o A STUDY OF THE PERIODIC ACID OXIDATION OF CELLULOSE ACETATES OF LOW ACETYL CONTENT By Franklin Willard Herrick 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 1950 ACOOTIBDGMENT Grateful recognition is given to Professor Bruce B. Hartsuch for his helpful guidance and inspiration throughout the course of this investigation. ********** ******** * * **** **** ** * TABLE OF CONTENTS I INTRODUCTION.................................... ........ Page 1 The Structure of Cellulose. ..................... The Present Problem.................................... II 1 2 GENERAL AMD HISTORICAL................................... 3 CELLULOSE ACETATE........................................ 3 PERIODATE OXIDATION OF CELLULOSE......................... DISTRIBUTION OF HYDROXYL GROUPS IN CELLULOSEACETATES.... III EXPERIMENTAL............................................. PREPARATION OF CELLULOSE ACETATE........................ 10 12 15 15 Materials.................................. .. ..... . 15 Preparation of Standard Cellulose...................... 15 Preparation of Cellulose Acetates of LowAcetyl Content 16 Conditioning and Cutting of Standard Cellulose and Cellulose Acetate.................... The Weighing of Linters ........................ 20 Analysis for Percentage of CombinedAcetic Acid........ 21 Tabulation of Analyses of Cellulose Acetate Preparations 23 Calculation of the Degree of Substitution and the Unit Molecular Weight of Cellulose Acetates......... 28 PERIODATE OXIDATION OF CELLULOSE ACETATES............... Materials............. Determination of the Periodate Content of a Solution of Periodic Acid....... Preparation and Standardization of Arsenite Solution... Preparation and Standardization of Periodic Acid Solu­ tion......................... Preliminary Oxidations: The Selection of Oxidation Conditions...................................... Oxidation of Cellulose Acetates: The Four Hour Re­ action. Long Time Oxidations and the Effect of Mixing.. Oxidation of Cellulose Acetates for 60 Hours with Standard Mixing......... Periodate Oxidation of Sulfuric-Acid Hydrocelluloses... Oxidation of Acetone Soluble Cellulose Acetates....... 31 31 31 32 34 36 43 55 64 83 86 19 Page 90 DISCUSSION THE NATURE OF FIBROUS CELLULOSE ACETATES OF LOW ACETYL CONTENT........................................ The Acetylation Reaction#•••*••••■•••••••••••••••##• The Fibrous Structure of Cellulose and the Acetyla­ tion Reaction................................. PERIODATE OXIDATION OF CELLULOSE.................... 90 90 92 97 The Mechanism of Periodate Oxidation............... 97 Periodate Oxidation of Cellulose#................ 99 Factors that Influence the Reproducibility of Re­ sults and the Rate of Oxidation of Cellulosic Materials.................................... 105 PERIODATE OXIDATION OF CELLULOSE ACETATES OF LOW ACETYL CONTENT................................. The Rate of Oxidation of Cellulose and Cellulose Acetates.#...................... The Course of the Acetylation Reaction............ jit********* 108 108 115 I IBTBDDUCTION The Structure of Cellulose During the past 50 years exhaustive investigations have been made in studying the structure of cellulose. sulted in the following conclusion^; These investigations have re­ cellulose is a high-molecular- weight polymer composed of anhydroglucose units joined together by 1,4-glycosidic linkages. The simplest structural formula for cellulose may be shown as follows; The chemical behavior of cellulose may be explained in terms of the reactions of the three hydroxyl groups in the anhydroglucose unit, and the sensitivity of the glycosidic linkage to cleavage by acids. In the past 25 years physical chemists have investigated tjhe manner in which the cellulose polymer is arranged in native cellulosic materials 2 7 61 * . The theory of the micellar structure of cellulose may be briefly stated as follows; the cellulose polymer is arranged in parallel bundles of chains in the greater portion of a fiber, but these areas of high degree of orientation are disrupted at fairly well defined inter­ vals with areas of relatively high disorder, i.e., random orientation of polymer chains. This theory is derived principally from X-ray studies of cellulosic fibers. The Present Problem There has been considerable argument among cellulose chemists regard­ ing the course of the reactions by "which various cellulose derivatives, such as esters and ethers, are prepared. The purpose of this investigation was originally stated somewhat as followsi "Prepare a series of cellulose acetates of low acetyl content and determine how the periodate oxidation of these preparations compares with that of the cellulose from which they were prepared. The comparison of the periodate consumption of a particular cellulose acetate with that of cellulose may lead to some conclusion regarding the average distribution of the acetyl substituents in this cellulose acetate. For example, if initial acetylation takes place at the primary hydroxyl group on carbon No. 6 of the anhydroglucose unit, then the periodate consumption of cellu­ lose and a partially acetylated cellulose should be the same. However, if acetylation takes place at either of the secondary hydroxyl groups on carbons No. 2 or 3, the consumption of periodate by the cellulose acetate would be less than that of cellulose." It was further suggested that conditions should be selected such that the rate of acetylation would be slow enough to permit the selective esterification (if any) of either primary or secondary hydroxyl groups. During the course of this investigation a serious attempt was made to control all possible factors that might affect either the rate of acetylation of cellulose, or the subsequent periodate oxidation of cellu­ lose and cellulose acetates of low acetyl content. ■2- II GENERAL AND HISTORICAL CELLULOSE ACETATE The acetic acid ester of cellulose is today one of the most important of the plastic materials. The acetates have found use in filaments, films, molding powders, and lacquer resins. The turn of the twentieth century marked the beginning of the cellulose acetate industry, and since that time extensive resear’ ch programs have been devoted to the study of these materials and to the improvement of acetate products. The commonly used materials for the preparation of cellulose acetate ares a cellulosic raw material, an acetylating reagent, a starter or swelling agent, and some diluent or solvent. Cotton fibers or linters have been used most extensively as raw material, although purified wood celluloses are becoming more important in the modern industry. The use of acetic anhydride as an acetylating agent is credited to Schutzenberger®®, who in 1865 prepared a cellulose acetate by means of acetylation in a sealed tube at 140° C. for several hours. Present day knowledge indicates that this product must have been considerably degraded under these severe acetolytic conditions. Other acetylating agents have been used. Acetic acid is unsatisfac­ tory and long refluxing produces only slight acetylation^. Acetyl chloride may be used if a base is present to react with the hydrogen chloride liberated, which otherwise might degrade the cellulose. -3- "Whol3^ introduced the use of pyridine in conjunction with acetyl chloride. The gas ketene (CI^CO) is mentioned as an acetylating agent in several patents33*^3*^6* Starters or swelling agents are used in the acetylating mixture to increase the rate of reaction. They are often called catalysts, The use of sulfuric acid was introduced by Franchimont 1fi in 1879, He observed that the rate of acetylation was increased by the presence of this acid. Cross and Bevan^ mention the use of such de­ hydrating salts as zinc chloride and sodium acetate, Hess^3 acetyl- ated cotton with acetic anhydride in pyridine medium and in the absence of sulfuric acid. In this reaction pyridine functions as a swelling agent, and Hess noted that the rate of reaction under such conditions was considerably accelerated. Furthermore, in pyridine medium the extent of acid degradation is very much reduced. ing to Kruger and coworkers, Accord- perchloric acid (HCIO^) is far super­ ior to any other catalyst, a very small concentration having a powerful effect, reaction, Eren water may be regarded as a catalyst in this Elod-^ found that very diy cellulose was acetylated with difficulty, but if the moisture content of the fibers were allowed to increase progressively, the reaction rate increased up to the point where water became a diluent, with added hydrolytic effect on acetic anhydride. The increased rate of acetylation, due to the presence of water in the cellulose structure, has been explained in terms of the swelling power of the water. -4- The diffusion of the acetylating mixture throughout the cellulose fibers is also facili­ tated by the presence of -water. Diluents or solvents that have been used in the preparation of cellulose acetates include glacial acetic acid, benzene, and pyri­ dine. Acetic acid is a solvent for cellulose acetates, whereas benzene and pyridine are nonsolvents. 'When it is desired to pre­ serve the fibrous structure of the original cellulose a nonsolvent medium is used. If glacial acetic acid medium is used, the end product is a clear viscous colloidal solution. The properties of a cellulose acetate depend on the kind and quantity of reactants, the acetylation conditions, and the subse­ quent treatment of the product. In the usual acetylation in glacial acetic acid, a 2.5 to 4.0-fold excess of acetic anhydride, and about 2 to 5 percent of concentrated sulfuric acid are used. The amounts of acetic anhydride and sulfuric acid are based on the weight of cellulose. Reaction temperatures are controlled between 30 and 40° C., and complete acetylation is ordinarily achieved in about eight hours. Mechanical stirring facilitates the formation of a homo­ geneous reaction mixture. The acetate is recovered from the react­ ion mixture by first diluting with more acetic acid, and then pouring the liquid in a thin stream into a large amount of water. It is important that the aggregation of large lumps of acetate be prevented, as these will retain the acetylating reagents under surface films. The precipitate is worked mechanically and washed repeatedly with -5- water until free from acid and a final boiling with water helps to remove the last traces of sulfuric acid. If benzene is used as a diluent, instead of acetic acid, the final product will be a fibrous mass similar in appearance to the starting cellulosic material. This reaction is more heterogeneous and generally requires a longer reaction time. Recovery of the acetate is made by filtration followed by thorough washing with water. The stability of the final product is determined largely by the thoroughness of the washing operation. Sulfuric acid, in this case'* reacts with cellulose to form mixed esters or sulfoacstates. Ost^, who is responsible for much of the research dealing with the acetyla­ tion of cellulose, concludes that all acetate preparations retain some small percentage of combined sulfuric acid. To obtain a stable acetate this sulfate content must be removed or reduced to a minimum by boiling with water. A product which has not been stabilized tends to degrade slowly, becomes brittle, and develops an odor of acetic acid. The degree of acetylation for any particular cellulose acetate may be determined by saponification with alcoholic alkali, and is expressed in terms of percent combined acetic acid or percent acetyl. These values may be used in a suitable equation (see page 28) to cal­ culate the unit molecular weight of anhydroglucose acetate. A com­ pletely acetylated cellulose, the triacetate, would have a combined acetic acid content of 62.50%, or an acetyl content of 43.75%, and 6- a unit molecular weight of 288.25. Lower degrees of acetylation may be obtained by either reducing the amount of acetic anhydride available for reaction, or by allowing the reaction to proceed for a shorter length of time. A third alternative consists of adding acetic acid (approximately 50%) at the triacetate stage of a react­ ion mixture, and allowing this mixture to ”ripen” for some time at a fixed temperature, usually not over 50° C. This process allows a limited amount of hydrolysis of the triacetate to take place, lower­ ing the combined acetic acid content to between 50 and 57%. Miles^, in patents obtained in 1903 and subsequent years, was the first to recognize the commercial possibilities of this technique. The partially hydrolyzed, so-called ’’secondary” acetate, was found to be soluble in acetone, a good, versatile and cheap solvent. The property of acetone solubility is associated with a certain limit in the amount of free hydroxyl groups present in the acetate. When the combined acetic acid content of an acetate falls below 50% or is above 57%, it is insoluble in acetone. The strength and toughness, and in part the solubility of a cellulose acetate, is dependent upon its degree of polymerization (D.P.). The higher the D.P. of the original cellulose, the stronger and tougher the acetate product will be. Effort is always made to reduce the amount of degradation of the cellulose chain by means of improved reaction conditions and controls. Low reaction tempera­ ture, rapid acetylation, and stabilization of end products are important. The average degree of polymerization and the molecular -weight of cellulose acetates may be determined by viscosity methods. Viscos­ ity and fluidity characteristics of cellulose acetates in solution are those of lyophilic colloids and are affected by such variables as concentration, temperature, solvent, instrument design, structure of polymer, and degree of association. The Ostwald type viscometer is commonly used, and the viscosity of a liquid is given by the equation: |7 * Ktd, where t is the time of flow of the liquid in the viscometer; d is the density of the liquid at some specific standard temperature, and E is a constant for the instrument. that the specific viscosity, The Staudinger rule®® states ^sp, of a very dilute solution of poly­ mer, is proportional to its molecular weight. The empirical rela­ tionship is written: *]r -1 : C where ^sp 'C Em M , J r is the ratio of the viscosity of the solution of linear polymer to that of the solvent; C is the molar concentration of the solution with respect to the structural unit of the polymer chain; M is the molecular weight; and Em is a constant for the cellulose derivative in a specific solvent at specific temperature and pressure. The value of Em must be calculated by substituting the value of the viscosity of a solution of cellulose acetate of known molecular weight in the equation. Osmotic pressure or ultracentrifuge methods are used as standards for the measurement of the molecular weight of high polymers. Viscosities are measured for solutions having concentrations of the order of 0.002 to 0.005 molar, A common solvent used in cellulose acetate viscosity measurements is m-cresol®®. Viscosity relationships are ideal only for very dilute solutions or at infinite dilution. For this reason it is customary to determine the viscosity of a series of dilute solutions of the same acetate. The value of *>p/C is then plotted against the concentration, and the curve extra­ polated to infinite dilution. The value of the specific viscosity at infinite dilution has been called the Hintrinsic viscosity”. The intrinsic viscosity is used in the Staudinger equation to calculate the molecular weight of a cellulose acetate. The degree of polymeri­ zation is obtained by dividing the molecular weight by the unit mole­ cular weight (288) in the case of cellulose triacetate. Kraemer^ has determined molecular weights as high as 250,000, and D.P. of 950 for cellulose acetates. Cellulose acetates of high acetyl content are usually soluble in chloroform, halo-hydrocarbons, aromatic amines and phenols, ali­ phatic acids, and some esters, ketones, and alcohols. Ostwald and coworkers^® have studied the solubility problem with respect to the molecular weight and degree of acetylation of cellulose acetate. They have classified solvents according to the dielectric functions / y £ > where /*is the dipole moment and € is the dielectric constant. For liquids having solvent activity the function has a value between 0.251 and 0.528. -9- PERIODATE OXIDATION OF CELLULOSE In recent years solutions of periodic acid and its alkali metal salts have been applied with considerable success to the analysis of 1,2-glycols in diverse fields of organic compounds^. The selective and quantitative aspects of the reaction have made it a valuable ana­ lytical and preparative tool, not only in applied but also in research chemistry. The reaction was discovered in 1928 by Malaprade®9*^9. and coworkers 11-14 Fleury were responsible for investigating the scope of the periodate oxidation, and found that many 1,2-type oxygen and nitrogen functional groups were attacked. Thus 1,2-diol, °<--hydroxy carboiyl, °0-hydroxy amino, and diketo compounds are reactive. The present discussion will be confined to the periodate oxidation of the 1,2-diol or glycol groups in cellulose, Criegee® has suggested the following mechanism for the oxidative cleavage of a 1,2-diol by periodic acids - c — oh (a) + HglOg — *- c — oh: - c - o - i o 5h4 (b) — — OH (c) -10- - i - o — io5h4 I - C — OH + HoO Jackson and Hudson^8 applied the reaction to cellulose. The secondary hydroxyl groups in the 2 and 3 positions are oxidized to aldehyde groups, with cleavage of the carbon-carbon bond between them: CHgOH «ch 2o h H H OH If cotton is oxidized with periodic acid, anhydroglucose rings will be opened at different points along the chain, and it will be seen that the polymer is no longer composed of successive anhydroglucose units, but at the points of oxidation, glyoxal and d-erythrose units will be present. Since this type of oxidation produces new aldehyde groups, periodate oxycelluloses have a high reducing property. sensitive to degradation by alkalies. They are also These properties were demon­ strated by Davidson8 who made a thorough investigation of the action of periodic acid on cellulose. According to Davidson the oxidation causes a weakening of the glycosidic linkages, making a periodate oxycellulose very susceptible to alkaline degradation. Pacsu8® has suggested that the glyoxal hemiacetal units may be enolized in basic medium, and that the degradation is due to an electrophilic attack by water on the hemiacetal bonds. Furthermore, a Cannizzarro type oxida- tion-reduction may occur between the aldehyde groups in basic medium. -11- Studies on the determination of aldehyde groups in periodate oxycellulose have been made by Harris and coworkers^. Mark and co- workers^-®, Jayme^-*-, H e a d ^ , and Timell*^ have also contributed to the knowledge of the periodic acid reaction on cellulose. The experimental technique for oxidizing cellulose with periodic acid may be described briefly. A standard sample of cellulosic material is treated with an excess of a solution of periodic acid or its alkali metal salts. The concentration of the oxidizing solution is usually less than 0.1 molar. Buffer solutions may be added to maintain a standard pH, usually pH 4. Reaction temperatures of be­ tween 20 and 30° C. are convenient, and a standard temperature is maintained throughout the reaction by means of a thermostat. At the end of the reaction time, the cellulosic material is filtered off and washed, and the filtrate is analized for unreacted periodic acid. The amount of periodic acid consumed by a certain sample may be ex­ pressed in milliequivalents per gram, or perhaps more suitably in moles of periodic acid per unit molecular weight (anhydroglucose unit). DISTRIBUTION OF HYDROXYL GROUPS IN CELLULOSE ACETATES The art of preparing cellulose derivatives is still far ahead of the theory. Although research has been able to point the way in many cellulose problems, there remains much to be done to explain the specific mode of attack of esterifying and etherifying reagents on cellulose. Reaction kinetics fail in this field where reaction rates follow heterogeneous patterns and are usually associated with adsorption -12- and diffusion factors. The preparation of certain acetates with specific properties, for example that of acetone solubility, has added further interest to the investigation of the distribution of hydroxyl groups along the cellulose chain. Any reaction that has reproducible selectivity with respect to the hydroxyl groups of cellu­ lose, and that does not destroy the substituents already present, might be of use in these studies. Acetone soluble cellulose acetates have been investigated quite extensively. Safcurada and Kitabatake^ condensed triphenylmethyl ' chloride with the acetate dissolved in pyridine. Assuming that this reagent reacts only with the free hydroxyl groups in the 6-position, they found that one-third of the total available hydroxyl groups were in this position. Purves and coworkers^»®»^ have Used p-toluene- sulfonyl chloride (tosyl chloride). They observed that the hydroxyl group in the 6 position reacted with the reagent approximately twelve times as rapidly as those in the 2 and 3 positions. Furthermore, the tosylated acetate could be treated with sodium iodide in acetone to replace the tosyl group with an iodine atom. These experiments indicated that at least 35% and possibly SO% of the available free hydroxyl groups in these acetates were in the 6 position. In order to determine whether any glycol groups were present in acetone soluble acetates, a glycol oxidation with lead tetraacetate in glacial acetic acid was performed. This reagent acts in the same manner as periodic acid, cleaving the carbon-carbon bonds between adjacent hydroxyls and oxidizing them to aldehyde groups. -13- Purves, using this method, could account for only one glycol group for every 100 to 150 glucose residues. Similar studies on the distribution of hydroxyl groups in cellu­ lose ethers have been carried out^®. Most recently periodic acid oxidation has been applied as a means of determining the glycol con­ tent of certain water soluble cellulose ethers^*®®»60. -14- Ill EXPERIMENTAL PREPARATION OF CELLULOSE ACETATE Materials 1. Celluloses Kodak filter cotton, Eastman Kodak Co, 2. Acetic anhydrides 3. Sulfuric acids 4. Benzenes C.P. grade. sp. gr. 1.84, 96.7$, C.P. grade. C.P. grade redistilled 79 to 81° C., dried and stored over sodium. 5. Washing solventss benzene, technical grade, and 95$ ethanol. 6. Waters laboratory distilled -water. All the chemicals used ^.bove) conform with A. C. S. specifications. Preparation of Standard Cellulose Kodak filter cotton was found to be a high quality long fiber cotton of high purity and uniformity. experimental use as followss The cotton was processed for 250 g. of cotton was placed in a 4-liter beaker and covered with 4 liters of water and allowed to stand in a hot room at 60° C. for 24 hours. The water was then squeezed out and the cotton covered with four liters of boiling water. When cool enough to work, the water was again squeezed out and the cotton was rinsed with another 4-liter portion of cold water. The thoroughly pressed cotton was then placed in a 2-liter beaker and covered with 2 liters of 95$ ethanol and allowed to stand as -15- before at 60° C. for 24 hours. Rinsing with 4-liter portions of boiling and cold water was repeated. This treatment was considered sufficient to remove pectic materials and waxes from the cellulose fibers. The purified cotton was pulled apart and dried for 24 hours at 60° C. Standard cellulose to be used for the preparation of cellu­ lose acetate samples was dried in a vacuum oven at 60° for 24 hours and stored in a large desiccator over concentrated sulfuric acid. The moisture content of the vacuum dried cotton was determined by heating a 5 to 10-gram sample, contained in a 100-ml. weighing bottle, at 110° C. in a vacuum oven for 24 hours® reduce the weight. Further heating did not In all cases the moisture content of the vacuum dried cotton was found to be less than 1% by weight. For example, 5.9380 g. of this cotton lost 0.0469 g. by this treatment, and there­ fore contained about 0.8% moisture* Storage over concentrated sul­ furic acid at room temperature was considered adequate to maintain, if not further reduce, the low moisture content of this standard cellulose. Preparation of Cellulose Acetates of Low Acetyl Content In this investigation it was desired to prepare a number of cellulose acetates of an acetyl content below 10%. This was accom­ plished by a method in which the acetylation reaction was interrupted. It was also desired that the acetylation reaction be slowed down somewhat, and that the fibrous nature of cellulose be retained. -16- Elod^ has shown that the rate of the reaction is decreased consider­ ably when cotton of low moisture content is used. The proportions of the acetylating reagents and cellulose used in these preparations 21 were established in a preliminary study . The reaction mixture used for the preparation of all cellulose acetate samples was as follows; cellulose, moisture content 10 1 0.1 g. of vacuum-dried standard K.1%; 205 ml. of anhydrous benzene; 55 ml. of acetic anhydride; and 0.8 ml. of concentrated sulfuric acid. The following figures show the molar quantities of each component of the reaction mixture; 0.0617 mole of cellulose 2.305 mole of benzene 0.578 mole of acetic anhydride 0.0147 mole of sulfuric acid The molar ratio in the same order would be; 1 ; 37.33 ; 9.37 ; 0.238. The concentration of sulfuric acid would be 0.0563 M based on 260.8 ml. of acetylation medium. Before starting the preparation of aiy cellulose acetate the proper quantities of benzene, acetic anhydride and sulfuric acid were mixed. Usually enough of this mixture was prepared to acetylate 5 samples of cellulose. This avoids small variations in the quantity of sulfuric acid used for each sample. At zero reaction time 10 g. of vacuum-dried standard cellulose was rapidly weighed into a dry 400-ml. beaker and immediately covered -17- with 260.8 ml. of the acetylating mixture. Since small variations in the total amount of this mixture are not important, it was measured in a 500-ml. graduated cylinder. Air bubbles were removed from the cellulose mass by pressing with a glass rod. The beaker was covered with a watch glass and the reaction mixture was allowed to stand at room temperature for a length of time that varied with the degree of acetylation desired. It was found that during the first 5 minutes of the reaction, the temperature rose approximately 2° C. and then slowly came back to room temperature. The temperature of the reaction was recorded from time to time and did not vary after the first half hour of reaction time. Average temperatures ranged between 27 and 29° C. At the end of the reaction time the cellulose acetate mass was transferred to a Buchner funnel-suction flask apparatus and pressed under suction with the bottom of a 250-ml. beaker. The mass was then returned to a 400-ml. beaker and washed with four 100-ml. portions of benzene and four 100-ml. portions of 95% ethanol, each washing being accompanied by agitation, and removal of the wash solvent by pressing under suction. At this point the acetate always tested acid to litmus. Water washings to remove all traces of acid were standardized as follows* two 4-liter washes with distilled water, the acetate mass being stirred in the second wash by means of an electric stirrer for 2 hours; one 2-liter rapid wash with dilute ammonia water (20 ml. of 14.7/6 NH4OH in 2 liters of water); two 4-liter washes with water; -18- one 24 hour steep in one liter of water at 60° C.; and finally one wash with 4 liters of water. The acid-free cellulose acetate was air dried for 48 hours. Conditioning and Cutting of Standard Cellulose and Cellulose Acetate. The cellulose acetates and samples of standard cellulose were conditioned for several days in a desiccator containing 36$ sulfuric acid in its well. The relative humidity of the atmosphere in this desiccator is approximately 65$ at 25° C. 66 in the laboratory also averaged about 65$. . The relative humidity Conditioning at 65$ rela­ tive humidity permitted handling and accurate weighing of acetate samples without too much danger of loss or gain in weight due to moisture transfer. The moisture content of conditioned samples was determined as described previously and was found to be about 5$ for both cellulose and cellulose acetate samples. The weight of all samples of cellulose and cellulose acetate recorded in the data tables is the dry weight, i.e., the weight of the conditioned sample cor­ rected for 5$ moisture content. In order to obtain a uniform sample suitable for oxidation stud­ ies, the long fiber's were cut to 20-mesh linters. This was done in a Wiley mill, which cuts the fibers by means of rapidly spinning knives that pass by stationary cutting edges. A 20-mesh screen allows the desired size of linters to pass into a receiving jar. No heating effects were observed during the cutting procedure, and the cut ends of the linters were sharp and clean. The cut samples were stored in the conditioning desiccator at 65$ relative humidity. The Weighing of Linters The method that m s used to weigh cotton and cellulose acetate linkers m s kept standard throughout this investigation. Samples were weighed on an analytical balance to the nearest 0,1 milligram. The samples were weighed in a specially constructed weighing tube. This tube consisted of an outer jacket and a plunger. The outer jacket m s made from a 1,5 x 9.5 cm, glass tube, which was fitted with a plunger which m s a 1.2 x 9,5 cm, test tube. A pair of forceps was used to transfer the linters from a storage bottle to the cylinder tube -which was held in a vertical position with one end against a piece of filter paper. The plunger m s then used to compress the sample into a compact mass. The depth of this compressed sample m s used to estimate the weight of linters present, and with a little practice, the desired weight could be judged quite easily. The sample could be ejected into the desired reaction flask by exerting pressure on the plunger. Adhering linters in the tube were dislodged by gently tapping the side of the flask with the tube and by manipulation of the plunger. weighing tube. Chamois finger protectors were used in handling the The weight of the sample was obtained by difference: the weight of the sample and tube minus the weight of the tube. This method of weighing m s found to be quite rapid. It pre­ vented the very lightweight linters from escaping into the balance housing, and furthermore, protected the linters from excessive moisture transfer. ■20- Analysis for Percentage of Combined Acetic Acid. The alcoholic alkali saponification of Eberstadt 34 as modified by Howlet and Martin^® was used to determine the combined acetic acid content of the cellulose acetate samples. for this method is t 0.1% HOAc^®. The accuracy claimed The concentrations of the reagents were modified slightly to add to the convenience of the analysis. Standard potassium hydroxide solution, approximately 0.2 IT (carbonate free) was prepared according to procedures used for the standardization of bases described by Willard and Furman 65 • Potassium acid phthalate was used as a primary standard, and phenolphthalein as an indicator of the equivalence point. Solutions of potassium hydrox­ ide, approximately 0.5 IT, and of sulfuric acid, approximately 0.6 IT, ■were prepared and the concentrations checked by reference to the standard 0.2 IT KOH, but it was not required that the concentration of these solutions be known, except for planning the proportions of solu­ tions to be used in the analysis. Three 0.5 to 1.0 g.-samples of the cellulose acetate to be ana­ lyzed were weighed in the -weighing tube and transferred to 250-ml. glass stoppered conical flasks. At least three flasks containing approximate equivalent amounts of standard cellulose were also pre­ pared for each group of analyses. determinations. These flasks were carried as blank Each sample of linters was wetted with 20 ml. of 95% ethanol, measured with a pipet, and allowed to stand for 15 minutes. Each sample was then treated with 20 ml. of approximately 0.5 IT KOH, -21- measured with a standard pipet, and the sample and solution were mixed thoroughly and allowed to stand for 24 hours at room temperature. At the end of this time each sample was treated with 20 ml. of approxi­ mately 0.6 IT HgSO^ measured with the same standard pipet. After mix­ ing, the sample was allowed to stand 4 to 6 hours so that the excess alkali could be removed from the fibers. During the 24 hour saponi­ fication period the potassium hydroxide swelled the cellulose acetate and diffused into its capillary structure. Howlet and Martin^® found that the addition of excess sulfxiric acid to such a mixture did not neutralize the base on the inside of the cellulose fibers unless the mixture was allowed to stand for a 4 to 6 hour diffusion period, after which the excess acid could be titrated with standard base and repro­ ducible resiilts obtained. The excess acid present in each sample was titrated with standard 0.2 IT KOH to the phenolphthalein end point. In this final step the contents of the flask was shaken repeatedly to insure complete neutralization of the acid. The calculation of combined acetic acid was made from the equa­ tion: per cent _ HOAc (T - t) (IT of KOH ) (0.06005) (100) . sample weight where T is the ml. of KOH to titrate the sample of cellulose acetate; t is the average of the blank determinations; and the figure 0.06005 is the milliequivalent weight of acetic acid. -22- Tabulation of Analyses of Cellulose Acetate Preparations. The preparation of cellulose acetates of low acetyl content under the conditions used required some preliminary experimentation. Samples of cellulose acetate, No* 1 to 7, resulted from this investi­ gation, and since they were not used in any oxidation studies, the data on these samples will be omitted. Cellulose acetate samples, No. 6 to 15, were prepared as a group and used in the study of the initial reaction (a 4 hour period) of periodic acid on cellulose ace­ tate, Samples, No. 16 to 25, were also prepared as a group and were used in oxidations over a period of 60 hours. Table I contains the data pertaining to the conditions of acetylation and the analysis of cellulose acetate samples. is used to designate cellulose acetate. The symbol ”CAc" The value of the average titer for blank determinations is listed at the bottom of the titer column of each analysis. The sample weight listed as 0.0000 g. may be interpreted as an equivalent quantity of standard cellulose in the blank. The progress of the acetylation reaction may be illustrated by plotting the percent combined acetic acid against the reaction time. Figure 1 shows the rate of acetylation for standard cotton cellulose (moisture content less than 1%) at 27 to 29° C. Although it was not intended that this study should reproduce the reaction curve of Elod'*'^, it can be seen that the rate of acetylation is the same even though the conditions of moisture content and temperature are slightly different. -23- TABLE I ANALYSES OF CELLULOSE ACETATE PREPARATIONS Analysis for percent combined ____________ acetic acid_____________ CAc No. Reaction Time, hrs. Tamp. °C. 8 2.5 27 sample wgt. g. 0.4782 0.4771 0.4798 0.0000 ml. of st. KOH 29.25 29.25 29.30 28.22 N of st. KOH HOAc % 0.2147 3.14 3.09 3.28 av. 3.17 9 5.0 27 0.4810 0.4864 0.4860 0.0000 29.85 30.00 30.05 27.40 0.2148 6.57 6.89 7.03 av. 6.83 10 4.0 28.5 0.4841 0.4869 0.4884 • 0.0000 29.65 29.65 29.75 27.40 0.2148 av. 6.06 ■ 11 7.5 28.5 0.4710 0.4895 0.4796 0.4768 0.0000 30.85 31.10 30.90 30.92 27.40 6.00 5.96 6.21 0.2148 9.45 9.75 9.41 9.53 av. 9.53 12 3.25 29 0.4760 0.4755 0.4776 0.0000 28.60 28.70 28.70 27.25 0.2148 3.66 3.93 3.92 av. 3.84 13 6.0 29 0.4772 0.4761 0.4758 0.0000 29.90 29.80 29.90 27.25 0.2148 7.17 6.92 7.20 av. 7.10 Cont’d next page -24- TABLE I - Continued Analysis for percent combined acetic acid CAc No. Reaction Time, hrs. Temp. °C. 14 4.25 28 'sample wgt. g. ml. of St. KOH N of st. KOH 0.9532 0.9505 0.9516 0.0000 10.48 10.60 10.54 9.49 0.1934 HOAc % 1.21 1.35 1.28 av. 1.28 15 7.0 28 0.5727 0.5744 0.5728 0.0000 12.05 12.10 12.08 9.49 0.1934 5.19 5.27 5.25 av. 5.24 16 2.0 29 0.9537 0.9554 0.9564 0.0000 10.00 10.02 10.05 9.49 0.1934 . 0.62 0.64 0.68 av. 0.65 17 4.0 29 0.4769 0.4800 0.4776 0.0000 11.55 11.65 11.58 9.48 0.1934 5.04 5.24 5.10 av. 5.13 18 6.0 29 0.4769 0.4770 0.4760 0.0000 12.88 12.85 12.85 9.48 0.1934 8.28 8.20 8.22 av. 8.23 19 8.0 29 0.4769 0.4804 0.4781 0.0000 14.44 14.50 14.50 9.48 0.1934 11.94 12.12 12.20 av.12.09 20 10.0 29 0.4796 0.4764 0.4796 0.0000 15.40 15.25 15.25 9.48 0.1934 14.34 14.07 14.01 av.14.14 Cont’d next page -25- TABLE I - Concluded Analysis for percent combined acetic acid CAc Ho. 21 Reaction Time* hrs. 2.75 Temp. °c. 28 sample ■wgt. g. ml. of st. KOH N of st. KOH 0.4821 0.4784 0.4813 0.0000 10.95 10.92 10.98 9.48 0.1934 HOAc % 3.39 3.35 3.47 av. 3.40 22 3.5 28 0.4775 0.4796 0.4807 0.0000 11.72 11.70 11.78 9.48 0.1934 5.30 5.23 5.40 av. 5.31 23 5.0 28 0.4804 0.4808 0.4780 0.0000 12.30 12.28 12.22 9.48 0.1934 6.66 6.62 6.50 av. 6.60 24 7.25 28 0.4777 0.4815 0.4768 0.0000 13.10 13.12 13.10 9.48 0.1934 8.65 8.63 8.66 av. 8.65 25 9.0 28 0.4792 0.4802 0.4802 0.0000 13.65 13.65 13.65 9.48 0.1934 9.96 9.93 9.93 av. 9.94 -26 16. 15 14 13 12 11 10 c / 8 7 6 5 4 3 2 1 0 5--- 1 10 11 Time, hours Figure 1 — Acetylation of cotton, <1$ moisture, at 27 to 29°C. --- Acetylation of cotton, O.lfo moisture, at 45°C., Elod and Schmidt-Bielenberg 10 -27- Calculation, of the Degree of Substitution and the Unit Molecular Weight of Cellulose Acetates. Combined acetic acid percentages may be converted to percent acetyl by multiplying by the ratio: molecular weight of an acetyl group to the molecular weight of acetic acid, i.e., 42/60. Peroent combined acetic acid may be converted to degree of substitution by means of a mathematical expression^ as shown below. When cellulose acetate is saponified the elements of HOH are added to yield cellulose (unit molecular formula CgH^QOg) and acetic acid. In terms of acetic acid content, the unit molecular formula of any acetate may be written: c6h10-2D.S.°5-d S (H0A-C)j) 3 » Y*iere D.S. is the degree of substitution of acetic acid per anhydroglucose unit. The percent acetic acid in cellulose acetate may then be ex­ pressed by: HOAc • 100 _ TTrt, ---------- Substituting the proper atomic weights in this equation and solving for the degree of substitution: D.S. = 162 (HOAC??) 6000 - 42(HOAc/) The unit molecular weight of cellulose acetate will always be that of an anhydroglucose unit plus the weight of the acetyl content. Knowing the degree of substitution, the unit molecular weight of any acetate may be expressed: Unit molecular weight - 162 + 42 D.S. The degree of substitution for a cellulose triacetate has the maximum value of 3. By the above equation its unit molecular weight is 288. -2 8- It was found convenient to calculate the unit molecular weight for cellulose acetates of 1 to 15% combined acetic acid, and to plot the HOAcjS against unit molecular weight. A large piece of graph paper (2 feet square) was used for this graph, and the value of the unit molecular weight of any acetate in this range of percent com­ bined acetic acid could be determined at a glance. The values of the acetyl content, degree of substitution, and unit molecular weight for each acetate sample are recorded in table II. -29- TABLE II DEGBEE OF SUBSTITUTION AND UNIT MOLECULAR WEIGHT OF CELLULOSE ACETATE PREPARATIONS CAc No. Combined HOAc % Acetyl content % Degree of substitution (calc.) Unit molecular weight (calc.) 8 3.17 2.22 0.0875 165.67 9 6.83 4.7e 0.1936 170.13 10 6.06 4.24 0.1706 169.16 11 9.53 6.67 0.2758 173.57 12 3.84 2.69 0.1066 166.48 13 7.10 4.97 0.2018 170.47 14 1.28 0.896 0.0348 163.46 15 5.24 3.67 0.1466 168.12 16 0.65 0.455 0.0177 162.75 17 5.13 3.59 0.1436 168.03 18 8.23 5.76 0.2360 171.91 19 12.09 B.43 0.3560 176.95 20 14.14- 9.90 0.4235 179.77 21 3.40 2.38 0.0940 165.94 22 5.31 3.715 0.1490 168.25 23 6.60 4.62 0.1868 169.84 24 8.65 6.05 0.2490 172.45 25 9.94 6.96 0.2880 174.10 -30. PERIODATE OXIDATION OF CELLULOSE ACETATES Materials 1. Standard cellulose (see page 15). 2. Cellulose acetates of low acetyl content (see page 16). 3. Paraperiodic acids HglOg, reagent grade, G-. Frederick Smith Chemical Co. 4. Sodium bicarbonate: 5. Potassium iodide: 6. Arsenious oxide: 7. Sodium hydroxide pellets: 8. Potassium iodate: 9. Hydrochloric acid: 10. Water: reagent grade. C.P. grade. reagent grade. C.P. grade. primary standard. sp. gr. 1.19, 37-38.5%, C.P. grade. laboratory distilled water. All of the above chemicals conform with A.C.S. specifications. Determination of the Periodate Content of a Solution of Periodic Acid. When periodic acid oxidizes cellulose it is reduced to iodic acid. The periodate content of a solution may be determined in the presence of the iodate ion. to iodate. In neutral solution periodate is reduced by iodide Thus, if a solution of periodic acid is neutralized with excess sodium bicarbonate, and then treated with excess potassium iodide solution, the reaction may be written: NaI04 + 2KI + 2NaHC03 __ * NalOg + I2 + Na2C03 + K2C03 ♦ h 2o The iodine in this solution may be determined by titration with standard 45 arsenite solution, using a starch indicator . -31- ft Davidson found that this method yielded the same results as the more commonly used but less direct method of Fleuiy and Lange14*15. The latter method uses an excess of arsenite to react -with the iodine, followed by a backtitration of the excess arsenite with standard iodine solution. The equation for the reduction of iodine by arsenite may be written: I2 + NaAsOg + 3NaHC03 Na2HAs04 *■ 2NaI + 3C02 + H20 Preparation and Standardization of Arsenite Solution. Standard 0.015 N sodium arsenite solution was prepared according to directions described by Willard and Furman55. Arsenious oxide, As20s analytical reagent, was dried at 110° C. for 24 hours, and used as a primary standard. one time. Seven liters of the solution were prepared at Exactly 5.1928 g. of the dried arsenious oxide was weighed out on a tared watch glass, transferred quantitatively to a 250-ml. beaker, and treated with a solution of 10 g. of sodium hydroxide pellets in 50 ml. of water. warmed slightly. The oxide dissolved readily when the solution was The solution was transferred quantitatively to a 1-liter volumetric flask and diluted with water to 500 ml. An excess of concentrated sulfuric acid (10 ml.) was added to the alkaline arsen­ ite solution. After thorough mixing the excess acid was destroyed by the addition of 10 g. of sodium bicarbonate added in small portions. Such a solution is saturated with carbon dioxide and contains excess sodium bicarbonate; therefore, it is a buffered solution. After thor­ ough mixing and cooling to 20° C., the solution was made up to the mark. It was then poured into a 7.5-liter bottle and diluted with six liters -32- of -water measured in the same volumetric flask. The storage bottle was fitted with a siphon apparatus and protected from the laboratory atmos­ phere with a calcium chloride-soda-lime tube at the air inlet. The formation of sodium arsenite from arsenious oxide may be writ­ ten: AsgOs + 2NaOH — ► 2 NaAsOg + HgO • The normality of the arsenite solution may be calculated from the equation: N (0,049455) g.of A s g O g ___ (ml, of solution) From this equation, the normality of 7 liters of solution prepared from 5,1928 g. of arsenious oxide would be 0,015. The normality of standard arsenite may be checked against a standard iodine solution or by an adaptation of the potassium iodate primary standard technique. The latter method was found to be more convenient. Three 0.25 to 0.30-g. samples of dried primary standard KlOg were weighed out. Each sample was dissolved in 50 ml. of water. This solu­ tion was transferred quantitatively to a 250-ml. volumetric flask, and diluted with water up to the mark. Twenty-ml. aliquots of these stand­ ard KlOg solutions were measured with a standard 20-ml. pipet into a 250-ml. glass stoppered conical flask. Each aliquot sample was diluted with 30 ml. of water and 5 ml. of 10%, KI solution. The iodine in this solution was liberated with 5 ml. of dilute HCl (1 part of 58% HC1 to 15 parts of water) according to the equation: KI03 + 5KI (excess) + 6HC1 — ► 6KC1 + 3I2 + 31^0 This solution is acidic and must be neutralized before the arsenite method of determining iodine can be applied. This requirement was ful­ filled by adding 2 g. of solid EaHCOg in small portions, the flask -33- being tilted so that no loss of solution occured during the evolution of carbon dioxide. The solution was chilled in ice water and titrated at once with approximately 0.015 N arsenite solution delivered from a 50-ml. buret, using 1 ml. of 1% starch solution as indicator. Calcu­ lation of the normality of the arsenite solution was made as follows: H of arsenite z (20) (g. of KIO3 ) _________ (250) (0.03567) (ml. of arsenite) . When checking the normality of the arsenite solution, three ali­ quots of three different standard KlOg solutions were titrated. The normalities obtained for these nine trials were usually in agreement within 2 parts per thousand. The best average was used as the stand­ ard normality of the solution. The normality of the arsenite solutions used throughout this in­ vestigation remained constant within 0.0001 N for several months. When the storage bottle was nearly empty, the concentration changed more rapidly, and therefore, when the solution was down to the 1 liter mark, a fresh solution was prepared. Preparation and Standardization of Periodic Acid Solution. Periodic acid solutions may be prepared by dissolving crystalline paraperiodic acid in the desired amount of water. The concentration of the periodic acid solution selected for the oxidation of cellulose and cellulose acetate was 0.03 N. The solution was made up in double strength (0.06 H) so that the linters of cellulose and cellulose ace­ tate could be wetted and dispersed in a standard volume of water before addition of an equal volume of approximately 0.06 N HIO4 solution. -34- To prepare a 6-liter volume of approximately 0.06 TT solution, 42.5 1 0.01 g. of HglOg was weighed and dissolved in 6 liters of distilled water. This solution was stored in a 7.5-liter bottle fitted with a siphon tube and protected from the laboratory atmosphere with a cal­ cium chloride-soda-lime tube. Paraperiodic acid is very hygroscopic and contains some water in excess of the formula HglOg. A trial solu­ tion was made up to establish the correct weight of the reagent to be used to prepare 0.06 N solutions. It was found that the reagent grade of HglOg contained about 5.5% water. Three 10-ml. samples of the stock HIO4 solution, measured with a standard 10-ml. pipet, were removed and transferred to 125-ml. Erlenmeyer flasks and neutralized with 10 ml. of saturated sodium bicarbon­ ate solution. utes. Each flask urns chilled in an ice water bath for 2 min­ Ten ml. of 10% potassium iodide solution was added and the liberated iodine titrated immediately with standard 0.015 II arsenite delivered from a 50-ml. buret. The arsenite was added as rapidly as possible at first, and then more slowly until the iodine color had almost disappeared or was a faint yellow. At this point 1 ml. of 1% starch solution was added as an indicator, and the titration was con­ tinued until the blue color of the starch-iodine complex just dis­ appeared. The starch solution was prepared by adding 1 g. of soluble starch dispersed in 20 ml. of water to 80 ml. of boiling water, and allowing this mixture to boil for 10 seconds. The periodic acid solutions were found to remain constant within 0.0002 II during any three day period. -35- After a period of 5 or 6 weeks the concentration of the stock solution decreased from ,06 K to 0.055 N. As in the case of the standard arsenite solutions, when the volume of periodate was down to the 1 liter mark, a fresh solution was prepared® The normality of the periodic acid solution was calculated from the equation: N of HIO4 = Preliminary Oxidations: (H of arsenite) (ml. of arsenite) (ml. of HIO4) The Selection of Oxidation Conditions. A number of preliminary periodate oxidation experiments were per­ formed on standard cellulose in order to select the proper oxidation conditions that would lead to reproducible results. It was found that mary variables required either accurate or at least approximate control. These factors are described and discussed in this section. The selection of the concentration of periodic acid solution to be used in the oxidation of cellulose and cellulose acetate was based on the work of Harris and coworkers 53 . This concentration varied between 0.0275 N and 0.03 H. All reaction flask equipment was scoured thoroughly with a deter­ gent scouring powder to remove oxidizable impurities or impurities that might catalyze the decomposition of periodic acid. After rinsing with water, the inside surface of each reaction flask was standardized by treatment with about 0.06 N HIO4 solution for 24 hours before use. After the flasks had been used in oxidation experiments they were cleaned by rinsing with water and were dried in an inverted position. •36- The physical form of the material to be oxidized was standardized by using 20-mesh linters of standard cellulose and cellulose acetate fibers (see page 19). Weighing of these linters was performed as des­ cribed on page 20. Before addition of the oxidizing solution, the linters samples were prewetted with a standard volume of water and allowed to stand for 1 hour. An equal standard volume of stock periodic acid was added and the time of addition noted. mixed. The solution and linters were carefully The reaction flask was then thermostated at 30° C. in a con­ stant temperature bath for the desired length of time. The constant temperature bath consisted of a battery jar, 40 cm. in diameter and 30 cm. high, fitted with Precision Scientific Co. elec­ tronic thermostating apparatus, together with a stirring motor, and a 50° C. thermometer calibrated in units of 0.1° C. A wire rack was sup­ ported in the jar by three chains that could be adjusted to hold the rack at any desired depth. The jar was filled with enough water to cover the thermostat control and the rack was adjusted to a depth 5 cm. below the water level. The thermostat control was adjusted for a bath temperature of 30 t 0.1° C. Snail flasks thermostated in this bath were weighted with squares of heavy lead sheet with the corners turned up to clamp around the bottom of the flask. No method of mixing was used in the case of these preliminary oxi­ dations. A flask containing at least 100 ml, of the stock periodic acid solution was thermostated with the reaction flask and was used for blank determination of the concentration of periodic acid. 37- Since the concentration of a periodic acid solution decreases slowly, due to the decomposition of the periodate ion, the standardization procedure was repeated for each series of oxidations. The pH of the reaction mixture was found to remain constant over a period of 72 hours. Since this pH was between 1.6 and 2.0 no buffer solutions were incorporated in the oxidation mixtures. At the end of the reaction time the mixture was analyzed for unre­ acted periodic acid by the same procedure used for standardizing periodic acid solutions. In some cases the linters were filtered from the oxidizing medium with a specially constructed suction-filter appara­ tus. This was made from a 2.5-liter wide-mouth bottle, 8.5 cm. in diameter at the top. This bottle was large enough so that a 250-ml. conical flask could be placed inside it. The bottle was provided with a rubber stopper holding a 6-cm. long-stem funnel and an 1-shaped glass tube. A Gooch crucible was supported in the funnel with a rubber adapter. T}ie funnel extended halfway into the receiving flask inside the suction apparatus, so that during the filtering procedure, the fil­ trate coming from the Gooch crucible passed into the flask in a quanti­ tative manner. Three methods of sampling to determine the excess periodic acid in an oxidation mixture were studied: (l) the oxidation mixture was fil­ tered and the sample removed from the filtrate; (2 ) the oxidation mix­ ture was filtered and washed, the combined filtrate and washings were made up to a standard volume, and an aliquot of the standard volume removed as a sample; (3) the entire reaction mixture (without filtering) was used as the sample. -38- Method 1. A 1 to 1.5-G. sample of standard cellulose linters mas weighed into a 250-ml. Erlenmeyer flask and suspended in 50 ml. of water measured with a standard pipet. Fifty ml. of approximately 0.06 N HIO^ solution, measured with a siiandard pipet, was added and mixed with the suspended linters. The flask was stoppered with a cork and thermostated at 30° C. for 30 minutes. At the end of this time the flask was plunged into ice water at 0° C. for 2 minutes. The reaction mixture was filtered through a Gooch crucible into a clean 250-ml. Erlenmeyer flask. Three 20-ml. samples of the filtrate were analyzed for periodate content. Method 2. Method 1 was followed exactly, except that after separation of the linters from the oxidizing solution, the linters were washed with 50 ml. of water added in about 5 ml. portions. The mixture of filtrate and wash water was transferred quantitatively to a 200-ml. volumetric flask and diluted to the mark with water. Three 50-ml. aliquots of this solution were analyzed for periodate content as usual. Method 3. In this experiment the oxidation mixture was analyzed directly. A 0.2 to 0.25-g. sample of conditioned 20-mesh standard cellulose linters was weighed into a 125-ml. Erlenmeyer flask and suspended in 10 ml. of water measured with a standard pipet. Ten ml. of approximately 0.06 N periodic acid was measured with a standard pipet and mixed with the suspended linters. The flask was stoppered with a cork and thermostated at 30° C. for 30 minutes. At the end of this time the flask was plunged into ice water at 0° C. for 2 minutes, and the oxidation medium was im­ mediately analyzed for periodate content. -39- At least, four separate samples of cellulose were used to test each method of sampling described. each sample was calculated. The consumption of periodate by This figure is expressed as milliequiv- alents of HI04 consumed per gram of cellulose, and may be calculated from the equation: HI04 , meq./ g. = (B - S) (W of arsenite) (F) (g. of standard cellulose) where B is the average ml. of arsenite used to titrate the blanks; S is the ml. of arsenite used to titrate the solution or aliquot fraction; and P is a multiplication factor that depends on the frac­ tion of the total solution analyzed. In method 1 this fraction was l/5, therefore the multiplication factor would be 5. In method 2 the factor would be 4, and in method 3, since the whole solution was titrated, the factor is 1. In methods 1 and 2 the three titrations made on each sample did not vary more than £ 0.05 ml. and the average was recorded. The data for the three methods of sampling are listed in Table III. The values of the blank titers are listed opposite the zero sample weights at the top of each group of analyses. From the data in column 5 of table III it can be seen that the variations in the periodate consumption, as determined by the three methods of sampling, are so small that it may be concluded that any one of the three would be satisfactory. It is difficult to decide the limits of reproducibility that should be considered as acceptable in the case of heterogeneous -40- TABLE III RESULTS OF THE ANALYSES OBTAINED BY THREE METHODS OF SAMPLING (Oxidations for 30 Minut es at 30° C.) Method of Sampling Sample Wgt., S» Arsenite Titer, ml. Normality 'of Arsenite hio4 Consumed meq./ g. 1 0.0000 1.4395 0.9364 1.0241 0.9738 39.65 38.30 38.80 38.60 38.60 0.015 0.0000 0.0704 0.0689 0.0769 0.0808 2 0.0000 1.0272 1.0202 0.9664 0.9906 49.55 48.42 48.25 48.28 48.25 0.015 0.0000 0.0654 0.0765 0.0788 0.0788 3 0.0000 0.1910 0.1881 39.70 38.80 38.80 0.015 0.0000 0.0707 0.0718 0.0000 0.1929 39.50 38.52 0.0147 0.0000 0.0753 0.0000 0.2028 37.45 36.50 0.0152 0.0000 0.0720 -41- reactions. It was thought that the maximum error in an arsenite- iodine titration (0.015 U arsenite) "was one drop (0.05 ml.) of the arsenite solution. This error would correspond to 0.00375 milliequiv- alents of HIO4 per gram of cellulose. Therefore, any method that gives periodate consumption figures which check within 0.004 meq./g. would be considered a satisfactory analytical method. Reproduci­ bility within this limit was easily obtained in short time oxidations. On the other hand, larger errors or variations in the periodate consumption by comparable samples may be encountered. Some of the factors to which these variations might be attributed are as follows: (1) Variations in the volumes of solutions delivered from a standard pipet might result in a periodate consumption that would be too high or too low. An error of one drop (0.05 ml.) of 0.06 N periodic acid would correspond to 0.015 milliequivalents of HIO^ per gram of cellulose. (2) The presence of impurities might catalyze the decomposition of periodic acid and result in a periodate consumption that would always be too great. This effect would be more significant in long time oxidations. (3 ) Differences in the swollen condition of the cellulosic mater­ ial might permit more rapid or slower diffusion of the oxidant into the interior of the fiber and thus result in a periodate consumption that would vary according to the rate of diffusion. also be more apparent in long time oxidations. -42- This effect would Oxidation of Cellulose Acetates: The Pour Hour Reaction. Cellulose acetate samples No, 8 to 14 were oxidized in these experiments. Since it was found that the determination of the excess periodic acid in an oxidation mixture by direct titration yielded reproducible results, this simple technique was used in the following experiments. At least two samples of each cellulose acetate and several sam­ ples of standard cellulose were oxidized for each of the following time periods: 0.5, 1.0, 1.5, 2, 2.5, 3, 3.5, and 4 hours. A few of the first oxidations were also recorded for time periods of 0.25 and 0.75 hours, but this practice was later abandoned. Time schedules that would allow the oxidation of 16 samples in one series were established. No two samples of the same cellulose acetate were oxidized for the same length of time in any one schedule. This rule was established so that the reproducibility of the results could be compared between experiments performed on different days and perhaps under slightly different conditions. The standard procedure used for all of these oxidation experiments is described below. A sample of cellulose or cellulose acetate approximately 0.2 g. in weight was weighed into a 125-ml. Erlenmeyer flask and suspended in 10 ml. of water measured with a standard pipet at 20° C. The flask was weighted and thermostated at 30° C. for a 1 hour wetting and swell­ ing period. A 250-ml. glass stoppered conical flask containing 200 ml. of stock periodic acid solution was also thermostated at 30° C. At zero time, 10 ml. of this periodic acid solution was measured with a standard pipet (calibrated for 30° C.) into the reaction flask and mixed with the suspended linters* Care was taken to prevent the lin­ ters from clinging to the sides of the flask and thus avoid contact with the oxidizing medium* The linters were allowed to steep, without any mixing, in the reaction flask for the desired length of time* At the end of the reaction time the flask was plunged into ice water at 0° C. for 2 minutes and analyzed immediately for unreacted periodic acid. The standard arsenite solution was delivered from a 50-ml* buret. After analysis of the sample was completed, or during a 2 hour period following the time of analysis, at least three 10 ml. samples of periodic acid solution were removed from the conical flask that was previouslythermostated with a pipet at 30° C, These samples were also measured calibrated at 30° C. andwere used as blank determina­ tions of the normality of the stock periodic acid solution. The oxidation data for cellulose and cellulose acetates No. 8 to 14 are recorded in table 17. The symbols S.C. and CAc are used to designate standard cellulose and cellulose acetate respectively. The consumption of periodic acid expressed in milliequivalents per gram of cellulose or cellulose acetate is also recorded in table IV. This cal­ culation was made for each sample from the equation: •n»*-—. / _ (B-S) (N of arsenite) ------------- — ■ HIOa meq./ g. : ~ 4 (g. of S.C. or CAc) • where B is the average ml. of arsenite solution used to titrate the blanks, and S is the ml. of arsenite used to titrate the oxidation solution. -44- TABLE 17 OXIDATION OF CELLULOSE AND CELLULOSE ACETATES (Oxidations up to 4 hours) Reaction Time, hrs. Temp. °C. Sample Wgt., Ars enite Titer, ml. Normality of St. Arsenite hio4 Consumed, meq./ g. S.C. 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 29.95 0.0000 0.1974 0.1965 0.2002 0.1944 0.1960 0.2010 0.1965 0.2005 39.50 38.65 37.90 37.40 36.82 36.40 35.90 35.55 34.90 0.0147 0.0000 0.0637 0.1206 0.1550 0.2050. 0.2340 0.2640 0.3003 0.3390 S.C. 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 30.00 0.0000 0.1929 0.1976 0.1995 0.1966 0.2009 0.1972 0.1994 0.2003 39.50 38.52 38.87 37.35 36.90 36.25 35.90 35.50 35.00 0.0147 0.0000 0.0753 0.1220 0.1590 0.1960 0.2400 0.2695 0.2970 0.3320 S.C. 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 29.90 0.0000 0.2028 0.1975 0.1983 0.1995 0.1986 0.1985 0.2006 0.2013 37.45 36.50 35.90 35.30 34.75 34.40 33.90 33.45 33.10 0.0152 0.0000 0.0720 0.1195 0.1650 0.2060 0.2340 0.2720 0.3030 0.3280 CAc No. 8 0.0 0.25 0.5 0.75 1.0 1.5 2.0 29.90 0.0000 0.1903 0.1941 0.1912 0.1901 0.1922 0.1885 38.25 37.75 37.40 37.05 36.85 36.27 35.75 0.0149 0.0000 0.0440 0.0651 0.0935 0.1096 0.1550 0.1980 Sample Number (Cont'd. next page) •45- TABLE IV - Continued Sample Humber Reaction Time, hrs. Temp. . °c. Sample Wgt., S» Arsenite Titer, ml. Normality of St. Arsenite hio4 Consumed, meq./ g. CAc No. 8 0.0 0.25 0.5 0.75 1.0 1.5 2.0 29.90 0.0000 0.1948 0.1925 0.1946 0.1922 0.1917 0.1943 38.25 37.63 37.35 37.00 36.75 36.20 35.70 0.0149 0.0000 0.0474 0.0696 0.0956 0.1160 0.1480 0.1950 CAc No. 8 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 30.00 0.0000 0.1983 0.1986 0.2019 0.2007 0.1987 0.1982 0.1998 0.1985 39.50 38.63 37.95 37.35 36.90 36.45 35.95 35.60 35.25 0.0147 0.0000 0.0655 0.1154 0.1575 0.1920 0.2270 0.2650 0.2900 0.3170 CAc No. 9 0.0 0.25 0.5 0.75 1.0 1.5 2.0 2.5 3.0 3.5 4.0 29.90 0.0000 0.2027 0.1940 0.1948 0.1995 0.2007 0.1998 0.1998 0.2016 0.1980 0.1942 37.80 37.28 37.00 36.75 36.60 35.90 35.55 35.05 34.60 34.25 34.05 0.0147 0.0000 0.0378 0.0607 0.0790 0.0887 0.1390 0.1653 0.2020 0.2330 0.2640 0.2840 CAc No. 9 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 29.90 0.0000 0.2016 0.2018 0.1971 0.1993 0.1969 0.1966 0.2023 0.1969 39.50 38.65 38.10 37.65 37.22 36.80 36.50 35.90 35.75 0.0147 0.0000 0.0625 0.1025 0.1405 0.1690 0.2035 0.2260 0.2640 0.2830 (Cont’d next page) -46- TABLE IV - Continued Sample Number Reaction Time, hrs. Temp. °C. Sample Wgt., g» Arsenite Titer, ml. Normality of St. Arsenite hio4 Consumed, meq./ g. CAc No. 10 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 29.80 0.0000 0.1957 0.1944 0.1931 0.1983 0.1961 0.1988 0.1982 0.1950 39.50 38.90 38.30 37.80 37.30 36.95 38.45 36.20 35.90 0.0147 0.0000 0.0456 0.0915 0.1300 0.1642 0.1920 0.2270 0.2460 0.2760 CAc No. 10 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 29.90 0.0000 0.1933 0.1966 0.1967 0.2012 0.2017 0.2003 0.2001 0.1952 39.50 0.0147 38.30 37.85 37.30 36.90. 36.43 36.20 35.85 0.0000 ... 0.0905 0.1240 0.1620 0.1910 0.2270 0.2440 0.2770 CAc NO. 11 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 29.90 0.0000 0.1957 0.2010 0.2010 0.1982 0.1970 0.2015 0.2024 0.1965 39.50 38.55 38.05 37.55 37.15 36.75 36.42 36.05 35.70 0.0147 0.0000 0.0715 0.1064 0.1433 0.1760 0.2060 0.22 60 0.2520 0.2860 CAc No. 11 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 29.90 0.0000 0.1995 0.1976 0.2002 0.1961 0.1991 0.1952 0.1919 0.1978 39.50 38.55 38.05 37.55 37.10 36.75 36.50 ----- 0.014-7 0.0000 0.0702 0.1085 0.1440 0.1810 0.2040 0.2272 (Cont'd next page) -47- TABLE IV - Continued Sample Number Reaction Time, hrs. Temp. °C. Sample Wgt., g* Ars enite Titer, ml. Normality of St. Arsenite hio4 Consumed, meq./ g. CAc No. 11 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 29.90 0.0000 0.2013 0.1989 0.2009 0.1965 0.2019 0.2005 0.1985 0.1951 37.45 36.50 36.05 35.60 35.20 34.80 34.45 34.20 33.75 0.0152 0.0000 0.0713 0.1065 0.1400 0.1740 0.2000 0.2275 0.2490 0.2880 CAc No. 12 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 29.90 0.0000 0.1986 0.1985 0.1998 0.2001 0.2019 0.2007 0.2028 0.2018 37.60 36.70 36.20 35.50 35.15 34.60 34.15 33.80 33.30 0.0152 0.0000 0.0690 0.1075 0.1600 0.1860 0.2260 0.2 610 0.2860 0.3240 CAc No. 12 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 29.90 0.0000 0.2025 0.2025 0.1986 0.1979 0.1983 0.1986 0.1988 0.1982 37.60 36.65 36.15 0.0152 0.0000 0.0730 0.1090 35.10 34.60 34.15 33.80 33.35 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 29.90 0.0000 0.2015 0.1982 0.1978 0.2018 0.2024 0.1995 0.1985 0.1995 37.45 36.65 36.05 35.60 35.20 34.60 34.40 33.80 33.60 CAc No. 13 0.1920 0.2300 0.2 640 0.2910 0.3260 0.0152 0.0000 0.0605 0.1070 0.1420 0.1690 0.2120 0.2320 0.2800 0.2930 (Cont'd next page) -48- TABLE 17 - Concluded Normality of St. Ars enite Reaction Time, hrs. Temp, 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 29.90 0.0000 0.1988 0.1969 0.2028 0.2011 0.2020 0.2000 0.2011 0.1979 37.45 36.65 36.10 35.55 35.20 34.80 34.40 33.90 33.62 0.0152 0.0000 0.0615 0.1040 0.1425 0.1695 0.1990 0.2310 0.2680 0.2940 CAc No. 14 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 29.90 0.0000 0.1999 0.2019 0.2004 0.2008 0.2021 0.2012 0.2000 0.2009 37.25 36.27 35.65 35.05 34.60 34.15 33.75 33.30 32. B5 0.0152 0.0000 0.0745 0.1210 0.1660 0.2010 0.2340 0.2 640 0.3000 0.3340 CAc No. 14 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 29.90 0.0000 0.2005 0.2012 0.2007 0.2029 0.2018 0.2020 0.2010 0.1991 37.25 36.35 35.65 35.15 34.70 34.10 33.70 33.30 32.85 0.0152 0.0000 0.0685 0.1210 0.1590 0.1920 0.2370 0.2680 0.2980 0.3360 Sample Number °c. Sample Wgt., g. -49- Ars enit e Titer, ml. m o C0XlSUffl’6Cl f meq./ g. In literature dealing with periodate oxidation the consumption of periodate is usually expressed in terms of moles of HIO4 consumed per unit molecular weight. The conversion of mi H i equivalents per gram to moles per unit molecular weight can be made by multiplying by the unit molecular weight and dividing by 2000, as shown below: Consumption of HI0A in moles f / % _ meq./ g. per unit molecular weight ^ (unit molecular weight) (2000) The average consumption of periodate expressed in meq./g. was con­ verted to moles per unit molecular weight for all the oxidation time periods recorded for cellulose and cellulose acetates. are recorded in table 7. These values Inspection of the figures for the consump­ tion of periodate expressed in meq./g. shows that there is a consistent difference between the consumption of periodate by cellulose and that consumed by the cellulose acetates. In general, the consumption of periodate decreases slightly as the per cent combined acetic acid in­ creases. However, the effect of an increasing degree of acetylation on the periodate oxidation of cellulose acetates cannot be seen when the consumption of periodate is expressed in meq./ g., since any cellulose acetate according to its acetyl content, contains less than 100^ cellu­ lose. On the other hand, the consumption of periodate expressed in moles per unit molecular weight shows this effect. In column 4 of table 7 it can be seen that the differences in the consumption of periodate between cellulose and cellulose acetates of low acetyl con­ tent are actually quite small for oxidation periods up to 4 hours. -50- The rate of oxidation may be illustrated by plotting the consump­ tion of HIO4 in moles per unit molecular weight against reaction time in hours. Such a graph shows the effect of increasing degree of acetyl- ation on the periodate oxidation of cellulose acetates, i.e., the acetyl substituents on the glycol groups of cellulose prevent cleavage of these glycol groups by periodate oxidation, and therefore, the higher the degree of acetylation the lower the consumption of periodate per unit of time. Figure 2 is a graph showing the rate of oxidation curves for stand­ ard cellulose and cellulose acetate No. 9 (6.83% HOAc). Curves for the other cellulose acetates are not shown because they are either the same as that for cellulose (CAc No. 14, 1.28% HOAc), or fall slightly below the curve for cellulose (CAc No. 8, 3.17% HOAc and CAc No. 12, 3.84% HOAc), or fall near the curve for cellulose acetate No. 9 (CAc No. 11, 9.53% HOAc, and CAc No. 13, 7.10% HOAc). -51- TABLE V CONVERSION OF PERIODATE CONSUMPTION FROM MILLIEQUIVALENTS PER GRAM TO MOLES PER UNIT MOLECULAR WEIGHT Reaction Sample No. Time, _______________ hrs. Periodate Consumption Milliequivalents Moles per Unit per Gram (av.) Molecular Wgt. 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.0703 0.1204 0.1600 0.2020 0.2360 0.2690 0.3000 0.3330 0.005695 0.009750 0.01295 0.01635 0.0191 0.0218 0.0243 0.0270 CAc No. B 3.17% HOAc 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.0667 0.1137 0.1535 0.1950 0.2270 0.2 650 0.2900 0.3170 0.005525 0.00942 0.01272 0.01616 0.01881 0.02198 0.02403 0.02630 CAc No. 9 6.83% HOAc 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 * 0.0616 0.0956 0.1402 0.1671 0.2027 0.2300 0.2 640 0.2835 0.00524 0.00813 0.01192 0.01422 0.01722 0.01955 0.02245 0.02410 CAc No. 10 6.06% HOAc 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.0456 0.0910 0.1270 0.1631 0.1915 0.2270 0.2450 0.2 765 0.00386 0.00770 0.01075 0.01380 0.01620 0.01920 0.02072 0.02340 S.C. (Cont’d next page) -52- TABLE V - Concluded Reaction Periodate Consumption Sample No. Time, Mi H i equ iva 1ent s Holes per Unit _____________ hrs.___ per Gram (av.) Molecular Wgt. CAc No. 11 9.53/? HOAc 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.0710 0.1071 0.1424 0.1770 0.2033 0.2270 0.2505 0.2B70 0.00616 0.00930 0.01236 0.01535 0.01765 0.01970 0.02174 0.02490 CAc No. 12 3.84% HOAc 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.0710 0.1082 0.1600 0.1890 0.2280 0.2625 0.2885 0.3250 0.00590 0.00900 0.01333 0.01574 0.01900 0.02185 0.02400 0.02705 CAc No. 13 7.10% HOAc 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.0610 0.1055 0.1422 0.1692 0.2050 0.2315 0.2740 0.2935 0.00520 0.00899 0.01212 0.01442 0.01747 0.01975 0.02338 0.02500 CAc No. 14 1.28% HOAc 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 0.0715 0.1210 0.1635 0.1965 0.2355 0.2660 0.2990 0.3350 0.00584 0.00988 0.01335 0.01605 0.01923 0.02172 0.02430 0.02735 -53- weight X 100 molecular unit per of HIO 4 , mols Consumption 1.0 Time, hours Figure 2 Periodate oxidation, initial reaction, 30*C. O Cotton cellulose + Cellulose acetate N o 0 9, 6 083$ H0Aco Long Time Oxidations and the Effect of Mixing. In the preceding section cellulose and cellulose acetates were oxidized for time periods up to four hours. During this more or less initial phase of the reaction, with respect to the over-all theoreti­ cal oxidation, the rate of reaction is fairly rapid, due to the greater accessibility of the glycol groups present in the amorphous and semi­ crystalline areas of cellulose. A study of the effect of partial acetylation on the periodate oxidation of cellulose would not be com­ plete unless the oxidations were carried out over a fairly long period of time. This section deals with oxidation experiments that were per­ formed for time periods of several days, and the reaction period that was selected for comparisons between cellulose and cellulose acetates was 60 hours. The first experiment in long time oxidation was performed on standard cellulose for varying periods of time up to 120 hours. Except for the volume of oxidizing solution and the method of sampling for excess HIO4, the oxidation procedure was the same as that used in the 4-hour oxidations. The total amount of available periodate in a 4- hour oxidation mixture was approximately 3 meq. per gram of cellulose or 0.243 moles of periodate per unit molecular weight. In 4 hours cellulose consumed 0.027 moles of HIO4 per unit molecular weight, and therefore, the concentration of the oxidation medium changed a little over 10% during the course of the oxidation. It was thought that the concentration of the oxidizing medium could be allowed to change as much as 50% without decreasing the rate ■55- of reaction, since in this heterogeneous oxidation the rate of oxida­ tion is dependent more on the rate of diffusion of periodate into the cellulose structure, than on the concentration of the oxidizing medium. However, in order to minimize any possible change in the rate of oxi­ dation due to a decrease in the concentration of the oxidant, a con­ siderable excess of periodic acid was used in these long time oxida­ tions. The volume of oxidant selected for use in the following experi­ ments was such that there were 7.5 milliequivalents of HIO^ for each gram of cellulosic material. A 0.2-g. sample of cellulose was weighed into a 125-ml. Erlenmeyer flask, wetted with 25 ml. of water, and 25 ml. of approximately 0.06 N HIO4 solution was added. After oxidation and cooling the mixture was filtered through a Gooch crucible into a clean 125-ml. flask. Two 20-ml. samples of this filtrate were analyzed for excess periodate. Blank determinations were carried out at 24 hour intervals. The aver­ age titration volumes of arsenite solution were used for the calcula­ tion of periodate consumption. The equation used to calculate the con­ sumption of periodate was as follows: HIO4 meq./g. Z (B-S) (N of arsenite) (g. of cellulose) (2.5) where B is the ml. of arsenite used to titrate the blank; and S is the average titer for the 20-ml. fractions of the filtrate solution. The results of these oxidations are recorded in table VI. -56- TABLE VI PERIODATE OXIDATION OF CELLULOSE UP TO 120 HOURS Reaction Time, hrs. 0 3 6 12 IB 24 36 48 60 77.5 96 120 Temp. °C. Sample Wgt., g» 29.85 0.0000 0.1997 0.1993 0.2005 0.2011 0.2002 0.2002 0.1985 0.1972 0.2000 0.2015 0.1996 Arsenite Titer, ml. 44.85 43.20 42.10 40.65 39.00 37.70 35.35 33.10 32.20 28.65 26.30 23.15 Normality of St. Arsenite hio4 HIO4 Consumed Consumed meq./ g. Moles per Unit Mol. Wgt. 0.01462 0.0000 0.2923 0.5040 0.765 1.064 1.305 1.735 2.165 2.344 2.960 3.350 3.710 0.00000 0.02318 0.04082 0.06195 0.0862 0.1057 0.1405 0.1754 0.1900 0.2398 0.2713 0.3004 In all of the previous experiments the oxidation mixture was not stirred or mixed in any way during the reaction period. Other investi­ gators studying the periodate oxidation of cellulose have mentioned the use of ’’frequent shaking” or ’’occasional stirring” as having been applied during the course of the reaction* Any procedure of this sort would be very difficult to standardize for a great number of oxidation mixtures, and therefore, it was thought that the problem of applying c controlled stirring during the reaction period should be investigated. Because of the limitations of space, equipment, and time, the application of stirring to a large number of small reaction flasks was considered to be unfeasible. A more efficient method of attack would -57- be to oxidize a fairly large amount of material with a proportionally larger volume of oxidizing solution, the mixture being stirred at a constant rate. Samples of the oxidation media could be removed at various time intervals, analyzed, and the decrease in concentration of the oxidizing solution used to calculate the periodate consumed by the cellulosic material. Sampling a reaction mixture that is a homo­ geneous solution in this way would only change the total volume of the mixture and would not change the concentration of the materials in solution. However, in any mixture of lirrters suspended in an oxidiz­ ing solution, it would be quite impractical if not impossible to obtain a homogeneous sample of the mixture. On the other hand, if a considerable excess of periodate is available at the beginning of an oxidation, small samples of the solution may be removed at intervals over a period of time without appreciably affecting the rate of the reaction. For example: in one particular oxidation reaction, the available periodate at the beginning of the oxidation was 7,18 meq./g., and at the end of 60 hours there was still 5.03 meq./ g. left in the reaction flask, even though four 10-ml. samples of the solution had been removed. This means that 70% of the initial periodate is still available, and therefore, the decrease in the amount of available periodate due to sampling would not have affected the reaction rate appreciably. In the next experiments oxidations were carried out with mechani­ cal stirring. The reaction mixtures were stirred with cone-drive electric stirring motors, and samples were removed from the reaction -58- mixture at various intervals of time. Several 1-liter 3-neck round bottom flasks were fitted with stirrers and filter sticks. The stir­ rer was a 6-mm. glass rod with a 45° bend 4 cm. from one end, and was held in place in the flask by a glass sleeve extending through a cork in the neck of the flask. The filter stick was made from a 7-mm. (inside diameter) x 25-cm. pyrex tube by shrinking the diameter of the tube in an oxygen flame to 1 mm. at a point 1 cm. from one end. A com­ pact mass of fine glass wool was pressed into this small cup and trim­ med even with the end of the tube. to about 250 revolutions per minute. Three stirring motors were regulated Three sets of the reaction flask- stirring apparatus could be placed in the constant temperature bath at one time. Oxidations were performed as followss a 2-g. sample of cellulose or cellulose acetate was weighed into the reaction flask and suspended in 250 ml. of water measured with a standard 50-ml. pipet at 20° C. This mixture was stirred for 1 hour at 30° C. At zero reaction time 250 ml, of approximately 0.06 II periodic acid was measured with the same standard 50-ml. pipet and added to the aqueous suspension of linters. The reaction temperature rose to 30° C. within five minutes. The available periodate in this oxidizing medium was 7.5 meq./g. or 0.607 moles of HIO^ per unit molecular weight. At intervals of time, according to a previously arranged schedule, 10-ml. samples of the oxidizing solution were withdrawn through the filter stick with a standard 10-ml. pipet calibrated at 30° C. The inside diameter of the filter stick was such that the pipet could be inserted to the -59- point of constriction in the filter stick. for periodic acid content. The samples were analyzed Blank determinations were made at 24 hour intervals. Several oxidations with stirring were performed on cellulose and cellulose acetate for periods up to 6 hours, with sampling at 0.5 hour intervals. This oxidation reaction was found to progress more rapidly than that previously observed during a 4 hour period without stirring. However, the results were not satisfactory from the standpoint of re­ producibility. Apparently a variation in the rate of stirring affected the rate of the reaction. The data for one sample of cellulose is recorded in table VIII so that a comparison may be made between the rate of reaction between stirred and non-stirred reactions. All other data obtained in these experiments will be omitted. The same oxidation procedure (above) was used to test the repro­ ducibility of results that would be obtained during a reaction where the oxidation period was as long as 60 hours. The stirring motors were very carefully readjusted and synchronized for the same rate of stirring. The reaction medium was sampled at 12 hour intervals. The data obtained for cellulose and two samples of cellulose acetate are recorded in table VIII. The calculation of the consumption of periodate by a known weight of cellulose is more complex in the case of a reaction mixture where samples of the oxidant are removed at intervals. The calculation may be simplified if, for each time of sampling the total periodate content of the solution is determined. For example, at zero time the normality 60' of the oxidizing medium (determined by a blank analysis) multiplied by the •volume of the solution yields the total milliequivalents of periodate present at this time. Since the concentration of the solu­ tion is also determined at each sampling time, the total milliequivalents of periodate in the volume of solution present before sampling may also be determined for each sampling time. But each sample re­ moved contains some periodate, and therefore, the sum of the milliequivalents of periodate removed in previous samples must be added to the milliequivalents present in the volume of solution at any sampling time. The difference between the periodate content of the reaction mixture at zero time, and periodate content present at any sampling time plus that present in the volumes of previously removed samples, represents the periodate consumption of the cellulosic material. This difference divided by the weight of cellulose or cellulose ace­ tate is the consumption of periodate in meq./g. Table VII is the record of a typical oxidation reaction that was stirred and in which the progress of the reaction was followed by sampling the oxidizing medium at 12 hour intervals. -61- TABLE VII RECORD OF DATA AND CALCULATIONS FOR A TYPICAL OXIDATION Cellulose sample, 1,9667 g. A Total Vol. of HI04 Solution ml. B Ars enite titer for 10 ml., ml. Normality (N) of arsenite, 0,01506 C D Normality of HIO Solution B x N 10 Meq. of HI04 re­ moved in 10 ml. samp. B x N E F Meq. of Meq. of HI04 in HI04 in Volume samples previous­ (A) D x A ly remov­ 10 ed 0 C onsumption of Periodate by the Cellulose Sample,Meq. Consump­ tion of H!0 meq./ g. Reaction Time, hrs. Samp. No. 00 0 500 21.35 0.03215 0.000 16.08 0.000 16.08 0.000 0.000 12 1 500 19.05 0.0287 0.267 14.34 0.000 14.34 1.740 0.885 24 2 490 17.35 0.0261 0.261 12.81 0.287 13.097 2.983 1.518 36 3 4B0 16.05 0.0242 0.242 11.60 0.548 12.148 3.932 2.000 48 4 470 14.75 0.0222 0.222 10.45 0.790 11.240 4.840 2.461 60 5 460 13.70 0.0206 0.206 9.48 1.012 10.492 5.588 2.e42 Meq. of HIO, in Vol.(A) plus Meq. of HI04 in samp.re­ moved (F) TABLE VIII OXIDATION OP CELLULOSE AND CELLULOSE ACETATES WITH STIRRING Samp. No. °c. Sample Wgt., g- Arsenite Titer, ml. N of St. Ars. 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 5.0 6.0 29.85 1.9079 22.42 21.95 21.65 21.50 21.35 21.25 21.10 21.00 20.87 20.65 20.40 0.01462 00 12 24 36 48 60 72 96 120 168 250 29.85 22.42 20.15 18.60 17.40 16.30 15.15 13.90 11.65 9.65 5.30 0.00 0.01462 00 12 24 36 48 60 29.85 21.35 19.05 17.35 16.05 14.75 13.70 0.01506 CAc ’ No. 18 8.23/ HOAc 00 12 24 36 48 60 29.85 21.35 19.65 18.25 16.95 15.90 15.00 0.01506 CAc No. 20 14.14/ HOAc . 00 ’ 12 24 36 48 69 29.85 21.35 19.68 18.60 17.55 16.60 15.80 0.01506 S.C. s .c . S.C. React. Time, hrs. Temp. 1.9149 1.9667 2.0013 1.9987 63- hio4 Consumed meq./ g. hio4 Consumed moles per U.M.W. 0 .0 0 0 0 0 .0 0 0 0 0.1940 0.2880 0.3740 0.4190 0.4492 0.5110 0.5423 0.5940 0.6590 0.7320 0.0157 0.0233 0.0303 0.0339 0.0364 0.0414 0.0439 0.0481 0.0534 0.0593 0 .0 0 0 0 0 .0 0 0 0 0.8780 1.479 1.910 2.300 2.706 3.132 3.900 4.560 5.950 7.600 0.0712 0.1198 0.1547 0.1864 0.2190 0.2540 0.3160 0.3695 0.4840 0.6160 0 .0 0 0 0 0 .0 0 0 0 0.885 1.518 2.000 2.461 2.842 0.0717 0.1230 0.1620 0.1995 0.2302 0 .0 0 0 0 .0 0 0 0 0.639 1.106 1.628 2.000 2.306 0.0549 0.095 0.140 0.172 0.198 0 .0 0 0 0 .0 0 0 0 0.630 1.023 1.415 1.750 2.023 0.0566 0.0918 0.12 72 0.1575 0.1820 Oxidation of Cellulose Acetates for 60 Hours with Standard Mixing. The proceeding oxidation experiments, in which standard cellulose and cellulose acetates No. 8 and 14 were oxidized for 60 hours with stirring, and in which the oxidation mixtures were sampled at 12 hour intervals to determine the decrease in periodate concentration, indi­ cated that there was an appreciable difference in the consumption of periodate between standard cellulose and cellulose acetates.over this period of time. The problem of synchronizing the electric stirring motors, so as to produce an equal rate of stirring, still remained to be solved. The only logical solution to this problem appeared to lie in the design and construction of a special standard mixing apparatus that would not only fulfill the limitations of space and the necessity of a constant temperature bath, but also permit the mixing of several reaction mixtures under exactly the same conditions. The thermostating and standard mixing apparatus is illustrated in figure 3. The specifications and construction of the standard mixing apparatus are described below. The main axis of rotation (2) was made from a 0.5-inch round steel rod, 46 cm. in length. A 1-cm. flat was machined along the top of this rod, except for 2 cm. at the points of support. The main axis was sup­ ported on the battery jar by two bearings that could be clamped firmly to the jar, and held in place by cotter pins at the bearings. Four motor arms, 20 cm. in length were made to fit on the axis, and could be fixed in any desired position by tightening a set screw on the flat of the axis. Three of the motor arms were fixed about 10.5 cm. apart on -64 the main axis. A 2-mm. hols was drilled in each end of these arms at a point 6 cm. from the axis. The fourth motor arm was fixed on the right end of the axis, and was connected to the source of power by means of a connecting rod attached 5 cm. off-center at the motor arm, and 1.5 cm. off-center on a wheel at the power source. The length of the connecting rod could be adjusted and was free to move around a bearing at the points of connection. The source of power (4) was a one-third horse power, 1725 r.p.m. electric motor coupled with a reducing gear (reduction ratio, 21.5) by means of a 0.5-inch V belt operating on 1-inch pulleys. The angular speed of the power source was 80 r.p.m., so that the up and down motion observed in the reaction flasks also took place at a rate of 80 cycles per minute. The reaction flask detail can be seen quite clearly in figure 4. Ten matched sets of this apparatus were assembled. The flask was a 250-ml. 3-neck round bottom pyrex flask, 10 cm. in diameter, and selected for uniformity of dimensions. The plunger was made of pyrex glass by fusing a 6-mm. x 12.5-cm. rod to a cup, 2.5 cm, in outside diameter and 3 cm. long. A No. 12 cork was fitted with a sleeve made from a 7-mm. (inside diameter) pyrex tube 5 cm. long, with a 1-cm. jacket of 6-mm. rubber tubing at one end. The glass rod portion of the plunger was put through this snugly fitting sleeve •which acts as a bearing in the up and down motion cycle of the plunger. A 1-cm. length of 6-mm. brass rod, containing a 2-mm. transverse hold drilled 2 mm. from one end was fitted with a 1 x 1.4-cm. steel wire staple bent in -65- a \_] shape. The bottom half of the small brass rod was connected to the top of the plunger rod with a 1.5-cm. length of tough 4-mm. rubber tubing. wire. This connection was wrapped with five turns of 1-mm. copper The top part of the plunger rod was lubricated with a small amount of silicone grease to facilitate the operation of the plunger in the bearing sleeve. The filter stick was made from a pyrex tube, 7 mm. in inside dia­ meter, and 14 cm. in length. The diameter of the tube was carefully reduced in an oxygen flame to 1 mm. at a point 1 cm. from one end. The top of the filter stick was jacketed with a 1.5-cm. length of 6-mm. rubber tubing. The filter stick was held in position in one of the side necks of the reaction flask by a No. 8 cork. A cork of the same size was used to stopper the other side neck. The flasks were held in a rigid position in the constant tempera­ ture bath by specially constructed clamps. These clamps were in turn held by universal clamps attached to a transverse rod as shown in figure 3. as follows: A flask clamp that would hold three 250-ml. flasks was made three holes 3 cm. in diameter were drilled with 10.5 cm. between centers along the width of a piece of wood 3 cm. thick, 5 cm. wide, and 27 cm. long. Eight-mm. holes were drilled transverse to the width and halfway between the centers of the large holes. The wood was then cut in half along the width, and each half was sanded down so as to fit comfortably between the necks of the flasks. was bolted to a 20-cm. x 0.5-inch steel rod. One of the halves The clamps when in posi­ tion around the center necks of the three flasks were held together by two bolts with wing nuts. -66- To assemble the standard mixing apparatus for an oxidation experi­ ment the main axis of rotation, with its motor arms, was mounted across the diameter of the constant temperature bath. Pieces of chamois leather were placed between the glass and metal and the supports were clamped firmly in place. This part of the apparatus place permanently if desired. could be left in The flasks containing samples to be oxidized were arranged in the flask clamps and were fixed temporarily in position in the bath. The motor arms were tilted up or down to per­ mit passage of the flask assembly, A plunger with cork was placed in each flask and the wire staple at the top of the plunger rod was spread apart and pressed into the small hole at the end of the motor arm, small pair of pliers was used in this operation. seated firmly in the neck of the flask. A The cork was then The flask clamps were adjusted to permit smooth operation of the plungers. The mixing apparatus was then connected to the source of power which had been carefully aligned with the axis of rotation. The flask clamps were readjusted if opera­ tion seemed noisy or caused excessive vibration. Cross rods (not shown in figure 3) were clamped permanently between the vertical rods support­ ing the source of power. These rods maintained the alignment of the apparatus and reduced the over-all vibration considerably. When adjust­ ed properly the apparatus could be left in operation for periods as long as a week without requiring any attention or servicing. While in operation the constant temperature bath was shielded from the electric motor by a thick asbestos sheet. motor operated at about 65° C, This was found necessary because the Bearings and plunger rods were occasion­ ally lubricated with silicone grease. -67 Figure 3 Thermostating and Standard Mixing Apparatus (1) Thermostat. (2) Axis of rotation with motor arms. (3) Reaction flasks. (4) Source of power. -68- II II f-'— n t -----) L _ ------------— h ------------- w~f Figure 4A Flask, 2S0 ml. J-neck R.B. B Filter stick C Glass plunger moving through distance D P'lexible connection E Motor arm, fixed on axis of rotation 1d 1 Oxidation experiments on cellulose and cellulose acetates No. 16 to 26 were performed in the standard mixing apparatus. were performed at the same time. Six oxidations The procedure was as follows: one-g. samples of standard cellulose or cellulose acetate were weighed into the reaction flasks and suspended in 150 ml. of water measured with a standard 50-ml. pipet at 20° C. The flasks were then assembled in the constant temperature bath, and the plungers were placed in position and connected with the motor arms. The electric motor was started and the aqueous suspension of linters was mixed for 1 hour. Minor adjust­ ments in the alignment of the flask apparatus was made during this time, so that the operation of the mixing apparatus was as smooth and quiet as possible. period. The filter sticks were also prepared during this A compact mass of pure glass wool was pressed into each fil­ ter stick cup and trimmed even with the end of the tube. The six re­ action flasks were numbered from 1 to 6. At zero reaction time, 150ml. volumes of approximately 0.06 N HIO4 solution were measured with a standard 50-ml. pipet at 20° C. and added to the reaction mixtures in the order: 1, 2, 3, 4, 5, and 6. minute intervals. The additions were made at 5 A filter stick was placed in each reaction flask. At each sampling time, i.e., at oxidation periods of: 12, 24, 36, 48, and 60 hoursj a 10-ml. sample of the oxidizing solution was removed from each flask through the filter stick with a 10-ml. pipet calibrated for 30° C., and analyzed for excess periodic acid. The 5 minute inter­ val between sampling of different reaction flasks was sufficient to allow a short chilling period (30 seconds) and completion of the 70 arsenite titration for any sample before the next one was removed. The standard arsenite solution was delivered from a 25-ml. buret. Blanks were determined at 24 hour intervals. At least two oxidations were performed for each cellulose ace­ tate No. 16 to 25. No two oxidations of the same acetate were per­ formed during any one 60 hour period or in the same reaction flask. This rule allows comparison of the reproducibility of results obtained on different days and perhaps under slightly different conditions. The data for oxidations of cellulose and cellulose acetates No. 16 to 25, for 60 hours with standard mixing, are recorded in table IX. Calculation of the periodate consumption was made in the same manner as described on page 60, and as illustrated in table VII. Table X shows the conversion of the average consumption of HIO^ in milliequivalents per gram to moles per unit molecular weight. Figures 5 and 6 are graphs of the consumption of periodate in moles per unit molecular weight plotted against reaction time. It can be seen that over a reaction period of 60 hours, the rate of periodate oxidation of cellulose and cellulose acetates is approximately the same, but the acetates, in the order of increasing degree of acetylation, con­ sume less periodate than is consumed by cellulose. -71- TABLE IX OXIDATION OF CELLULOSE ACETATES FOR 60 HOURS WITH STANDARD MIXING. Reaction Time hrs. Temp. °C. Sample Wgt., e* Arsenite Titer, ml. S.C. 00 12 24 36 48 60 2 9 .8 5 1.0012 S.C. 00 12 24 36 48 60 29.85 S.C. 00 12 24 36 48 60 S.C. S.C. Sample Number Normality of St. Arsenite hio4 Consumed, meq./ g. 1 9 .7 5 1 7 .9 5 1 6 .8 0 1 5 .7 0 1 4 .7 0 1 3 .8 0 0 .0 1 5 0 6 0 .0 0 0 0 0 .8 2 0 0 1 .3 1 9 0 1 .7 8 6 0 2 .2 0 0 0 2 .5 4 8 0 1.0009 19.75 18.00 16.75 15.70 14. 70 13.70 0.01506 0,0000 0.7800 1.3270 1.7750 2.1800 2.5990 29.85 1.0001 19.75 17.95 16.90. 15.90 14.80 13.90 0.01506 0.0000 0.8100 1.2700 1.6955 2.1362 2.5132 00 12 24 36 48 60 29.85 0.9990 19.675 17.87 16.65 15.68 14.70 13.80 0.01506 0.0000 0.8200 1.3510 1.7600 2.1600 2.5130 00 12 24 36 48 60 29.85 1.0023 19.675 17.90 16.60 15.55 14.55 13.65 0.01506 0.0000 0.8000 1.3660 1.8060 2.2200 2.5700 (Cont'd next page) -72- TABLE IX - Continued Arsenite Titer, ml. Normality of St. Arsenite hio4 Consumed, meq./ g. 0.01335 0.0000 Reaction Time hrs. Temp. °C. Sample Wgt., g. 00 12 24 36 48 60 29.e5 0.9973 00 12 24 36 48 60 29.85 1.0040 19.75 18.20 17.00 15.95 15.00 14.10 0.01506 0.0000 0.6980 1.2100 1.6600 2.0400 2.3950 CAc No. 16 00 12 24 36 48 60 29.85 0.9995 19.55 18.05 16.90 15.90 14.90 13.85 0.01506 0.0000 0.6800 1.1780 1.6040 2.0100 2.4200 CAc No. 1.7 00 12 24 36 48 60 29.es 1.0030 19.75 18.35 17.25 16.40 15.50 14.60 0.01506 0.0000 0.6380 1.1200 1.4700 1.8420 2.1900 CAc No. 17 00 12 24 36 48 60 29.85 0.9998 19.75 18.30 17.40 16.40 15.12 14.20 0.01506 0.0000 0.6600 1.0445 1.4725 1.8965 2.2490 Sample Number S.C. CAc No. 16 21.72 18.40 1.3400 16.35 2.1320 (Cont'd next page) 73- TABLE' IX - Continued Hormality of St. Ars enite hio4 Consumed, meq./ g. 19.75 18.45 17.27 16.30 15.32 14.42 0.01506 0i0000 0.6000 1.1020 1.5120 1.9147 2.2657 1.0027 19.75 18.50 17.48 16.60 15.80 15.00 0.01506 0.0000 0.5680 1.0080 1.3840 1.7150 2.0140 29.85 1.0000 19.75 18.40 17.50 16.77 15.80 15.10 0.01506 0.0000 0.6100 1.0030 1.3145 1.7070 1.9945 00 12 24 36 48 60 29.85 1.0005 19.75 18.65 17.70 17.00 16.30 15.60 0.01506 0.0000 0,5000 0.9180 1.2110 1.5050 1.7700 00 12 24 36 48 60 29.85 1.0017 19.75 18.60 17.70 16.95 16.25 15.53 0.01506 0.0000 0.5190 0.9080 1.2310 1.5150 1.8000 Reaction Time hrs. Temp. °c. Sample Wgt.., g* 00 12 24 36 48 60 29.85 1.0008 CAc Ho. 18 00 12 24 36 48 60 29.85 CAc Ho. 18 00 12 24 36 48 60 CAc Ho. 19 CAc Ho. 19 Sample Humber CAc Ho. 17 Ars enite Titer, ml. (Cont’d next page) -74- TABLE IX - Continued Normality of St. Arsenite hio4 Consumed, meq./ g. 19.75 18.60 17.62 16.90 16.10 15.45 0.01506 0.0000 0.519 0.938 1.252 1.578 1.830 1.0005 19.55 18.35 17.48 16.75 16.02 15.30 0.01506 0.000 0.540 0.923 1.230 1.531 1.810 29.85 0.9982 19.80 18.72 17.92 17.00 16.25 15.60 0.01506 0.000 0.482 0.840 1.220 1.535 1.780 00 12 24 36 48 60 29.85 0.9980 19.75 IB.85 17.70 16.90 16.10 15.40 0.01506 0.000 0.401 0.908 1.252 1.580 1.850 00 12 24 36 48 60 29.85 1.0006 19.55 18.35 17.60 16.90 16.10 15.45 0.01506 0.000 0.540 0.873 1.168 1.494 1.747 Sample Number Reaction Time hrs. Temp. °C. Sample Wgt., g» CAc No. 19 00 12 24 36 48 60 29.85 1.0018 CAc No. 19 00 12 24 36 •48 60 29.85 CAc No. 20 00 12 24 36 48 60 CAc No. 20 CAc No. 20 Arsenite Titer, ml. (Cont'd next page) -75- TABLE IX - Continued Normality of St. Arsenite HI04 Consumed, meq./ g. 19.80 18.20 17.15 16.15 15.40 14.45 0.01506 0.000 0.721 1.189 1.600 1.910 2.2 75 1.0013 19.75 18.10 17.15 16.20 15.30 14.35 0.01506 0.000 0.750 1.175 1.575 1.931 2.300 1.002 8 19.80 18.30 17.30 16.45 15.67 14.90 0.01506 0.000 0.669 1.110 1.460 1.782 2.085 1.0021 19.75 18.30 17.25 16.40 15.55 14.75 0.01506 0.000 0.659 1.112 1.471 1.904 2.130 1.0015 19.75 18.45 17.45 16.50 15.60 14.80 0.01506 0.000 0.640 1.030 1.427 1.800 2.103 Sample Number Reaction Time hrs. Temp. °C. Sample Wgt., g. Ars enit e Titer, ml. CAc No. 21 00 12 24 36 48 60 29.85 0.9986 CAc No. 21 00 12 24 36 48 60 29.85 CAc No. 22 00 12 24 36 48 60 29.85 CAc No. 22 00 12 24 36 48 60 29.85 CAc No. 23 00 12 24 36 48 60 29.85 * ■ ■K (Cont’d next page) -76- TABLE IX - Continued Normality of St. Arsenite hio4 Consumed, meq./ g. 19.75 18.40 17.35 16.45 15.60 14.85 0.01506 0.000 0.578 1.110 1.438 1.790 2.085 1.0021 19.55 18.15 17.10 16.30 15.40 14.50 0.01506 0.000 0.628 1.095 1.420 1.790 2.142 29.85 0.9989 19.75 18.58 17.50 16.65 15.80 15.05 0.01506 0.000 0.522 1.005 1.357 1.706 2.000 00 12 24 36 48 60 29.85 1.0028 19.75 18.35 17.50 16.60 15.85 15.05 0.01506 0.000 0.638 1.000 1.356 1.685 2.000 00 12 24 36 48 60 29.85 1.0003 19.55 18.20 17.28 16.50 15.70 14.80 0.01506 0.000 0.610 1.010 1.335 1.667 2.021 Sample lumber Reaction Time hrs. Temp. °c. Sample Wgt., g. CAc No. 23 00 12 24 36 48 60 29.85 1.0022 CAc No. 23 00 12 24 36 48 60 29.85 CAc No. 24 00 12 24 36 48 60 CAc No. 24 CAc No. 24 Arsenite Titer, ml. (Cont'd next page) -77- TABLE IX - Concluded Sample Numb er Reaction Time hrs. Temp. °C. Sample Wgt., g* CAc No. 25 00 12 24 36 48 60 29.85 1.0006 00 12 2436 48 60 29.85 CAc No. 25 Normality of St. Arsenite HI04 Consumed, meq./ g. 19.80 18.50 17.40 16.72 15.95 15.20 0.01506 0.000 0.580 1.070 1.360 1.670 1.956 19.75 18.72 17.60 16.72 15.85 15.10 0.01506 Arsenite Titer, ml. ■ 0.9994 -78- ' 0.000 0.460 0.960 1.335 1.683 1.964 TABLE X COWERS ION OF PERIODATE CONSUMPTION FROM MILLI EQUIVALENTS PER GRAM TO MOLES PER UNIT MOLECULAR -WEIGHT Periodate Consumption Reaction Time, hrs. Mi1li equ ivalents per Gram (av.) Moles per Unit Molecular Wgt. 12 24 36 48 60 0.8080 1.3290 1.7645 2.1741 2.5460 0.0655 9.1077 0.1430 0.1760 0.2063 CAc No. 16 0. 65% HOAc 12 24 36 48 60 0.689 1.194 1.632 2.025 2.407 0.0561 0.0972 0.1328 0.1647 0.1958 CAc No. 17 5,13% HOAc 12 24 36 48 60 0.633 1.090 1.484 1.881 2.235 0.0532 0.0916 0.1247 0.1582 0.1878 CAc No. 18 8.23% HOAc 12 24 36 48 60 0.590 1.005 1.349 1.711 2.004 0.0507 0.0864 0.1160 0.1471 0.1723 CAc No. 19 12.09% HOAc 12 24 36 48 60 0.5176, 0.9216 1.231 1.532 1.800 0.0458 0.0815 0.1089 0.1368 0.1592 CAc No. 20 14.14% HOAc 12 24 36 48 60 0.474 0.874 1.213 1.536 1.792 0.0426 0.0789 0.1090 0.1381 0.1611 Sample No. S.C. (Cont’d next page) 79 TABLE X - Concluded Periodate Consumption Reaction Time, hrs. Milliequivalents per Gram (av.) Moles per Unit Molecular Wgt. CAc No. 21 3.40# HOAc 12 24 36 48 60 0.735 1.182 1.587 1.920 2.287 0.0610 0.0981 0.1317 0.1585 0.1900 CAc No. 22 5.31# HOAc 12 24 36 48 60 0.6635 1.1110 1.465 1.843 2.107 0.0558 0.0935 0.1233 0.1552 0.1773 CAc No. 23 6.60# HOAc 12 24 36 48 60 0.615 1.080 1.435 1.793 2.110 0.0522 0.0917 0.1216 0.1522 0.1792 CAc No. 24 8.65# HOAc 12 24 36 48 60 0.590 1.005 1.350 1.686 2.007 0.0509 0.0867 0.1165 0.1455 0.1731 CAc No. 25 9.94# HOAc 12 24 36 48 60 0.520 1.015 1.347 1.676 1.960 0.0452 0.0882 0.1171 0.1458 0.1705 Sample No. -80- Consumption of hiu 4 , mods per unit moiecuiar weignt 0.20 0.10 O Cotton cellulose -+ Acetate No. 16, 0.65$ HOAc A Acetate No. 17, 5*13$ HOAc • Acetate No. 18, 8.23$ HOAc A Acetate No. 20,14.14$ HOAc 0.05 24 3 Time, hours 48 Figure 5 Periodate oxidation, 60 hours at 30°C., with standard mixing. -81- weight molecular unit per of H 1O4 , moIs Consumption 0.10 O Cotton cellulose + Acetate No. 21, 3.40^ HOAc 0.05 22, • 5.31% HOAc Acetate No. 24, 8.65% HoAc Acetate No. 15, 12.09$ HOAc Time, hours Figure 6 Periodate oxidations, 60 hours at 30°C., with standard mixing. Periodate Oxidation of Sulfuric Acid Hydrocelluloses. The cellulose acetates used in this investigation were acetylated in the presence of sulfuric acid. cellulose in two ways: This acid may react with (1) it may hydrolyze some of the glycosidic linkages of cellulose, and thereby decrease the degree of polymeri­ zation of the cellulose and increase the number of terminal groups (those that have three adjacent hydroxyl groups); (2) it may form sulfoesters with the hydroxyl groups of cellulose. In terms of the reaction of periodic acid with cellulose, an increase in the number of adjacent hydroxyl groups would increase the consumption of perio­ date, but, if some of the glycol hydroxyls are esterified with sulfuric acid, the consumption of periodate would be decreased. In order to find out which of the reactions of sulfuric acid with cellulose had the greatest effect on the periodate oxidation of cellu­ lose, the following experiment was undertaken. Four 5-g. samples of fibrous standard cellulose (numbered 1, 2, 3 and 4) were treated with 100-ml, portions of 0.5 M sulfuric acid and thermostated at 30° C. for 1, 2, 3 and 4 hours respectively. At the end of each reaction time, the sample was washed thoroughly until free from acid (ten 1liter portions of water). It was then steeped in 600 ml. of water at 60° C. for 24 hours, after which it was washed with four 1-liter por­ tions of water. The sample was pulled apart, dried in a vacuum oven at 55° C. for 4 hours, cut to 20-mesh linters, and conditioned at 6h% relative humidity for several days. -83- Hydrocellulose samples 2, 3 and 4 were oxidized for 60 hours with standard mixing in exactly the same manner as used for the oxi­ dation of cellulose acetates under these conditions. these oxidations are recorded in table XI. The data for The symbol H.C. is used to designate hydrocellulose. At the top of table XI the average consumption of periodate is listed for standard cellulose. If the periodate consumption of the hydrocelluloses are compared with that of standard cellulose, it can be seen that the periodate consumption decreases slightly for hydro­ celluloses according to increasing length of time that these hydro­ celluloses were steeped in 0.5 M sulfuric acid. The sulfate content for the hydrocelluloses was not determined, as it was only desired to find out which of the two reactions of sulfuric acid with cellulose was predominant. The preparation of hydrocelluloses involved the use of a total of 0.0308 mole of cellulose units and 0.05 mole of sul­ furic acid. The mole ratio would be 1 : 1.623, as compared with 1 s 0.238 used in the preparation of cellulose acetates. The concentra­ tion of sulfuric acid used in hydrocellulose preparations was 0.5 M as compared to 0.056 M used in the preparation of cellulose acetates. The reaction time was generally longer in the case of the cellulose acetates, but the sulfuric acid available for esterification was less, and the acid was more dilute. The data for the oxidation of the hydrocelluloses may show the effect of sulfuric acid on the consumption of periodate, but this data, due to the differences in conditions, cannot be accurately compared with that for the periodate consumption -84- TABLE XI PERIODATE OXIDATION OF SULFURIC ACID HYDROCELLULOSE Sample No. ^ * S. c« Reaction Time, hrs. Temp. °C. Sample Wgt., , g« Arsenite Titer, ml. N of Ars enite 00 12 24 36 48 60 hio4 Consumed meq./ g. hio4 Consumed Moles per Unit Mol. Wgt. 0.0000 0.8080 1.3290 1.7645 2.1714 2.5460 0.0000 0.0655 0.1077 0.1430 0.1760 0.2063 H.C. No. 2 00 12 24 36 48 60 29.85 0.9985 19.675 18.00 16.85 15.85 14.85 13.90 0.01506 0.0000 0.761 1.263 1.690 2.098 2.470 0.0000 0.0616 0.1022 0.137 0.170 0.200 H.C. No. 2 00 12 24 36 48 60 29.85 0.7304 19.675 18.55 17.60 16.90 16.80 15.4-8 0.01506 0.000 0.700 1.260 1.680 2.070 2.450 0.0000 0.0567 0.1020 0.1360 0.1680 0.1985 H.C. No. 3 00 12 24 36 48 60 29.85 1.0031 19.675 17.95 16.75 15.68 14.70 13.85 0.01506 0.000 0.778 1.300 1.752 2.150 2.483 0.0000 0.0630 0.1055 0.1420 0.1740 0.2010 H.C. No. 4 00 12 24 36 48 60 29.85 1.002 6 19.675 18.15 16.88 15.85 14.95 14.10 0.01506 0.000 0.688 1.239 1.679 2.043 2.372 0.0000 0.0557 0.1004 0.1360 0.1655 0.1920 * Standard Cellulose values are taken from X. -85- of cellulose acetates. Further periodate oxidations of sulfuric acid hydrocelluloses was left as a subject for future research. Oxidation of Acetone Soluble Cellulose Acetates. c tn The work of Purves and coworkers * on the lead tetra-acetate oxidation of acetone soluble cellulose acetates shows that a very small number of glycol groups may exist in these highly acetylated celluloses. It was thought that it would be interesting to determine whether or not the existence of such small amounts of glycol groups could be confirmed by the periodate oxidation method. Acetone solu­ ble cellulose acetates of comparable specifications were oxidized in the following experiment. Four Eastman Kodak commercial cellulose acetate samples were labeled as shown below. No. 4644: Acetone soluble, low viscosity, high acetyl. No. 4650: Acetone soluble, low-medium viscosity, high acetyl. No. 4655: Acetone soluble, high-medium viscosity, low acetyl. No. 2314: Cellulose triacetate. The samples were in the form of a coarse powder. Each sample was inspected for uniformity and cut in the Wiley mill to a 20-mesh or finer powder and conditioned for several days fet 65$ relative humidity. Sample No. 4644- was found to contain 3.5$ moisture. Each sample was analyzed for per cent combined acetic acid as described on page . The results are listed in table XII. Powdered acetone soluble cellulose acetate was very difficult to wet with water and floated on top of the aqueous oxidizing medium. -86- TABLE XII ANALYSIS OF COMMERCIAL CELLULOSE ACETATES Cellulose Acetate No. EK4644 Analysis for per cent Combined Acetic Acid ml. of N of HOAc Sample st. st. vrgt., % EOH KOH e« 0.5856 0.6253 0.5301 0.0000 36.65 38.40 34.15 9.40 0.1934 0.5406 0.5621 0.5779 0.0000 34.88 36.00 36.80 9.40 0.1934 0.6004 0.5086 0.5478 0.0000 37.20 33.05 34.75 9.40 0.1934 0.5369 0.5327 0.5311 0.0000 36.90 36.10 37.10 9.40 37.90 2.356 261.0 38.55 2.421 263,e 37.80 2.348 260.6 41.72 2.767 278J2 54.00 54.20 53.90 av. 54.03 EK2314 Unit Mol. Wgt. 54.90 55.10 55.30 av. 55.10 EK4655 Degree of Substitu­ tion. 54.00 54.10 54.40 av. 54.16 EK4650 Percent Acetyl 59.65 58.40 60.80 ■0.1934 av. 59.62 -87- The problem of wetting and dispersion was solved by the use of a very small concentration of detergent, A 0,01% aqueous solution of ’’Duponol Mn dry” was the smallest optimum concentration that would wet the powder in 10 minutes. A volume of this 0,01% solution of detergent, when mixed with an equal volume of approximately 0.06 II HIO^, did not consume any periodic acid when the mixture was allowed to stand for 72 hours at room temperature. The three acetone soluble cellulose acetates and the cellulose triacetate were oxidized in the standard mixing apparatus for 72 hours. A sample of standard cellulose and a blank made up with the detergent solution were oxidized at the same time. The method of oxidation was the same as used for the 60 hr. oxidation of cellulose and cellulose acetates with standard mixing, except that only three 10-ml. samples were removed from the oxidation reaction, and these at 24 hour inter­ vals. Over the 72 hour period the concentration of periodate in the reaction flasks containing the four commercial cellulose acetates re­ mained constant. This shows that no oxidation of the cellulose ace­ tates occurred. The suspended solid particles appeared to be swollen and translucent at the end of the oxidation period, and there was no reason to believe that the oxidizing solution had not been able to penetrate throughout each particle. The proportions of materials present during the lead tetra-acetate 5 17 oxidation of cellulose acetate as performed by Purves * were: 0.01 mole of cellulose acetate (unit mol. wgt. 259.e) dissolved in 100 ml. of glacial acetic acid that was 0.025 N with respect to lead tetra-acetate. -68- The reaction temperature was 25° C. After a reaction time of about 80 hours, a break in the oxidation rate curve was interpreted to mean that true glycol oxidation had ended, and at this point the consumption of oxidant corresponded to 0,01 mole of glycol per unit molecular weight. This is the same thing as saying that there is 1 glycol group in 100 glucose residues, all of the other glycol groups having been blocked by acetyl groups. The question remains whether this small amount (0,01 mole) of glycol could be detected over a 72 hour period of oxidation with 0,03 N periodic acid. The consumption of periodate by cellulose over a period of 72 hours under standard mixing conditions and at 30° C, was only ,227 mole per unit molecular weight, and therefore only about 22,1% of the total glycol content had been oxidized. On the other hand, a commercial acetate is far more amorphous in structure than native cellulose and therefore should be penetrated and oxidized more rapidly. If the value of 0,01 mole of glycol per unit molecular weight is converted to a figure that would represent a difference between blank and sample titrations with 0,015 IT arsenite (such as observed during the course of a periodate oxidation), this value would corres­ pond to 0,17 ml. Although a titration difference of this order could be observed, it is doubtful that a smaller difference, say ,05 ml, (that would correspond to only partial oxidation of a cellulose ace­ tate), could be determined by the method of analysis used. In conclusion, it was considered that the work of Purves had not been confirmed. IV DISCUSSION THE NATURE OF FIBROUS CELLULOSE ACETATES OF LOW ACETYL CONTENT The Acetylation Reaction. The reaction of cellulose -with acetic anhydride to form cellulose triacetate may he expressed asj C6H70 2 (0 H )3 + 3(CH3 C 0 )2 0 _____ ^ C6H70 2 (0C0CH3 ) 3 + 3CHgC00H This equation shows that at least three moles of acetic anhydride are required to completely esterify the hydroxyl groups of cellulose. The acetylation of cellulose proceeds very slowly unless a so-called catalyst is added to the reaction mixture. Sulfuric acid was used as the catalyst in the preparation of the cellulose acetates studied in this investigation. Some of the ideas that have been expressed concern­ ing the specific role of sulfuric acid in the acetylation reaction should be reviewed briefly. Franchimont'*'® considered that sulfuric acid reacted with acetic anhydride to form acetyl sulfuric acid. This acid is the acetylating reagent and reacts with cellulose to regenerate the sulfuric acids ,P ch 3 - ci 0 CH3-C^ 0 c 6h 9 o4 oh Ost 47 + HgS04 — + CHgCOOH + CH3C00S020H H0S02 0C0CH3 ______ C6H9 0 4 0C0CH3 +H2 S 0 4 studied the esterification of cellulose with sulfuric acid and the formation of cellulose sulfoacetates. His results indicate that the reaction of sulfuric acid with cellulose precedes the forma­ tion of acetyl cellulose, and that the sulfo group is replaced by the acetyl group. -90- Heuser^ considers that: ’’the chief function of the sulfuric acid is to aid in the swelling of the cellulesic material and to de­ grade it”. Degradation during the acetylation may reduce the degree of polymerization of the original material as much as 90%, depending on the severity and duration of the acid treatment. The effect of the moisture content of the cellulosic material on the rate of acetylation was studied by El8d and Schmidt-Bielenberg.^0 They measured the time of reaction necessary to obtain a product of a certain degree of substitution using celluloses that contained: 6.7%, and 24.4% water. 0.1%, The following figures are taken from the rate of acetylation curves of Elod: Combined Acetic acid in the product 10% 20% 30% 40% Reaction time (hrs.) for cellulose: 0.1% water B 16.5 25.5 36 4 9.5 15.5 24 0" 0” 6.7% wat er 24.4% water 1.2 2. The experimental conditions used by Elod were those for the preparation of fibrous cellulose acetates. Figure 1 on page 2 7, shows that there is no appreciable difference between the rate of acetylation at 45° C. of cellulose containing 0.1% water, and that for the acetylation of cellulose containing less than 1% water at 27 to 29° C., (the later contitions were used for the preparation of the cellulose acetate samples that were used in this study). -91- The Fibrous Structure of Cellulose and the Acetylation Reaction, Cotbon fibers are single plant cells that vary in length from 1 to 5 cm. and from 12 to 42 y- in thickness. Each cell is composed of a primary wall about 0.5 y thick, a thicker secondary wall, and a main central core called the lumen. Haw cotton contains about 0.5 to 1.0$ each of pectin and waxes that, are considered to be a part of the primary cell wall, and may be removed by boiling with dilute alkali. The proof of the chemical structure of cellulose and the arrange­ ment of the long cellulose chains in crystalline and amorphous patterns in the so-called micellar structure, are the results of almost 100 years of study. X-ray studies have shown conclusively that there is a considerable degree of order in fine structure of cotton fibers. In recent years the study of the proportion of crystalline and amorphous regions in cellulose has received considerable attention^**^. The mi­ celles that are made up of cellulose chains may have a certain perio49 50 dicity of structure * , in that some of the glycosidic linkages are apparently more vulnerable to attack by even the most mild acidic or oxidative treatment. With the concept of the micellar and plant cell structures of cellulose well in mind, one can proceed to discuss the reactions of the hydroxyl groups of cellulose. Any reagent that will react with these hydroxyl groups must either proceed gradually inward from the surface of the fiber, or promote some rapid transformation in the micel­ lar structure so that all of the hydroxyl groups can be reached. These are two extremes in reaction types, the first being called "topochemical” , -92- and the second "permutiodal". The nitration of cellulose is a good example of a permutiodal reaction, wherein the nitration of cellulose is complete within a short time after the reaction begins. The acetyl- 27 ation reaction proceeds quite slowly and in a topochemical fashion • Hess2® has referred to the acetylation reaction as a ’’micellarheterogeneous” reaction. The heterogeneous nature of the reaction may vary according to the procedure of acetylation. pA Herzog and LondbergG* have shown that an outer layer of cellulose acetate may be dissolved away from a partially acetylated fiber with nitrobenzene, to leave a residue of unchanged cellulose. Kanamaru®^ has made photomicrographs of partially acetylated cellulose fibers (acetic anhydride and benzene). These photomicrographs show that the acetylation reaction starts at several fairly equally spaced points along the fiber, and bands of acetylated material start to form around these points of attack, and finally the bands expand gradually until all of the fiber is acetylated. Kanamaru concludes that the reaction at the surface of the fiber pro­ ceeds to the triacetate stage. The attack of the acetylating reagent at definite intervals along the fiber has been compared to that observed by certain wood destroying fungi , and both proceed along certain planes referred to as planes of hydrolysis. The course of acetylation may be followed by X-ray techniques if the fibrous structure of cellulose is retained2^. Hess and Trogus2® found that the X-ray pattern of the original cellulosic material did not change upon acetylation until a degree of acetylation, corresponding to 43% combined acetic acid, had been reached. -93- At this point the pattern for triacetyl cellulose became noticeable and increased until at 53/$ combined acetic acid, the X-ray pattern of cellulose was no longer present. From this it might be surmised that the acetylation reaction may not proceed entirely to the triacetate stage until a fairly high degree of acetylation is attained. The course of acetylation and deacetyiation of cellulose fibers has been studied thoroughly by Vermaas and Hermans 63 . These investiga­ tors, being aware of the physical aspects of the cellulose structure, have attempted to prepare homogeneous partially acetylated cellulose filaments. According to the modern concept a swollen cellulose gel con­ sists essentially of a molecular network structure. "In the course of a substitution reaction like acetylation, the degree of substitution at any given moment and at a given locus, will be a function of the degree of accessibility at that locus". Standard filaments of regenerated cellulose were pre-swollen in an ether-acetic anhydride acetylating mixture without catalyst. At the beginning of the reaction the catalyst (HgSO^) was in a very small concentration which was increased after the reaction was started. A series of partially acetylated cellulose fibers prepared in this manner was found to be microscopically homo­ geneous, but of increasing thickness and decreasing birefringence with increasing duration of the acetylation. Vermaas and Hermans concluded from these experiments that acetylation took place from the beginning of the reaction in both the intererystalline (amorphous) regions as well as in the crystallites. The rate of conversion was considered to be greater in the amorphous regions of the gel. X-ray investigation showed that a gradual increase in the spacing in the cellulose crystal lattice occurred due to the substitution of the large ester groups /which force the glucose anhydride rings apart. Other dimensions in the lattice did not change. The foregoing discussion may be best summarized with the statement of Lorand and Georgias ’’There is evidence to support the view that cellulose reactions do not follow a single pattern, and that the type varies, depending upon the reaction partner, its concentration, the re­ action medium, temperature, etc. It is quite probable that all the re­ action models, worked out for micro-heterogeneous systems, such as surface reactions, topochemical macro-heterogeneous reactions, permutoid or quasi-homogeneous reactions, as well as the previously mentioned micellar heterogeneous reactions, may apply to cellulose in one case or another, depending on circumstances.” In conclusion, some of the factors that must be kept in mind when discussing fibrous cellulose acetates of low acetyl content, such as those prepared in this investigation, may be listed as follows: (1) The laboratory variables such as the moisture content of cellu­ lose, the concentration of acetylating agent and catalyst, the time of reaction, and the temperature of reaction, must be con­ trolled. (2) Sulfuric acid present in the acetylating mixture reacts with cellulose to form sulfoacetates and to degrade the cellulosic chain. ■95 (3) The acetylation of cellulose under these conditions is hetero­ geneous and will proceed faster in certain areas of the fiber. (4) Different degrees of acetylation will be present at different l points of attack, depending on the accessibility of the hydroxyl groups at these points. (5) A large proportion of the cellulose remains unchanged during the acetylation reaction, and there is a high probability that the proportion of such unchanged cellulose will be less in the amor­ phous regions of the cotton fiber. (6) Any reaction to which cellulose may be exposed will yield re­ sults that are only average values and do not truly express the stoichiometric relationships resulting from the reaction. -96- PERIODATE OXIDATION OF CELLULOSE The Mechanism of Periodate Oxidation The work of Criegee6 has been accepted by many authorities as a satisfactory explanation of the oxidative cleavage of 1,2-glycols by such reagents as lead tetraacetate and periodic acid. It is postulated that these reagents condense with the glycol group to form a cyclic intermediate that is unstable and decomposes by cleavage of the 2,3carbon-carbon bond in the ring. In the case of periodic acid, the inter­ mediate decomposes as follows: I -C-CK | io4h3 -- > 2 /C=0 + H20 + HI03 I n* Heidt and Purves have listed the requirements that must be met by any reagent that might conceivably perform this type of oxydation, (1) The central atom of the oxidant must have a diameter of about 2,5 to 3,0 5 which is large enough to bridge the space between the hydroxyl groups in a 1,2-glycol, (2) The central atom must be able to coordinate with at least two hydroxyl groups in addition to the groups already attached to it. (3) The valence of the central atom must be 2 units (and not 1 or 3) higher than the valence of its next lower stable state, (4) The oxidant should have- a standard E0 oxidation potential in the neighborhood of about -1,7 v. with respect to the next lower stable state. -97- Reagents that fulfill the preceding requirements arej HIO4 , UaBi03, anc* hydrated unstable Ag Xx«L . Pb(0C0CH3 )4 , The last two have been tested and found active towards glycol cleavage. Commonly used oxidiz­ ing reagents such as HaOCl, KgCrgOy, KMn04 , and HWO3, lack one or two of the above requirements. As opposed to the cyclic mechanism of Criegee, Waters 64 has present­ ed a free radical mechanism that also explains the oxidative cleavage of 1,2-glycols. The mechanism of Waters involves an auto-oxidation reaction that is catalyzed by the oxidizing agent, and leads to a series of chain reactions that end with glycol cleavage. The following mechanism was proposed for the lead tetraacetate oxidation of glycolss Pb(0C0CH3)4 H R-C-OH I R-C-OH I H 2 1? R-C-O* | R-0-OH H (a) + Pb(OCOCH3)2 *0COCH ' > ♦ H R-C-O* | R-C-OH t H (a) ? > R-C-OH | R-C-OH H 2 *0C0CH3 + ?. + HOCOCH, . R-C-O* I R-0-0* H (b) H -- *- 2 R-C=0 The acetate free radicals react with the glycol group to form new free radicals (a), which may collide to form the biradical (b). The biradi­ cal (b) is stabilized more easily by the symmetrical fission of a C-C bond than by further action involving a-O-H bond. Waters postulates that in the case of periodic acid a corresponding free radical reaction -98- may be started by ’’atomic hydrogen abstraction by an 1=0 bond”. The biradical is formed in one step: H /// | -C-0:H 1:0: / :6!.'i:0: H -o-o;h |" s' H -C-Q------ ► -o-o. A \ +■ H90 2 + HIO, H This mechanism would be in agreement with the requirements of space and configuration as outlined by Heidt and Punres 23 , but does not involve actual condensation or chelation of the oxidizing agent in a cyclic structure such as proposed by Criegee. Periodate Oxidation of Cellulose Jackson and Hudson30 (1936) were the first investigators to apply the so-called Malaprade reaction to cellulose chemistry. The work of Hudson and Jackson has been extended considerably and at the present time the periodate oxidation of cellulose is finding application as an analytical method in certain phases of cellulose chemistry ca Crt * . The literature cited for Jackson and Hudson, Jayme3^, and Pacsu30, deals mainly with the theoretical reaction and with the identification of oxi­ dation products, and does not furnish data that might be useful in a study of the rate of periodate oxidation of cellulose under different conditions. g Davidson has made an extensive study of this oxidation reaction and the properties of periodate oxycelluloses. He found that the ideal reaction of periodic acid with cellulose did not stop at the point where -99- 1 anhydroglucose unit had been oxidized by 1 mole of HI04 , as shown in the reaction; CH-OH c— 0 ?/H / 0 \oH / CHpOH a— 0 \ / c h /'A H HI04 \ / CH 0 c— 6 H E/S ► / OH \ / c=o c»o <. ^ H The actual oxidation, and especially during the later stages of the reaction, gave rise to measurable quantities of carbon dioxide, formic acid, and formaldehyde. Some of these products might be expected as a result of the normal oxidation of end groups along the cellulose chain. The periodate oxidation of an end groups may be written: HC=0 (l) HCOOH HC-OH (2) HCOOH HC-OH I GliO-CH (3) (4) HC-OH (5) HC=0 I Gl-O-C-H I HC=0 I HI04 I HC-OH H (6) H2C=0 The symbol Gl is used to designate an anhydroglucose residue. The forma­ tion of formic acid results from the oxidation of carbons Ho, 1 and 2, and the formation of formaldehyde, from carbon No, 6, Further oxidation of formic acid may account for the carbon dioxide. In a typical oxidation experiment, Davidson oxidized 1 g, of cellu­ lose with 100 ml, of a solution of 0,2 N HI04 . After about 350 hours, the cellulose had consumed 1 mole of HIO^ per unit molecular weight, and had produced COg, HCOOH, and HgC£Q, in the amount of: and 0.035 moles per unit molecular weight respectively. 0.096, 0.226, These figures correspond to a formation of these products to the extend ofs COg, 1 mole per 10 anhydroglucose units HCOOH, 1 mole per 4 or 5 anhydroglucose units, and HgC=0, 1 mole per 33 anhydroglucose units. It is obvious from these figures that these products did not arise solely from the end groups, but must have been formed as a result of side reactions. This particular experiment represents an extreme in side reactions. If periodate oxidations are carried out at pH values of between 2 and 5, and if the concentration of the oxidant is reduced to about .05 N, the production of COg, HCOOH, and HgC-0 is decreased considerably. Harris 53 and Purves 20 have studied periodate oxidation of cellulose and have determined the amounts of dialdehyde present in samples that were oxidized for various time periods. Purves found that a properly prepared oxycellulose contained at least 90% of the dialdehyde (as deter­ mined by the ainitrophenylhydrazone method) that would be expected from the periodate consumption. Purves further considers that the reaction (for starch) is only selective between the pH values of 2 and 5 and at temperatures below 20° C. "Oxidation of starch or cellulose with aqueous periodate takes place in a heterogeneous system and it is possi­ ble that some of the oxidant is dissipated in secondary reactions at -101- the surfaces while the interiors of the granules or fibers contain un­ changed starch or cellulose. The initial rate of periodate oxidation of cellulose is much faster than that observed for the over-all reaction. . Mark^ and Timell®® have shown that this initial rapid oxidation corresponds to the attack of periodate on the amorphous portions of cellulose. If a curve is drawn to show the rate of oxidation of cellulose over a considerable period of time, (see figure 9), it can be seen that in the early stages of oxida­ tion the observed rate of reaction is the net result of two reactions, one being the fast oxidation of the amorphous, and the other, the slower oxidation of the crystalline regions of cellulose, both of which take place at the same time. O Davidson has studied the effect of periodate oxidation on the X-ray diagram of filter paper cellulose. As the degree of oxidation increased the X-ray pattern of the original cellulose became more and more diffuse. This indicates that no new crystalline structure is pro­ duced, and that the original crystalline order is gradually destroyed as oxidation proceeds. The conclusion would be that the crystalline areas are gradually opened up during the course of periodate oxidation. Figure 7 shows the reaction rate curves for 8 periodate oxidations of cotton cellulose. Curves IV and VI were plotted from data obtained in this investigation; all other curves were plotted from data obtained from the references cited. An oxidation period of seven hours would cover mostly the rapid periodate oxidation of the amorphous regions of cellulose. -102- KIO4 pH 4.2 /KIO4 / pH 4 VI (p «) 0.03 N / HICU IV ( M 5) 0.03 N /HIO4 pH 2 0.03 0.02 0.089 N KIO 4 of RIO 4 , moles per unit molecular weight 0.04 VIII (*°) VII 0.05 N 0.1 N III ( ss) 0.021 N Consumption KIO 4 pH 4.6 0.01 0.01 N HIO 4 pH 2 Time, hours Figure 7 Periodate Oxida tio n of Cellulose Under Various Conditions. A study of figure 7 would lead to the following observations: (1) The rate of oxidation increases with an increase in the con­ centration of the oxidizing agent, as would be expected* (2) The effect of swelling of the fiber on the rate of oxidation is indicated very clearly by the difference between curve VIII and all others. Curve VIII was obtained by Purves 20 from data resulting from the oxidation of cotton linters that had been pre-swollen, and Purves claimed that after such swelling the internal surfaces, directly available to any aqueous solution, contained from 10 to 20% of all the hydroxyl groups present in the fiber. Therefore, in any oxidation study, one must take into account the effects of pre-swelling treatments, and also the swelling effects resulting from the presence of certain re­ agents in the oxidizing mixture. (3) The effect of vigorous stirring is shown by a comparison of curve VI with curve IV. All of the conditions were the same in both experiments, the only difference being that the re­ action mixture was vigorously stirred in the case of curve VI. It is clearly seen that the oxidation rate is much faster when the reaction mixture is stirred vigorously, and this could be considered as evidence that the availability of the glycol groups is increased. Vigorous stirring might also increase the diffusion gradient of the oxidant in the fiber, which in turn would cause a greater probability of reaction between periodate and glycol. -104- (4) Apparently the rate of oxidation is not affected appreciably by differences in pH between 2 and 5. Factors that Influence the Reproducibility of Results and the Rate of Oxidation of Cellulosic Materials. If conclusions are to be drawn from any series of periodate oxida­ tions of cellulosic materials, certain experimental conditions and fact­ ors must be kept in mind: (1) The reaction flask apparatus should be free from oxidizable impurities or impurities that might catalyze the decomposition of periodic acid. After glassware has been cleaned it should be filled or treated with dilute HI04 solution and allowed to stand for at least 24 hours. After oxidation experiments the flasks should be rinsed with distilled water and dried in an inverted position, (2) Oxidation experiments should be carried out on samples that have been standardized as to physical form, (a) The state of division of the materials should be approximately uniform, (b) The moisture content of the materials should be controlled so as to permit accurate weighing. (c) Any swelling treatment prior to oxidation should be standardized. (3) Extremes in reaction conditions should be avoided: (a) The concentration of the oxidant should not be much more than 0,1 K and not less than 0.02 N. -105- (b) If a measure of the true rate of reaction is to be obtained, the volume of oxidant per gram of material should be selected so that the reaction period can be completed without decreasing the concentration of the periodic acid to an extent that would effect the reaction rate. (c) The reaction media should have a pH between 2 and 5. A reaction mixture that is too acid yields an exces­ sive amount of by-products caused by side reactions. A mixture that is alkaline causes degradative re­ arrangements of the periodate oxycellulose. (d) All oxidation experiments should be performed at the same controlled temperature. Temperatures between 20 and 30° C. are suggested. (e) The selection of a reaction time should be governed by the nature of the cellulosic material and the con­ centration of the oxidant to be used. Short reaction periods show almost stoichiometric selective oxidation according to the ideal reaction, whereas, prolonged oxidation periods may involve side reactions. The amorphous areas of cellulose will be attacked rapidly during the early stages of the reaction, and as the oxidation of these areas becomes complete, the re­ action rate will be that corresponding to the oxidation of the crystalline areas of the cellulosic material. -106- (f) The contents of the reaction flask should be mixed at a moderate standard rate that permits the main­ tenance of a homogeneous suspension of the cellu­ losic material in the oxidizing solution, (4) The method of analysis for excess periodate should be stand­ ardized, and blank determinations of the concentration of the periodic acid should be made at least once every 24 hours during the course of a oxidation reaction. -107- PERIODATE OXIDATION OF CELLULOSE ACETATES OF LOW ACETYL CONTENT The Rate of Oxidation of Cellulose and Cellulose Acetates. Figures 5 and 6 show the reaction rate curves for a series of cellu­ lose acetate oxidations as plotted from data recorded in table X. All of the oxidation data resulted from experiments where the possible factors that might influence the rate of oxidation were kept standard. The con­ sumption of periodate, in moles per unit molecular weight, listed in table X is the average consumption for six experiments in the case of standard cellulose, and the average of at least two experiments in which each cellulose acetate was oxidized. The rate of the oxidation reaction has been shown by plotting perio­ date consumption against time. The progress of the reaction may also be shown by plotting the periodate consumption against the amount of free hydroxyl present in the cellulose acetates. The amount of oxidation of any cellulose derivative will be a function of the total number of free hydroxyl groups in that particular derivative. The percent of free hydroxyl in each cellulose acetate may be calculated from the equation: Percent of free hydroxyl K (S - D.S.) 100 3 where D.S. is the degree of substitution (table II), and 3 is the maxi­ mum degree of substitution for an anhydroglucose unit. The conversion of D.S, to percent free hydroxyl for cellulose acetates No. 17 to 25 is recorded in table XIII. -108- TABLE XIII CONVERSION OF DEGREE OF SUBSTITUTION TO PERCENT FREE HYDROXYL CAc No. 17 16 19 20 21 22 23 24 25 Percent Combined HOAc 5.13 8.23 12.09 14.14 3.40 5.31 6.60 6.65 9.94 Degree of Substitution Percent Free Hydroxyl 0.1436 0.2360 0.3560 0.4235 0.0940 0.1490 0.1668 0.2490 0.2880 95.21 92.13 88.18 85.87 96.87 95.03 93.77 91.70 90.40 -109- Figure 8 is a graph of the consumption of periodate plotted against the percent free hydroxyl for various cellulose acetates of low acetyl content. The values for cellulose are recorded at 100?£ free hydroxyl. The reaction rate curve for cellulose is shown in figure 9. This curve may be analyzed in terms of the oxidation of amorphous and crys­ talline regions of cellulose19*60. If the straight-line portion of the curve is extrapolated to zero time and displaced downward to the origin (line C) this straight line may be considered to represent the rate of oxidation of crystalline cellulose. If the differences between the act­ ual observed curve and this line for any series of time values are plotted on this same coordinated system, the result will be curve A, which represents the oxidation of amorphous cellulose. Curve A showB that the oxidation of amorphous cellulose is for all practical purposes com­ plete after a 24— hour oxidation period under the oxidation conditions used. The basic problem that must be solved at this point may be suggest­ ed by the question: How many hours would be required to complete the periodate oxidation of cellulose (the oxidation of all glycol groups in cellulose, i.e., one glycol group per unit mol. wgt.)? A few assumptions must be made: i (1) That the ideal reaction holds true during the 60 hour period under these very mild conditions of oxidation. (2) That the degree of periodate consumption at 24 hours (0.1077 moles per unit mol. wgt.) represents the end of the oxidation of amorphous cellulose. -110- 0„l5 0.10 Consumption of HIO., moles per unit molecular weight 0 o20 0.05 Oxidation Period o + ^ • a 60 48 36 24 12 hours hours hours hours hours 0 85 90 95 Percent free hydroxyl Figure 8 Periodate Oxidation of Cellulose and Cellulose Acetates at 30°C., with Standard Mixing. - I ! i- 100 3 o.i5 r-i 4-3 •H (U CU o 0.10 s o M