dimethylbenzene, 1,2-djnitro-3,4-dichlorobenzene, Ar,X-dimethyl-3,4-dinitroaniline, 2,2’-dinitrobiphenyl, 2,4’dinitrobiphenyl, 1,5-dinitronaphthalene, and 1,4-dinitronaphthalene. Of these compounds 1,2-dinitrobenzene, 1,4-dinitrobenzene, and 1,4-dinitronaphthalene were, by far, the best. The type of spectra obtained by the reaction of these three dinitro derivatives with 4Hcyclopenta [d,e,f]phenanthrene is shown in Figure 1. Considering the sensitivity arid the stability of the final colored products and the ready availability of the reagent, 1,2-dinitrobenzene was chosen for a more thorough study. The spectra of the reaction products obtained in the reaction of this reagent with fluorene, 2-nitrofluorene, and 2acetamidofluorene are s h o r n in Figure 2. One standard colorimetric procedure was used for all the reagents and all the test substances. The procedure
appears capable of application to the quantitative analysis of compounds containing the cyclopentadiene CH2 grouping. However, to obtain optimum results in the colorimetric analysis for any one particular compound, the conditions of the test would probably hare to be varied. Application. When t h e benzene extract of urban air-borne particulates is chromatographed through alumina with pentane, a fraction is obtained between the pyrene and chrysene fractions, whose ultraviolet absorption spectrum shorn points of resemblance t o t h e analogous spectra of benzo[a]and benzo[b]fluorene. The test applied to this fraction or to the benzofluorenes gave a grecn coior.,,,A, 730 mp. This result confirms the prcsence of a benzofluorenc in urban air particulates. The combined chromatographic fractions obtaincd after the
aliphatic fraction and before the pyrene fraction also gave a green color, Amax 695 mpj in the test. Apparently fluorene is also present in urban air particulates. LITERATURE CITED
( I ) Kuhn, R., Weygand, F., Ber. 69, 1969
11936). (2j Levy, K.$ Campbell, X., J . Chem. Sac. 1939, 1442.
(3) AIeisenheirner, J., Ber. 36, 4174 (1903). (4) Sanicki, E., Chemzst-Analyst 47, 9 11988). ( 5 ) SaAcki, E., Elbert, W., Ibtcl., 48, 68 (1959). (6) Sawicki, E., Miller, R., Stanley, T., Hauser, T., ~ ~ N A LCHEM. . 30, 1130 (19%). ( 7 ) Rislicenus, IT.,Ruthing, -A,, Be?. 46, 2770 (1913). RECEIVED for review Sovember 3, 1959. Accepted January 25, 1960.
Spectrophotometric Determination of Carboxyl in Oxidized Starch HERBERT C. CHEUNG, BENJAMIN CARROLL, and C. EDWIN WEILL Chemistry Departmenf, Rutgers, The State University, Newark, N. J. ,Methylene blue i s used in a spectrophotometric procedure to determine the carboxyl content of oxidized starches produced b y the hypochlorite oxidation process. The basis for the method i s the apparent constancy of the binding affinity o f methylene blue and carboxylate ion a t infinite dilution of the dye. Prior removal of inorganic cations usually present in carboxylated starches is not essential. For the special case of samples with exceedingly high carboxyl content, a simpler and more direct colorimetric procedure may b e used.
T
are several methods for the determination of the carboxyl content of oxidized starch. The two methods generally used are direct titration, which is applicable only to substances having relatively high carboxyl content, and titration of the acid liberated b y cation exchange, using a reagent such as calcium acetate (3, 6). The latter method yields accurate results, if the carboxyl groups are initially in the free acid form, and the sample is free from inorganic cations. Recently Hofreiter, R701ff, and Mehltretter ( 2 ) studied the determination of carboxyl content in dicarboxylated cornstarches by an ion HERE
8 18
ANALYTICAL CHEMISTRY
exchange procedure. This paper reports analytical results using the cationic dye, methylene blue, spectrophotometrically to determine the degree of carboxylation in starch. This method is more convenient and rapid than the conventional titration method, because it is based on one or two colorimetric readings as compared to a complete titration. For substances with lorn carboxyl content, the direct titration method is unreliable. I n previous attempts t o determine the carboxyl content of a carbohydrate using a cationic dye, the sample was kept in the solid form and vias usually washed or equilibrated n ith the dye solution. The uptake of the dye b y the solid was subsequently determined. A one to one correspondence is assumed for the dye-carboxylate ratio. This method was used by Davidson (1) to determine the carboxyl content of cotton cellulose. Attempts to use it for carboxylated starch yielded erratic results (a), although Schoch and Maywald (9) obtained qualitative information. The latter investigators used a wide variety of dyes to identify starch in granular form (9). I n the case of granular carboxylated starch not all the carboxyl groups are available for binding. There is also the possibility of nonspecific adsorp-
tion of the van der TTaals type. I n the method suggested in this paper, the carboxylated starch is dissolved and the uptake of the dye is determined directly in the solution. This is possible because the adsorption is accompanied by a spectral change. If the specific adsorption is extrapolated to infinite dilution, the ratio of the extrapolated values to the carboxyl content of the sample is a constant. EXPERIMENTAL
Buffers were prepared directly from reagent grade sodium hydroxide and monopotassium phosphate. -4histological grade of niethylene blue (Fisher Scientific Co.) 17-as used without further purification. Correction as made, however, for dye purity as given b y the supplier. T h e carboxyl contents of starches prepared by t h e hypochlorite oxidation process (3) Jvere determined by potentiometric titration, using the glass electrode. The starch samples were used without prior removal of the inorganic cations that were present. Apparatus. A Beckman Model G pH meter was used. A Beckman DU spectrophotometer equipped with cells of 1-cm. p a t h was employed for all absorbance measurements. Procedure. T h e starch solution was prepared in the usual manner bj7 first making a slurry, then pouring Materials.
WAVE LENGTH IN M r
Figure 1. Molar methylene blue PH 8 Ionic strength 0.0 1 Dye alone Dye plus 0.3% (0.84% carboxyl)
---
absorptivity
of
flask was 1 X 10-4V, and the ionic strength was approximately 0.01. The final starch concentration ranged from 0.1 to about 1%. A dye solution 1 X 10-4M at p H 8 prepared as described above was used to zero the spectrophotometer. Within approximately 10 minutes after addition of the starch, several measurements in rapid succession were taken at 580 mp for each sample. An alternative procedure would be to use a fixed concentration of starch and vary the total dye concentration. It was necessary to reneiv solutions from the volumetric flasks for each set of spectrophotometric measurements. Appreciable error may be introduced if the dye or dye-starch mixture is left in the cell for longer than 3 or 4 minutes, because of the adsorption of the dye on the cell n-alls. Appropriate correction must be made for the absorbance of the starch itself by subtracting the absorbance of the buffered starch solution (without dye) from that of the dyestarch mixture. RESULTS AND DISCUSSION
carboxylated starch
The effect of carboxylated starch on the absorption spectrum of methylene
Table I. Carboxyl Content of Oxidized Starches
yo Carboxyl Content Titration value 0.35
Binding constant valuea 0 33
0.44 0.i5 0.82
0.48 0.73 0.85
a Basis is three calibration samples iiot appearing in table.
constant value is taken as the absorbance of the bound dye. The fraction of dye nhich is frw can be expressed as (4)
F
=
-
Eb)/(cf
-
Eb)
nhere F is the fraction of free dye, E J is the molar absorptivity of the free dye, E b is the molar absorptivity of the bound dye. and is the molar absorptivity of the dye-starch mixture. Once F is obtained, the specific adsorption, Ro, nioles of bound dye per glucose unit of starch a t a particular free dye concentration, may be calculated. It has been shonn (6, 7 ) that for statistical binding, the follon ing relation holds : RIA
=
Kn - K R
where K is the intrinsic binding constant, n is the number of binding sites, A is the concentration of free dye, and R is the specific adsorption. Extrapolation of a plot of R/-I 21s. -4 to infinite dilution yields k l , the binding constant a t infinite dilution of the dye, 2
Figure 2.
4 AXIO’M
6
8
Adsorption of methylene blue and carboxylated starch
PH 8 lank strength 0.1 Total dye concentration ca. 1 X 10-*M 0 . 3 6 7 0 carboxyl
1.6770 carboxyl RG. Mole bound dye per glucose unit A. Free dye concentration
i t into boiling water. Boiling was continued for 3 minutes. After cooling t o room temperature, t h e solution was diluted t o a desired volume. A stock solution of 1 x 10-~.11 methylene blue, stored in a paraffined amber bottle, was used. Five milliliters of buffer of pII 8 and 5 nil. of the dye were added to each of a series of 50-ml. volumetric flasks. ( I t is most unlikely that phosphate will interfere with the adsorption of methylene blue, as phosphate is anionic.) Starch solutions varying from 1 to 25 ml. lvere added to each of these flasks, and the resulting mixture was diluted to volume. The final concentration of the dye in each
blue is shown in Figure 1. The change in intensity a t a fixed m v e length is a measure of the extent of binding of the dye and starch. To find the fraction of dye that is bound, it is necessary to know the absorptivities of the free dye and bound dye. The latter may be obtained b y plotting the absorbance us. starch concentration, keeping the total dye concentration constant. The absorbance d l rise with increase in starch concentration, and reach a value which virtually remains constant upon further increase in starch concentration. This
0
0.84% carboxyl
lim ( R I A ) = K n
=
121
A-0
A plot according to this equation for three samples is shonn in Figure 2. The intercepts on the Rc/d axis may be considered as the binding constants per glucose unit weight for these substances. The kl may be interpreted as being a direct measure of the degree of carboxylation, if one accepts the assumption that the predominant interaction between the Ptarch and the dye is electrostatic. A plot of the binding constant against carboxyl content should then yield a straight line. This linear relationship indicates that the binding affinity per carboxylate ion is an invariant. Once the binding constants of a few reference samples have been obtained, a calibration line may be constructed. The carboxyl contents of four carboxylated starches have been obtained from a calibration line using the binding constants of three reference samples. The results are VOL. 32, NO. 7, JUNE 1960
819
shown in Table I, with the values obtained from potentiometric titrations. The agreement is reasonably good. Choice of Wave Length and D y e Concentration. A relatively high concentration of t h e dye was used t o obtain a n appreciable spectral change. Optical measurements of this work nere made a t 580 mp, instead of a t a n-ave length corresponding to an absorption maximum of the dye. At this high concentration, sensitivity of spectral change was maximum a t 580 mp. Stability of Methylene Blue and pH Effect. T h e dye ryas unstable in alkaline medium, and t h e p H was, therefore, kept as close t o the acid side as practical. A suitable p H should be sufficiently high t o ensure the ionization of all carboxyl groups of the starch. p H 8 was satisfactory. The fading of the dye is also caused by certain kinds of surfaces. Thus i t fades more rapidly in a quartz cell than in an ordinary volumetric flask. For this reason, it is recommended that solutions fresh from volumetric flasks be used for each set of optical measurements. The influence of p H on the adsorption of methylene blue and oxidized starch may be further complicated by the fact that the molar absorptivity of the dye is also a function of pH. A threefold increase in the change of absorbance n a s observed when the p H increased from 6.5 to 7.5. However, the change n-as only about 2% betn-een pH 7.5
and 9.0. Thus a satisfactory pH value is around 8. Influence of Ionic Strength. T h e binding of methylene blue seems t o be susceptible t o changes of ionic strength. It decreases with increase in ionic strength. This may be attributed t o a competitive effect between t h e dye and the cations for t h e carboxyl groups of t h e starch. Under t h e specified conditions of this work, practically all dye ions were displaced when the ionic strength was increased from 0.01 t o 0.05. I n the recommended analytical procedure, the ionic strength is kept a t a n upper limit of 0.01,
for the dye and 0.01% for the starch. Once a calibration line has been obtained with reference samples, determination of an unknown involves only a single measurement of absorbance for the dye-starch mixture. ACKNOWLEDGMENT
The authors are indebted to T. J. Schoch for his interest in this work, and for supplying the starch samples and the potentiometric analysis of their carboxyl contents. They also acknowledge the financial support of the Corn Industries Research Foundation, Inc.
Effect of Heating Time of Starch Solution on Adsorption. The effect
of length of heating in t h e preparation of t h e initial starch solution as examined. Measurements were made of t h e same material at three concentrations. T h e heating times were 5 , 60, and 90 minutes, t h e solutions being refluxed during these periods. Within experimental error (+3%) , the resulting binding affinities r e r e the same. Special Case of Content Materials.
High-Carboxyl-
Substances having a carboxyl content above 0.7% may be analyzed b y a more convenient colorimetric procedure. For these substances a linear relationship exists between the carboxyl content and the absorbance of the dye-starch mixture. I n this procedure, the concentrations were fixed at 1 X 10-4X
LITERATURE CITED
(1) ~, Davidson. G. F.. J . Textile Inst. 41,
T3G1 (1950).
( 2 ) Hofreiter, B. T., Wolff. I. A., Mehltretter, C. L., J . Am. Chem. SOC.79,
6457 (1967).
(3) Kerr, R.,W., "Chemistry and Industry of Starch, Academic Press, Kew York,
1950. (4) ~, Klotz. I. 11 J . Am. Chem. SOC.68, 2299 (1948). ' ( 5 ) Klotz. I. 31.. I\'alker. F. M..' Pivan. R. B., Ibbid., 68; 1486 (194G). (6) Ludtke, M., Biochem. 2. 268, 372 (1934); 285, 78 (1946). (7) Scatchard, G. F., J . Am. Chem. SOC. 71, 660 (1949). ( 8 ) Schoch, T: J., Corn Pfoducts Co., Argo, Ill., private commnnication. (9) Schoch, T. J., Maywald, E. C., ANAL. CHEM.28,382 (1956). \ - I
~
RECEIVEDfor review May
20, 1958. Resubmitted December 9, 1959. Accepted March 28, 1960.
Improved Rapid Determination of Nickel in Soils and Laterites A. N. CHOWDHURY and B. DAS SARMA Chemical I aborafory, Geological Survey o f India, 27 Chowringhee Road, Calcutta, India
F A rapid geochemical method for the estimation of nickel i s based on the formation of a stable wine-red color. Nickel i s separated from most other metals b y extraction of nickel dimethylglyoxime b y a mixed benzeneamyl alcohol solvent. The separated organic layer i s shaken with strong alkali and more dimethylglyoxime, which extracts nickel again in the aqueous layer and develops a winered color. Air is the oxidizing agent.
F
several years the Geological Survey of India has been applying geochemical prospecting methods for OR
820
ANALYTICAL CHEMISTRY
location of new ore deposits and extension of old deposits. As a part of the program, geochemical prospecting for nickel, chromium, and vanadium in laterite and other residual materials has been started. A rapid method using a confined spot for estimation of nickel in plant ash was described by Reichen ( 3 ) . This requires a special apparatus, and like all confined spot methods, is not very reproducible. hfost of the reported procedures for the colorimetric estimation of nickel are based on the formation of a n intense wine-red color by nickel ion with dimethylglyoxime in the presence of a n oxidizing agent in basic media. This
was discovered by Feigl ( Z ) , first utilized by Rollet ( 4 ) , and later discussed by Sandell (6). Holyever, the color is unstable, varies in shade, and changes in intensity with time. An improved color reaction reported by Das Sarma and Ray (1) was based on the rapid development of a nine-red color when a nickel solution was shaken with a strong alkali solution and diniethylglyoxime, but in the absence of any oxidizing agent such as bromine or persulfate. The color is believed to be of the same chemical nature as in the earlier methods and may be due to the presence of nickel in a higher oxidation state, oxidized by oxygen from the air in