Nitrogen Gas and its properties

8/12/2019 Nitrogen Gas and its properties

1/4

urnal of Resea rch of the Na tional Bureau of Standards Vol. 53, No , October 1954 Research Paper 2543

Vapor Pressure of itrogenGeorge T. Armstrong

. . The va por pressure of nitrogen has been measured in the liquid range below the normalb O l ~ t n g pOl.nt and .can be represen te d b :y log P .(mm) = 6.49594- 255.821j T - 6.600) . Thenormal bOIling pOl.nt calculate d fr om t hIs equatIOn IS 77.364 K . Ni trogen vapor densitiesalong. th e sat l rat lOn ho e are represen te d by log p = 3.39858- 282.953 j T - 3 .83). ThefugaCity funct IOn In f jp for t he sa tura ted vapor is tabulat ed.

1. IntroductionIn t.he course . of a study of the vapor-liquid phaseof mIxtures of oxygen and nitrogen aurements has been made on the vaporpure liquid nitrogen. These measur el' tb e liquid range below the normal

2. Description of Cryostatrr:he c r y o s t ~ t was d n for liquid-vapor equistudIes of mlXtures and thus containsveral features not essential to the vapor-pressurees . The fo llowing description of the apparatusvers only those portions of the apparatus essentialthe measurements.The equilibrium vessel, in which the liquid nitrohaving walls of -in.and an inner diamet er of 1 in . The thermomer is in a .well suspended from the upper lid. Inr ~ the quantity of liquid nitrogenllsufficien t to touch the tb ermometer wellat reliance was placed on t he uniform-temp era ture

nt and the high thermal conductivity ofureat the thermometer and liquid were at the samerature. The q u ~ i b r i u m vessel is suspended byes wlthlll a copper can , which may beua ted or fIlled with helium gas for heat transfer.is can forms a constant-temperature enclosure fore equilibrium vessel. t is completely immersed ino n e ~ a t u r n i t ~ n bath. The temperare of thIS bath IS mallltallled at the desired operatmp erature by regulating the pr essure underit boils with the aid of a car tesian-divertat.To reduce losses of liquid nitrogen from the conerature nitrogen bath, and thus permitra tion before refIlling is required, the conall:t-te llperature b ath in?-mersed in a econdarygen bath , wInch IS allowed t o boil free lythe prevailing atmosph eric pressur e.The m a.nomet er tub e passes through each of theaths. The thermometer leads are broughtinto the helium-fIlled space surroundth e equilibrium vessel. To insure that they arethe bath temperature the leads are wound severalth e pumping tub e, which proj ects int oe hellUm-fIlled space, and are cemented to th e tube.This research was supported by the r Research and Development Com nd , Department of t he Air Force.

263

3 . Temperature MeasurementT ~ m p e r a t ~ r e were measured with a capsule-typeplatlllum r eSIstance thermometer immersed in a wellin th e lid of th e equilibrium vessel. This thermometer was calibrated against th e National Bureau ofStandards provisional temp era ture scale below th e

~ point [1] 2 and was checked at the oxygenPOll1t durlllg the cour e of the measurements.4. Pressure Measurement

. The manometer used is a version of one describedby Swindells, Coe, and Godfrey [2] modilied in sucha way as to make an absolute pressure-readinO' instrument . In thi manometer th e mercury u r f ~ c e s arelo cated by touching them with stainless-steel rods ofcalibrated lengths. The contacts arc det ec ted elecy in this instance , by observing the extinctionof a llghted neon bulb when a contact is made. Themanometer has one fix ed contact in the arm con

e c t ~ to the vapor-pressure apparatus. The oth er~ l m IS and evacuated. The detecting rods a remtroduced ll1to the arm through a mountingthat can > e moved vertIcally to bring the rod intocontact WI th the mercury . After a con tact is madethe position of the upper end of the movable rod

~ m i n d w i t ~ aid of a micrometer depth gagereadmg m millImeters. The manometer readsdirec tly to 0.01 mm , and it is pos ible to interpola teto about 0.002 mm . The manometer as used in tb esemeasurements did not provide readings of this accuracy , b ecause the control of the mercillYo l u ~ : m was not suffiCIent ly good. Errors as large as9. b mm or perhaps somewhat larger may have beenmtroduced at tImes because of uncertainty in themean temp erature of th e mercury column.The measuring rods were calibrated by the GaO'eSection of the t i o n a l B u r a u of Standards. Thirtyseven q u a r r - l l l h stalllless st eel rods differing inlength by lllcrements of 1 in. permitted completeof the In:essure range. The rods have a

~ l l m end WIth a rounded tip of approxima telyY a 2 ~ 1 l l radms. The ul?p er end of each is capped by a%ll' sphere of bearmg bronze, against which themlCrometer contact is made. The under surface ofthe sphere forms a vacuum tight bu t easily demoun table seal a conical opening through the movable mountmg at th e top of the closed arm of themanometer.

2 Fi gures in brackets indicate the literature references at the end of this paper.


2/4

Mercury heights determined in this manometerwere corrected to 0 C for thermal expansion of therods and of mercury, and to a stand Lrd gravity of980.665 cm sec- 2 The tube bore is 1 in., and so thenecessity for capillary corrections was eliminated.This diameter also insures that the mercury surface isflat enough that the centering of the longer rods doesnot have to be closer than about 2 mm .A small correction to the pressure was applied tocompensate for the pressure difference between themercury surface and the liquid-nitrogen surfacecaused by the greater density of the cold gas in thecryostat. This o r r e t i o ~ at most amounted to 0.08mm and was very nearly proportionn,l to the pressurein the system.

S. Material InvestigatedThe nitrogen used in the experiments. LindeAir Products Company standard high-purity drynitrogen. This was stated by the supplier to containless than 0.005 percent of argon. A calorimetricstudy of the melting point of a similar sample as a

function of the fraction melted indicated that liquidsoluble solid-insoluble impurities amounted to muchless than 0.01 percent. The material used in thelast series of vapor-pressure measurements was analyzed by mass spectrometer after the measurementshad been completed and was found to . contain approximately 0.01 percent of oxygen. This sample hadbeen in the vapor-pressure apparatus for approximately 2 weeks under reduced pressure, so it is probable that th e oxygen entered from the walls or byseepage through stopcock grease, and it may haveentered after the measurements were completed.In any case, the maximum effect produced by thisamount of oxygen would be 0.06 mm at 760-mm totalpressure.6 Experimental Procedure nd ResultsIn order to insure purity of the nitrogen introducedinto the system, the connecting lines to the highpressure cylinder were evacuated and filled severaltimes and left full of nitrogen at a pressure slightlygreater than atmospheric. The remainder of theapparatus was then evacuated overnight at a pressurebelow 10 - 4 mm. I t was then filled with nitrogen,and the cryostat was cooled. Approximately 2 litersof gas was then condensed into the sample holder.The amount condensed was varied in some of theearly measurements , and no effect on the measuredpressures was observed. After filling the apparatus,all parts except the manometer were closed off bymeans of stopcocks. .I t was impossible to keep the temperature absolutely steady . Drifts observed were of the order of0.01 deg in5 min at the lower temperatures and onehalf to one-third this rate near the normal boilingpoint. A series of alternate temperature and pressure measurements was made over a period of 10min to }6 hr. These were plotted as a function oftime, and for each pressure reading a correspondingtemperature was found by interpolation. Each

value thus determined has been treated as a separate

point, though in a sense the measurements of a seriesare not independent.Because there was no stirring in the nitrogen constant-temperature bath, a period of 1 or 2 hI wasneeded to fix the temperature of the bath at a newvalue and to allow equilibrium to be reestablishedI t was thought to be advantageous to start the measurements at a low temperature and to allow th etemperature to rise between measurements. Thisprocedure insured that the sample vessel, whichalways lagged the bath in temperature, would neverbe at a higher temperature than any part of the baththrough which the manometer tube passed.The measurements made on several different daysand using several different fillings of nitrogen, showedno consistent differences from one another. Almeasurements made in runs 1, 2, and 3 are shownin table 1. The only measurements not shown aresome earlier ones in which the bath level was noproperly controlled and in which the manostat regulating the bath pressure was not functioning properlyThey showed large and erratic fluctuations, whichdid not appear again when these two factors werecorrected.

TARLE 1. Vapor p eSSU e of liquid nitrog enL;TX103 L;T X Lng P TOK (observed Log P TOK (ubservedmm (observed) minus mm (observed) minuscalculated ) calculated)

Series 1 Series 32. 89075 77. 5578 - 1.1 2. 89479 77 . 6398 + 1 .42. 89040 7i. 553 5 1.5 2. 89486 77. 6406 + 0 . 82. 89037 77 . 552l + 0. 7 2. 89487 77. 6412 + 1 .22. 89025 77. 5495 + . 4 2. 89594 77. 6607 - 0. 42. 89119 77. 5693 + 1.7 2. 89611 I i 6654 . 92. 89111 77 . 5665 + 0. 5 2. lI060 64 . 9360 + 0. 72. 89114 77 . 5678 + 1.2 2. 11127 64.9452 + 1.02.89188 77. .5820 + 0. 9 2. 11159 64. 9495 + 1.02. 21273 66.3274 + 1.12. 21284 66. 3291 + 1. 2Series 2 2. 24894 66 .8354 - 0. 22. 24995 66. 8561 +. 22.25058 66 . 8576 - 1.22.83444 76. 4686 + 1. 0 2. 89618 77 6638 2.12.83450 76. 4697 + 0. 9 2.89619 77 6642 -1 . 92. 83453 76. 4704 + 1 .1 2. 8961 8 77. 6644 - 1.52. 79033 75 . 6344 - 1.6 2. 21036 66 . 2920 - 1. 32. 78979 75. 6267 +0 . 8 2. 21138 66 . 3088 + 1. 32. 78929 75. 6181 + 1. 5 2. 21457 66 . 3500 - 2. 02. 78910 75. 6130 - 0. 0 2.27468 67 . 1998 -3 . 12. 78944 75. 6186 8 2.27418 67 . 1959 + 0. 22. 56364 71. 6574 .8 2. 27402 67 . 1915 - 1.92. 56370 71. 6580 + . 8 2. 35593 68 . 3915 - 0. 72. 35740 68 . 4412 + 27 . 0 2. 35591 68 .3916 - .32. 35715 68. 4386 + 28. 2 2. 35609 68 . 3939 72. 35713 68. 4370 + 26. 9 2. 42898 69 . 5017 42. 27544 67 . 2135 + 0. 6 2. 42905 69 . 5036 . 42. 27547 67. 2137 - . 8 2. 42917 69. 5054 .32. 27556 67. 2137 - 2. 3 2. 50438 70 . 6901 22. 50457 70. 6933 12. 21875 66 . 4091 - 0. 5 2. 50473 70. 6966 . 72. 21881 66 . 4102 - 1. 7 2. 58980 72 . 0941 + 2. 22. 21894 66 . 4121 - 2. 5 2.58991 72 . 0952 +.I..J2. 11814 65 . 0373 +1.5 2.59002 72 . 0968 +1.32. 69836 73.9597 - 1 32. 11837 65. 0390 + 0. 1 2. 69871 73 . 9709 + 0. 72. 11864 65 . 0419 - . 6 2.69917 73.9797 + .1.32.05277 64. 1776 + 1 5 2.85437 76.8532 + 3. 22. 05289 64 . 1777 0. 0 2.85470 76 .8594 + 3 . 02. 85472 76 .8593 + 2.52.05296 64 . 1778 - . 8 2. 85465 76 . 8579 + 2. 5

Discarded by Chauvenet' s criterion.The data have been fitted by the Antoine-typeequation 1)

log P mm)= 6.49594 - 255.821/ T- 6 .600) . 1)26


3/4

Aside from Lhree points at 68.4, the measurementslie within a narrow band about eq (1) . . For thesethree points the observed and . c i l : pressuresdifle r by very nearly 1 mm , so It IS poss Ible tha t anerror was made in reading the micrometer depthgage. These three points were discard ed thebasis of Ohauvenet s criterion. The m ean devlatlOn1) of all measur ements except those spec ii ed above is 0.0012 deg K , or 0.063 mm . I t ISossible that a slightly. better fit could be o?t.ainedby the use of an additlOn al co nstan t or a dIff erentunctional form of an equatlOn because there appearsto be a slight cyclic trend of the deviations. Anestimation of the best fit in the form of a smoothcurve drawn through the deviations suggests thatthe mean dev iation could not be reduced below0.0010 deg K by any other simple eq uation. .The normal boiling point calculated from eq 1) IS77 .3 64 deo K. Some other experimen tal values arehown in table 2 [3 to 8]. The sta ndard dev iation inthis temperature, which was found to b e 0.001 3 degfor the presenL work, indicaLe onl y Lhe in ternal

precision of the data and does not grv e an.\T mdlCation of the reliability of the Lcmp erature scale. Thealues obLained by HClming and OtLo [6] and byKeesom and Bijl [7] are very clo e Lo presenLvalue. The value of 77. 34 deg reported by ] n edmanand White [8 ] is obtained from their equa tion. Theilvalue is subj ect t o ah uncer ta in ty of 0.05 to 0.07deg because of the deviation of t ~ e i r eq.uation ~ r ~ mtheir experimen tal value in the Immecha te vlclmtyof the boiling poin t.A comparison of the experimental daLa fromseveral laboratories with eq 1) is shown in fi gur e 1.The deviations shown in table 1 have been omittedfrom figure 1 in order to avoid a co nfusion of po in.tsnear the reference line. The present data arc I I Igood agreement wi th the data of o m and Bijl,showing only small sy tematlc de:, la tlOns. Th erath er la rge deviations of the da ta of H ennll1g andOtto are not easy to account for because they areerratic; on the other hand , the deviations of .the dataof Giauque and Olayton are very syst?matlC. ~ hystema tic deviations in the w o r ~ from v.anouses are probably due to chfIerences I I I th etemperature scales. I t is unlikely tha t any furthermprovement in the consist ency of the vapor-pressuredata of ni trogen will be mad e un tIl the m p e r a ~ u r ees used in various laboratones are brought mtoagreement in this region.

T 2. Normal boi ling point of ni tro genDatc Ts I1 ----:-;-1n vcst igators1927 Ii 361932 77 . 3461933 77. 32J936 Ii 3521937 77. 373

Dodge and Davis [3] - __________________ ____ ___ _______ _Heuse and OLto [4] - ___ ._____________ , _________________ _GiauQue and Clayton [5] ---------- - ---- --- -- _______ _~ ~ ~ I B q j \ t l 7 ~ ~ :::::: :::::::::::::: ::: _ :::::.-l ri cdm all and Whi Le [8] ______________________________ _ 1950 77. 34This rcscarc l ___ _______________________________________ 19M 77. 364

.02

.01

0

-. 0 1f- 0I 0

-.02 00 00 0-. 03 00 0- .04 0 0

0 0- .0 560 62 64 66 68 70 72 74 76 78 80T oK

F IG URE 1. Deviations (observed mim.s calr-ulated) of otherexperimental values from equation 16 , Kecsom and Bijl; 0 , GiauQlle and ClaY Lon ; 0 , Hcn ning and Otto.

7. alculation of Saturated Vapor Volumeand Fugacity\Vapor-pr essure data have at time been used forcalculating the la tent heat of vaporization, using theformula t::.H= T (17g - The vapor volumeof nitrogen is, however, probably no t as well knownas the other quan tities required in this formula.The only direct experimental measurements of thisq nan ti ty were those of M athias, Onne , and


4/4

T B LE 4. ugacity function o nitrogenl jp

O {64 - 0. 009666 012768 016370 020572 025074 - . 029776 - .0348.8 0409

In column 1 of table 3 are shown values of vapordensity calculated from eq 2). The values fromwhich this equation was derived are shown in co lumn2; those given by Mathias, Onnes, and Crommelinare listed in co lumn 3, and the revised values presented by Crommelin are shown in co lumn 4.The densities calculated from the vapor-pressuredata are seen to be in termediate between the observed and revised values of Mathias, Onnes, andCrommelin. I t should be noted that the revisedvalue for the vapor density given by Crommelin at64.80 K is incompatible with the behavior of a realgas near its saturation line, as it is less than ideal gasdensity at this temperature and pressure.Using eq 2) for calculating vapor volumes, thefugacity of nitrogen along the sa turation line has been

calculated and is shown in the form l lp in table 4.There are no experimental data for this quantityderived from PVT measurements below 80. However, extrapolations of higher- temperature data fromvarious sources give values that are in some caseslarger and in other cases smaller than those listedin table 4 .8. References

[1] R . J . R oge and F. G. Brickwedde, J . Re search NBS22, 351 1939) RP1l 88.[2] J . F. Swindells, J. R. Coe, and T. B. Godfrey, J. Re searchN BS 42, 1 1952) RP2279 .[3] B. F. Dodge a nd H . N. Da vis, J . Am. Chem. Soc. 49, 6101927) .[4] W. Heuse and J . Ot to, Ann. Ph ys. [5]14, 185 1932) .[5] W. F. Giauque a nd J . O. Clayton, J . Am. Chem . Soc. 55,4875 1933).[6] F. Henning and J . Otto, Ph ys. Z. 37, 633 1936).[7] W. H . Keeso n a nd A. Bijl, Physica 4, 305 1937).[8] A. S. Fr iedman and D . White, J . Am. Chem . Soc. 72,3931 1950).[9] E. Mat hia s, H. K. Onnes, and C. A. Crommelin, Koninkl. Akad . Wetenschap. Amsterdam 17, 953 1915);Commun . Leiden, 145C.[10] C. A. Crommelin, I nternationa l Cr itical Tables III , 204McGraw-Hill Book Co., In c., New York, N. Y. , 1930).[11] G. T . Furukawa and R . E. McCoskey informa.l communication).

W AS HINGTON , May 18, 1954 .

66 U, S. GOVERNMENT PRINTING OFFICE: 5

Documents

Nitrogen Gas and its properties