rate lots were purchased to a federal specification (Mil-A192A) and were certified to have a minimum ammonium perchlorate content of 99%. The potassium perchlorate was reagent grade material. As can be seen, a favorable comparison was obtained between methods for analyses of both AP lots. The slight discrepancy in the case of potassium perchlorate cannot be explained although the titanium chloride method gave results over 100%. An overall standard deviation of 0.073% was obtained for the potentiometric method. This compared favorably with the conductance determination of perchlorate for which a standard deviation of 0.074% was obtained ( 4 ) . Applications of the Potentiometric Perchlorate Titration. The primary use of this potentiometric precipitation titration is for assay of various perchlorate salts. In addition, testing of a wide range of other perchlorate-containing samples is feasible. For example, this titration was successfully used for rapidly measuring ammonium perchlorate in certain solid propellants. The perchlorate in both composite and composite-modified double-base propellant formulations was determined following a water extraction of ground samples. The use of aqueous solutions for analysis was necessary since most organic solvents attack the perchlorate ion electrode. Surprisingly, the perchlorate electrode was found to respond to permanganate, periodate, and dichromate. This was verified by preparing individual solutions of each of these ions a t various concentration levels and measuring the potentials with the perchlorate electrode. Plots of emf cs. log concentration are shown in Figure 3. No corrections were made to activities. As can be seen, the plots are quite similar to the perchlorate plot and are linear over about the same concentration range. The dichromate measurements were made a t pH 1. I n addition, solutions of these ions were quantitatively measured by a potentiometric precipitation titration with
MOLESlLlTER
Figure 3. Perchlorate electrode response curves for Cr207? -, C104-, IO4-, M n 0 4 A. B. C. D.
Dichromate 1.OM in acid Perchlorate Periodate Permanganate
tetraphenylarsonium chloride and the perchlorate electrode. Thus, the electrode might be useful as an indicator for potentiometric titrations involving these ions, RECEIVED for review December 4, 1967. Accepted January 12,1968.
Potentiometric Determination of Acid Groups in Acrylic Polymers and Fibers J. R a y Kirby and A. J. Baldwin Chenistrand Research Center, Inc., Research Triangle Park, Post Ofice Box 731,Durham, N . C. A procedure i s described whereby acidic groups in acrylics at the microequivalents level may be measured by nonaqueous potentiometry. A purified ethylene carbonate-propylene carbonate mixture is used as solvent. Tetramethylammonium hydroxide in this solvent provides a superior titrant. Use of platinum electrodes permits titration of acid groups in polymers in the presence of sodium and potassium salts, thus avoiding the problem of alkali metal poisoning encountered with glass-calomel electrodes. Titration of acid groups present as salts is performed after ion exchange treatment with appropriate resins. The technique is sufficiently sensitive to permit differentiation of acid groups of different strengths in the same polymer sample. Typical titration curves are given.
THE NEED to measure functional groups in polymers on a quantitative basis has assumed a role of increasing importance in recent years as the science of polymers has progressed. Quantitative determination of such functional groups is particularly difficult because of the extremely low level at which
they exist in the polymer. Polymers in a molecular weight range suitable for preparing synthetic fibers quite often have less than 100 peq of functional groups per gram of polymer. However, such analyses may play an integral part in understanding the mechanisms of polymer initiation and termination, the role of the catalyst, the degree of incorporation of functional groups and, in general, one’s ability to characterize the polymer. A broad knowledge of the types and amounts of functional groups is important to the dye chemist in order to develop suitable dyestuffs and dyeing techniques for yarns and fabrics. In the specific case of acrylics, the type and quantity of acidic groups present are of primary interest. A quantitative measure of the acid group content of acrylics is particularly useful in predicting the basic dye acceptance--i.e., the degree of acceptance of basic dyes (cationic dyes) by acidic sites on the polymer or fiber. In this paper the term acrylic polymer is to be considered synonymous with acrylic fiber unless otherwise specified. VOL. 40, NO. 4, APRIL 1968
689
Bacon ( I ) was among the first to use the persulfate-bisulfite redox system for acrylonitrile polymerizations. The mechanisms of free radical generation and subsequent initiation by this and other redox systems have been discussed in a review (2)which points out that initiation in this manner would result in strong acid end groups on the polymer. Polymer recipes for the incorporation of strong and weak acid monomers into acrylonitrile homopolymers or copolymers, which give acid groups along the polymer chain, may be found in the patent literature (3-5). The sulfonate and sulfate groups which arise, either from such free radical initiation or copolymerization of an acid monomer, may be present in either the acid or salt form; Le., -S03H or -S03M and -oS03H or -oS03M, depending on the conditions of polymerization and polymer recovery. Quite often the cations associated with such salt formation are sodium and potassium. The saponification of nitrile groups during polymerization to carboxyl groups has been suggested by Russian workers (6). Beckmann and Glenz (7, 8) were the first to report the titration of acidic groups on a n acrylic polymer. The technique used by them initially was a measurement of the quantity of alkali which could be bound by the polymer. DRALON was placed in aqueous NaOH or ammonia solution and later the amount of base “absorbed” by the polymer was determined by titration of the excess base with HC1 using methyl red as indicator. Results by this method showed a precision of + 20%. Next, direct titrations with base after dissolution of DRALON or PAN in dimethylformamide (DMF) were tried. The results by this procedure, using alcoholic NaOH as titrant and phenolphthalein as indicator, showed a precision of + 30%. Still later they reported a potentiometric method for acidic group determinations of DRALON and PAN. In was the improved method for which a precision of claimed, the polymer was titrated directly in D M F with aqueous NaOH using a glass-calomel electrode system. Their work also included titrations of copolymers of acrylonitrile and acrylic acid. The shape and potential break region of the potentiometric curves were similar for DRALON, PAN, and the copolymers. This suggests that the acid groups being titrated under the conditions of their experiments were, in all instances, carboxyl. Another approach reported for the titration of acidic groups on acrylic polymers is from the patent literature (4). In this procedure, the polymer is dissolved in D M F and ion-exchanged over first Amberlite MB-3 resin and then Amberlite IR120 resin prior to titration. Amberlite MB-3 resin is a mixed bed resin of IR-120 cationic resin in the hydrogen form and IRA-410 anionic resin in the hydroxyl form. The IR-120 portion of the mixed bed resin serves to convert salt groups of the polymer or residual catalyst salts to the corresponding acids. The IRA-410 portion serves to neutralize the nonpolymeric acids generated during exchange over IR-120. The usefulness of further treatment over IR-120 separately is not immediately apparent. The polymer solution was then (1) R. G. R. Bacon, Trans. Faraday Soc., 42, 140(1946). (2) R.G. R. Bacon, Quart. Rev. (London),9,287 (1955). ( 3 ) F. R. Millhiser (E. I. duPont de Nemours and Co.), U.S. Patent 2,837,501,June 3, 1953. (4) E. I. duPont de Nemours and Co., British Patent 823,345, November 11, 1959. (5) G. F. D’Alello, (Industrial Rayon Co.) U.S. Patent 2,531,408,
November 28, 1950. (6) E. S. Reskin, A. A. Kharkharov, and A. L. Shapiro, J. Appl. G e m . U.S.S.R., 32, 1601 (1959). (7) W. Beckmann and 0. Glenz, Melliand Textilber., 38,296 (1957). (8) Zbid., p 783. 690
ANALYTICAL CHEMISTRY
titrated, using alcoholic K O H as titrant, either with visual indicator or potentiometrically. The type of electrodes used was not specified. Using this procedure, it was claimed that alkyl sulfate and alkyl and aryl sulfonate groups on the polymer would titrate as strong acids to a first potentiometric inflection and carboxylic acid groups on the polymer would titrate to a second inflection. N o titration curves nor a statement of the accuracy or precision to be expected from this method was given. Titrations of polymer samples without ion exchange were not indicated. A differential infrared technique has been used to detect the presence of sulfonate, sulfate, and carboxyl groups in certain acrylonitrile polymers (9). Absorbance response to change in group concentration is too small for this technique to be particularly useful as a quantitative method. Tsuda ( I O ) has studied the persulfate-bisulfite catalyzed polymerization of acrylonitrile using NaHS3j03 and K2S23508tracers. A rather involved isotope dilution method was used to establish the sulfonate and sulfate content of the polymer. No mention of solvents other than D M F or titrants other than the alkali metal hydroxides for the titrations of acidic groups in acrylic polymers could be found in the literature. Cundiff and Markunas (11) found D M F not usable as a solvent in titrations involving strong acids. Three potentiometric inflections were noted in the titration of sulfuric acid, one being attributed to titration of formic acid from the hydrolysis of DMF. The use and advantages of quaternary ammonium hydroxides as titrants in nonaqueous titrations of strong and weak acids rather than alkali metal hydroxides has been reported (12-15). In particular, this type of titrant could be used in conjunction with the glass electrode in the highly alkaline region where sodium and potassium ions from the alkali metal hydroxide titrants normally cause interferences. The use of either special salt bridge systems or a modified calomel electrode was recommended for best results. Gibson (16) had previously investigated a number of solvent and titrant systems, including D M F and alcoholic KOH, for potentiometric determination of strong and weak acid groups in acrylic polymers. The following procedure resulted from this study. A mixture of ethylene carbonatepropylene carbonate (80/20 by weight) was used as solvent with tetramethylammonium hydroxide in isopropyl alcohol as titrant. Ion exchange of the polymer sample over mixed bed Amberlite MB-3 resin was used to remove catalyst residues and to convert polymer salt groups to their corresponding acids, Potential changes during titration were followed using a glass and modified calomel (KCI replaced with a quaternary ammonium chloride salt) electrode system. Recent efforts in this laboratory have been directed a t improvements of Gibson’s technique. Certain limitations had become apparent as investigations of acrylic polymer systems progressed. For one example, the mortality rate of electrodes was extremely high, The initial titration curves with a new pair (9) R. Yamadera, J . Polymer Sci. L , Issue 153, S3 (1961). (10) Y. Tsuda, J . Appl. Polymer Sci., 5, 104 (1961). (11) R. H. Cundiff and P. C. Markunas, ANAL.CHEM.,30, 1447 (1958). (12) G. A. Harlow, C. M. Noble, and G. E. A. Wyld, Zbid., 28, 787 (1956). (13) R. H. Cundiff and P. C . Markunas, Zbid., p 792. (14) Zbid., 30, 1450 (1958). (15) L. W. Marple and J. S . Fritz, Ibid., 34,796 (1962). (16) Snell-Hilton, “Encyclopedia of Industrial Chemical Analysis,” Vol. 4, Interscience, 1967, p 293ff.
of glass-calomel electrodes were generally well-defined with clear inflection points for both strong and weak acids. With use, however, the electrode response became sluggish, and the definition of the curves became progressively poorer. Much of this behavior was believed to be due to poisoning of the glass electrode by traces of alkali metal ions, despite attempts to keep such contaminants out of the system. The ineffectiveness of glass electrodes in nonaqueous systems containing mere traces of sodium or potassium ions has been reported by Harlow (12, 17, 18). Evidence of chemical attack on the modified calomel electrode by the quaternary ammonium chloride salt was also indicated. Apart from the above considerations, there existed a major limitation on research measurements. The desirability of knowing the free acid content of certain polymers prior to ion exchange treatment was obvious. Yet, with the glass-calomel electrode system, such measurements were precluded because of the presence of alkali metal salts in the polymers. Furthermore, even with ion exchange treatment of the polymer solutions, the response of the glass electrode was uncertain, particularly in the weak acid region. This necessitated the frequent replacement of electrodes to achieve consistent, reliable data. Consequently, the search for a new, more stable, electrode system was instituted. EXPERIMENTAL
Apparatus. A Corning Model 12 or equivalent pH meter is used, The reference electrode is a titrant delivery tip with an internal platinum electrode (such as No. S-29703, E. H. Sargent and Co.). The platinum-foil indicating electrode (such as No. S-30515, E. H. Sargent and Co.) is immersed in near-boiling 6N HC1 for 4 hours prior to use and after each 10 determinations. Just before each titration, the electrode is rinsed with water and immersed in chromic acid cleaning solution for 5 minutes. It is then rinsed with water. The electrode is connected as the anode and another piece of platinum as the cathode in 1 aqueous sulfuric acid solution, and 4.5-volts direct current is applied for 1 minute. The electrode is rinsed thoroughly with water and then with acetone and allowed to dry. It is now ready to use for one titration. After a titration, the residual polymer solution is removed by rinsing the electrode with acetone. A Gilmont syringe-type microburet equipped with a I-ml reservoir to deliver 0.0100-ml increments of 0.1N titrant is used. The chromatographic-type glass columns are approximately 24 inches long and have a coarse-fritted-glass disk and a stopcock on the lower end. Two columns are needed that have the following approximate inside diameters: 0.75 and 1.5 inches. Reagents. Activated carbon, Nuchar C-l90N (Matheson, Coleman, and Bell), is used to purify the solvent. Analytical reagent ion exchange resins are used--e.g., Amberlite IRA410, chloride form, and Amberlite IR-120, hydrogen form. Tetramethylammonium hydroxide pentahydrate (such as that obtainable from K & K Laboratories, 177-10 93rd Avenue, Jamaica 33, N. Y . ) should be used. Prepurified cylinder nitrogen is used. Ethylene carbonate, spectro grade, (Matheson, Coleman, and Bell) is recommended; practical grade material may be used but it requires additional purification. Practical grade propylene carbonate may be used. A 0.04% aqueous solution of bromocresol green sodium salt is prepared. Preparation of Columns and Solvent. Three parts (by weight) of ethylene carbonate is mixed with one part of pro-
(17) G. A. Harlow, C. M. Noble, and G. E. A. Wyld, ANAL.CHEM., 28, 784 (1956). (18) G. A: Hariow, [bid., 34, 148 (1962).
pylene carbonate. Ethylene carbonate is solid a t room temperature so that warming is needed to make the ethylene carbonate-propylene carbonate mixed solvent. The 3 :1 ratio prevents solidification at room temperature, but an 8 :2 ratio does not always accomplish this purpose. One pound of Nuchar C-l90N activated carbon is added per 15 kg of solvent (only the propylene carbonate needs treatment with carbon if spectro grade ethylene carbonate is used). The mixture is stirred for 24 hours with sufficient agitation to prevent the carbon from settling. It is filtered through Whatman No. 40 paper to remove most of the carbon and then filtered again through S&S No. 576 paper. Carbontreated solvent must be further treated with ion exchange resins before use as a titrant solvent or polymer solvent. One pound of Amberlite IRA-410 resin, chloride form, is transferred to a glass column (1.5-inch i d . ) . About 1 liter of 4 z aqueous NaOH solution is poured through the column. The solution is drained and washed with distilled water until the washings are neutral to phenolphthalein indicator. Distilled water is allowed to stand over the resin for 0.5 hour and then it is tested again. The procedure is repeated with fresh water until a neutral test is obtained, after which the water is drained off. Carbon-treated solvent (500 ml) is passed through the column at a rate of about 20 ml per minute. This treatment is necessary to prevent undesirable side-effects when isopropyl alcohol is added to the resin in the following step. Reagent grade isopropyl alcohol is poured through the column at a rate of about 15 ml per minute until the effluent shows no more than 0.2% water as determined by Karl Fischer titration; several liters of isopropanol may be required. The column is stoppered with an Ascarite absorption tube, the effluent end of the column is connected to a vacuum source, and air is drawn through the column for 1 hour. The dry resin is then stored. One pound of Amberlite IR-120 resin, hydrogen form, is treated in a similar fashion, but the treatment with NaOH solution is omitted. Methyl orange indicator is used instead of phenolphthalein. A solvent column (1.5-inch i.d.) is made by introducing the following materials in the order given: 5 inches of prepared IRA-410 resin, 1 inch of glass wool, 9 inches of prepared 1R-120 resin, and 1 inch of glass wool. The resins are introduced by slurrying with carbon-treated solvent. A sample column (0.75-inch i d . ) is made by introducing the following materials in the order given: 3 inches of prepared IRA-410 resin, 0.5 inch of glass wool, 4 inches of prepared IR-120 resin, and 0.5 inch of glass wool. The resins are introduced by slurrying with carbon-treated solvent. The solvent column and the sample column are each prepared for use in the following manner. Carbon-treated ethylene carbonate-propylene carbonate solvent is poured through the column at a rate of about 20 ml per minute and the effluent is tested for neutrality according to the following test. One drop of 0 . 0 4 z bromocresol green indicator solution is added to two 5-ml portions of effluent: 1 drop of 0.003N HC1, aqueous, must turn one portion yellow; 1 drop of 0.003N NaOH, aqueous, must turn the other portion deep blue. The column is now ready for use. Solvent level is never allowed to drop below the upper glass wool plug. Resin-treated solvent is prepared by passing carbon-treated solvent through the prepared solvent column at about 2 drops per second, collecting the effluent under nitrogen gas. Each 200 ml must be neutral according to the bromocresol green test given above. A new column is prepared when the effluent no longer satisfies the test, usually after treating 5 to 10 kg of solvent. Preparation of Titrant. Tetramethylammonium hydroxide titrant, 0.1N , is prepared as follows. Tetramethylammonium hydroxide pentahydrate (1.81 grams) is dissolved in 100 ml of resin-treated solvent, filtered through medium-porosity fritted glass, and stored in a polyethylene bottle. Exposure to atmospheric moisture and carbon dioxide should be VOL 40, NO. 4, APRIL 1968
691
minimized. It is standardized according t o the Titration Procedure against pure benzoic acid that is dissolved in resintreated solvent. Preparation of Samples. To prepare a sample for titration without ion exchange, 0.5 t o 1.0 gram of acrylic polymer, accurately weighed, is dissolved in 150 ml of resin-treated solvent. Heating may be used if necessary. To prepare a sample for titration after ion exchange, 0.5 to 1.0 gram of acrylic polymer, accurately weighed, is dissolved in 50 ml of resin-treated solvent. Heating may be used if necessary. The solution is passed through the prepared sample column at 80 to 100 drops per minute, collecting the effluent under nitrogen gas in a 300-ml Berzelius beaker. The column is washed with resin-treated solvent until 1 drop of the effluent added to 5 ml of water shows n o turbidity due to precipitated polymer. Final volume is about 150 ml. Before use each day, the sample column should be flushed with resin-treated solvent until a neutral bromocresol green test is obtained. The sample column efficiency can be tested periodically as follows. About 50 peq of titrant and about 50 peq of HCl are added to 50 ml of resin-treated solvent. The solution is passed through the column in the same manner as for a polymer sample. The effluent should be neutral according to the bromocresol green test. Titration Procedure. The buret is connected to the titrant delivery-tip. The indicating electrode, the electrode deliverytip, and a nitrogen delivery tube are immersed in the solution in the beaker and then stirred vigorously with nitrogen bubbles. At least 15 minutes should elapse for the attainment of stable potential readings. Increments of added titrant should be recorded together with the corresponding millivolt readings after electrode equilibrium is approachedusually within 1 minute after adding a n increment. The data may be plotted to locate end points, but it is advantageous to apply the second differential, or derivative, method for determining end points. A blank should be run on each lot of solvent and reagents. However, experience has shown that with proper solvent purification, a blank correction is seldom encountered. DISCUSSION
Solvent Purification and Ion Exchange Treatment. Preliminary titrations in this laboratory had shown that, for the concentration of polymer solution being used, the amount of ionic impurities in the solvent was equal to or in excess of the amount of acid groups attached in a salt form t o the polymer. It was apparent that most of the ion capacity of a resin would be exhausted by the solvent ionic impurities. Therefore, it seemed appropriate to resin-treat the solvent separately and to prepare a relatively large quantity of solvent essentially free of acidic and basic impurities. In this way the entire ion capacity of another resin column would be available for the treatment of polymer samples. Amberlite MB-3 resin was used in initial efforts to purify 5 - to 10-kg quantities of ethylene carbonate-propylene carbonate solvent. MB-3 resin in a column in the presence of ethylene carbonate-propylene carbonate solvent tended to separate into its components, which have different densities, IRA-410 resin being the lighter. Also, on standing in ethylene carbonate-propylene carbonate solvent, IR-120 resin evolved carbon dioxide. Effectively, the solvent or polymer solution was passing through an anionic exchange column first, and then through a cationic exchange column. Thus, with this order of resin treatment, evolved acidic carbon dioxide would not be removed. The difficulty was remedied by using the separate resins and introducing them into a column in the desired order. Thus, IRA-410 resin could be placed in the bottom of a column and 692
ANALYTICAL CHEMISTRY
held in place with a glass wool plug. The IR-120 resin could then be placed in the top of the column and the desired order of treatment could be achieved for solvent or for polymer solutions. A column prepared as described was quite effective in removing acidic and basic impurities from the solvent and in converting salt groups attached to the polymer to the corresponding acid groups. Because of the relatively high level of acidic and basic impurities in the solvent, the ion capacity of a column used for solvent purification was soon exhausted. By treating the solvent with activated carbon-Le., Nuchar C-l90N, Matheson, Coleman, and Bell-prior to resin treatment, the capacity of the column was greatly increased. The described treatment produces a solvent that is water-white and free of appreciable acidic or basic impurities. The treatment produces a solvent superior in purity to the solvent obtained by distillation or freezing. An ion exchange column (0.75-inch id.) composed of about 4 inches each of IR-120 (H+) and IRA-410 (OH-) resins was found efficient for treating more than 50 polymer samples, representing an exchange of more than 2000 peq of polymer salt groups in addition to any residual salts. A similar column (1.5-inch i.d.) packed with 5 inches of IRA-410 (OH-) resin and 9 inches of IR-120 (H+) resin was found efficient for treating 5 to 10 kg of carbon-treated solvent. The effluent from either column was caught under dry nitrogen and thereafter was allowed only minimal contact with the moisture and carbon dioxide of the atmosphere by appropriate use of nitrogen. Titrant. Harlow (19) states that tetramethylammonium hydroxide in isopropanol is by far the most stable of a number of quaternary bases (in isopropanol) which he studied, Evidence obtained in this laboratory indicates a drift in the normality of this titrant over a period of days. Solutions which are stored over a period of weeks at room temperature in polyethylene bottles discolor progressively through yellow to deep brown. Standard solutions, 0.1 N and O.O2N, of tetramethylammonium hydroxide in resin-treated ethylene carbonate-propylene carbonate solvent, which are stored over a period of months at room temperature in polyethylene bottles, hold their titer and remained water-white. These solutions seem eminently suitable as titrants. Electrode System. The usefulness of the platinumplatinum electrode pair for acid-base titrations, particularly for weak acids in nonaqueous solvents, has been demonstrated recently (20, 21). Harlow concluded that the main advantage of the platinum electrode over the glass electrode is freedom from alkali ion errors. Prepolarization of the indicating platinum electrode was recommended and was accomplished by applying 3 volts potential difference for 1 minute to two platinum indicating electrodes immersed in dilute (1 to 100) sulfuric acid solution. The electrode which served as anode was referred to as being anodically polarized. The other electrode was cathodically polarized. The simple platinum reference electrode of Willard and Boldyreff (22) was used. This electrode consists merely of a piece of smooth platinum wire sealed in the buret tip which dips into the solution being titrated. (19) ANAL.CHEM., 34, 1487 (1962). (20) G. A. Harlow, C. M. Nobel, and E. A. W. Garrard, Ibid., 28,784 (1956). (21) F. Pellerin and D. Demay, Ann. Pharm. Franc., 20,661 (1962). (22) H. H. Willard and A. W. Boldyreff, J. Am. Chem. SOC.,51, 471 (1929)
3
2
> E
LL
E
MI
TITRANT
Figure 2. Titration of weak acids 1. Benzoic (monobasic) 2. Malonic (dibasic) 3. Citric (tribasic) MI TITRANT
Figure 1. Titration of strong acids 1. Solvent blank 2. Perchloric acid 3. Sulfuric acid (curve shifted for clarity)
Superior resolution of weak acid inflections, and improved electrode response, was attained with either prepolarization treatment; but the anodic prepolarization gave the better performance. RESULTS
Using essentially the platinum-platinum electrode system described by Harlow, solutions of acrylic polymers dissolved in ethylene carbonate-propylene carbonate could be titrated before and after ion exchange treatment, as the presence of alkali metal ions caused no error using this electrode pair.
To be useful, a potentiometric curve obtained from the titration of an acrylic polymer must give a well-defined inflection for each of the different strengths of acid groups encountered. The primary purpose of this investigation was to attain such curve resolutions. In this work “strong acid groups” refer to polymer groups of the mineral acid type such as sulfonate or sulfate, whereas “weak acid groups” refer to those groups similar in strength to carboxylic acids. T o learn something about the selectivity of the method, for such strong and weak acid types, certain model compounds
0.1 ml
MI TITRANT
Figure 3. Acid groups in polyacrylonitrile 1. AIBN initiated, none (0.506 g polymer titrated) 2. HSzOs-/HS03- initiated (0.589 g polymer titrated) Strong: 34 req per g polymer Weak: 45 req per g polymer
Mi TITRANT
Figure 4. End point details of selected model compounds 1. Methallyl sulfonic acid
2. Lauryl hydrogen sulfate 3. Admixture of 1 and 2 4. Admixture of 1 and 2 plus acetic acid VOL. 40, NO. 4, APRIL 1960
0
693
.i 4
>
4
E
IL
I W
ml . I
LOLO '.
MI
,Ol
TITRANT
Figure 5. AIBN initiated polymers containingco polymerized acid 1. None (0.506 g polymer titrated) 2. Strong acid monomer, sodium-p-sulfophenylmethallyl ether, 10 peq per g polymer (0.627 g polymer titrated) 3. Weak acid monomer, methacrylic acid, 86 peq per g polymer (0.592 g polymer titrated)
ml -I
MI TITRANT
Figure 6. HS208-/HS03-initiated polymers 1. Weak acid monomer, methacrylic acid, strong: 19 peq per g polymer, weak: 160 peq per g polymer (0.504 g polymer
titrated) 2. Strong acid monomer, sodium-p-sulfophenylmethacrylamide ether, strong: 63 peq per g polymer, weak: 22 peq per g polymer (0.643 g polymer titrated)
were titrated. In Figure 1 are shown the titration curves for perchloric acid and for sulfuric acid, together with a solvent blank titration. Note that two inflections were obtained from sulfuric acid in this system. In general, curves shown in this paper are shifted for clarity. Figure 2 exemplifies the curve resolutions obtained for typical weak acids. Note that one, two, and three inflections resulted from the titrations of benzoic, malonic, and citric acids, respectively. Phenol has not been sucessfully titrated in this system. Titration results for acrylic polymers prepared with two quite different initiator systems illustrate well typical titration curves for polymers and the type of information that can be obtained using this method. The following reaction involving catalyst and activator for the persulfate-bisulfite redox system is the logical source of strong acid groups in the absence of an acidic monomer.
> E LL
I W
SzOs-
+ HS03'-
---*
SO4-*
+ HS03* + Soh'-
(1)
If this hypothesis is valid, the same acrylic polymer prepared with a catalyst not capable of undergoing an acid group generating reaction should show a difference in strong acid group content. Azoisobutyronitrile (AIBN) catalyst undergoes the following thermal decomposition : Figure 7. Titrations of a polymer before and after ion exchange 1. Before, strong: 2 Meq per g polymer, weak: 20 peq per g polymer (0.801 g polymer titrated) 2. After, strong: 35 peq per g polymer, weak: 20 peq per g polymer (0.806 g polymer titrated) 694
ANALYTICAL CHEMISTRY
Table I.
Titration Results for Commercial Acrylic Fibers (Ion Exchanged Samples)
Commercial fiber A
Microequivalents of acid per gram fiber Weak Total Strong 58 28 5-1 8-11 17-111
B C
60 0
D E
6 55
14 121-1
74 154
33-11 0 28
6 83
Table 11. Typical Results for an Acrylic Polymer (Ion Exchanged Samples)
Microequivalents of acid per gram polymer Strong Weak Total 54.7 19.7 35.0 54.1 19.0 35.1 55.0 18.7 36.3 53.9 19.1 34.8 54.4 18.9 35.5 54.5 18.4 36.1 54.9 19.8 35.1 19.1 54.5 35.4
Mean Standard deviation 1 0 . 6
10.5
MI
TITRANT
Figure 8. Titration of a commercial acrylic fiber Unidentified strong: 20 peq per g polymer Unidentified weak: 5,8, and 17 p e q per g polymer Total: 58 peq per g polymer (0.991 g polymer titrated)
10.4
Hence, polymers so-prepared would be expected to show no acid groups. Potentiometric titration curves of the two types of polyacrylonitrile are shown in Figure 3. Note that acid groups are present only for the redox-initiated polymer. The polymers had been ion-exchanged prior to titration. The existence of weak as well as strong acid groups for the persulfate-bisulfite prepared polyacrylonitrile was somewhat surprising and not easily accountable to the initiator system. Because sulfonate and sulfate groups may be present in the redox-initiated polymer, it was of interest to compare its titration curve with that of an aliphatic sulfonate and an aliphatic sulfate. Sodium methallylsulfonate and sodium lauryl sulfate were chosen. Solutions of both salts were converted to the acid form by passage over Amberlite IR-120, hydrogen form, ion exchange resin. The end points obtained for each and in admixture are shown in Figure 4. Additionally, the titration curve for the admixture plus acetic acid is shown. These results imply that differentiation of sulfonate and sulfate type acids is not possible by this method and that the weak acid inflection for the polymer could be due to carboxylic acid groups. Strong acid group values for polymers were found to agree well with sulfur content. One possible explanation for the existence of weak acid groups is that hydrolysis of nitrile groups to give carboxylic acid groups occurred with one initiator system but not the other. The AIBN polymer was prepared by solution polymerization in a nonaqueous solvent, whereas the persulfate-bisulfite polymer was prepared by aqueous suspension polymerization.
Examples of the same types of polymers when strong acid monomers (sulfonic acid derivatives) or weak acid monomers (carboxylic acids) are incorporated into the polymer, are shown in Figures 5 and 6. Titration of polymers before and after ion exchange confirmed that many of the acid groups are neutralized to give salt groups during polymerization or subsequent treatment to recover the polymer. Titration curves of a redox-initiated polymer before and after ion exchange to liberate free acid groups are shown in Figure 7. In the absence of ion exchange, essentially no strong acid is available for titration. After ion exchange, the acid level is 35 Meq per gram. Titration results by this method for a number of commercial acrylics are given in Table I. Titrations were performed after ion exchange treatment of the fiber. One of the more interesting titration curves in terms of number and types of acid groups detected is shown in Figure 8. Typical results for an acrylic polymer are shown in Table 11. In view of the extremely low concentration of acid groups measured, remarkable precision was attained. ACKNOWLEDGMENT
We gratefully acknowledge many helpful discussions with R. H. Heidner pertaining to this work. We thank Ronald H. Salmon and Mary G. Yount for their contributions in the performance of much of the laboratory work. RECEIVED for review December 4, 1967. Accepted January 22, 1968. Presented in part at the 12th Detroit Anachem Conference, Detroit, Mich., October 1964.
VOL. 40, NO. 4, APRIL 1968
695