expected.None of the product 2phenylamino-benzo[e] [1,3]oxazin-4-one 9 due to S N Ar via the O-centre of the anion was observed.Small amounts of 2-bromo-N-phenyl-benzamide 13 were isolated.The reaction conditions were variable and repeat reactions (4.5 -6.7 h) gave varying amounts of the quinazoline-2,4-dione 8 and 2-bromo-N-phenyl-benzamide 13 (51-63%).We suggest that the intermediate anion 10 undergoes a novel rearrangement via a 4-membered ring intermediate 11 which extrudes isocyanic acid to yield the anion of 2-bromo-N-phenyl-benzamide 12.The driving force is provided by the irreversible extrusion of a neutral molecule of isocycanic acid.In our earlier studies we found that Cu-mediated reactions gave good yields for intramolecular reactions. 11We therefore also studied the cyclization of 1-(2-halogenobenzoyl)-3phenylthioureas 1 and 2 using Cu(I)I which gave rearrangement to 2-bromo-N-phenylbenzamide 13 and 2-chloro-N-phenyl-benzamide respectively rather than S N Ar cyclization (Scheme 3).We propose a similar rearrangement with the S-atom complexed (e.g.intermediates 14 and 15) by the Cu(I) which hinders S N Ar.The nature of the intermediate anion is not clear.The formation of copper(I) thiocyanate would be a driving force for the rearrangement.When the amount of Cu(I) was lowered from one equiv.to 0.2 equiv.some of the 1-phenyl-2-thioxo-2,3-dihydro-1H-quinazolin-4-one 5 (9%) was formed along with the benzanilide 13 indicating that full complexation by copper is required to prevent S N Ar.The chloro analogue 2 also gave rearrangement to 2-chloro-N-phenyl-benzamide (46%).1-Benzoyl-3-phenyl-thiourea also gave rearrangement in a poor reaction (14%) and 1-(2-bromobenzoyl)-3-phenylurea 7 gave an intractable mixture of products.The N-methyl analogue 16 which is not able to cyclize via the N-atom of the intermediate anion was reacted under the same conditions and gave good yields of cyclization (65%) to the benzo[e] [1,3]thiazin-4-one 17 via the S-atom of the anion.Therefore, in the reactions with 1 the N-centre of the ambident anion cyclizes faster than the S-centre.When copper(I) (one equiv.) was added the yield dropped to 26%.The oxygen analogue, 3-(2-bromobenzoyl)-1-methyl-1phenyl-urea, did not cyclize and starting material was recovered.Use of Cu(I) or S RN 1 conditions with entrainment also failed.Finally, we investigated the equivalent cyclization in the vinylogous thiourea 18 with the hope that 8-membered ring, albeit unfavourable, may be formed by S N Ar but the expected conjugate addition onto the β-position of the α,β-unsaturated acryloyl-thiourea yielded the 6membered ring 2-thioxo-tetrahydro-pyrimidin-4-one 19 in good yield.[1,3]thiazin-4-one 17 can be synthesized in reasonable yield using S N Ar from simple precursors.The presence of a carbonyl group ortho to the leaving halide in the precursors facilitates the putative S N Ar reactions as reported in the literature. 5Interestingly, the anion (or dianion) of benzoylthioureas 1 and 2 and the phenylacryloyl-thioureas 18 undergo S N Ar and nucleophilic conjugate addition via the N-centre of the ambident anion rather than the S-center as commonly observed for thioureas and thioamides.The use of added Cu(I) leads to a novel rearrangement of the side chain thiourea and urea and blocks S N Ar .

Experimental Section
General Procedures.Melting points were determined with a Koffler block.Column chromatography was performed on silica gel (Merck 60, 70-230 mesh and 230-400 for flash chromatography) with the indicated eluent.TLC was carried out using aluminum backed TLC plates of silica gel 60 F254 (Merck, 0.2 mm). 1 H NMR spectra were recorded on a Perkin Elmer R32 spectrometer at 90 MHz and 13 C NMR spectra using a Bruker WP-80 spectrometer.All spectra were recorded in CDCl 3 with TMS as the internal standard.Chemical shifts were recorded in ppm and coupling constants J are given in Hz.Mass spectra (HRMS) at 70 eV using electron impact mode were performed on a Kratos MS 80 spectrometer.Irradiation of putative S RN 1 reactions was carried out using a Photophysics MLV18 irradiator with 12 lamps (25 W) emitting at 350 nm.All reactions were carried out under an atmosphere of nitrogen.Light petroleum refers to the bp 60-80 ˚C fraction.
Synthesis of thioureas 1-benzoyl-3-phenylthiourea was prepared by a literature procedure. 9-(2-Bromobenzoyl)-3-phenylthiourea (1).General procedure for thiourea synthesis.2-Bromobenzoic acid (25 g, 0.124 mol) and thionyl chloride (59 g, 0.496 mol) were refluxed for 3 h.The excess thionyl chloride was removed by distillation in vacuo to give a clear oil (27 g, 99%).The acid chloride was added dropwise to a mixture of ammonium thiocyanate (10.3 g, 0.135 mol) and dry acetone (75 mL).The mixture was stirred and heated under reflux for 5 min.Aniline (11.45 g, 0.123 mol) in dry acetone (25 mL) were added dropwise at a rate sufficient to maintain reflux.The resulting mixture was poured into cold water.The resulting crystals were filtered, washed with water, dried and recrystallized from ethanol to yield the thiourea

Scheme 1 .
Scheme 1. Synthesis of the quinazolin-4-one 5 by a possible S N Ar mechanism.