By W. H. KEESOM. A. BIJL and Miss H. VAN DER HORST. (Comm. No. 2I7a of the Physical Laboratory. Leiden.)

Size: px

Start display at page:

Download "By W. H. KEESOM. A. BIJL and Miss H. VAN DER HORST. (Comm. No. 2I7a of the Physical Laboratory. Leiden.)"

Jerome Moore
5 years ago
Views:

Physics. - Determination of the boiling points and the vapour-pressure curves of normal zydrogen and of para-hydrogen. The normal boiling point of normal hydrogen as a basic point in thermometry.

1 Physics. - Determination of the boiling points and the vapour-pressure curves of normal zydrogen and of para-hydrogen. The normal boiling point of normal hydrogen as a basic point in thermometry. By W. H. KEESOM. A. BJL and Miss H. VAN DER HORST. (Comm. No. 27a of the Physical Laboratory. Leiden.) (Communicated at the meeting of November ) 1. ntroduction. When BONHOEFfER and HARTECK 1) had shown. that the ordinary hydrogen is a mixture of ortho- and para-hydrogen. and also indicated a way to prepare almost pure para-hydrogen. it became of interest. to determine exactly the differences between the two kinds of hydrogen. Also it should be examined. if the boiling point of normal hydrogen (i.e. a mixture of para- and ortho-hydrogen in such proportions that it is in equilibrium at room temperature). is suited as a basic point for the temperature scale. Already in December 1929 we made preliminary measurements with the aid of Pt-thermometers which had been calibrated with the helium thermometer. t proved to be better. however. to repeat these determinations with the helium thermometer. 2. Preparation of the hydrogen. The norm al hydrogen used in the vapour-pressure apparatus was obtained by evaporating liquid hydrogen. The hydrogen had been in the liquid state for no longer than an hour. and kept at room temperature some days after that. so that we were quite sure that the hydrogen had its normal composition. The para-hydrogen was prepared after BONHOEFfER and HARTECK'S method 1) by adsorption of normal hydrogen at the temperature of liquid hydrogen. The composition of the mixture of para- and ortho-hydrogen was determined by measuring the heat conductivity after SCHLEERMACHER'S method. We found it advantageous to fix the current to a certain value, and then to measure the resistance of the tungsten wire. which was mounted in a glass tube. immersed in liquid hydrogen. The resistance of the wire was, in normal hydrogen Q. in hydrogen that had been absorbed the resistance decreased. and reached at last the value Q. We accepted. that the transformation was complete then. We used first chabazite to adsorb the hydrogen. To reach the minimum resistance of the wire. the hydrogen had to be adsorbed during 2 h 45 min. We replaced the chabazite by charcoal. but it did not increase the velocity of the transformation. When using norit. we reached a smaller time of transformation. namely h40 min We did not succeed in finding a coal. that 1) K. F. BONHOEFfER and P. HARTECK, Z. Phys. Chem. B *

2 1224 was so activated, that the trans format ion took place in so short a time, as BONHOEFFEH and HARTECK indicate (20 min. ). The para~hydrogen was prepared on the day of the measurements, or on the day before. The composition was always checked by means of the thermal conductivity. n some cases the hydrogen still contained a few procents of ortho~hydrog en. f so, a correction was applied. 3. Thermometers and vapollr~pressllre apparatus. The measure~ ments were made successively with two helium thermometers, one with the M A B Fig. 1.

3 1225 reservoir Tg (355 cc.) and one with the reservoir T 4 (108 cc.). These thermometers have been described already 1 ). Both the thermometers had a new filling, and the fundamental pressure coefficient was determined again. The result was for both thermometers: anhe = ). The differential vapour pressure apparatus (Fig. 1) may be described as a combination of two vapour pressure apparatus of the ordinary Leiden pattern ::: ) so that they have one leg of the manometer (M) in common. n this way the vapour-pressure difference of two substances at the same temperature is measured more accurately. The reservoirs were from 10 to 25 cm below the level of the liquid hydrogen of the bath. By a double stirrer uniformity in temperature of the liquid was established. n consequence of the small density of the hydrogen temperature differences, that could arise by hydrostatic pressure-differences, did not come into account. 4. Results. The results of the measurements with the large helium thermometer can be represented by the formulas : For normal hydrogen: t = (p - 760) X 10-6 (p - 760)2. () and for para-hydrogen : t= (p-760)-4.4 X 1O- 6 (p-760)2. (11) n table we give the results of the measurements. The two last columns give the differences between the observed temperature and the temperatures derived from the formulas and J. The measurements with the small helium thermometer are represented best by lhe following formulas: normal hydrogen: t = (p - 760) X 10-6 (p - 760)2 (a) para-hydrogen: t = (p - 760) X 10-6 (p - 760)2 (la) Table contains the results. ) W. H. K EESOJ\\ and Miss H. VAN DER HORST, Comm. Leiden N. 1B8a. These Proceedings ïO, ) This value differs somewhat from that, published in Comm. No. lbba. The cause of th is difference has not yet been sufficiently investigated. A later communication will deal on new measurements of th is pressure coefficient. :ti Cf. H. KAMERLNGH O NNES and C. BRAAK. Comm. Leiden No. 107a. These Pro.: , 333. Cf. further: G. H O.ST. Comm. Leiden No. 144c.

4 1226 TABLE. Vapour pressures of normal and of para-hydrogen. Date NO t lp (normal H2) 1 p (para H2) 1 t -t 11) oe mm Hg. mm HJ. ob. tob.-t (11) i O.OOi i TABLE 11. Date NO Vapour pressures of normal and of para-hydrogen. t lp (normal H2) 1 p (para H2) 1 t -t(la) t -t(la) oe mm Hg. mm Hg. ob. ob As a mean of all the measurements we take: For norm al hydrogen: t = (p - 760) X 10-6 (p - 760)2,

5 For para-hydrogen: 1227 t = (p - 760) X 10-6 (p - 760)2. The difference between the calculated temperature and the observed temperature nowhere exceeds 0.01 C. We find: Normal boiling point of normal hydrogen: of para-hydrogen: t C C. For the correction to the Celsius-Avogadro scale we used: Comparison with previous results. At the temperatures of liquid hydrogen neither the thermal expansion of the glass. nor the noxious volume of the thermometer still has a great influence on the determination of the temperature. So one should expect. that the differences that occur in temperature determination. will have their principal cause in differences of the pressure-coefficient used. At Leiden already three determinations of the boiling point of hydrogen have been done. The first determination was carried out by KAMERLNOH ONNES and KEESOM 1). with the aid of a platinum thermometer. compared with the hydrogen thermometer. As pressure coefficient anh, = was used. The result was: C. CATH 2) and later PALACOS and KAMERLNOH ONNES 3) made determinations with the helium thermometer. They used as pres su re coefficient anhe = and gave the values : c C. respectively. To compare these results with the result now obtained. to all these values a correct ion of C. has to be applied. because another coefficient of thermal expansion of the glass was accepted 4). Further the two last determinations must be recalculated with the pressure coefficient: anhe = (cf 3). The results become: KAM ER LNOH ONNES and KEESOM: C. CATH: C. PALACOS and KAMERLNOH ONNES : C. Our new determination for normal hydrogen concurs weil with them. ) H. KAMERL NG H ONNES and W. H. KEESOM. Comm. Leiden No. 137d. These Proc Cf. also H. KAMERLNGH ONNES and G. HOLST. Comm. Leiden No. 4la. These Proc ) P. G. CATH. Comm. Leiden No. 152a. These Proc ) J. PALA COS and H. KAM ERLNGH ONN ES. Comm. Leiden No. 156b. Arch. Néerland. (A) ) F. v. AGT and H. KAMERLNGH ONNES. Comm. Leiden No. 176a. These Proc W. H. KEESOM and A. BJL. Comm. Leiden No. 203a. These Proc

6 1228 At Berin two measurements had been done; by HENNNG 1) and by H ENNNG and HEUSE 2). They give for the boiling point: C. and C. During these measurements, a new determination was published by HEUSE and OTTO 3) that gives the result : C. with UnH. = Ve/ocity of the transformation of hydrogen in the /iquid state. During the measurements we could observe a decreasing of the difference in vapour pressure between normal and para-hydrogen. This is shown best by the measurements land 4 of (Tabie ) where this difference diminished with mm in 6 hours. This agrees with a velocity of transformation of 0'.5 % pro hour. We obtained a similar result, by measuring the composition of hydrogen evaporating from a vacuum glass. The composition of the vapour changed 0.4 % pro hour. We may conclude, that the velocity of the transformation of ortho- into para-hydrogen in the liquid state in a glass vessel is about 0.5 % pro hour. 7. The vapour pressure curve be/ow one atmosphere. These measurements except those on July 7th were done with the helium thermometer with large reservoir. The results are given in table 111. They agree very weil with the earlier determinations for normal hydrogen, made at Leiden 4). At the normal boiling point of normal hydrogen we find a difference in vapour pressure of 27.1 mm, in accordance with BONHOEFFER and HARTECK 5). The following formulas, proper for practical purposes, represent the results within 0.02 c.: Normal hydrogen: t = Q/og p Olog 2 p (p in cm) (A) Para-hydrogen: t = Q/ag p Olog2 p (p in cm) The differences between the temperatures observed, and those derived from A and B, are given in the two last columns of table 111. (B) ') F. H ENN NG, Zeitschr. f. Phys. 5, 264, ) F. H ENNNG und W. H EUSE, Z eitschr. f. Phys. 23, 195, ) W. H EUSE und J. O TTO. Ann. der Phys. (5) 9, 497, ') H. K AM ERL NG H O NNES and W. H. K EESO,\1. loc. cit.; P. G. CATH, loc. eit.; J. PALACOS and H. KAMERLNGH O NNES, loc. eit. 5) B ONHOEFFER und HARTEC K, loc. eit.

7 1229 TABLE 111. Date Vapour pressure of norm al and of para-hydrogen. t lp (normal H2) P (para H2) 1 t - t (A) t -t (B) oe cm Hg. cm H\}. obs ob L n table V we add some interpolated values of the vapour pressures at convenient temperatures. TABLE V. t p (normal H2) p (para H2) oe cm Hg. cm Hg Conclusion. The boiling point of norm al hydrogen can be used as a basic point. As according to 6 the normal boiling point of hydrogen in a glass vessel changes about degree per hour. we conclude that

8 1230 there is no objection a gainst using the normal boiling point of norm al hydrogen as a basic point in thermometry. to an accuracy of degree. if the boiling point is taken in a glass vessel less than 6 hours after liquefaction (provided one starts from normal hydrogen). Moreover care has to be taken for the total absence of such substances. that might catalyse the transformation. We express here our cordial thanks to Mr. B. G. DAMMERS and Mr. J. M. H. VAN MERLO. phil. cand.. for their valuable assistance at the measurements. Summary. A new determination of the normal boiling points of normal and of parahydrogen with a helium thermometer is given. The result is: normal boiling point of normal hydrogen: c.. 0 normal boiling point of para-hydrogen: C. There is no objection against using the normal boiling point of hydrogen as a basic point in thermometry. The vapour pressure below atm. is measured and expressed in a formula.

W2 = N (I + Pp + Qp2)

W2 = N (I + Pp + Qp2) Physics. - Measurements about the velocity of sound in hydrogen gas at liquid hydrogen temperatures. By A. VAN TTERBE EK and W. H. KEESOM. (Communication N0. 216c from the Physical Laboratory at Leiden).