Feb., 1921
T E E J O U R N r l L O F I N D U S T R I A L Anl-D E N G I L V E E R I N G C H E M I S T R Y
with phenols.’ By interaction of furfural with aniline alone or with acetone in t h e presence of alkalies, soluble resins are obtained which may prove useful in t h e varnish industry.2 Furfural has also found use as a solvent and insecticide.
135
have t o be obtained, ammonia is preferable. If ammonia is added t o formaldehyde or t o mixtures of phenol and formaldehyde, t h e ammonia disappears immediately and becomes hexamethylenetetramine:
FURTHER STUDIES ON PHENOLIC HEXAMETHYLENETETRAMINE COMPOUNDS3 By Mortimer Harvey and L. H. Baekeland LABORATORY OF T H S
DEPARTMENT OF CHEMICAL ENGINEERING,
COLUMBIA UNIVERSITY, NEWYORK,N. Y. Received May 12, 1920
The production of resins or resinoid substances of t h e Bakelite type4 by t h e interaction of phenols with compounds containing a n active methylene group has, of late, acquired considerable importance in t h e industry of coal-tar derivatives. The increasing number of applications of these products in t h e most diversified fields is stimulating research in many directions. T h a t this industry was born and developed in t h e United States, which to-day is still t h e leader in this branch of chemical industry, adds interest t o any subject of research which directly or indirectly may throw light on t h e unusually complicated chemistry of this subject. The theoretical interpretation of t h e different phases of t h e Bakelite reaction is not by any means an easy one, and considerable additional research work will be required before permitting ourselves t o do much beyond guessing at what really happens. I n t h e meantime, t h e careful study of t h e formation of intermediate products can render us considerable help in this subject. Among these intermediate products, t h e further advances are amorphous mixtures which are not amenable t o t h e usual methods of chemical purification or isolation. Therefore, it is more natural t o start first with t h e intermediates which are well-defined crystalline bodies of which t h e chemical composition can be determined b y wellestablished methods. The present research work was, therefore, confined t o some of t h e first phases of t h e reaction, and more particularly t o such bodies as are liable t o form when ammonia is used in t h e process, either as such or in t h e shape of hexamethylenet et r ami ne. I n the formation of these products of t h e Bakelite type t h e methylene-containing body may be commercial formaldehyde solution-known as formalin, formol, etc. This commercial product is practically a mixture of several bodies containing active methylene groups, as, for instance, methylal, formaldehyde, t h e polymers of formaldehyde, their hydrates, etc. The reaction is favored b y t h e addition of so-called condensing agents, or catalysts-whatever t h a t may mean. Acids, salts, and alkalies have been used for this purpose. I n some cases where particular effects 1 Beckmann and Dehn, Sitzb. A k a d . Wiss., Berlin, 1918, 1201; Chem. Abs.. 14 (1920). 642. 2 Meunier, “Application du Furfurol B Ia fabrication de resines B vernis,” Mat. grasses, 9 (1916), 4516. a Submitted by one of authors in partial fulfilment of the requirement for the degree of Doctor of Philosophy in the Faculty of Pure Science, Columbia University, New York, N. Y. 4 These substances are also known under other trade names, as for nstanse, Condens ite, Resinit, Sipilite, Redmanol, etc.
so t h a t all these reactions wherein formaldehyde and ammonia are used conjointly can be repeated by t h e direct use of hexamethylenetetramine. But in presence of phenol, t h e hexamethylenetetramine does n o t remain as such. I t combines with t h e phenol in t h e proportion of three molecules of phenol t o one molecule of hexamethylenetetramine and produces a welldefined crystalline product, hexamethylenetetramine triphenol, which has been described b y Moschatos and Tollens. I n 1909, Lebach2 pointed out t h a t whenever ammonia is used in t h e Bakelite reaction, hexamethylenetetramine triphenol is formed in t h e first stages of t h e process. Under t h e action of heat, this product undergoes a further decomposition and resinifies, emitting ammonia.3 Contrary t o t h e results of Moschatos and Tollens, who were unable t o prepare addition products of hexamethylenetetramine with any of t h e three cresals or with carvacrol or thymol, Baekeland had succeeded in his laboratory in preparing a corresponding crystalline cresol derivative, but inasmuch as this work had not been carried out with each one of t h e completely purified cresols and studied by itself, i t seemed desirable t h a t each one of t h e three homologs should be studied separately as t o its individual behavior. This research was also extended t o carvacrol and t h e results obtained thus far are set forth. Similar compounds obtained from other phenolic bodies are now under study. I n t h e meantime, the observations concerning the new cresol derivatives are submitted in t h e present paper. The reason of the non-success of Moschatos and Tollens in making t h e cresol derivatives of hexamethylenetetramine is, mainly, t h a t the isolation of these substances is incomparably more difficult t h a n in the case of phenol. The hexamethylenetetramine triphenol forms rapidly and visibly under almost all circumstances, and crystallizes very well from aqueous solutions or even from solutions when a considerable excess of one of t h e constituents is used. This is not t h e case with some of t h e cresol derivatives. Ann., 272 (1892), 271. Z. angew. Chem., 22 (1909), 1600; J . SOC.Chem. I n d . , 32 (1913), 559. 3 A resume of the literature on this subject is given by L. H. Baekeland, in “The Chemical Constitution of Resinous Phenolic Condensation Products,” THISJOURNAL, 6 (1913). 506. 1 2
T H E JOURNAL OF I N D U S T R I A L A N D ENGINEERING C H E M I S T R Y
136
The temperature a t which they form lies in some cases so close t o t h e temperature a t which they decompose t h a t their formation is almost sure t o be overlooked if proper precautions are not taken. Furthermore, some of those products have a tendency t o remain liquid in t h e presence of an excess of some of the reacting products or impurities. T h a t such products exist has been established beyond doubt by t h e present investigation. I n this work data were determined for the relationship of certain of t h e phenolic condensation products. The results are appended. HEXAMETHYLENETETRAMINE TRIPHEXOL
Xloschatos and Tollens made the easily prepared hexamethylenetetramine triphenol by mixing 6 g. of a concentrated water solution of hexamethylenetetramine with a concentrated solution containing 6 g. of phenol. The product isolated had the following composition:
Cza..
Calculated for CeHirN4.3CaHaOH Per cent 68.25 7.11 13.27
...........
........... ..............
€130.. N4
-Found 1
68.49 7.34
...
2
by M. and 3
Per cent 68.09 7.44
...
T.-
4
...
...
ij:&
ii:j7
All the phenols do not react with hexamethylenetetramine t o form a n addition product in which there are one mole of hexamethylenetetramine and three moles of the phenol. The various groupings about the benzene ring seem t o determine t h e extent t o which the addition takes place. The three cresols whose structural formulas are nearly identical with t h a t of ordinary phenol and whose properties are somewhat similar t o t h e latter should form addition compounds the same as does phenol. H E X A M E T H Y L E N E T E T R A M I N E DI-??%-CRESOL
The m-cresol addition product is the most easily obtainable. At first ordinary m-cresol was used in both dilute and concentrated alcoholic solutions; b u t t h e expected crystalline intermediate addition products did not appear. The alcoholic solutions were refluxed several hours and the concentrated solutions allowed t o stand several weeks t o see if t h e compound would crystallize out. N o crystalline product was obtained in this case. There must have been some impurity in the cresol t h a t hindered t h e formation, for with cresol purified according t o Fox and Barker1 t h e product crystallized out in 40 min. A mixture of 315 g. of m-cresol and 136 g. of hexamethylenetetramine was heated for an hour in 80 cc. of a 60 per cent (60 parts by volume of alcohol and 40 parts by volume of water) alcoholic solution. Too much heating caused t h e addition product t o decompose and pass over into the noncrystallizing resinous material. By withdrawing portions of the mixture from time t o time, and cooling slightly, it could be observed, by t h e formation of crystals, when t h e most favorable point was reached before resinification set in. On stopping t h e heating, crystals appeared even in the hot solution. T h e crystals were filtered off and pressed on a porous tile t o get rid of 1
J . Sac. Chcm. Ind.. 37 (1918), 260.
V O 13, ~ No. 2
t h e adhering sirupy material. The product was then dissolved in hot 95 per cent alcohol. On cooling, long, fine, needle-like crystals separated out. Analysis showed t h a t t h e substance was not formed on a 1: 3 basis as is the case with t h e ordinary hexamethylenetetramine triphenol, but was an addition product of 1 mole of hexamethylenetetramine and 2 moles of m-cresol. Calculated for I -F ound--CsI-Ii2Na.2CsHaOH.CHa 1 2 3 Per cent Per cent Czo.. . . . . . . . . 67.40 67.45 67.23 ... Ens . . . . . . . . . . 7.87 8.03 7.80 1 7 4 . . ......... 15.73 ...... 15:59
4
... 15180
Av. 67.34 7.96 15.69
Hexamethylenetetramine di-m-cresol has not a true melting point, since when the substance is held a t a temperature around its point of liquefaction, 90" C., it undergoes decomposition, passing over into the irreversible resinous stage. The compound is very soluble in hot 95 per cent alcohol, the solubility increasing with the temperature. A characteristic feature is t h a t when i t is placed in a sufficient amount of water or ether there is a very decided tendency towards a splitting of the product. In water the solubility of t h e hexamethylenetetramine shows up predominantly, as it is dissolved by the water leaving insoluble cresol as an oil. I n ether the solubility of the nz-cresol predominates, and the compound breaks up leaving t h e insoluble hexamethylenetetramine as a precipitate. The solubility in benzene is moderate, but increases with t h e temperature. Acetone has t h e same effect on t h e substance as has ether, t h a t is, breaking up the structure b y dissolving out t h e soluble cresol and leaving the insoluble hexamethylenetetramine. H E X A M E T H Y L E N E T E T R A M I N E DI-$-CRESOL
Pure $-cresol was first made from $-toluidine. When it was found t h a t a n addition product was formed with hexamethylenetetramine, a larger quantity of the material was made by the method given by Fox and Barker.l A mixture of 385 g. of #-cresol and 167 g. of hexamethylenetetramine in 150 cc. of 95 per cent alcohol was heated on a steam bath for 1 . 5 hrs. The same precaution must be observed here as in t h e case of the formation of the m-cresol compound. On allowing t h e liquid t o stand at room temperature, crystals separate out. The compound was recrystallized from 50 per cent alcohol. The addition product has no melting point, but begins t o resinify a t t h e temperature of liquefaction, 87.0" C. The decomposition is shown when the substance turns brown and partially resinifies upon heating in a sealed glass tube for 3 hrs. a t a temperature of 90" t o 10oo. Analysis shows t h a t it has t h e same proportion of t h e two constituents as the m-compound, namely, 1 mole of hexamethylenetetramine and 2 moles of #-cresol.
........ .......... ...........
Go..
Hzs N4
Calculated for C ~ H I Z N ~ . ~ C ~ H I O H 1, C H ~ 2 Per cent 67.40 67.18 67.35 7.87 8.20 8.01 15.73
1
LOC.cit., p. 268.
...
...
Found Per 3cent
... 15:82
4
.., ...
15.71
Av
.
67.27 8.10 15.76
Feb., 1921
T H E J O U R N A L O F I N D U S T R I A L A N D EiQGIiVEERING C H E M I S T R Y
The same qualitative solubilities as applied t o t h e m-cresol product apply t o t h e p-cresol compound. H E X A M E T H Y LE N E T ET R A MINE M 0S O - 0-CR E S 0L
........... .............
Cis.. Hzo Na
..............
Found 3 Fer cent 63.20 7.94 ,.. 22:?4
... 22:69
Calculated for C B H ~ ~ N I . C ~ H ~ O H . C ! HFound ~ Per cent Per cent 63.04 62.90 8.22 8,07 22.70 22 55
............................ .......................... .............................
From this i t appears t h a t there is ‘but one addition product of o-cresol and hexamethylenetetramine, and t h a t is with one mole of each of t h e two constituents present. HEXAMETHYLEKETETRAMINE HVDROQUINOL
Moschatos a n d Tollens2 give for t h e preparation of this compound 4 g. of hexamethylenetetramine in 5 g. of water mixed with 33 g . of hydroquinol in 4 g. of water. The product, purified b y washing with water and with ether, a n d drying over sulfuric acid, analyzed as follows: 1 2
Lac. cit. A n n . , 474 (1892-3). 283
T,-3
...
...
2i:i7
22:47
Calculated for CEHI?N&.C~H?(OH)Z Found Per cent Per cent . . . 57.60 57.35 7.20 7.11 22.40 22.46
... ...
On heating, part of t h e compound sublimed and part charred with very little melting. This behavior is similar t o t h a t of t h e hexamethylenetetramine o-cresol compound. HEXAMETHYLENETETRAISINE RESORCINOL
4
...
by M. and 2
This was checked up as follows: 5 g. C G H ~ ( O H ) ~ in 9 cc. of water were mixed with a solution of 6 g. of hexamethylenetetramine i n 10 cc. of water. The solution was heated on a water bath for 30 min., then allowed t o stand over night. Crystals washed with water, then ether, and dried over sulfuric acid. Analysis showed:
1
2
It was thought t h a t i t might be possible t o isolate a compound of o-cresol which would have t h e same proportions of t h e two constituents as have t h e p a n d t h e m-cresol intermediates. ’The crystals of hexamethylenetetramine were dissolved directly in t h e o-cresol, and with portions of this solution various runs were made in which t h e time factor of heating was t h e variable. Heating was accomplished on a water bath, t h e time varying from 2 t o :LO hrs. For t h e runs with a small amount of heating t h e solution was clear, while with t h e runs extending over 10 hrs. t h e solution was dark brown, showing t h a t a reaction had set in with t h e formation of t h e resinous material. After allowing t h e solutions t o stand several days t h e crystals were filtered off, pressed. on porous tile, a n d recrystallized from alcohol. I n all cases analysis of t h e crystals showed t h a t t h e product was a compound with a 1: 1 proportion of hexamethylenetetramine and o-cresol. Crystals obtained after 8 hrs.’ heating showed t h e following composition :
Xzo. N,
......................
...
Pure o-cresol was made according t o t h e method of Fox and Barker.l A mixture of 475 g. of o-cresol and 205 g. of hexamethylenetetramine in 100 cc. of 95 per cent alcohol was heated on a water b a t h for 2.5 hrs. On allowing t o cool a t room temperature, crystals separated out. These were recrystallized from 95 per cent alcohol. The compound behaves somewhat differently from t h e p - and m-cresol addition products, since on heating there was no sharp melting point t o t h e liquid stage, followed b y a final passing over t o t h e resinous material. A small portion seemed t o soften on heating and show signs of melting, b u t most of t h e substance either sublimed or charred. Analysis showed t h a t t h e proportion of hexamethylenetetramine t o o-cresol was 1: 1. Calculated for c C B H ~ Z N ~ . C E H ~ O H . C1H ~ Per ccnt 62.90 63.12 8.07 8.07 22.55 ...
Calculated for -Found C~HIZN~.CEH~(OH1 )~ Per cent Cia 57.60 57.20 H i s ...................... 7.20 7.77 Na ....................... 22.40
137
Moschatos and Tollens heating a mixture of 2 g. dissolved in 3 g. of water solved in 3 g. of water. compound was found b y be as follows:
Calculated for -Found C B H I Z N I . C ~ H ~ ( O H )1Z Per cent 57 60 57.14 7.20 7.43 22.40
........... .............
CIZ..
His
formed t h e compound by of hexamethylenetetramine and 3 g. of resorcinol disThe composition of t h e Moschatos and Tollens t o
.
N4 ..............
...
by M. and T.--2 3 4 Per cent 57.35 7.42 22:09 22:j2
...
...
...
By following t h e same order of procedure as outlined above, a precipitate was easily obtained. On analysis t h e composition was found t o be t h e same as t h a t represented b y Moschatos and Tollens:
............................ ............................ Na .............................. C~Z Hi8
Calculated Per cent 57.60 7.20 22.40
Found Per cent 57.39 7.10 22.34
This compound also shows no melting point, which is similar t o t h e hydroquinol and t h e o-cresol intermediates. I t seems t o be a characteristic feature of t h e hitherto observed phenol hexamethylenetetramine compounds t h a t i t is necessary t h a t there be at least 2 moles of t h e phenol t o 1 of t h e hexamethylenetetramine in order t h a t there be a well-defined point of liquefaction. HEXAMETHYLENETETRAMIXE CARVACROL
The carvacrol obtained for use in this work was made from cymene.l It ran 93 per cent pure, t h e other constituents being approximately 6 per cent thymol and 1 per cent thiophenols. The product was purified according t o t h e method developed by Mr Allan Leerburger: A very stiff paste of the carvacrol and lead acetate was allowed to stand a t room temperature for 30 hrs. The mass was broken up and the phenols extracted with petroleum ether. The carvacrol-lead acetate compound is soluble in the petroleum ether, whereas the thymol-lead acetate is insolubk in the ether, giving a means of separating the two phenols. After allowing the petroleum ether to evaporate, the liquid was washed with 1
Hixson and McKee, THISJOURNAL 10 (1918), 982
Vol. 13, No. 2
T H E J O U R N A L OAT I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y
138
H
a normal solution of mercuric chloride (using as the solvent 50 parts by volume of water and 50 parts by volume of alcohol).
This removes the thiophenols, leaving the carvacrol as the oil. The carvacrol then distilled in a 4-bulb fractionating column, the portion boiling between 237' and 239' C. being taken. One mole of hexamethylenetetramine in just sufficient 95 per cent alcohol t o dissolve t h e crystals was mixed with 3 moles of the purified carvacrol. The mixture was heated on a water bath for 40 hrs., then allowed t o stand a t room temperature for 1 wk. The uncrystallized mass was dissolved out by mixing with kerosene. The fine precipitate was filtered, and t h e crystals dissolved in hot 95 per cent alcohol. On cooling t h e alcoholic solution t h e compound crystallized out readily. Further purification was made by repeating t h e crystallization from hot 95 per cent alcohol. The compound shows a point of liquefaction at 148' C., a t which point i t resinifies quickly. It is very soluble in hot 95 per cent alcohol, but insoluble in t h e cold alcohol. An important point is t h a t it is very soluble in ether and acetone. Some of t h e cresol compounds, as has been stated, may be broken up in water, ether, and acetone, t h e two latter solvents dissolving out t h e easily soluble cresols and leaving t h e insoluble hexamethylenetetramine as a precipitate. This difference in solubility between t h e carvacrol and cresol compounds may be due t o t h e difference in linkings of different phenols with t h e hexamethylenetetramine. Analysis showed t h a t t h e compound was not of the same order of addition as were cresol and phenol products, which were in t h e proportion of 1 mole of hexamethylenetetramine t o 2 moles of t h e m or p-cresol, and 3 moles of phenol to 1 mole of hexamethylenetetramine. The composition was found t o be as follows:
...................
Carbon Hydrogen.. Nitrogen..
............... ................
Run 1 Per cent 76.95 9.20 4.92
Run 2 Per cent 77.10 9.27 4.99
Average Per cent 77.02 9.23 4.96
From the table below i t is clearly seen t h a t t h e hexamethylenetetramine is not directly added t o t h e carvacrol as it is in t h e case of the cresols and phenols. CsH1zNa.3CioH140 Found COHIZN~.CIOHI~O C~HI~NII.ZCIPH I~O 77.02 71.00 73.25 Carbon.. 66.25 9.50 4.96 Nitrogen. . . . . 19.35 12.75 9.23 9.10 9.16 Hydrogen. 8.96
....
...
The percentages found do not correspond t o any simple proportion of addition, as was shown in t h e case of the other phenols mentioned. However, if we assume t h a t a nitrogen is broken out of t h e structure of t h e hexamethylenetetramine t o form ammonia with hydrogens of t h e hydroxyls of 3 moles of carvacrol. and further t h a t 2 moles are taken u p additively by one or two of the other nitrogens t h e percentages of carbon, hydrogen, and nitrogen correspond exactly with the percentages as found. The smell of ammonia toward t h e end of t h e heating in the formation of this compound seems t o bear out this point t h a t ammonia is split out, ' b u t no quantitative determination has thus far been undertaken. The diagrams, in which R represents t h e radical part of t h e carvacrol structure, illustrate possible arrangements.
I
N-OR
N--CHr-OR
/ \ CHz 'CHz
or
/ I RO€€-N-CHz-N-CHz-OR / \ ROH
CHZ CHa
I I\
\
RQ-N-CH~N-CHe--BR
H CHz-OR
CHrOR
It is t o be pointed out t h a t although t h e interpretation of t h e structure of hexamethylenetetramine is thus far rather arbitrary, and although t h e correct one may be found t o be quite different from t h e above, t h e percentages of elements will in all cases be t h e same for t h e theoretical carvacrol compound. Calculated from Above Structure Per cent Carbon . . . . . . . . . . . . . . . . . . 76.95 Nitrogen. 4.84 Hydrogen.. . . . . . . . . . . . . . . . 9 . 1 0
................
Found Per cent 77.02 4.96 9.23
Deviation Per cent 0.01
0.12 0 13
All this becomes rather easy of interpretation if in t h e formation of the hexamethylenetetramine carvacrol compound there has been 1 mole of ammonia split out and there have been 2 moles of carvacrol added to one of t h e nitrogen. ENERGY RELATIONSHIP O F P H E N O L I C H E X A M E T H Y L E N E COMPOUNDS
t h e heat of combustion of the phenol, hexamethylenetetramine, and hexamethylenetetramine triphenol, a n Emerson adiabatic bomb calorimeter provided with a proper stirrer and a Beckmann thermometer graduated t o give an estimated For t h e heat of solution reading of O . O 0 l o were used t h e bomb was eliminated and t h e metal can for t h e water replaced b y a glass container. The substance whose heat was t o be determined was held i n a glassstoppered weighing bottle, t h e cover of which was removed b y small wires passing through t h e third hole i n t h e t o p of t h e calorimeter jacket. DATA-The water equivalent of t h e calorimeter was determined i n t h e ordinary way b y burning a material whose heat of combustion was known. Naphthalene from t h e U. S. Bureau of Standards laboratory was used. By weighing t h e separate parts of t h e bomb and accessories, t h e water equivalent of t h e bomb was found t o be 4 5 3 g.; without t h e bomb i t was found t o be 70 g . The error in the first number was *2 g., and t h a t in t h e second number was * 5 g . Checking these values against other standard substances, 6320 was obtained for benzoic acid, whereas the Bureau of Standards gives 6329 cal. per g. as t h e correct result. The second value was used in finding t h e heat of solution of as pure sodium hydroxide as could be made without wasting too much time. The following shows a comparison of t h e heat of solution of sodium hydroxide as determined by Thomsen, and b y Berthelot, and as obtained in this study: APPARATUS-FOr
Kg. Cal. Thomsen. 9.94 Berthelot.. 9.78 Present work., ............... 9 . 8 5
.................... ...................
The errors here would seem t o be due t o the varied purity of t h e NaOH used, rather t h a n t o t h e manipulation of the apparatus.
I
T H E J O U R N A L OF INDUSTRIAL A N D E N G I N E E R I N G C H E M I S T R Y
Feb., 1921
( a ) H e a t of solution of hexamethylenetetramine. Kg. Cal. 4.902 4.896
Run 1 Run 2
-
AVERAGE 4.8991 (where 1 Cal. = 1000 small calories) 1 Delepine (Bull., [3] 16, 1200) gives the heat of solution of hexamethylenetetramine at 15O C. as 4.8 Cal.
( b ) Heat of solution of phenol. Run 1 Run 2
Kg. Cal. -2.92 -2.87
AVERAGE -2.891 Landolt and Bornstein, 3rd Ed., p. 419, give for the heat of solution of phenol -2.6 Cal. f
Heat of reaction of phenol and hexamethylenetetramine in a n aqueous solution-In this observation t h e phenol was first added t o the water, then solid hexamethylenetetramine added in t h e manner stated above. The excess of rise of temperature above t h a t given by t h e hexamethylenetetramine would be due t o the reaction of t h e amine and t h e phenol. This is the weak point of this method of determining t h e heat of formation of hexamethylenetetramine triphenol, since i t is difficult t o obtain accurately t h e amounts of amine and phenol t h a t have combined in solution. After the reaction t h e hexamethylenetetramine solution was distilled t o obtain t h e phenol, the amount of which was determined by t h e tribromophenol method.‘ T h e error in this way would be in the dissociation of t h e triphenol compound on distillation of t h e phenol. It was found t h a t the energy reaction was (CH&N4Aq. 3C6HsOAq. = (CH&K4.3C&,o.Aq. 3.739 Cal. ( d ) Heat of f o r m a t i o n of hexamethylenetetramine. (c)
+
+
Heat of combustion of commercial hexamethylenetetramine: Run 1-7.380 Cal. per g. at constant volume Heat of combustion of hexamethylenetetramine resublimed in laboratory: Run 2 7.397 Cal. per g. at constant volume Run 3 7.399 AVERAGE 7.398
By means of t h e Hempel gas apparatus and freshly prepared solutions of sodium hydroxide and pyrogallol, t h e following results on the products of combustion were obtained: PRODUCTS Nz HNOa
............................. ..........................
Actual Result Grams 0.40 0.119 1.80
Theoretical Result Grams 0.38 0.112 1.88
cot ........................... T o represent t h e above results we can write the equation (CH?)cN4 18.55(0) = 6CO? 5.89H~O 1.89N2 0.22 HNOs f 0.1036.9 Cal. We know t h a t 6(C) $. 12(0) = 6C02 6 X 96.98 Cal. (1) (Land. and Born., 4th Ed., p. 855) 11.78(H) 5.89(0) = 5.89Hzo f 68.357 X 5.89 Cal. (2) (Land. and Born., 4th Ed., p. 850) 0 22(H) 0.22(N) 0.66(0) = 0.22HNOa 41.60 Cal. (3) (Land. and Born., 4th Ed., p. 854) Substituting these three equations in t h e found equation above, we have: 6(C) 12(H) 4(N) = (CH2)eNd - 43.18 Cal.
+
+
+
+
+ +
+
1
+
+
Therefore t h e heat of formation of hexamethylenetetramine = -43.18 Cal. ( e ) H e a t of f o r m a t i o n of phenol-Berthelotl gives the heat of combustion of phenol a t constant pressure and 18’ C. as 736.00 Cal. per mole. Required: 6(C) f 6(H) f ( 0 ) = C6HsO x Cnl. Found: C6H& 14(0) = 60 0 2 3H20 736.00 Cal. We know t h a t : 6(C) 12(0) = 60 0 2 96.98 Cal. (1) 3H2 30 = 3H20 f 68.36 X 3 Cal. (2) Substituting Equations 1 and 2 in the found equation we obtain 6(C) 6(H) f 0 = C6H6O 50.96 Cal. Accordingly t h e heat of formation of phenol is 50.96 Cal. per mole. This value is different from t h e one Berthelot gives2 because he uses t h e heat of formation of COa as 94.30 Cal. and t h e heat of formation of water as 69.00 Cal. These values are not considered correct and better values are used in t h e calculations above.3 The value 736.00 Cal. per mole for the heat of combustion of phenol is used here because i t represents t h e value obtained in this research. (f) Heat of solution of hexamethylenetetramine triphenol. Average value obtained was -10.671 kg. Cal. (g) Heat of f o r m a t i o n of hexamethylenetetramine triphenol. The heat of combustion of hexamethylenetetramine triphenol a t constant pressure was found t o be 3228.30 Cal. per mole. 4(N) = Required: 24(C) 30(H) f 3(0) ( C H & N ~ . ~ C O H ~xOCal. Found: ( C H ~ ) ~ N ~ . ~ Cf BH 30.73(0) &I = 24COz 14.71HzO f 1.71N~f O.58HN03 f 3228.30 Cal.
+
+
+
Allen’s “Commercial Organic Analysis,” 8th Ed., Vol. 3, p . 307.
+ +
+
+ +
-
139
+
+
+
+
+
+
We know: 24(C) 48(0) = 24(C02) f 96.98 X 24 Cal. (1) 29.42(H) 14.71(0) = 14.71Hz0 68.357 X 14.71 Cal. (2) 0.58(H) f 0.58(N) 1.74(0) = 58 HNOa 41.60 X 0.58 Cal. (3) Substituting these three equations in t h e above found equation and solving, we obtain t h e required equation: 24(C) 30(H) 3(0) -I- 4(N) = (CH~)CN~.~CBHBO f 128.76 Cal. per mole
+ +
+
+ +
+
+
The heat of formation of hexamethylenetetramine triphenol, starting with crystals of phenol a n d amine, is as follows: 3 phenol = Required: Hexamethylenetetramine H.T.P. x Cal. or (C&)&;c(crys.) f 3csH~O(CryS.) (CH,)oN4.3CeH~O(crys.)X Gal. We know: 6(C) f 12(H) 4(N) = (CHZ)&T~ - 43.18 Cal. (1) 18(C) 18(H) 3(0) = 3CsH60 3 X 50.96 Cal. ( 2 ) 24(C) 30iH) 3(0) 4(N) = ( C H Z ) ~ N ~ . ~ C ~128.76 H ~ O Cal. (3) Subtracting (1) and (2) from (3) we obtain: ( C H & N ~ ~ C ~ H=B(O CHZ)BN~.~C~H 19.06 ~ O Cal.
+
+
+
+ + +
+
+
+ +
+
1
Ann. chim. phys , (61 8 (1888), 326.
f
Vol. 2, p. 818.
8
Landolt and Bornstein, 4th Ed., p. 855.
T H E JOURNAL OF INDUSTRIAL A N D ENGINEERING CHEMISTRY
140
( h ) Heat of
combustion of
hexamethylenetetramirtt
di-#-cresol. AT CONSTANT VOLUME,20’ C. Cal. per G . Run 1 8.024 Run2 7.992 AVERAGE 8.008
(i) Heat of combustiolz of di-m-cresol. At constant volume and 20° C.
0‘) H e a t
of coinbuslion mono-resorcinol.
of
hexamethylenetetramine
-
8.010 Cal. per gram
hexamethylenetetramine
AT CONSTANT VOLUME AND 20’ Cal. per G. Run 1 6,730 Run2 6.700
c.
__-
AVERAGE 6.715
I n the addition reactions of hexamethylene tetramine with a phenol thus far investigated there does not seem t o be any definite rule by which one is enabled t o determine t h e number of moles of phenol t h a t will combine with t h e hexamethylenetetramine. Falk a n d Nelson1 have assumed t h a t in catalytic reactions there are binary and ternary compounds formed. Kendall* has called attention t o the import a n t general rule t h a t stable addition compounds are formed when there is a marked chemical contrast (acidic and basic) between t h e two reacting components. Thus in t h e additive compound formed between organic acids and phenols, t h e stability is very much greater when t h e organic acid is strong and t h e phenol weak, or vice versa, t h a n in t h e case in which both substances exhibit t h e same degree of acidity. A similar generalization holds for t h e addition compounds between two acids, or between an acid and a ketone, or an acid and an aldehyde. I n t h e case of t h e addition compounds formed in this work we have t h e phenol acting as t h e acid and t h e hexamethylenetetramine as t h e base. It might be assumed from this and from Kendall’s generalization t h a t t h e greater t h e chemical contrast t h e greater t h e stability of t h e compounds formed, and t h e greater t h e number of moles of phenol combining with t h e basic hexamethylenetetramine. This is not t h e case, however, in this instance. The degree of acidity seems t o have very little t o do with t h e extent of t h e reaction. Ordinary phenol, which is a weaker acid t h a n o-cresol, combines in t h e proportion of three moles of phenol t o one of hexamethylenetetramine, whereas the cresol combines in t h e proportion of 1: 1. Nitric acid, a very strong acid in comparison with t h e phenol combines only in t h e proportion of 1 mole of hexamethylenetetramine t o 2 moles of acid. Hence we cannot apply t h e generalization stated above t o t h e case of phenol addition products. Again, t h e three cresols have practically t h e same order of hydrogen-ion c ~ n c e n t r a t i o n ,b~u t with t h e p - and m-compounds there are 2 moles adding, whereas with t h e o-cresol there is only 1 mole adding t o t h e hexamethylenetetramine. 1
8
J . A m , Chem. Soc., 37 (1915), 1732. I b i d . , 86 (1914), 2498. Scudder, “Conductivity and Ionization Constants.”
Vol. 13, No. 2
Since t h e activity of phenol is greatly diminished in t h e case of t h e cresols, by t h e presence of a methyl group, i t might be said t h a t t h e more negative t h e benzene ring is made with negative groups (nitro and hydroxy) t h e greater t h e activity and t h e greater t h e number of moles uniting. With hydroquinol and resorcinol, where there are two hydroxy groups, t h e opposite is true. They react slowly with hexamethylenetetramine, and then only in the proportion of one mole of t h e phenol t o one of t h e amine. Picric acid, which contains three nitro groups and one hydroxy group, should represent a substance in which t h e benzene ring has practically t h e maximum of negative groups, and hence should have high combining properties. Moschatos and Tollens found t h a t t h e proportion was only 1: 1. As yet no rule can be laid down connecting t h e acidity, or t h e degree t o which t h e benzene ring is made negative by negative groupings, with t h e additive properties of phenols and hexamethylenetetramine. Why should two moles of p - and m-cresol unite with one mole of hexamethylenetetramine, while only one mole of o-cresol unites with one mole of amine? One difference lies in t h e structure assumed for t h e three cresols. The hydroxy group of t h e p - and mcresols has on each side of it a hydrogen, while t h e hydroxy group of o-cresol has a hydrogen on b u t one side. From this i t would seem t h a t t h e extent of addition depends upon t h e number and activity of the hydrogens adjacent to the reacting hydroxy group. The structure of hexamethylenetetramine as given b y
does not seem t o represent all t h e facts as presented by t h e addition products with phenols. Here t h e four nitrogens are all tertiary in character and we should expect t h a t hexamethylenetetramine would add four moles of an alkyl halide. A. Wohl’ found t h a t b u t one mole of methyl iodide was taken up additively. I n all t h e phenol addition compounds t h a t have been isolated there is not one case where t h e number of moles of phenol combining with one mole of hexamethylenetetramine is greater t h a n three. I n t h e carvacrol compound found in this investigation there is strong evidence t h a t one nitrogen is more reactive t h a n the others. This is shown by t h e fact t h a t ammonia has been split out with one of t h e nitrogens before one oE t h e other three has added any phenol. 1
Btu., 19 (1886), 1840
Feb., 1921
T H E J O U R N A L OF I N D U S T R I A L A N D E N G I N E E R I N G C H E M I S T R Y
Hexamethylenetetramine is formed from ammonia a n d formaldehyde. Tertiary amines are formed from ammonia and a n alcohol. Alcohol is a lower oxidation product t h a n is t h e aldehyde. Again, amides a r e formed from ammonia and a n acid, t h e latter being of a higher oxidation t h a n the aldehyde. From this we might expect t h a t hexamethylenetetramine should n o t be represented wholly as a tertiary amine, b u t should exhibit a character midway between t h e tertiary amine and a n amide. It does not seem from these considerations t h a t the structure of hexamethylenetetramine is properly represented when written in t h e above manner. STUDIES ON BAST FIBERS. 11-CELLULOSE FIBERS By Yoshisuke Uyeda
IN BAST
LABORATORY 08 AGRICULTURAL CHEMISTRY, UNIVERSITY OF CALIFORNIA AGRICULTURAL EXPERIMENT STATION, BERKELEY, CAL. Received June 9, 1920
In a previous paper,‘ t h e proximate analysis of Korean bast fibers, according t o t h e modified methods proposed b y Dore12for the analysis of wood, was discussed from t h e standpoint of textile chemistry. T h e scheme of analysis proposed by Dore was found t o be well applicable t o cellulose material other t h a n wood. Generally speaking, our knowledge of t h e nature of t h e substances which make up t h e structure of t h e materials belonging t o so-called compound cellulose, and of the forms in which these substances are present, as well as their special functions in t h e plant, is still very limited. One of t h e chief reasons for this is t h a t no accurate method for t h e analysis of cellulosecontaining material has been established, a n d consequently t h e results obtained b y t h e various investigators have not been directly comparable. From this point of view, t h e analytical studies recently made b y Dore may be regarded a s a forward step in cellulose chemistry. Cellulose is t h e chief constituent of t h e bast fibers, a n d t h e amount of it in the fiber is, t o a great extent, a measure of its industrial importance. For instance, t h e full bleached textile goods of t h e bast fibers may be considered chiefly composed of cellulose itself. I n determining t h e cellulose content of t h e bast fibers by t h e chlorination method, t h e alkali treatments before chlorination were found t o have a n important effect on t h e yields of cellulose as reported in t h e previous paper. I n t h e present paper, t h e effects of various preliminary treatments before chlorination on t h e yields of cellulose are presented, and t h e properties of the cellulose t h u s obtained are studied and discussed from t h e standpoints of cellulose and textile chemistry. S C O P E OF T H E W O R K
The original method of Cross and Bevan3 for t h e determination of cellulose is taken as a starting point of t h e work. Renker‘ published a critical s t u d y of THISJOURNAL, 12 (1920), 573 2 I b i d . 11 (1919), 556. 3 “Cellulose,” 2nd Ed , p. 95. 1
1
“Bestimmungsmethoden der Cellulose,” Berlin, 1910.
141
t h e determination of cellulose in various cellulose materials; and in his method t h e material was directly subjected t o chlorination b y t h e Cross and Bevan method without t h e preliminary alkali treatment. Schorgerl also confirmed t h e method of Renker by his experiments with woods. Johnsen a n d Hovey2 proposed a new method of hydrolysis, using a mixt u r e of glacial acetic acid and glycerol before chlorination in t h e cellulose determination. Recently Dore3 compared these three methods of treatment in the case of wood and decided in favor of t h e Renker procedure. Now, i t is of much interest t o determine whether t h e relation which was found by Dore is applicable t o bast fibers. T h e bast fibers, which belong t o t h e so-called pectocelluloses, differ much in composition from woods, which are classified as lignocelluloses, a n d contain substances which are easily converted into soluble forms by t h e action of alkali. I n t h e present work, t h e same three methods are used. After preparation by ( I ) no preliminary hydrolysis, ( 2 ) alkali hydrolysis, and (3) acid hydrolysis, t h e materials are subjected t o chlorination, according ta t h e improved method of Johnsen a n d Hovey, which was also recommended by Dore. But it is very obvious t h a t t h e yields of cellulose obtained by these three methods are not directly comparable. Whether a smaller yield may indicate a purer cellulose or may signify a partial destruction of t h e cellulose itself is very questionable. It is, therefore, very necessary t o standardize t h e purity of t h e cellulose t h u s obtained. I n this work two methods are available for this requirement. One is t o determine t h e quantity of a- or normal cellulose in t h e residue of t h e various cellulose processes by using t h e mercerization test of Cross a n d Bevan,* as recommended by D ~ r e of , ~this laboratory, a n d t h e other is t o estimate t h e furfural yield by Tollens a n d Kroeber’s method.6 EXPERIMENTAL
T h e Korean hemp fiber whose proximate composition was given in t h e previous paper’ is taken as t h e sample material for this investigation. T h e fiber is cut into small pieces, having a n average length of I cm., and preserved in a Mason fruit jar throughout t h e experiment. Portions of one gram each are weighed and dried for 16 hrs. in a constant temperature electric oven kept a t 100’ C., extracted for 6 hrs. with benzene, then for 6 hrs. with g j per cent alcohol, as described in t h e previous paper. After this treatment, t h e cellulose estimations are made in three ways, as described in the paper published by Dore.8 Results are given in Table I. From Table I i t will be seen t h a t t h e results for yield of cellulose by Method I are in good agreement, but those from Methods 2 and 3 vary considerably be1 2
THISJOURNAL, 9 (1917), 561. Paper, 21 (1918), No 23, 36.
a THISJOURNAL, 12 (1920), 264.
“Paper Making,” 1918, p. 97 THISJOURNAL, 11 (1919), 556. 8 J. Landw , 48 (1900), 357. 7 Uyeda, LOC.c i t . * T H I SJOURNAL, 12 (1920), 21%. 4
5
