-\pril, 1960
REACTIOSS OF H20(g) WITH YA*OASD L I ~ O
Finally, thrl fact that the adsorption was poisoned by the prewnce of adsorhatec: such a? oxygen or carhon dioside quggestc: a specificity in the nitrogen interaction n-ith certaiii qites that is not generally cvi4tlered to be characteristic of physiral adsorption. Tliu.; it wemq evident that the adsorption is chemical ii. nature and that, at -193", it is sufficient in magnitude t o coyer about 20% of the w.face. I: will be observed from the data in Table I1 that the chemisorbed nitrogen had no effect oii the exchange activity. Xppnrently, either the nitrogen TV:L' ac:>ociatedn ith qites not involved i:i the exchanne or the hydrogen displaced the nitrogen from the curfxce. Pome evidence for the latter powhility i g provicled I)y the fact that the reculting hrdrogfn-tlcuteriimi mixture Tvas found t o contain I race- of nitrogcw. Conclusions This WII 1; ha' confirmed aiid extended earlier ohserrations concerning the reacti7;ity of reduced chromia qurfaces at Ion- temperature.. Both carbon momxide aiid hydrogen are chemisorbed in diicieni quantities at - 193" t o cover essentially the entire surface, the carbon monoxide poisoning the sui,face both for the hydrogen-deuterium
457
exc.hange reaction and for the subsequent chemisorption of other gases. The data suggeit that the carbon monovide adsorption may involve a dual site merhanism. The adsorptions of oxygen at -1%" nnd carbon dioxide a t -78" are not as eutenqive, the former covering about (30% and the latter 80% of the total surface. The oxygen chemisorption is considered to be a surface oxidation with the resulting change in hydrogen exchange activity being a consequence of a change in the d-electron configuration of the surface chromium ions. Additional evidence has been obtained for the chemisorption of nitrogen on a reduced chrornia vrface, thi.; adsorption corresponding to a wrface coverage of some 20%. Following other norkers, it is suggested that the ability of this surface to chemisorb a variety of gaqes at theqe temperature? is due to the availability, a t the wrface, of a suitable number of unpaired d-electrons. Acknowledgments.-The authors take this opportunity to thank Dr. George F. Crable for making the mats spectrometric analysis. Thanks are also due to Drs. C. IT. Montgomery and Frank Morgan for their intereit and encouragement and t o Gulf Research & Development Company for permission to publiqh.
IIASS SI'ECTROXETRIC STUDY OF HIGH TERIPERATURE REACTIOXS OF H 2 0 ( g ) ASD €ICl(g) WITH Na20 ASD Li20'a'b BY RICHARD C. SCHOOXMAKER? ASD RICHARD F. PORTER Departnient of Chenristry, Cornell I'nii'ersity, Ithaca, Xew York Received October 26, 1959
The gaseous prodiicts formed tiy reactions of H?O(g) and HCl(g) ivith condensed I&O and Ka20 phases a t high temperatures have been identified mas? spectrometricnllp. At 1011- 15-ater pressures the major spe-ips produced by reaction of H 2 0 ( g )with Li?O(s) is LiOH(g). I n ?ia20-HrO(g) experiments, monomers and dimers of NaOH are formed. A small concentration of Sa013 trimer in the vapor is also indicated. An estimate of the relative heats of dimerization of LiOH and KaOH 1%-asobtained from the resu!ts of experiments with the Na,O-T,i,O-H?O(g) system. I n the systems Na& XnCI-HoO(g) and Sn?O-HCl(g) the mixed anion dimer, Say(OH)C1(g),is formed in addition to Na?(OH)!(g) and KalCl,(g). Recsuse of ambiguities due to the ion fragmentation patterns of Xan(OH)?.XasCl? and Ka:(OH)Cl the existence ?f the latter species is necesmrily inferred from the dependence of ion current intensities on flow rate of the react,ant gas. Similar behavior was noted for the system Sa&TaF-H?O(g). Some previously proposed processes of ion formation by electron impact of sotlium hydroxide vapors have heen confirmed. Comparison of ion fragmentation patterns obtained in studies n i t h pure IdC1 and with the Li?O-HCl(g) system indicates that Li' ions are formed from both LiCl(g) and Li?Cl?(g)molecules a t high i:lect,ron energies.
Introduction Several rnass spectrometric studies of the vaporization of alkali metal hydroxides3 and halides4 have been reported recently. Identification of the vaporizing specie? from condensed lithium hydroxide by mass spectrometric means is made dif(1) (a) This r?search 1 ~ 3 supported s by t h e U. 5. Air Force through t h e Air Force Office of Scientific Research of t h e Air Research and Development C o m m a n d under contract No. AF 18(G03)-1. (h) Part of a thesis presented by R.C.S. to t h e faculty of Cornell Univeisity in partial fulfillmeit of t h e requircinents for t h e degree of Doctor of
ficult because of the high decomposition pressure of water. In the present work, studies were undertaken to provide information concerning the stabilities of gaseous species in the Li-OH system, to investigate the existence of gaseous dimer. containing mixed anions (R12(0H)X(g),where X is a halide) and to provide additional information related to several mechanisms of ion forma t'ion which had been proposed previously. Experimental
Philosophy. (2) General Electric Prrdoctoral Fellow (1938-1950). (3) (a) R . F. P o r t e r a n d R. C. Schoonmaker, J . Chem. Phys., 28, 464 (19%); ib) R . C. scliuon:nnker and R. F. Porter, ibid., 31, 830 (19.59). (4) (a) J. Berkon-itz nnd TV. .I.Chupka. ibid., 29, 653 (1988); ( b ) R.F. Porter and R . C. Prhooninaker, ibid.,29, 1070 (1958); ( c ) Slilne, Klein and C u b i c c l o t t i , ibid.. 28. 718 (1968). ( 5 ) R. C. SchiionmaAer a n d R. E . Porter, ibid., 30, 283 (1959).
The experimental method involves the use of a mass spectrometer to identify the vapor species effusing from the orifice of n Knudsen cell containing a condensed phase over n.hich various reactant gases m a r be leaked from a source external to the instrument. Thermochemical data may be obtained from a studr of the ions produced and the dependence of ion current intensities on temperatures and leak rate. Effusing neutral gaseous molecules, formed by interaction
m5
Yol. (ik
RICHARD C. SCHOOXMAKER AND RICHARD 3'. PORTER
458
top except for a hole to pass effusing vnpors, TI:LS mounted between the furnace and shutter plntr ant1 coniit>ctrdthroiiglr heavy gauge copper rod and tubing to x i c s \ t i m d nietul Dewar flask. Conduction of heat from tht, m l r ~ d c rto this liquid nitrogen or ice-water fillcd reservoir TV maintain the ion source a t a temperaturc effects were not encountered. Runs were made with the system: SalO(s), Li2@(s), Na&KaF(e) and Na?O-SaCl(c), with uater vapor n.i leak gas; and. SazO(s) and Li20(s) 711th HC1 as leak gas. Further experiments were conducted TT ith purr Sac1 and LiC1. Samples of LieO and NazO were obt:iined l)\ thrrmal decomposition of the hydroxide and pcio\ide, I cxspect ivc.1~.
1
1
Fig. 1.-Schematic diagram of high temperature furnace assembly with leak system: A, crucible; B, electron bombardment shield; C, filaments; D, radiation shield; E, thermocouple; E', furnace pot flange; G, quartz spacers.
Results NazO(s)-H20(g).-Studies with Na20(s) and HzO(g) were conducted in the temperature range 534-866'. Ion species which were detected w r e Na+, NaOH+, Ka20+, S a 2 0 H + a i d in one 1'1111 Sa3(0H)z+. With the exception of S:i,(OT1):-, these ions have been observed iii .imil:ir r'xprwments3 with vapors from liquid S:LOH. Oil thci basis of these previous observation., the fornxtioll of these ions by electron impact in the mi PC was attributed to primary ionization proc ( L e . , by either simple ionization or f ragnieiitat The occurrence of second-order reactiom as, for esample, NaOH(g) NaOH+ + Xa20H' 011 habeen ruled out as a major process other alkali hydroxides3b it is asqi1rne.d that Li- a i d LiOH+ are formed by the re-pecti1-e mechanismq of dissociative and simple ionization of the gaqeous LiOH molecule. Clearly, condensed LiOH has an appreciable decomposition pressurr in this temperature range since the nater partial pressure due to the leak was never high enough to form any appreciable concentration of I,iOH(c~)(the LiOH+ ion intensity dropped nhriiptly to zcro upon termination of the leak). In thew .dudies the upper limit to the leak gas prc5wre whi('h is attainable in the effusion cell 1- determiiird by the pump out capacity of the ~ - a r ~ i uequipmelit m in the tem. If the evacuation rate i- t..;cecded, the effusing gas builds up a ?~ncbgroiincl prcwire in the mass spectrometer ith r o i i ~ t ' q i i ~ ~vattering it effects. The upper Innit i n thcb operational prewire of H20 in the effiiyion cell i. e*timnted to be about mm. 7'hr di-wriatioll prc-ure of LiOH(s) in the tcniperat (ire i3::nge investigated ir apparently high enougn that it is impossible with the present eciuipmeiit to nicreaqt' the xater vapor partial pres'lire to n v:iliw whcre T&(OH),(g) can he detected. Li,O(s) -Na20(s)-H20(g).-A mixture of Li?O(s)T;i?O(-) n-ith 13?O(g) has been investigated in ortlcr to o1)tnin information on the relative stab~litici of IJi2(OH)z(g) and Na2(0H)2(g)_. The ni:iior i n n y w i e i detected were Sa+, haOH+, ~
Xa20H+, LiOH+ and the mixed ion SaLiOH+.
A mass spectrum record of the ioti speck5 i. sh0TY-n in Table I. tTnder the condition.: of thew
experiments I N a 9 0 Et < l N a O I I + . Thi.: \ituativn probahly results from thc dwreaw i i i tllcrmodynamic activity due to miving and from the lo\\ water pressures employed. From the ion ciirrciit data obtained in these studies, it is po-hihlt. t o estimate the difference in heats of dimerization of KaOH(g) and LiOH(g). The mixed equilibrium constant, Keq = ( I S ~ L I O H +Iz, (IL~~OH +)(1 for the reaction Na2(OH),(g) Li2(OH)2(g) = 2KnLi(OH)2(g) is estimated to lie betireeii 4 and 8 from considerations involving the entropy change for the reaction and values for the correspoiiding reaction with lithium and sodium fluoride dimer,.. This assumption places the calculated intensity of LizOH+ between 0.8 and 1.4 units relative to the data in Table I. This small intensity would not be distinguishable above background a t Inass 31. -1 relative equilibrium constant calculation5 using total ionization cross sections, gives a value betiteen 4 and 6 kcal.jinole of dimer for the difference in enthalpies of dinierization for LiOH(g) :Ind S a O H (g). This magnitude for AH'L~OH - AH"s~oH (dimerization) has been predicted on the basis of the analogy between the alkali fliioridw and hydroxides. 3b
+
1 10s SPECIESFORVED WITH IJ& THE P R E S E ~OCF E HO(g) T4BLE
MASS SPECTRCMOF NapO(c)I N T. OK.
IT&+
I h OH+^
Ihin*o~-
I\,i
l o ~ l + II> O H -
1082 63 100 85 21 B 5 8 Ionizing electron energy = 100 volts, total intensities relative to NaOHT = 100.
TABLE I1 DEPEKDENCE O F 10s CURREKT INTEXSITIES OS F120X' OF HCl(g) OVERNalO(s) .\T T = 9 ' i R O h . Relative change in leak rate of H C l k )
ISaOH+ I S ~ O Ha -
Leak on 100 Decrease 95 Increase 82 Small Leak off Leak on 45 Increase 48 Ionizing electron energy =
I N ~ ~ o I HB~~ ~ c IIs- CI
2 1 3 0 1 2
27 14 68 0 0
2 5 5m:ill 1 3 23 100 volts.
I
00 30 110 0 0 25 57
Na20(s)-HCl(g), Na:O-NaCl(c)-H20 (g) and the Existence of Nan(OH)Cl(g) Molecules.-In the temperature range around 700", the ion specieq observed when HCl(g) was passed over Ka,O(s) lvere T a + , NaCl+, SaOH", Na2C1+ and S a 2 0 H - . The ?\'&el+ and T\iaOH+ were usually of about equal iiitensities n-hile HC1+ n-as virtually uncletected indicating that the uptake of gas is nearly complete. In most cases the ions formed from polymeric species were lower in intensity than those formed from the monomers. This behavior, which is the result of the activity effect, is t o be expected in these experiments and is in contrast to the observations of the vapors of pure S a C l and S a O H in which ions representing dimers are the major species. The interpretation of the mass spectrum
RICHARD C. SCHOONMAKER A N D RICHARD F. PORTER
460
T'ol. G I
TABLEIV is complicated in the present studies as a result of the fragmentation pattern of the dimers. The DEPEKDENCE OF I O N CURRENT ISTEXSITIES os FLOW OF presence of dimers containing mixed anions, Ka2Il,O(g) OVER Na20-NaF(c) AT T = 1066°K. (OH)Cl, should lead to the formation of Na2Cl+ Relative and Na20H+ ions on electron impact, but these rliange in INS.+ leah l a t e same species arise from fragmentation of Na2C12(g) of HzO(g) I Y ~ + ~ I N ~ ~ F + ~h aOH+ rY,,OE+ and Fas(OH),(g), respectively. Under no condi0 0 0 0 4.7 100 Leak off tions have Na2(OH)Cl+ ions been detected. For 38 28 31 2 6 Increase qystems with mixed cations and a common anion Isa+is corrected a Ionizing electron energy = 100 volts. (e.g., NaOH-KOH) the fragmentation pattern is intensity clue only to fragmentation of SaF(g). simplified because only one unique ion (e.g., SaI