EXCHANGE 
 
 P 
 
 
Rearrangements of Some New Hydroxamic 
 
 Acids Related to Heterocyclic Acids, 
 
 and to Diphenyl- and Triphenyl- 
 
 Acetic Acids 
 
 EXCHANGE 
 
 1922 
 
 CHARLES DEWITT KURD 
 
Rearrangements of Some New Hydroxamic 
 
 Acids Related to Heterocyclic Acids, 
 
 and to Diphenyl- and Triphenyl- 
 
 Acetic Acids 
 
 A DISSERTATION 
 PRESENTED TO THE 
 
 FACULTY OF PRINCETON UNIVERSITY 
 
 IN CANDIDACY FOR THE DEGREE 
 
 OF DOCTOR OF PHILOSOPHY 
 
 BY 
 
 CHARLES DEWITT KURD 
 
 \x 
 
 E ASTON, PA.: 
 
 ESCHENBACH PRINTING COMPANY 
 1922 
 

 
 
 REARRANGEMENTS OF SOME NEW HYDROXAMIC ACIDS RE- 
 LATED TO HETEROCYCLIC ACIDS AND TO DIPHENYL- AND 
 TRIPHENYL-ACETIC ACIDS. 
 
 A summary of the serious attempts to explain the mechanism of the 
 Beckmann rearrangement must necessarily include the work of Hooge- 
 werff and van Dorp, 1 Hantzsch, 2 Stieglitz, 3 and Jones. 4 ' 5 The types of 
 compounds generally assumed to undergo this rearrangement are the 
 azides, mono-substituted /3-hydroxylamines, monobromo-amines, oximes, 
 amidoximes, oximido acid esters, acid halogen amides and the hydroxamic- 
 acid derivatives. The azides are assumed to decompose as follows : 
 
 R CO N:N > R CO N: + N, > R N:C:O 
 
 R CHr-N : N, > R CHN : + N, > R N : CH,. 
 
 These two cases, together with the mono-substituted /3-hydroxylamines 
 presented below, furnish examples of the Beckmann rearrangement that 
 have never been explained successfully except by the theory of Stieglitz and 
 of Jones. 
 
 (O,H 6 )3C NHOH > (CH 6 ),C N: + H,O > (CH6)jC :NCH 6 . 
 
 Stieglitz was the first investigator to propose the hypothesis that univa- 
 lent nitrogen derivatives form the primary decomposition products in these 
 rearrangements. His extensive experimental work has demonstrated the 
 soundness of this postulate. 6 Some years later, Jones formulated the 
 reaction mechanism in a detailed manner by applying the theory of elec- 
 tron valence to the interpretation given by Stieglitz. The electron mechan- 
 ism was represented as follows. 4 ' 5 
 
 O 
 
 R +-C+-N + ^-[OC+tN-+R] >- OCJZN-+R. 
 
 Unless there were a driving force acting to cause the transfer of the 
 radical from carbon to nitrogen, it would still be difficult to imagine why the 
 
 1 Hoogewerff and van Dorp, Rec. trav. chim., 6, 373 (1887) ; 8, 173 (1889). 
 
 2 Hantzsch, Ber., 35, 228, 3579 (1902); ibid., 27, 1256 (1894). 
 
 3 Stieglitz, Am. Chem. J., 18, 75 (1896); 29, 49 (1903); Stieglitz and Earle, 30, 399, 
 412 (1903); Stieglitz and Leech, J. Am. Chem. Soc., 36, 272 (1914); etc. 
 
 Jones, Am. Chem. /.,48, 1 (1912); 50, 414 (1913). 
 
 5 /. Am. Chem. Soc., 36, 1268 (1914); etc. 
 
 6 The rearrangement of ketoximes, which was originally attributed by Stieglitz 
 to the formation of univalent nitrogen derivatives, is now regarded as an exception. 
 His present view assumes "the rearranging power of an intermediate hydrochloride of 
 a univalent nitrogen derivative, acting in place of the ordinary free univalent nitrogen 
 compound." See Montagne (Ber., 43, 2015 (1910)), Schroeter (ibid., 44, 1207 (1911)), 
 and Stieglitz. 
 
REARRANGEMENTS OF NEW 
 
 2423 
 
 reaction should proceed, even with the explanation offered. Such a driv- 
 ing force is to be found in the shifting of electrons within the molecule. 
 Jones stressed particularly that the free valences of univalent nitrogen 
 afforded the ''stage setting required to furnish a suitable environment in 
 which the essential action," viz., the transfer of the radical from carbon to 
 nitrogen might take place. In a footnote to an article published some 
 months later, 3 Stieglitz stated that, in regard to the most fundamental 
 questions of these rearrangements "postulating a shifting of electrons from 
 carbon to nitrogen, and a migration of a positive radical, Professor Jones 
 and the writer are happily in entire agreement." 
 
 As viewed from present day standards, this interpretation with very 
 slight modifications, still holds. In late years, worthy evidence 7 has been 
 submitted to show that the positive charges possessed by an atom are 
 centered in its nucleus. In compounds which are formed by the sharing 
 of negative electrons, a "bond" consists of a pair of electrons held in com- 
 mon between two atoms. This theory of a chemical bond requires sym- 
 bols in which positive charges cease to appear, except as they form an 
 integral part of the nucleus of any atom. Such an interpretation would 
 mean simply that a symbol, Ct+N, employed at a time when ion charges, 
 rather than electrons in the present sense, held the attention of chemists, 
 would now become C : N ; that is, the carbon atom and the nitrogen atom 
 share a single pair of electrons. This conception of paired electrons would 
 account just as readily for the driving force necessary to bring about 
 rearrangement. Thus, the system of symbols C : N > C : N is 
 equivalent to C*+N > C+IN previously employed. 
 
 Possibly the formula C+ZN contains an assumption not necessarily 
 implied by the formula C : N in which two pairs of electrons are shared, 
 since the former suggests that a compound which contains such a group 
 would not be exactly non-polar, although it should not be regarded as 
 polar in the sense in which Na+ Cl~ must be. To meet difficulties of this 
 kind, Lewis employs formulas in which the pairs of electrons are placed 
 nearer one symbol than another, e. g., A : B, which implies that the mole- 
 cule AB shows some polar characteristics. With these modifications to 
 adapt the old electronic formulas to present day practices, equations may 
 be given to represent the rearrangement of a univalent nitrogen deriva- 
 tive of a hydroxamic acid. 
 
 Formulas I, la and Ib represent the intermediate univalent nitrogen 
 derivative; Formulas II, Ila and 116 represent a transition stage, the re- 
 arrangement of the positive radical, R. It will be noted -that, in Formula 
 II, although the carbon atom still has its. "octet" completed, the nitrogen 
 atom has only 6 electrons in the outer shell. By the sharing of electrons, 
 both the nitrogen and the carbon atoms may complete their octets. This 
 7 G. N. Lewis, Langmuir and others. 
 
2424 7 
 
 : R 
 
 ,W- JOXES AND CHARLES D. KURD. 
 
 n 
 
 Fig. 1. 
 
 A A 
 
 / 
 
 ">->R 
 
 / /\ 
 
 7 /! A A A 
 
 A.--. .__x. 
 
 
 
 
 
 /ti / c 
 
 / 
 
 / \ 
 
 / Q / Q 
 
 / / ''" / /' N / 
 
 
 
 
 
 i 
 
 
 
 " N / 
 
 
 An 
 
 r 
 
 
 
 n- 
 
 Fig. 2. 
 
 m 
 
 R 
 
 _. + t + ^.^ .- . + _+ __^ 
 
 (16). (116). (Illfc). 
 
 is represented in Formulas III, Ilia and III6, the isocyanate stage in the 
 rearrangement. 
 
 With this conception, it is easy to understand why the radical R is able 
 to part company with carbon and attach itself to nitrogen. No hypothe- 
 sis has been offered, however, to explain why one radical R will do so with 
 much greater readiness than some other radical R'. To seek an explana- 
 tion of this factor was one of the motives that prompted the work which 
 follows. 
 
 Chief interest in the present paper centers upon the reactions of dihy- 
 droxamic acids. They have been shown to rearrange in the following man- 
 
 ner. 
 
 R CO NH OCOR ' 
 
 [R CO N] + R' COOH 
 
 > R N = C :O + R' COOH. 
 
 Either the action of heat, or of warm solutions of alkalies, will produce 
 this effect. In the former case, the isocyanate is formed by dry distillation 
 of the dihydroxamic acid, or, preferably, by heating a salt of the acid. 
 Usually there is a fairly definite temperature at which the decomposition 
 of the dry salt takes place. However, certain cases have proved that it 
 is not a reliable method to use when two similar compounds are to be judged 
 for comparative ease of rearrangement. If solutions rre employed, the 
 isocyanate generally reacts immediately with waiter to form the amine, or 
 the corresponding disubstituted urea. The behsvior of neutral solutions 
 
REARRANGEMENTS OF NEW HYDROXAMIC ACIDS. 2425 
 
 of the sodium or of the potassium salts in water seems now to furnish a 
 more accurate criterion by which to judge the ease of rearrangement of 
 these particular compounds 
 
 A few years ago, phenyl-acethydroxamic acid 9 was studied. The 
 benzoyl ester of this acid was capable of forming salts which possessed 
 unusual instability towards heat. Its solid potassium salt suffered 
 Beckmann rearrangement spontaneously at room temperature. Here, 
 
 C 6 H 5 CH 2 CO NK OCOC 6 H 5 > C 6 H 5 CH2 NCO + C 6 H 5 COOK. 
 
 Because of this rearrangement, it was not found possible to form a clear 
 solution of the salt in water, unless it was prepared immediately after the 
 isolation of the salt. We have recently repeated this experiment, and 
 found, in addition, that the clear solution, when left at room temperature 
 for 2 hours, did not undergo a noticeable rearrangement. A small white 
 precipitate of the urea collected in 10 hours, however. The similar po- 
 tassium salt of the benzoyl ester of acethydroxamic acid, CH 3 CO 
 NK OCOCeHs, did not possess this marked tendency to rearrange; so 
 the replacement of hydrogen by phenyl must have occasioned the decrease 
 in stability. With this in view, diphenyl- and triphenyl-acethydroxamic 
 acid were deemed important compounds to study. The sodium or the 
 potassium salts of their acyl esters should exhibit <a greater capacity for 
 rearrangement in solution than the similar compounds in the monophenyl 
 series. 
 
 The univalent nitrogen compounds, [R CO N], which have been 
 assumed to be the primary products of decomposition, have never been 
 isolated. The isocyanates, resulting from rearrangements which involve 
 readjustments of electrons, are obtained instead. It was not our object 
 in this investigation to try to isolate such derivatives. Indeed, with the 
 groupings [(C 6 H 5 ) 2 CH CO N], and [(C 6 H 5 ) 3 C CO N] the tendency 
 to rearrange to form isocyanates should be greater than in any case pre- 
 viously studied. 
 
 Tripheny] methyl is a group that is known to display a tendency to 
 exist as a free radical. Certainly, then, it would seem highly probable 
 that a derivative such as [(C 6 H 5 ) 3 C CO N], would be far more apt to 
 separate momentarily into [(C 6 H 5 ) 3 C ] and [ CO N] than a group such 
 as [H 3 C CO N ] in which the linking of carbon to carbon is conceived 
 to be stronger. The latter would give rise to [H 3 C] and [ CO N]. 
 By no means must it be considered probable that free radicals dis- 
 playing the great order of reactivity shown by triphenylmethyl could 
 ever be isolated while in the presence of the highly reactive univalent 
 nitrogen. 
 
 As a result of earlier experiments, and of those included in this article, 
 9 Thiele and Pickard, Ann., 309. 189 (1899): Jones, Am. Chem. /., 48, 6 (1912). 
 
2426 LAUDER W JONES AND CHARLES D. KURD. 
 
 we were led to formulate an hypothesis which may be stated as follows: 
 the relative ease of rearrangements of the Beckmann type is dependent 
 upon the tendency -for the radical R, in the univalent nitrogen derivative, 
 e. g., [R CO N], to exist as a free radical. 
 
 It must be noted that this hypothesis does not deal with the ease of 
 formation of the univalent nitrogen compound. It is essential that the 
 group NHX in the original molecules, R CO NHX, be identical 
 when two different compounds are compared. An extreme case of the 
 effect produced when different groups are substituted is displayed in the 
 acyl alkyl halogen amines, e. g., R CO NCI C & Hn. Stieglitz observed 
 that compounds of this type show no tendency to undergo rearrangement. 9 
 
 Triphenyl methyl, and other groups that are known to exist free, have 
 been shown to "exhibit polar characteristics in many of their compounds. 
 For example, triphenylmethyl bromide displays marked conductivity 
 when dissolved in sulfur dioxide; 10 this is one of the arguments which 
 leads to the belief that an ion, or a pseudo-ion, (Cel^sC*, must exist in 
 the solution. Furthermore, it has been shown that triphenylmethyl is 
 formed in the cathode chamber when triphenylmethyl bromide is elec- 
 trolyzed. In fact, appropriate solutions of triphenylmethyl, itself, conduct 
 the electric current. 11 These are important facts in support of our 
 hypothesis. 
 
 From the standpoint of our hypothesis, it would seem that the tendency 
 for the phenylmethyl radical, CeH 5 CH 2 , to exist free is greater than for 
 methyl, CH 3 . The allusion is to the experimental fact that the deriva- 
 tives of phenyl-acethydroxamic acid suffer rearrangement more easily 
 than those of acethydroxamic acid. On the basis of this hypothesis, also, 
 the observation made by McCoy and Stieglitz, 12 and supplemented by 
 van Dam, 13 that dibromo-salicylic-bromo-amide rearranges even at 12, 
 is evidence for the tendency of the radical C 6 H 2 Br 2 .OH to exist free. Other 
 bromo-amides that have been studied exhibit a more stable character. 
 
 From this same standpoint, it may be predicted that any substituted 
 acethydroxamic acid containing a tri-aryl methyl group will undergo 
 rearrangement with great ease. For example, tri-bisphenyl-acethydrox- 
 amic acid, (C 6 H 5 C 6 H 4 ) 3 C CO NHOH, should form derivatives 
 that are even more unstable than triphenyl-acethydroxamic acid. The 
 group tribisphenylmethyl, in contrast to triphenylmethyl, is one that 
 has been shown to exist exclusively in the monomolecular form at room 
 
 9 Stieglitz, Am. Chem. J., 29, 49 (1903). 
 
 10 Gomberg, Ber., 35, 2397 (1902). 
 
 11 Gomberg and Cone, ibid., 37, 2403 (1904); Schlenk, Weichel, and Herzenstein. 
 Ann., 372, 10 (1910). 
 
 18 McCoy and Stieglitz, Am. Chem. J., 21, 116 (1899). 
 
 13 Van Dam. Rec. trav. Mm., 18, 408 (1899); 19, 318 (1900) 
 
REARRANGEMENTS OF NEW HYDROXAMIC ACIDS. 2427 
 
 temperature. 14 Therefore, in the univalent nitrogen derivative, [(C 6 H 5 
 CeH4 )sC CO N], there should be a pronounced tendency for the 
 momentary existence of tribisphenylmethyl, and the ultimate formation 
 of the isocyanate. 
 
 The applications of this hypothesis need not be restricted to one class 
 of radicals that show a tendency to exist free. Since certain radicals 
 which contain divalent nitrogen 18 have been prepared, a hydroxamic acid 
 which could furnish such a group might also rearrange with considerable 
 ease. Illustrations would be, 
 
 (CeHsW CO NHOH > (C 6 H 5 ) 2 N N: C:O+H 2 O. 
 
 (CH 3 O C 6 H 4 )2N CO NHOH > (CH 3 O C 6 H 4 ) 2 N N : C : O +H 2 O. 
 
 Benzophenone when it is treated in ether solution with potassium 
 develops an intense color. 16 Since the boiling point of the ether is un- 
 changed after complete solution of the metal, a trivalent carbon radical, 
 (CeHs^CXOK) , is assumed to exist in solution. In this connection, 
 therefore, the hydroxamic acid of benzilic acid, (C 6 H 5 ) 2 C(OH).CO.NHOH, 
 should present a somewhat complicated, but, nevertheless, a highly inter- 
 esting case to develop. 17 
 
 Our assumptions were corroborated to a large extent by the experi- 
 mental evidence submitted in this paper. Unexpected difficulties presented 
 themselves in the attempts to prepare the sodium and potassium salts in 
 the diphenyl-acethydroxamic acid series. If the customary procedure, 
 viz., the addition of an alcoholic solution of sodium ethylate to an alcohol- 
 ether solution of the benzoyl ester of the acid, e. g., (C 6 H 5 ) 2 CH CO 
 NH OCOC 6 H 5 , was followed, no precipitation of the salt occurred, even 
 when a very large excess of ether was used. The existence of the salt in 
 solution was proved, but the salt could not be obtained 'pure. Evapora- 
 tion of the alcohol and ether in VQCUO always left a mixture of the salt with 
 its products of decomposition and rearrangement; viz., diphenylmethyl 
 isocyanate, diphenylmethyl urethane, sodium benzoate and, also, sym. 
 bi-diphenylmethyl urea, if any water was present. When this residue was 
 extracted with cold water, and the solution filtered and boiled, there was 
 an immediate precipitation of some sym. bi-diphenylmethyl urea, which 
 is the normal reaction for salts of this character. 
 
 Similar difficulties arose in the study of triphenyl-acethydroxamic 
 acid. The sodium or the potassium salts of the benzoyl ester could 
 not be formed pure. The acetyl ester seemed to yield a potassium 
 
 14 Schlenk, Weichel and Herzenstein, Ann., 372, 11 (1910). 
 
 16 Wieland, "Die Hydrazine," F. Enke, Stuttgart, 1913, p. 73. 
 
 16 Schlenk and Thai, Ber., 46, 2843 (1913); Schlenk and Weichel, ibid., 44, 1183 
 (1911). 
 
 17 Jones and Neuffer (J. Am. Chem. Soc., 39, 659 (1917)), have studied the re- 
 arrangements of lact-hydroxamic acid, Ch 3 CHOH CO NHOH, and of mandel-hy- 
 droxamic acid, C 6 H 5 CH(OH).CO.NHOH. 
 
2428 LAUDER W. JONES AND CHARLES D. KURD. 
 
 salt insoluble in ether; but subsequent tests showed that it was mixed 
 to a large extend with triphenylmethyl isocyanate. This isocyanate, 
 not previously described, possesses a remarkable stability, even in the 
 presence of boiling water. Thus, when the precipitate containing the 
 potassium salt was taken up in water, and heated, only the isocyanate 
 separated. The cause of its sluggishness may be attribu, ed to its insolu- 
 bility in water. It reacts normally with aniline in ether to form the urea 
 derivative. 
 
 (CH,),C NCO+C.HJNH2 > CJI 6 NH CO NH C(CH 6 ),. 
 
 The evidence obtained proves that there is an increase in the ease of 
 rearrangement of the molecule as more phenyl groups are added. In 
 solution, the potassium salt of the benzoyl or acetyl ester of monophenyl- 
 acethydroxamic acid was comparatively stable at room temperature; a 
 similar solution of the diphenyl compound became turbid in a short time ; 
 whereas the triphenyl derivative showed rearrangement almost immediately 
 when it was treated with water. In the diphenyl or in the triphenyl series, 
 it was found to be impossible to obtain a sodium or a potassium salt which 
 failed to show the effects of extensive decomposition. For this reason, 
 the temperature of decomposition of the pure dry salts could not be 
 determined accurately. 
 
 The silver salts, made by the action of aqueous silver nitrate upon the 
 cold ether solutions of the sodium or potassium salts were somewhat solu- 
 ble, 19 but precipated for the most part. Chromo-isomerism was dis- 
 played here. The silver salt of the benzoyl ester of diphenyl-acethydrox- 
 amic acid, when first formed, was bright yellow. In a short time, the 
 yellow substance changed to a pure white salt. The similar salt in the 
 triphenyl series precipitated as a white solid, but changed soon to a brilliant 
 yellow salt. 20 
 
 19 The exact cause of the solubility of the sodium and of the potassium salts in 
 alcohol with a large excess of ether is purely a matter of conjecture. There may be 
 tautomeric forms, one soluble and the other insoluble, such as (CeH 6 )jCHC O NK 
 OCOCH, and (CH 6 )tCH C (OK): N OCOCH,. It may be caused by the addi- 
 tion of a molecule of alcohol; thus, (CH 6 )CH C(OK)(OC 2 H 6 ) NH OCOCH 3 . 
 These are suggestions, which may be correct. Salts of the alkali metals which are 
 soluble in ether are very uncommon but not unknown. Sodium iodide behaves in this 
 manner. Loeb, /. Am. Chem. Soc., 27, 1020 (1905). 
 
 ao There are many chromo-isomers on record; such, for example, as silver violurate 
 (Hantzsch, Ber., 42, 969 (1909); Henrich. "Theorien der Org. Chem.," 1918, p. 364) 
 which is colorless when precipitated, and gradually changes through green to a dark 
 brown. Titherlcy (.7 Chem Soc., 71, 468 (1897); 79, 408 (1901)), reported silver 
 benzamide to t-xist in an orange and in a white modification. Jones and Oesper J. Am. 
 Chem. Soc.. 36, 2208(1914)), found chromo-isomeric silver salts among acyl derivatives 
 ofhydroxy-urethanes; c. g. C 6 H 6 CO O NAg CO OR. These salts, yellow when 
 prepared were easily transformed into white modifications, especially if R represents 
 i.w>-l>utvl. /.vo-amvl. or benzyl. 
 
REARRANGEMENTS OF NEW HYDROXAMIC ACIDS. 2429 
 
 Although the salts of the alkali metals could not be isolated pure, solu- 
 tions of them in water could be obtained without extensive decomposition, 
 if the silver salts were suspended in an ice-cold solution of potassium bro- 
 mide. In a few hours the reaction was complete. This constitutes a 
 helpful modification of a reaction never before applied to hydroxamic 
 acids. 
 
 An interesting relationship between chemical constitution and melting 
 points is to be gained from a study >of the following .table. The low melting 
 benzoyl ester of triphenyl-acethydroxamic acid is of particular interest 
 (See experimental part, p. 2439.) 
 
 M. Pt. Benzoyl ester. Difference. 
 
 c. c. c. 
 
 CH 3 CO NHOH 87 69 or 98 18 or +11 
 
 C 6 H 6 CH 2 CO NHOH 145 120 25 
 
 (C 6 H 5 ) 2 CH CO NHOH 173 140 33 
 
 (C 6 H 5 ) 3 C CO NHOH 178 44-47 131-134 
 
 Two new methods of preparation of hydroxamic acids were investigated. 
 
 First, the action of free hydroxylamine upon a ketene. With diphenyl 
 
 ketene, the compound formed was diphenyl-acethydroxamic acid, the 
 
 same in every respect as that prepared by other means. The equation is 
 
 (C 6 H 5 ) 2 C:CO + NH 2 OH + (C 6 H 5 ) 2 C-C :O 
 
 I I 
 H NHOH. 
 
 The reaction -is perfectly analogous to the addition of ammonia, or amines, 
 to ketenes, R 2 C: CO + H NHR' ^R 2 CH CO NHR'. Staudinger 21 
 predicted the formation of hydroxamic acids, but stated "Die Einwirkung 
 von Hydroxylamine, die zur Bildung von Hydroxamsauren fiihren sollte, 
 ist noch nicht untersucht." 
 
 There is another possible direction in which hydroxylamine might add 
 to diphenyl ketene. The amide of benzilic acid would be formed ; (C 6 H 5 ) 2 - 
 C:CO + NH 2 OH ** (C 6 H 5 ) 2 C(OH) CONH 2 . There are no cases 
 recorded in which the hydroxyl group adds to ketenes in such a manner. 
 Invariably, it combines with the carbonyl group. Hydrogen peroxide, 
 similar in many respects to hydroxylamine, would be forced to add hydroxyl 
 and form benzilic acid, if there were addition at all. However, Nicolet and 
 Pelc 22 recently reported that the amount of benzilic acid formed when 
 the two were mixed in anhydrous solvents was no greater "than when the 
 ketene itself, without the addition of peroxide was treated in the same way." 
 In the light of these results, it is not at all surprising that no trace of the 
 amide of benzilic acid was found. 
 
 A profitable field of research is made possible by this reaction. For 
 example, such ketenes as diphenylene ketene, (CiaH 8 ):C:CO; methyl 
 21 Staudinger, "Die Ketene," F. Enke, Stuttgart, 1912, p. 36. 
 
 2 Nirnlpf and Pplr. T A m C.hc.m Snc . 43. 935 C1Q21V 
 
2430 LAUDER W. JONES AND CHARLES D. KURD. 
 
 vinyl ketene, CH 2 : CH C(CH 3 ) : CO; ethyl ketene carboxylic ester, 
 OC : C(C2H 5 ) COOC2H 5 ; carbon suboxide, C 3 O 2 ; and ketene, itself, 
 with hydroxylamine, or substituted hydroxylamines, should lead to inter- 
 esting results. 
 
 The second new method of preparing hydroxamic acids is a modification 
 of the long established method in which acid chlorides are employed. 
 Heretofore, the chloride has been allowed to act upon an aqueous solution 
 of hydroxylamine. This always leads to side reactions, which lower the 
 yield and augment the difficulty of purification of the desired product. 
 When the acid chloride was dissolved in a neutral solvent, such as benzene, 
 and a trifle more than 2 mols of free hydroxylamine was added, it was 
 found that a quantitative yield of monohydroxamic acid resulted. 
 
 (C 6 H 6 ) 8 C COC1+NH,OH > (C,H 6 ) 3 C CO NHOH + NH,(OH)C1. 
 This reaction was modified later so that the preparation of free hydroxyl- 
 amine was avoided. Two equivalents of pyridine or of sodium carbonate 
 crystals was used with one equivalent of hydroxylammonium chloride 
 in a benzene or an ether solution of the acid chloride. A quantitative 
 yield was obtained here also. 
 
 Two heterocyclic hydroxamic acids were studied, one a deriva- 
 tive of furane, and the other of thiophene. The former (I) will be called 
 
 l|^ Jl CO NHOH ^ JJ CO NHOH / 
 
 O (I). S (II). 
 
 pyromucyl-hydroxamic acid ; and the latter (II), a-thenhydroxamic acid. 22 
 Previous work with heterocyclic hydroxamic acids is very slight. Pyro- 
 mucyl-hydroxamic acid was prepared by Pickard and Neville 23 , and by 
 Rimini, 24 but was not extensively studied. A few isolated examples in the 
 pyrone series are known. 25 Awto-ethyl-thenhydroximic acid, C 4 H 3 S 
 C(OC 2 H 5 ) : NOH, has been studied by Douglas. 28 It was not found possi- 
 ble to isolate this in the syn form. One object or our investigation was to 
 
 22 Other names such as a-thienyl formhydroxamic acid, or a-thenoyl-/3-hydroxyl- 
 amine suggest themselves. Both of these names have their shortcomings. Essentially, 
 the compound is a derivative of thiophene, not of formhydroxamic acid; to call it a 
 "hydroxylamine," conceals the acid nature of the substance. The difficulty could be 
 avoided easily, if thiophene-a-carboxylic acid possessed a simple name. It would be 
 entirely in keeping with both its chemical and physical properties to assign it a name 
 similar to benzoic acid. Inasmuch as the grouping "thenoyl," C 4 H 8 S CO is in com- 
 mon usage already, the name thenoic acid is suggested, The prefix "then" corresponds to 
 "benz," andjustasC 6 H 6 CONHOH is benzhydroxamic acid, so C 4 H 3 S CONHOH is 
 a-thenhydroxamic acid. 
 
 Pickard and Neville, /. Chem. Soc., 79, 847 (1901). 
 
 24 Rimini, Gazz. chim. itaL, [2] 31, 90 (1901). 
 
 S5 Oliveri-Mandala, /. Chem. Soc., 100, 916 (1905); 88, 428 (1911); Atti. 
 accad. Lincei, [5] 14, ii, 162 (1905). 
 
 Douglas, Ber., 25, 1312 (1892). 
 
REARRANGEMENTS OF NEW HYDROXAMIC ACIDS. 2431 
 
 study the chemical behavior of pyromucyl-, and of thenhydroxamic acids, 
 since they are typical members of a legion of unstudied heterocylic com- 
 pounds. 
 
 Pickard and Neville 23 stated "an aqueous solution of the sodium salt" 
 of the benzoyl ester of pyromucyl-hydroxamic acid, C 4 H 3 O CO NNa - 
 O COCeHs, "when boiled with water evolves carbon dioxide and an oil 
 (containing nitrogen) is obtained when the solution is evaporated. The 
 oil is presumably difurfuran-carbamide, but decomposes completely when 
 hydrolyzed. No better success was obtained on attempting to prepare the 
 carbamates by boiling the sodium salt with alcohols." 
 
 A year later, Curtius and .Leimbach 27 tried to isolate sym. difuryl urea, 
 CO(NH C 4 H 3 O) 2 , from the azide, C 4 H 3 O CO N 3 , and met with only 
 partial success. Crystals melting at 229, and at 220 were obtained. 
 The former was found to contain 12.01% of nitrogen, the latter 12.13%. 
 The calculated percentage for difuryl urea is 14.56%. 
 
 In the present study, it was found that when the potassium salt,C 4 H 3 O 
 CO NK OCOC 6 H 5 , was warmed gently in water solution, and cooled 
 at the first evidence of precipitation, the precipitate formed was the free 
 benzoyl ester, C 4 H 3 O CO NH OCOC 6 H 5 . This compound is pro- 
 duced by hydrolysis, not by rearrangement. When this same filtrate was 
 heated to boiling, much carbon dioxide was evolved, and a red resinous 
 mass precipitated; after purification, it melted at about 210. This 
 material is similar to the product found by Curtius and Leimbach. 
 
 Thenhydroxamic acid derivatives were observed to undergo a slight 
 hydrolysis also, but rearrangement to form sym. di-thienyl urea was a 
 simple matter. Curtius and Thyssen 28 obtained this same urea from the 
 azide, C 4 H 3 S CO N 3 . The properties of their compound check with 
 those of ours in all respects. The normal reaction, then, is as follows, 
 
 2C 4 H 3 SCONKOCOC 6 H 6 +H 2 O > CO(NH C4H 3 S)2+CO 2 +2C 6 H 5 COOK. 
 
 This behavior is entirely analogous to that of the salts of dibenzhydrox* 
 amic acid. 
 
 It will be noticed that the benzoyl ester of thenhydroxamic acid is 
 isomeric with the thenoyl ester of benzhydroxamic acid. The latter com- 
 pound was prepared., so that comparative properties of the two might be 
 observed. Melting points of the pure substance, and the temperatures at 
 which the potassium and the silver salts decomposed were for the former 
 144, 125 and 168; for the latter, 133, 135 and 165 : respectively. 
 The ease of rearrangement of the potassium or the sodium salts in aqueous 
 solution was nearly identical. Too much stress should not be laid upon the 
 temperature of decomposition of the solid salts. The figures are of im- 
 portance, but the method of applying heat to determine the temperature 
 
 27 Curtius and Leimbach, J. prakt. Chem., [2] 65, 37 (1902). 
 
 28 Curtius and Thyssen, ibid., [2] 65, 17 (1902). 
 
2432 LAUDER W. JONES AND CHARLES D. KURD. 
 
 of decomposition influences them very much. For example, when the 
 potassium salt of the latter compound is heated slowly there is no visible 
 action until about -160. However, if the tube containing the salt is 
 suddenly thrust in a bath at 135, there is violent decomposition. With 
 most of the salts, however, there is a fairly definite temperature at which 
 they explode when heat is applied gradually. 
 
 The physical and some of the chemical properties of thiophene com- 
 pounds are very similar to those of corresponding benzene compounds. 
 Thenhydroxamic acid is no exception. It melts at 124, while benzhy- 
 droxamic acid also melts at 124. It is of interest to note that di-thenhy- 
 droxamic acid melts much lower than di-benzhy droxamic acid . The former 
 was found to exist in two modifications, one melting at 105-107, and the 
 other at 83-86 . Dibenzhydroxamic acid is reported by Lessen 80 to melt at 
 145. Both of these heterocyclic hydroxamic acids resemble benzhydrox- 
 amic acid in that they form an acid ammonium salt, 31 (R CO NH O) 2 - 
 H.NH4, which is difficultly soluble in water. 
 
 Experimental Part. 
 
 1. Preparation of Diphenyl-acethydraxamic acid, 
 (GH 6 )jCH.CO.NHOH. 
 
 First Method. From Ethyl Diphenyl-acetate. Thirty g. of ethyl diphenyl-acetate 
 was dissolved in 180 cc. of methanol which contained a little more than the calculated 
 amount of free hydroxylamine. The hydroxylamine was liberated from 15 g. of its 
 hydrochloride by a solution of sodium methylate, wh ; ch -contained 4.8 g. of sodium. 
 To this mixture a solution of 3.3 g. of sodium in 60 cc. of methanol was added. After 
 10 hours, the mixture was diluted with one liter of water, and the hydroxamic acid 
 was precipitated with dil. sulfuric acid. The filtrate, separated from this precipitate, 
 contained a little diphenyl-acethy droxamic acid. By the addition of a solution of 
 copper acetate to this filtrate. 4 g. of the green copper salt was obtained. 
 
 A solution of sodium carbonate was used to purify the crude diphenyl-acethy drox- 
 amic acid, since it was found to dissolve all of the diphenyl-acetic acid present, but none 
 of the hydroxamic acid. The insoluble part was removed and washed several times with 
 water, and when dry weighed 35 g. The crude material melted between 145 and 168 
 Recrystallization from ethyl acetate formed needle-shaped crystals melting at 172. 
 
 Diphenyl-acethydroxamic acid is soluble in acetone, in ethyl acetate, in ethyl alco- 
 hol, and in a warm solution of sodium hydroxide. It is insoluble in water, in a solution 
 of sodium carbonate (hot), in ligroin, in benzene, in ether, or in chloroform. An alco- 
 holic solution yields the characteristic red color with ferric chloride. 
 
 Analyses. Subs., 0.1470, 0.1589: CO,, 0.3962, 0.4321: H 2 O, 0.0767, 0.0852. 
 Calc. for Ci 4 Hi 8 O 2 N: C, 73.98; H, 5.77. Found: C, 73.53, 7418; H, 5.84, 6.00. 
 
 Subs., 0.3042: N, 16.8 cc. (24 and 743.3 mm. (17.5)) 30% KOH sol. used. Calc. 
 N, 6.17. Found: 6.08. 
 
 Diphenyl-acethydroxamic acid could not be prepared in quantity by the action 
 of ethyl diphenyl-acetate upon free hydroxylamine. Enough was formed to give a 
 purple coloration with ferric chloride, but in order to obtain a satisfactory yield, one 
 mol. of sodium methylate (or its equivalent) seemed essential. 
 
 80 Lossen, Ann., 161, 347 (1872). 
 
 31 Lossen, ibid., 281, 172 (1894). 
 
REARRANGEMENTS OF NEW HYDROXAMIC ACIDS. 2433 
 
 Second Method. From Diphenyl-acetyl Chloride. Diphenyl-acethydroxamic acid 
 was also prepared from diphenyl-acetyl chloride, although the method was less satis- 
 factory because of side reactions. A solution of 1.9 g. of hydroxylammonium. chloride 
 in a small amount of water, was mixed with a solution of 2.8 g. of sodium carbonate. 
 After carbon dioxide had escaped, 6 g. of diphenyl-acetyl chloride crystals was added 
 and the mixture was shaken vigorously. When the reaction had apparently ceased, 
 the product was warmed to 60, then filtered to collect the solid. When the filtrate 
 was acidified, half a gram of diphenyl-acetic acid (m. p. 145) precipitated. 
 
 The residue obtained by filtration (5 g.) gave a double melting point, 155-160, 
 and 215-230. Recrystallization from ethyl acetate led to the separation of diphenyl- 
 acethydroxamic acid (m. p. 172), and sym. bi-diphenylmethyl urea (m. p. 269-270), 
 CO(NH.CH(C 6 H 6 )2) 2 , which will be described later. Yield of the hydroxamic acid, 
 2 to 3 g. 
 
 Third Method. From Diphenyl Ketene. Diphenyl ketene was prepared by Schroe- 
 (C 6 H 5 ) 2 C=C=0+NH 2 OH > (C 6 H 6 ) 2 C C=O 
 
 I I 
 H NHOH 
 
 ter's method. 82 Azibenzil, formed by the oxidation of 29 g. of benzil hydrazone dissolved 
 in 120 cc. of dry benzene, was warmed to 60 in a current of dry carbon dioxide for 
 about 3 hours, until evolution of nitrogen had ceased. 
 
 Two g. of freshly distilled hydroxylamine was suspended in a mixture of 50 cc. of 
 absolute ether and 20 cc. of ethyl acetate, which had been carefully purified to remove 
 traces of alcohol, water, or acetic acid. To this mixture, 100 cc. of the benzene solution 
 of diphenyl ketene was added, while ah* was excluded carefully by maintaining an 
 atmosphere of dry hydrogen gas. 33 Ether was employed to increase the solubility of 
 hydroxylamine, which is insoluble in benzene. A better yield would be obtained, no 
 doubt, if an ether solution of diphenyl l^etene were used, but in this preparation the 
 yield was of secondary interest. 
 
 After the mixture had been shaken thoroughly, the product gradually darkened. 
 From time to time, the stopper of the flask was lifted momentarily to release the pressure, 
 probably caused by the decomposition of some hydroxylamine. After two hours, 
 very little pressure accumulated; so the flask was left overnight. The solvent was 
 then distilled at the temperature of a water-bath and the residue was poured into an 
 Erlenmeyer flask to crystallize. The crystals secured in this way were crystallized 
 from ethyl acetate and petroleum ether Yield, 4 g. After one recrystallization, 
 the melting point was 169-172. 
 
 To establish the identity of this compound and diphenyl-acethydroxamic acid 
 prepared above, the benzoyl esters of both were made and found to possess identical 
 properties. Both preparations melted at 140. (See below.) . ; 
 
 Fourth Method. From the Copper Salt. Apparently, there are two forms of mono- 
 phenyl-acethydroxamic acid. By the decomposition of the copper salt with hydrogen 
 sulfide, Thiele and Pickard 9 obtained a compound melting at 121. Phenyl-acethy- 
 droxamic acid which melted at 145 was prepared by Jones by the interaction of free 
 hydroxylamine and ethyl phenyl -acetate. It was thought probable that a second form of 
 diphenyl-acethydroxamic acid might be obtained, if the copper salt should be suspended 
 in alcohol and decomposed by hydrogen sulfide, but such was not found to be the case. 
 
 Benzoyl Ester: (C 6 H 5 ) 2 CH.CO.NHO CO.C 6 H 6 . Three and five-tenths g. of di- 
 phenyl-acethydroxamic acid was dissolved in a warm solution of potassium hydroxide 
 
 32 Schroeter, Ber., 42, 2345 (1909); Staudinger, Ref 21, p. 144. 
 
 33 C0 2 reacts with NH 2 OH. 
 
2434 LAUDER W. JONES AND CHARLES D. KURD. 
 
 just sufficient to cause the solution of the acid. Benzoyl chloride (1.65 cc.) was added 
 in 4 portions. The reaction mixture was constantly agitated, while a stream of cold 
 water was played over the flask. This gave 6 g. of crude dry product. Crystallization 
 of this material from alcohol yielded plates which melted at 139.5-141. Three re 
 crystallizations caused the compound to have a sharp melting point, 140-140.5. This 
 ester is soluble :n acetone, in ethyl acetate, in chloroform and in hot alcohol; slightly 
 soluble in ether and m benzene, and insoluble in water, in ligroin and in a cold solution 
 of sodium hydroxide. When the solution of sodium hydroxide is warmed, rearrange- 
 ment takes place. When the ester was heated a little above its melting point, the odor 
 of isocyanate became noticeable at once. 
 
 Analysis. Subs.. 0.5417: N, 20.6 cc. (over 40% KOH, at 26 and 740.1 mm.) 
 Calc. for C;iH 17 O,N:N, 4.22. Found: 4.13. 
 
 Potassium Salt. (CeHs^CH.CO.NK.O.CO.CftHs. First Method. Alcoholic po- 
 tassium hydroxide was prepared of such a strength that 1 cc. = 0.113 g. of KOH. 
 One cc. of this reagent was added to a solution of 0.67 g. of the benzoyl 
 ester in a mixture of 15 cc. of absolute alcohol and 10 cc. of dry ether, previously cooled 
 to 15. There was no precipitation when a sample of the solution was diluted largely 
 with ether, or with petroleum ether. Part of the solution was treated with silver 
 nitrate to form the silver salt (see p. 2435). The rest was evaporated in vacua over 
 cone, sulfuric acid. For convenience, the residue left after evaporation of the solution 
 will be called "R." 
 
 This residue (R) was shown to consist of a mixture of the desired potassium salt, 
 together with diphenyl-methyl isocyanate, diphenyl-methyl urethane, and potassium 
 benzoate. The part of (R) which was soluble in ether was extracted and half of this 
 ether solution was evaporated. A residue remained which melted at 90 to 100. Two 
 recrystallizations of this material from benzene and ligroin resulted in the isolation of 
 the urethane, (CH 6 ) 2 CH.NH.CO.OCjH, melting at 124. The properties of this 
 compound were confirmed by its preparation from diphenyl-acetamide, (see below). 
 
 The other half of the portion soluble in ether was shown to contain the isocyanate, 
 (CeH 6 ) 2 CH.NCO, since an ether solution of benzhydryl-amine added to Ft, caused an 
 immediate precipitate of the urea, m. p. 268-270, according to the equation, 
 
 (C 6 H 6 ) 2 CH.N:C:O+(CH6) 2 CH.NH 2 > CO(NH.CH(CH 6 ) 2 ) 2 . 
 
 Benzhydryl-amine is without action upon an ether solution of the urethane. 
 
 The residue left from (R) after ether extraction was also divided into two parts. 
 One part was analyzed and was found to contain a much greater percentage of potas- 
 sium than that calculated for the potassium salt, (C6H 8 ) 2 CH.CO.NK.O.CO.CH|. 
 
 Analysis. Subs., 0.1385: KzSO*, 0.0441. Calc. for C 2 iH,ONK : K, 10.58. 
 Found: 14.30. 
 This would indicate the presence of potassium benzoate as an impurity. 
 
 The other part dissolved in water to give a clear solution. A portion of the solu- 
 tion was boiled; this caused an immediate precipitation of sym. bi-diphenylmethyl urea, 
 which confirms the presence of some potassium salt of the benzoyl ester. The re- 
 mainder of the solution, acidified with dil. hydrochloric acid, gave a precipitate which, 
 by fractional crystallization from alcohol, was resolved into the original benzoyl ester 
 of diphenyl-acethydroxamic acid, m. p. 140, and benzoic acid, m. p. 121. 
 
 Second Method. A solution of potassium ethylate prepared from metallic po- 
 tassium instead of potassium hydroxide gave no different results. 
 
 An interesting observation was made concerning the extreme solubility of the 
 potassium salt. When 0.33 g. of the benzoyl ester was suspended in a cold mixture 
 of 3 cc. of alcohol, and 6 cc. of ether, the undissolved ester went into solution when 
 alcoholic potash was added. 
 
REARRANGEMENTS OF NEW HYDROXAMIC ACIDS. 2435 
 
 Third Method. About 0.36 g. of silver salt, (CeHs^CH.CO.NAg.O.CO.CeHa 
 (see preparation, below) suspended in 4 cc. of ice-water, was treated with 0.12 g. of 
 potassium bromide, dissolved in a little water. The mixture was stirred frequently. 
 After an hour, the precipitate had assumed the yellow color of silver bromide. Ample 
 roof of metathesis was furnished when a little of the solution was filtered, and heated. 
 A heavy crystalline precipitate of sym. bi-diphenylmethyl urea (m. p. 268-270) 
 separated. There appears to be no record of the use of silver salts in the preparation 
 of alkali salts of hydroxamic acid. 
 
 The reaction mixture was kept at overnight. Little, if any, decomposition 
 occurred. One-third of the solution, after filtration, was acidified. The benzoyl ester 
 precipitated in quantity. When a second portion of this solution was allowed to stand 
 at room temperature, a gradual precipitation of the urea took place. 
 
 A portion of the solution containing the potassium salt was treated with silver 
 nitrate, copper acetate, and cobalt nitrate solutions. The colors of the three precipi- 
 tates were white, light green, and light pink respectively. These salts were not studied 
 further. 
 
 Sodium Salt. (CeHs^CH.CO.N.Na.O.CO.CeHs. A convenient solution of sodium 
 ethylate to employ is one in which 1 cc. =c= 0.023 g. of sodium. To a solution of 0.3 
 g. of benzoyl ester in 4 cc. of absolute alcohol and 15 cc. of ether, 0.9 cc. of the sodium 
 ethylate solution was added. (The ester was in slight excess.) Just as with the 
 potassium salt, here, also, no precipitate could be obtained. One portion of this solution 
 was saved for the preparation of the silver salt, (see below) while a second pbrtion 
 of the solution was evaporated rapidly, and the residue extracted with water. Fil- 
 tration from the insoluble matter left a clear solution that soon became turbid. A 
 precipitate of the urea derivative was formed by boiling the solution. 
 
 The remainder of the ether-alcohol solution was evaporated in vacua over cone, 
 sulfuric acid. Ether caused the extraction of a large amount of the urethane, m. p., 
 123-124. The portion insoluble in ether was shown by analysis to be chiefly sodium 
 benzoate. 
 
 Analysis. Subs., 0.0558: Na 2 SO 4 , 0.0251. Calc. for C^HieOsNNa: Na, 6.51. 
 Calc. for C 6 H B .COONa: Na, 15.97. Found: 14.56. 
 
 Silver Salt. (CeH^CH.CO.NAg.O.CO.CeHs. First Method: from the potassium 
 salt. The ether-alcohol solution of the potassium salt (see p. 2434) was treated with 
 an aqueous solution of silver nitrate. The ether layer instantly assumed a deep yellow 
 color, but remained clear. When this solution was shaken, a pure white precipitate 
 of the silver salt formed, and the yellow color of the ether layer disappeared simul- 
 taneously. 
 
 Analysis. Subs., 0.1164: Ag, 0.0285. Calc. for C 2 iHi 6 O 3 NAg: Ag, 24.62. Found: 
 24.49. 
 
 A little of the white salt, suspended in ether, did not cause the ether to assume a 
 yellow color, even when alcohol was added. Here, again, as with the sodium or the 
 potassium salt, there is evidence of the existence of two modifications, one of which 
 is soluble in ether and alcohol, and the other insoluble. Gradual rearrangement occurred 
 when the salt was heated gently; the odor of isocyanate was very marked. In a small 
 tube the salt decomposed with a puff at about 145. 
 
 Second Method: from the sodium salt. A similar procedure was followed with 
 an ether-alcohol solution of the sodium salt (see above). In this case, the ether 
 layer became yellow in color, and the precipitate which formed was yellow as well. 
 The color of the salt gradually changed to pure white, and, in so doing, formed needle- 
 shaped crystals. This change of color was hastened by scratching the precipitate 
 with a glass rod. That the color change commenced at the surface was proved; for 
 
2436 LAUDER W. JONES AND CHARLES D. KURD. 
 
 yellow clumps that had apparently turned white, were found still to be yellow at the 
 center. This material, when heated on a spatula, decomposed with a puff to form the 
 isocyanate and a residue of metallic silver was left after ignition. 
 
 Di-acetyl Ester, (C 8 H 5 ) 1 CH.CO.N(CO.CH,).O.CO.CH,. Apparently, the normal 
 reaction of acetic anhydride and diphenyl-acethydroxamic acid leads to the formation 
 of the di-acetyl ester, a trihydroxamic acid, instead of the mono-acetyl ester, even 
 when half a mol. of acetic anhydride is used. In this case, a mixture of mono- and di- 
 acetyl derivatives is obtained. By recrystallization from alcohol, the mixture was 
 separated only with the greatest difficulty. 
 
 To prepare the di-acetyl ester, diphenyl-acethydroxamic acid (3 g.) was dissolved 
 in a large excess of acetic anhydride (10 cc.). This solution was kept warm for 2 hours, 
 and the excess of acetic anhydride was then evaporated in a vacuum desiccator con- 
 taining powdered alkali. The solid ester, after recrystallization from alcohol, melted 
 at 95.5-97.5. It is soluble in hot alcohol, in ethyl acetate, in chloroform, in acetone 
 and in benzene. It is but slightly soluble in ether, and is insoluble in water, in ligroin, 
 or in a solution of sodium hydroxide. 
 
 Analysis. Subs., 0.5008: N, 19.8 cc. (over 40<? f KOH at 25 and 743.8 mm. 
 (26)) Calc. for Ci S H 17 O 4 N: N, 4.50. Found: 4.34 
 
 Two g. of the di-acetyl ester was distilled in a small flask heated by a metal bath. 
 A liquid that weighed 0.45 g. was collected when the temperature of the bath rose to 
 between 200 and 300. This liquid all redistilled between 120 and 138 ; this fraction 
 consisted chiefly of acetic anhydride. When the residue in the flask was heated with 
 a free flame, a liquid distilled at about 300 and left resinous materials in the flask. 
 When the distillate was treated with an ether solution of benzhydryl-amine to detect 
 any diphenylmethyl isocyanate and then diluted with an equal quantity of petroleum 
 ether, a tardy precipitate resulted; m. p. 136-137. This was identified as benzhydryl- 
 amine acetate; no sym. bi-diphenylmethyl urea could be found. 
 
 Benzhydryl-amine Acetate, (C^Hs^CH.NHaO.COCHs. A sample of benzhydryl- 
 amine acetate was prepared by mixing an ether solution of benzhydryl-amJne with 
 glacial acetic acid. W r hite crystalline plates separated, which melted at 138. The 
 salt dissolved readily in water. 
 
 Mono-acetyl Ester, (CH 6 ) 2 CH.CO.NH O CO.CH 8 . In order to avoid the for- 
 mation of the di-acetyl ester, precautions had to be tak^ri to destroy the acetic anhydride 
 as soon as the reaction mixture showed no mono-hy uroxamic acid. Ferric chloride was 
 used as an indicator. Three g. of diphenyl-acethydroxamic acid was dissolved rapidly 
 in an excess of warm acetic anhydride (10 cc.), and, after about 30 seconds, the re- 
 action was stopped by the addition of 50 cc. of cold water. A white precipitate, chiefly 
 the mono-acetyl ester, resulted. Two recrystallizations gave crystalline plates of 
 pure substance melting at 113-113.5. 
 
 The acetyl ester dissolves in acetone, in chloroform, in ethyl acetate in benzene 
 and in alcohol. It is insoluble in a cold solution of sodium hydroxide and in ligroin, 
 and only slightly soluble in ether. When the solution of sodium hydroxide is warmed, 
 rearrangement takes place. 
 
 Analysis. Subs., 0.3046: N, 14.1 cc. (over 40% KOH at 23.5 and 738.8 mm. 
 (19)). Calc. for CiHuON : N, 5.20. Found: 5.09 
 
 Potassium Salt. (CeH 6 )2CH.CO.NK.O.CO.CH 3 . One-third of a gram of acetyl 
 ester dissolved quite readily in 5 cc. of cold alcohol. The solution was diluted with 10 
 cc. of ether and was then chilled to 10 by means of a freezing mixture of ice and cone, 
 sulfuric acid. To this solution, 0.6 cc. of alcoholic potash solution (see p. 2434) 
 was added. Here again, no precipitate formed, even when an excess of ether was 
 
REARRANGEMENTS OF NEW HYDROXAMIC ACIDS. 2437 
 
 added. Part of the liquid was evaporated; the residue was extracted with ether, dried 
 and analyzed. 
 
 Analysis. Subs., 0.0929 : K 2 SO 4 , 0.0293. Calc. for CieHuOsNK: K, 12.72. Found: 
 14.15. - 
 
 This high result is explained by the presence of potassium acetate, a decom- 
 position product. 
 
 The remainder of the solution, which still contained the potassium salt, was diluted 
 with 10 cc. of petroleum ether. There was no precipitate. If it was shaken with a 
 dil. solution of silver nitrate in water, a white silver salt was obtained (see below). 
 
 Sodium Salt About 0.3 g. of the acetyl ester was dissolved in 2 to 3 cc. of absolute 
 alcohol, and diluted to 15 cc. with ether. This solution, cooled with ice, was treated 
 with Q.8 cc. of sodium ethylate solution (see p. 2435). There was a tardy precipi- 
 tation that gradually became quite heavy. The addition of ether hastened its for- 
 mation This material, collected and washed with ether, was dried and analyzed. 
 Its sodium content was found to be midway between that required for (CeHo^CH.CO.- 
 NNa.O.COCH, and that for sodium acetate. 
 
 Analysis. Subs., 0.0634: Na 2 SO 4 , 0.0346. Calc. for deHuOjNNa: Na, 7.90. 
 Calc. for CH 3 .COONa: Na, 28.05. Found: 17.67. 
 
 Silver Salt. The ether-alcohol solution of the potassium salt, when mixed with 
 aqueous silver nitrate, formed a pure white precipitate. It was found best not to wash 
 the salt with alcohol, since that frequently caused it to turn dark. The white pre- 
 cipitate was washed carefully with water, dried, and analyzed. When it was heated 
 in a small tube, it decomposed between 103 and 110. The isocyanate, recognized 
 by its characteristic odor, distilled. 
 
 Analysis. Subs., 0.0738: Ag, 0.0213. Calc. for CieHuOsNAg: Ag, 28.69. Found: 
 28.86. 
 
 In order to establish the identity of the products of rearrangement isolated in these 
 reactions, they were prepared in a different manner. Some of them represent new com- 
 pounds, not previously described. 
 
 Diphenylmethyl Urethane, (C 6 H 5 ) 2 CH.NH.COOC2H 5 . A method, similar to that 
 developed by Stieglitz and his students, 34 was employed successfully in the preparation 
 of this ure thane. A solution of sodium ethylate was made by the action of 3.42 g. 
 of sodium upon 160 cc. of alcohol. Nine g. of diphenyl-acetamide was suspended in it, 
 and caused to dissolve by the rapid addition of 4.1 cc. of bromine. Without filtration 
 to remove sodium bromide, the neutral mixture was boiled for 10 minutes, after which 
 the alcohol was distilled. The urethane was precipitated by the addition of an excess 
 of water. Yield, 8 g. Nothing except sodium bromide was found in the filtrate. The 
 precipitate was extracted with 50 cc. of hot benzene. One g. of material, later proved 
 to be sym. bi-diphenylmethyl urea, was insoluble. When the benzene became cool, 
 1.8 g. of unchanged amide (m. p. 161) precipitated. The benzene filtrate contained 
 nearly pure diphenylmethyl urethane. After evaporation of the solvent, the residue 
 was pressed on a clay plate. Yield, 4.2 g. It may be recrystallized from a mixture of 
 benzene and ligroin, or of alcohol and water; m. p. 122-123. It is soluble in alcohol, 
 in chloroform, in acetone, in ethyl acetate, and in benzene, but insoluble in water and 
 in ligroin. 
 
 Analysis. Subs., 0.4130: N, 20.69 cc. (over 40% KOH at 20 and 763.9 mm. 
 (16)). Calc. for CieHnOsN: N, 5.49. Found: 5.64. 
 Diphenylmethyl urea chloride, (C 6 H 5 ) 2 CH.NH.CO.C1. Two g. of the urethane 
 
 34 Jeffreys, Ber., 30, 898 (1897); Am. Chem. J., 22, 27 (1899); Folin, Am. Chem. 
 J., 19, 324 (1897). 
 
2438 LAUDER W. JONES AND CHARLES D. KURD. 
 
 and 1.63 g. of phosphorus pentachloride were dissolved in 10 to 15 cc. of dry chloroform. 
 This mixture was warmed, and a current of dry hydrogen chloride was bubbled through 
 the solution. This served to expel phosphorus oxychloride and also to prevent the 
 decomposition of the urea chloride which would yield the more volatile isocyanate. 
 Chloroform was added twice as the solution evaporated. Finally, when the solvent 
 was nearly exhausted, petroleum ether was added, and the entire solution was poured 
 through a filter. A solid mixed with oil resulted when the filtrate was evaporated. 
 
 Diphenylmethyl Isocyanate, (C 6 H 6 )iCH.NCO. The crude urea chloride was 
 extracted with benzene, and the solution filtered. Some calcium oxide was placed in 
 the flask, and the mixture was left in the stoppered flask for 3 hours. This solution 
 contained the isocyanate (see below). After filtration, it was not purified further, but 
 was treated at once with benzhydryl-amine to form the urea 
 
 Benzhydryl-amine, (C*H 5 ) 2 CH.NHj. This compound wasp repared by reducing 
 benzophenone oxime with metallic sodium and alcohol. It has been prepared from 
 benzophenone oxime by Goldschmidt, 34 who employed sodium amalgam, and also 
 by Michaelis 36 from-benzophenone-phenylhydrazone with metallic sodium and alcohol. 
 
 Five g. of dry benzophenone oxime was suspended in 50 cc. of absolute alcohol. The 
 mixture was poured through a reflux condenser upon a piece of sodium which weighed 
 12 g. The oxime dissolved at once, and enough heat was generated to melt the sodium. 
 A little alcohol was introduced from time to time. After 40 minutes, when all the sodium 
 had dissolved, most of the alcohol was recovered by distillation, and water was added 
 to the residue. After partial neutralization of the alkaline solution, benzhydryl-amine 
 was extracted by means of ether, and the ether solution was dried over potassium hy- 
 droxide. To precipitate benzhydryl-amine hydrochloride, it was only necessary to 
 add cone, hydrochloric acid to the ether solution. Between 5.5 and 6 g. of the salt 
 was collected. 
 
 Sym. bi-diphenyhnethyl Urea, CO(NH.CH(CH6)j)i. The benzene solution of the 
 isocyanate (see above) was treated with an ether solution of benzhydryl-amine. 
 There was a tardy precipitation of solid which was collected after a few hours and shown 
 to consist of a mixture of benzhydryl-amine hydrochloride, and the urea. The former, 
 produced because of some unchanged urea chloride, was soluble in very dilute hydro- 
 chloric acid; it melted at 276-280. The portion insoluble in the acid melted at 266. 
 It was rather soluble in hot acetone and in hot ethyl acetate. Recrystallization 
 from either of these solvents produced fine needle-shaped crystals of the urea which 
 melted at 269.5-270. 
 
 Analysis. Subs., 0.3246: N, 19.90 cc. (20 and 763.9 mm. (16)). Calc. for 
 Ci 7 H 24 ONj: N, 7.14. Found: 7.07. 
 
 2. Triphenyl-acethydroxamic Acid. 
 
 (C 6 H 6 ) S C.CO.NHOH. 
 
 Triphenyl-acetyl chloride, (C 6 H 5 )jC.COCl. Schmidlin and Hodgson 36 prepared 
 triphenyl-acetyl chloride from triphenyl-acetic acid, acetyl chloride and phosphorus 
 pentachloride. In the same year Bistrzycki and Land twig 37 described a method of 
 preparation in which phosphorus oxychloride was used as a solvent, instead of acetyl 
 chloride. 
 
 It may be prepared much more simply with thionyl chloride as the chlorinating 
 agent. Three g. of triphenylacetic acid and 10 cc. of thionyl chloride were placed 
 
 34 Goldschmidt, Ber., 19, 3233 (1886). 
 
 35 Michaelis, ibid., 26, 2168 (1893). 
 
 36 Schmidlin and Hodgson, ibid., 41, 438 (1908). 
 
 37 Bistrzycki and Landtwig, ibid., 41, 687 (1908). 
 
RE ARRANGEMENTS OF NEW HYDROXAMIC ACIDS. 2439 
 
 in a small flask equipped with an air condenser, and gently warmed until solution was 
 complete. The liquid was boiled for about 5 minutes after gas evolution (sulfur dioxide 
 and hydrogen chloride) had ceased. Then, it was cooled and poured upon crushed 
 ice to decompose any excess of thionyl chloride. (If larger quantities were employed, 
 it would be better of course, to recover the thionyl chloride by distillation.) The solid 
 triphenylacetyl chloride was filtered, washed with a little cold water, pressed upon a 
 porous plate, and then dried in a vacuum desiccator. Without further treatment, 
 the material obtained was pure enough for most purposes. It melted at 127. The 
 yield was practically quantitative. 
 
 Thionyl chloride readily dissolved both triphenylacetyl chloride, and ethyl 
 triphenylacetate, and it was without action upon either one. 
 
 Triphenylacethydroxamic Acid. Two and five-tenths g. of triphenylacetyl chloride 
 was dissolved in 40 cc. of dry benzene. This was poured into a flask containing 2.5 g. 
 of free hydroxylamine. The mixture was agitated at frequent intervals during an 
 hour, and then was put aside until the next day. Since there was no precipitation of 
 hydroxamic acid, the benzene layer was decanted from the excess of hydroxylamine, 
 placed in a distilling flask, and the solvent was distilled. An oil remained which crystal- 
 lized readily when it became cool. Petroleum ether caused an additional precipitate. 
 More than 2 g. of crystals was obtained. The crude substance melted at 172. 
 
 Triphenylacethydroxamic acid is soluble in a warm solution of sodium hydroxide, 
 in benzene, in acetone, in ethyl acetate and in alcohol. It is insoluble in ligroin, and 
 in water. Recrystallization from ether produced a pure product which melted at 
 175-176. 
 
 An alcoholic solution, treated with ferric chloride, gave the usual red color reaction. 
 Furthermore, copper acetate formed a light green copper salt. 
 
 Analysis. Subs., 0.3858: N, 16.14 cc. (over 40% KOH at 21.8 and 758.9 mm. 
 (18)). Calc. for C 2 oH 17 O 2 N: N, 4.62. Found: 4.75. 
 
 Attempts to prepare triphenylacethydroxamic acid by the action of free hydroxyl- 
 amine upon ethyl triphenylacetate, (CeHs^C.CO.OCaHs, resulted in failure. There 
 was no action, even when an excess of sodium ethylate was present. In all cases the 
 ester was recovered unchanged. 
 
 When triphenylacetyl chloride was employed with a mixture of hydroxylamine 
 and sodium carbonate in water, it was very difficult to purify the triphenylacet- 
 hydroxamic acid. 
 
 Benzoyl Ester, (CeHs^C.CO.NH.O.CO.CeE^. First Method. One-tenth g. of 
 triphenylacethydroxamic acid was fused with a considerable excess of benzoic anhydride 
 until a sample gave no reaction with ferric chloride. The excess of anhydride was then 
 extracted several times with warm ligroin (50). An oil was left which refused to 
 solidify in the course of a week. It was extracted with a very dilute, cold solution of 
 sodium hydroxide. This solution was filtered and acidified immediately. A white 
 solid precipitated, m. p. 40-50. When this process was repeated the benzoyl ester 
 was obtained in a somewhat purer state. It melted between 44 and 47 for the most 
 part, but there was a little that did not melt until the temperature reached 70. It 
 is highly probable that, even at this low temperature, the ester decomposed to a slight 
 extent into benzoic acid and triphenylmethyl isocyanate, which may account for the 
 anomolous melting point (see p. 2429). 
 
 This low-melting ester was found to be extremely soluble in organic solvents. 
 From benzene solution, petroleum ether precipitated it as an oil which could not be 
 made to crystallize, even after the removal of the solvent. 
 
 By careful hydrolysis with alkalies, it was possible to convert part of the ester 
 into triphenylacethydroxamic acid. The greater tendency by far, however, was 
 
2440 LAUDER W. JONES AND CHARLES D. KURD. 
 
 rearrangement of the salt in solution, not to form the urea derivative, as would be antic- 
 ipated, but to give the isocyanate, (CH 6 )jC.NCO. 
 
 Second Method. Because of the instability of the sodium salt of the benzoyl 
 ester of triphenylacethydroxatnic acid, this method (Schotten-Baumann reaction) 
 is less successful than the first method. One and six-tenths g. of triphenylacet- 
 hydroxamic acid was dissolved in 15 cc. of water that contained 0.23 g. of sodium 
 hydroxide. The solution was cooled in a stream of tap water, and 0.6 cc. of benzoyl 
 chloride was added in 3 portions. 
 
 No attempt was made to purify the sticky solid. Instead, it was washed thoroughly 
 with water, then dried, and converted into its salts. 
 
 Potassium Salt. (CeHO.C.CO.NK.O.CO.CeHs. The crude benzoyl ester was 
 dissolved in 1.5 cc. of absolute alcohol. This solution was treated with a solution of 
 potassium hydroxide (0.11 g.) in alcohol. Since 20 cc. of ether caused no precipitation 
 of salt, a solution of silver nitrate in water was added. This gave the silver salt. 
 
 Silver Salt. (CH6)iC.CO.NAg.O.CO.CH 6 . A white precipitate formed, which 
 turned canary-yellow in a short time. It was filtered and washed thoroughly with water, 
 and with ether. In a small tube, this salt became gray in color at 120; it was black 
 at 160; and at 195 it melted with much decomposition. 
 
 Analysis. Subs., 0. 1444 : Ag, 0.0300. Calc. for Ci 7 HjoOjNAg: Ag, 20.98. Found: 
 2085. 
 
 Acetyl Ester, (C 6 Hr) 3 C.CO.NH.O.CO.CH,. A large excess of warm acetic anhy- 
 dride (about 3 cc.) was used to dissolve 0.6 g. of triphenylacethydroxamic acid. In 
 about half a minute a test with ferric chloride showed that the reaction was complete. 
 In order to stop any further action which might lead to the formation of a di-acetyl 
 ester, 10 cc. of cold water was added to decompose any unused anhydride. In a short 
 time, the acetyl ester solidified to form a white solid. It was filtered and dried. Very 
 few impurities were present in this crude material. One recrystallization from a mix- 
 ture of 2 cc. of benzene and 10 cc. of ligroin gave a pure product that melted at 133.5- 
 134. The ester was very soluble in most common organic solvents. The yield was 
 quantitative. 
 
 Analysis. Subs., 0.2156: N, 7.62 cc. (20 and 762.2 mm.). Calc. for C**Hi,O|N: 
 N, 4.06. Found: 4.08. 
 
 Potassium Salt. (CH 6 )iC.CO.NK.O.CO.CH 3 . When 0.38 g. of the acetyl ester 
 was dissolved in a small quantity of absolute alcohol and neutralized with 0.55 cc. of 
 alcoholic caustic potash (0.062 g. of potassium hydroxide), a little ether caused the 
 separation of a gelatinous precipitate, which was collected upon a filter. The material 
 would not dissolve to give a clear solution in cold water. It was evident that the 
 unstable potassium salt suffered rearrangement in solution, because the solution, turbid 
 at first, rapidly became white and opaque. When the solution was boiled, the potassium 
 salt was decomposed completely; triphenylmethyl isocyanate was precipitated. The 
 dry salt decomposed slightly at 112. At 165 there was a vigorous evolution of gas. 
 
 Analyses. Subs., 0.0400, 0.1253: K2SO 4 , 0.0080, O.D252. Calc. for CH 18 O 3 NK: 
 K, 10.19. Found: 8.98, 9.03. 
 
 Sodium Salt. Sodium ethylate and ether did not cause a precipitate when added 
 to a solution of the acetyl ester in alcohol. Accordingly, the solution was treated with 
 an aqueous solution of silver nitrate to form the silver salt. 
 
 Silver Salt. This was thrown down at once as a yellow precipitate, but the quantity 
 was insufficient to warrant an analysis. When heated upon a spatula it decomposed 
 at a fairly high temperature and left a residue of metallic silver. 
 
 Triphenylmethyl Isocyanate, (CeH^sC.NCO. When a solution of the potassium 
 
REARRANGEMENTS OF NEW HYDROXAMIC ACIDS. 2441 
 
 salt of the acetyl ester of triphenylacethydroxamic acid was boiled for a short time 
 with water, an oil was formed. In the course of a few hours this oil solidified. It 
 was collected, and washed with water. The isocyanate was very soluble in organic 
 solvents, even in ligroin. It was dissolved in petroleum ether, filtered into an open 
 dish, and most of the solvent evaporated in a vacuum desiccator. White crystals 
 were formed in this manner, m. p. 85-87. 
 
 Analysis. Subs., 0.1518: N, 6.72 cc. (17 and 757 mm. (18)). Calc. for C 22 - 
 H 16 ON:N, 4.91. Found: 5.12. 
 
 The solubility in petroleum ether, and the low melting point are sufficient evidence 
 that this compound could not be sym. bi-triphenylmethyl urea, CO(NH.C(C6H 6 )i) 2 . 
 Although E. von Meyer and P. Fischer 39 claim to have prepared this urea, none of its 
 properties, not even the melting point, was described. It can be predicted, however, 
 that the melting point is very high. 
 
 In order to prove that the material in hand was the isocyanate, it was dissolved 
 in ether, and treated with the calculated quantity of aniline, also dissolved in ether. 
 Upon evaporation of the ether, a white solid was left which melted sharply at 242. 
 This was anticipated, since phenyl-triphenylmethyl urea, CH 5 .NH CO.NH.C(C6H 6 )i, 
 which has been prepared 39 previously, was known to melt at that temperature. 
 
 3. Pyromucyl-hydroxamic Acid. 
 I CO.NHOH 
 
 O 
 
 The procedure described by Pickard and Neville 24 for the preparation of pyromucyl- 
 hydroxamic acid employed ethyl pyromucate and hydroxylamine in alcoholic solution. 
 It was purified through the medium of the insoluble copper salt. Our modification 
 of the method consists in the addition of a mol of sodium ethylate, and the direct re- 
 covery of the hydroxamic acid without the intermediate formation of the copper salt. 
 
 Thirty g. of methyl pyromucate was poured into 150 cc. of a methanol solution of 
 free hydroxylamine which had been liberated from 19.5 g. of its hydrochloride by 
 a solution of sodium methylate containing 6.4 g. sodium. A solution of 5.4 g. of sodium 
 in 70 cc. of methanol was then added to the mixture and, after 10 hours, as light excess 
 of cone, hydrochloric acid was introduced. Sodium chloride was separated and the 
 methanol removed by distillation in vacuo. An oil was left which solidified after a 
 few hours in a vacuum desiccator. One crystallization from water brought the melting 
 point to 119-122. Twenty-four g. of this material was obtained. Three g. of the 
 copper salt was precipitated from the solvent used in recrystallization. 
 
 The acid is soluble in alcohol, in ethyl acetate, and in water. In benzene, in chloro- 
 form, in carbon disulfide, in ligroin, and in ether it is insoluble. 
 
 Ammonium Salt. (CJlsO.CO.NH.O^H.NHi Ammonium hydroxide caused 
 the precipitation of beautiful crystalline plates, when added to an aqueous solution of 
 pyromucyl-hydroxamic acid. (Cf. p. 2444.) This precipitate was soluble either in 
 an excess of ammonium hydroxide, or in dilute acid, which indicated that the salt 
 is the acid salt and not the normal salt. In the former case, the normal salt is un- 
 doubtedly obtained, whereas with acids, the soluble pyromucyl-hydroxamic acid is 
 restored. The salt melted at 130-131, with considerable gas evolution. The analysis 
 gave the same values whether the salt was prepared in this manner, or in the absence 
 of water, by the action of dry ammonia upon an alcohol-ether solution of pyromucyl- 
 hydroxamic acid. 
 
 39 von Meyer and Fischer, /. prakt. Chem., [2] 82, 521 (1910). 
 
2442 LAUDER W. JONES AND CHARLES D. HURD. 
 
 Analyses. (I) Prepared in the presence of water. Subs., 0.1114: N, 15.9 cc. (over 
 40% KOH at 27 and 737.9 mm.); (II) prepared in the absence of water; subs., 0.2126 : 
 N, 27.58 cc. (17.3, and 772.6 mm'. (15)). Calc. for CioHuOsN,; N, 15.49. Found- 
 (I) 15.41, .(II) 15.30. 
 
 Benzoyl Ester, C 4 H 3 O.CO.NH.OCO.CHi. Pickard and Neville 39 stated that this 
 derivative "is precipitated when an aqueous solution of the hydroxamic acid is shaken 
 with the calculated quantity of benzoyl chloride, and sodium acetate." Potassium 
 hydroxide was found to be far superior to sodium acetate. Three g. of pyromucyl- 
 hydroxamic acid, and 1.3 g. of potassium hydroxide were dissolved in 20 cc. of water, 
 and 2.7 cc. of benzoyl chloride was added in 4 portions. A quantitative yield of the 
 benzoyl ester resulted. After recrystallization from alcohol, the compound melted at 
 140. 
 
 Preparation of Salts. The following general method was employed in the prepara- 
 tion of the potassium and the sodium salts of the esters. Alcoholic solutions 
 of sodium ethylate, or of potassium ethylate were added to concentrated 
 alcoholic solutions of the esters. The salt usually precipitated at once. It was found 
 advantageous to add the alcoholate in a slight excess. Three or four volumes of ether 
 were added to effect a complete separation of the salt. 
 
 To prepare the silver salts, it was necessary only to add a solution of silver nitrate 
 in water, to an aqueous solution of the potassium salt. In most cases this could be 
 done at room temperature, but in a few cases a lower temperature was required to 
 inhibit any hydrolysis. 
 
 Potassium Salt. CH,O.CO.NK.O.CO.CH6. The appearance of the dry salt 
 was unchanged after 8 months. In a small tube, it puffed at 125. 
 
 Analysis. Subs.,0.4337: KjSO 4 , 0.1380. Calc. for C 12 H 8 O 4 NK:K, 14.52. Found: 
 14.28. 
 
 Sodium Salt. 41 This salt was white when first formed, but in 3 months the color 
 was distinctly yellow. 
 
 Silver salt. The silver salt was pure white. 
 
 Analysis. Subs., 0.1289: Ag, 0.0413. Calc. for CnH 8 O 4 NAg: Ag, 31.92. Found: 
 32.04. 
 
 All of these salts showed a decided tendency to undergo hydrolysis. The potassium 
 salt, dissolved in cold water, gave a ferric chloride color test in a very short time. Sim- 
 ilarly, a little of the silver salt suspended in boiling water, caused a separation of black 
 silver, a reaction characteristic of monohydroxamic acids. 
 
 When a 0.5 M solution of the potassium salt was warmed carefully to about 70, 
 crystals were precipitated which were unmistakably identified as the benzoyl ester. 
 By careful application of heat, the filtrate gave a still more abundant yield of the same 
 ester. If the filtrate from this precipitate was heated to boiling, a red resinous mass 
 accompanied by the evolution of much carbon dioxide, was formed. This was dried, 
 and then recrystallized twice from a mixture of alcohol and ether. After this treat- 
 ment, it was still colored. The material darkened at 180, and melted at 210 with 
 decomposition. This is presumably sym. difuryl urea, although enough was not obtained 
 for analysis. 
 
 If the solution of the potassium salt was heated immediately to boiling, the same 
 resinous mass was produced, but the separation of the benzoyl ester did not occur. 
 Acetyl Ester, C 4 H 3 O. CO. NH.O.COCH 3 . Three g. of pyromucyl-hydroxamic acid 
 
 39 Re/. 24, p. 848. 
 
 40 That obtained by Pickard and Neville melted at 134. 
 
 41 This salt was previously prepared by Pickard and Neville. 
 
REARRANGEMENTS OF NEW HYDROXAMIC ACIDS. 2443 
 
 and an excess (3 cc.) of acetic anhydride were warmed until solution occurred. The 
 excess of anhydride was evaporated in a vacuum desiccator. The white crystalline 
 mass was pressed upon a porous plate, and recrystallized from water. It melted at 
 95-96. This ester is soluble in alcohol, in ethyl acetate, in chloroform, in acetone, in 
 hot benzene and in hot water. It is insoluble in ether and in ligroin. 
 
 Analyses. Subs., 0.3526, 0.2920: N, 26.1 cc. (over 40% KOH at 24.5 and 740.3 
 mm. (26.5)), 21.6 cc. (26 and 743.8 mm.). Calc. for C 7 H 7 O4N :N, 8.28. Found: 
 8.09, 8.08. 
 
 Potassium salt. The appearance of this salt was unchanged after standing for 
 8 months. 
 
 Analysis. Subs., 0.3325: K 2 SO 4 , 0.1403. Calc. for C 7 H 6 O 4 NK : K, 18.87. Found: 
 18.98. 
 
 Sodium salt. Alcoholysis of the salt could not be prevented, even when it was 
 prepared at 0. The salt of pyromucyl-hydroxamic acid was shown to be present 
 when it was treated with a solution of ferric chloride, by the appearance of a deep 
 purple color. It was expected, therefore, that the percentage of sodium in the sample 
 would be higher than that required by the sodium salt of the acetyl ester of pyromucyl- 
 hydroxamic acid. 
 
 Analyses. Subs., 0.2667, 0.1064: Na 2 SO 4 , 0.1133, 0.0436. Calc. for C 7 H 6 O4NNa: 
 Na, 12.14. Found: 13.75, 13.27. 
 
 A still higher sodium content was found when methanol, at 0, replaced ethyl 
 alcohol in its preparation. After 3 months the salt was bright yellow in color. 
 
 Analysis. Subs., 0.1032: Na 2 SO 4 , 0.0470. Found: Na, 14.75. 
 
 Silver salt. It was necessary to prepare this derivative at When the salt was 
 prepared at room temperature, hydrolysis to give the monohydroxamic acid was in- 
 dicated by the formation of a black precipitate of metallic silver. 
 
 Analysis. Subs., 0.0982: Ag, 0.0382. Calc. for C^eO^Ag: Ag, 39.09. Found: 
 38.90. 
 
 Both the rearrangement and the hydrolysis of these salts were perfectly analogous 
 to the behavior shown by the salts of the benzoyl ester; except that in this case the 
 tendency to hydrolyze appeared greater. 
 
 4. Thenhydroxamic Acid. 
 
 jj CO . NHOH. 
 
 \ 
 
 S 
 
 First Method. A solution of free hydroxylamine in methanol was prepared by 
 mixing Solutions A, and B, composed, respectively of, 4.90 g. of hydroxylammonium 
 chloride in 50 cc. of methanol, and 35 cc. of methanol to which 1.67 g. of sodium had 
 been added. Sodium chloride was removed by filtration. 
 
 Nine and six-tenths g. of thenoic ethyl ester, C 4 H 3 S CO.OC 2 H 5 , previously purified 
 by distillation in vacua, was added to the alcoholic solution of hydroxylamine, followed 
 by one mol of sodium methylate (1.42 g. of sodium in 30 cc. of methanol), and the mix- 
 ture was allowed to stand overnight. Then 5 cc. of 'cone, hydrochloric acid was used 
 to precipitate the sodium as sodium chloride. The solvent distilled in vacua, after 
 filtration to remove the salt, left a thick liquid which was placed in a vacuum des- 
 iccator to crystallize. The solid obtained in this way, crystallized from water, melted 
 at 105-113. Recrystallization from toluene resulted in pure material, m. p. 123-124.5. 
 
 A water solution of tVienhvoVoyamio arin" is arid to litmus. With ferric chloride 
 
2444 LAUDER W. JONES AND CHARLES D. HURD. 
 
 and copper acetate, it reacts in the normal manner, to give respectively, a purple colored 
 liquid, and a grass-green copper salt. The acid is soluble in alcohol, in hot toluene, 
 in hot benzene, in chloroform, and in water. It is insoluble in ether and in ligroin. 
 With isatin, and warm sulfuric acid the red color characteristic of thiophene derivatives 
 is produced. 
 
 Analysis. Subs., 0.2259: N, 19.12 cc. (over 40% KOH at 18.5 and 757.8 mm.). 
 Calc. for CiH,OiNS: N, 9.78. Found: 9.74. 
 
 Isolation of thenhydroxamic acid through the medium of the copper salt and 
 hydrogen sulfide was found to be successful, but the procedure is not to be recommended. 
 
 The Preparation of a-Th noyl Chloride, C^aS.COCl. Thenoyl chloride has 
 always been prepared from thenoic acid by the action of phosphorus pentachloride. 48 
 A much simpler, and smoother preparation follows if thionyl chloride, SOClj, is employed 
 instead of the phosphorus compound. 
 
 Six g. of a-thenoic acid was warmed with 16 cc. of thionyl chloride until the gas 
 evolution showed that the reaction was complete, when the two chlorides were separated 
 by fractional distillation. Thenoyl chloride all distilled between 206 and 210; yield, 
 6g. 
 
 a-Thenhydroxamic Acid. Second method. A solution of 0.3 g. of sodium carbonate, 
 and 0.2 g. of hydroxylammonium chloride in 2.5 cc. of water was shaken with 0.3 g. 
 of thenoyl chloride. Di-thenhydroxamic acid CiHtS.CO.NH.O.CO.CiHiS (see p. 
 2445) precipitated from the solution, whereas thenhydroxamic acid did not. To recover 
 the latter, the filtrate was made slightly acid with acetic acid, and treated with copper 
 acetate solution. The light green copper salt precipitated, although the green color 
 of the solution showed that it possessed a rather high degree of solubility. The filtrate 
 still gave an intense ferric chloride reaction. 
 
 The precipitate of di-thenhydroxamic acid was recrystallized from alcohol and 
 water. It melted at 105-107. We did not try to prepare thenhydroxamic acid by 
 our new method, viz., by the action of free hydroxylamine and a benzene solution of 
 thenoyl chloride. No side reactions would be anticipated in this case. 
 
 Ammonium Salt. (C 4 H,S.CO.NH.O ) 2 H.NH 4 . This difficultly soluble acid salt 
 was prepared in exactly the same way as the similar furan salt, (p. 2441) and showed 
 the similar property of dissolving either in ammonia, or in hydrochloric acid. For analy- 
 sis, the salt was precipitated from an alcohol-ether solution bf thenhydroxamic acid by 
 dry ammonia. It melted, with decomposition, at 142-143. 
 
 Analysts. Subs., 0.1172: N, 13.95 cc. (21 and 769 mm. (20)). Calc. for do- 
 HtfQiNaSi: N, 13.85. Found: 13.75. 
 
 Benzoyl Ester, C 4 H S S.CO.NH.O.CO.CH5. Two g. of thenhydroxamic acid, and 
 0.8 g. of potassium hydroxide were dissolved in 10 cc. of water. This solution, agitated 
 with 1.7 cc. of benzoyl chloride, gave a precipitate which was recrystallized from alcohol. 
 The needle shaped crystals so obtained melted sharply at 143-144, with decomposi- 
 tion. It is soluble in alcohol, in acetone and in ethyl acetate, and is insoluble in ligroin, 
 or in cold water. 
 
 Analysis. Subs., 0.4336 g.: N, 21.97 cc. (at 21 and 758.6 mm. (at 19)). Calc. 
 for C, 2 H 9 O3NS: N, 5.67. Found: 5.77. 
 
 Potassium Salt. The general method described upon p. 2422, was used also to 
 prepare the salts in this series." The potassium salt was pure white. When it was 
 heated in a small dry tube, it became yellow at 121 and decomposed with a faint 
 puff at 125-127. 
 
 Analysis. Subs., 0.5313: K 2 SO, 0.1626. Calc. for Ci,H s ONSK: K, 13.71. 
 Found: 13.73. 
 
 " V. Mever and Kreis, Ber., 16, 2174 (1883); Peter. B.-r . 18, 543 (1- 
 
REARRANGEMENTS OP NEW HYDROXAMIC ACIDS. 2445 
 
 Sodium Salt. When slowly heated, the sodium salt decomposed mildly between 
 155 and 160. 
 
 Analysis. Subs., 0.2388: Na 2 SO 4 , 0.0631. Calc. for Ci 2 H 8 O 3 NSNa: Na, 8.54. 
 Found: 8.55. 
 
 Silver Salt. The silver salt decomposed between 163 and 168. Above this 
 temperature thienyl isocyanate distilled from the tube. 
 
 Analysis. Subs., 0.2364: Ag, 0.0723. Calc. for C^HgOsNSAg : Ag, 30.47. Found 
 30.58. 
 
 Rearrangement of the Salts. A 0.5M solution of the potassium salt in water was 
 boiled. The solution darkened, and di-thienyl urea precipitated as a gray- violet crystal- 
 line mass. 44 It was collected and recrystallized from acetic acid. The melting point 
 was 224-225. 
 
 The nitrate was acidified and the precipitate obtained was crystallized fractionally 
 from dilute alcohol. Crystals melting at 140-144 separated at first. This material 
 was the benzoyl ester, which must have been formed by hydrolysis of the salt. Benzoic 
 acid was left in the filtrate. The ratio of the urea to the benzoyl ester was about 4:1. 
 
 Acetyl Ester, C 4 H 3 S.CO.NH.O.CO.CH 3 . Three-tenths g. of pure thenhydroxamic 
 acid was warmed in an open glass dish with 0.2 g. of acetic anhydride. As soon as 
 solution had taken place, the dish was placed in a vacuum desiccator to remove the 
 excess of anhydride. The white solid obtained was recrystallized from benzene. The 
 melting point was sharp at 96.5-97 ; yield, quantitative. The ester is soluble in alcohol, 
 in ethyl acetate, and in hot benzene, but is insoluble in cold benzene and in ligroin. 
 
 Analysis. Subs., 0.3242: N, 21.47 cc. (over 40% KOH at 17.8 and 760.3 mm. 
 (15)). Calc. for C 7 H 7 O 3 NS: N, 7.57. Found: 7.66. 
 
 Potassium Salt. No precautions were necessary to prepare the potassium salt 
 
 Analysis. Subs., 0.3099: K,SO 4 , 0.1215. Calc. for C 6 H 6 O 3 NSK: K, 17.51. Found 
 17.49. 
 
 Sodium Salt. Ice-cold solutions were essential in the preparation of the sodium 
 salt. Otherwise, alcoholysis to yield the monohydroxamic acid occurred. 
 
 Analysis. Subs., 0.0609: Na 2 SO 4 , 0.0209. Calc. for C 7 H 6 O 3 NSNa: Na, 11.10. 
 Found: 11.11. 
 
 Silver Salt. Because of hydrolysis, a water solution of the sodium salt could not 
 be used to prepare the silver salt. A black precipitate of silver always resulted. With 
 the potassium salt, however, the preparation was entirely successful. 
 
 Analysis. Subs., 0.1481: Ag, 0.0548. Calc. for C 7 H 6 O 3 NSAg: Ag, 36.94. Found: 
 37.00. 
 
 These salts all puffed at a comparatively high temperature. A pronounced odor 
 of thienyl isocyanate was always noticed. When water solutions of the salts were 
 boiled, di-thienyl urea separated instantly. However, this was always accompanied 
 by an unavoidable hydrolysis to form the acetyl ester and a second hydrolysis to give 
 thenhydroxamic acid. 
 
 Thenoyl Ester, (Di-thenhydroxamic Acid), C 4 H 3 S.CO.NH.O.CO.C 4 H 3 S. One prep- 
 aration of this ester (Cf. p. 2444) which resulted in the high-melting modification (see 
 below) has already been described. 
 
 Three g. of thenhydroxamic acid, and -1.18 g. of potassium hydroxide were dis- 
 solved in 15 cc. of water. The ice-cold mixture was treated with 3 portions (1 g. each) 
 
 44 Sym. di-thienyl urea was found by Curtius and Thyssen 2a to be gray-violet in 
 color, even after 7 recrystallizations from solvents with animal charcoal; m. p. 224. 
 
2446 LAUDER W. JONES AND CHARLES D. KURD. 
 
 of thenoyl chloride. A precipitate which weighed 2.7 g. was collected ; m. p. between 
 80 and 88. 
 
 The filtrate was found to be acid in reaction, and to contain some unchanged 
 thenhydroxamic acid. It was therefore made slightly alkaline with sodium hydroxide 
 and treated with 2 cc. of thenoyl chloride. The resulting precipitate melted at 90- 
 100 and weighed 1.4 g. When the filtrate was acidified, an appreciable quantity of 
 thenoic acid separated. 
 
 The crude precipitates, (m. p. 80-100) were combined and recrystallized from 
 benzene. Crystals which melted between 83 and 86 were obtained. Later a re- 
 determination of the melting point of this same sample indicated that a transformation 
 to a higher-melting modification had taken place. The melting point ranged from 100 
 to 104. 
 
 When alcohol was used as the solvent, the high melting modification separated first. 
 There was an indication of softening at 95, but the melting point at 102-104 was 
 quite distinct. 
 
 Both of these compounds responded to all the reactions expected of di-thenhydrox- 
 amic acid. 44 They both formed salts; neither one gave a color test with ferric chloride, 
 but after hydrolysis both showed an intense purple color with this reagent. The purest 
 sample of the high-melting form melted at 105-107. Its isomer melted at 83-86. 
 Both were soluble in alcohol, in ethyl acetate, and in hot benzene. In water and in 
 ligroin they were insoluble. 
 
 Analyses. (I) Low-melting modification; subs., 0.3339: N, 16.90 cc. (16.5 and 
 762.0 mm.); (II) high-melting modification; subs., 0.4288: N, 19.60 cc. (18.5 and 769.1 
 mm. (16)). Calc. for CioH 7 O 8 NSj: N, 5.53. Found: (I) 5.90, (II) 5.35. 
 
 Potassium Salt. C^S.CO.NK O.CO.C 4 HsS. The salts were all prepared 
 from di-thenhydroxamic acid melting at 102-104 In a small tube, the temperature 
 of decomposition of the potassium salt was 122. 
 
 Analysis. Subs., 0.1988: K a SO 4 , 0.0584. Calc. for C.oHeO^NSjK: K, 1342. 
 Found: 13.18. 
 
 Sodium Salt. There was no evident alcoholysis, when this last was prepared at 
 room temperature. At 142 the dry salt decomposed. 
 
 Analysis. Subs., 0.0797: NajSO 4 , 0.0206. .Calc. for C 10 H 6 OjNS,Na: Na, 8.36. 
 Found: 8.37. 
 
 Silver Salt. An abundant white precipitate of the silver salt of di-thenhydroxamic 
 acid resulted when aqueous solutions of the potassium salt and silver nitrate were 
 mixed. It decomposed at 155. 
 
 Analysis. Subs., 0.1253 : Ag, 0.0376. Calc. for C, H e O,NSjAg: Ag, 29.96. Found : 
 30.01. 
 
 When boiled with water, both the potassium and the sodium salts gave a dense 
 precipitation of sym. di-thienyl urea. Water solutions of the pure salts did not readily 
 undergo hydrolysis or rearrangement. 
 
 5. Benzhydroxamic Acid. 
 
 The new method of preparation of hydroxamic acids, (see p. 2430) namely, 
 by the use of the acid chloride in benzene, was successfully employed in the preparation 
 of -benzhydroxamic acid from benzoyl chloride. 
 
 Hirst Method. Six and five-tenths g. of benzoyl chloride was dissolved in 70 cc. 
 of dry benzene. Upon the addition of 3.3 g. of free hydroxylamine a dense precipitate 
 of benzhydroxamic acid was formed. The reaction mixture was shaken vigorously, 
 
 44 Lessen. Ann., 281, 289 (1894). 
 
REARRANGEMENTS OF NEW HYDROXAMIC ACIDS. 2447 
 
 and then set aside for a few hours until the odor of benzoyl chloride had disappeared. 
 Less than half a gram of benzhydroxamic acid was recovered from the benzene filtrate. 
 One recrystallization of the solid from hot toluene removed hydroxylammonium chloride 
 and yielded a pure product, m. p. 124. This is identical with the material described by 
 Lessen. 29 No evidence of the higher melting di-benzhydroxamic acid was observed. 
 
 Second Method. Benzhydroxamic acid was also prepared in excellent yields, from 
 benzoyl chloride, dissolved in organic solvents, with hydroxylammonium chloride 
 instead of free hydroxylamine. An intimate mixture of anhydrous sodium carbonate 
 (2.1 g.) and hydroxylammonium chloride (1.4 g.) was suspended in 50 cc. of ether. 
 When 2.8 g. of benzoyl chloride was added there was little action, but it became more 
 vigorous when 3.5 cc. of water was added. In about half an hour the reaction was 
 complete. The ether solution was filtered into a distilling flask, and most of the solvent 
 was removed. When this residue was cooled, an abundant crystalline mass precipitated. 
 This was filtered, washed with a little ether, and pressed upon a porous plate. The 
 benzhydroxamic acid melted at 124-125 without further purification. Yield, nearly 
 quantitative. 
 
 Similarly, benzhydroxamic acid was obtained in quantitative yields when pyridine 
 was substituted for sodium carbonate. Benzene was used instead of ether in this case. 
 
 Thenoyl Ester of Benzhydroxamic Acid, C 6 H 5 .CO.NH.O.CO.C4H 3 S. Two and a 
 half g. of benzhydroxamic acid dissolved in 25 cc. of water was neutralized with 0.7 g. 
 of potassium hydroxide. Three portions of 1 g. each of thenoyl chloride were added 
 to this solution. The precipitate was filtered, and washed with water; yield, 2 g. 
 When it was recrystallized from ethyl acetate, it gave white needles which melted at 
 133.0-133.5. The ester is soluble in alcohol, in acetone, in ethyl acetate and in hot 
 benzene. It is insoluble in cold benzene, in ligroin and in water. 
 
 Analysis. Subs., 0.1547: N, 7.70 cc. (over 40% KOH at 15 and 763 mm.). Calc. 
 for Ci 2 H 9 O 3 NS: N, 5.67. Found: 5.85. 
 
 Potassium Salt. C 6 H 5 .CO.NK.O.CO.C 4 H 3 S. When 0.32 g. of the thenoyl ester 
 was dissolved in a mixture of 2 cc. of alcohol and 1 cc. of ether, the potassium salt was 
 precipitated by the addition of 0.7 cc. of an alcoholic solution of potassium hydroxide 
 (0.08 g. KOH). Five cc. of ether caused further precipitation. 
 
 A water solution of this salt became turbid in a short time, when left at room 
 temperature. When the solution was boiled, a crystalline precipitate of sym. diphenyl 
 urea, m. p. 236-238, was formed. 
 
 The dry potassium salt puffed vigorously at 135-140, providing the tube contain- 
 ing it was thrust into a bath previously heated to that temperature. However, when 
 the tube was gradually warmed, a mild decomposition took place. There was no 
 noticeable action until 160, although phenyl isocyanate was undoubtedly being evolved 
 at a temperature somewhat below that. 
 
 Analysis. Subs., 0.1788: K 2 SO 4 , 0.0538. Calc. for Ci 2 H 8 O 3 NSK: K, 13.70. 
 Found: 13.50. 
 
 Silver Salt. When silver nitrate, in water solution, was added to a solution of 
 the potassium salt in water, a voluminous white precipitate of the silver salt resulted. 
 It blackened near 165. 
 
 Analysis. Subs., 0.0897; Ag, 0.0274. Calc. for Ci 2 H 8 O 3 NSAg : Ag, 30.47. Found: 
 30.54. 
 
 Summary. 
 
 An interpretation of the mechanism of the Beckmann rearrangement 
 has been proposed, based upon the modern conception of chemical bonds 
 
2448 LAUDER W. JONES AND CHARLES D. KURD. 
 
 and electrons. The transfer of electrons for all stages in the process has 
 been pictured in a detailed manner. 
 
 The following hypothesis has been advanced: the relative ease of re- 
 arrangements of the Beckmann type is dependent upon the tendency for 
 the radical R in the univalent nitrogen derivative, e. g., (R.CO.N), to 
 exist as a free radical. 
 
 The hypothesis was tested successfully with the sodium and the potas- 
 sium salts of the acyl esters both of diphenylacethydroxamic acid and of 
 triphenylacethydroxamic acid. Rearrangement of the salts in solution 
 was found to be more reliable for purposes of comparison than rearrange- 
 ment of the solid salts. In agreement with the hypothesis, the relative 
 ease of rearrangement was greater with triphenylacethydroxamic acid 
 derivatives than with the similar compounds in the diphenyl series. When 
 water solutions of the salts in the triphenyl series were heated, the pro- 
 duct of rearrangement was triphenylmethyl isocyanate, and not sym. 
 bi-triphenylmethyl urea. This is a singular fact. The phenomenon of 
 chromo-isomerism was displayed by the silver salts of the diphenyl series, 
 as well as those of the triphenyl series. 
 
 Diphenyl ketene and hydroxylamine were found to react readily in a 
 neutral solvent to form diphenylacethydroxamic acid. Modifications 
 of this new reaction between ketenes and hydroxylamine, or substituted 
 hydroxylamines, should be valuable in future synthetic work. 
 
 The properties of pyromucyl-hydroxamic acid, and of a-thenhydrox- 
 amic acid have been determined. The latter resembles benzhydroxamic 
 acid very closely. The former is similar to benzhydroxamic acid in many 
 respects, but its derivatives differ materially in rearrangement. These 
 two compounds, which are representatives of the furane and of the thio- 
 phene types, prepare the way for further investigations of heterocyclic 
 hydroxamic acids. 
 
Accepted by the Department < f Chemistry, 
 Jum HUM 
 
Rearrange 
 
 new hydroxa 
 related to 
 
 iients of some 
 flic acids 
 beterocyclic 
 
 A7E8 
 
 UNIVERSITY QF CAUFORN1A LIBRARY