Researches upon Prospective Standard Substances for Bomb

---

RESEARCHES UPON PROSPECTIVE STANDARD SUBSTANCES FOR BOMB CALORIMETRY. VI

A STUDYOF SUCCINIC ACID L. J. P. KEFFLER

Thermochemical Laboratory, Inorganic Department, University of Liverpool, Liverpool, England Received August 18, 1999 I. INTRODUCTIOPI'

At a recent meeting, the Commission Permanente de Thermochimie (set up by the International Union of Chemistry) has adopted the principle of a secondary standard for bomb calorimetry and requested two of their number to investigate succinic and adipic acids, both of which appeared to be worthy of consideration as prospective secondary standards. The present paper is an answer to the Commission's request in so far as succinic acid is concerned. 11. PREPARATION O F PVRE SAMPLES O F SUCCIXIC ACID; MISCELLANEOUS

TESTS FOR PURITY AND STABILITY; PRELIMINARY COMBUSTIONS

Succinic acid, sold by the British Drug Houses (B.D.H.) as an analytical reagent guaranteed to satisfy their special standards of purity,' was used as the starting material. This was first recrystallized from conductivity water at a temperature kept below 70°C., in order to avoid the risk of some slight decomposition of the acid through loss of carbon dioxide which may possibly take place at higher temperatures.2 The crystals were filtered through a Buchner funnel and partially dried by passing a current of air through them for several hours. After standing between filter paper for two days, the acid was well crushed (in small portions at a time) in a mortar; it was then spread in thin layers in large Petri dishes and For some details, see section 111, paragraph 1. According to Lamouroux (4), saturated solutions of either oxalic or malonic acids give rise to very large bubbles of carbon dioxide (with simultaneous formation of acetic acid) when heated to 68°C.; similar decompositions may possibly take place with saturated solutions of the homologues of the above acids, though this was not actually observed by Lamouroux, perhaps merely because it was not sufficiently apparent. To get the acid quickly into solution without having to raise the temperature above 70°C., nearly four times as much water as would have been required a t 100°C. was used. The solubility of succinic acid is 1 per 0.81 part of water a t IOO"C., and 1 per 19 parts of water a t 15°C. 717 1

m m

JOURNAL OF PHYEICAL CHEMISTRY,

VOL.

xxxvxri, NO. 6

718

L . J . P. KEFFLER

dried for two further periods of 6 hours (each of these being followed by careful crushing), in a desiccator containing fresh phosphoric anhydride, under a pressure reduced to approximately 0.1 mm. of mercury. Two portions of about 4 g. each were now introduced in two weighing bottles and left to stand over phosphoric anhydride in an evacuated desiccator, the loss in weight being recorded from time to time so as to ascertain to what extent the drying had been carried out. The results given in table 1 were obtained. The loss in weight recorded during the last week, namely 0.05 to 0.08 parts per thousand, was so small that it might well be entirely accounted for by a very faint volatility of the substance when exposed to a low vacuum. But to make quite sure that if there were any water left occluded in the crystals it could only be in minute traces, the TABLE 1 Loss in weight of succinic acid (in powder) o n standing over phosphoric anhydride in vacuo Loss

WEIGHT AFTER 11 DAYS

grams

mg.

grams

grams

mg.

3.9407 3.8036

0.9 0.7

3.8029

3.9396 3.8026

1.1 1.0

INITIAL WEIGHT

~

LOSS

~ TOTAL LOSSo

LOSS I N PARTS PER ~ THOtBAND

1

0.25 per cent

. TABLE 2 Isothsrmal heat o j combustion of succinic acid, per gram weighed in vacuo, ut a n initial temperature oj 20°C. -

THE AUTHOR'S VALUES

1

VALUES OF VERKADE AND COOPS

samples were dried for a further 6 hours in vacuo, a t a temperature of 45°C. =t 5OC., and the heat of combustion determined directly afterwards in order to compare it with the value obtained by Verkade and Coops (8). The two groups of results are given in caloriesl5 in table 2, Verkade's values having previously been reduced to vacuo. It will be seen that the author's trial values are already very near to those of Verkade and Coops, more particularly to the first three which agree exceptionally well with one another while the fourth appears to be abnormallyhigh. If, on the other hand, it is observed that thevalues of the Dutch investigators (obtained by the ordinary method) seem to be usually somewhat higher than those of the author (obtained by the adiabatic method) , as illustrated before in the case of salicylic acid and naphthalene (3), it

E

STARDARD SUBSTANCES FOR BOMB CALORIMETRY. V I

719

will appear that the sample of succinic acid derived in the way described from the starting material was either completely pure and dry or a t least very nearly so. Having thus satisfied himself about the apparent degree of purity of the B.D.H. preparation, by some very sensitive and independent test, the author proceeded with the preliminary examination of succinic acid from the point of view of its resistance to decomposition on heating in air or in vacuo. The stability on heating is indeed an important quality to be expected from a standard substance, as it affords a sure method to ascertain whether the drying has been carried to completion, without running the risk of deteriorating the substance. It is indeed only after such stability tests have proved satifactory that an apparently suitable substance becomes really interesting as a prospective standard and that its energy contents are shown to be worthy of a systematic study. The stability tests now to be recorded are of two kinds: one series consists of the values obtained for the loss in weight experienced by two lots of pellets of the acid, of about 4 g. each (previously dried, when still in powder form, for four days at the ordinary temperature and then for 6 hours at a temperature in the neighborhood of 45°C.) on heating in vacuo over fresh phosphoric anhydride, to gradually increasing temperatures; the other series was obtained by determining the heat of combustion of various specimens of succinic acid after having subjected them under greatly reduced pressure to increasing temperatures. These data have been collected in tables 3 and 4. It will be seen a t once from the consideration of table 4 that the heat of combustion observed does not show any characteristic variation when the sample of succinic acid is gradually raised from the ordinary temperature to a temperature in the neighborhood of 130"C., the heating being combined with a reduction of pressure to about 0.1 mm. of mercury. All of the results collected in tables 1, 2, and 4 prove conclusively that: (1) Succinic acid can be completely dried by heating for 6 hours in a vacuum at a temperature of 45 h5"C. (2) The volatility of succinic acid is very small, even when it is heated in vacuo. (3) A statement recently published and according to which succinic acid, when heated in vacuo to a temperature of 55"C., would give an appreciable proportion of succinic anhydride is not justified. Indeed it can easily be calculated that the anhydrization of 1 gram of succinic acid to the extent of only 1 per cent would bring about an increase in the heat of combustion of 6.7 calories. No such change has obviously taken place during the heating; on the contrary, the mean values corresponding to the specimens which have been heated to temperatures ranging between 62°C. and 90°C. are, if anything, smaller than those corresponding to the temperature of 45°C. only. Again, a portion of the acid (after recrystallization from conductivity

720

L. J. P. KEFFLER

water) was dried for 7 hours under atmospheric pressure at a temperature of about 13OOC. (sample A) and another portion of the same sample to TABLE 3 Loss in weight of succinic acid (in pellet f o r m ) , on heating in vacuo (in presence of phosphoric anhydrids) to the various temperatures recorded

I

I

LOSS I N W E I G H T O F P E L L E T S

grama

w

mg

mu

3 8853 3 5610

02 0 3

0 5 0 2

0 4 0 4

TOTAL LOSS I N W E I G H T

11 0 9

2 5 28

TABLE 4 Isothermal heat of combustion of succinic acid per gram weighed in vacuo, after heating in steps u p to 130°C. TEMPERATURE

HEATING I N T E R V A L

[EAT O F COMBliBTION/

MEAN

SAMPLE ~

degrees C .

hours

caLls

1

45 f 5

6

3025.2 3024.6

45 It 5

12

3025.6 3026.3

3025.9

62 i 3

6

3023.9 3024.2

3024.2

68 f 2

4

3023.9 2023.7

3023.8

70 i 2

6

3024.3 -

90 It 10

2

120 i 10

4

3021'5* 3023.5

3025.1 -

1)

I] 1)

3024.9

3024.3 3023.5*

3025.1

General mean for these preliminary combustions* = 3024.6 cal.16;mean error for mean = 0.09 part per thousand. For the tenth of these combustions, it was observed that the bomb was slightly leaking at the end of the experiment, thus explaining the low- value (3021.5); this result will consequently be neglected.

about 14OoC.for a period of 6 hours (sample AI), with the following results for the heats of combustion:

STANDARD SUBSTASCES FOR BOMB CALORIMETRY. V I

721

Heal of combustion

Sample

A, dried a t 130"C., no phosphoric anhydride. . .

'

'

A,, dried at 140"C.,with phosphoric anhydride

3024 'j3024 8 (mean) 3024 7 / 3 0 2 5 ' 0 ) 3 ~ 2 5 , 2(mean) 3025.4

However, on heating sample A for 3 hours a t about 172"C., and subsequently to a still higher temperature, until the whole of the sample had melted, there was an unmistakable formation of succinic anhydride, as shown by the following values for the heat of combustion of these heated samples: H e a t oJ combustion

Sample

A', heated at 172"C,

3036'01 3035.6 (mean) 3035.2j

heated above melting point , , . . . . . . . . . . . . . . . . . . 3094

Anhgdride fomaatiorr

1 . 4 per cent

rl",

10 per cent (approximately)

111, CONFIRMATORY COMBUSTIONS WITH SAMPLES O F SUCCIKIC ACID OBTAINED BY RECRYSTALLIZATION FROM SEVERAL SOLVENTS

The results from the combustions recorded up to the present have been obtained under somewhat varying conditions (namely with different amounts of substance and cotton burnt, Le., for a different range of temperatures than the usual one, and also in connection with a small gradual variation in the value for the water equivalent); these differences will appear clearly from the comprehensive tables (7 and 8) summarizing the whole of the calorimetric data obtained with succinic acid. At this juncture it appeared advisable to check under more standardized conditions the results collected so far (Le., with succinic acid obtained by recrystallization from water), and also to find out whether the substitution for water of solvents with a lower boiling point,3 but possessing at the same time a heat of combustion per gram much larger than that of succinic acid,* would affect them in any way. 1. Samples recrystallized from water

The starting material for the new combustions was sample A (dried under ordinary pressure at a temperature of approximately 130°C. for 7 hours). This sample was now heated in a vacuum for a period of 7 hours a t a temperature of 52 i 2°C (without phosphoric anhydride); sample B 3 Thus more easily got rid of and with less danger of decomposition for the substance. 4 So that the presence of the merest trace of such solvents would quite easily reveal itself through a variation jn the heat of combustion, this time however of opposite sign to that which might have been expected from the presence of traces of water.

722

L. J. P. KEFFLER

obtained in that way was then heated in vacuo for 8 hours at a temperature of 80 f 2°C. (in the presence of phosphoric anhydride), giving sample C; samples B and C wexe now burnt in duplicate directly after heating, and in the case of the latter the combustions were repeated after the pellets had stood for one month over phosphorus pentoxide (see sample C'). On the whole the results show a very good agreement amongst themselves: those obtained with samples dried in the presence of phosphoric anhydride (see samples AI and C) are higher by a few tenths of one calorie than those (relative to A and B) obtained in the absence of a drying agent; the difference, amounting to about 0.5 calorie in both cases, is however so small as t o TABLE 5 Isothermal heat of Combustion o j succinic acid, per gram weighed in vacuo, and dried jar different lengths of time at the temperatures indicated SAMPLE

DRYINQ AQENT

degrees

A

ea. 130

A1

ca. 140

c.

HEAT OF COMBUSTION

nam.

ca1.u

"7.

760

3024' 3024.8 (mean) 3024.7)

760

,

B

52 i 2

0.15

C

80 i 2

0.14

PZOS

3024'9\ 3025.2 (mean) 3025.51

P206

C'

I

I

3025' 3\ 3025.8 (mean) 3026,31 3025.9 3025.4 (mean) 3025'0)

General mean for the ten combustions = 3025.3 caL16 per gram in vacuo. Mean error for mean = 0.05 part per thousand.

have only little significance. The final results obtained with succinic acid recrystallized from conductivity water have been collected in table 5. 2. Recrystallization from solvents other than water

The choice of a suitable solvent for succinic acid is very limited even if no additional conditions were required, as there are in the present case. Ether was selected in spite of its small solvent power for succinic acid on account of its large heat of combustion (8805), its very low boiling point, its inertness, and also its availability in a pure state. The small sample wl--derived from sample C and thus completely dry-was recrystallized, from ether (A.R.) and then entirely freed from traces of solvent by heating in a vacuum (0.15 mm.) for 4 hours at a temperature of 80 f 2°C.

STANDdRD SUBSTANCES FOR BOMB CALORIMETRY. VI

723

Another portion of the same sample C was recrystallized from purified acetone (boiling between 55.8 and 56.OOC.) which was selected as the second solvent for practically the same reasons as those given with reference to ether, with the useful difference that the solvent power of acetone is much more suitable than that of ether with respect to succinic acid. By heating the last lot of crystals for 5 hours in vacuo a t a temperature of 77 =t2"C., a sample w2 was obtained. The heat of combustion found for that sample (3026.5and3027.1;mean = 3026.8)washoweverlarger by 1.5calories than the mean just recorded for the ten final combustions with the samples prepared by recrystallization from water. This small discrepancywas probably due to the presence in thesamplewzof a faint trace of acetone. It can be calculated indeed that, owing to the relatively large heat of combustion of acetone (7351),it would suffice that 1 g. of succinic acid would contain only I mg. of acetone (i.e., one part per thousand) in order to increase the heat of combustion of the contaminated sample by 1.3 calories; this is just about equal to the divergence observed. As, however, the small discrepancy might have resulted from a slight measure of purification having taken place as the consequence of the substitution of acetone for water, it was considered worth while to repeat the crystallization from acetone with a fresh sample of succinic acid. The crystals from this new recrystallization were first dried for one night over filter paper in the air, then heated for 6 hours in vacuo to a temperature of 80 & 2"C.,with alternate crushing. The value now obtained for the heat of combustion of the new sample to2' (3025.6 and 3026.3; mean = 3025.8) being piactically identical with the general mean (3025.3) obtained previously, it was clear that the slightly too large value found for the heat of combustion of the last sample but one (from the first recrystallization with acetone) was really caused by the presence of a minute trace of acetone in the sample. As a final confirmatory test for this statement a further recrystallizabion of the last sample was carried out, the crystals being heated in vacuo for 5 hours at a temperature of 83 rt 2OC.; the new sample (a") was now used for a last combustion, the result of which (3025.5) was entirely in agreement with the general mean determined previously. The results obtained for the various combustions with samples derived from recrystallizations with ether or acetone have been collected in table 6. It can be seen that this general mean, obtained from the five combustions carried out with samples of acid recrystallized from solvents other than water, is exactly the same as that (3025.3) obtained previously when the solvent was water. The mean of means (for the whole of the twenty-six combustions carried out successfully with fifteen different samples of succinic acid) is equal to 3025.0 calorieslh: it differs only by +0.4, -0.3, and -0.4 calorie from the means referring respectively to the eleven preliminary and to the ten final combustions corresponding to solvent water,

724

L. J. P. KEFFLER

and to the five combustions corresponding to solvents ether or acetone. It can thus safely be concluded that the heat of combustion of succinic acid, per gram weighed in vacuo, and burnt isothermally at the initial temperature of 20°C. is genuinely equal to 3025 f 1 calories:e. IV. TECHNIQUE

1. Apparatus An adiabatic calorimeter with a stirrer of the propeller type was used in conjunction with a Landrieu-Malsallez bomb (with gold lining and platinum electrodes) ;the jacket, of the submarine type, was heated electrolytically by means of a current of about 25 amperes (at 230 volts). The temperature rise was measured with a Beckmann thermometer having a range TABLE 6 lsothermal heat o j combustion of succinic acid p e r gram i n vacuo, after recrystallization from ether (once) and acetone (twice) SOLYEN?

Ether, Acetone (once). . . . . . . . . . . . . . . . . . . . . .

Acetone (twice). . . . . . . . . . . . . . . . . . . . .

.I 83

& 2

1

5

1 3025.5

General mean = 3025.4 caLI6;mean error for mean = 0.1 per thousand.

.

of about 5°C. (calibrated by the P.T.R. a t each 0.1 division to the nearest thousandth of one degree). 2. Method

The "constant range method" first proposed by the author in 1926 (2) was usually adhered to. The normal range of 2°C. used in former work was again selected, starting from the same initial temperature of 19.7"C. In order to meet in advance the possible criticism that by using another range the results might be somewhat different, a larger range (2.6"C.) was used as well. V . GENERAL INFORMATION; CALIBRATIONS ; MAIN FEATURES O F T H E

COMBUSTIONS WITH SUCCINIC ACID

1, General information In order to comply with the recommendations just adopted by the Commission Internationale Permanente de Thermochimie at a meeting

STANDARD SUBSTANCES FOR BOMB CALORIMETRY. V I

725

held in Paris in April, 1933, the following data have been collected hereunder : (a) The mass of substance burnt was approximately 1.35 g. of benzoic acid and 2.80 g. of succinic acid for the range of 2"C., against 1.75 g. of benzoic acid and 3.65 g. of succinic acid for the range of 2.6"C. (b) The weight of auxiliary substance (cotton thread or wool) varied between 3 and 6 mg. (c) The mass of water introduced into the bomb was 1 g. in every case. (d) The oxygen was introduced under a constant pressure of 35 atmospheres ( + 5 per cent). (e) The bomb had a capacity of 450 cc. (f) The initial temperature was 19.7"C. in every instance, while the final temperature was either 21.7"C. or 22.3"C., according to whether the range selected was 2°C. or 2.6"C. The oxygen used in connection with all the combustions reported in this paper was manufactured (in the works of the British Oxygen Company Ltd.) by fractional distillation of liquid air. As will be shown elsewhere oxygen derived from that source contains traces of combustible impurities which are exceedingly difficult to detect (owing to the extreme precision required to bring this cause of error to light in a conclusive way), so much so that they do not make their influence felt to the same extent all through the life of the oxygen cylinder. This important and curious phenomenon has been studied very closely by the author during the last three years, using several different cylinders of oxygen. Its discussion and interpretation will form the subject of a separate paper to be published shortly. 2 . Calibrations

In order to establish once and for all for the water equivalent a fixed value of reference relative to absolutely pure oxygen,j a special calibration had been carried out four years ago with a sample of the preparation No. 39c of benzoic acid supplied by the Bureau of Standards, using pure oxygen (purified by slowly passing the gas over palladiumized asbestos raised to a red heat); the true constant thus obtained for the water equivalent, after reduction to an isothermal value, was 4277.1 caloriesls. The data relative to that calibration have been collected in table 7, combustions 1 to 5. Since however the combustions of succinic acid were to be carried out with a newly refilled cylinder of oxygen, it was necessary either to purify this oxygen (by preheating it under suitable conditions prior to filling the bomb) or to carry out a special calibration with it. Since the former operation is very tedious and troublesome, the latter alternative was preferred. Furthermore, since the range of 2.6"C. had not been investigated before and since, on t.he other hand, some doubt had been expressed recently from a continental laboratory as to the true identity of successive preparations of Thus allowing one, by comparison with the value found for the water equivalent obtained with some other oxygen, to estimate the approximate purity of the latter.

TABLE 7 Water equivalent (at the temperature of 80'C.) for pressure o j 36 atmospheres in the bomb and for various pressures in the cylinder of oxygen (in cal.la and in vacuo)

I

CORRECTIONS I N CALORIES

FOR EQUIYALENT

1. (1929) Calibration with preheated oxygen for range of 8°C. (mean error for mean = 0.11 per thousand

1 1.34062.03734.0205 139 -5 103 1.9845 8.15 10.2 39c - 4277.4 1.3444 1.9517 3.9447 140 1 123 1.9921 - 19.4 - - 4278.0

2 3 4 5

1.34321.92143.9178 - -1681.9901 8 . 4 1.34581.91353.9095 - -122 1.9953 8 . 6 1.34521.92443.9267 - -12001.9918 7.7

15.0 19.0 13.8

- - - -

!

4276.3 4277.1 4275.6 4278.0)

2. (1933) Calibration with new and non-preheated oxygen for range o,f 83°C. = 116 t o 113 atmospheres and 112 to 109 atmospheres. General mean (for combustions 6 to 13) = 4277.1 with mean error for mean = 0.06 per thousand

A. For pressure in cylinder

6 7 8 9

1.7550 1.9031 4.4861 181 0 492.5962 1 . 8 12.t 48a 116 4276.3 ) 1.75461.90594.4908 - - 772.5953 1.1 - - 115 4276.5 /4277,0 1.7543 1.9962 4.5810 - - 86 2.5943 2.2 - - 1144277.8 1.75311.90784.4935 - - 972.5941 1 . 6 - - 113 4277.4 4277.1

_ _ I -

10 1.7533 1.90704.4909181 11 1.7533 1.9021 4.4843 12 1.75241.90584.4847 13 1.75391.90174 4826 -

B. For pressure in cylinder 14 15 16 17

0 -

862.5934 712.5932 482.5922 582.5932

I

2.0 1.9 1.6 1.7

12.E 39b 112 4276.8 - - 1114277.0 4277,2 - - 1104276.5 - - 1094278.7J

93 to 87 atmospheres. per thousand

=

Mean error for mean = 0.09

1.7521 1.9036 4.4821 181 0 45 2.5921 1.5 15.0 39b 1.75301.89414.4738 - - 382.5940 1.1 - 1.74871.90924.4821180 1 482.5862 0 . 9 12.6 1.7533 1.89654.4766 181 0 51 2.5931 1 . 8 13.0 -

93 4276.7 884277.0

3. (1933) Calibration with same oxygen as in $ 2 , but for range o j 8.0"C. A. For pressure in cylinder = 66 to 64 atmospheres. Mean error for mean = 0.08 per thousand 18 19 20 21

-

1.34281.89733.8743138 0 1.34131.89763.8723 - 1.34561.89793.8798139 1.34661.89963.8826 - -1

B. For pressure in cylinder

=

421,9866' 441.9841 481.9910 531.9915

1.4 1.6 1.5 1.7

1

I

7 . 8 39b 8.2 7.0 6.6 -

54 to 39 atmospheres. per thousand

664275.1 654275.4 - 4274.2 644275.7)

I

Mean

= 0.02

39 4275.3 General mean (for combustions 18 to 24) = 4275.2 cal.>bwith mean error for mean= 0.04 per thousand 726

STAXDARD SUBSTANCES FOR BOXB CALORIMETRY, V I

727

standard benzoic acid issued by the Bureau of Standards, it was considered desirable to carry out two independent calibrations for the same range of 2.6"C., one with the preparation 39b and the other with the preparation 48a. The combustions made with the latter (see table 7, Nos. 6 to 9) gave for the water equivalent the value 4277.0, while with the former the value 4277.2 was obtained (see Nos. 10 to 13 of the same table). The mean for these eight combustions is 4277.1; it is thus identical with the value found four years earlier wth the same calorimetric system but for a range of 2°C. and with purified oxygen. This striking agreement establishes a very strong probability in favor of the purity of the oxygen used for the present research and at the same time proves that the preparations 39b, 39c, and 48a of standard benzoic acid supplied a t various times by the Bureau of Standards are really identical, at least in so far as their energy contents are concerned. But for the reason already stated that the value for the water equivalent changes appreciably as a rule when the pressure in the oxygen cylinder decreases, five more series of control calibrations were carried out for different pressures in the cylinder. Three out of these were especially carried out in connection and simultaneously with the actual combustions of succinic acid; they have for that reason been introduced in this paper (see table 7, Nos. 14 to 24). It will appear from a consideration of the last column of table 7 that in the particular case of the oxygen used for this investigation there has been a gradual decrease of some 2 calories as the pressure in the cylinder dropped from 116 to 35 atmospheres. 3. Combustions of succinic acid

Simultaneously with these calibrations, three series of combustions were carried out with succinic acid under conditions as identical as possible,. in particular with respect to the two selected ranges of 2.0"C. and 2.6"C. The data relative to these experiments have teen collected in table 8. VI. GENERAL CORRECTIONB

I , Reduction to vacuo of the weights of substance burnt (a) The density of benzoic acid was taken as equal to 1.266 (cf. International Critical Tables, Vol. I, p. 77), so that the corrective terms to be added to the weights in air (1.75 and 1.35 g.) of benzoic acid required for the two ranges of 2°C. and 2.6"C., are 1.4 and 1.1mg., respectively. (b) The density of succinic acid was taken as equal to 1.56 (same reference), so that the corrective terms for 3.65 and 2.80 g. of this acid are 2.3 and 1.77 mg., respectively. (c) The weights of cotton were so small that the vacuum reduction term was quite insignificant.

TABLE 8 Isothermal heat of combustion (at constant volume) per gram of succinic acid (weighed in vacuo) at a temperature of bO"C., expressed in ca1.16

1

T E M P E R A T U R E CORRECTIONS (CORRECTED I N TEN T H O U SANDTHS FORBORE)

! 1 99

CORRECT I O N S I N CALORIESFOR

1

-~ _ _ _ _ _ _ _ _ - - _ _

25 3.6545 1.9067 4.4826 180 1 51 2.5889 0 . 7 20.6 a! 26 3.6570 1.9061 4.4858 - 0 452.5932 14.9 19.4 -

--1-I I 98 3025.2 1 3024, 4276.6 97 3024.61 -

273.65951.90084.4802 - - 422.5932 1 . 0 22.0 p 283.65911.90494.4849 - - 452.5935 0.9 - -

95 3025.6) 3025,9 94 3026.3 -

29 3.65961.90844.4867 - - 482.5915 0 . 8 20.8 30 3.6590 1,90884.4861 - - 43 2.5910 0.6 19.4

~

-

31 3,6621 1,90664.4896 181 1 932.5919 1 . 6 14.2 6 32 3,5987 1.9061 4.4398 177 0 452.5469 0 . 8 - -

-

33 3,6581 1.8970 4.4721 180 1 392.5893 1 . 2 - 6' 34 3.6584 1.89184.4769 - 0 452.5886 0 . 7 13.4 E - 3812.5834 0 . 9 13.0 p 35 3.6492 1.8965 4.4657

-1

362.80581.90103.8819138 0 661.9881 0 . 6 114.6 A 372.80621.89983.8820 - - 761.9884 0 . 4 15.0 -

61 3024.8) 3024,84275.2 60 3024.7 -

38 3.6589 1.8987 4.4756 180 - 43 2.5906 1 . 2 14.2 Ai 393.65651.89734.4730 - - 452.5892 0 . 8 - -

-

402.8051 1.94883.9284138 1 61 1.9874 0 . 7 13.4 B 412.80461.90173.8819 - 0 631.9877 0 . 5 14.6 -

63 3024'g) 3025.2 62 3025.5

422.80701.90493.8871 - - 681.9892 0 . 6 14.2 C 432.80361.90013.8808 - - 701.9875 0 . 9 - -

57 3025'3 3025.8 56 3026.3

442.80671.89963.8808 - - 601.9890 0 . 7 15.4 C' 452.80671.90203.8827 - - 491.9896 0 . 6 - -

42 - 3025'0:/ 3025.9 3025.4

462.8068 1.8988 3.8786 138 1 48 1.9889 0 . 7 15.0 wi 472.80801.89863.8789 - 0 451.9896 0 . 6 - -

-

-

482.80631.89943.8818 492.80541.90223.8868

---

502.80411.89873.8791 - - 1

69 3025.0 3025.4

)

3025,

-

-

)

-

-

47 3024.9) 3024.7 4275.2 3024.6

721.9890 0 . 8 14.6 941.9890 1 . 0 15.0

w'g

-

40 41 3025'6} 3026.3 3025.8

-

651.9876 0 . 9 15.4

0''s

39 3025.5 3025.5

-

STANDARD SUBSTANCES FOR BOMB CALORIMETRY. VI

729

2. Reduction of the heat of combustion of succinic acid to i t s isothermal value

In a paper which will appear shortly elsewhere, the author has shown that, for a substance of gram-molecular weight M and heat of combustion (per gram) &, and whose empirical formula is C,H,,O,Na, the corrective term to be added to the value observed for the watsr equivalent (when determined as per international agreement by means of standard benzoic acid, whose heat of combustion is taken as 6319 caloriesl5 per gram) in order to give it the value which corresponds to an isothermal combustion for succinic acid is given by corr

Is

=

(-

1.7n

+ 7.75b + 2 5(c + d ) .-1 - 'A2)W.At . i

Q

631

where 4 t represents the rise in "C. and W stands for the water equivalent in calories per degree. Hence, for succinic acid (;M = 118; Q = 3025 calories per degree) whose formula is C4HC04, the corrective term (to be added to the water equivalent) is either 1.98 - 0.92 = 1.06 calories per degree (for At = 2.6"C.) or 1.5 - 0.7 = 0.8 calorie per degree (for 4 t = 2°C.). 3. Correction f o r the temperature coeflcient of the heat of combustion

The correction is nil in this case, since the value (6319) used for the heat combustion of benzoic acid refers to an isothermal combustion at the temperature of 20°C. and both series of combustions with either succinic acid or benzoic acid were carried out at practically the same initial temperature (19.7"C.).

4. Correction for the formation of natric acid For work of the highest precision it is necessary to reduce to as low a value as possible the correction for the formation of nitric acid, owing to the fact that this correction is always somewhat ill-defined. This result can be achieved by making use exclusively of nitrogen-free oxygen (or at least oxygen containing only very minute traces of nitrogen) and by sweeping the air out of the bomb prior to its filling. When, as is the case here, the volatility of the substance under investigation is very small, the same result may be obtained more easily by evacuating the bomb for a very short while under well standardized conditions. For the actual calculation of the correction allowing for the formation of traces of nitric acid, the heat of formation of the latter under the conditions of dilution normally prevailing in the bombwas taken as equal t o 230 calories per gram of acid (formed from N 0 HzO).

+ +

5 . Correction for the heat evolved b y the combustion of the auxiliary substance used for i g n i t i o n

For the present research, both cotton wool and cotton thread were used. Two auxiliary combustions were made in order to determine the heat of

730

L. J. P. KEFFLER

combustion of the two forms of cotton used for ignition: for the wool, the heat of combustion was found to be 3954 calories per gram, while the value 3990 calories was obtained for the thread. 6 . Corrections for incomplete combustion

Tests for complete combustion were carried out in a certain number of cases taken at random. The residual gases when bubbled at a very low7 speed through palladium chloride solution failed to indicate the presence of the merest traces of carbon monoxide.

7 . Correction.for the Joule effect in the ignition wire The platinum wire used for the present work was exceedingly fine (via., 0.03 mm. in diameter); the time required to fhse the length of wire (about 1 cm.) stretching between the two platinum electrodes was so small as to make the corrective term for the Joule effect entirely negligible. VII. COMPARISON BETWEEN T H E RESULTS OBTAINED FOR THE HEAT O F COMBUSTION OF SUCCINIC ACID BY T H E AUTHOR AND OTHER OBSERVERS

Stohmann, Kleber, and Langbein (6) found for the heat of combustion of succinic acid the value 3022 caloriesl6, with a maximum deviation of 1.1per thousand. This result is only smaller by about 3 calories than that of the author, the difference being remarkably small considering that the work of the German pioneers was done nearly half a century ago! Louguinine (5) found two values, differing by 12 calories from one another, the highest of the two being still lower than that of the author by 10 calories. These results of Louguinine, like much of his work are characterized by a large maximum deviation (4.2 per thousand in his latest series of combustions with succinic acid) and have no more than a historical value ! Verkade, Hartmann, and Coops (9), on the contrary, made some trustworthy determinations-unfortunately only four in number-with two different samples of succinic acid; the general mean for their four combustions was 3026.9 calorieslB,with a mean error of the mean of 0.25 per thousand and a maximum deviation from the mean of 0.75 per thousand. As already observed on page 4, the value of Verkade is higher than that found by the author by only 1.5 calories, in spite of the fact already mentioned that his fourth value (3029.1) is abnormally large and stands in sharp contrast with the three others, which are identical (3026.1). It can therefore be safely concluded that Verkade’s value (3026 to 3027) is as near to that of the author (3025 f 1) as can reasonably be expected, especially when it be remembered that the Dutch investigators used a different method (the non-adiabatic !) and worked otherwise under slightly different conditions,

S T A S D A R D SUBSTANCES FOR BOMB CALORIMETRY. V I

73 1

More recently still, Beckers (10) found for various speciniens of succinic acid values which vary between 3024.4 and 3027.4 caloriesls. The final value given in the present paper (3025.0 =t1) is thus lying well within that interval. VIII. SUCCINIC ACID POSSESSES THE MULTIPLE QUALITIES WHICH RENDER

IT SUITABLE AS A CALORIMETRIC STANDARD

The conditions which succinic acid will have to fulfill in order to be acceptable as a calorimetric standard will now be considered in turn. 1. It should be easily obtainable in an exceptioiaally p u r e state

Succinic acid will have no difficulty to establish its claim in that respect as mill appear from a rapid consideration of the specifications given by the British Drug Houses for their A.R. preparation. The maximum limits of impurities allowed are: 0.05 per cent of ash (on 1 g.); 0.003 and 0.005 per cent of chlorides and sulfates, respectively (on 1g. of acid in 20 cc. of water). Actually, the tests carried out by the author for ash, first on 1 g. and afterwards on 5 g., were entirely negative, as well as the tests for chlorides and sulfates. It may also be recalled here that the successive recrystallizations of the original preparation, first from water and then from ether and acetone (the latter solvent being used twice), failed to show any significant variation in the value for the heat of combustion, thus giving much support t o the claiin of absolute purity of the substance.

2. It should be easy to dry and should remazn so without requiring special precautions for keeping It has been mentioned already that succinic acid, even when recrystallized from water, could easily be obtained in a perfectly dry condition, either by heating to a temperature of 130 to 140°C. in air or to a temperature of about' 50°C. in vacuo. The proof that succinic acid is not at all hygroscopic is given by the following results, obtained by allowing two pellets of about 1 g. to stand over night in a moist atmosphere such as that provided by a desiccator containing a dish full of water. The gain in weight was only 0.01 and 0.00 per cent for the two pellets respectively, thus giving an average gain of 0.005 per cent in 24 hours; this is just within the experimental error relative to weighings. 3. I t should be perfectly stable in air, even at temperatures above 100°C., so as to allow of thorough drying

The proof that this condition is completely fulfilled by succinic acid lies in the fact that the heat of combustion of succinic acid does not change

732

L. J. P. KEFFLER

measurably when it is heated to 140°C. (for 6 hours) under atmospheric pressure or to about 120°C. (for 4 hours) under a pressure reduced to approximately 0.1 mm. of mercury.

4, I t should not be volatile It has been found that samples of about 4 g. of succinic acid, either in the form of finely crushed powder or in pellets, lost only one part in 10.000 after being left for a fortnight in an evacuated desiccator over phosphoric anhydride. The volatility of succinic acid, at ordinary temperature and even in vacuo, is thus exceptionally small. 5 . I t should be easily compressible into f i r m pellets

Whether the pellets of succinic acid were made after a preliminary grinding to an exceedingly fine powder or whether they were prepared directly from the larger crystals, there never arose any difficulty during the removal of the pellets from the pellet press or during subsequent manipulations. Dozens of them were seized with the forceps t o and from the balance pans without any loss of material being ever measurable; some were placed in the crucible inside the bomb and oxygen admitted swiftly into the latter, without any loss of weight being experienced. In short the pellets of succinic acid behaved in all respects quite as satisfactorily as those made with benzoic acid. 6. I t should be easily injlammable

The fact that some difficulty was experienced a t the beginning of the first series of combustions when cotton wool was used for ignition purposes cannot be considered, in the authors’ opinion, as an argument in favor of the contention that succinic acid may be somewhat deficient from the point of view of its inflammability, as has been stated about some other substances (sucrose in particular!) which possess only a small heat of combustion per gram, Indeed, succinic acid can be ignited with great ease when a suitable igniter is used, and it is not necessary t o employ an unduly large quantity of igniter, since less than 4 mg. of thread was found to be always quite sufficient. 7 . I t must be of such a nature as will favor complete combustion

As already mentioned, all tests for carbon monoxide gave negative results; nor were deposits of unburnt carbon ever left behind. As a corollary to the last condition, the substance should obviously not possess the tendency to project small particles of its own substance out of the crucible, before burning. The excellent agreement obtained between individual results for the heat of combustion of the acid from some thirty

STANDARD SUBSTANCES FOR BOMB CALORIMETRY. V I

733

combustions proves quite peremptorily that such an eventuality is not likely to arise with succinic acid. IX. CONCLUSION

Succinic acid has been shown to fulfill to a high degree all the conditions which may reasonably be required from a calorimetric standard: in particular, it has given for the heat of combustion of fifteen different samples values which show a remarkable agreement amongst themselves. Even when the values of different observers are considered, there is undoubtedly a definitely better agreement between the modern ones as there is in the case of the other prospective auxiliary standard, namely, salicylic acid : the first is markedly more stable on heating than the latter and consequently much easier to dry thoroughly and without fear of decomposition; further, the purity of succinic acid may always be tested by way of an accurate titration, which is not the case with salicylic acid. In conclusion, the author proposes to adopt formally succinic acid as a secondary calorimetric standard for bomb calorimetry and t o accept for its isothermal heat of combustion at the temperature 20°C. the value 3025 i 1 caloriesl&(per gram weighed in vacuo). The author desires to express his sincere thanks to Professor E. C. C. Baly, F. R. S., Head of the Inorganic Department, for laboratory facilities generously provided for the present investigation. REFERENCES (I) BECKERS:Bull. soc. chim. Belg. 40, 571 (1931). (2) KEFFLER:J. Phys. Chem. 31, 61 (1927); Rec. trav. chim. 49, 434 (1930), table 1. (3) KEFFLER:J. chim. phys. 28,463, 465 (1931). (4) LAMOUROUX: Compt. rend. 128, 999 (1899). (5) LOUGUININE: Compt. rend. 107, 597 (1888)) Ann. chim. phys. [6] 23, 196 (1891). (6) STOHMANN, KLEBER,AND LAKGBEIK:J. prakt. Chem. [2] 40, 202 (1889). (7) SWIETOSLAWSKI: Bull. soc. chim. 37, 84 (1925). (8) VERKADE A N D COOPS: Rec. trav. chim. 46, 378 (1926). (9) VERKADE, HARTMANN, A N D COOPS:Proc. Acad. Sei. Amsterdam 27,859 (1924).