PART I THE LOBRÏ de BRUYW TRANSFORMATION OF D-GLUCOSE AND 3 ,ii,6-TRI/4ETHïL-D-FRUCTOSE PART II KINETj.es OF THE METAL ION CATALYZED DEGRADATION OF DL-GLYCERALDEHYDE By Don Si' Mivada. A THESIS Submitted to the School of Graduate Studies of Michigan State College of Agriculture and Applied Science in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1953 Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. PJiRT I THE LOBRY de BRUYN TRilNSFORMATION OF D-GLUCOSE AND 3 6-TRIt4ETHXL-D-FRUCTOSE PART I I KINETICS OF THE M e TAL ION CATALYZED DEGR/.DAT10N OF DL -GLYC ERALDEH YDE By Don si' Miyada an /Ü3STRACT Submitted to the School of Graduate Studies of Michigan State College of Agriculture and Applied Science in partial fulfilLment of the requirements for the degree of DOCTOR 0j?‘ p h i l o s o p h y Department of Chemistry Year 1953 Approved Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Don S . Miyada THESIS ABSl’RACT Tlie base catalyzed degradation of D-glucose was studied kinetically in sodium hydroxi.de solutions by two methods of analysis of the reaction nii^ctures involving specific oxidation and the optical rotation of such mixtures. The implications of the results are discussed. The base catalyzed degradation of 3,h,6-trimethyl-D-fructose in dilute solutions of calcium, barium, and sodium hydroxides was investi­ gated, There appeared to be no difference in the trend of formation of the methylated products with the three bases; however, in solutions con­ taining alkaline earth bases, a steady increase in periodate consumption was observed wliicii was probably caused by the déméthylation of the methylated hexoses, A comparison of the effect by these bases in the enolization reaction indicated that hydroxyl ion catalysis predominates. A new set of values for the catalytic constants of acetic acid and acetate ion was obtained for the general acid-base catalyzed degradation of glyceraldehyde, and the dependency of the reaction rate on ionic strength was determined. The effect of metal ion catalyzed degradation of glyceraldehyde was investigated using the following metal ions : lithium, calcium, barium and magnesium. An acceleration in the rate of reaction was observed in all cases except with magnesium ion where in­ creased metal ion concentration produced a diminution of the pseudo first order rate constant. A mathematical treatment of the catalysis by calcium ion, based on (.1) incomplete dissociation of the species CaOAc^ - 1- Reproduced with permission o f the copyright owner. Further reproduction prohibited without permission. Don S , Miyada and (2) the limitation of catalysis by the calcium acetate system to calcium and acetate ions, was found to give pseudo first order rate constants which compared favorably with experimentally determined values as well as a reasonable dissociation constant for the species, CaOAc'*', — 2— Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. il ACKJMOWLEDÜMENT The author wishes to e^giress sincere appreciation to D r . John C . Speck , J r , for his assistance tliroughout the course of these investigations. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. Xll T/iBLE OF CONTENTS PAGE HISTORICAL INTRODUCTION.......................................... 1 The Lobry de Bruyn - vanEkenstein Transformation............. 1 The Glyceraldehyde-Dihydroxyace-bone-Pyruvaldehyde Transformation......................................... 3 EXPERIMENTAL..................................................... 9 PART I The Lobry de Bruyn Transfoimation of D-Glucose and 3 ,U,6-Trimethyl-D-Fructose EXPERIMENTAL METHODS............................................. 9 Materials................................................ Preparation of 3 ,U,6-trimethyl-D-fructose................... Apparatus................................................... Analytical Procedures....................................... 9 9 13 21 RESULTS AND DISCUSSION........................................... 2k D-glucose................................... 3 ,U,6-Trimethyl-D-fructose.................................. 30 SUMMARY.................................................... 3^ PART II Kinetics of the Metal Ion Catalyzed Degradation of DL -Glycer aldehyde EXPERIMENTAL METHODS............................................. Materials................................................... Apparatus................................................... Quantitative Analysis ofReaction Mixture..................... Analytical Methods.............................. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 36 36 36 36 IV TABLE OF CONTEOTS - Continued PAGE RESULTS ÂHD DISCUSSION............................................ 39 Order of Reaction........................... Calculation of CatalyticConstants....... Effect of Ionic Strength.................................... Effect of Calcium Ion....................................... Effect of Barium Ion ..................... Effect of Magnesium Ion........... Effect of Lithium Ion....................................... Mechanism of Metal IonCatalj'sis.............................. 39 39 LO U2 h9 149 SO Si SUMMARY........ SU LITERATURE CITED.................................................. SS I Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. HISTORICAl INTRODUCTICW Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. HISTORICAL INTRODUCTION The reversible transformation of aldoses and ketoses occurring in dilute alkaline solution was first described by Lobry de Bruyn and Alberda van Ekenstein in 1895. (1,2) These Investigators studied the interconversion of fructose, mannose, and glucose under the influence of dilute solutions of different metallic bases . Their later work in­ cluded investigation of similar transformations of galactose, melibiose, maltose, and lactose. (3) Other investigators have studied the effects of the metallic lydroxides on glyceraldehyde, (U) lactose, (5) glucoheptose, (6) xylose and arabinose, (7) and cellobiose. (8) Organic bases have also been used to catalyze the Lobry de Bruyn transformation, Fischer, Taube, and Baer conveirbed glycer aldehyde into dihydroxyacetone using the base pyridine. (9) Danilov and co-workers reported that glucose was isomerized to fructose without any admixed mannose when heated with either pyridine or quinoline; however, in either aqueous pyridine or aqueous alcoholic quinoline, the rearrangement did not proceed so smoothly and was accompanied by acid and mannose forma­ tion, (10) Midorikawa and Takeshiraa confirmed these observations (11) and extended them to include quinaldine as a catalyst, (12) Anhydrous organic bases, especially pyridine, have been of value in the prepara­ tion of many ketoses from the corresponding aldoses. (13-20) Lobry de Bruyn found that boiling water caused the Isomerization of fructose. Qarbutt and Hubbard demonstrated the interconversion of ^ucose, fructose, and mannose in boiling aqueous solutions buffered at Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. neutrality. (2l) Spoehr and co-workers observed this transformation in presence of neutral and slightly acid phosphate buffers, (22,23) The number of investigators applying different conditions to this basic isomerization reaction is legion, and only a few are mentioned here. Despite the large number of investigations of this reaction, precise data in the literature having to do with its kinetics or mechanism are strictly limited. This is probably due to the fact that the transform­ ation is complicated by fragmentation reactions, acid formation, and rearrangements, all of which were recognized very early. The generally accepted mechanism of the Itohry de Bruyn-van Ekenstein transformation postulates enediol intermediates and mi%r be represented schematically as follows: R R H I C= O I C — OH I R* 1 0= 0 I R I C HO 0 — H 1 R» OH G — OH R H R I R* I OH C HO I 0 — H 1 0= 0 I R» 0=0 I R* [ R - H, OHgOH, etc, ; R' - H, OHgOH, etc, ] Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. As early as 1900, Wohl and Neuberg explained the conversion of glyceraldehyde to dihydroaçyacetone in alkaline solutions on the basis of an enediol intermediate. (1;) Lobry de Bmjyn and, in his earlier writings, Nef, (2it) assumed tlie enediol to be formed by the alternate addition and removal of the elements of water. Later in 1910, Wef (25) assumed a molecular shift as the basis for enediol formation, and the research of Lewis and co-workers with tetramethyl-hexoses and triraethyl-pentoses supported the simpler concept of enolization, (26-30) Since no ketoses were found, it was suggested that only the 1-2 enediol can form in these methylated sugars. In addition, high iodine absorbing substances be­ lieved to be the enediol intermediates were present in alkaline solution but r^idly disappeared on acidification. The investigations of Michaelis and Rona (3l) and Qroot (32) also implied enediol formation. In a series of investigations of carbohydrate oxidations, Evans and his co-workers repeated and extended the observations of Nef by conducting quantitative studies of products formed-under more carefully controlled 8xpeidmental conditions. (33-U?) T h ^ accepted Nef*s postulation of an equilibrium between the sugars and a series of enediols in alkaline solution. Electroreduction of sugars in alkaline solutions by Wolfrom and co­ workers gave isomeric polyols indicating the presence of 1,2- and 2,3enediols. (U9-52) Dei^terium analysis has produced conflicting results. Early studies by Fredenhagen and Bonhoeffer indicated no incorporation of carbon bound deuterium with D-g^ucose at 25® C, (53) On this basis, a "dimer Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. intermediate* was postulated. Goto reported similar results. (51) In the hydroxyl ion concentration. Utilizing the initial rate method, search of an explanation for this reported isomerization without exchange Michaelis and Rona found that the rate of isomerization of D-glucose was of carbon bound hydrogen by deuterium, Bottmer-By and Gibbs employed directly proportional to the hydroxyl ion concentration, (3 I) 1-C“ -D-glucose to test the possibility of carbon chain rearrangement Spoehr and Strain studied the interconversion of glucose to fructose during the reaction. (55) No such rearrangement was observed, Topper in a slightly acidic phosphate buffer. (58) Ashmarin and co-workers and Stetten in a reinvestigation of the reactions of D-glucose in heavy studied the affects of acetate, formate, and sucoinate buffers on glucose water observed deuterium exchange in agreement with the enediol mechanism. and fructose, (59-61) The latter investigators found that, in all (56) From their results, they postulated the following mechanism. HO I 1 h - g - oh 1 H O -C -H i H-C-OH 1 H-C-OH 1 CHjOH D-glucose in catalyzing the transformation, Braun and Konnova reported similar H CHgOH 1 1 C=0 1 'g 1 1 H-C-OH 1 1 HO - G “ H A instances, these anions acted as bases according to the Bronsted theory r - T— H-C-OH 1 H-C-OH 1 1 CHjOH trsns-snediol HO-G-H 1 H-G-OH 1 1 H-G-OH 1 GH3OH D-fructose Y I H-C-OH C-H I HO-G-H I I HO - G - H I Using aqueous pyridine systems, Midorikawa observed that with in­ HO-C-H I H-C-OH I H-C-OH H-G-OH CH 3OH CHjOH cis-enediol findings with an acetate buffer, (62) h - g - oh I D-mannose Bowden and Schaffer (57) working with heavy water presented further evidence for deuterium exchange at 25 C. but concluded that fructose was creasing pyridine concentrations, the conversion of glucose to fructose ' increased to a maximum and subsequently decreased. In analogy with Lowry and Faulkner's results on the mutarotation of glucose, (6 3 ) Midorikawa proposed an acid-base catalysis, (61)) Employing methods based on periodate scission, Forist and Speck have not only demonstrated acidbase catalysis in the interconversion of glyceraldehyde to dihydroxy­ acetone but have determined the kinetics of this transformation under the not a necessary intermediate in the conversion of glucose to mannose. influence of acetate, fomate, and trimethylacetate buffer systems, (61i) The consensus appears to indicate the general acceptance of an enediol Several investigators have found variations in the Lobry de Bruyn intermediate brought about by an enolization of the sugars. The nature and the mode of action of catalysts have received little attention. The majority of the reactions have been carried out in alkaline solutions, and the activity of the bases has been attributed to transformation depending on the particular cationic species involved. Lobry de Bruyn and van Ekenstein reported that P b ( d )3 converted glucose to mannose with no detectable fructose, and that under these conditions fructose was not isomerized to the corresponding aldose. (3 ) Nef reported that enolization of hexose was not caused by calcium and lead acetates Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. or by câloium cliloride . In 1919, Murschauser, in a series of papers, (65) described the effects of alkaline earth carbonates on D-glucose and concluded that dextrose is converted to levulose and eventually into other levorotatory or weakly dextrorotatory sugars. In 1926, Kusin reported that the mechanism of isomerization shows cationic dependence in that different types of enolic intermediates are involved when alkalies containing, respectively, mono- or divalent cations are employed. (66, 6?) A comparative study of the action of calcium hydroxide and sodium hydroxide on D-glucose and D-fructose at 2 ^ C . indicated that with calcium hydroxide, the sugars showed reducing power which persisted for some time after acidification. Such results were not obtained with sodium hydroxide. Calcium hydroxide favored the formation of mannose from glucose, idiereas sodium hydroxide favored the formation of fructose from glucose. Hence, Kusin postulated that, at low temperatures, calcium hydroxide produced a cyclic enol without rupture of the pyranose ring, and sodium hydroxide produced an acyclic enol. At hi^er temperatures, no difference in the action of the two bases was detectable. Recently, Sowden and Schaffer have reported that the nature of the initial course of the isomerization of D-mannose is dependent on the cationic species present. (68) These investigators observed this effect with 0,5 N bases at however, at base concentrations of 0.035 N and at a temperature of 35°C., they found no differences when the cationic species were changed. Wind (69) and Ahlstrom and von Euler (70) reported that oxidations of glyceraldehyde in buffered solutions are catalyzed by heavy metals. Assuming the enediol as the oxLdlzable species, this may Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. indicate increased ability for enediol formation in the presence of these ions. A reaction ^diich complicates the Lobry de Bruyn reaction is the deîiydï'ation of sugars ^ this was observed by Deniges (71) and Fischer and Taube (72) with dihydroxy acetone. However, again, little infor­ mation is available concerning either the kinetics or the mechanism of the transformation of glyceraldehyde and dihydroxyacetone into pyruv alde­ hyde . Various investigators have suggested that a common enediol is involved in the conversion of the trioses to pyruvaldehyde. Among these were Evans and Cornthwaite, (Uo) Strain and Spoehr, (60) and Smith and Anderson. (73) In 193U, Dische and Robbins reported that the addition of phosphate or arsenate catalyzes the transformation of glyceraldehyde and dihydroxyacetone to pyruvaldehyde. (7U) The addition of other ions in the form of calcium chloride, sodium fluoride, sodium citrate, sodium sulfate, copper sulfate, ferric chloride, or ferric sulfate in appropriate concentrations had no effect on this reaction in neutral solutions. Recently, Forist and Speck have shown that a general acidbase catalysis exists for this reaction also. The present work is comprised of a reinvestigation of the alkaline degradation of reducing sugars, and the effects of metallic ions in the Lobry de Bruyn transformation by «application of new analytical methods capable of a high degree of precision to these reaction mixtures. In this, hexoses were chosen for Investigation of the reactions occurring in alkaline media for the reason that side reactions, such as aldoliz ation and dehydration, appear to be slower relative to enolization than Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 8 they are with simpler sugars. Moreover, the higher carboned reducing sugars possess large and well established specific rotations, thus making possible examination of the systems for stereospecific effects. In order to gain further insight to the Lobry de Bruyn transformation of a substance with which dehydration effects are minimal, the work has been extended to include examination of the isomerization of 3,U,6“trimethyl-D-fructose as catalyzed by sodium, calcium, and barium hydroxides, The investigation of the effects of metal ions in the Lobry de Bruyn transformation under acidic conditions has been carried out with glyceraldehyde in acetate buffers. This triose reacts at conveniently measurable rates under these circumstances at moderate temperatures and concentrations, and dehydration proceeds to a definite product, pyruvaldehyde, which permits a detailed analysis of the system. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. EXPüKlMia^TAL Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. PART I The Lobry de Bruyn Transformation of D-glucose and 3 ^6-trimethyl-D-fructose Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. EXPERIMENTAI METHODS Materials The D-glucose (76) and D-fructose (77) employed in these measure­ ments were prepared according to Bureau of Standards procedure. Matheson Company practical grade chromotropic acid was recrystallized from ethanol. Periodic acid, porchloratoceric acid, and nitroforroin were obtained from G. Frederick Smith Chemical Company. Matheson's practical grade dimethylsxilfate was vacuum distilled before using. Inulin was a Nutritional Biochemicals Corporation product. Pyridine and chloroform ifere Bakers analytical reagent grade chemicals. Acetic anhydride was an Eastman KodaJc Company white label product and phenylhydrazine was an Eastman Kodak Company yellow label product. Acetone, oxalic acid, calcium carbonate, calcium hydroxide, barium carbonate, barium hydroxide, sodium hydroxide, sodium bicarbonate, sodium sulfite, hydrochloric acid, anhydrous sodium sulfate, anliydrous magnesium sulfate, phosphorous pentoxide, arsenious acid, potassium iodide, and iodine were all C. P. reagents. Sulfuric acid was Merck reagent grade. Preparation of 3,1:,6-trimethyl-D-fructose In order to prepare 3,ii,6-trimethyl-D-fructose, the general procedure has been to methylate inulin, a natural fructosan, hydrolyze the methylated product, hydrolyze the resulting methyl-fructoside and then isolate and purify the subsequently formed 3,1^^6-trimethyl-D-fructose . There are Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 10 two distinct methods for methylating inulin; they will be designated method A and method B. Method A was that described by Hirst, Me Oilvray, and Percivsl (78) and involves an initial méthylation of inulin by dimethylsulfate followed by repeated méthylations with methyl iodide and silver oxide. In method B, the procedure of Haworth and Straight, (79) trimettylinulin is prepared by dimethyl sulfate methylation of triacetyl-inulin. Prelimjjiary experiments indicated that the latter method was the more practicable, The following is a description of the slight modification of Haworth's method which was used for preparing trimethyl-fructose. (1) Swelling of Inulin. One hundred grams of inulin was vigorously stirred with a liter of pyridine at 80°C. for approximately two hours. During this period, the mixture became a clear, yellowish-green solution, The solution was then cooled with continued stiiring. (2) Acétylation of Inulin. Two hundred milliliters of acetic anhydride was gradually added over a period of six hours to the above material, representing 100 g. of anhydrous inulin. During the addition, the re­ action mixture was stirred vigorously and maintained at 20°C. This was followed by the gradual addition of 370 ml. of acetic anlxydride with stirring at 20^0, The resulting solution was allowed to stand 12 hours, The supernatant liquid was then poured into 15 liters of ice water, A white precipitate of crude triacetyl-inulin formed immediately. This was filtered in portions and repeatedly washed with distilled water to Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 11 remove pyridine and acetic anhydride, The product was dried in air and then placed over potassium hydroxide for further drying. crude The yield of product was 202 g., 97 per cent of theory. (3) Preparation of Trimethyl-lnulin. Twelve grams of triacetyl-inulin was dissolved in 250 ml. of acetone in a two liter flask fitted with a Hirschberg stirrer. The temperature was kept at 55°C. during the gradual addition of 120 ml. of dim ethylsulf ate and 320 ml. of 3O per cent, aqueous sodium hydroxide. One-tenth of the volume of each reagent was added every ten minutes. After the second or third addition, an emulsion formed wiiich persisted until the end of the reaction. During the formation of the emulsion, the temperature rose rapidly unless the reaction mixture was cooled. Wlien the addition of dimetljylsulfate and sodium hydroxide was complete, ICO ml. of water was added, and the temperature of the reaction mixture was raised to 75°C. for a period of 15 minutes to distil off the bulk of the acetone. During this operation, trimethyl-inulin precipitated in the form of pale yellow pellets. This solid product was separated from the warm reaction mixture by filtration. It was then digested three times for two hours with 70-ml. portions of boiling water. After trituration of the resulting product with acetone and ether, the methylated inulin was obtained as a fine, white product. Methylated inulin is soluble in acetone, hence only 5 ml. of acetone was used in this operation. Repeated trituration with l5-ml. portions of Although recrystallization of triacetyl-inulin can be accomplished from hot methanol, it was found that the large losses involved made such purification iitqpractical at this stage of the preparation, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 12 ether removed the colored impurities. The above procedure appears to give the best results, inasmuch as a. batch using twice the amounts of reagents gave an incompletely methylated product. (79) The melting point of the product was li|0°C . vdiich is in agreement with values re­ ported by previous investigators. (79) The average yield was 7.1 g. of trimethyl-inulin, 97 per cent of the theory based on triacetyl-inulin, (ii) Hydrolysis of Trimethyl-inulin. Trimethyl-inulin was hydrolyzed in 5-g. portions by digesting with 2^0 ml. of 70 per cent alcohol in whicli was dissolved 2.5 g. of oxalic acid. The digestions were carried out at 80°C. for from eleven to seventeen hours. The shorter period gave a poor yield, possibly due to incomplete hydrolysis. The oxalic acid solution was tlien neutralized with calcium carbonate and the solution was allowed to stand for several hours. The supernatant liquid was decanted through a fluted filter paper and the residual mixture was centrifuged. The liquids were combined and evaporated under reduced pressure (l5 mm.) at iiO°C . The resulting light syrup was extracted repeatedly with chloroform. The chloroform extracts were combined, and dried over anliydrous sodium sulfate and again reduced to a syrup. Because of a possible presence of a fructoside, the syrup was digested with 0,25 per cent, aqueous hydrochloric acid for seventy hours at 20°C. The mineral acid was neutralized with barium carbonate, and the solution was filtered to remove traces of an unidentified substance which was found to be soluble in chloroform but insoluble in liquid trimethy1-Dfructose. The filtrate was reduced to a syrup by distillation at Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 13 reduced pressure, and the syrup was again extracted with chloroform. The extract was reduced again to a syrup. This syrup was dissolved in dry benzene and the benzene was distilled at reduced pressure in order to remove final traces of water. The residue was then distilled under reduced pressure, and the colorless fraction distilling at 11^°C.'at 0.2 mm. was collected. The average yield of trimethyl-D-fructose was 3.8 g. 76 per cent of the theory based on trimethyl-inulin. 23.1° in chloroform, [e< [ = (cone, = 1.02 g. per lOO ml. of solution.) a 28.9° in water, (conc. = 1.02 g. per 100 ml. of solution) The osazone prepared according to the procedure of Haworth and Learner (80) was a yellow product melting at 80°-8l°C. in close agreement with values reported by previous investigators .* (80 ,6l) The analysis based on the release of formaldehyde as a consequence of periodate scission, gave the expected result of one mole of form­ aldehyde per mole of 3 ,6-trimethyl-D-fructose , These results are shown in Table II, The results of periodate reduction, which also gave nearly the expected value for pure trimethyl-fructose are shown in Table I and Figure 1. Apparatus A Beckman Model B spectrophotometer equipped with matched Corex cells was used for optical density measurements . A Rudolph polarimeter * Optical rotation data do not compare as favorably with litera­ ture values, foC » 23.1 in chloroform. [oc » 26.3° (82), 27.7° (81), 25° (80) in chloroform. [ oc 28.9° in water. [ oc - 30,51° in water (83) . Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ih TABLE I PERIODATE OXIDATION OF 3 ,h ,6 TRIME1‘HYL-D-FRUGT0SE Oxidation Time (Hours) 2 Equivalents of 10% Consumed Per Mole of Sugar Theoretical Consumption Of 10% ^ Per Mole of Sugar 2.028 2.028 2.028 2.028 2.028 1.871 1.897 1.923 2,019 2.121 h 8 12 18 * All values are average of duplicate analysis. trimethyl-D-fructose. The word sugar refers to TABLE II FORMALDEHYDE AN/iYSIS OF 3 ,U j6 trimethyl-d-fructose Oxidation Time (Hours) 1.25 2.00 ■St Moles of Formaldehyde Found X 10® 0.98U l.OlU Theoretical Amount of Form aldehyde in Moles x 10® 0.99k 1.019 All values are average of quadruplicate analysis Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2.40 CD ■D O Q. C g Q. Z O ■D CD C/) C/) I- 2.00 0_ 3 ■8D ^ 1.60 Z O o 1.20 CD u 3 3. " CD CD ■D O Q. C a O 3 "O O CD Q. h- < o cr Lj J û- .40 ■D CD C/) C/) .00 0 4 8 12 HOURS ■Figure 1. Periodate consumption per mole o f 3 6- trim e th y lD -fructose in terms o f equivalents o f periodate per l i t e r versus o xid a tio n tim e . 16 15 equipped with a sodium lamp and a thermostat was employed for optical rotation measurements. Quantitative Analysis of Reaction Mixture In the Lobry de Brtyn reaction there are two principal products and the original carboliydrate in the reaction mixture . Such a mixture ob­ viously requires three analytical methods or three independent equations for its description. Assuming that the starting hexose is transformed into only the other two hexoses , the stoichiometry of the reaction pro­ vides one equation. Two analytical methods are then required to estimate the concentrations of the three principal products . The first analyti­ cal method employed involved optical rotation measurements. The exjjression for the rotation of such a mixture is as follows. [oC .C .d 100 [ oc + Glucose • c • d' 100 ■[ oc ]26 , c . d ■ 100 + Mannose = optical rotation of reaction mixture. Fructose By the substitution of 10c/M = moles per liter, the equation was reduced to [.036] [O] + [.036] [M] + [.036] [F] [^p]g = optical rotation of reaction mixture or optical rotation Cl) Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 16 [ oC = specific rotation of the sugar at 2S°C . using sodium light. The subscripts within the brackets refer to the first letter of the hexose. The values for specific rotations are ^2.7^, lL.2°, and -^2 .hP for glucose, mannose, and fructose respectively. C = grams of hexose in 100 m l . of water at 25°C. d = length of polarimeter tube in decimeters. [G], [M] and [F] refer to the concentrations of glucose, man­ nose, and fructose respectively in moles per liter. The additivity of optical rotations of dilute solutions of these sugars was tested, and in all instances, the added values were, within experimental error, in agreement with corresponding values calculated from specific rotation data, This principle has long been used in carbohydrate chemistry in the analysis of simple mixtures of sugars. The second of the analytical methods involved the Malaprade reaction in which the uptake of periodate by the reaction mixture was measured. The equation employed follows : [10.20] [G] + [10.20] [M] + [9.iiO] [F] = I (2) I is equal to the periodate uptake of the reaction mixture per millimol of sugar. duced The value 10.20 represents the equivalents of periodate re­ permole of glucose or mannose by the methods used in thisexperi­ ment . This value is an average obtained from a large number of determin­ ations and is assumed to be the value for mannose because both sugars give the same products in equal quantities on oxidation. Rechecks on Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 17 the reagents and the method of analysis consistently gave the value of 10.20 instead of 10.00 milliequivalents of periodate reduced per milli­ mol of sugar. The value, 9.^0, represents the equivalents of periodate reduced by one mole of fructose under the conditions of this eoq^erlment, During early experiments, oxidations made with periodic acid gave inconsistent results, and it was found that glyoxylic acid, a product of periodate oxidation of fructose, was oxidized subsequent to the addition of sodium bicarbonate in the determination of unreduced periodate. The oxidation of ^yoxylic acid by bicarbonate buffered periodate is described by Char gaff and Sprinson. (6ii) Accordin^y, all periodate oxidations of the hexoses were performed in bicarbonate buffer. in good agreement with literature values. The value, 9 ,hO, is Reeves 16S) obtained 1.7 moles of formaldehyde per mole of fructose, and Chargaff and Sprinson (6U) reported a periodate uptake of U.8 moles per mole of fructose. All of the concentration terms in the three simultaneous equations are in moles per liter. An attempt was made to replace periodate oxidation by perchloratoceric aicid oxidation as described by Forist, %>eck, and Neely. This appeared attractive since fourteen equivalents of Ce (86) are required to oxidize one mole of fructose, whereas twelve equivalents of Ce'*"*' are required to oxidize one mole of either glucose or mannose. However, this approach to the analysis of these mixtures gave results of a low order of reproducibility. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 18 The methods of analysis of reaction mixtures containing 3,U,6-tri~ methyl-D-fructose as substrate are similar to the ones just described. 3 ,14-,6-trimethyl-D-fructose affords a convenient analysis, for as indi­ cated in the equation below, two equivalents of periodate are consumed per mole of trimethyl-fructose, and one mole of formaldehyde is re­ leased. CHgOH I C- 0 I heO — C — H I H — C — Cî''ie I lO, excess HCO” I H — C - OH OH MeO - C - H I H — C ~ OMe I H - C - OH O HC H I CHgOMe CHgOMe Hence by analyzing aliquots of the reaction mixture for fonnaldehyde, after periodate scission, the rate of disappearance of the starting fructose derivative could be determined. The principal products, assum­ ing a Lobry de Bruyn type of reaction, are 3 and 3,^^6-trimethyi-D-mannose. ,6-trimethyl-D-glucose Neither of these substances release formaldehyde when oxidized with bicarbonate buffered periodate. However, both compounds consume two equivalents of periodate per mole, so that the integrity of the systems trimethy 1-fructose-trimethyl-mannose- trimethyl-glucose could be conveniently checked by determination of its periodate consumption. The expression used for calculating the concen­ tration of trimethyl fructose is given below. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 19 — JE IC) = F (3) a X = optical density Clog Iq/I) of the dye produced from the inter­ action of the reaction mixture with chromotropic acid. C = initial concentration of trimetlnyl-fructose in the reaction mixture expressed bs moles per liter, a = optical density of the dye produced by C concentration of tri­ methyl-fructose with chromotropic acid. F = concentration of trimethyl-fructose In moles per liter. The concentrations of trimethyl-mannose and trimethyl-glucose were calculated by solving two simultaneous equations, one involving optical rotation measurements and the other involving a stoichiometric relation­ ship between the concentrations of the thiree, methylated sugars. The equation involving optical rotation measurements was derived in a manner similar to that used for equation (l) and is given below. [0] [ cxC . [M] [=C„]g^° + [F] D = specific rotation of the sugar at 2$°C . using the sodium D line. The values used for [oC and { oC were 77.5°, 987) 8.2°, 188) and 28.9° respectively. The subscripts G, M, and F refers to methylated glucose, mannose, and fructose respectively. [G], [M], and [F] refer to the concentrations of trimethyl-glucose, trimethylMmannoae, and trimethyl-fructose, respectively, ex­ pressed as moles per liter. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 20 The stoichiometric equation assumes that the total concentration of the methylated sugars is equal to the initial concentration of trimethylfructose. In equation form, G + M + F = 1 The last two equations can be solved simultaneously for (5) the concen­ trations of 3^jL,6-trimethyl-D-glucose and 3 ,U,6-trimethyl-D-mannose by inserting the values for trimethyl-fructose which i-rere calculated previously, from formaldehyde analysis. In this manner, the concentra­ tions of the tïiree components in the reaction mixture can be determined. The base catalysed degradation of D-glucose and 3 ,lt,6-trimethyl-Dfructose were carried out in a blackened reaction flask which was placed in purified by passage through two a 25° Z .01^0. water bath. Nitrogen, gas wasliing bottles containing pyrogallol-sodium hydroxide solution and one containing water, was bubbled into the reaction mixture at a rate of approximately 100 ml. per minute. In this manner, photochemical effects and oxygen oxidation were minimized. Reaction samples were re­ moved periodically and quenched by the addition of an equal volume of dilute, standard acid. This step eliminated the possible presence of an enediol at the time of periodate oxidation. Sodium-2 ,6-dichlorbenzenone- Indophenol was the reagent employed in testing for enediols. Acid form­ ation during the alkaline degradation of the hexose was determined by neutralizing an aliquot of reaction mixture with an excess of standard acid and back titrating with standard base. In the case of methylated sugars, the stability of the metliylated products in aqueous alkaline Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 21 solutions was tested by determining periodate consumption of aliquots of the mixture which were removed from time to time. The concentrations of sodium, calcium, and barium hydroxide solu­ tions were determined by titration with standard base to the bromthyraol blue end point. The calcium ion concentration was also estimated gravimetric ally as the monohydrated oxalate, whereas the barium ion concentration was determined by weighing the sulfate. The following is a description of a typical experiment. The calcu­ lated amount of the sugar was weighed into a volumetric flask, and sufficient water was added to dissolve the substance. A volume of standard base was then added to give the required hydroxyl ion concen­ tration, The solution was mixed, placed in the water bath for fifteen minutes, and then diluted to volume. Nitrogen gas was passed tlirough the solution at the time of mixing and thereafter. Samples were removed periodically, quenched, and stored in the refrigerator. Usually, samples were removed at tiie beginning and at the end of the rate studies to test for acid and enediol formation. Analytical Procedures Formaldehyde Analysis The method used was devised by Forist and Speck. (75) A 2-ml. aliquot of the quenched reaction mixture was transferred to a 100-ml. volumetric flask. Two milliliter of 0.3 M periodic acid was added to the flask followed by 2 ml. of 1 M sodium bicarbonate. allowed to proceed for one hour. Oxidation was At the end of this period, excess Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 22 periodate was reduced by the addition of 5 m l . of 0.5 M sodium sulfite, and the solution was diluted to the mark. Duplicate 1-ml. aliquots of tiiis solution were pipetted into 50-ml. volumetric flasks followed by 1 ml. of 5 per cent cliromotropic acid and 5 m l . of 11: M sulfuric acid. The flasks were heated in a boiling water bath for 30 minutes, cooled, and diluted with approximately 1^0 m l . of distilled water. The flasks wore allowed to cool to room temperature and then diluted to the mark. Excess sulfur dioxide was removed by bubbling air saturated with water tlirough these solutions at a rate of 7oO-1000 ml, per minute. The optical densities of these solutions were then determined against a re­ agent blank at a wavelength of 570 myu. using a Beckman Model B spectro­ photometer and 1-cin . Corex cells for the measurement. The amount of formaldehyde was calculated from a standard curve prepared by a similar treatment of glucose. Determination of Periodate Consumption The oxidations were carried out in periodic acid solutions buffered with sodium bicarbonate. This was followed by determination of the periodate consumed according to the method of Fleury and Lange. (89) Five milliliters of the quenched reaction mixture was mixed with 20 m l . of standard 0,1500 W periodic acid and 10 ml. of saturated sodium bicarbonate. Oxidation was allowed to proceed for two hours at wïiich time 10 m l . of standard 0.1000 N arsenite was added. The solutions were allowed to stand overnight and were then titrated with 0,0100 N iodine solution to the starch end point. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 23 Cerate Oxidation The method used was described by Forist, Neely, and Speck. (86) A 5^-ml. sample of the quenched reaction mixture was oxidized by 15 ml. of standardized 0.1|.80 M Ce^ in 6 M perchloric acid; the oxidation was allowed to proceed for exactly one hour. The excess Ce duced with 15 m l . of standard 0,1000 N arsenite solution. was then re­ Two drops of osmium tetroxide in 0,10 M sulfuric acid was added to catalyze the re­ duction ■vrtiich was allowed to continue for eight minutes . The excess arsenite was then oxidized with 0,0120 N Ce** in 2 M perchloric acid to the nitroferroin end point. Optical Rotation Measurements Measurements were made on the quenched reaction samples at 25^0, with a Rudolph polarimeter equipped with a sodium lamp and a two decimeter, water jacketed polarimeter tube. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2h RESULTS AND DISCUSSION As a resTilt of preliminary studies as well as the probability that side reactions possess a higher order of dependence on hydroxyl ion con­ centration than the enolization reaction being studied, it was decided to maintain the liydroxyl ion concentration in the vicinity of 0.01 N for the isomerization of D-glucose. The data obtained using 0.020 N sodium hydroxide appear in Table III. Periodate consumption and the change in optical rotation with time are shown in Figures 2 and 3 respectively. The concentrations of the three sugars were calculated by taJking points from these curves and applying the three simultaneous equations . The concentrations of the sugars at different times appear in Table IV and are also plotted in Figure ii. The values were not calculated beyond the 32nd hour, at which time a negative mannose concentration was obtained. The results of one of two subsequent experiments with 0.010 N sodium hydroxide solutions illustrates the difficulty in obtaining reproducible data. The data are listed in Tables V and VI and their plots are shown in Figures ^ and 6. An alternate method of oxidation of the reaction mixture with perchlorato-ceric acid was attempted, and the results obtained from one of three runs are given in Table VII end Figures 7 and 8. The data from analysis of glucose degradation mixtures by the tech­ nique of specific oxidation indicate that this approach cannot be Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 25 G O -e <0 +3 m o tuO •H O C D ü 0> -H o OD co • t—I o ^ o I—I r— i _ f \ » • I— i I—I oo co <5 cA o • • I—f /—I co UA CN * o -G -G -C f c o c o OO c o r~ -s o • • • o o o c o -G "U A I-IA u a -G • * o o o> I A O g t 4 g •H “Sâ o g n o bO p»^ O 0 M M i O ex C cu U CD bfi CQ 0 a CO O o 4h o 1 o 'O^COONrHC^QcOrHi-IONÇSJ rH On _ = f CO GO v O - G - CM C~-TJA C A CMr-HfHi-HOOOOOOsOsCK * M o «M G O S I—1 . cr ID s ; # * # # * # # » * * » O O O O O O O O O C N C N C 7 \ fHfHrHrHiHrHrHiHiH Ih (Ü CX • c 4-4 o O •4Ha 0) Ü cd S E-« o i • • L #0 • • • • • • • •OQtoioâconioioo>*}co ^ Æ ^ è C|A^ ^ ^ è ^ ^ ^ CM 0 3 0 \ rH CM r 4 o 3 CO CM M3 0 3 0 0 CM t—( CM CM C A C' A —3 -OT U A CM o> iH rH C M C A o3'C A N O C ^aoO \O rH C M 1 en Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CD ■D O Q. C g Q. TABLE IV ■D THE LÛBRÏ de BRUW TRANSFORMATION OF D-GLUCOSE [0.02 N NaOH, 0.10 N D-glucose, 25*^0.] CD C/) C/) ■8D Time o f Reaction 0 h CD 3 3. " CD CD ■D O Q. C a O ■D O 8 12 16 20 2k 28 32 Meq. o f 10’ Consumed Per Mmol, o f Sugar 10.216 10.195 10.173 10.151 10.130 10.106 10,086 10.061 10.01+2 O ptica l Rotation Readings, ° 1.886 1 . 701+ 1.528 1.368 1.220 1 .081+ 0.972 0.860 0.768 * O p tica l ro ta tio n readings were taken using a 2-dm. Concentrations, Mmol, x 10^/M l. Glucose Mannose Fructose 1.000 0.933 0.881 0.81+1+ 0.811 0.787 0.763 0.779 0.788 0.000 0.013 0.066 0 .071+ 0.081 0.078 0.051+ 0.031 Neg. 0.000 0.051+ 0.053 0.082 0.108 0.135 0.163 0.190 - polarim eter tube. CD Q. ■D CD C/) C/) ro o 27 c o •H O •p «1 •H •73 "g O o 'H O O O O C V l Q C M - a - O c v j O t N I C C î O C V J - E f C O - E J C O r ^ f —1 '2 5 t > - a D - E j _ = t O t r v O n tJ A CÎN _ c r a O ^-N O N O P jr^C O f-IrH O O # # * # * » * * # # # #CKCOX3 * » #^ - n#O 1—4 f—I ( — iI—I I—I1—4 r—I I—I I—I I—IrH o o o O O QJ S w CO o o ro ■ë o o 1o L T V o ■ CM rH ë :a o M E-i 5 O CL, a> n o o G 1 —1 W) t O C3 s: o Pd rH E-c O to tG p; o PQ CD z: I tcO to G G to O O «M o 1 H< o M rH O «H Oi . G cr I CM O r—oo T-T\ nq co co p f .-? • • • • • • • • • • • • * • * * O O O O O O O O O O O O n O n O n O n O n c I CD (U Pi 1—1 O 03 ; .S I— I I—I I— I I—I I— 1 rH rH rH rH i—I rH -o 3 s nJ bfl LTNrH 0 \P f co y \ r-- rHI rH C — —N O \T\■rH C N J oo nO 1/n cnj o co c .—j I— I— I & —-C — CNJrHrHrHrHrHO OO O OCKO NONO NON n Cm o o •H P »ujntotonMmnn % .3 0> I —I i •S P rH CNJ r r \ _ : ; f l_ f\N O C ^C O 0 \ O r-HCNJ f ^ _ = f H r \ v O i - i r - l r - l r - t r - l r - l i —l S' co Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CD ■D O Q. C g Q. TABLE VI ■D CD i- THE LOBRY de BRUIN TRANSFORMATION OF D-GLUCOSE [0.01 N NaOH, 0.10 M D-glucose, 2$°C,] C/) 5 CD 3 3. " CD CD ■D O Q. C a O 3 "O O Time oi Reaction 0 il 6 12 16 20 2ii Meq. o f IO4 Consumed Per Mmol, o f Sugar 10.220 10.195 10,176 1C .159 lo .llo 10.121 10.103 O p tica l R otation Readings, ° 1.880 1.785 1.693 I .605 1.517 1.132 1.352 « A i l o p tic a l ro ta tio n readings were taken using a 2-dm. Concentrations, Mmol, x 10 / k l . Glucose Mannose Fructose 1.000 0.995 0.995 0.995 0.995 1.000 1.003 0.000 - 0.023 - 0.023 - 0.023 - 0.023 0.000 -0.115 0.000 0.026 0.028 0.028 0.026 0.000 O.lliÔ polarim eter tu b e , CD Q. ■D CD C/) C/) ro oo CD ■D O Q. C g Q. TABLE V n ■D CD THE LOBRY de BRUB TRANSFORMATION OF D-GLUCOSE [0,01 N NaOH, 0.10 M D-Glucose, 25°C,] C/) C/) +4 ■8D Sample 1 CD 2 3 3. " CD CD ■D O Q. C a O 3 "O O 3 il 5 6 7 8 9 10 Time o f Reaction 0 1 2 3 h 5 6 7 9 11 h r. h r. h rs . h rs . hrs . h rs . hrs. h rs . lir s . h rs . lieq. o f Ge Reduced Per Mmol, of Sugar O ptical R otation Readings, ° 12.189 1.865 1,81|2 1.830 1.B26 1.787 1.762 1.717 1,712 1.69ii 12.201 12.207 12.215 12.228 12.236 12.279 12.281 12.293 12 .315 1.656 CD Q. * O p tica l ro ta tio n readings were taken using a 2-dir., polarim eter tube. ■D CD C/) C/) ro \o 10.25 10.20 o CL CO z o o I0.05 ÜJ < Q I0.00 O cr LU CL 9.95 9.90 O 32 16 48 HOURS Figure 2. Periodate consumption in equivalents of periodate reduced per mole of hexose (based on initial glucose concentration) from a reaction mixture originally lOO M in D-glucose and ,0200 Min sodium hydroxide versus reaction time. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2.00 .80 O o f— d 40 I— o cr 20 < o 00 F— CL O 80 60 6 32 HOURS Figure 3. Optical rotation (2 dcm. tube) of a reaction mixture originally .100 M in D-glucose and .0200 N in sodium hydroxide versus reaction time. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 48 I 00 CD ■D O Q. C g Q. o ■D X 800 CD C/) o" 3 I cr LU 600 8 ■ 'D < H 400 3 . 3 " CD CD "O O Q. O 3 ■D O CD Q. (f) LU o 200 000 0 8 16 HOURS ■D CD C/) C/) Figure L . Concentrations o f hexoses in moles per l i t e r x 10^ in re a ctio n mixtures o r ig in a lly .100 M in D-glucose and .0200 N in sodium hydroxide versus rea ction tim e. # , ® j and Q re fe r to glucose, fru c to s e , and mannose re s p e c tiv e ly . 24 0.2 5 0 20 O Hû_ ID CO O CD 10 05 üü I— Q 10.00 o cr CL 9 95 9.90 64 32 96 HOURS Figure 5. Periodate consumption in equivalents of periodate reduced per mole of hexose (based on initial glucose concentration) from a reaction mixture originally .100 M in D-glucose and .0100 N in sodium hydroxide versus reaction time. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2.00 .60 O t— H 32 o o o o o Os I CM 1 o o 1 sO o o 8 o rr\ I G O I o O o e +-> O § G J Q 1C C ^ 1O O1 1O Og i > 1O s 1G 1 s g O* ■0iH pr, f-, 00 Q) 1 o3 id o o o o o 1 O C O so 1 o o o 1 CM 1 rH O o s o 1 o 1c — O I o D— S O -G I C O 1 lf\ 1 r— I rH 1rH 1 1— 1 O O O .5 ès cx c o •iH "ë X n Pti bo OC \ J G O r— O O 'O O I C M o G O r — v O i _ T \\ _ r \ _ z f c<\ I rr\ CM CM CM C\) CM CM CM rr\ C M CM s E-4 Ed G O •H O c d 4- ) • » CO O bO CM CM CO \r\ rH MD CM dO lQ> _Gf GO CM r— C f\—# _ = f vO Q CM Q m 3M3 I tn O +3 o I—I Q o 1 1 CO CNJ 1 CNJ 1 OO o o c^ 1 NO 1 c— o CNJ c— 1 r—J 1 GO 1 C-— 1 N O O o ■UN 1 C— UN 1 CNJ 1 rH On ■UN t t— 1 On o 1 CO 1 CO 1 1— 1 1 rH O o ^ o ! S o 1 5 O X r—I 0) O CO }i © o o C) 1— 1 -=t On O f c L 4 •H k © -»£ © ü o © Ü § ë 8 o o 1 o o 1 O o o N K O © o c P, K) A C o '-Stl 64” (2 w o PQ W I o C M O CO C^ I l rf^C^CNJCM "UA I r— O ’LTN I (N-\ r r \ c \] CNJ On r-4 CNJ eu CNJ CNJ 5 ^ èî o 0 • « a I 43 © 3 I S) Iro PQ § © •H O +3 (O «\ +3 © (§ ^ -ïi o © 1 3 m i£ s o ; On CO g S S 0) 1 c— 1 CP *V 3= +©3 g.g 'O e -\ t T-rv 1 C) rH e na ■ CNJ e -i O o 8 O I— l O J-PN C CNI <0\ 1-P\ 0 > 0"N - OO CN O 3 JX Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 2.0 0 O .80 O h< \— 1.60 O en < o hQ_ O .40 1.20 O 2 4 6 8 10 H O U R S Figure 9. Optical rotation (2 dcm. tube) of a reaction mixture ori^nally .100 M in 3 ,U ,6-trimethyl-D-fructo3e and .0200 N in sodium hydroxide versus reaction time. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 12 360 320 280 O O 240 200 0 2 4 6 8 10 HOURS Figure 10. Log optical density of formaldehydechromotropic acid dye obtained from a reaction mixture originally .100 M in 3,1*,6-trimethyl-D-fructoae and .0200 M in sodium hydroxj.de versus reaction time. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 12 CD "O O Q. .000 C g Q. "O CD C/) C/) o .800 X "8O .400 3 . 3 " CD CD "O O Q. C a O 3 "O O CD Q. en üj -I .200 O .000 0 4 8 HOURS "O CD C/) C/) Figure 11. Concentrations of triraethyl hexoses in moles per liter X 10^ from a reaction mixture originally .100 M in 3,^,6 trimethylD-fnictose and ,0200 N in sodium hydroxide versus reaction time. ^ ,3 j O refer to 3,L,^-trimethyl-D-(glucose, fructose, and mannose respectively. 12 2.00 1.80 O o h 3 j snd O i^F er to 3,U,6 trim e th yl-D -(g lu co se , fructose and mannose) re s p e c tiv e ly . 12 4.00 tn O 3.00 X LU < O 5 2.00 ai LU û_ Li_ O ai LU .00 o 0.00 o 4 8 HOURS Figure 18. Increase in periodate consumption in moles per liter x 10 of one ml. samples of reaction mixture versus reaction time. Molarity of trimethyl hexose was .100 M. Molarity of base was approximately .0200 N. ® , and O refer to transformations in the presence of calcium hydroxide, barium hydroxide, and sodium hydroxide respectively. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 12 3L sodium ion.* Barium hydroxide is in its effect intermediate between that by sodium aund calcium hydroxides. The increase in periodate con­ sumption is probably the result of déméthylation and is a striking instance of catalysis by alkaline earth metal ions. It should be pointed out that the change in periodate consumption complicates the analysis of the mixtures and may indicate that the results obtained by this scheme are not valid beyond the point where the periodate titer becomes greater by 5 per cent than that for the trimetliyl-hexoses. f' These results obtained with 3 6-trimethyl-D-f rue to se do not agree with reports of previous investigations (.66,68) of the isomerization of unmetliylated hexoses as regards the implication of stereospecific effects by metal ions . The data also substantiate that obtained with glucose in indicating that simple equilibria among reducing sugars are seldom realized in a]_kaline solutions. A yellowish-green indicator was formed wtiich was decolorized after acidification in the early samples of the reaction mixturej later the coloration remained even after acidification, Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 35 SUMMARY An attempt was made to follow the Lobry de Bmiyn transformation of D-glucose by a new scheme involving two methods of analysis and three simultaneous equations . The results indicated that side reactions occurred to a degree wliich greatly limited the usefulness of this approach, They also showed that alkaline glucose solutions are far more complex than was implied in the reports of many previous investigations. The course of the isomerization of 3 6-trimethyl-D-fructose under / the influence of the bases, sodium hydroxide, calcium hydroxide, and barium hydroxide, was followed by analyzing for formaldehyde after periodate oxidation and by optical rotation measurements . The results indicate no significant differences in the isomerization of 3 methyl -D-fructose in dilute solutions of these bases at 25°C. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 6-tri- PART II Kinetics of the Metal Ion Catal^'zed Degradation of Glyceraldehyde Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 36 EXPERIMENTAL METHODS Materials The DL glyceraldetiyde employed in these measurements was synthe­ sized by the method of Fisher and Baer. (90) ed from G. Frederick Smith Chemical Company. Periodic acid was obtain­ Matheson practical grade cïiromotropic acid was re crystallized from $0% ethanol. Standard 1 M sodium perchlorate was prepared by neutralizing standard 2 M perchloric acid to'pH 7.0 with a standard 2 M sodium hydroxide solution. Calcium, barium, lithium, and magnesium acetate solutions were prepared by neutralization of weighed samples of C.P. metal carbonates with acetic acid. Arsenious acid, sodium carbonate, sodium bicarbonate, sodium sulfite, potassium iodide, iodine, sodium hydroxide, acetic acid, and perchloric acid were all C.P. reagents. Sulfuric acid was Merck reagent grade. Apparatus A Beclcnan Model B spectrophotometer equipped with matched Corex cells was used for optical density measurements . A Beckman pH meter was used for pH measurements . Quantitative Analysis of Reaction Mixture The method of analysis is that devised b y Forist and Speck. C?5) If G, D, and P represents millimols of glyceraldehyde, dihydroxyacetone and pyruvaldehyde respectively per milliliter of solution analysed. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 37 Go represents initial glyceraldehyde concentration, and I represents milliequivalents of periodate consumed by one milliliter of reaction mixture, then the following equations may be written, G + D 2+G + 2D G = l/g - Go consequently, + P + = 2P Go = 16) I (7) (8) Hence, beginning with pure glyceraldeliyde, the glycer aide }iyde concentra­ tion , G, at any tnne may be calculated from the initial glyceraldehyde concentration and the periodate consumption, I. Since G can be deter­ mined at any time, a second method which would analyze for total trioses >/Quld permit an estimation of all tliree components . Total trio se con­ centration was determined by formaldehyde analysis ; by subtracting G, the glyceraldehyde concentration, from the total triose concentration, one obtains the concentration of dihydroxyacetone, D. The pjrruvaldeh.-joie concentration, P, is then calculated from equation (6), A typical experiment is described below. The calculated quantity of standard acid was mixed in a ^0-ml. volumetric flask with sufficient standard sodium hydroxide to provide the desired buffer ratio and concen­ tration. To this was added sufficient sodium perclilorate to give an ionic strength of O.U M in the final solution. This solution was placed in a water bath which was maintained at $0° 2 ,01 C . A sample of pure DL glyceraldehyde was weighed, dissolved in redistilled water, and trans­ ferred to the reaction flask. The solution was mixed, diluted nearly to the mark, and replaced in the water bath for fifteen minutes. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. It was 38 then dilnted to the mark, thoroughly mixed, and returned to the bath. Samples were removed periodically to determine periodate consumption and form aldehyde formation. Analytical Methods Determinations of periodate consumption and formaldehyde formation have been described previously in Part I of this thesis. The only dif­ ference is in the amounts of the reagents used. Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. 39 RESULTS AiMD DlSCUSSIOf^ Reactions were usually followed to the disappearance of 25 per cent of the initial glyceraldehyde concentration. A plot of the logaritlim of the glyceraldehyde concentration versus time indicated a first order reaction confirming the results of Forist, (75) See Figure 19. Thus, neglecting the foniabion of glyceraldehyde from dihydroxyacetone, the rate of the reaction may be expressed as. a iill = k (g) (9) dt wiiich on integration yields , - In G = kt + constant or log G = -k't + constant (10) The pseudo constant, kj_, embodies all of the catalytic species and in an acid buffer system, may be represented as follows; k* = [ k^^(h+) + k ^ (HA) + k^-(A-) ] (11) JMo terms for spontaneous water catalysis were included because there was no detectable reaction in the absence Allowing, ceriment with only acetic acid as the catalytic species revealed that the catalytic constant for acetic acid is approximately equal to only 1 x 10“* mol ^ min.^. Hence, the term, kg [HOAc] [QAc“ ] was included in the revised rate expression. ing catalysis by this pair, the value, 5 x 10“^ mol ^ min. In assum­ is obtained Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. CD "O O TABLE XI Q . C g Q . PSEUDO CONSTANTS — ACETATE SYSTEMS IT = 50°C., .10) ■D 3 CD C/) C/) o' o 3 3 Expt. [Go] [CHgCOOH] 19 0.20 O.iiO O.iiO — 210 i 29 0.20 O.W 0.25 0.15 136 n “n c 32 0.20 O.iiO 0.15 0.25 62 31 0.20 0.25 O.iiO — — 207 2ti 0.20 0.133 " O.iiO — — 201 26 0.20 0.20 Wi 0.20 O.iiO 25,27 0.20 0.10 0.30 [CHaCOONa] [WaClO*] k 'x 10® minT^ CD ■oD (O ' 2 3 CD 3 . 3 " CD O ■D O Q . C Q - o 3 ■D o 0.20 101 3 " c r CD Q . § o 3 " ■D CD 1 . 37 0.20 O.iiO ii 0.30 0.10 155 0.30 0.10 159 — - • C/) 22. 3 * This experiment was run at an io n ic strength of .20 M. A ll concentrations in moles per l i t e r . #H h2 for the catalytic constant, kg. The added term, kg [HOAc] [OAc""] fits very well into the rate expression at different concentrations of acetic acid and acetate ion, giving values for k' -vdiich are in agreement, within the limits of error, with experimental values. Thus, the e3q>res- sion for the pseudo first order rate constant in acetate buffers will for the remainder of this discussion be written as follows: k' = ki [HOCa] + kg [OAc"] + kg [HOAc] [OAc"] wiiere the values of k^, kg, and kg are 1 x 10“^, (lU) x 10 ® , and ^ x 10 mol"^ min.^ respectively. An experiment was performed to determine the effect of ionic strength on the rate. Instead of the usual ionic strength of O.UO M, a solution ifith an ionic strength of 0.20 M was used, and no difference in the value of the pseudo rate constant was observed. The result is given in Table XI, The Effect of Calcium Ion The data on the effect of calcium ion were obtained in extending the work of Forist who made the observation that calcium ion catalyzes the conversion of glyceraldehyde to dihydroxyacetone and- pyruvaldehyde. Forist*s investigation revealed that this catalysis, w M c h was first order in calcium ion, occurs only in the presence of bases, such as acetate ion; there was no measurable effect by solutions of calcium perchlorate. Since this reaction is catalyzed by both acetate ion and acetic acid in acetate buffers and since the calcium ion catalysis is acetate ion dependent in these solutions, the most reasonable expression Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. ii3 for the pseudo rate constant is as follows: k« = kj. [HOAc] + kg [OAc-] + ks [HOAc] [OAc"] + k^ [Ca"^"^] [OAc"] U5) Forist's data were obtained by varying the calcium ion concentration in an acetate buffer, the acetate and acetic acid concentrations of wliich were each 0.2 molar. The present data, which are listed in Table ZIIx, are the pseudo constants obtained at other acetate and acetic acid concentrations, and from them, the catalytic constant k^ may be evaluated. Plots of the variation of pseudo constant with calcium ion concen­ tration at different acetate ion concentration are shown in Figure 23. From this it will be seen that the slopes of the lines obtained over a one and one-half fold change in acetate ion concentration are nearly identical. It might thus appear that although acetate ion is necessary for the catalysis by calcium ion, tliis effect is independent of vari­ ations in the acetate ion concentration over the range investigated. However, the catalysis by calcium ion is complicated by incomplete dis­ sociation of calcium acetate. This is taken into account in the follow­ ing mathematical treatment of the data wliich appears to afford a satis­ factory ejqalanation of these observations as well as to permit calculation of the magnitude of k ^ . The assumptions which are made in this analysis are, (a) The dissociation of Ca(OAc)g to CaOAc^ and OAc" is virtually complete, whereas the dissociation of CaDAc^ is not. This appears to be justified by previous investigation of the Reproduced with permission of the copyright owner. Further reproduction prohibited without permission. )4i» ST On o «-< O CM + 'd c ^ A s CM CM I —I CM co nO r-H CM O •H + + -H ^A S a •H O •rH f-4 O C M •H O n CM + h-4 On On i-H •H 1-4 t-4 O CM O o + + + + t»0 % o iH hfl s NÛ o W) g; ro O o o O f o r-l O CM % a % s i -P CO ÎU O pL# cd S Cd H-) (9 E-4 ! i. I en o o a. â â O 34 a a O C3 3 3 o CM O CM O CM O CM o o 0> •p V-f »H O O rH S a R » .3 p -4 S 3 4 K> +:> 0)