FEBRUARY, I973 THE ANALYST Vol. 98, No. I163 Ionic Polymerisation as a Means of End-point Indication in Non-aqueous Thermometric Titrimetry Part I.* The Determination of Organic Bases BY E. J. GREENHOW AND L. E. SPENCER (Department of Chemistry, Chelsea College, University of London, Manresa Road, London, S . W . 3 ) A method for the thermometric titration of organic bases in non-aqueous solution, in which ionic polymerisation is used to indicate the end-point, has been evaluated for a range of aliphatic and aromatic amines, including substituted and polyfunctional amines, heterocyclic nitrogen compounds, amides and some basic sulphur and phosphorus compounds. Perchloric acid and boron trifluoride were used as titrants and a-methylstyrene and isobutyl vinyl ether were the monomers used, respectively, in conjunction with them.The precision of the method with titrants of molarities from 0.1 to 0.001 is of the order of 1.5 per cent. when a simple manual procedure involving the use of a thermometer to measure the temperature is adopted. More elaborate methods, in which the temperature is measured with a thermistor and recorded, give precisions better than 1 per cent. Sample sizes down to about 0.0001 mequiv, e.g., about 10 pg of morpho- line, which corresponds to 10 p.p.m. of morpholine in the volumes of sample solution titrated, can be determined with 0.001 M titrants. Calibration graphs show that the volume of titrant and amount of sample are linearly related in the range 0 to 3 ml of titrant. It is suggested that comparison of the results obtained when a base is titrated with the two titrants, a Brarnsted acid and a Lewis acid, can be used to investigate some of the properties of the base. THE determination of organic bases in non-aqueous solution by thermometric titration has been described by several workers.Keily and Humel studied the titration of bases in acetic acid while Forman and Hume2 used acetonitrile as the solvent; the titrants were perchloric acid in acetic acid and hydrogen bromide in acetonitrile, respectively. In both investigations temperature changes were slight when 0.1 M titrant was used, and a device reading to 0.001 "C was used for temperature measurement. Greater temperature changes, about 2 "C in the course of the titration, were achieved by Vaughan and Swithenbank3 when using a 5 M solution of hydrogen chloride in propan-2-01 as the titrant, which gave an endothermic heat of dilution at the end-point.In a method reported by Vajgand and co-worker~,~~~ the catalytic effect of the titrant, 0.25 M perchloric acid, on a mixture of water and acetic anhydride included in the sample solution causes temperature rises of about 0-5 "C at the end-point. In all of the methods described above, the molarity of the titrant or the temperature change obtainable at the end-point, or both, tends to fix a lower limit for the size of the sample that can conveniently be titrated, which, with the most sensitive method, would appear to be about 0.1 mequiv. Two recent papersss7 have briefly described procedures in which the rise in temperature that accompanies anionic and cationic polymerisation is used as a means of indicating the end-point in the non-aqueous titration of acids and bases, respectively.In these procedures, monomers that are capable of ionic polymerisation are used as solvents for the acids or bases to be determined. The temperature rise following the neutralisation of the sample by the titrant is a result of the catalytic action of free titrant on the monomer - solvent mixture. Catalysts for anionic polymerisation are, in general, strongly alkaline in character, and are therefore suitable as titrants for weak acids in non-aqueous systems. Similarly, cationic polymerisation can be initiated by catalysts with strongly acidic properties ; these catalysts, * For Parts I1 and I11 of this series, see pp.90 and 98, respectively. @ SAC and the authors. 8182 GREENHOW AND SPENCER : IONIC POLYMERISATION FOR END-POINT [AndySt, VOl. 98 in turn, are suitable reagents for the titration of bases. As many ionic polymerisations can take place at room temperature and require relatively small amounts of catalyst to initiate the highly exothermic chain reactions, sharp rises in temperature, often as great as 5 “C in 1 s, can be achieved near the end-point in these titrations. The present paper reports a detailed evaluation of the application of cationic polymerisa- tion to the thermometric titration of different classes of organic bases. Certain properties of the solvents used in ionic polymerisation studies, particularly their dielectric constants, are known to have an influence on the initiation of the polymerisation and on the rate of polymerisation.A number of solvents, covering a range of dielectric constants and “donici- ties,”8 have been investigated as solvents for the titrant and for the sample. Two titrant - monomer systems have been investigated. In the first, perchloric acid, a Brernsted acid, is used as the titrant and a-methylstyrene as the monomer-solvent. In the second, a Lewis acid, boron trifluoride, is the titrant and isobutyl vinyl ether the monomer - solvent, EXPERIMENTAL REAGENTS- a-Methylstyrene, isobutyl vinyl ether and the organic bases were laboratory-reagent grade materials and were used as received without purification. Toluene, 1,4-dioxan, 1,2-di- chloroethane and nitroethane were dried over 4A molecular sieve.Morfdioline-Analytical-reagent grade morpholine (BDH Chemicals Ltd.) , with a mini- mum assay (acidimetric) of 99 per cent., was used as a standard compound. C +I I- t Fig. 1. Thermistor bridge circuit: C, Mallory cell, Type RMlSR, 1.35 V ; D, Servoscribe potentio- metric recorder, Type RE 511; R,, sensitivity control, variable resistor, 50 R; R,, fixed resistor, 2000 R; R,, fixed resistor, 1500 R ; R, and R,, zero control, variable resistors, 1000 R ; S, switch; and T, thermistor, S.T.C. Type F23D (2000 a)February, 19731 INDICATION IN NON-AQUEOUS THERMOMETRIC TITRIMETRY. PART I 83 Potassium hydrogen @zthaZate-AnalaR grade. Dry for 2 hours at 110 "C before use. Dry acetic acid-Prepare by adding the theoretical amount of acetic anhydride to glacial acetic acid of known water content (determined by the Karl Fischer method) and allow the mixture to stand overnight.Boron trijztoride diethyl etherate-Mix 10 volumes of laboratory-reagent grade boron trifluoride with 1 volume of diethyl ether and add about 2 g of calcium hydride to each 100 ml of the mixture. Distil the dried mixture and take a middle fraction, the boiling-point of which is above 120 "C. Boron tri$uoride diethyl etherate, 0.1 M in dioxan-Standardise this solution against a solution of morpholine (20 mg) in toluene (2 ml). Perchloric acid, 0.1 M in dry acetic acid-Measure 8-5 ml of AnalaR grade perchloric acid (71.0 to 73.0 per cent.) into a 1-litre calibrated flask and add 100 ml of dry acetic acid and 15 ml of acetic anhydride slowly with constant stirring and cooling.Dilute to the mark with dry acetic acid and leave for 24 hours in the dark. Standardise the solution against solutions of potassium hydrogen phthalate in dry acetic acid and morpholine in toluene. Prepare other standard solutions of perchloric acid by adding the appropriate amount of the 0.1 M solution in dry acetic acid to the solvent of choice, and other standard solutions of boron trifluoride by diluting the 0.1 M solution in dioxan in a similar way. APPARATUS- A. Manual method-Use a -5 to +SO "C thermometer graduated in 0.1 "C, either a 10-ml Dewar beaker or a 50-ml tall-form beaker, a 5-ml burette graduated in 0.01 ml, with the tip drawn out to a capillary about 1 inch long, and a magnetic stirrer. B. Semi-automatic method-Use the apparatus of Method A with a thermistor for tem- perature measurement instead of the thermometer.The thermistor (S.T.C., Type F23D) forms part of a bridge circuit (Fig. l), which is connected to a millivolt chart recorder with full-scale ranges of 5, 10, 20, 50 and 100mV and a chart speed of 30 mmmin-l. C. Automatic method-Use the thermistor system for temperature measurement, a 10-ml Dewar beaker to contain the sample, and a motor-driven Agla micrometer syringe (0.5 ml) for the titrant feed. The switches for the syringe motor and the recorder-chart drive are coupled together to synchronise titrant addition with temperature measurement. PROCEDURE A. MANUAL METHOD- Prepare a solution of the base in a suitable solvent, e.g., toluene, acetone, nitroethane, 1,2-dichloroethane or acetic acid (dry).The concentration will depend on the titrant con- centration, thus 2 ml of the solution should contain about 0.2 mequiv of base when the 0.1 M titrant is used, about 0.02 mequiv with 0.01 M titrant, and so on. Transfer by pipette 2 ml of the sample solution into the beaker, add 10 ml of monomer (2 to 5ml if the 10-ml Dewar beaker is used), stir the solution, in which the thermometer bulb is immersed, for 1 minute, then add titrant at the rate of about 0.5 ml min-l to within 0.3 ml of the end-point, noting the temperature at &minute intervals, and complete the titration with addition of titrant at a rate not exceeding 0.2 ml min-1, noting the temperature at 15-s intervals. At rates significantly greater than 0.2 ml min-1, high titration values are obtained owing to overshooting of the end-point.Addition of titrant at rates slower than 0.2 ml min-l gives only a small improvement in precision with the more concentrated titrants, and with the less concentrated titrants can result in loss of end-point sharpness. When the temperature exceeds 35 "C, remove the thermometer from the solution so as to avoid damaging it, and pour the solution to waste after diluting it with about two volumes of acetone. B. SEMI-AUTOMATIC METHOD- Use a procedure that is similar to that used in Method A, but replace the thermometer by a thermistor and record the temperature changes on the millivolt chart recorder. Start a stop-clock synchronously with the recorder-chart drive in order to relate the titrant volume, which is read off at appropriate times, to the temperature on the recorder chart.In practice, it is necessary only to note the time and the volume corresponding to the required inflection point on the chart record.84 GREENHOW AND SPENCER IONIC POLYMERISATION FOR END-POINT [A?W&d, VOl. 98 C . AUTOMATIC METHOD- Use a procedure that is similar to that used in Method €3 but add titrant at a constant rate of 0.2 ml min-l from the motor-driven syringe. Calibrate the recorder chart in millilitres of titrant in order to read off the titrant volume directly. DETERMINATION OF END-POINT- The titrant volume at the end-point is measured in three ways. (i) As the volume corresponding to the "upturn" temperature where the titration curve leaves the tangent drawn to its horizontal part.In Method A a rise in temperature of at least 0.1 "C is needed in order to locate this end-point. (ii) From the intersection point of tangents to the two component parts of the tem- perature - volume curve. For precision, this measurement requires the parts to have linear sections in the vicinity of the end-point. (iii) As the volume corresponding to the temperature following the first temperature increase of 0-3 "C or more, in a period of 15 s, when the rate of titrant addition is between 0.1 and 0.3 ml min-l. Methods (ii) and (iii) give approximately the same result, and method (iii) can be used as a rapid means of end-point determination in conjunction with Method A. Method (i) is the preferred method when applicable, as the volume of titrant thus measured corresponds closely to that obtained by potentiometric titrimetry. RESULTS AND DISCUSSION In an earlier paper,6 it was shown that the perchloric acid-a-methylstyrene system could be used for the determination of primary, secondary and tertiary aliphatic amines, represented by n-butylamine, morpholine and triethylamine, respectively, and it was claimed that a solution of boron trifluoride in dioxan was also a suitable titrant when used in con- junction with isobutyl vinyl ether as the monomer.The applicability of the latter titrant to the determination of these amines has been confirmed. The precision of the manual method of thermometric titration (A) has been determined by using both titrants in concentrations in the range 0.1 to 0.001 M (Table I). In general, TABLE I RESULTS FOR PRECISION FROM THE THERMOMETRIC TITRATION OF MORPHOLINE WITH 0.1 TO 0.001 M SOLUTIONS O F PERCHLORIC ACID AND BORON TRIFLUORIDE Coefficient of variation, Titrant Method per cent.- of End- Mono- Morpho- No. of Mean titra- point mer/ line/ titra- titre/ Standard Single Mean Nominal molarity Solvent* tiont method? ml mg tions ml deviation points value Perehlorie acid (monomer : wmethy1styrene)- - 0.1 A A (ii) 10 10 5 1.27 0.016 1.29 0.58 0.1 A A (ii) 10 20 6 2.52 0.040 1-68 0.71 0.025 A A (ii) 15 7 3 3.91 0.056 1.42 0.82 0.01 A A (ii) 10 2 4 3.38 0.083 2-44 1.22 2 0.2 3 1.30 0.010 0.77 0.45 0.2 3 1.67 0.015 0.90 0.52 0.002 A + X (1 + 49) A (2) 0.002 A + N (1 + 49) €3 (2%) Boron trifluoride (monomer : isobutyl vinyl ethev)- 0.1 D A (ii) 10 10 5 1-04 0.011 1.09 0.49 0.1 D A (ii) 10 25 5 2.42 0.015 0.63 0.28 0.01 D A (id) 10 2 4 3.33 0.038 1-14 0.57 0.2 3 1-80 0.026 1-39 0.80 0.002 D + C (1 + 49) A (i) 2 0.2 4 1.44 0-034 2.36 1.18 0.1 3 1-62 0.030 1.85 1.07 5 10 4 0-867 0.005 0.68 0.29 2.5 5 0.436 0.002 74 0.63 0.28 0.002 D + N (1 + 49) A ($2) 0.001 D + N (1 + 49) A (2) 0.1 D B (4 0.1 D C (ii) * A = acetic acid; C = acrylonitrile; D = dioxan; N = nitroethane; and X = 1,2-dichloroethane.t See Experimental. Figures in parentheses refer to proportions by volume.February, 19731 INDICATION I N NON-AQUEOUS THERMOMETRIC TITRIMETRY. PART I 85 the titration values show coefficients of variation of single results about the mean to be of the order of 1.5 per cent. and coefficients of variation of the mean value to be less than 1 per cent.Results for precision have also been calculated from values obtained with the semi- automatic method (B) and the automatic method (C) ; these results are also shown in Table I. As might be expected, Methods B and C gave results with higher precision than those obtained with the manual method by using titrants of the same molarity. However, the improvement in precision is not as marked as one would expect, considering the large size of the bulb (20 x 5 mm) of the thermometer used in Method A, and the fact that with the 10-ml beaker the bulb is immersed only to one third of its length in the titration solution. The semi-automatic method offers some advantage over the fully automatic method in that, with the manually operated titrant feed, it is possible to add the bulk of the titrant rapidly and the remainder at a slow rate so as to avoid over-running of the end-point.All of the precision experiments were carried out by using standard solutions of morpholine in toluene; 1 and 2-ml aliquots were taken with grade A pipettes (B.S. 1583:1961), and sampling errors arising from their use will, of course, affect the over-all precision of the determination. Calibration studies with morpholine have shown that sample size and titre are linearly related in the ranges 0.05 to 0.3 mequiv of sample (about 5 to 30 mg) with 0.1 M titrant, 0.0005 to 0.003 mequiv of sample (about 50 to 300 pg) with 0.001 M titrant, and in the corresponding ranges with titrants of intermediate molarity.The morpholine samples for the calibration experiments were added to the monomer as solutions in 1 ml of toluene. It can be seen, therefore, that with the 0*001 M titrant a 50-pg amount of the organic base could be titrated as a 50 p.p.m. solution in toluene. With this amount, titres of about 0-5 ml were obtained, and it can be concluded that concentrations as low as 10 p.p.m. should easily be detected. Titrants more dilute than 0.1 M were prepared by adding solvents to the 0.1 M solutions, Further dilution of the 0.1 M boron trifluoride solution with dioxan, and 0.1 M perchloric acid solution with dry acetic acid, down to concentrations of 0.01 M, gave satisfactory titrants with respect to end-point sharpness, but at lower concentrations the end-points became less I I I I I Perchloric acid titrandm1 (1 division = 1 ml) Fig.2. Effect of titrant molarity and titrant solvent on the shape of the thermometric titration curve : perchloric acid titrants (morpholine sample). a b c d e f Rlolarity . . . . . . 0.1 0.01 0.002 0.002 0.002 0.001 Morpholine/mg . . . . 10 2 0.2 0.2 0.2 0.1 Diluent for 0.1 M HC10, in CH,COOH . . . . - A A N X X a-Methylstyrene/ml . . 10 10 2 2 2 2 A = acetic acid; N = nitroethane; and X = 1,2-dichloroethane86 GREENHQW AND SPENCER : IONIC POLYMERISATION FOR END-POINT [AfiUZySt, VOl. 98 satisfactory (see Figs. 2 and 3). This difficulty was overcome by using either nitroethane or 1 ,Z-dichloroethane as the diluent ; it was then possible to obtain significant rises in tem- perature at the end-point with 0.001 M titrants.Acrylonitrile also proved to be a satisfactory diluent, but only for the boron trifluoride titrant. Although the rise in temperature at the end-point with nitroethane as the diluent was usually greater than could be obtained with 1,Z-dichloroethane or acrylonitrile, the endotherm resulting from the dilution of the monomer made it difficult to determine the end-point accurately. In contrast, sharp end-points were obtained with titrants diluted with 1,2-dichloroethane, and titration curves of similar shape were obtained with titrants prepared by diluting the 0.1 M boron trifluoride titrant only with acrylonitrile. I t t s I I 1 I I I Boron trifluoride titranthl (1 division = 1 ml) Fig. 3. Effect of titrant molarity and titrant solvent 011 the shape of the thermometric titration curve : boron trifluoride titrants (morpholine sample).a b c d e f g h Molarity . . .. . . 0.1 0.01 0.002 0.002 0.002 0.001 0.001 0.001 Morpholinelmg . . . . 10 2 0.2 0.2 0.2 0.1 0.1 0.1 Diluent for 0.1 M BF3.(C,H,),0 in dioxan - D N X C D N X Isobutylvinyl etherlml . . 10 10 2 2 2 2 2 2 D = dioxan ; N = nitroethane ; X = 1,2-dichloroethane ; and C = acrylonitrile The effect of solvents on the rate of cationic polymerisation has been the subject of considerable investigation (see, for example, reference 9). Increasing the dielectric constant of a catalyst solution promotes the separation of ion pairs and allows the cation and monomer to react more readily. Consequently, highly polar solvents, such as nitroethane, might be expected to increase both the rate of initiation and of propagation of the polymerisation process.Although the rate of polymerisation is lower in dichloroethane, the absence of a large dilution endotherm, which tends to nullify any temperature rise, results in a sharper end-point inflection than is possible with nitroethane-based titrants. Thus with titrants diluted with 1,2-dichloroethane, the “upturn” temperature at the end-point [method (41 can be measured reasonably accurately, while with titrants diluted with nitroethane only the less reproducible tangent-intersection end-point [method ( 4 1 can be obtained. The ineffectiveness as titrants of 0.002 and 0-001 M solutions of boron trifluoride in dioxan probably results from the Lewis-base properties of dioxan, which reduce the activity of this catalyst.Acrylonitrile appears to have a similar effect on perchloric acid, but not on boron trifluoride for which it is a satisfactory diluent. The temperature rise indicating the end-point of the thermometric titration occurs when a critical concentration of catalyst is exceeded. Blank titrations show that thisFebruary, 19731 INDICATION IN NON-AQUEOUS THERMOMETRIC TITRIMETRY. PART I 87 concentration is reached with less than 0.05 ml of 0.1 M titrant under the conditions used, but with 0401 M titrant about 0.4 ml is required. The amount of titrant needed to reach a required concentration will depend on the volume of the titration solution, and it has been found that with 0.001 M titrants it is desirable to limit the volume of monomer to about 2 ml.The sharp end-points obtained with 0.001 M titrant (Figs. 2, f and 3, h) would suggest that even more dilute titrants, such as 0.0001 M concentration, could be used, and organic bases in amounts as low as 0.000 01 mequiv, e.g., about 1 pg or less of morpholine, could be determined. However, with 0.0001 M titrant, there would be difficulties in achieving the critical catalyst concentration needed to initiate polymerisation, and it would probably be necessary to use a volume of monomer of much less than 2 ml, thereby limiting the amount of heat available from polymerisation. At this level, then, the manual titration method would no longer be suitable and it would be desirable to have constant temperature conditions in the vicinity of the apparatus. 1 2 0.1 M Boron trifluoride in dioxadml Fig.4. Effect of water on the thermometric titration of morpho- line (10 mg in 1 ml of toluene plus 10 ml of isobutyl vinyl ether) with 0.1 M boron trifluoride in dioxan (titration Method A): a, no water added; b, + l o mg of water; c, +22 mg of water; d, +95 mg of water; e, +120 mg of water; f, +104mg of water + 0.5 g of molecular sieve 4A; and g, +73 mg of water (only 2 ml of isobutyl vinyl ether). Points on curves have been omitted for clarity Small amounts of impurities can have an adverse effect on polymerisation processes that proceed by chain-reaction mechanisms. In an earlier papera it was shown that both water and methanol inhibited the polperisation that is catalysed by perchloric acid, but that a fairly high proportion of methanol could be tolerated and water could be eliminated by addition of molecular sieve to the titration solution.The presence of water also affects the efficiency of the boron trifluoride catalyst. As shown in Fig. 4, it reduces the sharpness of the end-point inflection but not as markedly as with perchloric acid. In addition, there is a “drift” of the end-point to a higher titre; however, this drift is small even when the mass88 [Analyst, Vol. 98 of water is equal to that of the morpholine in the sample, and can be reduced by increasing the volume of the monomer, i.e., by decreasing the concentration of water in the monomer, or, of course, by reducing the sample size. Methanol has a much smaller effect on boron trifluoride than on perchloric acid, and sharp end-points were obtained in titrations of 1 + 100 morpholine - methanol mixtures. Ketones, aldehydes and esters do not appear to interfere in the titration procedure.A range of organic bases that were titrated by using the two titrant - monomer systems, with 0.1 M titrant in each instance, is listed in Table 11. For these determinations, solutions of titrants were standardised with solutions of morpholine in toluene. Perchloric acid was also standardised against a solution of potassium hydrogen phthalate in dry acetic acid, by using both an indicator (crystal violet) and the thermometric method to measure the end-point. The results obtained for nominally 0.1 M perchloric acid were similar for all three methods. GREENHOW AND SPENCER: IONIC POLYMERISATION FOR END-POINT TABLE I1 ORGANIC BASES TITRATED WITH 0.1 M PERCHLORIC ACID IN ACETIC ACID AND Conditions: 0.1 mequiv of base in 1 ml of solvent (toluene, dichloroethane, nitroethane, acetone or acetic acid) added to 10ml of the appropriate monomer in a 50-ml beaker and titrated by using Method A BORON TRIFLUORIDE DIETHYL ETHERATE IN DIOXAN Aliphatic amines- n-Butylamine (1 : 1 : 1) ; morpholine (1 : 1 : 1) ; triethylamine (1 : 1 : 1) ; tris(hydroxymethy1)- methylamine (1 : 1 : 0) ; and 6-aminocaproic acid (1 : 1 : 1) Pyridine (1 : 1 : 1) ; a-picoline (1 : 1 : 1) ; 4-vinylpyridine (1 : 1 : 1) ; quinoline (1 : 1 : 1) ; 8-hydroxy- quinoline (1 : 1-4 : 1) ; and 2,6-lutidine (I : 1 : 0.9) p-Toluidine (1 : 1 : 0-7) ; p-nitroaniline (1 : 1 :0) ; and P-hydroxyaniline (1 : 1.2 : 0) o-Phenylenediamine (2 : 2 : 1) ; m-phenylenediamine (2 : 2 : 1.3) ; p-phenylenediamine (2 : 2 : 1.5) ; 2-amino-4-methylpyridine (2 : 1 : 1.3) ; 8-aminoquinoline (2 : 2 : 1.3) ; and hydrazobenzene Pyridine derivatives- Aniline derivatives- Difunctional aromatic amines- (2 : 2 : 0) Heterocyclic nitrogen compounds- Imidazole (2 : 1 : 1.2) ; benzimidazole (2 : 1 : 1) ; benzotriazole (3 : 1 : 1) ; quinoxaline (2 : 1.4 : 0.5) ; and 2,3-dichloroquinoxaline (2 : 0.8 : 0) Acetamide (1 : 1 : 0) ; dimethylformamide (1 : 1 : 0) ; diethylformamide (1 : 1 : 0.6) ; NN-dimethyl- acetamide (1 : 1 : 0.4) ; dimethyl sulphoxide (1 : 1 : 0-7) ; and hexamethylphosphoramide ( 1 : 1.5: 1)* Figures in parentheses following the name of the base denote the theoretical number of basic functional groups in the molecule, the number of groups titrated with the 0.1 M perchloric acid and the number of groups titrated with the 0.1 M boron trifluoride, respectively.* On the basis of the amido groups the functionality will be 3. Amides and sulphur and phosfihorus derivatives- Both perchloric acid and boron trifluoride titrants can be used to determine primary, secondary and tertiary aliphatic amines, and simple pyridine and quinoline derivatives. Only perchloric acid reacts stoicheiometrically with the monofunctional aniline derivatives examined ; boron trifluoride gives titration values ranging from zero, with 9-nitroaniline, to about 70 per cent. of the theoretical requirement, with 9-toluidine. With difunctional organic bases the titration values obtained depend on the amount of interaction of the two groups, i.e., the influence each group has on the reactivity of the other.With perchloric acid both amino groups in o-, m- and 9-phenylenediamine are deter- mined, while with boron trifluoride only one of the groups in o-phenylenediamine is deter- mined and, under the titration conditions used, titration values with the m- and $-isomers are obtained corresponding to about one and a half amino groups, which suggests that the forces of attraction for the boron trifluoride exerted by the monomer and the second amino group of the m- and p-isomers are similar. Both titrants can be used to determine 'the amino group in 6-aminocaproic acid (6-amino- n-hexanoic acid) but application of the method to aliphatic amino-acids in general has been found to be influenced by the solubility of the acids in the solvent system and low resultsFebruary, 19731 INDICATION IN NON-AQUEOUS THERMOMETRIC TITRIMETRY.PART I 89 were obtained with glycine, for example. Boron trifluoride reacts stoicheiometrically, in a 1 : 1 ratio, with 8-hydroxyquinoline but does not react with p-hydroxyaniline. On the other hand, perchloric acid reacts with both compounds non-stoicheiometrically in excess of the 1: 1 ratio, indicating that some reaction occurs with the hydroxyl groups. Both groups in hydrazobenzene can be titrated with perchloric acid but boron trifluoride does not react. With 2-amino-4-methylpyridine, perchloric acid reacts stoicheiometrically in a 1 : 1 ratio, whereas with 8-aminoquinoline, in which the functional groups are on separate rings, the reaction is in the ratio 2 : 1.Rather surprisingly, boron trifluoride reacts non-stoicheio- metrically in a ratio exceeding 1 : 1 with both of these difunctional bases. The cyclic bases imidazole, benzotriazole, quinoxaline and 2,3-dichloroquinoxaline have been examined. Perchloric acid reacts in a 1 : 1 molar ratio with the first two compounds, but non-stoicheiometrically with the other two, the ratios being about 1.4: 1 with quinoxaline and only about 0 4 : l with the dichloro compound. Boron trifluoride does not react with this last compound, as might be expected with a compound containing two strongly de- activating halogen atoms. With quinoxaline the reaction occurs in a molar ratio of about 0.5: 1, but with imidazole the ratio exceeds 1 : 1, in contrast to the ratio of 1 : 1 with perchloric acid.In contrast, benzotriazole does not react with boron trifluoride. Some compounds that have high electron-pair donicitiess have been titrated. Dimethyl sulphoxide, diethylformamide and NN-dimethylacetamide react stoicheiometrically with perchloric acid but sub-stoicheiometrically with boron trifluoride. Hexamethylphosphor- amide reacts with perchloric acid and boron trifluoride in the ratios of 1 : 1.5 and 1 : 1, respec- tively, which accords with the reported high donicity of this compound. Dimethylformamide and acetamide, which are both lower on the donicity scale, react stoicheiometrically with perchloric acid but do not react with boron trifluoride.The reactions described in this paper are competitive, the organic bases competing with the monomer for the titrant - catalyst. With perchloric acid in combination with a-methyl- styrene, the reactivity of organic bases is, in general, so much higher than that of the monomer that stoicheiometric reaction of the base with the titrant occurs before polymerisation. With boron trifluoride in conjunction with isobutyl vinyl ether, however, the difference in reactivity between many bases and the monomer is not so marked and sub-stoicheiometric reactions are more common. It is clear that different results would be obtained in the sub-stoicheio- metric reactions if other monomer - catalyst combinations were used. We have examined styrene and isoprene as possible alternative monomers in combination with the present catalysts but both were unsatisfactory under the conditions used. When sub-stoicheiometric reactions occur, i.e., when the reactivities of the organic base and the monomer with respect to the titrant are similar, the process is kinetically controlled. This effect has been noted with 9-toluidine for which the reaction ratio with 0.1 M boron trifluoride solution is 0.7 when 10 mg of sample are titrated in 10 ml of monomer and 0-58 when 20 mg of sample are titrated in 10 ml of monomer; to obtain more exact information about the reactivities and related properties of the bases it will be necessary to carry out a detailed kinetic investigation. This thermometric method can therefore be used to study the basic properties and, indirectly, the structures of organic bases. The reaction ratios given in Table I1 are an approximate indication of the basic properties of the organic bases examined when non- stoicheiometric reaction has occurred. For example, the results show that e-toluidine is more basic than 9-nitroaniline, the ratios being 0.7 and zero, respectively, with boron trifluoride titrant, under the conditions used for the titrations. 1. 2. 3. 4. 5. 6. 7. 8. 9. REFERENCES Keily, H. J., and Hume, D. N., Analyt. Chem., 1956, 28, 1294. Forman, E. J., and Hume, D. N., TaZanta, 1964, 11, 129. Vaughan, G. A., and Swithenbank, J. J., Analyst, 1967, 92, 364. Vajgand, V. J., and Ga&l, F. F., Talanta, 1967, 14, 345. Vajgand, V. J., Kiss, T. A., GaAl, F. F., and Zsigrai, I. J., Ibid., 1968, 15, 699. Greenhow, E. J., Chew. & Ind., 1972, 466. Gutmann, V., Chem. Brit., 1971, 7 , 102. Plesch, P. H., Editor, “The Chemistry of Cationic Polymerisation,” Pergamon Press, London, 1963. Received May 16lh, 1972 Accepted September 25th. 1972 -, Ibid., 1972, 422.