observed in CHC13 and reaches the factor of 4 per carbon ‘I t om. The slight change of K D values ~ obtained a t 2’ and 25“ C. is indicative of either the small values of the heats of solvation of the chelates in the two widely different solvents or of their close similarity. ACKNOWLEDGMENT
The authors gratefu 11y acknowledge the financial assistance of the U. S. .ltomic Energy Commission in this work.
LITERATURE CITED
(1) Bjerrum,
J., Schwarzenbach, G., Sill& L. G., “Stabi1,i:y Constants of Metal Ion Complexes, Parts I and 11, Chemical Society, London, 1957. 12) ~, Dvrssen. D.. Division of Analvtical Che“mistry, ACS, Summer Symp&ium, Tucson, Ariz., 1963. (3) Dyrssen, D., Rec. Trav. Chim. 75, 753 (1953). 14) Dvrssen. D.. Svensk Kem. Tidskr. 64,313 (1952).‘ (5) Dyrssen, I]., Dyrssen, >I., Johansson, E., Acta Chem. Scand. 10, 341 (1956). (6) Fresco, J., Freiser, H., ANAL.CHEM., 36, 372 (1964). ( 7 ) Jankowski, S.,Freiser, H., Ibid., 33, 776 (1961).
(8) Johnston, W. D., Freiser, H., AnaZ. Chim. Acta 11,201 (1954). (9) Krishen, A., Freiser, H., ASAL. CHEM.31,923 (1959). (10) Lacroix, S., Anal. Chim. Acta 1, 260 11947). (11) Morrison, G. H., Freiser, H., “Solvent Extraction in ilnalytical Chemistry,” Wiley, New York, 1957. (12) Phillips, J. P., Elbenger, R. L., Merritt, L. L., J . Am. Chem. Sac. 71, 3986 (1946). (13) Schweitzer, G. K., Bramlitt, E. T., Anal. Chim. Acta 23, 419 (1960). (14) Schweitzer, G. K., Coe, G. R., Zbid., 24,311 (1961).
RECEIVEDfor review October 8, 1963. Accepted Sovember 26, 1963.
A n Autorniatic Method for Measuring Slopes of Rate Curves Applied to Quantitative Determination of Cystine HARRY 1. PARDUE Department o f Chemisfry, Pardue University, lafayette, Ind.
b A new method i s described for the measurement of reaction rates. In the method, the slope of the rate curve i s determined automatic:ally near zero reaction time. The measurement i s accomplished b y matching the slope of the output signal froin an integrator circuit with that of the signal from the chemical system. A servo i s used to compare the two signals and to vary the integrator input until the signals are changing a t the !same rate. The slope of the integrator output and therefore of the rate c:urve a t balance i s proportional to the integrator input which i s indicated b y the position of the servo pen or measured with a meter. The method has been successfully applied to the quantitative determination of cystisle a t the partsper-million range. The measurement i s made completely alJtomatically and an integral multiple of concentration i s read directly from a dial. The procedure consists siniply of injecting reactants into the reaction vessel and reading the numerical answer from a meter. Total measurement times are in the range of 30 secclnds with relative standard deviation of 1 to 2y0.
R
there has been an increasing interest in the automatic measurement of reaction rates with application to rapid chemical analyses. One general approach has been to measure the time required for the reaction of interest to prcduce or consume a predetermined amount of product or renctant (3-5). X second approach has been to measure the e,tent of reaction over a conbtant time nterval ( 1 ) . I n ECESTLT
each case the measurement step is simple and rapid and the quantities measured are easily related to the concentration of sought-for constituent. In this work, a new method has been developed for the measurement of reaction rates. The new method automatically measures the slope of the rate curve near zero reaction time. The measurement is accomplished by matching the slope of the rate curve with the output from an electronic integrator. -1 servomechanism compares the signals from the chemical system and integrator and adjusts the integrator input until the slopes of the two are equal. ;it balance the input to the integrator is proportional to the slope of the rate curve. The method has been developed for the quantitative determination of cy+ tine based on its catalysis of the reduction of iodine by azide. The rate of decrease in iodine concentration is detected potentiometrically ( 5 ) . Under controlled conditions, the response curve is linear with a slope proportional to cystine Concentration. L-nder the condition of balance described above, the integrator input is proportional to cystine concentration. Some of the important favorable characteristics demonstrated by the method are as follows: The measurement equipment is easily constructed from commercially available components. The measurement step is rapid usually requiring less than 20 seconds per sample. Rate data (slopes) can be presented continuously by strip chart recording or can be read numerically from a meter.
The operational features of the system are very simple. The measurement step consists simply of adding reagents and sample to the reaction vessel and reading a multiple of the cystine concentration from a meter or chart. The reproducibility of the method is good. Relative standard deviations are less than 2%. The method is sensitive. Cystine is determined a t concentrations down to 0.25 p.p.m. in a total volume of 2 ml. PRINCIPLES
OF THE METHOD
C e l l Response. Conditions for the potentiometric determination of cystine based on its catalysis of the reduction of iodine by azide have been established ( 5 ) . L-nder the prescribed conditions and for cystine concentrations below 2.5 p.p.m., the rate of decrease in iodine concentration can be represented by Equation 1
where k 1 is a rate constant depending upon solution conditions, [Iz],is the iodine concentration a t time t , and C is the cystine concentration. For the potentiometric detection of the rate of change in iodine concentration the time dependent portion of the cell voltage E , is given in Equation 2. Et
=
k In [IZIt
(2)
where k is a temperature dependent constant from the Nernst Equation. The rate of change of potential with time is then given by Equation 3. (3) VOL. 36, NO. 3, MARCH 1964
633
Substituting for the rate of change of iodine concentration from Equation 1 into Equation 3 gives
Equation 4 predicts linear response with a slope proportional to cystine concentration. Integrator Response. T h e integrator used in the conventional operational amplifier design (2). For the integrator with a n input signal e,, a n input resistor R , and a feedback capacitor C,, the rate of change of the output signal e, is given by Equation 5. de, dt
(5)
Combined Response. T h e condition of balance for the system is t h a t the slopes of the cell response curve and the integrator are equal.
readout of rate data is presented in the next section. CIRCUITDESCRIPTION.Basic units of the instrument are a Sargent SR potentiometric recorder (E. H. Sargent de, = Et and Co., Chicago, Ill.) and a Heathkit dt dt operational amplifier system (Model ELW 19, Heath Co., Benton Harbor, Combining Equations 4 and 5 and rehiich.). Minor modifications of the arranging gives recorder are required. Plug P1,which connects the recorder potentiometric input circuit to the chopper amplifier, is disconnected. Pins 1 and 2 of the male Amphenol plug are shorted Equation 7 gives the relationship together. This provides access to the between the cystine concentration and slidewire source at the red and black the integrator input signal. input jacks to the recorder. Resistors For cystine concentrations between R22and R23 in the input circuit are 5 and 25 p.p.m., the rate of change of shorted out so that the full 1.35 volts iodine concentration is approximately of the mercury cell are applied across proportional to the 1.24th power of the slidewire. This is large compared to the millivolt range drift encountered cystine concentration ( 5 ) . For this with the unstabilized operational amplirange the cystine concentration is given fier. by Equation 8. A schematic representation of the circuit is given in Figure 1. The slider of the zero adjust potentiometer of the recorder (red jack) is connected to ground. The slider of the servo conEXPERIMENTAL trolled potentiometer (black jack) is Instrumentation. CONCESTRATION connected to the input of a follower ( F ) (Amplifier 1 of the Heathkit system in CELL. T h e concentration cell used follower position). The follower isois that described earlier ( 5 ) . The reaclates the slidewire source from the low tion takes place in a test tube of about impedance of the inverter and meter. 4-ml. total capacity immersed in the The output from the follower is conelectrolyte of a reference electrode. An nected through an inverter ( I S V ) to asbestos fiber sealed in the bottom of an integrator ( I S T ) with an output the test tube provides electrical conslope of 100 mv. per second per volt tact between the sample and reference input. The integrator output is disolution. The potential difference bevided to provide slopes in the range of tween similar platinum electrodes im0.005 to 0.5 mv. per second. Tenfold mersed in these solutions is measured ranges of 0.005 to 0.05 and 0.05 to 0.5 to follow the course of the reaction. mv. per second are obtained by placing The sample electrode is rotated a t about 2 megohm and 200-kilohm (in paren2000 r.p.m. and therefore provides theses) resistors, respectively, in series efficient stirring in the sample compartwith a 1-kilohm potentiometer. ment. The cell is thermostated by The reference electrode (R) of the circulating water from a constant temconcentration cell is connected to a perature bath through the glass jacket voltage divider a t the output of the surrounding the cell. inverter. This arrangement provides Changes in the iodine concentration damping for the system as outlined in the sample compartment resulting below. The sample electrode (S) and from the chemical reaction result in a divided integrator output are connected change in cell voltage. The slope of the to pins 1 and 2, respectively, of a two cell response curve is determined by conductor Amphenol plug (80 l\l-male matching it with a signal from an cable plug, double contact type). The integrator. A circuit for automatically latter is plugged into the femal socket matching the slopes and providing direct 634
ANALYTICAL CHEMISTRY
Figure 1 . Circuit for automatic measurement of slopes of rate curves
of P I of the recorder which is connected to the chopper input of the servo amplifier. The servomotor (M) moves the potentiometric slidewire from right to left when pin 1 is negative with respect to pin 2 and from left to right when pin 1 is positive with respect to pin 2. The output from the follower, which controls the slope of the integrator output, is measured by a voltmeter (Heathkit VO31, Model 1131-1, Heath Co., Benton Harbor, Nich.), with a n ohmsper-volt rating of 20,000. Resistors R1,R2,and R3are linear potentiometers whose nominal values are shown in parentheses. R1 serves as a damping adjust to provide rapid response without overshoot or oscillation. R2and RBare coarse and fine sensitivity controls, respectively. R2 is used to select the fraction of the integrator output, which is compared with the cell voltage. RBis used to adjust the meter response so that its readout is a simple multiple of cystine concentrations. Switch Sw provides a means for discharging the integrator capacitor between runs. CIRCUIT OPERATION. The general operation of the system is as follows. The potentials a t pins 1 and 2 are changing continuously. K h e n the potential difference is zero the system is a t balance. A difference in slopes between the rate curve and integrator response for a short time results in a difference signal. The difference signal causes Jf to operate and move the slidewire. A portion of the slidewire signal is fed back through the divider at the output of the inverter to compensate for the difference which produced the change and to return the system to balance. I n addition, the slidewire signal causes the slope of the integrator output to change and correct for the difference in slopes which generated the off-balance signal. The detailed operation is described for the potentiometric measurement of the rate of the catalytic reduction of iodine bj- azide for which quantitative data are presented. The zero adjust is set so that with the servo slider a t its maximum displacement to the right, there is zero output. The conditions in the concentration cell are such that a t zero reaction time the sample electrode
is a few millivolts positive with respect to the reference electrode. This positive signal at pin 1 cau:,es the servomotor to move the slider to the right hand side of the slidewire so that there is zero output and the integrator output is zero. As the reaction proceeds, the potential at S decretses linearly ( 5 ) . Jl'hen the potential at S passes through zero, pin 1 becomes negative with respect to pin 2. This negative signal activates III to movch the slider from right to left producing a negative voltage at the input to t'le follower. The follower passes the sip nal unchanged to the inverter where its sign is changed. A fraction (R1/lOOK) of the positive signal a t the inverter output is applied between the cell and ground. The motor X continues t c operate until the positive signal developed across R1 exactly compensates for the difference existing between pins 1 and 2 so that balance is established. The positive signal a t the inverter output is integrated to give a neg,ttive ramp signal across the divider wiich is connected to pin 2. If the slope of the ramp signal matches t h a t of the cell response curve then the sydem remains at balance. However, if the slopes are not equal, after a brief period a n error signal is generated be1 ween pins 1 and 2 and the above seqLence is repeated until a true balance (equal slopes) is achieved. I n practice for proper adjustment of R1, the approach to the true balance conditions is stepwise, requiring less than ten seconds. Reagents. Reagents used in this work are those de-cribed in detail earlier ( 5 ) . Important working reagents are sodium azidll, acidified iodineiodide, standard cystine, and reference electrode solutions all prepared in deionized water. The reference electrode solution contains iodine, iodide, and sodium chloride as a n electrolyte. The acidified iodine-iodide solution contains sufficient hydrochloric acid to adjust the p H of the ;ample solution to 5.8. Cystine standards are prepared by dilution of a 50-p.p.m. solution. Procedure. T h e procedure is outlined for t h e determi nation of cystine within t h e range of 0.25 t o 2.5 p.p.m. Minor modificatior s required for higher concentrationr are then given. T h e discussion arsulnes t h a t the operational amplifiers have been warmed u p and properly balanced. PREPARATIOX O F l?,QUIPhIENT. The circuit is connected ab shown in Figure 1 with the 2-megohm resistor in the divider circuit at the ntegrator output. The damping control on the recorder is set a t the minimum damping position (completely clockwibl.). With the recorder pen displacec to the extreme right the recorder zero control is adjusted until the m e t u a t the follower output reads 0 volt., on the 1.5-volt scale. The reference electrode solution is added to the thermostated reference compartment througb a hole in the compartment cover. The hole is then sealed with a rubber stopper. The sodium azide, acidified iodine-iodide (in a glass-stoppered bottle) and sample
solutions are adjusted to the working temperature (25.0' C.) by immersion in a water bath. Reagents and samples are handled with tuberculin-type hypodermic syringes fixed with glass tips. Solutions are removed from the sample compartment by a n aspirator tube. Prior to the analysis step the damping and sensitivity controls are adjusted. With the reaction mixture in the sample compartment containing the appropriate amounts of reagents (see analysis step) and 1.00-p.p.m. cystine, the damping adjust potentiometer (Rt) is adjusted until the servo approaches the balance position rapidly without overshoot. With the meter sensitivity set at the 1.5-volt position and R3 set near its midpoint, R2 is adjusted until the meter reads near 5 on the d.c. scale calibrated between 0 and 15. R3 is then adjusted until the meter reads exactly 5.0. The meter readout is then multiplied by 0.2 to get the cystine concentration in the reaction mixture in parts per million. Each of the above adjustments should be made during the first 30 seconds of reaction time. Therefore the first time the instrument is put into operation for a given concentration range, these adjustments will require more than one calibration run. For cystine concentrations between 2.5 and 25 p.p.m., the concentration is proportional to the 1.24th root of the slope of the response curve. I n this range a scale calibrated to read a multiple of the 1 2 4 t h root of the meter reading is attached to the meter face. The multiplier is taken such that the scale reads 25.0 full scale. The 2megohm resistor in the integrator divider circuit is replaced by a 200kilohm resistor. The damping and sensitivity controls are adjusted as described above with 10-p.p.m. cystine in the reaction mixture. The sensitivity controls are adjusted until the meter reads 10.0 on the nonlinear scale. The meter readout is then equal to the cystine concentration in parts per million in the reaction mixture. -1NBLYSIS STEP. The sample compartment is rinsed with 1 to 2 ml. of deionized water. Then 0.500 ml. each of sodium azide and acidified iodine-iodide solutions are added. Then 1.00 ml. of sample is added. Immediately after adding the sample, is closed momentarily and switch SW-> then opened. When the meter response becomes constant, the reading is taken from the appropriate calibrated scale and multiplied by a constant to obtain the cystine concentration. RESULTS AND DISCUSSION
Quantitative Data. Table I shows d a t a for t h e low concentration range where t h e meter readout is proportional to cystine concentration. Cystine concentrations given represent final concentrations in the reaction mixture. These d a t a demonstrate the concentration range over which Equation 7 is valid. A t the higher
Table I.
Automatic Results for Aqueous Cystine Solutions
(Proportional range) Metera Cystine reading, concn. P.P,m. R Taken Foundb 1.20 2.45 4.96 7.60 10.0 12.9
0.25 0.50 1.00 1.50 2.00 2.50
0.24 0.49 0.99 1.52 2.00 2.59
Rel. error,
Rel. std. dev.,
%
70
-4.0 -2.0 -1.0 +1.3 0.0 +3.6
4.3 2.0
0.7 0.3 0.6 0.6
a The value of R given in each case is the average for five separate runs. * These data are computed by multiplying R by 2 X lo-'.
Table II. Automatic Results for Aqueous Cystine Solutions
(Exponential range) nfetera Cystine read- concn. P,P.m. ing, R Taken Found-
Rel. error,
Rel. std. dev.,
2 5 5 0 100 15 0 200 25.0
1 40 -20 +10 0 0 +05 +0.8
4 3 1 7 0 8 1 1 0 0 0.4
2.6 4 9 101 15 0 201 25.2
2.6 4 9 101 15 0 201 25.2
%
9%
a The value of R in each case is the average for five separate runs.
concentrations, the relative standard deviation of the results is in the range of 1%. T h e relative errors a t intermediate concentrations also are in the range of 1%. The poor reproducibility and accuracy a t the low concentrations result from the fact that the meter used could not be read better than ~k0.05units. Data obtained a t increased sensitivity betting? demonst>rated reproducibility in the range of 1,5% for cystine concentrations down to 0.125 p.p.m. The large positive error at 2.5-p.p.m. cystine results from the deviation of the chemical system from the simple proportionality predicted by Equation 7. Table I1 s h o w data for the high concentration range where a nonlinear scale i.; attached to the meter face. These data demonitrate the validity of Equation 8. The reproducibility and accuracy at the low end of the concentration range are limited by the reproducibility with which the meter can be read. -1fter the initial adjustments are made the total meaiurement time for a single sample, including cleaning the sample compartment, adding reagents and sample, completing the measureVOL. 36, NO. 3, MARCH 1964
* 635
ment, and computing the cystine concentration is within 30 seconds. Effects of temperature, pH, and potential interferences in this reaction have been discussed. This work was done a t a temperature of 25.0 =t0.1’ C. and a t p H 5.8. Although the measurement method has been developed and d e w i b e d for the potentiometric detection of the rate of a chemical reaction i t should be applicable
to other signal systems. Also the method should be applicable to systems with nonlinear response curves. These possibilities are being investigated.
tists,” p. 356, W. A. Benjamin, New
LITERATURE CITED
(3)York, Malmstadt, 1962. H. v., Hicks, G. p., A ~ cHEM. ~ 32, ~ 3g4. (1960). (4)Malmstadt, H. T., Pardue, H. L.. Ibid., 33, 1040 (1961). ( 5 ) Pardue, H. L., Shepherd, S. A., Zbid., 3 5 , 2 1 (1963).
( 1 ) BlaedeI, W. J., Hicks, G. P., ANAL. CHEW34, 388 (1962). ( 2 ) Malmstadt, H. Y,, Enke, C. .G., Toren, E. C., “Electronics for Scien-
RECEIVEDfor review October 3, 1963. Accepted Xovember 29, 1963. Research supported in part by research grant from the National Institutes of Health, USPHS.
Spectrophotometric Determination of p-Phenylenediamines and p-Aminophenols with Ninhydrin ROBERT SUFFIS, ADELE LEVY, and DONALD E. DEAN Shulton, Inc., Clifton, N. J.
b The use of ninhydrin as a spectrophotometric reagent for the determination of low concentrations of p phenylenediamines and p aminophenols is proposed. Under the prescribed conditions no color is produced by any of the other aromatic amines investigated. This method is applicable to the determination of p-phenylenediamines and p-aminophenols in the presence of their ortho and meta isomers. Synthetic samples in the concentration range 0.05 to 0.50% were prepared and analyzed by this technique. The results had a range of i10% from the mean.
-
N
-
has found wide utility as a reagent for the spectrophotometric determination of amino acids as well as primary and secondary aliphatic amines. However, there appears t o be no published \T-ork concerning the quantitative spectrophotometric determination of aromatic amines with this reagent. Aromatic amines are known to give colored condensation products with ninhydrin. Most of these compounds have a color intensity so low that the reaction is uqable only as a spot test and then only if the aromatic amine is present in high concentration. Of the compounds investigated only p-phenylenediamines and p aminophenols give colors sufficiently intense for the spectrophotometric determination of trace amounts. Sinhydrin has been known to react with aliphatic primary amines and amino acids since its discovery by Ruhemann in 1910 ( 4 ) . The product of this reaction is a compound known as Ruhemann’s purple, which has an absorption maximum a t 570 mp. d reproducible spectrophotometric method based on this test was developed for the deter-
636
IXHYDRIN
ANALYTICAL CHEMISTRY
mination of amino acids by Moore and Stein ( 2 ) . The reaction of ninhydrin with aniline and its derivatives was first studied by Ruhemann ( 5 ) and later by Moubasher ( 3 ) . The reaction product that forms with most aromatic amines is compound I.
Ortho- and m-aminophenols and phenylenediamines did not give the highly colored reaction products produced from the para compounds. A possible reason for the reactivity of the para derivatives is given below. ;1Schiff base (111) may form from compound I by the loss of a mole of water. This compound can exist in a stable quinoid form (IV).
I Aniline forms a different product than the substituted anilines investigated. I n this case 2 moles of aniline react with 1 mole of ninhydrin t o give compound 11.
0
I1 These reactions have been utilized by Barakat, Wahba, and El Sadr (1) for the qualitative identification of aromatic amines. When 10 mg. of aromatic amine were allowed to react with 10 mg. of ninhydrin in 2 to 3 ml. of solution, colors were produced in most cases. This study was undertaken to apply the method of Moore and Stein to the quantitative determination of aromatic amines, Gnder these conditions most aniline derivatives gave a color that was not significantly different from that of a reagent blank. The products formed had molar absorptivities less than 20. The only compounds that gave more intense colors were p-phenylenediamines and p-aminophenols, which gave blue or purple products with molar absorptivities of 760 t o 9000 (Table I).
The intense color of the p-phenylenediamine or p-aminophenol adduct with ninhydrin may be accounted for by the high degree of stability which an amino or hydroxy group imparts to the quinoid form. The quinoid form would be expected to absorb light a t a longer wavelength with a greater intensity than the nonquinoid form. The meta derivatives have a much lower color intensity, because a quinoid form is not possible. Ruhemann (6) described the reaction of o-phenylenediamine with ninhydrin to give a yellow quinoxaline (Ir).
The absence of color in the reaction of o-aminophenol with ninhydrin may be explained by the added stability of the noriquinoid form (VI) due to hydrogen bonding.
