Reaction Of Glyoxal With Boric Acid And Borate Ion PDF

Terrahdron. Vol. 25. pp. 4137 to 4145. Parymon Rem 1969. F’nntbd m Great Bntam REACTION OF GLYOXAL WITH BORIC ACID AND BORATE ION B. PEZTSKY andN.R. ELDRED Research and Developmen...

Terrahdron. Vol. 25. pp. 4137 to 4145. Parymon Rem 1969. F’nntbd m Great Bntam REACTION OF GLYOXAL WITH BORIC ACID AND BORATE ION B. PEZTSKY andN.R. ELDRED Research and Development Department, Union Carbide Corporation, Chemicals and Plastics, South Charleston, West Virginia (Received in the USA 1 November 1968; Received in rhe UKfor publication 15 January 1969) Ah&act-Like many cisdihydroxy compounds, glyoxal promotes the ionization of boric acid, apparently by withdrawing borate ions from solution The association constant of glyoxal with borate ion is 2-g x 10’. The conductivity increment, 1032 n-mhos in 05OOh4gfyoxal plus CGOOMboric acid at 25’, is unusually high. The data give an independent confirmation of the polycyclic structure of aqueous glyoxal. Salts of the new acid have not been isolated. Two ditferent methods used to determine the ionization “constants” gave different results which could bc partially reconciled by assuming that glyoxal reacts with borate ion in preference to boric acid. The constants correspond to molar ratios of glyoxal to borate ion of 32 and 2.7 at 25”. The latter number is probably the average degree of polymerization ofglyoxal in solution. DISCUSSION ALTHOUGH glyoxal is known to exist in hydrated form in water. its known reactions have largely been those of a dialdehyde.’ The discovery that borax increases the acidity of aqueous glyoxal solutions was therefore unexpected. A review of the behaviour of boric acid with bidentate ligands shows many instances of increased acidity, conductivity, and ionophoretic mobility.’ Coordination com- pounds which show such properties are derived from polyhydric alcohols, and phenols, hydroxy acids, and dibasic acids. Increases in acidity in boric acid solutions have generally been attributed to the following reaction : HO-R4Hf-0,e,oH HO-R-OH I&no, = “LO,B,OH + H,O + + H,O+ although it has previously been observed that much of the chemistry of the coordi- nation compound of glycols, at least in aqueous medium, does not involve boric acid as such, but rather its conjugate base, B(OH); (Ref. 2, p. 628) thus B (OH); + HO-R-OH - H”\e/o\ R + 2 Ha0 HOAB’d gives a better representation of the reaction. Glyoxal causes a greater conductivity increase in boric acid than any commercial, aliphatic dial. A few conductivity increments, selected from Steinberg’s text, are pre- sented in Table 1. It has been stated that the primary requirement for enhancement of the conduc- tivity of boric acid is a cis configuration and a planar O-C-C-O conformation.‘* 4 This is vividly illustrated by the fact that cis-l&cyclopentanediol, in which the OH groups are almost eclipsed, enhances the acidity of boric acid (d = 149) while cis-1,2- 4137 4138 B. PESESKY and N. R. ELDRBJ cyclohe~nediol in which the hydroxyl groups assume an equatorial-axial position, does not (A = O).*This difference between these two cyclic dials is further emphasized TABL@ 1. CONDUCTIVITY INCRLWZN~ OF NPICAL LIGANDS (Ref. 2) (L.&and and boric acid each at 0.5OOM, 25.0”) Polyol A Ethylene glycol 0 Glycerol 9 Mannitol 685 cis-1,2-Cyclohexanediol 0 Catecllol 516 Hydroxy acid Glycolic acid 441 (Glyoxal 1032) when we consider the 5000-fold difference in their rates of reaction with lead tetra- acetate which also involves a cyclic intermediate OH OH,k=40,000; OH ,k=8 6- OH 0 However, other factors are important in determining the degree of conductivity enhancement. There is an appreciable inductive effect. Electron donating groups on OH the carbons holding the OH groups reduce the enhancement (cis OH A = 114) while electron withdrawing groups increase the enhancement (cis A = 494). Increased functionality also enhances conductivity as evidenced by the series glycerol (A = 9), erythritol (A = 64) and sorbitol (A = 794). Furthermore, increased acidity enhances the conductivity as the high values for catechol (A = 5 16) and the cc-hydroxy acids (+CHCOOH, A = 19,303) indicate. It is quite probable I OH that all of these factors contribute to the enhancement of the acidity of boric acid by aqueous glyoxal. HO, OH CH’ HO-CH/“CHOH HO-CHOoL~HOok HOH I ‘i HO-CH , ,AHOH HO-L HO’ CHbH I 11 O * The conductivity increment, A, is the difference between the conductivity of the borate-glycol solution and the sum of the condudivities of the components, usually at 0500M concentration.  Reaction of glyoxal with boric acid and borate ion 4139 The planar O-C-C-O structure would not be expected to exist in the monomeric hydrate (I), but would exist in the hydrated dimer or trimer (II, III) if the dioxane ring assumes a “boat” form. The fact that the conductivity increment is very large (see Table 1) suggests that a large portion of the glyoxal exists in the dimer or trimer form. On the basis of the isolation of a derivative of the trihydrate, Raudnitz proposed that hydrated glyoxal exists as a trimer in aqueous solution.’ Simple glycols do not react with boric acid3 presumably because repulsion between adjacent hydroxyl groups in simple glycols gives a skewed conformation. In glyoxal, which in its aqueous monomeric form may be thought of as l,l&?-ethanetetrol (I), adjacent hydroxyl groups would still be expected to assume an unfavourable, skew conformation. The reduction of dissociation constants when aqueous glyoxal is diluted, suggests that. at least in the presence of boric acid, both the cyclic and monomeric forms exist. DETERMINATION OF CONDUCTIVITY INCREMENT The increase in conductivity of 0005M boric acid due to complex formation with glyoxal is shown in Table 2 which also shows that the conductivity increment (A) drops off after initial mixing. The value becomes constant after about 6 hr which is attributed to changes in the degree of polymerization of glyoxal on dilution. When a stock solution of 5*73M glyoxal was mixed with a stock solution of boric acid to give 05OOM glyoxal in 05OOM boric acid at 250”, the conductivity increment was 1642 ~.HI&OSimmediately after mixing and 1032 p-mhos in 64 hr. The decrease in conduc- tivity would be expected if part of the dimer or trimer depolymerized to the mono- meric form which should have a lower conductivity increment because of its confor- mation. To determine whether glyoxal or boric acid was responsible for the change in conductance with time, solid glyoxal (80% glyoxal-20% water) was dissolved and diluted to a concentration of @5M in water and allowed to equilibrate for several days at 25”. Then solid recrystallized boric acid containing 134% water was added to prepare a solution containing 05M glyoxal in 05M boric acid. The boric acid was dissolved as quickly as possible (in a period of about 3 min) and specific conductances were measured with time at 25”. The initial specific conductance, 1050 p-mhos, rose slightly to 1090 p-mhos in several hours. The dilution and subsequent equilibration of TABLET.Co~~~ct~vnues OFPURIWDGLYOXALANDBORICACID (25.0”) GlyoxaI +@50M Boric Acid Glyoxal + O5OMboric acid Molarity Glyoxal alone freshly mixed$ after 64 hr of glyoxal* conductancet Conductance pH Conductance A PH 1.00 57.2 - - - - - 0.83 52Q 34m 2.0 2820 2141 - @50 41.3 1710 2.4 1100 1032 2.6 0.25 29.1 940 2.1 450 394 2.9 0.125 194 480 29 185 139 3.1 0063 12.7 254 3.2 92 52 3-3 0.031 7.7 139 3.5 60 35 3.4 * Molecular weight of glyoxaI taken as 5804. The sample contained 006 per cent acid as acetic acid. t Specific conductance in pmhos. $ Specilic conductance of 05OM boric acid = 267 @OS. 4140 8. PESTSKY and N.R. ELDRBD the glyoxal and not the dilution of the boric acid is responsible for the large decreases in conductivity reported in Table 2. In an effort to determine the ratio in which glyoxal complexes with boric acid, solu- tions of boric acid were titrated with glyoxal and vice versa, and conductivities were determined and plotted. No conclusions about combining ratios could be reached from these curves because no sharp breaks occurred. It was therefore necessary to turn to more sophisticated techniques. DETERMINATION OF “IONIZATION CONSTANTS” With these boric acid complexes, the “ionization constant” is by no means constant because as the ratio of the boric acid to the mono- and di-ligand adduct changes with the changing ligand concentration, the ionizing species changes. This is further complicated by the apparent depolymerization of the glyoxal itself and the slow decrease in conductivity which occurs after the boric acid and glyoxal are mixed. The theory of formation of new acids by association of boric acid and glycols has been discussed by P. J. Antikainen6 who found that the “ionization constants” of such acids often obey an equation of the following form: zc**= C~KlK, + K1 The “ionization constant”, K**, varies with the concentration of the ligand (C,), and is independent of the concentration of the boric acid at concentrations below about O*lM. K, is the association constant of the boric acid and glycol, and K1 is the first ionization constant of boric acid. “Ionization constants” were determined by two different methods: the buffer capacity method of Kilpi” and van Slyke,’ and the half-neutralization or approxi- mate EMF method.g Attempts to determine the ionization constant by conductivity methods failed to yield clear results. The “ionization constants*’ were first determined using the buffer capacity method. TABLET.IONIWnONCONSTANISOPBORlCACIWLYOXAL (254 By half neutralization W5M HWh) (OG5MH,BOI, DlM KCl) MC? -logM, CM, -logCM, PK.’ M, -log& CM, -logCM, pK+* oam 0.000 9.25 0 00 922 @loo l& @040 I+% 7.26 0098 I%0 0.035 l.Y5 7.17 0.200 @697 0.14 0.85 6.21 @195 O-708 u13 @88 625 0.400 0.397 @34 0.47 5.05 0.390 0408 0.33 o-49 5.12 0600 0.221 054 Q27 440 0650 0186 c-59 0.23 4.34 O-800 0096 0.74 @13 3.97 - - - - 1Gw OGOO 0.94 0.03 3.66 0.975 0080 0.91 004 3.78 M, = Molarity of original glyoxal solution. M,, = Mohuity of glyoxal at intktion point. CM, = Corrected molarity, obtained by assuming 2.5 moles glyoxal react with 10 mole of borate- e-c text. pK** = pH of solution at half neutralization. PI = Obtained graphically (ace reference 8).  Reaction of glyoxal with boric acid and borate ion 4141 This procedure is easier to carry out in the presence of salt which reduces polarization of the electrodes of the pH meter and improves the precision of the reading, but the determination was also carried out in the absence of salt. The results are presented in Tables 3 and 4. In the absence of salt, the association constant, K, is 2.8 x lo5 which is obtained by plotting pK** against the negative logarithm of the glyoxal concentration and solving for K, in the straight line equation-log (K+* - K,) = -n log C2 - log K lK,. The slope of the line gives a value for n of 2.73, indicating the number of mole- cules of glyoxal associated with each boric acid molecule. These values were signifi- cantly affected by addition of 0 1M potassium chloride, which yielded K, of 3.6 x lo5 and n = 2.5. Thermodynamic equilibrium constants must be determined in solutions of ionic acitivity approaching zero. Determination of the dissociation constant by the buffer capacity method in the absence of salt requires the addition of some acid followed by back titration, resulting in an ionic strength about OW3 units higher than that resulting from B(OH); and H30+ alone. The amount of correction for the pK** value was only of the order of 0.2 to 04 per cent and was not applied. The “ionization constants” were next determined by measuring the pH of the half neutralized glyoxal-boric acid adduct (see Table 3). This value should be pK*“. and plotting pK** against the log of the glyoxal concentration should yield K, the asso- ciation constant, and n, the number of moles of glyoxal associated with each mole of borate ion. The results with 005 and OlOM boric acid as first calculated, differed greatly. It is apparent, however, that because of the high degree of association between borate ion and glyoxal, as the titration proceeds, generating borate ion, the glyoxal concentration is reduced. Data obtained in the buffer capacity method permitted a simple first approximation correction for the reduced amount of glyoxal, bringing the values obtained at the two different concentrations into reasonable agreement. The corrected data yield a line whose intercept corresponds to an association constant, K, of 4.3 x 10’ while the slope yields a value for n of 3.2, corresponding to 3.2 moles By bulk capacity (O.lM H,BOJ (O-OSM H,BO,. @lM KCI) MQ -log M, CM, -log CM, pIc** M,, -log&, P, x 10’ pK** oal a2 oao 03 9-21 oooo 0.360 891 010 l.CQO 0017 1.77 7.55 0100 1: 8.245 418 020 0697 0074 1.13 668 0199 M97 2068 539 CM0 0397 @27 056 5.36 0398 0.397 4800 466 - - - - - 0665 @175 8095 418 080 0096 0.67 017 4.11 - - - - la oooo 087 006 3.75 0995 OQO2 141 3.67  4142 B. PESIXKY and N. R. ELDR~D of glyoxal per mole of borate ion. As anticipated, addition of 0 1M potassium chloride made no significant difference. Data obtained by the buffer capacity method are believed to be more valid. In the first place, no correction is required to account for ligand removal from solution by the borate, and in the second place, glyoxal is more stable at the lower pH used in the buffer capacity determination. Both methods have been used to determine data in the literature, and authors using one method frequently disagree with authors using the other. Furthermore, data sometimes fail to yield straight lines.” TABLEZ 4. IONIZATION CONSTANTS BY BUFFT?R CAPACITY MJiTHOQ BORIC ACID-GLYOXAL COMPLEX. M) ADDKI SALTS (25.0”) Solution 1 2 3 4 5 6 H,BO,, Molarity 00MO 00500 OQ500 O.OMo 0100 00100 Glyoxal, molarity OMO 0.0993 0199 0993 0199 0199 -log Glyoxal molarity 1.002 0700 0002 0700 0.700 P, x 104 0.307 5.71 148 124 209 8.36 iv* 8.86 x 10-l” 3.08 x lo-’ 2.06 x 10-e 1.67 x lo-* 2.06 x 1o-6 3.30 x lo+ pK** 9-05 451 5.69 3.78 569 548 P, determined graphically, -log K, = 9.24, K, = 5.75 x lo-” The strong effect of borate concentration on the “ionization constant” is not seen with free boric acid. Table 4 shows that the ionization is independent of boric acid concentration when determined by the buffer capacity method (experiments 3,5 and 6). This is further confirmation of the observation that the coordination occurs with B(OH); rather than H,B09. Glyoxal can be used to sharpen the end point in titration of boric acid. Invert sugar or mannitol is usually used for this purpose but glyoxal gives a sharper end point. Because of the internal Cannizzaro reaction, HC==O H,C-OH I +OH-+ 1 HC=G o=C-c the pH of neutralized glyoxal drops spontaneously to about pH 5 in a few hours. A trace of acid in the commercial material prevents this reaction. This behaviour makes it necessary to neutralize glyoxal to pH 7 immediately before titrating glyoxal- boric acid mixtures. The titration should be carried out rapidly to avoid interference by the Cannizzaro reaction. The neutralized commercial glyoxal is satisfactory to sharpen the end point. The dependence of pH on glyoxal concentration permits the use of glyoxal and boric acid to prepare multi-range buffers. Fig. 1 illustrates the use of the buffers covering pH ranges of 3 to 4,4 to 6,5 to 7, and 6 to 8 prepared simply by changing the glyoxal concentration. Several solutions of half neutralized boric acid and glyoxal were aged to determine the stability of pH with time. The pH shifts by as much as 0.1 unit during the first 24 hr but remains stable afterwards. Decomposition of the glyoxal at the higher pH limits the stability of the buffer system.  Reaction of glyoxal with boric acid and borate ion 4143 Attempts to isolate copper, iron, barium, silver, pyridine, and ammonium salts were unsuccessful. Evaporation of stoichiometric solutions of glyoxakboric acid and selected metal salts or bases yielded boric acid crystals initially and no boric acid- glyoxal salts. 12 II IO 9 E ‘,7 6 5 p’ LEQEWD 0.67 4 - 0 Solution0 Contoinad 0.050 Y 1.00 Boric &Id plum Indkatrd Mdalty txyoxal qt etort d titrotlon 3 q --_ 0 1.0 Y Blyoxql. no boric ocld wow: lOOnI- of raoh rolutlon tltmtrd rHh 1.00 It, WoOH. 2 I 1 I I. I 2 3 4 5 6 7 8 Yllllrqulvotrnts NoOH FIG. 1. Titration of boric acid and glyoxal with base (25”). OTHER ALDEHYDES Glutaraldehyde and a-hydroxyadipaldehyde were investigated to determine whether these materials combined with boric acid to enhance its acidity. Commercial samples of aqueous glutaraldehyde and a-hydroxyadipaldehyde having pH’s of 2.9 and 2.8, respectively, were used. At 05M concentration in @5M H,BOJ both materials had a conductivity increment of - 3Ou-mhos which indicated no complexing. The negative values were probably due to the free acid in the aldehydes which decreased the ionization of the boric acid and reduced its contribution to conductivity in the mixture. It has been found in our laboratories l 1 that glyoxal has a greater affinity for cellulose than does glutaraldehyde or a-hydroxyadipaldehyde. Glyoxal also is more strongly hydrated by water. 4144 B. PESETSKY and N.R. ELDRED EXPERIMENTAL Determination of ionization constants. The method of Kil~i’~ was used to determine ionization constants by the minimum buffa capacity or dirrcrential potentiometric method. This method has the advantage of requiring very small additions of acid and base thereby causing a minimum disturbance in the equilibria between hydrated glyoxal monomer and polymer and between glyoxal hydrate and boric acid. Thus to 250 ml of equilibrated OalM boric acid and 02M glyoxal was added @l ml of@lN hydrochloric acid and the minimum buffer capacity was determined by addition of lO@l increments of @ICONsodium hydroxide. A plot of incremental change in pH vs. the increment of acid or base added yields the maximum pH change denoted by d@H),. Niisiinen (12) determined this value with a mathematical equation. The mini- mum buNer capacity is then calculated from the following equation: and P, = 4.606 ,/(K*+Ca) ” =V’iJ[AW),] where AV is the incremental volume of base added, B is the normality of the base, 6 is the volume of solution being titrated at the minimum buffer capacity. K l * is the “ionization constant”, and C, is the initial concentration of acid. K** is calculated by successive approximations. Ionization constants were also determined by measuring the pH of the half neutrahzed glyoxal-boric acid adduct as described by Glasstone.’ Solutions containing various amounts of glyoxal with OOSM and @lOM boric acid and with O.lOM boric acid containing DlM KC7 were titrated with l.ON sodium hy- droxide, recording the pH after each incremental addition of base. Solutions of glyoxal and boric acid were allowed to stand at least 16 hours before titration and were purged with nitrogen to remove carbon dioxide. This method gave pK values for the ionization of boric acid of 9.25922921 (see Table 3) which are in good agreement with the literature. The buffer capacity method is dimcult to apply to boric acid because of interference of carbon dioxide. It is more suitable for use with stronger acids such as the boric acid-glyoxal complexes described here. In many cases minimum buffer capacities are not easily determined during the titration. Addition of salts such as KCI and NaCl accentuates the point of minimum buffer capacity and generally increases the dissociation constant. Thermodynamic constants cannot be determined directly in the presence of salts. To obtain the thermo- dynamic constants from data obtained in the presence of salts, the e%ct of salt concentration on ionization constants is determined and the thermodynamic equilibrium constant is obtained by plotting dissociation constants as a function of the square root of the ionic strength and extrapolating to zero ionic strength. Constants in the Debye-Hiickel equation can then be determined.6*9 Boric acid pur$cation and an&is. Boric acid (Mallinckrodt AR grade) was recrystallized prior to use by dissolving 106 g of boric acid in 300 g of hot deionized water, then cooling The crystals were filtered and washed several times with cold deionized water. Conductivity of 05OOM boric acid in deionixed water before purification was 38.5 u-mhos, and after recrystallixation it was 275 u-mhos. Correction for con- ductivity of the water (@8)leaves a value of 267 for boric acid Boric acid analyses were made by standard techniques using neutralized invert sugar to enhance the end point and phenolphthalein as an indicator. Conductivity of the @5C@Mglyoxal was 41.3 u-mhos, and conductivity of the solution of 05OOM boric acid in 05OOM glyoxal was 1710 when freshly mixed and 1100 after standing 64 hr. The conductivity increments thus became 1710 - (41 + 27) = 1642 and 1100 - (41 + 27) = 1032 Glyoxal purification and analysis. Glyoxal WA solution from Union Carbide Corporation was purified by treatment with ion exchange resins and decolorixing carbon. This increased the pH from 2.3 to 3.1 and reduced tbe acidity from 05 to 006O/,, calculated as acetic acid. Glyoxal content is best determined by neutralizing the sample to phenolphthalein, adding excess standard base, allowing the solution to react for 15 min at room temperature, then back titrating to the phenolphthalein end point.’ Crystallized glyoxal can be prepared by concentrating Glyoxal #“/. to about W/, solids by warming to 60” in uucuo, then allowing the syrup to stand at room temperature for a few weeks. The crystals are tiltered and dried. The product analyzes about 80% glyoxal, corresponding to the trimeric hydrate.”  Reaction of glyoxal with boric acid and borate ion 4145 REFERENCES 1 Anon., General Chemistry ofGlyoxa/, p. 27 Union Carbide Corp.Bullctin 41296A, New York (1967). 2 H. Steinberg, Orgunoboron Chcmisfry, chapters 14, 15 and 16. Wiley, New York (1964). 3 J. Bccsckcn, Ado. Curbohydrare Chem 4,190 (1949). * H. Kwart and G. C Gates, 1. Am Chem Sot. 80,881-883 (1958). ’ H. Raudni& Chem & M. 327,366 (1944). 6 P. Antikaineq Ann. Acad. Sci. Fenn. A. II 56,361(1954). ’ o S. Kilpi, J. Am. Chem. Sot. 74.5296 (1952). b Z. physik. Chem. (A) 173,427 (1935). B D. D. van Slyke, J. Bbl. Chem. 52,525 (1922). 9 S. Glasstone, Introduction to Electrochemisfry, pp. 322-325. Van Nostrand, New York (1942). lo P. J. Antikaincn, Acto Chem. Scund. 13,312 (1959). I1 N. R. Eldrcd and J. C. Spicer, TAPPI 46 (No. lo), 608-612 (1963). If R.NMbn,Suomen Kemistilehti 21, 5 (1948). I3 J. C. Bondiou and B. J. Pctusseau, U.S. Patent 3,016,385 to Society Nobd Bezel, Jan. 9.1962.