A manomorommféwov or ACID-BASE :NDIcAjorzs IN LIQUID AMMO‘ZNIA Thesis {or flu: Degree of DB. D. MICHIGAN STATE UNIVERSITY R. Dean Hill 1962 (rhesus run v'n-ri p .f.-;*[ 'II‘O‘If'IWC‘IT\( \ I _ . , |cl|i.«.,l.. .rn'w‘ J::7l x’.‘:. Van“! EASI' LANSIi‘xELS,lV1iLfilCzAN ABSTRACT A SPECTROPHOTQMETRIC STUDY OF ACID-BASE INDICATORS IN LIQUID AMMONIA by R. Dean Hill The behavior of a weak acid in liquid.ammonia can be described by an ionization step followed by a dissociation step: RH + “Ha 1” R-3NH4+ (1) R mu: 1!. R“ + NIL,+ (g) The equilibrium constant for equation 1 was studied spectrophoto- metrically using indicators of the acid-base type. The dissociation reaction occurs to a negligible extent except in very dilute solutions. A method for the calculation of ionization constants from spectrophotometric data was devised for both indicators and other acidic or basic compounds which affect indicators. A low temperature cell was devised for spectrophotometric study of the ionization of such substances in the ultraviolet and visible region. Five indicators were examined in detail. Absorption bands present inwthe visible and ultraviolet spectrafiwere related to the molecular structure of each indicator. The extinction coefficient for each species was calculated. Four of the indicators studied were strong acids while the other was suitable for studying basic solutions. The indicators p—nitrsacetanilide, 2 ,h-dinitroaniline , phenolphthalein, and thymol blue were found to have first ionization constants greater R. Dean Hill than one in liquid ammonia. The effect of neutral salts upon these indicators was studied. The effect of temperature upon the phenolphthalein ionization was investigated in detail and ionization constants calculated for solutions of this indicator in the liquid range of ammonia. A relationship between ammonium ion concentration and the absorbency of phenolphthalein.was found. This relationship enabled calculation of ionization constants of some weak acids. Thiourea, methylthiourea, and trimethylthiourea were found to be acidic in ammonia even though they are basic in water. Ionization constants for these substances are reported. Urea and methyl-substituted ureas have no effect on indicators which change color in acidic liquid ammonia, but they show a weakly basic reaction to thymol blue in its third ionization step. Urea does not affect tropeoline 00 (which changes color in basic solution) and therefore urea probably is nearly neutral in ammonia. A SPECTROPHOTOKETRIC STUDY OF ACID-BASE INDICATORS IN LIQUID AMMONIA By R. Dean Hill A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of Dacron or mnosom Department of Chemistry 1 962 ACKNWLEDGEMENTS The author wishes to express his appreciation to Dr. Robert N. Hammer under whose direction this investigation was accomplished. To Rupert A. D.‘Wentworth, the author would like to extend his thanks for many helpful suggestions and interest during the course of this study and to Forrest E. Hood for construction of the apparatus employed. A deep sense of gratitude is extended to the author's wife, Jean, for her patience and understanding which have made this investigation possible. 11 INTRODUCTION. . . . . . EXPERIMENTAL. . . . . . Absorption Cell. . ' Spectrophotometer. RESULTS AND DISCUSSION. Introduction Weak A01“ 0 weak Bases . . WOOOOOOSO APPENDHIOOOOOOO Volumetric and Dilution'Vessels. Thermostats........... AmmoniaSystem......... Use of Vessels and Absorption Cell Chanicals............. Sodium Salt of Phenolphthalein . 0.0... O O O O C C O O O O O O O O O O O O O O O O O O O O O O O 0 iii eeeOeeOQ new or CONTENTS seesssee O O O O O O O O 0 00...... O soeseee-O_ Page 1h 1h 17 20 25 28 32 35 37 37 1:2 82 99 100 105 108 116 118 TABLE I. II. III. V. VI. VII. VIII. LIST OF TABLES Colors of Some Substituted Phenols and Anilines in mania O O O O O O O O O O O O O O O O I O O O O O 0 Colors of Various Indicators in.Ammonia. . . . . . . Salt Effect on Indicators in Aqueous Sodium Chloride Solutions...................... Liquids Used for Constant Temperature Refrigerants . Indicator Color Changes in Ammonia . . . . . . . . . Ionization Constants and Extinction Coefficients for Various Indicators in Ammonia at -77'C . . . . . . . Ionization Constants of Phenolphthalein at Various Taperatures.................... Ionization Constants of Some Thioureas . . . . . . . is Page 10 to Sh 76 79 FIGURE 9. 10. '11. 12. 13. 1h. 15. LIST OF FIGURES Low Temperature Spectrophotometric Cell. . . . . . . . Low Temperature Spectrophotometric Cell (Schematic). . spOCtrOphOtomethe e s e e e e e Volumetric Flask . . Dilution Flask . . . Dilution Flask . . . Wheatstone Bridge. . Resistance Vtrsus Temperature. Ammonia Handling System. . . . FilternastQQOoecoso Log of Concentration of Ammonium Log 3'. of Phenolphthalein . . . . 0 O O 0 O O O O O O 0 Perchlorate Versus O O O O O O O O O O 0 Log of Concentration of Ammonium Bromide Versus Log 3 Of PhenOIPhthalein s e s s e e 0 Log of Concentration of Ammonium ofPhenolphthaleln.................. C O O O O 0 O O O O O Iodide Versus Log 5 Log of Concentration of Ammonium Nitrate Versus Log 5 ofPhenolphthaloin.................. Absorption Spectra of Some Hethylsubstituted Ureas iantersseeeeeseeesseesssssese 16. Absorption Spectra of p—Nitroacetanilide in Ammonia . . 17. 18. 19. Absorbancy Versus Concentration of p-Nitroacetanilide atBBS‘POOOOOOOOOOOOOOecease... Absorbancy Versus Concentration of p-Nitroacetanilide at M6 m P O O O O O O O O O O O O 0 O O 0 O O O O O 0 Log of £1 Versus Concentration of p-Nitroacetanilide at hh6 m P O O O O O O O O O O O O O O O O O O O O 0 O V Page 15 16 18 21 23 2h 27 29 3O 36 1:3 his 115 146 he 50 S1 52 53 LIST OF FIGURES Continued FIGURE 20. 21. 22. 23. 25. 26. 27. 28. 29. 30. 31. 32. 33. 3h. 35. 36. 37. Absorbancy of pritroacetanilide Versus Concentration of Potassium Perchlorate . . . . . . . . . . . . . . . . Absorption Spectra of 2,h-Dinitroaniline in Ammonia. . . Absorbancy Versus Concentration of 2,h-Dinitroaniline at 536mP C C C O . 0 . C O O C C O C C C O o O O O C . Absorbancy Versus Concentration of 2 ,h-Dinitroaniline atBBSmP. . O C. C O O O . C O C C C C O C O C O O O Log-gi Versus Concentration of 2,h-Dinitroaniline at 536mf......................... Log K3 Versus Concentration of 2 ,h-Dinitroaniline at 385mp.............,............ Absorbancy of 2,h-Dinitroaniline Versus Concentration of Potassium Perchlorate . . . . . . . . . . . . . . . . Absorbancy of 2,h-Dimitroaniline Versus Ammonium Perchlorate at Constant Ionic Concentration. . . . . . . Absorption Spectrum of Phenolphthalein in Ammonia. . . . Absorption Spectrum of Phenolphthalein in Aqueous monia. O O O O O O O O O O O O 0 O O O O O O O O O O O Absorbancy Versus Concentration of Phenolphthalein at SYSmPand'77oceesssesesesessssseee gbsorbancy Versus Concentration of phenolphthalein at YSnPO O O O O O 0 O O O O O O O O O O O O O O O O O O Absorbancy Versus Concentration of Phenolphthalein at 575 m P. O O O O O O O O O O O O O O O O O O O O O O O 0 Log I, Versus Concentration of Phenolphthalein at S7Smr. O O O O O O O O O O O O O O O 0 O O O O O O O 0 Log K, Versus Concentration of Phenolphthalein at S7SmpO_O-_OOOOOOOOOOOOOO0.0.000.0 Log K, Versus Temperature for Phenolphthalein at 575 n ’1 Absorbancy of Phenolphthalein Versus Concentration of P0u881unlsaltaats7SM’Jssesseesseesses Absorbancy of Phenolphthalein Versus Concentration of SodiumSaltsatS7SMIJ.................. vi Page 55 58 6O 61 62 63 65 68 7O 71 72 73 7h 75 77 78' LIST OF FIGURES 38. 39. 140. L1 . 112. 1:3 . 1:5 . 1:6. 1:7. 118. 1:9. 50. S1. FIGURES Continued Absorbancy of Phenolphthalein L__ersus Concentration of Some Thioureas at Constant Potassium Iodide Concentration. Log _I_(_' Versus Concentration of Some Thioureas. . . . . . . 1 Absorption Spectra of Thymol Blue in Ammonia . . . . . . . Absorbancy'zgrgug Concentration of Thymol Blue at 390 m.p. Absorbancy 12.13.19. Concentration of Thymol Blue at 301 m p. Absorbancy V_e_§_s3_s_ Concentration of Thymol Blue at 620 m )1. Loggia 'L__ersus Concentration of Thymol Blue at 620 m p. . . LogK '_V___ersus Concentration of Thymol Blue at 390 m ,1. . . Absorption Spectra of Tropeoline 00 in Ammonia . . . . . . Absorbancy Versus Concentration of Tropeoline 00 at h6omPO O O O O O O O O O O O O O O O O O 0 O 0 O O O O O Absorbancy Versus Concentration of Tropeoline 00 at S9Om.PO O O O O O O O O O 0:. O O 0 O O O O O O O O O O 0 Absorbancy of Tropeoline 00 Versus Log of Potassium AmideConcentration.................... Absorption Spectra of Thymol Blue in the presence of ore. at 390nP0 O O O O O O O O O O O O O O O O O O O O 0 Absorption Spectrum of Potassium Amide in Ammonia. . . . . vii Page 80 81 8h 86 87 88 89 90 91 92 93 95 96 98 INTRODUCTION Anonia in the liquid state as .s ionizing solvent and parent system of compounds has been extensively studied by many investigators (1, 2, 3, h, S, 6), usually in reference to its similarities to water. It is of interest to note that ammonia dissolves mam ionic compounds (7, 8, 9, 10, 11) and as such has been utilised as a reaction medium for new substances which are obtainable by few other means . Ammonia as a parent solvent undergoes the autoionisation: 2m, :- 1m.” + NH; (1) The equilibrium constant for this reaction is reported to be 10'"33 at -SO°C (12). In the solvent system concept, amonium compounds act as acids in amonia and amides as bases (13). ' Not only do ammonium compounds act as acids but an potential proton donor can undergo amonolysis to produce amonium ions: mums m.*+x’ (_2_) Thus new substances which are weak acids, neutral compounds, or even basic substances in aqueous solution become acids in amonia (1h). It has been shown that in solvents of medium dielectric constant, the coulombic force between ions is great enough to overcome the influence of the solvent and form ion pairs. Even higher aggregates tons in liquid ammonia solutions more concentrated than about 10'3 molar (15). 2 This is a function of charge density of the ions as pointed out by Bjerrum (16) 1 b"! 213.29 "' " +. s (2) z. 1000 s. 2.1232 where K is the equilibrium constant of an ion pair dissociating into -d its ions, 1 is the Avagadro umber, g; and g, are the charges on the respective ions, 3 is the electronic charge, 2 is the dielectric constant of the medium, _1_t_ is the Boltzmann constant, 3 is the absolute temperature, and i is the distance of closest approach of the ions. Bjerrum made the quite arbitrary definition that an ion pair is an two oppositely charged ions at a distance such that the electrostatic energy does not exceed 21:2 (17). For water this distance is 3.57 I at 25-0 and for ammonia it is 9.2h i at are for a 1:1 electrolyte. Thus a substance which exists as an ionic species in the solid state is not completely dissociated in liquid amonia. The number of unassociated ions in solution is then a function of cation-anion size and charge. Kraus, Hawes, and others have measured the conduct- ances of now common electrolytes in ammonia and calculated dissoci- ation constants of the order of magnitude of 10"" (18, 19, 20, 21). This small value for K -d concentration of solvated ions in ammonia except in very dilute solutions indicates that there is never a high and that the solution chemistry of this solvent must take into consid- eration ion pairing. In the consideration of am ionic equilibrium, the effect of ionic atmosphere must be taken into account. Debye and Hu‘ckel (22, 23) considered this in terms of Coulomb's law and the Boltzmann distribution law and observed that the true thermodynamic equilibrium constant of 3 a reaction must be stated in terms of activities and is related to concentrations by the activity coefficients. The expression in dilute solution for a given ionic activity coefficient is given by me 4°27, - 1,2,3“,— (u) where _Z_i is the charge on the ion, 71 is the activity coefficient of that ion, A is a constant given by 1.82 x 10-6/ (132)3/ 2 and is equal to 0.509 at 25‘C for water and 5.26 for monia at -77'C, _B_ is h.8 x 109/ (E)1/ 2, i is the distance of closest approach of ions, and p is the ionic strength which is defined by P ' '15! 9.1%.: (5) where 91 is the concentration of the ions. Equation g is valid for solutions up to 10'1 molar in high dielectric constant solvents and up to 10'3 molar in solvents of lower dielectric constant (2h). Since there is no experimental way to separate the positive and negative ion activity coefficients, equation 1, is used to express the mean activity coefficient, 7,, which is defined by (25): _l_i_ logx + n log): 2*}! 108 7; - (Q) where 7’, is the mean activity coefficient of a pair of oppositely charged ions of charge 3 and g. The effect of the ionic atmosphere on the nature of the solution for ion pairs is less well defined owing to the interaction of the coulombic forces between the ions involved. For a given ionic strength, the deviation from equation .1; in ammonia is greater than that in water. For almonia however, the ion h association which occurs tends to offset this such that equation .1; remains valid as a first approximation in dilute solutions (26). In amonia as in other solvents, there is the possibility of a molecular species such as a weak acid undergoing an ionization step such that an ionisation equilibrium is set up: ax + an, a waf‘u' (_7_) The ion pair formed can then participate in a dissociation equilibrium which is independent of the ionization step and can be described by: NH‘+3X- 7* NH: "' I- (Q) The overall. equilibrium constant, 5*, for the total.reaction has been defined by Bruckenstein (27, 28) and Lagowski (29) as: K 1 + £1 where E1 is the constant for the ionisation equilibrium in equation 1 and Ed is the constant for the dissociation equilibrium in equation 8 . Owing to the greater basicity of monia compared to water, may weak acids in water are enhanced in acidity in ammonia; i.o. they undergo complete ionization though not complete dissociation. Nearly all acid-base indicators which are used in aqueous solution are themselves weak acids, usually one color in the molecular form and another color in the ionic form. Mary of the familiar indicators dissolve in ammonia to form colored species, some of which are different from their aqueous counterparts (30). Neglecting ion pairs, 5 the equilibrium may be expressed in general as film + g at 3": + In‘ (1.9.) Color I Color II where g is the solvent in question. The equilibrium constant for this motion is : 5-H __-.'_§_______"'s ‘In" _ [H+s][In'] . 7’IIIsYIn' (1_1_) In gRIn—s [HIn][s] 7/HIn76 The absorption of the solution, i.o. the ratio of color I to color II is: determined primarily by the ratio of concentrations of [HIn] to [In-1. For water this becomes simply: [In-1 . 1 pH p511], + 103 [Mn] (13) where p511“ is defined by pEHIn + log 3:11! . In nonaqueous solvents film where pH does not have the significmce that it possesses in aqueous solution, the situation is@ more complex. Considering ion pairs in light of aquatics: 12 and 11 the ratio of acid to anion indicator form is (31). zfifln _ [m1 + [3315] (1;) [In 1 [In-1 but no 'simple relationship exists between this ratio and hydrogen ion concentration. By appropriate choice of indicators and acid strength, Kolthoff and his students studied acid-base equilibria in acetic acid (32), acetonitrile (33), and dimethylsulfoxide (3h) . Various indicators have been used in mania to follow neutralization 6 reactions (35) or to compare acid strengths (36) {affine weak acids. The types of indicators which show reversible eqdilibria are the phthaleins represented by phenolphthalein (6) , triphenylmethane as the sodium salt (37), azo and hydro compounds (38, 39) , nitrophenols, and nitroanilines (10,141). It is of interest to note that sodium and potassium salts (1.12) of the various nitroanilines have been isolated. This gives some credence to the existence of a nitroaniline anion. Electrolysis of an ammonia solution of m-dinitrobenzene (h3) has yielded hydrogen at the cathode. This evidence points to acid-base behaviorfsr thistype of indicator. Schattenstein (ho) reported the colors (Table I) of several of these compounds in ammonia. Qualitatively observed colors of some other indicators of various types are listed in Table II (29). Spectrophotometric studies of indicators in aqueous solution have become laboratory routine, but the difficulties; involved in handling liquefied gases have been the subject of investigation for some time. Generally, the solution was placed in metal containers or sealed tubes and studied at room temperature (M). In low temperature studies, moisture condensation on cell windows has been prevented by several means. Cells have been protected by double windows with vacuum insulation or by a stream of dry gas (115, 146). A unique light-pipe cell using this property of polymethylmethacrylate has been utilized in the visible region by Lagowski (29). Modified polarimeter tubes have been used along with additional studies which mention no experimental details (117, 118, 119). In aqueous solution, acid-base indicators are obviously affected by salts which undergo hydrolysis, and the effect of neutral salts TABLE I COLORS OF SOME SUBSTITUTED PHENOLS AND ANILINES IN AMMONIA (140) Indicator g-nitrophenol p-nitrophe nol g-nitro phenol 2 , 5-dinitrophenol 2 ,h-dinitro phenol 2 , h , 6, -trinitrophenol g-nitro aniline p-nitro aniline g-nitroaniline 2 , h-dinitroanilins 2 ,h, 6-trinitroaniline p-benzeneazoaniline p-benzeneasodimethylaniline p—tolueneazodimettwlaniline Neutral Acid Base orange orange cherryered yellow yellow red yellow yellow yellow red red precipitate yellow yellow precipitate yellow yellow precipitate yellow yellow green yellow yellow orange yellow yellow red red red -..- orange orange precipitate yellow yellow red yellow yellow violet yellow yellow violet 8 TABLE II COLORS OF VARIOUS INDICATORS IN AMMONIA (30) Indicator methyl violet thy-cl blue tropeoline 00 methyl orange bro-ophenol blue alisarin.red S bromocresol purple methyl red bruotkwmol blue neutral red turmeric phenolphthalein alkali blue alisarin.yellowIR methyl blue clayton.yellow basic fuchsin triphenyhssthans 2,h-dinitropheny1hydrisins ‘pynitrophenylhydrazine penitroacetanilide genitroacetanilide Neutral Acid Base colorless orangeebrown green blue blue-green blue-green yellow yellow violet yellow yellow red-brown purple purple redébrown purple purple colorless purple purple yellow I'llow' yellow red blue blue yellowbgreen yellow yellow blue orange red yellow pink violet colorless red-orange red-orange yellowbgreen orange yellow-orange purple red red blue-green redporange red-orange blue-green yellow orange blue colorless colorless red greenayellow greeneyellow purple orangeeyellow orange-red red colorless yellow-green. yellow yellow yellow orange 9 on these indicators has been the subject of investigation.by several people (So, 51, 52, 53). Since the equilibrium.constant for an indicator acid (equation 11) involves activities, it can be seen that the concentrations of the colored forms will change as the activity coefficient ratio changes. Activity coefficients are directly related to the ionic atmosphere in the solvent-~that is, to the ionic strength--as seen from equation 1,. Hence, if the ionic strength of the solution changes, the color of the indicator also should change. This has been observed in water with several types of indicators in sodium chloride solutions (50). Rearranging equation 12 and defining PERIn as log aH+8-—log L19:l [HIn] gives: 931:; " pEHIn+1°g In (11;) :;/Bln Not only is this equation.valid for uncharged acids, but it also applies to acids of the types film. and HIn+. Then.for the three cases, equation.lh_can be combined with the Debye-Hfickel limiting law (22) to give: 4» .. HI“ “'3‘“ 3P§dln'P§uIn'§- WT (a) a + -— HIn II B +In gpg'IIn-II pEHIn-BA fil- (b) (15) HIn+ :- H” +In ,pggh+- gran“; ,5? (c) These equations are of the nature of first approximations in that the ions are considered point charges in a medium.of uniform dielectric constant (51). They do, however, point out the effect that ionic strength has upon various types of indicators in water. Table III TABLE III SALT EFFECT CORRECTIONS FOR INDICATOR pfia VALUES FOR AQUEOUS SOLUTIONS OF VARIOUS IONIC STRENGTHS* (50) Indicator Ionic Strength methyl orange 0.02 0.01. 0.01. 0.01. methyl red 0.00 0.00 0.00 0.00 thynol blue [5,] 0.00 0.00 0.00 0.00 thymol blue [5,] -0.0h -0.11 -0.16 -0.35 chlorphenol red -o.02 -0.10 -0.15 -0.3h phenolphthalein -0.02 -0.09 -0.1h -0.30 p-nitrophenol -0.02 -0.03 -0.06 -0.22 triphervlcarbinol --- 0 .08 0 .1 7 --- *Tabulated corrections are p_K_"-- FEHIn values and are to be added to the observed p53. A negative correction indicates that the salt solution appears, from the indicator color, to be more basic than is actually the case. 11 shows a correlation between equations 12 and experimental data reported by Kolthoff (50). The nitrophenols are of the type found in equation 15; and show a negative salt effect. The sulfonephthaleins show the largest effect in accordance with 12 and tripherwlcarbinol shows a positive effect as 1_‘_5_c_:_ predicts. The behavior of the indicators which show little or no effect have been explained by Klots (52) in terms of the Debye-Hu'ckel effective radius of ionic atmosphere. These indicators in their charged forms are switterions; hence if the ionic atmosphere is large enough to shield the charges fro. each other, the attraction between them can be neglected and the salt effect shouldbe very mall. The effect of a change in temperature upon ionisation constants of both indicator acids and nonindicator acids has been studied extensively in aqueous solution and reviewed by Everett and Wynne-Jones (5h). The ionization constants of acids which are neutral molecules (e.g. acetic acid) or negatively charged (e.g. bicarbonate ion) rise with temperature up to a certain maximum, then fall as the temperature is increased further. The main factor causing the fall in ionization with rising temperature is the effect of rising temperature on the dielectric constant of the medium, which is generally to decrease it. This is important in ionic dissociations where electric charges are separated, for the greater the dielectric constant, the easier it is to separate charges. _//In an acid-base reactioninwhich there is no separation of charge.-e.g., in the ionization of positively charged acidsuthe dielectric constant is of minor importance and the equilibrium constants for such ionisations generally increase steadily with rising temperature (55, S6). ‘12 0f the various types of indicators, only a very few observations have been.made concerning temperature effects in liquid ammonia, whereas the majority of such observations have been limited to aqueous solutions. Franklin (57) noted that neutral ammonia solutions of phenolphthalein are colorless at room temperature while Lagowski (29) observed the same indicator to be violet-colored with an absorption Peak at 575 m p at ~77'C. From the scanty'information concerning temperature effects on various types of indicators in water, the following facts stand out. The azo indicators show a considerable increase in ionisation as the temperature rises. Such an indicator is of the positively charged acid type in which there is no separation of charge as the proton is released: + + R "I“ NH(CH3)’ * R - N(CH3)3 "’ H . (E) while it is true that the R.group often contains a negatively charged group at the end opposite the amino group, this charge is quite far removed from.the nitrogen atom and has little influence in restraining the removal of the proton (58). In pynitrophenol, where ionisation of the phenolic group causes a separation of charges, ionisation changes very little with temperature because of compensating effects. ‘With increasing temper- ature, the dielectric constant falls and ionisation decreases, whereas the increased thermal energy counters by increasing ionisation. This effect takes place in only one temperature region which happens to be near room temperature for this type of indicator. The phthaleins and sulfonephthaleins should be analogous, yet the 13 phthaleins show an appreciable change in ionisation with temperature, probably owing to the fact that the effect of decreasing dielectric constant is greater than the strictly thermal effect upon the ionisation constant. EXPERIMENTAL Spectrophotcmetry in liquid ammonia can be carried out either at room temperature in cells which will withstand the pressure exerted by amonia (9.89 atmospheres at 25’0) (S9), or in cooled cells maintained somewhere between the boiling point and freesing point of ammonia. Cells of this later design should be closed to the atmosphere as ammonia is hygroscopic and water changes the spectrum of may substances (60) dissolved in ammonia. The principal objection to the use of conventional cells is the formation of frost on the windows; hence some means must be devised for elimination of water from the cell faces. Since absorption spectroscopy in amonia is limited to the visible and ultraviolet parts of the spectrum due to strong absorption bands at 235 m '1 and at 1ND m p, (60) a cell can be constructed of standard borosilicate glass or some other material that is transparent to visible radiation for use in this region, or of silica for use in the ultraviolet and visible ranges of the spectrum. ABSOIU’TION CELL A low temperature spectrophotometer cell (Figures 1 and 2) was constructed around a camaercial" cylindrical fused silica cell with optically flat windows and a path length of 10.000 In. Two 7 mm inlet and outlet tubes were attached 180’ apart. This cell (A) was suspended in a closed Pyrex tube by means of graded seals (_B_) and the *American Instrument 00., Silver Springs, Maryland. 1h 5 Low Temperature Spectrophotometric Cell _ Figure 1. H G Photocell H Pizzzhromator §\s—)’ Housing N [ [r Plate ‘N “’2. B \ A fl/ E Q F g , g h I ’ . :1) {—1 V C E \ e \ bracket? \\\\\\\\\\\\\\\\\\\\\\\\\ . ”.ring . \ S stand base Figure 2. Low Temperature Spectrophotometric Cell (Schematic) 17 enclosing envelope was provided with two T-seals (Q) at 180’ along the light path (2) . The resulting cross was fastened to the inside of a 90 mm Pyrex tube in such a manner that the light path did not go through Pyrex. The entire assanbly was sealed inside a 105 mm tube to form a Dewar-type jacket. Two 16 mm tubes (32;) were sealed to the outer wall of the flask in alignment with the light path. After these were cut off and ground square on a Carborundum wheel, two circular silica windows (_F) 20 mm in diameter“ were attached with black wax. A vacuum ; stopcock (9) enabled the Dewar-space to be evacuated. This design resulted in a total span of approximately 15 cm between the outermost windows. The cell is filled through tubes (1!) extending out of the Pyrex envelope and beyond the top of the flask. These tubes connect to the amonia line with ground glass ball joints and sockets, but were closed by an interconnecting U-tube when the cell was removed from the line. The space between the double walls and the envelope containing the silica cell could be filled with a cooling liquid. In order that the contents could be seen, the flask was not silvered. The total volume of the silica cell is less than five milliliters when filled to the level of the cooling bath. Beaker clamps were used to attach the cell assembly to a ring stand modified as shown in Figure 2. SPECTROPHOTOHETER A Beckman Model DU spectrophotometer was modified to accept the .. (Figure 3) low temperature cell and was used in all ’measurements in this study\. ‘Quaracell Comparw Inc., New York 18 See 325 aceoasofiohoocm .n 9:63 ago cognhwmwg nauseonmoocox law;- 19 The monochromator was separated from.the photocell compartment and the conventional cell holder compartment was removed from the instrument. The ring stand containing the cell was visually aligned such that the light emitted from the monochromator slit passed through the cell and onto the photocell shutter. Longer bolts were constructed which passed through the photocell compar‘bnent, through the plates which were bolted on the ring. stand, and fastened to the monochromator housing. For insulation.purposes, Bakelite plates were placed between thegmonochromator housing and the ring stand plate as well as between the opposite ring stand plate and the photocell compartment. Dowel pins and holes insured reproducible positioning of the low temperature cell with respect to the monochromator and the photocell. A light-tight bag constructed of black poplin and rubberised cloth covered.the cell during operation. Drawstrings were used to close the two side arms and the bottom. The ring stand, plates, insulating plates, etc., were painted black to eliminate any reflection of light. A tungsten lamp was used in the region of the spectrum from 320 m p to 1000 m p and a violet filter was added for measurements from 320 to 1400 m F. For the 230 m p to 350 m ,1 region ultraviolet radiation was obtained from a hydrogen lamp. in electronic power supply. (mm-noon 11033700) was used. This was set in position five when the tungsten lamp was used and in position seven‘with the hydrogen lamp. The monochromator slit widths used with these photomultiplier settings varied between 0.800 mm and 0.030 mm. To establish the thermal equilibrium necessary for reproducible instrument readings within an.hour,the lamp housing was cooled by circulating tap water. Since this instrument'works on a single beam.principle, it was necessary to fill the low temperature cell with pure liquid ammonia 20 and to determine the slit width opening for 100$ transmittancy at each wavelength used. These values were recorded at comtant sensiti- .vityb Average'sof three readings per wavelength were used as the slit width program. These readings were checked periodically and gave results within the precision of the instrument (1 1%). The effective path length of the cell was determined with a 0.1 1! solution of cobalt anunonium sulfate in water according to a method given by Hellor (61). Pure water was employed to calibrate the slit width and then was replaced with the standard solution. The path length, l, was found using Beer's law from the expression A l . _S__- W x 1.000 (11) 0.17h2 where 5(0138) is the observed absorbancy of the cobalt solution at 510 m P and 0.17h2 is the known absorbancy in a 1.000 on cell. The effective path length was found to be a consistent value of 1.00 on during this investigation. Volumetric and Dilution Vessels For preparing an ammonia solution of known concentration, a one liter volumetric flask (A) was modified by fastening two side arms (_B_) on the upper neck and joining a 2h/h0 ground glass joint (9) to the top (Figure 1:). A three way stopcock (Q) was placed on a long glass tube which reached to the bottom of the volumetric flask. With the tube in place, the volume of the flask was determined from the weight of water required to fill it to a reference mark. A weighed amount of solute was placed in the flask and ammonia condensed up to the mark 21 To dilution flask NH; am I; I6 Io Na SEW To exhaust +2— reference mark Figure h. Volumetric Flask 22 by immcrsion of the entire vessel in a large Dewar flask filled with coolant. The resulting solution was essentially anhydrous as the inlet tube was not withdrawn and exposed to moisture present in the atmosphere. Two dilution vessels similar to the one used by Lagowski (29) were used with slight modification. The first vessel consisted of two 200 m1 round-bottom flasks connected at their necks such that a reference mark (A) was between them. A short section of a 50 ml graduated cylinder (B) was placed on the opposite end which acted as the neck of the vessel. A 2h/h0 standard taper ground glass joint (9) was placed on top and a 111/35 standard taper ground glass joint (13) holding an addition pistol was sealed on at an angle at the side of the neck. A 7 mm glass tube (E) was connected on the opposite side of the neck (Figure 5). The joint at the top of the flask accepted the upper piece (2:) which carried a three-way stopcock as well as inlet and outlet tubes. The longest of the two tubes (9) extended to the bottom of the flask and was used to condense ammonia, introduce solution from the volumetric clask, withdraw solution to the spectrophotometer cell, and stir the solution by allowing dry nitrogen to pass through it. The shorter tube (g) extended only to the reference mark between the bulbs and was connected to the aspirator; through it solution could be withdrawn to the intermediate mark. Solid samples were introduced into the flask by means of the addition pistol (_I_).. The other outlet was used as a vent to the exhaust manifold and also as a means of applying pressure to force solutions out of the flask. The second vessel used was identical except that the bottom bulb had a volume of only 50 ml (Figure 6). The flask was used when a 23 NH; andNa T° cell To aspirator .F. a To exhaust 2 g ll]! 200 ml vessel E cf reference mark A 200 m1 vessel L , Figure 5. Dilution Flask lb I0 a P. 200 ml vessel C::_g reference mark _A_ .mnumi anenmmumk 25 dilution factor greater than one-half was desired. These flasks were calibrated gravimetrically using distilled water with the glass tubing installed. By removing solution down to the middle reference marks, dilutions of approximately one-half and one-fifth could be made by addition of ammonia from the volumetric flask or by Condensing ammonia to the proper reference mark on the neck. This method insured an essentially anhydrous system as the tubing did not have to be removed to make a measurement. Thermostats All measurements were carried out at constant temperatures with refrigerant baths of a suitable liquid cooled with Dry Ice. These mixtures maintained a given temperature when the Dry Ice was in excess. The cooling compartment built around the spectrophotometer cell was filled with the refrigerant mixture at all times. Because this cell was unsilvered, frost slowly formed on the outer walls of the vessel owing to radiant energy transfer, but any mnuber of runs could be made with no appreciable change in temperature (I 0.90). Using this system,frost did not appear either on the outer silica windows or on the windows of the inside cell itself. The majority of the work was done at -77°C. This temperature was maintained with methyl Cellosolve (ethylene glycol monometh ether) as the heat transfer medium. For the other temperatures at which measure- ments were made, a series of organic liquids were used at their freezing points. These are listed in Table IV along with the temperature at which solid is in equilibrium with liquid. 26 TABLE IV mourns ‘USED FOR consnm' TEMPERATURE mmmms (”Slush Baths") Liquid Freezing Point, ('0) Ethylene dichloride -3S.6'c Chlorobenzene 4:5 .2°c ‘ Chloroform -63.S‘C Hethyl Cellosolve -77 .0‘6 A typical constant temperature bath was made up by adding powdered Dry Ice to the liquid until a thick slurry was present in the refrigerant portion of the cell. Addition of excess Dry Ice to the liquids was not necessary and proved to be undesirable. Temperature measurements were made by mans of a temperature dependent resistor (thermistor? which was inserted into a thermowell compartment which extended into the Pyrex envelope in proximity to the spectrophotometric cell. This well was sealed out of contact with the refrigerant by means of black wax. The thermistor was connected to a Wheatstone bridge as one leg of the resistance (Finn 7). The actual measurements were performed by balancing the two legs of the bridge to a null point by mans of a ten-turn Helipot“ resistor and recording the resistance needed to balance the bridge at a given temperature. The null point indicating device used was a Pyem *Victory Engineering Corp. , Union, N.J. "Bookmanlnstruments, Inc. , Fullerton, Calif. H. G. Pye 00., Cambridge, England. 27 dab-IOU» 1000 ohms 10,000 ehne Varible aeeietcr ' 5000 ohm Thermistor (m0 358) Pye Optical Galvanometer Dry Cell Battery (6.0 volts) Figure 7. Vheatstone Bridge 28 galvanometer. A 5000 ohm thermistor was found to have the correct resistance for tanperatures between the boiling point and freezing point of ammonia. The thermistor .was calibrated (Figure 8) by inersion in a liquid-solid slush bath of known temperature. When measurements were being made with the cell containing slush baths rather than excess Dry Ice, the variation in resistance was 1’ 55 ohms or t 1.2‘0. Other cooling baths employed in condensation and transfer of the ammonia solutions were contained in large mouth Dewar flasks filled with Dry Ice-metivl Cellosolve mixture maintained at ~77’0. Ammonia System The system used to handle the ammonia consisted of an inlet manifold and an exhaust manifold (Figure 9). The inlet manifold was closed and evacuated by means of a stapcock 2. Synthetic ammonia from a tank was introduced at _B_ while dry nitrogen was connected at Q. The open-ended manometer at _A_ was used to determine the pressure of the inlet gases. Stopcock at E was usually in the off position such that D and E in the open position allowed the gases to pass through 9, the drying chamber. This chamber was half filled with ammonia in which was dissolved sodium metal; it was cooled by a Cellosolve--Dry Ice mixture in a Dewar flask. The amonia and nitrogen were passed through this solution at a steady rate which insured anhydrous conditions. Stopcock E was opened when the drying chamber entrance tube became clogged with sodium snide in the middle of a run. The valve at g was the outlet to the exhaust manifold. A stopcock at g was connected to the three way valve 2: 29 33.20989 among 083363 mowed—Gosh .o Pang Asimov eognwnom mom? mSm moot I e a q . ~ - e u - on: 0m: 3: om: o, exhaustive}, scream mimosa 333 .e 9:62 31 while I was connected to the upper neck tube on the flask. Stopcock 5 connected the exhaust manifold to the upper neck tube on flask g. Three-way stopcocks E, _Q, and g were connected to the dilution flask Q and were used in conjunction with the volumetric flask g and the spectrophotometer cell 2. Valve 2 was connected to valve _L_ by means of silicone rubber tubing’ which maintains its elasticity at -77'C. The three-way stopcock at g was used as a gas inlet from the inlet manifold and also as an outlet to the exhaust manifold. Valve 9 connected the short tubing extending into the dilution flask to the aspirator for removing solution down to the reference mark between the bulbs. This stopcock was also used to introduce ammonia into the flask; the gas was condensed to the upper reference marks on the neck. Three-way stopcock g was used to introduce liquid ammonia from the dilution flask into the cell T, and also as an inlet to the manifold for the cell. Valve E1. was the cell exhaust connection. The one-way valve at V prevented water from backing up into the system if the internal pressure dropped below atmospheric pressure during condensation of the mmonia. Tw0 large carboys, LI and _I_, were used as traps for getting rid of excess mania and as ballast for the pressure in the system. The carboy 5 was filled with water and connected to E which contained the gas present in the system. Thus if the pressure decreased, the water would flow from _I_ to E and not be forced into the ammonia system. The outlet frm E was vented into a hood. The water~ in I also served as a ‘ ”flow meter” to show how much nitrogen was being passed through the system. All flasks and vessels were connected to the system by means of Tygon or silicone rubber tubing and ground glass ball Joints which alleviated aw strain between the flasks and the manifold and prevented *Dow Corning Corporation, Midland, Michigan 32 breakage. Flasks _Q_ and g were cooled by Dewar vessels containing methyl Cellosolve-g-Dry Ice refrigerant. . Use of Vessels and Absorption Cell The system described above made it possible to perform operations of condeming, stirring, diluting, filling, etc. , in an essentially anhydrous system and to effect quantitative transfer of amonia solution from one vessel to another. A typical run included making up a standard solution, dilution of this solution by a known amount, filling the spectrophotometer cell, and taking readings on the Beckman DU. These operations were carried out as follows. The inlet manifold was sealed by closing the stopcocks and evacuating the system with an oil vacuum pmp connected at I_J_. Mean- while a solution of sodium metal in amonia was prepared in the drying chamber 0 by adding liquid ammonia from a siphon tank to sodium metal placed in the chamber. This solution was cooled in a Dewar flask and connected by means of ball and socket Joints at _D and E. The nitrogen ‘valve was slowly opened after valve 2 was closed and _I_)_ and {were opened. As nitrogen passed through the drying solution, {1 was opened and the system flushed with dry nitrogen. After the volumetric flask had been heated in a drying oven at 110‘!) for two hours, a weighed mount of solute was quickly introduced into this flask and the flask and contents were connected to the system at _I_, l, and 5. Valve E was turned such that when E was opened, it allowed nitrogen to sweep through the flask. Valve g was closed; 5 was allowed to cool to room temperature while being flushed with nitrogen. The flask was then cooled to -77’C by surrounding it with a Dewar flask filled with 33 refrigerant. Ammonia from tank 2 was introduced and began to condense immediately. As the level of the liquid reached the reference mark, 11 was opened, the ammonia turned off, and the level checked. when this indicated one liter of solution, .1 was closed and .L. rotated to line up with .13- In order to stir the solution and hasten equilibrium, nitrogen was bubbled through the solution at a vigorous rate. The dilution vessel _Q_ was dried in an even and attached at g, Q, and 2. It was cooled in the same manner as was flask g. Valve 1 was opened to nitrogen pressure which forced the solution from g through 12. and E into _Q_. As the. reference mark was reached, it was opened and I. was closed. The solution in the dilution flask was used to fill the spectre- photometer cell in a like manner. with 9 off and g rotated, valve 2 was aligned with § to allow nitrogen pressure to force the liquid through I. When the cell was full, nitrogen pressure was released by opening g. With g and 5 closed, the cell was disconnected; The closed U-tube was placed over the opening and the entire assembly attached to the spectrophotometer. Dilutions were carried out by aligning stopcock Q with the aspirator comm ction which emptied the dilution vessel to the middle referencemark. Antonia was condensed in the flask to the upper reference mark and nitrogen was bubbled through the solution to stir it. The dilution flask can be used in conjunction with the volumetric flask to change the concentration of one of the species in solution while concentrations of other solutes are held constant. Solute on be added to the dilution flask by means of the addition pistol. The volume of the first dilution flask to the middle reference mark was 207.9 ml with the glass tubes installed. The upper reference 3h used gave a volume of 1318.? ml which resulted in a dilution factor of 0.h96. The second dilution vessel had a volume of 5h.90 n1 when filled to the lower reference mark and a total volume of 258.0 ml. Its dilution factor was 0.211. Not only could pure ammonia be condensed in the dilution flask, but a known solution could be added to the mark from the volumetric flask as described above. Thus, all possible combinations of indicator, neutral salt, and ammonium salt concentrations could be varied or held constant with respect to each other. Chemicals Synthetic amonia*was used from two types of tanks. One was a siphon tube tank which delivered the liquid at the spout and was used in all synthetic work undertaken in this study when ammonia was used as a solvent. The other type was a regular gas delivery tank from which ammonia was distilled into the proper flask after passing through a sodimn-ammonia drying solution. Purified dry nitrogen“ was further dried in the same manner as the monia prior to use. All inorganic salts employed were reagent grade, dried at 110°C for several days, and stored in desiccators containing Drierite or anhydrous magnesium perchlorate. Organic indicators were recrystallized from the appropriate solvents to constant melting points and stored in desiccators until used. Experimental details for preparation of other organic compounds used in this stuw are given in Appendix I. *netheeon, Inc., Joliet, Illinois. "General Dynamics Corporation. 35 Sodium Salt of Phenolphthalein A reaction vessel for the preparation of alkali metal salts in amonia was constructed by sealing a coarse porosity sintered glass frit to the bottom of a 300 m1 three-necked flask. A stopcock was ' ettechad below the frit (Figure 10). A etirrer was fitted in the center neck along with an addition pistol and drying tube in the other two openings. One hundred milliliters of ammonia was introduced into this flask from a siphon tank. With the bottom stopcock closed, the ammonia remained above the glass frit. Phenolphthalein [11.5 grams (1.1; x 10'3§)] was dissolved and upon addition of 0.055 gram (1 .h x 10‘3y_) of sodium enide the red-purple color intensified. After a few minutes of stirring , the lower stopcock was opened into a 250 ml Erleuneyer flask end the stirrer, drying tube, end pistol ell were replaced with stoppers. As the ammonia evaporated, the pressure increase in the flask forced the solution to filter rapidly through the glass frit. Since the sodium salt is soluble in amonia, the filtrate was evaporated to dryness in a stream of dry nitrogen and the salt recovered in quantitative yield (5.0 grns). 36 Ci. 300.1 flask ("2 coarse frit Figure 10. Filter Flask REULTS AND DISCUSSION Introduction Compounds which contain a labile Ivdrogen can act as acids toward ammonia even though they are bases in water: m+mi,ea‘+un.* (1g) Acid1 Base 2 Base1 Acid 2 If a color change occurs when Acid 1 becomes Base 1, the substance can be used as an indicator in ammonia. The concentration of Base 1 depends upon the equilibrium constant of equation 1_§, which in turn is dependent on the acidic nature of the compound and the basic nature of the solvent. For the general acid-base reaction A + s a B" + H; (12) the ionization constant is is" as; g - —— - K K (20) 1 91A. ‘ -(acidity) A —(basicity)8 — —-a where £(acidity) of any protonic acid (i.e. RE) is: -e + 1 "B - * (31) 31A §(basicity) 5(acidity) - Thus equation _2_0 shows the role of the basicity of the solvent in determining the ionization of an acidic compound. The extent to which 37 38 this compound ionizes is then a measure of its acid strength. Because of the difficulties in evaluating the constants £(acidity) and §(basicity) present in equation 29, indirect methods must be used to evaluate acid strengths. If the ionization constants of two acids A; and A, are compared in ammonia, then from equation.gQ: (Ki):l . £(acidity)A1 (22) (51)? 5-( acidity) A2 where the basicity of the solvent cancels out. Comparison of two acids in one solvent therefore should give the same ratio of (Ei)1 to (Ed)3 as in any other solvent. Equation 29 does not describe completely the acid-base equilibrium as there is not only the ionization of the species to consider but its dissociation also. These steps can be represented by two separate equilibria: RH + We * nun." (22) 113an a: R" + rm,+ (all) where equation g;_is the ionization step and equation.gh is the dissociation step. In solvents of high dielectric constant the dissociation step may be essentially complete, but in ammonia with its dialectic constant of about 23, the distinction is important. Because no further ionization can occur when an ionic crystal dissolves, dissociation is the only dissolution reaction which ionic substances undergo. Dissociation in solutions of many ionic compounds 39 has been studied conductometrically (3, 1S), and Ed has been found to have an approximately constant value of 10'h for one-one electrolytes. Hence in ammonia ion pairs predominate except in very dilute solutions of the order of 10‘3 g,or less. Higher aggregates such as ion triplets and quadruplets also exist (62), but only at greater concentrations than those encountered in this study. Conductivity--like electromotive force measurements--cannot be used to obtain ionization constants of acids because it does not distinguish between ion pairs and molecules. The intrinsic acidity of a compound can be determined by its ionization constant and can be compared with other compounds in this respect. Many organic compounds which act as indicators in aqueous solutions also form colored solutions in ammonia. Since these indicators are acids in water, their acidity is enhanced when dissolved in ammonia. Thus most indicators show little or no change in going from acid to neutral ammonia solutions while many more display a change from neutral to basic solution. Several of the more promising indicators, i;g; ones showing a definite acid to neutral change, are listed in Table V together with their color changes. or the various experimental methods available for studying solutions, absorption spectroscopy seems to be best suited for investigating indicator systems. Indicator color changes afford easy to man” INDICATOR COLOR CHAMES IN mm Indicator Acid Neutral Base p—Nitroacetanilide colorless yellow A yellow 2,h-n1 nltroaniline orange red yellow Phenolphthalein pink purple colorless Thymol Blue blue blue-green blue-green Trapeoline 00 yellow yellow purple access to very useful data; the molecular form of the indicator acid absorbs at a wavelength different fraa o. ionized fora, an! spa ctrophotonetric measurements permit calculation of ionisation constants. ' There are several things to consider in applying this method to mania. Even though the effects of salvation upon the electronic energy levels of a molecule are not well known, it has been assumed that the extinction coefficient of an ionic species in a solvent such as a-Ionia is the same whether the ions exist as ion pairs or as discrete ions. If Beer's Law is valid, the expression for the absorbancy due to a given indicator anion at a single wavelength for unit path length is l-eg,+ e9, (.22) where 5 is the absorbancy, e is the molar extinction coefficient (assumed to be the same for associated and dissociated ions), 9,, is the concentration of indicator anion pairs, and g, is the concentration m of dissociated anions. Factoring equation gé gives: A - as + 9.2) (2.6.) Solving for the total concentration (91 + 93) gives the expression used to calculate anion concentrations: (91 4' Ca) ' " (27) Equation 21 assumes the spectrum of ion.pairs to be the same as that of free ions; that is, the spectral contribution of free (dissociated) solvated ions cannot be distinguished from the ion pair absorption. Ammonia is so basic that even very weak proton donors are converted to ammonium ions, the strongest acid species that can exist in liquid ammonia solution. This phenomenon is illustrated by acetic acid, a weak acid in water but a strong (i.e., completely ionized) acid in.ammonia. The indicator acids studied in this research were found to have first ionization constants greater than one. These indicators, therefore, are strong acids in liquid ammonia. The color changes of acid-base indicators in.liquid ammonia do not have the same significance as in aqueous solutions. In water the ratio of acid form to base form is directly related to the activity of the hydrogen ion. When a pH measurement is made in aqueous solution, the difference between the value of an experimental parameter in unknown and reference solutions is attributed entirely to a change in activity of the hydrogen ion. The activity is assumed to be h2 unaltered by exchange of one anion for another. In water, these assumptions are valid as first approximations, but when applied to ammonia, the effect of ion pairing must be considered. Hydrogen ion activity cannot be measured directly as all experimental methods give mean activity rather than the activity of a single ion. Since Fuoss and Kraus (15) have demonstrated experimentally that -h dissociation constants are of the order of 10 for salts in ammonia, and since they have shown that dissociation constants vary with the anion present, it follows that a correlation of the type made in aqueous solution is not valid in ammonia, In liquid ammonia no single relationship holds between the concentration or activity of the ammonium ion and the ratio (:9 of indicator anion to molecular indicator. This investigation.shows that plots of the logarithm E 125223 the logarithm of the ammonium ion concentration do not give straight lines (Figures 11 through 1h). These figures also show that the functions are highly dependent on the nature of the anion.present. 'Weak Acids According to the Franklin concept (6) of the nitrogen system of compounds, guanidine is the analog of carbonic acid while urea is the mixed aqua-ammono compound. 8 f’ 'i'.“ no -C-0H Han-é-m'la Halt-C -NH3 carbonic acid urea guanidine Carbonic acid is a weak acid in water, urea is weakly basic, and guanidine is«a strong base in this medium. In ammonia, on the other to hand, it has been postulated that both urea and guanidine are weak 133 fine on I gonna 2E no a no." some: ensuoamaom Bane! no 3.355300 no man . p... one»: 68.5% w I D .W o.o. _ 3.0.. .1801 «.0: «.0 aegannaononm no M won eoaue> soda—ohm enamel-a Ho m3 9935230 we won . up shaman a n u p q d J ‘ 1‘ CW 1 l . 4.0: s L No? l m. a _J l o I «.0 . 40° J 1:5 floqganoaaofi no m. m3 neon: oHooH leases—.4 no someone—noose". no men 1M Hogan my N .2 9:63 3.0.0.. . «.o 1:6 anon-5a H e o h on make sash: I33 3e neon meg 3m . a on M m b a o! no no .3 00 Ho gong—um . a m :2. Eur.— so. so. NCO-I .J «.0 h? acids, and indeed from the structure of these compounds, amphiprotic behavior might reasonably be expected: snafu!“ (gs) minions; (_2_9) 0 NH ’ where R is the Bali-EL or Hall-(L group. Evidence of the acidity of these two compounds in ammonia is that: (1) potassium salts can be prepared using potassium amide in ammonia, and (2) hydrogen is evolved when elemental potassium is used in the same reaction (61:, 65). The spectra of urea and metrvlsubstituted ureas in water show an absorption band in the far ultraviolet which shifts to longer wavelengths as the number of methyl groups substituted for hydrogens is increased (Figure 15). Because ammonia absorption bands occur in the region where these compounds absorb, direct spectrophotometric study of these compounds in solution is blocked. Indicators therefore have been used to investigate the acid-base properties of substituted ureas. In a preliminary study, several indicators were selected on the basis of the difference in color which they display in acid and neutral monia solution. The simplest of these indicators is p-nitroacetanilide. Its yellow-green color in pure liquid anaemia is greatly intensified by the addition of potassium amide. p-Nitroacet- anilide contains one acidic hydrogen atom in the molecule and presumably reacts with ammonia according to equation ]_8_: 9 9 H» lé-CH; .u/éws 1'01: H0. 335 m ,1 absorption “:6 m p absorption .35 5 825 35333333. 88 mo snooze 33.8% 593E a a .geaerrz he o:« 09». o~« 03 oo« Q 1 1 I I d . q . Q D‘l t I l i a .36 . L . s ‘ ”0° J o l «J l seamen: asoaaofisoosoo seuaHona—ehvea u o _ I 0., «toaafioflnnai . o some—”have: u a _ .. 1. o.~ foueqxosqv 1:9 The ultraviolet and visible spectrum of this indicator (Figure 16) shows bands at 335 I p and M6 3 p which are attributed to the molecular acid species and the amoniun ion pair respectively. Amoniun perchlorate added to the solution depresses the band at 34146 n p and increases the 335 n ,1 band as would be expected from the equilibrium indicated in equation 3. (Figures 17 through 19). 51 for this reaction was calculated (Table VI) by the method described in Appendix II and found to be 17. Such a large value of the equilibrium constant is not surprising in a nediun as basic as amonia where acidity is . ' enhanced and ionization is extensive. The extinction coefficients (Table VI) for molecular and ionized p-nitroacetanilide were calculated by the ndhod described in Appendix II ; the greater value for the molecular" fora eXplains ulv the absorption bands approach the same size even though the equilibrium favors the monium-anion pair. By analogy to the behavior. of phenols in water solutions, p—nitroacetanilide should have a rather large salt effect and this is observed. (Figure 20). his can be explained in terns of the theraody-aaic equilibrium constant. The ionization constant for an indicator acid is ‘ [33111-1 99-3111" K - -—-——— -— (31) _i [HID] 3 7m 01‘ x - x‘fit’h- (32) -i -i 711m E1 is the concentration ratio. If the ionic atmosphere changes- as by addition of a neutral salt--the activity coefficient factor in where SO easel: a.“ ouaageoeohfizum no «50 cam 5598.54 .3 shaman & a . 5394.053 om; 8.. L omn 8n om... . - J «E who I'll / \ .3338 dbpoz ‘ .. / 1K\ mauo. u a; . toeeoHEH .3 0.0 New 0.. o.~ :4. Louvqaoeqv 1&3 flegfiamconm no M no.» menu: 3.03388 532.4 no nofigaeonoo no mod . up one»: 665 a m N P - Ill-J J. J a q i d J . a L 000! J 4:1? m. a .J «.0: L0 aegpnnaoaeam no M m3 unuueb e355 5.38:: no n3 #935300 no m3 .wp shrug N p ‘93 n l “‘0' + .1301 1.5 goinaamaoaosm no M m3 even; 0363 5303—4 no 8393538 no won .9 shaman 35% a n . . u p d d‘ q q 1 ‘ 1 1.0.0.. ‘1. 14—0? M. 8 J I «00' — l .I o l .. ~.o aognnaaonm no M .moA 25.8» 393.3 1332.5 no nofiufihooaoo Mo mod 4:. 99m: 1:6 .62..sz , a n . a P .. fl 1 - I q q , 1 * L . L, l 000! 1 3... l m. 8 1 «6:... . L o .. «.0 h? acids, and indeed from the structure of these compounds, mphiprotic behavior might reasonably be expected: an 1! Run“ (.2_a_) RH +11+ I! RH; (32) O NH ’ where R is the Halli!- or Han-H- group. Evidence of the acidity of these two compounds in amonia is that: (1) potassium salts can be prepared using potassium snide in mania, and (2) hydrogen is evolved when elemental potassiun is used in the same reaction (61;, 65). The spectra of urea and metlwlsubstituted ureas in water show an absorption band in the far ultraviolet which shifts to longer wavelengths as the number of methyl groups substituted for hydrogens is increased (Figure 15). Because ammonia absorption bands occur- in the region where these conpounde absorb, direct spectrophotometric study of these compounds in solution is blocked. Indicators therefore have been used to investigate the acid-base properties of substituted ureas. In a preliminary study, several indicators were selected on the basis or the difference in color which they displq in acid and neutral mania solution. The simplest of these indicators is p—nitroacetanilide. Its yellow-green color in pure liquid amonia is greatly intensified by the addition of potassium amide. p—Nitroacet- anilide contains one acidic hydrogen atom in the molecule and presumably reacts with ammonia according to equation j_8_z 9 .° 11‘“ l&-CH; -u/é'CH3 4- NHa * 3 . NH: (29) “on 302 335 n ,3 absorption M6 m ,1 absorption h8 new: 5 825 Epipfifiabzom as» no .38& «Santana .m— 9:53 5552: unofivwnpseosoo eeuwHPaeasuaea n o .fifihfiofin...z.z .. p schnahfiez .- e 03 k a afibaeaeblp CNN 05 .oou .2. ed «.- o; o.~ lousqaoeqv 1:9 The ultraviolet and visible spectrum of this indicator (Figure 16) shows bands at 335 m p and 1046 m p which are attributed to the molecular acid species and the ammonium ion pair respectively. Anonium perchlorate added to the solution depresses the band at M6 m p and increases the 335 m p band as would be expected from the equilibrium indicated in equation L2”. (Figures 17 through 19). £1 for this reaction was calculated (Table VI) by the method described in Appendix II and found to be 17. Such a large value of the equilibrium constant is not surprising in a medium as basic as amonia where acidity is - ' erflxanced and ionization is extensive. The extinction coefficients (Table VI) for molecular and ionized p-nitroacetanilide were calculated by the mdhod described in Appendix II 3 the greater value for the molecular“ form explains wtw the absorption bands approach the same size even though the equilibrium favors the monium-anion pair. By analog to the behavior, of phenols in water solutions, p-nitroacetanilide should have a rather large salt effect and this is observed. (Figure 20). This can be explained in term of the thermodynamic equilibrium constant. The ionisation constant for an indicator acid is ' [1143111.] 79311:“ K - —— --——-- (31) —1 [H111] . 7m 01‘ x x" 79,111- I (32) -1 " 47;: where 5;. is the concentration ratio. If the ionic atmosphere changes-— as by addition of a neutral salt--the activity coefficient factor in 50 3:25 5 ocflaésooaofiszh no «at a... 539334 .2 93mg om... a a «findedebez co: 0mm 8n . 0mm «an .56 'Il nosaflom whee. Ni: x a; .. Lassa H :6 m6 N; 0.9 o.~ a." Lousqaosqv S1 1 .- mmm we evaageoeouflzum Ho deflebneonoo anon; 359.334 .5 enema.— mmop M engageoeohfizum no define—"830 8 m. on m o la a . . - q , ..\|4 .e\.\ . .\ .. \ i .\e. i .\ .\ .15 .\Q .N; .\ if b\ . e\ If .\ . o.~ 0\ 6ng mtoll.‘ .. .1 fauvqmsqv 52 1 I 0.3 as gadgeoeohdzh no 33933300 henna; 359.323 .2 earn mac. n .3338. Bough no uoafibfionoo m N a a Q Q I u d 1 L 11.0 fia I a ILAeO A: .‘ ‘QOO \. l \ \ ‘ ”CO H \ew «an m to III. \ 8358 338: III 1 busqxosqv 53 I -i he a I l j 0 1 I 2 Concentration of p-Nitroacetanilide x 101! Figure 19 . Log : of 5i m Concentration of p—Nitroacetanilide at M6 m p Sh TABLE‘VI IONIZATION CONSTANTS AND EXTINCTION COEFFICIENTS FOR VARIOUS INDICATORS IN AMMONIA AT -77‘C A _I' Indicator Ionisation Constant Extinction Coefficient ' - - Wavelength , Step 51 m p p-Nitroacetanilide K1 17.0 1.07 x 101‘ 335 , ms x 103 11116 2,11-Dinitroacetanilide K; '1.59 1.80 x 10h 385 x, 8.3 x 10’3 1.55 x 101‘ 535 Phenolphthalein 1:, oo 5 .86 x 10h 575 x, 211.0 Thylol Blue x, 00 1.56 x 10" 301 x, ' 0.1.2 1.25 x 101‘ 390 x. 11.26 x 10“3 7.60 x 105 620 Tropeoline oo 6.67 x 10“ 276 3.05 x 105 1160 1.57 x 105 S90 5'5 _ 330.398." luau-3o." Ho soaaeavneonoo 333 3383qu Ho hunenuoefi .81 ,. shaman $3.365 .oGu no 83.5888 n a ma III. a a 83 mi: a 3.. .. usuaefi fousqaosqv 56 equation.2g_decreases. As a result,.§i must increase in order to maintain the equilibrium constant. The ionization of the indicator therefore increases; thus the concentration of ammonium.ion pairs and free ions increases while the concentration of unionized acid decreases. The reaction can be pictured as one in.uhich each ion or ion pair in the solution is surrounded by an ionic atmosphere arising in the following manner. Consider a small volume element surrounding an indicator ion at a given point in the solvent. This volume element should have a net positive or negative charge, depending on the charge character of the central ion, and the probability of finding ions of opposite charge in the space surrounding a given ion is greater than the probability of finding ions of the same sign. Charge density in the ionic atmosphere decreases outward from the central ion and the net charge of the entire atmosphere is equal in magnitude but opposite in sign to the charge of the ion (65). J The ionic atmosphere is changed by addition of neutral salts. As the neutral salt density increases around an ammonium ion--indicator anion pair, it becomes increasingly favorable for dissociation to take place. Then in order for the equilibrium governing the indicator acid to be maintained, ionization of the acid molecule must increase and the absorbancy in turn changes. At the same time the charge density of the ionic atmosphere becomes greater in the solution and the potential 37 energy needed to separate charges becomes less. Thus, the sane effect is achieved as results from increasing the dielectric constant of the solvent. At about 0.111 neutral salt, the absorbancy of p-nitroacetanilide at M6 m p goes through a maximum and then begins to decrease at higher concentrations. At salt concentrations above those at which the Debye-Hfl'ckel expression (equation g) no longer holds (i.e. above approximately 0.1g), the activity coefficients increase after passing through a minimum. Hence is; must then decrease. At these concentrations, the Onsager extension (214) must be applied in order to be in agreement with the observed data. 2,14-Dinitroaniline , an mono analog of phenol, has more than one replaceable hydrogen. The monoalkali metal salts of this compound have been prepared (h3) but the dialkali metal salts have not been isolated. A neutral ammonia solution of 2 ,h-dinitroaniline is red, an acidic solution is orange, and a basic solution is yellow. The spectrum: of a neutral solution of this indicator shows bands at 385 m P and 536 m ’1. The tailing off of the 385 m )1 band into the visible region along with the decrease of the 536 m P band upon addition of an ammonium compound results in the observed orange color in acidic medium (Figure 21). Also a new band arises in acidic solution at 350 m p which masks the 385 m ,2 band above 0.02}: in monium ion. The 385 I uband persists in basic solution and may be attributed to the doubly-charged anion. The reactions mich dinitroaniline are believed to undergo, together with the band assigments, are: Hm,“ N02 5, no2 + NH: * " m (2) 350m}: 536m}: 58 com a .Eolwifl 335.0333; go 8335 83.55. . .8 26E 1 .- .fihsedersm O . mm mom we“, m2 Jomn .Bn \Ol/O fl NR1 //. \.l / ./ v \ «ma— nta I'll / 68.8. mic . . //I .3358 138a ‘ mmk: n ax - Basses «.0 , 0.0 o; 4.. . COP NJ Asusqaosqv ‘ 59 H " N’ ' N no2 _IS; noa 4» Mia t + NH: (3h) n 3 N a 536 m p 385 m p The extinction coefficients for these species and the ionization constants for these reactions were calculated by the method of Appendix II from data of Figures 22-25. Results are reported in Table VI (page Sh). E; for equation.;2 shows that dinitroaniline is a strong acid and is largely converted to its red anion by the basic solvent. The small value of the equilibrium constant for the second ionization step (8.3 x 10-3) and the fact that the extinction coefficients are nearly the same lends support to the absorption band assignments. The salt effect upon 2,h-dinitroaniline is quite marked (Figure 26) as is eXpected. Because the ionic strength of the solution affects the absorbancy of this indicator, the total ion concentration (as a summation of concentrations of both ion pairs and solvated ions) was kept constant. Thus any change in the absorbancy was due to ammonium ions (Figure 27). In these solutions of constant ionic concentration, it was assumed.as-a first approximation that the dissociation constants of potassium salts are the same.as the dissociation constants of the ammonium salts owing to the similarities of the cation sizes and charges. This assumption is supported by the conductivity work of Fuoss and Kraus (15). The spectrum of phenolphthalein in ammonia is characterized by'a single absorption band at 575 m p (Figure 28) and is similar to the spectrum obtained from this indicator in basic water solution which has l a 0mm as 05.350533 no soaaefisoosoo lmsmuob hosenuomnm . «w 95mg We. a ofinfiohan no 83.5888 cm or up - . . J4 q . 0 . w c J O ‘ ”CO I .Nop w; Ian sqmo sqv 61 k a man a.- ofidfi-ohgn no 5339:0300 aw homenuoend . nu sash We. a 8:23.3an no 835888 ON or Np w a o d ‘ I 4 1 fi I 4. Li 0 ' :00 L ' Q00 1 New .» ‘ 00F «Eu mic III. A noapnaom .3582 4 o.~ Aousqaosqv 62 '3 .0 -2.0 . 0 - 1.0 .. . 1 ' 1 1‘ I 4 ‘1 '1 L 1 , I o h . 8 : 12 16 20 Concentration of Dinitroaniline x 105! Figure 21.. Log I; Versus ,Ooncentration of Dinitroaniline at 536 m P 63 44.0 k I: ‘300 I- on .3 -200 b 4 A l a j 1 a 1 1 l o 2 h 6 8 1o Concentration of Dinitroaniline x 105! Figure 25. Log K}, Versus Concentration of Dinitroaniline at 385 m p. 3.93393." linemen no soapsupsoosoo mosh; 331930.339 no hoses—hound . cm 93»: Abe—Acme .38 «o catsusupo m.o zoo Moo woo —.o 0 l1 . n m I J 1 u i 1 q - Q ”00 m6 .— a a For New .. m a «mm L n; Imlop H mmfi. .- Reagan—u Asmqmsqv sodvgseosoo odsoH .5338 we 3.93.3090.“ 53053 some: efidfiofign no hoses—.334 .5 0.33pm mac. n .6845 go 83.3838 65 m a” m a P o q - u q . 4 q 4, m . -. . 0°. La! Le! a w c e m a .3. m .b n d1 l ‘ mmm fl ,. oeP a .- 0mm 6 Imsop H m.m .- 93338” 3:85 fl fiodfifiaoaoé no Shear. cage}: .8 9:52 .. 0mm 1. l I .awq—uohul. mm: 8*. - an - 1 u 4 mi: N 3.. .. 3.883 «.0 3.0 v.0 ed 0; N.— laquocqv 67 a single. band at 560 m P (Figure 29). The shift resulting from solvent change can be eXplained in terms of the structure of the molecule. The indicator acid, obtained from stock in the lactone form, undergoes ring opening in the presence of base to yield a carbinol form in water or an amide form in aqueous ammonia. It is proposed that phenol- phthalein in anhydrous ammonia exists as the amide form (I) in equilibrium with structures II, III, and IV: HOi iOH BO 2W3 '. \ C < NHz * - C - 0- ll 0 I II 32 fi (35) .0i 0. '0. i; / 1" \‘\ // 233:.“ \C-NI-Ia ‘3 2113:." (it +NH ’0- ’0- C\\O C\\o IV III The formation of I apparently proceeds as rapidly as solution occurs and the imediate formation of II gives the solution its characteristic purple color. As base (i.e. potassium amide) is added, the color intensifies. A shift to III is indicated since it has a higher 68 33an 28.33 5 odoaefinnaononm no gooam 333034 .mm earn a I . finnedegz com 8: com d d A d d 1 L to n «.o Kamqaosqv 69 extinction coefficient than II. As the addition of base is increased, the solution turns colorless as IV is formed. 'With the addition of ammonium ion to the neutral purple solution, the color turns light pink but never disappears entirely. Since absorption bands are by nature symmetrical (66), presumably there are two bands present in this region (dotted lines, Figure 28). The small hump observed at the lower wavelength of the large absorption band can be attributed tentatively to form.II which may well have a smaller extinction coefficient than form III because of its electronic structure. The extinction coefficient for the 57S m.p band (Figures 30-32) calculated by the method in Appendix II is 5.86 x 10h, which is in good agreement with 7.7 x 10h calculated by Lagowski (29) using a different method. Franklin observed (57) ammonia solutions of phenolphthalein to be colorless at room temperature, whereas this study shows that ammonia solutions of this indicator have a distinctive purple color at -77‘C. As the temperature is raised, the color gradually changes to a light pink at the boiling point of ammonia, and as the solution is allowed to warm to room temperature in sealed tubes, the color disappears completely. The loss of color occurs in the vicinity of ~10'C, the exact temperature depending upon the concentration of indicator present. If the first reaction in equation gg represents the equilibrium in question, it can be seen that increasing the temperature should shift the position of equilibrium to the left in the following manner. Since there is no separation of charge in this reaction, the equilibrium constant should be little affected by temperature. On.the other hand, if this process consists of two steps, the equilibrimm constant should vary with temperature, as was found to be the case (Figures 33-35, Table VII). Presumably the equilibrium.measured spectrophotometrically is that represented by E; in 70 9:... use a a msm we eweaefinnaoeenm Hm nosehvneonoo noun; hogan: .om 933$ do limo" H fifiefisnaoeoam no 53.935300 m 4 a n q u q q a .rP .39 m3. .Inll. 8338 H.582 so fCOO «.9 0.9 0.“ l a.“ bueqaosqv 71 a a mu m a. 55.5 a s Hannah no noggaaonoo 26."; 33.3 . 2a . R 85E a m - a F .l. p H a Mo Ugazfiofifi «0 aoapubaooaoo q . m c - q - - 1 - :6 woo «J 3. com .3“ bquoaqv 72 a a m . sm .91 33 Eafiaoc 2E no 3395588 upon; .53 . 3.334 . «« 25$ . . h “0 _ q u 1 a q q . - 0 0.3. \W DOOM... Ill-l «.0 l’ I. J00 nmd ' 00° laquoaqv floaafinnaoqufi you tsunami...“ upon: am. 33339 05. .mn 25w: “Eugen 333.33 73 on: .2... cm: can on: o. ‘ munch“; 4:6 ‘0 __ _ -36°C . l L . 1‘ l o 1 2 3 h Concentration of Phonolphtlnlein x 106g Figurc 31;. Log I; Versus Concentration of Phonolphthnloin at. 575 n )1. -008 zs -770c / , __...63-c 7 O ‘ 1 n 1 l L l I L 1 1 L - 2 3 h _ f 5 Concentration of Phenolphthalein x 105g Figaro T5. Log 1'. Versus concentntion of Phonolphthnloin at 575 n p 76 TABLE VII IONIZATION OONSTANTS OF PHENOLPHTHALEIN AT VARIOUS TEMPERATURES Temperature ‘6 Log £2 E; .36 00225 1 s69 .hé o.h7o 2.95 -63 0.991 9.80 -77 1 .380 2,400 equation.2§. As the temperature is raised, forn III is converted to II and the equilibrius represented by'gg is in turn shifted to the left, resulting in production of a colorless species. The salt effect upon phthalein indicators is quite pronounced in water solution (58) and this is also observed with phenolphthalein in ammonia. As seen in.Figure 36 and 37, the effect of the neutral salts depends upon the anion.nore than the cation. This is attributed to the greater polarizability of the anion and its subsequent effect upon ion pairing and dissociation constants. The indicators so far discussed change color in acidic solutions. Such indicators can be used to measure the acid.strengths of compounds which contain an annonium.ion or produce an ammonium ion by solvolysis. Even though thiourea and its methyl-substituted derivatives are bases in water, these compounds are acids in liquid ammonia and amnenolyze to for: ssneniun ions. 77 3......» undue-pom Mo soapsbnoonoo an.” sodafinmdononm no kudos—hog . wn shaman Q3305 pawn a3. uneven no savanna-once m6 :6 . no r «.o x to o 4 q d 1 q a h a q d I? A? w a a m. A. a m. 3 ‘ b 1” El... ‘ ‘o woo , a .29 0 H“ q 3 .5 B \ .32. 6 to... a _- mu 9. 3d.» 538 23m «0 nonpafieooooo .. as...» 5.15a38oé no engage .R 9:82 Abflqnomv 33 538 no 5: vengeance m6 .3 n6 «6 . . .6 o q u d A w ~ . 1 I ”~00 l r I. 8 7 _ .0! :30 I? Q a 3.3: .588 0 333 838 a . NJ 3.8228 532.. O 79 The reaction of thiourea in ammonia was investigated by examining its effect upon phenolphthalein (Figure 38). It was observed that thiourea reacts as an acid by lightening the color of the indicator solution. Measurements of the absorbancy of phenolphthalein in various concentrations of thiourea (Figure 38) allowed calculation of g; by the method of Appendix II. A plot of log g; 22:22:. concentration extrapolated to infinite dilution (Figure 39) gave the ionization constant for thiourea. Since the value obtained is 7.1 x 10-3, thiourea can be considered to be a weak acid which presumably reacts as: S S HJ£4m+Nfir¢HJ£Jme+ Q9 Substitution of methyl groups for hydrogens in thiourea should decrease the acid strength; this is shown to be true by the ionization constants which were obtained in a similar way for some methyl-substituted thioureas (Table VIII). TABLE VIII IONIZATION CONSTANTS OF SOME THIOUREAS Compound Log Ei $1 (I 103) Thiourea -2.15 7.1 Methylthiourea -2.h2 3.8 Trimethylthiourea -3.25 0.18 Methylthiourea is slightly less acidic than thiourea and trimethyl- thiourea even less acidic. Trimethylthiourea still affects the indicator 80 soaasuasoonoo 3.33 543.38 vgnoo as noon—Sana seem no soapshpnoonoo g fioasfinnaaofi Ho Meghonn.‘ 6m «99mg mnop N 39535. no codasbusoonoo sonpoflvahfioz Q zéosfiahaofih. G 2 a e a N o 4 . q J . q a q q i .B sl 40H .aF 3.335. B lousqaosqv 81 Q Thiouru A Methylthiouru [g Trim‘ethylth iouroa -1.2 -1.6 -208 l J J l o 1 2 3 1: Concentration of Thiourua x 103g Figure 39, Log Ki m Concentration“ Some Thiouroaa 82 sufficiently to permit calculation of its ionization constant. In the determination of the acid strengths of these weak acids, it had to be considered that neutral salts have effects on the indicator of the same order of magnitude as the weak acids and also that plots of the logarithm.of {_325322 the negative 1ogarithm.of the ammonium.ion +) do not give straight lines for all anions NH. (Figures 11 to 1h). In the investigation of these weak acids, concentration (pC potassium iodide was used in constant concentration because the plot of pCNH‘I— versus the logarithm of r approached a straight line in the concentration range used. Any ammonium ion produced by solvolyais would react toward phenolphthalein as ammonium iodide owing to the large excess of iodide ion present. Potassium iodide was present in large enough excess in all cases to give a constant salt effect. weak Bases Because ammonia is strongly basic, solute acidity is enhanced and the majority of bases in liquid ammonia are weak. Indicators which are useful in studying basic solutes exhibit different colors in neutral and basic solutions. For spectrophotometric study of base strengths in ammonia, two approaches were used. In the first, a polyprotic acid indicator was chosen and one of the latter steps in its ionization was used as the indicator reaction with bases. This method is especially advantageous where very weak bases are concerned as their effect upon strongly acidic indicators is negligible. The second method was to choose an indicator which changes color on the basic side of neutrality in ammonia and whose ionisation constant can be found in terms of the amide ion concentration. 83 The indicator thyml blue (tlvmol sulfonephthalein) is of the first type mentioned above. It has, in addition to the sulfonic acid group, two replaceable hydrogens which comprise the second and third ionisation steps. The sulfonio acid group issuch a strong acid that in amonia it is completely ionized. The measurable ionisation steps are: ,CH, on; on, on, ”ca ‘ H.‘cn Ka '\c{1m, 5 \c/ a r .. I 50.0 so.o‘ ‘ 301 m 620 n p P g,» (37) K0113 .033/ CH, O :0 + NH: \| 30,0“ 390 n p Solutions of this indicator in ammonia are blue in an acidic media and a blue-green color in both neutral aid base. It has absorption bands at 361 n p, 390 n p, and 620 n P (Fism‘O ho). In acid solution the band at 390 n '1 decreases in intensity, giving the solution its characteristic blue color owing to the 620 m p band. As the solution because more basic, the 390 m ’1 band increases, giving the blue-green color as the yellow component is increased in intensity nanosad ad spam Hoshga no ehaoomm soapnhonn< .ml shaman A I .npmueaepes om: co: «.5 m tollll .eosemz a to I.| uoavndom Hehvsez mics a 2... . hostess—H moo :6 New £9usq109qv 85 (Figures 1.1 and h3). The band at 301 m ,2 follows Beer's Law (Figure 1.2) at all concentrations measured and is assigned to the species formed by the loss of the sulfonic acid proton. The 390 m ,u band is assigned to the structure resulting from ionization £3 in equation 11 (loss of the third proton). Spectrophotometric measurement of this band should give the data needed to calculate ionization constants of weak bases. The ionization constants and extinction coefficients for the species in equation.21 are listed in Table VI (page Sh) and were determined by the method of Appendix II from the data of Figures hh and hS. Tropeoline 00 (the sodium salt of diphenylaminoazobenzenefp- ~sulfonic acid) is an indicator which has a different absorption band in basic solution from that in acidic or neutral solution. The Spectrum of a neutral solution of tropeoline 00 has two absorption bands at 276 m.p and h60 m,P (Figures h6-h8). The latter band is responsible for the intense yellow color observed for this indicator in neutral ammonia solution. Acidic solutions yield the same Spectrum as the neutral solution, as would be expected for azo type indicators in such a basic medium. In aqueous solution the acid form of an azo dye is protonated on the nitrogen--nitrogen (azo linkage) giving aquinonediimine structure, while in basic solution the color is due to the azo structure itself. The color shown in neutral ammonia is the same as in aqueous base. The deep violet color of amide solutions of this indicator is attributed to a different reaction from those found in aqueous solution. Schattenstein (b1) postulated that the color change arises from addition of the amide ion to the azo linkage: l G Can as spam Hog Ho cogehvaoonoo esnnop 359983 . a3 93mg 2mg N spam Hog no capehasoosoo ‘0 o \ octagon Hence; a u q q as m 3 III 87 1 a 5m seesaw H255. mo noaveuaneocoo ensues. 355.32: ”mo— K spam Hog Ho aofigsoosoo m o a. a q i 1 u d d . q .3 enemas lousqaosqv as K a own an 3.8 H235. no qoavauanoonou 3nho> hoganuound . m: 953E .mmop H 3.8 H255. no defipananooqoo -m m m u p ..o q 1 fl q q q d N; o.— o.w buniocqv 89 .008 P Ooh "‘ 0.8 I- 1 I L o 1 Concentration of Thymol Blue 2: 1055 NI- Figuro h)... Log g; Eel-g3; Concentration of Thymol Blue at 620 n p L085 _ -O.6 -1 .0 '1 oh ‘1 08 -2.2 -206 - I l J l l l l I I l o 1 2 3 h 5 Concentration of Thymol Blue x 10514 Figure 145. Log K' Versus Concentration of Thymol Blue at 390 m .3 _ P 91 dado—Bi 5. 00 25330.5 no «.3025 aoapmhonnd .0... 03mg a E «gang-H053 omo 8o 0mm 8m cm: 8; omm 8m fl — q 4 . u a q . a . d . / x / \ / \ / , \ / \ / , , , , \\ , \ l(.\\ «a mtolll. cannon H58: HTS n a: .. usuaufl «.0 4.0 0.0 m6 0.? NJ buuqu sqy 1 a 00: an 8 05.30th Ho douvuhnoonoo opnuo> hoawnuoua .5... 0.33m mmop H 8 034.0303. no aoaaobcoocoo 92 Q or. P o M .o 10.9 hmqaosqv 93 L a 0mm an 8 0300939 no noaaahanoonoo may: 559334 .3 953$ «an Ed 5 We, a 8 ofiaoafia Ho aoflanbnoonoo o A N 9h H H ’ M3» m . Dig-31,20 h60m}: §90mp (_3__8_) The band at 590 m P is present only in basic solution and is thus assigned to the structure in equation _3_8_. From basic to acidic solution (i.e., on addition of an ammonium compound), a very sharp color change from purple to yellow occurs. All other indicators studied showed a gradual transition from one color to the other. Figure ’49 shows a steep slope at the transition point of tropeoline. 00 for both bands. The salt effect upon azo indicators is in general negligibly small and this was observed for tropeoline 00. A reasonable explanation lies in the fact that the azo ion is quite large and could well be of approximately the same size as the ionic atmosphere, thus minimizing the salt effect. Urea does not react with the indicators which have been found useful for determination of the strengths of weak acids (phenolphthalein, 2,h-dinitroaniline, and p-nitroacetanilide). Urea does, however, react as a neutral or basic compound toward them. It was observed by Lagowski (29) that urea and several other compounds postulated as weak acids in ammonia have no effect upon phenolphthalein. On the other hand, the effect of urea upon thymol blue is as a very weak base G'igure SO). . Urea does not affect tropeoline 00, however, and conse- quently its acid strength must be very close to that of a neutral ammonia solution. Though potassium salts of urea have been prepared (61;), these reactions were done in potassium amide solution which is a considerably stronger base than ammonia. The sodium salt of urea can be 9S coapmfinooqoo 035 Sawmmmpom we won g 8 can—“023.5 no 5059.325. . ma 0.33m «mzmoc & a om: III as 0mm mnop x ma.~ . noeeeeeaH 2 mm «.0 {o 0.0 m6 o; w; Koueqmo sqv 96 Indicator - 1.23 x 10'5” Neutral Solution ~ —-——_ 0.1g Urea 0.21; . / \\ -—._ 0.1g NH‘ClO‘ 0.08 b- \ j l ‘ | I 380 390 hoo Wavelength, m P Figure 50. Absorption Spectra of Thymol Blue in the Presence of Urea at 390 m In 97 prepared but it has only been obtained using elemental sodium (63). These data can be reconciled if urea is considered to be an amphiprotic substance as shown in equations g§ and 22(p.h7). The species present in neutral ammonia could be that represented in equation 22, which would affect a basic indicator such as thymol blue but not tropeoline 00. Also in the presence of a stronger base such as amide ion, urea could react in the manner indicated in equation 28. The effect of substituting methyl groups for hydrogens in urea should increase the base strength, although the magnitude observed is not as great as would be predicted from their structures. The order of base strength is: urea < methylurea < 1 ,3-vdimethylurea < 1,1—dimethylurea < trimethylurea < thiourea The order of urea and thiourea are the same in ammonia with thiourea displaying acidic character while urea is very weakly basic. Guanidine should then be more basic than urea in ammonia and this could be studied by use of basic indicators of the azo type such as tropeoline 00. 2. When an ammonia solution of m-dinitrobensene. is electrolysed, hydrogen is evolved at the cathode (113) and the nitro group is reduced. This suggests formation of an ammonium compound and use as an acid- base indicator. It was observed during the course of this study that trinitrobenzene also forms highly colored solutions and further study of it is recommended. 3. The ion product of water can be determined directly by spectrophotometric means, while in ammonia the value of 10'33 for its ion product was determined by electrochemical means. Because ion pairing is extensive in ammonia, a spectrophotometric determination of the ion product may be more valid as this method measures concentrations and not activities. 1.1. when a liquid monia solution of 2,11-dinitroaniline in a sealed tube was warmed from -77°C to room temperature, the deep red 116 117 color became orange. Upon cooling, a.red compound precipitated and redissolved upon standing. Since evaporation of ammonia solutions of 2,h-dinitroaniline always gives a yellow compound, perhaps the red precipitate could be the ammonium salt of dinitroaniline. The spectrophotometric study of temperature effects on.this is recommended. 5. The indicator tropeoline 00 has a very sharp color change from acidic to basic solution (or 1L0: 1219.!) in ammonia. This could be used in acid-base titrations in ammonia as a visual indicator to show the equivalence point of a neutralization reaction. Other azo dyes may have this preperty, and because the salt effect is negligible for this indicator, investigation is warranted. 6. In the two-step equilibria established in{ammonia by addition of a molecular solute, the ionization step is more dependent upon the nature of the solute than upon the solvent. Since this is the equilibrium which can be studied spectrophotometrically3 a look at some indicators in other solvents such as pyridine or amine-type solvents and a comparison of acidities should lead to a fundamental characteristic of acids and bases. BIBLIMRAPHI 1. E. 0. Franklin and c. A. Kraus, Am. Chen. J., g, 1 (1899). 2. E. 0. Franklin and 0.1. Kraus, Am. Chen. J., 23, 277 (1900). 3. c. A. Kraus and w. c. Bray, J. Am. Chem. Soc., 3 1315 (1913). 11. R. Taft, J. Chem. Educ. 12, 311 (1933). S. L. 1". Audrie'th and J. Kleinberg, Non-A eous Solvents, John Wiley and Sons, Inc., New Iork, 1953, pp. 53-93 6. E. C. Franklin, The Nitrogen §zstem of Cgfigunds, Reinhold Publishing Co. Inc., ew ork, , pp. - . 7. 11. Linhard and n. Stephan, z. phwaik. Chfllo, _A_1_§_3_, 85 (1933). 8. h. Linhard and M. Stephan, z. phwaik. Chem, 5161, 87 (1933). 9. H. Hunt, J. Am. Chem. 800., fly 3509 (1932). 10. w. 0. Johnson and 0. r. Krumbolts, z. prmik. Chem, A167, 2117 (1933). 11. G. W. Watt, W. A. Jenkins, and C. V. Robertson, Anal. Chem, 22, 330 (1950). 12. V. A. Pleskov and A. H.1Honosson, Acta. thsicochim. 0.3.8.3., l, 713 (1935). 13. H. A. Laitinen and C. J. Nyman, J. Am. Chem. 800., 19, 3002 (19118). 111. 0. 11. Watt, D. u. Sowards, and w. a. McBride, J. Am. Chem. 800., 11, 5835 (1955). 15. R. H. Fuoss and C. A. Kraus, J. Am. Chem. Soc., 25, 2391 (1933). 16. N.(Bje1‘:1)'um, Kgl. Danske Videnskab. Selskab Nat. fys. Hedd., _9_, 7 192 . 17. F. Baaolo and R. G. Pearson, Mechanisms of Inor anic Reactions, John Wiley and Sons, Inc., New TorE, 1953, p. 3119. 18. c. A. Kraus and R. u. Fuoss, J. Am. Chem. 300., 35, 1315 (1913). 19. C. A. Iraus and U. U. Ems, J. Am. Chem. 800., _5_5_, 2776 (1933). 20. 11. U. Hawes, J. Am. Chem. 800., £5, 11112! (1933). 118 119 21. c. A. Kraus and P. B. Bein, J. Am. Chem. 800., 55, 3609 (1933). 22. P. Debye and E. Hiickel. Physik, z., 31;, 1855 (1923). 23. J. 11. Williams, Chem. Rev., 8, 303 (1931). 1 211. J. Sedley and T. DeVries, J. Am. Chem. 800., _7_3_, 5808 (1951). 25. 0. N. Lewis and H. Randall, J. Am. Chem. 800., 133, 1112 (1921). 26. H. A. Leitinen, Chemical Analysis, HoGraw-Hill Book Co. Inc., New York, 1960, p. 13f 27. 8. Bruckenstein, Dissertation Abstr., 12, 3113 (1955). 28. S. Bruckenstein, Private Comunication 29. J. J. Lagowski, Dissertation Abstr., 2_0., 35011 (1960). 30. A. I. Schattenstein and N. _H. Dykhns 2, 1.6 (19118); c. A., gig, 6658 (191181. 31. 1E. E.)I. Hummelstedt and D. N. Hume, J. Am. Chem. 800., 8_3_, 1561.1 1961 . 32. I. H. Kolthoff and S. Bruckenstein, J. Am. Chem. Soc., 1g, 1 (1956). J. Five. Chem. (0.8.5.3.), 33. I. M. Kolthoff, S. Bruckenstein, and H. K. Chantooni, J. Am. Chem. 800., £32.: 3927 (1961). . 3h. I. M. Kolthoff and '1'. B. Reddy, Inorg. Chem., 1, 189 (1962). 35. H. Lemaire and H. J. Lucas, J. Am. Chem. 300., 73, 5198 (19511 36. A. H. Corwin and J. 0. Reinheimer, J. Am. Chem. 300., 13, 11811 (1951). 37. c. A. Kraus and 11. Rosen, J. Am. Chem. Soc., 1g, 2739 (1925). 38. I. Shurman and w. c. Fernelius, J. Am. Chem. 800., 5_2_, 21125 (1930). 39. 0. r. Hhite and x. H. Knight, J. Am. Chem. Soc., E5, 1780 (1923). 110. A. I. Schattenstein,1Acta Physicochim. 0.3.3.8., _12, 121 (1939). 1.1. A. I. Schattenstein, Acta Pmioocmn. 0.8.8.8., 1, 1.01 (1937). 112. A. 0. Green and F. 11. Rowe, J. Chem. 800., 193, 508 (1913). 143. J. D. Parr C. C. Bard, and G. W. Hheland, J. Am. Chem. 800., 11, 2013 (191191. 1111. 11. L. Jolly, U.S.A.E.C., N.S.F., Nash. 0.0. 00111., 3, (1952). 120 115. 0. E. Gibson and w. L. Argo, J. Am. Chem. 800., 11_0, 1327 (1918). 116. G. E. Gibson and H. L. Argo, Phys. Rev., 1, 33 (1916). 117. N. Dykhno and A. I. Schattenstein, Acta Physicochim., 0.8.8.8., 2_0, 61.5 (19115). 1.18. E. Vogt, Z. Electrochem., 112, 597 (1939). 119. R. A. Ogg. J. Chem. Phys., _1_1_1_, 1111 (19116). 50. I. H. Kolthoff, J. Phys. Chem., _3_2_, 1820 (1928). 51. H. F. Walton, Chemical Analysis, Prentice-Hall, Inc. , Englewood Cliffs, New Jersey, 19 2, p. 1. 52.. I. 11. ”non, Chem. Reva.,_11_1_, 373 (19117). 53. A. Thiel and 0. Coch, z. anorg.u. allgem. Chem., 211, 353 (19311). 511. D. H. Everett and W. F. K. Wynne-Jones, Trans. Faraday 800., 33, 1380 (1939). \ 55. H. S. Harned and N. D. hbree, J. Am. Chem. 800., 56, 1050 (19311). 56. E. c. Baughan, J. Chem. Phys" 1, 951 (1939). 57'. E. 0. Franklin, The Nitrogen §ystem of Cgfigunds, Reinhold Publishing Co. Inc., ew or , , pp. , , 273. 58. H. F. Walton, Chemical gflsis, Prentice Hall, Inc., Englewood Cliffs, New Jersey, , p. 3. 59. c. D. Hodgman, Editor, Handbook of Chemisg and 235E111“, 37th Edition, Chemical 0 r s ., eve an , Ohio . p. 2161. 60. A. Prikhotko, Acta Physicochime. U.R.S.S., 13, 565 (19110). 61. H. 0. Heller, A%a%1_. ical Absorption Spectroscopy, John Wiley and Sons, New York, 9 0, p. . 62. a. h. Fuoss and c. A. Kraus, J. Am. Chem. 8oe., 55, 2387 (1933). 63. J. 8. Blair, J. Am. Chem. 800., 1333.: 96 (1926). 61.1. R. A. Jacobson, J. Am. Chem. Soc., £8, 19811 (1936). 6S. 8. Glasstone ”Win for Chemis‘lgsJ D. vanNostrand, Inc. , New York, 19117, p. . 66. A. Head, Trans. Faraday 866., 32, 1055 (19311). 67a 68. 69. 70. 71 . 72. 73. 7h. 75. . 121 E. Spinner, Spectrochim. Acta, lg, 95 (1959)- 0. Fraenkel and c. Franconi, J. Am. Chem. 800., 5g, hh78 (1960). M. L. Anderson, Private Communication.r . 1 1 ‘Wurtz, Canpt. rend., 5g, h1h (1851). '. ’ . T. L. Davis and K. 0. Blanchard, J. Am. Chem. soc., 51, 1790 (1929). Leuckart, J. prakt. Chem.,, _2_1_, 29 (1880). 0. Schroeter, Ber., [43, 3356 (1909). H. z. Leoher and K. Gubernator, J. Am. Chem. Soc., 15, 1089 (1953). E. C. Horning’, Editor, Dr anic S theses John Wiley and Sons, Inc., New York, 1955, Collective Vol. E, p. 317. CHEIMSTnY LIBRARY \\ LB "71111111111111fifijflflflflhflflmflu1111'“ 606