V O L U M E 2 7 , N O . 6, J U N E 1 9 5 5
1019 OR‘IFICE
SAMPLE LINE
above D. The stopcock below D is then reopened, and flow of liquid past the colorimeter is resumed. The head for the reductor, E , was enlarged so that temporary removal of the measured liquid would not affect the pressure in the system.
AMALGAMATED ZINC (‘20-MESH)
SUPPORT FOR SECOND ABSORPTION CELL
OiR E 6
COLORIMETER
Figure 1.
O x J g e n analJzer
h 1 0 ~L), and one in the gas litif, aliove D. Cloaing the fiibt ctopcock vauses liquid to acrumulate in the gas separatoi until it overflows into the graduated measuring tube: the time needed to collect 10 ml. can be measured with a stop watch. The liquid I‘ returned to the system hv opening stopcock SI nest to the reductor and applying gar pressure by throttling the stopcock
In actual operation, the observed colorimeter reading a t the measured liquid rate can be converted to transmittance a t one of two standird liquid rates by means of charts. The oxygen content of the sample is then obtained from plots of oxygen concentration os. transmittance. The latter plots ivere calibrated a t the two chosen liquid flow rates and covered the two desired ranges of oxygen concentrations. Two other modifications are shown in Figure 1. To simplify changing the liquid rate, parallel orifices were installed in the line next to the zinc reductor, E, one for each range of oxygen concaentration. A second absorption cell was inserted in the liquid line between the zinc reductor and the gas-liquid junction. With this cell, the intensity of the reduced anthraquinone solution can be checked periodically; if i t differs from the reading on which the calibration charts were based, the appropriate correction can be applied to observed readings. With these modifications, highly satisfactory results have heen obtained. Precision is about 1 p.p.m. of oxygen in the range 0 to GO, and about 5 p.p.m. in the range 0 to 400. LITERATURE CITED
(1) Brads-, L. J.,
.\NAL.
CHEM.,20, 1033 (1948).
RECEIVED for review duglist 12, 1934. -4ccepted Deceniher 22, 1954.
Colorimetric Dete rmination of N-Phenyl4 -naphthylamine in New and Used Oils W. S. LEVINE and W. A. MARSHALL Socony-Vacuum O i l Co., lnc., Brooklyn 22,
N. Y
4 rapid colorimetric method has been del eloped for determining JV-phen>l-l-naphth> lamine, an oxidation inhibitor, in new- and used industrial oils. The oil is dissolved in acetone to which diazotized-p-iiitroaniline is added to produce a blue-Fiolet color. 4 single determination can be made in 18 to 20 minutes. The accurac> is within =klO7cc,which is adequate for blending plant control w o r k and routine anal?sis.
T
HE manufacture of lubricating oils today continues to emphasize the use of additives to give the oil4 special properties. One of the aromatic aniines which function a s oxidation inhibitors is .~r-phenyl-l-naphth?-laiiiine. To study the behavior of AV-phenyl-l-naphthylaniinea3 :til oxidation inhibitor, a method for its determination in new and used oils was needed. One procedure ( 4 ) began \vit,h tlie precipitation of the amine as the hydrochloride by passing hydrogen chloride gas through the oil. The precipitate was filtered to f r w t,he oil and impurities, and finally weighed as the free amine. However. the method required about 24 hours to complete :i determination and a t times gave erratic results. Thus. the ohjrct of this work was to develop a more rapid method to nieasiiw the S-phenyl-1-naphthylamine content of new and used oil?. .1 possible approach was due to Palit ( 3 ) . N-Phenyl-1-iiaphthylamine, being a very weak base, might titrate in a noiiaqueouP medium with perchloric acid as the titrant. Hence, the amine dissolved in an ethylene glycol-isopropyl alcohol mixture wa,c
titrated x i t h perchloric acid in the same solvent with methyl red as the indicator, but an end point’ could not be detected. Titrations of N-phenyl-1-naphthylamine in glycol-chloroform mixtures ( 3 ) and in glacial acetic acid ( 2 )were also unsuccessful. .-ilthough 1-naphthylamine gave good end points in the latter titrations, N-phenyl-1-naphthylamine appeared to be too weak a hase to be titrated in the solvents tried. Possible colorimetric methods were studied nest,. One of the qualitative tests for aromatic amines, described by Burchfield and Judy (I), proved to be excellent for detecting N-phenyl-liiaphthylamine in oils. The amine reacted with diazotized p nitroaniline (p-nitrobenzene diazonium chloride) in a weakly acid acetone solution to form a bluish-purple diazoamino compound. ( ‘I
o.s \~ --,
+
IHC1
The extreme color seiiaitivith-, t,he i,agidity, and the relLttive hiplicity wit,h lvhich the colored system formed made it a good reaction on which to try to base a quantitative reaction. The greatest stumbling block to its use was the stability of the color formed, which intensified rapidly with time. However, this difficulty has been overcome. The color was developed .so that its intensity increased for about 10 minutes and then
ANALYTICAL CHEMISTRY
1020 very slowly decreased with time. Under the conditions described, the color intensity proved to be constant for a t least 20 minutes after the maximum color intensity was reached. The method finally developed was rapid, requiring only 18 to 20 minutes for a single determination. A nontechnical man can easily analyze nine samples in 2.5 hours. The method has an accuracy of about &IO%, which is considered satisfactory for blending plant control work on new oils and for used oil analysis. While the method gives good results on used oils in normal service, high results may be obtained with very badly oxidized oils.
Colorimeter. Any spectrophotometer or photoelectric colorimeter may be used. If a spectrophotometer is used, set the wavelength dial to 555 mp. If a colorimeter, use a green filter having maximum transmittance around this wave lengt,h. Colorimeter Cells. Matched test tubes with a diameter of 10 t o 20 mm. are suitable.
50-ml. volumetric flask. Add 30 ml. of acetone and shake well to dissolve the sample. (Some oil samples may not dissolve completely in the acetone. This causes no trouble, since the undissolved oil quickly settles in droplets to the bottom of the flask, leaving a clear solution with all the amine in the acetone.) Complete the color development and the measurement of the color intensity, described for the calibration curve preparation. Oil Blanks. The transmittance of a sample due to the color imparted to the acetone solution by the oil stock or other additives present varies somewhat with the type of oil blend. Hence, the oil blank may be different for each type of oil to be analyzed for N-phenyl-1-naphthylamine. When analyzing new oils, prepare, if possible, 100 ml. of the same oil without N-phenyl-I-naphthylamine. Then, take 0.1 gram of this oil blank through the above procedure as if it were a regular sample. Once the blank transmittance is found for a given blend, it need not be rechecked unless the stock oils used in making the oil blends change. If a used oil of known type is not too badly oxidized, its new oil blank mag be used to correct the transmittance reading of the sample. If a used oil of unknown type, or a badly oxidized known oil is tested for the additive, the oil blank is prepared by dissolving 0.1 gram of the oil in acetone and diluting to 50 ml.
REAGENTS
CALCULATIONS AND REPORT
@-Nitroaniline Solution. Dissolve 1.0 gram of the reagent in 50 ml. of (1to 1)hydrochloric acid and dilute to 100 ml. with distilled water. Sodium Nitrite Solution. Dissolve 2.5 grams of the reagent in 100 ml. of distilled water. Diazotized &Nitroaniline Reagent. Mix 5 ml. of the p-nitroaniline solution and 1 ml. of the sodium nitrite solution. Prepare the mixture fresh each day a t least 15 minutes before use. Acetone. C.P. and technical grades are suitable. Standard Stock N-Phenyl-1-naphthylamine Solution. Keigh 100 1 mg. of the reagent into a 150-ml. beaker, and dissolve in about 25 ml. of acetone. Transfer the solution to a 100-ml. volumetric flask, rinsing the beaker with several small portions of acetone. Fill the flask to the mark with acetone and mix well. Dilute Standard N-Phenyl-1-naphthylamine Solution. Pipet 1 ml. of the standard stock solution into a 100-ml. volumetric flask, dilute to the mark with acetone, and mix. Each milliliter of this solution contains 10 y of N-phenyl-1-naphthylamine.
Correct the per cent transmittance or absorbance of the sample for the oil blank in the following manner:
APPARATUS
*
PREPARATION OF CALIBRATION CURVE
Transfer 0, 2, 4,5, 7 , 10, and 15 ml. of the dilute standard solution to 50-ml. volumetric flasks, add about 30 ml. of acetone to each, and swirl to mix. Using a I-ml. graduated pipet, add carefully 0.2 ml. of the diazotized p-nitroaniline reagent. Shake well for about 30 seconds, dilute to the mark with acetone, and mix. Allow the solutions to stand at leafit 10 minutes to attain msximum color development. Partially fill a test-tube cell with the solution to which no Ai-phenyl-]-naphthylamine was added. This will serve as the reference solution. With the cell in position, adjust the colorimeter to 100% transmittance (zero absorbance). Partly fill a matching test tube with one of the standards and measure its transmittance (or absorbance). Take three readings of the transmittance, checking and, if necessary, readjusting the 100% transmittance setting of the colorimeter with the reference solution. Record the average of the three readings and the corresponding per cent of N-phenyl-I-naphthylamine in the standard. Since the method requires a 0.1-gram sample for analysis, the standards are equivalent in amine content to oil samples containing 0.02 0.04, 0.05, 0.07, 0.10, and 0.15% of N-phenyl-lnaphthvlamke, respectively. Repeat these steps for each of the standaFd solutions.. Plot the average transmittance (or absorbance) readings aga.inst their corresponding percentages of N-phenyl-l-naphthylamine. Use semilog paper if plotting transmittance readings, and linear coordinate paper if plotting absorbance readings. Draw the best straight line between these points and the origin of the paper.
Corrected per cent transmittance reading =
where A is per cent transmittance of the sample and B is per cent transmittance of the oil blank. Corrected absorbance reading = C D where C is absorbance of the sample and D is absorbance of the oil blank. Using the corrected sample transmittance or absorbance, read the per cent of N-phenyl-I-naphthylamine in the sample from the calibration curve. Report results to the nearest 0.01% of N-phenyl-l-naphthylamine.
-
ABSORBANCE CHARACTERISTICS OF COLORED SYSTEM
A curve of wave length us. absorbance for the purple color formed by the reaction between Ai-phenyl-1-naphthylamine and the diazotized p-nitroaniline is shown in Figure 1. The data were obtained with a Beckman DU spectrophotometer using a sample containing 100 y of reagent. The color formed shows maximum absorbance a t 555 mp.
0400
0.300
W
\
PROCEDURE
Qualitative Test. Put 2 to 3 drops of the oil sample in a Coorn
KO.0 porcelain crucible, add 10 ml. of acetone, and stir the mix-
ture for 30 seconds. Add a drop of the diazotized p-nitroaniline solution and stir. A bluish-purple color that increases rapidly in intensity shows that N-phenyl-1-naphthylamine is present in the oil. Quantitative Method. Weigh 100 3 mg. of sample into a
A B
I
500 6 00 WAVE LENGTH Mp
I 2
700
Figure 1. Wave length-absorbance curve for colored system
V O L U M E 27, N O . 6, J U N E 1 9 5 5 Table I.
1021
Stability of Color Developed with Oil
(Containing O.lO??o S-phenyl-1-naphthylamine)
%
Time, Minutes 3
Transmittance 54.0 50.0 48 2 46.5 44.9 43.3 42.6 42.4 41.8 41.8 41.5 41.6 41 9 41.8
tla 7 9 10 12 14 16 18 20 25 30 42
EFFECT O F SOURCE O F "V-PHENYL-I-N4PHTHYLAMINE
Table 11. Effect of Age of Diazotized p-Nitroaniline on Color Intensity and Stability Time, Minutes
0.25 50 2
9 11 13 15 20 25 30 35 40
48 44 43 42 42 41 41 41 40 40 40
7 6
3 3 0 5 1
2 2
5 J
Age of Reagent, Hours 2 Yo Transmittance
18
48.7 45.7 43 6 43.2 42.7 42 5 42.2 42 3 42.5
44.8 40.8 38.7 37.5 37.1 37.0 36.8 37.1 37.2 37.2
..
.. ..
n-ith two drops of reagent than with one drop. The use of a third drop of reagent onlv slightly increases the color compared to that developed by 2 drops. The results of these tests indicate that the best balance of time, color intensity, and color stability, would be attained by using 0.2 ml. of reagent prepared at least 0.25 hour before use. Samples must stand a t least 10 minutes after the reagent is added before color readings are made.
-
Different batches of the reagent from a number of sources were noted to vary in color and perhaps in purity. A study was made of the possible error that would result if the darkcolored form was used to prepare the calibration curve and the light-colored form to make oil blends. Therefore, solutions of 0.1% of N-phenyl-1-naphthylamine in an oil blend were prepared using reagent from four different sources. Another sample contained only 0.04% of N-phenyl-1-naphthylamine. On nnalysis, these samples gave the results shown in Table 11'. The results show little variation in amine content accoiding to source of the reagent. The differences are not significant. Thus, N-phenyl-1-naphthylamine from one source can be used to prepare a calibration curve and N-phenyl-1-naphthylamine from another source to make an oil blend.
..
ANALYSIS O F N E W AYD USED OILS
..
The colored system formed follows the Beer-Bouguer law very well over the concentration range of 0 to 0.15% of N phenyl-1-naphthylamine. No attempt was made to determine the amine concentration range for which the Beer-Bouguer law would still hold. When samples contained more than 0.15% of additive, the sample size xl-as decreased. STABILITY O F COLOR
The stability of the color formed was first checked on an oil containing 0.10% of N-phenyl-1-naphthylamine. The method given \vas used to develop the color, except that only one drop of the color-forming reagent was added. The transmittance of this solution was measured a t regular time intervals starting with the timeof reagent addition. The results are summarized in Table
I. The color increases in intensity-decreased transmittancefor about 15 minutes and becomes relatively constant for the next 0.5 hour. The stability was much better than had been originally expected from qualitative work. In some preliminary tests, the color sometimes stabilized in less than 15 minutes and was more intense. Therefore, the effect of reagent age and volume on the colored system was studied. Tests were run in which the reagent stood 0.25, 2, and 18 hours, respectively, before use. The effect of reagent age on color intensity and stability is shown in Table 11. One drop of reagent was added to each of these samples. The sample containing the 0.25-hour reagent required 30 minutes to reach maximum color development, while that developed with the 18-hour reagent stabilized in 10 minutes. Therefore, the longer the reagent aged, the quicher the color became stable. Also the color intensified with the age of the reagent. Two sets of samples were prepared using 2 and 3 drops of the diazotized p-nitroaniline reagent. Colors in one set were developed a i t h the 0.25-hour reagent, and in the other with the 18-hour reagent. Table I11 shows how the transmittance of these samples varied with time. Thus, nith 2 drops of the 18-hour reagent, the color becomes stable within 5 minutes. With the 0.25-hour reagent it stabilizes in ahout 10 minutes. In both cases, the color intensity is stronger
Using the described method, a number of new and used oils were analyzed for their ,V-phenyl-1-naphthylaminecontent. In Table V the results are compared with those found by t,he previous gravimetric method ( 4 ) . The oils used in t'his tabulation were selected a t random from :t group of samples of different formulations and used for different purposes. They were largely indust'rial oils. In a few cases results by the t'wo methods do not' check well. However, values by the colorimetric method show no t,endency to be higher or l o m r than by the gravimetric method. When rechecks were run, duplicates by the colorimetric procedure have shown the better repeatability and reproducibility. Besides being much faster, the colorimetric procedure is believed to be more precise than the gravimetric one. APPLICABILITY O F X E T H O D TO OILS CONTAINING N-PHEYY L-2-NAPHTHY LAMIIVE
Some work has been done to apply this method to oils containing .V-phenyl-2-naphthylamine. I t , too, imparts antioxidation properties to an oil blend. It reacts similarly to N-phenyl-l-
Table 111. Effect of Reagent Volume on Color Intensity and Stability Time, Minutes 3 5 7 9 11 13 15 20 25 30 35
0 25-Hour Reagent One Two drop drops 50 2 38 3 35 6 48 7 44 6 35 6 43 3 35 7 42 3 35 7 42 0 35 5 41 5 35 2 41 1 34 7 41 2 34 8 40 2 35 1 40 i
18-Hour Reagent Two Three drops drops 35.6 35.5 33 3 36.1 33 1 35.3 33 2 35.4 33 4 35.7 33 4 35.9 33 4 36.5 34 1 36.6 33 2 33 6 ~
One drop 44.8 40.8 38.7 37.5 37.1 37.0 36.8 37.1 37.2 37 2
Tahle IV. Analysis of Sample Prepared with Different Batches of .V-Phenyl-1-naphthylamine
_ _ _ _ N-Phenyl-1 _ ~ -naphthylamine
Source A A B
c
D
% added
% found
0.10 0.04 0.10 0.10 0.10
0.10 0.04 0.11 0.11 0.11
1022
ANALYTICAL CHEMISTRY
Table V. Comparison of Results on New and Used Oils b>Colorimetric and Gravimetric Procedures
5 .I--Phenyl-1-naphthylamine SRln),le
.iddcd
Found by Nethod C'olorinietric C;ravinietric
\yere obtained with a Beckman DE spectrophotometer using 10-mm. cell?. However, separate calibration curves for both .\~-ph~n~-l-2-iiaphthylamineand S-phenyl-1-naphthylamiiie stnnt1:irtii :ire necessary. INTERFERENCES
S E R - OILS
A B
0 10 10 10 10 10 10
0 0 0 0 0
C
D
E l-
0 10 0 09 0 09
0 0i 0 00 0 10 0 12 0 09
0 OH
0 10 0 0') 0 00
0 09 0 10
L-SED Orre 0 10 0 09' 0 07
1
2
0 Oi 0 13
3
0 0 0 0
4
5
0 0 0 0 0 0
0
13 12
09 09 09 10" 09 OH
10
09 0 09" 0 10 0 11 0 007" 0 07 0 09
11 12
0 04 0 04
7
8 9
a
10
0 07 0 I1 0
o
17 n7
0 010
o
This method is not specific for S-phenyl-1-naphthylamine. The basic react,ion involved in the met>hodis a general one for :ironxitic amines. If any other aromatic amine besides 1l'-phenj.11-naphthylamine i9 present, it ma>-interfere. However, aromatic amines are not likely to be present escept as additives. Therefore. errors due t,o the piesence of other arcmatic amines are not likelx. to occur in knoTi-n oil blends.
in
0 l i 0 03 0 I?
0 08 SI1
0 08 0 01) 0 08 0 03
Results obtained bj- nontechnical man trying method for firit tiiiie.
naphthylamine hut the color foimed is reddish purple instead of bluish purple. With a little experience, it is easy to distinguish between them. The color formed with N-phenyl-2-naphth\ 1amine shows mavimuni absorbance of light around 540 nip instead of 555 mp for .$--phenyl-1-naphthylamine. Of interest too is the fact that 0.1% solutions of both amines have about the mnie absorbance, 0.192 xnd 0.198, respectivelv. The valutb.
SUMMARY
This paper describes a rapid method for determining t>he A*-phenyl-1-naphthylamine content of new and wed industrial oib. A single determination can be run by a nontechnical man i n 18 to 20 minuter. The prerision of the met,hod is about =t10%. which is adequate for blending plant, control n-or1.r and routine analvsis of used oils. The iiiethcd is applicable even n-hen the amine comes from different producers and is purchased a t different times. The method is not specific for X-phenyl-1-naphthylamine. Other aromatic amines, if present, may interfere, but are not likely to be present in known oil blends. LITER4TURE CITED
H. P.. a n d Judy. J. S . . AI,. CHEM.,19, 786 (1947) (2) Fritz, J. S..Zbid..22, 1028 (1950). (3) P a i i t , S. R., IND.ENG.CHEM.,AN.AI..ED., 18, 246 (1916). (4) Petrus. P., private communication. (1) Rurchfield.
RECEIVED for review May 2 0 . 1954. .iccepted December 4, 1954. Presented hefore the Anieriran Petroleunr Institute. Houston, Tex., 1954.
Polarographic Determination of lead in Beryllium R.
W. BANE
Chemistry Division, Argonne N a t i o n a l Laboratory, Lemont,
4 rapid polarographic niethod for the determination of small amounts of lead in beryllium metal is described. Yo separations are necessary. A hydrochloric acid solution of beryllium acts as the supporting electrolyte. In the range of 150 to 800 p.p.m. of lead the precision is 10% or better. The simple procedure is readily adaptable for routine use. Evidence for the feasibility of the simultaneous determination of lead, cadmium, and zinc is presented.
I
CONKECTIOS with the analysis of beryllium metal for impurities or alloying constituents it was necessary to investigate methods for the determination of lead in beryllium. The classical colorimetric method using dithizone was deinonstrated to be successful for the determination of microgram quautities of lead in relatively pure beryllium metal. However, when the lead and the impurit.ie.3 were present in the order of several hundred parts per million, t,he colorimetric determination p r o v d t o be tedious and not feasible for routine analysis. Even though the procedure can he simplified by a, sulfate precipitation of the lead using strontium as a carrier prior to the dithizone estrartion (S), i t is still time-consuming and requires considerable manipulation. An electrolytic-polarographic method has been I\;
111. used to determine microgram amounts of lead and other metali in elements, such as barium, beryllium, calcium, magnesium, sodium, and uranium, which are not deposited in the mercury cathode ( 2 , 6 ) . This separation-concentration technique is particularly useful when the element desired is below the polarographic range and solubility limits of the matrix material prohibit adequate concentration. The lead is separated by electrolysis with a mercury cathode and then isolated by volatilizatioii of the mercury, prior to the polarographic determination. The limit of detection has been reported as probably 0.002 mg. of lead (f ). However, t,his electrolytic-distillation separation is not recommended for lead in the milligram range (8). -4lthough this technique has been used successfully in this laboratory for the determination of trace impurities in uranium, it, does not lend itself to rapid or routine analysis since the electrolysis time is 10 hours or longer. Esperiments have now indicated that, lead can be determined directly I-, polarography in the presence of large amounts of beryllium, and a procedure has been developed for the rapid determination of lead in the range of 150 to 800 p.p.m. Advantage is taken of the speed and simplicity of polarographic analysis. S o cheniical or electrochemical separations are required. The procedurr is suitable for routine analysis. Although only the
