ACKNOWLEDGMENT
atmosphere total pressure in air or nitrogen, PAN does not depart seriously from the Beer-Lambert relationship. A portion of the 10-cm. cell contents was removed, the remainder pressurized to one atmosphere, and the spectrum rerun to reduce the absorbance of the carbonyl band. Absorbance a t each of the six major bands h,&ddecreased by a factor of five. The absorptivities of PPN and PBN measured in the 10-cm. cell were calculated assuming the value at 12.58 microns to be the same as that found at 120-meter path. Because of its lower vapor pressure the P13N spectrum was weaker than the others and therefore somewhat less accuraiie.
John Locke of Scott Research Laboratories and Eugene Cardiff of the University of California at Riverside assisted with the preparation, purification, and measurement of many of these sa ples.
8
LITERATURE CITED
(1) Darley, E. F., Kettner, K. A., SteDhens, E. R., ANAL.CHEM.35, 589-91 (1963 j. (2) Scott, W. E., Stephens, E. R., Hanst, P. L., Doerr, R. C., Proc. A P l 37 111, 171-83 (1957). ( 3 ) Stephens, E. R., “Symposium on
Chemical Reactions in the Lower and Upper Atmosphere;” Interscience, New
York. 1961. (4). Stephens, E. R., Darley, E. F.,
Taylor, 0. C., Scott, W. E., Intern. J . Air Water Pollution 4 (1/2),79-100 (1961). (5) Stephens, E. R., Darley, E. F., Taylor, 0. C., Scott, W. E., Proc. A P l 40 111, 325-38 (1960). (6) Stephens, E. R., Hanst, P. L., Doerr, R. C., Scott, W. E., Znd. Eng. Chern. 48, 1498 (1956).
EDGAR R. STEPHENS Scott Research Laboratories, Inc. P. 0. Box 2416 San Bernardino, Calif. Air Pollution Research Center University of California Riverside, Calif.’ 1 Address to which requests for reprints should be mailed. This work supported in part by the Air and Water Conservation Committee of the American Petroleum Institute.
Use of Glass Reference Electrode in Potentiometric Titration of Halides with Silver Nitrate in LiN03-KN03 Eutectic Melt SIR: The use of glass reference electrode in fused salts has been demonstrated ( 2 ) . It has also been shown that the silver halides are sparsely soluble in fused alkali metal nitrates (1, 4). Therefore, during the potentiometric titrations of halides with silver ion, sharp end points are observed on account of precipitation reactions. The purpose of this investigation was to show the use of glass reference electrode in the potentiometric determination of a mixture of halides in a molten salt solution such as the Li-K nitrate eutectic.
tion a s well as reagents were described previously (4). Procedure. Essentially the same procedure was used as previously reFor the potentiometric ported (4). determination of halides, the procedure consisted of dissolving appropriate single or mixed halides in the melt. The titration with silver nitrate was then carried out. The e.m.f. of the cell was measured after each addition of AgN03. I n an entirely analogous manner the silver ion added first as silver nitrate was titrated with various halides. All measurements reported in this study were carried out at 165.0’ C.
EXPERIMENTAL
RESULTS A N D DISCUSSION
Apparatus and Reagents. The electrolytic cell consisted of a 250-ml. borosilicate glass beaker. Other details of the cell and electrode prepara-
A typical titration curve for a halide mixture in molten nitrate solution is shown in Figure 1. Sharp potential changes a t the equivalence point were
Table 1. Typical Titration Results of Single Halides in Li-K Nitrate Melt
Taken, Halide CI Br I
Found, mg.
me. 137.8 97.2 117.9
136.4 96.6 118.5
observed in all cases. The order of the observed change is I->Br->Cl-. Examples of some experimental results obtained by the use of the present method for melts containing single and mixed halides are given in Table I and Table 11, respectively. The results presented in Tables I and I1 are averages of at least three separate titrations. In cases of sinele halides the relative error as calculated from e.m.f. data is 0.8%. For mixed halides in the melt the relative error is 1%. In carrying out the titration in the reverse manner, Ag(1) was determined in a series of nine runs (three for each halide) a t various initial silver ion concentrations. The relative error is less than 1% over a concentration range of 50 mg. to 1.9 grams per 1000 grams of melt. This is not surprising, however, since the silver electrode has been shown to be reversible in these melts (1, 4). As is well known in aqueous solution (3).if an anion forms a snarselv soluble compound with the metal, such as the silver halides in the present case, the metal can serve as an anion indicator electrode also. That this is so may be
-
,500
b Figure 1. Titration of a halide mixture with silver nitrate in Li-K nitrate melt First inflection indicotes equivalence point for I-, 2nd Br; and 3rd CI-
400
->
300
5
200
LL
I
100
W
0
-100 -200[
0
I
I
100
200
I 300
I 400
Weight of AgNO, added. rng
, 500
I
VOL. 36, NO. 4, APRIL 1 9 6 4
929
seen from the familiar solubility product relationship : KBp= [Ag+][X-] or [Ag+] = Ka, [X-I
(1)
Table II. Typical Titration Results of Mixed Halides in Li-K Nitrate Melt
Halides c1 Br
c1
I Br I
c1 Br I
Taken,
Found,
mg. 52.2 63.9 39.4 12.2 110.5 38.2 25.9 109.3 57.7
mg. 51.7 63.3 40.1 11.6 110.5 37.2 25.2 108.5 56.4
The modified Nernst equation, then, may be written as follows:
where E” is the sum of the appropriate constants. Plots of e.m.f. os. log[X-] give straight lines having slopes predicted by Equation 2. In a typical example, the experimental slope is 86.7 (for the chloride) as compared with the calculated slope 86.9. This is true, of course, only in cases where the melts have been saturated with silver halides. Hence, the present study has demonstrated that the silver electrode functions like a specific halide electrode in the melt under these conditions, and it may therefore be used for such purposes in certain fused salt solutions.
ACKNOWLEDGMENT
The author thanks Donald 0. Rudin, Department of Basic Research, Eastern Pennsylvania Psychiatric Institute, Philadelphia, for his kind permission to use some of the facilities of the institute. LITERATURE CITED
(1) Flengas, S.N., Rideal, E., Proc. Roy. SOC.(London)A233.443 11956). (2) Harrington, G. ‘W., Tien,’ H. T., J. Phiis. Chem. 66, 173 (1962). (3) Kolthoff, 1. M., Laitinen, H. A., “DH and Electro Titrations,” pp. 114L16, Wiley, New York (1941.). (4) Tien, H. T., Harnngton, G. W., Inorg. Chem. 2, 369 (1963).
Department of Chemistry Northeastern University Boston 15, Mass.
H. Ti Tien
Determination of Trimethylolpropane in Polyesters and Polyurethane Foams SIR: Very little is found in the literature on the identification and determination of polyols in synthetic polyesters. Similarly, there is a void in the literature on the identification and determination of components of polyurethane foams. In the analysis of polyesters, the conventional approach has been the saponification of the polyester followed by recovery, identification and estimation of the products of the saponification. In 1954, Kappelmaier, Mostert, and Boon (3) introduced a new approach to polyester analysis. The alkyd resin was reacted with excess amine to free the poly01 and convert the acid component to the corresponding amide. More recently, Esposito and Swann (2) qualitatively determined polyols by reacting esters prepared from phthalic anhydride with butylamine to free the poly01 which was then acetylated to convert the poly01 to the polyol acetate. The acetate was then identified by a procedure utilizing gas chromatography. A semiquantitative determination of polyols in polyesters
Table 1.
Trimethylolpropane Results on Polyesters
Sample A
B
C D
Trimethylolpropane, % 1.22,1.13 2.56,2.41 2.98,2.56 2.39,2.63 1.41,l.G
Lab prepareda Known trimethylolpropane content of
1.6270.
930
ANALYTICAL CHEMISTRY
by methanolysis is presented by Percival
(4).
The following work was initiated in an attempt to provide a quantitative procedure for the determination of polyols in polyesters particularly the small amounts of trimethylolpropane found in polyester formulations used in the production of flexible polyurethane foams and for the determination of trimethylolpropane in the polyurethane foams made from these polyesters. EXPERIMENTAL
Gas Chromatography Unit. The unit used to obtain the chromatograms was a ChromAlyzer 100 (Dynatronics Instrument Corp., Chicago, Ill.) equipped with a Sargent SR-30 recorder. The following parameters were used: Column size, 14-foot copper tubing .(3/16-inch 0.d.). Column packing, 20% ’Ucon 50HB2000 on 30- to 60-mesh Celite 545. Column, injection, and detector temperature 220’ C. isotherm. Helium gas flow, 130 ml. per minute for polyester analysis; 33 ml. per minute for polyurethane analysis. Detector cell current, 280 Fa. Attenuator setting, super sensitive. Reagents. All chemicals used were reagent grade. Procedure. An accurately weighed 5-gram sample of resin or foam was refluxed with 10 ml. of phenethylamine (b.p. 198’ to 200’ C.) for 3 hours. For the reaction of the foam it was first necessary to bring the amine t o reflux temperature before adding the foam which had been shredded. This brought about instant physical breakdown of the foam structure. If this is not done the foam absorbs the amine within its cell structure and, upon
physical breakdown of the foam, the amine is released onto the bare flask surface and is badly burned. After reflux, the reaction mass was cooled so that 25 ml. of acetic anhydride and 2 drops of water could be added. The mix was refluxed for another 90 minutes. The mix was cooled and 35 ml. of water were added. The mix was refluxed for another 5 minutes and, then, was cooled. The heterogeneous mix was then quantitatively transferred to a separatory funnel. It was estracted once with 20 ml. of chloroform and then twice with 10 ml. of chloroform. The collected chloroform layers were then washed 5 times with 50-ml. portions of water. The chloroform solution was then diluted to 50.00 ml. in a volumetric flask. A pinch of anhydrous sodium sulfate was added to dry the chloroform solution. A 30.0-pI. injection was made into the gas chromatography unit. The area under the elution curve for the trimethylolpropane triacetate was determined by means of a mechanical integrator in the work on polyesters. For the work on the foams, a greater degree of precision was desired because of the lower trimethylolpropane content. In this case, the area was determined by cutting out and weighing the paper forwed by the elution curve and the base line. Because of tailing of a previously eluted compound, the base line was not horizontal and the lower precision of the mechanical integrator can be understood. To further increase the precision of the foam analysis the helium gas flow rate was reduced to 33 ml. per minute. This greatly increased the detector sensitivity. Dimbat, Porter, and Stross ( I ) state that the area is inversely proportional to the gas flow for a given detector assembly. The retention time for the trimethylolpropane triacetate was 14.9
