Ryan and Dunlea — Condensation of Aldehydes with fi-Difcetones. 69 



acetone would probably be a tri- vanillin condensation product of the 



following formula : — 



CO . CH : CH . C 6 H 3 (OCH 3 ) (OH) 

 / 

 (CH,0)(OH)C 6 H,.CH : C 



\ 

 CO . CH : CH . C 6 H 3 (OH) (OCH 3 ). 



This is borne out by Heller's own analytical data. Thus for iso-curcumin 



he gives — 



found . . C 69-38 69'35 H5"52 5-63 



(C 2l H ao 6 ) calculated . 6848 — 543 



difference . 09 0-87 0-09 0/2 



There is here a considerable discrepancy in the carbon column. On our 

 tri- vanillin acetylacetone hypothesis the results would be as follows : — 

 found . . C 69-38 H 5-52 



(C 29 H, 6 8 ) calculated . 69-32 5-17 



difference . 0-06 0-35 



We might also mention that in the introduction to his paper Heller 

 says that Knoevenagel [Ber. 37 (1904) 4480] obtained a vanillin-acetyl- 

 acetone " of obviously the following formula " : — 



(CH 3 0) (OH) C 6 H 3 . CH : CH . CO . CH 2 . CO . CH 3 . 



Knoevenagel, however, gives no structural formula for this compound, but 

 to anyone reading the communication it is apparent that the type of formula 

 which Knoevenagel had in mind was 



CH 3 . CO . C . CO . CH 3 

 CH 

 C 6 H 3 (OH)(OCH 3 ). 



From what we have just said, it is evident that to obtain a substance 

 of the type attributed to curcumin it would be necessary to protect the 

 methylene group of the diketone before attempting the condensation. This 

 we did by converting the acetylacetone into its mono- or di-methyl 

 derivative. 



Using di-methyl-acetylacetone, the condensation with benzaldehyde in 

 the presence of anhydrous hydrochloric acid yielded bright red crystals, 

 which on exposure to the air gradually split off hydrochloric acid and 

 became colourless. The colourless compound was found on analysis to have 

 the formula C 25 H 23 2 C1. When heated with pyridine it was converted into 

 another crystalline substance of the formula C 25 H 22 2 . The latter body 

 reacted with hydroxylamine to form a compound C 2i H 23 2 . jSTHOH, and 

 with bromine to form a derivative C 23 H,-0 2 Br 2 . 



[K*] 



