the sample size. Dilution quenching is the reduction of light emission due to dilution of the scintillation phase with saniple. Sinc,e most of the materials which would partition in the aqueous ]]has(>of these eniulsions have very poor energy transfer properties, it is expected that very few or none of the sample molcculw, which are excited by primary interaction with p particles, will transfer their escitation energy to phosphor moleculcs. Thus, light emission will be reduced in proportion to the decrease in 1)otentially productive primary excitations. The limitations imposed by thcw quenching phenomena of course become greater as isotopes with weaker p's are count,ed,but in the case where the avei'age path lrngth of the p particle is not large compared to the droplet size of the dispersed sanii)lr, self absorption becomes a further limiting factor. Under these conditions, a dispro. large fraction of the energy lost in the droplet in which they originate. This effect can be niinimized by adequate dispersion of samlile and scintillation phases, which is morr easily achieved by preparation of chemical1~-stabilized emulsions of liquid phases, than by other heterogeneous counting techniques. Even under ideal conditionb of sample dispersion and chemical quenching elimination, these phenomena will ultimately limit the proportion of aqueous phase that can be counted with reasonable efficirncy. As espected, our results show that the counting efficiency of H 3 is reduced to a
much greater extent by dilution than i!: that of CI4, since a much greater proportion of the H3 D's produces barely detectable numbers of photons in undiluted scintillation phase. But, in spite of this reductmion.tritium can be counted with reasonable efficiency (10i27c) in mixes containing a much higher percentage of aqueous phase than is possible with any previously described syPtem. Carbon-14 and other isotopes which emit more energetic p particles, can be counted with much higher efficiency in mises which contain even greater percentages of aqueous phase. Though no observations have been made, it is assumed that the emulsions described in this paper are of the water in oil type, since the organic phase is always present in greater quantity than t,he aqueous phase and since the viscosity of each emulsion increases n-it'h the amount of water. Recause of such increases in viscosity, it is not practical t o increase the water content any higher than described here, while using toluene, Triton X-100 water emulsions. Only a small number of emulsifiers were tested before Triton X-100 was found to have desirable counting properties and the search was stopped. h systematic evaluation of the many commercially available emulsifie hould yield several systems which are scintillation counting, including oil water emulsions. Emulsions of this type would allow the susp relatively small amounts of sc phase in aqueous solution, thus further
increasing the quantity of water which can be counted in a standard vial. As discussed above, the results which we have obtained suggest that, with presently used scintillators and counting equipment, it is not pract,ical to increase the aqueous phase for tritium counting, but some increase in water content is apparently feasible for counting kot.opes with more energetic betas. Since light emission by the scintillators, and photon detection by the photomultipliers, are both relatively inof theoretical), an efficient (-10% increase in the ratio of aqueous phase to scintillation phase should be possible if the efficiency of either or both of these processes \wre significantly improved. Thus, under theoretically ideal conditions, emulsion counting should be preferable to most other methods of counting aqueous solutions because it allows the incor1)oration of a high percentage of aqueous phase with reasonably high counting efficiency, and, except for colored materials, the efficien relatively unaffected by quenche the aqueous phase. LITERATURE CITED
(, 1,) IIeade. R. C.. Ht,iditz R.. Intern. J . A p p l . Rddintion'lsot&es' 13,' 11 (1962). (2) Rapkin, E., l b i d . , 1 5 , 69 (1964). ( 3 ) Schram, E.) "?rganic* Scintillation
Detectors," Elsevier, Amsterdam-London-Sew Tork, 1063. RECEIVED for review October 9, 1964. Accepted April 16, 1965. This work war slipported in part by Grant Yo. (>SI 10317 from the Tationd Institutes of Health .
Gas Analysis by Geiger Pulse Attenuation FREDERICK W . WILLIAMS and ARTHUR F. FlNDElS School o f Chemistry, University o f Alabama, University, Ala.
b Nucleonics counting gases possess electronic characteristics which are markedly altered by the admixture of additional components. Alteration of the composition of the common counting gas, 1.3% butane-98.7% helium, results in a change in gas amplification. The presence of all gases, except argon, investigated in this work resulted in a decrease in the pulse amplitude when the counter is operated in the Geiger region of gas amplification. The decrease in pulse amplitude is a linear function of the amount of gas injected in the stream for many gases over wide ranges of sample injections. In most cases the pulse amplitude variations are additive. Linear calibration curves were obtained for many binary mixtures. In those cases where one of the components had a high electron attach-
ment probability, nonlinear but reproducible data were obtained. The gases investigated were ammonia, nitrogen, oxygen, argon, ethylene oxide, carbon dioxide, hydrogen, sulfur dioxide, low molecular weight hydrocarbons, and binary mixtures of the above.
T
are many ways of detecting gases, one being by ionization of a gas. .i gas may be detected if ionized by bombardment with high-speed charged particles, y- or x-rays, ultraviolet photoionization, or by direct application of a sufficiently high electric potential. Deal et al. (I) descr ibea detector which depends on the differences in the ionization cross section of different molecular species for its action. .1 10-pc. source of stronHERE
tium-90-ytt'rium-90 was used in this detector. Harley and Pretorius (2) designed a glow discharge detector which will detect mole per second of a gas. A sensit,ive ionization gauge detector developed by Ryce and 13ryce ( 8 ) makes use of the fact that the ionization potential of the helium carrier gas (24.5 volts) is very much greater than that of most other volatile substances. A sensitive detector whose method of operation is based on the uniqudonization properties of ai'yon was developed by Lovelock (4). Lovelock and Lilisky ( 5 ) developed an electron absoiytion detector. -4 stream of inert gas is passed through the chamber and the i~otentialis adjusted so as to collect all of the electrons liberated from the gas hy ionizing radiation. -\ new, simple, and highly VOL. 37, N O . 7, JUNE 1965
a
857
+-I,.
This work was initiated to evaluate the effects of impurity gases on the gas amplification of the commercial counting gas, 1.3% butane-98.7'y0 helium. The reproducibility of pulse attenuations suggests its application to gas analysis.
0 3 l l U CONNECTOR
EXPERIMENTAL
,
B A I L AND ENDS, LUCITE
l Y l L T\;STEN
CENTER WIRE
,
4
U
+; I+
Figure 1 .
Foil flow counter
sensitive detector system for analysis of both gases and organic vapors has been developed by Shahin and Lipsky (9). The limit of detection approaches mole per second. The Montgomery and hlontgomery curve (6) typifies the gas amplification regions over which the previous detectors operate. A chamber consisting of a cathode and anode may operate as an ionization chamber, or proportional, limited proportional, or Geiger counter, depending on the gas amplification and the geometry of the chamber. The detector discussed in this work is a Geiger counter. The starting potential of a Geiger-Muller counter has been measured by Weisz (11) for mixtures of argon with hydrocarbons and helium with hydrocarbons. The results were interpreted in light of the known ionization potentials and the energy levels of the gases in the mixtures. The start of the Geiger region is defined by Korff (3) as the voltage at which pulse size equalization sets in. Rochester and McCusker ( 7 ) demonstrated that the discharge in a Geiger-Muller counter is quenched by three processes: the formation of a positive ion sheath which so reduces the field in the neighborhood of the center wire that further ionization by collision could not take place; the prevention of emission of photoelect'rons from the cathode due to rapid absorption of the photons by the polyatomic vapor followed by the dissociation of the vapor; and the predissociation of the organic molecule preventing t'he emission of secondary electrons from the cathode by the positive ion bombardment. dccording to Rochester and McCusker the role of polyatomic vapor in the Geiger tube suggests that increasing the pressure of the vapor should improve the performance of the counter, since the efficiency of each quenching process would seem to be increased by increasing the quantity of vapor. 858
ANALYTICAL CHEMISTRY
Apparatus and Reagents. A 541-A Tektronix cathode-ray oscilloscope was used. T h e high voltage supply for the Geiger counter was supplied by Hamner Electronics Co., Inc., 500- to 5000-volt output, Model N-4050. T h e linear amplifier, 5-37'1, decade scaler, 5-240, and electronic timer, N-804, were obtained from Hamner Electronics Co., Inc. A gas proportioner, No. 665, was obtained from T h e Matheson Co., Inc., and used to mix the gases. The Geiger counter was constructed in this laboratory using much the same design as t h a t of Sugihara, Wolfgang, and Libby (10). Figure 1 shows the counter used in this study. d block diagram of the apparatus is shown in Figure 2. All of the gases that were studied-ammonia, oxygen, nitrogen, carbon dioxide, argon, methane, ethylene, cyclopropane, acetylene, neopentane, propylene, ethylene oxide, ethane, propane, butane, hydrogen, and sulfur dioxide-were c. P . grade, 99.070 minimum purity from The Matheson Co., Inc. The flow gas for the Geiger counter was the mixture, 1.3y0 butane98.7y0 helium, obtained from The Matheson Co., Inc. Any radioactive source yielding a bright oscilloscope trace may be used. I n this work, a 0.1-gc. source of RaZz6was used. Procedure. Pulse attenuation data were obtained in the following manner. T h e sample gases were delivered from high pressure cylinders through a two-stage regulator into a calibrated flowmeter connected in series with the gas sampling tube. T h e samples of gas were taken from t h e glass sampling tube through a rubber septum sealed in the glass. T h e sampling was done with a hypodermic syringe. The gases were introduced through another rubber septum directly into the flow stream of the nucleonics gas supplying the counter. A %foot section of '/r-inch copper tubing was placed between the injection port and
0.04 0.08 M L . O F GAS
OD4 ML.
Figure 3.
0.08
0.12
0.16
INJECTED
0.12
0.16
OF G A S INJECTED
Pulse attenuation curve s
the counter. The flow of the gas for the Geiger counter was regulated with a flowmeter at 76.9 ml. per minute. Composition of binary mixtures was controlled by the use of the gas proportioner. Samples were taken as described above. The pulse attenuation was observed on the oscilloscope as the samples passed through the counter. The sweep time of the oscilloscope was 5 microseconds and the pulse amplitude was adjusted to 80 volts with the linear amplifier when pure counting gas was flowing through the counter. This value was arbitrarily chosen for the measurements in order to utilize the full scale of the oscilloscope screen at minimum sensit'ivity. The voltage applied to the Geiger counter during the measurements was 1200 volts, d . c., a value about, halfway on the plateau. RESULTS
INJECTION
Figure 2. Block diagram of apparatus
Each gas that was studied showed unique properties in the Geiger counter. The pulse attenuation, as monitored on the oscilloscope, was characteristic of each gas that was introduced into the flow stream of the Geiger counter as exemplified in Figure 3. Xot only do the gases show characteristic pulse attenuations, but a linear relation over wide rmges of sample size esists for the pulse attenuation when plotted against the volume of the foreign gas introduced into the flow gas.
01 02 03 04 ML OF G A S INJECTED
Figure 4. Pluse attenuation for oxygen at selected voltages on counter plateau
As expected, the pulse amplitude decreases with a decrease in the potential applied to the counter, but the actual pulse attenuation remains the same for a given volume of foreign gas introduced into the Geiger counter, as shown in Figure 4 for oxygen. These data were taken directly from the oscilloscopic display without amplification of the pulse from the counter. The counter
was connected to the high voltage supply through a 1-megohm load resistor and the pulse coupled to the oscilloscope directly by means of a 0.001-Ffd. capacitor. The amount of gas passing through the counter can be arrived a t by the use of a scaler in conjunction with a sensitive discriminator. If the discriminator is set to reject pulses a t a n amplitude less than the amplified 80-volt pulses obtained from the detector, the count will cease as the sample passes through the counter if the pulse amplitude becomes less than the discriminator setting. The duration of the loss of counts is a function of the amount of gas injected in the stream of the counting gas at a given discriminator setting. The counter was exposed to a constant source of radioactivity and the counts were recorded for a time interval which would cover the arrival and subsequent passage of the sample gas through the counter. Twenty seconds were sufficient for this time interval and was a convenient preset time interval on the scaler. Samples were injected and the loss in counts for the 20-second interval can be related to the amount of gas injected in
ML. OF G A S INJECTED
ML. OF
Figure 6. A.
8. C.
Nitrogen-oxygen mixtures Nitrogen-methone mixtures Argon-nitrogen mixtures
D. E. F.
GAS
-
3 z
0.2 0.4 0.6 ML. OF O2 INJECTED
0.8
Figure 5. Discriminator study of count losses in oxygen injection
the counting gas stream. If the sample profile becomes Gaussian in shape as it traverses the tubing prior to the arrival a t the counter, then the loss in counts will be reflected in a Gaussianshaped calibration curve. T h a t this is so is exhibited by typical data plotted in Figure 5 . In this particular experiment, the discriminator was set to reject all pulses of less than 78-volt amplitude. These data were not corrected for coincidence loss and were fitted to the curve by use of normal probability graph paper.
M L . O F G A S INJECTED
INJECTED
Pulse attenuation curves
Methane-ethylene mixtures Nitrogen-ethylene mixtures Nitrogen-cyclopropane mixtures
G. H. 1.
Cyclopropane-ethylene mixtures Carbon dioxide-ethylene oxide mixtures Oxygen-neopentane mixtures
VOL. 37, NO. 7, JUNE 1965
859
The general shape of the curve is a function of the discriminator setting. If the discriminator was set just below the amplitude of the amplified pulses from the counter, any injection of a gas resulting in an attenuation of the pulse would cause a loss in counts. This was the case for the data plott'ed in Figure 5; thus the abscissa starts a t 0 ml. If the discriminator was set a t some lower value, no counts would be lost until the amount of gas injected was sufficient to result in a pulse which was attenuated to a value less than the discriminator setting. Such amounts are readily obtained from the pulse attenuation curves shdwn in Figure 3. I n such a case, there is no loss in counts until the pulse amplitude drops below the discriminator setting. The plot of t'he counts per 20 seconds is constant until the volume of gas injected exceeds the corresponding volume threshold of the discriminator. The working part of the curve, however, does not start a t 0 ml. injected, but is of the same Gaussian shape as indicated in Figure 5 if one plots the milliliters injected only from the point at which counts were lost. The principal change is in the value of the abscissa. This point is readily obtained from a plot on probability paper. The statistics of the measurement of radioactivity do not make this approach favorable for precise measurements of small injections of gases. When binary mixtures of the gases were introduced into the Geiger counter, the pulse attenuation in most cases was an additive function of the percentages of the gases. The pulse attenuation for a 50% mixture of oxygen and nitrogen was halfway between the attenuation for the pure gases, oxygen and nitrogen, as shown in Figure 6. This fact held true for most of the conceivable binary mixtures of the gases that were studied in the pure state. Data for selected mixtures are plotted in Figure 6. The nitrogen-oxygen curves (Figure 6, A ) were obtained with the amplifier set to 40 volts rather than 80 volts. In some of the mixt,ures the attenuation curves of the pure gases were too close for precise differentiation of mixtures. This is true in the cases of ammonia and cyclopropane, ammonia and methane, ammonia and carbon dioxide, propylene and ethylene, and a few others. The binary gas mixtures which cannot be analyzed can be predicted from the pulse attenuation curves of the pure gases. Since the attenuation curves are reproducible, linearity is not a requirement for precise calibration and measurement. The sensitivity of the method depends on the degree of complexity of the molecule and the tendency for negative ion formation. Of the gases investigated in this work, sulfur dioxide 860
ANALYTICAL CHEMISTRY
was found to give the greatest pulse attenuation per milliliter injected into the stream. If one assumes that a pulse attenuation of 1 volt is detectable, this corresponds to an injection of 0.001 ml. of sulfur dioxide or 4.1 X lop7 mole. The pulse attenuation is very little affected by the distance the sample of gas travels in the 1/4-inch copper tubing before it reaches the counter. Up to 20 feet of 1/4-inch tubing has no effect. After 20 feet of tubing there was a slight spreading of the gas sample, which is shown by the diminished and slower rate of pulse attenuation. The comparison of the inherent properties of the Geiger counter before and after the termination of these studies showed very little effect on the starting potential and plateau of the Geiger counter. DISCUSSION
The factors which influence the pulse amplitude when a foreign gas is introduced into a Geiger counter are dilution, negative ion formation, increase in t'he number of degrees of freedom-Le., change in entropy-and change in the metastable state population. Positive ion formation is the favored for most gases. The pulse attenuation for a gas mixed with the commercial counting gas, 1.3% butane-98.7% helium, is a function of the complexity of the molecule. The order of pulse attenuation observed in this study was: argon < hydrogen < oxygen < nitrogen < carbon dioxide < ammonia < cyclopropane < ethylene oxide < acetylene < propylene < methane < ethane < ethylene < propane or neopentane < butane < sulfur dioxide. The increase in the number of degrees of freedom-that is, the increase in vibrational, rotational, etc., energy levels with increasing complexit,y of the molecule-accounts for the pulse attenuation when a homologous series of hydrocarbons is considered. The normal hydrocarbons through butane show the effect of excitation rather than ionization being predominant. In bhe case of argon, the increase in the metastable state population results in an increase in pulse amplitude. Hydrogen in small amounts exhibits much the same characteristics as argon, but the effect is not as pronounced. Kitrogen, oxygen, and carbon dioxide show effects of dilution with a small contribution from negative ion formation. Sulfur dioxide exhibits a strong tendency for negative ion formation. The binary 50y0 curves show the additive effect of each gas. This is observed for most 50% mixtures; thus, the resulting curve is an average of the pure gases. Nitrogen and hydrogen do
not yield a mean pulse attenuation for 50% mixtures; the curve is closer to the pure nitrogen curve. The pulse attenuation of ammonia falls below the nitrogen curve. This indicates compound formation or a t least gives evidence of an ion-molecule or other reaction which may be interpreted in light of the synthesis of ammonia. The standard deviation of over 100 measurements of pulse attenuation was j ~ 0 . 5 7volt by direct visual observation of the oscilloscope trace. This value reflects the error in repetitive injection as well as the error in the observation of pulse attenuation. The standard error in the measurement of binary mixtures is 10.81 volt. Thus, the per cent standard error for binary mixtures depends on the difference in the pulse attenuations of the pure gases making up the binary mixture. I t is clear that the greater the separation in pulse attenuations for the pure gases, the greater the precision of measurement of the composition of binary mixtures of the gases. For example, when 0.2 ml. of a 50% mixture of nitrogen and methane (see Figure 6, B ) was injected into the counting gas, the pulse attenuation was 18.5 volts greater than that observed for a 0.2-ml. sample of pure nitrogen. The standard deviation in the pulse amplitude measurement resulted in a standard deviation of d ~ 4 . 4 7in~ the measurement of the composition of the mixture. If the difference in pulse attenuations of the pure gases comprising the binary mixtures were less, the error in the measurement would be increased. The identity of an unknown gas may be assigned on the basis of the pulse attenuation for a known sample injection, provided the gas is one for which pulse attenuation parameters are known. This is so if the pulse attenuations differ by a sufficient amount for each gas. If the pulse attenuations are identical, it is not possible to distinguish between the gases. The attenuations must differ by a t least 1 volt after amplification, in order to make a positive assignment. The composition of the counting gas, nominally 1.3y0butane98.7% helium, changes throughout the life of a cylinder. This causes a drift in the measurements of pulse attenuation, since the gas becomes enriched in butane. This is particularly a problem when only 10% of the gas remains in the cylinder. Frequent calibration checks are necessary as the cylinder becomes exhausted. LITERATURE CITED
( l ) T D e a l ,C. H., Otvos, J. W., Smith, V. h., Zucco, P. S., ANAL. CHEM. 28, 1958 (1956). (2) Harley, J., Pretorius, V., Nature 178, 1244 (1956).
(3) Korff, S. A., “Electron and Nuclear Counters,” p. 15, Van Nostrand, New York, 1946. (4) Lovelock, J. E., J . Chromatog. 1, 35 (1958). (5) Lovelock, J. E., Lipsky, S. R., J . Am. Chem. SOC.82, 431 (1960). (6) Montgomery, C. G., Montgomery, D. D., J . Franklin Inst. 231, 447 (1941).
(7) Rochester, G. D., McCusker, C. B. A., Nature 156,366 (1945). (8) Ryce, S. A., Bryce, W. A., Ibid., 179,
(11) Weisz, P. B., Phys. Rev. 74, 1807 (1948).
(9iShahin, h l . M., Lipsky, S. R., ANAL. CHEM.35, 467 (1963). (10). Sugihara, T. T., Wolfgang, R. L., Libby, W. F., Rev. Sci. Instr. 24, 511
RECEIVEDfor review June 1, 1964. Resubmitted February 25, 1965. Accepted A ril 12, 1965. Division of Analytical 8hemistry, 144th Meeting ACS, Los Angeles, Calif., April 1963.
,541 11957).
(1953).
Adsorption of Methyl Orange and Ethyl Orange on Tailored Silica Gels GEORGE H. REED’ and
L. 8. ROGERS
Chemistry Department, Purdue University, lafayetie, Ind.
b To explore factors that might influence the selectivity of adsorption, a study has been made of the effects of small changes in the structure of the azo dye coprecipitated with silica gel on the ability of that gel to adsorb methyl orange and ethyl orange. The capacity of the gel was always increased (40 to 350%) by the coprecipitation step, even when the selectivity was reversed or erased, despite the fact that a visible amount of coprecipitated dye could not b e removed by washing. Absence of a sulfonate group in the coprecipitated dye led to greater ease of removal b y washing. In two cases, selectivity was eliminated, a result that might b e useful in proving that two or more adsorbates had certain structural features in common.
S
is an intriguing and promising aspect of adsorption for which a number of applications seem imminent. Dickey ( 8 , 9 ) was the first to report that silica gels formed in the presence of methyl orange (MO) or ethyl orange (EO) showed an increased capacity for adsorption of that particular dye. More important, he found that the “natural” selectivity of silica gel for 310 relative to EO could be reversed by forming the silica gel in the presence of EO. Those results have been confirmed by others (11, 16). Dickey ( 8 ) has also shown that propyl orange can, in the same way, be selectively adsorbed relative to MO, EO, and butyl orange. I n addition, Bernhard (4) prepared a gel using a dye that had a p’-sulfonamido group in place of the sulfonic acid on the XI0 structure. His result seemed to indicate that there was little or no charge effect involved in the general phmomenon of specific adsorption. I n similar studies, Frlenmeyer and FECIFICITY
Present address, Chemistry Department, I-niversity of Wisconsin, Madison,
47907
Bartels (10) have observed, using thin layer chromatography (TLC) , specific adsorption for N,N-dimethyl- and N,Ndiethyl-anilines on their respective gels, but those gels did not differentiate between MO and EO in thin layer chromatograms and barely so in adsorption isotherms. However, Morrison et al. (21) reported enhanced sorption of MO and EO on a gel tailored with sulfanilic acid or p,p’ diaminodiphenyl. Going one step further, “stereoselective” adsorbents have been reported for compounds such as quinine, quinindine, cinchonine, and cinchonidine by Beckett and Anderson (2, 3) and for camphorsulfonic acid by Curti and coworkers (6, 7 ) . Other work closely related to specific adsorption has been done by Basmadjian and coworkers (1) concerning the effect of dissolved polymers on pore size and surface area of silica and alumina gels. The exact reason for the increase in adsorption capacity that can sometimes reverse the selectivity is still a matter of debate. Some workers favor the “imprint” theory (9, 15). Waksmundzki et al. ( 1 7 ) attribute the effect to a change in pore structure (diameter). Others believe that the portion of coprecipitated dye that remains in the gel after
exhaustive extraction is responsible for the phenomenon (11). Snyder (14) has proposed an important role for site type and topographical distribution. The present study is concerned chiefly with adsorption of -MO and EO on silica gels prepared in the presence of similar azo dyes. The effects of coprecipitated dyes having no sulfonate group or having methyl and ethyl groups in different positions were examined by means of adsorption isotherms and thin layer chromatograms. EXPERIMENTAL
Reagents. DYES. Except for azobenzene (Matheson Coleman & Bell) and methyl orange (Mallinckrodt), the organic reagents used to synthesize three of the dyes and the remaining dyes themselves were obtained from Eastman. Except where noted, chemicals were used as received. N , N - Dimethyl - p - phenylazorn-toluidine was prepared according to the literature (12). The melting points of 68” C. were identical. N - Ethyl - N - methyl - p - phenylazoaniline was prepared by coupling freshly distilled aniline and N-ethyl-N-methyl aniline, available as the hydrochloride (13); p - N , N - Diethylamino - K , N dimethyl - p - phenylazo-aniline was
Table I. Equilibrium Behavior of Gels toward Methyl Orange and Ethyl Orange’ -Adsorbate __
Tailored gel
I. Control 11. Methyl orange 111. Ethyl orange
MO
Ratio EO- MO/EO
7.0
IV. N ,A’-Dimethyl-p-phenylazo-aniline S‘. N,K-Dimethyl-p-phenylaso-m-toluidine . . _ . VI. Azobenzene S‘II. p‘-Al’,Al’-Diet hylamino-N,N-dimethyl-p-phenylazo-aniline VIII. .V-Et hyl-A’-methyl-p-phenylazo-aniline
14.8 24.5 15.0
11.0 13.5 22.0 20.0
1.50 1.59
0.82
1.93 1.73 1.04 1.05
a Concentration ratios (moles per Kg./moles per liter) taken from adsorption isotherms for methyl orange and for ethyl orange at an equilibrium solution concentration of 2 . 0 0 ~ M.
WE.
VOL. 37, NO. 7, JUNE 1965
861