V O L U M E 2 7 , N O . 8, A U G U S T 1 9 5 5
1215
Under certain conditions, use of two platinum electrodes is possible where one acts essentially as a reference electrode, thus facilitating titrations where the usual reference electrodes are not desirable-e.g., in nonaqueous media. The disadvantages are similar to those of amperometric titrations. Titrations in the negative potential region must be performed in the absence of ox?-gen, TTith a wait between additions of titrant during the measurements, and the end point requires a graphing of the results. Precipitation t'itrations could not be performed easily, for erratic results are obtained unless the electrode surface is renen-ed before each run. The method is more sensitive than amperometry to minor convec.tion and vibr:tt,ion of the solution. Titrations that require
heating would necessitate more iigorous temperature control to eliminate convection current. ACKNOWLEDG3IE\;'l
This research was supported hy the L7nited States Air Force through the Office of Srientific Research of the Air Research and Development Command. LITERATURE CITED
(1) Delahay, P., and 3Iamantoi-, G., As \r.. CAEM.,27, 475 (1955). (2) Hall, ,J. L., Gibson, J. -I., TTilkiiiioli, P. I i . , and Phillips, H. O., I h i d . , 26, 1484 (1954). (3) Reilley, C . S . , Everett, G . TV., and John;, R. H., Ibid.,27, 483 (1955). (4) Sand, II. J. S.,Phil. Mag.>1 , 45 (1901). R E C E I V Efor D r e v i e x Fehroary 23. 1R2,i. .\ccepteii .lpril 2 1 , 193:.
Analysis of Reaction Products by Isotope Dilution Procedure Examination of Acrylic Acid-Ethyl Alcohol Reaction Mixtures for Diethyl Ether Formation J E R O M E G. BURTLE College of St. Thomas, St. Paul, M i n n .
JOHN
P. RYAN
Central Research Department, Minnesota M i n i n g
& Manufacturing Co., St. Paul, M i n n .
Small amounts of peroxide-forming impurities affect the polymerization characteristics of many monomers. This paper describes an investigation of the acrylic acid-ethyl alcohol reaction product to determine the amount of by-product ethsl ether present. -4s no chemical tests are available for the quantitative detection of sniall amounts of ether in the presence of large amounts of ester, the isotope dilution procedure using Carbon-14-labeled diethyl ether as the tracer was used. A number of different catalysts were used and small but definite amounts of diethyl ether were found in the reaction mixtures. The method is highly sensitive as applied to this system.
T
1-1 1.; n:itui.e of by-products eiicountered during the esterification of alpha-beta unsaturated carboxylic acids is a matter of considerable concern when these esters are to be used in subsequent polpinerizat~ions. This is especially true of by-product' ethers which. because of peroxide formation, are capable of iiitroducing a i i riiicontrolled factor into peroxide-catalyzed polymerization rc~:ictionsinvolving these esters ( I , ~7~6. 8,10-12, 1.5. 18. 20). .I stlawh of the literature hrought to light no published report of ether iolmation (-luring esterification procedures. Severtheless, estrrific:ition reactions involving the use of alcohols in the preserirc~of :icid wtalysts (l+ 1 8 ) do provide ideal conditions for ether forni:ttion (I.?, 14, 21). Therefore, it seemed advisable to investigatc~t,hc extent of ether formation in a typical esterification involving n polynierizable acid. Many chemical tests for ethers are known ( 2 6 , Zi)! but they fail to detect small amounts of these compounds in t,he presence of large aniouiits of esters. Physical methods of analysis (such as infrared j :is npplied in this laboratory failed to indicate the presence of ether in the reaction products studied in these experiments. This implies that such impurities, if produced a t all, are present in very small quant.ities. In view of t,hese facts, it was decided to investigate the reaction betreen acrylic acid and ethyl alcohol and to use the isotope dilution procedure with diethyl-l-C-14*her ( 2 ) as the tracer t o analyze the reaction product.
1lETIIOI)
The esterification reaction between wrylie acid and ethyl alcohol may be driven to completion in a number of tvays. The technique w e d in this study was to carry out the reaction in the presence of a large amount of benzene ( 2 2 ) . T h e benzene acts as a rrater-entraining fluid and both water :md benzene are removed from t,he reaction zone by distillation. The water is separated in a phase separator and t.he benzene returned t o the reaction mixture. When nomore water is observed in the benzene distillate, the reaction is considered complete. The amount of ether in the reaction mixture is then determined by applying the isotope dilution technique ( 3 ) . Briefly, this technique consists of adding t o the reaction mixture a measured amount, of the ether to be determined containing a known amount of radioactivity. T h e ether is then isolated in a pure form and the activit8ymeasured. The dilution of activity by the nonradioactive ether in the mixture is a nieasure of the amount of t h a t compound present. The important factor in this technique is t,hat o d > - cnough material need be isolated bo prepare samples for radionrtivity measurements. The st,eps in the procedure vere: 1. -1measured amount of radioactive ether of known activity 11-as added to the reaction mixture.
2. Diethyl ether was carefully distilled from the mixture by means of an efficient fractionating column. 3. The activity of the ether thus isolated was determined. 4. The dnta obtained were substituted into the standard isotope dilution equation
b = a ( 2 - 1) where b relmsents the amount of inactive mat,erial present, a the amount of active material added, and 2 the ratio of the specific activity of the active ether to that of the ether sample isolated. Five esterification reactions v-ere carried out (Table I). T h e first t,wo runs were 1-mole sj-nt.hesesand in each the reaction mixture was predistilled t o a l l o r isolation of approximately 75 ml. of the lowest boiling portion of the mixture. T h e radioactive ether %-as added t o this distillate and the distillation procedure described in step 2 was applied. Thus, the first two experiments
ANALYTICAL CHEMISTRY
1216
Table I. By-product Ether Formed during Direct Efiterification of Acrylic Acid
1 2 3
1 1 2
tf
2 2
H290r H?SOn H2SOn p-Toluenesuifonic acid Sulfonir acid ion exchange resin Amberlite I R 106
5740 5.530 6780 6780
0,8331 1 3396 1.4191 1.4178
5360
5’20 ,1770 6230
1.079 1.174 1 089
7 .3 7.3 14.8 8 2
0.0683 0.1058 0.2464 0.1262
0.07 0.11 0.12 0.06
6780
1.4215
6510
1.042
1.0
0.0597
0.03
1.084
All radioactivity measurements were made in a windowless. gas-flow Geiger-MCiller rounter. Counting rates were dpterrriined for infinitely thirk layers of tiariiini rarbonate. Therr values. proportional to the specific activity of the samples. mere used directly in all c*alrtilations.
Known Values in Evaluation of Experimental Tectiniqiies
Table 11. Rrnzenr. 0. 430 4.50 450
Inactive Ether Added, G. 0,6995 0.3217 0,1333
Radioactive Ether .\dded,
Activity of Radioactive Ether, C./JI.
1 3944 1 3981
6780 6780 6780
G.
1 4164
i:iclrided a preliminary distillation involving a conderisate which was predominantly benzene nnd \vhich froze on the total contlrnser when solid carbon dioside i v : i ~used as the refrigerant. This necessitated the m e of ice in the condensing unit and it \xis ftslt t h a t under these conditions a portion of the et,her might have (.wiped through the condenser. For esperiments 3, 4. and 5 the procedure n-as changed by doubling the size of each batch and ntlding the radioactive ether directly t o the reaction mixture. RESL‘LTS AYD DISCUSSIOS
T h e results of the isotope dilution analyses (Table I ) show that the amount of ether formed in these preparations is very small. T o reduce the esperimental data to a common basis. the amount of ether present was calculated as percentage based on the theoretical yield of eater (100 grams for a 1-mole batch). On this basis, :dl results are roughly of the s m i e order of magnitude. Because of the high volatility of diethyl ether and the numerOUY uncontrolled factors involved iii these esperiments. it does not necessarily follow t h a t t,he differences in amounts of ether rioted in esperiments 3, 4, and 5 mirror solely a difference in activity of catalyst used. T h a t this is undoubtedly a factor, honever, is evidenced by the close agreement between all the rsperimrnts using sulfuric acid. T h e sensitivit,y of the method used is a function of the batch sizr. This is shown by comparing the results of experiment 3, a 2-mole batch. n i t h those of experiment 2. a 1-mole run. -1device for increasing the sensitivity of the method, therefore, lies in increasing the size of the batch as far as is practical, or increasing the specific activity of the tracer ether. T h e isotope dilut,ion procedure is a very effective anal),tical method for detection of ethers in the presence of esters. The results are well outside st,atistical variation, b u t it was felt that’ a series of “knowns” should be run t o determine whether the differences noted could be due t o esperimental manipulation. Accordingly, a volume of benzene approximately equivalent t o the finished volume of a 2-mole run vias placed in a distilling flask and a known amount of nonactive ether added. Radioactive ether in known quantity was then introduced, t h e mixture was carefully distilled, and a pure sample of ether was isolated. T h e :ictivity difference between the radioactive et,her added and the clthrr isolated was noted in the usual way and the amount of nonactive ether present x a s calculated by the standard isotope dilution equation. The data (Table 11) indicate t h a t observed differences in counting rates are very close t o the calculated variations. T h e esperimental differences presented in Table I are greater than the “technique” differences shown in Table 11, n.hich tends t o confirm the presence of small b u t definite amounts of ether in the esterification reaction mixtures.
Artivit,y of Isolated Ether D~R,, Found.
-
45ro 3300 6210
C./M.
nc
4420 5400 (1289
1 2 011 f1.82 -1.13
Ether Present fBased on Total Benzene), % Calcd. Found 0.16 0.16 0.07 0.08 0.03 0.02
Careful distillation studies on pure ether, benzene-ether mixtures. and reaction product-ether mixtures showed no azeotrope formation. Ether distillation from the reaction, therefore, is obviously similar to the distillation of ether from benzene. Active ether distilled from benzene as in the knowns above n-as collected and oxidized. and the results were compared with undistillrd :ictive ether as follows: Undistilled ether Ether distilled from b e n i e n ~
6680 c./rn. 6680 c./m.
As the same distilling column !vas used for all distillations. cuts of both active and diluted materials were directly comparable. These data wtahlish the puriti- of the ether collected. EXPERIhlENTA L
Preparation of Esters for Isotope Dilution. For the preparation of a 2-mole hatch, a 1-liter round-bottomed flask mas charged with 144 grams ( 2 moles) of 99% acrylic acid, 50 grams of 9570 ethyl alcohol, 1.0 gram of pyrogallol, and 0.4 gram of cuprous liromide as polymerization inhibitors, 200 ml. of benzene, and :in appropriate amount of catalyst. The latter amounted t o 1 gram when sulfuric acid was used, 4.0 grams when p-toluenesulfonic acid monohydrate vas employed. and 33 grams of resin containing F14.57~ moisture when -1mberlite IR-105 was the catalyst. The flask !vas then fastened t o a phase separator and cyclicly distilled with intermittent removal of a water-rich layer of distillate. T h e density of this layer was taken and its water content \vas calculated by means of the equation: Grams of water = (density of layer - 0.79) X grams of mater-rich layer 0.21 .is water evolution decreased. ethyl alcohol was added in increments of 20 grams until a total of 135 grams had been introduced into the reaction. Distillation was continued until 90 to 100% of the theoretical amount of water had been taken out of the system. T h e misture \vas then cooled in preparation for isotope dilution analvsis. Isotope Dilution and Isolation of Ether. T o the reaction misture, or t o the distillate from the reaction, an appropriate, weighed amount of radioartive ether was added in a sealed ampoule. This ampoule was broken under the surface of the liquid in the flask and the resulting sj-stem was carefully distilled using a concentric tube fractionating column ( 7 ) provided with a distilling head having a cold finger charged with solid carbon dioxide. T h e column was allowed to operate a t total reflux until it reached equilibrium (apnrosimately 2 hours) and the distillate coming over a t 33” to 34’ C . wis collected for activity evaluation. For the investigation of the knowns, the same procedure was used. escept that the distillation flask contained 450 grams of pure benzene in lieu of reaction misture. T o this was added a known amount of nonactive ether (also by means of sealed ampoules) before the introduction of the radioactive tracer. Radioactivity Measurements of Ether Samples. I n order t o measure the radioactivity of the ether samples, the compounds
V O L U M E 2 7 , N O . 8, A U G U S T 1 9 5 5 \vere oxidized using the \-an Slyke-Folch reagent (9,1 9 ) arid the carbon dioxide was collected in carbon dioxide-free sodium Ii>-droxide and precipitated as barium carbonate. T h e counting rates were determined for "infinitely" thick uniform layeis of barium carbonate. Since these values are proportional to the specific activity of the burned fractions. they were used directly i n all activity comparisons. All counting measurements were made in a windowless, gas-flow Geiger-Muller counter connected to a conventional scaler. T h e count,ing time for each sample was made long enough to give a "reliahle" error (9/10) of less than 17,. ACKNOW LEDGMENT
The authors are indebted to the Research Corp. for a grant to the College of St. Thomas for eiipport of t,his investigation. They also wish to thank the Central Research Department, Minnesota Mining (s: 3Ianiifacturing Co.. Lvhich kindly allowed the use of eciuipmeri t. LITERATURE CITED
Ramett, H. .J., C-. S. Patent 1,942,531 (Jan. 9, 1934). Ijurtle, J. G . , aiid Turek. IT. S . , J . A m . Chein. Soc., 76, 2498 (1954).
Calvin. AI., Heidelberger. C., Reid, J. C.. Tolbert, €3. AI.. and Yankwich, P. E., "Isotopic Carbon," pp. 278-82, Wiley, S e w York, 1949. Gilniaii. H.. and Blatt, A. H., ed.. "Organic Syntheses," Coll. Vol. 1 , 2nd ed., p. 261, Wiley, Sew York, 1941.
1217 Hermann, IT.O., and Baun, E.. Can. Patent 257,808 (l'eh. 2, 1926). (6) Hill, J.. Brit. Patent 423,790 (Feb. 4. 1933,. ( i )Saragon. E. d.,aud Lewis, C. J.. Isu. Esc;. CHE\I.,.Is.\r..E t ) , , (8,
18, 448 (1946). (8) Sobel
Francaise. French Patent 750,348 ( d u g . 3, 1938). (9) Pregl. F., and Grant, ,J., "Cuantitative Organic llivroanalysi,~," 4th English ed.. p. 6 2 , Ulakiston, Philadelphia. 1946.
(10) (llj (13) (1.3)
Ilohrn. 0.. CIICIIL. E I L OS. P W 31. S . 4 t 8 (1953,. I h i d . , p . 560. Scheideinandel. H., Ger. Pat,ent 615,995 (July 17. 1938) Schorigen, P., and llakarov-Zernliriati~kii. J.. H e r . , 65, 1293
(1932). (14) Schroeter, G., and Sondag, W., I b i d . . 41, 1924 (1908). (1.5) Scorah. L. \-.D., and Wilson. J., Brit. Patent 422,697
(.Jail.
14,
1935).
R . L., and Fuson. It. C.. "Identification of Organic Compounds." 3rd ed., pp. 103, 186 -8.Wiley, S e w York. 1949. ( t i ) Yiggia, 9.. "Quantitative Organic: ,Inaly-h via Functional Groups," pp. 27-30, 59-64, 106-7, Wiley. Sew York, 1949. (1s) Thielepape, E., Ber.. 66, 1454 (1933). (191 Van Slyke, D. D.. aiid Folch; J., J . K i d . C ' h e i n . , 136,509 (1~,340), ( 2 0 ) W e r n t a , J. H., J . z4in, Chcna. Soc., 57, 204 (t935). (21) Weygand, C., "Organic: Preparations." pp. 11% 9 . Interscirncte. S e w York, 1945. (16) Shriner,
( 2 2 ) Ibid., pp. 171-5.
RECEIVED for review
Deceniber 2 , 1954.
Accepted
May 3 195:.
Measurement of Atmospheric Pollution by Ultraviolet Photometry D. J. TROY Engineering Research Laboratory,
E. 1. du Pont de Nemours & Co., Wilmington, D e l ,
.4 sensitive portable ultraviolet photometric analyzer is an extremely convenient device for determining the concentration of toxic gases and vapors in the atmosphere, but instruments described in the literature have generally been limited to one anal>-ticalwave length, 254 m p , and data on the absorptivity of gases and vapors, on which anal>-zercalibrations could be based, have been scanty. -4 portable ultraviolet gas and vapor analyzer, based on Hanson's design, has been gradually improved during several >-ears'experience with a number of these instruments at several chemical manufacturing plants. The wave-length range of the analyzer has been extended by relatively simple techniques, thereby enhancing sensitivity and minimizing interference. -4 convenient gas-mixing apparatus for calibrating sensitive gas analpers under simulated operating conditions at concentrations from 2070 to 0.1 p.p.m. has been developed. The extended wave-length range and the use of the calibrating apparatus have materially increased the utility and precision of the portable ultraviolet analyzer in the anal>-sisof toxic gases and vapors i n the atmosphere.
I
S THE measurement of atmospheric pollution by toxic gases and vapors, ultraviolet photometry ha- a number of import a n t advantages over man:, alternative methods ( 2 ) . It 13 extremely sensitive: Considerably less than the maximuin alloir able concentration of many atmospheric contaminants can be detected. Moreover, the equipment is relatively simple, and its rapid response permits measurenients t o be made in less than 1 minute. Over the past 15 years, a number of sensitive ultraviolet photometers for atmospheric pollution analysis have been described, notably by Hanson ( 3 ) )N o t a and Dole (4),and Silverman (8). T h e great utility of such instruments has apparently not been widely appreciated, however, as commercial equipment of this type has been unavailable until very iecently.
Portable ultraviolet analyzers previously degcrihed have provided only one analytical wave length, usually 25.4 mw; convenient methods of calibrating them at low concentrations have been lacking: and published data on gas and vapor absorptivity, on which a calibration might be based, have been wanty. These limitations have made these analyzers difficult to apply to specific problems. DESCRIPTION OF INSTRUXIEXT
T h e t,ype of analyzer described is similar in principle to one previously described by Hanson ( 3 ) but has been modified extensively to make it upefiil for the detection of a greater number of substances and to permit easier operation and maintenance. T h e principal parts of the instrument are a n ultraviolet soiirce having 80% of its measurable radiation at 254 mp, a double-tirani optical system, Type 935 vacuum-phototube detectors, and a sensitive, stable electronic niicroaninieter. .i riniilar instrument was for a time manufactured by the Mine Safety hppliances Co., but was discontinued several years ago. The instrument is Engineering & Equipcurrently manufactured by 1Iariufactni ment Corp., Hatboro, Pa. The analyzer is shown in Figure 1. The sample inlet and outlet are conveniently accessible on the front. On the top are the meter which indicates the absorbance of the sample, the range selector switch which permits the selection of full scale absorbances between 0.0005 and 0.05, the zero adjust for setting the zero of the microammeter, the scale adjust for setting the mirroammeter sensitivity, and coarse and fine optical balance controls for zeroing the instrument on a sample of nonabsorbing ga3. Figure 2 is a schematic diagram of the optical syst,em and the photometric circuit. The right-hand or measuring phototube views the lamp through t h e sample cell, while the left-hand or reference phototube views the lamp directly, or through a wiro screen used t o achieve approximate balance of the phototube responses. Two screiv-operated metal rods, a thick one in front of the reference phototube, and R thin one in front of the measuring phototube can be advanced across the faces of t h e phototubes, t o serve as coarPe and fine optical balance controls, resprctively.