Picric A c i d M e t h o d for Determination of Aromatic Content OF Aviation Gasoline C. M. GAMBRILL AND 1. B. MARTIN', Ethyl Corporation, Detroit, M i c h . Dobryanskii and Tikhonov-Dubrovskii utilized the difference in solubility of picric acid in aromatic and nonaromatic hydrocarbons to develop a method for determining the aromatic content of aviation gasoline. The method has been modified to increase its accuracy, precision, and rapidity. The solubility of the picric acid in gasoline i s determined at some temperature between 2O"and 30"C. A p p l y -
ing a temperature correction to obtain the solubility at 2 5 " C., the percentage of aromatics is obtained b y reference to a standard curve. The presence of olefins and naphthenes necessitates a slight correction: -1.4% for each 10% of olefin and -0.6% for each 10% of naphthenes. The method is accurate to h i % in the range of 0 to 2570 of aromatics.
A
where t is the observed temperature in degrees Centigrade. The percentage of aromatics in the gasoline is then read from one of the curves presented with this report, as discussed below.
METHOD for determining the aromat,iccontent of aviation gasoline by utilizing the difference in solubility of picric acid in aromatic and nonaromatic hydrocarbons was reported by Dobryanskil and Tikhonov-DubrovskiI ( 2 ) . Their method f o ~ , diitermining the dissolved picric acid was tedious, and they did iiot consider the efect of variations in temperature on t,he solubility. The authors have developed a more accurate, more precise, and more rapid method for determining the dissolved picric acid. By making a temperature correction, the solubility of the xrid may be determined a t any convenient temperature bettveen 20" and 30" C. (a xider range may be used if needed). The per rent of aromatics in the gasoline is then obtained directly from :i (xrve of aromatic content versus solubility. The procedure is most accurate in the determination of a siugle Aromatic hydrocarbon in paraffin hydrocarbons, but is sa&f:i.ctory for those mixtures of aromatics which may be found in aviation gasoline. The presence of olefins and naphthenes in gasoline necessitates a slight correcbion: -1.4y0 for each 10% of olefins and -0.6% for each 10% of naphthenes. For high :iccuracy, therefore, the bromine number of cracked gasolinw .tiould be determined and the approprintc, correction made. METHOD
60-ml. sample of the gasoline is measured into a 125-nil. Squibb-type separatory funnel containing 1 gram of pure picric acid. Commercially available C.P. or reagent quality picric acid containing up to 10% water gives precise results, but for a large number of determinations the dry picric acid is more convenient. The sample is brought to room temperature, which should be between 20" and 30" C., and is shaken for 5 minutes, either by hand or by means of a mechanical shaker. If shaken by hand, the flask is held in the finger tips, so that the heat from the hand will not cause a change in temperature during the shaking period. The temperature of the sample is then recorded and the sample is filtered into a 125-m1. Erlenmeyer flask through a Whatman S o . 1 or other fast filter paper. A 50-ml. sample of the filtrate is pipetted into a 250-ml. iodine flask containing 40 ml. of distilled water. Three drops of 771-cresol purple indicator are added, alkali is added to the sample from a buret, while the rled gently, until the color of the indicator is changed to purple and approximately 1-ml. excess of alkali has been added. The mixture is then shaken vigorously for one minute. The two layers are allowed to separate, after which the excess alkali is titrated with 0.01 AT hydrochloric acid to the complete disappearance of the reddish-purple color of the indicator as viewed under a tungsten filament lamp. The dissolved picric acid in 100 ml. of a saturated solution of picric acid in the gasoline is expressed by the equation:
EXPERIMENTAL
The divelopment of the method included the following investigation: Titration of the picric acid dissolved in the gasoline. Effect of temperature on the solubility of picric acid. Satisfactory shaking time to saturate gasoline with picric acid. Effect of purity of picric acid on its solubility in gasoline. Solubility of picric acid in blends of aromatics with iso-octane (2,2,4trimethylpentane, Rohm and Haas). Effect of olefins and naphthenes upon the solubility of picric acid. SELECTION OF TITRATION METHOD. I n this p,)rtion of the work the following points were studied: The most suitable indicator for the titration. Direct titration of the picric acid compared with back-titration of an excess alkali. Size of the gasoline sample. Concentration of alkali. A blend of 10% benzene in iso-octane was prepared and 500 ml. TTere shaken with 5 grams of picric acid for 30 minutes. Samples of 10 to 15 ml. of this gasoline, saturated with picric acid, were pipetted into 125-m1. iodine flasks containing water and the acid was titrated: (1) by adding 0.01 or 0.05 LValkali with long shaking until an end point was reached with phenolphthalein indicator, or (2) by adding an excess of alkali as evidenced by the color change of thymol blue, bromothymol blue, phenolphthalein, or m-cresol purple, and back-titrating with 0.01 1\1 acid.
These tests shoxed that thymol blue did not give a sat.isfactoi-y end point,, probably because it was soluble in the gasoline; that bromothymol blue had a green to yellow end point that was very difficult to detect in the picric acid solution; and that a higher solubility was obtained by using phenolphthalein than by using rn-cresol purple, This difference may be attributed to differences in the pH range of the two indicators, to the fact that more carbon dioxide may be titrated as picric acid by using phenolphthalein, and that the m-cresol purple end point was more easily detected than the phenolphthalein end point. The selection of m-cresol purple made it necessary to use the back-titration procedure, since it M as impossible to extract all the picric acid from the gasoline phase a t the alkaline pH range of this indicator (pH 7.4 to 9.0). The back-titration procedure had a further advantage in that the solution required less shaking to reach the end point. I n varying the size of the gasoline sample for Mg. of picric acid = 458.2 X (ml. of S a O H X A- KaOH - ml. of HC1 X zV HC1) 100 ml. titration from 10 to 50 ml., it was noted that the corrected to the value for 25" larger the sample the lower the apparent solubility of picric The measured soluhllity by the equation: acid, This effect was attributed to interference by small amounts of carbon dioxide dissolved in the water. Although it mg. of picric acid [l 4-0'035(250 Solubility a t 25 ' C. = nould be possible to correct for this aliquot error empirically, it 100 ml. seemed advisable to use the larger gasoline sample. In 'Omparing Obtained by using 0'01 1 Present address, Agricultural Cheinistry Department, Purdue C'niversity, alkali, it was found that the more concentrated alkali gave reLafayette, I n d
c.
c'
689
.
690
INDUSTRIAL
AN^
ENGINEERING CHEMISTRY
sults of higher precision. It. is probable that the 0.01 11: alkali contained sufficient carbonate to cause the poorer results. Therefore, 0.05 N alkali is recommended. DETERMINATION OF TEXPERATC'RE COEFFICIEXT.Measurements of the solubility of picric acid were made a t room temperature and a t approximately 0" C., in blends of aromatics with technical grade iso-octane and in Houdry process gasoline. These solubilities were inserted into the ideal solubility equation to obtain a constant by which solubilities determined near 25" C. may be corrected to give the solubility a t 25" C.
I
I
P
Vol. 18, No. 11
25-
-
2c-
-
?
-
B 2
-
-
I5
0
0 0
-
Numbers refer l o mixlurcs of benzene toluene and xylene in the g h n prohrlionr.
In using ideal solubility Equation 1, weights W1 and W Zin milligrams of picric acid per 100 ml. of gasoline have been substituted for mole per cents nl and n2,since
o Thlr invesflgolion. 0 From data by Dobryanskli and Tikhmov- Dubrovokii.
the solubilities are low. The constant, AH/2.303R, has been designated by S, giving the equation the following form:
0
I
I
I
50
100
150
I 200
I
I
250
300
350
rng. Picric Acid/100 ml. at 25'C. Figure I. Solubility of Picric A c i d
After solving for S in Equation 2, using the experimental values from Table I, the fractional increase in solubility per degree a t
25" C.,
i ,has been calculated from Equation 3,
(3) Since the variation of solubility with temperature over a restricted range will be practically linear, solubilities determined between 20 ' and 30 O C. may be corrected to 25 O C. by the relation:
"
[ + w (25 -
Solubility a t 25 C. = observed solubility 1
I
t)
(4)
where t is the observed temperature in degrees Centigrade.
EFFECTOF PURITY OF PICRIC ACID ON ITSSOLUBILITY.Various stocks of picric acid were found to differ in solubility in the same gasoline. Recrystallization increased the solubility of impure samples, but did not change that of the reagent grade material. The solubility of the picric acid is thus a function of its purity. As shown in Table 11, drying the acid has no significant effect on the determination. SOLUBILITY OF PICRIC ACIDIN BLENDSO F AROMATICS. Blends of aromatics with iso-octane were prepared and the solubility of picric acid in these blends was determined by the method described above. The results of these determinations are given in Table 111, and are plotted in Figures 1 and 2. All compositions given in this paper are reported on a volume per cent basis.
Table I gives the data obtained and the calculated valueh of
w/W.
.
A value of 0.035 for w/W is recommended for correcting solubilities to 25' C., but should not be used for solubilities measured outside the range of 20" to 30" C. For work of high accuracy it is recommended that the temperature be maintained a t 25" C. by means of a constant-temperature bath. DETERMIXATION OF SHAKING TIME. A study was made to determine the minimum shaking time in which a solution of picric acid in gasoline would reach equilibrium. It mas found that for a fuel high in aromatics, such as Houdry process fuel, the amount of picric acid dissolved in the gasoline was almost a t a maximum after 5 minutes of shaking, being only 1 to 273 greater after 20 minutes, For the iso-octane (07' aromatics), however, a 2minute shaking period was as adequate as 5 minutes; only a slight increase in solubility was noted after a longer shaking period. A 5-minute period has been selected as the most satisfactory.
Table
I.
Table
TI
U'l
Tz
Ti.*
Solubility of Picric Acid Undried Dried M g . / I O O nil.
63.2 61.4 297.1 301.4 301.5
B
Table 111.
70 Aromatics 0 Benzene 5 10 20 25 Toluene 3 10 20
(1
Xylene (commercial) 3 10 20 25 l:l:linixa 2 5 10 15 25 1 : 3 : 2 mixD 2 5 10 15 25
+ $)
23.6
10.9
1.4
4.7
1.035
26.6 26.0
59.1 262.3
$0.5 0.7
22.0 119.3
1.036 1.029
26.6 26.2 25.5
74.2 334.6 167.1
0.5 0.5 0.5
29.2 167.4 43.8
1.034 1.025 1.051
63.5 62.1 299.2 298,9 298.2
Determination of Aromatics in Blends
Aromatic Blends with Iso-octane
Temperature solu- Coefficient, bility, Equation 31
Gavoline Iso-octane Benzene and iso-octane blend l O 7 ! benaene 2370 benzene Xylene and iso-octane blend 10% xylene 257, xylene Houdry gasoline
Effect of Drying Picric A c i d on Its Solubility in Hydrocarbon Blends
Blend A
Temperature Coefficient Data 801uTempera- bility, T e m p e k ture, O C . , hIg., ture, C., JIg.,
II.
Solubility of Picric Acid, 25' C. .Ug./100 ml. 11.0 26.9 55.8 163.0 253.1 23.1 69.3 198 8 24.0 70.2 206.5 320.5 17.3 31.5 64.0 115.4 286.2 18.5 32.8 67.3 119.8 297.8
A mixture of benzene, toluene, and conimercial xylene in given proportions. Q
ANALYTICAL EDITION
November, 1946 Table IV.
Solubility of Picric A c i d in 10% Blends of Aromatics in Iso-octane
Aromatic Compound Possihle aviation gasoline constituents Benzene Toluene Xylene (commercial) o-Xylene m-Xylene ni,p-Xylene (mixture) Ethylbenzene Cumene Improbable aviation gasoline constituents hlesitylene Diethylbenzene Triethylbenzene m-Diisopropylbenzene p-Cymene n-Butylbenzene sec-Butylbenzene tert-Butylbenzine sec-Amylbenzene tert-Amylbenzene
Solubility of Picric Acid 3fg./100 nil.
(1:3:2Curve) %
Aromatics
56 69 70 82 81 78 48 36
8.7 10.4 10.5 11.8 11.7 11.3 7.6 5.7
99 41 41 22 29 33 26 27 20 24
Table
V.
Effect of Olefins and Naphthenes on Solubility A c i d in Gasoline
B.P. (Theoretical)
80.1 110.8
A.
...
144 138.8
Naphthenes present
10% benzene, 9% methylcyclohexane, 81% iso-
59.1
10.4
10% toluene, 9% methylcyclohexane, 81% iso-
73.3
10 5
10% xylene, 9% methylcyclohexane, 81% iso-
77.7
10.6
10% l'l:la,957, methylcyclohexane, 81% iso-
69.7
10.5
71.7
10.5
85.8
13 8
98.8
13.0
106.4
13.4
94.5
13 2
23.1
3.0
11.3 12.6
0 0.3
76.0 100,s 77.7 97.1 82.9 102.3 79.3 98.7 13.5 22.2
11.4 13.7 11.5 13.4 11.8 13.6 11.4 13.3
94.8
13.2
99.0
13.3.
octane octane
130:2 153
13.4 6.5 6.5 3.0 4.4 5.1 3.9 4.1 2.6 3.5
octane
octane 10% 1:3:Za,9% methylcyclohexane, 81% isooctane lORbenzene, 45% methylcyclohexane, 45% iso0. F. t ._ I I ". P. 10% toluene, 45% methylcyclohexane, 45% isooctane 10% xylene, 45% methylcyclohexane, 45% isooctane 10% l:l:la, 45% methylcyclohexane, 45% isooctane 0% aromatics, 50% methylcyclohexane, 50% iso-octane 0% aromatics, 25% cyclohexane 0% aromatics, 25% o-dimethylcyclohexane
164.8 182 218 204 176 180 174 169 189 190
B.
Olefins present 1 : 1:In, 10% amylene, 80% iso-octane 1 ' 1 ' la 2 5 7 amylene 65% iso-octane 1;1: la' 1 0 4 diamylehe, 80% iso-octane 1 : 1 1.: 25% diamylene, 65% iso-octane 1 :3:2", 10% amylene, 80% iso-octane 1:3.2" 257 amylene, 65% iso-octane 1:312": 10% diamylene 8 0 7 is0 octane 1:3:Za,25% diamylene: 65% iso:octane aromatics, lOy0 cyclohexene aromatics, 257, cyolohexene
C. Naphthenes and olefins present 10% 1:1:15, 25% methylcyclohexane, 10% diamylene, 55% iso-octane 10% 1:3:2",25% methylcyclohexane, 10% diamylene, 55% iso-octane
matics, not especially purified, are given in Table VI. ANALYSISOF AVIATIONGASOLINES. Since the different aromatic compounds give slightly different solubility curves, it is necessary in applying the picric acid method to gasoline to make some assumption regarding the composition of the aromatics in the gasoline. The authors have assumed that a 1 to 3 to 2 mixture of benzene, toluene, and xylene gives a good average representation of the aromatics in most aviation gasoline, and have
Solubility of Picric Acid G./lOO ml. aromatic
Temperature
Benzene Toluene Xylene (commercial)
7.550 8,950 9.810
22.4 22.3 22.8
B
C D
I
I
I
I
IS0
200
250
Figure 9.
mg. Picric AcidA00 ml. Solubility*of Picric A c i d
I N O
1
VI. Solubility of Picric A c i d in Undiluted Aromatics
Aromatic
Gasoline A
100
2.9
SOLUBILITY OF PICRIC ACIDIN PUREARO.\~ATICS. The results of measurements of the solubility of picric acid in undiluted aro-
Table
'
0.6
0 A mixture of benzene, toluene, a n d commercial xylene in given proportions.
Table
50
of Picric
Aromatics Calculated Solubility from Picric of Picric Acid Solubility Acid Mg./lOO ml. R
c.
A plot of the values obtained b_v DobryanskiI and TikhonovDubrovskiI is included in Figure 2 for the sake of comparison, I t is believed that the differences in intercept and in slope betlyeen their curve and the authors' are due to (1) the presence of olefins or naphthenes in their base stock, (2) the presence in their xylene of some impurity, such as ethylbenzene, in which picric acid is less soluble, and (3) the fact that their determinations were made a t temperatures some 10' lower than the authors'. Besides the three aromatics for which solubility values are reported in Table 111, a number of other aromatics were tested in 10% blends with iso-octane. These results are given in Table IV. EFFECT OF OLEFINSAND SAPHTHENES. Blends of iso-octane were prepared with cyclohexane and its methyl and o-dimethyl derivatives and with amylene, diamylene, diisobutylene, and cyclohexane. Various aromatics were added to these blends, and the solubility of picric acid in the mixture was-determined. The apparent content of the aromatic compound or mixture used was then calculated from the curve given by that aromatic in mixture with only iso-octane. The results of these tests, given in Table V, show that both naphthenes and olefins cause a small increase in the apparent aromatic content, the effect of the olefins being greater. The error in terms of per cent aromatics is approximately +1.4y0for each 1Oyoof olefins and not more than + O . S S for each 1Oyonaphthenes.
o!'
691
E F G
350
H a
c.
VII. Determination of Aromatics in Aviation Gasoline Per Cent Aromatics Picric acid Ultraviolet method absorption ( I ) 0 0.2
