INDUSTRIAL AND ENGINEERING CHEMISTRY
1388
Conclusions The results of studies on the effect of air flow, air pressure, and the degree of agitation on the production of sorbose from sorbitol by submerged growths of Acetabacter suboxydans show that these factors are of great importance in determining the duration of the fermentation period. The combined effect of the proper adjustment of these factors is emphasized by the results shown in Figure 4, curves R and S. Highly aerobic conditions are most satisfactory for the rapid conversion of sorbitol to sorbose. The use of 20 per cent sorbitol solutions in the submerged growth process provides a commercially feasible method for sorbose production. The method developed for the large-scale preparation of a highly active inoculum overcomes one of the chief obstacles encountered in the industrial application of such a process. Translation of the small-scale results to semiplant scale operation is planned for the near future.
Acknowledgment The culture of Acetobacter suboxydans used in thie work was kindly furnished by W. H. Peterson of the University of Wisconsin, and the sorbitol sirup by the Atlas Powder Company.
VOL. 29, NO. 12
Literature Cited Bernhauer, K., and GBrlich, B., Biochem. Z., 280, 375 (1935). Bertrand, G., Compt. rend., 122, 900 (1896) ; Ann. chim. phge., 8, 3 (1904). Boeseken, J., and Leefers, J. L., Reo. trav. chim., 54, 861 (1935). Fulmer, E. I., Dunning, J. W., Guymon, J. F., and Underkofler, L. A., J . Am. Chem. Soc., 58, 1012 (1936). Herrick, H. T., Hellbach, R., and May, 0. E., IND. ENO.C H ~ M . , 27, 681-3 (1935). Kluyver, A. J., and de Leeuw, F. J., Tijdschr. Vergelijk. Geneeekunde, 10, 170 (1924). Maurer, X., and Schiedt, B., Biochem. Z., 271,61-3 (1934). May, 0. E., Herrick, H. T., Moyer, A. J., and Wells, P. A., IND.EN@. CHEM., 26, 575 (1934). Moyer, A. J., Wells, P. A., Stubbs, J. J., Herrick, H. T., and May, 0. E., Ibid., 29, 777-81 (1937). Sazerac, R., Compt. rend., 137,90 (1903). Shaffer, P. A., and Hartmann, A. F., J. Biol, Chem., 45, 365 (1921). Vissert Hooft, F., Dissertation, Delft, 1925. Waterman, H. J., Centr. Baht. Parasitenlc., (11) 38,451 (1913). Wells, P. A., Lynch, D. F. J., Herrick, H . T., and May, 0. E., Chem. & Met. Eng.,44, 188 (1937). Wells, P. A., Moyer, A. J., Stubbs, J. J., Herrick, H. T., and May, 0. E., IND. ENG.CHEM.,29, 653-6 (1937). RECEIVED September 9 , 1937. Presented before the Division of NIedicinal Chemistry at the 94th Meeting of the American Chemical Society, Roohester, N. Y., September 6 to 10, 1937. Contribution 278 from the Industrial Farm Products Research Division.
CHERUBS IN THE ALCHEMIST’S WORKSHOP By David Teniers, the Younger
P
Again we return to that reliable old standby of the collector of alchemical prints, the younger and more famous of the two David TE
The painter, in a lightkr vein than usual, has here portrayed the beneficent cherubs furthering the workofthe absent alchemist, who,on his return to work, will no doubt find a goodly amount of gold in his crucible after a forced cessatiorj, of work because of “lack of funds. Please note thepaucityof equipment and a paratus needed by the cherubs to \ring about the transmutation.
e A detailed list of Reproduotiom Nos. 1 to 60 appeared in our imue of January, 1936 page 129 and the list of Nos. 6 1 to 7 2 appeared in January 1937 page 74, where also will be foun’d ReGroduction No. 73. Reproduction No. 74 a pears on page 168, February issue, $0. 75 on page 345,. March issue, No. 76 on p>ge 459. April issue, No. 77 on page, 504, M a y issue No 7 8 on page 710 June issue N o . ’ 7 9 bn page 776, Jul; issue, No. ’SO on page 945, August issue No 81 on page 1039 Septemb‘er issue No. 82 on page( 1134, October issue,’and No. 83 on page 1276. November issue.
i
Dummy page Please scan
VOL. 29, NO. 12
INDUSTRIAL AND ENGINEERING CHEMISTRY
1390
The stripping nwnsurcments were inado by the flow method in all casea where the total pressure of mercaptan and water waa bclow atmospheric. The sodium mercaptide solutions were charged t o a series of four glass bubblers, about 50 cc. to each vessel. Connections between bubblers were all glass, and the whole unit was immersed in a water bath heated to the desired temperature. A measured volume of dry nitrogen IYM passed through the series, and the exit gas ui&s scrubbed in a weighed absorption train containing caustic solution and solid calcium chloride. The relation between volume of nitrogen passed and volume of solution tested was such that the change in mercaptan concentration of solution was negligible; further assurance on this point wns the se arntion of the four bubblers and practical establishment of equitbriurn in the first unit. Actual proof that equilibrium was established bctneen liquid and gas was obtained by testing ure liquids of known vapor pressure. The totar increase in weight in the absorption train represented mercaptan plus water. hIcrcnptan was determined by titration with iodine, and water was determined by ditference. The principal source of error in these measurements WBB the temperature, no attempt being made to hold closer than 1’ F. to the value reported. The point of particular interest, however, was the ratio of mercaptan to water vapor pressure; therefore minor errors in tern erature mcasurement should cause only a negligible error in the {nal interpretation of results. In the single instance where the total prcssure wns above atmospheric, the stripping vapor pressures were measured by the static method, the sodium mercaptide solution being charged to a steel bomb which \VRJ then immersed in an oil bath heated to t h e dcsired temperature. As the tcmperature within the bomb increased, the air was relensed and finally only mercaptan and wnter vapor remained. When thermal equilibrium was well estnblished, as evidenced by a pressure gnge, a snmple of the vapors was absorbed in a weigtied caustic-calcium chloride train and analyzed as above. The sample release valve was immersed in the hot oil bath, thus avoiding any chance of refluxing during sampling.
Derivation of Equilibrium Equations
Using partial pressures as a measure of the concentrations of RSIi and H,O, and assuming Dalton’s law to hold for mixtures of steam and mercaptan vapor:
a -X, [(OH)-] 40 32 9
Substituting Equations 3, 4, and 5 in 2: X. 18Y ET
32
K($-%> Solving for Y and simplifying:
Y-(
40
18K a
- 1.25X,
+ RSH = RSNa + 1320
(7)
Equation 7 is the equation of the equilibrium curve for stripping mercaptan from caustic solution with steam.
Absorption Equilibrium Again using partial pressures as a measure of concentrations of IiSH and H,O, and assuming that mercaptan dissolved in naphtha follows Raoult’s law:
IRsHl (HsO]
- PXnM 32dPw
(8)
Substituting Equations 8, 4,and 5 in 2:
s,
Equilibrium concentrations of methyl or ethyl mercsptan, whether between cniistic solution and steam or between cnustic solution m c i naphtha, may be cnlciilnted from thc equilibrium constnnt for the following reaction:
XaOH
(5)
Solving for
Xmand simplifying:
or, written in ionic form,
(OH) - -t RSH
=
(RS)-
+ Hi0
The eqiiilibrium constant for this reaction may be written:
The brncketed symbols represent molal concentrations of the vsrious reactants. T h e symbols used in the equations have the following meanings:
X. X.
-
mercaptan sulfur per gal. of naphtha, lb. mercaptan sulfur pcr gal. of caustic solution, Ib. Y = mercaptan sulfur per It). steam, lb. a KaOH per gal. of sulfur-free caustic solution, lb. P, = vapor prwaure of water over caustic solution, Ib. per =
sq. in. abs.
vapor pressure of pure mercaptan, 11). pcr sq. in. nbs. .ti = av. mol. weight of naphtha d = dcnsity of naphthn, Ib. per gal.
P,
=
Stripping Equilibrium Yrorn Jkjtr~tiori1:
Let
xm
Ip
(a
- xe 1.25 )
Equation 12 is the equation of the equilibrium curve for absorbing mercaptan from nnphtha with caustic solution. The relation between K and K‘ is given by Equation 11.
Results Experimental results of the absorption tests are presented in Table I. T h e values of a shown were calculated from d a t a (7) on specific gravity of caustic solutions. Values of K’ were calculated from Equation 12. In T a b l e I1 the arithmetic average of values of K’ for each mercaptan is s h o r n . Corresponding values of K were calculated from Equation 11. Values of d and d f were calculated from the hydrocarbon analyses of tho two naphthnss. Valuca of P, and P , were obtained from thc Chemical Engineers’ Ifandbook (f3)and the International Critical Tables (s), rcspectively. Data on stripping equilibria are recorded in Table 111. Values of K were calculated from Equation 7. Tho numerical value of the equilibrium constant, K , depended primnrily upon the incrcnptnn consicicred, thc strength of caustic usmi, and t h e tempemturc. Conscqucntly, in Table 111, wlicn sevcral determinations mcrc mnde with a given nmcnptRn, using the samc strcngtli cnustic nnd n constant tcmpemtrire, nn arithmetic nvwngc K hna I m n computcd,
DECEhlHEII, 1937
INDUSTIIIAL 2LND ENGINEERING CIIEMIS'rRY
1391
WTAN W U R
us7*L.
I t is to be cspcctc.tl th:it o i w short rangcs of tcniperatiire, whcm the natiiral log:ir.itlini of cquilibriuui coiistnnt li is plottcd against t h e reciprocal of the nlsolute'ternpcraturc, n straight-line relntion will result. In Figure 1 the vslues of I\' i n Tnblcs I1 snd 111 ivcre plotted in this way. The dntci show c 1car1y t hn t t,I i e rcl n t ion lxt wee n cq 11i I ibr i ti 11 1 cons t:i 11 t r.1
r
CAUSTIC
FIGURE2
1
where 7'
-
nbs. temperature,
O
K.
In l%gir.c 2 the c:Llculated equilibrium curvc nntl dntri for :huorption uT ctliyl mercaptan frorn pentme-pcntenc in 12' 138. cbnuxtic :it 75' F. are presented. The curve was cnlculated from Figure 1 and Equations 11and 1 using appropriate values of o, d, IJ.,P,, and ;M (8,12). Fig! i re 3 also presents similar d n t s for stripping ethyl mercaptan from 12'BB. caustic a t 190" T?. In lmtli cases the close agreement between tlic 11nt:t :inti the cnlcdnted cquilibrinm curve indicntes the vulidity u l tlic :tlpd.iraic form of eqirilibriuin Equations i nnd 12.
Design of Equipment 'l'lic
lIcsiKii iiictliotls iised wen? (iwt*rilmI l ~ y Iivirns (/t) nntl LJ. \Vdlxr, Imvis, mid XfcXtlniiis ( 1 8 ) . 1.c.t 115 :isftii~iietlitil it in rlenirctl to clwiun :L r n u s t i c - t r t w t i i i r sptciii to retluw tlic iiiercrlptan sulfur c*oiitciit of :I liqiiirl pviitmrjxi1t:iiic ciit froiii 0,2:3 to U.006 \wiglit per ceiit. 'I'rartinr is carried oiit in :L couiiterctirrcnt, picked towcr. I h i r i i is Imcd 011 tlic fr~llowingassiiii~j~tiuns:( ( I ) iisc uf 12' 116. railstir :wcl (6) :I truiting tenipcratiil-c of 73" F. 'rlie rlcsigii clin~rnrnTor ttic prcsciit csai~iplei y givcii i i i tlw tipper part of 14gure 3. 'l'he equilibrium line is calculntcd as described n b o w Tlic o l w n t i n g liiic! is Kscd 112' tlic iiilct aut1 the outlet nicrcnpt:iij coiiccntrntioiis. Tlic stoicliioiiictric jy:ipliiv:tl
INDUSTIITAI, AND ENGINEEIUNG CHFSMISTRY
1393
units eniployetl result in a strnight operating line. In locating the operating linr, soiiic jiiclgnient is rcquired. In view o f practicnl liinitntions on tlic hcight of the treating tower, it is probnblc t h i t riot over tiyo or tlirce tlicorcticnl plates should b e used. 1111;igiire 8 two tlirowticnl plates hnvr bccn nssuincd. The slope of the o1)ernting lirw gives tlic required cnristic circulntion ratc--O.@~X gnllon 1)cr gnllnii of nnplitlia.
-
5 WtORETKAI. R A T E S STEAM RATE 4.7 LB/GAL CAUSTIC
ool
have alrcady bccii sct ill the nbsort)er dingrain of 1Jigiirc :3--1ianicly, thc tncrcriptnii sulfur i n tlic caustic entering n i i t l Ica14ng the stripper. Vivc thcowticnl platw htivc 1)ceii nssiiniccl. Ttic corrcspontling stcnrn rate rcqtiircd is 4.7 poilids of stearn pcr gallon of C:iuhtic, 11s obtainctl froiii the slope of tllc opcrrhng liiic. Iiigurc 3 reprcscnts onc ol)rrol)lcsolutioii to the prol)lcn) a& statecl-iianicly, to rcducc the Iiicrcapt:in s111furcclrlteilt of :i liquid pcntenc-pcntnne cut from 0.23 t o 0,006 weight pcr cent. Assuming t l i n t tlic equip~ncnthns Imn built and tliat thc number of theorctical p l a t c ~avnilahle ic tlir alisorbcr and stripper is known, the caustic circulation rate nrid corrcsponding s tri 1) pi II g s tcnm I-cquirenicnt s ~ n n y n r i c t 1 over fairly witlc limits. Thc optimum values (Icpciid so lurgcly 011 local costs t h a t their dctermination is beyond the seopc of this paper. In thc csainplc givcn hcrc thc cicsign nicthotls have bceii :tpplictl to ethyl incrcaptan. T h y are cqually applicnble to tlic rcmoval of methyl mcrcaptnn from butzinc-butcnc fraction. H O I V C V C cthyl ~ , niei*cnptnii is niorc diffcult to rciiiovc from n a p h t h by caustic scrubbing, wl~ercasmethyl mcrcnptan is morc difficult to strip from thc caustic. It may be clcsircd to rcniovc both mctliyl atid ethyl mercaptan from :i 1)~it~iic-l)crit~iric Ilnction. In this CRSC, coinputation of sitnultruncoris scrilihing or stripping of Imtli ~ ~ i c r c n p t a ni ss somcwhat lril)orioiiq, I t is suggcstctl th:it for this case, absorbcr tlcsign he txiscci on cquilibriuin for ethyl iiicrciiptfiti, and strippcia tlcsigii 011 cqrlilibriuln for inctliyl i ~ c r e : ~ p t a n , of tvliicli
Literature Cited
o
00s
Kc
010
01s
010
I 16 MERCAPTAN SULFUR
GAL.CAUSlIC
l~lclvl
