Detection of Surface-Active Agents Containing Polyoxyethy Iene or Polyoxy propy Iene Grou p By Pyrolysis with Phosphoric Acid MILTON J. ROSEN Department of Chemistry, Brooklyn College, Brooklyn,
Although a number of qualitative tests for the polyoxyethylene group have recently appeared in the literature, they suffer from a lack of specificity. The present test depends upon the thermal decomposition of the polyoxyethylene linkage in the presence of phosphoric acid to yield acetaldehyde, which produces a blue color with sodium nitroprusside and diethanolamine. The polyoxypropylene group, under the same conditions, yields propionaldehyde and its polymers, which produce orange colors. Positive results are obtained in the presence of the ester, alkylaryl, sulfide, sulfonate, sulfate, amino, amido, and phosphate groups. The only compounds which give positive results in the absence of the polyoxyethylene or polyoxypropylene groups are glycerides which, under the conditions of the test, decompose to acrolein, which also gives a blue color with sodium nitroprusside and diethanolamine.
A""
result of the prevalent use of the polyoxyethylene group as a hydrophilic linkage in many of the newer surface-active agents, a number of qualitative tests for this functional group have recently appeared in the literature (8-4, 9). All of these tests, however, are based upon either the precipitation of the oxonium salt of the polyoxyethylene compound by a large anion, such as 1 3 - > cobaltothiocyanate, or phosphomolybdate ion, or the appearance of turbidity upon heating an aqueous solution of the surfactant, due to the inverse temperature-solubility relationship of certain polyoxyethylene compounds. The former type of test suffers from the fact that it is essentially a modification of the generally used precipitation test for cationic surfactants by means of B large anion ( I , 6, 9) and therefore is given not only by polyoxyethylene compounds, but by all cationic surfactants, with or without the polyoxyethylene group in the molecule. On the other hand, the latter type of test is applicable only to water-soluble compounds and gives negative results with sulfonated polyoxyethylene compounds and the polyoxyethylene glycols, among others (9). In a continuation of the search for simple, definitive tests for the functional groups present in commonly used surface-active agents ( ~ 7 )it~ has been found that all types of compounds containing the polyoxyethylene group may be detected very simply by pyrolyzing them in 85% phosphoric acid and leading the volatile products into an aqueous solution of sodium nitroprusside containing a water-soluble secondary amine, such as diethanolamine. The polyoxyethylene group, under these conditions, decomposes to yield acetaldehyde fOCH2CH2f,
Hf 2 nCH3CHO + +OCH,CH,-f,
N. Y. hydrophobic character to the molecule. Here, the polyoxypropylene group decomposes under the conditions of the test to yield propionaldehyde (which can be isolated as the 2,4-dinitrophenylhydrazone from the water-soluble fraction of the pyrolysis products) and its polymers, which produce orange colors with sodium nitroprusside and diethanolamine. +OCHzCH-f,
1
CH3
H+
7n CH3CH2CH0+
+
( C H B C H ~ X O ) , fOCHzCH)z-(n+m)
I
CH, PROCEDURE
Place 200 mg. (or 4 drops) of anhydrous surfactant and 1 to 1.5 ml. of 85% phosphoric acid in a 5- or 6-inch test tube and agitate thoroughly for a few seconds. Insert a 0.5-inch plug of absorbent cotton into the mouth of the test tube (to prevent condensed water vapor from falling back into the hot reaction mixture with consequent violent spattering during the pyrolysis) and attach a glass delivery tube with a 60" angle bend by means of a one-hol! rubber stopper. Clamp the test tube a t an angle of about 30 from the horizontal, so that the main portion of the delivery tube is vertical. The end of the delivery tube should pass beneath the surface of the "detecting solution" contained in a 4-inch test tube placed 011 a white surface to facilitate observation of any color change. The detecting solution consists of 1 ml. of water to which have been added 2 drops of sodium nitroprusside solution (20 grams of sodium nitroprusside dihydrate dissolved in 50 ml. of water and diluted with 450 ml. of methanol) and 1 drop of diethanolamine. Heat the mixture of phosphoric acid and surfactant with a 2-inch flame until the mixture turns dark brown. If foaming is excessive, discontinue heating momentarily to allow foaming to subside somewhat, then heat the entire test tube surface strongly just above the level of the foam and then move the flame down to the upper level of the foam. Ss the foam disappears under the intense heating, move the flame down the tube, keeping it always a t the upper level of the foam until the liquid portion of the mixture is once again being heated by the flame. Continue the pyrolysis until either a blue color or an orange color (which may or may not change to dark brown) appears in the detecting solution (positive result) or for a maximum of 5 minutes if no blue or orange color appears (negative result). DISCUSSION O F RESULTS
Every type of surfactant tested which contains either the polyoxyethylene or the polyoxypropylene group, or both, gives a positive result with this test. This includes anionics (Triton X-200, hlipal CO-436), cationics (Ethomeens, Priminoxes, Hyamine 10-X), and nonionics (Igepals, Ethomids, Tweens, Pluronics, Victawet 12), containing such other functional groups as the ester linkage (Emulphor VN-430, Sterox CD), the alkylaryl group (Igepals, Triton X-loo), the sulfide linkage (Sterox SE and SK), the sulfonate group (Triton X-200), the sulfate group (Alipal CO-436), the amino group (Ethomeens, Priminoses), the amide linkage (Ethomids), and the phosphate group (Victawet 12). Surfactants containing only the polyoxyethylene group (Table I ) produce a blue color; those containing only the polyoxypropylene group (Table 11) produce an orange color; those containing both the polyoxyethylene and the polyoxypropylene groups (Table 11) produce an orange color which quickly turns dark brown.
- ,,
which produces a blue color with the sodium nitroprusside and the secondary amine. This latter reaction is a reversal of the Simon test (?'), in which secondary amines are detected by the blue color (of unknown structure) which they produce when treated with sodium nitroprusside and acetaldehyde. In addition to detecting the polyoxyethylene group, this test can also be used to detect the polyoxypropylene group, which recently has been used in a number of surfactants to confer
787
ANALYTICAL CHEMISTRY
788
Table I.
Reactions to Pyrolysis in 85% Phosphoric Acid of Surfactants with and without Polyoxyethylene Groups
Product
Source
Structure"
Color Produced
Result
0 Eniulphor VN-430
Antara
II
H(OCzH4)zOCR (oleic acid ester) 0
Royal blue
+
I1
Sterox C D
Rlonsanto
H(OCZHI)ZOCR(tall oil ester) 0
Royal blue
1 .
Diglycol oleate L Tergitol T h I N b
H(0CzHdzOCR (oleic acid ester) H(OCzH4)zOR (triinethylnonyl ether)
Royal blue Royal blue
I
Einulphor ON-870
Kessler Carbide & Carbon Antara
H ( O C Z H ~ ) ~ O(oleyl R ether)
Royal blue
4-
Igepal CO-530
Antara
Royal blue
I
Igepal CA-710
Antara
Royal blue
+
Triton X-100
Rohm & Haas
Royal blue
L
Red-purple then bluec Red-purple, then blue C Royal blue Roval blue
I1
+
Gterox SE
Monsanto
Sterox SK
Monsanto
Tween 40 Tween 80
Atlas Atlas
Triton x-200 b
Rohm d: Haas
Royal blue
+
Alipal CO-436 b
Antara
R o y d blne
+
Ethomeen C/15
Brmour
Royal blue
I
Polyoxyethylene sorbitan mono-palmitate Polvoxvethvlene sorbitan mono-oleate
Ethomeen 18/60
Armour
Royal blue
I
Priminox 32 Priminox 43
Rohm & Haas Rohm & Haas
Royal hlue Royal blue
4
Ethomid C/15
Armour
Royal hlue
I
Ethomid HT/60
Armour
Royal blue
+
Hyamine 10-X
Rohm & Haas
Royal hlue
Hyamine 1622
Rohm'& Haas
Royal blue
Victawet 12 Kino1 HA10 Ninol 201
Victor Ninol Ninol
H(OCzHdzOP(0R)z 1: 1 C0c.F.A.-diethanolamine condensate 1 :2 oleic acid-diethanolamine condensate 0
Royal blue Royal blue Royal blue
++ +
Glyceryl monostearate S
Glyco
HOCH~CHOHCH~O~R 0
Royal blue
+
NH4OSO?OCHzCHOHCHzOCR Triglyceride Triglyceride Sorbitan monopalmitate Sorbitan mono-oleate Sorbitan sesquioleate
Royal blue Royal blue Royal blue Royal blue Yellow Yellow Yellow
ROOzOXa
Pale pink
+
0
11
Arctic Syntex M Castor oil Cottonseed oil Phosphated castor oil Span 40 Span 80 Arlacel C
Victor' Atlas Atlas Atlas
Santomerse D
Monvanto
Colgate
...
II
.....
' '
V O L U M E 2 7 , NO. S, M A Y 1 9 5 5 Table I.
789
Reactions to Pyrolysis in 85% Phosphoric .4cid of Surfactants with and without Polyoxyethylene Group (Continued)
Product
Source
Color Produced
Structurea
Result
Alkanol WXKb
Du Pont
RoSOzONa
Beige
-
Ultrawet S K b
Atlantic
R(JsOzoxa
Cloudy white
-
Alkanol B
Du Pont
R o 3 S O z O N a
Beige
-
ArePkap 100
Monsanto
I?(J-()OH)SOzONa
Beige
-
Duponol LIE
Du Pont
ROSOzOKa
Beige
-
Beige
-
Light amber
-
Yellow
-
0
0
I!
I1
rlmer. Cyan.
Aerosol O T
b
ROCCHzCH(S0zONa)COR
e I
Igepon T-73
Antara
Alkaterge-C
Coml. Solvents
Glucaterse-2S b Aninionyx A0 b
Coml. Solvents Onyx Oil
RCK(CHa)CHz(CHOH)rCHzOH CizHzeN(CH3)20
Pale yellow Pale yellow
-
Hyarnine 2389 b
Rohm & Haas
[ROCH~N(CH~)Z]+C~-
Pale yellow
-
RCN(CH~)C~H~SOAON~ N-CH(CHg)CzHpOH RC’ ‘0-CHz
I
0
I1
/ a
b c
R = alkyl or alkenyl group. Dried s t 115’ C. Initial red-purple color may be due to volatile sulfides produced in pyrolysis.
Group to Table 11. Reactions of Surfactants Containing the Oxypropylene or Pol~-oxypropylene Pyrolysis in 85% Phosphoric Acid Product G-9 17 G-923 G-2800 Onyxol 368 Pluronic FGS Pluronic LG2 Tergitol X C
Source Atlas Atlas Atlas Onyx Oil Wyandotte Wyandotte Carbide & Carbon
Structure Propylene glycol monolaurate Propylene glycol mono-oleate Polyoxypropylene mannitol dioleate Lauric isopropanolamide Propylene oxide-ethylene oxide polymer Propylene oxide-ethylene oxide polymer Alkyl ether of propylene oxide-ethylene oxide polymer
The absence of a blue color in the case of surfactants containing both the polyoxyethylene and the polyoxypropylene groups is presumably due to prior decomposition of the polyoxypropylene group, producing the observed initial orange color. Upon subsequent decomposition of the polyoxyethylene group, instead of a blue color, dark brown is observed as a result of the combination of the initial orange color with the blue from the decomposing polyoxyethylene group. The test is not limited to surface-active agents containing these functional groups, but is of application to derivatives of ethylene oxide or propylene oxide in general with tlie exception of ethylene glycol and propylene glycol. The results obtained with these derivatives are given in Table 111. The only type of compound tested which gives a blue color without containing the polyoxyethylene group is glycerides. This is due to the fact that glycerides, upon pyrolysis in acid medium, yield acrolein, which, when treated with sodium nitroprusside and secondary amines, produces a blue color similar to that produced by acetaldehyde (8). -411 other types of commonly used surfactants give negative results in the
Color Produced Orange Orange Orange Orange Orange, then brown Orange, then brown Orange, then brown
Result
++ ++ ++ +
Table 111. Reactions of Derivatives of Ethj-lene Oxide and Propylene Oxide to Pyrolysis in 85% Phosphoric Acid Product Ethylene glycol Diethylene glycol Polyethylene glycol 4nn C a i i i w a x 4000 Cellosolve Tetraethylene glycol dimethyl ether Propylene glycol Dipropylene glycol Polyglycol P-400 Polyglycol P-750 Dowanol 50-B
Source Carbide & Carbon Carbide & Carbon Carbide & Carbon
Structure H(0CHzCHz)zOH (x = 1) H(0CHzCHz)zOH (x = 2 ) H(0CHzCHz)zOH (x N 9)
Color Produced Purple Blue Blue
Carbide & Carbon Carbide & Carbon Ansul
H(OCHzCHd,OH (x N 90) HOCHzCHzOCHa CH~(OCHICH~)~OCI-I~
Blue Blue Blue
Carbide & Carhon
H(0CHCHz)zOH (z = 1)
Pale yellow
Dow
&HI H(0CHCHz)rOH (z = 2 )
Orange
Dow
&Ha H(OCHCHz).OH
7)
Orange
Dow
CHs H(OCHCHz),OH (z
13)
Orange
Dow
CHa CHI(OCHCHZ)ZOH
Orange
Carbide Carbide Carbide Carbide ~ ~
&Ha HOCHzCHzNHz (H0CHzCHz)zNH (H0CHzCHz)rN HOCHCH2NH1
Blue Blue Blue Orange
I
(t N
I
N
I
Monoethanolamine Dietha nolamine
~ ~
~
$
$ i
~
~ i
Triisopropanolamine
8: Carbon & Carbon & Carbon &$ Carbon ~ e
Carbide ~ & Carbon ~
AHa ( H~O C H C H ~ ) ~ N H
Carbide & Carbon
&HI (H0CHCHz)sN
L Ht
~
~ Orange Orange
~
~
790 absence of (Table I).
ANALYTICAL CHEMISTRY polyoxyethylene
or
polyoxypropylene
groups
LITERATURE CITED
(1) Flanagan, T. L., Jr., Drennen, T. J., and Goetchius, G. R., Soap Sunit. Chemicals, 24, No. 4, 163 (1948). (2) Gnamm, H., “Die Losungs und Weichmachungsmittel,” p. 330, Stuttgart, 1941. (3) Haakh, H., v. Candie, D., and Mobus, m7., MelLiand TestiZber., 32,699 (1951).
(4) Newburger,
(5) (6) (7) (8) (9)
S. H., J . Assoc. Ofic. A v . Chemists, 34, 109 (1951). Rosen, M. J., ANAL.CHEM.,27, 111 (1955). . Proc., 1950. Shiraeff, D. A., Chem. Specialties M ~ T sAssoc. Simon, L., Compt. rend., 125, 536 (1897). Ibid.vP.1105. Wurzschmitt, B., 2. anal. Chem., 130, 105 (1950).
RECEIVEDfor review October 2 5 , 1954. Accepted December 20, 1954. Presented before the Division of Polymer Chemistry a t the 126th Meeting of the AMERICAN C H E M I C A L SOCIETY, New York.
Identification of Petroleum Refinery Wastes in Surface Waters A. A. ROSEN and F. M. MIDDLETON Robert A. Taft Sanitary Engineering Center, U. S.
Public Health Service, Cincinnati,
Because the most significant pollution effect of petroleum refinery wastes is the production of tastes and odors in receiving waters, there is a need for nonsubjective methods of identification with sensitivity comparable to odor judgments. The aliphatic and aromatic hydrocarbon fractions separated from the wastes of five refineries showed corresponding infrared spectral patterns sufficiently characteristic to suggest their use in identification. The organic materials in three samples of surface waters containing varying amounts of refinery wastes were concentrated with active carbon. Hydrocarbon fractions, separated by chemical and chromatographic procedures, were identified by the resemblance of their infrared spectra to the spectral patterns of refinery waste hydrocarbons. This method provides chemical evidence, independent of odor evaluations, of the presence of low concentrations of petroleum refinery wastes in surface waters.
B
ECAUSE of their large volumes and high odor intensities (6, 7 , 17), the most significant pollution effect of petroleum refinery process water effluents is the production of tastes and odors in receiving waters (1, 6, 6, 15). Indicative of the nature of this problem is the active current research on pollution abatement and the development of necessary analytical procedures (8). The major organic component of petroleum refinery wastes is the neutral group, consisting of hydrocarbons and closely related compounds which do not form salts with acids or bases. il large proportion of the odorous organic components is contained in the neutral group ( 1 7 ) . Such compounds are resistant to biological and chemical action (1, 5 ) ; consequently, the odor effects persist beyond the immediate vicinity of the waste discharge. Even in instances where the petroleum odor in water is recognizable, i t is seldom possible by nonsubjective methods to establish the presence of petroleum products. Gross pollution has been detected by the presence of an oil film on the incoming water basins ( 1 ) and by paper chromatography of oil recovered from harbor slicks (9,18). Melpolder, Warfield, and Headington ( I S ) have recently published a very sensitive procedure for the identification and determination of volatile hydrocarbons in Tvater. The method is limited to hydrocarbons boiling up to 400” F. and requires the use of a mass spectrometer. Descriptions have been published ( 2 , 1 7 ) of the use in this laboratory of active carbon filters for the recovery and characterization of organic materials present in very low concentrations in surface waters. An application of these procedures ( 1 4 ) has been partially successful in demonstrating the presence of refinery wastes in water supplies drawn from a lake in the vicinity of the dis-
Ohio
charge points of a number of refineries. In this method, organic substances were recovered from the water by adsorption on activated carbon and subsequent elution, then were compared with a sample of refinery waste materials on the basis of elemental analysis, physical and chemical properties, and infrared spectra. The presence of oxygenated compounds, arising both from oxidation of the petroleum waste and from the presence of other types of pollutants, obscured the hydrocarbon properties serving as a basis for identification, particularly the infrared spectral characteristics. Petroleum hydrocarbons, consequently, were not clearly recognizable. To minimize this interference, the method described in this paper utilizes adsorption chromatography on silica gel to remove oxygenated substances. In addition, the hydrocarbons are separated into an aliphatic and an aromatic fraction. Infrared spectra of corresponding hydrocarbon fractions of five refinery wastes showed a remarkable degree of similarity, suitable for use in identification. Similar infrared spectra were shown by the corresponding chromatographic fractions of organic compounds recovered from surface waters polluted n-ith refinery wastes. DEVELOPMENT OF METHOD
Chromatography. Adsorption chromatography is regularly applied to petroleum products, but its application to undistilled materials has been less frequent (4,10-12). Redgewood and Cooper ( 1 9 ) have detected polynuclear hydrocarbons in gas works waste by a combination pf chromatography and ultraviolet spectroscopy. I n developing the present procedure, preliminary experiments to evaluate adsorbents and operating conditions were performed. -4 sample containing equal weights of n-octadecane, 1-methylnaphthalene, and methyl stearate was chromatographed on five adsorbents: silica gel (Davison Codes 912,923, and 950), alumina (Fisher A541/2), and carbon (Nuchar C-190 unground). Effluent portions of constant volume were collected separately, the quantity of residue in each fraction wag estimated after evaporation of the solvent, and the composition of each residue was estimated from its infrared spectrum. Silica was the best adsorbent. The three silica gels were nearly equivalent in separation efficiencv, but on the basis of flow characteristics, minimum discoloration, and economy, the Code 950 gel was slightly preferahle. Separation was also improved by wetting the adsorbent with the first solvent before adsorption of the sample. Optimum volumes of eluting solvents 11-ere determined in similar experiments. Recovery of Chromatographic Fractions. The quantitative recovery of petroleum materials from solution in volatile solvents is generally recognized to be a complex problem (8). The conditions finally adopted resulted in complete removal of solvent (as %vasshown by infrared spectra) with the attainment of con-