DE POLSR AND NOK-POLAR VALENCY. 127 XIV.-Polar. and Norc-polas. VaZeracy. By RAJENDRALAL DE. IN a paper recently published Briggs (T. 1917 111 267) asks: “What is the valency of cobalt in chloropentammine cobaltic chloride C 0 F H s CI,?” This question has suggested the view set forth in the present paper. From the study of the optical properties of the tetraethylene-diamminep-aminonitrodicobaltic salts, 0 en,Co/ 'been X, \NO,/ NH 1 Werner (Ber. 1913 46 3674) concludes that there is no essential difference between the principal and auxiliary valency bonds. Evidently this conclusion may be applicable only to the bonds within the complex radicle (that is bonds employed in uniting radicles co-ordinated with the cobalt atom). The bonds outside the complex radicle are employed in combining radicles which exist as ions.Their nature is thus entirely different from that of the former ones. A distinction ought therefore to be maintained between the bonds outside the complex radicle and those within it. We can find an explanation of the valency outside the ccunplex radicle which is polar in type from Sir J. J. Thornson’s theory of valency. To understand the mechanism of it let us picture the structure of an atom derived by Thomson. According to him the atom consists of corpuscles moving in a sphere of uniform pmitive electrification and its valency depends on the ease with which corpuscles can escape from or be received by the atom. Difficul-ties however arise in explaining the valencies within the complex radicle in the above manner they being non-polar in type.During the disintegration of radioactive substances the negative charges of electricity are carried by &rays and the positive charges by a-rays. The &rays consist of expelled particles-not atoms of matter but free atoms of negative electricity or “ electrons.” An a-particle however consists of two atomic charges of pcrsitive electricity combined with a helium atom-a substance inert in the chemical sense. It may therefore be assumed that the pmitive electricity can have an attraction for the mass itself even if there be no charge of negative electricity on it. Thgmson (“Rays of Positive Electricity and their Application to Chemical Analysis,” p. 40) also observes that molecules with positive! charges are quit 128 DE POLAR AND NON-POLAR VALENCY.co8mmon whilst those witah negative charges of elect~city are very rare. This property which the positive electricity possesses affords an explanation of the phenomena of the valencies inside the corn-plex radicle of a complex salt. I n order t o explain the phenomena of the above valencies we shall conside'r Rutherf o'rd's view as regards the constitution of an atom. According t o him a positive nucleus is situated in the centre whilst electrons move around it in various concentric rings. We shall conceive this nucleus as having a binding capacity for the radicles which am co-ordinated with a metallic atom in the case of complex salts. It is significant that no positive radicles, such as ammonium tetramethylammonium etc.which can exist as cations have been observed t o combine with a metallic atom forming a complex radicle. Negative radicles such as C1 (chloro-), NO (nitro-) etc. however do form a complex radicle with a metallic atom. These1 negative radicles also carry negative charges of electricity when they exist as anions. Let us form a picture of the mechanism as conceived above. We have the positive nucleus of the metallic atom (capable of forming a complex radicle) in the centre and around it there are various concentric rings along which the electrons move. We may assumel that adjacent t o the outermost ring of electrons constituting the atomic structure there are the neutral molecules f o r example, NH, H,O etc. or the negative radicles f o r example C1 NO,, etc.or b0t.h these neutral and negative radicles held by the influence of the positive nucleus of the metallic atom concerned. Accordingly in the case of tetraethylenediamzine-p-aminonitro-dicobaltic salts radicles within the complex radicle may be sup-fmed to be1 abtached to the positive nuclei of its cobalt atoms and thereby the valencies within the1 complex radicle being taken to be all alike the two cobalt atoms become linked to1 the two groups in the middle namely NO and NB, in a similar way. The conditions favourable to the formation of complex salts may now be stated. The number of concentric rings in the struc-ture of an atom gro~ws large as the atomic weight increases and, thereby the structure also becomes more complex According to Rutherford however (Soddy " The Chemistry of the Radio-elements," 1914 Part 11 p.39) the mass of an atam is concen-trated in an exceedingly small central nucleus. Hence with the decreasei in the atomic volume only i.he rings will decrease in size, and the outermost ring will approach nearer to the nucleus. We have already supposed the radicles co-ordinated with a metallic atom to be placed adjacent to its outermost ring of electrons and also bound by its nucleus. Evidently t.he attracti-on of the nucleu D E POLAR AND NON-POLAR VALENCY. 129 fo8r the mass of the radicles would increase where there are possi-bilities of their being placed near to the nucleus that is to say, where the atomic volume is small. I n fact metals that. are cap-able of forming complex salts as for example chromium iron, manganese cobalt nickel copper ruthenium rhodium palladium, osmium iridium platinum gold etc.are situated on the troughs of Lothar Meyer’s atomic volume curve. Here it may also be mentioned that Ephraim ( B e y . 1912 45 1322; 1913 46 3103; 1914 47 1828; Zeitsch. physiknl. Chenz. 1913 81 513 539; 83 196) from his st’udy of the strength of the auxiliary valencies of various metals has drawn the conclusion that the strength of the auxiliary valencies falls with the increase of the atomic volume of the metal concerned. We thus find a further support f o r the abo’ve assumption. It will be observed that the metals which form complex salts are mostly found both in the (‘ ous ” and the (‘ic’’ state of their ionic condition as for instance we have Cr” (chromous) Cr”’ (chromic) Co” (cobaltous) Co”’ (cobaltic) etc.It may also possibly be that the “ous ’’ condition of the metallic ion is more favourable to the formation of a complex radicle. During the reduction of a metallic ion from the “ i c ” to the “ ous ” state there is an alteration in the electric charge of its rings and the proba-bility is that this alteration is confined to the outermost ring (Soddy ibid.). Evidently in the (‘ ous ” condition of the metallic atom there is a less number of electrons in its outermost\ ring. Keeping in view the structure of an atom it would be natural to expect that ordinarily the outer rings of electrons would offer themselves as a shield against the attraction of the positive nucleus f o r thel radicles which may be co-ordinated with a metallic atom.The case is however different in its ‘( ous ’’ state for there being produced a weakness in the shield due t o a less number of electrons in the ring the attraction of the1 nucleus will obtain an oppor-tunity of manifesting itself by forming a complex radicle. It is known that chromic chloride has to be reduced t o the chromous state for the preparation of chrom-ammonia salts (C’hristensen, J . pr. Chem. 1881 [ii] 23 54). Similarly in the preparation of cobalt-ammonia and platinum-ammonia compounds (Gerhardt, Annalen 1850 76 307) the starting materials are the ( ( ous” salts of the metal concerned. Last-ly it is found that the formation of a complex anion is a more general phenomenon than the formation of a complex cation ; for example there’ are compounds of the type1 [M(C,O,),]R, where M may be Vd Cr Mn Fe Co Rh R1 As Sb or Bi (Werner, (‘New Ideas on Inorganic Chemistry,” p.116 ed. 1911). I 130 DE POLAR AND NOR-POLAR VALENCY. seems poasible that there is a cmnexion betwwn the increase of attraction of a metallic atom for these negative radicles (co-ordinated with it) and the cause which occasions the presence of negative charges of electricity on them when they exist as anions. We have already supposed that t'he valencies outside the comples mdicl0 are caused by electrons of the outermost ring constituting the atomic structure. These electrons may therefore be termed walenee-dectrons. Obviously the number of the valence-electrons of a metallic atom corresponds with that of its maximum valencies outside the complex radicle.I n the case of the complex metal-ammonia compounds this maximurn valency is exhibited when all the radiclm co-ordinated witK the metallic atom are NH or H20, and when a negative radicle is introduced into the above complex radicle the number of valencies outside the complex one is decreased (that is the number of valenceelectrons appears to become less). We may call those valence-eleckons which seem t,o have vanished in this way bozcnd valence-electrons and those which have caused the appearance of valencies outside the complex radicle free wdemce-electrons. We may also notice that the maximum number of free valence-electrons (that is electrons which can escape from a metallic atom forming a complex cation) is the same as the maximum number of electrons which can be received by the atom in addition to its own valence-electrons when it' forms a part of a complex anion.For illustrating this point we may cite the com-pounds (i) [Co(NH3),]C13 where the complex radicle is a ter-valent cation (ii) K,[Co(NO,),] where the complex radicle is a tervalent anion and (iii) [ (NH,),Co(NO,),] a nm-electrolyte. In the third compound no electron has escaped or been received by the metallic atom but all the three valencz-electrons along with the three univalent negative radicles have been bound by its posi-tive nucleus. Regarding the question a t hand namely that of the number of valencies in chlorqpentamminecobaltio chloride it may be said that here the cobalt atom contains three valence-electrons-one is bound along with the univalent chlor+radicle mordinated with cobalt and the remaining two have caused the appearance of valencies of polar Oype outside the complex radicle.Besides these three valenceelectrons the metallic atom possesses six valency bonds of non-polar type caused by the attraction of its positive nucleus. The structure of the complex radicle as conceived above explains also the phenomenon of the directional nature of the auxiliary valency bonds indicated by the stereoisomeric compounds of com-plex metal-ammonia salts for the nucleus being centrally placed in the structure of an atom (metal) has an advantage in exertin DE POLdR AND NON-POLAR VALENCY. 131 its attraction along different directions which the electrons moving in their orbits cannot possibly have.The assumption of the nuclear attraction however need not be confined to these cases of complex derivatives alone. The phenomena of non-polar valencp may in general be considered to have arisen from this attzaction. The kind of valency exhibited in organic compounds is a typical non-polar one. I f w0 compare ths valencies of the carbon atam with those of metallic ones employed in co-ordinating radicles with them we find that both these two kinds are non-polar and direc-tional in nature (shown by the stereoiscmerides of the carbon compounds and those of complex metal-ammonia derivatives). This similarity in their character may indicate the probability of their being brought about by the same cause namely by the attraction of the positive nucleus of an atom.The assumption receives further support from the small at.omic volume of carbon (as shown in Lothar Meyer’s atomic volume curve). It has already been supposed that ths attraction of the nucleus for radicles should increase as the atomic volume decreases and this should tend ta a maximum when the volume becomes very small. Nernst is of opinion that the forces by which the carbon atoms in a crystal of diamond are held together are identical with the attraction of its four valencies called into play in the formation of organic compounds that is to say (‘the forces of cohesion are identical in nature with the forces of chemical affinity” (“The Theory of Solid State,” p. 6). The cohesive forces are found to increase with.the decrease of the atomic volume of elements. They may therefore arise from tfie very same nuclear attraction of atoms mentioned before. An inspection of the behaviour of carbon and its compounds might help in giving some light in this direction. I n the process of the synthesis of diamond an immense pressure is brought about by the contraction of iron in which carbon has been dissolved. Evidently for effecting union (that is saturation of the valencies of carbon atoms) the atoms are required to be brought very near to one another. It may also be noted that in the case of the carbides of metals the carbides Li,C, CaC, etc., where the metals are of large atomic volumes are decomposed by water whilst the carbides Cr,C, Cr,C etc. where the metals in combination are placed on a trough in the atomic volume curve (that is are of small atomic volumes) form stable compounds.Taking into consideration that the mass of an atom is concentrated in an exceedingly small central nucleus in the structure it seems possible that carbon atoms would exert only a very feeble attrac-tion when placed a little apart or when combined with metals of large atomic volumes. The various rings of electrons around th 132 DE POLAR AND NON-POLAR VALENCY. nucleus which have already been compared to shields may also occasion a further hindrance in their union. Further light will bel thrown on the above if the followiiig view is taken of the constitution of triphenylmethyl. It is found that in solution triphenylmethyl has a molecular weight corresponding with the formula (CPh,) (Gomberg and Cosne Ber.1904 37, 2033). This is what may be expected from its mode of preparation : 2CPh,Br + 2Ag = 2AgBr + CPh,*CPh,. As the compound is very reactive even a t a low temperature it has been assumed that i t is rather a derivative of tervalent carbon. Exposure to the air even for a very short time is sufficient to transform it into a peroxide' of the constitution CPh,*O*O*CPh,. We may however represent the constitution of triphenylmethyl as CPh ... CPh,. The weak attraction between the two carbon atoms is shown by the dotted line. The feebleness of their attrac-tion may be due to the inability of the carbon atoms to approach very near t o each other on account of the hindrance caused by the large phenyl groups attached to them.The hindrance referred t o may be of the type1 similar to that of steric hindrance (Wegscheider, Monntsh. 1895 16 148) and their reactivity may be due to the possibility of their drawing small atoms very near t o them. View-ing the constitution given for the oxidation product" it' is seen that by the intervening of two' oxygen atoms the large radicles have been placed apart' and by the union of the' two carbon atoms with two oxygen atoms (small indeed compared with the triphenyl-methyl radicle) a stable compound has been formed. According t o our assumption radicles bound by the positive nucleus should not show any polar character. Alt,hough the valency of carbon is ordinarily non-polar there are a few organic compounds where it seems t o function as polar as for instance i n hydrogen cyanide and in organic acids where we have thel radicles *CIN or *NiC and R.CO,* respectively besides hydrion.There are also sodium acetylide CHiCNa silver acetylide C2Ag, and cuprous aceltylide C2Cun where' the hydrogen atoms of acetylene have been displaced by metallic atoms. I n order to explain this anomaly we may consider Sir J. 3. Thornson's observation that "when the discharge tube cont'ains such gases as CH, CO, CO, where there are 110 bonds between two carbon atoms in the mole-cule we find negatively charged carbon atoms but no negatively charged molecules. When hocwever we use compounds such as acetylene HCiCH ethylene H,C:CH, o r ethane H,C*CH, where, according to the usual interpretation of the constitution of these subst4ances there are bonds between carbon atoms in the molecule, then we find molecules as well as atoms of carbon with the negativ DE POLAR AND NON-POLAR VALEBTCY 133 charge” (Zoc.cit.). He is also of the opinion that on account of the unsaturated valencies of the carbon atoms in the molecule it has been possible for the negative corpuscles to become attached to them (ibid.). A similar explanation may be applicable in the above cases. I n them more than one bond of carbon has been occupied with the other element combined with it and a corpuscle received from an adjacent hydrogen atom may become attached t o the remaining part of the compound thus giving rise to their polar character. Compounds such as LiH H,O NH, etc. Ni(CO), Co2(CO),, etc.and also groups of atonis forming radicles such as CO, NO,, SO, etc. being formed by non-polar valencies may have their origin in the nuclear attraction. We may also ascribe the forces by which atoms and molecules in a crystal are held together to the same attraction. These forces have been supposed to be caused by residual valency which has also been assumed to bring about the solution of a substance in a solvent (Baly “Spectroscopy,” 1912, 11. 487). The phenomena of solution should necessarily be con-sidered t o be due t o the same attraction of the nucleus. I n these cases the size of the molecules may account. for the feeble character of their binding. Lastly all catalytic substances which are employed in gaseous reactions may be supposed t o owe their cata-lytic action to the positive nuclei of the atoms in them.Indeed, the study of the dissociation of the hydrogen molecule into atoms, and other similar studies have convinced Langmuir ( J . Amer. Chem. Soc. 1916 38 2221) that prior t o the dissociation absorp-tion of hydrogen by tungsten wire due t o its secondary valency, does take place. We see therefore t.hat the nuclear attraction plays a great part in all chemical phenomena. Regarding the number of valencies of the non-polar type for different elements it may be noted that carbon (placed in the first trough of Lothar hleyer’s atomic vdume curve) has four valencies, whilst other elements (placed in subsequent troughs of the curve) generally have six. I n the case of the complex platinum-ammonium salts however the derivatives of the platinous salts, for example [(NH,),Pt]Cl, tetra-amminoplatinous chloride show the number of auxiliary (lion-polar) valencies t o be four whilst those of the platinic salts for example [(NH,),PC]Cl, hexa-amminoplatinic chloride the number is six. The increase of two non-pollar valenciea in the latter case has been attended with an increase of two polar ones. Also the directions of these valencies, in the former case lie in a plane whilst’ in the latter case there are two additional directions lying in the same line perpendicular to the above plane. Whether and how the electrons constitutin 134 DUICRANT THE INTERACTION OF the atomio structure influence the number and directions of the non-polar valencies of different elements awaits further study. My best thanks are due to Prof. P. C. RSy for his kind help and encouragement. PRESIDENCY COLLEUE CALCUTTA. ISLAMIA COLLEGE PEBHAWAR INDIA. [Becedved October 2nd 191 7.