Table 111. Analysis of Aqueous Extracts of Airborne Particulate Samples
Nitrate/ml. ____.Extract 1-Aminopyrene 2,4-Xylenola 4 . 3 i 0.16 3.8 11.0 i 0 . 2 9.4 10.3 i 0.1 9.8 - 8 . 9 0:i 8.3 7.6 f 0.1 8.1 1.2 f 0 . 1 1.5 1 . 8 0.1 2.0 4.3 f 0.1 4.8 2.6 f 0 . 0 2.5 a These results were taken from a series of routine analyses obtained with the help of a Beckman DU spectrophotometer by n group under the direction of Norman Huey of the Air Quality Network at the Robert A. Taft Sanitary Engineering Center. Duplicate determinations. pg. ______.
cedure as compared to its reaction in the brucine, xylenol, and aminopyrene procedures. Formaldehyde is an interference. The main difficulty in the chromotropic acid procedure is associated with the interference of the blank. The chromotropic acid in the blank has strong bands (with decreasing intensity) at 347, 331, and 313 mp, while the band at which analysis takes place is a t 357 mu.
Application. Aqueous extracts of airborne particulates were analyzed by the 1-aminopyrene procedure and the 2,4xylenol method ( I ) , Table 111. The latter method is the one used routinely in the local laboratories. The comparative procedure times of the two methods emphasize the simplicity of the 1-aminopyrene procedure compared to the 2,4xylenol method. The xylenol method involves about 17 steps consisting of a 30-minute heating period, two liquid-liquid extractions, and a filtration, all of which tend t o make the method much more tedious and prone to error in routine analysis than the 1aminopyrene procedure. Consequently, the latter procedure is preferred. Since the aqueous extracts of the airborne particulates did not contain the nitrite ion, the step involving destruction of nitrite was not necessary. Although airborne particulate samples contain a fairly large amount of iron, the aqueous extracts contained only minute traces and therefore ferric salts were not a serious interference in the analyses. The 1-aminopyrene procedure is especially recommended where a simple, sensitive, and direct method is needed for the determination of nitrate in mixtures containing little or no nitrite and ferric salts. It should also be useful in
Determination of Phenolic Substances UItravioIet Difference Spectrometry
the determination of total nitrate and The nitrite nitrogen in mixtures. method does not have the Beer's law relationship or the reproducibility of the 2,6xylenol method (4-6)but does have a much greater sensitivity. LITERATURE CITED
( 1 ) Barnes, H., Analyst 75, 388 (1950). (2) Fisher, F. L., Ibert, E. R., Beckman, H. T., ANAL.CHEM.30, 1972 (1955).
(3) Greenberg, A. E., Rossum, J. R., Moskowitz, N., Villaruz, P. A , , J . Am. Wufer Works Assoc. 5 0 , 821 (1958). (4) Hartley, A. M., Asai, It. I., ANAL. CHEM.35, 1207 (1963). ( 5 ) Ibid., p. 1214. (6) Hartley, A. M., Asai, R. I., J . Am.
Water Works Assoc. 52, 255 (1960). (7) Holler, A. C., Huggett, C., Rathmann, F. H., J. Am. Chem. SOC.72, 2034 (1950). (8) Hora, F. B., Webber, P. J., Analyst 85, 567 (1960). (9) Johnson, C. M., Ulrich, A., ANAL. CHEM.22, 1526 (1950). (10) Montgomery, H. A. C., Dymoclr, J. F., Analyst 87,374 (1962). (11) Swain, J. C., Chem. Ind. Lmdon 1957, 470. (12) Swinehart, B. A., Brandt, W. W., Proc. Indiana Acad. Sn'. 63, 133 (1953). (13) Taras, M. J., ANAL.CHEM.22, 1020 (1950). (14) West, P. W., Lyles, G. L., Anal. Chim. Acta 23, 227 (1960).
RECEIVEDfor review June 13, 1963. Accepted August 16, 1963.
bY
ARTHUR S. WEXLER Dewey and Almy Chemical Division, W.
b The selective determination of phenols in the presence of nonionizing interfering substances by ultraviolet difference spectroscopy is discussed. The difference spectrum of the alkaline form of a phenolic substance in water or alcohol is recorded directly in an ultraviolet recording spectrophotometer against an identical concentration of the substance in neutral or slightly acidified sofvent. The resulting difference spectrum is a characteristic and useful indication of the concentration and chemical identity of the phenolic substance. Possible interferences due to nonionizing, nonphenolic species are usually canceled out in the difference spectrum. Applications to the polymer field of analysis are discussed.
T
CHANGE in the ultraviolet spectriim of phenolic materials as a function of basicity ha.: been the basis of determinations of the phenolic content of gasolines by Murray ( Y ) , waste water HE
1936
ANALYTICAL CHEMISTRY
R. Grace and Co.,
Cambridge 40, Mass.
by Schmauch and Grubb (8), rubber by Wadelin (IO), surfactants by Smullin and Wetterau (9), and food products by Englis and Wollerman (4). The basis of these determinations is the well known bathochromic shift of the long wavelength maximum of phenols in alkaline solutions due to the formation of phenolates as shown by Coggeshall and Glassner (a) and by Doub and Vandenbelt (3). The differenre spectrum obtained by subtraction of the absorbance of the neutral solution of phenolic type materials from the absorbance of the alkaline solution has been utilized in the study of types of lignin and lignosulfonates by Aulin-Erdtman (1). Goldschmid (5) and Maranville and Goldschmid (6) developed difference spectra methods for the analysis of lignosulfonates and of tannins (polyphenols). I n this paper the identifications and quantitative estimation of some phenols by mwns of their ultrnviolet difference sj)~(+r:i me re1iortmI. While the method
was developed for the rapid and direct estimation of phenolic antioxidants in synthetic rubber latex and in rubber products, it is obvious that the reported procedure is applicable to a wide variety of materials and problems involving the detection and identification of phenolic materials. EXPERIMENTAL
Apparatus. All ultraviolet determinations were made on a Beckman DK-2 recording spectrophotometer. Absorbances and wavelengths were occasionally checked on a Beckman D U spectrophotometer. Both instruments were calibrated for wavelength with a mercury source and a holmium oxide standard. Matched 1-cm. and 0.2-rm. silica cells were used. Materials. Commercial and reagent grade phenols were used without purification. Rragent grade potassium hydrolide and reagent grade methanol were used. Procedure. One per cent stock solutions of phenols ill methnnol were
Figure 1 . A
Ultraviolet spectra of phenol
- - - - 30 P.p.m. phenol in 1 N potassium hydroxide in methanol -.....- 30 P.p.m.
B C-
phenol in methanol difference spectrum with alkaline solution in sample beam and methanol salution in reference beam Conditions: 0.20-cm. cell paths
30 P.p.m.
Figure 2. Ultraviolet difference specira of phenol
-InBlank o f 0.1N potassium hydroxide methanol nethanol B - - - - 5 P.p.m. difference spectrum C - *. 12.5 P.p.m. diff Erence spectrum D . ....... 25 P.p.m. difference spectrum
A
VI.
*
E
F
*
9-
-.-
-..._.-
Conditions:
37.5 P.p.m. difference spectrum 50 P.p.m. difference spectrum
0.20-cm. cell paths (matchedqto within 6 parts in 2000 a t 238 mp), 0.1 N alkali solution in sample! beam and methanol d u t i a n in reference beam
Figure 3. Ultraviolet spectra of pmethoxyphenol (methyl ether of hydroquinone)
- .... .- 50 P.p.m. in 0.1N potassium draxide salution in methanol 8 - - - - 50 P.p.m. in 1.ON potassium droxide solution in methanal c - .- .- .- 50 P.p.m. in methanol A
D-
hyhy-
50 P.p.m.
difference spectrum with 1N alkaline solution in sample beam and methanol solution in the reference beam Conditionr: 0.20-cm. cell paths
HO-CT>-OCH. KIYFIFNOTH,
M
p
VOL 35, NO. 12, NOVEMBER 1 9 6 3
1937
prepared. Further dilutions in methanol were made as required with 5.0 or 100 X Hamilton syringes. The precision of a dilution was generally within 5 parts per thousand. Potassium hydroxide 1N solution in methanol was prepared by dilution of 12.5N concentrated solution in water with methanol (12.5 fold dilution). Solutions, 0.1N1 were prepared by 125-fold dilutions of the concentrated potassium hydroxide solution with methanol. RESULTS AND DISCUSSION
WAVELENGTH. M j l
Figure 4.
Ultraviolet spectra of 4,4’-rnethylene bis(2,6-di-fert-ButyI phenol)
- - - - 50 P.p.m. in 1N potassium hydroxide solution in methanol - . . . , .- 50 P.p.m. in methanol
A B C-
50 P.p.m. difference spectrum with alkaline solution in sample beam and methanol solution in reference beam Conditions: 0.20-cm. cell paths CHn
CH3
CH3-C-CHs
CHa-C-CHa
I I
I I
Figure 5. Ultraviolet spectra of 4,4’thiobis(6-terf- butyl-rn-cresol)
- 50 P.p.m. in 1N potassium hydroxide solution in methanol - .....- 50 P.p.m. in methanol
A---
,701
50 P.p.m. difference spectrum with alkaline solution in sample beam and the methanol solution in the reference beam Conditions: 0.20-cm. cell paths C-
CHa
CH:
HO-&S-b-OH
I I
CHs-C-CHa C Ha
1938
(Santowhite)
I I
CHrC-CHs CHa
ANALYTICAL CHEMISTRY
0301
Ultraviolet spectra were obtained over the region 340 mp to 220 mp. Instrumental settings were as follows: time constant 0.2 minute, slit setting 0.08 mm. a t 280 mp; scale in “absorbance.” Speed of scan was slow enough to record maxima accurately. Measurements were usually made within a few minutes after dilution of stock solutions. Repeat measurements were made 1 hour later to detect any changes. Unhindered monohydric phenols were stable in this time period. Polyphenols in certain cases yielded time varying spectra. Certain hindred phenols yielded variable results as follows: full ionization obviously was not achieved in 1N solution of phenols with both ortho positions occupied by a tertiary butyl group. A 1N solution of potassium hydroxide was sufficient to cause about 60 to 70% ionization in such cases. Results are reproducible and the difference spectra are sufficiently well defined for characterization. Figures 1, 3, 4, and 5 show the changes occurring on ionization of phenols listed as 1, 2, 10, and 11 in Table I. In each case a shift in the long wavelength maximum in the 270 to
no
=Op, v
~
VJ
li:
60
t
I
i v
[
I
I
I
r0 0 - J
I
50
e
401
1
33,
20
-
io -
W A V E L E N G T H , M,U
Figure 6.
---- . .. .-
Ultraviolet difference spectra of phenols
A p-Methoxyphenol B . Benzyl ether of hydroquinone CHydroquinone Conditions: 50 P.p.m., 0.20-cm. cell paths, 1 N potassium hydroxide solution in methanol in the sample beam and methanol solution in the reference beam
I
c
t
i
t
4
I
I 1
m
m
e
2 5 (Y
0
h
i
u)
VOL. 35, NO. 12, NOVEMBER 1963
0
1939
290 mp region of 15 to 20 rnfi is observed upon ionbation. The shift is in all cases toward longer wavelengths. The difference spectrograms in each case exhibit 2 maxima and 2 minima. For comparison, difference spectra of 12 phenols are shown in Figures 6, 7, and 8. Each spectrum is sufbiently unique to distinguish and identify each phenol. Difference spectra are useful in making quantitative estimates of phenols as shown by the data in Table I1 and Figure 2 for phenol itself. Close adherence to Beer’s law is observed for the Merence spectra peaks found a t 288 and 236.5 mw. Data on difference spectra absorptivities are reported in Table I11 for 13 phenolic substances.
I\\
!
Table 1. Ultraviolet Spectra for the Following Phenols Are Reported in Figures 1 through 14
1. Phenol Merck 2. p-Methox phenol EKa 3. Benzyl etler of hydroauinone R. T. Vanderbilt 4. Hydroquinone EK 5. 2,GDimethylphenol EK 6. 2,GDimethoxyphenol EK 7. 2,6-Di-terl-butylphenol EK 8. 2,6-Dimethyl014 methylphenol EK 9. 2,6-Di-tert-butyla-dimethylamino-pcresol Ethyl 703 10. 4,4’-Methylenebis(2,6-di-tert-butylphenol) Ethyl 702 11. 4,4’-Thiobis(Gtert Monsanto Santobut 1-m-cresol) white 12. 4,4’-&iobis(G-terG butyl-o-cresol) Ethyl 736 13. 2,2’-Methylenebis(emethyl-6American Cyante7t-butyIphenol) amid 2246 a EK = Eastman Kodak.
WAVELENGTH,
Figure 8.
- .. ... c ----D---E -. - .- . A
B
Conditionrr
M)l
Ultraviolet difference spectra of phenols
2,6-Dimethylphenol 2,6-Dimethoxyphenol 2.6-Di-fed-butylphenol
2,6-Dimethylol-5-methylphenol 2,6-Di-ferf- butyl-2-dimethylamlno-p.cre~l 50 P.p.m., 0.20-em. cell paths, 1N KoH/methanol
VI.
A
B
H 0
H
H
0
0
C H I .
I
C f i O b O C H *
C Y J H - ( ) - ? - C f iC Y
D
E H
H
?
methanol
C
c
?
c
CHI
I
N
Table II. Ultraviolet Difference Spectra Absorptivities of Phenol
Absorptivities 288 288 236.5 236.5 P.p.m. mra mpb mro mp‘ 5 33.5 25.6 110 102 12.5 33.9 26.0 112 107 25.0 33.9 26.8 111 104 37.5 34.2 26.8 114 107 50.0 33.5 26.1 111 104 a Measured from the minimum a t 268 mP. Measured from the minimum at 340 mr. Conditions: Cell paths 0.20 cm. (cells matched to within 7 parts in 2000 a t 236.5 mp). Solvents 0.1N potaasium hydroxide in methanol in the sample beam and neutral methanol in the reference beam. Absorptivity is in units of litera per grams-’ per centimeters-].
1940
ANALYTICAL CHEMISTRY
i ’\
CY CY (Ethyl 703)
Table 111. Ultraviolet Difference Spectra Data for Phenols Sub- Wavelength, Absorp Wavelength, AbsorpMiniPeak stance mp tivitya mr tivity“ mum, mr ratio 1 236.5 112 289 47 269 2.38 2 239.5 81.5 310.5 34 285 2.40 3 242 45 310 17 287 2.65 30 294 1.30 4 244 39 316 5 241 71.5 290 31.5 269 2.27 6 248 52.5 285 23.5 320 2.23 7 251.5 43 297 19.0 330 2.26 8 245 38 305 25.5 279 1.49 9 258 41 300 17 330 2.40 283 6.75 54 297.5 8 10 256.5 52 295 26 242 2.00 11 269 44.5 297 9 247 4.94 12 272.5 20 306 22 278 0.91 13 250 a Absorptivity is in units of liters per grams-’ per centimeters-l.
Amp 52.5 71 68
72 49 37 46
GO
42 41 26 24.5 56
100 I
I
I
I I
0 90
I
I I
I 0 80 I I I
4
I
i
070
Figure 9. Ultraviolet spectra of alkyl benzene sulfonate (Ultrawet K)
I I
I
060
Am--C
--
I
I w
p
: A
050
I I I 1
e Y) 0 040 I
I \
I
1 $
I
0 30
170Difference spectrum with alkaline
( 1 N potassium hydroxide in methanol)
solution of Ultrawet K on the sample beam and methanol solution of Ultrawet K in the reference beam Conditionr: 1 %, 0.20-cm. cell paths
I
4
1% Solution in methanol
I
I
1 I
5'
! 0.20
1
',
I
I I
C
!
010
0
WAVELENGTH,
hip
I 00
090
OBU
0 70
0 60
0.50 Irl
;
040
C> 0: 0,I
0 30
i' 20
0 IO
0
L*O
250
240
260
270
280
290
x)O
310
320
330 34!J
W A V E L E N G T H , M )J
Vigure 10. Ultraviolet spectra of mixtures of 50 p.p.m. o f 4,4'-thiobis(&fertbutyl-rn-cresol) and 1% alkyl benzene sulfonate I\ €1 (:
-- .-..-..-- Mixture Mixture dissolved in 1N potassium hydroxide in methanol dissolved methanol -Difference spectrum o f the mixture with alkaline solution in the sample beam and methanol solution in the reference beam
Conditions:
in
50 P.p.m. of antioxidant, 10,000 p.p.m. of surfactant, 0.20-cm. cell paths VOL 35, NO. 12, NOVEMBER 1963
1941
Figures 9 through 14 are examples of the value of the difference spectrum method in the determination of mixtures containing other ultraviolet absorbing substances which would interfere in the direct determination of phenols and such mixtures. Figure 9 shows spectrograms of a 1% solution of alkyl benzene sulfonate (Ultrawet K). Curve A shows an absorbance greater than one a t 270 mp and a peak maximum greater than 2 a t 262 mp. For clarity, the spectrum of alkyl benzene sulfonate in 1N potassium hydroxide has been omitted. Its curve is almost identical with that of curve A. A small difference is observed as shown by the difference spectrogram of curve
C.
WIVELENGTH,
Figure 1 1.
htp
Ultraviolet spectra of disproportionated rosin soap
- - - - 0.23% -.....- 0.23%
in 1N potassium hydroxide solution in methanol in methanol CDifference spectrum with alkaline solution in sample beam and methanol solution in reference beam Conditions: 0.23%, 0.20-cm. cells paths A
B
W A V E L E N G T t i , M,LI
Figure 12. Ultraviolet spectra of mixtures of 0.005% 4,4’-methylenebis(2,6di-terf-butylphenol) and O.23y0 disproportionated rosinate A B C
- - - - Mixture dissolved in 1 N potassium hydroxide in methanol - . . . . .- Mixture dissolved in methanol -Difference spectrum with alkaline solution in the sample beam and methanol solution in the
reference beam Conditions: 0.00570 Antioxidant and 0.2370 rosinate, 0.20-cm. cell paths
1942
ANALYTICAL CHEMISTRY
Ultraviolet spectrograms of a mixture of 1% alkyl benzene sulfonate and 50 p.p.m. of Santowhite (substance 11 in Table I) are shown in Figure 10. The direct spectra of curves A and B appear typical of alkyl benzene sulfonate with some hint of a second substance. The difference spectrogram shows almost complete cancellation of the spectrum of the surfactant and reveals quite clearly both quantitatively and qualitatively the presence of Santowhite (compare curve C of Figure 10 with curve C of Figure 5 ) . Figures 11 and 12 illustrate the use of the differential technique in the detection and the determination of Ethyl Corporation antioxidant 702, (substance 10 in Table I) in the presence of a much larger amount of disproportionated rosin soap. Curves A , B, and C in Figure 11 are, respectively, the spectrograms of the soap alone in one normal alkali, in neutral methanol, and the difference spectrum of the alkaline and neutral solutions. I n Figure 12 are shown spectrograms of a mixture of 50 p.p.m. of antioxidant and 2300 p,p.m. of rosinate soap. The direct spectra A and B are typical of the soap and reveal little evidence of the presence of the antioxidant. The difference bpectrum curve C shows almost complete cancellation of the surfactant spectrum. Comparison of curve C of Figure 12 with curve C of Figure 4 shows the antioxidant to be Ethyl 702 and also shows the quantity to be close to 50 P.P.?. Figures 13 and 14 illustrate the detection of 50 p.p.m. of p-methoxyphenol in a mixture with 1000 parts per million of an aromatic surfactant, Triton X-100. The spectra of the neutral and alkaline solutions of Triton 9-100 are almost identical lis shown by Figure 13. The spectra of the mixture of p-methoxyphenol and Triton 5-100 as shown in Figure 14 (curves 4 and B ) indicate tlie presence of two components in the direct spectrum. 1t would be easy to detect Triton X-100 ab shown by the presence of two peaks a t
100
4 090
Figure 13. Ultraviolet spectra of isooctyl phenyl polyethoxyethanol (Triton X-1 00)
0 80
0.1% in 1N potassium hydroxide solution in methanol B 0.1% in methanol Conditions: 1000 P.p.m., 0.20-cm. cell paths
A-
----
Of0
0 60 w
z 0
2
around 280 mp. The identification of the other component would be difficult in the direct spectrum. The difference spectrogram curve C shows almost complete cancellation of the spectra of Triton X-100. Comparison of this curve with curve C of Figure 3 established both qualitatively and quantitatively the presence of p-methoxyphenol. The ultraviolet difference spectrogram of the ionized and neutral forms of phenols is of value in identifying the phenol even in the presence of large quantities of interfering materials which mask or completely obscure the direct spectrum. Difference spectrum analysis is of value in determination of phenols in mixtures which are time-consuming or difficult to separate. Interferences are then reduced to the small number of substances which also undergo significant changes in spectra with changes in basicity (enols, thiophenols). The difference spectrum is of value in structural analysis of phenols particularly if taken in conjunction with level of basicity as in differentiation of hindered from unhindered phenols or detection of degree of steric hindrance.
050
LL 0 v)
4 m
040
030
020
0 10
0
W A V E L E N G T H , M,U
LITERATURE CITED
(1) Aulin-Erdtman, G., Chem. Znd. London 1955, 581. (2) Coggeshall, N. O., Glassner, A. S., Jr., J . Am. Chem. SOC.71, 3151 (1949). (3) Doub, L., Vandenbelt, J. AI., Ibid., 69, 2714 (1947). (4) Englis, D. T., Wollerman, L. A., ANAL.CIIEM.29, 1151 (1957). (. 5.) Goldschmid. O., Ibid.. 26. 1421 (1954). (6) Maranville, L. F., Goldschmid, O., Zbid., p. 1423. (7) Murray, M. J., Ibid., 21,941 (1949). (8) Schmauch, L. J., Grubb, H. M., Zbid., 26,308 (1954). (9) Smullin, C. F., Wetterau, F. P., Zbid., 27, 1836 (1955). (10) Wadelin, C. W., Ibid., 28, 1530 (1956). I
WIVELENGTH,
Mp
Figure 14. Ultraviolet spectra of mixtures of 0.005% p-methoxyphenol and 0.1 00% Triton X-1 00
- - - - Mixture dissolve3 in 1N potassium hydroxide in methanol Mixture dissolved in methanol c - . . . . .- Difference spec'rum of mixture with alkaline solution in the
A B-
Conditions:
solution in the reference beam 0.20-Cm. cell Flaths
.
RECEIVEDfor review April 18, 1963. sample beam ond methanol
Accepted August 5, 1963. Pittsburgh Conference on Analytical Chemistry and Applied Spectroscopy, Pittsburgh, Pa., 1963.
VOL. 35, NO. 12, NOVEMBER 1963
0
1943
