Document KRJQJyk99L59zxr2r5MKvLrNQ

Nov-, 1937 BOMB riTYXIC-U, PROrERTlBS OP VINYL QILOKIDB 2S01 Z mixture u.iium of ijtU'AUItU .-.itc'.iard's solutions. .'ausclow's .nivitr rocoefficient t available wcivu* (c^) ok lor the rausly supilations are ' ,U<1. :>< 7i3 7V2 770 7*2 *.; .S23 To it mod. C k QA 07V r.d the **pr. renin; pou-' ontcat; tb^ are mult'. V.simal dat-. cocfficw::by 6*.' . Tbe ap*-` 1.0 -0-OSi 4.0 sucnt is quite satisfactory. Similar data for solutions up to 0.1 if have iircn discussed by Scatchard.** The writer wishes to express his indebtedness and gratitude to the KUzabeth Thompson Science Fund for providing mechanical aid in the jalcularinua. A report b made f the dctennhntiaa of the freezing points of tur^tt solutions of hydrochloric acid up to a concentration of 1.0 sndal, c*c of a platinum tltcrmometer as a temperature measuring device. The activity coefficients are calculated and compared with timer de rived from electromotive force measurements. Tuns Cmians, MAfascnusana yrninoms non ms Lasokavoky o ms Loros Am Paoaucia Company) SOME PHYSICAL PROPERTIES OF VINYL CHLORIDE . By L. I. Dana, J. K. Bussteu axb A. C. Jenkins Bamves Ivmwt IS, U0 IWan* Know t. 1B37 Although vinyl diloride (CrIIjCI) has been known as a dtemieal com v,,sd for a number of years, owing to the difficulties attending its prepara.-<5 and purification, relatively little exact work has been done with it. Uuvnt progress in chemical synthesis has made this material available in entity commercially and therefore it has seemed worth while to review : wvicus work on the subject in order to ascertain the present state of . r. wlcdgc of the properties. licyond a few scattered values of the boiling point of vinyl diloride, little has been published concerning its physical properties. Re- i'.y the desirability of utilizing vinyl chloride as a refrigerating fluid -iling print --13.0) has been suggested, and lienee it baa appeared l -irable to conduct an experimental determination of its most important seal properties. In this investigation the measurfmeiits of the vapor: :.-*urc curve, liquid densities and freezing point are reported; in - :`.w of its other physical properties have been estimated from theory. Description of Experiments Vrpor-Pressure Curve.--Commercial samples of vinyl diloride1 were in small steel cylinders. This liquid was purified by fractional Pinion in a vacuunf insulated, low-temperature Hempel column 1 was further subjected to a fractional distillation in a dosed, low- "Scuchaid, Tars Journal 7, Ml (1035). 8m also LaMtr, Trow. Amtr. M. Sit, fl, (1627). P*P*iat No. AS; aad Van Lear, Z. saarg. attf/m. Qret,, U9, ac:i). ' Matwfal obtained bn tha Carbide and Carbon n--t--1 Corporation, New -i Ctf. 2502 BABA, IDtnaC AMD JBNKtNS Vo' temperature fractionation apparatus, operated with a Langmuir pump in aeries (or the purpose of removing the traces of moisture and ail non-condensable gases. Finally, the samples were tested /or boiling range and pressure drop in the vapor-pressure apparatus mentioned below. On off over 85% of the sample at constant temperature, the jaiisiur drop was fund to be'has than 2 mm. b 7(10 mm. of Hg, and on jfisritfiqgrf at mastant pressure the temperature drop was less than OHS*. It is probable that the parity was higher than 99.9%. Tfaa vapor pressures wen measured over the temperature range from --2S to 60*, corresponding to a pressure range of 40 to 700 an. of Hg absolute or from 0b to 10 atm. absolute. Below and around one atmos phere a mercury manometer, with a vacuum ova one arm, was employed for the pressure measurement. From one to five atmospheres a tall mercury manometer was used, while above this pressure a dead-weight piston gage served to measure the pressure. Virions forms of containers lor the sample were used. In general the experimental technique was similar to that used b the measurement of the vapor pressures of some peraflb hydrocarbons,* to which work reference is made for experimental details. The boding point at one atmosphere was determined a number of times and was found to be --13.9, 10.1*. Biltz* gives the boding range as --IS to --15*, which would seem to be much too low temperatures and too wide a range to correspond with the pure materiel tested b this in vestigation. In the following table ere given the values of the; and corresponding temperatures observed for the vapor -28.78 --28.02 -16.61 '-18.61 - 8.82 - 1.57 TakxX Varna Psatsoxss os Voro Cmosms S% T*J- 89.56 51.80 67.76 `78.78 84.90 122.48 4.01 8.53 . 16.21 . 25.72 83.58 89.72 149.06 158.2 223.8 802.7 878.9 449.2 89.72 48.80 84.87 60.84 00-84 Si*, ---X* 450.0 843.4 967.6 756.6 760.8 The experimental observations have been combined into the wellknown Kerast4 equation of the form toe t rn m + >/r + 1.7* h* r + tT b which the pressure p is expressed b atmospheres absolute and the tem perature T b degrees centigrade absolute. The values for the constants am as follows ' __________ - 08(20; - --1130.9; e - --0.002415 * Dana, Jenldn*, Burdick and Timm, sg,, 12, 2S7 (1986). B3U, Btr., IS, 2525 (1902). 4 Ncnut. "Thewetittl Cbenu*try,M Mwmfllan Co* New York, 1923, p. 815. * On the average by tlie fapmih Liquid Dcr.L the liquid den*,: and is similar the Bureau o ammonia.1 Fr meats were firs pycnometer in coefficient of liquid was me weight of liqv b the pycnor certained. T xn.ats were pycnometer, constructed o After callhr; meter carciV the sample and the pycr Headings of Stem were 1 temperature! in a thermo-: the tempera1 a platinum meter. By meter full a rected wcigl was found Enough dat able for con: density. "3 coefficient < to v with the V Correct;. . for the qua meter. X however, s ever, the it * Cra;c Ucc 047427 Vnl. JO, A pump re and all 'or boiling mod below, mature, the Ig, and on than 0.05*. ange from cm. of Hg me atmos. employed :rcs a tall :ad-wcight container! vas similar ic paraffin ail details, tr of times ; range as .lures and in this in* pressures suras. lie * 4.10.0 aia.4 Atr.o 7.18.0 700.3 the weB- ;*:e tomcoaitant? p. 815. Xov., 102? sown nivsiCAt: rRorcancs op vinyi, chlorjdb 2S03 On the average the agreement of the observed results with those computed by the formula is within Utc experimental error, 0.3%.. Liquid Densities.--Tlic method employed for the determination of the liquid densities has been described in an earlier paper by the authors,* and is similar to that, used fay the Boican of Standards on tuuoonia.* Preliminary experi- meats mtfiatmade in a Pyrtx pycnometer which only the coefficient of expansion of the L r* $ liquid was measured, met the weight of liquid vinyl chloride wl** in the pycnometer was not as* certamed. The final measure* meats were made with the pycnometer, shown in Fig. 1, constructed of Jena 1GIU glass. After calibrating the pycno meter carefully with mercury, -- ft" * rpLj.TI*j $ *1 Zx 7 % C**tujt*r . Tub*** the sample was condensed in and the pycnometer scaled off. Readings of the liquid in the $stem were observed at various temperatures from --13 to 00* in a thenno-rcgulatcd bath and the temperatures were read by o 3 #* 9 a platinum resistance thermo meter. By weighing Uic pycno- meter full and empty, the cor* reeled weight of vinyl chloride was found to lie .1.9710 g. rvzo "tiaa Oolo Enough data were then avail able for computing the absolute density. The results for the coefficient of expansion agreed to 0.1% with those obtained Jiy with the Pyrcx pycnometer. Correction has to be applied Ftf. 1.--Picromctcr for liquid dentine*. ;ar the quantity of vapor in the vapor space above the liquid in the pycno meter. Xo experimental values for the vapor densities are available, however, so that these values were computed theoretically. Since, how ever, the miximum value for this correction is about 0.15% and an accuracy * Cn;oe and Harper, Bureau of Standards, Sdentidc Papar, No. 430. ucc 047428 2SW daxa, mnencK axd jbnbks Vol49 of 0.1% is claimed for the results, it is evident that an accuracy of 00% for the vapor density correction is sufficient. It is believed that the vapor density can be computed with much higher accuracy and as a matter of interest this has been done and the results am given later on. Resisnwce ThrAKtortrre* The liquid densities of vinyl chloride am shown in Table II. It is believed that the results am food to 0.1%. An empirical formula representing the results from --13 to 60* is as follows i - 0.M71 - 00017461 - 0000003341* TKSHT SAL in which i is the liquid density in grams per cc. and l is the temperature in de grees centigrade. The data agree with the formula to better than 0.1%. Ua<stD M//r Freezing Point.--Since relatively large quantities of vinyl ehoride, puri fied by fractional distillation for the vapor-pressure measurements, hod I . .. \----/j ^^ ' been prepared, it was decided to de termine the freezing point by direct immersion of the thermometer in a Chlok/oc large sample. The apparatus is shown in Fig. 2. ' Throughout the measure ments the sample was handled in a dosed system to keep out moisture. Liquid air was employed as the cod ing medium. With a single tube the rate of cooling was too rapid; hence a double-walled tube was used. This also bad the effect of producing a uni form temperature gradient in the freezing-point sample. Preliminary determinations with a calibrated, propane filled, low temperature tberrnometer resulted in the value --13?*. The feu; and most accurate determina l-"e. *--Fri39i-r.wt arpamtus. tion was nude with a platinum resis tance thermometer winch had previ ously been calibrated at the freeaing point of mercury, the sublimation point of carbon dioxide and the boiling point of oxygen.5 The period of freezing lasted about seven minutes and the freezing temperature was Nov, 1087 some rnv constant to within 0.1* at it is safe to claim an accu tVm, C. --12:0G 1.32 13.43 38.11 Ltj; Li" Estimated Since the vapor dens; densities, it appeared w so that the values mig the vapor-pressure cunwith the aid of Clapc Unfortunately the c so that the law of co: but van Laar* has do by means of which tht serve in other relation tained. By these me while the estimated In the following ta vapor, estimated by Estimated Smsch TP, -30 -30 -10 0 +10 To compute the 1. were used and the graphically. The later: hi.*.:,spending stale*, er; chloride. "At the e while at SZ* they : * Vas Laar. "1. r Holst, Ccmv;, ' Ucc 04*29 it the vapor a matter of -yl chloride results ue epresentiag 60" is as .KW00324I* it}- in grams lturc in dc- . %.l agree with 01 relatively aridc, purii.*>:i for tlic cuts, bad ded to de. by direct neter in a cs is shown e mcaiure adied in a : moisture. ^ the cool ie tube the d; hence a ed. This cing a unii: in the reliminary rated, pro 's thenr.o;e -- .Z.terminar.unj resishid previ-blimation $ period-of -ature was .. . /,, 1027 SOUS PHYSICAL PROPERTIES OP VINYL CHLORIDE 2805 ` i coastant to within 0.1* at --159.7*. Considering the purity of the sample it is safe to claim an accuracy of 0.1* for this freezing point. , Tasls n ** ' Lipuio PKKsmas os Vnm. CnuMum -'-Sf** -c. UoM iluAlr. tJn. Two*. C. Uo*M dMaitr. t/*. -12:06 O.OOS2 80.57 0.8733 1.32 .0*43 48.20 .8555 13.40 .0223 50.01 .8310 36.11 .8065 Estimated Vapor Densities and Latent Heats Since the vapor densities were required for correction terms to the liquid densities, it appeared worth while to estimate them as accurately as possible so that the values might be of use elsewhere. At the same time, siuec the vapor-pressure curve was measured, the latent heats could be estimated with the ail of Clapeyron's equation. Unfortunately the critical constants of vinyl chloride arc tint known ! ' so that (he law of corresponding stales could not be employed directly; t but van I.aar* lias described methods based on empirical relationshijn by means of which the critical constants may be computed, which in turn -* serve in other relationships from which the specific volumes can be asccrj tained. By these methods the estimated critical pressure is 52.2 atm. while the estimated critical temperature is 142*. t la the following table the estimated specific volumes of the saturated sapor, estimated by the method of van Laar, are given. Tails IQ * Estimates Srxcmc Volumes or Saturates Va* or Viktl Chloride Ymc,, Srm.Jtt.*l. c. ft*7-s*.*. -30 635 30 105.4 -20 413 30 70.7 -10 284 40 60.3 0 100 50 46.3 +10 143.8 60 36.2 To compute the latent heats the above values for the specific volumes were used and tlic slopes of the vapor-pressure curve were determined graphically. The latent heats* were also computed by means of the law of corre sponding states, comparing with the values given by HolstT for methyl chloride. CAt the boiling point the values by the two methods agree to 1%> while at 50* they agree to 5%. I Van Laar, "Die Zastasdscldchahc." Leopold Von, Leipzig. 1924, pp. 186,21. 1 Hoist. Ctmm. Pkyt. Ltb. Unb. Lridc*. No. 144c. ucc 047430 2S0G * . KGTB VoL 40 XT -S* -JO 0 +10 30 ao 40 '50 Taiujc IV IisnMATKD I^kTinrr IIbat* or Vimyl Cnuwroit moUwtl,IWI fUBa f IIa/^Cs 35.4 35.4 47.1 01.8 70.7 100.3 134 151 H* HClfWcwliKroIa'a W.4 m.i 81.3 70.8 77.7 70.3 73.3 70.1 camal--o'rrn KV7 HI. 1 KI.O 81.7 80.3 78.5 70.0 74.4 Summary The vapor pressure of vinyl chloride mu measured from --28 to GO*. The normal boiling point was found to be --13.0, *0.1*. I The liquid densities were measured over the temperature range --13 to GO*. The freezing point was found to be --130.7, *0.1*. * Estimated values of the vapor densities and latent heats were computed from theoretical relationships. Bumno, Nsw Ymuc NOTE Calculation of Heats of Combustion.--Kharasch and Slier1 have applied the electronic conception of valence to the calculation of heats of combustion of organic compounds. They find that the energy change which accompanies the displacement of one valence electron from the posi tion which it occupies in methane, a carbon chain or the benzene nucleus to the position after combustion in carbon dioxide is 20.03 kg. cal. per mole. Thus the heat of combustion of a hydrocarbon is given by the product 26.03 X N, where N is the number of electrons displaced.* Electrons associated with bonds other than C--C and C--II are accompanied by an additional energy change during combustion. To the terra 20.03 N, Kharasch and Sher add 2G kg. cal. for each aliphatic ether group, 13 kg. cal. for each ethylenic linkage, primary alcohol, ester or aromatic ether group and G.5 kg. cal. for each secondary alcohol or ketone group. No correction is added for tertiary alcohol, phenol or add groups. In most cases this method gives good results, but for aromatic compounds the cal culated values are uniformly higher than the experimental values and the 1 Kharasch sad Shcr, J. Pkys. Cheat., 20, 025 (1025). * Each bead corrwpondi to two electrons. In oxygen compoondr the two oxygen bonds arc disregarded. Thus A* is 8 for methane, 80 foe benzene, 5S foe phenyl benzoate, etc. < \ 9 \ * Nov., 1027 I I deviations liccomc grrali necessitates a correction for this correction follow i *17ic riortrofis avmrml' \ (ft ckflrmr.) will I*- nv: i tlte olii'T twelve (X i * 20.05 kg. Col. Jkt moli:,: tlic compound befog bur at constant pressure of is given by the crjuatio group + IflJ* for each I primary alcohol or cstc + 3.2 for each phenol ketone group. Of the 57 aromatic < whose heats of combu* new method gives result i tire tame in 7 cases, mental and calculated shows per cent. (lilTen t for 171 aromatic compt \ f Dhyvsunci: sctvkix C. Aromatic nupiinili caulainiac C awl II, j . ar C. II awl Saturated hydrocarbons I Ethylenic hydrocarbons nwncis Ethos Aldehydes Ketones Acid* Ester* Total CoxrainoTion non TO! ] DerAKTMEXT or Chemist: Tnz Fcxxsylvama Stati State CoLLcer, Pcnxsw. Rrcn'rD M 'rcn ll. *.' I Puwus.icb NovtXLS" i. * See Pauling, T:s i 4 X and s arc the w s , there are no a electrons at benzene nucleus contain* i ti anthracene; X 40, a - f ucc 047431 Vol \ SV7 $4.: S3, o $1.7 $">.2 73.3 73.a 74.4 2S to CO*, ffe --13 to computed her* have - of heati *y choose : the posi:c s.dcuf per ai Ic. product Electrons icd by an 2J.<V* .V, ?. 13 kg. ii;c ctlicr Nr. I:: mo.*; i the cal* i s.zi the :vra oxy;rn V benzoate. ; Nov, ' ' r'< in 2S07 ' T-tovialk^-^cconic greater as i.c number of benzene nuclei increases. This necessitates a correction for each benzene nucleus. The method proposed or this correction follows. ' The electrons associated with three of the bonds of the benzene nucleus (fi elections) will be assumed \> differ in energy relations from tlmsc of the other c.vclve (X elect, ons).* .rhere is an energy cliange of 21.So and > 51.05 kg. cal. per mole, respectively, for each a and X electron present in the compound being burned. The heat of combustion in kg. cal. per mole at constant pressure of compounds containing C and H, or C, H and O, h given by the equation4 Qt - 26.05X + 24.S5p + 22.7 for each ether group + 19.5 for each aldehyde group + 13 for each ethylenie linkage, primary alcohol or ester group + 6.5 for each secondary alcohol group + 3-- for each phenol or tertiary alcohol group + 0.0 for each add or ketose group. Of the 57 aromatic compounds containing C and H, or C, H and 0, whose heats of combustion were calculated by Kharasdi and Sher, the 1 new method gives results closer to the experimental values in 37 cases and | the same in 7 cases. The average per cent, difference between experi( mental and calculated values is lowered from 0.52 to 0.32%. Table I : shows per cent, difference between calculated and experimental values i for 171 aromatic compounds of different types. 1 Tabl* X DrrrEXExcz sctweex Calculated ajcp Exteehouctal Heats or CouausTtox Aronuik mipundi renumiiir C ! H. f C, H ft4 O Xe.eS wUculvilcM DMTtmce betwen celceleMO wl tiperimaUI nlim Teul % Anna % Saturated hydrocarbons so 8.4 0.32 Ethylcnic hydrocarbons rheneis Ethos Aldehydes Ketones Acids n 2.7 0.25 24 10.7 0.44 23 8.0 0.30 12 7.8 0.61 15 5.5 0.37 33 13.3 0.40. Kut* 38 8.3 0.30 T.nul 171 04.2 0.37 CilXTS'lU-TlilK WnM Tlll( ni:i\tKTii!NT or CnUMisTNV, Vi::: 1\:\nsvi.vama Stati: Colucoh, S;.\Ti: Ci >;.u:i:i:, Piinxsvlvakia Minrii II. IVit btuiuib NvsHdiiH A, 1927 J,!{. Snvimm See Pauling, This JounXAt, 48, 1133 (1020). * X and a are the numbers of X and a electrons present. In aliphatic compounds < :lee ire no a electrons and X is the same es .V in Kbamsch and Shcr's equation. Each benzene nucleus contains 6 a electrons. Thus X 33 for hexane; X 48, a m 18 for iI sathneese; X 46, a * 13 for phenyl benzoate, etc. ucc 047432