Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Cathodic halogen and a systematic study of the preparation of alkyl chlorides from the corresponding… Streight, Harvey Richard Lyle 1929

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


831-ubc_1929_a8_s7_c2.pdf [ 40.9MB ]
JSON: 831-1.0059409.json
JSON-LD: 831-1.0059409-ld.json
RDF/XML (Pretty): 831-1.0059409-rdf.xml
RDF/JSON: 831-1.0059409-rdf.json
Turtle: 831-1.0059409-turtle.txt
N-Triples: 831-1.0059409-rdf-ntriples.txt
Original Record: 831-1.0059409-source.json
Full Text

Full Text

U . B . C . U 8 R A R Y ) CAT ACC. MO. "f CATHODIC HALOGEN and A SYSTEMATIC STUDY OF THE PREPARATION of ALKYL CHLORIDES: FROM THE CORRESPONDING ALCOHOLs by Harvey Richard Lyle Streight A Thesis submitted for the Degree of -HAOiiLff. 'Ji' jr^IYiO in the Department of CHEMISTRY THE UNIVERSITY OF BRITISH COLUMBIA April, TABLE OF CONTENTS. General Introduction Page 1. i Part 1. Theories of Valence. The Electronic Conception of Valence. 2. The Shared Electron Bond Theory or The Electron Displacement Theory. 6. (c) Summary of Theortical Discussion. 14. Part 2. Positive Halogen. fa) Hypohalous Acids & Positive Halogen. 1&. (bl Positive Halogen Compounds. 21. (c) Electrical Nature of Cyanogen Halides. 26. Part 3. Previous Electrical Work. 2?. Part 4. Experimental Work on Cyanogen Halides. 'a) Preparation of Compounds. 31* b^i Titrations for Positive Halogen. 32. _c) Electrolysis. 33. Part 3. Electrolysis of Halogenyl Amides. Properties etc. of Halogenyl Amides. 37. Electrolysis of 8 Halogenyl Amides. 39. Discussion of RSsults of Parts A & 44. Summary of Parts 4 & 5. 43. Part 6. A Systematic Study of the Preparation of Alkyl Chlorides from the Corresponding Alcohols. (a) Introduction & Discussion of Former Methods. 4 b) Experimental Work. 32. c) Discussion of results. 37. (d) Relative Costs of Various Methods. &1. Summary. 62. Bibliography. 64. CATHODIC HALOGEN AND A SYSTEMATIC STUDY OF THE PREPARATION OF ALKYL CHLORIDES GENERAL INTRODUCTION. There are several electronic conceptions of the struc-ture of molecules, many of which have had to be radically modified or entirely discarded. The theory of chemical union as is now generally accepted is based on Bohr's conception of the atom under conditions defined by the quantum theory. Part I discusses the two important theories of valence - the electronic or dualistic theory and the electron bond theory. It gives explanations for the polarity of elements and radicals and in the union of two atoms or groups suggests which plays the part of a positive radical and which plays the part of a negative radical. (1) According to one of these theories put forward by Fry if the valence of an element is 'n' it may function in (n+1) different ways. Thus univalent halogen may function as negative halogen (Hal) and positive halogen (Hal)*. Part ill of this paper is concerned with the electrolysis of the cyanogen halides and halogenyl amides in order to see if it were possible to isolate halogen on the cathode. According to the products of hydrolysis iodocyanogen should give cathodic iodine while chlor and brom cyanogen should give anodic chlorine and bromine respectively. The chemical reactions of (1) Electronic Conception of Valence; Longmans Green & Co.London 1921. - 2 -the halogenyl amides show them to contain a so-called "positive halogen atom". The products of electrolysis of the cyanogen halides also has been determined in a number of solvents. Part 'V deals with a systematic study of the preparat-ion of alkyl chlorides, both by methods already in the litera-ture and also by two new modifications, namely - thionyl chloride, using benzene as a solvent, and phosphorus penta chloride, using zinc chloride as a catalyst. A review of the literature showed excellent methods for the preparation of alkyl bromides and iodides with good yields. In the case of the chlorides, however, several methods are mentioned for a few definite compounds and often yields have not been recorded. The preparation of 28 alkyl chlorides with hydrochloric acid, phosphorus tri- and penta-chlorides, and thionyl chloride, with various catalysts and solvents has been conducted in this laboratory. Yields and relative costs are indicated in Part VI.* PART I. The Electronic Conception of Valence. The electronic conception of valence is a modern con-ception of the old dualistic theory of Berzelius. In fact, the theory is an extension of the widely accepted dualistic theory of polar inorganic substances to non polar organic substances. (1) Abegg stated that chemical action was due to a transference of (1) Abegg - Z. anorg. Chem. 39, 33O, 1904. - 3 -electrons from one atom to another and tried to apply the full theory of electrochemical dualism to organic compounds. Sir (1) J. J. Thomson regarded valency as a tube of force emanating from a valency electron, either ending on the positive charge with-in the atom or on that of another atom. So in a molecule of hydrogen, for every tube of force sent out from the electrons of one atom, the latter must be the recipient of a second tube of force sent out from a second atom. Thus the atom of hydrogen must be divalent and possesses one positive and one negative valence which Thomson represented by arrows. H H (2) This idea was expanded by Falk and Nelson to all organic compounds. H. S. Fry extended the theory regarding the electrostatic forces holding atoms together in a molecule as set up by a difference in the electrical properties of the two combined atoms. Thus a negative element or radical attracts a positive element or radical. An atom A may change its sign by the loss or gain of electrons. Thus if A is positive and it gains an electron it becomes neutral, by the gain of another electron it becomes negative and vice versa. + e — ^ x x + e — x ' This is similar to anodic oxidation and cathodic reduction of electrochemistry. 1) J.J. Thomson - Phil. Mag. (6) 27, 737, (l$l4). 2) J. Amer.Chem.Soc.,(a) 32 1637 1910; (b) 33, 1810, 1913; (c) 37, 1732 1913. - 4 -Also, all univalent atoms or radicals can function both positively as well as negatively and vice versa. However they have a tendency to behave as either + or -. Thus in HC1, hydrochloric acid, the hydrogen is considered positive and the chlorine negative; in NaH, sodium hydride, sodium is positive and hydrogen negative; in R0C1, hypochlorous acid, hydroxy! is negative and chlorine positive. This theory gives a plausible explanation of the electrostatic forces holding diatomic gases in combination. H* 'H , C1+ '01 . In diatomic molecules of different elements it is possible to have two different electronic compounds with the same chem-ical formula. Thus AY may be A*—"Y or —+Y. Clark and (1) crozier have shown that there are two "ortho nitro toluenes" possessing different physical properties. One is relatively unstable and with the application of heat passes over to the more stable form. The equilibrium set up between these "electronic isomers" or "electromers" is almost entirely in one direction - so we recognize only one isomer. The A*—Y** isomer may go over to the X"—Y* isomer by an oxidation - reduc-tion process. Each step represents a definite state of oxid-ation of the atom. (1) Trans. Royal Soc. of Canada. 3rd. series, Vol.XIX.137(1923) See also survey by Clark and Carter, ibid Vol. XXI, 323 (192?) on "The Replaceability of Nitro Groups from the Nucleus of Various Aromatic Compounds." one hundred and ten in number. - 7 -X+-Y" - (Y+e)+X+ = Y+X = (Y*+e)+X = Y++X' = Y+- X" . This has been applied to the carbon atom by Fry (3 states of (1) oxidation) and to the nitrogen atom by Jones. The latter explained the relations between various organic nitrogen com-pounds and the Beckmann rearrangement occurring in them by this theory. To convert the theory into a working hypothesis it is necessary to assign to the atom in a particular compound its characteristic sign. This is done (^ a) by considering the reactions of electrolysis - which determines the charge on the atoms; ^b) by hydrolysis, in which the hydroxy! group will attach itself to the positive constituent and the hydrogen to the negative constituent; and (c) by considering the process of oxidation-reduction, which may be represented by the loss or gain of electrons. These terms will be discussed later with reference to the cyanogen halides. The theory correlates in a simple fashion the several ways in which atoms and radicals react, positively, negatively, and amphoterically, in strict conformity with the actual chem-ical behavior of the molecule - as amply illustrated by the well established chemical reactions actually being dealt with as ionization and electrolysis, hydrolysis, and oxidation-reduction processes. In fact, practically all chemical phenomena may be classified under these types. It does away (1) Amer. Chem. Journ. 30, 414. - 6 -with any ambiguity attached to the structure of the atom itself. It gives valuable information regarding probable chemical properties of compounds and suggests methods of preparation. It has also been used to explain the properties of straight chain compounds, the Beckmann rearrangement and is still more fruitful in the case of aromatic compounds (which lies outside the scope of this paper) in explaining the Crum Brown Gibson rule, ortho and para vs. meta substitution, direct substitution (a) (2) (3) etc. See Fry Higgins and Crocker. Shared Electron Bond Theory or Electron Displacement Theory. The theory is based on Bohr's structure of the atom and valence as exhibited in the periodic table. The atom consists of a positive nucleus surrounded by planetary electrons arranged in shells or energy levels. The number of electrons in the outer shell is closely connected with the valence of the element and the outer shell tends to take on or give up electrons until it reaches the structure of the nearest rare gas, thus accounting for positive and negative valence. One difference is to be noted between Lewis-Langmuir and Rutherford-Bohr hypothesis. In the former, the electrons are arranged in electronic shells while in the latter, they are arranged in electronic orbits. This leads to a certain amount of confusion. (11 Electronic Conception of Valence; Longmans,Green & Co. 1$21. (2) Higgins - (a) Science 33, 674 1?22; (b)J.Amer.Chem.Soc. 44, 1607, 1922. (3) Crocker - J. Amer. Chem. Soc., 44, 1618, 1$22. The theory as proposed by Lewis and afterwards modified (2) by Langmuir is based on the theoretical considerations that two atoms may conform to the rule of 8 or octet rule, not only by the transfer of electrons from one atom to another, but also by the sharing of one or more pairs of valence electrons. These electrons are held in common by the two atoms and are considered to belong to the outer shell of both atoms. Previous to this, (3) (4) (3) (6) Stark, Bohr, Kossel, and Parsons arrived at the conclusion that two atoms are held together by groups of electrons held in common by the two atoms. It is of interest to note that of some hundred thousand known compounds all but a few contain an even number of such electrons. Thus it is an almost universal rule that the number of outside valence electrons in a molecule is a multiple of two. uertain metallic vapors at high temperatures, monatomic hydro-gen and halogens, chlorine dioxide, nitric and nitrous oxides, and triphenyl methyl are exceptions. These "odd molecules" show a great tendency to combine with other like or unlike odd molecules to form molecules with an even number of electrons as IC1, (NOgJg etc. The chemical bond thus consists of two electrons coupled together and lying between the two atomic centres and held (1^  J. Am. uhem.Soc., 38, 762, 1^16. (2) J. Am. uhem.Soc., 38,222, 1?16; (b) 41, 868, I343. (3) "Prinzipien du Atomdynamik, Die Elektrizitat im chemischen Atom" Rurzel Leipzig I9I3. (4) Phil. Mag (6) 26, 837. 3) Ann. der Phys. 4$, 22^ (l?l6) 6) Smithsonian institute Publ. 63, fJo. 11, I913. jointly in shells of two atoms. When two atoms of hydrogen join to form a diatomic molecule each furnishes one electron of the pair which constitutes the bond. Representing each valence by a dot we have the following graphic formula for hydrogen H : H. When hydrogen,with its one electron, unites with chlorine with its seven electrons they produce together a * . molecule represented by R : CI : Similarly for methane and carbon tetrachloride. H : CI: R : C : H : ci : C : CI : H : CL : Negative elements or radicals are defined by this theory, as those which draw towards themselves the electron pairs which constitute the outer shells of all neighboring atoms, whereas electropositive groups are those which repel or attract these electrons to a less extent. For example, in the hydrogen molecule the electron pair is symmetrically placed between the two atoms. In sodium hydride,the electron pair is closer to the hydrogen than the sodium making the hydrogen relatively negative; while in hydrogen chloride the bond is shifted towards the chlorine leaving the hydrogen relatively positive and in the presence of a polar solvent the chlorine assumes full possession of the bonding pair and we have complete ionization. We can express this as follows:-H : H , Na : H , H ;C1: ,(H)*+ ( :C1: ) - ? -A symmetrical molecule like H^ or may become polar-ized in one direction or the other, - for example, iodine on heating to a high temperature breaks in such a way to sever the bonding pair and forms two uncharged iodine atoms. : I : I : : 1 * . I : .. .. In liquid iodine some molecules break apart in another manner -the bonding pair remains intact but remains the exclusive property of one atom forming iodide and iodous ions. This (1) experiment was suggested by Sir J.J. Thomson as a proof for positive iodine. : I : I : : I : + 1 : iodide iodous ion. ion. In the union of two electronegative elements as ul-Cl, C1-0H, CI-NH2, HO-OH, HgN-OH, H^N-NH^, etc., Lewis does not decide which atom is positive and which is negative, as does the dualistic theory. He says that both atoms are in a more positive condition than when combined with electropositive elements and that the state of tension can best be relieved by breaking the bond between the two atoms and attaching to each atom a more naturally positive element . In order to give direction to the bond and to give it a (2) point of attack on the valence ring Sommerfield's idea of elliptic orbits is applied. As the orbits have direction Lewis "Corpuscular Theory of Matter" page I30 (1907). (2) "Atomban und Spektrallinien". Vieweg & Sohn. Braunschweig. 1^22. - 10 places his electrons in fixed positions corresponding to the extremities of the ellipses and in the case of a tetra valent atom, like carbon, the paired ellipses extend from the atom to the four corners of a tetrahedron. Sommerfield's spectroscopic work gives direct evidence that the pairing of electrons is the same sort of configuration of two magnets, of such character as to eliminate mutually their magnetic moment. This would account for the unsaturation of "odd molecule" compounds. There is one point of the theory which needs further discussion - the physical interpretation of the bonding pair. In the interest of clarity it has been proposed to confine the term "polar valence", associated with polar bond, to the designation of valences which are actually ionic because of the complete transfer of an electron. The "semi-polar valences" of oxygen in SO radical, the amine oxides, etc., are 1/2 non polar or covalence. It has also been suggested that non polar valences which have an electrical moment and whieh permit atoms held together by a covalence to separate readily in ionic form "potentially polar". The distinction between polar and non polar bonds is sharp and distinct while there can be all (1) (2) Wrurcs: - -r,!nr<3-!!r DV^ Tz7T 390 (f) Harrison, Kenyon & Phillips, J. Chem. Soc., 12 (1 (g) Noyes Proc. Nat. Acad. Sci. 13, 377, (1927) 926) - 11 -degrees of "potentially polar" valences due to the shifting of the bonding pair. (1) Noyes, in a recent paper, considered this a logical development of the positive and negative valences proposed (2) (3) partly erroneously and vaguely in 1901 and 1904 by himself (4) (5) (6) and advocates of Professor Stieglitz, Falk and Nelson, Fry, (7) L.W. Jones and others. It also furnished an almost complete reconciliation with the Lewis-Langmuir theory. (8) (9) (10) Knorr, Pauling, and others have given a physical inter-pretation for the distinction between polar and non polar bonds, based on the orbits of the bonding pair. A polar bond is due to a pair of electrons rotating about the kernel of some negative atom or to a deficiency of one electron in a positive atom. This is equivalent to the electronic picture of Jones, Stieglitz, and Fry. A non polar bond is due to a pair of electrons (1) Noyes - J. Amer. Chem. Soc., 30, 2902. (1928). (2) Noyes & Lyon-J. Am. Chem. Soc., 23, 463, (1901). (3) Noyes - Chem. News. 90, 228, (1904). 4) Stieglitz - J. AM. Chem. Soc., 23, 796, (1901). 3) Falk and Nelson ibid (a) 32,1637,(1910); (b) 33,1810,(1913); (c) 37, 1732,(1913). (6) Fry - Zeit.physik. Chem. 76,383,398,391(1911). J. Amer. ahem. Soc., 34, 664, (1912). 7) L.W. Jones - ibid 36, 1268 (1914). 8) Knorr - Zeit. anorg. allgem. chem. 129,109 (I923). 9) Pauling - J. Amer. Chem. Soc., 48, II32, (1926). (10) (a) Noyes - ibid 39, 879 (1917). b) Sidgwick - Trans. Faraday Soc., 19,469 (I923). c) Paulus - J. Amer. ^hem. Soc., 48, II32 (1926). d) Grimm & Sommerfeld - Zeit. Physik., 36,32 (1926). el Glocker - J. Amer. Uhem. Soc., 48, 202 (1926). f) Noyes - Proc. Nat. Acad. Sci., 13, 377 (1927). - 12 -describing a double elliptical orbit including the kernels of the two combined atoms. In this way, if the enclosing orbit is small, the bond is very stable and unreactive but, if the enclosing orbit is elongated, the bond is more or less unstable and has a tendency to change to a polar bond - the resulting polarity depending upon which of the two atoms retains the bonding pair. This explanation serves to more strongly differentiate between polar and non polar compounds and explains the "potentially polar" compounds. I will give a few illustrations showing how the bonding pair is shifted by the substituent introduced into an organic compound. In the saturated compound, ethane - a non-polar compound - the bonding pair of electrons is halfway between the carbon atoms, if one pow introduces a negative group, like chlorine into the molecule the bonding pair of electrons of the carbon and chlorine would be displaced towards the chlorine atom and the electron pair between the two carbon atoms would be shifted away from the carbon atom attached to the halogen. That this is the directioh taken by the displacement is shown by considering ethylene; in this compound the electrons forming the two bonding pairs are somewhat midway between the two carbon atoms. If the chlorine derivative of this compound be consid-ered - vinyl chloride - the position of the extra bonding pair of electrons between the two carbon atoms can be readily ascertained by treating the compound with a halogen acid. Upon treating vinyl chloride with hydrochloric acid ethylidene - 13 -chloride (unsymmetrical dichloroethane) is formed, showing that the bonding electrons shifted away from the carbon atoms to which the chlorine is attached and caused the other carbon atom to assume a relatively electro-negative condition as the positive hydrogen of the hydrochloric acid became attached to this carbon atom. This conception has been used by R. harasch, Stieglitz, and others to explain many organic reactions. Lewis put forward his theory strictly as a hypothesis without any experimental data, other than that already described in the literature, to prove the same. Re showed that the effect of substituents upon the strength of organic acids could be accounted for on the basis of electron displacement and he pointed out that the electronic displacement, due to any substituent at one end of a chain of atoms, extends throughout (1) the entire chain. Latimer and Rodebush made use of this idea in accounting for the ionization strengths of some of the acids (2) of phosphorus. A series of papers by Lucas and co-workers have shown that the electron displacement theory in carbon compounds explains the mechanism of addition of hydrogen halides to ethylene homologs, to allene homologues, and the rearrangement of the alkyl bromides much better than the alternate polarity of (3) the carbon atoms in open chain compounds as put forward by cuy (1) J. Amer. uhem. Soc., 42, 141?, (1920). (2) J. Amer. Chem. Soc. (a) 46, 2473 (1924); (b) 47, 1439(1923); (c) 47, 1462 (1?23); (d) 30, 1711(1928). (3) Cuy - J. Amer. Uhem. Soc., 42, 303 (1920). ' - 14 -(1) and others. For example, on the basis of alternate polarized carbon atoms the reaction between 2 pentene and hydrogen bromide should yield 2 bromopentane chiefly, while on the theory of electronic displacement it should yield 3 bromopentane as the main product. Experimentally 78% of 3 bromopentane and 22% of 2 bromopentane was obtained. Alcoholic potash solution in 2 bromobutane resulted in formation of cis and trans 2 butene and not 1 butene as suggested by Cuy's alternate polarity theory. (2) Other cases are given. Lucas also applied Lewis' theory to (3) Pauling's model of the benzene molecule, which ^ies outside the scope of this paper. To summarize the theoretical discussion - the Lewis-Langmuir theory offers a picture of the state of a molecule, based on the physicist's idea of the atom, before reaction takes place and the probable effect of an added substituent, and whether it will cause the compound to become more or less polar. If the compound is polar we have ionization and we can treat the molecule by the ordinary electronic formulae. The electron bond theory differs from the electronic theories in that the atoms are not bound by shared electrons and any atom may change its polarity by the gain or loss of electrons, it (1) (a) Lapworth - J. Chem. Soc. 121, 416 (1920). (b) Kermack & Robinson ibid 121, 427 (1920). (c) Stieglitz J. Amer. Chem. Soc. 44, I3OI (1922). (d) Hanke & Koessler ibid 40, 1726 (1918). (2) J. Amer. Chem. Soc., 48, 1827, 1926. (3) ibid 48, H32, 1926. is easy to realize that with such accomodating electrons most reactions can be satisfactorily explained. In a recent paper Fry urges the acceptance of the single electronic formula with these words: "The system of electronic valence notation is designed to indicate and correlate the actual chemical behavior of the atoms and the radicals of the molecule as displayed under specific experimental conditions, rather than with the imposition upon structural formulae, the suppositions based on the unsolved problems of the constitution of the atom and the disposition of the hypothetical electron shells or the elusive electron orbits of its valence electron. He also explicitly states that the Lewis-±sohr doublet theory of the union of atoms is accepted so far as the nature of the bond is concerned but he insists on polarity resulting from the definite approach of doublets to certain atoms - in other words he regards all compounds as polar and uses plus and minus signs to designate the polarity. THESE PLUS AND MINUS SIGNS MUST NOT BE INTERPRETED AS IF IONIC POLARITY WERE REPRESENTED, as was done by physical chemists, including Lewis. Both theories show the possibility of an unstable isomer or electromer - Lewis' theory showing the reason for such an isomer (position of bonding pair of electrons), the electronic theory giving ionic formulae for the same to explain their (1) Fry. uhemical Rgviews, 4, 337, 1^28. - 16 -chemical reactions. In the future as we are dealing with polar compounds, we can represent the molecule most simply by the usual electronic symbols. PART II. hypohalous Acids and Positive halogen. The halogen atoms in certain inorganic and organic compounds have the property of being replaced on hydrolysis by hydrogen, and appearing combined with the hydroxy! group as hypohalous acid, in these compounds it is claimed that the halogen is electropositive or positive; in fact the criterion for positive halogen is that it shall be eliminated on hydroly-sis as hypohalous acid. The reasons for believing that the hypohalous acids contain positive halogen are numerous. The suggested electronic formula for hypochlorous acid is H* '0' *C1. It is very reactive and unstable, is an oxidizing agent and in acid solution it immediately undergoes an oxidation-reduction change to hydrochloric and chloric acids. 2 H0C1 + H0C1 = 2 HC1 + HCIO^ 2(+l) +1 2(-l) +5 (small figures represent state of oxidation of chlorine.) It is stable at low temperatures for a considerable time in alkaline solution as there is no negative ion other than hydroxyl for the positive chlorine to oxidize. - 17 -In a very careful and extended experimental study published in 1899, Jakowkin demonstrated that chlorine is hydrolyzed by water to hydrochloric and hypo-chlorous acids and he showed it was a reversible reaction. Clp + HOH = H0C1 + HC1. He determined the partition coefficient of chlorine between carbon tetrachloride and pure water, and between carbon tetra-chloride and solutions of strong and weak acids, chlorides, and hypo-chlorous acid. He also determined the conductivity of pure hydrochloric acid formed by the hydrolysis and showed that the hypochlorous acid is ionized scarcely at all under these con-ditions. Finally he showed that hypochlorous acid volatilized more readily than hydrochloric acid for a dilute solution of chlorine in water and that the difference in rates reaches a maximum at 90°. The simplest explanation of Jakowkin's results is that on hydrolysis chlorine separates into positive and negative ions and that these combine with the ions of water. ul* *C1 + H* *0H = R+Cl' + CI* "OH. (2) This conclusioh was stated by Stieglitz. As the reaction is ionic for water and hydrochloric acid and as the reaction is reversible, it must be ionic for chlorine and hypochlorous acid. (3) Pelouze showed that a concentrated solution of hypochlor-11 Z. physik. Chem. 2p—513 (1893). 2) Stieglitz - J. Amer. Chem. Soc., 23, 796, (1901). 3) Ann. Chem. phys. (3), 7, 126, (I843). - 18 -ous acid gives on treating with calcium hitrate, chlorine mon-oxide, ClpO which he called "hypochlorous acid" in accordance (1) with the nomenclature of the day. Noyes and Wilson confirmed his work by other experiments and showed that chlorine monoxide and not hypochlorous acid escapes from a solution of acid when a current of air is passed through, indicating positive chlorine ions in such a solution and showing the amphoteric ionization of the acid - that while hydrogen and hypochlorite ions pre-dominate enough positive chlorine ions are present so that they combine with the hypochlorite ions to form chlorine monoxide. ul* + CIO* = cl + '0* *C1. It is interesting to note that according to the Lewis-Langmuir hypothesis the union between oxygen and chlorine in hypochlorous acid is "non polar" and in terms of that hypothesis only positive ions as hydrogen, lithium, glucinium, and carbon of the positive ion of triphenylmethyl have less than 8 electrons in the outer shell - while Noyes' and Wilson's work demonstrates the existence of positive chlorine having only 6 electrons in the outer shell. If we regard the decomposition of hypochlorous acid as a dehydrating reaction (2 HC10 = Cl^O + H^O) we have only one explanation possible in accordance with the electron bond hypothesis as in the following diagram: H.t :0:ci: H:0:!ci: .. !.. (l) J. Amer. Chem. Soc.; 44, I63O, (1?22). - -This means that water is formed of a positive hydrogen ion from one molecule of hypochlorous acid and a negative hydroxyl ion from another molecule of hypochlorous acid. Assuming the pair of electrons is stable, the two electrons between oxygen and chlorine must go with the oxygen of the negative hydroxy! ion leaving only 6 electrons with a chlorine atom. The latter would therefore be positive. There is very little difference between such an explanation and that of actual ionization giving rise to s positive chlorine ion. Hypochlorous acid is also a chlorinating agent - in fact we attribute the chlorinating power of chlorine itself to the formation of hypochlorous acid in solution. This brings up the question of halogenation. Bromination in aqueous solution is usually considered to be due to the presence of hypobromous acid since it is much faster than bromination in aqueous media; also aqueous bromination is markedly slower in presence of hyd-robromic acid, which would reduce the concentration of hypo-bromous acid; and finally, the reaction of ethylene bonds with bromine water frequently gives considerable amounts of bromo-(1) hydrin compounds. Cofman on studying iodination of phenols concluded that hypoiodous acid is the active iodinating agent and that it dissociates into positive iodine which brings about (2) iodination. Baines came to the same conclusion regarding hypobromous acid on bromination of phenols. Re showed that (1) J. Chem. Soc. 113, 1040 (1919). (2) J. Chem. Soc.,121, 2810 (1922). - 20 -hypobromous acid and phenol give tribromphenol and emphasized the fact that in the presence of concentrated hydrobromic acid, (1) bromine is only slowly absorbed by phenol. Soper and Smith concluded that hypochlorous acid is not the cause of chlori-nation as chlorine reacts more quickly with phenol than does (2) hypochlorous acid. Francis recently studied the action of bromine on aniline, toluidines, etc., and concluded that hypobromous acid was not the halogenating agent as the brominat-ion of meta nitrophenol by means of free bromine water was 1000 times as fast as the bromination with hypobromous acid from which free bromine was removed by silver sulphate. This ratio was found to hold for other organic compounds and even for inorganic substances as potassium iodide. Re concluded that the active agent coiisists of bromine atoms with a positive charge. A review of the whole question of halogenation shows that the active agent is the positive halogen ion formed either by the amphoteric dissociation of hypohalous acids or by the dissoci-ation of diatomic halogen molecule. Just as hypochlorous acid contains a positive chlorine ion we can also conclude that hypobromous and hypoiodous acids also contain positive bromine and iodine respectively. In a discussing halogen compounds, if they give hypohalous acids on hydrolysis, they can be considered to contain positive halogen. (11 J. Chem. Soc. 12$, 1382 (1$>26). (2) J. Amer. Chem. Soc., 47, 234O ( W 3 ) - 21 -Positive Halogen Compounds. Considerable chemical evidence has been presented in the literature for the existence of compounds containing positive halogen. In general, organic halogen compounds in which the halogen has replaced the hydrogen atom in the amino, imino, and hydroxy groups exhibit extraordinary activity. Unlike the halogen of the ordinary types of alkyl and acyl halide linkages which are reduced with more or less difficulty the halogen in the above mentioned types shows actual oxidizing power in the presence of such reducing agents as aqueous solutions of ferrous salts or metallic iodides. They are also unstable and many decompose with explosive violence. (1 ) As far back as 1863 Schutzenberger prepared alkyl hypo-= 0 chlorites, (R - C _ Q ^ ), and he suggested that the chlorine in these compounds played the part of an electrophsitive atom. (2) In 1892, Selivanow pointed out the hypochlorous character of chlorine in monochloro and dichloroamines (RN HCl^ RNCl^) and in nitrogen trichloride. These aliphatic halogen derivatives of primary and secondary amines hydrolyze to hypohalous acids and have been called "halogenyl" compounds by Selivanow. (3) Wm. Lauder Jones compared the action of bromine upon tertiary amines to the action of the same on lead dibromide. The bromides obtained were treated with potassium hydroxide and (1) Ann,- 120, II3. (21 (a) Ber.-23 36I7 (18?2); (h) Ber. 26, 423 (1893). (3) Amer. Chem. Jour. 30, 414. - 22 -altogether different products were formed although from theoretical considerations similar products would be expected. The only satisfactory explanation for this interesting case was shown by Jones to be due to a difference in the polarity of one bromine atom in the two compounds. The chlorimido ketones, prepared by Stieglitz and (1) Petersen,were recognized as derivatives of positive chlorine, according to their method of preparation and especially accord-ing to their behavior. They are prepared either by treating the corresponding hydrogen compound with hypohalous acid, or with an alkaline solution of halogen at low temperatures. The compounds are very unstable, sensitive to light, and their solutions have strong oxidizing power. (2) W. L. Jones has written two articles, reviewing the work done on polarity up to 1^13* (3) In 1^20 Howell and Noyes showed that iodine in mono and di iodoacetylene was positive. For example diiodoacetylene was prepared by the^action of iodine chloride on calcium or copper carbide. 2 IC1 + CaC^ = 1C=CI + CaCl^ It reacted with ammonia to form nitrogen triiodide and oxidized a potassium iodide solution to free iodine. An interesting (11 J. Amer. Chem. Soc. - 273 1?14. (2) Amer. Chem. Journ.- (a) 48, 1?12; (b) 30, 426. (3) J. Amer. Chem. Soc. 42, 1^20. * 23 * explanation of the poisonous properties, irritating effects, and disagreeable odor of the compounds was given - due to positive iodine instead of "dyad carbon atom" according to Nef's methylene view. The structures of thedilodocompound are I-C^^I and 1^C=C<. These compounds readily add on bromine, hydrobromic acid, etc., and decompose again to givediiodoacetylene. They may add on to the + iodine atom, which has a tendency to become trivalent instead of<*0 bond. Inorganic compounds which are similarly unstable and possess strong oxidizing powers are the halogenated ammonias, (1) NCI? and NI?.NE-. The work of W. A. Noyes and co-workers on 3 3 3 nitrogen trichloride and Chattaway and co-workers on nitrogen triiodide and others on the same compounds led to the assumption that halogen is combined as a positive atom. In his first paper Noyes tried to prepare the electromer of nitrogen trichloride, containing negative chlorine which would hydrolyze normally as the chlorides of non metallic elements do. N+ + +G1 ***" + 3 HOH = HONO + HOH + HC1. 3 Experiments were carried out with nitrosyl chloride and (1) (a) Noyes & Lyon. J. Amer. Uherp. Soc.- 22, 460 (1^01) (b) Noyes, ibid, 33,767 (W3); (c) ibid 39,903(1917); (d) ibid.42, 2167 (1920); (e) ibid, 42, 2173 (1920). (2) (a) Chattaway, J; Chem. Soc.- 70, I372 (1396); (b) Amer. Chem. J.- 23,^63 (1900); (c) ibid 24,1^9 (1900); (d) ibid, 24, 318 (1900); (e) ibid, 24,331 (1900); (f) ibid, 24, 342 (1900). - 24 -phosphorus pentachloride and of thirty gas mixtures analyzed twelve gave evidence of very small quantities of this isomer. (.12 - .22 milligm. mols in 70 c.c.). In his later work JNoyes doubted his early results and gave much evidence to show that nitrogen tri chloride containedpositive chlorine as its preparation - action of anhydrous or aqueous ammonia and chlorine, action of ammonium chloride and chlorine or hypo-chlorite or hypochlorous acid. (1) It is interesting to give Noyes explanation for the formation of nitrogen trichloride from ammonia and chlorine. He first assumed an addition of chlorine to ammonia (or if ammonium chloride was used addition took place to ammonia given + — off on its decomposition) forming NH=G1 CI which.dissociates into chloroamine N^^l and HC1. The chloroamine would add another molecule of ul*-"ul giving dichlorammonium chloride NH^Cl^^Cl" and a repetition of the dissociation and addition of chlorine and a third dissociation and addition gave 1NC1 . The 3 action of hypochlorous acid on ammonium salts is similar -addition of ammonium hydroxide gives chloramnonium hydroxide NH al*0H**, exactly as it adds to ethylene to give ethylene 3 chlorohydrin CH CI. CH OH. This dissociates to chloramine 3 2 NH^cl and water and repetition would give mCl^. The above hypothesis was checked experimentally by detecting AiH^ Cl etc. in the formation of j^ Cl^ . (1) Noyes & Lyon. J. Amer. Chem. Soc. 42, 2173 (1920) - 27 -Chattaway prepared NI^NH^ by the action of ammonia on iodine, ammonium hydroxide and iodine chloride, and ammonia and an alkaline solution of potassium hypoiodite. The last two methods point to positive iodine. Chattaway and Orton showed quantitatively that, in the third reaction, the triiodide is formed at the expense of the potassium hypoiodite in the solution. Regarding its properties, he reports it to be explosive, very sensitive to light, giving ammonia and hypo-is iodous acid on hydrolysis and'a strong oxidizing agent. A somewhat similar compound azoic iodide JN-1 was shown (i) 3 to contain positive iodine. Other inorganic compounds containing positive halogen are iodine chloride and iodine bromide. They hydrolyze to give hypohalous acids, are very reactive, and undergo an (2) oxidation-reduction change to iodic acid and free iodine. These compounds will be mentioned later when discussing their hydrolysis. (3) (4) Jones and Werner and i^ ef have shown that one halogen atom in carbon tetrachloride, tetrabromide, and tetraiodide was different from the other three as on hydrolysis with alkalies they gave chloroform, bromoform, and iodoform respectively. Also there are many aromatic compounds which contain (1) Abegg. "HancLbuch der anorg. cb.em," Atr,3. Bd.3. 5.213. (2) Orton, Blackmann - J. Chem. Soc. 77, 830. 1^00. (3) J. Amer. Chem. Soc. 40, 1237 1919-(4) Ann. 308, I73. - 26 -positive halogens, but this lies outside the field of this work. ELECTRICAL NATURE OF CYANOGEN HALIDES. The compounds chlor, brom, and iodo cyanogen have been purposely left to the last in order to give the reader a general idea of the type reactions and properties of positive halogen compounds. (1 ) Nef observed that iodocyanogen on boiling with potash formed potassium iodate whilst the corresponding brom and chlor compounds yielded chloride and bromide respectively, as shown in the following equations. (a) 1+CN" + 2K*0H"* K*CN' + K*lfl+ + H+llH^ 3K0I = KIO^ + 2KI 3 (b) Br"CN+ + 2K*0H" = R*"0"CN+ + K+Br' + (c) Cl'CN* + 2K+0H** = R.*"0'CN+ + K*Cl" 4 The reactions show that the iodo compound contains positive iodine while the brom and chlor compounds contain negative halogen. (2) uhattaway and Stevens showed the positive nature of iodine in the reactions 3HC1* = 31+Cl' + NH_ 3 3 NI + 3HCN* = 31+CN" + NH 3 (3) 3 Chattaway-Wadmore also concluded iodine to be positive by its (1) Ann.- 287, 316. 1893. (21 Chattaway. J. Chem. Soc. 24, 331 (1^00). (3) J. Chem. Soc. - 81, 1?3, (1?02). - 27 -action on the hydrogen halide acids as, 1"CN* + H+Cl' = R*CN" + I+Cl'. With hydrobromic acid it formed iodine bromide (I+Br") and with hydriodic acid formed free iodine (1*1**). It is also a strong oxidizing agent oxidizing sulphites to sulphates, and sulphides to free sulphur. It might be well at this point to draw attention to a number of mistakes which appeared in the paper of chattaway and Wadmore. (l) In discussing the action of cyanogen iodide on a solution of sulphurous acid he found .2<?60 gms. of cyanogen iodide oxidized 38.7 c.c. N/lO J^ a^ SO^ . Now 1000 c.c. of 1 N. J^ a^ SO^  soln. contain 63 gms. 38.7 c.c. of 1/10 N. Na-S0^ " " 38.7 x 63 ^ 3 10,000 and 38.7 x 63 Na S0_ acted on .2% gms, ICR. 10,000 2 3 Then 1 Mol. Wgt. of Na SO would act on 126 x .2% x 10,000 2 3 3H.7 X 63 = 133 gms. I33 gms. is the molecular weight of ION. Thus one molecule of Na SO reacts with one molecule of ICN. 2 3 His equation 2CNI + H^SO^ + H^O = 2CNH + H^SO^ + Correct " ICN + H-SO^ + H-0 = CNH + H-SO, + HI. 2 3 2 <-4 He found iodine to be liberated when sulphurous acid was slowly added to cyanogen iodide but that it ceased with the addition of more acid. The first equation probably takes place - 28 -but at equilibrium the oxidation-reduction goes further with the liberation of hydriodic acid. (2) His equation 2CNI * H^S = 2HCN + + S. obtained from expt., .2896 gms of ICN oxidized 37.9 c.c. of N/lO H^S similarly should read: ICN + H^S = HCN + HI + S. He also noted that when hydrogen sulphide was slowly added, free iodine was first liberated but the reaction ceased when more hydrogen sulphide was added. These reactions show that oxidation proceeds further than indicated in his equations. (Iodine is reduced from 0 valence in to - valence of one in HI.) (1) Grignard and Bellet in studying the action of the cyanqgen halides on organic magnesium compounds found that iodocyanogen gave the iodide of the organic radical while chlorvanogen gave little or no alkyl or acyl chloride but alkyl or acyl nitrile. (1) I*UN" + R"Mg**Ral' = RI + + (CN)*Mg^(Hal)' (2) C1**CN* + R*Mg'*Hal" = + (Cl)*Mg**(Hal)' Assuming magnesium to have a positive valence of two then in (1) cyanogen is negative and hence iodine is positive while in (2) chlorine is negative and cyanogen is positive. (jotppare equation (1) with the normal action of alcohol on any organic magnesium halide. (1) (a) Compt. rend. 133, 44, (15*12); (b) ibid,158,437,(1914). - 29 -K'OH+ + Rl*Mg++Hal" = (RO)"Mg++Hal" + R'^H*. Bromcyanogen was found to react in most cases like iodocyanogen. Compare our experimental titration results of brom and iodo cyanogen. 189.2 and 92% positive halogen. PART III. PREVIOUS ELECTRICAL WORK In spite of the frequent references in the literature to positive halogen, only three references could be found to investigators who have reported halogen as deposited on the cathode in electrolysis and their conclusions have been disputed. In I827 de la Rive electrolyzed iodine-bromide in an aqueous solution containing starch and claimed from the colour changes that iodine was obtained at the cathode and bromine at (2) the anode, in I834 Faraday passed an electric current through water free iodine chloride and molten silver chloride and reported chlorine at the positive and iodine or silver at the negative pole. Solly, in I830, on the electrolysis of dry liquid iodine chloride with platinum electrodes, reported that the "anode was corroded but clear and that the cathode was encrusted with a black matter very much like iodine in appear-(4) ance". In 1901, Stieglitz expressed his belief in the existence (1) De la Rive - Phil. Hag. (2) 3l I43, (21 Faraday - Ostwalds Klassiker 86, 43, (1834). (3) Solly - Phil. Mag. (3) 8, 401, (I836). (4) Stieglitz - J. Amer. Chem. Soc., 23, 796, (1901). - 30 -of positive halogen ions and his intention of searching for them by the electrolysis of the hypochlorites. As he has reported nothing further, we may assume that his efforts were unsuccessful. The results of these early investigators were called in (1) (2) question by Bruner and Bekier and Bruner and Galecki. The former electrolyzed dry molten iodine bromide and iodine chloride between silver Electrodes, and obtained bromine and chlorine at the anode but found no definite evidence of iodine at the cathode. However, they conclude that the separation of iodine at the cathode is exceedingly plausible. Bruner and Calecki studied the electrolysis of iodine bromide and iodine chloride between silver electrodes in sulphur dioxide and in nitro benzene as solvents. Their experiments were conducted without the exclusion of moisture and all the halogen was liberated at the anode. At this point it is well to remember that anodic hydrogen has been obtained by the electrolysis of molten (3) hydrides. In 1920 Moer found that lithium hydride on electrol-(4) ysis gave anodic hydrogen and in 1922 Bardwell obtained anodic hydrogen in electrolyzing a solution of pure dry calcium hydride in a fused mixture of potassium and lithium chlorides. (1) Bruner & Bekier, Zeit. Elec. Chem. 18, 36$, (1912). (2) Bruner & Galecki, Zeit. Phys. Chem. 84, 313, (I9I3). (3) Z. Anorg. Chem. 113, 179. (1920). (4) J. Amer. Chem. Soc. 44, 2499 (1922). - 31 -Iron electrodes were used. No gas was given off at the cathode while hydrogen was given off at the anode in amount correspond-ing to that calculated by Faraday's Law. PART IV. EXPERIMENTAL PART The object of the present investigation was to isolate and recognize the products of electrolysis of the cyanogen halides, employing various solvents, in the hope of obtaining definite proof of halogen being deposited on the cathode. Preparation of Compounds. Chlorcyanogen was prepared according to Jennings and (1) Scott. A drying tube containing lime was inserted after the calcium chloride tube to remove traces of hydrogen chloride which induced polmerization. After drying, the chlorcyanogen was redistilled directly into the cell, which was immersed in an ice bath. Bromcyanogen was prepared according to Grignard and (2) Crouzier. It was recovered by distillation, the whole operat-ion being carried on out of doors. After several redistillat-ions bromcyanogen kept without the formation of the para compound. M.P. Iodocyanogen was also prepared according to u-rignard and crouzier with modifications. The crude product,was filtered, (1) Jennings & Scott.- J. Amer. Chem. Soc. 41, 3241, (191$). (2) Grignard and Crouzier. J. Chem. Soc. A (1) 404, (1921). - 32 -dried on a porous plate, and purified before use by sublimation. It was obtained as long silky white needles. This method was found to give a purer product, with a better yield, than the suggested method, by extracting with ether. Titrations for Positive Ralogen. (I) Selivanow has shown that compounds containing positive halogens exert per atom of halogen twice the oxidizing power ^2) as the same amount of elementary halogen. VJ. A. ^oyes records on (3) the action of nitrogen trichloride "arsenious acid; Chattaway, the action of reducing agents on nitrogen triiodide; and Aiicolet and co-workers the action of stannous chloride on positive aromatic halogens in alcohol solution. They found that one molecular weight of stannous chloride was necessary for the removal of each positive atom of halogen replaced from aromatic compounds. Preliminary measurements were made by ^icolet and (3) Sampey's method on brom and iodocyanogen to determine the percentage of positive halide present in these compounds under the conditions of the experiment. The stannous chloride solution was stored under an atmosphere of hydrogen and the titrations carried out in an atmosphere of nitrogen. An excess (1) (a) Ber. 23 3617 (18^2); (b) Ber. 26, 423. (18^3). ,2) Noyes & Lyon. J. Amer. Chem. Soc. 22, 460 (1^01). 3^) Chattaway. Amer. Chem. J. 23, 363 (15*00). ,4) J. Amer. Chem. Soc. 49, 284§, (1$27). '3) Nicolet & Sampey. - J. Amer. Chem. Soc. 4p, 2849, (1927). - 33 -of stannous chloride was added to the weighed compound and refluxed for one hour at 70°, the excess of stannous chloride being then titrated with standard iodine solution, on the ass-umption that a positive halogen requires one mol. of stannous chloride per atomic weight of positive halogen; the bromcyan-ogen repeatedly titrated as if it contained 89*2% positive bromine, whilst the iodocyanogen gave results corresponding to positive iodine. These values were not increased by additional refluxing. Electrolysis. On account of the fact that iodocyanogen sublimes at a much lower temperature than its melting point, solutions of the cyanogen compounds were employed. The electrolysis of molten (1) cyanogen bromide has been carried out by Glocker, who did not analyze the products formed. All the solvents employed were most carefully dried after being thoroughly purified. Solutions 0.1 molar were used for the most part. Solvents with both high and low dielectric constants were employed. Graphite, copper, and silver electrodes were employed, the latter, being found the most satisfactory, were finally used throughout. These were made of silver foil 0.3 x 1.3 cm. in cross section welded to silver wires. The solutions were placed in a V shaped cell having a stop-cock of large bore at the bottom. A small plug of carefully dried cotton wool was placed in the bore of the (1) Glocker, Nat. Acad. Sci. 12, 322, (1526). - 34 -stop-cock to prevent diffusion of non-adhering precipitate. The electrodes were placed in each arm of the cell and were connected by mercury cups to a circuit having a milliamnieter and water resistance in series and a voltmeter in parallel across the electrodes. The water resistance gave a range of 2-80 volts. The arms of the cell were closed with tightly fitting corks. After a run the stop-cock was turned and the total precipitate in each compartment analyzed for halide, cyanide and paracyanogen. As low a voltage was employed in each case as would cause precipitation. The actual voltage depended upon the dielect ic constant of the solvent and thickness of the cotton plug. The currents registered on the milliammeter were usually so small that accurate current densities could not be calculated. Owing to the fact that all the precipitates were partially non-adhering, that the solubilities of the precipi-tates in most of the solvents were unknown, and the difficulty in depositing the last trace of cyanogen halide, qualitative methods only have been used. The precipitates were freed of organic matter by washing with acetone and then with water and were then tested for halogen and cyanide. Part of the precipitate was treated with conc. amm. hydroxide, or heated with conc. nitric acid, which dissolved the silver cyanide. This was confirmed by heating some of the precipitate on the electrode, cyanogen gas was given off and brownish red paracyanogen left b hind on the electrode. The test for paracyanogen outlined in Allen's Commercial Organic Analysis was employed. The silver halide was dissolved in potassium cyanide solution and reprecipitated with nitric acid. The resulting precipitate was tested for halogen with manganese dioxide and conc. sulphuric acid. Solute Solvent Dielectric Const.1 of solvent at Temp, indicated. Temp.of 3xpt. Anode° Deposit Cathode Deposit C1CN acetonitrile 33-82° 0° chloride cyanide BrCN acetonitrile 35.82° room bromide cyanide nitrobenzene 36.5^ room bromide cyanide methylalcohol 33.3''-' room bromide cyanide ethylene glycol 41.2 room bromide cyanide furan . . . . non conductor pyridine^ 12.421 room tarry precipitate aniline 7.3^ room bromide cyanogen gas ICN acetamide 3?.277 83°' iodide cyanide benzene 2.3IS room iodide cyanide acetonitrile 33.82° room iodode cyanide nitrobenzene 36.31-8 room iodide cyanide methylalcohol 33.3^-4 room iodide cyanide ethyl alcohol 24.814-7 room iodide cyanide nitro-methane 39.42° room iodide para-cyanogen (1) Allen's Commercial Organic Analysis, Vol. 7, 439. - 36 -1 8 pyridine 12.4 room cyanide iodide aniline 7*3 room iodide cyanogen gas and para-cyanogen. 4 thiophene 2.76 room iodide quinoline room iodide pyrrol reacted chemically furan non conductor. 1+ Laudolt Physiklischen Tabellen. 2. Cyanogen bromide breaks the pyridine ring forming a tarry mass which would not be electrolyzed. Compare KOnigs (Sidgwick-Organic chemistry of nitrogen, p. 378). 3. Cyanogen iodide started decomposing after short time on heating. 4. E.M.F. 220 volts employed. Conductance even then slight. a ) The pyridine used was dried according to MacClung with modifications and distilled, in order to remove all traces of water; the fraction coming over at 114° was mixed with freshly ignited barium oxide and shaken in an automatic shaker for two days. The solution was filtered and distilled over additional barium oxide. On addition of iodo-cyanogen to pyridine the solution had a slight yellow tinge and after applying an E.M.F. of 60 volts for five hours turned green and later red. The precipitate formed on the cathode was shown to be silver (1) "Solvents in Synthetic Organic Chemistry", p. 108. D. Van. Nostrand Co. (1923). - 37 -iodide, by liberating the violet iodine vapours which turned starch paper blue. JNo iodine was deposited on the anode. The experiment was also carried out using a small test tube for a cell, when 20 volts were sufficient to cause precipitation of silver iodide on the cathode. Concentrations much greater than 0.1 molar became viscous and black and were not investigated. On the addition of water to the pyridine solution, electrolysis gave iodine on the anode only; which might account for the results obtained by Bruner and Galecki with iodine bromide and iodine chloride in moist nitrobenzene. From the change in colour, hoted in dissolving iodocyanogen in pyridine, one might expect that an addition product had formed, which ionized to give cyanogen and complex pyridine iodide ions, the latter migrating to the cathode and depositing the iodine. Pyridine is a base and unsaturated, consequently other similar solvents, such as, aniline, thiophen, and quinoline were employed, which might readily form corresponding addition products, in each of these cases however, iodine could only be detected at the anode. PART V. ELECTROLYSIS OF HALOCENYL AMIDES. Other cases of bromine and iodine migrating to the - 38 -a ) cathode on electrolysis were found in this laboratory. On electrolysis, using silver electrodes, the halogen deposited on the anode in all solvents except acetonitrile. in this solvent the halogen from freshly purified bromyl succinipide deposited wholly on the cathode, while the halogen from iodyl succinimide and bromyl acetamide, in the same solvent, repeatedly deposited on both electrodes. At the present time other similar compounds are being investigated such as the halogenyl amides - so called by (2) Selivanow - in which a hydrogen of the amino group has been =0 replaced by a halogen. Thus bromyl acetamide is These compounds are made either by treating the corres-ponding hydrogen compound with hypohalous acid, or with an alkaline solution of halogen at low temperatures or by acting on the alkali or silver salt of the imide of a dibasic acid O ) with free halogen. Houben gives a very complete summary of these compounds, as to their individual preparations and properties. Their chemical properties confirm their relation-ship to the hypohalous acids; like them, they are very unstable and sensitive to light. Their solutions also have strong oxidizing power, - an aqueous solution of bromyl acetamide *5 1) Clark & Ball, J. Amer. 31ectro-Chem. (1228). 2) (a) Ber. 23 3617 (1892); (b) Ber. 26, 423 (1??3). 3) Houben "Die Methoden der Org. Chemie. Vol. 4, 414, (1?24). - 3? -being able to oxidize primary and secondary alcohols, potassium iodide solution, hydrogen sulphide, etc. 1. Bromyl Acetamide. ( 1 ) Bromyl acetamide was prepared by Hofmann's method and was purified by several crystallizations from absolute ether. The compound was dried in an oven at 30°^. over night, giving a product with a melting point of 104°c. Using a stop-cock cell with a cotton plug in the bore, - as in the previous work on the electrolysis of the cyanogcn halides, and silver electrodes 12 m.m. apart, with acetonitrile as a solvent, a precipitate of silver bromide was obtained on both electrodes, more forming on the anode than the cathode, in this work .1 Molar solutions were used and an e.m.f. of 110 volts, in a solution of carefully dried pyridine a stringy precipitate formed on the cathode, reaching toward the anode. This, on analysis, proved to be silver bromide. Free bromine also came to this side as shown by a deepening in the color of the cathode solution. On the anode a small crystalline precipitate formed after the cathode precipitate had formed. It gave no test for halogen and was shown not to be silver acetamide. 2. Iodyl Acetamide. (2) This compound was prepared according to Boismenu by (1) Hofmann, Ber. 13, 412, (1882). (2) Boismenu. Compt. rend. I33, 948, (1$11). - 40 -adding iodine, in small amounts, to a solution of acetamide in ethyl acetate. The color was removed by shaking with dry silver hydroxide and the process was repeated until the cal-culated amounts had been added. The iodyl acetamide was puri-fied from chloroform and had a melting point of 143°c. In an acetonitrile solution silver iodide was formed on both electrodes, using an e.m.f. of 110 volts. In pyridine, however, silver iodide formed entirely on the cathode. A small white crystalline precipitate formed on the anode after precipitation on the cathode took p^ace. This precipitate contained no halogen - further work on its identification is taking place. The cathodic precipitate when first formed was very dark in appearance but lightened in color on standing. The precipitate was also quite stringy in appearance, extending from the base of the cathode towards the anode. The original solution was a pale straw color but after running for twenty-four hours the cathode liquid became quite reddish, while the anode side became colorless. It appears that iodine is liberated at the cathode, perhaps a small amount in the free state which colors the precipitate and finally colors the cathode solution by going in solution, the rest combining with the silver electrode to form silver iodide. 3+ Iodyl Formamide. Iodyl formamide was prepared in the same way as iodyl acetamide. After washing with chloroform the product had a - 41 -melting point of On electrolysis, in a pyridine solution, using an e.m.f. of 110 volts, silver iodide formed entirely on the cathode. Precipitation took place faster than in the case of iodyl acetamide and the precipitate was lighter in color. A slight color change was noted after a twnety-four hour run, the cathode liquid turning a pale straw color, no precipitate formed on the anode. 4. Iodyl Propionamide. Iodyl propionamide was prepared in the same manner as iodyl acetamide. After washing with chloroform the product had a melting point of 128°U. On electrolysis in an acetontrile solution silver iodide was deposited on both electrodes, using a pyridine solution, silver iodide first formed on the cathode (stringy precipitate) and then, after several hours running, on the anode. At the end of a ten hour run the amounts formed on each electrode were approximately equal. Analysis of the precipitates showed each to be silver iodide. The yellow halide precipitate was crystalline in appearance and adhered firmly to the surface of the electrode. On further electroly-sis the precipitate nprtly dropped to the bottom of the cell. Iodyl n-Butyramide. Iodyl butyramide has not been described in the literature. It was prepared in this laboratory for the first time by the same method as used for iodyl acetamide. it was separated from butyramide and free iodine by dissolving these substances in a - 42 -little chloroform, iodyl butyramide was a white solid having a melting point of 97°^. it showed signs of decomposition at its melting point. An analysis was made for iodine by dissolving a weighed sample in dilute potassium hydroxide. Potassium iodide solution was added and the whole acidified with dilute hydrochloric acid. The liberated iodine was titrated with a standard sodium thiosulphate solution. Found 32.397", calculated 32.38%. iodyl butyramide immediately decomposes with water precipitating free black iodine. it was found to be insoluble in chloroform and benzene; slightly soluble in ether; and soluble in alcohol, petroleum ether ^solution turned dark red -probably due to water in technical grade of ether), acetone, and acetic ester. On electrolysis, in a pyridine solution a black precipitate formed on the cathode as for iodyl acetamide only a longer time was required for precipitation. Roth electrodes were eventually covered with silver iodide as was iodyl pro-pionamide. The cathode liquid deepened in color, turning dark red, while the anode liquid was almost clear. 6. Iodyl n-valeramide. An attempt was made to prepare this compound for the first time by the same method used for iodyl acetamide. The product obtained was a viscous, readily decomposing liquid. - 43 -7. Bromyl propionamide. Bromyl propionamide was made as bromyl acetamide. it was obtained, after recrystallizing from absolute ether, as pale straw colored needles melting at 80°U. Before purification the product melted between 68-89°^. After three hours a reddish stringy precipitate formed on the cathode, in a pyridine solution using a 110 volts emf. On further running the whole anode liquid clouded and a precipitate formed through-out the same. This precipitate was probably due to a secondary reaction and is being analyzed. 8. Uhloryl succinimide. (1) Chloryl succinimide was prepared according to Render from bleaching powder and dilute acetic acid. On electrolysis in a pyridine solution a stringy dark precipitate formed on the cathode and a very small precipitate was found on the bottom of the anode liquid. The anode precipitate was found to be insoluble in conc. ammonium hydroxide, acetone, alcohol, petroleum ether and sodium cyanide; and slightly soluble in ether. It had no definite melting point. This is the first recorded case of cathodic chlorine -previous work giving cathodic bromine and iodine. Other similar compounds are being prepared and electrolyzed at the present time. (1) Bender. Ber. 1$, 22?2, (1886). - 44 -DISCUSSION OF RESULTS OF PARTS IV AND V. From Nef's investigation of the products of hydrolysis of the cyanogen halides and Nicolet's and Sampey's quantitative method for the determination of positive halogens as applied in this laboratory, one might have expected that the iodine in iodo-cyanogen would have reacted electropositively in practically all solvents and migrated to the cathode, yet this was found to be the case in only one solvent. $he cyanogen halides, would appear to behave as "electromers" 1+CN' = I**CN* If electromic isomers, judging from their behaviour on electrolysis, the concentration of the two electromers present in solution would not appear to be directly dependent upon the dielectric constant of the solvent, as is the case of ordinary tautomers. The electrolysis of seven halogenyl amides, chlor-succinimide; bromyl acetamide and propionamide; iodyl formamide, acetamide, propionamide, and n-butyramide, with pyridine as a solvent, showed that these compounds contain electropositive halogen as predicted by their physical and chemical properties and their methods of preparation. The electrolysis of these compounds offers strong support for Fry's Electronic Theory of Valence and for the assumption of "electromers." - 43 -SUMMARY. 1. The products of electrolysis of the cyanogen halides between silver electrodes have been determined in various solvents. 2. In all cases the halide travelled to the anode with the exception of iodine from iodocyanogen in pyridine, which went to the cathode. 3. Iodocyanogen behaves as an electromer, the iodine reacting negatively, on electrolysis, in most solvents but entirely positive in a pyridine solution. 4. The products of electrolysis of seven halogenyl amides in acetonitrile and pyridine solutions have been investigated. 3. The first recorded case of cathodic chlorine was obtained by electrolyzing chloryl succinimide in a pyridine solution - the chlorine going entirely to the cathode. 6. Cathodic bromine was obtained on the electrolysis of bromyl acetamide and bromyl propionamide between silver electrodes, using pyridine a^ a solvent. 7. Cathodic iodine was obtained on the electrolysis of iodyl formamide, acetamide, propionamide, and n. butyramide between silver electrodes, using pyridine as a solvent. 8. Halogenyl formamide and acetamide precipitated the silver halide much more quickly than halogenyl n. propionamide and n. butyramide, the time increasing in the order indicated. The higher halogenyl compounds also formed silver halide on the - 46 -anode - this was noticeable in the case of iodyl n. propionamide and n. butyramide. Iodyl n - butyramide was prepared for the first time. It has a melting point of 27°c. PART VI A SYSTEMATIC STUDY OF THE PREPARATION OF ALKYL CHLORIDES FROM THE CORRESPONDING ALCOHOLS The preparation of the alkyl halides is of particular importance when one considers the number and variety of the reactions in which they enter. Satisfactory methods for the preparation of alkyl bromides and iodides have appeared in the (1) literature. Norris and his co-workers obtained excellent yields of the pure halides by heating the alcohols in an open vessel with an aqueous, constant boiling solution of hydro-bromic or hydroiodic acids. Many type reactions have appeared in the literature for the preparation of alkyl chlorides. In many instances the reactions were limited either to the preparation of the more simple members or to the preparation of a certain class of chlorides e.g. tertiary or aromatic chlorides. The yields were often recorded as small, quantitative, or were not given. This investigation had for its object a systematic study of the (1) Norris (a) Amer. Chem. Jour. 38, 627,(1907); (b) J. Amer. Chem.Soc. 38,1071,(1916); (c) ibid 42, 2023, (1920). - 47 -preparation of 28 alkyl chlorides from the corresponding alcohols by the better methods in the literature, and by two methods found in this laboratory, comparing the relative yields and costs. Aliphatic, polyhydric, unsaturated, and aromatic alcohols were used and the best reagents for chlorination determined. (1) A review of the existing methods appeared in 1907 and afterwards in 1923* Alkyl chlorides have be^n prepared by direct chlorination of paraffinic hydrocarbons; the addition of dry gaseous hydrogen chloride and hydrochloric acid to olefinic hydrocarbons and alcohols; and by the action of phosphorous oxy,- tri,- and penta-chlorides on the alcohols, lio mention was made in either review to the action of thionyl chloride on alcohols as a means of preparing the chlorides. General objections to these methods were found in the smallness of the yields or inconvenience in handling sealed tubes. Special objection to the use of phosphorous tri- and penta-chlorides was found in the violence of their action and the tendency to form phosphorous - containing by-products, as ^ (4) phosphorous-esters, or suboxides, the latter being formed by the action of phosphorous halide on the phosphoric acid first formed. (1) Dehn and Davis. ^mer. (Jhem. aoc.,29,,132H, (l^C?). (21 Organic Synthesis Vol.3,27. John Wiley & Sons, New Y§rk(1923). (3) *.7urtz. Ann. 38,72; Schiff.An,,. 102,334; 103,164; Menschutkin, Ann. 139,343; Chambox, Jahresberichte 1876,203; Barlarew,Zeit,anorg. Uhem. 99,188; Uarre, (Jompt.rend. I36, 736,(1903); & Bull.Soc.Chim.(111)27,264,(1902). (4) Kraut, Agn. 138,332; Pantier,Cor:pt.rend.76,49;Risson,ibid. 123,1032. - 48 -(1) The addition of zinc chloride to hydrochloric acid or phosphorous trichloride was found to increase the yields of the chlorides. Norris and Taylor converted alcohols into their chlorides by heating with hydrochloric acid (38%) and zinc chloride. (1 mole alcohol: 2 moles acid: 2 moles zinc chloride). Yields varied from 60% to 82%. With hydrochloric acid alone tertiary and some secondary alcohols furnished satisfactory yields of chlorides while saturated primary alcohols and other secondary alcohols would not react - although chlorides with benzyl and allyl alcohols were formed. uehn and Davis obtained the chlorides of n. propyl, iso-butyl, and iso amyl alcohols, in 24%, 83%, and 88% yields with phosphorous trichloride and zinc chloride. (3 moles alcohol: 1.6 moles phosphorous tri-chloride and 6 moles zinc chloride.) Other catalysts such as calcium and aluminium chlorides, zinc oxide, and aluminium, were tried without success. Other catalysts as - stannous, aluminium, and calcium chlorides: zinc oxide: 3% glacial acetic acid: and 3% syrupy phosphoric acid - were used in this investigation in place of zinc chloride both with hydrochloric acid and phosphorous trichloride. The addition of these com-pounds gave little or no chloride, the yields, in all cases, being reduced. (See experiments on n. propyl, n. butyl, and iso-butyl alcohols.) Norris and Taylor considered zinc chloride (1) (a) Groves, J. Chem.Soc. 12,636, (1874)- (b) Norris & Green, Amer.Chem.Jour J 26,307(1901)? (c) Norris & Taylor, J. Amer. chem.Soc. 46, 737,(1924). - 4? -to act as a catalyst, which functioned through the formation of a molecular compound of the alcohol with the metallic chloride, and not as a dehydrating agent. In this work it was found that the addition of zinc chloride to phosphorous penta-chloride increased the yields of chlorides produced. This method of preparation has been thoroughly studied. See tables and discussion of results. The action of thionyl chloride on alcohols, as a means of preparing alkyl chlorides has not been fully investigated. The most frequent use of this reagent is for the conversion of an acid into its chloride as it is easier to handle than (1) (2) phosphorous penta chloride. In a recent article by Silberrad, a summary of the work upon thionyl chloride is given, it has been used to introduce chlorine in place of various groups, e.g. OH, SH, NO^, SO^ H, H, 0; to introduce sulphur aloRe or combined with oxygen as the thionyl group (S)); as a catalyst; a dehydrating agent, a condensing agent, and a reagent for removing hydrogen from mercaptans. When thionyl chloride reacts in equivalent amounts it converts certain alcohols into (3) their chlorides. Thus Stahler and Schirm prepared methyl and (4) ethyl chlorides: McKenzie and Tudhope, d-B chlorooctane from l-sec. octyl alcohol; McKenzie and Clough, chlorethyl benzol (1) McMaster & Ahmann in a recent article (J.Amer.Chem.Soc. 30, 143,(1928) investigated the products farmed by the action of thionyl chloride on various acids. (2) Silverrad, J. Soc.Chem.Ind. 43,37,33,(1926). (3) Ber.44, 319,(1911). (4) J.Biol. Chem. 62,331,(1924). (3) J. Chem.Soc. 103, 687, (1913). from methyl benzyl carbinol; and Barger and Ewins - piperonyl ( 2 ) chloride from piperonyl alcohol. In 1911 Darzens found that an excess of thionyl chloride in the presence of a tertiary base as dimethylaniline, diethylaniline, or pyridine facili-tated the formation of the chloride. Thus isoamyl chloride, trichlorglycerol, benzyl chloride and cinnamyl chloride were prepared in good yields - compounds which he found difficult to prepare by phosphorous halides. In 1926 Majima and (3) Simanuki studied the action of thionyl chloride on polyatomic alcohols and obtained various sulpho and chlor sulpho esters. With an excess of thionyl chloride alone, Darzens obtained OR sulpho esters (S0<Q-^ ) and not the chlorides. Other investi-gators have obtained sulphurous esters of ethyl and certain (4) (3) (6) other fatty alcohols, fencyl, and triazoethyl alcohols. (7) Phenols in general formed such esters in pyridine, but others (8) as catechol without it, the reaction usually being conducted in a carbon disulphide solution. Similarly alizarin formed (1) J. Chem. Soc. 103, 733, (1?08). (2) Compt. rend. 132, I3I4, 1601. (1911). (3) Proc. Imp. Acad. (Japan) 2, 344, (I926). (4) (a) Carius, Ann. 70, 297, (184?); 106, 330, (I838J; and 111, 93,(1839). (b) Michaelis & Wagner Ber. 7, 1073,(1874); (c) Rosenhain & Sadow Ber. 38, 1298, (1903); (d) Arbusow, ahem,centre. 11, 684, (1909); (e) Hunter, J. uhem. Soc. 123, 1392, (1924). (3) Ruzicka & Liebl - Helv. Chim. Acta. 6, 267, (1923) or J. chem. Soc. A (1) 473, (1923). (6) Mckenzie & Clough J. Chem. Soc. 97, 2249, (1910). (7) Richter, Ber. 49, 2339, (1916); O-.P. 303, O33; J. uhem. Soc. A (1) 297, (1918). (8) Anschutz & Posth. Ber. 2?, 273I, (I894). - 31 -(1) "thionyl-alizarin" or anthraquinone sulphite. G^^H^O^SO^. When thionyl chloride acts on an excess of ethyl and other ^2) alcohols in the cold, as on several hydroxy acids, the inter-mediate and very unstable chlorosulphinyl compound e.g. Etc.Soul can be isolated, whence the chloride is obtained by removal of SO^, usually by means of heat. The action of thionyl chloride on alcohols as a general means of preparing alkyl chlorides has been further investi-gated and it has been found that alkyl chlorides can thus be prepared in the pure state with good yields, it was found to be an exceptionally good reagent for the preparation of h^gh boiling point chlorides as all the products of the reaction, except the chloride, are gaseous. up to the present, no work has been done to compare the relative yields and costs of chlorides prepared by the various methods. In many laboratory guides phosphorus pentachloride is the sole chlorinating agent, although hydrochloric acid or phosphorus trichloride and zinc chloride give a cheaper product. Chlorides of 28 alcohols have been prepared by the four best methods in the literature:- viz. (l) action of hydrochloric acid and zinc chloride (Norris method); (2) action of phosphorus trichloride and zinc chloride (Dehn and Davis method); (3) action of phosphorus pentachloride with zinc chloride (New method) (4) action of thionyl chloride. Various catalysts and solvents fl) Green, J. Chem. Soc. 123, 1430, (1224). (2) McKenzie & Barrow, ibid, 1910,(1211); Franklin & Garner, ibid, 103,1101,(1214); Blaise & Montagne, Compt. rend. 174, 1333, (1922). 32 were used in an effort to increase the yield and purity of the chlorides. The action of thionyl chloride on alcohols as a general preparative method has also been studied. The alcohols used in this investigation were obtained from the Eastman Kodak Co. The simpler aliphatic alcohols, methyl to amy! inclusive, and the polyhydric alcohols were carefully dried. Pure n-amyl alcohol and diethyl carbinol were obtained from the corresponding technical grades by a !* hly efficient rectifying column designed in this laboratory. The n-amyl alcohol fraction coming over at 137° " I39°c and the diethyl carbinol fraction coming over at 116° - 118°C were used. Other fractions gave isomeric alcohols, mostly iso-amyl alcohol. In all preparations, except where noted, one quarter mole of alcohol was used. All yields were based on the weight of chloride obtained on a second distillation after it had stood a day over anhydrous calcium chloride. The chlorides were not purified by refluxing with concentrated sulphuric acid (2) (3) as suggested by Norris. A recent paper has shown that this acid splits off halogen acid forming small quantities of olefines which then polmerize, the polymers adding on halogen acid to form unstable products and violet colors - giving an fl) Marshall, Ind. & Eng. Chem. V 20, 12, 1379 , (1928). (2) Norris & Taylor, J. Amer. Chem. Soc. 46, 737, (1924). (3) McCullough & Cortese. J. Amer. Chem. Soc. 31, 223, (1929). EXPERIMENTAL WORK - 33 -increase in the high boiling products. It was found sufficient to wash the chlorides with a 10% sodium hydroxide solution, then with a 13% sodium chloride solution, which minimized emulsion formation, and finally with pure water. Norris and Taylor's method was followed in preparing the chlorides with hydrochloric acid and zinc chloride. Briefly the procedure consisted in adding 2 moles of cooled, fused anhydrous zinc chloride to 2 moles of 38% hydrochloric acid solution. 1 mole of alcohol was added and the mixture was refluxed for a few hours or allowed to stand in the cold for a day, with frequent shaking (see benzyl chloride). The chloride was separated, or if necessary distilled, washed and dried as above. The zinc chloride, recovered by evaporating the residue, was used again. (1) Dehn and Davis' method was followed in preparing the lower aliphatic chlorides with phosphorus trichloride and zinc chloride. The apparatus consisted of a round bottom flask containing the previously heated anhydrous zinc chloride (2 moles per mole of alcohol) and one half the calculated amount of phosphorus trichloride (1;6 moles per mole of alcohol). The neck of the flask was fitted with a cork stopper holding two dropping funnels, containing the alcohol and remaining phosphorus trichloride respectively, and a distilling column. The column was closed by a cork containing a thermometer and (1) Dehn and Davis. J. Amer. Chem. Soc. 29, I328, (I9O7). was attached to the top of an inverted spiral condenser, down which water trickled from an automatically filled dropping funnel - to dissolve the hydrogen chloride evolved in the reaction. A suction flask packed in ice water or water acted as the receiver. The alcohol and phosphorus trichloride were slowly added, drop by drop; the flask in the meantime being gently heated. The chloride was finally distilled, separated, washed and dried as above. In preparing the higher aliphatic and aromatic chlorides where smaller quantities of the alcohol (l/lO or 1/20 mole) were used, a small distilling flask, with the side arm sealed to a small condenser, was used. The neck of the flask was fitted with a cork stopper holding a separatory funnel. The alcohol was added drop by drop from the funnel to the phosphorus trichloride and zinc chloride in the flask. The condenser was first inverted and the evolution of hydrogen chloride allowed to take place by gently warming the flask on an aniline bath. After gentle refluxing for half an hour the chloride was finally distilled, washed, and dried as before. By using a larger distilling flask and condenser, the chlorides were prepared as in the last paragraph from phosphorus pentachloride (1.1 mole), anhydrous zinc chloride (.2 mole), and the alcohol (1 mole). The mixture was finally distilled; the chloride, separated from the phosphorous oxy-chloride formed, was washed and dried as before. For the higher boiling chlorides and the polyhydric alcohols the mixture was - -first refluxed on an oil bath for 2 or 3 hours and then poured into a distilling flask and distilled. After distillation tarry phosphoric residues remained in the flask. In preparing the chlorides with thionyl chloride and a solvent, the alcohol and tertiary base (1:1) were mixed in a round bottom flask attached to an upright condenser. The flask was surrounded by a water bath and the thionyl chloride (2 moles) added, drop by drop, from a dropping funnel suspended in the top of the reflux condenser. The whole operation was conducted in a fume cupboard and the sulphur dioxide evolved escaped through the top of the condenser. After the addition of the thionyl chloride the flask was warmed and the whole refluxed for half to one hour, and then was cooled. If pyridine was used as the solvent the chloride was separated from the pyridine base -for other solvents a distillation was necessary. The excess thionyl chloride was recovered by distilling the crude chloride and the chloride was washed with water to remove any traces of thionyl chloride, dried, and distilled as before. Owing to the volatility of methyl and ethyl chlorides another apparatus was required for their preparation. The following apparatus was used in all their preparations. A distilling flask was used - its side arm bent sharply upwards and attached to an inverted condenser, (to condense alcohol vapor) which in turn was connected to four bubbling bottles. The first and third were empty and acted as traps while the second was filled with water and the fourth with ethyl alcohol. m pj (y ^  H e ca W-H > 4) 4^  ad o f-t 3 <6 a ^ Ssw ^  # w !L) H -St-* m E4 O cj 4^  Pf -H a o -H O C S-t M c <d o o OJ M * ^ ^ H -H d m }> a (!) <0 o o S-) Tl -H o Cv) CvJ c— C-- Lf\ co <X[ -st - -d-# * * H H H r-! OJ <\1 os CO - H tr-00 CO * * so tf\ C\) r-t H ) r-t ) OJ C--H CO ! H 0 00 tr- ) ^ t--co 00 co ! op 00 * * 0 0 43 0 0 (X) 0 -H t CO 0 C9 * r-t Lf\ * H 0 H a PH ! * o\ ! H * Lr\ CO Lf\ t PL) p) PQ c- CO * H H H r-t H ! O H 0) H # a o e * a a o a 43 o < ' a 0 O -H 0 (\) 43 d .a a ^ * 43 & m CO a ^ H H H 4s -H a Q * <D s M 0) 43 o O a o ^ o o 4 o 43 a) OS H o\ M 'r-t -st- CO H t- t- -H a OO 00 CO CO CO CO 'H * < * * * * pq a? 01 *H (D ^ 43 o 0) M <0 The water solution was first saturated with methyl (or ethyl) chloride at room temperature to dissolve the hydrogen chloride gas given off during the reaction. The distilling flask and last two bubbling bottles were placed in an ice bath. A current of dry air, introduced at the base of the condenser, was passed through the apparatus to prevent the contents of the bubbling bottles from sucking back. The neck of the flask was closed with a cork bearing a dropping funnel by which the alcohol or thionyl chloride was added, drop'by drop. The contents of the flask were warmed and the chloride slowly distilled, which dissolved in the alcohol. An analysis of the alcohol solution was made for chloride and the yield calculated. TABLES 1 & 2. Amounts indicated at column heads were those used per hydroxy! group. Thus for the polyhydric alcohols twice or three times as much of reagents were used per mole of alcohol. S.T. Abbreviation for sealed tube. "Same" refers to the same amount of alcohol and chloride used as chlorinating agent but with a different catalyst. Alote 1. 1/20 mole of alcohol used. Boiling point of chloride made by various reagents ranged from 91° - at mm. pressure. The chloride prepared with thionyl chloride had the sharpest B.P. B.P. 92° -33mm. Note 2. 1/40 mole of alcohol used. Benzene added as a solvent. Reaction was allowed to take place in an open beaker, the the excess thionyl chloride and benzene evaporated on a hot plate. Precipitate was dried between filter papers to remove phosphoric acid, etc., then recrys-tallized from ether. With thionyl chloride pure - 37 -prismatic crystals were obtained on evaporating off thionyl chloride. Note 3. n. Nonyl chloride has been prepared for the first time, the purified product boiling at 102° - 108°u. at 27 mm. pressure. Analysis for chloride gave 21.36%; calc. 21.82%. Its density and index of refraction is being determined. DISCUSSION OF RESULTS. Alcohols with the following radicals were studied:-methyl; ethyl; n. and iso. propyl; n., sec., iso., and ter., butyl; n., iso., sec., ter., amyl; diethyl carbinol; n. hexyl; n. heptyl; n. octyl; capryl; n. nonyl; cetyl; allyl, and cinnamyl. Polyhydric alcohols as ethylene, propylene, and trimethylene glycols; glycerol and aromatic alcohols as benzyl, phenyl methyl carbinol, and triphenyl carbinol were studied. The yields of the various methods verified Riorris' (1) observation that "an increase in the molecular weight in a parrafin series decreases the activity of the alcohol, and that secondary alcohols react more readily than primary alcohols while tertiary alcohols react quantitatively in the cold". Thus capryl alcohol or methyl n. hexyl carbinol, a secondary octyl alcohol reacted more vigorously than did n. octyl alcohol itself. Sec. n. amyl alcohol and diethyl carbinol reacted less vigorously than and iso amyl alcohols probably because very small amounts of sec. n. amyl alcohol were used and because the two ethyl groups in diethyl carbinol, joined to the same carbon atom, more or less protected the carbinol (1) J. Amer. Chem. Soc. 38, 1071, (191^). - 38 -group. Polyhydric alcohols reacted less vigorously than did monatomic alcohols. Trimethylene dichloride was obtained in higher yields than was propylene dichloride - probably because the secondary alcoholic group in the latter was protected by the groups on either side. Glycerol on chlorination usually formed the monochlorhydrih although the dichlorhydrin was sometimes formed. Unsaturated alcohols, as ally! and cinnamyl gave lower yields than the corresponding saturated compounds, usually forming tarry addition products on distillation. in a number of cases very small amounts of the alcohols (3 to 3 gms.) were employed, so large yields could not be expected. The results obtained have led to practical methods of preparation. Only those chlorides were prepared by the action of concentrated hydrochloric acid and zinc chloride which had not been previously made in this manner. N. butyl and iso amyl chlorides were age in prepared as a check on previous work. The yields of the lower aliphatic chlorides (see n. butyl) and others (see benzyl) were obtained by shaking the reagents (1) together in the cold. Lucas claimed that this method gave a purer product and a better yield - at least for the higher aliphatic alcohols. This observation was not cinfirmed for n. hexyl, n. heptyl, and n. octyl alcohols. This method is recommended for the preparation of large amounts of the lower (1) Lucas, J. Amer. Chem. Soc. 31, 248, (1?29). - 39 -aliphatic chlorides and benzyl chloride. The zinc chloride recovered could be used over again. All tertiary alcohols -trimethyl carbinol, dimethyl ethyl carbinol, and triphenyl-carbinol, gave yields of 83% - 100% with hydrochloric acid alone in the cold. The second method of preparing chlorides by the action of phosphorus trichloride and zinc chloride on the alcohol gave fair yields for primary and secondary chlorides (60% -80%) and still better yields for tertiary chlorides (92% - 98%). The apparatus was slightly complicated and required constant attention. The yields were not as high as those obtained with thionyl chloride. Often the chlorides formed were impure -yellow phosphoric compounds separated out while drying. When benzene was used as a solvent with the polyhydric alcohols the amount of tarry material was reduced. This method is recom-mended, next to the hydrochloric acid - zinc chloride method, for the preparation of the lower aliphatic chlorides and has the advantage that it can also be used for the preparation of higher aliphatic, unsaturated, and aromatic chlorides. The third method introduces a new means of preparing alkyl chlorides. Phosphorus penta-chloride, alone, has long been used for the preparation of chlorides in laboratory manuals although the yields were relatively poor and the com-pound very unpleasant to work with. In the literature many chlorides have been prepared with phosphorus pentachloride but - 60 -(1) only in one instance has the yield ever been recorded. The addition of a small amount of zinc chloride was found to increase the yields of the chlorides. Thus iso butyl alcohol gave a 76% yield of the chloride with zinc chloride and only 43% without it; likewise isoamyl alcohol gave a 71% yield of the chloride with zinc chloride as compared to 30% without it. This method was studied as a general means of preparing alkyl chlorides from 28 alcohols and was found to give yields of 70% - 80% with most aliphatic alcohols. Note -.intermediate aliphatic alcohols from n. hexyl to n. nonyl gave slightly smaller yields of chlorides. Cetyl, cinnamyl and aromatic alcohols formed tars. The action of thionyl chloride on alcohols proved to be a practical method of preparing pure alkyl chlorides with good yields. For the lower aliphatic alcohols it was found necess-ary to use a solvent, otherwise impure chlorides, mixed with sulpho esters and chlorosulphinic compounds, were formed with poor yields. Pyridine was shown to be the best tertiary base for the lower aliphatic chlorides but with the higher boiling chlorides tars were formed from which the chloride could not be obtained. The unsaturated alcohols, aromatic alcohols, and higher polyhydric alcohols also resinified when pyridine was used as a solvent. Dimethyl and diethyl anilines gave lower yields than pyridine (see experiments on isobutyl and isomyl (1) Pierre, Puchot. Ann. I63, 266, (1872). o R H M PQ M ^ M H w o M o PQ ^ o o M E-i t,Q CO C3 M > H ^ M o <D > r-t O 03 CM H c\j 3 o^ o 73 a H o H M H CD o -t cUEr-t O c! to CM H a tS! r-t CO Pi it ^ r-t ) co PL) ti M CU CO a r-t CO M CO o o o H ^ it it r-t CO CM H :o a tS] -t t^ t^ O -t & t H CO o E-t CM H o o co co sO S-t <0 o -d <-t -w^o P) K) o o o r - l O O O OS sO o sO it U it t oo 00 r-t OO co yO MA It I It It CM CO sO CO sO i it it it CM sO cr-r-t O CM co sO o r-t r-t >3-r-t O CD oo co sO o oo OO Os It t t t OS oo so r-t O CO oo OS CM co H t H i o r-t r-t OS OO o it t I u co Os CO CO OJ * CO r-t r-l S ^ CO ^  o - co o * M ^O OO # r-t CM CO Os sO Os it i i i <-) C r-t C— <D OO Os t i ti it <-t OS cr-CO sT CO CM CM O ti i t ti CO OS O r-t r-t S -M Cj o * o ^ * C sO +3 ^  O ^ p-l < CO CM r-t CO r-t CO c sO CM it it il ' (0 CO sO CM rW r-t CO r-t O^ Lf\ CM t It i r-t ^ CO P) * O CO c- p^ o r-t r-t CO CM IX-sO I t ti i 3 r-t CM os CO CO O O r-t r-t en S tR r-l >3 CO CO O ^  * CM OO r-t CO eo O OO OJ CO OO r-l H I i it Ii t ti I Os C\J CM CO OJ CO CO CO CM it i Ii It OS OJ O s CM r-t CO CM g -r-l CO rM ^ CO .^t CJ O sO CO p) CO r-t CO MX CO CO r-t CO i ti i I I sO MA r-t CM sO OS CM CO H ti Ii i ti E— r-t CO sO 03 O CO ti I It It (D O -W r-t .,-) O ^  -{ O.CCO >3 O CJ CD -Hr-t'ClUA to b.3 —^  ^ O s sO CO L-A Ii it i t Ii it i r-t sO Os 0^ CO CS sO It Ii It r-t C'J r4 CO a * - 61 -alcohols) and also tarred with the higher alcohols. Aniline was found to be unsuitable as a solvent. For the higher ali-phatic alcohols, unsaturated, aromatic, and polyhydric alcohols, thionyl chloride alone or thionyl chloride with benzene as a solvent gave satisfactory yields. Here an excess of thionyl chloride was generally used (3*4 moles per mole of alcohol) and the excess recovered for further use by distillation. All the chlorides, prepared by means of thionyl chloride, as above, were very pure with sharp boiling points. The yields varied from 80y. - 297° f°r methyl-amyl chlorides inclusive while n. nonyl alcohol gave 62% yield by this method. The yields of aromatic chlorides were also good (82% - 97%); the yields of the unsaturated and polyhydric alcohols being somewhat less. Diethyl carbinol, again, was an exception and on repeated preparation gave only a 42% - 44% yield. This method is recommended for preparing small quantities of a chloride when only a small amount of the alcohol is available and a pure product is demanded. JMo phosphoric residues remain after the reaction, the other products of the reaction being gaseous. TABLE 4. The relative costs of preparing one kilogram of six typical alkyl chlorides, by each method, have been determined -using the yields obtained in their preparation. Owing to the high cost of n. octyl alcohol, the relative costs of preparing 100 gms only, were calculated. The seven alcohols chosen - n. - 62 -propyl, iso-amyl, n. octyl, cinnamyl, ethylene glycol, and benzyl, are typical of the 28 alcohols investigated. In the case of ethylene glycol, costs were calculated on preparing one kilogram of the dichloride. The prices of the alcohols were taken from Eastman Organic Chemicals List, No* lp, October, 1$28. The costs of other chemicals were taken from recent trade lists. The following costs per pound were used:- zinc chloride, 80%; hydrochloric acid, 23%; pyridine and benzene, §2.23; phosphorus trichloride, §1.30; pentachloride, §1.30; and thionyl chloride, §6.00. The cheapest method for all chlorides but one (n. octyl) is the hydrochloric acid - zinc chloride preparation; Owing to the high cost of thionyl chloride as compared to the chlorides of phosphorus, the chlorides prepared with it are more expensive, except in a case where the cost of the alcohol is very high (n. octyl). If a good yield of pure chloride from a small amount of alcohol is desired then it is advisable to use thionyl chloride. Sm-S'.IARY. 1. Alkyl chlorides of 28 alcohols have been prepared by four methods - hydrochloric acid and zinc chloride, phosphorus trichloride and zinc chloride, phosphorus penta-chloride and zinc chloride, and thionyl chloride and pyridine or benzene. - 63 -2. The yield for each method has been determined. 3. The use of thionyl chloride as a general means of preparing alkyl chlorides has been studied. 4. Alkyl chlorides haveibeen prepared with phosphorus pentachloride using zinc chloride as a catalyst. - the yields being higher than if phosphorus pentachloride was used alone. The limitations of each method can be readily seen by consulting Tables 1 and 2. 6. The effects of structure on the yields obtained have been verified and extended from former observations. 7. Other catalysts than zinc chloride have been used without success. 8. N. nonyl chloride has been prepared for the first time and its physical properties determined. 9. The relative costs of preparing seven typical alkyl chlorides have been calculated. 64. BIBLIOGRAPHY. 1. Abegg. Z. anorg. Chem. 39, 330.(1904). "Handbuch der anorg. Chem."Abt.3,Bd.3,S215. 2. Allen. Commercial Organic Analysis. Vol. 7, 439.' 3. Anschutz & posth. Ber. 27,2731,(1894). 4. Arbusow. Chem. Centre. 11, 684,(1909). 3 . Baines. J. Chem. Soc. 121, 2810, (1922). 6. Bardwell. J. Am. Chem. Soc. 44, 2499, (1922). 7 . Barggr & Ewlns. J. Chem. Soc. 103,733, (1908). 8. Barlew. Zeit. anorg. Chem. 99,188. 9. Bender. Ber. 19, 2272, (1886). 10 . Blaise & Montagne. Compt. rend. 174, 1333, (1922). 11. Bohr. Ph&l. Mag. (6) 2b, 837. 12. Boismenu. Compt. rend. 133, 948, (1911). 13. Bruner & Bekier. Zeit. Elec. Chem. 78,369, (1912). 14. Bruner & Galecki. Zeit. Phys. Chem. 84, 313, (1913). 13. Carius. Ann. 70, 297, (1849); 106, 330, (1838); 111, ?3, 16. Chambox. Jahresbpriohte (1876), 203. . (1859). 17. Chattaway.* J. Chem. Soc. 70, 1372, (1896); Amer. Chem. See.J. 23, 363, (1900); 24, 139, (1900); 24, 318, 331, 342, (1900). 18. Chattaway & Wadmore. 8l, 193, (1902) J. Chem. Soc. 19. Clark & Ball. J. Amer. Elec. (1928). 20. Clark & Carter. Trans. Royal Soc. Canada. 3rd. Series, Vol. 21, 323, (1927). 21. Clark & Crozier. ibid. Vol. 19, 157, (1923). 22. Cofman. J. Chem. Soc. 115, 1040, (1919). 23. Crocker. J. Amer. Chem. Soc. 44, l6l8, (1922). 24. Cuy. ibid. 42, 303, (1902). 23. Darzens. Compt. rend. 132, 1314, 1601, (1911). 26. Dehn & Davis. J. Am. Chem. Soc. 29, 1323, (1907). 27. De la Rive. Phil. Mag. (2)3, 143, (1828). 28. Falk & Nelson. J. Am. Chem. Soc. 32. 1637, (1910); 33, 1810, (1913); 37, 1732, (1913). , 29. Faraday. Ostwald's Klassiker. 86, 43, (1834). 30. Francis. J. Am. Chem. Soc. 47, 2340, (1923). 31. Franklin & Garner. J. Chem. Soc. 103, 1101, (1914). 32. Fry. " Electronic Conception of Valence"; Longman's Green & Co. Ltd. London. (1921). Chemical Reviews. 3, 4, 337, (1928). 33. Garre. Compt. rend . 136,736, (1903).; Bull. Soc. Chim. (Ill), 27, 264, (1902). 34. Glocker. Nat. Acad. Sci. 12, 522, (1926); J. Am. Chem. Soc. 48, 202, (1926). 35. Green. J. Chem. Soc. 125, 1430, (1924). 36. Grignard & Bellet. Compt. rend. 135, 44, (1912);158, 457 37. Grignard & Crouzier. J. Chem. Soc. A.(l),404(1921%1914). 38. Grimm & Sommerfeld. Zeit. Phys&k. 36* 32, (1926). 39. Groves. J. Chem. Soc. 12, 636, (1874), 40. Hanke & Koessler. J. Am. Chem. Soc. 40, 1726, (1918). 41. Harrison, Kenyon & Phillips. J. Chem. Soc. 129, 2079,(%926) 65. 42. Higgins. Science. 55 , 674, (1922).;J.Am. Chem. Soc.44, 1607, (1922). 43. Hofmann.Ber. 15, 4121 (1882). 44. Houben* "Die Methoden der Org. Chem." Vol. 4, 414,(1924). 45. Howell & Noyes. J. Am. Chem. Soc. 42, 991, (1920). 46. Hunter. J. Chem. Soc. 125, 3:392, (1924). 47. Jakowkin. Zeit. physik. Chem. 29, 613, (1899). 48. Jennings & Scott. J. Am. Chem. Soc. 41, 3241, (1919). 49. Jones. Am. Chem. J. 48, 1912; 39, 426; J. Am. Chem. Soc. 36, 1268, (1914). 50. Jones & Werner, ibid.. 40, 1257, (1919). 51. Kermack & Robinson. J. Chem. Soc. 121, 427, (1920). 52. Kossel. Ann. der Phys. 49, 229,(1916). 53. Knorr. Zei&. anorg. Chem. 129,109, (1923). 54. Kraut. Ann. 138, 332. 55. Langmuir. J. Am. Chem. Soc. 38, 222,(1916); 41, 868,1543, (1919). 36. Lapworth. J. Chem. Soc. 121, 416,(1920). 37. Latimer a Rodebush. ibid. 42, 1419, (1920). 58. Lewis, ibid. 38, 762, (1916). 59. Lowry. Trans Faraday Soc. 19, 486, (1923); J. Chem. Soc. 123, 822, (1923). ^ 60. Lucas. J. Am. Chem.Soc, 46,2473,(1924);47,1439,(1923); 47,1462,(1923);48,1827,(1926);30,1711,(1928);51,248,(1929). 61. Majima & Simanuki. Proc. Imp. Acad. (Japan)2,344, (1926). 62. MacClung. "Solvents in Synthetic Organic Chemistry." Van Nostrand & Co. 1925. page 108. 63. Marshall. Ind & Eng. Chem. Vol. 20, 12, 1379,(1928). 64. McKenzie & Clough. J.Chem. Soc.97,2249,(1910);103,687,(1913^ 63. MCKenzie & Barrow. J. Chem. Soc. 99,1910,{1911). 66. MCKenzie & Tudhope. J.Biol. Chem. 62, 551, (1924). 6 7 . HcCullough & Cortese. J. Am. Chem. Soc. 51, 225,(L929). 68. MOMaster & Ahmann. ibid. 50, 145, (1928). 69. Menschutkin. Ann. 139, 343-70. Michaelis & Wagner. Ber. 7 , lo?3, (1374). 71. Moer. Zeit. anorg . Chem.113, 179, (1920). 72. Nef. Ann. 287, 316,(1895);308,173. 7 3 . Nicolet & Sampey. J. Am. Chem. S6c. 49,2849, (1927). 74. Nofris. Am. Chem. J. 38, 627, (1907); J. Am. Chem. Soc. 38, 1071, (1916); 42, 2093, (1923). 73. Norris & Green. Am. Chem. J. 26, 307, (1901). 7 6 . Norris & Taylor. J. Am. Chem. Soc. 46, 737, (1924). 77. W.A.Noves. Chem. News. 50, 228,(1904);J. Am. Chem. Soc. 35, 767, (1913); 39, 903, (19R7); 42, 2i67,(i920$;47, 3027,(1925); 50, 2902,(1928); Zeit. Physik.Chem. 130,329, (1927); Proc. Nat. Acad. Sci. 13, 377, (1927). 78. Noyes & Lyon. J. Am. Chem. Soc. 23, 463, (1901). 79. Noyes & Wilson, ibid. 44, 1630, (1922). 80. Orton & Dlackmann. J. Chem. Soc. 77, 030, (1900). 81. Pantier. Compt. rend. 76, 49. 82. Parsons. Smithsonian Inst. Publ. 65, No.11,(1913). 83. Pauling. J. Am. Chem. Soc. 48, 1132;fl926). 84. Pelouze. Ann. Chem. phys. (3)7, 126,(1843). 66. 85. Pierre & Puchot. Ann. 163, 266, (1872). 86. Richter. Ber. 49, 2339, (1916); G.P. 303,033; J. Chem. Soo. A.(l) 297, (1918). 87. Risson. Compt. rend. 125, 1032. 88. Rosenhain & Sadow. Ber. 38, 1298,(1905). 89. Ruzicka & Liebl. Eelv. Chim. Acta. 6, 267, (1923) or J. Chem. Soc. A. (l) 473, (1923). 90. Schiff. Ann. 102, 334; 103, 164. 91. Schutzenberger. ibid. 129, 113. 92. Selivanow. Ber. 23, 3617,(1892); 26, 423,(1393). 93. Sidgwick. Trans. Faraday Soc. 19, 469, (1933). 94. Silverrad. J. Soc. Chem. Inl.43, 37, 33, (1926). 95. Solly. Phil. Mag. (3) 8, 401, (I83M. 96. Sommerfeld. " Atombau und Spektrallinien" Vieweg & Sohn. Braunschweig (1922). 97. Soper & Smith. J. Chem. Soc. 129, 1382, (1926). 98. Stabler & Schirm. Ber. 44, 319, (1911). 99. Stark. "Prinzipien du Atomdynamik, Die'Elektrizitat im chemischen'Atom." Hurzel. Leipzig. (1913). 100. Stieglitz. J. Amer. Chem. Soc. 23, 796, (1901); 44, 1301, (1922). 101. Stieglitz & Petersen, ibid. 36, 273, (1914). 102. Sugden. J. Chem. Soc. 123, 822, (1923). 103. J.J.Thomson. Phil. Mag. (6), 27, 737, (1914); "Corpuscular Theory of Matter." P. 130. (1907). 104. Wurtz. Ann. 58, 72. 


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items