Phase diagrams of reciprocal molten salt systems. Calculations of

Chem. , 1974, 78 (11), pp 1091–1096. DOI: 10.1021/j100604a009. Publication Date: May 1974. ACS Legacy Archive. Cite this:J. Phys. Chem. 1974, 78, 11...
0 downloads 0 Views 837KB Size
Phase Diagrams of Reciprocal Molten Salt Systems

1091

!sof Reciprocal Molten Salt Systems. Calculations of ~ i q ~ i ~ - ~ i q u ~ d Phase Bia ~ b i ~ i la~ ~ aboungi,*lb H . Schnyders, M. S. Foster, and M. Blander Chemical Engineering Division. Argonne National Laboratory. Argonne, lllinois 60439

(Received January 2, 7974)

Publication costs assisted by Argonne National Laboratory

Liqui(3-liquid miscibility gaps of ternary reciprocal molten salt systems may be calcuiated (z priori from confoirmal ionic solution theory utilizing data for the four subsidiary binary systems and the four pure constituents. lising a modification of the second-order theory, we calculated miscibility gaps for hypothetical systems containing two cations, A+ and B+, and two anions, C- and D - ; the results show that the sizes of miscibility isotherms depend on the standard free-energy change for the metathetical reacBC as well as on the binary interaction parameters whereas the asymmetry in tion AC d- BD F! AD the shape and position of miscibility isotherms depends only on the binary interaction parameters. The caiculated sizes and positions of miscibility isotherms in the Na, T1/IBr, NO3 system are in good agreement with measurements

+

Introduction The conformal ionic solution (CIS) theory2a has been shown to lead to predictions of liquidus temperatures a priori which are in gosod agreement with measurements for a variety of reciprocal salt ~ y s t e m s . This ~ b ~ lends ~ ~ ~ confidence in the theory for predictions in systems which exhibit fairly large deviations from ideal solution behavior. A preliminary discussion has been given of miscibility gaps and upper consolute temperatures2b as well as some limitations of the 12neory.~However, no detailed examination of the theory has been undertaken for systems in which deviations from. ideality are so large that liquid-liquid miscibility gaps foccur. A large number of such systems are known5 arid it is important to examine the applicability and limitations of this second-order perturbation theory for such systems. In this paper we present a test of the theory for the system Na, T1 1 ) Br, NO3. In addition, we discuss the influence of particular parameters and ionic i.nteractions on the shapes and position of miscibility gaps. An empirical generalization of the theory is suggested to eliminate one of its limitations. heory for Ternary Reciprocal Systems2a A ternary reciprocal system contains two cations (A+, B'-) and two anions ( ( C - , D - ) and is represented by the symbol A, B / I C, D. There are four constituents (AC, AD, BC. BD), any three of which may be chosen as independent The activity coefficients, yi,can be calculated from the CIS theory. For BD, for example, the activit:y coefficients art: given by

RT In yBD== X,.XcAG" -t- X,X',(Xc- - X,,)X, X(.lX..IIYD -1- X B X C ) X B -I- X*(X*XLl X,X,)X, X,Xr(X4 -- . X B ) X C iX,JC(X.AX" i- XBX,

+

+ -

xBxLI)Ad

+ (1)

+

where Xi is the ion fraction of ion i ( e . g . , X 4 = n,/(n,A nH),X c = nC/(nC + n.D) where n's are number of moles of ions indicated), 9G"is the standard molar Gibbs free-energy change for the reaction AQl) -i- EID(1)

e AD(1)

+ BC(1)

(2)

A, is an energy parameter in the second-order equation for the excess free energy of binary systems with the common ion i

AG,4E ACBE AGcE AGDE

=

XcX,XA

(3) (4) (5) (6)

XCXJB = X,XBA, XAXRAD

=

and A is approximated by

h

-(AG")'/%ZRT

(7) All the parameters may be estimated from available data, AGO from data on the pure salts, X, from data on the binary systems, and Z is a coordination number and has been taken as 6. Consequently. the theory utilizes data obtained from lower order systems to calculate properties of ternary systems. Equations similar to eq 1 can be obtained for the activity coefficients of the other constituents AD, BC, and AC by changes in the subscripts in eq I and the constituents of equilibrium (2) as follows: for AD, chaQge A B and B A; for AC, change A B, I3 A, C - + D, and D -* C; D and D - * C. The chemical potenfor BC, change C tial of BD is given by p B D - p B D a = RT In ~ B D= R T In XJ,?;,, = ApBD (8) where hRno in the chemical potential of pure liquid BD and u H is~the activity of BD in solution.

-

=

-

+

-

-

Calculation of Liquid-Liquid Miscibility Gaps The thermodynamic conditions for defining miscibility gaps for ordinary ternary systems are relatively ~ i m p l e . ~ Because ternary reciprocal systems contain not three but four entities and have four possible constituents, the conditions for defining miscibility gaps must be stated differently although the fundamental equations are the same. If one defines the three components as then the molar Gibbs free energy of mixing, L I G , ~is, given by the expression The Journal

of Physical Chemistry, Vol. 78, No. 11, 1974

1092

AG,,

M. L. Saboungi, H. Schnyders, M. S. Foster, and M . Blander ==

-XC&F.C -t x ~ A p . 4+~ ( X D - XA)APBD (9)

where the cbem+ai poterltials of any constituent, as for example BD, may he calculated by equations as6 ~ B D (i’nncrn/anidn, 4- (dnAGm/anD)n, (10) where n is the total number of moles ( n = n 4 + n13 = nc nD), and the derivative is taken a t constant T and P. If at constant T and P one considers the surface for AGm with the magnitude of AG, plotted in the coordinate perpendicular to the composition square, then one may define the conditions for miscibility gaps geometrically in terms of Ihir: surface.7 Any two compositions on this surface which are in cbquilibrium with each other must be in a single plana whicki is tangent to the AG, surface a t two points k’ and k” which correspond to these two compositions. If the two points are represented by a prime (’ for h’) and doutile prime (” for k”) then we can express this condition by ?,heequations

+

The calculation utilizes a method for minimizing the sum of the squares of nonlinear functions of several variables developed by P0we11.*~~ (See paragraph at end of text regarding supplementary material.) Using this computational method, we calculate many pairs of points at fixed temperatures which define the isothermal cut of the dome of immiscibility. As the temperature is increased, the area of these isothermal cuts decreases until it shrinks to a point a t the upper consolute temperature, T,. Above this temperature the salts are completely miscible. The calculation of upper consolute temperatures from the CIS theory has been discussed.2b An approximate expression is

An examination of the theory3 for a hypothetical case in which XA = AB = Ac = AD = X has shown that an artifact is introduced into the calculations of miscibility gaps from the theory for Z 5 5.3 when X = 0. This artifact also occurs for cases in which the binary interaction parameters, A,, are negative enough so that the ratio of the term in A to the others in eq 1 i s greater than the ratio when Z = 5.3. This artifact in the calculations occurs because the theory has only been carried out to second order and the dependence of this artifact on values of X makes the self-consistent treatment of miscibility gaps impossible for some systems without modification of the equations. The calculations made previously3 suggest a modification to the term :I in eq 1 which leads to a self-consistent treatment of miscibility gaps such that the artifact will not appear in the calculations for very negative values of X. This modification i s

If one expands (23), the form of the added terms in (23) appear to be consistent with third- and fourth-order terms in the CIS. The term in (AG”)(ZA,) i s probably related to some of the third-order terms and the term in (ZA,)2 is probably related to some of the fourth-order terms. The From eq 11-1’7 one deduces the well-known conditions purpose of this paper is to test the eq 1, 7 , and 23 and we &AC’ = AP..W” (18) will show that we obtain a good representation in the Na, & A d = APADI) (191 T1 / I Br, NO3 system. In addition, we will calculate miscibility gaps in hypothetical systems t o gain an underApt,,' = BpuRc’’ (20) standing of the influence of the different parameters in Allg; = ApB/’ ( 21) the theory on the location, shape, and size of the miscibility gaps. These considerations provide some insight into which is statement that any constituent of the two soluthe interpretation of the topological characteristics of mistions in equilibrium with each other has the same chemicibility gaps and illustrate features which are observed. cal potential in boich solutions. Only three of these four equations are indepiendent. Consequently, any three of the Miscibility Gaps in Hypothetical Systems four eq 18-21 or the three eq 11-13 may be utilized in conIn this section, we calculate miscibility gaps from eq 1 junction with eq I, 8, or 9 to calculate the pairs of immisand 23 to see what influence the parameters AGO, Ab, and cible compositions which are in equilibrium with each Z have on the shape and size of the miscibility gaps and other. In our computer calculations we generally utilized on the positions of the tie lines which connect individual three of th.e eq 18-21 and occasionally checked the selfpairs of points which are in equilibrium. All the calculaconsistency of the results from eq 11-13. tions in Figure 1were made for AG” = 18.0 kcal mol-1 For any given temperature, the problem of finding conThe calculations in Figure l a were for the case in which jugate pairs of liquid compositions which are at equilibriall X,’s are zero and Z = 6, in Figure I b when all Xi’s are um with each other ( ; . e , ,tie lines) is reduced to the probzero and Z = 5.5, in Figure I C when three Xi’s are zero, A 4 lem of find.ing a composition which simultaneously satis= 3500 cal mol-I, and Z = 6, and in Figure Id when three fies three of the four eq 18-21. An analytic approach to X,’s are zero, X 4 = -1500 ral mol-1, and Z = 6. Figures l a the problem i:s precluded because of the logarithmic terms and one must .resort t o nunierical methods. aed Pb are universal curves which are the same for any The Journal of Physicai C!iernistry, Vol. 78, No. 7 7 , 7974

Phase Diagrams of RecirJrocal Molten Salt Systems

1093 BC

0 00 BD

1

0 20

0 40

0 60

0 80

I AD

BC

AC

IO0

0 80

060

040

0 20

0 20

000 BD

040

0 60

080

IO0 AD

0 00 000

BD

0 20

0 40

0 60

0 80

IO0 AD

= Ac = AD = 0; (b) AGO = Figure 1. Calculated miscibility gaps for hypothetical systems: (a) AG” = 18 kcal mol-’, 2 = 6 , XA = 18 kcal mol-’, Z 5.5, XA = hc = AD = 0; (e) AGO = 18 kcal mol-’, Z = 6 , AA = 1.5 kcal mol-’. Ag Ac z- AD = 0; ( d ) AGO = 18 kcal mol-’, Z = 6 , XA = -1.5 kcal mol-’. AB = Xc = XD = 0. ._^

other value ~f A G O for temperatures such that values of (AG”/RT)are tile same as for the curves in Figures l a and 1b. The miscibility domes in Figures l a and l b rise steeply a t low reduced temperatures and flatten out a t higher temperatures. At low temperatures there is little difference between Figures 1a and l b . The flattening of the top of the dome O C C U ~ S a t slightly lower temperatures for Z = 5.5 and is more pronounced than for Z = 6 with the isotherms at higher temperatures being narrower and more elongated for Z = 5.6 than for Z = 6. The shapes are somewhat elliptical except a t low reduced temperatures where the shapes tendl to straighten so as to conform to the two corners of the composition diagram near the sta13’6)). ecause of symmetry, the tie lines ble pair (AC for these two cases in which all A’s are zero are all parallel to the stable diagonal zonnecting the AC and BD corners. In Figure IC for the case in which 2 = 6, XA is positive and Xu = hc = XI, = 0 , it can be seen that the isotherms subtend a larger area than in Figure l a ; they are also dis-

+

placed toward the AC corner and are asymmetric in a way such that the miscibility gap subtends a somewhat larger area in the AC-BC-BD triangle than in the AC-AD-BD triangle. In Figure Id for the case in which Z = 6, AA is negative and AH = Ac = AD = 0, the isotherms subtend a smaller area than in Figure l a ; they are also displaced toward the BD corner and are asymmetric such that they subtend a somewhat larger area in the AC-AD-BD triangle than in the AC-BC-BD triangle. Further, the tie lines are no longer parallel to the diagonal AC-BD, but are displaced in a manner as is illustrated in Figures l c and Id. These reflect the complex changes in relative solution properties for these idealized systems and indicate that for a proper thermodynamic analysis one needs measurements covering the entire composition square rather than only along the stable diagonal. These considerations provide us with an insight into the influence of particular ionic interactions on miscibility gaps and are significant for the interpretation of measurements. The Journal of Physical Chemistry, Voi. 78, No 1 7 . 7974

Saboungi, H. Schnyders, M. S. Foster, and M. Blander

M. L. NaBr

TlBr

I nn

Nn Qr 74'

436

60

'" 1

a

h

I

/ / /

o,80!

///

/

040-

02000

29'

f

306 NIINO~

164

NaBr

TINO,

09 OOOJ 000 NaN03 TlBr

I

I

I

I

I

I

040

0 20

Id

1

o ao

060

IO0 TINOS

NaBr

TlBr

I

oao-

c 060I

-

040-

0 20-

I

koNO,

NaNO,

I

I

I

I

I

I

I

TIN03

Figure 2. Experimental and calculated miscibility gaps for the reciprocal system (Na, TI 11 N 0 3 , Sr): (a) measured (ref 10); (b) calculated miscibility isotherms with AGO = 8720 - l . I T ( K ) cal mol-' and Z 6; (c) calculated miscibility isotherms with AG" = 9240 - 1I; T(K) Gal mol.-' and Z = 6; ( d ) calculated miscibility isotherms for AG" = 8720 - 1 . l T ( K ) cal mol-' and Z = 5.5.

TABLE I: Vislues ,of the Interaction Parameters Used in the CalcuXatiorm

Miscibility Gaps in the Na, TI I I Br, NO3 System The system Na, '61 1 1 Br, NO3 has been studied by Sinistri, Franzosini, and Flor*O who have measured isothermal liquid-liquid miscibility gaps at 440, 480, 520, and 560" given in Figure 2a as well as the intersection of the miscibility dome with liquidus surfaces. This work was chosen as it constitutes the most complete and detailed study of immiscibility in reciprocal systems which i s suitable for our calcn1atSons. The values of the binary interaction parameters for this system are small or positive. The The Journal ol Physicel Chemistry. Vol. 78. No. 7 7. 7974

values of A, given in Table I were taken from the calorimetric data of Kleppa and Meschelll for the TlN03-T1Br binary and calculated from the phase diagrams for the NaBr-NaNOs and NaN03-TlNOa systems.lz There are no measurements for the NaBr-T1Br system and the phase diagram cannot be utilized because of solid solutions. We have assumed that X B ~= 0. This is not too different from predictions of small values of AB,. one makes from theoretical correlations.6J3 A G O was calculated from available tabulated data14J5 and is represented by AGO = 8720 - 1.1T(K) cal mol-I. The miscibility isotherms calculated from eq l and 23 with Z = 6 are given in Figure 2b and are to be compared with the measured isotherms, For comparative purposes, the isotherms for the case in which AGO = 9240 - l.lT(K) cal mol-* and Z = 6 are given in Figure 20 and for the case in which AGO = 8720 - l.lT(K) cal mol-1 and 2 = 5.5 are given in Figure Zd, The correspondence between the measured isotherms and the calculations represented in Figure 2b is remarkably good. The shapes and positions of the measured and

Phase Diagrams of Recliirocal Molten Salt Systems

1895

TABLE 11: Cahculated and Measured Compositions Along the Miscibility Gaps at the Stable Diagonal in the Mole fraction of T1Br

__ __-_____--

440 480

520 560

__-

Measd

T OC

0 . IO 0.13 0.1!3 0.28

Calcd (Figure 2b)

0.95 0.92 0.88 0.80

0.10 0.14

0.93 0.90

0.20 0.30

0.87 0.80

calculated isotherms are very close. The isotherms are asymmetrically displaced somewhat in the direction of the TlBr corner and subtend a somewhat larger area in the NaNOS-NaB r-TIBr triangle than in the TlBr-TlN03NaN03 triangle This i s consistent with the fact that AT] is the largest and predominant interaction parameter influencing the shapes and position of the isotherms. Since is positive, this system is analogous to the system illustrated in Figure IC with AT, being analogous to XA. Thus we see a confirmation of the predicted displacements of the position of the miscibility gaps which result from the influence of interactions related to the binary interaction paramelers. Discussion The corresportdence of the measured and calculated phase diagrarns illustrates the potential of the theory in predicting miscibility gaps. This is further exhibited in Table 11 where we list measured and calculated compositions along the stable diagonal. The correspondence is remarkably good. The calculations are sensitive to the value of AGO As can be seen m Figures 2b and 2c, a change of 500 cal rno1-l makes a significant difference in the size of the miscibility gap %hen temperatures are not too far from the upper CQnS0)Utetemperature. We see that the size of the gap depende on AG"and on the interaction parameters whereas the asymmetry depends only on the relative values of the interaction parameters. Fairly small uncertainties in ,%Gomay lead to significant uncertainties in the size of the gap. The influence of the size of 2 is illustrated in 'Table T I and in Figure 2 with the gap being somewhat smaJ1er for Z = 5.5 than for 2 = 6. We have utilized eq 23 for A . Had we utilized eq 7 with Z -- 6 the gap n7ould have been wider. We obtain results similar to Fiigurc 2b rising eq 7 with a value of 2 = 5. When hi's are positive as in this case, the size of the term A in eq 1 relative to the other terms is smaller than when the X,'s are zero if eq 7 is used for A . It is this relative size that leads i.0 the artifact in the equations discussed previously3 with 1 he artifact appearing when A gets relatively large. This artifact is manifested by a double maxiinum in the calculated miscibility gap. Thus, in this case, the artifact does not appear when eq 7 i s used even when 2 = 5 whereas it appears when Z < 5.3 when all the A,'s are zero. Conversely, when the A,'s are negative, such artifacts may appear e w i for values of Z much larger than 5.3. For example, we havie performed calculations for the Li, Cs 1 1 F, C1 systena in which there are large negative deviations from ideality in the LiF-CsF and LiC1-CsC1 binaries ltisirig eq '7 with Z = 6 led to difficulties in calculation and the calculations we obtained were characterized by two (rather than one\ maxima in the miscibility dome. With the use of eq 23 these artifacts were suppressed. Our calculations in this system led to a calculated miscibility gap which, within the uncertainties in the parameters, was as long and asynmietric as the measurements of Bu-

~

Calcd (Figure 2 c )

0.08

_

_

_

I

_

_

Cakd iFigiire 2d)

0.95 0.93 0.90

0.IQ

0.92

0.10 0.14

0.15 0.23

0.19

0.87

0.40

0 90 0.85 0.78

khalova and SementsovalG and Sholokhovich, et al., i 7 along the stable diagonal (LiF-CsCl) but was much wider than indicated in the measurements along the unstable diagonal (LiCl-CsF). The reasons for this discrepancy remain unclear. The measurements in the Li, Cs / / F, C1 system are only of the intersection of the miscibility dome with the liquidus curves. A full test of' the theoretical calculations requires much more detailed data on the entire miscibility dome. Our calculations illustrate that the theory :s in accord with the most completely studied system suitable for our calculations. Only one other system, the Rb, TI / / Nos, Br system appears to have been studied in detail adequate for a similar comparison.5 Unfortunately, this system is very close in character to the one we studied. Clearly, more data on systems which are decidedly different from the Na, TI I I Nos, Br system are necessary for a real test of the range of applicability of the theory. It appears that the use of eq 1 and 23 with Z = 6 are adequate for the calculation of miscibility gaps and should also be suitable for the calculation of liquidus temperatures. This is the only example of which we are aware for which the entire w.iscibility gap may be predicted a priori from solution theory. Acknowledgment. One of us (M. L. S.)would like to acknowledge support received from the C.N.R.S. of Lebanon. Supplementary Material Available The description of the method of calculation will appear following these pages in the microfilm edition of this volume of the journal. Photocopies of the supplementary material from this paper only or microfiche (105 x 148 mm, 24x reduction, negatives) containing all of the supplementary material for the papers in this issue may be obtained from the Journals Department, American Chemical Society, 1155 16th St., N.W., Washington, D. C. 20036. Remit check or money order for $3.00 for photocopy or $2.00 for microfiche, referring to code number JPC-74-1091. eferences and Notes (a) This work was largely supported by the U . S. Atomic Energy Commission, (b) On a fellowship from the C.N.R.S., Beirut. Lebanon. (a) M . Blander and S. .J. Yosirn, J . Chem. Phys.. 39, 2610 (1963); (b) M. Blander and L. E. Topol, Inorg. Chem.. 5, 1641 (1966) M L. Saboungi and M . Blander, High Temp. Sci.. in press M . L. Saboungi, C . Vallet, and Y . Doucet, J.Phys. Chem.. 77, 1699 (1973) C. Sinistri, P Franzosini. and M. Roila, "An Atlas of Miscibility Gaps in Molten Salt Systems," Institute of Physical Chemistry, University of Pavia, Italy, 1968. M . Blander, "Thermodynamic Properties of Molten Salt Solutions." in "Molten Salt Chemistry," M. Blander, Ed.. Interscience, New York. N. Y . , 1964. I. Prigogine and R. Defay, "Chemical Thermodynamics," Translated by D. H.Everett, Wiley. New York, N. Y., 1954, M . J. 0 . Powell, Comp. J.. 8, 353 (1965). Quantum Chemistry Program Exchange, University of Indiana, Bloomington, I nd. C. Sinistri. P. Franzosini, and G . Flor. Gazz. Chim. (Rome). 97, 275 119671,

0 J Kleppa and S V Meschei, J Phys Chem . 87,688(1963) The Journal of Physical Chemistry. Val 78, No 1 1 1974

1096

J. C . Tou, L. B. Westover, and L. F. Sonnabend

( 1 2 ) P. Frarizo5.ini and 12.Sinistri. Ric. S c i . . A3, 439 ( 1 9 6 3 ) . (13) J, 1-urnsden, "Themodynarnics of Molten Salt Mixtures," Academic Press, London. 19156, (14) D. 13, Stul: and H . Prophet, Nat. Ref. Data Ser.. Nat. Bur. Stand.. No. 37 (1971).

(15) F. D. Rossini, eta/.. Nat. Bur. Stand. Circ.. No. 500 (1952). (16) G . A. Bukhalova and D V. Sementsova, Russ. J. lnorcl. Chem. 10. 1027 ( 1965). (17) M . L. Sholokhovich, D. S. Lesnykh, G. A . Bukhaiova, and A . G . Bergman, Dokl. Akad. Nauk. S S S R , 103, 267 (1955).

inetic Studies of Bis(chlorornethy1) Ether Hydrolysis by Mass Spectrometr J.

c. Tw,*

L. B. Westover,

Analytical Laboratories

and L. F. Sonnabend Designed Polymers Research, Dow Chemical U.S. A , , Midiand, Michigan 48640 (Received October 10, 1973) Publication costs assisted by the Dow Chemical Co.

The hydrolysis of bis(chloromethy1) ether (bis-CME) was studied in 2 N NaOH, 1 N NaOEI, water, 1 N HCl, and 3 N HC1. From the measured values of A&* and E*, it was interpreted that the mechanism of the hydrolysis of bis-CME was S N 1 in character in basic solution and shifted to an S N like ~ mechanism in acidic solution.

htroductior.. 160-cc vessel filled with the appropriate solution in which the hydrolysis rate was to be determined. Assuming no Ghloroniethy! methyl ether (CMME) is a commonly prior hydrolysis took place, the concentration so prepared used chlorornethylating agent in the manufacturing of was ca. 1 ppm. The vessel was placed in a thermal bath nic resins. One of the impurities present in and completely filled with the appropriate solution so that en found to be bis(chloromethy1) ether (bisno air space remained into which bis-CME might volatize GME). B.is-CME 'has been shown recently to be a very thus changing the concentration of the solution. The solustrong carcic.ogen in experiments involving skin painting,l tion was constantly stirred. The syringe used for injecting subcutaneow iajection,l and inhalation.2 Since these the bis-CME-acetone solution served for sealing the openfindings. possible industrial exposure to this simple moleing used for the injection. To avoid any bis-CME escaping cule has been of very much concern. Analytical techniques into the atmosphere, the syringe was not removed until for analysis of ppb levels of bis-CME in air have then after the bis-CME in the aqueous solution reached an unbeen developed quite r e ~ e n t l y . ~ , ~ Unlike CMME, which has been extensively ~ t u d i e d , ~ detectable level. The concentration of bis-CME in the aqueous solution was monitored using a CEC 21-110 douthe hydrolysis of bis-CME in solution has not been invesble focusing mass spectrometer coupled with a 15-head tigated to the best of our knowledge. In this report, the rates of bisCME hydrolysis were determined in 2 AT semi-membrane silicone fiber probe. The development of the fiber probe was published elsewhere by Westover, NaOH, 1 N NaOFI, water, 1 N HC1, and 3 N HC1. The Tou, and Mark.6 In order to detect the anticipated weak Arrhenius expressions for the hydrolysis rates were also signals, an amplifier with a gain of about 30 was installed determined. The systematic changes of the determined to amplify further the mass spectrometer output. values of A&* and E* from basic solution to acidic soluA few milligrams of benzene in a 500-cc reservoir at Lion are disc'assed in terms of the changes of the hydroly200" was bled int,o the mass spectrometer through a mosis mechanisms. lecular leak. The purpose for this was twofold. One was The least-squares linear fits of the rate and arrhenius for cali.bration as an internal mass marker for the peaks to plots were achieved with use of an IBM 1130 computer. be monitored. The second was for checking whether or not Experimentill Section there were variations of the instrum.ent sensitivity during Due to thl? high toxicity of bis-CME, any exposure to each experiment. The resolution of the mass Spectrometer this molecul? must be avoided. The rate determinations was adjusted to ca. 2000; which was enough to resolve a were, therefc're, carried out at very low concentration levdoublet peak at m / e 79 due to C5P3CHs+and C2H4QC1+-. els of the orlder of I ppm. Bis-CME dissolves in aqueous The former ion is the I3C molecular ion of benzene. The solution very slowly. To facilitate a fast dispersion of bislatter ion, C2H40C1+, which was monitored, is the most ChlE in aqui-.ous solution, a 1% bis-CME-acetone solution intense ion in the mass spectrum of bis-CME and is genwas used. &re, acetone acted as a carrier for dispersion. erated from the molecular ion, ClCM2QCH2C1.t The hydrolyhi apparatus is shown In Figure 1. The bisClCHzQ+=CHz C .C1. The intensity of the C2H40CI-k peak is directly proportional to the concentration of bisCME-acet,one solution (20 pl) was injected into a sealed

-

The Joilr-rai o f Pi:vs,cai Chernstry. Voi. 78. No 7 7 . 7974