V O L U M E 23, NO. 5, M A Y 1 9 5 1
763
where pb
= barometric pressure in inches of mercury
V , = volume of the adsorbent section when empty (27.5 ml. for the author's apparatus) DISCUSSION OF METHOD
The method has been applied to a variety of materials having surface areas ranging from 2 to 700 square meters per gram. This method is particularly advantageous for mat,erials having specific areas in the range from 10 to 200 square meters per gram because of short time requirements. For accurate ineasurernent of materials less than 10 square meters per grain, it xould be advisatile to reduce the dead space. BET C' values (estiniat'ed from the pressure a t 1 = '/zk.J were found in all cases to be greater than 30. -1check run was carried out on many samples inmediately after the first run. Excellent reproducibility (within 2%) was invariably obtained, as illustrated in Figure 9. When check runs were oarried out a t a later date, reproducibility was within 4%. The average deviation from the mean of check determinations was 1.2%. The amount of error in area measurement result,ing from various causes on a 2OO-square-nleter sample is believed to be within t.he following limits: Measurement of flow rate Hydration of sample from exposure t o atmosphere during weighing Calibration of volumes Dead m w.-~ a c _e correotion .. _..~ ..~~.~ ~ eighing and transfer of sample to adsorbent tube Measurement of flow time during adsorption to end point Temperature fluctuations
1% 1%
1% 1V"
2%
1%
1%
Less than 15 minutes of working time is generally required for a surface area determination. ALTERNATIVE PROCEDURES
Although this laboratory has adopted the procedures detailed herein, there are other alternative procedures and usages as follows:
1. The increment method of gas addition can be used if desired and the amount of gas introduced measured by the time of aow ,
2. The gas can be allowed to equilibrate with the adsorbent at any desired pressure in the range from 0 to 33 inches of mercury by proper setting of valves 2, 5, 6, and 7; the desorption can be measured Fith a wet test, meter or by the pressure developed on gage 3. 3. The continuous flow method can be used for pore volume determination as well as area measurement. The Helicoid gage has both high and low vacuum contact points. However, equilibrium cannot, be approximated closely enough for accurate results (Figure 6). 4. Other more convenient gases such as n-butane a t 0" C. might be employed for area and pore volume measurement. However, this has not been tried as yet. 5. B vacuum recorder with a short time cycle can be used in place of the Helicoid gage to record the adsorption isotherm directly (uncorrected for dead space). 6. A manifold with several adsorption tubes might be employed t,o increase the capacity of t'he apparatus, or the same 6 pound-per-square-inch nitrogen pressure sources may be used for several units of the type described. ADDENDUM!
Improved operation has now been realized by use of a special low-flow lfoore flow controller; by use of a low pressure, pressure regulator (Conoflow Corp., Philadelphia, Pa.) in series with the cylinder regulator, and replacement of the relief valve with a needle valve. A Precision Scientific Time It timer is now used. The new Moore controller gives a constant flow rate up to saturation pressure with flows less than 5 ml. (STP) per minute. ACKNOWLEDGMENT
The author wishes to t,hank K. D. Ashley for his interest and encouragement, and the American Cyanamid Co. for permission to publish this work. LITERATURE CITED
(1) Barrett, E. P., and Joyner, L. G., ANAL.CHEY.,23,791 (1951). (2) Barrett, E. P., and Joyner, L. G., .J. Am. Cliem. Soo., 73, 373 (1951). (3) Brunauer, S., Emmett, P. H., and Teller, E., I b i d . , 60, 309 (1938). (4) Emmett, P. H., Ibid., 68, 1784 (1946). ( 5 ) Livingston, H. K., J . Colloid Sci., 4, 478 (1949). (6) Ross, E., J . Am. Chem. Soc., 70,3830 (1948). RECEIVED June 8, 1950. Presented before the Division of Colloid Chemistry a t the 118th hleeting of the AXERICAS CHEMICAL SOCIETY, Houston, Tex.
Polarographic Investigations of Reactions in Aqueous Solutions Containing Copper and Cysteine (Cystine) Amperometric Titration of Traces of Cysteine and Cystine with Cupric Copper as Reagent I. M. KOLTHOFF
AND
WALTER STRICKS, Vniuersity of Minnesota, Minneapolis, Minn.
From a polarographic study of the reaction between cupric copper and cysteine in ammoniacal medium in the presence of sulfite, it was inferred that this reaction could be made the basis of a rapid amperometric titration of traces of cysteine and cystine, using the rotating platinum electrode as the indicator electrode. Procedures are given for the rapid and accurate determination of traces of cysteine and cy+
A
MPEROMETRIC titrations of sulfhydryl groups in amino acids and proteins with silver nitrate have been reported (1,2 ) . No analytical use has been made of the reaction between cupric copper and cysteine in the presence of sulfite. From a polarographic study (3) it was inferred that it should be possible t o make this reaction the basis of a rapid amperometric titration. It is shown in this paper that with the rotating platinum electrode as indicator electrode traces of cysteine can be titrated
tine at concentrations between 4 X and M. Cadmium, zinc, and iodide do not interfere. Cobalt interferes. A n accurate and rapid determination of traces of cysteine and cystine is of great significance in the investigation of biological materials like normal and pathological blood sera. The new method is more specific and precise than methods that ha\-ebeen previously used.
simply, rapidly, and accurately. The end point can be detected more sharply with cupric copper than with silver as a reagent. The diffusion current involved in the titrations with cupric copper corresponds t o the reduction of cupric to cuprous copper, and not t o a reduction to the metal. Copper salts catalyze air oxidation of cysteine ( 5 ) , but it is possible t o obtain accurate results with copper solutions containing air if titration mixtures of the proper composition are used.
ANALYTICAL CHEMISTRY
764 Cupric copper reacts with cysteine and sulfite in ammoniacal medium to give cuprous copper and cysteine sulfonate (3) (designated as RSSOs-). As an intermediate, cuprous cysteinate (designated as RSCu) is formed from cysteine and cuprous copper, which under the proper conditions is oxidized further by cupric copper (see Equation 5 ) . The partial reactions can be represented by the following equations; Equation 6a represents the over-all reaction, which is the basic reaction in the amperometric titration. 2Cu(II) 2RS- = RSSR 2Cu(I) 2RS- = 2RSCu 2Cu(I) RSSR SO3-- = RSRSSO32Cu(II) 3RSSo,-- = 2RSCu RSSOI(4) 2RSCu 4Cu(II) 2S03-- = GCu(1) 2RSSO3(5) GCu(I1) 3RS3So3-GCu(1) 3RSS03(6) or 2Cu(II) RS- SOs-- = 2Cu(I) RSSO3(sa)
+ + + + + ++
equal. Figure 2 represents a polarogram a t the rotating electrode of a reaction mixture containing cysteine and cupric copper in the mole ratio 1 to 2 and of an ammonia and sulfite concentration of 0.1 and 0.15 M ,respectively. It is seen that this solution does not give an appreciable current between -0.3 and -0.0 volt.
+
+ + ++
+
+ + ++
I
MATERIALS USED
Cysteine hydrochloride was a Paragon reagent grade product. Cystine was a Merck reagent grade product. Directions for weighing cysteine hydrochloride and the preparation of cysteine and cystine stock solutions were given in a recent aper ( 2 ) . The copper metal used for the preparation of the stan&rd cupric copper solutions was an Eimer and Amend, hydrogen-reduced e . ~ . product. Directions for the preparation of the copper solutions are given in the procedures. EXPERIMENTAL METHODS
The equipment used for obtaining current voltage curves was a Heyrovskf self-recording polarograph. The amperometric titrations were carried out with a manual apparatus and circuit described by Lingane and Kolthoff ( 4 ) .
I
I
I
I
-.I - . 2 -.3 -.415 -.6- 7 - . 8 -.9
Figure 2.
I
40
J
I
I
-1.2
,
-14
E, VOLTS VS. S.C.E. Current Voltage Curves
At rotated platinum electrode of 50 ml. of 10-4 M cysteine (0.1 M "8, 0.1 M NHdCl, 0.15 M NaZSOd after addition of 2 ml. of 5 X 10-8 i W cupric copper solution
The products of the reaction between cysteine and cupric copper are cysteine sulfonate and cuprous copper. Cysteine sulfonate does not give a wave, and cuprous copper under the esperimental conditions is reduced a t about -0.65 volt a t the rotating platinum wire electrode. The addition of an excess $of cupric copper to the above solution gives rise to the appearance o i the wave of the cupric ammino ion, Cu(I1) e +Cu(1). The height of this wave is proportional to the concentration of the cupric copper in the solution. Cse is made of these observations i n the amperometric titration of cysteine and cystine with cupric copper. The best results were obtained a t an applied potential of -0.4 volt (US. S.C.E.).
+
TlTRATlOSS W'ITIi ROTATING PLATISUJI W I R E ELECTRODE A S ISDICQTOR ELECTRODE
-.I -.2 -3 -.47 5 76 : 7 -.8 3 -1.0 -1.1 E, VOLTS VS. S.G.E.
Figure 1. Current Voltage Curves 10-4 M Cupric copper in ammonia buffer (0.1 M "3, 0.1 M NH,NOs, 0.05 M Nags@) at rotating platinum electrode
A synchronous motor with gears provided rotation of the platinum wire electrode a t 600 r.p.m. A constant speed of rotation of the platinum electrode is required only in the determination of current voltage curves. For routine titrations a synchronous motor is not necessary. -411 potentials are expressed versus the saturated calomel electrode (S.C.E.). A 5-ml. semimicroburet divided into 0.01 ml. and a Gilmont ultramicroburet of 0.1-ml. capacity were used in the titrations. Oxygen was removed from the solutions with a stream of nitrogen which was passed through two wash bottles containing ammonia buffers of the same composition as that in the titration mixture. This was necessary in order to maintain a constant ammonia concentration in the titration mixture during the analysis. CURRENT VOLTAGE CURVES WITH ROTATING PLATINUM WIRE ELECTRODE
It has been shown ( 8 ) that cysteine, dissolved in an ammonia buffer, does not give an anodic diffusion current up to $1 volt if electrolyzed a t the rotating platinum wire electrode. An ammoniacal cupric copper solution gives two well-defined cathodic waves a t this electrode (see Figure 1). The first wave corresponds to the reduction of cupric to cuprous copper and starts a t -0.1 volt, while the second wave is due to the reduction of cuprous copper to copper; it starts a t about -0.6 volt (OS. S.C.E.). The diffusion currents of the two waves are practically
Titrarions ivere carriel1 out under varying coti.i:tioiis. The effects of the conwiit r:,tioii of aninioniu. aninwnilini :litrate, or chloride, oi sodium ; u l t i r c ~ . :ind oi thc cysteine :dncentrfition were studied. .\I1 titrations ivere carried out nhile nitrogen was bubbled through thta titrdtion misture. C'y-tcinc is airosidized relatively r:ipiJi). i n :likali:tt, .GolJrion. 11: :iw ;,rc..t.i:ce oi copper the :-y.teni is estremcly st1:isitive to ~ i r - ~ > ~ i ~ h t i s It 2ii. is found that osidhtion of c.yitc.irir I)?. a i r Jissul\.t-l in 1k.i cupric copper solutiuii. u5t.J a,< titr.,tiiig q y ! i t , is .ippw-i ii tht. reaction betwen cupric copper :ind cj.;teinr proceed. :!,),,\ ly. The rate o i thip rc-sctiun is affect, i by 3 1 1 ~ n : ~ ) I: i i t i i sulfite concentrations. Thc rcaction ra:e incrcnses n i r l i ILt.c:.e:tiing ainmonin hnd incrcsGing sulfitr cooncentration ;see .5 . I i the reaction is slow, the current mc:,surcd at -0.4 voir incre:ws abruptly upon e3ch addition of cupric copper and then dccrcnses slo\vly until a constant value i.2 attained (this is of tile order of 0.05 to 0.1 p . l . before the end poilit ;. In titrations w i t h misrurcs of a coniparativc.1.v 3 1 1 . l i l l sulfite concentration the rutc o i the reaction is normal at t l w bcgiiiiiing o i the titration but beconws markedly smaller a3 :tie reaction according to Equation 4 rcachc? completion. On iurilitt:. ahlition oi cupric copper the reacrion (0sid:ition o i RSCu TI) Cu- and R3503-) proceeds ia5tt.r agdin rncl soon reaches i t + x i t i a l rate, which persist3 throughout the reniaii,ing part tho titration. I1isturt.s of cysteine with zinc or cadmium ion rer+cti,ister and a t a uniform rate iii titrations: w i t h cupric copper. T!I%ia found 1: 5 . for mole ratios oi initial cystrine to metal from l : O . . j The simplest way 10 obtain a uniiorm and greu: r+action rite is the proper adjustment of sulfite and amnlonia conxn:ratinii. Solutions containiug 0.15 .\I sodium sulfite and 0.05 .If ammonia arc found to give the L c i t r t d t s iir titrations with Air-cont:ii!iing i1116,
V O L U M E 23, NO. 5, M A Y 1 9 5 1 cupric copper solutions. Under t'hese conditions the reaction is so fast that oxygen in the titrat'ing agent does not interfere under the experimental conditions. With large ammonia or low sulfit,e concentrations the errors can amount to -25%. Titratioiis repeatedly done with 50 ml. of 3 X lo-( to 8 X Jf cysteine solutions which were 0.5 JI in ammonia and 0.05 JI in sulfite and with air-saturated 5 X JI cupric copper solution (total volume added, 4 . i t o 1.6 nil.) gave end points a t which t,he mole ratio RSH : Cu(I1) was found to be about 1 to 1.5 instead of 1 to 2. This was found polarographically (see 3) as n-ell as in amperometric titrations with the rotating platinum wire electrode. Because the amount of oxygen added to the titration nlixture decreases with decreasing volume of the copper solution added, titration n.ith relatively concentrated copper solutions is recommended. This is particularly important in titrations of very dilute cysteine solutions. If an air-free cupric copper solut.ion is used as titrating agent, aninionia and sulfite concentrations can be varied widely without affecting the results. For example, correct results were obt,ained with 0.5 to 0.05 JI ammonia concentrations and wit,h 0.05 to 0.2 AI sulfite concentrations even though the reaction was very s l o a~t high aninionia and l o a sulfite concentrations. I n the presence of ammonia, cupric copper solutiolis can be niade air-free by the addition of sodium sulfite. ilmmoniacal cupric copper solutions which were 5 X to Jf in copper, 1 to 0.5 '1.I in ammonia, 0.1 to 0.05 AI in ammonium chloride, and 0.05 to 0.025 111 in sodium sulfite were found to be stable for about 2 hours after preparation. After this time a slow reduction of cupric copper to the cuprous state could be observed polarographically. Experiments carried out under various conditions showed that the ammonium salt concentration of the titration mixture has hardly any effect on the result. Correct results are also obtained in the absence of aninioniuni salts.
765 of cystine will use up 1 instead of 2 moles of cupric copper. Place about 5 ml. of a cystine solution, which should be at least 2 X 10-3 M in cystine and about 0.05 N in hydrochloric or sulfuric acid, in a test tube. Add a few drops of 1% sodium amalgam and let stand for 1.5 hours with occasional shaking. Treat an accurately measured portion of this solution further as in procedure 1. 3. Analysis of Mixtures of Cysteine and Cystine. For the analysis of mixtures of the two amino acids two amperometric titrations are required. One titration is carried out according to procedure 1. A second titration is done after previous treatment of the mixture n.ith sodium amalgam, as in the modified procedure for cyst.ine. If the number of milliliters of cupric copper solution consumed in the first and second titration are A and B , respectively, the difference ( B - A ) corresponds to the niolarity of cystine present in the mixture of the two amino acids. ( B - A ) ml. of 0.01 Jf cupric copper is equivalent to 1.20 ( B 7 A ) mg. of cystine. A O T E . If the concentration of cysteine is small as compared to the cystine concentration, the cysteine is better deterniined by argentometric amperometric t8itration in the absence of sulfite ( 2 ) ,preferably adding the silver from an ultramicroburet. PREPARATION OF STASDARD CUPRIC COPPER SOLUTlON
Dissolve an accurately weighed amount of C.P. metallic copper in 6 N nitric acid. After complete dissolution of the copper, add a few drops of concentrated sulfuric acid and evaporate until the residue is nearly dry. Dissolve in distilled mater, transfer int,o a volumetric flask, and make up to volume. The copper titer can be checked by iodometric titration or by electrolysis. Standardized cupric sulfate solutions may be used as well. PREPARATION O F AIR-FREE AMMONIACAL CLTPRIC COPPER SOLUTIONS
Place an accurately measured volume of a concentrated standard cupric cop er solution (0.05 to 0.01 J-1in copper) in a volumetric flask. Adfammonia and sodium sulfite to make the solution 0.5 to 1 &I and 0.05 to 0.025 JI,respectively, in these constituents. Sulfite must be added after the ammonia. Make up the solution to volume with air-free dist'illed water. Copper solutions prepared in this way must be used within 2 hours after preparation. RESULTS
PROCEDURES
1. Determination of Cysteine or Cystine. Introduce 25 to 100 ml. of a 0.05 to 0.025 M ammonia solution into a 150-nil. beaker which is provided with a rubber stopper with holes for electrode, buret, salt bridge, and inlet tube for nitrogen. Immerse a platinum wire electrode in the solution. Remove air with purified nitrogen, which is passed through during the entire titration. T o the air-free ammonia solution add enough cysteine or cystine solution to make t'he misture 4 X to 10-5 Jf in KSH or RSSR. If the cysteine or cystine solution is acid, neutralize it with ammonia. Make the mixture 0.15 to 0.1 hf in sulfite by adding the proper volume of 1 M sodium sulfite solution. Immerse the salt bridge and tip of the buret in the solution. For titrations of 5 to 0.5 mg. of cystine ( 4 X to 8 X M) a semimicroburet with 0.01 to 0.002 -11cupric copper solutions can be used, while quantities smaller than 0.5 mg. of RQSR (8 X 10-5 to U ) must be titrated with an ult'raniicroburet and more concentrat'ed cupric copper solutions (0.05 Jf). Titrate a t an applied potential of -0.4 volt us. S.C.E. Allow the cupric copper solution to flow into the titration mixture a t such a rate that the current remains practically constant (0 to -0.1 PA.) before the end point. After the end point the current increases rapidly. Measure the current after the end point after addition of three to four small increments of cupric copper solution. Plot the results and find the end point graphically. If carried out correctly, the time of titration proper is about 2 minutes. XOTE. h new indicator electrode mu8t be cleaned with concentrated nitric acid and rinsed with distilled water. It can be used immediately after cleaning. After each titration the electrode is rinsed with distilled water. If not in use, the electrode should be ke t in distilled water. If kept in this way, a used electrode neefnot be cleaned with concentrated nitric acid. According to Equation 6a, 1 ml. of 0.01 Jf copper solution corresponds to 0.788 mg. of cysteine hydrochloride. One mole of cystine in the presence of sulfite gives 1 mole of cysteine. One milliliter of 0.01 M copper solution corresponds to 1.20 mg. of cystine. 2. Modified Procedure for Cystine. Cystine can be reduced with sodium amalgam before the titration is carried out by the above method. After reduction with sodium amalgam 1 mole
Tables I, 11, and I11 give the results of amperometric titrations of cysteine and cystine as obtained with air-containing cupric copper solutions as titrating agent. Contrary to the argentonic.tric titrations ( 2 ) ,it is seen that the cysteine concentration has hardly any effect on the results. This is found with RSH and RYSR a t concentrations varying and 10-5 .If. I t is seen in Table I that at high between 4 X ammonia and low sulfite coiicentrution large negative errors are found, due to oxidation of cysteine by oxygen contained in the copper solution. The error is reduced to practically zero by using a more concentrated copper solution and adding it from an ultramicroburet (see Table 111)or uqing air-free copper solutions (Table 117). Under the experimental conditions given in the procedure it is necessary to add the cupric copper from an ultramicroburet or to use air-free copper solutions when the cysteine or cystine concentration is less than 8 X 10-5 AI, With an ultramicroburet cystine quantities as low as 60 micrograms in 25 ml. of titration mixture ( X ) can be determined with an accuracy of -1 to -2%. Gsing a semimicroburet 0.5 mg. of cystine (8 X .If) can be determined with an accuracy of -0.4%. The last experiment in Table I1 shows that the error becomes considerably greater if smaller quantities of cystine are titrated with a semimicroburet. From Tables I to 111 it is seen that best results with copper solutions containing air are obtained w-ith ammonia and sulfite concentrations of 0.05 -If and 0.15 t o 0.1 M , respectively. If lower or higher ammonia concentrations are used, the sulfite Concentration should be sdjusted in approximately the same proportion. Thus it is also possible t o obtain reasonably accurate results with solutions which are 0.1 or 0.025 M in ammonia if the corresponding sulfite concentrations are about 0.25 and 0.1 M, respectively. If the sulfite molarity is markedly lower than 0.1 Jf the reaction is slowv,eveu if the
ANALYTICAL CHEMISTRY
1166
Table I.
KO.of Detns.
Cysteine Hydrochloride Taken
Cysteine Hydfochlorlde Found
Average Error
M
Mo.
M g.
x
10-4
1.58
1.58
$0.3
1
2
x
lo-'
1.58
1,5l
-4.4
1
2
x
10-4
1.58
1.58s
-0.2
1
2 2
x
10-4
x
10-4
1.58 1.58
1.63 1.54
+3.2 -2.7
1 . 6 X 10-4
1.26
1.26
x
lo-'
1
1.2
1
1 . 6 X 10-4'
0.945
0.938
0.630
0.630
0.0
-0.7 0.0
2
2
x
10-4
1.58
1.19
-25.0
1
2
x lo-'
1.58
1.15
-27.0
1
ax
0.630
0.488
-23.0
10-6
Volume of titration mixture 25 ml.
Composition of Electrolyte
70
2
1
a
Init. RSHConcn. of Soln. Titrated
2
1
ing the diffusion current of the cupric copper flattens o f f in the absence of alkali, indicating the d i s a p p e a r a n c e of the cupric copper due to reduct'ion to cuprous copper. In the presence of alkali a straight line is obtained. In an amperometric titration of cystine or cysteine in a solution which is 0.01 M in potassium hydroxide and 0.1 A l in sodium sulfite, the slope of the excess reagent line is less than that obtained with an ammoniacal solution. From a practical point of view, it is advisable to carry out the titrat,ions in ammoniacal medium, whenever p o s s i b l e . The reaction is faster and the end point sharper in the presence of ammonia.
Amperometric Titration of 50-MI. Cysteine Solution with 0.005 M Air-Containing Cupric Copper, Using 5-M1. Semimicroburet
0 . 0 5 .M KHs. 0 . 0 5 M N H ~ K O I , 0 . 1 5 M XarSOs 0 . 0 5 AM "3. 0.05 M . NHdOa, 0 . 1 A4 XatSOa 0 . 0 5 AI NHa, 0 . 0 5 M SHdNOa, 0 . 2 M KarSOa 0.0 0 .52M 0.05 M 5 M"3, NazSOa 0 . 1 M "3, 0 . 1 .VI 0 . 2 5 X NatSOa 0 . 0 5 M NHs, 0.05 .M 0 . 1 5 M NatS03 0.06 M "3, 0.05 M 0.15 X NazSOs 0 . 0 5 M "3, 0.05 M 0 . 1 5 M NazSO3 0 . 5 J4 "3, 0.1 M . 0 . 0 5 AM h-atS03 slow) 1 0.)I. 0 5"3, M 0NazSOa . 1 Jf very slow) 0 . 5 .tr N H ~ ,0 . 1 M 0 . 0 5 Jf KatSOa
KHdOz, .\-"
