摘要:

238 DAWSON: ON THE NATURE OF POLYIODIDES AND THEIR XX.-On the Nature of Polyiodides and their Dissociu tion in Aqueous Solution. By H. M. DAWSON, Ph.D., B.Sc., late 1851 Exhibitioner. THE nature of the equilibrium i n solutions of potassium iodide con- taining dissolved iodine has been the subject of frequent investigation and discussion. From the fact that strong solutions of potassium iodide, when saturated with iodine, deposit the latter almost completely on dilution, Dossius and Weith (Zeit. Chem., 1869, 5, 379) concluded that no compound exists i n the solution. A similar conclusion was drawn by Baudrimont (Compt. rend., 1860,51, 287), who found that by means of carbon disulphide the whole of the iodine could be abstracted from a solution of potassium iodide. Le Blanc and Noyes (Zeit.physikal. Chern., 1890, 6, 385) found that ,the freezing point of a solution of potassium iodide is but very slightly altered on the addition of iodine; on the other hand, the electrical .conductivity of the solution was observed to be considerably smaller, These facts can only be satisfactorily explained by assuming that an additive compound is formed in the solution. By making use of the Jaw regulating the distribution of a substanco between two non-miscible solvents to determine the amount of uncombined iodide in the potassium iodide solution, Jakowkin (Zeit. physikal. Chem., 1894, 13, 359 ; 1896, 20,19) showed that for dilute solutions, the application of the law of massDISSOCIATION IN AQUEOUS SOLUTION. 239 action to the experimental numbers gave results in agreement with the assumption of a dissociation equilibrium represented by the equation, HI + I, KI,.Noyes and Seidensticker (Zed. physika2. Chem., 1898, 27, 357) finally have shown that in the limiting case where the Jilute potassium iodide solution is saturated with iodine the amount of halogen dissolved by solutions of different concentration is also in complete accordance with the above conception of the equilibrium in solution. I n Jakowkin’s experiments, a quantity of iodine was shaken up with carbon disulphide and a solution of potassium iodide of known strength, in a stoppered bottle at a constant temperature of 25O, until equilibrium was attained. From the titration of the carbon disulphide layer with sodium thiosulphate, the concentration OF the free uncom- bined iodine in the aqueous layer could be calculated, the total amount of iodine being obtained from the titration of the aqueous layer. I f A denote the iodine concentration in the carbon disulphide layer, c the iodine concentration in the potassium iodide solution, and k the distribution coefficient of iodine for CS,-H,O, the fraction of the iodine uncombined is If, further, v is the volume of the aqueous solution containing one gram-mol.of iodine as titrated by sodium thiosulphate, a the number of gram-mols. of potassium iodide in volume v, and if in the solution we have a dissociation equilibrium which can be represented by the equation then, neglecting electrolytic dissociation, there will be present in volume v, (1 - x) gram-mols.of the complex compound and (a - 1 + x) gram-mols. of potassium iodide. Application of the law of mass action to the above equation leads to the equilibrium constant, k = KI + I, K13, (a - 1 +x)x u ( l - x ) * Jakowkin found that the expression thus obtained for k is constant, provided the concentration of the iodine is not very great. On account of the electrolytic dissociation, the equilibrium in solution is, however,not quite so simple as the above equation would suggest, and i n order to set up a strictly applicable formula this dissociation must be taken into account. Assuming that a compound of the formula KI, exists in the solution, and that this undergoes electrolytic dissoci- ation with the formation of I, ions, then if the I, ion is one of the components of the equilibrium in which the free molecular iodine240 DAWSON: ON THE NATURE OF POLYIODIDES AND THEIR takes part, we may represent the equilibrium in solution by t h e following dissociation isotherms : t - (1).HI, z k + I,. (2). k1 K 3- I. f - - - (3). 13 = I + I,. Let k, and IC, denote the dissociation constants of the states of equi- librium represented by equations (1) and (2), and suppose that the degree of electrolytic dissociation of the triiadide and of the iodide in their common solution is y and a respectively. If, further, v, a, and x have the signification previously assigned to them, the constant, k, of the dissociation equilibrium in which, according to the above formula- tion, the uncombined iodine takes part is : jp(1-4 ~ 4 a - 1 +XI%, or V V2 (1); .. . . . &,x (a - 1 + x ) x Y V - X ) From this equation y and x might be eliminated by writing down Since the concen- the values for the dissociation constants k, and k,. trations of undissociated KI, and KI are - Y)(l - and 2, Ll -%)(a - 1 +x) f - - and those of the K? I,, and I ions respectively 21 y(i - xj + z(a - i 3- 4, p( 1 - q, and - + %>, we obtain for k, and k, V V by application of the law of mass action to the dissociation equilibria, the following expressions : - Y r Y ( l - 4 + z ( a - 1 + ~ ) ] k, -- (1 - Y b k,=- z[y(l - x) + z(a - 1 3- 4, (1 - x)w The resulting expression for k would obviously only contain the electrolytic dissociation constants k, and k2 in addition to the factors x, o, and a, which are determined directly by experiment.The ordinary method of determining the electrolytic dissociation constant by conductivity measurements fails, however, in the case of strong electrolytes. This cannot be taken as evidence or proof of the non-existence of such a constant in these cases, since it is a direct consequence of the application of the law of mass action, and the dissociation formulae not only result from the application of mole- cular kinetic conceptions, but rest on the sure basis of thermodynam- ics. It must rather be inferred that the ratio of the molecular con- ductivities&/&, does not give the truevalue of the degree of dissociationDISSOCIATION IN AQUEOUS SOLUTION. 241 in the case of strong electrolytes, According to Jahn (Zeit. physikal. Chem., 1900,33, 545), the reason of this is to be found in the experi- mental fact that in the case of strongly dissociated substances, the ionic mobility not only depends upon tho intrinsic nature of the ions, but also upon the concentration of the solution.This is the case even for very dilute solutions. Arrhenius (Zeit. phy~ikal. Chem., 1899, 31, 197) attributes the increase in the value of the Ostwald constant with increasing concentration of the solution to the stimulating action of the ions themselves in promoting electrolg tic dissociation, comparing it with the action of neutral salts in increasing the dissociation of weak acids. Until the anomalous behaviour of strong electrolytes is satisfactorily explained, the values of the dissociation constants of salts cannot be determined.The empirical constant of Rudolphi or van't Hoff has no theoretical signification and cannot be regarded as a dissociation constant. For the compounds under consideration in this paper thevalues of Ic, and k, are therefore not known, and the general equation for the equilibrium in the haloid salt solution cannot be directly subjected to the test of experiment. The experimental results can, however, be applied in an inverse manner to enable conclusions to be drawn with regard to the nature of the components taking part in the equilibrium. I f we assume that the triiodide and the iodide in their common solution are electrolytically dissociated to the same extent (y=z), which is equivalent to the assumption that the dissociation constants k, and k, are equal, the formula (1) for k reduces to the simpler form ( a - 1 +x)x u(1-x) k = (2) .. . which is identical with the dissociation formula set up by Jakowkin The experiments of Jakowkin, carried out at 25O, show that for solu- tions up to 1/4 normal in regard to potassium iodide and to 1/10 normal with respect to iodine, the values of k calculated from equation (2) are very constant. In order t o confirm these results and so justify any conclusions drawn from them and further to determine the influence of temperature on the dissociation, a series of experiments was carried out similar to those of Jakowkin but a t a different temperature. The temperature chosen was 13*5*, the lowest which could be conveniently maintained in the laboratory for considerable periods of time.For the calculation of the degree of dissociation of the polyhaloid compound, it was necessary to know the distribution ratio of iodine between carbon disulphide and water at 13.5'. Three experiments were carried out at 1 5 O (the ratio at this temperature being required for another investigation) the values obtained being 616.8, 619.6, and 624.0, mean=620. The value obtained by Jakowkin at 25' for the (loc. C i t . )242 DAWSON: ON THE NATURE OF POLYIODIDES AND THEIR smaller concentrations of iodine, such as were used in my experiments, was 585. By extrapolation, the value of the distribution ratio at 13.5' obtained is 625. The distribution coefficient varies considerably therefore with the temperature, decreasing by about 0.6 per cent. for a rise of lo.The experiments to confirm the validity of equation ( 2 ) were carried out as follows. About 30 C.C. each of carbon disulphide and of potassium iodide solution of known strength were introduced into a stoppered cylindrical separator, a weighed quantity of iodine added, and after shaking for five minutes and allowing to stand at 13.5' until the two layers had completely separated, the concentration of the iodine in each was determined. The small changes resulting from the mutual solubility of the two media, and the small influence of the presence of dissolved salt in the aqueous phase on the value of the distribution coefficient of the iodine, could be neglected. The following tables contain the experimental data and the calculated values of k from equation ( 2 ) : c is the iodine concentration in the aqueous solution, A that in the carbon disulphide ; x the fraction of uncombined iodine in the aqueous solution = A/625c, v the volume of the aqueous solution containing 1 gram of iodine and a the molecular ratio KI/12.A. G- I-- 118 N potassium iodide : 1 *255 2 *283 4'425 7 -185 6.627 z:; 48'10 114 N potassium iodide : 1 '229 2'154 4'231 4-313 6'469 8'898 12.66 3.205 5-751 11 -39 11-63 18'31 25.89 39'68 N potassium iodide : 5.731 11.10 19.14 54 -63 3'657 7.062 12'68 35'94 2. 0.008449 0'008893 0 *OO 9 24 9 0'01071 0 -0041 7 3 0 *004272 0*004306 0 '004315 0.004529 0 '004655 0*005015 0*001020 0'001018 0*00106 0'001052 2'. 202.1 111.1 57.33 35-31 206'4 117.8 59.96 58-82 39 '22 28-51 20'04 44.27 22.86 13-26 4.644 a.L. 25-27 13-89 7.166 4'414 51'60 29.45 14 '99 14-70 9-805 7.128 5.01 44.27 2 2 ~ ~ 13'26 4.644 0 *001 G21 0'001042 0'001006 0*00105 0.001027 0 '001037 0*00101 0'00101 0 -001 021 0'001005 0*001008 0 -0009984 0 '0009744 0*0009811 0 -000827DISSOCIATION IN AQUEOUS SOLUTION. 243 The value of k as calculated from equation (2) is constant provided the concentration of the halogen is not too great, although there appears to be a slight gradual decrease as the concentration of the potassium iodide increases, which can probably be explained by the influence of the dissolved salt on the value of the distribution co- efficient, which has been assumed to be constant. The mean value of the constant (0.001015) is, however, much smaller than that determined by Jakowkin at 2 5 O , namely, 0*00138.That this difference is merely due to the difference of temperature is shown by the result of two experiments made with l/S normal potassium iodide at 2 5 O , the values of k obtained being 0*001361 and 0.001351, which agree with the constant determined by this author. Even with 1/160 normal potass- ium iodide solutions, k at 2 5 O retains the same value as at the higher concentrations. The following numbers show the influence of temperature on the degree of chemical dissociation of the polyhaloid compound, the per- centage of uncombined iodine at the concentrations given in the table having been interpolated from curves representing the experi- mental data ; Concentration of iodine. 1/4N potassium iodide i+B molemlar, Degree of dissociation of the polyiodide at 13.5" 25" 0.439 per cent.0,575 per cent,. 0.452 ,, 0.592 ,, 0.481 ,, 0.628 ,, 0.905 ,? 1-16 ,, 0.97 ,) 1.25 ), 1-14 ,, 1-44 ), The dissociation of the polyhaloid compound obviously increases with rise of temperature, the relative increase for a rise of temperature from 13.5O to 2 5 O being in each of the above cases about 25 per cent. From the fact that the simplified equation (Z), represents the ex- perimental results within the above limits in such a satisfactory manner, it may be concluded that the assumption made in the de- duction of this equation from the general one is justifiable. In accordance with this assumption, namely, that the degree of electro- lytic dissociation of iodide and triiodide in the s-olution containing both is the same, potassium triiodide must be considered as a salt the electrolytic dissociation constant of which is the same as that of potassium iodide ; its character is therefore that of a normal salt of potassium.It should be noted that equation (I), would also represent the experimental results with the same amount of exactitude if instead of y/x=l, we assumed y/x=k' any constant, but such a244 DAWSON: ON TBE NATURE OF POLYIODIDES AND THEIR relationship between the degrees of dissociation of electrolyticallJ dissociating compounds for varying relative and absolute concentra fions is not known. The result thus arrived at regarding the nature of potassium tri. iodide is perhaps what might have been expected, if the latter bc regarded as a substance having the character of a true salt correspond. ing to an acid of the formula HI,.The case of the corresponding acid is, however, quite different. A p i o r i , it is impossible t o say whether this acid belongs to the weak acids, to the half electrolytes, or the strong acids. For the equilibrium in a solution of hydriodic acid containing iodine, dissociation isotherms can, however, be written down analogous to those employed in the case of the potassium iodide solution, and the application of the law of mass action leads as be- fore to the equilibrium constant k = for the reaction in which the free molecular iodine is supposed to take part, y being the degree of electrolytic dissociation of HI, and x that of the HI in the same solution. If the two halogen acids are not of the same order of magnitude in respect of their chemical affinities, the ratio z/y will vary very con- siderably with varying concentration of the solution and with the relative masses of the two compounds. If, as before, however, we assume that the electrolytic dissociation constants of HI, and of HI are equal, then y=x, and the equation reduces t o the simpler form represented by equation (2).A series of iodine distribution experiments similar to those pre- viously described, was carried out with solutions of hydriodic acid. The results are contained in the table on page 245, the numbers in the several columns corresponding to those in the previous table for solu- tions of potassium iodide. As was found by Jakowkin at 25', the value of k calculated from the simple equation remains constant within the limits of experi- mental error for all the concentrations investigated.I n agreement with his results also, the equilibrium constant for the acid solution is somewhat less than that for the salt solution, the numbers at 13.5' being 0*00098 and 0.001015 and at 25' 0*00134 and 0*00138 respectively. The constancy of k shows that the assumption made in the deduction of the simple formula is justified. The electrolytic dissociation con- stant of HI, is therefore of the same order of magnitude as that of hydriodic acid; in other words, the compound HI, belongs to the group of strong acids, and in aqueous solution at moderate dilutions is (apart from the chemieal dissociation) almost completely dissociated into its ions H and I,.x, (a- 1 +%)% Y 4 1 - 4 + -DISSOCIATION 1N AQUEOUS SOLUTION, 245 l j 8 N hydriodic acid : 0.730 I 3.662 2'260 11'66 5'001 28.84 7'615 1 15'21 1/16 iV hyclriodic acid : 0 $246 8 *128 2'046 1 22.11 118 2 N hydriodic acid : 0'5815 11.74 0'864 1 18'32 0.008026 0'008254 0-008646 0 -009227 0'01013 '0 001 609 0.01729 0'0323 0'03394 v. 347.6 112'2 84.94 50.73 33.32 394'8 124 *o 436 2 293.6 a. 43'45 14.02 10'62 6'311 4-165 24'88 i 9 5 13.63 9*Ti5 o '000 988 8 0-0009658 0-000!)878 9 '0009830 0 *00097 54 0~0009812 0 * 00 0 9 6 02 c)*000969 0 '0009822 dccording t o the theory of Abegg and Bodlander (.&it. anorg. Chem., 1899, 20, 453), every complex ioii may be supposed t o consist of a simple ion combined with a neutral component. -4s a result of this formation of a complex ion, the electro-affinity-that is, the aflinity of the atom or atomic group for the electric charge-of the ion is increased.The greatest tendency towards the formation of such complex ions by addition of a neutral group of atonis is exhibited by weak ions, that is, those of small electro-affinity. In the case of the complex ion I,, the neutral component I, has the same chemical composition as the simple ion I, analogous to which we have the complex divalent mercurous ion Hg,, which results from the mercuric ion Hg by addition of a neutral atom of mercury. I n the case of weak ions, the formation of complex ions has an enormous effect in increasing the electro-affinity, compare the S ion with the SO, ion, but in the case of strong ions which, as a rule, have a much smaller tendency to form complex ions, the increase is relatively small, com- pare the Cl, ClO,, and C10, ions.The electro-afinity of the latter ions is not very different. The iodine ion I is already a strongly electro- negative ion, and the formation of a complex ion by the addition of a molecule of iodine would not be expected to produce any remarkable increase in the electronegative character. The result obtained above regarding the relative degrees of dissociation of hydrogen iodide and hydrogen triiodide is in harmony with this conclusion. - - + + + + - - - _ - - - - VOL. LXXJX. 8246 DAWSON: ON THE NATURE OF POLSIODIDES. With regard to the velocity of migration of the complex ion I, under the influence of a difference of potential, some light may be obtained from the measurements of Le Blanc and Noyes (Zoc.cit.) of the conductivity of potassium iodide solutions containing iodine. The following numbers are taken from their paper, v denoting the volume in litres containing one gram-molecule of KI, the moleoular conductivities having been multiplied by lo7 : Percentage diminution of 2’. pt, (III). P ’ ~ , (KI t 4/1OIj ptU compared with pC. 2 112.7 101.6 9.9 8 121.1 110.4 8.8 32 128.2 117.0 8.7 54 130-7 121.5 7.0 128 133.2 125.4 5.8 25 6 135-4 128.9 4.8 At all the concentrations investigated, the molecular conductivity is considerably diminished on the addition of iodine. The migration velocity of the complex ion would appear to be smaller than that of the simple ion. This diminution of the velocity of the ion with increasing complexity is in accordance with the regularities observed regarding the dependence of the velocities of organic acid ions on the number of the atoms contained in them when this number does not exceed a certain limit represented by about twelve atoms, I n connection with the views brought forward in this paper, some older observations of Johnson (this Journ., 1877, 31, 249) regarding the nature of potassium triiodide are of interest.By slow evaporation of a potassium iodide solution containing dissolved iodine over sul- phuric acid, he found that lustrous, dark blue prismatic crystals were deposited which had the composition required by the formula KI,. The specific gravity of these crystals was found to be 3.498, from which the atomic volume of the triiodide = 120.1, Now the atomic volume of potassium + three times the atomic volume of iodine = 122.2, whereas the atomic volume of potassium iodide + twice the atomic volume of iodine =105.6, “so that it appears that the tri- iodide of potassium does not consist of a molecule of potassium iodide united with a molecule of iodine, but of an atom of potassium united with three atoms of iodine without condensation ,” The existence of a large series of polyhalogen compounds of cssiurn and rubidium, in which the metals appear to be ter- and quinque- valent, has been shown in recent years by Wells (Zeit. anorg. Chenz., 1892, 1, 85) and Wheeler (Zeit. anorg. Chem., 1892, 1, 442 ; 2, 255), and by Erdmann. Without entering into the merits of the dis- tinction, the admission, by a large number of chemists, of the possi-SODEAU: THE DECOMPOSITION OF CHLORATES. PART 111. 247 bility of differentiation between so-called ‘ molecular ’ and atomic ’ compounds, forces upon us the question as to whether these polyhaloid derivatives of czeesium and rubidium belong to the first or to the second class. If the analogy between these compounds and the less stable potassium triiodide and the evidence in favour of the latter being an ‘atomic ’ compound be considered, then there are grounds for the conclusion that at any rate the trihaloid derivatives of cadurn and rubidium are also ‘ atomic ’ as opposed to ‘ molecular ’ compounds. THE YORESHIGE COLLEGE, LEEDS.

ISSN:0368-1645    

DOI:10.1039/CT9017900238

出版商:RSC    

年代:1901

数据来源: RSC

摘要:

SODEAU: THE DECOMPOSITION OF CHLORATES. PART III. 247 XXI.-The Decomposition of Chlowctes. Part 111. Calcizim Chlorate wid Silvey- Chlorate. By WILLIAM H. SODEAU, B.Mc. PURE calcium sulphate, silver sulphate, and barium chlorate were prepared,and by carefully balancing the barium salt with the sulphates, pure calcium and silver chlorates were respectively obtained ; these chlorates were subsequently recrystallised. When employing silver sulphate the barium chlorate should be added little by little, never exceeding the amount required to precipitate the sulphate already in solution, as an excess practically stops the reaction. Calcium Cldomte. By heating a t 160-165O, the crystals of the dihydrate Cla(C1O3),,2H,O were slowly rendered anhydrous. The product, which mas very deliquescent, contained only a faint trace of chloride and was neutral to methyl-orange ; unlike commercial specimens, it was free from magnesium.Potilitzin (compare Abstr,, 1892, 62, 1275) has studied the decomposition under atmospheric pressure with reference t o the formation of a small proportion of perchlorate and the relation oE temperature t o rate of decomposition. Exnnzination of the Decomposition Pyoducts.-The solid residue from the partial decomposition of 1 gram of chlorate was dissolved in water with the addition of a little methyl-orange, some N/50 nitric acid mas then added, and the slight excess titrated back with N/50 soda. The amount of calcium oxide or of liberated chlorine having been thus ascertained, the neutral liquid was diluted to 50 c.c., and the calcium chloride determined by titrating two 10 C.C.portions with s 2248 SODEAU ! THE DECOMPOSITION OF CHLORATES. N / l O silver nitrate, using potassium chromate as an indicator. TO test for peroxide, 20 C.C. were treated with 0.2 C.C. of iV/lOO perman- ganate and a little N/60 nitric acid ; the pink colour was not dis- charged even after standing for some minutes, and titration with N/lOO thiosulphate, after adding potassium iodide and starch, indicated no appreciable reduction of the 0.2 C.C. of permanganate. Had the whole of the oxide been present as CaO,, 1.0 to 7.3 C.C. would have been bleached. The oxide present in the residue must, therefore, be CaO (probably as oxychloride), but it may be formed by reduction of the peroxide, as calcium chlorate tends to be converted into perchlorate.The numerical results stated in Table I (p. 249), were calculated from the analyses of the residues, but accidental errors were guarded against by observing the loss of weight, volume of oxygen, and (in decom- positions conducted under atmospheric pressure) the amount of iodine liberated from potassium iodide. The potassium iodide did not become alkaline, hence the gas contained no appreciable amount of ozone or of oxides of chlorine. As in previous papers, the chlorine has been stated per 100 parts contained in the weight of chlorate actually decomposed. Decomposition under varied Conditions.-The chlorate in portions of 1 gram was decomposed in bulbs of 30 mm. diam. blown at the end of soda glass tubes of 8 mm.bore (placed in a bath of fusible metal to- gether with a thermometer) ; in other respects, these experiments were conducted in the same manner as the decompositions of lead chlorate (Trans., 1900, T7, 718). Calcium chlorate undergoes fusion at the beginning of the decom- position, and its great tendency to spirt has rendered it uncertain whether reduction of pressure affects the rate of decomposition. A slight retardation observed on reducing the pressure to 4 mm. in ex- periments 155 and 156 is prabably explicable by the fact t h a t the average temperature as well as the pressure must have been reduced, for under 4 mm. the greater part of the chlorate left the lower half of the bulb and a portion formed a broad ring in the 8 mm. tube. I n the slow decompositions, the temperature of the bath was about 295-320° during the greater part of the time, being slowly raised in order t o render the rate fairly uniform.In experiments 157 and 160 the average amount of chloride present during the decomposition was more than double that in experiments 155a and 156a, yet rather less free chlorine was obtained in the two former experiments ; this appears to show that no appreciable amount of chlorine was displaced from previously formed chloride, for otherwise the proportion should have been much greater in experiments 157 and 160. Experiments 155b and 1566 were similar to 157 and 160, except asPART III. CALCIUM CHLORATE AND SILVER CHLORATE. 249 154 158 159 155a 156n regards pressure ; their mean gives 0.58 per cent. of free chlorine at atmo- spheric pressure as compared with 0.54 per cent.under 4 mm. pressure. Since reduction of pressure caused no increase of free chlorine, there was no appreciable action between chlorine and calcium oxide, for this would have varied with the concentration (or pressure) of the gas. In experiments 154 and 158 the greater part of the decomposition About 0.1 ti) 120 140 TABLE I.- Culcium chlorate. Duration No* (minutes). I - 157 1 105 160 100 _- I 155bI 120 156bi 140 Proportion decomposed Pressure. (per cent.). 1 I I 95 9 I 1 atmosphere 94.1 78-5 j :: I -__ 31 '8 1-i mm. 28.3 1 I , 76 -0 1 1 atmosphere 80'5 i ,, I decomposed portion 1 Chlorine (total in I '=loo): j Mols. chloride , Mols. oxide. As 1 97.65 1 41.5 :% 1 97.98 48.4 1-52 98.50 1 65 0.51 99 -49 197 0'66 1 99'34 I 151 The '' proportion decomposed" under 4 mm.pressure would doubtless be much higher if ascertained for the portion remaining a t the bottom of the bulb. took place in a few seconds, the thermometer bulb being then at about 355' ; rapid decomposition greatly increased the proportion of free chlorine, the mean being 2.2 per cent. as compared with 0.58 in slow de- composition. Experiment 159 gave a lower result because the bath wa s heated too slowly and a fair amount of the chlorate decomposed before the rush occurred. The further significance of these results is discussed in the last sec- tion of the paper, Silver Chlwate. This chlorate is anhydrous and sparingly solubIe. After drying a t 100' it dissolves in water t o a practically clear solution neutral t o methyl-orange.When the salt is wrapped in paper, a tap with a hammer readily produces an explosion, but on a clean anvil I have not250 SODEAU : THE DECOMPOSITION OF CHLORATES. been able to explode it by a blow. If rapidly heated in a glass tube, i t explodes with a yellow flash and sharp report when the thermometer in the bath has reached a temperature of about 345O, the increase in the rate of decomposition giving very little warning of the approach of the explosion. Analyses of the solid residues resulting from three such experiments showed that the proportions of free chlorine mere respec- tively 5.1, 8.7, and 6.2 per 100 parts of chlorine present in the sub- stance. Some metallic silver was formed, probably because the temperature rose beyond that at which silver oxide ceases t o be even moderately stable.Silver chlorate undergoes slow decomposition a t about the same temperature as calcium chlorate and with far more difficulty than lead chlorate (Zoc. cit.), yet the violence of its decomposition when rapidly heated is much greater than is the case with either of these chlorates. XEow Decomposition wade$* Yavied Conditions.-The mode of working resembled that adopted in the case of calcium chlorate (p. 247) except that plain tubes of 18 mm. bore, sealed a t one end, were substituted for bulb tubes. By employing potassium iodide solution, the gas was shown to be free from ozone and oxides of chlorine, as in the case of other chlorates. The solid residue was completely soluble in ammonia and hence contained no appreciable amount of metallic silver, although, for example, a tenth of the residue consisted of silver oxide.Each residue was thoroughly disintegrated in N/10 nitric acid ; after standing for twenty-four hours methyl-orange was added and the excess of acid titrated with N/50 soda; the difference gave the amount o€ oxide in the residue and hence the amount of chlorine remaining free. The residual silver chloride * was weighed in order to determine the ‘‘ chlorine as chloride,” and the filtrate titrated with N/10 sodium chloride as a check upon the other results. The ‘‘ proportion decomposed” was calculated from the sum of oxide and chloride. Variation of pressure did not markedly affect the rate of decom- position, but it will be seen from Table 11.(p. 251) that, as in the case of lead chlorate (Zoc. cit.) each decrease of pressure caused an increase in the amount of chlorine remaining free. There was, therefore, considerable secondary action between the chlorine and the silver oxide, this being lessened on diminishing the concentration (that is, pressure) of the gas. The residues obtained under atmospheric pressure were practically white, whilst those formed under reduced pressure were dark brown, owing to the large proportion of oxide. I n experiment 168 there was a second tube in which the chlorate was * When this was dissolved in ammonia and reprecipitated with nitric acid no silver remained in solution, hence the whole of the silver oxide had been dissolved by the standard acid.PART 111. CALCIUM CHLORBTE AND SILVER CHLORATE.251 __- decomposed under atmospheric pressure whilst a current of dry air was led t o the bottom of the tube and up through the substance; the volume of air was nearly fifty times that of the gas evolved, and the pro- portionof free chlorinewas2.3 percent.-about ten times that ordinarily obtained under atmospheric pressure but only about a third of that found under 20 mm. pressure. The silver chlorate fused and remained fairly liquid until near the stage a t which these decompositions were stopped ; this will explain the relative inefficiency of the air current, the gas in many of the bubbles doubtless escaping dilution. I n experi- _____ TABLE I I. - Xi her ch lomte. 172 I 6 7 165~6 166n 168 110 120 130 140 110 Proportion iecomposec (per cent.) -_ ~ i 8 .5 i l . 8 50.0 63 *i 8 8 5 87.8 73.2 69 *6 72'4 60'4 68 '4 - Prcssnre. ___ - 1 atmosphere 9 Y 9 ? tbout 20 mm. 20 mm. 6 3 9 4 9 > 32 7 ) 3 1 ) 21 a ) > 1 atmosphere > 9 9 9 Chlorine (total in decomposed portion = 100). Free. -_-___ 0*15* 0*14* 024* 0.25* 7.76 6-55 16.8 19.1 21.1 22.6 18.4 -___- 5 *1 8.7 6'2 AS chloride, 99 85 99 '86 99.76 99.75 92-24 93'25 80 -9 i 8 '9 i 7 - 4 a3 ' 2 a1 -ti _.___ 94 -9 91.3 93.8 Equivs. chlorilie. h'quivs. oxide. 670" 690" 420" 400* 11'9 13-8 4.96 4-44 4'24 3.74 3 *42 18.5 10'4 15.1 * In the numbers marked with an asterisk, only the first figure is really significant. ments 167 and 172 there were comparison tubes each containing 1 gram of silver chlorate together with half a molecular proportion of well washed dry silver chloride, this addition more than doubling the average amount of chloride present.The chloride had no marked influence upon the rate of decomposition, but slightly more chlorine remained free, the (total) amounts being 20.9 per cent. a t 4 mm., and 18.1 per cent. a t 6 mm. As silver chloride appears t o dissolve in the fused chlorate, the increase, which mas about the same as that caused by lowering the pressure to the extent of 1 mm., is sufficiently accounted for by25 2 SODEAU : THE DECOMPOSITION OF CHLORATES. the decrease in reabsorption brought about by the dilution of the silver oxide. The conclusion is therefore drawn that silver chloride is practically unaffected by decomposing silver chlorate and hence that no appreciable proportion of the free chlorine accompanying the oxygen has been displaced from previously formed chloride.Duration and average temperature of decomposition did not vary t o any very great extent in the series of experiments at present under con- sideration, and increase of duration mas necessarily coupled with decrease of temperature, so that such variations as occurred tended to compensate for each other as regards the secondary action between chlorine and silver oxide, hence this action was mainly a function of the pressure of the gas. Plotting the observed pressures (ranging from 2.5 to 760 mm.) as ordinates and taking as abscissae the number of equivalents of silver chloride to one equivalent of oxide(Tab1e II., last column), it was noticed that all points lay close to a straight line, as might have been expected from theoretical considerations.On producing this line, it was found to cut the axes in the points (1.8, 0 mm.) and (0, - 3 mm.) respectively. It thus appears that if a perfect vacuum could be main- tained above the liquid, the observed ratio of chloride to oxide would be 1.8 : 1, or in other words 36 per cent. of the total chlorine mould be obtained in the free state,* To eliminate reabsorption, it is, however, necessary to put the average pressure in the 6ubbZes equal to zero. Under reduced pressure, the bubbles are large a t the surface, hence their (unknown) average internal pressure caused by surface tension cannot be great. The pressure axis is cut at - 3 mm., hence the supposition that the chlorate first of all decomposes entirely into oxide, chlorine, and oxygen requires that the average internal presure should be 3 mm.Summas. y of Resu Its. Considerations resembling those given in Parts I. and 11. (Trans., 1900, 77, 137, 717) will apply to calcium and silver chlorates. It follows that during the slow decomposition of the chlorates of (potassium T), barium, calcium, lead, and silver, the whole of the free chlorine is produced directly, that is, it is not displaced from previously formed chloride, and the primary cause of the production of oxygen and chlorine is the occurrence of two simultaneous and independent reactions which may be represented by the equations 2M(C1O3), = 2MC1, + 60, and 2M(C10,), = 2M0 + 2C1, + 50, $ where M is written for Ca, Ag,, &c.* This mode of elimination could not be employed with lead chlorate, as decrease of pressure necessitated a considerable increase of temperature. Jr Chlorine vanishingly small. $ Lead chlorate yields PbO, instead of PbO.PART 111. CALCIUM CHLORATE AND SILVER CHLOHATE. 263 With calcium chlorate, the ‘‘ chloride ’’ decomposition proceeds a t about 180 times the rate of the oxide ” decomposition, and the free chlorine has no appreciable action on the oxide or oxychloride; this chlorate resembles barium chlorate (Zoc. cit., p. 138), the main differ- ence being that the calcium salt yields much more free chlorine. I n the case of silver chlorate, the velocity of the ‘‘ chloride ” is less than 1.8 times that of the “ oxide” decomposition (hence more than 36 per cent.of the total chlorine is first liberated), but simultaneously the reaction Ag,O + C1, = 2AgCl+ 0 proceeds to an extent determined by the conditions, all but about 0-2 per cent. of the total chlorine recom- bining under atmospheric pressure. The latter results resemble those obtained with lead chlorate (Zoc. cit., p. ‘i17), but are more striking although less complete. As silver chlorate resembles the barium and calcium salts in fusing during decomposition, the proof of the absence of any recombination of chlorine during the decomposition of the two latter chlorates is supported by the behaviour of this chlorate in a more satisfactory manner than by the results obtained with (non- fusing) lead chlorate. The ‘6 chloride ” decomposition evolves much heat, whilst it is practically certain that the “ oxide ” decomposition is an endothermic reaction, and hence might be expected to gain upon the ‘‘ chloride ” decomposition when the temperature is greatly raised, as in violent decomposition.This will account for a much larger pro- portion of free chlorine being obtained when calcium chlorate is rapidly decomposed; in the case of silver chlorate, the increase may partly depend also upon the alteration in the conditions under which the chlorine acts upon the silver oxide. Lead chlorate and silver chlorate yield large quantities of oxide on decomposition, although in either case the oxide has a low heat of formation, and is rapidly attacked by the chlorine simultaneously produced; this accords well with the accepted view of the structure of the chlorate molecule, namely, that the met’al is joined directly t o oxygen. The decomposition of the chlorates of calcium, barium, and potassium has, however, been shown to consist almost entirely (and that of lead chlorate to the extent of about 13 per cent.) of the direct production of chloride and oxygen without the intermediate formation of oxide. This would appear to show either that these chlorates are transformed a t high temperatures into isomerides in which the metal is directly attached to chlorine, or else that the main decomposition is a reaction taking place between two molecules, t,he chlorine of one uniting wit,h the metal of the other.

ISSN:0368-1645    

DOI:10.1039/CT9017900247

出版商:RSC    

年代:1901

数据来源: RSC

摘要:

254 SENIER AND GOODWIN: THE ACTION OF ETHYLENE By ALFRED SENIER and WILLIAhf GOODWIX. AMONG the reactions announced by Hofmann in his classical researches on amines is that between ethylene dibromide and aniline whereby important bases are formed each of which has been fruitful as leading to subsequent discovery. Of these, two are especially notable, namely, diphenylethylenediamine and diphenyldiethylenediamine. The latter is. now called diphenylpiperazine, a name suggested by Merz and Mason (Bey., 1887, 20, 267), which shows in a happy manner its relation to other heterohexacyclic compounds. The f ollowing experi- ments were undertaken with the object of applying this reaction t o the production of the related xylyl and cumyl bases, and of obtaining from them derivatives homologous with those which Mills has de- scribed (Trans., 1900,77, 1020). We have also endeavoured to extend this reaction with the view of preparing methylenediamine derivatives and have succeeded in obtaining a series of well crystallisable methyl- ene compounds.The study of these as well as their behaviour with phenylcarbimide me propose to continue. Di~?/Z?/Zeth?/Zenedinmine, C,H,(NH*C1,H,BIe,),. Pure xylidine of commerce (b. p. 212') which consists chiefly of the unsymmetrical meta-isomeride was employed. Four mols. of the base were brought together with 1 mol. of ethylene dibromide in presence of sufficient sodium carbonate to neutralise the hydrogen bromide set free. In order t o moderate the reaction, the base was mixed with the sodium carbonate in a flask fitted with a reflux con- denser and the ethylene dibromide gradually added through the latter.The reaction was completed by heating in an oil-bath for three hours a t 145-150'; the mass being frequently shaken. The contents of the flask were shaken into cold water, which removed the sodium salts and left a soft, resinous mass. This was mashed with cold rectified spirit and afterwards extracted with the hot solvent. From this solution on cooling, white feathery crystals separated ; after recrystallising three times from absolute alcohol they melted at 74-75', On analysis : 0.2094 gave 0.61136 CO, and 0*1710 H,O. C18H,,N, requires C = 80.59 ; H = 8.95 per cent. DixyZ?lleth?lZenecmi~e nitrate, C,H,(NH* C,K3iWe,)2,2.KN0,, was prepared by adding dilute nitric acid to an alcoholic solution of the C-80.56; H=9*06.DIBROM IDE ON STLIDISE ASD 1Jr-CUUIDINE. 255 base.Rise of temperature was prevented by surrounding the vessel containing the solution with cold water. On standing, small colour- less needles were deposited which, after washing with water, melted a t 165". These were recrystallised from dilute alcohol and then melted at 166". 0.3982 gave 47.6 C.C. moist nitrogen at 13' and 747.5 mm. C,,H2,N,,BHN03 requires N r= 14.21 per cent. On analysis : N = 13.91. Dix?/lyleth?/Ze?zecEici~ni~ze Pkc&nicldomkie, U,H,( NH*C,H,Me,),,H,PtCl,, separates as bright yellow, glistening crystals when a solution of platinic chloride in alcohol is added to an alcoholic solution of the base. The salt was dried at 130°, and, on analysis : 0.0543 gave 0.0157 Pt.C,,H,,N,,~,PtCl, requires Pt = 28-79 per cent. Dixy Zy Zethylenedicmine iYercu~ichZoride, C,H,( NH* C,H3M e&, Hg Cl,, is deposited as pale yellow, well-formed crystals by allowing a mixture of alcoholic solutions of the base and mercuric chloride, in molecular proportion, to stand for 34 hours; it may be recrystallised from dilute alcohol. The melting point is not dehnite, but is reachedat about 118". On analysis, 34.3 per cent. of Hg was obtained, whilst the formula re- quires 37.1 per cent. There can be no doubt, however, that the formula given is the correct one, for it is very difficult to obtain these rnercurichlorides without admixture of either mercuric chloride or the base. Tetranitrodixy ZyZethy Zenedicintine, C,H,[ NH* C,H (NO,),Me,],, was obtained by treating a solution of the base in glacial acetic acid with an excess of concentrated nitric acid.On warming, a brisk reaction commenced, which, for its completion, required the application of heat. Pale yellow crystals made their appearance, increasing in quantity as the liquid cooled. The crystals were collected and washed, first with acetic acid and then with water. After drying, they melted a t 220'. On analysis : 0.3542 gave 57.2 C.C. moist nitrogen at 12' and 752.5 mm. N = 18-98. After the separation of the tetranitro-derivative, the solution, on addition of water, deposited a small quantity of an orange-red, semi- plastic mass, which became crystalline and melted at 52-53O. There was not sufficient of the substance to enable a further study to be made of it, but from the analogous experience of Nills (Trans., 1900, 77, 1021), in the case of ethylenediphenyldiamine there can be no doubt that this compound is an isomeric tetranitro-derivative.An attempt was next made to prepare a dinitro-derivative. In the Pt = 28.96. C,,H,,08N, requires N = 18.75 per cent.256 SENIER AKD GOODWIN: THE ACTION OF ETHYLENE first experiment, the base was dissolved in cold concentrated nitric acid, and when the reaction was complete the resulting liquid was thrown into cold water. An orange powder was precipitated, which melted sharply at 191-192'. N = 17.9, On analysis : 0,1723 gave 26.6 C.C. moist nitrogen a t 16" and '750 mm. C,,H,,O,N, requires N = 17-37 per cent. This substance, therefore, may be a tl.initrodix~lyZet~yle?aedianai?ze. I n a second experiment, an alcoholic solution of the base was heated to boiling with concentrated nitric acid.The solution was thrown into an excess of water, when brownish-yellow crystals separated which melted at 152--154O. N = 13.43. On analysis : 0.1598 gave 15.3 C.C. moist nitrogen a t 13" and 767 mm. C,,H,,O,N, requires N = 13.41 per cent. The dinitro-compound sought was not obtained, as the derivative just described has a composition corresponding to that of a mononitro- dixylyleth ylenediantine, C,H,Me,*NH* C2H4*NH C6H,Me2 NO,. Dixyzylpiperawhe, C6H3i\fe, N: ( C2H4),: N*C,H,Me2. This base was produced by using no alkali, but a larger proportion of ethylene dibromide, than in the preparation of the diamine. The reaction is more violent, and it is necessary to add the dibromide care- fully in small quantities.Gentle warming is required to start the reaction, which is completed by heating for 2 hours in an oil-bath a t 150O. The piperazine, after repeated recrystallisation from absolute alcohol, was obtained in shining, white leaflets which meIted a t 151". On analysis : 0.2374 gave 0,7113 CO, and 0*1997 H,O. C = 81.72 ; H= 9.34. 0.2802 ,, 0.5401 GO, ,, 0.2335 H20. C=81*7'7 ; H = 9-25, C,oH26N2 requires C = Sl.63 ; H = 8-84 per cent. Di-~-cunayZet?iyZenedicc.milte, C2H4(NH -C6H,Me,)2. The $-curnidine employed melted a t 67". The proportion of base to ethylene dibromide, and the method followed generally, were the same as described for the homologous xylyl compound, save that the recrystallisation was effected from benzene. A variation of this method, in which no sodium carbonate was used, but the mass, after washing with rectified spirit, was treated with potassium hydr- oxide solution before recrystallisation from benzene, gave a good yield.The diamine recrystallised twice from benzene was obtained in the form of brilliant, colourless needles melting at 168'. These dissolveDIBROMIDE ON XYLIDINE AND $P-CUMIDINE. 257 readily in benzene, but. only slightly in hot alcohol or ether. analysis : On 0.2550 gave 0.7556 CO, and 0.2190 H,O. C2,H2*N2 requires C = 81.08 ; H = 9.45 per cent. Di-JI-c2cnz~ZethyZenecliccnz.ine J i t m t e , C2H,(NH*C,H,~~e,),,2HN0,, was obtained in the same manner as described above for the corre- sponding xylyl salt.It consists of colourless needles melting at 154". On analysis : 0 = 80.81 ; H = 9.52. 0.2074 gave 22.4 C.C. moist nitrogen at 17'and 755 mm. N = 12.5. Di-$-cuml/Zetl~ylenedicciizine Platinichloride, C,,H,,O,N, requires N = 13.2 per cent. C,H,( N H*C,H21$e,),,H,Pt Cl,, was prepared as in the case of the corresponding xylyl compound, and precipitated as a buff-coloured powder. 0.3451 gave 0.0939 Pt. Pt=27-49. Di-~-cunL~Zetl~~lenediccmine Mercurichloride, On analysis : C20H,,N,,H,PtCl, requires Pt = 27-64 per cent. C,H,(NH*C,H2Me,)2,HgCl,, corresponds in method of preparation to its xylyl homologue. The crystals were a little darker in colour and melted at 150'. The analysis also gave a lower percentage of mercury than that calculated from the formala.Found Hg = 33.2 ; required Hg = 35.2 per cent, Dinitl.odi-~-c~,)~~letl~~lenediacmine, C2H,(NH*C,HMe,-N02)2.-It is remarkable that, working in exactly the same manner as for the preparation of the tetranitrodixylylethylenediamine, we obtained with the cumyl base, not a tetranitro-, but a dinitrocderivative. The product melted at 97-98', and was precipitated as a pale yellow powder. On analysis : 0.1694 gave 20.5 C.C. moist nitrogen at 163 and 7535 mm. N = 14.0. C2,H,,0,N, requires N = 14.5 per cent. An iaomeride obtained by dissolving the base in concentrated nitric acid with application of heat, gave, on throwing the liquid into water, a compound melting a t 49-50'. N = 14.43. On analysis : 0.2069 gave 26 C.C. moist nitrogen at 18' and 769.6 mm. C,,H,,O,N, requires N = 14.5 per cent. Di-$-cu.flzyZ~i~eraxine, C,H,Me,*N: (C,H,),:N*C,H,Me,. $-Cumidine (2 mols.) was mixed with ethylene dibromide (1 mol.), and heated in a closed tube at 200' f o r 3 hours. The contents were treated with boiling water and the residue dissolved i n258 SENIER AND GOODWIN : THE ACTION O F YHENPLCARBIMIDE hot rectified spirit ; from the solution, on cooling, crystals were de- posited which melted at 148-150’. Recrptallisation from ether gave large, brilliant needles which had the same melting point. They were readily soluble in chloroform, but less so in light petroleum. On analysis : C = 81.4 ; H = 9.28. 0.2338 gave 0,6984 CO, and 0.1954 H,O. C,,H,,N, requires C = 81 *9 1. H = 9-25 per cent. We desire t o express our indebtedness to MI-. Thomas Walsh, B.A., for assistance during the course of‘ these experiments. QUEES’S COLLEGE, GALWAY.

ISSN:0368-1645    

DOI:10.1039/CT9017900254

出版商:RSC    

年代:1901

数据来源: RSC

摘要:

258 SENIER AND GOODWKN : THE ACTION O F PHENPLCARBIMIDE XXII1.-The Action of Phenylcarbim.ide on Diphenyb-, Dialphyl-, and ~iiaa~lEthyl-clia~n~nes, By ALFRED SENIER and WILLIAM GOODWIN. WHEN the two ethylenediamines described in the previous paper are treated with phenylcarbimide, an interesting reaction, resulting in the formation of ethylenediamines of the type C,H,(NR'*CO*NHPh),, is brought to light. Combination is effected between each of t h e alphyl amino-groups of the diamine and the carbonyl groups of t h e carbimide by the severance of the double linking between the carbon and nitrogen of the latter. The reaction is of the same type as t h a t between ethylenediamine hydrochloride and silver carbimide, whereby ethylenediurea, C,H,(NH*CO*NH,),, was obtained by Volhard (Amrtalen, 1861, 119, 349), and is analogous to the general reaction between phenylcarbimide and primary and secondary amines.The compounds are also homologous with the trimethylenecarbanilide of Hanssen, C,H&NH*CO*NHPh), (Bey., 1887, 20, 783). The latter experimenter did, indeed, by a different reaction, that between ethyl- enedicarbanilic chloride and aniline, produce a compound which was doubtless an impure specimen of the dicarbanilidodiphenylethylenedi- amine described below, but it melted at 170" and resisted all attempts a t purification. We have succeeded in extending this reaction t o diphenyl- and several dialphyl-diamines, and its general character is further proved by its application t o the naphthgl group. In the latter case, however, unsymmetrical or mono-ureas have so far only been obtained.05 DIPHEX'TL-, DIALPHYL-, ASD DISAPHTHTL-DIAMINES.25 9 D i c a i . b a ? 2 i Z i d o d ~ ~ 3 ~ ~ e ~ ~ ~ ~ e t ~ ~ Zenedimtine (Eth yle.necliccwbanilide), C,H,(NPh*CO*NHPh),. Diphenylethylenediamine (1 mol.), carefully dried by standing over aulphuric acid in a vacuum, WAS heated in a closed tube with phenyl- carbimide ( 2 mols.) at 120-1 30' for 3 hours, The product, of a steel grey colour, was washed with ether and crystallised from absolute alcohol. The crystals thus obtained were, however, still impure, and needed to be recrystallised several times from benzene, and finally from absolute alcohol. They consist of pale, bluish-grey needles which melt a t 220O. On analysis : 0.2371 gave 0.6512 CO, and 0.1306 H,O. C=74*9 ; H=6*12.C,,H,,O,N, requires 0 = 74.66 ; H = 5.77 per cent. DicarbccniZidoditolyZetT~~lEenedianaines, C,H,[N(C,H,Me)*CO* NHPh],. The ditolylethylenediamines (1 mol.) were heated in closed tubes with phenylcarbimide (2 mols.) in the manner above described, except that, in the case of the para-derivative, a somewhat higher tempera- ture, 150°, gave the best yield. I n the preparation of the ditolyl- ethylenediamines employed, the method described by Mills (Trans., 1900, 77, 1020) mas followed in the case of the ortho-derivative, that of Francis (Trans., 1897, 71, 426) for the meta-, and of Gretillat (Monit. Scient., 1873, [iii], 3, 383) for the para-compound. Dicaybanilidodi-o-tolylethylenediainine appeared on opening the tube as a mass of pink crystals which, after being washed with ether and recrystallised from absolute alcohol, retained R faint pink colour.I t melted a t 195-196'. C == 74-87 ; H = 6.6. On analysis : 0-2765 gave 0.7591 CO, and 0.1 654 H,O. C,H,,O,N, requires C = 75-31 ; H = 6.27 per cent. Dicarbanilidodi-m-toZyZet~ylenedi~naine crystallised in the form of brilliant, colourless needles melting at 181.5". On analysis : 0.3047 gave 0.8402 CO, and 0.1788 H,O. C30H3,0,N, requires C = 75.31 ; H = 6.27 per cent. Dica~banilidodi-p-toZ~ZEet~yZe~e~ia~i~2e was purified in the manner described for the ortho-derivative, except that after washing the tube contents with ether, the residue was recrystallised several times from benzene before the final recrystallisation from absolute alcohol. It separated in the form of fine, colourless needles which melted a t 186'.On analysis : C = 75.2 ; H = 6-52. 0.2990 gave 0.8236 CO, and 0.1766 H,O. C = 75.03 ; H = 6.58. C,oH,oO,N, requires C = 75-31 ; H - 6-27 per cent,260 THE ACTION OF PHENYLCARBIMIDE ON DIAMINES. Dicurbaieilidodix?/Zyleth~Zenediamine, C,H,[N(C,H,Me,)*CO*NHPh],. Dixylylethylenediamine (this vol., p. 254) was heated in a closed tube with phenylcarbimide for 3 hours at 135-140'. The proportion used was one molecule of the base to two of the carbimide. The semi-crystalline mass which resulted was washed with a small quan- tity of rectified spirit, and the residue was recrystaliised several times from absolute alcohol. Well formed, colourless needles melting sharply at 167' were finally obtained.On analysis : 0,3037 gave 0.8453 CO, and 0.1912 H,O. C = 75.9 ; H = 6.99. C,,H3,0,N, requires C = 75.88 ; K = 6 a72 per cent. Dicavbanilidodi- JI-cz~?ltylet~~ylelzediccmine, C,H,[N( C,H,Me,) CO*NHPh],. Di-+-cumylethylenediamine (this vol. , p. 256) and phenylcarb- imide, in the proportion of one molecule to two, were heated in a closed tube for 4 hours at 180-190'. The contents were well washed with ether and recrystallised several times from alcohol. To obtain the compound in R pure condition, i t was, however, necessary to re- crystallise again several times from benzene, and finally from ether, in which latter solvent it is only slightly soluble. It consists of glistening, colourless needles which melt at 191'. On analysis : 0,1615 gave 0.4532 CO, and 0.1108 H,O.C=76.53 ; H=7.62. C,,H,,O,N, requires C = 76.4 ; H = 7-11 per cent. Monocarbunilidodi- a-napltth ylethylenediucmine, C,,H7 NH C,H, N ( C,,H7)*CO* NHPh. The ethylenedi-a-naphthyldiamine was prepared according to the directions of Bischoff and Nastvogel (Ber., 1890, 23, 2039), and melted at 126'. It was carefully dried and heated for 5 hours with phenylcarbimide in the proportion of one molecule to two molecules. The tubecontents were washed with small quantities of ether and crystallised twice from benzene. As the melting point was still not definite, further recrystallisation from chloroform was resorted to, and after twice treating in this way, small, colourless needles melting at 266O were obtained. The crystals are readily soluble in benzene, less so in chloroform, and dissolved only sparingly in alcohol or ether. On analysis : 0.2534 gave 0.7530 GO, and 0.1486 H,O. C = 81.04 ; H = 6.51. C,9H2,0N, requires C = 80.74 ; H = 5-80 per cent. C'36H,o02N, ,, C=78.54; Hz5.45 ,,USE O F PYRIDINE FOR MOLECULAR WEIGHT DETERMINATIONS. 26 1 The analysis proves the compound formed to be a monocarbanilido- derivative, thus differing from the phenyl- and alphgl-derivatives de- scribed above. Several a t tempts were made, using larger proportions of phenylcarbimide and varying the temperature, but in each instance only the monocarbanilido-derivative was produced. QCKEN’S COLLEGE, GALWAP.

ISSN:0368-1645    

DOI:10.1039/CT9017900258

出版商:RSC    

年代:1901

数据来源: RSC

摘要:

USE OF PYRIDINE FOR MOLECULAR WEIGHT DETERMINATIONS. 26 1 XXIV.-iVote on the use of Pyi-idine foT Moleculai* Weight Detemlinations by the Ebullioscopic Method. By WILLIAM Ross INNES, M.Sc., Ph.D. PYRIDINE has a very remarkable solvent power for both organic and inorganic compounds ; many substances which are insoluble, or only sparingly soluble in the organic solvents in general use, dissolve readily in pyridine. An organic liquid in which so many substances are soluble should be very valuable in the investigation of the physical properties of solutions, as a large number of comparable results can be obtained. For such a purpose, i t is important to know whether the molecular weights of dissolved substances are normal i n this solvent. As is well known, substances containing hydroxyl groups generally give abnormally high molecular weights both by the freezing and boiling point methods, when hydrocarbons such as benzene, naphthalene, and phenanthrene are used as solvents.The molecular weight is found t o increase with the concentration, and is usually nearly normal in dilute solution. Von Laszczynski and von Gorski (Zeit. EZektrocJLenz., 1897, 4, 299) have shown that pyridine solutions of a number of inorganic salts have marked electrical conductivity, and deduce from this that the salts are ionised to a considerable extent. If this is the case, pyridine should give normal molecular weights with organic hydroxy-compounds, as an ionising solvent does not favour the formation of polymerised molecules. Werner (Zeit. anovg. Chenz., 199’7, 15, 1) has determined the mole- cular weight of a number of inorganic salts in pyridine.The concen- trations used were small (about 0.5 to 3percent.). The molecular weights found were constant within the limits of experimental error for the various concentrations and agreed closely with those calculated from the formulz, in some cases being a few units above, in others slightly below the normal. The only substances which gave values much VOL. LXXIX. tr262 INNES: NOTE ON THE USE OF PYRIDINE FOR MOLECULAR below the normal were mercuric iodide and cyanide. Both these sub- stances are ionised to an unusually small extent in aqueous solution. Ttiere is therefore no evidence of ionisation in pyridine to be drawn from the molecular weights of dissolved salts.* Consequently there was a possibility that pyridine might give abnormal results with organic hydroxy-corn pounds.As the author wished to examine in a polymerising solvent, sub- stances which are soluble in pyridine, but insoluble in all the solvents known t o have this action, the following determinations were carried out to test this point. The method described by Jones (Zeit. physikal. Chem., 1899,31,114) was tried; the boiling tube was somewhat smaller than that used by him, but otherwise the apparatus was the same. The boiling point of the pyridine was not of satisfactory constancy. It varied considerably with changes i n the gas presswe, and draughts affected i t very much. This may, co doubt, be ascribed to the direct manner of heating recommended by Jones, and to the absence of a vapour jacket.A determination of the constancy of the boiling point was therefore made in an ordinary Beckmann apparatus, using a platinum cylinder in the way described by Jones. A glass vapour mantle was used, and a plain glass boiling tube without a platinum rod fused into the bottom. It was found that pyridine acted rapidly on cork, leaving a shrivelled, hard, wood-like residue ; the condenser tube was therefore fused directly to the boiling tube and the boiling tube made longer than usual above the condenser. The cork holding the thermometer was covered with tinfoil, Hardly any pyridine vapour came into contact with it, owing to the length of the tube. The platinum cylinder was '7.5 cm. long, 1.9 cm. in diameter, and made conical at the top so as t o fit the thermometer closely.The boiling tube was filled to a height of about 2.5 cm. with glass beads of the usual size, the platinum cylinder slipped down, and very fine beads poured in to occupy about 1 cm. The cylinder was then brought into the middle of the tube by tapping, and some pieces of platinum foil placed in it. The small beads hold the cylinder more firmly than the large ones, and mix the vapour and liquid more thoroughly. The result was very satisfactory. * Joneshas suggested (.%:oit.physikaZ. Chem., 1899,31,114) that the apparent absence of ionisation indicated by osmotic methods in this and similar cases may perhaps be accounted for bycombination of the dissolved substance with the solvent. It is evident, however, that combination with the solvent should lower, not raise, the apparent molecular weight.The number of molecules of dissolved substance would be the same, but the amount of solvent would be smaller than that used in the calculation,WEIGHT DETERMINATIONS BY THE EBULLlOSCOPIC METHOD. 263 The pyridine used was Kahlbaum’s ‘‘ gereinigt ” ; it was dried over solid potassium hydroxide and fractionated, a dephlegmating column being used. The fraction boiling a t 115.2--115*5O mas taken for the experiments. Two series of determinations were made with Kahlbaum’s ‘‘ pure ” pyridine ; the constant obtained agreed perfectly with that given by the much cheaper specimen. A Jena glass Beckmann thermometer with a very small bulb was employed. A constant temperature was only obtained after the liquid had been boiling for several hours.The following readings show the apparent variation of the boiling point due to changes in the thermometer : Liquid began to boil at 11-30, Time. Readiii g. Time. Reading. 11-45 2-45 3-25 2.990 12.15 2-59 3-40 2.990 12.30 2.740 4.0 3.060 2.10 3.835 4-25 3-070 3.0 2.885 4.35 3.070 3.7 2.940 By leaving the liquid boiling all night, a boiling point of satisfactory Time. Reading. Time. Reading. 5.10 P.M. 3.130 10.50 AX 3.272 6-10 ,, 3-1 48 1.5 P.M. 3.270 During both these determinations the atmospheric pressure varied very little. I n the experiments, of which the results are given, the pyridine was left boiling all night, th6 boiling point observed several times during an hour in the morning, and if it did not vary more than 0.01’ the determinations were then carried out.A regulator was used to keep the gas pressure constant. I n calculating the results, 0.2 gram is deducted from the weight of solvent actually taken. Rosenheim and Woge (Zeit. anorg. Chem., 1897, 15, 315) used pyridine to determine the molecular weight of beryllium chloride. They state that, with ordinary German glass, constant results could not be obtained, as the pyridine attacked the glass. They found i t necessary to use Jena glass. No effect of this sort had been noticed in my experiments, Twenty grams of pyridine were heated for sixteen hours in the constancy was obtained. 10.20 A.M. 3.278Wt. of solvent. 15.72 9 , ? 9 9 , 18.59 ?, > ? 9 ) 9 ) 18-40 ?, 9 7 264 INNES: NOTE ON THE USE OF PPRIDINE FOR MOLECULAR boiling tube, fitted up as if for a molecular weight determination.The pyridine was then evaporated t o dryness in a weighed platinum dish. There was no residue, and the weight of the dish was the same to four places of decimals. Determination. of the rnolecukcw vise of boiling point for pyricline : Wt. of subst. 0.2480 0.5970 0.926 1.259 0.1592 0.4236 0.8406 1.450 2-063 0.2282 05278 1.7008 Kise of Grams subst. per Mols./100 per b. p. 100 g. solvcnt. 100 g. solvent. Constant. Benxophenone, C,,H,,O = 183. 0-268O 1.60 0.88 30.5 0.648 3.85 2.12 30.7 0.98s 5.97 3.28 30.1 1.318 8.1 2 4-46 29.6 Phenantlwene, C,,Hlo = 178. 0*140° 0.87 0.49 28.8 0.392 2-30 1.29 30-3 0.740 4.57 2.57 28.8 1,240 7-88 4.43 28.0 1-7-58 11.21 6.30 27.9 Diplwzylarnine, C,,H,,N = 169.0*21fi0 1.23 0.73 29.0 0.536 2.84 1.68 30.9 1.596 9.34 5.53 28.5 ‘The mean of all the determinations is 29.5. The constant has been determined by Werner (loc. cit.) with diphenylamine in concentrations varying from 0.93 to 9.69 per cent. The mean value was 30.1, the numbers varying from 28.6 to 31.5. The constant has also been determined by Rosenheim and Woge (Zoc. cit.), who used diphenyl, triphenylmethane, and ptoluidine as dissolved substances in concentrations of from 1.7 to 3.7 per cent. The mean was 30.7. The numbers obtained for the constant were very concordant, the greatest difference being only 1.3, but the determina- tions were nearly all carried out at about the same concentration, and the greatest differences of concentration were small. Werner Calculated the constant from the Trouton-Schiff rule C = 0-00096Tm = 29.5 This method only gives an approximate value in most cases, The constant may be calculated with considerableWEIGHT DETERXINATIONS BY THE EBULLIOXCOPIC METHOD.265 accurltcy by van't Hoff's formula C=- , which has been shown to apply to the boiling point by Arrhenius (Zeit. physikal. Chenz., 1889, 4, 550). The heat of vaporisation of pyridine has been determined by Louguinine (Compt. rend., 1899, 128, 366), who found i t to be 101.39 at 115.5". 0.0198P Introducing these values, we get 0.0198 x (388.5)z 101.39 C = ~ _ _ _ _ =29*47 ; 29.5 is therefore taken as the const,ant for pyridine. Molecuular weight of hydroxy-compounds in pyridine : Wt. of solvent, 17-17 ?7 ?7 $ 7 19.60 I > ,# 18.26 9 , 9 9 3 , 16.89 2 9 ?) Wt.of Rise of Grams subst. per Mols./100 per subst. b. p. 100 g. solvent. 100 g. solvent. S ~ ~ c i ~ a i c G C C ~ ~ , C,H,(CO,H>, = 11 8. 0,2104 0.278' 1.18 1 *oo 0.4316 0.575 2-43 2-06 0.8524 1.215 4.80 4.06 1.492 2.205 8.40 7.1 1 Tartaric acid, C2H,(OH),(C0,H), = 150. 0.3542 0.335O 1.81 1-20 1.1530 1.065 5.88 3.92 2,372 2.175 12.10 8.07 Salicylic acid, CGH,(OH)*CO,H = 138. 0.1782 0*18So 0.98 0.71 0.5021 0.615 2 ~ 7 5 1.99 1.1978 1.390 6.56 4-75 2.193 2.630 12.00 8.70 Resorcinol, C,H,(OH), = 110. 0.0968 0.1 4 5 O 0.57 0.53 0.3050 0,497 1 *so 1 *64 0.6968 1.153 4.1 7 3.99 hlol. wt. 127 126 117 113 160 164 165 123 133 140 136 124 107 105 Tartaric ditoluidide, C,H2(0H),(C0 *NH* C,H,* CH,), = 328. 16.99 0.3094 0.110' 1.23 0.38 331 9 , 0.4112 0.218 2.42 0.74 330 Y ? 0-6234 0.308 3.67 1.12 354 VOL. LXXIX. U266 FRANKLAND, WHARTON, AND ASTON: THE AMIDE, @-Naphthol, C,,H7* OH = 144. Wt. of Wt. of Rise of Grams subst. per Mols./100 per solvent, subst. b. p. 100 g. solvent. 100 g. solvent. Mol. wt. 1854 0.1'734 0.176O 0.93 0.65 158 9 9 0.4284 0.465 2.31 1.60 148 f ? 0.9694 1.060 5.23 3.63 147 I? 1.342 1.470 7.22 5-01 146 I 9 1.798 2-000 9.68 6.72 144 The molecular weights found are in every case near the normal, and there is no regular increase with concentration. Pyridine may therefore be used to find the normal molecular weight of hydroxy- compounds. BIRMINGHAM UNIVERSITY. CHEMICAL LABORATORIES.

ISSN:0368-1645    

DOI:10.1039/CT9017900261

出版商:RSC    

年代:1901

数据来源: RSC

摘要:

266 FRANKLAND, WHARTON, AND ASTON: THE AMIDE, XXV.-The Amide, Anilide, and 0- and p- Tohidides of Glyceric Acid. By PERCY FARADAY FRANKLAND, FREDERICK MALCOLM WHABTON, and HENRY ASTON. IN pursuing the systematic study of the effect onrotatorypower produced by substitution in optically active compounds, we have recently been directing our attention to active amides and their substitution products. The information at present available concerning the rotatory power of compounds of this class is summarised in the following table. In this table only such amides have been included as are derived from acids of which the rotatory power of several esters is also known, and the rotations of these esters are also given so that the relative rotatory effects of the amino- and alkyloxy-groups are rendered apparent.It will be seen that the existing material for such comparison is very scanty, being practically limited to the derivatives of malic, tartaric, and mandelic acids. Mcmdelic acid (Walden, Zeit. physikd. Chem., 1895, 17, '706) C. r .I?. [MID. Free acid Acetone.. . . . . 2.50 - 148*0° - 225.0' Methyl ester ,, 3.33 - 110.2 - 182.9 Ethyl I ? 9 ) 5.81 - 90.6 - 168.1 Amide ? ? 2.60 - 66*6 - 100.6 7 ) ? ? 1 *50 - 66'7 - 100.7 ?? >? 9 , lb16 - 87.1 - 156.8ANILIDE, AND 0- AND P-TOLUIDIDES OF GLPCERIC ACID. 267 Ma& acid (Walden, ibid., 24s). Dimethyl ester Liquid Diethyl ?, ?, Diisopropyl ,, 7, Diisobutyl ,, ,? Diarnide Water Dipropy1 ,9 ,? I7 7, Dianilide Glacial acetic acid Di-o-toluidide 7 ) Di-p-toluidide 9 9 /3-Naphthimide ,, A d * Water Y ? ? I 9 ) 7, C.- 4-32 8.65 1.50 0.75 2-00 1 -00 1.00 1 *oo 0.2258 [ 4y. - 6-85' - 10.18 - 11.62 - 10.41 - 11.14 - 37.6 - 38.0 - 60.66 - 58.66 - 65.0 - 66.5 [ ~ ] g - 70.0 [a]:- 51.5 [.IJj - 33.95 [ Af]D* - 1 1-l0 - 19.3 - 25.3 - 22.7 - 27.4 - 49.6 - 50.2 - 172.3 - 166.6 - 203.8 - 207.5 - 218.4 - 124.1 - 59.4 d- Yurt aric acid. C. [ a goo. 1 M ID. Dimethyl ester Liquid - + 3-14" + 3.8' ? ? + 7-66 + 15.8 Diethyl ,, - Dipropy1 9 7 9 , - + 12.44 + 29.1 Diisopropyl ,, 9 9 - + 14.89 + 34.8 Diamide Water - [a]j+133*9 [M]j+198.2 [ a l D * [MID. Methyltartrimide 4" Water 7.31 + 194.2 + 281.6 9 , ? P 12.94 + 192.6 + 279.3 Ethyltnrtrimide T ,, 5-76 + 164.9 + 262.2 9 , 7, 8.57 + 168.2 + 264.3 A glance at the table is sufficient to show that the rotatory effect of the NH, group is qualitatively similar to that of the alkyloxy- (OA) group, but quantitatively greater.Thus, in the case of man- delic acid, the replacement of OH by OA has the effect of dimin- ishing the lzvorotation (the ethoxy- having a greater effect in this respect than the methoxy-group), the replacement by NH, leading to a st,ill greater depression of the latter. I n the case of malic acid, the alkyloxy-groups successively raise the * Bisch off and Walden, Stcreochemie, 855. t Ladenburg, Ber., 1896, 29, 2710. u 2268 PRANKLAND, WHARTON, AND ASTON: THE AMlDE, laevorotation, the latter being very greatly further elevated by the replacement of OA by NH,. Again, in the case of d-tartaric acid, the effect is exactly similar, only that here it is the dextrorotation which is raised by these sub- stit utions.On substituting the hydrogen in the amino-group by phenyl or tolyl, it is seen that the rotatory effect of the group is still further increased. Again, if the amino-group becomes an imino-group leading to the formation of a ring compound, the rotatory effect is also in- creased, as in the cases of malic P-aaphthimide and of methyl and ethyl tar trimides. The present paper deals with the preparation and properties of the amide, anilide, o-toluidide, and p-toluidide of glyceric acid. With the exception of the o-toluidide, these compounds have had their rotation determined in the liquid state, whilst the rotation of all of them has also been taken in methyl alcohol solution. So far as we are aware, these are the only amides which have hitherto had their rotation determined in the fused or liquid condition.The rotations of these new compounds should be compared with those of the esters of d-glyceric acid given below (P. Frankland and Price, Trans., 1897, 71, 270) : Methyl glycerate ......... [a]: - 4-80' Ethyl ,, ......... - 9.18 Propyl ,, ......... - 12.94 isoPropy 1 ,, ......... - 11.82 Butyl , , ......... - 13.19 isoButyl ,, ......... - 14.23 Octyl ), ......... - 10.23 Amy1 ,, ......... - 14.12 Heptyl , , ......... - 11.30 [M]r - 5.76' - 12.30 - 19.15 - 17.49 - 21.37 - 23.05 - 24.85 - 23.05 - 22.28 The rotation of glycerylamide is much greater, [ MILw0 = - 41*98O, of glycerylanilide still greater, [ M]bwo = - 72*36O, whilst that of glyceryl-p-toluidide is intermediate between that of the two latter, [ Mlg,8" = - 67.65'.The lxvorotation of the o-toluidide in the fused state is doubtless lower than that of the para-compound, as in methyl alcohol solution it is very much lower than that of the latter ; indeed, it is about the same as that of glycerylamide in the same solvent at the same concentration. The laevorotation of the amide, anilide, and p-toluidide is very con- siderably greater in methyl alcohol solution than in the liquid state ; the o-toluidide, as already mentioned, could only be optically examined in solution, but it may safely be assumed that its rotation is simi- larly affected by this solvent.ANILIDE, AND 0- AND P-TOLUIDIDES OF GLYCERIC ACID. 269 It is further noteworthy that the rotation of the amide, anilide, and p-toluidide in each case diminishes with rise of temperature; in methyl alcohol solution, increase of concentration diminishes the rota- tion of the amide and anilide, whilst i t increases that of the o-toluidide.The rotation of the p-toluidide is almost independent of the concentra- tion within the limits examined, there being, however, a very slight diminution with increase of concentration, The corresponding compounds of inactive glyceric acid were also prepared, and their melting points compared with those of the active substances : Me1 ting point. Active. Inactive. Glycerylamide.. . . . . , . . 99 *5-100° 915- 92" Glycerylanilide . . . . . . 11 3 -1 13.5 91 Glyceryl-o-toluidide.. . 89 - 89.5 129 -129.5 Glyceryl-p-toluidide .. 131 -131.5 120 -120.5 Thus, in all cases, excepting that of the o-toluidide, the melting point of the active is higher than that of the corresponding inactive compound.We are continuing the investigation of the effect produced by these and similar subs ti t u t ion s. EXPERIMENTAL. 1. Glycerykamide (Inactive). Dry ammonia was passed into inactive ethyl glycerate surrounded with a freezing mixture of ice and salt. The saturated liquid was allowed to stand for 24 hours, the excess of ammonia and alcohol being then taken off under reduced pressure at the ordinary tem- perature of the air. When nearly the whole of the ammonia had been removed, the residue crystallised out. It was recrystallised from methyl alcohol ; the prismatic crystals melt a t 91.5-92O. 0.0411 gave 0.0054 nitrogen.N = 13.14. C3H703N requires N = 13 *33 per cent Gkycerylarnide (Active). This was prepared in exactly the same way from active ethyl glycer- 0.0402 gave 0.0052545 nitrogen. ate.* It forms large plates or prisms melting a t 99*5--100°. N = 13-07. 0.0409 ,, 0.005397 ,, N = 13.20. C3H703N requires N = 13.33 per cent. d 13Oo/4O= 1.3099. d 140°/40= 1.3016. d l5Oo/4O= 1.2932. * QD= - 11'35" ; Z=1.270 FRANKLAND, WHARTON, AND ASTON: THE AMIDE, Rotation of glycerylarnide. Temp. Length of tube. d t0/4". a,. r a ID. [M 100' 44mm. 1.3347 - 23.47' - 39-98' - 41.98' 109.5 1.3268 - 23.06 - 39.50 - 41.48 136 1.3048 - 21.88 - 38.11 - 40.02 182 1.2666 - 19.91 - 35.73" - 37.52 9 ? 9 , ?, [20° calculated from rotations a t 100' and 136' above - 44.14 - 46.35 3 11. Glycerylanilide (Inactive).Ten grams of inactive ethyl glycerate were heated with 8 grams of aniline (7 grams is the theoretical quantity required) for 4 hours at 140'. The temperature was then gradually raised to 160' and main- tained there for 4 hours more. On cooling, the mixture solidified to a brown, pasty mass. The colour was removed by boiling in alcoholic solution with animal charcoal, and crystallisation was then effected from chloroform, small plates melting a t 91' being obtained. 0,1745 gave 12.2 C.C. moist nitrogen at 12' and 729 N-7.95. 0.1584 ,, 11.0 C.C. ,, ,, 12.5 ,, 725.5mm. N=7% CQH1lOSN requires N = 7.73 per cent. mm. Glycevylanilide (Active). This substance was prepared from active ethyl glycerate as described above for the inactive compound. It was first crystallised from alcohol, being obtained in long, flat blades. On subsequently crystal- lising from chloroform, in which it is but slightly soluble, it was ob- tained in small needles.On melting, these yielded a dark liquid, which could not be satisfactorily examined in the polarimeter ; it was therefore boiled with animal charcoal in alcoholic solution, and finally crystallised from this solvent. The crystals melted at 113-1 13*5O, and yielded a clear and colourless liquid. 0.2288 gave 15.6 C.C. moist nitrogen a t 14Oand 736 mm. N = 7.77. 0,1993 ,, 13-2c.c. ,, I9 12' ,, 751.4mm. N=7.77. C9H1,0,N requires N = 7.73 per cent. d 12oo/4O= 1.1914. d 14Oo/4O= 1.1'744. * This value is somewhat doubtful, as the glycerylamide underwent partial de. composition during the polarimetric measurements at this temperature,ANILIDE, AND 0- AND P-TOLUIDIDES OF GLYCERIC ACID.271. Rotation of glycerykcmikide. Temp. Length of tube d to/4". a,. [a],. 100' 49.85 mm. 1.2084 - 24.09' - 39.98' -72.36' 109 9 9 1.2008 - 23.42 - 39.12 -70.81 139 ? I 1.1752 - 21.19 - 36.16 - 65.45 180 $ 9 1.1404 - 18-15 - 31.93 - 57-79 [20' calculated from rotations at 100' and 139' above. - 47.82 - S6.553 111. Gly cevy I-0- t oluidids (Inactive). Ten grams of inactive methyl glycerate were heated with 15 grams of o-toluidinefor 4 hours at 140°, the temperature being then raised slowly to 160' ; finally the excess of toluidine was distilled off under 14 mm. pressure, On cooling, the contents of the flask nearly solidified to a dark brown mass, On crystallising from benzene, the substance was still much coloured, and it was therefore boiled with animal charcoal in alcoholic solution, It was then several times recrystallised from benzene until of constant melting point, and finally from chloro- form by which the melting point was not altered.The crystals are colourless, have a somewhat silky appearance, and melt a t 129-129.5O. The liquid obtained by melting the crystals does not colour in con- tact with air, and in this respect differs from the active compound (see below). 0*2000 gave 0.4499 CO, and 0.1219 H,O. 0.1536 ,, C = 61.35 ; H = 6-77, 9.4 C.C. moist nitrogen a t 145Oand 757.4 mm. N= 7.15. C1,H,,O,N requires C: = 61.54 ; H = 6.66 ; N = 7-18 per cent. Glyceyy I-o-to Zuidide (Active), This was prepared similarly to the inactive substance, 15 grams of active methyl glycerate * being heated with 22.5 grams of o-toluidine.Although the colour of the product was removed by boiling in alco- holic solution with animal charcoal, it reappeared during the exposure t o air necessitated by the subsequent filtration. Thus pink crystals were obtained from benzene, but finally an almost perfectly colourless separation was obtained by using chloroform as the solvent and wash- ing the crystals with ice-cold chloroform, The crystals had a constant melting point of 89-89.5'. On melting these crystals, however, the liquid obtained was of such a deep-red colour, when viewed by trans- mitted light, that it could not be used for polarimetric observation. The rotation could therefore only be determined in solution, and this * a,= -6.20"; Z=1,272 FRANKLAND, WHARTON, AND ASTON : THE AMIDE, necessitated determining the rotatory power of all the other substances referred to in this paper in solution also.The values for the rotatory powers in solution are given later (see p. 273). 0.2002 gave 0.4509 CO, and 0.1213 H,O. C=61*42 ; H=6*73. 0.2004 ), 0.4515 CO, ,, 0.1219 H,O. C = 61.44 ; H= 6.75. 0.2030 ,) C,,H,,O,N requires C = 61.54 ; H = 6.66 ; N = 7.18 per cent. 12.6 C.C. moist nitrogen at 12' and 746.4 mm. N = '7.23. IV. Glyceryl-p-toluidide (Inactive). This was prepared from inactive ethyl glycerate and ptoluidine (m. p. 44.5'). After heating together as described in the case of the o-toluidide, the mixture became semi-solid on cooling. The brown mass was treated with benzene to extract the excess of base, and the residue then crystallised from chloroform, in which it is only very slightly soluble.Crystallisation was repeated from this solvent until the melting point was constant at 120-120.5'. 0.2009 gave 12.9 C.C. of moist nitrogen a t 14' and 734 mm. N = 7.29. 0,2013 ,, 12.6 ,, 9 9 ,, 13' ,, 744.5 mm. N= 7.24. C1,H,,O,N requires N = 7.18 per cent, Glyceryt-p-toluidide (Active). This was similarly prepared and purified by cry stallisation from The constant melting point was 131-131.5'. 0*1001 gave 6.5 C.C. moist nitrogen at 23Oand 754.8 mm. N = 7.27. 0.2009 ,) 12.9 C.C. ,, ), 14' ,, 734 mm. N=7*28. chloroform. C,oH,,03N requires N = 7.18 per cent. d 140°/4' = 1.1735. d 150'/4'= 1.1643. Rotation of Glyceryl-p-toluidide.Temp. Length of tube. d tQ/4". a,. [ale. 136 9 ) 1.1772 - 19.03 -32.38 109 9 , 1.2020 - 20.41 - 34.02 98 l ? 1.2121 - 20.99 - 34.69 [20° calculated from rotations at 98' and 136' above - 39.43 179" 49.92 mm. 1.1376 - 16-82' -29.62' [ M ID. - 57.76 - 63.14 - 66.34 - 67.65 - 76-89] V. Rotation of Glycerylanilide and of Glycevyl-o- and -p-toluidide in Methyl A lcohol Solution. Owing to the impossibility of takiug the rotation of the fused glyceryl-o-toluidide, it was determined in solution, the rotation of theANILIDE, AND 0- AND P-TOLUIDIDES OF GLYCERIC ACID. 273 Weight of solution, grams. p-compound and of the glycerylanilide and amide being also deter- mined under the same conditions for the purpose of comparison. The glycerylanilide is only soluble in methyl alcohol to the extent of a little more than 6 per cent., and the glyceryl-p-toluidide t o the extent of 6.61 per cent., so the rotations were determined in solutions of about 2.5, 5, and 6 per cent.concentration respectively. I n benzene, glycerylanilide is only soluble to the extent of 0.15 per cent. Glycerylamide is even less soluble in methyl alcohol than the other compounds, and hence could only be optically examined in the two lower concentrations. Weight of substance, grams. Rotation in methyl alcohol solution at 20° ( I = 100.45 rnm.). Glycerylamide. * 4'7619 1 0'8122 1 -2'40 2'4390 0.8022 - 1'24" - 63-09' -61.77 8.1877 7'7469 Grams sub- 1 I I 0.1997 0,3689 stance in 100 grams 20"/4"* solution. 1 j a2Do* 1 10*2500 lO?iOOO 10'6000 0.2500 0*5000 0*6000 I 2.4390 4'7619 5.6604 0.8035 - 1'42" - 72.13" 0.8118 -2.70 -69.53 0.8155 -3.12 -67*29 [ M ]:On.- 66'24" - 64.86 2.4390 4'7619 5'6604 [Rotation in liquid state (calculated) [ a Jioo = - 44*14", [MI","" = - 46-35"]. 0.8039 - 1-26' - 63.97" 0%108 -2.49 -63'95 0.8140 - 2.95 -63.74 10.2500 10*5000 10.6000 0.2500 0-5000 0-6000 - 130-55" - 125'85 - 121.79 2'4390 0'8025 - 0.64" 4.7619 0.8117 -1.37 5'6604 0'8137 -1.72 [Rotation in liquid state (calculated) [a]:on = - 47-82', [ M ]ioo = - 8655'1. - 32-55' -35.41 -37'18 G Zy cer y 2- p- t oluidide. 10.2500 10'5000 10'6000 0.2500 0'5000 0'6000 - 124.74" - 124'70 - 124'29 [Rotation in liquid state (calculated) = - 39-43", = - 76-89']. - 63.47" - 69 '05 - 72'50 * A G per cent. solution could not be prepared, owing to the limited solubility.274 CHATTAWAY AND ORTON: THE PREPARATION OF V1. Molecular Volumes. I n the following table, the experimental molecular volumes are compared with the values calculated by Traube's formula : Molecular volume at 15". Calculated. Experimental. ~- 105 - 74.7 G1 y cerylamide . . . . . . . , . . . . 85.1 1.4053 Glycerylanilide , , , , . , . , . 143.7 181 1-2806 - = 141.3 195 1.2885 Glyceryl-p-toluidide . . . 159.8 - = 151.3 Constants used : C = 9.9 ; H= 3.1 ; 0 (in CO) = 5.5 ; 0 (in OH) = 0.4 ; IS"'= 1.5. Co-volume = 25-9 ; deduction for benzene ring = 13.8. Thus in the case of all three compounds the molecular volume points to association, which is greatest in that of the amide and least in that of the anilide. I n carrying out this investigation we have been assisted by funds placed at the disposal of one of us by the Government Grant Com- mittee of the Royal Society. THE UNIVERSITY, BIRMINGHAM.

ISSN:0368-1645    

DOI:10.1039/CT9017900266

出版商:RSC    

年代:1901

数据来源: RSC

摘要:

274 CHATTAWAY AND ORTON: THE PREPARATION OF XXV1.-The Preparation of Acet;ylchEoraminobenxene and some relcrted Compounds. By F. D, CHATTAWAY and K. J. P. ORTON. THE criticism to which Armstrong (Trans., 1900, 77, 1047) bas recently subjected our work makes it evident that several points need fuller consideration. In a reference which he makes to the omission in our earlier papers of all mention of Slossen’s work (Ber., 1895, 28, 3265), he fails to remark that we have twice expressed our regret (Proc., 1900, 16, 2, and Trans., 1900,77,134) that at the time these papers were published we had unfortunately overlooked the communication and were con- sequently in entire ignorance of it. Whilst very gladly acknowledging Slossen’s contribution to the subject, we may perhaps again repeat that our observations were made quite independently.Armstrong also notes in our papers the absence of all reference toACETYLCHLORAMINOBENZENE. 275 his views upon substitution, I n his first communication to the Society on the subject (Trans., 1887,51, 258), he-advocates the view originally put forward by Kekul6 that in all cases of substitution additive com- pounds are first formed by the union of the interacting substances. He endeavours to develop a general theory, and writes : ‘( Passing now to the consideration of the formation of di-derivatives, I may a t once point out that the difference under discussion in the behaviour of the two classes of mono-derivatives * is, in my opinion, attributable to the different manner in which they form additive compounds.I n the case of those which furnish the ortho- and para-derivative, it may be held that the additive compound results from the union of the reacting molecule with the carbon atom to which the radicle R’ is attached” -and again : L L . , . . my opinion being that the additive compound which generates the meta-derivative is formed by the combination of the reacting molecule with the radicle R’ of the mono-derivative, and not with the carbon atom which carries that radicle.” ? Forster Morley, in reply (Trans., 1887, 51, 579 ; Proc., 1887, 3, 61), directed Armstrong’s attention to the case of the anilines, and instanc- ing their behaviour on nitration, insisted that his assumptions as to the particular atoms where addition takes place were opposed to known facts and ‘( in the very highest degree improbable.” $ After considering these objections, Armstrong, although he does not in any may withdraw his previously quoted generalisations, appears to have modified his views, for in his answer (Trans., 1887, 51,583) he states : “ The behaviour of amido- and hydroxy-derivatives, which is in many respects peculiar, deserves minute study from the above point of view, several highly suggestive observations having been recorded, from which it may be inferred that the phenomena of * The formation of para- and ortho- and of meta-derivatives respectively.t Armstrong earlier in the paper refers to the facts that potassium phenyl sul- phate on heating becomes potassium phenol-p-sulphonate, and that the nitrosamine of methylaniline is converted by digestion with an alcoholic solution of hydrogen chloride into p-nitrosomethylaniline, not as in any way throwing doubt on his theory, but as showing that where para-compounds are produced it is not necessary t o assume that the reacting molecule attaches itself to the carbon atom occupying the para-position relatively to the one to which the radicle R’ is attached.He also refers to the transforniations of sodium and potassium phenyl carbonates as instances, among others, showing that the proportion in which ortho- and para-derivatives are formed varies with the nature of the transforming compounds, $ Forster Morley pointed out that, if Armstrong’s views were correot, compounds containing the NH, group, which can combine with nitric acid, ought to produce nieta-derivatives, whilst the chief product when aniline, for example, is nitrated is p-nitroaniline.He argued that when aniline yields a meta-derivative on nitration in presence of a large excess of sulphuric acid, it is the circumstance that the nitric acid, HNO,, cannot then combine with the NH, group which causes the formation of the meta-compound. This explanation, which is now generally accepted is the exact opposite of Armstrong’s view.276 CHATTAWAY AND ORTON: THE PREPARATION OF substitution are less simple than is perhaps commonly supposed, and that isomeric change is by no means of infrequent occurrence in the formation of derivatives of amines and phenols. ” He then mentions several well-known cases of isomeric change, among them the trans- formation of acetylchloraminobenzene, of the nitrosamine of methyl- aniline and of potassium phenyl sulphate, and adds : ‘( These examples, although few, are perhaps numerous enough to justify the inquiry whether the formation of para-derivatives of primary or secondary amines and of phenols is not always preceded by that of an isomeric compound formed by the displacement of the aminic or hydroxylic hydrogen. Whether ortho-derivatives are formed in a similar manner remains to be aecertained .. . .” Since 1887,* much work throwing great light on the mechanism of substitution has been done. For example, Bamberger (Ber., 1894, 27, 359; 1895, 28, 399; 1897, 30, 1248, 2247) has shown that nitro- amino-derivatives of benzene are transformed readily into o- and p-nitroanilines, and that phenylsulphamic acid easily passes into o- and p-sulphanilic acids.Hantzsch (Ber., 189’7, 30, 2334) also has shown that in brominated diazonium chlorides the chlorine can ex- change places with bromine atoms in those positions. These facts clearly demonstrate that the NO, group and halogen atoms when associated with a nitrogen atom can readily interchange with hydrogen atoms in the nucleus in ortho- and para-positions relatively to the carbon atom t o which the nitrogen atom is attached, and in these only. It had been shown much earlier that CH, and C,H, groups could wander from the nitrogen of secondary anilines only into these positions in the ring (Hofmann, Ber., 1874, 7, 5 2 6 ; Limpach, Ber., 1888, 21, 641). In 1899 we submitted to the Chemical Society a paper in which our views on the substitution of chlorine and bromine in anilines and anilides were clearly staked, accompanied by the experimental evidence on which we based them.This paper was read on June lst, and an abstract of i t containing all our experimental results, which admitted but of one explanation, was printed in the Proceedings of that date (Proc., 15, 152). In the next number of the Proceedings, containing an account of the meeting of June 15th (Proc., 15, 176), Armstrong published a short note containing an unreserved withdrawal of his previous opinions, and a substitution of views in the main identical with those to which our work had led us, all the important facts of Bender in 1886 (Bey., 19, 2272), a year before the appearance of Armstrong’s first paper, showed that the chlorine in acetylchloraminobenzene easily passes into the para-position in the nucleus, and this observation was extended in 1890 by Comstock and Kleeberg (Amer.Chem. J., 12, 500), who proved that the iodine atom in formyliodaminobenzene behaves similarly.ACETYLCHLORAMINOBENZENE. 277 this work having been published a fortnight previously. Armstrong’s more complete communication to the British Association was made two months later. I n neither place does he refer to our work, although his changed views are as direct a consequence of it as of the work of Bamberger, of Hantzsch, and of others. We have never in any way claimed originality for the view that substi- tution is probably preceded by the formation of an isomeride which subsequently undergoes isomeric change ; this directly follows from much earlier work.What we do claim is that in the particular case of the substitution of halogen in anilines and anilides we have con- verted a probability into a certainty by showing that all the compounds ordinarily formed by substitution can be produced by isomeric change from previously formed nitrogen chlorides and bromides, and that the degree of difficulty with which transformation is effected corresponds with that experienced in effecting apparently direct substitution. EXPERIMENTAL. Acetylchloramirzobenxerne. Armstrong (Trans., 1900, 77, 1051), referring to our method of preparing acetylchloraminobenzene from acetanilide, states : ‘(Accord- ing t o my experience, whatever excess of hypochlorite be taken, and however long the stirring be continued, the chlorination is almost always imperfect, and eventually at most 60-70 per cent.of p-chlor- acetanilide has been obtained.” We have in consequence reinvestigated our original method, namely, the action of a solution of bleaching powder on acetanilide suspended in excess of potassium bicarbonate solution. A mixture of finely powdered acetanilide and potassium bicarbonate was added to an excess of a O*3Nsolution of bleaching powder. After standing for some hours at the ordinary temperature, the solid was filtered off and thoroughly extracted with chloroform, in which both acetanilide and the nitrogen chloride are soluble. After the solution had been dried by calcium chloride, the solvent was completely evapor- ated away in a vacuum.An analysis, made by titrating the iodine set free from acidified potassium iodide by a weighed quantity of the solid residue, showed it to be pure acetylchloraminobenzene. 0.1 126 liberated I = 13.25 C.C. N/lO iodine. C,H,*NCl*CO*CH, requires C1 as : NCl = 2009 per cent. I n our later work (Trans,, 1900, 77, 789, &c.), we have used a solu- tion of potassium hypochlorite containing potassium bicarbonate * for * Prepared by mixing solutions of potassium bicarbonate and bleaching powder and filtering off the calcium carbonate. C1 as : NC1= 20236.278 CHATTAWPAY AND ORTON : THX PREPARATION OF the preparation of chloramino-derivatives, when it was necessary to avoid the presence of acids. With this agent, pure acetylchloramino- benzene is very easily and rapidly prepared in any quantity.Twenty gkams of acetanilide were added to 300 C.C. of a soliltion of potassium hypochlorite (0-3 to 0.4 N ) containing potassium bicarbonate. After standing for half an hour at the ordinary temperature of the laboratory, some of the solid was filtered off, mashed with water, and dried in a vacuum over sulphuric acid. The product was well crystal- lised in small plates. 0,1655 liberated I = 18.45 C.C. NllOiodine. C1 as:NCl= 19.8 per cent. Further portions of the solid were filtered off at intervals, dried, and analysed : After 3 hours the product contained C1 as : NC1= 20.26 per cent. ?, 5 ¶ 9 1 , ,, C1 as : NC1=20*81 ,? ,, It is seen, therefore, that in half an hour more than 95 per cent.of the acetanilide becomes converted into the chloramino-derivative, and that the conversion is finally absolutely complete. We do not find it an advantage to work at the higher temperature recommended by Armstrong. It must be carefully noted that a too concentrated solution of bleach- ing powder cannot be used, If potassium bicarbonate is added to a solution of bleaching powder above 0.5 N, chlorine is always set free, no matter what excess of bicarbonate is present. Under such circumstances, we find that pchloroacetanilide is invariably formed. If a sufficient excess of bleaching powder is employed, the whole of the acetanilide will become converted into p-chloroacetanilide, and this, in its turn, into its chloramino-derivative. Five grams of acetanilide were mixed with a considerable quantity of potassium bicarbonate and added to a large excess of a normal solu- tion of bleaching powder.Free chlorine was liberated, and was unmistakably present in the carbon dioxide evolved. After standing for 3 days, the solid was filtered off and extracted with chloroform. The residue left, after evaporating away the chloroform, was pure acetyl- chloramino-pchlorobenzene. 0.2395 liberated I = 23.5 C.C. N/10 iodine. C6H,C1*NC1*C0*CH, requires C1 as : NC1= 17.38. We have never observed the formation of p-chloroacetanilide when no chlorine is developed ; moreover, acetylchloraminobenzene has been kept completely unchanged for a week under dilute hypochlorite solution, and apparently may be so kept for a long period.Acetylchloraminobenzene, from whatever solvent it separates, C1 as : NCl= 17.39.ACETYLCHLORAMINOBENZENE. 279 crystallises, according t o our observations, in plates or short, four- sided prisms, When needles are present, they are undoubtedly the isomeric p-chloroacetanilide, and can be shown not to be the nitrogen chloride by the following method, The crystals are placed on a slide under a microscope and a few drops of a strong solution of potassium iodide added. The needles of anilide remain untouched, whilst the plates and prisms of the chloramino-compound vigorously react with the iodide, and melt to an oil which afterwards solidifies. This substance is readily obtained in large plates melting at 92' from a mixture of chloroform and petroleum (b. p.50--80°), which we still believe to be the best solvent. It dissolves easily in cold glacial acetic acid, and by adding water may be reprecipitated in a pure state as small plates or prisms. When it is dissolved in hot dilute acetic acid, needles of p-chloroacet- anilide frequently separate, as we have noted.* We have, however, also recrystallised i t from dilute acetic acid, although we do not regard this as a suitable solvent. Armstrong has taken exception to our statement concerning the general reaction with alcohol of substances containing the nitrogen halogen linking, We were not referring in this statement to the transfownation which acetylchloraminobenzene undergoes under cer- tain circumstances in alcoholic solution, This transformation had already been observed by Bender.Owing to the readiness with which all such substances react with alcohol, and in some cases undergo transformation when dissolved in it, we cannot recommend it as a solvent. Some other observations of Armstrong will be more appropriately considered at another time. Benxo ylch Zoraminobenxene. As a confirmation of our previous observations, we have again pre- pared this substance in order to show that benzanilide also is com- pletely transformed into its chloramino-derivative by a dilute solution of potassium hypochlorite in the presence of bicarbonate. The reaction is very much slower than in the case of acetanilide. After 3 days standing, Some of the solid was filtered off, dried, and analysed, and shown to be pure benzoylchloraminobenzene. 0.1474 liberated I= 12-75 C.C.N/10 iodine. C1 as : NCI = 15.33. C6H,*NCl*CO*C,H, requires C1 as : NCI = 15.33 per cent. After another week's standing, it was found that the substance had * Bender describes i t as separating i n needles from this solvent. He does not give the melting point of these needles ; most probably they werep-chloroacetanilide, which invariably crystalliees in this form.280 PATTERSON AND DICKINSON: THE PREPARATION OF undergone no change, and on analysis it again gave C1 as : NCl == 15.33 per cent, Acetylchloramino-2 : 4-dicldorobenxene, Hitherto this substance has been obtained by Witt, Jackson and Wing,* and us, by the action of bleaching powder on an acetic acid solution of acetanilide or 2 : 4-dichloroacetanilide. From this method of preparation it might. be inferred that it could only be formed through. the agency of hypochlorous acid. We find, however, that it can very easily be prepared by the direct action of chlorine. Acetanilide (1 part) is dissolved in glacial acetic acid (4 parts) and excess of anhydrous sodium acetate is added. Dry chlorine gas is then led into the cooled solution until no more is absorbed. On adding water, the nitrogen chloride separates as an oil which rapidly solidifies. It can be obtained pure by one crystallisation from petroleum, In the absence of sodium acetate, although some acetylchloramino-2 : 4- dichlorobenzene is always formed, the conversion of the 2 : 4-dichloro- acetanilide into the nitrogen chloride is not perfect, owing to the liberation in the reaction of free hydrochloric acid. CHEMICAL LABORATORY, Sr. BARTHOLOMEW’S HOSPITAL AND COLLEGE, E. C.

ISSN:0368-1645    

DOI:10.1039/CT9017900274

出版商:RSC    

年代:1901

数据来源: RSC

摘要:

280 PATTERSON AND DICKINSON: THE PREPARATION OF XXVIL-The Prepuration of Estersfvow8 other Esters of the same Acid. By T. S. PATTERSON and CYRIL DICEINSON. SOME time ago we commenced some experiments with the object of preparing methyl ethyl tartrate, a comparison of the rotation of which with that of methyl and of ethyl tartrate could not fail to be interest- ing. In this work we were encouraged by the hope that methyl ethyl tartrate would prove to be a solid at the ordinary temperature, in which case its purification would present no great difficulty. Our first attempt to prepare it was made by boiling tartaric acid with ethyl alcohol for some hours, then removing the excess of alcohol first by distillation and afterwards by the evaporation of the residue in a vacuum over sulphuric acid.The very viscous substance thus obtained (83 grams) was then dissolved in excess of methyl alcohol (100 grams) and the solution saturated with hydrogen chloride. After removal of the * This substance, as already noticed by us, was not obtained by these chemists in a pure state, but was regarded by them as a liquid additive product with hypo- chlorous acid.ESTERS FROM OTHER ESTERS OF THE SAME ACID. 281 alcohol and hydrogen chloride and distillation of the residue under low pressure, the ester obtained had a rotation of + 5.72' (I= 2, t = 17'), that is to say, consisted of nearly pure methyl tartrate. The method of preparation was therefore modified. One hundred grams of tartaric acid, after being dried and powdered, were heated in a flask under a reflux condenser with 25 grams of methyl alcohol (the calculated quantity being 22 grams) on a water-bath for about 8 hours.The acid ultimately went into solution. The syrup after cooling was then saturated with hydrogen chloride, after which 33 grams (calc. 31 grams) of ethyl alcohol were added and hydrogen chloride again passed into the solution to saturation and the mixture allowed to stand over- night. The ester obtained after removal of the excess of alcohol and hydrogen chloride and distillation of the residue under reduced pres- sure was found to boil at 144-152' (bath 190-210°, pressure about 12 mm,) and weighed 50 grams. It had a rotation of + 11.14' (I= 2, This substance was then redistilled under reduced pressure. I t boiled a t 157-158' (bath 194-197', pressure 15 mm.).Three Erac- tions were collected, the first being only a small one. The rotation of the second was + 10.9' and of the third + 11.0" ( I = 2, t = 15'). This is approximately intermediate between the rotations of methyl tartrate and ethyl tartrate, so that the substance obtained was probably composed mainly of methyl ethyl tartrate, but it showed no tendency to crystallise nor did any solid separate out when a few crystals of methyl tartrate were added. No further purification could be expected by distillation, and as any other method, except crystallisation, would probably involve more labour than the result would justify and, in addition, be still uncertain, the attempt to prepare this substance in a pure condition has been for the present abandoned.The result of the first experiment, however, seemed interesting enough to warrant further investigation, since it appeared from it that ethyl hydrogen tartrate had been almost completely converted into metbyl tartrate by the action of methyl alcohol and hydrogen chloride. Now in certain cases the preparation of an ester from the acid and the alcohol presents difficulties which could be overcome by the use of another ester in place of the acid. It is therefore of importance to ascertain whether such a method will yield a pure product. t = 180). Preparation of Ethyl Tartrate from Methyl Tartrate. This was carried out by the method recommended by Fischer and Speier (Ber., 1895, 2'7, 3252), except that methyl tartrate was sub- stituted for the acid.Hydrogen chloride was passed into 120 grams of ethyl alcohol until VOL. LXXIX. X252 PATTERSON AND DICKINSON: THE PREPARATION OF 5 grams had been absorbed; 30 grams of methyl tartrate were then added, and the mixture boiled for about 4 hours under a reflux con- denser. The alcohol and hydrogen chloride were then removed, and the ester remaining was distilled in a vacuum. Three fractions were collected, the last of which had an observed rotation of + 8.498' ( J = 1, t = 1S96O). Under the same circumstances, pure ethyl tartrate would have a rotation of +9*15'. Since pure methyl tar- trate has a rotation of about + 2.8' a t the same temperature, and we may assume without any great error that the rotation of a mixture of ethyl and methyl tartrates is the sum of the rotations of each constituent separately, it follows that about 90 per cent.of the original methyl tartrate had been converted into ethyl tartrate. The three fractions of impure tartrate were then mixed and treated anew with 120 grams of ethyl alcohol containing 5 grams of hydrogen chloride, After the alcohol and hydrogen chloride had been removed and the ester fractionated several times, a product was obtained having a rotation of +9.25' (Z=l, t=18*8'), which is exactly the same as that of pure ethyl tartrate. It is thus clear that methyl tartrate may be completely converted into ethyl tartrate by the method described. The reverse reaction was then tried, Prepratim of Methyl Tartrate from Ethyl Tartrate. Sixty grams of methyl tartrate were boiled for about 5 hours with 240 grams of methyl alcohol containing 15 grams of hydrogen chloride.On subsequent removal of the alcohol and distillation of the residue in a vacuum, a very viscous liquid was obtained of observed rotation + 6.249' (1 = 2, t = 16*5"), so that about 87 per cent. of the ethyl tar- trate had been converted into methyl tartrate. Instead of treating this again with methyl alcohol and hydrogen chloride, a few small crystals of methyl tartrate were added. By next morning, the ester had almost completely solidified, only a small quantity of liquid refusing to crystallise. The whole mass was spread on porous earthen- ware. When dry, the solid weighed 37 grams. This represents a yield of 71 per cent, on the 60 grams of ethyl tartrate used, which, considering the loss of methyl tartrate incurred in the removal of a first fraction in distillation, and of the portion dissolved in the ethyl tartrate which was absorbed by the porous plate, agrees as well as could be expected with the 87 per cent.found polarimetrically. The substance obtained mas then dissolved in benzene and allowed to crystallise ; 22 grams separated out. The crystals were dried on porous earthenware and then distilled. After the first small fraction had been removed, the residue boiled at 150' (bath 175', pressure aboutESTERS FROM OTHER ESTERS OF THE SAME ACID. 283 15 mm.). The distillate soon solidified; its melting point was not very sharp, and lay a t 55-56O, which is considerably higher than that usually given, namely, 48O. Its density, relative to water a t 4', was examined at various temperatures, the following figures being obtained : Temperature .. . . . . 2g3 39.8O 58.2' 7 7.3O Density ....,. ... ... 1,3187 1.3074 1.2878 1.2676 These figures plotted relatively to temperature are found to lie upon a straight line. By extrapolation, the density a t 20' is found to be 1.3286, Pictet (Jahresber., 1882, 856) giving 1.3284, and at looo, 1.2433, Pictet giving 1.2500. As the rotation of methyl tartrate appears to have only been examined at 18O, ZOO, and looo, we have obtained the following figures for the rotation of this specimen at intermediate temperatures, from which a curve may easily be constructed. to. 18.2 12.7 36.1 99 82.1 79-8 71.4 64.2 5 8.6 52.5 28.1 Rotation (100 mm.). + 2*910° 2,250 4.766 8.063 7.580 7476 7.025 6.650 6.305 5.958 4,030 Density. 1.3306 1.3363 1.3113 1.2444 1.2624 1.2649 1.2738 1.2815 1.2874 1.2937 1.3198 [ q:. + 2-19' 1.68 3.63 6-48 6.00 5.91 5.51 5.19 4-90 4.6 1 3.05 Since, from the experiments described above, it appears that methyl tartrate can be converted into pure ethyl tartrate, and vice versci, by Fischer's method, and as the first experiment mentioned shows that the same change can be accomplished by the saturation method a t a low temperature, i t seems probable that any other ester may be obtained in a pure condition by an analogous process, it being best, of course, to choose as starting point a substance whose physical condition is different from that of the ester which it is desired to prepare. YORKSHIRE COLLEGE, LEEDS. x 2

ISSN:0368-1645    

DOI:10.1039/CT9017900280

出版商:RSC    

年代:1901

数据来源: RSC

摘要:

284 LEE: NOTE ON TECOMIN. XXVIK-Note on Tecomin, u Colouying Mutter derived from the Heart- Wood of Bignonia Tecoma. By THEODORE H. LEE. Bignonia Tecoma is a fairly common tree i n the uplands of Minas, Brazil. When full grown, it is about thirty feet high t o the first branches; the crown is umbrella-shaped and a further twenty feet high. The bark is smooth and the trunk usually crooked. IE September, just before the rains set in, B. Tecorna is a very striking object, as it is covered with a mass of brilliant yellow flowers resembling the azalea, and showing up all the more strongly on account of the absence of leaves, which do not appear till later. The tinctorial properties of the wood have long been known t o natives, who mix the sawdust and shavings with slaked lime, and heat the mass with water.The resulting bath is used t o dye cotton cloth. A paste of lime and the sawdust with water is used to stain lighter coloured woods a deep brown. The writer’s attention was recently drawn to the fact that, on rubbing the sawdust with soap and water, a pink colour was developed. Preliminary experiments showed that a yellow colouring matter was contained in the wood, which was coloured red by alkalis, and a clearer yellow by acids. Fifty grams of the sawdust were exhausted by 85 per cent. alcohol and the combined alcoholic filtrates concentrated to about 100 C.C. in a retort. On transferring t o a dish and cooling, a plentiful crop of shining chrome yellow crystals with a nacreous lustre appeared. These were collected, washed with cold alcohol, and dried between blotting-paper, The filtrate was evaporated to a paste, taken up with aqueous ammonia, boiled, and filtered to separate a resinous residue.The ammoniacal filtrate, which is of a magnificent crimson colour, was acidified with hydrochloric acid, the dull yellow precipitate washed, extracted with hot alcohol, filtered, and the alcohol evaporated. A further crop of the yellow crystals was obtained ; these, however, were of a darker colour than the first, and required recrystallising from alcohol to free them from traces of resin, which is obstinately retained. The yield of the yellowing colouring matter was approxi- mately 5 per cent. Of the resin, 2 per cent. was separated, and a further quantity remained in the mother liquor from the second crystal- lisation.A small quantity of the yellow dye, provisionally named tecornin, is retained in this mother liquor, but the amount is small, asFOWLER : IRON NITRIDE. 285 the colour reaction with alkalis is almost masked by the brown colour of the resin. From the sawdust which has been exhausted by alcohol, hot dilute caustic soda extracts a deep brown dye (which is that utilised by the natives). It is precipitated by acids as a flocculent, brown solid, soluble in alcohol, from which it separates on evaporation as an amorphous mass. It is slightly sensitive to alkalis probably on account of a little tecomin retained. The crystnllised tecomin is not very soluble; a saturated solution in S5 per cent. alcohol contains about 0.8 per cent. of it. When a couple of drops of this solution are let fall into 50 C.C.of water, a faint opalescence is produced. On addition of a drop of a dilute mineral acid, the solution clears and takes a pale yellow tint. Weak acids do not affect tecomin, which therefore can be used in the cold for titrating carbonates, silicates, sulphides, sulphites, borates, and cyanides. Solutions of acetates, oxalates, citrates, &c., are alkaline in their reaction, and the red colour is discharged slowly by strong acids, but the end-point is indistinct. The indicator answers excellently for alkalimetry, using the alkalis, alkaline earths, and mineral acids as reagents, Calcium, barium, and magnesium carbonates react with the yellow solution, giving the red colour. The sensitiveness is extreme. I n a bulk of 50-70 C.C. coloured with two drops of the alcoholic solution, 0.2 C.C. of N/100 acid or alkali causes a sharp end reaction. MINAS, BRAZIL. A little alkali changes this to a full rose-colour. Organic acids affect the colour reaction, but indefinitely, MORRO VELHO,

ISSN:0368-1645    

DOI:10.1039/CT9017900284

出版商:RSC    

年代:1901

数据来源: RSC

摘要:

FOWLER : IRON NITRIDE. 285 XX1X.-Iron Nitride. By GILBERT JOHN FOWLER, 1SI.Sc. (Vict.). ALTHOUGH the chemistry of the nitrides is of considerable interest from many points of view, little is definitely known concerning these compounds. Since the discovery of azoimide, which reacts with metals to form a class of nitrides, i t became of interest to compare the pro- perties of the latter with nitrogen compounds of the metals obtained by other means. Iron nitride, from its fairly easy decomposability, appeared likely to present some interesting reactions and possibly t o allow of a determination of its heat of formation. The earliest work on the combination of iron and nitrogen was286 FOWLER : IRON NITRIDE. done by Berthollet, and by Thenard (2’~aitd de Chimie, 1834, i, 434), who allowed ammonia to act on iron wire a t a red heat and found that the weight of tho iron scarcely increased, but that the metal became brittle. Savart (quoted by Stahlschmidt, Pogg.Ann., 1865, 125, 37) men- tioned that after ammonia had passed over iron for a long time the metal became softer, whilst‘after an action of only two hours it behaved like steel and could be hardened. These observers did not recognise the formation of iron nitride. According to Desprets (Ann. Chim. Phys., 1829, 42, 122), iron increased in weight from 7 t o 111.5 per cent. by heating in ammonia and ohanged to a white, brittle mass, which he recognised as a com- pound of iron and nitrogen. Buff (Annalen, 1852, 83, 375) found an increase of 6 per cent. and RBgnault of I2 to 13 per cent.in this reaction. Fremy (Compt. rend., 1861, 52, 322) prepared iron nitride by heating anhydrous ferrous chloride in a stream of dry ammonia. The compound thus obtained, according to Fremy, contained 9.3 per cent. of nitrogen, from which he calculated the formula Fe,N2. The 11.5 per cent. of nitrogen found by Despretz corresponded to that required for the formula Fe,N. From these results, it might be concluded that nitrogen, like carbon, was able to combine with iron in several proportions t o form com- pounds which would alloy with free iron, Stahlschmidt (ZOC. cit.), however, came to the conclusion that nitrogen combines with iron in a perfectly definite ratio to form a substance of the formula Fe,N, and that all the nitrides previously prepared containing less nitrogen than required by this formula must be looked upon as mixtures of this nitride of definite composition with pure iron.According to Rogstadius (J. pr. Chern., 1862, 86, 307) and Briegleb and Geuther (AnnaZen, 1862, 123, 228), finely divided iron prepared from ferrous oxalate or by the reduction of ferric oxide in hydrogen, takes up about 2 per cent, of nitrogen at the ordinary temperature or by continuous heating in nitrogen. Stahlschmidt was not able t o prepare iron nitride in this way. By the electrolysis of a mixture of a ferrous salt and sal ammoniac, a shining or spongy mass separates which, according to Kramer (Arch. Pharm., 1861, [ii], 105, 284), is an iron nitride containing 1.5 per cent. of nitrogen ; according to Meidinger (Dingl.PoZyt. J., 1863, 163, 283), it is an iron ammonium alloy. The substance is-said to give off a smell of ammonia and liberates hydrogen from boiling water. It should be mentioned, finblly, that Allen (Chem. News, 1880, 41, 231), by heating soft iron wire in ammonia, obtained a product con-FOWLER : IRON NITRIDE. 287 taining 24 per cent. nitrogen, and concluded, from comparison of the behaviour of this compound with iron and steel containing nitrogen, that this element existed in iron and steel combined with the iron to form nitride, The experiments on iron nitride described in the present paper may be divided under the following heads :- 1. Preparation. 2. Analysis. 3. Determination of physical properties, 4. Investigation of action of various reagents on the substance.5. Determination of the temperature of decomposition and heat of formation of iron nitride (see Fowler and Hartog, this vol., p. 299). The experiments have confirmed the results of Stahlschmidt in almost every particular, and various new facts have been ascertained, conclusions from which have been dealt with below. Preparation of Iron Nitride. This was accomplished by three methods : ( a ) By the action of ammonia on ferrous chloride or bromide. ( b ) By the action of ammonia on finely divided iron. ( c ) By the action of ammonia on iron amalgam. Preparation of Ammonia.-& large quantities of ammonia were required in this research, a ready method of preparing the gas pure and dry was necessary. The best plan was found to be a modification of one described by Neumann (J.pr. Chem., 1888, [ii], 37,342). A strong solution of ammonia (sp. gr. 0.88) is allowed to drop from a tap funnel upon solid caustic soda contained in a glass tower, from the bottom of which the solution of the alkali can be drawn off from time to time. It was found best to place a pad of glass wool under the end of the tap funnel, as the evolution of ammonia is apt to be too violent if the sdution is allowed to fall directly on to the caustic soda. The ammonia gas, thus evolved in a fairly dry state, from contact with the solid caustic soda, was passed through three towers, filled partly with sticks of caustic potash and partly with granulated soda lime which had been previously heated. The gas, after passing through the tube in which the nitride is pre- pared, was led into an ordinary nitrogen measuring tube containing hydrochloric acid, which is prevented by mercury from being sucked back.By this means, it was easy t o see when the ammonia was free from air and also to note the progress of the reactions described below. (a) Action of Ammonia on Ferrous Chloride.-The course of the reaction was carefully observed, especially with a view t o discover whether any hydrazine derivatives were formed. The ferrous chloride,288 FOWLER : IRON NITRIDE. prepared by heating iron wire to redness in a current of hydrogen chloride (the unattacked iron was removed by means of a magnet) was submitted to the action of ammonia at various temperatures up t o the melting point of lead, and the products of the reaction, both solid and gaseous, were examined.The course of the reaction appears t o be as follows (Fowler, Chem. News, 1900, 82, 245). At low temperatures, one mol. of ferrous chloride rapidly absorbs six mols. of ammonia to form a voluminous white mass which gives off ammonia again a t a temperature of looo. On further heating reduction of the ferrous chloride takes place, resulting in the formation of ammonium chloride, iron nitride, and a small quantity of nitrogen. No formation of hydrazine compounds or other intermediate products could be observed. To prepare the nitride in quantity by this method a temperature of about 600° appears to be necessary. B’em-0~8 Bvomide was prepared and similarly treated, the temperature of reaction was not sensibly lowered and the products of analysis were analogous (Fowler, Zoc. cit.).I n appearance, the nitride obtained from either ferrous chloride or bromide was sometimes like that obtained from reduced iron; a t other times, when a higher temperature was used, it was obtained in silvery pellicles as described by Stahlschmidt and others. It was found difficult to obtain it in quantity free from intermixed ferrous and ammonium chloride or bromide ; it was therefore thought better t o pre- pare the nitride directly by the interaction of iron and ammonia. (b) Preparation of Iron Nityide by Interaction of Ironand Ammonia. -For this purpose, iron reduced from the oxide obtained by precipi- tation and drying was found to be most suitable. As this iron tends t o be pyrophoric, it was found necessary to complete the reduction of the oxide in hydrogen and the conversion of the iron into nitride without exposure to air.Ferric hydroxide is reduced by hydrogen in a wide glass tube to each end of which narrower glass tubes are fused which can be readily connected up with the ammonia apparatus and measuring tube outside the furnace. One of the narrower tubes is bent i n such a way that this part of the apparatus can be conveniently detached and suspended on the balance. The hydroxide must be heated in hydrogen until no further loss of weight occurs, as it is found that if the reduction is not complete, subsequent heating in ammonia at a temperature sufficient to form nitride is not sufficient completely to reduce the oxide of iron. The last portions of oxide mixed with alarge excess of iron apparently require a higher temperature for their reduction.It was found in one case, in which care had not been taken to completely reduce the oxide in hydrogen, that, after heating in ammonia until no further increase in weight occurred, there wm still a large amount of oxide present, as evidenced by the fact thatFOWLER : IRON NITRIDE. 289 a black, insoluble residue consisting probably of magnetic oxide was left on treating with dilute acid, and also that on heating further in ammonia at a higher temperature water mas given off. As this steam decomposes iron nitride, the presence of oxide of iron acts prejudicially to the formation of the latter. When the iron is once reduced, it is important that air be as far as possible excluded until i t has been completely converted into nitride.By using pinchcocks upon the indiarubber connections, it is possible to connect and disconnect the tubes and insert stoppers without admitting an appreciable amount of air. After each weighing, the contents of the tube were shaken so as to expose a fresh surface. The best temperature for the preparation of iron nitride is that stated by Stahlschmidt, namely, just above the melting point of zinc (414’). A t this temperature, hydrogen is rapidly given off on passing ammonia over finely divided iron. The almost complete cessation of the evolution of hydrogen shows the reaction to be complete. If the temperature is raised, it is found, in confirmation of Stahl- Schmidt’s results, that the percentage of nitrogen in the nitride decreased, even if ammonia be passing over.This is explained by the fact that iron nitride is decomposed by hydrogen at the same temperature as that a t which it is formed, ammonia being again pro- duced. On raising the temperature, the percentage of hydrogen in the gas above the substance is increased by dissociation of the ammonia, and the tendency is for it to take up nitrogen from the nitride. By allowing hydrogen to pass over the nitride heated in a tube, Stahl- Schmidt showed (Zoc. cit., p. 47) that ammonia is re-formed at this temperature, and this result was confirmed. As soon as the tempera- ture of the furnace reached approximately that at which nitride is formed, ammonia was evolved, and recognised by its precipi- tating Nessler’s solution, placed in connection with the end of the apparatus.The temperature of formation in ammonia, and decomposition in hydrogen, of the nitride being the same, the preparation of the com- pound is difficult and tedious, and some practice is needed to keep the current of ammonia passing a t a sufficient pace, as, if hydrogen is allowed to accumulate, the formation of nitride ceases. It was a t first thought that this reaction might be a suitable one to study quantitatively at different temperatures as an example of mass action. The temperature, however, at which the reaction takes place, and that at which ammonia is dissociated in presence of n surface such as powdered iron affords (Ramsay and Young, Trans., 1884, 45, 88), are too close to admit of satisfactory results being obtained, It was found indeed, that even at the temperature of the experiment some ammonia is in all probability dissociated.Some of the gas, after290 FOWLER : IRON NITRIDE. passing over the iron and iron nitride formed, was collected in the nitrometer and analysed by exploding with air in a Hempel pipette, with the following results : Expt. I. Expt. 11. Gas taken .............................. 11-22 C.C. 11.4 C.C. Air added .............................. 50-28 58.0 Contraction ........................... 16.0 16.4 hydrogen ,.. 10.7 10.9 residua1 gas.. 0.52 0.5 Composition of gas AS the gas for analysis mas all collected in the space of about 10 minutes, and as, at the close of the experiment, on allowing the apparatus to cool in a current of ammonia for an hour or more after driving out the hydrogen from the apparatus, practically the whole of the gas was absorbed by the acid in the nitrometer, the residual gas above-mentioned can scarcely be nitrogen derived from air present in the apparatus, but must be due to the dissociation of ammonia.It seems probable that on passing ammonia over reduced iron at a temperature rather above 400°, ammonia is dissociated, the nitrogen at the moment of its liberation combining with the iron. A certain small proportion, however, will escape combination thus, and pass on uncombined along with the hydrogen. Ramsay and Young (Zoc. cit.) give the temperature at which ammonia begins to dissociate under the most favourable circumstances as a little below 500’.The above experiments tend to show that dissociation begins a t a lower tempera- ture under the conditions described. In confirmation of another of Ramsay and Young’s results, it may be mentioned that on heating ammonia in a sealed tube to a temperature approaching redness no decomposition ensued. (c) Preparation of I r o n Nitride from Iron ArnaEgarn.-Becent in- vestigations have shown that iron and other metals, when liberated from their amalgams, have specially energetic properties (Guntz, Compt. rend., 1892, 114, 115 ; Maquenne, ibid., 25, 220). Accordingly, some iron amalgam was prepared by Joule’s method (Mem. Mun- Chester, Lit. and Phil. Xoc., 1865, [iii], 2, 115). A saturated solution of ferrous sulphate was electrolysed by a weak current, a bundle of iron wire constituting one electrode, the other being a piece of plat- inum foil attached to a stout platinum wire, which was inserted into a glass tube filled with mercury, by which connection was made with the battery.The platinum foil dipped under mercury contained in a small, deep porcelain dish. On passing the current, iron amalgam is formed round the platinum. It is important to use a weak current, or the iron is deposited in a powdery state, and does not amalgamate with the mercury. The excess of mercury is squeezed from theFOWLER : IRON NITRIDE. 291 amalgam, which is obtained of a buttery consistency, and contains about 10 per cent, of iron as determined by ignition. Joule's soft amalgam contained 11.8 per cent. of iron. This amalgam rapidly blackens in air, even Then carefully washed free from acid ; it blackens also under water, On drying some of this amalgam and heating in ammonia, the mercury was driven off and iron nitride gradually formed from the porous iron left.The latter, however, seems to offer no advantage over ordinary reduced iron for the preparation of the nitride, Doubtless this is to be expected from the character of the reaction, there being always the same tendencies to break up the nitride formed. A determination of nitrogen in the substance formed by long-continued action of ammonia gave 9.16 per cent., showing that only partial conversion had taken place. That this percentage was taken up by degrees was evident from the fact that after the mer- cury was expelled the weight gradually increased on continuing the action of ammonia, Also the iron, obtained after driving off the mercury, was readily attracted by a weak magnet, which is not the case with iron nitride.The percentage of nitrogen mas found t o vary somewhat between the more and less finely divided portions of substance. Before leaving the subject of the preparation of iron nitride, it should be mentioned that Stahlschmidt unsuccessfully attempted to make this compound by heating iron and nitrogen together, and also by passing a mixture of nitrogen and hydrogen over heated ferrous oxide. His results are therefore at variance with those of Briegleb and Geuther and of Rogstadius (Zoc. cit.). This point was investigated in the present research as follows. About 7.6 grams of reduced iron mere sealed up in a glass tube 30 cm.long and 2 cm. diam. in an atmosphere of nitrogen. After heating at 180Ofor 6 hours, and again a t 330' for 8 hours, the tube was opened under mercury. There was no indication that absorption had taken place. On resealing and heating a t 420' for 4 hours, and at 470' for 6 hours, no indication of absorption was noted on opening under mercury. It may be concluded, therefore, that under these conditions iron and nitrogen do not combine. No attempt was made to obtain iron nitride by electrolysis, as the experiments of Classen (Quantitative chemische Ancdyse durch Elektrolyse, 1886, p. 56), confirmed by laboratory experience, show that pure iron is obtained on electrolysing a solution of ferrous ammonium oxalate, Analysis of Iron Nitride.That all the nitrogen is converted into ammonia on solution, and that the substance consists solely of iron and nitrogen, was proved by292 FOWLER : IRON NITRIDE. analysis of a sample which was used for a determination of the heat of formation. I n this case, the compound was dissolved in sulphuric acid, the iron oxidised by nitric acid, precipitated by ammonium hydroxide, and the ferric hydroxide washed, ignited, and weighed. I n this case, the solution in the calorimeter was diluted to one litre, and 500 C.C. taken for analysis. 0.6625 gave 0.8466 Fe,Os. The nitrogen was determined by distilling the nitride with caustic soda, absorbing the evolved ammonia in hydrochloric acid, and es- timating as platinum salt. Fe = 89.4 per cent. Two determinations were made : 0.1714 gave 0.2836 (PU’H,),PtCI,.N = 10.46 per cent. 0.3126 ,, 0.5791 (NH4)2PtCI,. N = 10.5 ,, The composition therefore is Fe = 89.4; N = 10.48 (mean). Total = 99.88 Although this sample still contained a small quantity of free iron, the analysis shows that the nitride most probably dissolved according t o the following equation : 2Fe,N + 6H,SO, = 4FeS0, + 2(NH,)HSO, + H,. This is rendered certain by further experiments. Another sample was analysed in the following manner. A weighed quantity was dissolved in hydrochloric acid, evaporated on a water- bath with platinic chloride to almost complete dryness, and taken up with absolute alcohol. On filtering and washing with absolute alcohol, the ammonium plstinichloride can be obtained free from traces of iron, if care be taken to leave a trace of free hydrochloric acid in the evaporating basin before adding alcohol.This method was used throughout as allowing of greater accuracy with small quantities than a method involving distillation. The following result was obtained : 0.2020 gave 0.3537 (NH4),PtC16. Ee,N requires N = 11.1 1 per cent. The percentage of platinum found after ignition of the double salt was 43.77, the calculated percentage for the ammonium salt being 43.79. This result shows that only ammonium compounds are formed when the nitride is dissolved in acids. This conclusion is confirmed by a determination of the hydrogen evolved when the nitride is dissolved in sulphuric acid. The determin- ation was made by enclosing a weighed portion of the substance in a small tube and quickly inserting under a gas measuring tube standing in a small glass trough.Both tube and trough were filled with dilute siilphuric acid. The gas was prevented from escaping from beneath the measuring tube by means of a funnel, Correction was made for N = 11.07 per cent.FOWLER : IRON NlTRIDE. 293 the amount of air enclosed in the small tube containing the substance ; it amounted to 0.3 C.C. The calculated amount is 24.4 C.C. assuming the nitride to dissolve according to the equation : 0.275 gave 23-1 C.C. hydrogen (N.T.P.). 2Fe,N -t- 6H,SO, = 4FeS0, + 2(NH,)HSO, + H,. Allowing for the fact that a little hydrogen will dissolve, these figures are in sufficient agreement with the equation, and show plainly that no hydrogen enters into the composition of the nitride and that no nitrogen compounds other than ammonium salts are formed by treat- ment with acid.(a) By treatment with nitric acid and subsequent ignition : * The iron in the same sample was determined. 0.4215 gave 0,5325 Fe,O,. (b) The nitride was dissolved in sulphuric acid and the solution diluted to 250 C.C. and titrated with potassium permanganate (1 C.C. = 5.505 mgms. Fe), after reduction of any oxidised iron by means of zinc : 0.4833 required 15-6 x 5 C.C. KMnO, (mean of 3 titrations). F e = 88-82, 0.5185 ,, 16.6 x 5 C.C. KMnO, ,, 9 9 Fe = 88.1. Fe,N requires F e = 88.88 per cent. No percentages of nitrogen above 11.1 could be obtained, whilst any percentage below that could be got according to the time during which the iron had been exposed to the current of ammonia, These results are fully in agreement with those obtained by Stahlschmidt, and confirm his conclusion that only one iron nitride exists, and that i t has the composition expressed by the formula Fe,N.Fe-88.43 ; calc. 88-88 per cent. Physical Properties of Iron Nitride. Appearance and Maynetic Qualities. Iron nitride prepared as above described, is a grey powder, or' a rather duller tone than reduced iron. One of the difficulties in work- ing with this substance is that, it is impossible from its appearance alone to tell whether it is even approximately pure. The powder is gritty, and the particles under the microscope have a vitreous appear- * Satisfactory results could not be obtained by simply igniting in air, and weigh.in,a the oxide of iron farmed owing to the varying percentages of magnetic oxide produced. This source of error seems to have escaped notice by Stahlschmidt, who de- termined his iron by simple ignition. By heating with nitric acid and decomposing the resulting nitrate, better, but not entirely satisfactory, results are obtained. As i t is the practice in some commercial analytical laboratories to incinerate filter papers together with an adhering precipitate of iron oxide, whereby iron is reduced and subsequently oxidised largely to magnetic oxide, this source of error deserves mention,294 FOWLER : IRON NITRIDE. ance differing from reduced iron, which appears powdery, Although not attracted by a strong bar magnet, a moderately strong electro- magnet readily attracts it.The magnetic property of iron is therefore still apparent, even when this element is combined with nitrogen. Xpecz$c Gmuity.-This was determined by means of the specific gravity bottle, using benzene and water as liquids for comparison, The quantity of iron nitride available for the determination was rather small, being about 0.5 gram, so that two results obtained with benzene varied from 6-55 to 6.065. Using water as the liquid for comparison, the value 6.25 was obtained. A deter- mination made a t the same time of the specific gravitgof reduced iron gave the value 7.89. The above value for the sp. gr. of iron nitride is rather higher than the number 5 given by Despretz for the nitride described as a white, brittle mass. It is of interest to compare the following sp.gr., taking in all cases the highest value recorded : Reduced iron., ............. 7.S9 Taking the molecule of iron to be diatomic, we get the following In each case, concordant weighings only were taken. Magnetic oxide, Fe,04 ... 5.2 Ferric oxide, Fe,O, ....... 5-17 Iron nitride, Fe,N ........ 6.25 values for the atomic volumes : Iron ........................ 7.1 0 in Fe,O, .................. 5.8 Nitrogen in Fe,N ......... 5.9 0 in Fe,O, .................. 5.6 Chemical Reccctions of I~on Nitride. Oxidation.-Air or oxygen dried by siilphuric acid was passed over the nitride contained in a boat placed in a L. Meyer constant temperature furnace. The escaping gas was passed into potassium iodide. On heating to 120°, a slight liberation of iodine took place (about 5 milli- grams), and there was a gain in weight of about 3 milligrams on half a gram of the substance. Oxidation became visible a t ZOO0, brown ferric oxide being formed; there was no increase, however, in the amount of iodine liberated.The nitrogen therefore appears not to suffer appreciable oxidation under these conditions, the slight libera- tion of iodine being possibly due to nitric peroxide produced by the oxidation of a trace of ammonia occluded in the nitride. ChZorine.-The action of chlorine was examined by placing some of the nitride in a boat enclosed in a tube, and passing chlorine over after displacing the air by carbon dioxide. Any nitrogen given off could then be recognised by passing the gases through strong potash solution contained in an ordinary nitrogen measuring tube.The nitride takes fire in chlorine on slightly warming, or sometimes spon-FOWLER : IRON NITRIDE. 295 taneously, ferric chloride being produced and nitrogen evolved. There is no evidence of the formation of nitrogen chloride, said to be produced when chromium nitride is similarly treated. Bromine.-Bromine attacks the nitride only slowly, even on warming, and as ammonium bromide was always found in solution after action had taken place, it is probable that the reaction is chiefly brought about by the trace of hydrobromic acid contained in the bromine. Iodine.-An ethereal solution of iodine has no action on iron nitride. Hydrochlo& and Sulplmric Acids.-The action of these acids in the diluted condition has already been described, namely, the nitride is dissolved, ferrous and ammonium salts being formed, and hydrogen liberated.I t s action was not exhaustively studied, as it must vary much according to the strength, The main reaction, doubtless, is the formation of ammonium nitrate and the products of the action of nitric acid on iron, nitrous or nitric oxides being formed according to the strength of the acid. No great evolution of nitrogen peroxide takes place on warming with moderately concentrated nitric acid. Hydrogen, ChZorids.-It is interesting to note that, whilst hydro- chloric acid rapidly attacks the nitride in the cold, gaseous hydrogen chloride only begins to attack it at about 220°, the action becoming rapid at 350°, complete conversion into ferrous chloride and ammonium chloride taking place.Nit& Oxide.-The action of nitric oxide was examined in order to compare the temperature of oxidation with that a t which free oxygen begins to act on the substance, and also to see whether the nitrogen which would be evolved in such a reaction differed in its properties from ordinary nitrogen. Nitric Acid.-This acid, even if strong, acts only slowly. It seemed possible that the following reaction might take place : Fe,N + 2N0 = 2Fe0 + N,. In view of the opinion put forward by Berthelot and others that argon may be condensed nitrogen of the Formula N,, it became of interest to examine the gas given off in the experiment to see whether such a condensation occurred under these conditions. Oxidation began at about the same temperature as in oxygen, namely, at about 120', becoming rapid at 170' ; the gas evolved was collected, and, after removal of the nitric oxide by ferrous sulphate solution, was sparked with oxygen over potash, being practically all absorbed.It may be concluded therefore that the gas produced when nitric oxide acts upon iron nitride is ordinary nitrogen. Action, of Ccccbon Monoxide.-On heating the nitride to a tempera-296 FOWLER : IRON NITRIDE. ture approaching redness in a current of carefully purified carbon monoxide, freed from moisture, oxygen, and carbon dioxide by passing successively through strong sulphuric acid, red hot copper gauze, potash, and strong sulphuric acid, a, slight loss of weight took place and the substance darkened somewhat in colour. On passing the gas into lime water, a precipitate of calcium carbonate showed that carbon dioxide had been formed.This may either result from the reduction of a trace of oxide in the substance or from the decomposition of the carbon monoxide, the darkening above-mentioned being due to de- position of carbon as in the experiments of Lowthian Bell. On passing the escaping carbon monoxide into potash solution, no cyanogen com- pounds could be detectedin the latter, so that the nitrogen does not appear to com bine with carbon under these conditions. Possibly some iron cyanide may be formed, but this was not looked for at the time. Carbon Dioxide.-It was thought at first that the temperature of decomposition of the nitride might be found by heating in carbon dioxide. Accordingly, some of the compound was heated in this gas in a tube surrounded by vapour of mercury, and afterwards by that of boiling sulphur.A t the temperature of the latter, 448O, the nitride was oxidised, as could be judged by the change in appearance and the increase in weight. I n another experiment, the compound was strongly heatedin a boat placed in n tube through which a current of carbon dioxide was passed, tho escaping gas being collected over potash. It was found to consist chiefly of carbon monoxide, some nitrogen, however, being left after absorbing the carbon monoxide with ammoniacal ccipper solution. No definite equation can be given for this reaction, as the increase in weight observed in the nitride may be due partly to deposition of carbon as well as to oxidation.Assuming, however, magnetic oxide to be formed on oxidation, the ratio between the nitrogen and carbon monoxide should be rather more than 1 to 5. Out of 17.8 C.C. taken for analysis, 3.4 were not absorbed by copper solution, giving approxi- mately this ratio. The temperature a t which nitrogen is given off in this way was roughly determined by noting the behaviour of salts of known melting point placed in small pieces of platinum foil on the heated tube. As the temperature was above the melting point of lead chloride and about that of calcium chloride, 530' may be taken to re- present the temperature at which iron nitride is oxidised by carbon dioxide. It is possible that the nitride used contained small quantities of free iron, but as more than 10 per cent. of nitrogen was present and as any finely divided iron present is rapidly oxidised in air, the amount of free metal present could be but small.Action of Hydrogen Xu&hide.-On heating with hydrogen sulphide,FOWLER : IRON NITRIDE. 207 increase in weight took place, and a strong smell of ammonia was noticed in the soda-lime used to absorb the excess of hydrogen sulphide at the end of the tube. The hydrogen sulphide was obtained pure by heating a solution of magnesium hydrosulphide. The substance left in the boat after heat- ing in hydrogen sulphide contained iron sulphide, as it evolved hydrogen sulphide on treatment with an acid. The reaction taking place when hydrogen sulphide is heated with iron nitride is evidently expressed by the equation : 2Fe2N + 6H2S = 4FeS + 2NH,HS + H,.The temperature at which this reaction occurs was approximately ascertained by heating the nitride in a current of hydrogen sulphide in a test-tube immersed in a bath of sulphuric acid. The gas, after passing over the nitride, was conducted into a neutral solution of ferrous sulphate. As soon as ammonia was formed, ferrous sulphide was precipitated. This took place at about 200". Hydrogen sulphide alone was found not to precipitate neutral ferrous sulphate. Action of Stearn.-When the nitride was heated in a current of steam, kept a t 100' b y a bath of boiling water, a very slight evolution of ammonia was noticed, the litmus paper being slowly but continu- ously turned blue. Behaviour with Carbort.-On heating iron nitride with carefully purified sugar charcoal (freed from hydrogen by chlorine) in a tube closed at one end and passing any evolved gas into potash, no reaction for cyanides could be obtained in the latter.The nitrogen in the nitride appears therefore to have no tendency t o combine with carbon either in the free state or as carbon monoxide. I n presence of sodium, however, the two elements combine and sodium cyanide is formed. On heating a small amount of the nitride with pure charcoal and a small piece of freshly cut sodium, a reaction takes place, and on dis- solving out in water a strong Prussian blue reaction is given on adding a mixture of ferrous and ferric salts and acidifying. The reaction was not given after heating the charcoal and sodium alone.Copper Su&hate.--Neutral copper sulphate solution is without action on iron nitride; on slight acidification, copper is precipitated, the nitrogen going into solution as ammonia. Phenol.-In order to compare the action of phenol, as a typical substance of slightly acid nature, with the behaviour of ordinary mineral acids, the nitride was heated with phenol in a sealed tube at 220O. No appreciable change took place, no pressure was created in the tube, and the substance after extracting with ether and washing was unaltered in appearance. The weight of nitride obtained after the treatment differed only slightly from the weight taken for experi- VOL. LXXIX. Y29 8 FOWLER : IRON NITRIDE. ment. It may be concluded therefore that phenol and compounds of a like nature are without action on iron nitride.Ethyl Iodide.-In order to ascertain whether amines mould be pro- duced by the action of ethyl iodide on the nitride, the two compounds were heated together in a sealed tube. Up to 150°, no action took place, judging by the absence of pressure in the tube, on opening after it had been exposed to that temperature. After heating for some time to 200-230°, considerable action took place, blackish, iodine-like crystals being formed and much gas produced. This gas was collected by breaking the end of the tube under a piece of indiarubber tubing communicating with a laboratory tube. On analysis of the gas in the Hempel apparatus, it was found to consist of olefines to the amount of one-third and a slight trace of paraffins, in addition to the air pre- viously in the tube.On boiling the crystals above mentioned with caustic soda, ammonia was evolved and iron hydroxide formed. On distilling with caustic soda, collecting the escaping gas in hydrochloric acid, and evaporating this solution with platinic chloride, a platinum salt was obtained which on ignition gave Pt = 44.9 per cent. (NH,),PtCI, requiring Pt = 43.7. No amines are therefore formed in this reaction, which is probably as follows : 2Fe,N + lOC,H,f = 2NH,I + lOC,H, + 4Fe1, + H,. Action of Hydrogen Peroxids in Presence of Acid.-It was thought of interest to study the simultaneous action of an acid and an oxidising agent on iron nitride, in order to see whether in such a case the nitrogen would be wholly converted into ammonia, become partially oxidised to hydroxylamine, or partially escape in the free state. Weighed quantities of nitride, containing 9 to 10 per cent. of nitrogen, were therefore treated with a mixture of hydrogen peroxide and sul- phuric acid, excess of hydrogen peroxide being used. The nitrogen was determined in solution after dissolving, by distilling with caustic soda, collecting the evolved gas in hydrochloric acid, and determining the ammonia as platinichloride. The nitrogen was also determined in a similar sample dissolved in acid without hydrogen peroxide. The results obtained show conclu- sively that most, at any rate, of the nitrogen is not oxidised. In on0 comparison, the percentages of nitrogen found in the two experiments were almost exactly the same, namely, 8-82 without and 8 -73 with hydrogen peroxide. Analysis of the gas evolved on solution of iron nitride in a mixture of sulphuric acid and hydrogen peroxide showed that it consisted chiefly of hydrogen and oxygen. The small amount of residual gas was not enough to lead to the conclusion that nitrogen had escaped conversion into ammonia.HEAT OF FORMATION AND CONSTITUTION OF IRON NITRIDE. 299 Temperature of Decomposition of Iron Nitdde.-A minimum value for this was obtained by heating the substance in a current of nitrogen. Up to the temperature of boiling sulphur, 448', no change in weight took place. A very slight loss occurred at a temperature approaching dull redness. The temperature of decomposition must therefore cer- tainly be above 600'. THE OWENS COLLEGE, MANCHESTFP.,

ISSN:0368-1645    

DOI:10.1039/CT9017900285

出版商:RSC    

年代:1901

数据来源: RSC