A STUDY OF OPTICALLY ACTIVE DYES, MECHANISM 
 OF DYEING AND ABSORPTION SPECTRA 
 
 BY 
 
 WALLACE REED BRODE 
 
 B.S. Whitman College, 1921 
 M.S. University of Illinois, 1922 
 
 FEE Gifvany. ar rue 
 
 PHESIS 
 
 - SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS 
 FOR THE DEGREE OF DOCTOR OF PHILOSOPHY IN CHEMISTRY 
 IN THE GRADUATE SCHOOL OF THE UNIVERSITY 
 OF ILLINOIS, 1925 
 
 Reprinted from the JOURNAL OF THE AMERICAN CHEMICAL SOCIETY, 
 - Vol. XLVI, pages 581 to 596 and. 2032 to 2043 
 A 
 
(ese. Gh abi ihe wiih on? mci 
 Wis eat ae ee es x 
 . pat \ NS Bs i sci tas 
 Breit ths # ce ns ba ee Rie Sadie a MEE ee . 
 ager Rae.) a ee ke Me) sii’ ie dye Meet Saice 
 
 Be ewes he. se “poe ca Bassi Me a 
 Bocinnrieh we Bo! en eaahio Ka Ra eins SROmay By Hey 
 PET LON E nid het Re ie, saad He Hi ie. Ee aS as 
 its ee ae Santee’ Writ. ey eal ok are 23) mak ed apps Pee qlee ATT 
 7 Rk Gifted slp aie ch as a nr ee ee et LY nak ot She 
 oot ea APRS Clik SFP ates See er yaey Fi aie aie tee tt sie Benes te 
 
 ESS RE Bia wT eusiissite sis MOLVA Sah denny Velibagomt SET 
 ka Leati yh ak aare mia vane =ae SSR ao slmate, we Bi aoe ed RPE apa + eee i ete Vise ie y 
 
 grasses MUTE Wane 
 ar eHetar 4 Srerar We © titer sf aS Ms ste ORI aes aad eae 
 
 pits Mia eiag Be. Pers th Rete Oe ies 5 Ee 
 \, T abet <t ne a a a se en ia’ WA en re ea a ig ~ a hy eee ie i 
 eat k ieee. AP eames ae SE GR PsP vet TORS TAS eet Ey, fh SPADA He Re 
 
 Hip PREREAON C-te Beitetudy id tie senchinian of dyeis bo the temercl nection: 
 ee pyrene Saicgebeat cele dick esa anet: AWE MOBy “otras stil 
 = Paiied. PP ied a sheB RIG AVITOA TIGA DITIAO = OH OK 
 i “worcatos VA ARTOL1ES. WOITAQKE A morrasozan: Hi 
 is i oda mas ef eakey Bien ct a Wrew cae PSN iw AP a ra 
 [oe ae theeeans reas Garena © ‘eer ie. aattdrigiS Tie ras oP a 
 “ bnilion: rin ites “ps c tamslk te ape 8 Hattie Spor Pe s rial 
 gant Barks okt MOTs is ia eg ‘to 1 HoHereqoty wilh ‘ol ‘padivie “3h ine 
 : ee abe mounts : a sido syeitiitliy ev) ay — iotesey 
 q & ) * (Reset Aapiegrits ew edi Jo gate rh to meine 
 ad? to forhsqmot.s subs lerartel\ bao b arn; oot SiN ied 
 is »ehecis. od. bluon toni. ani, ved sol quezds: ebieges ah emoilplen fags 
 wnge-mty yedsmiislog eqedieg, 10 tetomiveios: 8. Meenesiy 
 Si te boregnne- ad blues emotion i bes &. badepndxe yiigittaq 
 Eisen p orseed ear aolioqnmer:s oft! 2n0tnios lenigne edt dir bas 
 marion giig gy sre. iO) ened: binow.eguaead sostig Uy das - 
 
 Yih = 
 
 ahs Puree 
 
 i gt 
 
 1S ID cape AB bere WO HAB Baw. aed, etd sd eodtqoete lags 
 ah ren ilies soot doe. gas 10 _nektoser Jasimedd |e Aj, baod 
 Misr ici Abe ba io monpieebs lo e¢te to eaeresh ol) ,bediweba 
 dapat Pas lenin, RH: tapoose nciinow heb 
 aban: pia ates _ beens 2Ew. ~ 
 a dora aeilueon Pm. ti tot ammo oe. sabi This * 
 
 nes Pa ee : 
 
fReprinted from the Journal of the American Chemical Society 
 Vol. XLVI, No. 9. September, 1924.] 
 
 [CONTRIBUTION FROM THE CHEMICAL LABORATORY OF THE UNIVERSITY OF ILLINOIS] 
 
 OPTICALLY ACTIVE DYES 
 II. ADSORPTION, ABSORPTION SPECTRA AND ROTATION 
 
 By WALLACE R. BRoDE! WITH ROGER ADAMS 
 RECEIVED OCTOBER 38, 1923 PUBLISHED SEPTEMBER 5, 1924 
 
 In a previous article by Ingersoll with Adams? a very convenient method 
 was described for the preparation of pairs of enantiomorphic dyes. ‘The 
 research had as its ultimate object the determination of whether the mech- 
 anism of dyeing of fibers was a physical process, a chemical process, or 
 both. With the pure d and / forms of a dye, a comparison of the two equiv- 
 alent solutions as regards adsorption by the fiber could be made. By 
 means of a colorimeter or, perhaps, polarimeter, the concentration of the 
 partially exhausted d and / solutions could be compared with each other 
 and with the original solutions. The assumption was that compounds 
 that are mirror images would have the same dyeing properties provided 
 the adsorption by the fiber was a purely physical process. On the other 
 hand, if a chemical reaction of any sort took place when the dye was 
 adsorbed, the degrees or rates of adsorption of the d and / forms might be 
 different on account of the optical activity of the substance of which the 
 fiber was composed. 
 
 In order to be sure that, if such results: as have bere mentioned were 
 
 1 This communication is an abstract of a portion of a thesis submitted by Wallace 
 R. Brode in partial fulfilment of the requirements for the Degree of Doctor of Philosophy 
 
 in Chemistry at the University of Illinois. 
 2 Ingersoll with Adams, THis JOURNAL, 44, 2930 (1922). 
 
Sept., 1924 OPTICALLY ACTIVE DYES. I 2033 
 
 obtained, it would be proper to draw the above conclusions, certain as- 
 sumptions are necessary. First, the physical adsorption of the d and / 
 forms would be equal under like conditions. Second, the absorption 
 spectra of the d and / forms are the same, otherwise it would be impossible 
 by means of a colorimeter to determine the amount of dye adsorbed from 
 the solution of the two dyes. Third, the change in the hydrogen-ion 
 concentration as a result of the addition of an adsorption medium to the 
 dye solution would not affect the rate of adsorption by the fiber or the color 
 of the solution within experimental error. If an appreciable change in 
 the hydrogen-ion value of the solution took place during treatment of the 
 solution by the fiber, it would complicate the quantitative determination 
 of the amount of dye in solution and necessitate the control of the hydrogen- 
 ion concentration in the solution when the polarimeter and colorimeter 
 readings were made. 
 
 Before preparing a number of pairs of optically active dyes and studying 
 their adsorption by fibers, it seemed advisable to study one pair and de- 
 termine whether experiment would substantiate these assumptions and 
 warrant a detailed study of the mechanism of dyeing by the general method 
 planned. In this communication the B-naphthol derivatives of the diazo- 
 tized d, 1 and dl-phenyl(p-aminobenzoylamino)acetic acid, HO,C—CH— 
 (CsH;s) NHCOCsHiN==NCioH,OH, were chosen for study, since this set 
 of dyes is easily prepared and purified. The work was divided into four 
 parts: (1) the adsorption of these dyes by various inactive physical ad- 
 sorbing reagents; (2) the determination of the absorption spectra; (3) the 
 effect of hydrogen-ion concentration upon the absorption spectra; and 
 (4) the determination of the optical rotation of these dyes. 
 
 Adsorption 
 
 The adsorption of a dye by charcoal or an inorganic adsorbing medium 
 has been conceded by many’® to be of a physical nature although the point 
 has not been generally decided. If such adsorption were chemical, the 
 fact that the charcoal adsorbs equal amounts of the d and / does not 
 conflict with the deduction that the action of the fiber in adsorbing differ- 
 ent amounts of the d and / dyes is chemical. Equal adsorption by char- 
 coal and inorganic adsorbing agents even if chemical in nature, would be a 
 question of adsorption by a symmetrical substance, whereas fibers are un- 
 symmetrical substances. 
 
 A number of different adsorbing media were tested, including Tennessee 
 ball clay, kaolin, silica, alumina, aluminum hydroxide, Sil-O-Cel, fuller’s 
 
 ’ Wood, “Chemistry of Dyeing,” D. Van Nostrand Co., 1913. Dreaper, Trans. 
 Faraday Soc., 6, 31 (1910). W. C. McC. Lewis, Phil. Mag., 15, 499 (1908). Liddell, 
 “Chemical Engineering Handbook,’ McGraw-Hill Book Co., 1923, Vol. II, p. 780 
 (Jerome Alexander). Gee and Harrison, Trans. Faraday Soc,, 6,42 (1910). Freundlich, 
 Z. physik. Chem., 57, 385 (1906) ; 59, 204 (1907). 
 
2034 WALLACE R. BRODE WITH ROGER ADAMS Vol. 46 
 
 earth and charcoal. Silica and alumina adsorbed the dyes only slightly, 
 not enough to give good results. Sil-O-Cel and fuller’s earth adsorbed 
 satisfactorily, but difficulty was experienced in the centrifuging of the 
 material. The adsorption by Tennessee ball clay, charcoal and kaolin 
 was not complicated by these difficulties and gave satisfactory results, 
 the first two of these adsorbing agents being used for the adsorption ex- 
 periments. 
 
 For the absorption experiments, the d, / and dl forms of the dye were 
 dissolved in molecular equivalents of alkali solution; the hydrogen-ion con- 
 centrations of these solutions were determined, to be certain that they were 
 the same. Although a change in hydrogen-ion concentration of the solu- 
 tions by the addition of acid or alkali was claimed by Abderhalden‘ to 
 have no effect on the adsorbing power of charcoal, the solutions of the dyes 
 in this investigation were prepared with the same hydrogen-ion concentra- 
 tion to remove any possibility of change of color or solubility which would 
 thus affect the quantitative determination with a colorimeter of the amount 
 of dye adsorbed. 
 
 Adsorption by Tennessee Ball Clay.*—Dilute solutions of the d, / 
 and dl dyes, each containing 0.01428 g. of the dye in 50 cc. of water con- 
 taining a molecular amount of alkali, were well shaken with 0.0304 g. of 
 clay. The mixtures were allowed to stand for three weeks in tightly corked 
 containers; then the clear supernatant liquids were pipetted off and cen- 
 trifuged for half an hour to remove any fine suspended particles. The 
 original solutions and the exhausted solutions were then compared in a 
 Duboscq colorimeter in all possible combinations. The actual difference 
 in any solutions before and after treatment could be determined indirectly 
 through fifteen sets of readings, thus giving a very close check on the read- 
 ings obtained. ‘The errors in the readings were much less when solutions 
 of approximately the same concentration were compared, such as the orig- 
 inal d and the original J, or the exhausted d with exhausted ]. There was 
 considerably more variance when the original d, / or dl was compared 
 with the exhausted d, 1 or dl. The color intensity of these solutions was 
 a measure of the quantity of dye present. 
 
 Adsorption by Charcoal and by a Mixture of Charcoal with Silica.—Since pre- 
 liminary experiments showed that the amount of dye adsorbed by powdered silica 
 was negligible, the presence of this material did not affect the adsorption value of the 
 charcoal, but merely diluted the dye and made possible more accurate determination of the 
 amount of charcoal used. The ordinary animal charcoal was employed. The mixture 
 
 4 Abderhalden and Foder, Fermentforschung, 2, 74 (1917). 
 
 5 Clay used for this experiment was obtained from the Ceramics Department of the 
 University of Illinois. It gave an approximate analysis of silicon dioxide, 47%}; alu- 
 minum oxide, 38%; titanium oxide, 1%; iron oxide, 1%; calcium and other oxides 
 (0.5%; moisture, 12%; organic matter, 0.5%; 98% of this clay passed through a 200-mesh 
 ALEVE. 
 
Sept., 1924 OPTICALLY ACTIVE DYES. II 2035 
 
 with silica consisted of 1 g. of silica with 0.1 g. of charcoal. From this mixture weighed 
 portions were used for each experiment. In each case, the weighed amount of adsorbing 
 agent was added to the solutions which were then thoroughly shaken and allowed to 
 stand at room temperature for a period of five hours in the case of pure charcoal and ten 
 hours in the case of the silica-charcoal mixture. The liquid was then pipetted off and 
 centrifuged for one hour at high speed. ‘The amount of dye in the clear, supernatant 
 liquid was then determined as previously described under the experiments with Ten- 
 nessee ball clay. 
 
 Determination of the Colloidal Nature of the Dye.—Fifty cc. each of the d, 1 and dl 
 solutions as used in the adsorption experiments was filtered through a collodion filter 
 such as is used in ultrafiltration. Filter paper of a quantitative grade was soaked ina 
 3-4% collodion solution in glacial acetic acid and then immersed in water and kept under 
 water until used. The dye solutions were filtered under an air pressure of about four 
 atmospheres and the percentage of color change in the two solutions was determined by a 
 colorimeter as in the preceding experiments. ‘The amount of dye not retained by the 
 membrane was thus determined. The experiments showed that more of the d] form ex- 
 isted in a colloidal state than of the d or / solutions, a fact which would be expected from 
 their solubility. 
 
 TABLE I 
 ADSORPTION OF DYES 
 Tennessee ball clay Charcoal Charcoal and silica Contained colloids 
 0.0304 g. used 0.0200 g. used 0.0150 g. used 
 0.01428 g. of dye 0.00625 g. of dye 0.00625 g. of dye Colloids 
 50 ce. of soln. 25 cc. of soln. 25 cc. of soln. 0.0125 g. of dye 
 H 8.4 H 7.2 Pu 7.2 50 cc. of soln. 
 Adsorbed © Error Adsorbed Error Adsorbed Error Colloids <Error® 
 0 sk 0 ct 7% fod % =a 
 d 35.9 0:59 41.96 0.45 19.05. 0.61 73.50 0.29 
 L 390.8 94 41.93 Ph 19.09 61 74 .66 14 
 dl 41.7 48 46 .25 .22 27.05 49 84.71 12 
 Results 
 
 The above results bring out two very important points. In the first 
 place, they show that the active forms are adsorbed to the same extent 
 in the case of inactive adsorbing media where there is no possibility for 
 a true chemical action. ‘This does not mean, however, that the adsorption 
 by these media is necessarily physical in character, since equal adsorption 
 might be expected in the case of chemical combination with a symmetrical 
 adsorbing agent. The conclusion can be drawn that if any difference in 
 the adsorption of the dyes by symmetrical and asymmetrical adsorption 
 agents could be detected, it would be due to chemical reaction. 
 
 In the second place, there is shown the necessity of having absolutely 
 pure dyes with which to carry out these experiments, as the racemic form 
 has a higher adsorption. ‘This would naturally be expected from the fact 
 
 6 The probable error in this experiment and with others following was calculated 
 
 SE eba Seed, in which 2(-+v) is the sum of the deviations of all 
 
 n 4/ n—1 
 individuals from the mean, without regard to the sign and u is the number of individuals. 
 Mellor, Means and Average Errors in “Higher Mathematics for Students of Chemistry 
 and Physics,’’ Longmans, Green and Co., 1913, pp. 524-531. 
 
 from the formula 
 
2036 WALLACE R. BRODE WITH ROGER ADAMS Vol. 46 
 
 that it has a much lower solubility and a higher percentage in a colloidal 
 state, than the active forms do. The presence of any racemic form in 
 one of the active dyes would interfere seriously with adsorption experi- 
 ments with fibers. 
 
 Absorption Spectra 
 
 If the enantiomorphic forms did not have the same absorption spectra, 
 it would be impossible to make accurate quantitative determinations of 
 the amount of dye in the solutions by means of a colorimeter. ‘The identity 
 of the absorption spectra of various stereo-isomers has been investigated 
 by a number of writers’ and the general results conform with the rule given 
 by Kayser® that “substances with similarly related structure give similar 
 molecular vibration curves, whereas substances with different structures 
 give different curves. This applies without exception to aromatic com- 
 pounds as well as to alkaloids, dyes and colored substances.’ ‘The re- 
 sults obtained in this paper agree with this general rule and with the facts 
 obtained by various other workers, that the optical enantiomorphs show 
 the same frequency of vibration. The previous work, however, as a rule 
 has included only two of the possible three forms, usually the d and dl 
 as with corydaline;” or else compounds were used in which the color ion 
 was separate from the group containing the asymmetric carbon atom as in 
 the case of copper tartrate."4 The effects of color and rotation in 
 compounds such as the latter are not necessarily connected. 
 
 The solutions of the d, / and di 6-naphthol dyes for the determination 
 of the absorption spectra in the visible contained 0.01 g. of dye per 
 1000 ce. and in the ultraviolet contained 0.025 g. per 1000 cc. When- 
 ever the three dye solutions were compared, care was taken that the 
 hydrogen-ion concentration in each of the three was the same. ‘The so- 
 lutions used had a Sérensen value (PH) of 7.8. 
 
 The method for plotting the curves showing the transmission for differ- 
 ent wave lengths was similar to that used by the Bureau of Standards? 
 in which —logi of the transmittancy was plotted against the wave length 
 or frequency, depending upon whether the measurements were in the 
 visible or in the ultraviolet. ‘The extinction coefficient in this case was k, 
 
 where k = = (—log T), in which b is the length of the tube, ¢ is the con- 
 
 7(a) Dobbie and Lauder, J. Chem. Soc., 83, 605 (1908). (b) Gadamer, Arch. 
 Pharm., 239, 648 (1901). (c) Stewart, ‘‘Stereochemistry,”’ Longmans, Green and Co., 
 1919, p. 34; J. Chem. Soc., 91, 199, 1537 (1907). (d) Cotton, Compt. rend., 120, 989, 
 1044 (1895). (e) Hartley and Huntington, Proc. Roy. Soc., 31, 1 (1881). (f) Magini, 
 J. chim. phys., 2, 407 (1904). 
 
 8 Kayser, “Handbuch der Spectroscopie,”’ Hirzel, Leipzig, 1905, Vol. III, p. 213. 
 
 ® “Spectral Transmission of Dyes,’’ Gibson and others, Bur. Standards Bull,, 440, 
 Vol. 18, p. 128 (1922). 
 
Sept., 1924 OPTICALLY ACTIVE DYES. II 2037 
 
 centration, J the transmittancy which equals [/I’ where J and I’ represent 
 the intensity of light entering and leaving the cell. 
 
 The absorption spectra in the ultraviolet was determined at the Bureau 
 of Standards on a Hilger sector photometer.!° The apparatus used is 
 described in detail in Bull. 440 of the Bureau of Standards.* Photographic 
 exposures were made at various transmittancies and from these plates 
 the curve was determined. ‘The absorption bands in the visible were de- 
 termined in part by the Koenig-Martens spectrophotometer at the Bureau 
 of Standards, 1! and in part by the Keuffel and Esser spectrophotometer 
 at the University of Illinois. The Koenig-Martens instrument employs a 
 Wollaston prism as a polarizer and uses Nicol prisms to measure the 
 
 GENUS ae ae TENN NON | ot es reesei ye | 
 
 ea SSE 
 > 
 
 SEar canine 
 PPRSEEELLAL 
 Sana 
 ee a gig ae ete 
 
 Logio transmittancy. 
 XN 
 8 
 
 SY 
 towerdee ON eee 
 
 ¥ NS 
 NX 
 
 E 
 E 
 g 
 E 
 s 
 Z 
 E 
 : 
 
 ve 
 
 oe 
 ae 
 us 
 
 Spectral [rorrsim/(ssior? or 
 [emt &¢, and Ai SB Nagethe/ 
 rAd Sw il Dyes ina Concentration | 
 
 OF $ cg. per Liter: 
 
 ified (a! ay 
 5 9 nada a a a Sd 
 500 600 700 800 900 /000 H0O0 1200 [300 
 Frequency = vibrations + (seconds X 1012). 
 
 Fig. 1. 
 
 Ele 
 one 
 idee. 
 mnlA 
 als 
 
 intensity of the polarized light coming through the sample and the sol- 
 vent. The Keuffel and Esser spectrophotometer employs sector disks as 
 a means of controlling the light passing through the solution containing 
 the dye. 
 
 All three dyes gave bands of the same frequency both in the visible and 
 the ultraviolet. In the case of the d dye a small difference was observed 
 
 10 The authors are indebted to the colorimetric section of the Bureau of Standards for 
 help and suggestions in carrying out these determinations. 
 1 Ann. Phystk [4] 12, 985 (1908). 
 
2038 WALLACE R. BRODE WITH ROGER ADAMS Vol. 46 
 
 in the height of one of the bands, which is indicated by a dotted line on the 
 graph (Fig. 1). This difference was so small that it did not affect the shade 
 of the color sufficiently to be detected in the Duboscq colorimeter since this 
 colorimeter does not employ spectroscopic dispersion by the use of mono- 
 chromatic light of wave length of the maximum absorption as the means of 
 determining the color intensity. It is not out of the question that a slight 
 impurity was the cause of this difference in one of the bands. 
 
 The band in the visible was of a wave length of 486 mu or a frequency of 
 625. ‘The bands in the ultraviolet appeared at frequencies of 788, 956 
 
 Ppl wi bivehial ieALt obeew YA alede el ous [inate th fail 
 Seger 4 ee ee Ps be Pe 
 
 OF 
 06 
 ~ 
 © 08 
 s 
 Ew 
 a 
 § 
 re \ 
 8 Ban Eiek wisi ie) Spectral Trarsmissior ar ge! 
 
 ~ 
 A 
 
 Fee Eee lhe Vorious Corpcertrariors it hey, 
 file baa BS «HCI ok eae hal 
 OE eS ee a ee 
 PL A VRS ededealae wemlcoiste coe ee tee ae 
 2) LE CSS Se Rie Sie Ss es 
 [ec ie VA ind) are helt ene eRe NE AS 
 s BEBE BEE Why ish SS ee 
 
 400 300 700 
 Wave length, ana wT ~ 1078 
 Fig. 2. 
 
 16 
 
 /8 
 
 and 1108. ‘These bands are approximate multiples of the definite funda- 
 mental frequency of 158, the first band in the spectrum corresponding to 
 the fourth period and the others to the fifth, sixth and seventh periods. 
 The light transmission of these dilute solutions of the dyes were shown to 
 vary in accordance with Beer’s law, the logarithm of the transmission being 
 directly proportional to the concentration (Fig. 2) and the thickness 
 (Fig. 3). In determining the transmission of various concentrations, 
 solutions of 1, 2, 4, 10, 100 and 1000 cg. per liter were used. ‘Transmission 
 at various thicknesses of the solutions was determined by using tubes of 
 1, 2, 4, 7 and 10 cm. in length. 
 
Sept., 1924 OPTICALLY ACTIVE DYES. I 2039 
 
 If the optical isomers, especially the two active forms, had not had the 
 same frequency of absorption, the comparison of the two by means of a 
 colorimeter would have been impossible; a difference in the intensity of 
 the two bands would also complicate the determination of the dyes in the 
 solutions by the colorimetric method. As has been shown in these results, 
 however, the three enantiomorphic forms had bands of the same intensity 
 throughout the visible spectrum. In a quantitative determination of the 
 absorption of the dye by adsorbing media, the validity of Beer’s law was 
 assumed. This law was shown to hold for the solutions of the strength 
 used in these experiments. 
 
 Te SNS dca a 
 is inl GR SY 74a a 
 Se in Su B/N A 7S 
 Pe ee i a TN Za 
 he Naa ie ee a 
 Pees oa SR ay as 
 Ne SIN aelarive 
 God Sale A elle 
 i ona eT oo 
 ee Zo ea 
 S74 HS 
 Risa 
 
 Logio transmittancy. 
 “~ 
 S 
 
 14 
 16 
 18 
 far 
 Sy See eee 
 
 Wave length, mu= millimicrons = meters < 107°. 
 Fig. 3.—Visible spectra transmission at various thicknesses in cm. 
 
 The Effect of Change in Hydrogen-Ion Concentration upon the 
 Absorption Spectra 
 The addition of an adsorbing agent to a solution almost invariably 
 changes the hydrogen-ion concentration of the solution, the character 
 of the change depending upon whether the adsorbing medium is of a nega- 
 tive or positive character.!* It was, therefore, essential to determine the 
 color change which is effected by the change in hydrogen-ion concentra- 
 12 Bancroft, ‘“Applied Colloid Chemistry,’’ McGraw-Hill Book Co., 1921, p. 155. 
 
2040 WALLACE R. BRODE WITH ROGER ADAMS Vol. 46 
 
 tion. Thestrength of the solutions used for this experiment was the same as 
 those used in the determination of the absorption spectra in the visible. 
 The hydrogen-ion concentration of the solution was determined electro- 
 lytically. ‘The change in color with the change in hydrogen-ion concentra- 
 tion in the case of the 6-naphthol dyes is very slight between Pu 3 and 8.5, 
 but beyond 8.5 a sharp drop in the band takes place and secondary bands 
 appear at 443 and 516 mu (Fig. 4). A similar effect is shown when the 
 solution is made strongly acid. At PH 1.2 there is a sharp change with the 
 appearance of two bands at 522 and 553 my (Fig. 5). This curve does not 
 change in shape but simply drops in height, due possibly to the formiauen 
 of larger colloidal particles. 
 
 ROA SE RS | 2679 2 a | IETS |) 
 Fi <BR RE A862 ZZ PP TP 
 | lercess Att tt 
 Hal. NR Nils Sha A 
 
 OF 
 
 Logiy transmittancy. 
 
 Wave length, my = millimicrons = meters * 1079. 
 Fig. 4. 
 
 It has been shown in the experiments described above that the absorption 
 band does not change in wave length but rather in intensity with a change 
 in hydrogen-ion concentration. This. does not lead to the conclusions 
 of Smith and Boord,'* that the band shifts with an increasing alkalinity. 
 The change in intensity without change in wave length has been applied 
 to the study of indicators. A method has been developed by which hydro- 
 gen-ion concentration could be determined by taking this fact into account.!4 
 
 From the results obtained, it is seen that the band reaches approximately 
 its maximum height at Pu 7 to 8 and a change of 1 or 2 in the Sorensen 
 value on either side does not change the color intensity of the solutions 
 sufficiently to be detected by a colorimeter of the type used in this in- 
 
 18 Smith and Boord, Tuts JouRNAL, 44, 1449 (1922). 
 14 Brode, tbid., 46, 581 (1924). 
 
Sept., 1924 OPTICALLY ACTIVE DYES. II 2041 
 
 vestigation. The hydrogen-ion concentration in all these experiments 
 did not vary within the experimental error and did not complicate the data 
 in this research. ‘The solutions used in the adsorption experiments had 
 a Soérensen value of 7 to 8 and the actual hydrogen-ion change, after treat- 
 ment, did not exceed 0.5 Pu. It was permissible, therefore, to assume 
 that the color change due to the change in hydrogen-ion concentration by. 
 adsorption was negligible. 
 
 Poet Dididotade alee aby 50] WAGro] bos] ves fons] en | 
 a earner let tea] 
 [ousted mecca accel ed td Sd 
 onal icc auc Of AeA ee ene en a 
 Coa alco ie dnd dic A oe a Ae 
 [or \ Jace hte ee i on es es el 
 [ol os a me ee aco he in ll Oc 
 wo AN ab oy Al Sos a le 
 IN 4 Cie Effect of Charge of Pr i 
 | LUNA | oo ah Land at BNaptho! |_| 
 Iga gn a eh ea Ne Seem Dyes, (Acid Range) at 
 |e cede > bere pe UR eb 8 eel 
 [cas tacvat. <4) 4 0 0p 9 me era Oe ered ples had 
 
 S 
 \ 
 
 S 
 4 
 
 S 
 » 
 
 Togio transmittancy. 
 
 400 500 600 700 
 Wave length, my = millimicrons = meters X 107°. 
 Fig. 5. 
 Rotation 
 
 Rotation of colored substances at different wave lengths presents a 
 characteristic curve due to the proximity of the absorption band. ‘This 
 relation between the rotation and absorption bands has been studied by 
 Lowry, Akermann, Drude, and others,’ and the curves presented in this 
 paper are of the same type as those obtained by previous workers. 
 
 The rotation of these dyes was very difficult to obtain, due to their 
 physical properties, several of the difficulties being almost impossible to 
 ~ overcome. The dyes being red, monochromatic light sources such as 
 sodium and mercury were eliminated for the determination of the rota- 
 tions. Previous workers have described apparatus for the determina- 
 tion of rotations at various wave lengths other than those furnished by 
 - monochromatic light sources.1® ‘The best results for these dyes were 
 
 Lowry, Trans. Faraday Soc., 10, 106 (1914); J. Chem. Soc., 107, 1196 (1915). 
 Lowry and Austin, Trans. Roy. Soc. (London), 222A, 272 (1922). Pickard, J. Chem. 
 Soc., 123, 435 (1923). Akermann, Ann., 420, 1 (1920). 
 
 16 Bur. Standards Bull., 44, 17 (1918). Bates, Astrophys. J., 8, 214 (1898); Phil. 
 Mag., 18, 320 (1909). Patterson, J. Chem. Soc., 109, 1143 (1916). 
 
2042 WALLACE R. BRODE WITH ROGER ADAMS Vol. 46 
 
 obtained by the use of a 7-ampere, 20-volt straight condensed filament 
 lamp. The filament was about 16 mm. long and, on overloading, gave 
 about 200 candle power. ‘The filament image was projected on the slit 
 of a spectroscope and by making both the entrance and exit slits as narrow 
 as possible, nearly homogeneous light of a mono-chromatic nature was 
 obtained. The spectroscope used was one attached to a polarimeter 
 which was a large Franz Schmidt and Haensch instrument capable of 
 reading to 0.0002°. Due to the construction of the spectroscope, how- 
 ~ ever, as the far red end of the spectrum was approached, the dispersion 
 
 fig ea lioe dices caddie ideo Gee ane acc 0K ta 
 
 Specitic Kotatior of thé 
 
 Ee! AEF, 
 
 eae dW and LENapthd lyes pad Pgh hed ih ik te a 
 ie eee” Beem fee be Ph 
 Fee is | Pa as ae ey Aes Pe gee oe 
 OP HH ls A ea 
 400 300 
 
 Wave length, haga (aha wae ei ee 
 Fig. 6. 
 
 became less so that the readings obtained beyond a wave length of 650 . 
 muy are of no great quantitative value. The rotation of the d dye (that is, 
 the one obtained by diazotization and coupling of d-phenylamino-benzoyl- 
 amino-acetic acid) in this case was negative and that of the / dye was posi- 
 tive (Fig. 6). These dyes act similarly to amino acids in that on long stand- 
 
 ing with sodium hydroxide, racemization takes place, and in the case of satu- 
 rated solutions of the active forms the dl, which is less soluble, precipitates. 
 The probable reason for the failure to obtain rotations of the dimethy]l- 
 aniline dye in sunlight!’ was that the rotation of the dye approached zero 
 
 1 Ingersoll with Adams, THis JOURNAL, 44, 2930 (1922). 
 
Sept., 1924 OPTICALLY ACTIVE DYES. II 2043 
 
 at the wave length where the absorption curve crossed the maximum ¢en- 
 sitivity curve of the eye. At other wave lengths, however, rotations could 
 be observed, the rotation being positive on one side and negative on the 
 other. ‘The rotations by sunlight involve various wave lengths and since 
 the rotation is dependent on the wave length, the results obtained by sun- 
 light, or any heterogeneous source which is not treated so as to obtain a 
 light of monochromatic nature, are of no great quantitative value. 
 
 Summary 
 
 1. The adsorption of enantiomorphic forms of dyes by charcoal, clay 
 and other inactive adsorbing agents is the same. 
 
 2. The adsorption of the racemic form of these optically active dyes is 
 different from that of the active. 
 
 3. The active forms exist more nearly in the form of true solutions 
 than does the racemic modification. 
 
 4, ‘The absorption spectra of the optical enantiomorphs give bands of 
 the same frequency. 
 
 5. The change in hydrogen-ion concentration of the dye solutions does 
 not change the wave length of the band but rather the intensity of the 
 various bands. 
 
 6. Within the range of hydrogen-ion concentration change involved 
 in the use of common adsorbing agents, the change of color of these par- 
 ticular dyes is negligible and the possible difference in the intensity of the 
 two forms is negligible as far as the use of colorimetric determinations is 
 concerned. 
 
 7. Beer’s law is applicable to the dilute solutions used in these experi- 
 ments. 
 
 8. As far as studied, the rotation of these dyes and intermediates is 
 normal. 
 
 From these statements it is apparent that these dyes are suitable for 
 future experiments in connection with the possible mechanism in the dyeing 
 of wool and other fibers containing active constituents. 
 
 URBANA, ILLINOIS . 
 
Sia hierte ane hy OU WEI CRRA aGO ee 
 
 oS: 
 
 , ri SEELEY Te SO Sop SURE Shp TTR ets ORY Porte 
 : cua 8 by i RE Be Soot ae a, 
 : at A 3 Mit Mb Aji aae RTE eae 
 + i ae h A 
 ¥ , P : 4 : ‘ » pe 
 2 rey ee eee me oT Hoty 
 ; r. bape 
 ’ UIE Sis Aor Feat 
 ; 
 riset ye ' 4 aed i , a eT i Ese ae GOTT TOE- ig’ 
 ' : Y 
 
 : 
 * ake 4 
 aa 
 - 
 E 
 
 4 ) fy he ‘ at ge me or ae ee ee mbt an 
 ~ Sit Ft Vit289 ; te 2 Bt JEILD  £3ecFa8 Bei be BDF Meee De. eae iF 
 eC, arene “yteecel Se eee ee tani swe) ter euler Lage ht 
 prov MmoV ET bene & > ee 1 See ae 1 Oe 28 o | MeNEat Owe) Mea, J Abaly ee 8 
 
 ioe yee i). ster). je TOT | oe Bie Be 3 DIR TOEDL POTEET OY Te 
 
[Reprinted from the Journal of the American Chemical Society, 
 Vol. XLVI, No. 3. March, 1924.] 
 
 (CONTRIBUTION FROM THE CHEMICAL LABORATORY OF THE UNIVERSITY OF Inurors] 
 
 THE DETERMINATION OF HYDROGEN-ION CONCENTRATION 
 BY A SPECTROPHOTOMETRIC METHOD AND THE 
 ABSORPTION SPECTRA OF CERTAIN INDICATORS! 
 
 By WALLACE R. BRODE 
 RECEIVED NOVEMBER 23, 1923 
 
 In the study of the absorption spectra of dyes and other colored organic 
 substances in aqueous solutions, knowledge of the hydrogen-ion concen- 
 tration of the solutions is quite essential, due to the variation of the ab- 
 sorption band in shape and intensity with various hydrogen-ion concen- 
 trations. Several methods have been suggested for the colorimetric de- 
 termination of hydrogen-ion concentration of colored solutions, but all 
 of them are more or less complicated and in most cases give only approx- 
 imate results.? 
 
 All of these methods require the use of a colorimeter or other special 
 apparatus, whereas the suggested spectrophotometric ‘nethod employs 
 no apparatus in addition to that necessary for the determination of the 
 absorption band of the colored solution, and may be made in a compara- 
 tively short time after the completion of the determination of the ab- 
 sorption band. , 
 
 It was observed in the spectrophotometric study of certain dyes at 
 various hydrogen-ion concentrations,* that on the change of the hydrogen- 
 ion concentration of the dye solution the absorption band did not shift 
 gradually in wave length but merely changed in height. In the same 
 manner the absorption bands of the indicators studied did not shift in 
 ~wave length with a change in hydrogen-ion concentration but varied in 
 intensity while, in most of the cases observed, another band proceeded 
 in the opposite direction at the same time. By using the same concen- 
 
 1 This paper is the result of suggestions derived from a paper given by Buswell and 
 Smith on ‘‘Color Standards in Water Analysis,’ at the 1923 Spring Meeting of the 
 American Chemical Society, in which the absorption spectra curves of phenol red for 
 various hydrogen-ion concentrations were determined in a manner somewhat similar 
 to the method used in this paper, and a paper by Baker and Davidson (see Ref. 4d) 
 on “Spectroscopic Measurements of the Hydrogen-ion Concentration Color Changes 
 in Recent Indicators.” 
 
 2 (a) Gillespie, J. Bact., 6, 399 (1921). (b) Meyers, J. Biol. Chem., 50 [Proc.] 
 22 (1922). (c) Everz, Analyst, 46, 393 (1921). Other references as given in (d) Clark’s 
 “The Determination of Hydrogen Ions,’”’ Williams and Wilkins Co., 1922, pp. 66-73. 
 
 3 Brode with Adams, Forthcoming article in Tu1s JOURNAL, 46 (1924). 
 
582 | WALLACE R. BRODE Vol. 46 
 
 tration of indicator in all solutions the hydrogen-ion concentration of an 
 unknown solution may be determined by comparing the height of its 
 absorption band with those obtained from solutions of known hydrogen-ion 
 concentrations. Any absorption due to the color of the original solution 
 may be counteracted and its effect removed by using the original solution 
 in the comparison cell. Since the height of the band varies, within certain 
 ranges, with the hydrogen-ion concentration of the solution, it is necessary 
 only to use a standard quantity of indicator solution to give a colored 
 solution, the height of whose absorption band may be compared against - 
 the standard curves presented in this paper and the hydrogen-ion concen- 
 tration determined. 
 
 The method proposed is not new or distinctive but very few data has 
 been prepared on it and a large part of these data have been determined 
 with a simple spectroscope, rather than by using a spectrophotometer, 
 so that only the edge and not the peak of the band was determined.* A 
 study of the series of curves presented in this paper will show that such a 
 treatment does not lead to accurate or conclusive results in regard to the 
 effect of hydrogen-ion concentration on the absorption spectra of these 
 various indicators. Baker and Davidson have determined the absorption 
 spectra of a number of the Clark and Lubs indicators at various Sérensen 
 (PH) values, the curves being determined photographically and not photo- 
 metrically. Their data, although not of the same degree of accuracy as 
 attainable by a spectrophotometric method, confirm the fact that the 
 band changes in intensity but not in frequency with a gradual change of 
 the Sérensen value. ‘The object of their work was to devise a new method 
 to determine Sérensen values by spectrophotometric means. ‘The de- 
 termination of the dissociation constant of phenolphthalein by means of 
 the change in height of the absorption band of that indicator> presents a 
 method of determination which is quite similar to the proposed method 
 for the determination of hydrogen-ion concentration. ‘The measurement 
 of the change in the amount of color content of one particular type in 
 certain indicators with a change in PH is an approach at the same type of 
 measurements as carried out in this paper.® 
 
 A number of indicators were tried so as to ascertain the best for the 
 determination of certain hydrogen-ion ‘concentrations and to obtain a 
 range of curves that would cover a hydrogen-ion concentration of 1 to 
 10 Pu, inclusive. The most suitable type of indicator for this purpose is 
 one which has a narrow sharp band in the middle of the spectrum, whose 
 
 4(a) Ref.2d. (b) White and Acree, THis JouRNAL, 40, 1092 (1918). (c) Tingle, 
 J. Soc. Chem. Ind., 37, 117(t) (1918). (d) Baker and Davidson, Phot. J., 62, 375 (1922). 
 (e) Holmes, THis JourNAL, 46, 208 (1924). ; 
 
 5 Hildebrand, Z. Elektrochem., 14, 349 (1908). 
 
 6 Bjerrum, Samml. Chem. Tech. Vortrage, 21, 80 (1914). 
 
March, 1924 ABSORPTION SPECTRA OF INDICATORS 583 
 
 color change covers a fairly wide range of hydrogen-ion change, whose 
 band at its maximum is quite intense, and the secondary band, which 
 appears on the reduction of the primary band, is far enough removed from 
 the latter so that it does not affect the determination of the height of the 
 primary band. If possible, the secondary band should also be in the visible 
 portion of the spectrum so that further continuous observations can be 
 made on the indicator. An indicator satisfying all of these conditions 
 might not necessarily be an ideal indicator, as it must have chemical 
 properties which. will cause it to resist certain reagents. 
 
 The indicator which comes nearest to satisfying this requirement is 
 probably thymol blue, although the range that it covers is rather un- 
 common. ‘The intermediate range within which thymol blue does not 
 give any change can be covered by a mixed indicator of methyl red and 
 bromothymol blue. This gives two indicator solutions capable of covering 
 the entire range from 1.0 to 10.0 Pu. The curves formed by these two 
 solutions do not overlap to any great extent, and since the accuracy de- 
 creases as the maximum or minimum height of the curves is approached, 
 other indicators were studied which covered these ranges, making it 
 possible to obtain more accurate data at these values. 
 
 On account of their brilliancy, permanence, the shapes of their absorption 
 _ bands, and the positions of the bands in the spectrum, the Clark and Lubs 
 indicators’ seem quite superior to almost all of the common indicators, 
 with the possible exception of phenolphthalein which is a phthalein dye 
 like the Clark indicators. Methyl red, which is not a phthalein dye, is 
 not as satisfactory as the other indicators of this series, but must be used 
 as it covers a range that is not covered by any other of these indicators. 
 
 Apparatus 
 
 The spectrophotometer was a Keuffel and Esser Model C direct reading 
 color analyzer, with a lamp housing containing the standard magnesium 
 carbonate block holders in a direct line with the solution tubes, rotating 
 sector discs and constant deviation spectroscope. ‘The light source con- 
 sisted of two 400-watt projection Mazda lamps. ‘The tubes supplied with 
 the instrument proved unsatisfactory for the purpose of this research, 
 as they were made of copper alloy and were affected by the alkali or acid 
 in the solutions to be tested. A much simpler tube was used in this work 
 which gave quite satisfactory results. It was so constructed that the 
 only surface in contact with the solution was glass and it was possible 
 to empty, wash and refill the tube in a very short time. For the ex- 
 periments described in this paper, tubes 5 cm. in length were used. The 
 
 7 Ref. 2d, p. 74. 
 
 8 Recently some new sulfonephthalein indicators have been prepared, among which 
 is bromocresol green, which almost perfectly covers the hydrogen-ion concentration of 
 methy] red. (a) B. Cohen, Public Health Repts., 38, 199 (1923). 
 
584 WALLACE R. BRODE Vol. 46 
 
 rubber caps on the tubes have a decided advantage over metal in that 
 they prevent to a large degree the warming up of the solutions due to 
 conduction of heat. 
 
 Inasmuch as the curves which were being studied were sharp and pro- 
 nounced, it was found much more satisfactory to set the transmission for 
 
 DISPERSION PRISM 
 
 TO VACUUM VENTILATOR Ere Sut 
 
 MOTOR DRIVEN AUXILIARY 
 SECTORED DISK 
 
 Peat EV er" FROM STANDARD 
 THRY EYE SLIT 
 LIGHT FROM SAMPLE 
 
 SAMPLE HOLOER FOR 
 
 REFLECTION MEASUREMENTS WAVE LENGTH SCALE 
 
 HOLOER FOR 
 STANDARD SAMPLE 
 
 COLOR ANALYZER 
 Fig. 1—Optical system. 
 
 certain definite values and adjust the wave length until the two halves of 
 the field matched, rather than set the wave length at a definite value and 
 adjust the transmission. Values for the transmission settings were so 
 chosen as to give direct values for the ““—log.’’ of the transmittancy or the 
 extinction coefficient, at intervals of 
 0.10 from 0.00 to 2.00. In this way 
 twenty different settings were made 
 of the transmission discs and the posi- 
 < ea tion of the band noted for both sides 
 RUBBER CAPS of it throughout this series of settings, 
 Fig. 2.—Absorption cell. or as far as the band extended. In the 
 case of a substance which gave a broad flat band this method was not as 
 satisfactory or did not give as many readings as one which required the 
 setting of the wave length at definite intervals and the adjusting of the 
 transmittancy to obtain the necessary data. 
 
 ® Gibson and others, Bur. Standards (Sci. Paper, 440) 18, 124 (1922). 
 
March, 1924 ABSORPTION SPECTRA OF INDICATORS 585 
 
 The readings were all made within 15 minutes of the time that the in- 
 dicator was introduced into the solution and in most of the cases observed, 
 the curves were completed within ten minutes. Data for the entire curve 
 were obtained for each solution and the points obtained recorded directly 
 on special graph paper so that any discrepancy caused by misreading a 
 number on either of the two discs could be rectified before the solution was 
 discarded. In practice, however, for the determination of the hydrogen- 
 ion concentration in solutions, it would not be necessary to determine the 
 entire curve but merely the peak of the curve, so that the determination 
 could be done in a much shorter time. 
 
 Experimental Part 
 
 The buffer solutions used with the indicator solutions to give definite 
 values of hydrogen-ion concentration were for the most part made up 
 according to the Clark and Mcllvain buffer standards.!° By this means 
 buffer solutions were obtained covering a range of 1 to 10 Pu at intervals 
 of 0.2 PH and in some cases two different buffers were obtained which 
 covered the same values, so that a check reading could be made on them. 
 
 The following strengths of indicator solutions'! were made up after 
 first determining the maximum height of the band produced when 1 cc. of 
 the initial indicator solution, made according to the directions of Clark 
 and Lubs, was added to 50 cc. of solvent, and then this indicator solution 
 was diluted to a final strength which would give an absorption band that 
 approached an extinction coefficient of 2.00 at its maximum. 
 
 TABLE I 
 STRENGTHS OF INDICATOR SOLUTIONS 
 
 Initial Final Ratio of final Pu range 
 
 Indicator g.per100cc. g.per100cc. to initial of indicator 
 Thymol blue 0.04 0.04 1.0 1.2—- 3.4 
 Bromophenol blue .04 02.4 4 0.5 2.4— 5.6 
 Methyl red .02 .015 75 3.4- 7.0 
 Bromocresol purple .04 .032 8 4.8- 7.6 
 Neutral red .02 .0266 1.33 5.6— 8.4 
 Bromothymol blue .04 .04 ee] 5.8- 8.4 
 Phenol red .02 .01388 0.66 6 .2- 9.2 
 Cresol red .02 .016 4 6.8- 9.6 
 Thymol blue 04 .032 8 7 .4-10.2 
 Phenolphthalein .02 01 5 8 .0-10.6 
 
 Other indicators, such as congo red, methyl orange, benzopurpurin 4B, 
 methyl violet, etc., were tried, but the band was too broad, the secondary 
 band too near the primary, or the indicator so unstable that the change 
 in the height of the band was small or difficult to determine. For the 
 solutions mentioned above, water was used as a solvent except in the cases 
 
 10 Ref. 2d, pp. 111-116. 
 11 Ref. 2d, p. 80. 
 
586 WALLACE R. BRODE Vol. 46 
 
 of methyl red, phenolphthalein and neutral red. In the first two of these 
 95% alcohol was used, and in the latter 50% alcohol. 
 
 To 50 cc. of the buffer solution of known hydrogen-ion value, 1 cc. of 
 the prepared indicator solution was added, the solution thoroughly shaken 
 and the absorption spectra immediately determined. ‘This process was 
 repeated for each indicator covered by that particular hydrogen-ion value 
 and for all hydrogen-ion values at intervals of 0.2 Pu from 1.2 to 10.0 Pu. 
 From these series of curves, obtained by the determination of the ab- 
 
 i Za 
 
 aa 
 
 Transmittancy. 
 
 IN 0.50 
 
 5 0.20 
 50. 
 F 0.10 
 = i 
 ae Gh ieee 
 51.20 | 
 B ykstiedecl-blle kad ake lecloaGd ac alachae ae aaa 
 i RECS Y QR 
 1.60 
 
 ee Ree kee 
 i bol | tot olka eet | LS a 
 
 500 660 
 Sivan length mp = ta Ma = aes x 10 30 
 Fig. 3.—Transmittancy curves for bromophenol blue from 2.2 to 5.4 Pa: cell thick- 
 ness (b) = 5cm.; concentration (c) = 0.04 g. per 100 cc. 
 The curves for phenolphthalein were similar to the primary band only of bromo- 
 phenol blue (there being no secondary band); maximum at 553 mu, range (8.4 — 10.0 
 Pu) b = 5cm.; c = 0.01 g. 
 
 sorption spectra of each of these solutions, another set of curves was 
 derived, by plotting the height of the absorption band in terms of the ex- 
 tinction coefficient, against the hydrogen-ion values for each of these bands. 
 These curves which correspond to dissociation curves®* are similar for 
 all of the indicators tried and afford a more condensed form for comparison 
 of the heights of the bands, and a quicker method for determining the 
 hydrogen-ion concentration of an unknown solution by the spectronio 
 metric method. { 
 In using the condensed graph for this determination, it is not necessary 
 to determine the entire curve of the absorption band, but simply the peak 
 
_ March, 1924 ABSORPTION SPECTRA OF INDICATORS 587 
 
 of the curve. For this reason it is convenient to know the wave length 
 of the absorption bands for the indicators used, the values given being 
 determined from the curves described in this paper. 
 
 TaBLE II 
 WAVE LENGTHS OF ABSORPTION BANDS 
 Wave length Wave length 
 
 Indicator my Indicator my 
 Thymol blue (acid) 544 Cresol red O72 
 Bromophenol blue 592 Phenol red 558 
 Methyl red 530 Thymol blue (alk.) 596 
 Bromocresol purple 591 Neutral red 533 
 Bromothymol blue | 617 Phenolphthalein 553 
 Thymolphthalein 598 
 
 For work involving a limited range of hydrogen-ion change or where 
 the hydrogen-ion concentration of the solution to be tested is approximately 
 known, the proper indicator may be selected for that range so that the 
 change in the absorption band will be between extinction coefficient values 
 of 0.50 and 1.50, in which range the greatest change in the band takes 
 place and also the greatest accuracy is attained in measuring the height 
 of the band. Where the hydrogen-ion concentration is unknown or where 
 a wide range is to be compared, the best indicators are thymol blue (with 
 a range of 1.0 to 3.5 and 7.5 to 10.0 Pu) and a mixed indicator consisting 
 of methyl red and bromothymol blue, in the same concentrations as in the 
 separate solutions of each of them. ‘This indicator will cover a range of 
 3.5 to 8.3 PH although in the middle of this range the accuracy is not as 
 great as might be desired. Values obtained in this part of the curve may — 
 be checked by the use of bromocresol purple. 
 
 To determine the hydrogen-ion concentration of an unknown solution 
 2 cc. of the indicator solution is added to 100 cc. of the unknown solution, 
 the mixture shaken thoroughly, and the height of the absorption band in 
 a 5cm. cell determined. The hydrogen-ion value is then determined by 
 comparing the height of this curve against the standard curves given here. 
 In using cells of other lengths than 5 cm. it may be assumed that Beer’s 
 law holds for these solutions and the values obtained may, therefore, be 
 reduced to the equivalent of a 5cm. cell. With the help of a color chart”? 
 or standard samples it is possible to determine the approximate hydrogen- 
 ion concentration and whether the correct indicator has been used. In 
 the case of colored dye solutions the 2 cc. of the indicator solution are 
 added to the 100 cc. of the colored solution and the absorption band de- 
 termined in the usual manner, except that in place of a colorless solvent 
 in the standard or compensating cell of the spectrophotometer, there will 
 be the dye solution of unknown hydrogen-ion concentration, without the 
 
 12 Ref. 2 d, p. 52. 
 
588 WALLACE R. BRODE - Vol. 46 
 
 indicator solution. For extreme accuracy, the standard solution should 
 be diluted with 2 cc. of the original colorless solvent for each 100 cc. so as 
 to compensate for the dilution of the dye solution by the indicator solution. 
 
 Discussion of Results 
 
 The curves obtained in the investigation of the effect of hydrogen-ion 
 concentration on the absorption spectra of these indicators are of interest 
 from a theoretical as well as a practical point of view. ‘They show clearly 
 that for at least two classes of dyes, namely, phthalein and azo dyes, the 
 
 cs 1.00 
 Ht 
 LE 
 
 Wi! 
 ok ft 
 Ht 
 f f} 
 I] 
 fl 
 0.40 i 
 pe. vr 
 . Uh); 
 Ht 
 } 
 } 
 if 
 | g j i CEPT 
 V il } 
 i 
 f 
 ty 
 
 0.00 | 
 
 Log transmittancy. 
 Transmittancy. 
 
 ee ee 
 mn oo ohae 
 EEE ++} a 
 1.20 
 che coe 
 7‘ fs ball 
 osha tala 
 
 Stop 4 tales 
 
 3880 420 460 500 540 580 #620 660 700 # £740 
 Wave length mp = millimicrons = meters X 107°. 
 Fig. 4.—Methyl red (3.2 — 5.4 Pu); b = 5cm.; c = 0.015 ¢. 
 The curves for neutral red were similar to those for methyl red; maximum 533 
 my; intersection of curves at 480 my; 0.10 transmittancy; range (5.8 — 8.6 PH) b = 5 
 cm.; c = 0.0266. 
 
 absorption band does not shift in wave length but simply changes in in- 
 tensity with a change of hydrogen-ion concentration. ‘This rule does not 
 hold for all classes of dyes; for example, the absorption band of methyl 
 violet, a triphenylmethane dye, changes in wave length with only a slight 
 change of intensity, upon a change of hydrogen-ion concentration. ‘This 
 effect is being further investigated and agrees with the data of Holmes 
 and with unpublished observations by Appel and Brode of the United 
 States Bureau of Standards on the fact that the triphenylmethane dyes 
 do not necessarily obey Beer’s law and that their color production appears 
 to be of a different type from that of the azo dyes.. The curves presented 
 13 Holmes, Ind. Eng. Chem., 16, 35 (1924). 
 
March, 1924 ABSORPTION SPECTRA OF INDICATORS 589 
 
 here also show that a secondary band appears on the disappearance of the 
 primary band and that at certain hydrogen-ion concentrations there is 
 an equilibrium between the heights of the two bands. 
 
 The existence of absorption bands of definite wave lengths and of an 
 equilibrium between two types of color-producing molecules, is in accordance 
 _ with recent color theories!* that a definite configuration and frequency of 
 vibration exists for each color of a certain substance. ‘The way in which 
 the fading of the band takes place fails to confirm theories that intermediate 
 
 Log transmittamcy. 
 Transmittancy. 
 
 | WY 0.20 
 ai PES: 
 a EEEECEEL AMEE 
 
 SBSSBA1 SRBSRBBEL s 
 
 1.60° 
 
 2.00 
 
 380 420 460 500 540 ~~ 580 620 660 700 86740 
 Wave length my = millimicrons = meters X 107°. 
 
 Fig. 5.—Bromocresol purple (4.8 — 7.6 Po); b = 5cm.; c = 0.016 g. 
 
 The curves for thymol blue (acid) were not completed within the range of Sérensen 
 
 values studied. They were similar in shape to the bromocresol purple curve but had a 
 
 maximum at 544 my and an intersection of the curves at 485 my and 0.30 transmit- 
 
 tancy; range (1.2 — 3.4 Pu); b = 5cem.; ¢c = 0.04 g. 
 
 colored products exist between any two principal colored modifications, © 
 as the curve retains the same general continuity and does not change on 
 one side of the center of absorption more rapidly than on the other. An- 
 other interesting point is that in practically all of the curves studied, there 
 was one point at which all curves of various Sérensen values crossed in 
 changing from one colored form to the other, this point being about midway 
 between the centers of the two bands. It-is hoped that it will be possible 
 to study the ultraviolet absorption spectra of these indicators with special 
 
 4 Moir, J. Chem. Soc., 121, 1555 (1922). Hantzsch, Ber., 48, 158 (1915). Wat- 
 on, ‘‘Color in Relation to Chemical Constitution,’ ongmans, 1918. Ref. 2 d, p. 54. 
 1 Moir, J. Chem. Soc., 123, 2792 (1923), and Ref. 14. 
 
590 WALLACE R. BRODE Vol. 46 
 
 reference to the relation between the structure of the dyes, the number, 
 shape and order of appearance of the various bands and the hydrogen-ion 
 concentrations required to produce or diminish the various bands. 
 
 The determination of acidity or alkalinity by spectroscopic means was 
 suggested by Tingle“ in which the lateral shifting of the band was meas- 
 ured by a simple spectroscope. ‘The conclusion was drawn that “the 
 absorption bands of colored solutions in most cases, shifted in wave length 
 rather than the substitution of one band for another’ with a change of 
 hydrogen-ion concentration. By using a simple spectroscope and a cell 
 
 Pu 
 
 Zan 
 
 ae 
 
 CB. 
 
 Log transmittancy. 
 
 S 
 ) 
 oa 
 
 | = 
 tated id 
 
 0.02 
 
 2.00 ** [t+ eran | —tl 901 
 380 420 460 500 540 580 620 700 740 
 Wave length my = millimicrons = meters X 107°. 
 Fig. 6.—Cresol red (6.4 — 9.4 PH); b = 5cm.; c = 0.016 g. 
 The curves for phenol red were similar to those for cresol red; maximum 558 my; 
 
 | 
 
 + 
 
 tt 
 i 
 r 
 
 intersection of curves at 481 my; 0.45 transmittancy; range (6.2 — 9.2 PH); b=5cm.; 
 
 c=0.0133 g. ; 
 
 of constant thickness, one obtains the same effect as that which results 
 from the following of the curves here presented at one definite extinction 
 coefficient... The apparent effect thus obtained is a lateral shift in the 
 band, due of course to the simultaneous change of the height of the peak. 
 This effect is especially true with azo dyes (Tingle used methyl orange) 
 where the secondary band is. near the primary and the bands are quite 
 broad. The same conclusions were reached by Smith and Boord,’® in 
 regard to the position of the peak of the absorption band of certain azo 
 dyes in solutions of various bases and different strengths of solutions 
 1% Smith and Boord, THis JouRNAL, 44, 1449 (1922). | 
 
 (=) 
 
 _ 
 
 o 
 Transmittancy. 
 
 \ 
 c 
 
March, 1924 ABSORPTION SPECTRA OF INDICATORS 591 
 
 Although they used a spectrophotometer they did not measure the peak 
 of the band, but simply the side of the band near the base and from that 
 concluded that the addition of alkali (or acid) caused a ‘‘decrease in the 
 frequency of the center of the vibration producing the absorption band,”’ 
 and that the band varied in wave length rather thaninintensity. Baker 
 and Davidson** have measured photographically the absorption spectra 
 of a number of the Clark indicators at various hydrogen-ion concentrations 
 and found that the ratio between the two different color components of 
 the indicator is dependent solely on the hydrogen-ion concentration of the 
 
 0.00 
 TL LESS |g 27 AM 
 
 S Ber 
 IBES> \-ScSuc77 OE 
 Cf ~ c¢/ jae 
 
 1.00 
 
 Vs ageie' 
 
 eS ft 
 
 ia ZN 
 
 eRe? ABER NNN) ER 
 aa Al i Ned MGA Are 
 ee eeeeee a eee 
 : SRRERANY/ SRR 
 100 sotedlel ea ah \ Wa) ee il ede 
 
 co eee 
 leer ee 
 
 0.01 
 
 380 420 460 500 540 580 620 660 700 £740 
 Wave length mz = millimicrons = meters X 107°. 
 Fig. 7.—Bromothymol blue (5.8 — 8.4 Pa); b = 5cm.; c = 0.04 g. 
 The curves for thymol blue (alkaline) were similar to those for bromothymol 
 blue; maximum at 596 my; intersection of curves at 500 mu; 0.47 transmittancy; range 
 (7.4 — 10.0 Po); b = 5cm.; c = 0.0382 g. 
 
 solution and not on the amount of indicator present. Their data also 
 confirm the fact that the absorption bands of these indicators change in 
 intensity rather than in wave length with a change of hydrogen-ion concen- 
 tration. 
 
 It is noticeable that the band after Peaening a certain hydrogen-ion 
 value which gives it a maximum height, remains constant at this value 
 over a considerable range of hydrogen-ion concentration and then in certain 
 cases, on further treatment with alkali or acid, as the case may be, it again 
 begins to drop and a different band appears. The shape of the phthalein _ 
 and sulfonephthalein curves particularly adapts them to use as indicators, 
 
592 ; WALLACE R. BRODE Vol. 46 
 
 as they are narrow, sharp and show a relatively large change in the position 
 of the absorption-band maximum in relation to the change in the hydrogen- 
 ion concentration of the solution. ‘The azo dyes, on the other hand, have 
 broad curves with a secondary band so near the primary that it adds to 
 the effect of the primary and reduces the amount of possible change in the 
 transmissive index, with a change in the hydrogen-ion value of the solution. 
 The use of methyl red, however, was necessary, due to the fact that in this 
 investigation there were no phthalein dyes available which covered the 
 particular hydrogen-ion range covered by methyl red.** 
 
 . RSE 
 Ww || Fos 
 
 Transmittancy. 
 
 Log transmitta 
 
 : YUL 
 Seo * BAL | fo20 
 WAAC 
 1.20 HOA VALDEa\ Ga } 
 | ChCN ote 
 SLANE 
 AC 
 
 2.00 : 
 1.0 2.0 3.0 4.0 5.0 sal OPH 
 Hydrogen-ion ee 
 Fig. 8.—Condensed graph showing the relation between the extinction coefficient of 
 the peak of the bands (—log transmittancy) and the hydrogen-ion concentration of the 
 indicator solutions. 
 
 The dissociation or ionization constants of the indicators studied may 
 be determined in a manner similar to the method applied to phenolphtha- 
 lein, methyl orange, etc.,!”>5 the constant being taken as the middle 
 point on the condensed curve between the maximum and minimum heights 
 of the band, or approximately on the line representing an extinction coeffi- 
 cient of 1.00. ‘The values obtained agree closely with those obtained by 
 Clark.'8 
 
 Application and Limitation of Method 
 
 Unknown solutions were made up from various buffer solutions and the 
 hydrogen-ion concentrations determined both by this method and by the 
 
 7 Prideaux, “Theory and Use of Indicators,” Constable, 1917, p. 127, 
 #® Ref, 2d, p. 90, 
 
March, 1924 ABSORPTION SPECTRA OF INDICATORS | 693 
 
 TABLE III 
 DISSOCIATION CONSTANTS OF INDICATORS 
 : Spectrophoto- Clark’s Spectrophoto- Clark’s 
 Indicator metric value value Indicator metric value value 
 Thymol blue (acid) 175 14 Bromothymol blue 7.10 a0 
 Bromophenol blue 4.05 4.1 Phenol red 7.90 7.9 
 Methyl red 4:95, 5.1. Cresol red 8.20 8.3 
 Bromocresol purple -6.30 6.3 Thymol blue (alkaline) 8.90 8.9 
 
 electrometric method. ‘The electrometric determinations were made with 
 a palladized electrode in a special type of quartz hydrogen electrode de- 
 veloped by Dr. T. E. Phipps. 
 
 TABLE IV 
 DETERMINATIONS BY THE SPECTROPHOTOMETRIC METHOD, AND ERRORS INVOLVED 
 Soln. no. Pu (elec.) Px (spectra) Error Indicator used 
 1 1.62 1.59 —0.03 TB: 
 2 3.42 3.42 .00 BPO B: 
 3 3.44 3.43 — 01 B. P. B. 
 4 4.40 4.42 + .02 Bar. bp: 
 4 4.40 4.48 + .08 M.R. 
 5 6.60 6.57 — .03 BECP: 
 5 6 .60 6 .58 — .02 Nor 
 5 6 .60 6..57 — .03 Bull B. 
 6 7.58 7.68 — .05 N.R. 
 6 7.58 7.59 + .01 Balen, 
 6 7.58 7.58 .00 PR. 
 6 1.05 7.66 + .08 COR 
 cf 7.85 7.92 + .07 Par: 
 7 0.85 7.87 + .02 Car. 
 8 8.19 8.29 + .10 PER: 
 8 8.19 8.29 + .10 CoaRs 
 9 8.44 8.50 + .06 Poe 
 9 8.44 8.42 — .02 Cons 
 10 8 .67 8.61 — .06 TB. 
 11 8.80 8.76 — .04 Cok. 
 12 9.20 9.15 05 EPP: 
 12 9.20 9.22 + .02 babe 
 13 9.39 9.36 — .03 PyP: 
 13 9.39 9.32 — .07 TI 
 Av. error 0.042 Pu 
 Mean error = (0 .0054 PH 
 
 ‘The method may be applied somewhat better to cloudy or milky solutions 
 than to colored solutions. Intensely colored solutions cannot be deter- 
 mined by this method, but dye solutions of the strength usually employed 
 in spectrophotometric analysis, where the peak of the band is to be de- 
 termined, can be used. A suitable concentration for the determination 
 of their absorption bands and one which permits the use of this method is 
 about 0.2 to 0.25 cg. of dye per liter of solution. In using cells of other 
 
594 WALLACE R. BRODE 7 | Vol. 46 
 
 thickness this concentration may be varied accordingly. ‘The error in- 
 volved in the determination of the hydrogen-ion concentration of a colored 
 solution is somewhat greater than the above, due to the reduction of the 
 intensity of the light transmitted through the cell, especially if the ab- 
 sorption band of the colored substance is near that of the indicator. The 
 error, however, is not as great as that which would probably be incurred 
 in measuring the small amount of indicator solution. “In the determination 
 of hydrogen-ion concentration by this method it is suggested that check 
 observations be made on buffer solutions or on other solutions of known 
 
 SIRS ape: ne 27 ames 
 Hie ya: oor hua 
 Te 1 Nef) MBER: 
 pe We Ra 
 Leiden, 
 
 ttancy. 
 
 Log transmittancy. 
 
 ahs 
 = 
 =) 
 or 
 Transm 
 
 380 420 460 500 540 ~=580 700 740 
 Wave length mu = Se seas = adie SoLOiee: 
 Fig. 9.—Transmittancy curves for the mixed indicator of methyl red and bromothymol 
 blue (8.2 — 8.6 PH); b = 5cm.; cof M.R. = 0.015 ¢.; cof B. T. B. = 0.04 g. 
 
 hydrogen-ion concentration, in order to eliminate any possibility of error 
 in the preparation of the indicator solutions or other mechanical errors 
 in the spectrophotometer. 
 
 In this experiniental work no attempt has been Apa to correct for any 
 “salt effect.” The work of Acree’? has shown, however, that in the case 
 of buffer solutions of the concentrations of the Clark solutions, the salt 
 effect is small; for example, the correction for a 0.04 N phosphate mixture 
 was 0.02 PH and this concentration was greater than those of nearly all 
 of the solutions used in these experiments. Inasmuch as this method is 
 
 19 Brightman, Meackatn and Acree, J. Bact., 5, 169 (1920), and other articles by 
 Acree and co-workers in the Am. Chem. J., and THis JouRNAL. 
 
 1 
 
March, 1924 ABSORPTION SPECTRA OF INDICATORS 595 
 
 intended for the determination of the hydrogen-ion concentration in 
 solutions containing only 0.25 cg. or less of dye per liter, it would be per- 
 missible to neglect the salt effect, providing the buffer solvent for the dye 
 did not contain a salt concentration greater than that in the Clark buffer 
 standards. 
 
 The author wishes to express his appreciation of the suggestions given 
 by Professor Roger Adams of the University of Illinois, and Dr. C. E. 
 Waters and Dr. W. D. Appel of the United States Bureau of Standards, 
 during the progress of this work. 
 
 P 0.20 | 
 50.80 Sy 
 ~ q 
 ~~ 3 
 ‘g 0.102 
 a q 
 $1.20 
 bo 0.054 
 i 
 
 2.00 
 
 ae; -ion bie ets 
 
 Fig. 10.—Condensed graph showing the relation between the extinction coefficient of 
 the peak of the bands and the hydrogen-ion concentration of the indicator solutions of 
 thymol blue and methyl! red-bromothymol blue; b=5cm.; c of T. B. =0.0266; cof M. R. 
 = O:015; c of B. T. B. = 0.04. Data on M. R. Cicer at riage 
 
 Summary 
 
 A new method has been proposed for the rapid determination of the 
 hydrogen-ion concentration of dye solutions and other solutions which 
 are being studied spectrophotometrically. 
 
 Data are included, showing the effect of the change of hydrogen-ion 
 concentration on a number of common indicators and that the height 
 of the absorption band of these indicators is, within certain limits, a func- 
 tion of the hydrogen-ion concentration of the solution, 
 
 From the curves obtained it has been shown that on the change of the 
 hydrogen-ion concentration the center of the absorption band (the peak) 
 does not change in wave length, but rather in intensity. 
 
596 WALLACE R. BRODE - Vol. 46 
 
 A secondary band is in equilibrium with the primary band and on the 
 decrease of one of the bands the other increases in height. 
 The dissociation constant for the indicators studied was determined by 
 a photometric method. 
 URBANA, ILLINOIS 
 
VITA 
 
 The writer was born in Walla Walla, Washington, June 12, 
 1900. His grammar and high school education was secured at 
 the public schools and high school of Walla Walla. From 1917 
 to 1921 he studied at Whitman College where he received the 
 degree of Bachelor of Science in Chemistry in 1921. From 1921 
 to 1925 he has been a student in the Graduate School of the 
 University of Illinois, where he secured the degree of Master of 
 Science in Chemistry in 1922. Fora part of the time from 1923 
 to 1925 he has been a Junior Chemist and Assistant Chemist at _ 
 the United States Bureau of Standards, Washington, D.C. He 
 has been an assistant in Chemistry at the University of Illinois 
 from 1921 to 1924. 
 
 PAPERS AND PUBLICATIONS 
 
 The Determination of Hydrogen Ion Concentration by a 
 Spectrophotometric Method and the Absorption Spectra of 
 Certain Indicators. J. Am. Chem. Soc., Vol. 46, p. 581, 1924, 
 (with Roger Adams). 
 
 The Spectrophotometric Determination of Chroptorrabe 
 10 B. Ind. and Eng. Chem. 16, 797, 1924, (with W. D. Appel). 
 
 A Spectrophotometric Study of the Blue Color of Halite 
 Crystals. Presented at the 67th meeting of the American 
 Chemical Society, (with T. E. Phipps). 
 
 Optically Active Dyes II, Adsorption, Absorption Spectra 
 and Rotation, J. Am. Chem. Soc., 46, p. 2032, 1924. — 
 
 The Effect of Solvents on the Absorption Spectrum of a 
 Simple Azo Dye. Presented at the American Chemical Society 
 meeting in Baltimore, 1925. 
 
 The Spectrophotometric Determination of Agalma Black. 
 
 Presented at the American Chemical Society meeting in Balti- 
 more, 1925, (with W. D. Appel). 
 
30112 072908 
 
 ee