THE STRUCTURE OF DISALICYL ALDEHYDE AND THE 
 FORMATION OF DIOXANS 
 
 BY 
 
 CLARENCE WILLIAM KREGER 
 A. B. Miami University 1919 
 
 THESIS 
 
 Submitted in Partial Fulfillment of the Requirements for the 
 
 Degree of 
 
 MASTER OF ARTS 
 IN CHEMISTRY 
 
 IN 
 
 THE GRADUATE SCHOOL 
 
 OF THE 
 
 UNIVERSITY OF ILLINOIS 
 1922 
 
K ^ 
 
 UNIVERSITY OF ILLINOIS 
 
 THE GRADUATE SCHOOL 
 
 January 1 6 1 92 3_ 
 
 I HEREBY RECOMMEND THAT THE THESIS PREPARED UNDER MY 
 
 supervision by Clarence W. Kreger 
 
 ENTITLED Th e Structure of Di salicyl Aldehyde and 
 
 The Formation of Dioxans. 
 
 BE ACCEPTED AS FULFILLING THIS PART OF THE REQUIREMENTS FOR 
 THE degree OF Master of Arts in Ch emistry 
 
 Recommendation concurred in* 
 
 Committee 
 
 on 
 
 Final Examination* 
 
 ^Required for doctor’s degree but not for master’s 
 
 _ 1 
 
Digitized by the Internet Archive 
 in 2016 
 
 https://archive.org/details/structureofdisalOOkreg 
 
ACKNOWLEDGMENT 
 
 I wish to extend my sincere thanks 
 and appreciation to Professor Roger Adams 
 for the suggestion of this problem, and 
 for the interest he has shown and the help 
 ful suggestions which he has given during 
 the experimental work. 
 

 
 
 
 
 
 
 
 
 
 
 
 
 ' 
 
the structure of disalicyl aldehyde 
 
 AND 
 
 THE FORMATION OF DIOXANS. 
 

 TABLE OF CONTENTS. 
 
 Page 
 
 I. 
 
 INTRODUCTION 
 
 1 
 
 II. 
 
 HISTORICAL PART 
 
 3 
 
 III. 
 
 THEORETICAL PART 
 
 6 
 
 
 (l) Disalicyl Aldehyde 
 
 6 
 
 
 (3) Aliphatic Bis-Dioxans 
 
 10 
 
 
 (3) The 1,3 Dioxans 
 
 13 
 
 IV. 
 
 EXPERIMENTAL PART 
 
 18 
 
 
 Preparation of: 
 
 
 
 (l) Disalicyl aldehyde 
 
 18 
 
 
 (2) Salicyl aldehyde triacetate 
 
 19 
 
 
 (3) 5, 6-benzo-2 -phenyl -1 ,3-dioxan 
 
 30 
 
 
 (4) Benzyl idene diacetate 
 
 21 
 
 
 (5) 5, 6-benzo-3-p-chlorophenyl-l ,3-dioxan 
 
 33 
 
 
 (6) 5, 6-benzo-2-p-bromophenyl-l ,3-dioxan 
 
 33 
 
 
 (7) 5,6-benzo-3-m-nitrophenyl-l ,3-dioxan 
 
 35 
 
 
 (8) Saligenin 
 
 36 
 
 
 (9) Nitro-saligenin 
 
 37 
 
 
 (10) Bromomethyl alcohol 
 
 38 
 
 V. 
 
 . SUMMARY 
 
 30 
 
 VI. 
 
 BIBLIOGRAPHY 
 
 33 
 

 
 
 
 
- 1 - 
 
 PART I 
 
 INTRODUCTION 
 
 One of the very common derivatives of salicyl aldehyde, 
 and one which is very easily formed is the so-called disalicyl 
 aldehyde. This compound has been known since the year 1845 and 
 many chemists have worked with it since that time. Despite these 
 facts no satisfactory formula has as yet been suggested for it. 
 
 The work described in this paper was taken up with the idea 
 of definitely proving that disalicyl aldehyde has the following 
 formula: 
 
 It was proposed to show that this formula accounts for all 
 the known properties of the substance* that a possible mechan- 
 ism for the formation of such a structure from salicyl aldehyde 
 can be devised and that from this mechanism conclusions can be 
 drawn as to new methods by which it should be possible to prepare 
 disalicyl aldehyde from salicyl aldehyde: that from the mechan* 
 
 ism of the formation of disalicyl aldehyde, conclusions can be 
 drawn as to the formation of a compound from saligenin and ben- 
 zaldehyde which should possess similar properties to disalicyl 
 aldehyde :and that the reaction of the formation of this compound 
 from saligenin and benzaldehyde is not limited to these two 
 substances, but is a general reaction, yielding similar compounds 
 with substituted benzaldehydes and substituted saligenins. 
 

 
 
 
 
 
 
 - 
 
 . 
 
 . 
 
 . . . 
 
 
 
 ; . *Z . 
 
 
 
 
 
 
 
 : 
 
 _ : 
 
 - j *: - • 
 
 : 
 
 ■ 
 
 l . • • - 
 
 
 
 
 
 
- 3 - 
 
 It should be mentioned here that in the inaugural dissertation 
 
 of Walter P. Bradley for the degree of Doctor of Philosophy at 
 
 Goettingen in 1890, the formula given above was mentioned as a 
 
 possibility for the structure, of disalicyl aldehyde. No proof 
 
 was given by Bradley, however, that this structure was correct, 
 
 (i) 
 
 and in publishing an abstract of his thesis, he did not even 
 mention this formula. 
 

 
 
 . 
 
 
 
 : 
 
 . 
 
 
 
 
 , ■ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 * 
 
 
 
 
 
 
 
- 3 - 
 PART II 
 
 HISTORICAL PART 
 
 (3) 
 
 As early as 1845, disalioyl aldehyde was found along with 
 
 salicyl aldehyde when the copper salt of salicyl aldehyde 7/as 
 
 (3) 
 
 distilled. In 1851, Cahours obtained this same product by the 
 
 action of benzoyl chloride upon salicyl aldehyde while attempting 
 
 ( 4 ) 
 
 to prepare the benzoyl derivative. Perkin, a little later, 
 showed that acetyl chloride or euccinyl chloride on salicyl alde- 
 hyde yielded the same product. Since that time, it has been 
 
 shown that practically all types of acid chloride cause this 
 
 ( 5 ) 
 
 same reaction to take place; thus phosphorus trichloride, 
 
 ( 6 ) (?) 
 
 urea chloride, phosgene in pyridine, and oxalyl chloride in 
 
 ( 8 ) 
 
 pyridine. Investigations have. indicated that the product is 
 
 readily prepared by the action of acetyl chloride upon salicyl 
 
 aldehyde in glacial acetic acid as a solvent: moreover, many of 
 
 the substituted salicyl aldehydes yield substituted disalioyl 
 
 (9) 
 
 aldehydes by the same treatment. 
 
 The history of the dioxans does not date back so far. Only 
 a few are known and these are of rather recent origin. The 
 dioxans are six membered ring bodies with two oxygen atoms and 
 four carbon atoms in the ring. The possible formulae for these 
 rings are:- 
 
 i 
 
- 4 - 
 
 There are no known examples of compounds having a nucleus 
 
 structure of type I. Several examples of compounds having a 
 
 nucleus structure of types II and III are known. Those of type II 
 
 are of particular interest in this paper because the dioxans 
 
 prepared by the author are of this type. These are 1,3 dioxans, 
 
 ( 10 ), 
 
 according to the nomenclature given in Meyer and Jacobson, 
 
 the numbering of the ring being as shown above. 
 
 The simplest known compound of thia type is the methylene 
 
 ether of trimethylene glycol or 1,3 dioxan. It was prepared by 
 
 ( 11 ) 
 
 Henry, in 1903, and has the following structural formula:- 
 
 C Hz 
 
 
 C H* 
 
 The methylene ether of saligenin is also a 1,3 dioxan, and 
 
 this compound is known as the hitro derivative. It was first 
 
 (13) 
 
 prepared by Borsche and Berkhout. It is made by the action 
 of formaldehyde on p-nitrophenol in the presence of dilute sul- 
 furic acid. It is 5 nitro saligenin methylene ether or 5,6-p- 
 nitrobenzo-1 ,3-dioxan. 
 
 >c Hz. 
 
 Compounds containing two 1,3 dioxan nuclei are also known. 
 

 
 . .. 
 
 
 - 
 
 . . 
 
 
 
 
 
 
 
 
 
 - . 
 
 
 
 * 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 i 
 
 
 
 
 
 
 
 
 
 
 
 
 • 
 
 
 
 
 
 
 
 
 
 - . . • 
 
 J 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 . 
 
 • • - 
 
 
 
 
 
 
- 5 - 
 
 These are of the general type: 
 
 O — CH* c 
 
 / 3 H- \ / * 3 \ 
 
 Tj-fiO * C d ' rCH-Tf 
 
 \ ’ / \ U ! / 
 
 ® — C// £//* — O 
 
 (13) 
 
 and were prepared by Read, by the action of aldehydes (RCHD) 
 on pentaerythritol (C(CH s OH) 4 ) 
 
6 
 
 PART III 
 
 THEORETICAL PART 
 
 Disalicyi Aldehyde . 
 
 The empirical formula and properties of disalicyi aldehyde 
 
 have been determined in part by many of the investigators who 
 
 ( 14 ) 
 
 obtained this product. It has the formula C 1 ^H 1 o0 3 , correspond- 
 ing to two molecules of salicyl aldehyde with a molecule of water 
 eliminated. It gives no reaction for an aldehyde or ketone group 
 with such reagents as phenyl hydrazine, hydroxyl amine or sodium 
 bisulfite. It gives no reaction for a hydroxyl group with such 
 reagents as ferric chloride, acetyl chloride or acetic anhydride, 
 and is insoluble in sodium hydroxide solution. Disalicyi aldehyde 
 is extraordinarily stable to alkali, being undecomposed after long 
 boiling with a concentrated potassium hydroxide solution or even 
 fusion with potassium hydroxide. On the other hand, concentrated 
 sulfuric acid dissolves it upon warming to give a red solution 
 which upon careful dilution yields salicyl aldehyde. Bromine gives 
 a mixture of bromo-salicyl aldehyde and dibromodisalicyl aldehyde. 
 Strong nitric aoid decomposes it and yields a certain amount of 
 dinitrosalicylic acid. 
 
 In an attempt to write structural formulas which would account 
 for the properties of the disalicyi aldehyde, two have been found, 
 (1) and (2) : 
 
 6 ) 
 
 n - 
 
 (?) 
 
 y\ 
 
 — OH — O - 
 
 — o — C H- 
 
 
- 
 
 
 
 
 
 . 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 ■ 
 
 . 
 
 
 
 
 ' ' 
 
 
 
 
 
 
 
 ■ 
 
 
 
 
 
 
 
 
7 
 
 These structures contain no hydroxyl or carbonyl groups: more- 
 
 over, being acetals, they would be expected to be stable to 
 alkalis and unstable to acids. Formula (l) would seem unlikely 
 on account of the four merabered rings which are ordinarily diffi- 
 cult to form and rare. It was, therefore, eliminated and the 
 investigation started to determine whether formula (2) was the 
 correct structure for disalicyl aldehyde. 
 
 A possible mechanism by which such a substance could be 
 produced from salicyl aldehyde is as follows: 
 
 
 V 
 
 
 - G- 
 
 O — 
 
 
 \ 
 
 0(H 
 
 \ 
 
 X- 
 
 
 - h*Q 
 
 — > 
 
 
 CH - G — 
 
 \ 
 
 o 
 
 
 — CH- 
 
 It may be seen that the first step involves the formation of an 
 acetal between the hydroxyl of one molecule of salicyl aldehyde 
 and the aldehyde group of the second molecule. This reaction 
 corresponds to that which takes place when phenols or naphthols 
 react with benzaldehyde. Phenols or naphthols do not react with 
 benzaldehyde unless a mineral acid is used and then the reaction 
 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 • 
 
 •« 
 
 . 
 
 
 
 
 
 
 
 
 
 • 
 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
- 8 - 
 
 goes readily with or without a solvent such as glacial acetic 
 
 (15) 
 
 acid. Ordinarily, triphenyl methane derivatives are obtained 
 (16) 
 
 but Claisen has shown that by working at low temperatures, 
 the acetals from certain of the phenols and naphthols could be 
 isolated as intermediate products. Whereas ordinary phenol benz- 
 aldehyde acetals tend chiefly to rearrange to triphenyl methane 
 compounds, this salicyl aldehyde hemiacetal of salicyl aldehyde 
 has a tendency to condense as shown in steps (3) and (3), which 
 involve respectively acetal formation and dehydration to form 
 di salicyl aldehyde. 
 
 If this explanation is correct, it may be understood why 
 various acid chlorides cause the reaction to take place. A small 
 amount of acid chloride reacts first with the hydroxyl of the 
 salicyl aldehyde, giving hydrochloric acid. This mineral acid 
 being the catalyst usually used for the formation of an acetal 
 causes steps (1) and (3) to occur. The remaining portion of the 
 acid chloride now acts as a dehydrating agent to remove water 
 and thus gives step (3). This deduction leads to the conclusion 
 that any very small amount of mineral acid in the presence of a 
 dehydrating agent should cause this reaction to take place. Exper 
 iments fully substantiated this. Salicyl aldehyde is unaffected 
 by cold acetic anhydride even after long standing. If to this 
 mixture, however, a small drop of concentrated sulfuric acid is 
 added, an immediate reaction takes place. The mixture turns red 
 and heat is evolved sufficient to boil the acetic anhydride if it 
 

 
 
 
 
 
 
 
 
 ' 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
- 9 - 
 
 is not cooled. Inside of three minutes after the sudden reaction 
 is over disalicyl aldehyde separates from the cooled reaction 
 mixture in yields which amount to over 80^> of the theory. In 
 place of sulfuric acid, phosphoric acid or hydrochloric acid may 
 used. 
 
 Acids are catalyzers for the decomposition as well as for the 
 formation of acetals. Treatment of disalicyl aldehyde with con- 
 centrated sulfuric acid to yield salicyl aldehyde has already 
 been mentioned. Warming with strong aoid is rather drastic treat- 
 ment , however. A very simple experiment showing the decomposition 
 of disalicyl aldehyde with acids consists in treating pure di- 
 salicyl aldehyde with an excess of acetic anhydride. No reaction 
 takes plaoe, even on long boiling; if, however, a drop of sulfur- 
 ic acid is added to the cold solution and the mixture allowed to 
 remain at room temperature, large, heavy, transparent crystals 
 separate from the solution within twenty-four hours. These 
 crystals are salicyl aldehyde triacetate. 
 
 -f- 3 (CH ^ C °)z O 
 
 O ch(ococHo ) 2 
 
 sCHs C -l 
 
 OCOCHs 
 
 The structure represented by disalicyl aldehyde involves a new 
 type of nucleus and consequently it is necessary to outline 
 
- 10 
 
 a convenient system of nomenclature. The following formula repre 
 sents the nucleus and numbered as in the figure should be called 
 a 1,3,7 bis-dioxan: 
 
 I 
 
 By using this system of nomenclature, disalicyl aldehyde is then 
 4,5-8 , 9-dibenzo-l , 3 , 7 -bi 3 -dioxan. 
 
 Aliphatic Bis-Di oxana 
 
 The formation of a bis-dioxan from salicyl aldehyde is of 
 more than ordinary interest because of the fact that salicyl aide 
 hyde should resemble in its reactions aliphatic ^-hydroxy-alde- 
 hydes. These latter substances should, therefore, be expected to 
 form bis-dioxans. It is a well-known fact that aldol, which is 
 a representative of the class of ^3 -hydroxy-aldehydes, enters 
 into a vigorous reaction on the addition of a small amount of 
 sulfuric acid, giving off maeh much heat and thus resembling the 
 action of acid on salicyl aldehyde. A certain amount of croton 
 aldehyde forms, but in addition, a large amount of by-product, 
 the nature of which is being studied. 
 
 Aldol ( / 3 -oxy-butyraldehyde) is prepared by treating acet- 
 
 , , (!■?), 
 
 aldehyde with dilute hydrochloric acid (Wurtz ). If this 
 mixture is allowed to stand for some time, water separates, and 
 

- 11 
 
 the so-called "Dialdan" is obtained. This is a crystalline body 
 melting at 139°-140°, and has the empirical formula C 8 H 14 0 3 . 
 
 This corresponds to two molecules of aldol with a molecule of 
 water eliminated. 
 
 ~H 3 0 
 
 3(CH 3 CH(0H)CH 2 -CH0) - ■■ > C 8 H 14 0 3 
 
 * (IS) 
 
 In an earlier paper, Wurtz states that this compound is the 
 anhydride of aldol: (C 4 H 7 0) ^0. 
 
 „ ( 19 ) 
 
 In a later paper Wurtz changes his opinion and assigns the 
 following structure to dialdan :- 
 
 CH - CH(OH) - CH 2 -CHO 
 
 ll 
 
 CH - CH a - CH(OH) - CH S 
 
 tl 
 
 Despite the fact that Wurtz offers several reasons for 
 assigning this structure to this so-called dialdan, its structural 
 formula is always written with a question mark following it, and 
 it has never been proven definitely just what the structure is. 
 
 Since aldol resembles salicyl aldehyde in its action toward 
 acid, and since it is a representative of the -hydroxy-alde- 
 
 hydes which should resemble salicyl aldehydes in their reactions: 
 it is logical to believe that this dialdan is none other than 
 the bis-dioxan of aldol. If the reaction of two molecules of 
 aldol losing one molecule of water is explained in the same manner 
 as was the formation of di salicyl aldehyde from salicyl aldehyde, 
 we get the following mechanism: 
 

 
 ' 
 
 
 . 
 
 
 . ... 
 
 
 
 
 
 
 
 . ... 
 
 
 
 
 
 
 
 
 I - • 
 
 * 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 , l • 
 
 * 
 
 ' 
 
 . 
 
 
 . 
 
 . 
 
 . 
 
 
 
 
 
 
 
 
HzC 
 
 
 h-c~ oh 
 
 C Hz 
 
 1 
 
 + I 
 
 GHz 
 
 1 
 
 Lc 
 
 ho-g- h 
 
 
 1 
 
 
 C H 3 
 
 - 13 - 
 
 C Ha 
 
 I 
 
 H —C — OH 
 
 GHz 
 
 1/ 
 
 H-C“ 
 
 O 
 
 
 o 
 
 gHz 
 
 c- H 
 
 I 
 
 C H 3 
 
 This gives a 1 ,3,7-bis-dioxan very similar to disalioyl 
 aldehyde and using the same nomenclature as was used in naming 
 the nucleus of disalioyl aldehyde - the so-called dial dan be- 
 comes 4, 8-dime thyl-1 ,3,7-bis-dioxan. The empirical formula for 
 the bis-dioxan is C 8 H 14 0 3 - corresponding exactly to that given 
 
 rf 
 
 by Wurtz for dial dan. 
 
 Experimental work is now under way to prove that the compound 
 which Wurtz called dial dan is really a bis-dioxan corresponding 
 to disalioyl aldehyde in structure. The proposed procedure is 
 to treat aldol in identically the same manner as salicyl alde- 
 hyde was treated in the formation of disalioyl aldehyde. (This 
 paper, experimental part). This should give a compound whose 
 physical properties are similar to those of dial dan; but whose 
 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 : 
 
 
 
 
 
 
- 13 
 
 chemcial properties correspond to those of a bis-dioxan and which 
 will lead to definite proof of the fact that the so-called di- 
 al dan is really 4,8-dimethyl* 5 ! ,3,7-bis-dioxan. 
 
 The 1,3-Dioxans 
 
 If the steps (3) and (3) described in the mechanism for the 
 formation of disalicyl aldehyde are correct, it should be possible 
 to form with ease a compound of similar properties to disalicyl 
 aldehyde by the condensation of salignnin with benzaldehyde. The 
 analogy is best shown by the two following equations: 
 
 It is true that the hydroxyl group in the saligenin is an alcohol 
 while that in the intermediate step of the formation of disalicyl 
 aldehyde is a hemiacetal hydroxyl, nevertheless the reactions 
 should proceed in somewhat a similar manner. The experiments 
 came up to expectations. The reaction between saligenin and ben- 
 zaldehyde does take place with the greatest ease and doss give a 
 
 compound which is exactly analogous to disalicyl aldehyde in 
 
 it 
 
 properties. In fact, the reaction goes so readily that/is merely 
 necessary to warm together these two substanoes in the presence 
 of a very small amount of acid without the use of a solvent or 
 

 
 
 
 
 
 
 * 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 * 
 
 . • 
 
 ■ 
 
 . ... 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ■ • “ ' 
 
 ; 
 
 * i <. 
 
 .1 . Li • ■ • 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 . • ' * t 
 
 > : - 
 
 
 
- 14 - 
 
 dehydrating agent. To get the best results in forming disalicyl 
 aldehyde, a high temperature must be avoided: likewise a high 
 
 temperature must be avoided in the formation of this new product. 
 The compound forms with such ease that even benzoic acid acts as 
 a catalyser and in fact it is preferable to use this rather than 
 hydrochloric acid, since the latter tends to cause saligenin to 
 condense with itself to a certain extent, thus lowering the yield 
 of desired product. 
 
 This new condensation product of saligenin and benzaldehyde 
 is 5,6 benzo-3-phenyl-l,3-dioxan or phenyl methylene saligenin. 
 Both names are derived from nomenclatures given in Meyer and 
 Jacobson. The dioxan nomenclature is probably the best because 
 it gives a distinct idea of the structure of the compound. 
 
 / . 
 
 The 5, 6-benzo-2-phenyl-l ,3-dioxan is a well crystallized 
 compound, unreacted upon by hydroxyl or carbonyl group reagents, 
 insoluble in and extremely stable to alkalis. It dissolves in 
 warm concentrated sulfuric acid to a red solution, which on 
 dilution gives benzaldehyde and a resinous substance, presumably 
 a saligenin condensation product. These properties are exactly 
 similar to those of disalioyl aldehyde, and the compound decomposes 
 

 
 
 
 
 
 
 
 
 
 . 
 
 
 ; 
 
 
 . ^ 
 
 
 
 
 
 
 
 . 
 
 
 
 
 . 
 
 
 
 
- 15 
 
 in a similar manner to disalicyl aldehyde when dissolved in 
 acetic anhydride and treated with a drop of sulfuric acid. Ben- 
 zylidine diacetate is readily obtained, together with a saligenin 
 
 The ease with which the condensation of saligenin and 
 benzaldehyde took place led to the conclusion that other alde- 
 hydes should react in the same way with saligenin to give similar 
 products. Experiments quickly verified this conclusion. 
 
 Para-chlorobenzaldehyde and para-bromo -benzaldehyde condense 
 very readily with saligenin to give the corresponding dioxans: 
 
 These are respectively 5, 6-benzo-3-p-chloro-phenyl-l ,3-dioxan 
 or p-chlorophenyl methylene saligenin, and 5, 6-benzo-3-p-bromo- 
 phenyl-1 ,3-dioxan or p-brcmophenyl methylene saligenin. The 
 condensations to form these dioxans take place so readily that 
 it is merely necessary to warm the mixture of saligenin and one 
 of the above aldehydes until a uniform melt is obtained. On 
 
 7*esi n 
 

- 16 
 
 standing at room temperature, the entire mass solidifies into a 
 quantitative yield of the dioxan. The halogens in the aldehyde 
 molecule activate the aldehyde to such an extent that the reaction 
 goes without the use of a cond.ensing agent such as benzoic acid 
 or hydrochloric acid. If the reaction mixture is heated longer 
 than is necessary to obtain a uniform melt, resinificaticn of the 
 saligenin takes place and the yields of the dioxan are correspond- 
 ingly lower. The condensation of saligenin with the halogen 
 substituted benzaldehydes can also be carried out in benzene as 
 a solvent. Long heating is necessary, however, and the yields 
 are not so good. 
 
 These dioxans crystallize in beautiful colorless needles of 
 varying length up to one half inch. Their properties are exactly 
 similar to those of 5,6-benzo-2-phenyl-l,3-dioxan and of disalicyl 
 aldehyde. 
 
 Saligenin also condenses with m-nitrobenzaldehyde to give 
 the corresponding dioxan. This condensation does not take place 
 so readily and benzoic acid must be used as a condensing agent. 
 
 The yields are not very good and the product could not be obtained 
 in any definite crystalline form. This condensation product is 
 5,6-benzo-2-m-nitrophenyl-l ,3-dioxan or m-nitrophenyl methylene 
 saligenin: ^ 
 
 

 
 
 t 
 
 
 
 
 
 
 
 
 . 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 • 
 
 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
 . 
 
 ... 
 
 
 . 
 
 ... • \'j. 
 
 . , ' 
 
 
 
 
 
 
 
 
 
 
 ■ 
 
 
 
 ... . . 
 
 
 
 
 
 
 V 
 
 . 
 
 < ' • . - : • • ' • • 
 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
- 17 
 
 and it possesses properties exactly similar to those of the other 
 dioxans prepared. 
 
 A striking example showing the similarity of these dioxans 
 
 and their preparation to that of disalicyl aldehyde was brought 
 
 out when an attempt was made to condense nitro-saligenin with 
 
 benzaldehyde. The nitro-saligenin was prepared by the method 
 
 ( 30 ) 
 
 suggested by Hart and Hirschf elder , and every known means to 
 
 condense it with benzaldehyde were tried without success. In 
 
 their work on the preparation of disalicyl aldehyde from substi- 
 
 ( 9 ) 
 
 tuted salicyl aldehydes, Bradley and Dains found that nitro- 
 salicyl aldehyde prevented the condensation to the corresponding 
 disalicyl aldehyde from taking place. These authors found that 
 the introduction of negative groups hinders or entirely prevents 
 the formation of the disalicyl aldehydes. f/hether this is entire- 
 ly true in the formation of the dioxans from substituted saligenins 
 and substituted henzaldehydes must be proven by further experiment- 
 al work. The behavior of nitro-saligenin would lead to the con- 
 clusion that the analogy was very close. 
 
W1 
 
 
 
 
 .. jfcf,' l 
 
 : , Qfrti -tfe;; 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
- 18 
 
 PART IV. 
 
 EXPERIMENTAL PART. 
 
 Preparation of disalicyl aldehydes (4, 5-8 ,9-dibenzo-l ,3,7- 
 bis-dioxan) . - Thirty grams (1 mole) of salicyl aldehyde and 
 35 g. (1 mole) of acetic anhydride are mixed and cooled to 0° 
 in an ice and salt mixture. One drop of concentrated sulfuric 
 acid is added and the mixture immediately stirred to male it homo- 
 genous. Within ten seconds a deep cherry-red color has developed 
 and a vigorous reaction takes place and is complete within two or 
 three minutes, after which time the mixture practically solidifies. 
 The mass is stirred thoroughly in order to be sure that all of the 
 salicyl aldehyde has reacted, and it is best to allow the mixture 
 to stand in the ice for five or ten minutes longer. By suction 
 filtration and washing with water, then drying, 34 g. of disalicyl 
 aldehyde is obtained which corresponds to over 85% of the theory. 
 There is even a small amount more of product formed by allowing 
 the filtrate from the crystals to stand for ten to fifteen minutes 
 longer in ice. 
 
 The product is best purified by recrystallization from alcohol 
 from which it forms white, prismatic needles melting at 130°(Corr. ) 
 The melting points previously reported in the literature vary from 
 137° to 130°. 
 
 If a larger amount of acetic anhydride is used, the reaction 
 takes place in essentially the same way but not so good a yield 
 results. It is also necessary to take certain precautions in using 
 larger amounts of acetic anhydride: the mixture should not be 
 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
 
 - . 
 
 
 
 
 • 
 
 . 
 
 
 
 
 ■ 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
- 19 - 
 
 allowed to stand too long before filtering the disalicyl aldehyde 
 since the latter reacts with acetic anhydride in the presence of 
 acid to form salicyl triacetate. With the smaller amount of 
 acetic anhydride, the product precipitates out so quickly that 
 there is no need for longer standing. 
 
 If the reaction mixture is not very carefully controlled in 
 an ice and salt bath, the heat evolved is so great that the mix- 
 ture will boil, and on cooling, a poorer yield of disalicyl alde- 
 hyde is obtained. 
 
 In place of concentrated sulfuric acid a drop of glacial 
 phosphoric acid may be used. Hydrochloric acid is also a catalyst 
 but it is necessary to alter the procedure slightly. One-half of 
 the acetic anhydride is saturated with dry hydrochloric gas. This 
 saturated solution is then added to the mixture of salicyl alde- 
 hyde and the other half of acetic anhydride at room temperature. 
 
 The reaction develops heat but not so much as in the case of the 
 other acids. A somewhat longer time is required for the precipita- 
 tion of the disalicyl aldehyde and the yields are not quite so 
 high. 
 
 Trichloroacetic acid may also be used as a catalyst but the 
 results are still poorer than in the cases already mentioned. 
 
 Conversion of disalicyl aldehyde to salicyl aldehyde triacetate — 
 Five grams (l mole) of disalicyl aldehyde is dissolved in 14 g. 
 
 (6 moles) of acetic anhydride and one drop of concentrated sulfuric 
 acid is added. This mixture is allowed to stand at room temper- 
 ature for a day or two. The triacetate of salicyl aldehyde separ- 
 
. 
 
 
 
 
 t 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
 
 
 
 . 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 « 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 . ' 
 
 
 
 
 
 
 . . 
 
 
 
 
 
 . 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
- 30 - 
 
 ates in large, heavy crystals of rhombic form, some of which are 
 
 one-fourth to one-half an inch on a side. They melt at 100-101° 
 
 and do not lower the melting point of salicyl aldehyde triacetate 
 
 (31) 
 
 (M. P. 100-101°) prepared according to the method of Barbier. 
 
 Preparation of 5 , 6-Benzo-3 -phenyl -1 ,3-dioxan (phenyl methylene 
 saligenin) . - Ten grams (l.l mole) of saligenin and 10 g. (l mole) 
 of benzaldehyde which has previously been saturated with benzoic 
 acid are heated for two hours on a steam cone. A light yellow 
 resinous-looking product results. The reaction mixture, after 
 standing at room temperature for two hours, is treated with 300 cc. 
 of a 5% sodium hydroxide solution and cooled in Ice so that the 
 product will solidify. The white solid is filtered and washed 
 with water. It is crude 5, 6 -benzo -3 -phenyl -1 ,3-dioxan and corres- 
 ponds to practically a quantitative yield of product. It is best 
 purified by dissolving 3 g. in 70 cc. of 95% alcohol, adding 30 cc. 
 of water, warming until a clear solution results and then allowing 
 to stand in an open beaker at room temperature. Within half an 
 hour a white crystalline precipitate forms. After two crystal- 
 lizations the pure material is produced, melting at 54°. 
 
 A drop of concentrated hydrochloric acid may be used as a 
 catalyst in place of the benzoic acid. It is then merely necessary 
 to heat the reaction mixture of saligenin, benzaldehyde and a 
 
 drop of hydrochloric acid until solution takes place. The mixture 
 is allowed to stand several hours at room temperature. The product 
 is isolated in the same manner as described above but the yields 
 

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 < 
 
 . 
 
 
 
 
 .. 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 
 
- 21 
 
 are not so good as with the benzoic acid method. If the reaction 
 mixture, using hydrochloric acid, is heated longer, the yields are 
 in general still poorer, due probably to a partial resinif icat ion 
 of the saligenin. 
 
 Subs. - 0.1343 g: C0 2 , 0.3885 g. : H 2 0, 0.0715 g. 
 
 Subs. - 0.5957 g. : 0.4810g. : C 6 H 6 43.9 g. 43.9 g. ATF© 0.319° 
 0.357° 
 
 Calc, for C 14 H 1S 0 2 : C, 79. 3.4: H, 5.66: mol. wt. 212. 
 
 Found: C. 78.97: H, 5.93: Mol. wt. 312,313. 
 
 5, 6-Benzo-3 -phenyl -1 ,3-dioxan is soluble in all the common 
 organic solvents. It does not dissolve in alkali and is unchanged 
 even after long boiling with 20% sodium hydroxide solution. When 
 treated with concentrated sulfuric acid an immediate decomposition 
 takes place with the formation of benzaldehyde and red lumps of 
 resinified saligenin. The product gives no aldehyde or ketone 
 test with sodium bisulfite, phenyl hydrazine or hydroxyl amine: 
 it gives no test with hydroxyl group reagents such as ferric chlor- 
 ide or acetic anhydride even after long boiling. | 
 
 Conversion of 5, 6-benzo-3-phenyl-l ,3-dioxan to benzyl i dine di - 
 acetate and saligenin resin . - Ten grams (l mole) of 5,6-benzo-3- 
 phenyl-l,3-dioxan are mixed with 30 g. (6 moles) of aoetic anhydride 
 and one drop of concentrated sulfuric acid at room temperature. 
 
 The mixture is then allowed to stand for several days. At the end 
 of this time the acetic anhydride is distilled off in vacuo. 
 
- 33 
 
 the residue taken up in ether, washed with water and then with 
 dilute sodium carbonate solution. The ether solution is then 
 dried over anhydrous sodium sulfate, the ether distilled, then 
 the residue distilled in vacuo. A constant boiling, colorless 
 fraction comes over at 135° at 4 mm. pressure which on cooling 
 and inoculating with a crystal of benzylidene diacetata almost 
 completely solidifies. The melting point is 45° which agrees with 
 that found in the literature for benzylidene diacetate. The 
 residue in the distilling flask consists of a light yellow, heavy 
 oil which does not distill at 350° at 4 mm. pressure, and at or- 
 dinary temperatures forms a resinous-like mass which is without 
 question a saligenin resin. 
 
 Preparation of 5.6-Benzo-3-p-chlorophenyl-l , 3-dioxan (p-chloro - 
 phenyl methylene saligenin. ) - Two grams (1 mole) of saligenin 
 
 and 3.35 grams (l mole)of .p-ehlorobenzaldehyde are heated on a 
 steam Cone until all has dissolved to a clear liquid. This re- 
 action mixture is allowed to stand at room temperature for two 
 or three hours, when the entire mass solidifies. It is then 
 treated by shaking thoroughly with 100 cc. of a 10% sodium car- 
 bonate solution and filtered by suction. The whits solid reaction 
 product is dissolved in 100 cc. of hot 95% ethyl alcohol: water is 
 added until the solution becomes murky: the solution heated until 
 clear and allowed to stand. Within fifteen minutes a heavy white 
 precipitate forms. This is the crude dioxan and corresponds to a 
 quantitative yield of the product. The dioxan is purified by 
 

 
 
 
 
 
 
 
 
 
 
 
 
- 23 
 
 several recrystallizations from alcohol - each time dissolving the 
 white solid in a slight excess of hot 95$ ethyl alcohol, adding 
 water until solution is murky, heating until clear and allowing | 
 to stand in an open beaker. On the seconcL:recrystallization, 
 the product forms in beautiful, colorless needles, about one-fourth 
 inch in length and melting at 107°-107.5° (oorr. ) 
 
 Analysis : 
 
 Substance: 0.1303 g. : C0 2 - 0.3247 g. : H 2 0, -0.00577 g. 
 
 Calc, for C^HnOaCl: C - 68.15: H - 4.46. 
 
 Found: C -67.96: H - 4. 43. 
 
 5, 6-benzo-2-p.-*chlorophenyl-l,3-dioxan is soluble in all the 
 common organic solvents. It does not dissolve in alkali, and is 
 extremely stable to alkali. Concentrated sulfuric acid decomposes 
 it immediately giving the red resin of saligenin and p-chlorobenz- 
 aldehyde. 
 
 The dioxan gives no aldehyde or ketone test with sodium bi- 
 sulfite, phenyl hydrazine or hydroxyl amine: it gives no test 
 
 with hydroxyl group reagents such as ferric chloride or acetic 
 anhydride even after long boiling. 
 
 Preparation of 5,6-benzo-3-p-bromophenvl-l .3-dioxan fn-bromophenvl 
 methylens saligenin ). - Two grams (l mole) of saligenin and 3 grams 
 
 ( 1 mole) of p-bromobenzaldehyde are heated together on a steam 
 cone until all has dissolved to a clear solution. This reaction 
 mixture is allowed to stand at room temperature for eight or ten 
 hours, when the entire mass solidifies. The solid mass is broken 
 
- 24 - 
 
 up and treated with 100 cc. of a 10fo sodium carbonate solution 
 and filtered by suction. The white solid is dissolved in 100 cc. 
 of hot 95 $ sthyl alcohol, water is added until the solution be- 
 comes murky, and it is then heated until clear. On standing in 
 an open beaker, a heavy white precipitate forms within a half 
 hour. This is the crude dioxan and corresponds to a quantitative 
 yield of the product. It is purified by several recrystallizations 
 from ethyl alcohol, each time dissolving the white solid in a 
 slight excess of hot 95$ ethyl alcohol, adding water until solu- 
 tion is murky, heating until clear and allowing to stand in an 
 open beaker. On the second recrystallization, the dioxan forms 
 in beautiful, colorless needles of varying length up to one-half 
 inch and melting at 117°-117.5° (corr. ) 
 
 This dioxan as well as the chlor substituted one can also be 
 condensed in benzene as a solvent. Similar amounts of saligenin 
 and aldehyde are dissolved in 35 cc. benzene and the solution re- 
 fluxed for fourteen hours. The benzene is then distilled off 
 and the sticky white solid is taken up in and recrystallized from 
 ethyl alcohol in the same manner as described above. The yields 
 are not so good - the best obtainable being about 75$ of theory. 
 
 Analysis : 
 
 Substance: 0.1183 g. : C0 2 , 0.3507 g. : H s 0, 0.004355 g. 
 
 Calc, for C^H^OsBr: C - 57.73: H-3.78 
 
 Found: C,~57.79: H-3.68 
 
-35- 
 
 5,6-benzo-3-p.-bromophenyl-l,3-dioxan possesses properties 
 exactly similar to those of 5,6-benzo-3-p-chlorophenyl-l ,3-dioxan. 
 
 Preparation of 5 , 6-benzo-3-m-nltrophenyl-l ,3-dioxan (m-nitro- 
 phenyl methylene saligenin) . - Two grams (l mole) of saligenin 
 and 2.4 grams (l mole) of m-nitrobenzaldehyde are heated together 
 on a steam cone for two hours in the presence of a small amount 
 of benzoic acid as a condensing agent. The reaction mixture is 
 allowed to stand at room temperature for a day. A heavy oil 
 is obtained. This oil is treated with 100 cc. of a 10$ sodium 
 carbonate solution, and is then dissolved in 50v cc. of hot 
 95$ ethyl alcohol. Just enough water is added to produce slight 
 murkiness: the solution heated until clear and allowed to cool 
 
 in an open beaker. After an hour a flucculent precipitate will 
 form. If an oil should appear in this solution while it is hot, 
 it will solidify on cooling. When dry, the dioxan has a yellowish 
 tinge. The yield is never better than 50$ of theory. The dioxan 
 can be recrystallized from ethyl alcohol in a similar manner to 
 the other dioxans, but no definite crystalline structure can be 
 obtained. A light flocculent product is always obtained which 
 melts at 86°-89.5° (corr.) 
 
 Analysis : 
 
 Substance: 0.1815 g. : N gas - 15.45 cc. (corr.) 
 
 Calc, for C 14 H ai 0 4 N: N - 5.447 $ 
 
 Found: N - 5.33$ 
 
- 36 
 
 5, 6-benzo-3-m-nitrophenyl-l ,3-dioxan possesses properties 
 exactly similar to those of the other dioxans prepared. 
 
 Preparation of Saligenin (o-hydroxy benzyl alcohol) . 
 
 The saligenin used in the preparation of the dioxans is prepared 
 
 (33) 
 
 according to the method of Heckel. This method involves the 
 reduction of salicyl aldehyde by hydrogen gas, platinum black 
 having been used as a catalyzer and the reduction carried out in 
 alcohol as a solvent. Two parts (60 cc. ) of ethyl alcohol and 
 one part (30 cc. ) of salicyl aldehyde are placed in the container 
 of a shaking machine, and one gram of platinum black is added. 
 
 The hydrogen is led in under pressure in order to determine the 
 amount absorbed. The reduction goes quantitatively. The alcohol 
 is distilled off in vacuo and the saligenin crystallized from hot 
 benzene. 
 
 Heckel prepared the platinum black by the method of Will*- 
 (33) 
 
 statter. , i.e. , by the reduction of chlor-platinic acid with 
 formaldehyde in the presence of alkali in the cold. If the 
 calculated amount of chlor-platinic aoid to give one gram of 
 platinum black is added to 4-6 grams of sodium nitrate, and the 
 mass fused, a platinum black is obtained, by washing away the 
 soluble salts, which causes the reduction of salicyl aldehyde 
 to saligenin to go 10-13 times as fast as when Willstatters plat- 
 inum black is used. This method for the preparation of platinum 
 black given above was suggested by Hr. Vorhees in this laboratory. 
 

 
 
 
 
 e 
 
 
 
 
 
 
 
 ' 
 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 * * t 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 . « 
 
 
 
 
 
 
 
 
 
 
- 37 - 
 
 Preparation of nitro-saligenin (3-nitro-6-hydroxy benzyl alcohol ) 
 
 Nitro-saligenin is prepared by several methods. Hart and Hirsch- 
 (30) 
 
 felder, give a foot-note to their article on the preparation of 
 
 p-hydroxy-m-nitrophenyl carbinol, in which they suggest that 
 
 nitro-saligenin may be prepared in a similar manner to the above 
 
 carbinol, i.e. , by the action of bromomethyl alcohol on p-nitro- 
 
 phenol in the presence of a small amount of fused zinc chloride. 
 (34) 
 
 Fishman prepares the 3-nitro-4-hydroxy-benzyl alcohol by the 
 
 action of formaldehyde on o-nitrophenol in the presence of a 
 
 large amount of concentrated hydrochloric acid. When p-nitrophenol 
 
 is used instead of o-nitrophenol, nitro saligenin should result. 
 
 The method was found to give very poor yields. 
 
 (35) 
 
 Einhorn, Bischkopff and SzelinskJ., prepare nitro-saligenin 
 
 by treating m-nitro-o-oxybenzyl amine with nitrous acid. 
 
 „ (36) 
 
 Eichengrun, prepares it by the action of chlormethyl 
 
 alcohol on p-nitrophenol. This is of course, similar to the 
 
 (30) 
 
 method suggested by Hart and Hirschf elder. 
 
 Nitro-saligenin is prepared according to the method of Hart 
 and Hirschf elder as follows: 
 
 30 grams of pure p-nitrophenol are mixed in the cold with 
 35 grams (an excess) of bromomethyl alcohol and 5 grams of powder- 
 ed fused zinc chloride are added. The mixture is kept cold for 
 twelve hours and then allowed to run at room temperature for 
 several days. The p-nitrophenol first dissolves in the bromo- 
 methyl alcohol and then the entire mass solidifies. One liter 
 

 
 
 
 
 
 , 
 
 
 
 
 ’ 
 
 
 • • >c ; • 
 
 
 * 
 
 
 
 ' 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 • * 
 
 
 
 
 
 
 . 
 
 
 
 
 . 
 
 
 
 
 
 
 
 • . . ; . t . BbT' I 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
- 28 - 
 
 of water is added and the mixture steam distilled to remove any 
 
 unchanged p-nitrophenol. The steam distillation is carried on 
 
 for four hours, at the end of which time, the liquid remaining 
 distilling 
 
 in the /flask is poured into a beaker, and allowed to cool. An 
 oil separates which solidifies on further cooling into colorless 
 needles, of nitro-saligenin. A gummy tar remains in the distill- 
 ing flask as a by-product. The nitro-saligenin is recrystallized 
 from hot water and forms in almost colorless crystalline needles. 
 The yield is about 32$ of theory. The product was recrystallized 
 several times from hot water. The purest product obtained, 
 melted at 135°-136° (corr. ) The melting point recorded in the 
 literature is 126°. This melting point is regarded as erroneous 
 in the light of the present work, and the correct melting point 
 for nitro-saligenin is given as 136°. 
 
 Preparation of bromomethyl a lc ohol - BrCH a QH . 
 
 Bromomethyl alcohol is prepared according to the method of 
 
 (37) 
 
 Henry, by treating formaldehyde with hydrogen bromide in the 
 cold. The hydrogen bromide is prepared by passing a mixture of 
 hydrogen and bromine vapors over heated platinum, and the entire 
 preparation of hydrobromic acid gas and bromomethyl alcohol can 
 be carried out in one continuous apparatus set up. Hydrogen 
 from a tank is bubbled thru concentrated sulfuric acid and then 
 thru dry bromine. The bromine vaporizes sufficiently at room 
 temperature and the mixture of hydrogen and bromine vapors is 
 

 
 
 
- 29 - 
 
 passed into a combustion tube containing a spiral of platinum 
 wire heated to a dull red. The catalytic union of hydrogen and 
 bromine to form hydrobromi.c.. acid gas goes quantitatively, based 
 on the bromine present. The hydrogen should be passed in at 
 such a rate as to insure an excess of it over the bromine. 
 
 The hydrogen bromide is then led through a trap cooled in 
 ice to collect any unchanged bromine and also to cool the gas. 
 
 It is then passed directly into a 40 $ solution of formaldehyde 
 cooled in a freezing mixture of ice and salt. 135 cc. of 40$ 
 formaldehyde are treated with approximately 370 grams of hydrogen 
 bromide in the course of seven hours. After some time, a heavy 
 red liquid begins to separate. This is essentially bromomethyl 
 alcohol and is formed in an amount corresponding to an 85$ yield. 
 When the required amount of hydrogen bromide has been passed into 
 the formaldehyde solution (ascertained by the fact that no more 
 hydrogen bromide is absorbed), the alcohol is quickly separated 
 from the supernatent aqueous HBr by means of a separatory funnel. 
 The bromomethyl alcohol is kept in a tightly stoppered bottle in 
 the cold, as it decomposes at 30° or more. 
 

 
 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
- 30 - 
 
 PART V 
 SUMARY. 
 
 1. Disalicyl aldehyde has been shown to have the following 
 formula: 
 
 and may be called 4,5-8,9-dibenzo-l,3,7-bis-dioxan. It accounts 
 for all of the known properties of this substance. 
 
 3. A possible mechanism for the formation of such a structure 
 from salicyl aldehyde is given and from this mechanism conclusions 
 have been drawn as to new methods by which it should be possible 
 to prepare disalicyl aldehyde from salicyl' aldehyde. 
 
 3. Disalicyl aldehyde may be formed almost quantitatively 
 by the action of a drop of mineral acid upon a solution of salicyl 
 aldehyde in acetic anhydride. 
 
 4. The disalicyl aldehyde may be decomposed in the presence 
 of an excess of acetic anhydride by a drop of mineral acid to 
 form salicyl triacetate. 
 
 5. The suggestion is made and proof outlined whereby it should 
 be possible to show that the so-called dialdan - the product re- 
 sulting from two molecules of aldol with a molecule of water elim- 
 inated - is really a bis-dioxan with a structure and properties 
 very similar to disalicyl aldehyde: 
 
 

hi 
 
 1 
 
 H -C 
 
 N~€ — 
 
 I 
 
 C«3 
 
 - 3l - 
 
 c — o 
 
 ‘V 7 
 
 o 
 
 CH. 
 
 c- 
 
 
 5 
 
 o 
 
 X 
 
 w 
 
 <&- 
 
 0 
 
 M 
 
 and which may be called 4, 8, dimethyl-1 ,3, 7-bis-dioxan. 
 
 6. It was concluded from the mechanism of the formation of 
 disalicyl aldehyde that a compound should form from saligenin and 
 benzaldehyde which should possess similar properties to disalicyl 
 aldehyde. This substance does form and may be looked upon as 
 5, 6-benzo-3 -phenyl -1 ,3-dioxan, or phenyl methylene saligenin, 
 forming white plates with a melting point of 54° and having the 
 structure : 
 
 h. 
 
 X\/ c \ 
 
 0 
 
 1 
 
 c — 
 
 W Vi 
 
 7. It was concluded from the ease with which 5,6-benzo- 
 2 -phenyl -1 ,3-dioxan was prepared that other 1,3 dioxans could be 
 prepared from saligenin and substituted benzaldehyde3 with the 
 result that the following 1,3 dioxans were prepared: 
 
 5, 6-benzo-3-p-rChlorophenyl^l ,3-dioxan, forming in colorless 
 needles; melting at 107°-107. 50° , and having the structure; 
 
 /\A" 
 
- 33 
 
 5, 6-benzo-3-p-bromophenyl-l ,3-dioxan, forming in colorless 
 
 needles. 
 
 melting at 117°-117.5°, and having the structure: 
 
 5 , 6-benzo-3-m-nitrophenyl-l , 3-dioxan, 
 crystalline form: and melting at 88°-89.5 
 structure : 
 
 having no definite 
 °, and having the 
 
 8. New information regarding the preparation of saligenin 
 is given, and details for the preparation of bromomethyl alcohol 
 and nitro-saligenin are given. 
 
 9. The melting point of nitro-saligenin given in the liter- 
 ature an 136° was found to be erroneous, and 136° is shown to be 
 the correct melting point. 
 

 
 
 
 
 
 
 ■ 
 
 
 . 
 
 , . 
 
 . 
 
 
 
 
 
 
 
 
- 33 - 
 
 BIBLIOGRAPHY. 
 
 1. Bradley, Ber. 23, 1134 (1889) 
 
 3. Ettling, Ann. 53, 77 (1845) 
 
 3. Cahours, Ann. 7J3, 328 (1851) 
 
 4. Perkin, Ann. 145 , 399 (1868) 
 
 5. Zwenger, Ann. Spl. J5, 43 (1870) 
 
 6. Gatterman, Ann. 344 , 46 (1888) 
 
 7. Einhorn, Ber. 38. 3830 (1905) 
 
 8. Adams, J. Am. Chem. Soc. 37 , 2719 (1915) 
 
 9. Bradley, Ber 23, 1134 (1889): Am. Chem. J. 14, 393 (1893): 
 
 Dains, A zn. Chem. J. 16, 634 (1894). 
 
 10. Meyer and Jacobson - Lehrbuch der organischen chemie. 
 
 Vol. 3, Pt. 3, Ab't. 3, pp. 1145-1149. 
 
 11. Henry, Cent. Bl't. 1903 II - 939. 
 
 12. Borsche and Berkhout, Ann. 330 , 83-107 (1904). 
 
 13. Read, J. of the Chem. Soc. 101 , II, 2090 (1913). 
 
 14. Previous references and also Rivals, Compt. rend. 134 , 
 
 368 (1897). 
 
 15. Russanow, Ber. 32., 1944 (1889). Michael, Am. Chem. J. £,130(1886 
 
 16. Olaisen, Ann. 337 , 369 (1887). 
 
 17. Wurtz, Bull. Soc. Chim. 28, 169 (1877). 
 
 18. Wurtz, Liebigs Jahresbericht 1873, 448. 
 
 19. Wurtz, Liebigs Jahresbericht 1876, 484. 
 
 30. Hart and Hirschf elder , J. Am. Chem. Soc. 43, II, 2683 (1930) 
 

 
 
 f 
 
 
 
 
- 34 - 
 
 31. Barbier, Bull. soc. chim. (3) 33, 53 (1880). 
 
 33. Heckel, n The Synthesis of Saligenin," University of Illinois 
 Thesis, 1931. 
 
 33. Willstatter, Ber. 54, 113 (1931). 
 
 34. Fishman, J. Am. Chem. Soc. 43, II, 3388 (1930) 
 
 35. Einhorn, Bischkopff and Szelinski, Ann. 343 , 343 (1905). 
 
 36. Eichengrun, Verh. d. Vers. Deutsch. Ntf. u. Arzte, 1901 -II, 
 
 140: Chem. Centr. 1903 - II, 894. 
 
 37. Henry, Ber. 37 R , 336 (1894): Bull. acad. roymed. Belg. 
 
 (3), 36, 615. 
 
U |i*i* VERSrrY 0F ,LL 'NOIS-URBANA 
 
 3 0112 108856755