tassium iodide solution. The orange-red, liquid product of the decomposition was purified by distillation. One fraction, amounting to a yield of about 20%, had b.1,. 80" (17 mm.), n% 1.610 arid (ES4 1 .3.These propcrtics wlicii voinpared with those previously reported3 itieritify the conipound as benzenesulfeiiyl chloridc. Hydrolysis of Sulfur Trichlorides.--iVhcii added to '1 cold sodium bicarbonate solution ethvlsulfur trichloride formed sodium ethanesulfinate which reacted on boiling with benzyl chloride t o form benzyl ethyl sulfone. The latter melted a t 83" m d the mcltirig point was unchanged when the sainple >vas mixed with benzyl ethyl sulfone prcp:ire\l h y the oxidation of benzyl ethyl srilfitlr. Unsuccessful Attempts to Prepare Sulfur Trich1orides.--Benzyl disulfide ( 5 g.,) was chlorinateti :it --HIo, hut no preripitate formed. The reaction mixture was distillet1 under reduced pressure a n d there w:is obtained 2 g. of liquid with the properties: 11.11. 89" (37 nun.), n% 1.540, il?", 1 . 1 1, ivhich itlentify t.hc produci :IS hc~ri~yl rhloriric.. I)EPARTMENT OF CHEMISIRY ~,-XIVERSITY OF o R o s o , IIATNF
u\
h1.41~
The Structure of Isoamidone. I R F A ~ T OS4MUFT Y, 1 WE1 SON' A V D VFI MT'R B
YTI SOY
srsri R E C F I V r D rZPR1L
24, 1952
.47icii. Calctl. for C21FI.5SOI: S, 3.20. Found: S , 3.17, 3.10. Oxalate (from absolute ethanol), m.p. 161.5-.1Mo; rcported for Isoannidoiie I oxalate,? 158-160°. *,lmi. Calcd. i o r C20€f2jS0.C&04: N,3.65. Found: S ,:3.60, Picratc (froin ethanol) 1n.p. 1:3-1-136°; reported for Isoainidone I picrate,? 131-133". .-ltrci/. Calcti. for C.oH2?NiOh: S, 10.68. Found: S, 10.48, 10.32.
Acknowledgment.-The authors wish to thank Dr. Joseph R . Stevens, Technical Director of the 1. T . Baker Chemical Company, for permission to publish the results of some experiments carried out in the research laboratory of the J. T. Raker Chemical Company, Phillipsburg, New Jersey. ~ V M H. . CHAXDLER CHFMISTRV LABORATORY LEHICHUXIVERSITY BETHLEHEM, PESNA.
The Rate of the Reaction between Thallium(II1) and Iron(I1) in Perchloric Acid Solution B Y OTTO L. FORCHHEIMER~ A N D ROBERTP.
EPPLE
RECEIVEDJUSE 30, 1952
While carrying out some work on 4,4-diphenyl-6W'hile preparing a paper describing our work2 dimethylamino-3-hexanone we had occasion to prepare several derivatives. The melting points of on the kinetics of the Fe(I1)-TI(II1) reaction, the these compounds agreed very closely with those article by Johnson3 on the same subject appeared. reported2 for Isoamidone I, and the melting points We are therefore limiting our report to results of the hydrochloride of this substance, when taken which amplify or supplement those reported by in the same melting point apparatus was the same him. Measurements made under the same condias that of the hydrochloride of 4,4-diphenyl-6- tions agree quantitatively. Experimental dimethylamino-3-hexanone. p\'o depression in The preparation of the thallous, thallic, ferrous and ferric melting points was observed on mixing. These data prove that Isoamidone I is 4,4- perchlorates has been previously described.' The peracid was G. F. Smith Chemical Company's 7070 diphenyl-6-dimethylamino-3-hexanone. =1 logical chloric "purified." Spot tests as recommended by Feig16 showed explanation for the finding of this conipound in the that this acid had a ferric ion content of less than 3 X 10-6F mother liquors from the synthesis of Amidone is and a chloride ion concentration of less than 3 X that the propylene chlorohydrin, which was used in S o chlorate could he detected. The ferrous perchlorate solution contained traces of chloride ion (for quantitathe preparation of the dimethylaminochloropro- stock tive information see reference 4), but not in sufficient pane, was contaminated with ethylene chlorohydrin. amounts to affect the results of this investigation. Sodium Experimental The 6-dimethylammo-4,4-diphenyl-3-hexanone mas prepared by the procedure3 of Dupre, et al , in which 4-bromo2,2-diphenylbutanenitrileis aminated with dimethylamine and then the ketone is prepared by the usual Grignard reaction. The ketone was usually isolated as the hydrobromide and then converted t o the desired derivative Preparation of Derivatives of 6-Dimethylamino-4,4-diphenyl-3-hexanone.-The hydrobromide was dissolved in warm water and the solution was made basic with sodium hydroxide The oil was extracted with ether and the ether was dried over magnesium sulfate and the solvent distilled. Concentrated hydrochloric acid and ethanol were added. The mixture was concentrated t o dryness and crystallized on scratching It was recrystallized from acetone and melted at 173.5-175°,4 not depressed on an admixture of Twamidone I hydrochloride of m p. 173 5-175' .Inn? Calcd fnr C20H~BNOCl: C1, 10 68 Found: C1, 10 92 Methiodide (from absolute ethanol), m.p. 199 ,5-200 5"; reported for Isoamidone I methiodide2 195-196'.
and lithium perchlorates were prepared by adding the reagent grade carbonate to perchloric acid. Basic magnesium carbonate was used in the same manner t o prepare a magnesium perchlorate solution. Manganese perchlorate solutions were prepared by dissolving the reagent grade metal in perchloric acid. These solutions all contained less than the above limits of iron and chloride ion. All other chemicals used were Baker and Adamson reagent grade. The reactants, in separate glass-stoppered bottles, were allowed t o come t o uniform temperature in a constant temperature bath. .The order of addition did not affect the results. All experiments were carried out a t 25 f 0.1". The time of starting a run was taken at the time when one half of the second reagent had been added t o the first. Samples for analysis were taken with a pipet and discharged into an acidic chromate solution which stopped the reaction by oxidizing iron(I1) and precipitating thallium( I). The acidic precipitate was metathesized into the stoichiometric thallous chromate by a previously described m e t h ~ d . ~ Although chloride ion has been mentioned as a catalyst for the dichromate oxidation of thallium( I),O tests showed _ _ ~ (1) University of Chicago, Chicago, Ill.
(2) 0. L. Forchheimer, Thesis, Brown University, 1951. (3) C . E . Johnson, Jr., THISJOURNAL. 74, 959 (1952). of the requirements for the degree of doctor of philosophy. (4) 0. I*, Forchheirner and R . P. Epple, Anal. Chem.. 23, 1415 (2) N. R . Easton, J H Gnrdner and J. R . Stevens, THISJOURNAL, (1951). 70, 70 (1918). (5) F . Feigl, "Qualitative Analysis h y Spot Tests," Nordemann (3) D . J . Dupre, J . Elk%,D. A IIernu. K . K, Speyer a n d R . hl. Evans, Publishing Co., Inc , Pjew York, N. Y . , 193 .I Chrm. . S o c . , 500 (19491. A S I T ~ rd,i ~ ~ Tqbring ~ ~ ~ il,,rbov (4) 3f I l r i ~ v t t i i i i x l i l : < f t ( l L: l ~ 1 ~ r I b : a r t ..1 , o r . 661. .>2 ~ I ~ I 1 S 1li4 , 173 (1) Abstracted in part from the thesis submitted in partial fulfillment
1 ?.:
NOTES
Nov. 20, 1952
577.7
that this analytical method gave good results in the presence of the chloride ion concentrations used in this study.
Results The specific rate, .k, is obtained from the integrated rate expression k
b(a - 2 ~ ) = -___ 2.3 log ~t ( a - 2b) n(b - x)
where t designates the time of reaction, a the initial concentration of the iron(II), b the initial concentration of thallium(II1) and x the experimentally determined amount of thallium(1) formed. k has been calculated and listed for all experiments except tliose in the presence of added chloride ion where the above equation does not apply. The specific rate was observed to decrease slightly after the reaction was 50-707, complete. In the absence of oxygen this decrease was not observed. Since the effect of oxygen was small, no effort was made to exclude it. The reported rate constants are the averaged rate constants of the first 50--SO~o of the reaction. The results of the analytical work in reference 5 indicate that x is subject to a probable error of 0.27 mg. and to a maximum error of 0.7 mg. Differentiating k partially with respect to x we obtain a probable error for k of 0.015 liter-F.M;.-l-minute-l. This amounts to a probable error of 2-3% of the k values obtained in this study. I n Table I are listed some representative results obtained a t an acid concentration of 2.0 F. Run 11 demonstrates that the products do not affect the rate. I n the various experiments k was shown to be constant over a four-fold concentration range in thallium(II1) and a three-fold variation in the concentration of iron(I1). This variation is slightly greater than that studied by Johnsonj3 and the experiments as a whole were conducted a t higher concentrations than he used, extending the concentration ranges over which this reaction has been found to be second order. The nature of the analytical method limited the concentration range investigated here. The magnitude of k found here a t an ionic strength of 2 cannot be directly compared with the results of Johnson a t an ionic strength of 3. SOMERESULTSAT 2 0
[Cl-]
RUN 18
[Cl-]
:
El-]
= 0.00076F
0
0 RUN 19 0
0 Fig. 1.-The
RUN 20 [W]
0.0002SF
:
0.0027SF 0.01F
B 8 10 Time in hours. effect of chloride ion on the rate a = 0.0201, b = 0.0075, HClO, = 1.5 F. 2
4
for which Harbottle and Dodson found that thallium(II1) moved toward the cathode carrying chloride with it, and observed no chloride ion moving toward the anode. This shows that under such conditions (the concentrations of thallium(II1) and chloride ion being essentially equivalent) the thallium(II1) is capable of complexing all the chloride present. Benoits reports a dissociation constant of for TlCl++. Silverman and Dodson report that chloride ion catalyzes the iron (11)-iron(II1) exchange r e a ~ t i o n . ~I t is very probable then that we are dealing with a reaction between iron(I1) and a monochlorothallate or perhaps a dichlorothallate ion. The rates obtained when chloride less than equivalent to the thallium(I1l) concentration was present (runs 18 and 19) approximate the second order rate expected if the thallium(II1) concentration is replaced by the quantity ([Tl(III)]-[Cl-1). However, the rate was somewhat faster than calculated using this “reduced concentration” and the chloro complexes must account for a fraction of the reaction. The effect of varying the concentration of perchloric acid is shown in Fig. 2 . The ionic strength was not kept constant in this series of runs and is essentially equivalent to the acid concentration. At acid concentrations between 0.75 and 1.5 F TABLE I the rate of the reaction decreases with increasing F PERCHLORIC ACID CONCENTRATION acid concentration, illustrating the important part k, which hydrolyzed species play in the reaction, as F.W. -
a,
Run
F
b
[Fe(ClOl)a],
[TICIOI],
liter-’ minute-’
6 7 8 11
0.0201 .00715 ,0262 ,00715
0.0083 ,00342 .0128 ,00342
... .. .... . . ....
..... .....
0.00767
0.00301
0.69 .67 .72 .67
6
F
F
.....
Figure 1 illustrates the effect of chloride ion as an inhibitor. This effect is similar to the action of chloride ion on the rate of the electron exchange between thallium(1) and thallium(II1) studied by Harbottle and Dodson.’ They found that chloride ion in small concentrations decreased the rate of the exchange enormously. The conditions prevailing in run 20 are similar to those in a solution (7) 0. Harhnttle end H. W~ Dodnon, THISJOURNAL, 1 9 , 2442 (1961).
B RUN 12
8 TAKEN
AT IO% COMPLETION SO% COMPLETION
0 TAKEN AT
0.0 1.0 2.0 3.0 4.0 Perchloric acid concentration in F.W. per liter. Fig. 2.-Variation of rate with acid concentration ( 8 ) R. Benoit, SOC.Chim. France, 5-6, 518 (1949). (9) I . 8ilverman and R. W,Dodaon, J , Phya, Chrm., 56, 846 (1RGZ).
577'2
NOTITS
quantitatively shown by J ~ h n s o n . ~Between 2.0 and 4.0 F, the rate of the reaction' increases with increasing acid concentration, due, perhaps, to a kinetic salt effect, or perhaps the perchlorate ion concentration itself has a positive effect on the reaction due to ion pair formation.1° Attempts were made to study the reaction a t lower acid concentrations, but the hyclrolysis of the thalliurn(II1) solutions interfered.I1 A direct comparison is afforded with Johnson's results by the run in 3.0 F HC104. The experimentally obtained value of k is 0.77. Using Johnson's f ~ r m u l aapplicable ,~ at :HI ionic strenqth of 3, we get k = 0.74. The results of some runs in the presence of certain cations are listed in Table 11. The experiment in the presence of a small quantity of manganese was performed in order to see whether the presence of this ion could provide an alternative reaction path to the system. Shaffer'? has shown mmganese(I1) to be a good catalyst for some reactions between two-electron reagents and one-electron reagents, as it can form oxidation states of 2 , 3 and 4 in solution. No catalytic effect was found in this system. The other salts increased the rate, as was expected, in view of the results obtained a t high acid conceiitrations. These results are, however, insufficient to draw any conclusions about the effect of ionic strength, perchlorate ion concentration or any specific ionic effect. TABLE 11 EFFECTO F S O M E C A T I O N S O V T I E RATE 0.0201, [HClOd] = I 5 F b = 0 0113, u
Run
a
k 0 67
Added salt a n d its concentration
13 Kone 21 0 30 FLiClOl 22" .011 F?rln(C104)1 23 .20 F big( ClO,), 24 .SO F Sac104 Acid concentration in this run 1.79
81 6s
. SR .87
r.
One can account for the kinetics of the reaction by a simple two-step mechanism Tl(II1) Tl(I1)
+ Fe(I1) --+ Tl(I1) f Fe(II1) + Fe(I1) -+ Tl(I) + Fe(II1)
or alternatively
+
+
+
H20 Tl(II1) Fe(I1) ---+FeO++ Tl(1) -I-2I-I FeO++ Fe(I1) 2 H + --+ 2Fe(III) HzO
+
+
+
'
where the first step in each case is the slow, ratedetermining reaction. Since the specific rate is independent of the concentrations of the reactants (over the range studied) and also independent of the concentration of the products, no measurable equilibrium can be involved in the first step. On the basis of the kinetics no choice can be made between iron(1V) and thallium(I1) as an intermediate. However, the path through thallium(I1) (10) Cf A R Olqen and T R Simonsen, J . Chcm P h y s , 17, llG7 (1949). (11) Thallic hydroxide precipitates only slowly from solutions when the solubility product is not greatly exceeded. The solubility product of TI(0H)a was found t o be 8 8 X 10-4' b y letting a small excess of TI(0H)a precipitate from B 0 5 F perchloric acid solution This value is in good agreement with the value 1.5 X 10-44 ralriilated b y Sherrill and Haas (THISJOURNAL, 68, 952 (1936)) (12) P A. Shnffer, i b t d , 66, 2169 (1993)
Vol. 74
is preferred because an intermediate which can act as a reducing agent is necessary to explain the intluccd oxidation by oxygen.
.
METCALF CHEMICAL LABORATORIES BROWNUNIVERSITY PROVIDENCE, I
