Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Polarity effects in organic compounds. The quantitative replacement of halogen atoms substituted in the… Ball, Ralph Henry 1928

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

Item Metadata


831-ubc_1928_a8_b2_p6.pdf [ 24.57MB ]
JSON: 831-1.0060848.json
JSON-LD: 831-1.0060848-ld.json
RDF/XML (Pretty): 831-1.0060848-rdf.xml
RDF/JSON: 831-1.0060848-rdf.json
Turtle: 831-1.0060848-turtle.txt
N-Triples: 831-1.0060848-rdf-ntriples.txt
Original Record: 831-1.0060848-source.json
Full Text

Full Text

r*  U.B.C.  LIBRARY  CAT LEs (i !i I *CC. «*0.  /3m* fls-  POLARITY EFFECTS IN ORGANIC COMPOUNDS.  The Quantitative Replacement of Halogen Atoms Substituted in the Benzene Nucleus.  A Search for the Halogen Cation .  - by RALPH HENRY BALL.  A Thesis submitted for the Degree of MASTER OF ARTS in the Department of Chemistry.  THE UNIVERSITY OF BRITISH COLUMBIA APRIL.  1928.  .!•! II Ml  i  TABLE OF CONTENTS* POLARITY EFFECTS IN ORGANIC COMPOUNDS  Part I -  Theories of Polarity.  A. Introduction. B. General Theory. A development of the two fundamental electrical theories of valence, as applied to aliphatic compounds. Part II - The Quantitative Replacement of Halogens Atoms Substituted in the Benzene Nucleus. A. Theoretical Part. 1. Extension of the theories of valence to include the structure of benzene. 2. The application of the "electronic" structure of benzene. B. Review. Previous work on the replacement of benzene substituents considered in the light of this structure of benzene. C. Experimental Part. 1. Preliminary experiments. 2. Final Method. 3. Results. 4. Discussion of results. Part III - A Search for the Halogen Cation. A. Introduction. 1. The reasons for believing that hypohalous halogen is positive halogen. 2. The compounds which contain positive halogen. B. Review. Previous work on the electrolysis of positive halogen compounds. 6. Experimental Part. 1. Work on carbon tetrabromide. 2. Work on bromyl acetamide. 3. Work on bromyl succinimide. D. Discussion of Results.  POLARITY EFFECTS IN ORGANIC COMPOUNDS  PART I. A. GENERAL INTRODUCTION. In recent years several authors have proposed different "electronic" theories of valence, "based more or less on the ideas of chemical combination suggested by Bohr's theory of the structure of the atom. These theories predict certain relations between the reactivity of a group and its polarity, which are supported by considerable chemical evidence. In view of these predictions it was decided to investigate the reactivity of variously situated halogen atoms substituted in the benzene nucleus. The results of this research are recorded in Part II of this paper. A further investigation was undertaken to determine by electrolysis if the halogen atoms in certain aliphatic halogen compounds, wherein the halogen was predicted to be positive, could be separated as a positive ion. The results of this research are recorded in Part III of this paper. B.GENERAL THEORY. The Theories of Valence. Hlhe results of both Parts II and III of this paper have R.  been interpretedthrough a consideration of the polarity of  the radicals concerned, ^ence it is necessary at this point to outline "briefly the various theories regarding the polarity of the chemical "bond; that is, in the union "between two atoms or groups, which of the two plays the part of a positive radical and which negative. The fundamental assumption in any theory of the polarity of the chemical bond is that the bond is due to electricstatic force , or to the agency of electrons in some way. Based on this assumption there are two major theories: (l) the so-called mdidern dualistic or electronic theory, which assumes that each atom is the source of a definitely directed force; and (2)the electron-bond theory, originated by Lewis, which assumes that the bond consists of a pair of electrons shared by the valence ring of both combining atoms, and which is more directly based on Bohr's theory of the v atom. As these two theories lead to similar predictions in many cases, while in others they are opposed, it is necessary to proceed to a brief description of their main points. The Electronic Theory of Valence. The basis of this theory is that the atoms in a chemical compound are held together by electrostatic force set up by the difference in the electrical properties of the two combined atoms. Hence the bonds have direction. This idea originated with fir J.J. Thomson, and was developed from a  3. n  2  3  chemical point of view byjFry, Stieglitz, Falk and Kelson , a 6 6 Lapworth ,1.W. Jones , Cuy , and others.The fundamental assumptions of the theory are as follows: (a) when an element X becomes an ion X?" it does so by losing an electron, thereby becoming the centre of an electrostatic force. It may also gain an electron and become X, in which case it is the centre of an electrostatic force of the opposite sign to xt It is assumed that all univalent elements or groups can either acquire or lose an electron, but that thay usually have a tendency to behave as either positive or negative. The validity of this assumption may be illustrated by hydrogen and chlorine. From electrolysis we infer that the hydrogen in HC1 is H + , but from the hydrolysis of NaH we must conclude that the hydrogen here is H~. Similarly from electrolysis of the HCl we see that the chlorine is Cl~. But in hypochlorous acid we believe that we have  rj  i.e., H0C1 . Furthermore, by these assumptions we have a very plausible explanation of the forces which hold the atoms of the diatomic gases in combination • !ghus we may represent H 2 as H V H , or Clg as CI—"01. (b) Now consider a reaction between two diatomic molecules:-  1. Fry. Electronic Conception of Valence. 2. Stieglitz. Am. ,Ghei*i> SPBr.J. 18, 756. 3. Falk and Kelson. J. Am. Chem. Soc. 38, 1637. 4. lapworth. J. Chem. Soc. 181, 416. 5L.W. 3ones. Am.-Chem. 50, 414. 6. Cuy. J. Am. Chem. Soc. 42, 503 7. Stieglitz. J. Am. Chem. Soc. 83, 797.  4 H ^  H +  H~  c i ^ c r + c f jr  jr  H-Ci" H - C I h From such a reaction we should obtain two electronic isomers(electromers). ®ut we ordinarily assign one structure to the result of such a reaction, and recognize only one product. This difficulty is overcome by the assumption of electronic tautomerism;i.e., a dynamic equilibrium between the two electyomers.  ^ — ^  - — X  ^Y  •^s the electronic properties of each group are usually quite pronounced, the equilibrium is almost entirely in one direction, and so we recognize only one electromer. (c) Now apply these ideas to a divalent atom. As each bond may posses direction, we have the following possibilities.  X H  X I Z  X  ^  This is generalized in the statement that if an element has a valence of n, it may function electronically in n+1 different ways. (d) This idea of electronic valence finds its chemical analogy in oxidation-reduction. From anodic oxidation and cathodic reduction in inorganic chemistry, we are accustomed to consider oxidation as the loss of electrons by an atom,and reduction as the gain of electrons. Applying this to the different electronic ways in which an atom may function, we see that each way represents a definite state of oxidation of the atom. This has been applied to the carbon atom by Fry as f0}.l0W8 .  As in As in As in As in formic As in methane.methyl alchoi„ formaldehyde. &oid. carfeon dioxide The tracing of these steps by Bone"1" in the slow combustion of methane is a strong point for the theory. It is also interesting that Jones2 has worked out the same idea for nitrogen, and thus explains the relations between the various organic nitrogen compounds, and the Beckmann rearrangements occuring in them. (e) Finally, to convert this theory into a working hypothesis it is necess&ry to assign to the commoner groups and atoms their characteristic polarity. This is done by considering their reactions on electrolysis, on oxidation and reduction, and on hydrolysis. In this way electronic formula have been worked out, which often prove very fruitful ig interpreting and predicting the reactions of a compound. For instance, from the method of making acetbromamide, and its analogy to nitrogen trichloride, we believe that the halogen is positive. Hence, we would expect it to be very reactive, and to give hypobromous acid on hydrolysis,which is verified on experiment. Electronic formula have been employed also in explaining the properties of straight chain compounds and the mechanism of the Beckmemn rearrangement. However, since the electronic theory is used  1. Bone, see Fry, "Electronic Conception of Valence." p. 17. 2. Jones. Am. Chem. d . 60, 414.  6. in this paper chiefly to interpret the reactions of aromatic compounds, I shall leave a more detailed account of its application for Part II. The Shared-electron Bond Theory* This theory is based on Bohr's model of the atom, which is now very widel^4ceepted. It was originally proposed by G.N. Lewis , and has since been elaboratedAmodified slightly 2 3 4. by Langmuir , Kermack and Robinson , Lucas and Jameson , 5 fi 7 8 Q 10 Lowry , Crocker , fiuggins , Knorr , Pauling , Berliner and others. The fundamental difference between this and the "electronic" theory is that it denies that each atom in combination exists in a definite oxidation stage, and instead claims that a bond may change gradually from a p o l a r — b o n d to a polar  bond, passing through the intervening non polar  stages. One of the chief contributions of this theory is that it offers a physical interpretation for the mechanism of chemical combination, which is in accord with the most generally accepted structure of the atom. (a) This mechanism is developed by combining Bohr's theory of the atom with valence as exhibited in the periodic table. Bohr claims that the atom consists of a nucleus, with a net 1. G.N. Lewis. J. Am. Chem. Soc. 38, 762. 2. Langmuir. J. Am. Chem Soc. 38, 222; 41, 868; 1543. 3. Kermack and Robinson. J. Chem. Soc. 121, 427. 4. Lucas and Jameson. J. Am. Chem. Soc. 46,2475; 47, 1459. 5. Lowry. J. Chem. Soc. 123, 822. 6. Crocker. J. Am. Chem. Soc. 44, 1618. 7. Huggins. J. Am. Chem. Soc. 44, 1607. 8Knorr. Z. Anorg. Chem. 129, 109. 9. Pauling. J. Am. Chem. Soc. 48, 1132. 10. Berliner. J. Phys. Chem., Frt.1928, 293.  7 positive charge represented by the atomic number, surrounded by planetary electrons arranged in shells or energy levels. #y subtracting the total number of planetary electrons(given by the atomic number), from the number contained in the inner levels, we find that the number of electrons in the outermost shell is closely connected with the valence of the atom. This connection is summed up in the statement that there is a strong tendency for the valence shell to contain eight electrons; i.e., to take on the electron structure of the nearest noble gas. For example, chlorine has seven electrons in its valence shell. It has a strong tendency to acquire one electron and become a negative ion, thereby attaining the stable configuration of eight electrons in its outer ring. On the other hand, sodium has one electron in the valenee ring, and eight electrons in the ring within this. Hence, sodium is very willing to part with its one valence electron and become a positive ion. likewise magnesium having two valence electrons, readily parts with them to become a divalent positive ion; and carbon^ in the middle of the table,, with its foijr valence electrons, can either gain or lose four electrons to attain the stable configuration of eight. (b) For reasons which are beyond the scope of this paper, Lewis also claims great stability for the group of two electrons, and produces evidence to show that even in the valence ring of an atom, the electronsy4re paired; i.e.,their orbits are in some way similar. (c) Now in order to give direction to the bond, and to  8 give it a point of attack on the valence ring, Sommerfeld's idea of elliptic orbits is applied. From this idea , and that of paired electrons, it follows that for a tetravalent atom like carbon, the paired ellipses will extend from the atom towards the four corners of a tetrahedron. (d) Finally we come to Lewis's conception of the chemical bond. He believes that the chemical bond consists of the sharing of two valence electrons between the two combined atoms. As the two valence electrons concerned are rotating in orbits, he must consider that the actual forces of combination are electromagnetic, although he does not make this point clear. In representing this bond, because of the fact that these orbits have direction, Lewis places his electrons in fixed positions, corresponding to the extremities of the ellipses. Some of his formula are as follows:-  H  H:H  :ci:ci:  ''  "  H*.C:H  Li  H  H'-C::d: - O'  H'X:::c:H  H  In this theory the bonding pair may lie anywhere between the two bound atoms, depending on the intrinsic positive or negative characteristics of the atoms  concerned. In the case  of polar bonds the bonding pair is assumed to be in complete possession of one of the atoms so joined,In this way the Lewis theory includes the "electronic" theory as a special case. Consider A B, bound by an electron pair. It may exist as A :  B, In which A is negative and B is positive, and they  are held together by electrostatic force; or as A polar; or as A  :  B, non  :B , in which A is positive and B is  negative. According to Lewis, all stages in between ( limited  9 by quantum relations ) are possible. In my opinion this is open to critieism, for in shifting his electron bond in this fashion, Lewis seemes to have forgotten that he is not shifting electrons but orbits, and he does not explain what chemical or physical significance this has. This particular difficulty is surmounted by the interpretation of the electron bond given by Knorr\ and supplemented by Pauling . These authors differentiate between the polar and non polar bond. In the polar bond they assume that She valence electrons of the positive element are held by the valence ring of the negative element, and thus the force holding the two in combination is electrostatic. For instance, sodium chloride is represented thus:-  A/q+  -CI:  This is equivalent to the electronic formulae proposed by Fry, Stieglitz, Jones, etc. But in order to represent the non polar bond, and to place chemical combination on a physical basis, Knorr and Pauling realize the necessity of the shared electron bond; and they interpret this by a double elliptical orbit enclosing the nuclei of both combined atoms. In this way, if the enclosing orbit is small the bond id very stable and unreactive, whereas if the orbit is elongated the bond is more or less unstable and has a tendency to change to a polar bond, the  1. Knorr. 2. Anorg. Chem. 129, 109. 2. Pauling. J. Am. Ohem. Soc. 48, 1132.  10 polarity of the resulting atoms depending upon which of the two atoms retains the electrons. Even this interpretation is open to criticism in this particular - that in chemical reactions we are dealing with the breaking of the bond, and as the bond always breaks on one side or the other of the bonding pair, the bond in breaking must become polar. So that it would seem that, no matter which theory is concerned, we are justified in indicating the course which our reaction will take by the customary positive and negative signs. Thus  the theory of the shared electron bond gives us  a very plausible mechanism for chemical combination, and also for the transfer of basic or acidic properties along a carbon chain. It has also been used to explain induced alternate polarity in a carbon chain, and in the benzene ring. This latter will be dealt with in Partll. To sum up, the usefulness of this theory is limited to the picture it gives us of the stale of a molecule before reaction takes place, and thus it helps to predict the reactivity of the various groups. But it is of no value in representing the actual course of a reaction, as an atom or group, on entering into reaction, must become polar. In this way a negative group can only be replaced by a negative group without oxidation, and vice versa, so that we are justified in writing our positive and negative signs as representing a definite state of oxidation of the group.  11  PART II. THE QUANTITATIVE REPLACEMENT OF HALOGEN ATOMS SUBSTITUTED IN THE BENZENE NUCLEUS.  A. THEORETICAL PART. Theories of The Structure of Benzene.  There have "been many theories advanced to account for the Crum-Brown Gibson rule for benzene substitution , and for the influence of certain groups on the mobility of substituents in the benzene nucleus. These may be divided into two parts; (1) Those offering a mechanical explanation, such as various 1 2 methods of addition, proposed by Holleman , Weinberg ,and in various modifications by many other authors; (2f Those which aecount for the rule by the electrical properties of the substituting groups and the benzene ring, 3 4 advanced r by Fry, Vrrdlander ,Q Flurscheim , Obermiller^, 7 q  Lapworth , Huggins , Crocker , Pauling . As the electronic theories are much more useful in correlating the extensive datd of benzene substitution, and offer a logical explanation of the Crum-Brown Gibson rule, I shall consider only them in 1. 2. 3. 4. 5. 6. 7. 8. 9.  Holleman. Rec. Trav. Chim. 33, 1. Weinberg. Ber. 52, B.C. (1919) 1501. Vordlander. Ber. 52, 263. Flrrscheim. J. Pract. Chem. 66, 321; 71, 497. Obermiller. Organic Chemistry. Volume 1,+a"1 edition.Cohen. Lapworth. Organic Chemistry. Volume I, 4- edition. Cohen. Huggins. J. Am. Chem. Soc. 44, 1607. Crocker. J. Am. Chem. Soc. 44, 1618. Pauling. J. Am. Chem. Soc. 48, 1132.  12 the review of previous work and the interpretation of my own results. I have previouslyshown that the electrical theories of valence may he divided into two classes, the "electronic" theory, upheld chiefly by Fry and Stieglitz, and the "electron hond" theory, brought forward by Lewis and Langmuir. These may both be developed into a theory for benzene, which gives the alternate carbon atoms different electrical properties. The Electronic Theory. Attention was drawn previously to the fact that the electronic theory postulates that the carbon atom exists in five different states of oxidation. Fry shows that on combining these into a formula for benzene of the centric type, only six differdnt arrangements can be made; and furthermore that all these have one peculiarity, namely that the carbonhydrogen bond is alternately  , z-JL around the ring.  Therefore the hydrogen atoms in benzene are alternately positive and negative in the benzene nucleus. In applying this to substitution and replacement in aromatic compounds it may be abbreviated as follows, where the + and—sign indicate the polarity of the hydrogen atoms:-  13 The Electron  Bond Theory.  In extending this theory to include "benzene, we must first apply it to the conjugated double bond. This has been done by Kermack and Robinson1, who produce strong arguments to show that the conjugated carbon atoms share three electrons in commen, instead of the customary two. This has been 2 developed into a workable benzene theory by Crocker and g Berliner . Berliner states to thus: "There is a ring of six carbon atoms, each singly bonded by pairs of electrons to its neighbor on either side, amd to hydrogen. The remaining six electrons are placed between the carbons in the plane of the ring, thus forming an ictet for each carbon atom. Substituents of the hydrogen would cause a shifting in the position of these latter six aromatic electrons, 'i'he direction and degree of this shift dependc on the electrical nature of the substituent." This may be clarified by an abbreviated formula.  The bars Represent the valence pair joining the carbon to the substituent, and the dots the "aromatic" electrons. Berliner assumes that when the carbon- hydrogen bonding pair is forced away from the nucleus the hydrogen is loosely 1. Kermack and Robinson. J. Chem. Soc. 121, 427. 2. Crocker. J. Am. Chem. Soc. 44. 1618. 3. Berliner. J. Phys. Chem. (1928), 293.  14 bound and thus reactive. if it be assumed that the hydrogen which is loosely held is positive, this formula is equivalent to assigning alternate positive and negative polarity to the hydrogens in the substituted benzene ring, although it does not do so for benzene itself. The electron bond has been applied to benzene in a slightly different way by Pauling1, and amplified by Lucas 2 . Using Knorr's idea that the bond consists of a double orbit surrounding both combined atoms, Pauling represents the formula for benzene as follows  this diagram merely represents a mechanism for the effects assumed in that of Berliner. For if a hydrogen is replaced, the orbit joining the substi&uent to the carbon atom is distorted - that id, either shortened or elongated. This may be supposed to affect the other orbits of th6 carbon atom, and so to influence directly the ortho and para position. Thus it may make these positions either more or less reactive than the meta position, although it has no direct meta connection.  1. Pauling. J. Am. Chem. Soc. 48, 1132. 2. Lucas. J. Am. Chem. Soc. 48, 1827.  15.  Summary of Theories. In summing up the different formula for benzene presented above, I have shown that they should be regarded as comple mentary rather than opposing. The shared electron bond theory jrovides a mechanism for the influences which are indicated by the positive and negative signs of the electronic theory. The only place in which the former theory does not agree with the latter is in assigning positive and negative charges to the hydrogen atoms in unsubstituted benzene, a point which is not uniformly agreed upon even by the supporters of the electronic theory.^In this paper I have interpreted my results and reviewed the literature by assigning alternate positive and negative oharges to the hydrogens in the benzene nucleus. Because of the similarity which I have shown  to  exist between the different formula for benzene I feel that this interpretation cannot be criticized by the supporters of either the electronic or electron- bond theory of valence. Application of the Theories. In proceeding to interpret benzene substitution and replacement by the above theory, it is necessary to consider the polarity of the commonly substituted groups. This has been done in a satisfactory manner by Fry, who used Holleman's "promotion-retardation" series as a basis. o Holleman measured the rate of substitution of various 1. Stieglitz. J. Am. Chem. Soc. 44, 1301. 2. Substituenten in den Benzolkern.(1910).  16. benzene derivatives with different reagents, and, in accordance with his addition theory, drew up a promotionretardation series, giving the relative effects of the different groups in influencing substitution. The "promotion" groups oause ortho-para products to be formed, while the "retardation"  radicals induee meta substitution. His series  is:-  OH,  HalogenS,CH>, (H), COOh, CHO, SQH,n. promote retard  Fry claims that "promoting" groups have a greater tendency to function negatively than positively, the strength of this tendency increasing in the direction of the arrow. The reverse is true of the"retarding" groups. This may be represented thus:-  OH, NHt, Halogens, C H3,  (H),  +  +  + CQQH, CHO,  t  H, NO,.  Also Fry shows that any group which enters the nucleus directly must do so by replacing a positive hydrogen. This may be illustrated by the case of nitric acid. It can only have the formula  • that is  HO—NO,  so that iJOg must  2  be put in as a positive group. This is further indicated by the fact that the hydrogen which is replaced unites with th<3 OHfljroupto form water. The same is true of sulphonating, While in the case of the halogens the hydrogen unites with the CL or Br of Cl"- CI or Br - Br to form HC1 or HBr, the positive halogen being substituted in the nucleus. It must be remembered, however, that once the group has been  IV; substituted it may take on its characteristic polarity if there are no other substituents of stronger polarity which prevent it from dGing so. Hence the electrical polarity of the substituent already in the nucleus determines the formation of either ortho and para, or meta derivatives when a second substituent enters the nucleus. To illustrate this principle of substitution, let us consider some familiar examples. When benzene is nitrated, the products are nitrobenzene, meta di-nitrobenzene, and 1 3 5 trihitrobenzene. Direct nitration will not produce any more highlj^itrated derivatives. As we have assumed that the nitro group tends to be positive, this reaction has the following electronic interpretation:JSST  No,  N4  - r > +  +  %  X v ^ W  (The heavy arrow shows the direction of the equilibrium.) On the other hand, on chlorinating benzene we get ortho and para dichlorbenzene.  cr  C|+  Again on nitrating toluene, as the CHg group is weakly characterized as negative, we get a mixture of ortho, meta, and para di-derivatives. Here the equilibrium between the two electromers of toluene has not a very pronounced direction,  18. so that the reaction may proceed with "both elegtromers. Similarly this theory may he applied to the great mass of data on benzene substitution. There is another application 6f the theory, which has so far received very little attention, namely the replacement of groups substituted in the benzene nucleus. The electronic theory predicts that the substituents in the nucleus have a polarity determined hy their own intrinsic tendency and that of other groups present. Mow it has previously been shown that no matter which valency theory is supported, a positive group can only be replaced by another positive one, and Vice versa. Hence in acting on aromatic compounds with KOH or NaOMe only negative groups should be replaced. It was with the purpose of investigating to what extent this is true  of  the aromatic halogen compounds that this investigation was undertaken.  B_;_ REVIEW . The Work of Holleman. In reviewing the work for previous or somewhat similar work, it was found that Holleman had investigated the problem of benzene substitution and replacement mast completely. While he does not agree with the theory of alternate polarity, much of his data can be explained and olarified by considering it from this point of view. Most of Holleman's work has been done by putting sub-  19. on  stituents into the nucleus rather thanAreplacements, and the experiments in both cases have been with the object of finding velocities, and thus comparing the relative strength of the different groups in directing a new substituent. However in the course of his researches, Holleman has made several determinations on the replacement of the halogens. These do not coincide with those required by the electronic theory, but as he was measuring velocities, and not the maximum replacement without decomposition, these values do not constitute a serious criticism of the theory. In many cases he has only succeeded in replacing one halogen atom where, from ray own experience, I believe that two or three could be replaced; While in other cases, where he has obtained higher values than the theoretical, from experiments in the laboratory at the temperatures at whiiSh he worked, I feel sure that decomposition must have taken .place. (a) Holleman refers in his paper1 to the experiments of DJagornow2 on the velocity of replacementof bromine in the bromonitrobenzenes by diethylamine. He quotes the percent replaced under ortho meta para  similar conditions as:- 78.Q°/o - 54.5u/o - 0 °/o  These results are quite compatible with the electronic theory. (b) Holleman and W. J. de Mooy  IX  report the following  replacements on treating the dichlornitrobenzenes with sodium methylate:1. Holleman. C.A. 10, 1509. 2. Jtfagornow. J. Russ, Phys. Chem Soc. 29, 699. 3. W.J. de Mooy. J. O h m Soc. 110, Ab I, 22.  20  Not  hX  cr  ci-  The other chlorine can be drawn out on prolonged treatment.  In conjunction with J. Ter We el,-^Holleman gives results of similar experiments with chlorodinitrobenzenes, which can be interpreted like those above. His general conclusion is that the presence of a nitro group in the ortho and para to a chlorine atom facilitates the replacement of the latter, while a nitro group in the meta position has the opposite effect. This conclusion is quite in harmony with the predictions of the electronic theory, (c) In the case of the trichlorbenzenes it is difficult to predict how they will replace, for there is a possibility of electronic tautomerism, thus: ci4 cr  a 1. Holleman and J. Ter Weel. J. Chem. Soc. 110, Ab I, 23.  21. This is borne out by Holleman's experiments1 on treating the trichlorbenzenes with sodium methylate.  As far as these replacements go, they are in accord with the theory. 2 (d) Holleman also has some data on the replacements in the trichlornitrobenzenes. He reports:-  1. Holleman. J. Chem. Soc. (1918) Ab I, 216. 2. Holleman. J. Chem. Soc. (1921) Ab I, 167.  As "before, these replacements, as far as they go, are favorable to the theory of alternate polarity. (ei Holleman has also performed experiments with C van der Hoeven and F.E. van Haeften on the tetrachlorbenzenes, pentachlorbenzene  and hexachlorhenzene. On  treating these with sodium methyl&te at 180° G they obtained the following results:-  23.  In the esse of the 1 2 4 6 tetra and pentachlorbenzene, where electronic tautomerism would alter the products formed, the electromer ahosen is considered the most probable because it gives the maximum number of chlorine atoms their characteristic negative polarity. That this assumption is justified is shown by the products which ^oiieman reports. In connection with these experiments Holleman heated these compounds with sodium methylate to determine the relative velocities of reaction. He obtained the following results 1 2 4 5 tetraohlorbenzene 1 2 3.4 tetraohlorbenzene 1 2 4 6 tetraohlorbenzene pentachlorophenol hexachlorbenzene  -  72% of total Chlorine liberated. 60,8% 83% 77.8% 71%  The fact that these values do not agree with the products isolated confirms my opinion that decomposition must have taken place. (f) Holleman has also performed some experiments on the nitration of the mixed dihalogen benzenes.1 In the case of para chlorobromobenzene, he found on nitration a mixture of 2 5 and 5 2 chlorobromonitrobenzenes. In the former the 1. Holleman. C.A. 9, 2385.  24. <rrtho-para directing influence of the chlorine predominate; i.e., the chlorine was negative, while in the latter the bromine was negative. On treating the 2 5 isomer, 98% of the ohlorine was replaced and the bromine left intact,while on similar treatment the 6 2 isomer gave up 98°/o bromine and no chlorine. This is very favorable evidence for the electronic theory. From the preceding discussion it is apparent that Hollemanfs work is not adverse to the electronic theory, but, as far as it goes, corroborates it. Where the replacements he records have fallen below the tbeoreticel value, he was not trying to obtain maximum displacement. On the other hand, where the replacements are above the theoratical value, I feel sure, from my own experience with the temperature at which he worked, that decomposition took place. Work of Other Men. Limbricht1 experimented dm the removal of bromine and the sulphonic group by hydrolysis. The following is an interesting comparison between the relative strenghts of the NHg and SOgH groups;S<?>H  1. ilimbricht. Ber. 10, 1541.  25.  So3H  SO}H  SOjH  NHJJ  This is an example of the replacement of positive halogen hy hydrogen. Prom this reaction it would appear that the amino group is a very strongly characterized negative group which is quite in accord with its position in Holleman's. series. Burton and Kenner 1 studied the removal of halogen atoms during the reduction of nitro-halogen benzene compounds to amino benzene compounds with hydrochloric acid and stannous chloride. They prove that the reduction to the amino compounds precedes the removal of the halogen, fhis makes an electronic interpretation of their work possible.  HCJ  n o  iodine  Z152  1. Burton and Kenner. J. Chem. Soc. 121, 675  replaced  £6. The fact that no iodine is liberated from the meta isomer may be accounted for by the assumption that the rate of reduction of the nitro to the amino group is a great deal faster than the rate of elimination of the positive iodine. The demonstration of the course of the above reaction by Burton and Kenner gives us the key to the explanation of the similar work by filanksma1. Blanksma investigated the reduction of nitro-halogen benzene compounds with tin and hydrochloric acid. He concluded that the halogen is eliminated only when it is ortho or para to the nitro group. As the amino compound is formed first, we must conclude that the halogen is replaced by hydrogen only if it is positive. Blanksma also stated that the elimination is facilitated if hydroxyl, amino, or methyl^roups are in the meta position to nitro group. This will have the effect of making the halogen more positive, i/typical reaction of Blanksma may thus be explained by the electronic theory:NO„ Br  S>f CH, B-i NO, A somewhat similar research is that of Shoesmitb and Slater* on the reduction of halogen phenols with hydriodic acid. They  1. Blanksma. Eec. Trav. Chim. 24, 320. 2. Shoesmith and Slater. J. Chem. Soc. 125, 1317.  27. found that the meta isomers will not reduce at all. We may account for this by the fact that the halogen meta to the hydroxyl group is negative, and cannot be replaced by hydrogen They also report that when the halogen is in the para position it is more easily replaced than when it is ortho. This conclusion is opposed to most of the evidence; on replacement. Finally these experiments show that the iodo compounds react quite readily, the bromo fairly readily, and the chloro with difficulty. This is in accord with the relative polarity of these atoms. By searching the literature it is possible to obtain numerous isolated reactions which may be interpreted very conveniently by the electronic theory. These have not been included in the review, because they are largely isolated cases, and, unlike the papers just dealt with, they do not draw general conclusions about the influence of the different positions on replacement. This review will be concluded with the mention of a very interesting set of papers which have appeared since the experimental work of this research has been completed. Nicolet and his collaborators1 adopting the electronic point of view, have investigated the removal of positive halogens from aromatic compounds by hydrolysis with hydrochloric acid and stannous chloride. In the compounds investigated the halogen was situated ortho or paramo the strongly negative  1. Nicolet. J. Am. Chem. Soc. 43, 2081; 49, 1796.  28. hydroxyl or amino group. These men found that only halogens situated in this way are removed and that many halogens so situated (particularly in the case of iodine) are readily found  removed quantitatively. They alsoAthat halogen atoms which we should expect to be negative are not replaced by this treatment. This is all in fine agreement with the electronic theory. However, flicolet proves a new point, which I believe is a very necessary modification of the. electronic theory; namely, that meta influenee plays practically no part in making a group reactive. We are accustomed to assume that a halogen atom meta to an electropositive group will react as positive halogen, while if it is meta to a negative group it will give the reactions of a negative halogen. This is called into question by the following two experiments of Uicolet, the first of which is an ideal case of meta positive influence and the second of meta negative influence  NH  Thi3 would seem to indicate that in predicting whether a certain group is replaceable or not we must consider chiefly the ortho and para influences. From the above review we may draw the conclusion that the replaceability of a group in the benzene nucleus may be pre-  29. dieted "by considering its polarity, but that in so doing we must take account of the fact that the meta position has little influence on reactivity,  C. EXPERIMENTAL PART, An investigation of maximum replacement in aromatic halogen compounds is reported by Clark and Crozier1. In this research I have continued along the same experimental lines, except that in trying to obtain maximum replacement without decomposition I have varied the strength of the reagent rather than the time of  heating and the temperature.  Preliminary Experiments. An attempt was made to replace the halogens in hexachlorbenzene by refluxing with strong alkali. The temperature reached variec^vith the strength of the reagent, ranging from 105' to 118°C. However, due to the low temperature, and the insolubility of the compounds in the reagent, the replacements recorded were negligible, and the method had to be abandoned. The compounds were next sealed in a pyrex combustion tube with sodium methylate as a reagent, and the tubes were heated in a tube gas furnace at a temperature of 150°to 155*C for thirty hours. This method gave quite satisfactory replacements, but caused considerable delay and risk due to the sealed tubes exploding. This was attributed to defects in the seal and to 1, Clark and Crozier. Trans. Royal Soc. Canada, 19, (1925),153  30  the large variation in the temperature with the changes in the gas pressure. Also, as the capacity of the furnace was four tubes, the method was very slow. In order to speed up the work and also to control the temperature, a large electric oven was obtained, equipped with a constant temperature device. This was losded with eighteen sealed tubes, each surrounded by an iron pipe to prevent the explosion of one tube from damaging the others, ©pposite the ends of the pipes was placed sheet iron, to protect the oven. Unfortunately, after the first twelve hours run there was a violent explosion, which destroyed the oven and ten of the eighteen tubes. It is believed that the constant temperature device was not sufficient to take care of the large increase of voltage during the night. Final Method. The apparatus  finally used for this work was a large  autoclave made to withstand a pressure well over two hundred pounds. It was cylindrical, about two feet long and ten inched in diameter, with a detachable cap on one end holding a lead seal. The temperature was measured by an oil well, which projected about two inches within the cylinder. The machine was heated by gas flames, so that in order to hold the temperature constant it was necessary to keep the gas pressure constant. This was done by leading the gas from the tap through a reservoir where it was kept under a constant head of water. To accomplish this the flow of gas into the  31. reservoir was controlled by a ground glass valve whose stem held a bob floating in the water. The gas from the tap to the reservoir, and from thence to the burners was controlled by screw-cocks. By this means it was possible to keep the temperature of the autoclave constant to within two degrees by inspecting it egery two or three hours. The experimental procedure was as follows: From 0.3 to 0.5 gran^bf the halide compound was sealed in a pyrex combustion tube with 30 c.c. of sodium methylate reagent. This reagent was prepared by dissolving sodium in absolute methyl alchol and titratingagainst standard acid. From twelve to fifteen tubes were heated in the autoclave for fifty to sixty hours at 155*G. To counteract the pressure within the tubes, Methyl alchol was placed within the autoclave, therby setting up a counter pressure on the outside of the tubes of from 150 - 180 pounds. This prevented all explosions. After heating, the tubes were cooled and opened, and the contents washed out into a beaker, the total volume of the reaction mixture and washings being about 300 c.c. This solution was made acid with nitric acid and boiled to expel the methyl alchol. After cooling it was filtered to remove any undissolved organic matter, and the halide in the filtrate precipitated with tenth normal silver nitrate solution. On digestion the silver halide coagulated and was filtered out. In order to free this precipitate from organic matter, it was dissolved in concentrated ammonium hydroxide and  32. reprecipitated by adding nitric acid. After this precipitate had coagulated it was filtered through a tared Gooch crucible, dried at 120° c for forty minutes, desiccated and weighed. The percent of halogen replaced was calculated from the weight of silver halide and the weight of the total halogen in the compound used. Results. The experimental results are given in the following table. Under theoretical replacement is given the percentage of halogen that might be expected to be replaced, according to the electronic theory, from the more stable electromer. fhe normality of the alkali employed is given in column three, while column four records the replacement found in percent of total halogen present. Under remarks are given any results recorded on the same compound by previous investigators . COMPOUND  (1) Phenyl chloride.  (2) O-dichlorbenzene.  (3)  m-dichlorbenzene  Theor- Norm% etical ality halogen reReof place- NaOMe.placed. ment. 100%  2.0 2.5  3.6 7.4  50  1.5 2.0 2.5  5.34 7.1 12.3  100  1.5 2.0 2.5 3.0  6.2 18.2 26.2 33.0  REMARKS.  Blau, Monatshefte 7, 622, 635. states phenyl bromide yields some phenol and anisole. Holleman, Abst. J. Am. Chem. Soc. 10, 1509 reports O-chlorphenol. Holleman, loc. cit. reports mchlorphenol.  33.  Compound.  [Theor-Norm % etical ality halogen Re- Of replace-NaOMe placed. ment.  (4) p-dichlorbenzene... (5) 1,2,4, trichlor benzene...  50 66  1.5 2.27 6.2 2.5 3.0 11.9 1.42 37.8  100 (6) sym-tribrombenzene...•  1.5 2.0 2.5  32.5 43.5 46.2  (7) 1,2,4,5, tetra chlorbenzBne..  50  0.6 1.42 1.5 2.0  23.7 40.6 55.4 61.0  (8) hexachlorbenzene....  50  1.5 2.0  57.0 85.4  (9J o-chlorphenol..  0  1.5 2.0 1.5 2.0 2.5 2.0 1.5 2.0 1.5  0.0 0.0 12.a 11.7 10.9 0.0 9.1 13.5 52.3  2.0 2.5 1.5 2.0 2.5 2.0 2.5 1.5  34.9 49.5 38.4 69.3 68.2 1.7 2.7 14.4  1.5  0.0  1.5 1.7 1.5  14.0 24.5 0.0  2.0 2.5  65.2 70.5  (10) m-chlorphenol. .100 (11) p-ehlorphenol. 0 (^.2) 2, chlor, 5, b y o droxytoluene.. • (13) 3,5, dibrorn100 o-cresol (14) tetrabrom-050 cresol.... (15) pentabrom40 phenol...• (16) 3, brom, 4, aminotoluene.. (17) o-brom benzoic acid....... (18) m-brom benzoic acid (19) p-brom benzoic  0% 100 0 or 100 100  (20) 2,5, dichlor50 aniline.... (21) 2,5, dichlor50 nitro-benzene. •  Remarks.  Holleman, loc. cit. reports p-chlorphenol. Holleman, Abst. J. Am. Chem. Soc. 13,569 reports 2,5 dichlorphenol and a displacement of 27.7% chlorine. Holleman, loc. cit. reports symdichlorphenol and a displacement of 57.4% chlorine. Holleman, Abst. J. Am. Chem. Soc. 15,705 reports 2,4,5, trichloranisole and a 77% displacement of chlorine. Holleman, loc. cit. re« ports pentachlorphenol and 71% displacement of chlorine.  4-chlor, 2 nitro-anisole-Chem.Zent.I,112.  34. Discussion of Results. Sn^nterpreting these results we must watch for two points. In the first place, we must consider whether or not we have maximum replacement without decomposition. This offers considerable difficulty, for on increasing the strength of the reagent the replacement value becomes greater, but it is difficult to tell whether this is due to actual replacement or to decomposition of the compound. It is evident that, particularly with the more highly halogenated compounds, the slightest decomposition will raise the replacement value considerably. In the second place, we must consider to what directing group we are looking for our polarity, whether it is ortho, para, or meta. If we find that the negative sign on the halogen comes from a meta negative group, the halogen will not be active as a negative substituent, and the replacement will be difficult. In applying these considerations to the compounds in the table, it is seen that: fa) In the case of compounds 3,5,6,7,8,10, the influencing group is CI or Br in the meta position, which would not make the halogen replaceable. (b) In the case of compounds 2,4,5,6,7,8,14,20,21, the influencing group is CI or Br in the ortho or para position, which wecannot expect to have a strongly activating influence because it has not its customary polarity. (c) In compounds 9,11,12,16, the influencing group is OH or  35. HH2 in the orthe or para position, which we would expect to make the halogen strongly positive and not replaceable. This is upheld by experiment, except for 2 chloro 5 hydroxy tolh&ve  uene, where we A a replacement incompatible with the theory. (d) In compounds 10,13,14,15, the influencing group is OH in the meta position, which, due to its position, will not make the halogen easily replaceable. (e) In the case of the brom benzoic acids, as the carboxyl group is weakly characterized, we cannot expect it to have a strong activating influence. The replacements uphold this. (f) In the case of 2,5, dichloramline the amino group, because of its meta position, will not make the halogen replaceable, as the experiments prove. The above explanations account very well for the halogen atoms which have been found to be difficult to replace. But there are certain cases where the replacement is higher than the theoretical value. Here we must assume either that the theory is untenable, or that a certain amount of decomposition took place. The detection of the latter is extremely difficult and in the cases of the tetrachlorbenzene, hexachlorbenzene, or pentabromphenol, a small amount of decomposition would increase the value greatly. Hence, in spite of the lack of experimental evidence, I believe that in these cases it is permissable to assume that a certain amount of decomposition has taken place and raised the value accordingly. In the case of 2 5 dichlornitrobenzene, decomposition may be considered very likely, as the color of the reaction mixture id so dark  36. that it would he almost impossible to detect a small deposit of carbon. It is evident that the aromatic halogen compounds which contain negative halogen are not as a whole, suitable for measurements of maximum replacements, particularly those which depend upon negative groups to determine the polarity. This is largely due to the position of the halogens near the centre of Holleman*s series. The most likely compounds fofc these measurements are those in which the negative halogen atom is ortho or para to strongly positive group such as the nitro group. In conclusion, the measurements tabulated in Part II of this paper do not provide any conclusive verification of the electronic theory. They indicate instead that the meta position plajrs very little part in making a negative halogen atom replaceable. I believe that this latter ppint is a very necessary modification to the electronic theory.  37. PART III. A SEARCH FOR THE HALOGEN CATION. A.  INTRODUCTION. Positive Halogen.  It has been shown in Part I that, no matter whether the electronic or the electron- "bond theory of valence is considered, when the chemical hond is broken it must become polar. During reaction, therefore, a positive group can only be replaced by a positive, and a negative by a negative group, unless oxidation or reduction takes place. Now it is generally oonceded that in hydrolysis hydrogen is positive, as the hydrolized group usually appears in combination with the hydoxyl radical. It is on this fact that the chemical evidence for positive halogen is based. The halogen atoms in certain organic compounds have the property of being replaced on hydrolysis by hydrogen, and appearing combined with the hydroxyl  group as hypohalous acids. In these compounds it  is claimed that the halogen is positive. Hence the criterion for positive halogen is that it shall be eliminated on hydrolysis as hypohalous acid. Conversely, in order to show that these compounds contain positive halogen it is necessary to show that hypohalous halogen is electronically positive. The reasons for believing this ane numerous and quite convincing. Let us take hypochlorous acid as an example. In the first place it is very reactive and unstable. In acid  38. solution it immediately changes to chloride and chlorate. It is only stable at low temperatures and in alkaline solution. Under these circumstances, since there is no negative ion other than hydroxyl for the positive chlorine to unite with, the solution will keep for some time. In the second place hypoohlorous acid is an oxidizing agent, and its oxidizing power is what we would expect of positive chlorine. Its mutual oxidation-reduction in acid solution to chloride and chlorate has been mentioned. This may be represented as follows:-  _ -t  - +  . /Ja -M,^"  2HOCi + HOC I — > 2 HCi -t HO -Cu_o  One of the molecules has been oxidized to chlorate, gaining four positive valences; i»e., losing four electrons. These electrons have been gained by the two molecules which have been reduced to chloride. Therefore we may write,  CI +2 electrons = Ct ,  -  d + © *C1 , C\ + e>*Cl .  Thus according to the electronic theory hypochlorous acid must contain positive halogen. In the third place hypochlorous acid is a chlorinating agent. In fact many authors attribute the chlorinating power of chlorine itself to the formation of hypochlorous acid. This has been proved for iodinating by Soper and Smith1 . From the velocity constants of the action of iodine on phenol, these men conclude that hypoiodous acid is the iodinating agent. Thus there is quite good evidence that the halogen in hypohalous acid is positive.  1. Soper and Smith. J. Chem. Soc. Nov.(1927), 2767.  39.  There is , however, an even more direct proof than anymentioned above. Jakowkin1, in studying the action of chlorine on water, proved that it is a reversible reaction, and that it is ionic so far as the HCl is concerned.  HOCl + HCl  2  CK.  This result is discussed by Stieglitz , who claims that, since the reaction has been proved ionic for HCl, and is recognized as being ionic for water, the hypochlorous acid _  -+  -  and chlorine must both be ionized also into HO and CI and CI  -h  and CI. Hence he claims that the halogen in hypochlorous acid is positive, and that the positive halogen ion can exist. Positive Halogen Compounds. I shall now proceed to give a brief account of the compounds in which we find positive or hypohalous halogen. 3 Howell and W. A. Noyes  consider positive #alogen as that  which replaces hydrogen in the OH, NH£, NH and HC =CH groups. Evidence has been already been discussed which proves that certain aromatic halogens are positive. It is further believed that one of the halogens in carbon tetrachloride tetrabromide and tetraiodide is positive; and that in compounds of the halogens with themselves one is positive and the other negative. The reasons for these conclusions will now be given, (a) Type HOX - Besides the hypohalous acids, there belong to this class the hypochlorites of the type T? - C^ QC j 1.Jakowkin. Zeit. Phys. Chem. 29, 613. 2. Stieglita. J. Am. Chem. Soc. 23, 797. 3. Howell and W. A. Noyes. J. Am. Chem. Soc. 42, 991.  40. prepared "by Schutzenberger^". These hydrolyze to hypochlorous acid and are unstable and difficult to prepare. (b) Type -HZ. This is the largest class of positive halogen compounds. To it belong nitrogen trichloride and triiodide and the halogen substituted amines and amides. Both the methods of making these compounds and their reactions indicate that they contain positive halogen. Nitrogen tri-iodide has been fully investigated by Chattaway2. It may be made by ammonia and iodine, by ammonium hydroxide and iodine chloride, and by ammonia and an alkaline solution of potassium hypoiodite.The last two methods in particular point to positive iodine. It has slso been 3 shown quantitatively  for the third of the above reactions  that the nitrogen iodide is formed at the expense of the KOI in the solution. AS for the teactions of this compound, Chattaway reports that it is explosive, very sensitive to light., gives ammonia and hypoiodous acid on hydrolysis, $s a strong oxidizing agent, and as such it exerts per atom of iodine twice the oxidizing power of free iodine. Thus there can he no doubt that this compound contains positive iodine. The properties of the halogen substituted amines and amides have been studied by Selivanow4. He indicates that the halogen is joined to the nitrogen, by calling these compounds halogenyl amines and amides. These compounds are all 1. 2. 3. 4.  Schutzenberger. Ann. 120, 113. Chattawpy. Am. Chem. J. 23, 363. Chattawy and Orton. Am. Chem. J. 24, 324. Seliianow. Ber, 25, 3617; 26, 423.  41. made either by the action of hypohalous acid direct, or by adding halogen slowly to the cold alkaline solution of the amine, amide or imide. These halogenyl compounds readily exchange their halogen for hydrogen, hydrolyzing with water _to form hypohalous acid. Their solutions have  strong ox-  idizing and halogenating properties. Hence they are compounds of positive halogen. (c) Miscellaneous cases:- There are several detached cases of positive halogens, (l) Because of its oxidizing power, and other peculiar properties, Howell and Noyes 1 claim that the iodine in di-iodo acetylene is positive. (2) Also Jones and 2 Werner  produce evidence to show that one halogen atom in  carbon tetrachloride, tetrabromide and tetraiodide is positive and different from the rest. This is further shown by the work of Nef^ who found that each of these three compounds on hydrolysis with alkalies gave chloroform, bromoform and iodoform as the primary products. This indicates that one halogen atom is positive. (3) Again, it is generally concluded that the iodine in iodine chl&ride and iodine bromide is positive. An example of this is their well known iodinating properties, as well as their ready hydrolysis. (4) Finally, from hydrolysis evidence  4  it is believed that the  iodine in iodocyanogen is positive. Except for the aromatic  1. 2. 3. 4.  Howell and Noyes. J. Am. Chem. Soc. 42, 991. Jones and Warner. J. Am. Chem Soc. 40, 1257. lief. Ann, 308, 173. Chattaway and Wademore. J. Chem. Soc. 81, 191.  42 compounds, this concludes the chief organic "bodies containing positive halogen. If the halogen in these compounds is electropositive, as we have supposed, it should be possible to isolate it as a positive ion on electrolysis. But in spite of the frequent references in the literature to positive halogen, there are only four investigators who have reported halogen at the cathode, and thfeir conclusions are disputed. Hence it was decided to undertake a search for the halogen cation among the different compounds containing hypohalous halogen. B. Review. The first case of cathodic halogen was reported in 1827 hy de la Rive 1 . On passing an electric current through an aqueous solution of iodine bromide, containing starch as an indicator, de la Rive noticed color changes around the electrodes. As he obtained "blue at the cathode and orange at the anode he concluded that iodine was separating as a positive ion and "bromine as negative. 2 We also find that Faraday  in 1834 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. A similar experiment was performed by Solly2 in 1836 on the electrolysis c£ dry liquid iodine chloride with platinum 1. de la Rive. Phil. Mag. (2), 3,86, 145. 2. Faraday. Ostwald's Klassiker. 45. 3. Solly. Phil. Mag. (3), 2, 130.  43. electrodes. He reported that the "anode was corroded but clear and that the cathode was encrusted with a black matter very much like iodine in appearance." The results of these experimenters are questioned by Bruner and Bekier^. These men 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 Kathode is rendered exceedingly plausible." A further search for the halogen cation was undertaken 2 by Bruner and Galecki . They studied the electrolysis of pure iodine, bromine, chlorine, iodine bromide and iodine chloride between silver electrodes in sulphur dioxide and in o nitijbenzene as solvents. They found that the elementary halogens in these dry solvents are non conductors, while iodine bromide and chloride are poor conductor®. Ad these latter two compounds were such poor conductors, these experimenters performed their electrolysis without the exclusion of moisture, in order to increase the conductivity. Hence we are not surprised to find that all the halogen precipitated at the anode. I have found no other references to the electrolysis of  2)  positive halogen compounds except that of Stieglitz .In this  1. Bruner and Bekier. Zeit. Elect. (1912) 18, 368, 2. Bruner and Zeit. Phys. Chem. (1913) 8£ 3. Stieglitz. J. Am. Chem. Soc. 23, 797.  513.  44. paper he expressed his "belief in the existence of the halogen ion and his intention of searching for it by the electrolysis pf the hypochlorites. As he has reported nothing further, we may assume that his efforts were unsuccessful. C. Experimental Part. The first compound experimented upon was carbon tetrabromide. As it had an extremely high resistence both in the molten state and in solution, conductivity measurements were resorted to. Solutions of carbon tetrabromide in ethyl alco- . hoi, carbon disulphide, nitromethane, nitrobenzene, acetone, acetonitrile, and ether were prepared, and their conductivity compared with that of the pure solvent. As no appreciable increase was noted in the conductivity of the solution over that of the solvent, it was decided not to investigate the compound further. The next compound chosen was bromyl acetamide. It was made by Hoffinann's method^" and recrystallized from benzene. The product had a melting poin£ of 104° . It was first tried in purified acetamide as a solvent at 100° C, with graphite electrodes. The color changes and evolved gases were observed on . passing 110 volts through a five per cent solution. This method did not produce any conclusive evidence, because the halogen liberated is so reactive that none of it is evolved as a gas, and hence its detection is impossible. To overcome this difficulty silver electrodes were used. Bruner and 1. Hoffmann. Ber. 15. 412.  45. Golecki1 had previously shown that the halogen in a solution may he completely removed as silver halide by using silver electrodes. In this way the following results were obtained with bromyl acetanide: fa) With acetamide as a solvent at 100° C and with silver electrodes a quarter of an inch apart, a white non adherent precipitate was formed at the anode with 6 volts, which appeared to be silver bromide. (b) With methyl cyanide at room temperature- In a cell with very close parallel silver electrodes, on applying 6 volts a heavy white precipitate formed on the cathode and dropped to the bottom. No precipitate could be seen forming on the anode, but the electrodes were so close that it was difficult to tell. In order to verify this result the experiment was repeated in a V  cell with a large bore stop-cock in the  middle. The electrodes were about half an inch apart, and it was necessary to apply over 100 volts to get an appreciable precipitate. This time the precipitate of silver halide formed at both electrodes, heavier at the anode. (c) With nitro methane the precipitate under all circumstances formed at the anode. The compound which proved most successful in this work was Vromyl suocinimide. This was made by the method of Lengfield 2 and Stieglitz . It was found necessary to modify their method slightly. The procedure id as follows: 20 grams of succin1.Bruner and Galecki. Zeit. Phys. Ohem. 84, 513. 2. Lengfield and Stieglitz. Am. Chem. J. 15, 215.  46.  imide is dissolved in a solution of 8 grams of sodium hydroxide in 50 c.c. of water, and cooled in an ice hath. To this is slowly added 32 grams of bromine, which precipitates the bromyl succinimide. The reaction mixture is filtered hy suction and the precipitate placed on a porous plate to allow the excess bromine to volatilize. In order to extract the product from this filter cake it is necessary to wash it out with 50 c.c. of water until it breaks up and becomes a mass of glistening crystals. This is then filtered and dried and extracted with benzene. For the preliminary experiments this was recrystallized twice from benzene, giving a product of melting point 170°to 172°C. In the following experiments with this compound the solvents were very carefully dfied and purified, and the apparatus cleaned, dried at 120" and desiccated. Silver electrodes were used throughout. (a) With furfural as a solvent no precipitate took place at either electrode. There was a color change on adding the compound, and it is believed that the compound and solvent interacted. (b) In acetamide a white precipitate formed at the anode only. (c) In nitro methane the precipitation took place entirely at the anode. fa) In nitrobenzene the halide precipitated formed at the anode only. (e) In glycol also oonly anode precipitation took place, (f) In methyl cyanide- Preliminary experiments were tried with parallel silver electrodes at a distance of half an inch  49. with a potential difference of from 6 to 10 volts. In these experiments there was a certain amount on non adherent precipitate at the anode, "but mo4t of the precipitation took place at the cathode, forming as a heavy stringy mass, hanging from the electrode. This precipitate gave all the reactions of silver "bromide, being soluble in concentrated ammonia and potassium cyanide solution, and precipitated from both on the addition of nitric acid. In performing this experiment on a larger scale much difficulty was experienced with the formation of a precipitate at the anode. This always appeared first, and, as it was non adhereni^nd diffusedt through the solution, it made it difficult to determine for certain that halogen was precipitating at the cathode, and not reaching it by diffusion. At the time of this trouble the bromyl succinimide had been made for about a month, and it was thought that the anode precipitation might be due to halide coming from the decomposition of the compound. As the ihelting point had fall* en from 172° to 166°C • this is probably the correct explanation of the trouble. Therefore a fresh lot of the compound was made and recrystallized six times from benzene to bring its melting point up to 172*C. The cell used for the final experiments was a V c e l l with a stop-cock in the middle. In order to prevent diffusion of any non adherent precipitate, a loose cotton plug was placed in the bore of the stop-cock. As this plug increased the resistence considerably, precipitation took place only very slowly with a voltage of 110.  48. The solution was  0.15 molae, and the length of time 15 hours.  The electrodes were of silver, about one halflinch apart. Tb» precipitates in the^-final experiment (which was duplicated without any difficulty) were as follows:- f l ) On the anode were deposited well formed transparent crystals very much the shape of sugar granules. This precipitate is insoluble in water and ammonium nitrate solution. Therefore it cannot be silver cyanide. It is soluble in concentrated ammonia but does not reprecipitate on adding nitric acid. Therefore it cannot be silver bromide. It is soluble in potassium cyanide solution and reprfccipitates in voluminous white curds on adding nitric acid. It gave no bromine vapor on heating with concentrated sulphuric acid and manganese dioxide. Therefore, although this preoipitate has not been identified, it may be concluded that it does not contain bromine. (2) On the eathode was piled a heavy grey mass of non crystalline material, very similar to the stringy material observed on the cathode in the preliminary experiments. This precipitate is soluble in concentrated ammonia and potassium cyanide solution, add reprecip itates from both in heavy yellowish curds on neutralizing with nitric acid. It also liberates bromine on heating with concentrated sulphuric acid and manganese dioxide. Hence the cathode precipitate  is silver bromide.  D. Discussion of Results. In the case of carbon tetrabromide it is not difficult to see why it would give non conducting solutions. In the first  49 place, although it forms hromoform in hydrolysis, no one has detected the formation of hypohromous acid in this reaction. In the second place, carbon tatrabromide is HO symmetrical that we cannot expect it to be polar. With bromyl acetamide the results were not at all conclusive, although they were indicative of positive bromine. Prom these indications I feel sure that, if the compounds had been purified as carefully as the bromyl succinimide, this compound too would have precipitated bromine entirely on the cathode. As it was, the compound changed color quite quickly, and we are justified in assuming that considerable decomposition had taken place before the experiments on it were completed. With the bromyl succinimide there can be no doubt that the halogen was precipitated on the cathode. In case this result may be criticized because of the high voltages used, it may be well to point out that the same products were formed in the trial cell with 6 - 1 0 volts. Unfortunately* in order to investigate the products formed, the stop-cock and plug had to be introduced, thus raising the resistence. Therefore, because of the results of the trial experiments, it cannot be argued that the high voltage decomposed the compound. Finally, whether or not this precipitation is due to the halogen cation is still open to question. It is hoped to identify the product precipitated on the anode, and so determine the course of the electrolysis. Until the actual  50 current carrier is determined no claim can "be made for the . isolation of positive bromine. This is particularly so, because the only solvent in which the electrolysis goes in this direction is methyl cyanide. Since the other solvents tried have approximately the same dielectric constant, this difference must fee due to some special property of methyl cyanide.However, the author wishes to point put that even if the compound forms an addtion product with the solvent, such + Br as CH — GN^l  , the only way in which the bromine can reach  the cathode id for the succinimide to become a negatiiie ion. In this way even that interpretation would indirectly assume the existence of the positive halogen ion. In conclusion, the author wishes to gratefully acknowledge the many helpful suggestions and the keen interest of Dr. R. H. Clark in this research.  BIBLIOGRAPHY  Berliner, J.F.T.- Journal of Physical Chemistry, 32(1928), p. 293. Blanksma . -  Recueil des travanx chimiques des .  Pays.Bas. 24 (1905) p. 32G,  Bruner and Bekier - Zeitschrift fur Electrchemie, 18(1912) p. 326. Bruner and Galecki - Zeitschrift fur Physicalische Chemie, 18 (1913) p 513. Burton and Kenner - Journal of the Chemical Society, London, 121 (1922) p. 675. Chattawjay- American Chemical Journal. 23 (1900), p. 363. Chattaway and Orton - ibid. 24 (1900), p. 342. Chattaway and Wadmore - J. Chem. Soc. 81 (1902), p. 191. Clark and Crozier - Transactions of the Royal Society of Canada, 19 (1925), p. 153. th Cohen - Organic Chemistry, Vol.1, 4 , edition. Crocker - Journal of the American Chemical Society, 44,(1922) p. 1618. Cuy, E.J. - J. Am. Chem. Soc. 42, (1920), p. 503. Falk and Nelson - ibid. 32, (1910), p. 1637. Faraday, M. - Ostwald's Klassiker, 86, (1834), p. 43. Flurscheim - Journal furpractische Chemie. 66, (1902),p.321. ibid. 71, (1905),p.497. Hofmann - Berichte der deutschen chemischen Gesellschaft. 15, (1882), p. 412. Hilleman, A.F., Substituenten in den Benzolkern. (1910). Rec. Trav. Chim. 33, (1914), p.l. Chemical Abstracts, vol. 9, p. 2385. ibid. 10, p. 1509. ibid. 13, p. 569. ibid. 15, p. 1705. Abstracts, Jour. Chem. Soc. London,(1918),pt.I p. 216. ibid. (1921), ptl, p. 167.  II. Holleman and de Mooy -  ibid.  Holleman and Ter Weel -  ibid.  vol. 110.(1916) pt. I, p. 22. vol. 110.(1916) pt. I, p. 23.  Howell and Noyes - J. Am. Chem. Soc. 42 (1920), p. 991. Huggins -  ibid.  44 (1922), p. 1607.  Jakowkin - Zeit. Phys. Chemie. 29 (1899), p. 613. Jones, W.L. - Am. Chem. Jour. 50 (1913), p. 414. Jones and Werner - J. Am. Chem. Soc. 40 (1918), p. 1257. Kermack and Robinson - J. Chem. Soc. London, 121,(1922) p.427. Knorr -  Zeitschrift fur anorganische Chemie, 129(1923)p. 109.  Langmuir- J. Am. Chem. Soc. ibid.  38 (1916) p. 222. 41 (1919) p. 868.  Lapworth - J. Chem. Soc. London, 121 (1922) p. 416. Lewis. G.N. -mValence and the Structure of Atoms and Molecules, (1923), p. 71. J. Am. Chem. Soc. 38 (1916) p. 762. Lengfield and Stieglitz - Am. Chem. Jour. 15(1893), p. 215. Limbricht - Berichte  10 (1877), p. 1541.  Lowry - J. Chem Soc. London, 123 ( 1923) p. 822. Lucas and Jameson - J. Am. Chem. Soc. 46 (1924), p. 2475. ibid. Nef -  Liebig's Annalen -  Nicolet - J. Am. Chem. Soc.  Pauling-  47 (1925), p. 1459. 308  (1899), p. 173.  43 (1921),  p. 2081.  ibid.  49 (1927),  p. 1796.  ibid.  48 (1926),  p. 1132.  Rive, A. de la - Phil. Mag. (2) 3, (1827), p. 145. Selivanow -  -  Berichte  25 (1892),  p. 3617.  ibid.  26 (1893),  p. 423.  Schutzenberger - Annalen  120 (1861), p. 145.  Shoesmith and Slater - J. Chem. Soc. London 125 (1924),p.1317 Stieglitz -  Am. Chem. J.  18 (1896), p. 756.  III. Steiglitz - J. Am. Chem. Soc. ibid. Solly -  Phil. Mag.  23 (1901), p. 797. 44 (1922), p. 1301. (3) 8,  (1836), p. 130.  Soper and Smith - J. Chem. Soc. London. (1927), p. 2757. Vorlander Weinberg -  Berichte. ibid.  52 (1919), p. 263. 52B (1919), p. 1501.  


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